Page tree
Skip to end of metadata
Go to start of metadata

Click here to view a PDF version of this release notes document.

Table of Contents

About GSX Release Notes

This document provides information about the Sonus GSX9000 and GSX4000 Open Services Switch Release 09.00.13R001 software. The term GSX is used to indicate that the description applies to GSX9000, GSX4000, and SBC 9000 unless explicitly specified. This release introduces functionality and corrects several problems encountered in earlier versions of this product.

For additional installation requirements, known limitations, restrictions, and other GSX Release Note documentation, refer to the Related Documentation listed below.

To view or download product documentation produced by Sonus Technical Publications, navigate to and login to the online Documentation and Support portal at https://support.sonus.net/display/PORTAL/Home.

First time visiting Documentation and Support portal, or forgot your password? Click here to activate your account or request a new password.

Release Notes Use and Distribution

These Release Notes are protected under the copyright laws of the United States of America. This work contains proprietary information of Sonus Networks, Inc., Westford, MA01886USA. Use, disclosure, or reproduction in any form is strictly prohibited without prior authorization from Sonus Networks, Inc.

Problems or Questions

For problems or questions, contact the Sonus Technical Assistance Center (TAC) via telephone, fax, or Web access:
Worldwide Voice:+1 978 614 8589
USA Toll-free:1 888 391 3434
Worldwide Fax:+1 978 614 8904
Web: www.sonusnet.com

Related Documents

The following GSX documentation is associated with this release:

Title

Part Number

GSX Documentation

 

GSX9000 and GSX4000 Open Services Switch, Version 09.00.00R000

550-05636

GSX9000 and GSX4000 Operations Guide, Version 09.00.00R000

550-05637

GSX9000 Installation and Upgrade Guide, Version 09.00.00R000

550-05640

GSX4010 Hardware, Installation, and Upgrade Guide, Version 09.00.00R000

550-05641

GSX SBC 9000 Application Guide, Version 09.00.00R000

550-05639

GSX Alarm and Troubleshooting Guide, Version 09.00.00R000

550-05638

GSX9000 Field Replaceable Units

 

FRU Doc. CNA01 - T1 Redundancy Circuit Network Adapter

550-00065

FRU Doc. CNA02 - E1 Redundancy Circuit Network Adapter

550-00078

FRU Doc. CNA03 - T3 Redundancy Circuit Network Adapter

550-00066

FRU Doc. CNA04 – 36-Span T1/E1 Circuit Network Redundancy Adapter

550-01426

FRU Doc. CNA05 - 12 Span E1 Redundancy Circuit Network Adapter

550-00100

FRU Doc. CNA06 - High Density 3xT3 Redundancy Circuit Network Adapter

550-00221

FRU Doc. CNA07 - Circuit Network Adapter

550-00303

FRU Doc. CNA08 - High Density 3xT3 and Optical Circuit Network Redundancy Adapter

550-01425

FRU Doc. CNA10 - T1 Circuit Network Adapter

550-00019

FRU Doc. CNA20 - 120 Ohm E1 Circuit Network Adapter

550-00060

FRU Doc. CNA21 - 75 Ohm E1 Circuit Network Adapter

550-00061

FRU Doc. CNA25 - 12 Span E1 Circuit Network Adapter

550-00099

FRU Doc. CNA30 - T3 Circuit Network Adapter

550-00021

FRU Doc. CNA33 - SONET OC-3 Circuit Network Adapter

550-00097

FRU Doc. CNA40 - 36 Span T1/E1 Circuit Network Adapter

550-01423

FRU Doc. CNA60 - High Density 3xT3 Circuit Network Adapter

550-00220

FRU Doc. CNA70 - Circuit Network Adapter

550-00302

FRU Doc. CNA80 - High Density 3xT3 Circuit Network Adapter

550-01424

FRU Doc. CNA81 - High Density Optical Circuit Network Adapter

550-01421

FRU Doc. CNS40 - 36 Span T1/E1 Circuit Network Server

550-01416

FRU Doc. CNS45 - 36 Span T1/E1 Circuit Network Server

550-01417

FRU Doc. CNS60 - High Density 3xT3 Circuit Network Server

550-00219

FRU Doc. CNS71 - Circuit Network Server

550-00299

FRU Doc. CNS80 - High Density 3xT3 Circuit Network Server

550-01419

FRU Doc. CNS81 - High Density Optical Circuit Network Adapter

550-01414

FRU Doc. CNS85 - High Density 3xT3 Circuit Network Server

550-01420

FRU Doc. CNS86 - High Density Optical Circuit Network Server

550-01415

FRU Doc. CPIM71 - Circuit Pluggable Interface Module

550-00301

FRU Doc. CPIM81 - Circuit Pluggable Interface Module for CNS86 and CNS81

550-01422

FRU Doc. Fan - Fan Tray

550-00014

FRU Doc. MNA10 - Management Network Adapter

550-00016

FRU Doc. MNA20 - Management Network Adapter

550-00634

FRU Doc. MNA21 - Management Network Adapter

550-00635

FRU Doc. MNA25 - Management Network Adapter

550-00831

FRU Doc. MNS20 - High Density Management Network Server

550-00633

FRU Doc. MNS21 - High Density Management Network Server

550-01814

FRU Doc. MNS26 –High Density Management Network Server

550-02871

FRU Doc. MTA10a - Management Timing Adapter

550-00093

FRU Doc. MTA20a - Management Timing Adapter

550-00094

FRU Doc. MTA21a - Management Timing Adapter

550-00087

FRU Doc. MTA30 - Management Timing Adapter

550-00304

FRU Doc. PNA10 - Ethernet Packet Network Adapter

550-00023

FRU Doc. PNA30 - Gigabit Ethernet Packet Network Adapter

550-00127

FRU Doc. PNA35 - Packet Network Adapter

550-00380

FRU Doc. PNS40 - Dual Gigabit Ethernet Packet Network Server

550-01119

FRU Doc. PNS41 - Dual Gigabit Ethernet Packet Network Server

550-01813

FRU Doc. PNA40 - Dual Gigabit Ethernet Packet Network Adapter

550-01120

FRU Doc. PNA45 - Dual Gigabit Ethernet Packet Network Adapter

550-01121

FRU Doc. SHF - 16-Slot GSX9000 Chassis

550-00013

FRU Doc. SPS70 - High Density Signal Processing Server

550-00297

FRU Doc. SPS80 - High Density Signal Processing Server

550-01418

FRU Doc. SHIELD – GSX9000 Protective Shield

550-01180

GSX4000 Field Replaceable Units

 

FRU Doc. GSX4010-CHAS_T1_E1_Chassis

550-00957

FRU Doc. GNS15_Gateway_Network_Server

550-00958

FRU Doc. GNA10_Gateway_Network_Adapter

550-00959

FRU Doc. GSX4000-FAN

550-00960

FRU Doc. GSX4000-FILTER

550-00961

FRU Doc. GSX4000-AC_DC_Power_Assembly

550-00962

Network File Server/DataStream Integrator Documentation

 

DSI Administration and Maintenance Guide, versionV09.00.00R000

550-05632

DSI Installation and Upgrade Guide (Solaris), version V09.00.00R000

550-05630

DSI Installation and Upgrade Guide (Linux), version V09.00.00R000

550-05631

GETS Software

This release supports the High Probability of Completion (HPC) network capability intended to provide enhanced probability of call completion to authorized Government Emergency Telecommunication Service (GETS) and Wireless Priority Service (WPS) users during times of network stress and/or congestion. HPC features include GETS Call Routing, Office-Wide Call Queuing, and SIP Resource Priority Header. For more information about these features, refer to the GSX9000 and GSX4000 Series Open Services Switch Operations Guide.
The use of the GETS SOFTWARE is restricted in the U.S. and U.S. TERRITORIES to NS/EP users authorized by the Office of the Manager, National Communications System (OMNCS). Telecommunications deployment of GETS SOFTWARE must be coordinated with the OMNCS at the following address:
Office of the ManagerNational Communications SystemAttn: GETS Program Office701 South Courthouse Rd.Arlington, VA22204-2198Email: gets@ncs.gov

About GSX

The Sonus GSX Open Services Switches are carrier-class packet telephony gateways that provide all of the capabilities required for world-class telephony service on a packet backbone. The GSX9000 Open Services switch is a high density switch, while the GSX4000 series of Open Services Switches are low- to mid-density switches based on GSX9000 technology. The GSX4000 series extends a service provider’s network by creating more broadly distributed IP-to-TDM and IP-to-IP network connections.

Sonus Networks offers a comprehensive suite of products that enable carriers to implement and fully realize the promise of global-scale packet telephony. The Sonus Packet Telephony suite is based on the Sonus Open Services Architecture, the only open blueprint for packet telephony innovation that lets carriers and independent developers quickly create and deliver competitive new services.

Compatibility with Sonus Products

To view the minimum compatible release versions of each Sonus device that can interoperate with GSX, see GSX 9000 Interoperability Matrix.

When upgrading your network, be sure to upgrade each product to the most current release in order to take advantage of the latest features, enhancements and fixes.

For complete interoperability details between the various Sonus products, including backwards compatibility, see Ribbon Product Compatibilities.

Hardware and Software Requirements

Unsupported GSX Cards

The following GSX cards are not supported in GSX release V09.00.13R001:

  • MNS11
  • PNS10
  • PNS20
  • PNS30
  • PNS30a
  • CNS20
  • CNS30
  • CNS30a

CNS10 and CNS25 Card Performance Advisory

The end of product life (EOPL) CNS10 and CNS25 cards (the "EOPL Cards") will execute R9 software subject to the following changes in performance characteristics and supportability:

  • Performance of R9-enabled EOPL Cards may be lower than the performance the customer observed with previous SW releases (for example R6, R7, etc.) on the EOPL cards
  • Performance of the cards may be lower than observed with other CNS cards (e.g. CNS40, CNS80, etc.) still in production.

Other software features newly introduced in R9 (that is not present in R7 or earlier releases) may not operate or operate correctly in an EOPL Card.

Sonus does not guarantee the availability or correct operation of these new features in an r9-enabled EOPL card, nor does Sonus assure repair of any such new features should the customer raise a trouble report.

GSX Firmware Versions

The following GSX firmware versions are currently recommended for the following card types:

Table : Firmware Versions

CardBinary NameVersionAdditional Information
MNS20, MNS21, MNS26mns20b.binV01.20.00R002-
CNS8x, CNS4x, SPS80cns80b.binV01.16.00R001This version disables ECC HW circuitry, which has been found to cause erroneous software problems on rare occasions.
PNS40, PNS41pns40b.binV01.18.00R006This version adds support for new server module types for the GSX and other Sonus products.
All others (default)N.A.V01.12.00R003-

Sonus recommends that you perform firmware upgrades during a low-volume maintenance window, and refrain from disturbing the system during the upgrade.

Scripting Support - Optional Sonus SPE Software

The Sonus Service Profile Editor (SPE) Version 06.03.05 Build 11 is an optional, additional software package that can be used to develop custom application scripts and edit seeded processing scripts for the Trunking (PSX) V09.00.00R000 Release. SPE documentation is available separately with the optional software; the following SPE documentation is associated with this release:

Title

Sonus Part Number

SPE User Guide, Version V06.03.05

550-01243

SPE Version 06.03.05 Release Notes

550-02178

Upgrade Notes

For detailed information on GSX installation and upgrade procedures, refer GSX4010 Installation Upgrade Guide or GSX9000 Installation and Upgrade Guide.

Supported GSX Upgrade Paths

This section provides information on the supported upgrade paths in GSX release V09.00.13R001.

 The highlighted upgrade paths are explicitly tested by Sonus.

Table : Supported GSX Upgrade Paths

Release

Series

Version

V09.00

R

V09.00.00R000, V09.00.01R000, V09.00.02R000, V09.00.03R000, V09.00.04R000, V09.00.05R000, V09.00.06R000, V09.00.07R000, V09.00.08R000, V09.00.09R000, V09.00.10R000, V09.00.10R001, V09.00.11R000, V09.00.12R000, V09.00.12R001, V09.00.13R000 

 

F

V09.00.00F001
V09.00.02F001
V09.00.03F001, V09.00.03F002
V09.00.04F001
V09.00.05F001, V09.00.05F002, V09.00.05F003
V09.00.06F001, V09.00.06F002, V09.00.06F003, V09.00.06F004, V09.00.06F005, V09.00.06F006, V09.00.06F007, V09.00.06F008, V09.00.06F009, V09.00.06F010
V09.00.07F001, V09.00.07F004
V09.00.08F001
V09.00.08F002
V09.00.09F001, V09.00.09F002
V09.00.10F001
V09.00.11F001
V09.00.12F001, V09.00.12F002, V09.00.12F003, V09.00.12F004, V09.00.12F005

 

S

V09.00.02S001, V09.00.02S002

V08.04

R

V08.04.00R000, V08.04.01R000, V08.04.02R000, V08.04.02R001, V08.04.03R000, V08.04.04R000
V08.04.04R001
V08.04.05R000, V08.04.05R001
V08.04.06R000
V08.04.07R000
V08.04.08R000
V08.04.09R000
V08.04.10R000
V08.04.11R000
V08.04.12R001
V08.04.13R000
V08.04.14R000
V08.04.15R000

 

F

V08.04.01F001, V08.04.01F002, V08.04.01F003, V08.04.01F004, V08.04.01F005, V08.04.01F006
V08.04.02F002, V08.04.02F003, V08.04.02F004, V08.04.02F005, V08.04.02F006
V08.04.03F001, V08.04.03F002, V08.04.03F003, V08.04.03F004, V08.04.03F005
V08.04.05F001, V08.04.05F002, V08.04.05F003, V08.04.05F004, V08.04.05F005, V08.04.05F006, V08.04.05F007, V08.04.05F008
V08.04.06F001
V08.04.08F001, V08.04.08F002, V08.04.08F004, V08.04.08F005
V08.04.10F001
V08.04.14F002

 

S

V08.04.04S001, V08.04.04S002, V08.04.04S003, V08.04.04S004

V08.03

R

V08.03.00R000, V08.03.01R000, V08.03.02R000, V08.03.03R000, V08.03.04R000

 

F

V08.03.00F001, V08.03.00F002
V08.03.01F001, V08.03.01F002, V08.03.01F003, V08.03.01F004
V08.03.04F001, V08.03.04F002, V08.03.04F003, V08.03.04F004, V08.03.04F005

V08.02

R

V08.02.01R000

V07.03

R

V07.03.00R000
V07.03.01R000, V07.03.01R001, V07.03.01R002, V07.03.01R003, V07.03.01R004, V07.03.01R005, V07.03.01R006, V07.03.01R007, V07.03.01R008, V07.03.01R009, V07.03.01R010, V07.03.01R011
V07.03.02R000, V07.03.02R001, V07.03.02R002, V07.03.02R003, V07.03.02R004, V07.03.02R005, V07.03.02R006, V07.03.02R007, V07.03.02R008, V07.03.02R009, V07.03.02R010, V07.03.02R011, V07.03.02R012
V07.03.03R000, V07.03.03R001, V07.03.03R002, V07.03.03R003, V07.03.03R004, V07.03.03R005, V07.03.03R006, V07.03.03R007, V07.03.03R008, V07.03.03R009, V07.03.03R010
V07.03.04R000, V07.03.04R001, V07.03.04R002, V07.03.04R003, V07.03.04R004, V07.03.04R005, V07.03.04R006, V07.03.04R007, V07.03.04R008, V07.03.04R009, V07.03.04R010, V07.03.04R011, V07.03.04R012, V07.03.04R013
V07.03.05R000, V07.03.05R001, V07.03.05R002, V07.03.05R003, V07.03.05R004, V07.03.05R005, V07.03.05R006, V07.03.05R007, V07.03.05R008, V07.03.05R009, V07.03.05R010
V07.03.06R000, V07.03.06R001, V07.03.06R002, V07.03.06R003, V07.03.06R004, V07.03.06R005, V07.03.06R006, V07.03.06R007, V07.03.06R008, V07.03.06R009, V07.03.06R010, V07.03.06R011
V07.03.07R000, V07.03.07R001, V07.03.07R002, V07.03.07R003, V07.03.07R004, V07.03.07R005, V07.03.07R006, V07.03.07R007, V07.03.07R008, V07.03.07R009, V07.03.07R010

 

F

V07.03.01F001, V07.03.01F002, V07.03.01F003, V07.03.01F004, V07.03.01F005, V07.03.01F006, V07.03.01F007, V07.03.01F008, V07.03.01F009
V07.03.02F000, V07.03.02F001, V07.03.02F002, V07.03.02F003, V07.03.02F004, V07.03.02F005, V07.03.02F006, V07.03.02F007, V07.03.02F008, V07.03.02F009, V07.03.02F010, V07.03.02F011, V07.03.02F012, V07.03.02F013, V07.03.02F014, V07.03.02F015, V07.03.02F016, V07.03.02F017, V07.03.02F018, V07.03.02F019, V07.03.02F020, V07.03.02F021, V07.03.02F022, V07.03.02F023, V07.03.02F024, V07.03.02F025, V07.03.02F026, V07.03.02F027, V07.03.02F028, V07.03.02F029, V07.03.02F030
V07.03.03F001, V07.03.03F002, V07.03.03F003, V07.03.03F004, V07.03.03F005, V07.03.03F006, V07.03.03F007, V07.03.03F008, V07.03.03F009, V07.03.03F010, V07.03.03F011, V07.03.03F012, V07.03.03F013,
V07.03.04F001, V07.03.04F002, V07.03.04F003, V07.03.04F004, V07.03.04F005, V07.03.04F006, V07.03.04F007, V07.03.04F008, V07.03.04F009, V07.03.04F010, V07.03.04F011, V07.03.04F012
V07.03.05F001, V07.03.05F002, V07.03.05F003, V07.03.05F004, V07.03.05F005, V07.03.05F006, V07.03.05F007, V07.03.05F008, V07.03.05F009, V07.03.05F010, V07.03.05F011, V07.03.05F012, V07.03.05F013, V07.03.05F014
V07.03.06F001, V07.03.06F002, V07.03.06F003, V07.03.06F004, V07.03.06F005, V07.03.06F006, V07.03.06F007, V07.03.06F008, V07.03.06F009, V07.03.06F010, V07.03.06F011, V07.03.06F012
V07.03.06F013, V07.03.06F014, V07.03.06F015, V07.03.06F016, V07.03.06F017, V07.03.06F018, V07.03.06F019, V07.03.06F020, V07.03.06F021, V07.03.06F022, V07.03.06F023, V07.03.06F024, V07.03.06F025, V07.03.06F026, V07.03.06F027, V07.03.06F031
V07.03.07F001, V07.03.07F002, V07.03.07F003, V07.03.07F004, V07.03.07F005, V07.03.07F006, V07.03.07F007, V07.03.07F008, V07.03.07F009, V07.03.07F010, V07.03.07F011, V07.03.07F012, V07.03.07F013, V07.03.07F014, V07.03.07F015, V07.03.07F016, V07.03.07F017, V07.03.07F018, V07.03.07F019, V07.03.07F020, V07.03.07F021, V07.03.07F022, V07.03.07F023

 

S

V07.03.02S001
V07.03.04S001, V07.03.04S002, V07.03.04S003, V07.03.04S004, V07.03.04S005
V07.03.04S100, V07.03.04S101, V07.03.04S102, V07.03.04S103, V07.03.04S104, V07.03.04S105, V07.03.04S106, V07.03.04S107, V07.03.04S108, V07.03.04S109, V07.03.04S114, V07.03.04S115
V07.03.04S200
V07.03.05S402
V07.03.07S001, V07.03.07S002, V07.03.07S003, V07.03.07S004, V07.03.07S005, V07.03.07S006, V07.03.07S007, V07.03.07S008, V07.03.07S009, V07.03.07S010, V07.03.07S011, V07.03.07S012

V07.02

R

V07.02.00R000, V07.02.00R001
V07.02.01R000
V07.02.02R000, V07.02.02R001, V07.02.02R002, V07.02.02R003, V07.02.02R004, V07.02.02R005, V07.02.02R006, V07.02.02R007 V07.02.02R008, V07.02.02R009, V07.02.02R010
V07.02.03R000, V07.02.03R001
V07.02.04R000, V07.02.04R001, V07.02.04R002, V07.02.04R003
V07.02.05R000, V07.02.05R001, V07.02.05R002
V07.02.06R000
V07.02.07R000, V07.02.07R001, V07.02.07R002
V07.02.08R000
V07.02.09R000
V07.02.10R000, V07.02.10R001
V07.02.11R000

 

F

V07.02.02F003
V07.02.02F001, V07.02.02F002, V07.02.02F004, V07.02.02F005, V07.02.02F006, V07.02.02F007, V07.02.02F008, V07.02.02F009, V07.02.02F010, V07.02.02F011, V07.02.02F012
V07.02.05F001, V07.02.05F002, V07.02.05F003
V07.02.06F001, V07.02.06F002, V07.02.06F003
V07.02.07F001, V07.02.07F002, V07.02.07F003, V07.02.07F004
V07.02.10F001

 

S

V07.02.01S001, V07.02.01S002, V07.02.01S003
V07.02.02S301, V07.02.02S302, V07.02.02S303, V07.02.02S304
V07.02.06S001

V07.01

R

V07.01.00R000, V07.01.00R001, V07.01.00R002, V07.01.00R003, V07.01.00R004, V07.01.00R005, V07.01.00R006
V07.01.01R000, V07.01.01R001, V07.01.01R002, V07.01.01R003, V07.01.01R004. V07.01.01R005, V07.01.01R006, V07.01.01R007, V07.01.01R008
V07.01.02R000, V07.01.02R001, V07.01.02R002, V07.01.02R003, V07.01.02R004, V07.01.02R005, V07.01.02R006
V07.01.03R000
V07.01.04R000, V07.01.04R001, V07.01.04R002
V07.01.05R000
V07.01.06R000, V07.01.06R001, V07.01.06R002, V07.01.06R003
V07.01.07R000
V07.01.08R000
V07.01.09R000
V07.01.10R000

 

F

V07.01.00F001, V07.01.00F002, V07.01.00F003, V07.01.00F004, V07.01.00F005, V07.01.00F006, V07.01.00F007, V07.01.00F008
V07.01.01F001
V07.01.02F001, V07.01.02F002
V07.01.03F001, V07.01.03F002, V07.01.03F003, V07.01.03F004
V07.01.04F001, V07.01.04F002, V07.01.04F003, V07.01.04F004
V07.01.06F001, V07.01.06F002, V07.01.06F003
V07.01.07F001, V07.01.07F002, V07.01.07F003

 

S

V07.01.04S001, V07.01.05S001, V07.01.05S002
V07.02.02S301, V07.02.02S302, V07.02.02S303, V07.02.02S304

RTCP must be disabled before upgrading from GSX release V08.03.04R000 to V09.00.xx.

Note the following before performing Live Software Upgrade (LSWU):

  • If Layer3 IP redundancy is configured using OSPF, then OSPF mode must be set to PASSIVE mode prior to performing the LSWU.
    If OSPFv2 is being used for Layer 3 redundancy, execute the following command to set the mode to passive:

    CONFIGURE OSPF SHELF shelfSTATE DISABLED
    CONFIGURE OSPF SHELF shelf MODE PASSIVE

    In case OSPFv3 is being used for Layer 3 redundancy, execute the following command to set the mode to passive:

    CONFIGURE OSPFV3 SHELF shelf STATE DISABLED
    CONFIGURE OSPFV3 SHELF shelf MODE PASSIVE
  • If the setup has SPS70, perform the following steps before the LSWU:

    • LSWU should be done on low call volume on the SPS70 cards. Alternatively, the SPS card should be dried up before initiating LSWU.
    • Before upgrading SPS70 card through LSWU from the older releases (07.0x.0x), assign Overload Profile to the SPS70 active slot. This Overload Profile ensures that SPS70 does not run out of memory, execute the following commands to configure the Overload Profile to SPS70 card:

      CREATE OVERLOAD PROFILE <profilename>
      CONFIGURE OVERLOAD PROFILE <profilename> THRESHOLD CLEAR MEMORY THRESHOLD 10
      CONFIGURE OVERLOAD PROFILE <profilename> THRESHOLD set MEMORY THRESHOLD 20
      CONFIGURE OVERLOAD PROFILE <profilename> STATE ENABLED 
      CONFIGURE CONGESTION SERVER SHELF 1 SLOT <slot> MODE OUTOFSERVICE 
      CONFIGURE CONGESTION SERVER SHELF 1 SLOT <slot> OVERLOAD PROFILE <profilename> 
      CONFIGURE CONGESTION SERVER SHELF 1 SLOT <slot> MODE INSERVICE

      The slot number can be between 3 and 16.

    • Once GSX is upgraded successfully, execute the following command to revert previously configured Overload Profile for SPS70 card:

      CONFIGURE CONGESTION SERVER SHELF 1 SLOT <slot> MODE OUTOFSERVICE CONFIGURE CONGESTION SERVER SHELF 1 SLOT <slot> OVERLOAD PROFILE <profilename>

      If nothing was configured explicitly, set it to defaultSps.

    • Before upgrading to this GSX release V09.00.xx, ensure that in the pre-upgrade configuration, every NIF/SIF is configured only in one NIF GROUP.

    • The AMR and EFR VAD in case of LSWU for the stable call have VAD OFF for the upgraded version. By default, VAD will be ON for the new call.

    • A new input prefix value for IPv6 (IP6ADDRESS) is added to the GSX CLI. Ensure that all the configuration TCL files using short form for IP Address (i.e. "ip") need to either change it to "ipaddress" - denoting IPv4 address or, "ipv6address" - denoting IPv6 address.

    • Prior to performing the software upgrade from GSX Release 07.03.xx to this GSX release V09.00.xx, ensure that the light gray policer is disabled on all the NIFs. Light Gray policer information is configured at NIF level and can be retrieved by executing the following command:

      SHOW NIF ALL ADMIN

      For all the NIFs that have light gray policer enabled, execute the following command to disable it:

      CONFIGURE NIF <NIF NAME> POLICER LIGHTGRAY STATE DISABLED

      After the software upgrade is done, execute the following command to enable the light gray policer on all the NIFs for which it was disabled during upgrade.

      CONFIGURE NIF <NIF NAME> POLICER LIGHTGRAY STATE ENABLED

Master Trunk Group Upgrade Procedures

Pre-requisites Before Upgrading

MTRM need not be turned off unconditionally. If you are using release V07.03.07R008 or later or V08.04.05R000 or later which contain the MTRM updates, you can upgrade to V09.00.10 without changing the MTRM configuration.

The following steps must be performed before upgrading to this GSX release V09.00.xx.

  1. Disable all MTRM functions (for example, remove MTRG from IPTG configuration) and readjust IPTG local resource pool, if necessary.
    1. Increase the CAC limit for IPTG (if required).
    2. Remove all IPTG associations by using the following command:

      CONFIGURE TRUNK GROUP <IPTG Name> TRUNK GROUP NONE
    3. Delete Master Trunk Group configuration by using the following commands:

      CONFIGURE MASTER TRUNK GROUP <Master Trunk Group Name> MODE OUTOFSERVICE
      CONFIGURE MASTER TRUNK GROUP <Master Trunk Group Name>STATE DISABLED
      DELETE MASTER TRUNK GROUP <Master Trunk Group Name>
    4. Delete Master Trunk Group connection port by using the following commands:

      CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 1 MODE OUTOFSERVICE
      CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 1 STATE DISABLED
      DELETE MASTER TRUNK GROUP CONNECTION PORT 1
  2. All IPTGs operate in their local capacity with no MTRM resource management.

  3. Upgrade each GSX9000/SBC 9000 using the normal procedure.

  4. No explicit LSWU support for ICM messages as all MTRM functions are shut down.

    Post-upgrade Procedures
    The following steps must be performed after upgrading to this GSX release V09.00.xx:

  1. Create and enable all connection ports using the following commands:

    CREATE MASTER TRUNK GROUP CONNECTION PORT 1 ROLE SERVER
    CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 1 IPADDRESS <Connection Port IP> INTERFACE <Interface Name>
    CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 1 MODE INSERVICE
    CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 1 STATE ENABLED 
    CREATE MASTER TRUNK GROUP CONNECTION PORT 2 ROLE CLIENT
    CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 2 IPADDRESS <Connection Port IP> INTERFACE <Interface Name>
    CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 2 MODE INSERVICE
    CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 2 STATE ENABLED 
    CREATE MASTER TRUNK GROUP CONNECTION PORT 1 ROLE CLIENT
    CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 1 IPADDRESS <Connection Port IP> INTERFACE <Interface Name>
    CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 1 MODE INSERVICE
    CONFIGURE MASTER TRUNK GROUP CONNECTION PORT 1 STATE ENABLED
  2. Create and enable all master trunk groups. Configure the role as client/server using the following commands:

    CREATE MASTER TRUNK GROUP SERVER <Master Trunk Group Name>
    CONFIGURE MASTER TRUNK GROUP SERVER <Master Trunk Group Name> CONNECTION PORT INDEX 1
    CONFIGURE MASTER TRUNK GROUP SERVER <Master Trunk Group Name> mode in
    CONFIGURE MASTER TRUNK GROUP SERVER <Master Trunk Group Name> STATE ENABLED 
    CREATE MASTER TRUNK GROUP CLIENT <Master Trunk Group Name>
    CONFIGURE MASTER TRUNK GROUP CLIENT <Master Trunk Group Name> CONNECTION PORT INDEX 2
    CONFIGURE MASTER TRUNK GROUP CLIENT <Master Trunk Group Name> MODE INSERVICE
    CONFIGURE MASTER TRUNK GROUP CLIENT <Master Trunk Group Name> STATE ENABLE
  3. Configure all IPTG associations using the following command:

    CONFIGURE TRUNK GROUP sipDefault MASTER TRUNK GROUP <Master Trunk Group Name>
  4. Create co-resident client connection ports, if necessary.

  5. Create co-resident client Master Trunk Groups, if necessary.

Software Upgrade Considerations

Consider the following issues when upgrading to this release.

PSX Device Profile

The Device Profile object on the PSX Policy Server is not supported by the GSX in this release. Endpoint registration timeouts must now be specified on the GSX SIP Service object.
To specify the timeout for the default SIP registration keepalive mechanism, execute the following command:

CONFIGURE SIP SERVICE <sipservice_name> REGISTRATION MAXEXPIRES

To enable the protocol-specific timeouts for registrations established over TCP or UDP connections respectively, execute the following commands:

CONFIGURE SIP SERVICE <sipservice_name> REGISTRATION MAXEXPIRES NATOVERTCP
CONFIGURE SIP SERVICE<sipservice_name> REGISTRATION MAXEXPIRES NATOVERUDP

To enable non-nat timeouts, execute the following command:

CONFIGURE SIP SERVICE <sipservice_name> REGISTRATION MAXEXPIRES NONNAT

For more information on this feature,refer to "SIP Service" section in GSX9000 and GSX4000 Open Services Switch Operations Guide (550-05375).

Overload Control Profile Settings

The following releases and later support the System-Level Adaptive Overload feature: V06.04.02R000, V06.05.00, V7.x (all), and V08.04.04R000.

The MC3 overload profile call rate default is now set = 0 for new installations (per SONUS00040601).Existing customers who upgrade to this release must manually set the call rate set/clear thresholds to 0 in the default MC3 overload profile.

For more information on the Overload Profiles, refer to the "Congestion Control" section in the GSX9000 and GSX4000 Open Services Switch Operations Guide (550-05375).

IP Filters

Link integrity checking requires that ICMP replies from the next hop peer be allowed into the GSX. You must verify that there are no IP filters that would discard such replies. Otherwise, the link will be declared as unavailable and, if so configured, the card will switch over. The IP filter configuration should be checked even if link integrity is currently working, since a bug that incorrectly permitted ICMP replies through (SONUS00059718) has since been corrected.

IP Path Routing

As of release 6.3, the GSX enforces the correct path routing for signaling. In most cases this is fully backward-compatible for existing deployments. However, in releases prior to 6.3, SIP or Gateway signaling port configurations were allowed to be inconsistent with the IP routing tables. If SIP or Gateway signaling ports are configured to use a PNS interface, IP route table entries must direct SIP or Gateway signaling packets to a PNS NIF. Check your routing configuration before you upgrade.

New Default Value for Bandwidth Request Variable

The default value of the MIB variable sonusTgMtrgBWPerReq (range 2480 to 248000 in units of 1K bits per second) and the Trunk Group parameter "mbwperreq" was changed to 12400 in GSX 6.1 and later. TRM requests more bandwidth when the available trunk group (IP type MTRG) bandwidth is <1x this value, and frees bandwidth when the available bandwidth is >2x this value.

Default Carrier Codes

The default carrier codes for trunk groups are configured on the PSX, and not on the GSX.

Logical Signaling Addresses and ARP Requests

The 'ARPable logical address' feature introduced in 6.x allows configuration of signaling addresses (SIP/GW/H323) in the same subnet as packet NIFs, and expect GSX to respond to ARP requests for those addresses with appropriate ARP replies. By default ARPable logical address feature is disabled. Therefore, if logical signaling addresses are already in the same subnet as packet NIFs (in prior releases), the GSX does not automatically respond to ARP requests when upgrading to this release.

Additional Information

CAM/CDR Field Changes

No CAM/CDR field changes in this release.

NIF Group

Before upgrading to this GSX release V09.00.xx, ensure that in the pre-upgrade configuration, every NIF/SIF is configured only in one NIF GROUP.

If a NIF/SIF is configured in more than one NIFGROUP, the upgrade operation gets aborted.

Changes in Default Behavior

This section describes the changes made that affect the default behavior of the GSX.

  • Out of Dialog Refer will be routed by default if PSX has appropriate route. This behavior is not configurable and is available by default. In prior releases, we used to reject such a Refer.
  • Changes in Default DNS behavior:
    • DNS Servers will now be attached to a Zone. The default Zone of a newly created DNS Server is INTERNAL. Also, after an upgrade, existing DNS Servers will belong to INTERNAL Zone.
    • Configuration of more than 8 DNS Servers is supported in this release. However, maximum number of DNS Server supported per Zone is 8.
    • RFC5452 support:
      • Source Port of DNS queries will be randomized. The port rollover happens every 60 seconds.
      • Query ID of the DNS message will be randomized (all 16 bits).
      • DNS query responses will be subject to Query Matching Rules and will be dropped if the query matching fails. No update to local DNS Records cache will be made in this case.
  • It is no longer possible to configure one NIF/SIF in multiple NIF Groups.
  • In Local Policy mode, GSX will not relay in-dialog Refer and GSX would process and act on this Refer if Refer relay is not enabled.
  • PNS41 card supports a maximum of 20000 Relay Dialogs (REFER/SUBSCRIBE/OOD messages).
  • SBC will be compliant with RFC 3261 by mapping the 503 response code received on the egress call leg to 500 response code onto the ingress call leg for initial INVITE and REGISTER messages.

GSX CDR Streaming Configuration

When using the CDR streaming feature, one DSI can support live event/CDR data streaming from multiple GSXs. This means that during LSWU a DSI must be able to support multiple CAM versions, from multiple GSX devices running different CAM versions. Therefore, to ensure correct behavior during GSX upgrade, you must upgrade the active and standby DSI devices and load the correct streaming configuration files onto the DSIs before upgrading the GSX.

If the GSX is being upgraded as a part of a major planned release, then it is very likely that DSI too has been upgraded to the same release. When the DSI upgrade is complete, all the selectable CDR configuration files—both old and new—that are needed to support the GSX should be present on the DSI.

However, if the GSX is being upgraded in association with a specific GSX maintenance release, there may not be a planned DSI upgrade. It is especially important in this case to confirm dependencies of the new GSX maintenance release for possible new and required DSI selectable CDR configuration files to support the GSX maintenance release. Make sure that the selectable CDR streaming configuration files for the old and new CAM versions are present on the DSI before LSWU of GSX devices. If the specific required selectable CDR configuration file is not present on the DSI servers, then upgrade must be postponed until the necessary configuration files can be obtained and loaded on the DSI servers.

For this GSX release V09.00.xx, ensure this additional selectable CDR streaming configuration file is present on the DSI servers: BCAMSelRecord.gsx.76.0.0.stream_config
This file can be found in the /export/home/dsi/files directory on the DSI server along with the similarly named file(s) for whichever version(s) of GSX CAM is/are presently being supported before upgrade of the GSX. If this file is not present, you should contact Sonus TAC to get the necessary configuration files to support your upgrade.

New Features and Enhancements

This section provides an overview of new features and enhancements included in GSX 09.00.xx releases.

New Features and Enhancements in V09.00.13R001

There are no new features or enhancements included in in V09.00.13R001 release.

New Features and Enhancements in V09.00.13R000

The following enhacement is included in V09.00.13R000 release:

GSX-175—VPAD ECAN Error Handling Enhancements

For reasons unknown, the echo canceller processing returns error. Invoking this function further results in DSP hanging in the processing function and leads to DSP coredump. This issue exists only on CNS cards.

As a workaround, reinitialize the echo canceller module when the echo canceller processing function returns error. If the echo canceller module does not recover even after reinitializing, then DSP is inactivated and this eventually results in coredump.

The workaround is applicable only on the following  CNS cards:

  • CNS40
  • CNS45
  • CNS60
  • CNS71
  • CNS80
  • CNS81
  • CNS85
  • CNS86

New Features and Enhancements in V09.00.06R000

The following new feature is included in V09.00.06R000 release:

IPV6 Link Local Nexthop Support

With the current implementation, GSX was unable to use Link-Local Nexthop for IPv6. Support has been added to configure IPv6 Link Local address as Nexthop in static routes configured on MGMT NIF/NIF/SIF.

New Features and Enhancements in V09.00.05R000

End to End ACK

As per the current behavior of GSX, receiving a "200 OK" from a SIP side is responded with ACK almost instantaneously. This can cause inconsistencies in billing of the call if the ACK from other side has not been received yet.
As part of this feature, the behavior of sending an ACK to be End to End is implemented, i.e.; if configured, receiving a "200 OK" from Egress side will not be responded with ACK until an ACK is received from Ingress side.
This configuration is a flag in PSX: 'End to End Ack' in Egress IPSP of Ingress Trunk Group.

  • If the flag is enabled, the behavior will be End to End.
  • If the flag is disabled, the current behavior prevails, i.e.Instantaneous ACK will be sent on receiving "200 OK". This is also the default value.

    This behavior/feature is valid only for initial INVITE of the call.

Need alarm off Dynamic Blacklist Entry

Currently, there is no way to easily get the status of a dynamic blacklisted entity since there is no alarm generated and no log entry unless on INFO level logging.
This feature introduces a new SNMP trap, which is generated when DBL watch or blacklist threshold for an endpoint is reached. This trap will get generated every time a new EP is identified for Dynamic Blacklisting and will be delivered in best effort basis. The trap will also get logged in SYS logs at MAJOR level.
Following are the trap parameters:

  • Endpoint IP
  • Slot
  • Port
  • LIF Index
  • SIP service group name
  • DBL profile name
  • DBL threshold
  • Blacklist/watch duration
  • Reason (invite malformed or register authentication failure)

New Features and Enhancements in V09.00.04R000

The following features are included in V09.00.04R000 release:

This feature enhances the support of PIF tool and verifyAnnFile for DSI box (Linux).

  • PIF Tool: This tool (compiled and linked with snPifLib.a) allows a user to perform a variety of operations on a PIF file.
    Syntax:

    piftool <mode> <param file> [<text file>] [<options>]

    Mode

    Description

    read

    Read a PIF file and dump all objects as parsable records.

    browse

    Dump contents of parameter file in a form suitable for browsing.

    summary

    Dump summary of data from Sonus header and PIF header and trailers.

    checksum

    Reads parameter file to determine if file-wide checksum is valid.

    version

    Print Sonus software version that generated schema.

    genkey

    Produce MD5 key that uniquely identifies a schema.

    schema

    Schema Details.

    Option

    Description

    -flat

    When used in conjunction with "read" mode, it produces flat (non-hierarchical) output.

    -onlyobj

    Confine read/browse to loading specified obj-id or name.

    -fieldcopy

    When reading from file, copy "object" fields using PIF COPY API. This copies the field data into the space specified, otherwise, the PIF GET API is used which allocates memory and returns the value.

    -enumerations

    Translate numeric enumeration values into their textual form.

    -fieldasnnames

    Translate numeric field IDs into their associated MIB object names.

    -objnames

    Print textual name of object type.

  • verifyAnnFile: This tool allows a user to verify user created announcement files.

    Syntax:

    verifyAnnFile <wave file> [<-v>]


Lawful Intercept, RADIUS message buffering limitation removed

GSX uses packet cable PKT-SP-EM-I07-030815 specification to send LI message to the mediation device. This specification uses RADIUS Accounting protocol. RADIUS Accounting protocol is based on client/server model with the client sending a request and server sending a response to it. The GSX has a frame work, where it sends a LI message as a request and waits for a response before it sends the next message (request). The framework allows only one outstanding message at a time. If there are multiple messages to be sent, then the message are buffered and sent one by one.
The GSX was buffering a maximum of 256 messages when there is a burst of LI RADIUS messages that need to be sent. With this feature, this limit is removed and the GSX will buffer as many messages (more than 256) that need to be sent to the LI server.

AIS TIMER Call Drop

Currently, in the event of a fiber cut, the GSX declares AIS alarm on the E1s that are multiplexed within the fiber before the Automatic Protection Switching completed, thereby failing all the calls carried by the affected E1s.  GSX has an AIS timer hardcoded at 100 ms as verses 2.5 s, resulting in bringing down the E1 port in 100 msec. The AIS timer expires in 100 msec and the ports (E1 ports) are brought down.  The stable calls are cleared.
PCR 8172 addresses the configuration requirements for AIS timer providing flexibility for the customer to be able to configure AIS timer through CLI commands without bringing down the E1 ports.
As part of this PCR a new configurable control under E1 is introduced, which provides choice to the customer to choose between timer option values 100ms and 2500ms.
The CLI command is as follows:

CONFIGURE E1 <e1> AISTIMER
<millisec100> <millisec2500> 

The default value for AIS Timer is <millisec100> value.

A new configurable control under E1 is introduced, which provides choice to customer to choose between timer option values 450ms and 2500ms.
The CLI command is as follows:

CONFIGURE E1 <e1> RAITIMER
<millisec450> <millisec2500> 

The default value for RAI Timer is set as <millisec 450> value.

In order for GSX to declare AIS alarm on the E1s in 2.5 seconds, both AIS and RAI timers must be configured to value of millisec 2500.

GSX discards SID or CN Packets with Payload Size Greater than 10 Bytes  

Currently, GSX had a restricted SID payload size of a maximum of 10 bytes (1 byte of noise level and 9 bytes of reflection co-efficient). Any SID payload whose size is greater than 10 bytes gets dropped and DSP does not generate corresponding comfort noise.
This feature addresses the issue by increasing the maximum SID payload size supported to 36 bytes. This fix is applicable on VPAD resources on CNS cards.

Sonus proprietary SID is no more supported.

RFC 5806 not followed for a Single Diversion Header

Currently, on receiving a SIP INVITE with a single diversion header for a SIP to ISUP/ISDN/SIP-I call, the Diversion header was mapped to both the Redirecting Number and Original Called Number on the egress leg. This conflicts with RFC 5806, which describes the interworking between SIP-ISUP.
As part of this feature, a new flag is introduced in the SIP Service to control the new behavior:

CONFIGURE SIP SERVICE <ingress-sip-service> POPULATE ONLY_RN FOR SINGLE DIVERSION HEADER ENABLED
CONFIGURE SIP SERVICE <ingress-sip-service> POPULATE ONLY_RN FOR SINGLE DIVERSION HEADER DISABLED

The flag status can be checked using: SHOW SIP SERVICE <text> ADMIN
By default, this flag will be disabled and the behavior will be the same as prior to this feature.
Whereas, on enabling this flag, if a single Diversion header is received in the SIP INVITE, this will be matched only to the Redirecting Number and not to the Original Called Number.

Call Flow Transcoding fails between Inband and 2833 

While bridging two legs for Two B-channel Transfer (TBCT) (ISUP-A ----> GW1<->GW2 --- ISDN ----> ISDB-B ----> SIP-C), if original call legs do not have DRES (XPAD/PXPAD), GSX will now bring in new DSP resource for the bridged call legs. It will perform transcoding between In-band DTMF and 2833 DTMF based on the options selected in GW PSP.

Support for MAJOR logs to Notify Media Failure due to unreachability of Nexthop

This feature addresses the need to change the log level of two of the existing logs from MINOR to MAJOR.

 

New Features and Enhancements in V09.00.03R000

The following features are included in V09.00.03R000 release:

Currently, the Charge Message Propagation flag in the Signaling Profile screen on the PSX is not used if there is no Charge Band Entry configured.
This PCR allows for the charge message propagation control to be passed back in the policy response even when the charge band information is not provisioned/present. This allows the operator to specify that the CRG message in India can be propagated even though the charge band is not provisioned.

LCQ 8397 -Called Number incorrect SIP Redirect is followed by ISUP Redirect

Currently, in a call scenario where a SIP redirection is followed by ISUP redirection, the Called Directory Number sent in the final IAM contained the original called number from the first IAM instead of the called number from the Intermediate (second) IAM. LCQ 8397 addresses this problem by setting the Called Directory Number to the Called Number from the Intermediate (second) IAM.

PCR 8260 - GSX CANCEL behavior for SIP-T

Currently GSX can generate CANCEL method for INVITE before receiving 100 Trying for SIP-T. This is not compliant with the SIP standard RFC3261, but it is intended for SIP-T/ISUP.
The requirement is to provide a flag to disable the current behavior (generate CANCEL immediately) and wait for 100 Trying back from the peer before issuing a CANCEL.
A Flag called "Enable 3261 Cancel Handling" is added in "IP Signaling Profile" under "Egress IP Attributes" -> Flags.

  • When this flag is enabled, it indicates to the gateway initiating an INVITE that it can only send a SIP CANCEL after receiving a SIP provisional response i.e. SIP CANCEL must conform to RFC3261.
  • When this flag is disabled, this constraint does not apply and the CANCEL can be sent anytime. This is the default behavior.

LCQ 8450 - Need alarm off Dynamic Blacklist Entry

Currently, there is no way to easily get the status of a dynamic blacklisted entity since there is no alarm generated and no log entry unless on INFO level logging.
LCQ 8450 introduces a new SNMP trap, which is generated when DBL watch or blacklist threshold for an endpoint is reached. This trap will get generated every time a new EP is identified for Dynamic Blacklisting and will be delivered in best effort basis. The trap will also get logged in SYS logs at MAJOR level.
Following are the trap parameters:

  • Endpoint IP
  • Slot
  • Port
  • LIF Index
  • SIP service group name
  • DBL profile name
  • DBL threshold
  • Blacklist/watch duration
  • Reason (invite malformed or register authentication failure)

PCR 8817-GSX, SBC Gateway - Gateway Support for the PCRs 7643 and 6896

PCR 8817 addresses the following two issues with GW-GW call using GSX 9.0.x release and SBC 4.0 release.
In SIP to SIP call, when GSX receives JIP (Jurisdiction parameter) in INVITE, GSX drops JIP field's country code and global number flag, so SBC does not send these fields in the JIP parameter of egress INVITE.
In SIPI to SIP call, when GSX receives INVITE with SGD in IAM MIME and PSX flag "Map SGD digits to P-SIG-INFO header" is enabled, GSX fails to send the PSX indication "Map SGD digits to P-SIG-INFO" to SBC via GW-GW. SBC does not send P-SIG-INFO header with SGD in egress INVITE.
The mentioned issues are due to the down conversion routines of GSX. These conversion routines are updated to make the above scenarios work in GSX, SBC GW-GW calls.

PCR 8095&8096 - ISDN Progress Indicator parameter and STATUS message issue

As part of PCR 8095 & PCR 8096 the following issues have been addressed.

  • GSX currently does not add progress indicator=#8 to ALERT message receipt upon ACM/CPG in some cases. Some PBX cannot open its voice path without receiving progress indicator #8 in ALERT message. In that case, caller (ISDN user) will be connected to callee without hearing the ring back tone.
  • When GSX sends SETUP message with Cause of no ID parameter for TTC-ISUP, some PBX send back STATUS message as correct response since they do not support to receive the Cause of no ID parameter. GSX should not echo back STATUS message when it get STATUS message. This results in tearing down the call by some PBX.

The following fix has been provided and the behavior after the PCR changes is as follows.

  • Upon receiving the ACM/CPG for the JAPAN PRI to TTC ISUP, these scenarios are reported as an issue. GSX shall support sending out progress Indicator set to PI=#8 in ALERT message on the ISDN PRI.
  • For this TTC-ISUP to Japan PRI call scenarios reported, when GSX receives the STATUS message back from PBX, GSX shall consume the STATUS message received and shall not echo back STATUS message to PBX.
CLI Configuration

A new configurable control under ISDN SERVICE will be introduced which will take care of whether the requirements will be enabled and will ensure backward compatibility. A new CLI configuration will be introduced under ISDN SERVICE to enable the requirements for restricting echoing of the status message .This will ensure backward compatibility for the status echoing behavior.

Default configuration and command
 CONFIGURE ISDN SERVICE isdn_se SEND PROGRESS INDICATOR INBAND IN ALERT <supported> <unsupported>  
 CONFIGURE ISDN SERVICE isdn_se RESTRICT ECHOING STATUS MESSAGE <supported> <unsupported>  

Default is set as "unsupported".

PCR 8559: Enhancement to GSX-DSI fail-over mechanism in degraded IP NW conditions

This PCR describes the implementation of the functionality of GSX NFS related software. Currently, in case of degraded IP NW scenarios, GSX receives TIMEDOUT errors and the write operation to active NFS fails. When this network impairment persists, CDR records are dropped as GSX fails to notice anything wrong with the network and continues to send data to active NFS.
In order to enhance the response of the external N/W. GSX NFS module functionality is enhanced to keep track of the number of non-fatal NFS. NFS Switchover is performed when the following conditions are met.

  1. When the number of non-fatal write-timeouts reach "x" within "y" seconds threshold.
  2. Secondary NFS is tested to be ready. Secondary NFS is tested periodically at the interval of 15 seconds.

If secondary NFS is also bad at this time, the switchover is avoided and the current NFS is used. An NFS switchover failure trap is generated for this condition.EMS support for the traps will be provided in future.

When the secondary link shows the same degraded N/W behavior then a problem is reported.

Additional implementation of a new feedback mechanism for the dropped packets to the NFS FSM is currently not considered as part of this document.

Newly added CLIs:

configure nfs  shelf  1  timeout period   <in secs>                 
configure nfs  shelf  1  timeout threshold <> 
configure nfs  shelf  1  timeout state enabled 

Show nfs shelf 1 timeout admin
Show nfs she 1 timeout stats

Newly added trap:

NFS_SWITCHOVER_FAILURE( ULONG prevActiveServer, ULONG newActiveServer, SOCKADDR serverIp, STRING serverMount, ULONG reason)

Enhanced existing trap:

NFS_SWITCHOVER( ULONG prevActiveServer, ULONG newActiveServer, SOCKADDR serverIp, STRING serverMount, ULONG reason) 

New value for reason: sonusNfsReason_serverLinkNotStable

New Features and Enhancements in V09.00.02R000

The following features have been added in GSX V09.00.02R000 release:

This PCR addresses the need to enhance the GSX9000 to interwork between BT IUP to SIP-I(UK-ISUP) using GW-GW SBC5200 on the other side.
ISUPSG module has been updated for BT IUP changes on GSX. SS7LIB modules have been updated for the SIPI (UK ISUP) interworking on SBC.
ISUPSG parses the BT IUP messages and BT IUP parameters get mapped to internal CPC parameters. The whole IUP payload is tunneled to the egress SG and far side/egress ISUP SG maps BT-IUP payload to an ISUP payload.
In order to make this work over SIP-I, IUP payload has to be mapped to CPC parameters along with the bit bag and transported over Gateway.
For interworking between SIP-I (UKISUP) terminated on SBC5200 and IUP on the GSX using GW-GW, SBC5200 shall support encapsulating the whole UK-ISUP mime payload.
SBC5200 will also support new interworking functions to use the BT ISUP raw payload received from the other side to create the ISUP MIME in the SIP SG context instead of directly calling the existing generic library processing functions.

LCQ 8667 - SIP ONHOLD/OFFHOLD Race Condition Issues

With the current implementation, call on-hold/off-hold may teardown due to internal SDP offer/answer race conditions. Sometimes additional on-hold is generated while the call is already on-hold. LCQ 8667 addresses this issue such that additional on-hold is not generated while the call is already on-hold to prevent race conditions.

New Features and Enhancements in V09.00.01R000

The intent of 9.0.1 release is to qualify GSX4010 platform on V09.00.0R000 software code.

For a summary of features in the base release V09.00.00R000, refer to the GSX V09.00.00R000 release notes, Sonus part number 550-05636.

GSX4000 Unsupported Features and Functionalities

In this release the following GSX4000 features and functionalities are not supported:

  • ACLs, IP input filters
  • NICE recording
  • IPv6
  • IPSec
  • Wildcard Policers
  • Local disk support for GNS
  • PK headers
  • Registration admission control
  • Call Park/pickup changes (P2P)
  • SIP over SCTP and SCTP
  • Conformance to SIP Connect, Comedia
  • REFER in P2P scenarios
  • SIP Registration Fraud
  • SIP-I to SIP-I flows
  • SIP over TLS
  • SIP Transparency
  • H323 to SIP/SIP-H323 interworking
  • SUBSCRIPTION Relay
  • Direct Media (in P2P flows involving both SIP and H323)
  • RCS
  • DOS protection
  • Any ASX to Any SBC (Any to Any)
  • NAT/NAPT
  • Endpoint CAC
  • Policers TG based, per VLAN
  • SIP Registration
  • SIP-SIP
  • ISUP – SIP-I
  • PER 4062 - Minimize Media Changes Flag changes
  • PCR 2366 - Trigger conditional transcoding when payload types are different
  • PCR 3289 Existing DOS protection testing
  • PCR 4042 - Maintain From tag for register after 401
  • PCR 4044 - ITX / TXA Spirou message support and logged in CDRs
  • PCR 5139 Add NAPT handling cases for which NAT usage cannot be determined

Problems Resolved

GSX 09.00.13R001 Problems Resolved

The following are the issues fixed in GSX V09.00.13R001.

Issue ID

Severity

Category

Problems Description

Resolution

GSX-55616

1

Coredump

MNS coredump occurs post upgrade from GSX 7.3.5 to 9.0.13 due to ICM leak in SIPFE module.

The code has been modified to avoid ICM leak in SIPFE module and prevent coredump.

GSX-55638

2

SIP

Unexpected reinvite after upgrading to 09.00.13R000.

The GSX behavior has been modified as follows: GSX sends re-invite for transcoded calls, only when the peer PSP contains DTMF, and, when there is no DTMF in NRMA selected entry.

GSX-55691 / GSX-55567

2

GSX

If ARS has blocked contact, later contacts in 302 redirect message are not being routed to.

If there is no port number received in contact header from 302 redirect message.GSX fetches blacklist contact with default SIP SIGNALLING PORT number. 

 

GSX 09.00.13R000 Problems Resolved

The following are the issues fixed in GSX V09.00.13R000.

Issue ID

Severity

Category

Problems Description

Resolution

GSX-53144 1 SIP Refresh Re-INVITE SDP offer does not mirror original SDP offer when IPSP flag is unselected. The code has been modified such that GSX sends all codecs in refresh Re-INVITE after switchover when PSX flag 'Send Selected Codec in Session Refresh' is not selected.
GSX-53208 1 Signalling MNS20 and MNS21 unable to coexist.
Unable to have a MNS21 in one slot and a MNS20 in another slot. Once the MNS21 becomes active, the MNS20 refuses to reboot until the card is upgraded.
Removed check of security and cavium chip logic to get rid of MNS20/MNS21 co-existence issue
as it is not required during Hard card upgrade/downgrade.
GSX-55091 1 DSP When a CNS85/86 is covering for a CNS80/81 after switchover, the excess DSPs are not removed from the pool. The code has been modified to remove the invalid/unwanted card specific DSP pools during the transition from standby to active in N:1 ratio for redundancy of cards CNS80/85:CNS85.
GSX-51061 2 GSX9K CIC deletion causes SGP link to bounce. While processing ASP inactive message, if routing context of the message does not exists in GSX, then the GSX sends the error message with invalid routing context. This caused SGX link bouncing.
This fix ensures that GSX does not send the error message if routing context of the message does not exist in GSX.
GSX-51077 2 Coredump Invalid pointer in CcAcctBuildPcdrInfoString() caused CNS60 coredump. Code has been modified such that if call state is CC_NULL_ESCR_NULL then GSX does not store the event CC_EV_ASG_MAKE_CALL in pending events list while call is tearing down. This fix prevents invalid memory access.
GSX-51811 2 SIP SIPSG - video DPM determination missing important logic. The code has been modified to support video port 0 if "Minimize relaying of media" flag is enabled.
GSX-52354 2 Media(Voice) RFC311 - UPDATE err handling violation. The code has been modified to give preference to NTWK PSP when there is change in both NTWK PSP and NRMA PSP.
GSX-52467 / GSX-54264 2 NBS9K SIP over TLS - GSX in TLS client role selects wrong local certificate. The code has been modified such that the DN is compared only if the certificate present in the certificate database and the configured client certificate in TLS profile are same. If they are not same, the next certificate is retrieved from the certificate database. This ensures that DN comparison is done only if the certificate name is same.
GSX-52563 2 Call Processing Active MNS slot 1 resets after executing TRUNK GROUP MODE OUTOFSERVICE. The code has been modified to introduce defensive check and to capture additional information/logs related to this issue in order to find out the root cause.
GSX-52820 2 GSX9K Called Number in RURI is missing on egress. The code is modified to read "Forced Local Ring Back Tone" from 6th octet of ingress parameter flags and "Enforce Max Digits for Valid Phone Number" from octet 7 of Ingress Parameter Flags so that both features work correctly. This ensures that if PSX is new (where it writes both the features on their respective correct octet) or PSX is old (Where it write both the features on 7th octet),GSX
behaves accordingly.
GSX-52824 2 Platform Due to a memory leak observed in CPC and SIPSG modules, multiple PNS cards were coring at multiple sites. The code has been modified to free an unfreed pointer to fix the memory leak.
GSX-52858 / GSX-54366 2 Platform Server module software failure in shelf 1 slot 5 with error code 0x00000fb7. This issue was caused when multiple active cards were configured to a redundant one, the Policer configuration gets overlapped and goes to standby which leads to PNS card crash.
The code is now modified to allocate separate memory for Policer on each active card to avoid overlap and crash.
GSX-52878 2 SIP OOB set to 1 by the GSX even though the peer does not allow INFO method. The code has been modified to set DTMF out of band (OOB:1) capability to TRUE/ENABLE only if peer allow INFO in the request, so that both peers can be in sync.
GSX-52913 2 CLI CLI command "snmp step <OID> ..." returns error when executed second time. The rList which stores IP related information is initialized wrongly to -1 which is causing the CLI responses for CLI step command to wrongly append -1 as well to the OID and thus causing the issue. The rList should be initialized with 0 for all the indexes that will not be used.
GSX-53010 2 GSX9K HPC Related Call Failure. For GW-GW calls ,when egress peer is SIP, there is a check for SG_TYPE and if SG_TYPE is GW, then the function returns TRUE and 182 is not being sent leading to call failure.
The code has been modified to set isHpcCall flag in SgFsmE15S6 function and in SG_TYPE check added another check if it is HPC call, then return FALSE.
GSX-53066 2 Coredump Software failure with error code 0x00000051 in slot 3 of PNS40. The code has been modified to fix the Quarantine timer failure.
GSX-53099 2 ISUP When egressing UKISUP, the IAM contains a PCLI in the message even though a CLI is present. The code has been modifed in Ss7LibProcessSendingIAM() to include PCLI in egress IAM message only if there is no CLI is present.
GSX-53168 2 SIP Cancel received by GSX is not finding DIALOG, GSX returns 481 transaction does not exist. The code has been modified to start the cleanup timer only if GCID in msg (received by ICM)and GCID in callBlockPtr are same.This ensures to not delete the callBlockPtr of the new call.
GSX-53297 2 Coredump Server module software failure with error code 0x0000112d in shelf 1 slot 3 of sjkgsx02. This issue was happening due to an uninitialized access in frame 5 and a dereference before null check in frame 6.
The code has been modified to initialize before access and also added a null check.
GSX-53324 2 Platform Memory leak: GSX hit memory congestion and becomes unresponsive in CLI and starts dropping calls. Code has been modified to fix the memory leak.
GSX-53383 2 GSX9K software failure occurred with error code 0x0000112d in slot 3 of PNS41. Added a check to verify the number of entries for Direct media.
GSX-53401 2 GSX9K Memory Congestion after upgrade from V09.00.06 F002 to V09.00.06 F007. Code has been modified to avoid duplicate entry in sipFeTGNameReplyWaitList and to handle the ICM Leak.
GSX-53413 2 SIP One way audio - PNS40 coredump analysis required. Data stored for Port info on the redundant card was reset and hence after switchover if the port earlier used for SRTP call is used for a non SRTP, it tried to decode/encode and failed.
Code modified to handle the Security Decoding/Encoding for Active/Standby cards.
GSX-53550 2 GSX9K NFS stuck in unmounting state. The code has been modified to send an ICM to all the respective tasks if NFS unmount fails with E_MNTOPEN so as to close all the files opened by the tasks.
GSX-53666 2 GSX9K NRMA modify request on already deallocated resource. The resChain is not made NULL since it checks the reference in later point deallocation and allocation.
GSX-53694 2 Call Processing Applying Advanced Routing Business Trunking traffic causes call outage. The outage was cleared after failing over the MNS card. The code has been modified to fix the DSP leak which was occurring when "Tone and Announcement Profile" and "Use LRBT for Ingress" was enabled in the configuration.
GSX-53695 2 Call Processing After the upgrading to GSX V09.00.06.F008 the memory utilization of the PNS cards keeps increasing. The code has been modified to remove the memory leak in GWSG which was observed in late media with direct media call scenario.
GSX-53822 2 GSX9K SIPSG mem leak causing CNS core.
System congestion, rejected calls and CNS coredumps due to a SIPSG memory leak.
Fixed the following memory leaks:
1. While storing the called party or calling party information into the sipCallHandle we are overwriting the puchLocalBuffer. Freed the previous buffer to clear this leak.
2. Modified few assignment operations to MemCpy.
3. Nice Call fix for GSX-53855 is also included.
GSX-53879 2 SIP GSX overwrites destination IP address in REFER + 3xx redirection scenarios. The code is modified to avoid overwriting of destination IP returned from PSX for 3xx with refer-to IP of initial refer with checking flag transferred due to 3xx.
GSX-53916 2 GSX9K Overlapping subnet route table issue. This issue was due to the changes made in GSX-52540 which has now been reverted and reworked to fix it.
GSX-53990 / GSX-54396 2 Media(Voice) LRBT resChain re-built - silence RTP. When UPDATE request comes with LRBT enabled, the tone gets halted and cannot hear the tone. The code has been modified such that when a Update request comes with LRBT enabled, it binds the parameters when the Allocated channel get activated and continues to play the tone.
With numcadence parameter set to 1, the tone plays uniformly.
GSX-54007 2 Media(Voice) ISDN to SIP transferred call - LRBT issue. The code has been modified to not reset the flag CC_CALL_PROG_TONE_PENDING_BIT when it is multiparty (specific states) so that LRBT is played.
GSX-54016 2 SIP GSX9K Bug- Adding phone-context=private when "Disable 2806 compliance" flag enabled. The fix avoids adding double phonecontext in RURI to egress based on behaviour of "Disable 2806 Compliance" and "Transparency of Request URI" flags.
GSX-54039 2 Call Processing DSP usage too high for call load. Reached 100% causing failures. The code has been modified such that DSP resources are allocated only when 'tone-announcement' and 'lateMedia' flag is TRUE in function NrmaSetupModifyServiceActions().
GSX-54042 2 Coredump Multiple cards were getting cored if there is ICM message corruption by one card. The code has been modified such that, the card which sends corrupted ICM message is located and coredumps only that card.
GSX-54058 / GSX-54778 2 GSX9K SIPSG and CC mem leak. In case of SIP I, bitbag allocated in Ss7LibPerformSendingUPV method is not freed.
To address this issue, the code has been modified such that only when clientidentifier is ISUP, bitbag is allocated.
GSX-54062 / GSX-54785 2 DSP Server module software failure in shelf 1 slot 8 with error code 0x00001040. The server module software failure was due to a bug in
DSP which caused it to crash in AMR-WB decoder while
processing received SID packet with Frame Quality (FQ) bit set to ‘0’ (BAD_SID) packet.
The code has been modified to address this issue.
GSX-54088 / GSX-54328 2 GSX9K LNS ROUNDROBIN no longer works after LNS is bounced.
If a Local Name Service configured with order as Round Robin bounces the LNS, the order is reset to default value Priority. Even though the CLI SHOW command shows Round Robin.
The code has been modified such that irrespective of the order the value is configured, after bouncing the LNS the order is maintained as Round Robin and does not reset to the default value of Priority.
GSX-54099 2 GSX9K MNS mem leak causes call failures. The code has been modified to free up the buffer to avoid memory leak.
GSX-54129 / GSX-54767 2 Coredump GSX rebooted with dumped a core file for MNS slot 1. The code has been modified to add a NULL check for the security parameter index(SPI) and return failure from the crash function if the SPI is NULL.
GSX-54136 2 SIP Double mem free due to missing port of SBX-10251 The code has been modified in SipSgPrepareContactsForCcRedirection() to null the cpcMsgInfoPtr pointer after freeing the memory.
GSX-54137 2 GSX9K The CLI commands to make a linkset inservice appears to have caused coredump and mns switchover. The code has been modified to check whether the route array is null or not to avoid coredump.
GSX-54157 / GSX-54399 2 PES GSX stops sending keepalive messages and policy queries to PSX. The code has been modified such that after the
server is reconnected and is active, the validity of keepalive timer is checked and start new keepalive timer by resetting timer. Also updated the code to restart the timer properly if the socket was re-opened, that is if disconnected by PSX and GSX will maintain another reconnect timer by opening the new socket connection.
GSX-54158 / GSX-54414 2 SIP T.38 (fax) - 200 Ok never (no) ACK'ed. In Packet-Circuit call, if PSP is configured with "fax failure treatment" as "continue", and "fax tone treatment" as "fax relay". The GSX sends a re-invite with T38 on detecting FAX tone. On receiving 200 OK with media port as '0'(Answer reject), it does not send ACK.
With this fix in place, GSX sends ACK for this particular scenario.
GSX-54182 2 SIP SIPSG - IP:port validation in SipSgGetOaEventFromSipMsgBody() - call rejected (Rejecting packet). GSX rejects the re-invite, if it contains the connection IP/media port that was actually provided by GSX.
This Hold SDP must not be rejected by GSX, if the DPM is inactive. Hence, added a check for media direction, before rejecting the packet.

Now, Hold SDP with same IP/Port and DPM inactive will not be rejected by GSX.
GSX-54319 2 GW gwCtrlPtr->sortedCrvListsPtr is null causing MNS core. After processing an open ACK message, the Gateway Forwarding Engine (GWFE) changes the state as “Connected” and clears the sortedCrvList.
The code has been modified to ignore the second open ACK message if the state is GWFE_SL_STATE_CONNECTED.
GSX-54377 2 DSP DSP - g729AB RBT from peer is malformed on VPAD side (PSTN to IP call). A saturation check for input samples is added at ALAW encoder to address this issue.
GSX-54382 / GSX-54426 2 SIP Porting SBX-41515 to GSX.
SBC does not receive routes from PSX and also LNP_INVALID script in the policy response. But there is an LI indication (Ingress leg as target). Since ingress is intercepted, SBC sends a SigStart message towards LI server try to access translated number from route and dumps core.
The code has been modified to prevent coredump.
GSX-54388 2 Coredump CNS86 coredumped 3 times. The code has been modified to ignore error code -1 from ECAN module.
As error -1 is benign and happens when ECAN module is trying to converge under load.
GSX-54432 2 GW-GW GSX gets egress SG_RELR_RALL_DIND state OA_NTWK_OFFER_RPY, 102. The code has been modified such that, during activation, before moving Acked PSP to Active, check if active is a tone context & prevAcked PSP holds a more recent PSP as compared to Acked. If its true, move the more recent PSP to Acked.
GSX-54464 2 SIP After LSWU and reboot - lots of SIP sig ports are OOS - incorrect binding Under certain sequences of events triggered by the messaging between PNS and MNS, the slave SIPFE on PNS did not report the port UP status change to master SIPFE on MNS. So the signaling service on that port was not enabled in GSX.
To address this issue, you need to switch over the cards.
GSX-54469 2 GSX9K GSX-52878 fix not working in all cases The code has been modified to address this issue.
GSX-54653 2 SIP Mem leak when running ISDN->SIP call load. The memory leak was observed in the ISDN->SIP call flow while sending the FACILITY toward ISDN. The REFCOUNT was not getting decreased, due to which the memory was not released.
This has been addressed by making changes in the code so that REFCOUNT becomes 0 and memory is released.
GSX-54793 2 Coredump CNS81 switchover occurred due to software failure "0x00000087" on TOKGSX02. The code has been modified to avoid double mem free which was resulting in a coredump.
GSX-54861 2 SIP GSX throwing syntax error for candidate attributes in SDP The fix resolves the following:
- Candidate optional attributes in INVITE is passing without throwing the syntax error.
- GSX can now accept Multiple optional candidate optional attributes.
GSX-54906 2 Coredump The rolling node reboots after upgrading to 9.2.6R0. The fix done as part of GSX-54099 is reverted.
Memfree is moved to higher frame to fix the double memfree issue caused by GSX-54099 and to resolve the memory leak.
GSX-54954 2 SIP SIP Signaling port status:OOS even state:enabled and mode:inservice Modified sipFeMsgProc.c to deallocate SIPFE sigPortPtr in Standby MNS to avoid referring old slot number when SIP signaling port slot number changes through CLI after MNS switchover.
GSX-55047 2 GSX9K Drying up CNS card results in loss of new calls - even though resources are still available on other slots. The changes are done to select the DRM based on card mode state which inturn selects the active INSERVICE card to allocate DSP's.
GSX-55050 2 GSX9K Detected "sonusRedundGroupSwitchOverNotification"(Reason: SOFTWAREFAILURE.) on SJKGSX08(Slot2 -> Slot1). The code has been modified such that MNS does not core while storing telnet session information. 
GSX-55126 2 SIP GSX includes wrong value of RPH in BYE requests The code has been modified such that "dsnValue" is assigned with "SIP_RPH_DSN_NOT_PRESENT" because it is an HPC call.
Also, a new CLI command "Send ETS In Response" has been added under SIP SERVICE RESOURCE PRIORITY PROFILE which will be set to disabled by default.
It should be enabled if "ETS" value needs to be copied to responses.
GSX-55263 2 GW-GW OTG value is incorrect when using GW2GW protocol between GSX’s. The code has been modified such that, the function to set OTG parameter in Egress SIP call leg checks if the call was a GW-GW call and the CPC - "CPC_OP_TGRP_STR" is available.
If yes, then display 'otg' parameter value from the CPC 'CPC_OP_TGRP_STR'  else display the Gateway's own inbound trunkgroup.
GSX-51943 / GSX-54262 3 CLI While configuring nexthop for static route, if nexthop and NIF/SIF IP address are in different subnet mask, the following error message is displayed: "error: The value is bad / value: 10". The code has been modified to display the correct
error.
"error: The NIF's IP address and nexthop can't be on a different subnet / value: ENABLED".
GSX-52218 / GSX-54497 3 SIP SIP Accept: header is not accepted if greater than 16 parameters. Increased the number of accept-parameters the system can handle by 50% (increased from 16 to 24).
GSX-52347 3 CLI The SHOW <port> PERFMON CURRENT FIFTEENMIN does not show any statistics or at least the "Elapsed time" still shows 0. Removed changes done for GSX-45279. Added changes in order to ensure the current interval level does not stop counting up.
Fix is to add code that zeroes out all the old PRM data received from the far end at the point where GSX missed 4 consecutive PRM messages and are setting the interval data false flag to true (the flag is a double negative). The added code also clears two FarEndEvent flags that are derived from the PRM data.
GSX-52555 3 CLI GSX allows bad value in Simulated Facility Group - show admin then cores MNS. The code has been modified to add a CLI check to validate the range of Facility Code Value (FCV). Range of facility code value should be from 0 to 7F (hex value).
GSX-53169 3 GSX9K GSX should ignore media connection address in 200 if received in 18x. A PSX flag is introduced - Ignore SDP After Offer Answer Completed in Egress IPSP, Egress Attribute,if : Enabled - GSX ignores the SDPin 200 OK if OA completed by 1xxRel.
Disabled (default)- GSX does not ignore SDP.
GSX-53239 / GSX-54470 3 SS7 SS7Lib Decoder always sets OBCI to in-band info available. The decodetool is not processing the inband indicator value in the OBCI parameter correctly, it is displaying the present value instead of the parameter value.
This has been modified to display the parameter value.
GSX-53409 3 GSX9K Incorrect error when NIF 'state enable' attempted on already enabled NIF. To address this issue for all scenarios, certain
conditions are removed from nrsComNif.c
GSX-53856 / GSX-54258 3 SIP SIPS - UasUpdateRequestCmd() - UPDATE no SDP results in 500. While processing UPDATE request method in offer answer state(SIP_SDP_OFFER_RCVD), added a condition to check for the presence of SDP body. If SDP body is not present, then GSX responds with 200 OK.
GSX-53911 / GSX-54260 3 Coredump Software failure in shelf 1 slot 15 with error code 0x00000051. While disabling the ISUP Service circuit state, the coredump happened due to NULL pointer.
To address this issue, the NULL pointer check for "pFsmData" pointer has been added for the processing of ISUP service circuit state change.
GSX-53986 3 GSX9K CNS71 core dump with error code 0x00000051. Added preventive fix in the code to address this issue.
GSX-54119 / GSX-54270 3 GSX9K CNA60 in slot 6 reporting CNA0XPWRFAIL after CNATEST in two systems. The code has been modified to perform verification of Cna06 during StopOnlineRedundancy test when there are rleDataPath failures.
GSX-54199 / GSX-54403 3 CLI GSX keep deleted entries in 'ip icb' even after deleting the SIP Signaling from CLI.
Details of SIP signaling port do not populate on executing the "ip icb" command.
Details of SIP SIGNALING PORT are added and deleted in NRS if host type is LOGICAL.
GSX-54686 3 SIP MAJOR report SipCallTerminateTimerExpiredCmd fires for each call The code has been modified to release the held dialogue after the subscription timer expires.
GSX-55333 3 DSP Data Sections are excluded from CheckSum. The code has been enhanced to  include DTMF Tables ("d2data" constant data section) for "checksum".
GSX-54257 / GSX-54434 4 H.323 Fail to OOS H323 Signaling port only if action force and dryup are used. The code has been modified to set the mode parameter from CLI command. The mode parameter has to be passed along with action in snmp set command.

 

GSX 09.00.12R001 Problems Resolved

This release addresses the inclusion of GSX release V07.03.04S115 for LSWU support.

Now, LSWU from V07.03.04S115 to V09.0012R001 is supported.

There are no issues fixed in this release.

GSX 09.00.12R000 Problems Resolved

The following are the issues fixed in GSX V09.00.12R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Issue ID

Severity

Category

Problems Description

Resolution

GSX-53249

1

SIP

GSX fails to pick the trunk group when PSX returns
Fully Qualified Domain Name (FQDN).

The code has been modified such that when FQDN in SIP server is configured, trunk group is selected based on the IP address and zone returned by PSX, instead of default trunk group.
IP Signaling Peer Group in PSX as well as IP NST in GSX for the egress trunk group must be configured with IP as 0.0.0.0 when FQDN is configured in SIP Server.

GSX-53257

1

SIP

GSX takes 60 seconds to respond to BYE in refer scenario - then clears the call 30 seconds later.

The queued DISC_IND which came before CUTTHRU/GETACTIVELEGS/MULTIPARTOP/CONNECT is dequeued once the CONNECT RPY is received.

GSX-52449

2

SIP

After enabling Local Name Service (LNS) and disabling DNS server, the DNS lookup fails.

The code has been modified such that when an LNS entry is enabled for a particular FQDN, the GSX deletes all records of this FQDN from the DNS client and agents cache. So that the new record populated in the DNS client cache has TTL=0 and DNS agent picks the new record from the client cache upon next call where a
DNS lookup is needed.

GSX-52546

2

GW-GW

SIP Globalization not occurring on calls originating on GSX.
PRI Calls received on GSX9000 sent to SBC 7000 via gateway to gateway signaling are not properly globalizing the SIP header sent from the SBC.  Other calls from the SBC are sending proper SIP globalized headers.

The root cause for this issue was a bug in the code which prevented globalization from working correctly. The JIP parameter was added when the parameter CPC_OP_TYPE_JURISDICTION_INFO is found in a message going to an older version of code (which overrides the CPC_GLOBALNUM_TYPE_CALLEDNUM).
The older version of code does not check to see whether the message already contains a CPC_OP_TYPE_GLOBALNUM_PARM parameter or not. Whenever this happens, any other CPC_OP_TYPE_GLOBALNUM_PARM may be ignored.

The code has been modified to add the JIP information to the existing CPC_OP_TYPE_GLOBALNUM_PARM (if there is one already in the message).

GSX-52955

2

SIP

GSX Local Name Service (LNS) - the new IP address is
not propagated to slot 14 after updating an LNS entry.

After changing LNS entry (changed IP address from 199.255.15.231 to 173.205.13.80), the slot 14 still shows the previous entry which leads to call failures.

Few agent lists maintained in cache were getting deleted during multiple switchovers of MNS.

The code has been modified to fix this issue.

GSX-53025

2

GSX9K

When GSX sends RFC 2833 in SDP offer but no RFC2833 in the response, GSX sends re-invite.

The code has been modified to not trigger re-invite even if the MLD is disabled.

GSX-53026

2

Coredump

CNS60 cored with the error code 0x00000051.

The code is modified to fix core dump in SIP 302 redirect case with diversion parameter.

GSX-53063

2

Signalling

Media Capture is not working when there is a fast answer, i.e. INVITE followed by 200 OK.

The code has been modified such that MCT works for SIPI fast answer scenario.

GSX-53094

2

SIP

Failed to add logical address in the PNA stack. error  0x5037.
After rebooting 9.2.3R0 release using 8.4.11F1 parameter file, the PNS switchedover due to excessive link failure and few SIP Signaling ports are not reachable from outside.

The root cause was a memory leak which occurred in route structure when we enable and disable the signaling port.

The code has been modified to fix this memory leak.

GSX-53121

2

SIP

AOR Registration failure after upgrading from 9.2.2R000 to 9.2.3R000.

The collision drop logic of user name for the refresh REGISTER that creates AOR is modified.

GSX-53135

2

SIP

SIP Message Manipulation (SMM) rule deletes characters from the end of  SDP.

The logic for calculating the length of the message body modified to avoid deletion of characters from the end of SDP.

GSX-53139

2

Coredump

High memory usage on on PNS/CNS and caused Slot-3 coredump with error code 0x00000074.

The code has been modified to fix the memory leak specific to GCR feature.

GSX-53357

2

H.323

In 9.2.3F1, only 1-way audio is available for calls traversing
H.323 trunk, the same call flow in 7.3.7S10 has 2-way audio.

The root cause was setting of IP address as 0.0.0.0 while exchanging the TCS message in H323SG. The code has been enhanced and an additional condition is implemented to overcome this issue.

GSX-53365

2

SIP

GSX does not include m=video line in the re-INVITE SDP when a 2-stage script is used.

The root cause of the issue is that the SDP is not merged in some places in stack when it is queued.
This fix merges the SDP before queuing in all places and reverts the resultant receiveSDP after merge to previous SDP so that the previous SDP out is sent.

GSX-53679

2

Platform

CRITICAL.EVLOG: Shutting down event file logging for log type (5) (reason: ADMINISTRATIVESHUTDOWN) on shelf 1. Automatic recovery timer is (0)!

The alarm has been removed to address this issue.

GSX-53823

2

GSX9K

CHBgsx08 slot3 PNS41 switchover.
In case of an ME timeout, the GSX does not crash to avoid any service interruptions.

The code has been modified in XRM to issue a crash in case of an ME timeout.

 

GSX 09.00.11R000 Problems Resolved

The following are the issues fixed in GSX V09.00.11R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Issue ID

Severity

Category

Problems Description

Resolution

GSX-39984

3

SIP

CDR Ingress IP Circuit End Point 0.0.0.0 for Full Late Media Call 100rel.

TO fix this issue a new function has been introduced in SIPSG which sends the Ingress IP to CC for late media calls.

GSX-43366

2

Statistics

Statistics 'Max & Total Packet Outage' are reporting abnormal value.

The debug logs have been modified to fix the issue.

GSX-51066

2

Platform

After upgrading, unable to add nif subinterface in NIFGROUP.

The code has been modified to properly sync up the nifgroup count after the GSX is restarted or upgraded.

GSX-51067

3

GSX9K

Doc change request - Request for removing some obsolete traps for ARS.

Few obsolete traps of ARS were deprecated.

GSX-51941

2

GSX9K

NRMA is unable to handle some specific call flow and NrmaDetermineModifyServiceCmdOfferActions() returns NRMA_ALLOC_REJECTED.

The code has been modified to handle egress reverse offer case of H323.

GSX-51963

2

SIP

SBC9000: CIT param is not exist on ACM MIME of 180.

The code has been modified to fix this issue, ported the Fix from SBX-10607 (SONUS00128838).

GSX-52001

3

CLI, Documentation

The CONFIGURE SIP SIGNALING REGISTRATION REGISTRAR command is deprecated and is no longer in use. But CLI still shows REGISTRAR configuration and deletion options.

The registrar related code has been removed.

GSX-52110

3

Coredump

GSX coredumps when extended ASCII characters are encountered in SIP signaling.

The code has been modified to fix this issue.

GSX-52395

3

GSX9K

CDR packets lost and packet outage sometimescontain random values.

The packet outage computation for single packet received cases has been modified to address this issue and also made some enhancements to media metrics monitoring code.

GSX-52423

2

GSX9K

Local egress Signaling IP has unspecIP instead of valid IP address in CDR

The code has been modified to convert unspecIP & nullIP to empty.

GSX-52434

2

SIP

ISUP to SIP Interworking for one diversion sets the wrong redirection reason. In an ISUP-> SIP interworking situation where the ISUP IAM contains the Redirecting Number, Original Called Number, Redirecting Reason and Original Redirecting reason for a case where one a single diversion has taken place, the Original Called Number and Original Redirecting Reason are used to map into the SIP diversion header.

The code has been modified such that when there is one diversion then the reason for diversion will be taken from redirection information.

GSX-52540

2

GSX9K

Enabling the fourth NIF interface leads to GSX rebooting and coring both MNS slots. In case of duplicate routes on different interfaces, only route with lowest preference is added to radix tree (which also adds radix mask to the mask tree) and remaining routes are maintained in the list. Masks for these routes in the list is not stored in radix node. So when the head route is deleted from radix tree and the next route is added in the list to the radix tree, the mask value or mask of zero is not being used to insert this. Hence these routes were not getting deleted.

The code has been modified and optimized to use the mask of head route when it is being deleted from radix tree while inserting the next route in this list, so that mask is preserved for all routes.

GSX-52624

2

Coredump

CNS60 Core dumps with error code 0x00000070.

The structure variable is made NULL after free, since it results in dangling pointer when it is used in respected functions.

GSX-52625

2

SIP-I

SBC9000 does not send 18x (ACM) until receiving 200 OK (ANM) from the succeeding office if INVITE (IAM) from the proceeding office is 64K unrestricted or SIP-Payload is clearmode.

Introduced a new flag for data calls. This flag is set for data calls when 18x is received.

GSX-52824

2

Coredump

Due to a memory leak observed in CPC and SIPSG modules, multiple PNS cards were coring at multiple sites.

The code has been modified to free an unfreed pointer to fix the memory leak.

GSX-52889

2

Platform

Server module software failure in shelf 1 slot 15 CNS 86 with error code 0x00000021.

The code has been modified to fix this issue.

GSX-52896

2

SIP

GSX-38486 breaks all VSC calls.The SIP trunk has "allow VSC enabled", but when an invite is received with * in the request, GSX sends back 400 (Bad request).

The code has been modified to fix this issue and removed changes done as part of GSX-38486.

GSX-52898

2

DSP

For IP to PSTN call, if a modem tone is detected during Inband DTMF generation it results in corrupted DTMF digits.

The code has been modified to fix this issue.

GSX-52918

3

DSP

Check-sum enhancement on VPAD.

The code has been modified to enhance the checksum repair functionality.

GSX-52924

2

CLI

Logical mgmt IP not working for SNMP requests.

The code has been modified to fix this issue.

GSX-52947

2

Coredump

CNS71 cored in ISDNSG.

GSX cored when it accessed an unexpected NULL pointer while printing a log message.
Code has been modified to fix this issue.

GSX-53052

3

ISUP

Invalid Location Number NPI prints wrong UPV log.

Checked the Numbering Plan Field and ensured that the selected Numbering Plan is Not "1" and also performed variant specific checks to fix this issue.

GSX-53117

2

SIP

GSX does not ACK an OK to a media lockdown re-invite when it has sendonly.

This issue is due to GSX holding the ACK for re-Invite if any media port change is encountered in the respective 200 OK. This should happen only when both the legs are SIP and should not happen when one of the leg is PSTN.Added a condition to hold the ACK only in SIP-SIP cases.

GSX-53153

2

Platform

asgUtils - Error 0x240f / asgMsgProc - Error 0x241d

Provided code changes for defensive fix to free ASG Blocks.

GSX-53154

1

Performance

BRM errors post LSWU cause GSX to slowdown and stop processing of new calls requests when NICE call recording is ON.

Fixed the problem in NP, the changes are tested with splitter enabled trans-coded active call load switch-overs.

GSX 09.00.10R001 Problems Resolved

The following are the issues fixed in GSX V09.00.10R001. The implications of not upgrading to include a particular fix have been provided where appropriate:

Issue ID

Severity

Category

Problems Description

Resolution

GSX-52896

2

SIP

GSX-38486 breaks all VSC calls.
In GSX-38486, the issue was GSX removes the * in the end of the INVITE request number and sends this invalid number to PSX. To resolve this issue, the code was modified such that if '*' is found in the number, call will be dropped with 400 BAD request.

Deleted the changes made based on GSX-38486.

GSX-52898

2

DSP

For IP to PSTN call, if a modem tone is detected during Inband DTMF generation it results in corrupted DTMF digits.

The code has been modified to fix this issue.

GSX-52918

3

DSP

Checksum Enhancement on VPAD.

The code has been modified to enhance the checksum repair functionality.

GSX-52924

2

CLI

Logical mgmt IP not working for SNMP requests.

The code has been modified to fix this issue.

GSX 09.00.10R000 Problems Resolved

The following are the issues fixed in GSX V09.00.10R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Issue ID

Severity

Category

Problems Description

Resolution

GSX-30 / GSX-52842 / GSX-52843 / GSX-52847 / GSX-52849

2

NBS9K

DSP crashes while handling G.729 media. The third party G.729 library was found to access an invalid memory region as a result of which BUSERR* kicked in eventually leading to a DSP core-dump.

(*BUSERR, a hardening feature available on DSP, detects any invalid memory access and freezes the corresponding DSP.)

Affected Cards : CNS-40/45/85/86, SPS80

This has been fixed by the third party vendor. And BUSERR has been re-enabled.

GSX-38486

2

GSX9K

When GSX receives INVITE request with * in the end of the number, it removes the * from the number and sends this invalid number to PSX.

The code has been modified such that if '*' is found in the number, call will be dropped with 400 BAD request.

GSX-45376

3

GSX9K

ISUP circuits can be created even if the service name is not correct.
Due to this issue, some PSTN trunk groups are missing in "PSTN TrunkGroup/ Report Used Ports" on EMS.

The code has been modified to choose the exact match of ISUP service name while configuring ISUP service and while creating ISUP circuits.

GSX-50778

2

GSX9K

In H.323 to SIP call, the DTMF interworking is failing on egress SIP side as GSX is sending DTMF either as SIP INFO or inband whereas the expected is RFC2833.
The DTMF type capability exchange happens in TCS-TCS ACK message exchange when the offer-answer is complete. So SIP will not be advertising the DTMF capability in the offer-answer which happened. When the minimize media relay changes flag is set, the re-INVITE for DTMF type will not be advertised and the call fails. The same thing happens when the DTMF type selected is either 2833 or OOB or both.

To address this issue, SgOaMinimizeMediaChanges has been modified to include the DTMF types - 2833 or OOB or both.

GSX-51076

2

GSX9K

GNS switchover causes refclk binding status go to INVALID, only delete/regrow or node boot can restore them.

The code has been modified to fix this issue.

GSX-51089

2

Coredump

FKHgsx01 CNS86 slot11 switchover with error code 0x112d.
The issue was with the FSM manager where the FSM instance table had a valid address but with NULL contents in it.

The code has been modified to check for the availability of data inside FSM instance table.

GSX-51821

2

GSX9K

RTCP NAPT learning not enabled. GSX is responding to RTCP but not to the NATTed Port thus customer's device is not able to receive RTCP response.  NAPT media and RTCP is enabled on the ingress call leg. But GSX uses remote RTP port + 1 as the remote RTCP port instead of learning the remote RTCP port from received RTCP packets.

The code has been modified so that RTCP packet will be trapped and RTCP port will be learnt.

GSX-51844

2

SIP

When making SIP calls over TLS w/ SRTP, the FAX tone switch to FAX (NRMA/SgFSM) is OK, however SIPSG sends INVALID SDP data that leads to call failure on peer receiving T.38 re-Invite. While populating the cryptos, GSX fills cryptos on the basis of configuration shared from PSX.

The code has been modified to check for FAX such that it does not fill cryptos to SDP received and proceed accordingly.

GSX-51950

2

SIP

Sonus's T.38 implementation is not converting received T30_IND messages into a 2100 Hz (CED) tone on the PCM side. Due to this issue it is not able to receive fax from many different locations.

The code has been modified to fix this issue.

GSX-51983

2

GSX9K

Random calls fail with "DRE Activation / DeActivation".
This is a race condition that is impacting calls.
When re-invite comes before the tone, resources are released, NRMA tries to apply modify offer on tone resources that are being released which leads to call failure.

The code has been modified in common SG to queue answer message towards ingress peer if tone or announcement resources are not released by NRMA. The answer message will be sent towards ingress peer after tone resources are released. As re-invite arrives after tone resources are released, modify offer gets processed successfully

GSX-52079

2

DSP

The GSX does not count RFC2833 (DTMF) packets in the received packet and received byte statistics.

The code has been modified to count the DTMF packets and bytes.

GSX-52085

2

DSP

For a scenario with dial tone present in the packet to circuit direction, and DTMF digits being sent in the circuit to packet direction, the NLP's non-linear processor is squelching (removing) some of the DTMF digits. Due to this issue the call set up will fail because not all DTMF digits will be relayed.

The echo canceller NLP was adjusted to allow digits to pass in this scenario.

GSX-52143

2

GSX9K

Marlin errors is triggering inbound call failures which is fixed by Coredump.

The code has been modified to capture more information for this issue so that It will give more inputs in fixing if it reoccurs.

GSX-52210

2

SIP

PNS CPU load increases over time and halves after coredump for the same call load. As CPU load goes higher than 85%, call setup is impacted.
SIP MM rules and IP route messages were causing memory fragmentation and effecting overall performance of the PNS when allocating and freeing memory.

The code has been modified for memory allocation of these messages to retrieve memory from existing allocated bin to avoid fragmentation.

GSX-52263

2

GSX9K

Routing using FQDN as SIP server does not work on GSX. When routing to a SIP server (NOT using iptg) on a GSX, if the SIP server has FQDN information, the GSX fails to find an egress TG.

The code has been modified to fix this issue.

GSX-52359

2

SIP

CANCEL Request should match the INVITE transaction it is trying to cancel. When a call is canceled prior to answer, but after an 18x is received, the CANCEL that is sent out by the SBC contains Route-headers.
Due to this issue, it is not possible to onboard customers on to the SBC platform.

The SIP stack is modified to store the INVITE routeset in the call cb and will be retrieved while sending CANCEL and ACK to a negative response.

GSX-52444

2

Media(Fax)

Memory fragmentation still persists after GSX-52210.
Cached Route, Cached ARP and Packet Service Profile entries were causing memory fragmentation and effecting overall performance of the PNS when allocating and freeing memory.

The code has been modified for memory allocation of these messages to retrieve memory from existing allocated bin to avoid fragmentation.

GSX-52462

2

Platform

SBC9K does not globalize the Request-URI in egress SIP-I

When UNDO LNP flag was set in PSX, globalization was not happening. The code has been modified to fix this issue.

GSX-52513

2

SIP

SIPSG is getting terminated by FM due to 99% load in CPU.

The dequeuing logic in SIPSG has been modified to avoid queueing and dequeuing the same message repeatedly thereby resolving the repeated queueing of SG_EV_CC_EST_RSP.

GSX-52562

2

Platform

SUBSCRIBE KPML message gets truncated if transport is TCP. KPML does not work when the "Event" header is placed before the "Content-Length".

The code has been modified to fix this issue.

GSX-52623

2

Media(Fax)

Parse error, if Warning Header with host port is received.

The code has been modified to fix this issue.

GSX-52793

2

GSX9K

PortFix GSX-40 : Logical Mgmt IP for SNMP Response

The code has been modified to fix this issue.

GSX-52422

3

GSX9K

DNS responses from DNS servers are intermittently discarded by the media policer.
The response from the DNS server is not handled by GSX leading to call failure.

The code has been modified to fix this issue.

GSX-52428

3

GSX9K

GSX is creating invalid CDRs due to which Netscore fails to load CDRs into the database.

The code has been modified to fix this issue.

GSX 09.00.09R000 Problems Resolved

The following are the issues fixed in GSX V09.00.09R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Category

Problems Resolved in GSX V09.00.09R000

Effect if Not Fixed

GW-GW

GSX-52234: GW-GW calls from SBC to GSX take an exceedingly long time to setup.

 

The code has been modified to fix this issue. There was a bug in the code that handles TCP Window Scaling under certain conditions.

A symptom of this bug was that it will cause GW-GW calls from SBC to GSX to take an exceedingly long time to setup.

Severity 1: Under certain unusual conditions, GW-GW calls from SBC to GSX may take an exceedingly long time to setup – or may fail to setup entirely.

GSX9K

GSX-50600: PNS coredump with error code 0xffffffff.

 

The code has been modified to handle creation of NAPT sessions.

Severity 2: Card fails over to standby.

GSX9K

GSX-50716: Calls fail after SIP-I [CON] message followed by [CPG].

The GSX discards the re-invite decoded CPG message (which is generated on successful call establishment between ISUP to SIP-I) and throws error.

 

The code has been modified such that GSX accepts the re-invite decoded CPG message and process the message without any errors.

Severity 2: Call failure and GSX throws error.

Core Dump

GSX-50926: Coredump in multiple slots.

 

To address this issue, the checksum repair functionality has been enhanced.

Severity 2: DSP triggers coredump.

Core Dump

GSX-51077: CNS60 coredump in CcAcctBuildPcdrInfoString().

 

Previous code modification was to avoid core dump and capture more information related to this issue.

 

Now the code has been enhanced to prevent generation of unnecessary logs.

Severity 2: Unnecessary logs are generated.

SS7

GSX-51816: REFCLK is used as a timing source even when it is DISABLED.

 

The code has been modified to fix this issue.

Severity 2: Calls get dropped or poor call quality.

Coredump

GSX-51850: Server module software failure occurred with error code 0x00000051 from slot3(PNS40).

 

To address this issue the "reschain" structure variable is made NULL after free.

Severity 3: PNS core dumps.

GSX9K

GSX-51852: GSX does not put a call on hold when datapathmode passthru is enabled on the SIP Service and the hold SDP has 0.0.0.0 IP address and a=sendrecv attribute.

 

The code has been modified with added condition sendrecv to check direction offered from INVITE and setting the direction to INACTIVE in SIP Service.

Severity 2: Issues for call on hold.

Call Processing

GSX-51854: Configurable solution for TDD.

 

The ‘CONFIGURE RESOURSE PAD SHELF 1 DTMFTDD ENABLED/DISABLED’ configuration is introduced to avoid filtering of TTY/POS analog modem.

Severity 2: The DTMF/modems tones are filtered off causing DTMF error.

Coredump

GSX-51890: Slot 10 coring 0d003801 - 0D03801 Generic Exception code - Marlin POST Failure.

 

The code has been modified to introduce a check for standby card where it will check dspNum which is calculated on the basis of received resId. If standby card receives the wrong dspNum beyond the numDsps limit for that card, it will print a log with the detail and free the msg pointer and return it. So that it will prevent the invalid memory access using wrong dspNum at the time of switchover and DID mapping.

Severity 2: If Fixed is not present. And any memory corruption happens  in CNS card due to which dspNum value changes and goes beyond the DSP limit for that card and switchover happens that time during DID mapping on Standby card, it will try to access invalid index of DSP context structure table using that invalid dspNum. Due to this invalid memory access, card will get crash if fixed is not present.

SIP

GSX-51906: Call flow hits a NRMA collision during reconfig for T38.

The code has been modified to fix this issue.

Severity 2: NBS is not responding to the new media port in the 200 OK, even though an ACK is sent.

ISDN

GSX-51964: A single ISDN call crashes the CNS card.

 

To overcome this issue the code has been modified to dynamically allocate the memory.

Severity 2: When an ISDN/ISUP to SIP call is made and the route contains an Fully Qualified Domain Name (FQDN),

CNS card core dumps during DNS lookup for FQDN.

Core Dump

GSX-51970: PNS41 cored error 0x51 when preferred codec flag is enabled and there is no match found in active psp and peer psp.

 

The code has been modified to fix this issue.

Severity 2: Calls get dropped when card is reset.

Call Processing

GSX-52077: Support for 2 stage calls with transparency.

 

The code has been modified such that while the announcement is being played, the cached message is not deleted before the re-invite reaches GSX.

Severity 2: Irrespective of the codecs and status of transparency flags, the 2 stage calls will fail with the error log “SipSgCopyTransparencyContent: Unable to fetch message from the cache”.

 GSX9K

GSX-52084: Re-Invite SDP offer/answer at the GSX.

A call is originated from web client of ICMN user through VNI. The initial SDP offer has audio and video. The SDP response indicates audio only, and call is set up. The web client triggers Call Move. The first step in Call Move is for WCG to put the far end (VNI PSTN in this case) on hold.  SIP reINVITE is sent and SDP offer is sendonly, but WCG includes non-zero port for both audio and video (it does not maintain the accepted SDP that included video port=0). In response to SDP offer with sendonly, if accepted, the other end must include recvonly (or inactive) in SDP answer.  Expected VNI would respond with recvonly for both audio and video.  Instead, it responds with non-zero audio port as recvonly, but video port=0 as sendonly. Since the SDP answer did not contain valid hold response, WCG fails the Call Move (the indication that there was a failure was not included in the tracker when created).

 

The code has been modified to send media direction as inactive for any rejected streams.

 

Note: This fix changes GSX behavior when it sends rejected streams in the answer with media port set to 0(when multiple streams are present). The direction attribute in the streams is marked as inactive in all cases. This also applies to active stream marked as sendrecv. But additional streams are marked as inactive with port set to zero.

Severity 2: Call is passed to VNI SIP PSX from an IIP SBC. PSX sends call to GSX. The messages sent between the IIP SBC and the GSX are not captured. The Re-Invite SDP offer/answer at the GSX is missing. The SDP visible at the GSX is same as at the edge of the IIP network.

GSX9K

GSX-52099: Parse error when boundary uses quotes.

When boundary parameter is sent with double quote in "Content-Type: multipart/mixed" of INVITE msg, 400 bad request is sent as parsing error instead it should accept the msg.

The code has been modified where Sip parser logic is made to accept double quote in boundary "BOUNDARYCHAR".

Severity 2: Call cannot be established.

Core Dump

GSX-52203: Multiple PNS core dump.

 

The code has been modified to avoid PNS core dump.

Severity 2: PNS card crashes with error.

Coredump

GSX-52237: PNS card software failure with error code 0x00000051.

 

The code has been modified to fix this issue.

Severity 2: PNS core dumps.

SIP

GSX-52187: GSX to support reuse of existing TLS Session ID by multiple connections.

 

The code has been modified to support reuse of existing TLS Session ID by multiple connections

Severity 2: Re-use of session ID of an active TLS session fails.

Routing

GSX-52233: DNS lookup lockdown function request.

 

To address this issue, a new flag USE CONFIGURED DNS SERVER is added. If this flag is enabled, the default server list configured in GSX is loaded.

Severity 2: Outbound calls will fail when a Fully Qualified Domain Name (FQDN) is configured as egress peer and Domain Name System (DNS) response contains a Name Server (NS) record of authoritative server that is not reachable.

Performance

GSX-50794: SNMP performance statistics fail at peak times.

 

The code has been modified to increase the performance statistics request timer.

Severity 3: Fails to generate SNMP performance statistics at peak time.

GSX9K

GSX-50797: SHOW ISDN SERVICE ADMIN command returns incorrect output.

 

The code has been modified to use Exact Matched Search.

Severity 3:  

Fails to show the requested ISDN Service.

GSX9K

GSX-50928: Network Processor Bad DISABLED DSP RID.

 

To address this issue, the Alarm level for DISABLED DSP RID is changed to MINOR.

Severity 3: The DISABLED DSP RID alarm gets displayed as MAJOR.

GSX 09.00.08R000 Problems Resolved

The following are the issues fixed in GSX V09.00.08R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Category

Problems Resolved in GSX V09.00.08R000

Effect if Not Fixed

Call Processing

GSX-51018: Unable to configure the second SIF. The following error is displayed:

“Different subnet mask found for overlapping ipaddress”.

 

The code has been modified to avoid overlapping subnet mask associated with the IP address both in NIF and SIF while configuring SIF.

Severity 2: Configuring the second SIF fails.

GSX9K

GSX-50804: When creating an output filter for DSCP markings, provisioning application received an error from the CLI, stating the filter cannot be created because it already exists; it states not found.

 

The maximum number of IP output filters is set to 64. Modified code to accurately report errors after the 64 limit is reached and properly update the filter counters on standby MNS.

Severity 3: Random failure on creation of IP output filters.

Call Control

SBX-22455: Number of current TCP connections appears to not decrement when the TCP session is closed.

The output of the following command will show an incorrect value for the number of active SIP over TCP connections:

SHOW SIP SIGNALING PORT 1 CONNECTION STATISTICS

 

Modified code to display the correct value for the number of active SIP over TCP connections.

Severity 3: The output of the following command will show an incorrect value for the number of active SIP over TCP connections:

SHOW SIP SIGNALING PORT 1 CONNECTION STATISTICS.

SIP

SBX-27267: The CDR field 122 was populated with wrong value for successful call (500 – Server Internal Error) giving false indication of call failure.

 

The code has been modified to fix this issue.

Severity 2: CDR indicates call failure due to wrong value in CDR field 122, even for successful call.

Coredump

GSX-50924: xNS cards coredumped for double memory free.

Added checks to avoid freeing the memory the first time if the call data has been copied to the CDR pending list for later processing.

Severity 2: xNS cards could coredump.

Call Processing

GSX-51128: When more than 20 Gateway Service Groups were configured, GW-GW calls failed after an upgrade from 7.3.7R10 to 9.0.6R0 release.

 

The code has been modified to fix this issue.

Severity 2: When more than 20 Gateway Service Groups are configured, GW-GW calls may fail after an upgrade from 7.3.7R10 to 9.0.6R0.

GSX9K

GSX-50909: Docs allow 23 characters for the PingPathName name, but the CLI output only allows 12 characters for the SHOW PATHCHECK PING PATH SUMMARY command.

 

The code has been modified to fix the issue.

Severity 3: SHOW PATHCHECK PING PATH SUMMARY Truncates PingPathName Column.

GSX9K

GSX-51008: The ISUPSG call history buffer size needs to be increased.

 

Code has been changed to increase the ISUPSG call history buffer size from 50 to 100.

Severity 3: IUP call flows are experiencing logging issues.

 Call Processing

GSX-51093: GSX REFER scenario – GSX did not include OriginatingGwInfo (OGW) and Originating Trunk Group in Policy Request, and caused PSX lookup to fail after the SIP REFER procedure gets initiated with GSX to transfer the call to the transferee.

Call failed because PSX returned NO_ROUTE due to the fact that it could not find ingress TG.

 

The code has been modified to fix this issue.

Severity 2: Call could fail.

Call Flows

GSX-44987: When DML flag was checked in IGRESS side, Re-invite was unexpectedly sent, which shouldn’t occur. The Re-invite was sent with G.729 codec, which was not supported in the egress side and therefore disconnecting the call. The codec re-ordering was happening due to Send Route PSP precedence flag which was enabled in the ingress side.

 

The issue can be resolved by disabling "Send Route PSP precedence" on PSPs where “Honor Remote Precedence” is enabled.

Severity 2: One-way audio. Customer faces audio issues.

 Media

GSX-51771: Call established during LSWU had no audio. A call that was established when active MNS was running on 8.4.11R0 and the active CNS was running 7.3.2 had no audio.

 

The code has been modified to fix this issue.

Severity 2: There is no audio during the LSWU period (2-10 seconds) when MNS and CNS card are in different build. Once the upgrade is complete, the issue no longer exists.

Call Processing

GSX-51135: CallSuccessWithDspResAllocated, CallSuccessWithDspResAllocationError and CallSuccessNoDspRes counters are incorrectly populated. CallSuccessWithDspResAlllocationError and CallSuccessNoDspRes counters were getting incremented even when they had a DSP resource in the chain.

The code has been modified to populate the counters correctly.

Severity 2: The effect is with the counters and not on the call. The DSP resource was properly allocated.

GSX9K

GSX-50908: Syntactically incorrect Contact header is received where the URI username indicates a password, but password is not included after the ":".

 

Parser has been modified to accept a blank password in userinfo field after “:”, sent in contact header of SIP-INVITE message.

Severity 2: INVITE message is getting rejected due to a parse error with the Contact header.

Media

GSX-50338: Setting DTMF to Either in the PSP causes passthru call to transcode when both ingress and egress support G711U-2833.

Severity 2: In effective usage of resources (DSP resources will be used in case of transcoded which could have been avoided).

GSX9K

GSX-50379: SIP signaling messages packets disappeared during peak times.

 

The code has been modified to fix this issue.

Severity 2: The corresponding FE entries are missing in IXP.

GSX 09.00.07R000 Problems Resolved

The following are the issues fixed in GSX V09.00.07R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Category

Problems Resolved in GSX V09.00.07R000

Effect if Not Fixed

SIP

SBX-27055: When Relay REFER is enabled, and REFER with embedded Replaces header is received, Sonus responds with “500” status. 

The code has been modified to fix the issue.

Severity 1: REFER fails.

Signaling

GSX-50575: IPv6 distributed SIP signaling ports do not work properly after node reboot because ARP entry is not populated correctly.

 

The code has been modified to fix this issue.

Severity 1: After a node reboot, outgoing calls from this SIP signaling port may fail until the ARP cache is updated.

Signaling

GSX-50715: After an LSWU, customer was unable to add NIF sub-interface in NIFGROUP.

 

The code has been modified to fix this issue.

Severity 1: Customer will be unable to add NIF sub-interface in NIF group.

Call Processing

GSX-50622: GW-GW calls which rely on Global Charge Reference information (GCR), in order to provide call correlation between different network elements, are failing from an SBC to GSX (if the GSX is running a version higher than 7.3.7).

 

The code has been modified to correct this problem.

Severity 1: GW-GW calls which rely on Global Charge Reference information (GCR), in order to provide

call correlation between different network elements, will fail from an SBC to GSX (if the GSX is running a version higher than 7.3.7).

Memory

GSX-50702: Memory leak occurred in NRS. The issue was that the interface was not detached properly when disabling a SIF.

 

The code has been modified to fix this issue.

Severity 1: Memory leak could occur in NRS.

Signaling

GSX-50701: pNA table does not have NIFGROUP ID138 whilst NRS has.

 

A SIP port consumed many FE XRES due to the pNA missing NIFGROUP index 138, which is the NIFGROUP that this SIP port is bound to.

 

The root cause of SIP port 800 having extra FE entries was the bug fixed by GSX-50161. The code has been modified to fix this issue.

Severity 1: SIP port will consume many FE XRES due to the pNA missing NIFGROUP index 138.

SIP

GSX-50657: SIP ports did not come online.

 

The issue was that SIPFE sent 3 ADD_LADDR requests to NRA while transitioning to active, but SIPFE only received one reply because NRA did not send a reply message for port 250 and 550 as their FE XRES were not at ALLOCATED state.

 

This prevented slave SIPFE from reporting port status to master SIPFE. The code has been added in NRS to reply to caller when FE XRES is at IDLE state.

Severity 1: SIP ports cannot come online.

Call Processing

GSX-50875: ICM buffer leak.

An ICM (inter-card message) buffer is left allocated (leaked) for each RTCP DSP Scaling trap (“Packet Loss Exceed” or “No RTP or RTCP Packets”) indication sent to the application code from the DSPs. This leak will occur whether or not trap generation is ENABLED.

 

The code has been modified to free the ICM message buffer after use.

Severity 1: Without this change any DSP server (MNS and PNS cards are not included in this category) configured to send such RTCP traps may leak enough ICM buffers such that call processing will fail and a coredump will occur.

Call Processing

GSX-50555: After unexpected EV_PRM_DEALLOC_RPY event CICS lock up and become unusable, thereby failing new calls.

 

The code has been modified to fix this issue.

Severity 2: All calls on the stuck CIC will fail.

Signaling

GSX-50564: Mixed IPv4 and IPv6 SIF - MNS switchover - no ARP entry for GW/Nexthop. ARP failure was due to the incorrect interface route bound to the SIF. Also found a bug in logical signaling address binding route where boundSlot was incorrect.

 

The code has been modified to correct these two errors.

Severity 2: ARP failure - no ARP entry for GW/Nexthop.

Coredump

GSX-45060/SONUS00158515: MNS card failed over while processing PATH MTU discovery ICMP packet.

 

The code has been modified to prevent a switchover during PATH MTU discovery.

Severity 2: MNS/PNS card may fail over while processing a PATH MTU discovery ICMP packet.

Call Processing

PSX-16836: When trying to run INVITE with Error Response suite on codenomicon, there were repeated failures and PSX SIPE process was changed.

 

The following changes are made as part of this fix:

  • The definition of YY_FATAL_ERROR is made common for PSX and GSX. It is currently supported only for non PSX.
  • Corrected the logical error to send out a failure when a “accept_contact” header is received.

Severity 2: SIPE process could change when trying to run INVITE with Error Response suite on codenomicon.

Call Processing

GSX-50316: New CV required for GW-GW connection failure to initiate call CrankBack at the ingress GW.

 

New CLI introduced to handle GW Link Failure. The following highlighted portion was added to the “configure crankback profile” CLI as part of the fix:

 

CONFIGURE CRANKBACK PROFILE <crankback_profile_name> REASON <reason_code> ADD LOCATION GWLINKFAILURE

Severity 2: Call drops due to GW link failures (GSX to GSX/SBC).

Call Processing

GSX-50549: T38 call setup failed, GSX reverted top G711A because of double detect of fax tones on the ISUP side.

 

The code has been modified to ignore the double detect from the circuit side. The double detect will be ignored and no INVITE with “silence suppression” Off will be sent to the peer side.

Severity 2: Calls could fail.

CLI

GSX-50661: GSX CLI allowed IPv6 SSP to be bound to different IPv6 subnet LIF. The following CLI error appeared:

 

“IPv6 SIP Signaling port should be in the same subnet as any of the LIFs in the NIFGROUP. / value: ENABLED”

 

The code has been modified to fix this issue.

Severity 2: CLI allows users to create IPv6 SSP port from different subnet than LIF, whereas this should not be allowed.

Signaling

GSX-50734: There was a discrepancy between Static route table and full route table.

 

The Static route table has different configured entries than IP route table.

 

The root cause is in the getNext routine which is used by both SNMP getNext and the show command. Since there are mixed IPv4 and IPv6 static routes in the list, the getNext function lookup method was incorrect and caused certain static routes missing from the display.

 

The lookup method is modified to return the correct route entry for the getNext operation.

Severity 2: Discrepancy in the configured entries between Static route table and IP route table.

Coredump

GSX-50735: Invalid MIME body with null terminator caused coredump.

 

The code has been modified to fix the issue.

Severity 2: Invalid MIME body could cause coredump.

Coredump

GSX-50752: System experienced multiple cores on CNS and PNS cards when received unsupported video (RTP/AVPF).

 

The core occurred due to invalid pointer access. The code has been modified to fix the issue.

Severity 2: Multiple cores on CNS and PNS cards.

Signaling

GSX-50530: Network outage was caused by some SIP ports not coming up due to the XRM out of FEres. This happened after the maintenance work on NIFs/SIFs - removal of SIF/NIF IPv6 nexthops.

 

The code has been modified to fix this issue.

Severity 2: Network outage could be caused.

SIP

GSX-50758: User-Agent hard limit was hit.

 

On receiving SIP ACK messages having more than 10 server-val parameters in the User-Agent header, GSX runs into a parsing error. RFC 3261 does not specify any such hard limit on server-val parameters of User Agent Header.

 

As part of this CQ, the MAX limit for accepting the server-val parameters in User-agent header has been increased to “20”.

Severity 2: Parsing error could occur when there are more than 10 server-val parameters in the User-Agent header.

Coredump

GSX-50533: PNS card coredumped when doing a redundancy switchover. The core was the result of receiving a bad CRV in a packet.

 

The code has been modified to fix a corrupted string when reporting a bad CRV received in a packet.

Severity 2: PNS card could coredump during a redundancy switchover.

Call Processing

GSX-50700: PSTN to IPv6 SIP calls failed as the INVITE did not leave GSX.

 

The issue was that the distributed SIPFE was not supported for IPv6. The support has now been added.

Severity 2: PSTN to IPv6 SIP calls could fail.

Signaling

GSX-50408: SIF was out-of-sync.

 

The code has been modified to properly handle multiple mode OOS and IS while SIF is at drying up state.

Severity 2: SIF can be out-of-sync.

 

Coredump

GSX-50664: When received REFER with embedded REPLACES but empty, core occurred due to invalid pointer access.

 

The code has been modified to fix the issue.

Severity 2: SPS cards can coredump with SOFTWAREFAILURE and with error code 0x00000051.

Call Processing

GSX-50653: Stable SIP-H323 calls failed following an LSWU.

 

The code has been modified to fix this issue.

Severity 2: Stable SIP-H323 calls can fail after LSWU.

Call Processing

GSX-50516: SYS ERR - Error 0x0 on ISUP to H323 call scenario. The max socket ID limit for a task was set to max system limit. This would lead to array bound overrun (which was fixed by GSX-45103), or SYS_ERR when the max socket ID came.

 

The per task socket ID limit is reduced to fix this issue.

 

Note: This issue is fixed for PNS40. PNS30 etc. are no longer supported (EoL).

Severity 2: SYS ERR - Error 0x0 on ISUP to H323 call scenario.

SIP

GSX-50833: GSX omits Contact header in “202 Accepted” for REFER message.

 

After GSX relayed REFER from the server, the 202 Accepted message from the server can be relayed, but the Contact header is missing.

 

The code has been modified to fix the issue.

Severity 3: Contact header will be omitted in “202 Accepted” for REFER message.

Call Processing

GSX-50644: TrmIpLookupTgSelectionTable: ipAddr is NULL.

 

After an upgrade, DBG log was flooded with “ipAddr is NULL” messages.

 

The call with NULL IP peer was established fine. Code has been added to downgrade the message reporting the remote address is NULL.

Severity 3: DBG logs could be flooded with “ipAddr is NULL” messages.

Call Processing

GSX-50571: After deleting a NIF group and its interfaces, and then creating a new nifgroup with different interfaces, a manual PNS switchover may result in the newly active PNS server coredumping as well causing a PNS outage for that redundancy

group.

 

The code has been modified to correctly handle mirroring of NIF group and interface data.

Severity 3: If this event occurs, stable calls will be torn down and an outage may occur especially if this is the only PNS redundancy group on the chassis.

 

CLI

GSX-50615: “show event filter shelf 1 slot 1 log CLI class all admin” command has format issue.

 

Removed the CLI option in SHOW EVENT FILTER and CONFIGURE EVENT FILTER commands as it is an invalid option.

Severity 3: Invalid command output.

Signaling

GSX-50834: When “Multiple Contacts per AOR” is disabled and multi IADs (same AOR) register at the same time, major log "RCB not found" was triggered and failed to mirror to standby. Also, there were some registration leaks on the standby.

 

The code has been modified to properly mirror the RCB and leaks on the standby.

Severity 3: Not able to mirror RCB on standby with major log "RCB not found" when registration complete and standby RCB may leak.

Configuration

 

GSX-50463: In the case of interface and sub-interface addresses being IPv6 only (all IPv4 addresses are NULL), a CLI error will be reported when the configuration is done out of order as follows:

 

With IPv6 configuration alone (i.e NIFs and SIFs configured with IPV6 only) % CREATE NIFGROUP NIF-IN % CONFIGURE NIFGROUP NIF-IN ADD INTERFACE GENET-1-11-1 % CONFIGURE NIFGROUP NIF-IN ADD SUBINTERFACE SIF-129 error: Overlap LIFs must be part of disjoint nifgroups % CONFIGURE NIFGROUP NIF-IN STATE ENABLED

 

For sub-interface (NifType==NRS_GEN_NIF_TYPE_SIF), validation for IP address was missing.

Added the relevant changes to validate IP address for the sub-interface.

Severity 3: Overlapping LIF error while creating NIFGROUP.

CLI

GSX-50585: Overlapping SIFs/NIFs CLI validation was not fully working.

 

The code has been modified to avoid overlapping subnet mask associated with the IP address both in NIF and SIF while configuring SIF.

Severity 3: Overlapping SIFs/NIFs CLI validation may not work as expected.

Coredump

GSX-50757: PNS card coredumped. The coredump was caused by accessing a null pointer.

 

A defensive check has been added for the pointer.

Severity 3: PNS card could coredump.

Coredump

GSX-50757: PNS card coredumped. The coredump was caused by accessing a null pointer.

 

A defensive check has been added for the pointer.

Severity 3: PNS card could coredump.

Call Processing

GSX-50122: MODIFY offer failed after successful REFER.

 

When MODIFY request comes, the call context is referring to a stale call leg, the call leg that was removed due to REFER. As a result, MODIFY offer fails.

 

The code has been modified to delete stale call leg and associate active call legs.

Severity 3: The ability to modify offer after successful REFER is impacted. 

GSX 09.00.06R000 Problems Resolved

The following are the issues fixed in GSX V09.00.06R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Category

Problems Resolved in GSX V09.00.06R000

Effect if Not Fixed

Signaling

GSX-45521/SONUS00162961:.MNS switchover due to link detection loss resulted in GSX loss of connection to SGX 4k. Customer had default route on packet NIF and MGMT NIFs.

 

When customer brought down MGMT port 1, GSX sent HEARTBEAT on packet NIF due to default route and this resulted in SGX sending an ABORT, and GSX closed the SCTP connection when it received an ABORT.

 

To address the issue, code has been modified to do route lookup based on source address, which would be MGMT NIFs in this case.

Severity 1: GSX could lose connection to SGX 4k and calls could fail.

Media

GSX-2635/GSX-45439/SONUS00161887: For INVITE without 100rel GSX established the two stage call without the Announcement or Digit Collection while late media was set to PASSTHRU.

 

“INTERNAL RTP GENERATION” flag which is based on 100rel received in INVITE was not reset after receiving SDP from peer. Code has been added to reset this flag.

Severity 1: GSX could set up the call without playing the announcement or digit collection.

Call Processing

GSX-50578: “REFER” was not working as expected in 9.0.

 

The issue is that TRM lookup finds default (catch all) IPTG, but reports wrong status. This triggers SIPSG to request PSX lookup for call transferring.

 

The code has been modified to fix the issue.

Severity 2: “REFER” may not work as expected.

Coredump

GSX-50428: Server module coredump. A caret (^) character caused a SIP parser watchdog timeout coredump when received in the FROM Header.

 

The code has been modified to fix this issue.

Severity 2: Both the active and standby server cards could coredump as multiple offending invites are received, causing a server outage.

Signaling

GSX-50355 (CQ 161797): XRM_RES_TIMEOUT after LSWU. All SIP-GSX-SIP calls got dropped. The root problem was that NRS did not actually send out the ARP request in the code, which caused PNS missing proper ARP entries.

 

Note: This only happens on some of the NIF/SIF and depends on how the LSADDR list was constructed.

 

The code has been modified to fix this issue.

Severity 2: Calls could fail after upgrade or PNS switchover.

Call Processing

GSX-45467/SONUS00162229: All ISUP calls failed just after MNS switchover and CIC status was mismatched.

 

The issue was that when the Standby SS7FE took the role of ACTIVE and when it tried to send the ASP_ACTIVE message towards SGX4K and further while sending the Application messages (i.e. ISUP message), it found that pConnMgmt pointer is NULL. After the code analysis, there were no scenarios where the pConnMgmt pointer is explicitly set to NULL. So one of the reasons for the pConnMgmt set to NULL is due to “corruption”.

 

As a precautionary measure, new trap (critical) is introduced when the Standby SS7FE assumes the ACTIVE role or when the redundancy messages are received from ACTIVE SS7FE and the pConnMgmt pointer is already NULL.

Severity 2: Calls may fail after MNS switchover.

Upgrade

GSX-50315: After upgrade to 8.4.10, quotes around boundary in Content-Type not allowed due to backward compatibility issues.

 

The code has been modified to fix the issue.

Severity 2: Calls may fail.

CDR

GSX-45587/ SONUS00159086: CDR field “Call Setup Delay”: Attempt: 172, Start: 153 and Stop: 198 could get populated with out-of-range values.

 

A timestamp was incorrectly being cleared in preparation for a possible crankback scenario. To guard against crankback scenarios, the logic is modified to no-longer accept timestamp updates once it has been set.

Severity 2: CDR field (Call Setup Delay) value could be outside the expected range.

DSP

GSX-50142: Upon reception of a compound RTCP packet DSP validates the packet before processing it. In case, if DSP receives partially corrupt RTCP packet, based on its payload data, it is quite possible that DSP enters an infinite loop due to the bug in the code which validates compound RTCP packet. This leads to either watch dog time-out or DSP health check (DHC) time-out and eventual core-dump.

 

Affected cards: CNS4x/8x, SPS80 when RTCP is enabled.

Severity 2: DSP could crash.

DSP

GSX-45435/SONUS00161850: The ingress peer negotiates G729 with PT “0” and continues to stream G729 media with RTP PT “0”. After few 100s of milli-seconds, it negotiates correct PT and now sends RTP media with PT 18 (static PT associated with G729). On DSP, this translates into a DSP channel being activated with G729 codec and compressed media PT as “0”. After a while, this DSP channel receives a “modify” command to change the Rx PT to “18” and receives G729 RTP media with PT 18. DSP crashes in the process. Incorrect handling in play-out buffer after PT change causes DSP to crash.

 

Affected cards: SPS70, SPS80, Transcoding resources on CNS85/6. After RX PT changes, playout buffer should be flushed to ignore any old un-played frames. The “PT” in playout should be set to the new Rx PT.

Severity 2: DSP could crash.

Call Processing

GSX-50343: In some Race conditions, SBX fax/modem failed to fallback.

 

The code has been modified to fix the issue.

Severity 2: Fax/modem calls could fail.

Media

SONUS00160025 (GSX-45246): T.38 Fax failed to send RTP packets. A high degree of echo on the line can cause false detection of V.21.

 

There are periods during the fax transmission, where the demodulator needs to look for both V.21 and page data. In this case, there was an echo of a prior V.21 signal that the demodulator assumed was an actual V.21. As a result, the relay assumed it was receiving V.21, and was unable to detect the incoming page data.

 

Logic is added to measure the power level of the first good received V.21 message. This power level measurement is then used as part of the decision criteria when running the V.21 detector.

 

Severity 2: Faxes can fail on lines with loud echo that cannot be adequately controlled by the echo canceller.

Call Processing

GSX-50326: GSX connects calls in “Inactive” or “Send-Only” state.

 

On Ingress IPTG, the "Minimize Relaying Of Media Changes From Other Call Leg" flag is checked and the "Relay Data Path Mode Changes To The Other Leg" is UNCHECKED. After receiving ACK, the re-INVITE sends out “ONHOLD” and when egress resumes the call, GSX still answers “ONHOLD.”

 

The code has been modified to not send re-INVITE ONHOLD on ingress and to answer re-INVITE OFF-HOLD properly on egress.

Severity 2: GSX connects calls in “Inactive” or “Send-Only” state.

Signaling

GSX-50132: IPv6 SIF/NIF has bad MAC of GW after node reboot.

 

The issue was that the IP address was passed instead of the MAC address for the existing MAC field and so the message showed “bad address family.”

 

The code has been modified to send out neighbor solicitation message to resolve the MAC address properly.

Severity 2: IPv6 SIF/NIF has bad MAC of GW after node reboot.

CLI

GSX-45394/SONUS00161414: GSX CLI showed a massive amount of duplicate records for SCTP associations.

 

The code has been modified to pick the first entry only when both local address and local port are both “zero”.

Severity 2: Massive amount of duplicate records for SCTP associations.

Memory Utilization

GSX-50329: High memory utilization across all PNS slots (Active and Standby).

 

There was a memory leak when issuing the command: "show ip netstat tcp connection table status" command when the TCP connection table is empty.

 

The fix is to free the memory when TCP connection table is empty.

Severity 2: Memory leak can occur when issuing the command: "show ip netstat tcp connection table status" command when the TCP connection table is empty.

 

 

 

Signaling

GSX-50311: NRS bug - interface indices NULL.

 

The problem was that the NIFGROUP interface data on xNS, mostly for SIF on CNS/SPS, was out of sync with MNS.

 

Under CQ134388, slot field was added in NIFGROUP interface data structure to resolve the issue on CNS/SPS since SIF does not exist on those cards by design. But if the user upgrades from an older release with missing SIF information on CNS/SPS to the release with CQ 134388, the current logic does not correct it properly.

 

The code has been modified to recover the NIFGROUP interface data on xNS during the upgrade if MNS has it.

Severity 2: NRS bug - interface indices NULL.

Upgrade

GSX-50377: NEXT HOP for TRUSTED and UNTRSUTED VLAN was not reachable after upgrade to 9.0.5. This lead to one way audio on all calls across a SIF. It was necessary to bounce the SIF in order to return service to normal.

 

In NIF FSM, NIF did not move to enabled but mode OOS SIF to proper state when processing XRM Reply messages.

 

The code has been modified to move SIF to proper state if it is enabled.

Severity 2: One way audio on all calls across a SIF.

Signaling

GSX-45526/SONUS00163003: PNS card coredumped with error code 0x0000112d.

 

The code has been modified to delete the cached SIP message after calling the SIP stack. This makes sure that SIP stack has access to correct pointers while populating the media information.

Severity 2: PNS card could coredump.

Media

GSX-45241: For a SIP to ISUP call, when the "Early Media Cut through" control on ISUP SERVICE is set to "DISABLED", the remote media was not getting played.

 

As part of this CQ, the cut-through direction flag is set properly when the first backward message is received.

Severity 2: Remote media is not played (example: RingBack Tone received from remote end) when the "Early Media Cut through" flag is set to "DISABLED" for a SIP-ISUP call.

Coredump

GSX-50478: Server module software failure in CNS card with error code 0x0000102d (SE_SMA_SWITCHOVER_TIMEOUT).

 

The code has been modified to remove the check for media profile based cards when updating the standby media profile.

Severity 2: Server module software failure in CNS card with error code 0x0000102d.

Signaling

GSX-50161/50158: SIP signaling ports remained out-of-service until NIF sub-interfaces were bounced.

 

The code has been modified to fix this issue.

Severity 2: SIP signaling ports will remain out-of-service until NIF sub-interfaces are bounced.

Call Processing

GSX-50473: For Incoming call connecting indirectly such as through PSX to GSX, a subsequent messages received directly from far end may cause GSX to pick up the wrong TG.

 

Similar issue for outgoing calls.

 

The code has been modified to fix the issue.

Severity 2: Some of the features may not work due to picking up the wrong TG.

Signaling

GSX-50105: GSX network side implantation processes the PROGRESS message stopping T310 and violating INS1500.

 

The code has been enhanced. When configured as network side, INS1500 processing of the PROGRESS message can be prohibited by configuring the remote switch type as “noprogress”.

conf isdn ser is152 remote switch type “noprogress”.

Severity 2: T310 may be inappropriately stopped.

DTMF

GSX-50515: The feature "Interwork DTMF OOB-2833 without Transcoding" was not working correctly in the 8.4 and 9.0 branches.

 

The code has been modified to correct this issue.

Severity 2: The feature "Interwork DTMF OOB-2833 without Transcoding" will not work correctly.

Call Processing

GSX-50365: Bad BRM status was reported and calls failed until the card was cored.

 

The issue was that the RID was not deactivated correctly which triggered all the RID_IN_USE errors when trying to activate it later.

 

BRM functions have been modified to fix this issue.

Severity 3: Bad BRM status and calls can fail.

Coredump

GSX-50488: CNS card coredumped with 0x00000051 error code. The coredump occurred due to a null pointer access.

 

The code has been modified to fix this issue.

Severity 3: CNS card could coredump.

Coredump

GSX-45500/ SONUS00162661: CNS card coredumped due to error code 0x00000074 (out-of-memory).

 

The code has been modified to fix this issue.

Severity 3: CNS could coredump with error code 0x00000074.

Logging

GSX-45189: Undesired MINOR log messages from DIAMC.

 

Added case for NRM_CALL_CLEANUP_CMD on standby message processing that silently drops this message to avoid printing error log.

Severity 3: Confusing log messages on normal condition.

Logging

GSX-45592/ SONUS00164730: DBG log is filled with errors - continuation of SONUS00149434.

 

The code has been modified to fix this issue.

Severity 3: DBG log could be filled with errors.

Call Processing

GSX-44989/ SONUS00157667: MasterTG: client not able to make call. Unable to debug in the field because logging is at minor level.

 

The code has been enhanced so that major level logging can be enabled by adding 1000 to the lob mask. For example, major level logging can be enabled for MTRGs 1, 2, 3, and 4 as follows:

  • admin debugSonus
  • mtrm debug 1007
  • mtrm mtrgLogAllowed 1 2 3 4

Severity 3: Debugging could be hampered.

Signaling

GSX-42981: Cause of No ID” parameter is not passed to PSX in the Policy request.

 

When GSX receives “Cause of No ID” parameter in ingress IAM, a DM/PM rule based on “Cause of No Id” can be applied on PSX. But the DM/PM rule failed because GSX did not pass the “Cause of No ID” parameter to PSX in the policy request.

 

The code has been modified to send the “Cause of No ID” parameter to PSX in the policy request.

Severity 3: DM/PM rule based on "Cause of No ID" will fail, since the parameter is not reaching PSX.

CLI

GSX-50351: Additional changes needed for "Need more meaningful error message for incorrect command sequence."

 

Note: This is additional change for CQ 158090/GSX-45021.

 

Added check for standby NIF to return meaningful error messages.

Severity 3: Generic error could be returned.

Configuration

GSX-50463: In the case of  interface and sub-interface addresses being IPv6 only (all IPv4 addresses are NULL), a CLI error will be reported when the configuration is done out of order as follows:

 

With IPv6 configuration alone (i.e NIFs and SIFs configured with IPV6 only)

% CREATE NIFGROUP NIF-IN

% CONFIGURE NIFGROUP NIF-IN ADD INTERFACE GENET-1-11-1

% CONFIGURE NIFGROUP NIF-IN ADD SUBINTERFACE SIF-129

error: Overlap LIFs must be part of disjoint nifgroups

% CONFIGURE NIFGROUP NIF-IN STATE ENABLED

 

For sub-interface (NifType==NRS_GEN_NIF_TYPE_SIF), validation for IP address was missing. Added the relevant changes to validate IP address for the sub-interface.

Severity 3: Overlapping LIF error while creating NIFGROUP.

Signaling

GSX-3269/GSX-45483/SONUS00162370: GSX cannot use Link-Local Nexthop for IPv6.

 

Added support to configure IPv6 link local address as Nexthop in static routes configured on MGMT NIF/NIF/SIF. For information on this feature, refer IPV6 Link Local Nexthop Support.

Severity 3: GSX cannot use Link-Local Nexthop for IPv6.

Performance Monitoring

GSX-45279/SONUS00160396: When the GSX stops receiving T1 performance monitoring messages, it continues to increment error counts that may no longer exist.

 

The code has been modified to stop incrementing error counts when the other end stops sending performance monitoring messages.

Severity 3: GSX may report far end UAS erroneously.

Show Screen

GSX-44877: ISUP circuit status screens showed blank columns when CICs were disabled.

 

The code has been modified to fix this issue.

Severity 3: ISUP circuit status screens will not work correctly when the relevant ISUP circuits are disabled. This results in incomplete CIC status.

Signaling - SIP

GSX-50532: NRS FE XRES missing after bounced SIF.

 

The problem is that when the NIF/SIF was disabled, NRS ONLY removes the associated “ifNum” if FE XRES is found. So the ifNum is not removed when FE XRES is missing. When NIF/SIF is re-enabled, NRS finds the ifNum in the LADDR's ifset and skips it.

 

The code has been modified to update LADDR's ifset properly.

Severity 3: NRS FE XRES could be missing after bounced SIF.

Coredump

GSX-50296: PNS card failed with error code 0x0000112d.

 

The code has been modified to validate the length before setting up pointers to the appropriate fields of the J7 CAR Structure.

Severity 3: PNS card could fail with error code 0x0000112d.

Coredump

GSX-45385/SONUS00161293: PNS switchover occurred with error code 0x00000074. Multiple memory leaks were found in the CC/SIPSG/SS7 areas.

 

Added codes to free up the leaking memory from CC (accountingInfo) and SIPSG/SS7 (Japan chargeInfo).

Severity 3: PNS could switchover with error code 0x00000074.

Call Processing

GSX-45589/SONUS00164566: Fast connect call is not working with Sonus provided Comfort Ring script.

 

Code changes are made in ASG script engine so that call control processes the “200 OK” message correctly in a fast connect scenario.

Severity 3: The call fails in a scenario where the Comfort ring script is played to the caller and “fast connect” is involved. No “200 OK” message is sent to the ingress after fast connect.

Signaling

GSX-45547: SS7 Node – SGX4000 parameter.

 

The new SS7 node parameter SGX4000 should only be enabled for server protocol M3UA, but by default it gets enabled when a node is created and then it is hard to disable it.

 

The code has been modified to fix this issue.

Severity 3: The new SS7 node parameter SGX4000 is DISABLED by default, but it gets enabled when a node is created and then it is hard to disable it.

 

Signaling

GSX-50535: XRM RES timeout problems revisited - follow on to GSX-50450/GSX-45427/CQ 161797.

 

The code has been modified to select proper interface with overlapped IP address.

Severity 3: XRM timeout can occur.

CLI

GSX-50534: When a configured active PSX is operationally DOWN and the configured standby PSX is operationally ACTIVE, the correct CLI “STOP” warning is not displayed when administratively configuring the configured standby PSX as Out of Service.

 

Enhanced checks to include Softswitch Operational state and Automatic/Manual Switchover configuration mode when determining which CLI warning to display.

Severity 4: GSX will not display STOP!! CLI warning even in some cases where services will be interrupted.

GSX 09.00.05R000 Problems Resolved

The following are the issues fixed in GSX V09.00.05R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Category

Problems Resolved in GSX V09.00.05R000

Effect if Not Fixed

Logging

SONUS00161837: Some CICs on a trunk group were stuck in a RELEASING state and new inbound calls were failing.

 

The code has been modified to fix this issue.

Severity 1: Calls could fail.

Signaling

SONUS00151014: LCQ 8088 Call is not crankedback when "Enforce Max Digit" flag is enabled.

 

Note: This is overlap-enbloc scenario with overflow number. Configuration in PSX for “Min digit for valid phone number” and “Max digit for valid phone number” in prefix-profile was not being processed properly in GSX. Also there was a bug because of which "Enforce Max Digit" flag is disabled in PSX, but considered “enabled” in GSX during TRIGGER request response.

 

The code has been modified to fix the two issues.

Severity 1: Calls could fail.

Signaling

SONUS00161584: For transcoded call and multi codecs, SIP Session Refresh response “200 OK” may have wrong codec entries.

 

The code has been modified such that the first codec entry matches the first codec offered by the peer.

Severity 1: Peer may teardown the call.

Signaling

SONUS00160790: GSX intermittently sent re-INVITE without changing Min-SE from “422” response. As a result, there were call failures.

 

The code has been modified to fix the issue.

Severity 2: Calls can fail.

Call Processing

SONUS00153319: SIP STACK issue with UPDATE. GSX sent 200 OK INVITE right after UPDATE, causing peer misbehavior.

 

The fix is to wait for the peer to respond to “UPDATE” before sending “200 OK”.

Severity 2: Calls could be disconnected due to no audio.

GW Signaling

SONUS00164754: GW to GW calls could result in a coredump during codec negotiations.

 

The code has been modified to zero out a memory pointer after being freed to avoid attempting to free it twice or access it again.

Severity 2: Server cards may core.

Signaling

SONUS00161797: XRM_RES_TIMEOUT after LSWU. All SIP-GSX-SIP calls got dropped.

 

The root problem was that NRS did not actually send out the ARP request in the code, which caused PNS missing proper ARP entries.

 

Note: This only happens on some of the NIF/SIF depends on how the LSADDR list was constructed. The code has been modified to fix this issue.

Severity 2: Calls could fail after upgrade or PNS switchover.

Signaling

SONUS00150781: GW-GW Fax calls failed due to a race condition.

 

The egress peer is very fast in recognizing the CED tone and re-Invites immediately after the ACK (voice call established). This causes the ingress GSX to release call as T38 PAM was received from egress GSX before the ingress GSX was able to serve MCS ANM (answer).

 

The code has been modified to fix this issue.

Severity 2: Certain Fax calls can fail.

Call Processing

SONUS00160992: Direct Media calls may not work correctly after a switchover when there is more than 1 Direct Media Profile configured. (This problem does not affect call that arrived after the switchover.)

 

The code has been modified to fix this issue.

Severity 2: Direct Media calls may not work correctly after a switchover when there is more than 1 Direct Media Profile configured.

Signaling

SONUS00158330: NRMA double modem tone detection.

 

The following changes are made as part of this fix:

 

  • On detection of Modem tone, GSX transitions to FALLBACK, further if the modem tone is detected once again, it should be ignored. This logic has been implemented in SIPSG as part of SONUS00053709. Same logic has been implemented for GWSG.
  • Added logic in NRMA to ignore Modem/Fax tone detection event if:

 

  1.  Fax/Modem tone is detected for the first time and call is already found to be transitioned to fax.

 

i.e.: Switch to fax was due to Re-Invite and then a fax/modem tone is detected.

 

  1. Double modem or fax tone detection events.

 

Severity 2: FAX fails to proceed using T38 protocol.

Redundancy

SONUS00158972: GSX sent spurious re-INVITE while the transaction was in progress, i.e. before ACK w/SDP was received. In case of a late-media call, where 100rel is not supported in the initial INVITE, it would not be possible to play any announcement/media.

 

The code has been modified to not initiate any announcement events.

Severity 2: GSX can send spurious re-INVITE before ACK w/SDP is received for late media when 100rel support is not enabled.

Call Processing

SONUS00160052: H323/GWSG issues -The “Progress Indicator” field in a PROGRESS message sent to the ingress peer contained an incorrect value.

 

The issue is that the GSX is receiving a PROGRESS from egress peer with Progress Indicator of 4, but it sends a PROGRESS to the ingress peer with Progress Indicator of 1 (instead of 4).

 

The code has been modified to fix this issue.

Severity 2: If the GSX/SBX receives a PROGRESS message from the egress peer in which the Progress Indicator field is set to 4, the GSX/SBX may send a PROGRESS message to the ingress Peer with the Progress Indicator field incorrectly set to 1.

Call Processing

SONUS00159829: Multiple cards cored with error code 0x00000021. The issue is that peer sends an Invite with large number of codecs and repeating FMTP lines that trigger memory overwrite boundary when calculating checksum.

 

The code has been modified to ignore the previous FMTP line and use the latest one.

Severity 2: Multiple cards can coredump.

Coredump

SONUS00159943: For multimedia, and after call ONHOLD, when the GSX receives a re-INVITE without SDP, it responds with a 200 OK still ONHOLD, causing peer not able to initiate OFFHOLD.

 

The code has been modified to fix the issue.

Severity 2: One way audio.

Signaling

SONUS00160642: PATHCHECK SIP OPTIONS messages over TCP not working.

 

The code has been modified to support PATHCHECK over TCP.

Severity 2: PATHCHECK might not work over TCP.

Signaling

SONUS00159815: Unable to cleanup ISUP call.

 

The code has been modified to fix this issue.

Severity 2: Repeated attempts to cleanup call results in extensive logging.

Signaling

SONUS00147445: Due to possible Race condition, SIP to CPC cause mapping may be incorrect.

 

The code has been modified to handle the Race condition properly.

Severity 2: SIP to CPC Cause Mapping may incorrectly map SIP causes 300-306.

Signaling

SONUS00159538: SRTP was not working for single leg calls. GSX sent “AVP” instead of “SAVP” in case of late media for TG, which is configured as secured RTP.

 

The code has been modified to fix this issue.

Severity 2: SRTP might not work for single leg calls.

Performance Management

SONUS00162785/160482: Custom SPE script had an issue with “18x” with respect to SDP not clearing resources.

 

Call control code has been modified to clear the resources properly after the script is played.

Severity 2: The script works but under a light load, this can result in call failures.

Call Control

SONUS00150246: Re-Transmission of non-INVITE message was being treated as new transaction.

 

The code has been modified to fix this issue.

Severity 2: The system incorrectly treats the retransmitted ingress SIP MESSAGE as a new request, and generates two SIP dialogs.

Media

SONUS00157346: SIP allowed a carried code beyond the max length of 5. For example “cic=123456”.

 

This has been fixed. Now domestic codes with more than 4 digits (.i.e. cic=12345) or international codes with more than 5 digits (.i.e. cic=+112345) are discarded.

Severity 2: Tools to read the CDR stream may truncate CDRs.

Signaling

SONUS00161757: When GSX initiated call with video and peer response video port 0, GSX triggered re-Invite even when “Only Selected in Refresh Codec” was ON.

 

The code has been modified to fix the issue.

Severity 2: Unnecessary re-Invite after call gets connected on egress.

 

Billing

SONUS00151260: The value of "Call Setup Delay" field in the CDRs was not correct.

 

The expected behavior is that the call setup delay timers must be logged properly and within the defined range.

 

The issue was that there was a huge delay from the server, so the latency timers overflowed and exceeded the maximum possible range as documented. Overflow checks have been added before logging to the CDRs. So, in case of an overflow in the value of the timers, only the maximum values are logged.

Severity 2: Value of "call setup delay" field in the CDRs could be incorrect.

Signaling

SONUS00161674: For video call and call “on-hold,” late media Re-INVITE responses had video inactive.

 

The code has been modified to fix the issue.

Severity 2: Video call leg will always be “on-hold.”

Signaling

SONUS00161630: GSX-2374-LCQ - Port of PCR3667 to GSX (End2End ACK).

 

Refer End to End ACK for information on this feature.

Severity 3: Billing issues could be caused.

Call Processing

 

SONUS00160959: Under high load, re-INVITE may be sent before ACK due to ICM message drop internally. Peer could not handle it and responded with error.

 

The code has been modified such that ICM message will not drop internally.

Severity 3: Calls can fail.

Call Processing

SONUS00147204: GSX tears down the call after 491 Glare condition. The Retry timer did not stop properly when received “491” response.

 

The fix is stop Retry timer as soon as new request is received so that subsequent requests will not hit the invalid state.

Severity 3: Calls could fail.

Call Processing

SONUS00151758: “SHOW ST SOURCE” command display was incorrect on GSX4k.

 

The code has been modified to set the ST status correctly for GSX4k.

Severity 3: Command display can be incorrect.

Show Screens

SONUS00148041: Under certain conditions the Current Enforcement Counts returned by the following CLI command may be incorrect: SHOW DBL SIP_SERVICE “sipIpSrvcGroup” STATISTICS

 

The code has been modified to fix this issue.

Severity 3: Current Enforcement Counts returned by the CLI command may be incorrect.

Statistics

SONUS00158090: Need a more specific error message than generic for incorrect command sequence.

 

This CQ is to replace the generic error messages with meaningful error messages.

  • Defined one new COM error message for the mode in-service case.
  • Modified NRS COM code to use the new COM error for mode change and to use an existing error message for state change.

 

Severity 3: Generic error could be returned.

CLI

SONUS00161002: GSX may register two PSXs with same session ID. This may prevent the GSX from accurately identifying which PSX it is communicating with.

 

Added logic on GSX to reject PSX registration responses if a duplicate sessionId is found. This triggers the PSX to change sessionId value for following registration requests.

Severity 3: GSX cannot accurately determine which PSX it is communicating with. This may affect PSX statistics and registrations.

Signaling

SONUS00161840: Unexpected T9 expiry messages filling the logs. The issue was that the function just restarted the timer and did not check the state or if the timer is already running.

 

The code has been modified to verify that the timer is running before starting it.

Severity 3: Lots of unexpected T9 expiry messages could fill the logs.

Billing

SONUS00151611: LCQ 8450: Placeholder for code checkin.

 

Refer Need alarm off dynamic blacklist entry for information on LCQ 8450.

Severity 3: Since there is no alarm generated and no log entries unless on INFO level logging, there is no way to easily get the status of a dynamic blacklisted entity.

Signaling

SONUS00153611: “SHOW RESOURCE PAD USAGE SHELF 1 SLOT ALL PERFORMANCE STATISTICS CURRENT" showed Slot 0 as a valid DSP resource, whereas it is invalid.

 

The code has been modified to fix this issue.

Severity 3: Slot 0 can be shown as a valid DSP resource, whereas it is invalid.

CLI

SONUS00160574: SLS rotation violates JT-Q763. Currently, bottom bit of the SLS is rotated to the top and the other 3 or 4 bits are shuffled down by one bit before sending it to the SGX4K.Current GSX implementation of shuffling the last bit of CIC to create SLS is violating the JT-Q763.

 

A new CLI command is added that overrides the existing SLS rotation functionality:

% CONFIGURE SS7 NODE a7n2 SLSROTATION DISABLED ENABLED

The default setting is “Enabled”. This setting preserves the current operation.

 

Note: If the customer wants to disable M3UA SLS rotation, they may set this new CLI variable to “Disabled.”

Severity 3: SLS rotation violates JT-Q763.

Protocol Conformance

SONUS00144638: SNMP OID for the Redundancy Group Status Index returns incorrect index.

 

The code has been modified to return the correct value (redundancy group index).

Severity 3: SNMP OID for the Redundancy Group Status Index returned all zeros (i.e., instead of actual indices).

SNMP

SONUS00153686/162175: Stable IP to GW-GW calls were dropped during switchover of MNS and during LSWU.

 

The code has been modified to fix this issue.

Severity 3: Stable IP to GW-GW calls could be dropped during switchover of MNS and during LSWU.

Logging

SONUS00159603: The following error messages in the SYS log did not have enough data to investigate the issue:

 Error 0x3a30 Line 15495 File /software/src/marlin/SIPSG/sipsgUtils.c.

 

The code has been modified to log as Major level in DBG along with GCID for better investigation.

Severity 3: SYS log is meaningless.

Signaling

SONUS00160699/156218: Protocol Variant showed SIP in CDR for GW-GW call.

 

The Egress Protocol Variant Specific Data (PVSD) field in the CDR for some GW-GW calls may have some incorrect information in some cases due to a race condition when the first route attempt is not successful and the call is cranked back. The code has been modified to fix this issue.

Severity 3: The Egress Protocol Variant Specific Data (PVSD) field in the

Call Processing

SONUS00159721: When using a mixed redundancy group and redundant card is active, SHOW RESOURCE PAD SUMMARY Compression Utilization statistics was wrong. The issue is due to the mixed redundancy groups with a CNS card containing DSPs backing up one that does not contain them.

 

The fix is to initialize the value correctly for the redundant card.

Severity 3: When CNS card is redundant, resource pad summary can be corrupt.

Billing

SONUS00162773: REFER CDR changed after upgrade causing CDR mediation issue. 

 

After a call is referred the information recorded in STOP:36 and STOP:53 may not be correct.  Software has been modified to correct this issue for only when REFER is initiated from the egress leg.

 

Note: REFER initiated from Ingress Leg will be addressed in a future release.

Severity 2: CDR fields STOP:36 and STOP:53 may not contain correct information after REFER.

Performance Management

SONUS00160516: GSX did not report the performance statistics that were added in 9.0 (PER5270), after an LSWU.

 

A timer is created to make sure the statistics are updated when standby card is active.

Severity 3: Performance statistics added in 9.0 (PER5270) might not be populated.

GSX 09.00.04R000 Problems Resolved

The following are the issues fixed in GSX V09.00.04R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Category

Problems Resolved in GSX V09.00.04R000

Effect if Not Fixed

Signaling

SONUS00159270: “503” response from Registrar was not able to relay to IAD.

 

The code has been modified to fix the issue.

Severity 1: “500” response is sent to IAD.

Call Processing

SONUS00159093: On egress GW, if a SIP call transfer takes place to ISUP or to the other GW, the call may fail.

 

The code has been modified to fix this issue.

Severity 1: Call can fail.

LSWU

SONUS00159180/ 148332: The “defaultEmptyNifGroup” was not generated after LSWU. The expected behavior is that regardless of the parameter mode being BINARY or DISABLED, defaultEmptyNifGroup should always be seen in version 8.x after the upgrade. This may result in a logical signaling address binding failure.

 

The code has been modified to fix this issue.

Severity 1: Calls may fail after LSWU completion.

Call Processing

SONUS00159203: When the only administratively active PSX is administratively changed to out-of-service, the CLI warnings displayed are not severe enough to provide proper warning.

 

The current CLI warning is replaced:

From:

WARNING: Services WILL be interrupted if this command is executed

 

To:

STOP!! If this action is confirmed, ALL new incoming calls WILL FAIL on the entire GSX system

Severity 1: CLI warning is easily mis-understood and ignored.

Signaling

SONUS00159250: DSPs which encounter high jitter and packet loss can leak DSP buffers, eventually leading to DTMF Inter-working failure. DSP maintains a global pool of digit buffers in the form a linked list. There is a head pointer and a tail pointer.

 

When a new digit buffer is requested by a channel, the digit buffer pointed to by head of the list is given. Likewise when a digit buffer is released, tail pointer is decremented. Under some conditions it is possible that a channel may get released prior to all the egress digit buffers being played out affecting the total number of buffers available in the global buffer pool. The defect is not freeing the buffer correctly, so effectively the number of digit buffers available in the global pool reduces over time. If this scenario repeats, it will eventually result in a condition where there are no digit buffers available. If no digit buffers are available, no inter-working can be done.

 

Affected cards are:

  • CNS6x/4x/8x
  • SPS 7x/8x

 

The code has been modified to fix this issue.

Severity 1: DSPs which encounter high jitter and packet loss can leak DSP buffers, eventually leading to DTMF Inter-working failure.

Signaling

SONUS00157308: Some IP endpoints remained blacklisted and did not recover automatically.

 

The code has been modified to prevent IP endpoints from getting stuck in blacklisting forever. Further, defensive code has been added to force automatic recovery of endpoints whose internal recovery algorithm data may be missing.

Severity 1: In very rare edge condition circumstances, a blacklisted IP endpoint may not recover automatically.

Coredump

SONUS00156848: MNS failed over with 0x00002302 during an upgrade. The number of default ISUP Signaling profiles in Rel8 and upward has increased by 2, but the value of SIG_PROFILE_INDX_MAX (128) did not increase accordingly. This resulted in the coredump.

 

The SIG_PROFILE_INDX_MAX value has been increased to 130 to accommodate the additional two default ISUP Signaling profiles.

Severity 1: MNS can coredump with 0x00002302 during the upgrade.

SS7/CLI

SONUS00152051: MNS switched over multiple times continuously. It is possible for an out-of-service/disabled SS7 node to contain a NULL pointer reference to a Gateway.

 

When in this configuration state, the following CLI show screens will cause the MNS to core:

  • SHOW SS7 NODE <node> STATUS
  • SHOW SS7 NODE ALL STATUS

 

This bug only affects STATUS screens because the software that processes the status request was missing a NULL pointer check. A NULL pointer check has been added to fix this issue.

Severity 1: Multiple MNS cores could cause an outage.

Configuration

SONUS00156281: There was no logical bearer IP address after upgrade to 8.4.5F007. The root problem was that the logical bearer address binding entries were deleted on standby MNS when processing the PNS failure NOTIFY.

 

The code has been modified to insure the logical bearer address binding entries are updated correctly.

Severity 1: No logical bearer IP address after the upgrade.

Signaling

SONUS00158619: Registered user transition calls failed with “503” during LSWU. During LSWU for multiple distributed SIPFE pairs, received an INVITE on a slot of older version and sent a SETUP to a different slot of higher version where registered AOR reside. The higher slot version failed to do conversion and dropped the message. Subsequently “503” was received.

 

The code has been modified to fix the issue.

Severity 1: Registered user transition calls could fail with “503” during LSWU.

LSWU

SONUS00156788: LSWU disabled Early Media. The LSWU code from pre-8.x to post-8.x did not initialize the “EarlyMediaCutThru” flag to ENABLED as intended.

 

This issue has been fixed. LSWU initiating from 8.x code or later will leave the flag setting unchanged.

Severity 1: The “EarlyMediaCutThru” flag will get reset to DISABLED for LSWU from 7.x to 8.x or later. This may result in the failure of ring back tone to the ingress peer.

Signaling

SONUS00155304: One way audio was observed when the refer leg response was on hold in the initial 200OK.

 

The code has been modified to fix the issue.

Severity 1: There can be one way audio during a refer call on hold.

Call Processing

SONUS00157783: High Probability Of Completion (HPC) Limit Max rate issue. HPC Call Limit policer failed to work correctly for all values and failed completely for configured call rates less than 30 calls per minute. Also, the CLI command created errors in the SYS logs when making changes to the call limit.

 

The code has been modified to fix this issue.

Severity 1: Without this fix, the HPC limit feature does not work and will not allow any HPC calls when configured for call rates less than 30 calls per minute.

Media

SONUS00157817: VPAD discards SID or CN packets with payload size greater than 10 bytes.

 

For information on this feature, refer GSX discards SID or CN Packets with Payload Size Greater than 10 Bytes.

Severity 2:

Coredump

SONUS00158927: PNS coredumped in NrsSigHandler().

 

The check to validate system wide sockets supported on PNS card was incorrect. This resulted in out of bound socket array access, leading to the crash. Made necessary changes to address the issue

Severity 2: It could lead to PNS crash when the number of sockets opened on PNS card reaches max limit supported.

Signaling

SONUS00159529: If late media configured passthru, and “Minimize Relaying Of Media Changes From Other Call Leg" flag is checked, GSX may not able to send 200OK to the late media re-INVITE.

 

The code has been modified to fix this issue.

Severity 2: Call may be torn down by peer.

Signaling

SONUS00156004: XRES and NRMA failed when trying to bring TG into service. The root cause of the failure was that NIF/SIF data out of sync between MNS and PNS triggered by IXP ME timeout.

 

The code has been modified to properly sync up PNS with MNS.

Severity 2: Calls can fail.

Signaling

SONUS00155932: SIPSG ROUTEMESSAGEVALIDATION bug. In some cases, SIPSG internally clears the call due to failed ROUTEMESSAGEVALIDATION.

 

The code has been modified to fix the issue

Severity 2: Calls could fail.

Coredump

SONUS00158587: SIP stack writing to low memory area 0x30. Corrupted SIP message might cause SMM overwrite to lower memory.

 

The code has been modified to fix the issue.

Severity 2: Card could core.

CDR

SONUS00153659: NCR data was cleared from CDR after crank back or redirection. NCR data should remain in call if it was set by ingress call leg.

 

The code has been modified to add restrictions to when NCR data is cleared after call redirects.

Severity 2: CDR fields for Network Call Reference in ATTEMPT:163, 164 and also in analogous START/STOP/INTERMEDIATE records may be inaccurate.

Call Processing

SONUS00156177: In overlap to overlap call scenario, when the called party released the call, the CDR’s ‘EXTRA CALLED NUMBER ADDRESS DIGITS’ field got cleared unexpectedly. However, if the calling party released the call, the CDR field was populated correctly.

 

The code has been modified to fix this issue.

Severity 2: CDRs might not be processed correctly.

Coredump

SONUS00159656: CNS card coredumped while performing Continuity Test (COT) after an upgrade.

 

Crash happens when:

  1. A regular voice channel is disabled while in the middle of playing/receiving DTMF digits. This causes DSP to purge DTMF buffers.
  2. A COT channel is then enabled on the same DSP channel which experienced #1. When this channel is now disabled, the DSP crash occurs.
  3. When a voice channel gets disabled the used DTMF buffers gets released to global pool. However the channel linked list "take" and "put" pointers are not reset. These pointers are actually initialized while activating a voice channel.
  4. When a COT channel gets activated the DTMF related put and take pointers does not get initialized as DTMF does not get used. However, while de-activating the channels a common routine is used which purges DTMF buffers.

 

Since the channel linked list take pointer is non-zero, the unused DTMF buffers are purged again leading to a circular linked list scenario and DSP never returns from this loop. This leads to health check/watchdog timeout and eventually core-dump.

A fix is now implemented which checks for only voice channel to purge DTMF buffers.

 

Affected Cards : CNS6x/7x/8x

 

Severity 2: Coredump can occur while performing COT after an upgrade.

Call Accounting

SONUS00150538: The issue was that during a "bulk" call release, the processing of call accounting data can be purposefully delayed. The bug is that some of the memory required by call accounting was freed before the accounting data was processed.

 

Added logic to not free the memory required by call accounting until the data is processed.

Severity 2: Coredump could occur.

Signaling

SONUS00159606: DTMF were not being passed.

The INVITE going out to the egress peer has G729/G711 with DTMF events.  The PSP has G711 configured with no RFC 2833.

 

SBC changes for CQ 135423 are ported to take care of this issue.

Severity 2: DTMF not being passed.

Media

SONUS00158395: This problem occurred when the DSP was set up for a transcoded call with EVRC on one leg. When the DSP received some erroneous G.711 packets on the EVRC leg, it switched to G.711 mode, and treated the EVRC leg as a G.711 leg for the duration of the call.

 

With this fix, the EVRC leg will not switch to a G.711 leg even if G.711 packets are received on that

Severity 2: EVRC calls could fail.

 

Signaling

SONUS00158330: NRMA issue (modem/fax related), where the customer double-detected the modem tone (ANSam/) within its duration as the resChain is rebuilt in the middle by peer (ingress SIP) re-INVITE w/ T.38.

 

On detection of Modem tone GSX transitions to FALLBACK (i.e. depending on Modem tone treatment) . Once transitioned, subsequent Modem tone detection events should be ignored .Logic is updated to do the same.

Severity 2: FAX can fail to proceed using T.38 protocol.

Call Processing

 

SONUS00159787: GSX coredumped because 200 OK was corrupt with some invalid data.

 

Error parsing has been modified to fix this issue.

Severity 2: GSX can coredump due to invalid SDP.

CDR

SONUS00156928: GSX generated abnormal CDR records. The issue is that when a CDR size is approximately 1800 bytes or greater, the CDR data is divided into multiple messages by accounting subsystem before sending it to EVENTLOG subsystem.

 

The EVENTLOG subsystem transmits its memory buffer when it reaches the configured queue size. The configured queue size limit can be reached after receiving a CDR fragment and therefore data transmitted to file could contain a CDR fragment.

Normally this is not bad. However, if a file should roll at this time, then the last CDR of the file closed and the first CDR of the new file opened will contain a CDR fragment. Added logic in EVENGLOG to hold off transmitting its accounting memory buffer if the last message in the buffer is a CDR fragment.

 

The buffer will be transmitted as soon as EVENTLOG receives the remaining CDR fragments.

Severity 2: GSX could generate abnormal CDR records.

Configuration

SONUS00159287: Softswitch state was stuck in “DRYUP”. 

 

The code has been modified to add a timer when marking the Softswitch into “DRYUP.”  After specified time, mark Softswitch into the “DOWN” state.

Severity 2: GSX will show the Softswitch in state DRYUP and no lookups will occur to that Softswitch.

Signaling

SONUS00159969: In a Gateway-Gateway scenario, the HOLD/RETRIEVE  INVITE was being relayed to the other leg despite the "Minimize relaying of media changes from other call leg" and "Disable media lockdown" flags being checked and "Relay data path mode changes to the other leg" unchecked in the IP Signaling profile.

 

The code has been modified to fix this issue i.e. HOLD/RETRIEVE INVITE will not be relayed if the above mentioned flags are set accordingly for a Gateway-Gateway scenario.

Severity 2: In case of Gateway-Gateway scenario, the HOLD/RETRIEVE Invite would have been relayed despite “Minimize relaying of media changes from other leg” and “Disable media lockdown” flags checked and “Relay data path mode changes to the other leg” unchecked in the IP Signaling Profile.

Signaling

SONUS00153765: GSX did not process re-INVITE received from SBC in GW-GW scenario.

 

The code has been modified to convey "silenceSupp: off" towards egress (in re-INVITE) if it is being received from ingress SBC (in re-INVITE) and it is considered as VOICE_MODE call. This is done for established call. SBC release version should be 4.0 onwards.

Severity 2: Incorrect GSX behavior.

Signaling

SONUS00158351: IPSEC could not be established when remote is a subnet and not a single host.

 

The code has been modified to add logic to separate address and subnet address for local-ID and remote-ID.

Severity 2: IPSEC can fail when remote traffic selector is a subnet.

CDR

SONUS00157787: Latency CDR statistics were not being populated.

 

The code has been modified to calculate the Latency and publish in CDR.

Severity 2: Latency CDR statistics is not populated.

Signaling

SONUS00158392: Possible ICM usage leak – inUseCount was excessive and did not recover.

 

The code has been modified to fix this issue.

Severity 2: Memory leak and ICM leak can occur, which could cause call failures when it reaches the maximum.

Signaling

SONUS00158162: SPS cards ran into MAX CCB across all cards. However, resets of the cards resolved the issue. In some cases, the call did not release resources properly when terminating.

 

A fix is to add failsafe timeout to clean up the call.

Severity 2: Memory leak could occur.

Signaling

SONUS00159262: Memory leak occurred in a CNS card. Memory is leaked when:

 

  • the ISUP service is configured with revision INDIA
  • there is an EGRESS call that gets the CRG message
  • and CRG message propagation is disabled (PSX signaling profile).

 

The code has been added to free unused memory when the CRG is not propagated.

Severity 2: Memory utilization up to 84%.

Signaling

SONUS00157569: Disconnect Treatment Profile on Ingress gets associated- MNP call. There was a conflict between existing behavior and new MNP call flow.

 

While initializing the Disconnect Treatment for the ingress leg configured with Disconnect Treatment Profile, additional check is performed to make sure that the Disconnect Treatment is not executed when the “JAPAN MNP TRANSFER” control is ENABLED and the egress end is SS7, i.e. primarily not to perform the Disconnect Treatment for MNP transfer calls.

Severity 2: MNP (Transit calls) can fail.

Coredump

SONUS00153412: MNS coredumped due to software failure. SIPFE CCB was leaking and flooded with ICM messages during call audit.

 

The code has been modified to fix the issue.

Severity 2: MNS could coredump.

Fault Management

SONUS00144377: EMS was not parsing sonusArsEndPoint trap event description properly. Alarms were wrongly displayed on FM after the upgrade.

 

Added zoneName to EVENT description and trap but never wrote the function to get zoneName from non-MNS cards.

Severity 2: Monitor facilities could be impacted.

Signaling

SONUS00158317: During ISUP-SIP interworking, if the received IAM contains an invalid parameter and the PSX uses a DM rule to modify Called Number NOA (Nature of Address) from 0x78, the “NO OBCI FOR SIP APN” Signaling Profile option fails.

 

The code has been modified to correct this behavior.

Severity 2: OBCI could be sent to PSTN switches (when it is set in the ISUP Signaling Profile) when it is not expected to be sent.

 

Signaling

SONUS00155642: Path check code relies on SIPSG to determine transport. SIPSG was not accounting for TCP TLS operating on one port higher than TCP for path-check options pings, whereas it did make the adjustment for call traffic INVITE, etc.

 

The code has been modified to check transport type and increments sin_port on the basis of transport being TLS.

Severity 2: Path check will not work over TLS.

Call Processing

SONUS00158659: REFER from SIP to SS7 failed.

 

The code has been modified to allow the call flow to work correctly.

Severity 2: REFER from SIP to SS7 can fail.

Call Processing

SONUS00158086: SIP-GW-GW-SIP calls fail when “Accounting Send Egress CDR Info On Receipt” is set to TRUE in the Accounting configuration. This issue was due to a bug which prevented any backwards messages from being sent from the egress GSX to the ingress GSX.

 

The code has been modified to fix this issue. A workaround for this problem is to set “Accounting Send Egress CDR Info On Receipt” to FALSE.

Severity 2: SIP-GW-GW-SIP calls can fail when “Accounting Send Egress CDR Info On Receipt” is set to TRUE in the Accounting configuration.

Protocol Conformance

SONUS00156828: If there is a SIP-I to SIP call that is established without “180 Ringing”, and the call is put on hold by the egress leg with an INVITE, the INVITE propagated to the ingress leg includes an ACM (a protocol violation). This problem is not observed if “Minimize Relaying Of Media Changes From Other Call Leg” is selected in the PSX “IP Signaling Profile”.

 

The fix is for the INVITE to include a CPG (instead of ACM).

Severity 2: The protocol violation may result in the call being released.

Call Processing

 

SONUS00155685: Second re-INVITE gets incorrectly changed Media Attribute.

 

The code has been modified to call a new function, which populates the Receive Payload Type. The function compares the selected codec with the codec entries passed by SG/NRMA and populate the receive PT with the matched entry.

This is done for non-transcode call flows because NRMA was not setting the Receive PT for the call in the selected codec entry. Note that the fix for CQ148545 is removed.

Severity 2: The re-INVITE has incorrect media attributes and peer might tear down the call.

 

SIP

SONUS00157346:  SIP allowed a carried code beyond the max length of 5.  For example “cic=123456”.

 

This has been fixed. Now domestic codes with more than 4 digits (.i.e. cic=12345) or international codes with more than 5 digits (.i.e. cic=+112345) are discarded.

Severity 2: Tools to read the CDR stream may truncate CDRs.

Signaling

SONUS00158214: Users could not retrieve logical bearer address with “snmpwalk” in 8.4. This is due to a bug introduced in the NEXT function by the Ipv6 changes.

 

The code has been modified to correct the problem in the NEXT function.

Severity 2: Users cannot retrieve logical bearer address with “snmpwalk” in 8.4.

Signaling

SONUS00157322: GSX returned SNMP sonusDspSlotStatG711Total value zero.

 

The problem is that CLI command shows G711 Total value of 3634. However, the corresponding value of sonusDspSlotStatG711Total in dwGSXDspSlotStats EMS table shows as 0.

 

There was a mapping issue in EMS due to a confused MIB description.

 

The following changes were made to minimize the effect of MIB:

  1. In GSX: Changed the MIB description for sonusDspSlotStatG711Total
  2. In EMS: Changed the header column name “G711TOTAL” to “G711NOECHOCANCEL”, and “COMPRESSION_TOTAL” to “G711_COMPRESS_TOTAL”.

Severity 2: Wrong counter for G711 DSP PAD resources.

Protocol Conformance

SONUS00151213: RFC 5806 not followed for a single diversion header.

 

For information on this feature, refer RFC 5806 not followed for a Single Diversion Header.

Severity 3:

Call Processing

SONUS00150273: TBCT call flow transcoding fails between in-band and 2833.

 

For information on this feature, refer Call Flow Transcoding fails between Inband and 2833.

Severity 3:

GSX Tools

SONUS00159151: Verify annfile and piftool scripts fail on the linux DSI servers.

 

For information on this feature, refer Porting GSX Tools (PIF Tool and verifyAnnFile) to Lintel platform.

Severity 3:

Call Processing

SONUS00158141: Development of PCR 8172.

 

For information on this feature, refer AIS TIMER Call Drop.

Severity 3:

Signaling

SONUS00148337: RADIUS Identifier Limitation.

 

For information on this feature, refer Lawful Intercept, RADIUS message buffering limitation removed.

Severity 3:

Media

SONUS00158476: PCR 8787 – Check-in into 8.4.x branch for 8.4.9R0.

 

For information on this feature, refer Support for MAJOR logs to Notify Media Failure due to unreachability of Nexthop.

Severity 3:

Signaling

SONUS00158696: The ISDN feature “Call Rate Limit Release Cause” fails when the incoming SETUP has a Channel ID IE marked “exclusive”.

 

When the call rate goes above the value of “Maximum Ingress Sustained Call Rate”, some calls will be rejected and the disconnect reason is expected the configured “Call Rate Limit Release Cause”. That is the case if the Pref/Excl flag in the Channel ID of the received SETUP is “Preferred”, but if the flag is “Exclusive” (no B-channel negotiation), the disconnect reason is “44”, which is not desired.

 

The code has been modified to fix this issue.

Severity 3: A connected switch will not realize that the GSX has reached the call rate limit.

Coredump

SONUS00157713: A Race condition in SIP code caused a PNS core.

 

The code has been modified to handle the Race condition properly and to prevent the core.

Severity 3: A Race condition in SIP code may cause a PNS core.

Signaling

SONUS00156866: Japan ISUP feature “Sequential GRS” broken. The reset of transient calls after MTP suspend/resume was not being done when the Japan ISUP feature “Sequential GRS” is enabled.

 

The problem occurs when GRS applies (.i.e. when there is a range of CICs to be reset). Individual CICs are properly reset. The code has been modified to fix this issue.

Severity 3: CICs might be unavailable after an MTP suspend/resume.

Coredump

SONUS00157553: CNS card coredumped and eventually caused a healthcheck timeout.

 

The issue was caused by an ICM buffer leak in DS where the ICM message was not freed up properly from the pending ICM list in case of failure and hence depleting the ICM buffers pool. Added codes to free the resource properly.

Severity 3: CNS card could coredump.

CDRs

SONUS00119737: Crankback ATTEMPT records was logging incorrect value for Route Attempt Number field.

 

The code has been modified to print correct Route Attempt Number in ATTEMPT record.

Severity 3: Customer that use the 'crankback attempt' capabilities for traffic engineering and troubleshooting purposes may be mislead by the data

Configuration

SONUS00158144: PNS card cored due to Healthcheck Timeout while the customer was reconfiguring Trunk Groups.

 

The code has been modified to fix this issue.

Severity 3: Enable or Disable of a Trunk Group may cause a PNS core in certain rare circumstances.

Fault Management

SONUS00156009: Despite setting the following traps to ENABLED, the traps were not sent when a module reset with a coredump.

  • coreDumpNotGenerated
  • coreDumpGenerated  

 

The code has been modified to fix this issue.

Severity 3: No coredump traps will be sent.

CLI

SONUS00158780: When using DIRECTION option, no output was displayed on SHOW IP NETWORK SELECTION table.

 

Added code to correct the problem in the lookup logic.

Severity 3: No output may be displayed on SHOW IP NETWORK SELECTION table.

Coredump

SONUS00153736: PNS card coredumped after upgrade. Root cause of the core is a double memory free. This happened due to the registration data not being removed properly on the standby when the registration was removed from the active.

 

The code has been modified to ensure that there is no double free memory.

Severity 3: PNS card can coredump.

Signaling

SONUS00158942: PNS switchover due to software error – lost logical binding.

 

The root problem was that if standby MNS has no logical bear address binging entry before LSWU, then the logical bear address binding will not be recovered after LSWU. Added logic to recover logical bear address when L3 NIF/SIF switches over.

Severity 3: PNS could switchover due to software error.

Common Agents

SONUS00153607: snmpgetbulk command for sonusSctpAssocTable generated a SYS log error indicating that it “Cannot read TCP table. The sonusSctpAssocTable only applies for MNS and PNS cards and hence failed for other cards.

 

Added codes to block non-MNS/PNS cards from accessing this table.

Severity 3: snmpgetbulk of sonusSctpAssocTable can fail with error 0x502d.

Call Control

SONUS00156918: Change the default value of calltriggercount from zero to non-zero (10) in order to limit the number of traces by default. If the user sets the value to zero, the tracing will go on indefinitely as it does now for triggerCount = “0”.

 

Changed the default value of call trace triggers per slot from zero to 10 to prevent operator error from leading to indefinite call trace logging in the event the operator enters a wild card number or something that causes excessive logging.

Severity 3: The default value of zero will indefinitely record traces to the TRC log for the calls that meet the configured criteria for that call trace.

CLI

SONUS00152751: After an LSWU from 7.3, output of show mgmt nif command may show uninitialized IPv6 addresses as an empty space instead of using colons to represent no address configured.

 

The code has been modified to fix this issue.

Severity 3: The output of show mgmt nif command can be different than expected.

Fault Management

SONUS00148028: The description of the alarm "sonusEvLogShutdownReason" is missing the explanation of reason (11) for field in the GSX_8.4.1_AlarmTS_Guide_1.

 

The code has been modified to use the existing sonusEvLogShutdownReason (1) instead of (11) for PFM event logtype for sonusEventLogFileShutdownNotification trap, as both reason (1) and (11) indicates file write failure.

Severity 3: The sonusEventLogFileShutdownNotification trap might get generated with reason code (11), which will not have corresponding Reason description in the Alarms Guide.

GSX 09.00.03R000 Problems Resolved

The following are the issues fixed in GSX V09.00.03R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

GSX 09.00.02R000 Problems Resolved

The following are the issues fixed in GSX V09.00.02R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Category

Problems Resolved in GSX V09.00.02R000

Effect if Not Fixed

Call Processing

SONUS00137639: Wrong dataPathMode may be sent when the “Relay Data Path Mode Changes To The Other Leg” (IP Signaling Profile flag) is selected.

 

The code was modified to address dataPathMode race conditions.

Severity 2: May result in no audio or 1-way audio, in some on-hold / off-hold call scenarios.

 

Signaling

SONUS00149826: Calls failed after a GSX upgrade due to GUARD TIMER expiry. The peer violated the RFC by sending 180 (prack req) and 200OK simultaneously, so GSX was not able to send 200OK for the re-INVITE after the call was connected.

 

The code has been modified such that after the call is connected, PRACK pending is no longer required for a dialogue which is established.

Severity 1: Calls could fail after the GSX upgrade.

 

Signaling

SONUS00152479: There was an 8 second delay before executing the DIALMF SSP token after receiving the wink from the far-end.

 

The code has been modified to fix a bug in the CAS stack.

Severity 1: CAS Call functionality may be delayed or failed.

 

IP Routing

SONUS00150302: Significant MNS20 Cap Degradation between V7.3.1F9 and V8.4.6R0 SIP-S.

 

While processing incoming UDP packets, a call to function in_ifbroadcast() is made to check if the BROADCAST handling is enabled on local address. This function iterates through all the broadcast addresses available on that interface ONLY and makes a comparison with destination address of the incoming packet.

 

The issue was that the name of the function was wrongly modified to in_broadcast(). This function iterates through all the IP addresses in the system and compares it with destination address of the incoming packet.

 

 The code has been modified to replace in_broadcast with in_ifbroadcast call.

 

Severity 1: There could be degradation in performance during call load.

 

Signaling

SONUS00147628: Multiple PNS failure occurred.

For SIP-SIP calls and additional SDP offer trigger from one leg sometimes caused memory corruption on the other leg while accessing invalid memory.

 

The code has been modified to fix the issue.

Severity 1: There can be total outage.

 

IP Routing

SONUS00148354: GSX PNS sent IPv6 packets to incorrect NextHop MAC Address.

 

Appropriate changes have been made to select the best route based on the input interface.

Severity 1: IPv6 packets could be sent to incorrect NextHop MAC Address.

 

Coredump

SONUS00148156: While running the debug command traceroute6 to check network connectivity to remote IPv6DNS server connection to GSX was lost. MNS rebooted.

 

 The code has been modified by adding a defensive code.

Severity 1: MNS can coredump.

 

Coredump

SONUS00149698: Software failure occurred with error code 0x00000fb7 during LSWU. PNS coredumped.

 

 The code has been modified to fix the issue.

Severity 1: LSWU can stall.

 

Coredump

SONUS00148916: MNS switchover caused node reboot while attempting to come online with MANY SIFs. This happened because of a bug in SIF FSM wherein in case of ARP timeout, the state of the parent NIF was not being checked before sending out NIF_RES_ALLOCATE request to SIF.

 

The code has been modified to fix this issue.

Severity 1: MNS switchover can result in a node reboot.

 

Coredump

SONUS00147962: CNS cards continually coredumped after an upgrade. For a BT IUP to ISUP call, if the first backward ISUP message does not indicate subscriber free, the BT IUP code uses a null pointer.

 

The code has been modified to fix this issue.

Severity 1: CNS cards could coredump after the upgrade.

 

Call Control

SONUS00138181: After the upgrade from 7.3.2 to 8.3.1, the Content-Type in BYE message is changed by GSX from the ingress multipart or mixed to egress multipart or related. The call flow is ingress could be anything, egress is SIP-I. The Content-Type in other messages was not changed by GSX.

 

The function UacByeRequestCmd is modified to copy the flag from the transparency data provided by the application.

Severity 1: The content type in BYE may be processed incorrectly and thus calls may fail.

 

Call Number

SONUS00146270: In a call scenario where a SIP redirection was followed by ISUP redirection, the Called Directory Number sent in the final IAM contained the original called number from the first IAM instead of the called number from the Intermediate (2nd) IAM.

The code has been changed to set the Called Directory Number to the called number from the Intermediate (2nd) IAM (for the call scenario where a SIP redirection is followed by ISUP redirection).

Severity 1: The Called Directory Number can be a different one from what was expected by the customer.

 

Call Processing

SONUS00148106: Call transfer failed in the case of an H323->SIP call transfer, where SIP peer replied to re-INVITE with multiple codecs.

 

 The code has been modified to handle this specific case of H323->SIP call transfer correctly.

Severity 2: Call transfer may fail.

 

Alarms

SONUS00151976: NFS errors and GSX alarm “sonusEventLogDirectoryCreateErrorNotification” notifications were received.

 

When NFS Client Network Access is set to defaultMgmtNifGroup, NFS Timeouts occur intermittently causing automatic rollover of eventLog files.

 

Code was modified to fix this issue.

Severity 2: NFS timeouts will continue to happen.

Call Control

SONUS00146462: Second redirection number was not copied in to the accounting record on the first gateway.

 

The code has been modified to fix the issue.

Severity 2: Redirection number may be incorrect in accounting record.

 

Signaling

SONUS00149732: GSX was intermittently unable to offer multiple codecs when placing a call.

The issue was due to a media loopback feature that did not clean up properly when the call was torn down. The next time the resource was used, the GSX still treats it as a loop back media call and is not able to offer multiple codecs if needed.

 

The code was modified to properly cleanup media loopback call.

Severity 2: Calls can fail.

 

Signaling

SONUS00146713: While creating two SIP ports there were issues on VLAN on SBC 9000. The two ports that were created did not come into service.

 

The code is modified to eliminate the duplicate delete request to MNS PNA.

Severity 2: Signaling ports might not be available.

 

Call Processing

SONUS00150885: Certain H323 Slow Start call scenarios failed.

 

The code has been modified to fix this issue.

Severity 2: H323 Slow Start calls can fail.

 

SIP

SONUS00148395: Server card cored. This occurred due to a race condition in a Refer with Replaces call scenario.

 

 The code has been modified by taking an alternate approach at fixing a Refer with Replaces race condition.

Severity 2: Server card may core.

 

Call Processing

SONUS00148106/152350: Call transfer failed in the case of an H323->SIP call transfer, where SIP peer replied to re-INVITE with multiple codecs.

 

The code has been modified to handle this specific case of H323->SIP call transfer correctly. Note: This is a modification to the fix for 148106.

Severity 2: H323 Call transfer may fail in certain call scenarios.

 

Coredump

SONUS00149772: PNS cored due to invalid pointer access.

 

The code has been modified so that it does not use an invalid pointer.

Severity 2: PNS can coredump due to invalid pointer access.

 

Call Processing

SONUS00150690: Two-way audio was not established after the call was taken off-hold.

 

The code has been modified to fix this issue.

Severity 2: One way audio.

 

Coredump

SONUS00147790: xNS could coredump when IDLE is at 100% and there is an edge condition that incorrectly cannot find the highest CPU task other than IDLE.

 

The code has been modified to fix this issue.

Severity 2: Possible xNS coredumps during low activity.

 

Coredump

SONUS00148974: CQ123879 introduced a memory leak. If processing NOTIFY and relay flag are set in the dialog event package then memory can leak.

 

The code has been modified by changing the function to free memory after processing to fix the issue.

Severity 2: Call processing can fail.

 

Coredump

SONUS00147984: PNS switchover was followed by MNS core/switchover.

 

NIF code was modified to fix a NIF resource de-allocation issue.

Severity 2: MNS can core.

 

Database

SONUS00149041: For SIP to Japan ISDN calls that include a calling or called party sub-address, the Japan ISUP Variant Sub-field:58 and/or Sub-field:59 of CDR are invalid.

 

This ISDN bug was fixed.

Severity 2: CDR has invalid data.

 

Database

SONUS00149057: For SIP to Japan ISDN calls that includes UUI in the INVITE, the SIP Variant Sub-field:24 “UUI Sending Count” is invalid.

 

 This SIP bug was fixed.

Severity 2: CDR has invalid data.

 

LSWU

SONUS00147627: During LSWU to 8.4.5R000, the MNS20 link detection failed due to excessive processing of “rt_free” errors sent to the console port.

 

The root cause of these rt_free errors has been addressed preventing negative reference counts for routes within the PNA (routing) software.

Severity 2: LSWU may halt due to excessive link failure switchover of the ACTIVE MNS.

 

LSWU

SONUS00149232: After upgrading, NBS was not generating link local address.

 

The code has been modified to fix the issue.

Severity 2: After upgrade NBS may not display link local address.

IP Routing

SONUS00149168: After reboot, IPv6 configuration was lost for MGMT NIFs 3 and 4. Code to load IPv6 addresses for MGMT NIFs from param file was missing.

 

The code has been added to fix this issue.

Severity 2: After reboot, lost IPv6 configuration for MGMT NIFs could be lost.

 

Signaling

SONUS00147002: Possible valid Record Route was causing GSX to return 400 Bad requests. The NBS was unable to parse the Record-Route in its current form that was separated by multiple commas.

 

The code has been modified to handle the presence of multiple commas.

Severity 2: Calls can be rejected by NBS.

 

Call Control

SONUS00147748: Call trace filters configured with a "%" wildcard in the first position for the match criteria were triggered by calls with no incoming calling number.

 

 The call trace code has been modified to fix the issue.

Severity 2: No calling number can trigger a call trace filter using the % wildcard.

 

Signaling

SONUS00148016: For SIP-GW-GW-SIP call, GSX could not delete the redirection capability parameter (0x4e) and the redirection forward information parameter (0x8b) when the egress side was SIP-T trunk.

 

The code has been modified to fix the issue on Egress GW.

Severity 2: GSX cannot delete the redirection capability parameter on egress SIP-T.

 

Signaling

SONUS00146282: GSX killed the call on receipt of 183 with mime. This was due to unrecognized base version.

The code has been modified to fix the issue. As long as GSX can find a base or version in received message, GSX should try to carry on and decode it.

Severity 2: Calls can fail.

 

CLI

SONUS00146315: SIPSG/SIPS logging/trace was broken. Due to this important SIPS/SIPSG messages were not logged.

 

The code has been modified to fix the issue.

Severity 2: SIPSG

logging/trace can be missing.

 

 

Coredump

SONUS00148635: Standby PNS coredumped while traversing the RCB.

 

A defensive code has been added to deal with the null list case to fix the issue.

Severity 2: PNS can coredump.

 

Signaling

SONUS00147339: After PRACK timed-out, Q850 cause and location was not included on SIP.

 

The code has been modified to fix the issue.

Severity 2: Q850 cause and location may not be included at the time of PRACK(SIP).

Call Control

SONUS00136889: GSX failed to find egress service group. GSX had an event where all PNS cards cored. When the cards came back online certain GW-GW calls could not be established.

 

Debug code has been added to help isolate the problem.

Severity 2: Call cannot be established.

 

SONUS00148231: Refer LCQ 8667 - SIP ONHOLD/OFFHOLD Race Condition Issues for more information.

Severity 2

Coredump

SONUS00149864: Two CNS cards coredumped with different error codes due to DSP failure.

 

The issue was that the CNS cards with RTCP inactivity detection were causing some ICM buffers to leak. As result of the leak, no ICM RX buffers were available for the card to receive health check requests and card was reset due to health check timeout.

 

The code has been modified to free the ICM message buffer.

Severity 2: CNS cards could coredump due to DSP failure.

Call Processing

SONUS00147963: 400 bad request is generated when multipart SIP-I message received with CRLF-- Or Cr-- data in ISUP mime body.

 

The code has been modified to fix the issue.

Severity 3: Call can fail with a syntax error.

 

IP Routing

SONUS00148118:

While setting up the IPv6 addresses on the MGMT NIFs, one of the ports did not generate a link local IP address in Dualstack Mode.

 

The code has been modified to fix this issue.

Severity 3: GSX MNS port could fail to auto-generate IPv6 Link Local IP Address in DualStack Mode.

 

Signaling

SONUS00149778: CNS60s switched over/cored after AIS/RED alarm.

 

The code has been modified to fix the issue.

Severity 3: Particular slots can fail.

 

SONUS00149906: Refer PCR 7117- BTIUP to UKISUP mime Interwork  for more information.

Severity 3

DBG Logging

SONUS00149434: DBG logs rolled over too quickly (once every 20-30 minutes) and were filled with the following messages:

· “Unable to fetch message from the cache” message.

·   “Unable to copy Transparency content”.

 

Apart from the flooding of this major log, CANCEL message was missing the REASON header on egress.  

 

The code has been modified to fix these issues.

Severity 3: DBG logs can roll over too fast.

Call Control

SONUS00149904: Unwanted GwFeGetSigLinkStat messages were logged at MAJOR level in DBG log.

 

Some debug messages that are logged during processing of CLI command to get GW Sig Link Stats was changed to INFO level (from MAJOR level) to fix the issue.

Severity 3: Unwanted GwFeGetSigLinkStat messages can be logged at MAJOR level in DBG log.

Signaling

SONUS00151068: When GSX received ISUP IAM with respect to Called Directory Number (CdDN), GSX interworks this to INVITE with npdi and rn parameter in R-URI.

 

The code has been modified to fix the issue. If the Called Directory Number is not received, GSX will not populate “npdi” and “rn” parameter in the Request-URI.

Severity 3: NPDI and RN parameter can be sent via the INVITE even though “Called Directory Number” is not received.

Signaling

SONUS00151113: GSX did not send Called Directory Number in INVITE's Request-URI from incoming IAM with CdDN.

 

When GSX received IAM with Called Directory Number (CdDN), GSX interworks this to INVITE without Called Directory Number.

 

On the SIP side, when the "rn" and "npdi" were populated the population of Request URI with the Called Directory Number (CdDN) was added to fix the issue.

Severity 3: GSX may not send Called Directory Number in INVITE's Request-URI from IAM with CdDN.

 

Routes

SONUS00140456: Error 506E was encountered while deleting addr/mask/nexthop of route 0/0/0. This resulted in a major NRS event.

 

The issue has been fixed by not sending the second delete request to MNS PNA.

Severity 3: Major log “Can't delete route for 0.0.0.0, error 506E” can get printed inaccurately.

Signaling

SONUS00149773: After an upgrade, Logical bearer address could not be pinged and there were call failures.

 

 The code has been modified to fix this issue.

Severity 3: Logical bearer address cannot be pinged and calls can fail.

 

CLI

SONUS00147013: Light Gray Policer alarm counts did not increment.

 

The light gray and vlan stats were modified to fix the issue.

Severity 3: Light Gray Policer alarm counts may not increment.

 

SIP Signaling

SONUS00149484: Call Id hash table could contain stale entries, which could lead to an MNS core.

 

 The code has been modified to fix this issue.

Severity 3: MNS could coredump.

 

Media

SONUS00149166: DSP generates RTCP packets with incorrect value of LSR.

 

According to standard, Last SR (LSR) Timestamp is the middle 32 bits out of 64 in the NTP timestamp received as part of the most recent RTCP sender report (SR) packet from remote source.

 

However, LSR value in the generated RTCP packet by the DSP was updated with the time instant of the most recently received SR packet from the remote. This time was derived from an internal timer to DSP. This has been rectified and made standard compliant.

Severity 3: LSR is used in calculating total round trip delay and this will be incorrect.

 

Management

SONUS00149123: Logging interval for FTP based interval stats should be same as SNMP.

The timer value for FTP based interval statistics has been modified to fix the issue.

Severity 3:

The granularity of statistics for FTP based statistics collection should be equal to or multiple of performance statistics interval size. There will be discrepancy in statistics if the statistics intervals are not configured accordingly.

CLI

SONUS00150180: Network Selector Table did not allow default IPv6 address.

 

The code has been modified to fix this issue.

Severity 3: Network Selector Table might not allow default IPv6 address.

CLI

SONUS00134326: GSX Tone package allowed profiles with duplicate package identifier.

 

Code has been added to validate the given tone package ID for duplication.

Severity 3: There can be discrepancies with package identifier.

 

Performance

SONUS00149813: PNS accumulated a high ICM InUseCount with a slow memory leakage since the card was up for a long time.

 

A code was added to release the ICM buffer properly to fix the issue.

Severity 3: Performance can become slow.

 

Interoperability

SONUS00150688: ISDN Status message had wrong call state for optional IE.

 

The code has been modified to return the correct Call State in the Status message.

Severity 3: Calls can fail.

Call Control

SONUS00151508: GSX sent REL after receiving policy response in the case of LRNS.  The call was being released.

 

This flag should only cause a call release when the ingress side is SIP and the PSX query is due to a redirect.

 

The code has been modified to fix the issue.

Severity 3: Calls can get disconnected.

Performance

SONUS00149813: PNS accumulated a high ICM InUseCount with a slow memory leakage since the card was up for a long time.

 

 A code was added to release the ICM buffer properly to fix the issue.

Severity 3: Performance can become slow.

 

Signaling

SONUS00148116: While configuring the IPv6 addresses on the MGMT NIFs, “LIF not found" error occurred when the debug command “ping6” was executed.

 

Logic is added to check for the MGMT NIFs which were missing from the NRS debug "ip ping6 -N" command.

Severity 3: GSX debug command “ping6” will not work for MGMT NIFs.

 

Redundancy

SONUS00149609: If you have a PSTN to SIP call, followed by a SIP Refer, followed by an MNS switchover, the trunk group will consider the PSTN leg available for hunting. The busy PSTN leg may be repeatedly hunted and the calls fail with cause code 135. The failures stop when the call releases however long duration calls require maintenance.

 

This redundancy bug was fixed.

Severity 3: Calls can fail.

 

Signaling

SONUS00149394: If you unblock a range of CICs, and one of the higher numbered CICs are unassigned at the far end, the trunk group will consider all the CICs available for hunting. Unassigned CICs may be repeatedly hunted and the calls fail with cause code 135.

 

 This ISUP bug was fixed.

Severity 3: Calls can fail.

 

Statistics

SONUS00146635: The latency field in the ingress/egress RTCP statistics were populated as 0 and 65535 regardless of any latency introduced in the RTP stream.

 

The latency field in the code has been modified to fix the issue.

Severity 3: Inaccuracy in the latency field can occur.

 

NRMA

SONUS00147199: PNS failed with accessing invalid memory in NRMA resulting in error code 0x00000051.

 

The code has been modified to fix the issue.

Severity 3: PNS failure can occur.

 

Signaling

SONUS00148440: On Egress SIP, after GSX sent ACK for initial Invite and Re-Invite right after it, if ACK was dropped on the wire, GSX did not resend ACK for the rexmit 200 OK of previous transaction.

 

The code has been modified to fix the issue.

Severity 3: Peer may teardown the call.

 

Call Control

SONUS00144865: GSX received 2000OK with Record-Route. ACK/BYE sent missing Route headers.

The code has been modified to fix the issue.

Severity 3: GSX may send ACK/BYE to the correct destination.

 

Network

SONUS00145115: Due to network outage different slots did not sync correctly. During switchover, MNS/PNS might not able to sync up due to registration hash table corruption.

 

The code has been modified to fix the issue.

Severity 3: Calls can be lost.

 

Coredump

SONUS00145060: PNS40 card faulted with error code 0x00004a11 during Switchover.

 

The code has been modified to correct the issue where filter entry was not freed properly which caused error during switchover.

Severity 1: Cannot upgrade the network to 8.4.4F004 release.

 

Signaling

SONUS00148724: GSX sent empty FMTP attributes when the egress codec is AMR/AMR-WB with Bandwidth Efficient (BEff) option.

 

The code has been modified to format the FMTP in case of AMR/AMRWB codec.

Severity 3: AMR-WB HD codec may not work with 7.3.6F11.

 

Signaling

SONUS00146679: When GSX responds with 200OK to the received Invite/Update, the display name in Contact header was missing double quote.

The code has been modified to fix the issue.

Severity 2: When the far end gets the 200OK formatted incorrectly, it can retransmit multiple times until GSX finally responds with 500 internal server errors.

Call Processing

SONUS00147908: “Supported” header of “200 OK” message got changed after an upgrade. The issue was that "Empty option” tag was present in the header.

 

The code has been modified to fix this issue.

Severity 2: “Supported” header of “200 OK” message could get modified after an upgrade.

 

Signaling

SONUS00148802: GSX rejected T.38 re-INVITE with 488 response because the maxBitRate was greater than currently supported.

 

The code has been modified such that if GSX receives an INVITE with maxBitRate greater than currently supported, GSX will answer with the maxBitRate currently supported.

Severity 2: GSX could rejectT.38 re-INVITE with 488 response because the maxBitRate was greater than currently supported.

 

CDR/ Binary Streaming

SONUS00147838: When the GSX is configured for Binary Streaming, CDRs that are larger than 2056 bytes will get stuck in the CDR Spooler on the GSX.

 

GSX code has been modified by increasing the MAX size allowed to be transmitted to a binary accounting server to match the max size of a CDR.

Severity 2: Binary CDRs greater than 2056 bytes will never get transmitted to the accounting server.

 

GSX 09.00.01R000 Problems Resolved

The following are the issues fixed in GSX V09.00.01R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Category

Problems Resolved in GSX V09.00.01R000

Effect if Not Fixed

Coredump

SONUS00147962: CNS cards continually coredumped after an upgrade.

 

For a BT IUP to ISUP call, if the first backward ISUP message does not indicate subscriber free, the BT IUP code uses a null pointer.

 

The code has been modified to fix this issue.

Severity 3: CNS cards could coredump after the upgrade.

GSX 09.00.00R000 Problems Resolved

The following are the issues fixed in GSX V09.00.00R000. The implications of not upgrading to include a particular fix have been provided where appropriate:

Category

Problems Resolved in GSX V09.00.00R000

Effect if Not Fixed

Regression

SONUS00142204: GSX rebooted three times and it experienced congestion before rebooting.

The code has been modified to update the correct length of memory for different variants, ITU/JAPAN/CHINA.

Severity 3: GSX will run out of memory due to use of over-sized memories.

Alarms

SONUS00136596: The alarm “sonusCallFailureNoDspExceedThresholdNotification” has been deprecated from the GSX 8.4.4 and 9.0 Alarm Troubleshooting Guide.

The code has been modified to deprecate “sonusCallFailureNoDspExceedThresholdNotification” alarm.

Severity 3: The deprecated alarm will still be displayed in EMS GUI.

Coredump

SONUS00139127: MNS coredumps when SIP Signaling port is brought in-service.

The code has been modified to fix the memory corruption issue in radix tree implementation.

Severity 2: Rebooting the GSX in binary mode with multiple host routes configuration will lead to MNS crash.

CLI

SONUS00140749: When a ping path is configured using SipOptions protocol and then enabled, it works. Disabling and Enabling it again does not work. To make it work, it must be disabled/enabled again. This is seen when a single path is configured.

There was a logical error while disabling the ping path.

The code has been modified to fix this issue.

Severity 3: Path disable/enable does not work using SIP OPTIONS pathcheck ping.

 

Regression

SONUS00137843: PCR 1917/2010 – Disconnect Treatment (DT) profile announcement is not being played.

The code has been modified such that the DT profile announcement will be played.

Severity 3: DT profile announcement may not be played.

 

Security

SONUS00129391: Double free vulnerability in Open SSL 0.9.8 before 0.9.8s, when X509_V_FLAG_POLICY_CHECK is enabled, allows remote attackers to have an unspecified impact by triggering failure of a policy check.

As per US-CERT/NIST vulnerability Id: CVE-2011-4109, it allows unauthorized disclosure of information, unauthorized modification, and disruption of service.

The code has been modified such that it stops policy check failure from freeing same buffer twice.

Severity 2: It will allow unauthorized disclosure of information, unauthorized modification, and disruption of service.

 

Coredump

SONUS00141006: There were multiple coredumps on PNS resulting in error 000051. This was because the candidate attribute sent to GSX was anything other than “UDP and HOST”. SRFLX cannot be used.

 

The code has been modified to change the logic that fixes this issue. For more information, see WBA Warning-12-00018675.

Severity 1: PNS card can coredump.

Coredump

SONUS00128818: PNS41 and MNS21 core observed while disabling and enabling the NIF after an upgrade. The core dump happens when the configured routes contain an IP address of the signaling port of the same GSX.

The code has been modified such that the routes are removed.

Severity 2: PNS41 and MNS21 core dump.

Coredump

SONUS00133550: During LSWU upgrade from V07.01.04 R001 to V07.03.07 R0, the PNS cards core dumped.

LSWU has been modified to fix this issue.

Severity 3: PNS card during LSWU could coredump during upgrade from V07.01.04 R001 to V07.03.07 R0.

Call Processing

SONUS00138203: IP input/output filters will not work correctly.

Wrong filters are selected for processing.

The code has been modified to select the proper IP input/output filters.

Severity 2: Wrong filters are selected for processing.

Signaling

SONUS00137949: When upgrading from a pre-DXRM release to latest release, part of distributed XRM update was missing.

The code has been modified to fix this issue and to distribute the interface information properly.

Severity 1: LSWU from 6.5 to 7.3.6 NrsInterfaceSelect will fail.

Configuration

SONUS00132375: The CDR Maximum Record Sequence was configured to 32bits, but the actual sequence numbers in the records were changed back to 16bits after a GSX reboot.

Configuration was not loaded properly when reading configuration from the parameter file.

The code has been modified to fix this issue.

Severity 3: Billing data can be incorrect.

Call Processing

SONUS00143350: In a call-hold scenario, calls resulted in dead air. This issue occurred when the data path mode changed for on-hold that goes through a late media Re-INVITE from the remote side. The call processing state machine was not updating the active PSP for the call to be sent in the answer message for subsequent offers. Thus, GSX sent wrong SDP in the answer messages.

The Call Processing state machine has been modified to update the new active PSP in answer messages so that GSX would send the correct SDP.

Severity 1: No audio in on/off hold scenarios.

Signaling

SONUS00139192: MTP2LINK comes up only after GSX restart, when deleted and re-created.

The logical sequencing of deleting the MTP3 link has been modified.

Severity 2: When the MTP3 link is deleted and re-created, the link will not come back to INSERVICE/AVAILABLE state.

Coredump

SONUS00144573/144572: PNS cored with error code 0x00000021.

The GSX SIP stack has been modified to allocate sufficient memory for media proposal structures.

Severity 2: Coredumps can occur frequently throughout the network due to memFree failures of large media proposal structures.

Coredump

SONUS00137535: During LSWU upgrade, server cards including MNS may core.

ARS messages between older and newer software have been modified to address IPv6 modifications.

Severity 3: If ARS contains black listed entries, GSX LSWU upgrade will fail and service may be affected.

Signaling

SONUS00147087: When an active non-IXP card (MNS20s, PNS40s, CNS4x/8x and SPS80s are IXP-based cards) was mirroring the NIFGROUP interface data to standby using the local lookup by ifindex, the local lookup failed to find the slot of the SIF and NRS printed “No interface found”.

The code has been modified to get the slot information from the NIFGROUP interface object instead of local lookup by ifindex.

Note: This happens only during redundancy mirroring. Synchronization between servers is not affected.

Severity 2: Calls can fail and “No interface found” error could be printed in the logs by the NRS subsystem.

Performance

SONUS00136245: NRMA requested XRM resource an out-of-service (OOS) CNSx card on boot up.

The code has been modified to detect the card which is out-of-service during the first boot up.

Severity 4: Calls can fail, wasting CPU cycles.

Signaling

SONUS00138846: Initially, GSX tried to route the call to the blacklisted server.

The code has been modified to send calls directly to secondary server.

Severity 2: Initially, GSX can try to route the call to the blacklisted server.

Management

SONUS00137783: The port 3 MGMT IF on standby MNS were in OOS_LINKDOWN in NRS context. But it is in LINK_UP state in IXP context.

The code has been modified to reset lastAnnouncedLinkState in IxpNiInit().

Severity 3: The standby management NIF OOS-LINKDOWN will occur after switchover.

Call Processing

SONUS00138276: Occasionally during call audit processing, stable calls could be erroneously terminated. An incorrect message being reported during the audit could trigger the problem.

 

The software reporting the incorrect message has been fixed.

Severity 1: Many stable calls could get dropped with Disconnect Reason 132.

Call Control

SONUS00138181: After the upgrade from 7.3.2 to 8.3.1, the Content-Type in BYE message is changed by GSX from the ingress multipart or mixed to egress multipart or related. The call flow is ingress could be anything, egress is SIP-I. The Content-Type in other messages was not changed by GSX.

The function UacByeRequestCmd is modified to copy the flag from the transparency data provided by the application.

Severity 1: The content type in BYE may be processed incorrectly and thus calls may fail.

Logging

SONUS00138632: ACM is received with no optional parameter position 17 of 0x00.

The code has been modified such that it displays an error message when a message without optional parameter pointer and optional parameter is received.

Severity 3: Prints wrong log, if ACMis received without optional parameter pointer.

DP_XCODE

SONUS00136886: DTMF digits queued and eventually stopped generating. Sometimes, digits get queued due to bursts of digits and these queues may get stuck due to a timestamp change.

The code has been modified to skip timestamp adjustment when digits are active.

Severity 2: DTMF digits are queued and may stop eventually generating in case of a timestamp switch resulting from change in time-stamp base in incoming RTP stream.

Call Processing

SONUS00139056: 200OK is not ACK'd after 2 REFERs. The issue is seen in a multi-stage call transfer scenario load run. This results in call transfer failure.

The call control code is fixed to handle 200OK in certain race condition scenarios that involve multi-stage transfer.

Severity 2: GSX may not send back the ACK message.

Signaling

SONUS00136578: CNS card kept coring.

The issue was in handling syntactical wrong “a” attribute in SIP message. SIP parser was stuck in infinite loop.

The code has been modified such that Error handling is added for syntactical error cases.

Severity 3: CNS/PNS/SPS card will coredump.

Signaling

SONUS00138349: GSX 7.3.6R6 DSP resource issue. DSP leak occurred in a blind transfer case in which first leg of the call was transcoded. This issue is introduced by CQ134525. The issue was that the old active SP contexts were not freed by NRMA when a new context became active in case of transfer call.

The fix for CQ135425 is modified to delete the memory allocated for the old active context.

Severity 3: Calls can fail.

Fault Management

SONUS00133320: TCP sessions were getting "quietly" rejected. Need a new debug counter in "ipnetstat -s" and a major event log to show TCP connection attempts exceeding the maximum socket queue limit.

A new debug stat counter has been added in "ipnetstat -s tcp" and a more specific major event log from PNA to FM for the TCP connection attempts exceeding the maximum socket queue limit.

Severity 3: TCP sessions may get rejected.

Coredump

SONUS00139700: PNS coredumped following error code 0x0000022.

The code has been modified such that error checking is enhanced to avoid coredumps.

Severity 3: Coredumps could occur due to memory corruption.

Signaling

SONUS00139102: ACM sent too early in IUP-SIP interworking case.

BT IUP has to delay sending of the ACM until ringing as it has no option to send CPG type messages.

The code has been modified to add a CLI command to control whether the BT service must wait for the SIP ringing message.The default behavior is unchanged. The CLI command is as follows: CONFIGURE BT SERVICE<profilename> SIGNALONLYSIPALERT.

Severity 2: A SIP 183 Session Progress will cause an ACM and there is no further indication when the ringing occurs.

ISUP Signaling

SONUS00139798: The field ‘NetworkInd’ is displayed twice under 'ISUP SIGNALING GROUP ansi'.

The code has been modified to fix this issue.

Severity 3: The field ‘NetworkInd’ will be displayed twice.

Signaling

SONUS00138530: The ISDN to SIP call gets cleared because the GSX sends PROGRESS after ALERT. The network side sends a STATUS and this causes the GSX to clear the call because it is already in transitioned state (and the call state in the STATUS does not match).

The code has been modified to fix this issue.

Severity 3: GSX sends PROGRESS after ALERTING.

Fault Management

SONUS00137283: Call policing warning alarm was not generated.

The code has been modified such that the call policing alarm is now generated.

Severity 3: Call policing warning alarm could not be generated.

Fault management

SONUS00122371: CNS card has multiple switchovers and active and standby CNS get into an oscillation mode. The problem is FM declaring a DSP failure wrongly. Additional counter was added to detect DSP failure correctly.

The code has been modified to read the online cards on the Fault Management (FM) and the Maker/IXP counters to report as an event to MNS FM. EVENT_MAKER_BAD_SDU_LEN is added to notify the error counters to the FM on MNS card.

Severity 3: CNS cards may switchover repeatedly and can cause to coredumps.

ISDNSG

SONUS00138866: Multiple CNS cards were coredumped this was due to ISDN task was utilized 99% of CPU.

The code has been modified to add isdnsgdchannel operation state wait delayed establish state in dchannel state machine to delay datalink establishment events by 1 second, thereby to prevent isdn task from hogging CPU time and watch dog timer failure core dumps.

Severity 2: The core dump can be due to high CPU utilization.

Signaling

SONUS00118165: Cached entry containing MAC did not get updated when ARP changed.

The code has been modified to forward all ARP changes to XRM. XRM has been modified to detect MAC changes in these ARP messages. If the MAC has changed, XRM will proceed to process these changes as it did using the existing code that processes route change notifications.

This results in an immediate removal of this old MAC address from the XRM cache and the tear-down of all calls on that particular route using the NRM call cleanup mechanism. The next new incoming call to that route will force a lookup of that route, retrieve the new MAC address, and populate the XRM cache with the new MAC address entry.

Severity 2: Leaving the old MAC address unchanged will result in one-way audio for all the existing calls on the NIF.

Coredump

SONUS00129433: Need to add defensive code to avoid accessing or freeing NULL pointer in association with CQ120600.

The code has been modified such that a defensive code has been added to fix this issue.

Severity 3: Potential MNS coredump could occur.

Coredump

SONUS00140995: Due to the earlier implementation of CQ118165, a PNS coredump or reset without coredump was observed in some cases when a MAC address change and subsequent ARP change fault was raised.

The XRM code has been modified to correctly process the resource cleanup due to the route change fault (or ARP change fault).

Note: This CQ was found internally, but this is mentioned here because it is exposed by an earlier CQ 118165.

Severity 2: Due to the earlier implementation of CQ118165, a PNS coredump or reset without coredump could occur in some cases when a MAC address change and subsequent ARP change fault is raised.

Signaling

SONUS00140232: This is a BT IUP interworking bug. If a previous exchange delivers more digits than requested (by a SND (propagated backward from the egress leg)), the GSX may deliver digits out of sequence or may deliver multiple FAMs.

The code has been modified to fix this issue.

Severity 2: The call fails to complete.

Call Processing

SONUS00141378: ISDN to SIP calls failed after an upgrade.

The code has been modified to fix this issue.

Severity 2: This bug could prevent any call from using the MNS packet NIF.

Signaling

SONUS00140380: When a T1 channel goes out of service any resident static call should be cleared. If you re-configure an endpoint of a static call to be on a different channel, this mechanism breaks down for that channel.

The code has been modified to fix this issue.

Severity 2: A static call may not be cleared when its T1 goes out of service.

 

SYS Log

SONUS00140562: When GSX is configured to run SIP to SIP call of 25 CPS with CHT of 100 seconds, MNS coredump (in sensitive mode) is observed for softswitch modifications in a scenario when softswitch is deleted before dryup timeout.

The code has been modified to fix this issue.

Severity 2: MNS will coredump (in sensitive mode) or SYS error will be logged for softswitch modifications as mentioned.

 

Control

SONUS00134250: The following message (which may be logged during congestion) doesnot have enough information:

.TRM : GatewaySG Failed: gwName=registered SG Handle not found, shelf=1 slot=1

The code has been modified to include the sgType in this message.

Severity 2: None

Coredump

SONUS00141181: When an SA establishment failed as PFS is not supported in GSX, message was dropped & cleanup was invoked. The core happened during the cleaning up process when a null exchange name was being referenced.

The code has been modified to delay the cleanup

after the exchange name is being filled up properly.

Severity 2: GSX will restart while establishing IPsec.

Call Control

SONUS00138944: In certain very specific call flows which use Overlap Addressing and a Trigger Script, the Egress IAM called number may be incorrect.

The code has been modified to send the correct called number.

Severity 2: Overlap to enbloc with overflow number may be broken in certain very specific call flows in which a Trigger Script is run if egress side returns a specific release code.

Signaling

SONUS00139658: UPDATE message was triggered when SDP was included in the subsequent 18x message even when there was no change to the SDP.

The UPDATE message was sent because the local SDP in the first 18x and newly formed SDP was different. The newly formed SDP had only audio and it got queued. The queuing of the SDP was the problem.

The code has been modified to merge the local SDP sent and newly formed SDP in case of multiple m-lines.

The SDP would be queued and sent later (if

different after the merge). In this case UPDATE will not be sent because the SDP has not changed.

Severity 2: UPDATE could get triggered even when there has been no change to the SDP, in case if the offer had multiple m-lines.

Call Control

SONUS00139965: GSX was not updating the master key with the new one and therefore fails to decrypt the encrypted audio packets received from the ingress peer thereby causing one way audio.

The code has been modified to update the master key, if it is changed in a session modify.

Severity 2: GSX may ignore master key changes in the session refresh re-invite.

Coredump

SONUS00139889: PNS cards coredumped at multiple sites.

There were two problems in this CQ, one in IPM and the other in NRMA.

The cores occurred due to a corruption in the rogue media port quarantine list and due to an invalid access of the NRMA callLeg resource chain.

The code has been modified to avoid the coredump.

Severity 3: PNS cards could coredump.

Call Processing

SONUS00140999: Stale route leftover post PNS Re-IP caused call failures.

The issue was that the admin route delete request was not sent to XNS because the NIF was already disabled.

The code has been modified to allow the admin route delete request being sent to XNS even NIF/SIF is not at REMOTE_CREATED state.

Severity 3: Calls could fail.

Signaling

SONUS00140165: SYS ERR in parameter loading.

The SIF parameter loading function has been updated to allow SIF creation when parent NIF's VLAN tagging state is DISABLED and SIF admin state is DISABLED as well.

Severity 3: SYS ERR in parameter loading.

Coredump

SONUS00142204: GSX rebooted after congestion.

The crash was triggered by trying to allocate over-sized memory.

The code has been modified to fix this issue.

Note: This is the same issue that CQ106815 has fixed. But CQ106815 fixed only ANSI. This is happening for ITU.

Severity 3: Coredump could occur.

Media

SONUS00141217: On the upgrade, traffic load sharing was broken for the IP calls. Because of wrong gateway address returned from route lookup.

The code has been modified such that PNA route lookup all function to return the correct gateway address.

Severity 2: System can exhaust resources results to

failures of premature call.

 

Signaling

SONUS00140093: P-Charge-Info header is not generated in the egress INVITE in SIP-GW-to-GW-SIP call flow.

The code has been modified to fix the Egress Gateway.

Severity 2: P-Charge-Info header is not generated in SIP-GW-to-GW-SIP call flow.

Signaling

SONUS00140782: GSX does not send 18x and 200, but it received ACM and ANM message from GWGW. When remote GW did a crankback, the ingress SGFSM leg on the local GW had reinitialized its resources.

The code has been modified to reinitialize resource when the crankback perform for the local GW only.

Severity 2: Calls can Fail.

SIP Signaling

SONUS00140037: GSX8.4 and SBC 5200 2.2 sends 400 Bad Request to incoming SIP source port.

The code has been modified to fix this issue.

Severity 3: GSX8.4 and SBC 5200 2.2 will send 400 Bad Request to incoming SIP source port.

Call Processing

SONUS00134605: GSX sends 183 and receives PRACK, after that GSX sends 481 instead of 200 OK.

The code has been modified to fix this issue.

Severity 3: GSX sends wrong messages.

Signaling

SONUS00140227: Re-Invite with incorrect codec resulted in call teardown. This is for SIP-GW-GW-ISUP. Where SIP is late media, and transcoded only.

The code has been modified not to send reInvite if the codec selected does not match peer.

Severity 2: Re-Invite with incorrect codec may result in call teardown.

LSWU

SONUS00134701: Generating DTMF from SIP to PSTN not working after upgrading to 7.3.7R4.

The code has been modified to fix this issue.

Severity 3: Upgrade can rollback.

Signaling

SONUS00138505: There was call-hold behavior change after an upgrade, wherein the call flow changed without changing the configuration.

The code has been modified such that after the call on-hold, GSX does not send re-Invite lockdown (on-hold again).

Severity 2: After an upgrade, the call flow could change without changing the configuration.

Call Processing

SONUS00137790: Out of the two application servers (AS's) in a cluster, when AS1 was blacklisted, the calls were not routing out on the second contact/IP in the 302 move temp message. When 302 message contains contact address in the blacklist, GSX still tries to route to the blacklisted address.

The code has been modified to skip the address in the blacklist.

Severity 3: Call processing.

Call Control

SONUS00140469: Inconsistencies occurred in mssgthrottling, callspersegment, and intersegtimer due to incorrect ISUP values. The ISUP service "call cleanup data" should be the same on all ISUP services.

The code has been modified to ensure that the ISUP service data is consistent on all services.

Severity 3: Customer configuration can show inconsistent ISUP service data.

Coredump

SONUS00142190: The MNS card coredumped after an upgrade.

The ARM code has been modified to fix the list of corruption issues that were aggravated by, but not caused by CQ119518.

Severity 2: MNS may coredump after the upgrade.

Coredump

SONUS00142356: When deleting an MTRM client from the GSX, the MNS card may coredump due to the code incorrectly referencing a list structure.

The code has been modified to properly delete the client and not core the MNS.

Severity 1: MNS card could core dump when attempting to delete an MTRM client on the GSX.

Media

SONUS00142529: When cos/tos was enabled in PSP it resulted in one way audio.

The header and replay functions have been modified in the code to fix the issue.

Severity 3: Can result in one way audio.

Signaling

SONUS00136630: Post LSWU, SIFs remained Out of Service (OOS).

The code has been modified to bring back the SIFs to service.

Severity 3: SIFs could remain OOS post LSWU.

Signaling

SONUS00125973: DTMF detection failed due to "double detects" when using RFC2833 with remove on CNS cards.

The code has been modified to fix this issue.

Severity 3: DTMF detection could fail due to “double detects” when using RFC2833 with remove on CNS cards.

Media

SONUS00125785: Multiple DSP (Digital Signal Processors) core dumps.

The code has been modified to perform several additional checks to properly qualify the RTP SID payload type.

Severity 2: Temporary loss of redundancy.

Signaling

SONUS00141979: ISUP status of SGX4k node was incorrectly reported as “Failed”. This occurs after the first MNS switchover after a previous M3UA migration. An additional MNS switchover clears the problem and it will not re-appear.

The code has been modified to fix this issue.

Severity 2: Calls may work fine but administered CICs do not come back into service.

Signaling

SONUS00142698: For ISUP to BT IUP calls, any PSX parameter manipulation of the Forward Call Indicator is ignored.

The code has been modified to use the Forward Call Indicator provided by the PSX.

Severity 3: Customer loses the ability to modify the Forward Call Indicator for BT IUP to BT IUP calls.

Call Processing

SONUS00143636: Some static calls did not recover after a node reboot. Due to an apparent race condition, the DSP resource manager improperly initializes and will not allocate resources. This affects all call types (not just static calls).

The code has been modified to fix this issue.

Severity 2: A card may initialize with the DSP resource manager unable to

allocate resources. A card reset is required to clear the problem.

Coredump

SONUS00120482: MNS may switchover when adding an IP input filter or a new TG to the configuration.

The code has been modified to avoid the coredump.

Severity 2: MNS could coredump when adding an IP input filter or a new TG to the configuration.

Call processing

SONUS00141223: SBC stopped processing SIP calls.

The code has been modified to support up to 65535 interfaces and sub interfaces.

Severity 3: Calls may not get processed.

 

Memory Management

SONUS00143691: Switchover occurred on lagsx00 with error code 0x00000087. It was a memory double free issue.

The code has been modified to add a defensive check in SIPSG to prevent this from happening in the future.

Severity 3: Coredump could occur.

Signaling

SONUS00141824: IP NST issue -Users had to delete and rebuild IP NST entry for the calls to complete.

The code has been modified to put out a warning message when user trying to add or delete a number in the IP network selector table.

Severity 2: Calls could fail.

Signaling

SONUS00144141: CAS “Maint Busy” feature broken. With “Maint Busy” enabled, an out of service CAS channel signals busy to the far end to prevent incoming call attempts. The bug is that the busy signal does not survive a T1/E1 alarm. After an alarm idle is signaled.

The code has been modified to fix this issue.

Severity 3: The GSX may get ingress call attempts on blocked CAS channels.

Signaling

SONUS00142436/141891:

The parameter “rport” value was irregularly populated in the “Via” Header.

Generic parameters in “Via” header were not initialized properly.

The code has been modified to initialize the parameters properly.

Severity 3: The parameter “rport” value could be irregularly populated in the “Via” Header.

 

CLI

SONUS00140981:When the command “SHOW SIP SIGNALING ALL_REGISTRATIONS HOST ANY PHONE-CONTEXT NONE USER <telNumber> STATUS” is issued the requested output is not obtained. If AOR has ‘+’ sign, the show screen returns nothing.

The code has been modified to fix this issue.

Severity 3: Incorrect AOR details can be displayed on the screen.

Signaling

SONUS00142430: When PRACK received a new offer, 200 OK INVITE was sent prior to 200 OK PRACK message.

The code has been modified to fix this issue.

Severity 2: Calls may fail due to this.

Signaling

SONUS00142675: SBC did not send 200 OK to incoming UPDATE message.

The code has been modified to fix this issue.

Severity 3: SBC may not send 200 OK to the incoming UPDATE message.

Coredump

SONUS00143190: Multiple CNS 85 cards coredumped due to server module software failure in different slots. [ST] When relay messages are on for sip gw-gw sip calls, GSX failed to detect the sgtype.

The code has been modified to detect the gw-gw sgtype correctly.

Severity 3: Coredump can occur.

Coredump

SONUS00135479: SIPSG was deleted but was not removed from hash table corrupting it.

The hash table code has been modified to fix the issue.

Severity 2: The hash table can get corrupted, resulting in coredump.

Signaling

SONUS00144334: SIPS needs more defensive checks as it sent re-Invite in “invalid” state, whereas it should not.

The code has been modified to fix this issue.

Severity 2: Calls might tear down.

SIP

SONUS00144290: Multiple link failures (switchover) distributed SIP signaling ports, which did not get updated to MNS properly.

The code has been modified to update the SIP signaling ports to MNS properly.

Severity 1: SIP Signaling ports may not be available after upgrade causing massive call failures.

Coredump

SONUS00142889: When direct media calls and fax were involved, GSX cored due to duplicate memory free.

The code has been modified to fix this issue.

Severity 2: The card may core.

M3UA SS7 Migration

SONUS00144688: The SS7 node could not be enabled after completion of the M3UA migration process.

The SS7FE code has been modified to preserve the nodeServerProtocol to prevent this issue from occurring.

Severity 2: The SS7 node would need to be deleted and rebuilt. This can impact service.

Interoperability

SONUS00144145: “COT on previous circuit” not being signaled for HONGKONGIC variant.

The code has been modified to fix this issue.

Severity 2: When the COT message is sent on the egress side, the call is cleared.

Overlap Signaling

SONUS00144829/ LCQ 8088: Following issues are addressed as part of this CQ:

  • When overlap digits are received from ingress and GSX does crankback, the overlap digits must be appended while crankback to next route. But GSX appends overlap digits twice in called number. This is seen with the flag “enforceMaxDigitsForValidPhoneNumber” enabled in PSX.
  • In case of overlap call, “Called Number” in CDRs is not correct. Overlap digits are appended in “Called Number” also with “Extra Called Address Digits”. This is seen with the flag “enforceMaxDigitsForValidPhoneNumber” enabled in PSX.
  • The unnecessary merging of overlap digits (with PCR 5442 fix) in some scenarios like digit completion is determined and after that more overlap digits are received.
    • The PDD (Post Dial Delay) timer related issue was seen with the flag “enforceMaxDigitsForValidPhoneNumber” enabled in PSX.
    • Waiting for more overlap digits even though there are sufficient digits. For example, a scenario in which PSX dip fails due to insufficient digits and sufficient digits are received after the PSX query although the digits should have been processed and should have sent another PSX query. This is seen with the flag “enforceMaxDigitsForValidPhoneNumber” enabled in PSX.
    • Crankback call with multiple SAM messages was not working. When call is released from the first route and if SAM was received before crankback happened, it may be ignored.

The code has been modified to fix the above mentioned issues.

Severity 3: Mentioned issues will be seen in overlap call scenario.

Signaling

SONUS00137299: GSX PNS redundancy switchover resulted in all the ports being down. The “Signal Received” bit on the FPGA register was misinterpreted as “signal received and auto-negotiation complete or timed out” for auto-negotiation enabled mode or “signal received” for auto-negotiation disabled mode. It turned out that “Signal Received” only indicated “signal synchronized” and had nothing to do with auto-negotiation complete or timeout.

 

The code has been modified to check for auto-negotiation complete after LINK and SYNC has been established on the MAC using a timer. One known impact for this change is that it will increase the switchover time for a PNS40 port by 40 msec if the route/switch is set to auto-negotiate disabled. To get around this problem, the user would need to set the router/switch to auto-negotiate enabled and ensure that the router/switch adheres to the IEEE 802.3e Clause 37 standard.

Severity 3: GSX PNS redundancy switchover can result in all the ports being down.

Signaling

SONUS00138148: Standby NIFs failed to become active and stable after Layer 3 PNS switchover.

The code has been modified to fix this issue.

Severity 2: Standby Layer 3 NIFs may fail to take over after PNS switchover during an external router outage where both the active and standby NIFs are toggling in and out of service.

Signaling

SONUS00144898: GSX ignored the “expires” value in subsequent 200OK from the registrar.

The code has been modified to update the internal expires timer value when received subsequent 200OK from the internal registrar.

Severity 2: If Registrar receives the “expires” value in subsequent 200OK for refresh, GSX may not send refresh registration in time to the registrar.

Coredump

SONUS00142356/145035: When deleting an MTRM client from the GSX, the MNS card may coredump due to the code incorrectly referencing a list structure.

The code has been modified to properly delete the client and not core the MNS.

Severity 1: MNS card could core dump when attempting to delete an MTRM client on the GSX.

Static Call

SONUS00141097: If the GCID of a static call is zero (it is unclear how GCID became zero as zero is not a valid GCID) then MNS1 may core when the status of a static call is retrieved thru CLI commands or SNMP get requests.

The logic of the code has been modified to gracefully handle an invalid GCID of zero in a static call.

Severity 2: MNS switchover can occur and possibly a node reboot.

Coredump

SONUS00140352: PNS/CNS/SPS cards can leak memory on an edge condition SIP call flow. 100 bytes of memory leaks for this edge condition.

The code has been modified to clean up the call leg information.

Severity 1: PNS/CNS/SPS cards can leak memory on a specific unique call flow and will require resets to clear memory.

Coredump

SONUS00136696: MNS20 coredump results from ping command syntax error.

The code has been modified to fix this issue.

Severity 3: Coredump could occur.

Signaling

SONUS00136899: GSX did not resend UPDATE after 491 response in ss7 to SIP call.

UacUpdateRetryExpiredCmd has been modified to check for correct offer answer state for processing the Retry Update Timer.

Severity 2: UPDATEs may not be resent if 491 Response is received.

CLI

SONUS00131783: The command sh alarm bt circuits showed the same output as sh alarm isup circuits. In btCktAlarms.scr, the variable is compared with "" instead of " " (which is returned when there is no data) and it resulted in printing ISUP service circuits also.

The fix has been provided such that variable is compared with the right value.

Severity 3: The output of both the commands comes out to be same when ISUP circuits are configured and BT circuits are not configured.

PXPAD

SONUS00139667: Server card coredumped due to a null pointer reference.

The code has been modified to handle the null pointer.

Severity 2: A switchover will occur if redundancy is configured for the server card that coredumped. If redundancy is not configured, the server card will reset and calls may get dropped.

 

ISUP

SONUS00141875: Coredump occurred due to potential memory leak in ISUP during glare conditions.

The memory leak is resolved by freeing the old memory before allocating new memory.

Severity 3: Memory leak could occur, resulting in a coredump.

Trap

SONUS00142882: GSX did not generate “Inservice" event/trap.

During a network outage that affects connectivity to both Standby and Active PSXs, the GSX will incorrectly generate two Inservice Traps for the configured Standby PSX when both PSXs reconnect and return to service.  This issue only occurs if the configure Standby PSX reconnects first.

The issue was resolved by correctly sending a trap for the appropriate PSX that is returning to service.

Severity 3: "Inservice" trap for PSX could be missing.

IP Routing

SONUS00137531: After an MNS26 switchover, two empty routes are created on the MNS management ports which deny access to OAM network until deleted.

The code has been modified such that a check has been added to ensure 0.0.0.0 IP addresses as destinations for routes disqualifies them and prevents their erroneous addition to the routing table.

Severity 2: MNS26 switchover can cause empty routes to be created.

Signaling

SONUS00143403: When GSX responded with 488, Disconnect reason was different between Reason header and MIME (REL).

The code has been modified to fix this issue.

Severity 2: Disconnect reason between “Reason header" and MIME (REL) may not be the same.

Signaling

 

SONUS00144474/ 144171/ 143490: SBC sends SIP 488 with malformed SDP in response to re-INVITE that changes codec.

The code has been modified to send SIP 488 without SDP.

Severity 2: SBC may send SIP 488 with malformed SDP in response to re-INVITE that changes codec.

Signaling

SONUS00138704: If both the GSX and the peer endpoint simultaneously initiate separate H.245 connections, neither H.245 connection will be established.

The code has been modified to fix a bug that prevented the H.245 from being established.

Severity 2: If both the GSX and the peer endpoint simultaneously initiate separate H.245 connections, neither H.245 connection will be established.

 

CDR

SONUS00145313: In a Overlap call, the START/STOP CDRs had incorrect values for Called Number. Overlap digits were added in "Called Number" as well "Extra Called Address Digits" in CDRs.

Note: This happened when enforceMaxDigitsForValidPhoneNumber flag was enabled in PSX.

The code has been modified to avoid having overlap digits in both fields "Called Number" and "Extra Called Address Digits" in CDRs.

Severity 3: In case of Overlap call, START/STOP CDRs can have incorrect values for Called Number.

Call Processing

SONUS00145418: MNS stopped working and there was no way audio during the calls.

The code has been modified to correct the maximum monitor range.

Severity 1: MNS can stop working and “no way audio” could be experienced during the calls.

Call Processing

SONUS00144976: There were call failures after MNS and PNS swaps.

The code has been modified to ensure the signaling address is bound to the proper interface according to the configuration.

Severity 1: Calls could fail.

Configuration

 

SONUS00142263: The following log message was confusing to the customers:

  “Route destination gateway cannot be

    ZERO (0), rejected route.”

 

The log message has been modified as follows:

“Management NIF nexthop ZERO(0)

   detected - automatic default gateway

   route was NOT added for interface XX”.

Severity 3: None

Call Processing

SONUS00145354: Calls failed after an upgrade.

The code has been modified to set the correct header length when COS is specified by the user.

Severity 1: Calls can fail after the LSWU.

Coredump

SONUS00145208: LSWU from 8.4.1R000 to 8.4.4S001 failed and MNS coredumped when running in SENSITIVE mode.

The code has been modified to avoid the coredump.

Severity 1: LSWU could fail when running in SENSITIVE mode.

CLI

SONUS00138993: The command SHOW POLICER OFFENDERS SERVERPROTECT SUMMARY clears each time the command is run. 

The code has been modified to add CLEAR option in CLI commands for the user to decide if they want to clear the past data.

Severity 3: It can be hard to troubleshoot live issues.

Coredump

SONUS00145627: Standby PNS card coredumped.

This coredump was related to accessing a null pointer.

The code has been modified to add a defensive check for the null pointer.

Severity 3: PNS card can coredump.

Signaling

SONUS00146364: GSX did not delete registration even after Expiration Time. It took almost double the value of External Expiration time to delete the AOR.

Note: This happened only when internal expire time was less than 1000 seconds.

The code has been modified to add a check to ensure that the external timeout value is always less than internal timeout value.

Severity 3: GSX may not delete registration even after the Expiration Time.

CLI

SONUS00145520: TLS sessions could not be deleted. The following error message was received:

error: The object instance does not exist.

This happened because GSX could not find the TLS session while deleting it with the command: DELETE SIP TLS SESSION SIGNALING PORT ... PEER ... SESSION ...

The code has been modified to fix this issue.

Severity 2: TLS sessions cannot be deleted.

Logging

SONUS00144915: Post-upgrade, "UacReleaseCallCmd' event was filling up DBG log set to ‘MAJOR’ level.

The code has been modified to fix this issue.

Severity 2: The logs could be filled up by these messages.

Call Processing

SONUS00146124: SSRC Randomization was not taking effect on the egress leg upon egress re-INVITE.

The code has been modified to randomize the SSRC irrespective of the call leg, if PSP is configured to randomize and when res chain is re-built.

Severity 2: SSRC randomization may not occur on the egress leg even if it is configured on the PSP.

 

Signaling

SONUS00144502: MTRM functionality was not working correctly. When two client MTRGs on the same SBC initiated connection/registration towards the server MTRG, there was an issue that the subsequent registration request from the same SBC node was treated as a re-registration and not taken as a new connection request.

This is fixed by checking additionally against the IP address and port of the source. This prevents the server from treating a new connection request as refresh registration.

Severity 1: MTRM functionality may not work correctly.

Call Processing

SONUS00143350: Calls from Comcast resulted in dead air. This issue occurred during a call hold scenario when the data path mode changed for on hold that goes through a late media Re-INVITE from the remote side. The call processing state machine was not updating the active PSP for the call to be sent in the answer message for subsequent offers. Thus, GSX sent wrong SDP in the answer messages.

The code has been modified to fix the issue.

The Call Processing state machine is modified to update the new active PSP in answer messages so that GSX would send the correct SDP.

Severity 1: No audio in on/off hold scenarios.

Signaling

SONUS00132510: In certain network configurations, PNS cards exceeded the ARP table cache size of 256 entries, causing a PNS switchover.

The table size is increased to 1024 entries to accommodate this type of network configuration.

Severity 1: PNS card could switchover.

Signaling

SONUS00143151: After MNS switchover, ISUP messages were not sent out from SS7 links directly terminated to GSX.

The code has been modified to fix this issue and ISUP messages are sent out after MNS switchover.

Severity 2: After MNS switchover, ISUP messages are not sent out.

M3UA Routing Key Registration

SONUS00144284: GSX could not register to SGX when active MNS links bounced.

The code changes done for fixing the previous M3UA related CQs (126905, 134907, 141478) are reverted and used BICC check for the code that is introduced for BICC.

Severity 2: CICs could remain unregistered.

Signaling

SONUS00143317: Content-Type for INFO message was incorrect for SIP-I. The issue was that the SIP stack was not copying the multiparty flag from the SIPSG application layer.

The code has been modified to extend the fix of CQ138181 to SIP request messages such as INFO, PRACK, UPDATE, ACK, and Re-INVITE, which can send message bodies. The flag multiparty Flag is copied from the transparency data provided by the application while formatting these messages.

Severity 2: Content-Type for INFO message can be incorrect for SIP-I.

CPU Congestion

SONUS00138209: CPU congestion occurred on two GSXs.

The code has been modified such that if CPU goes over 85%, FM task of card will dump the highest offending task.

Severity 2: CPU congestion could occur.

 

Signaling

SONUS00144207: GSX incorrectly interworked the cause value from REL to CANCEL.

The code has been modified to fix this issue.

Severity 2: The REASON header in CANCEL request can contain the wrong cause value.

Resource Usage

SONUS00145503/ 146039/ PCR 8376: For G711-G7XX call if both the call legs have 2833 configured and DTMF remove is enabled, 2 PXPADs were getting used.

The code has been modified to use single PXPAD.

Note: Prior to 7.3.7 GSX was using single PXPAD for such calls.

Severity 2: Unnecessary additional resources (PXPAD) can be used.

 

Signaling

SONUS00139192: MTP2LINK comes up only after GSX restart, when deleted and re-created.

The logical sequencing of deleting the MTP3 link is modified.

Severity 2: When the MTP3 link is deleted and re-created, the link won’t come back to INSERVICE/AVAILABLE state.

Coredump

SONUS00112619: GNS coredumped with edge condition related to SIP translating the host address.

The code has been modified to fix this issue.

Severity 2: xNS card can core on an extreme edge condition related to SIP translating the host address.

 

LSWU

SONUS00145493: Users could not enable client and server connection ports post upgrade. Prior to upgrade client and server connection ports were disabled.

The code has been modified to fix this issue.

Severity 3: Cannot establish connection between client and server connection ports.

 

Protocol Conformance

SONUS00109277: 404 with Reason: Q.850; cause=26 failed to map correctly. Cause value 26mapped to some random cause value that it tried to insert into the ISUP message. Since this random number is outside of the Spec rage of 1 to 127, CC111 protocol error is received.

The code has been modified to initialize the cause value correctly.

Severity 3: Cause=26 can fail to map correctly.

Signaling

SONUS00143404: This CQ fix  enhances GSX to send H.245 Address in CALL PROCEEDING/ ALERTING/CONNECT/PROGRESS messages whenever “Don’t Send Facility Message” option is enabled in IPSP and H323 peer advertises support for H.245 Tunneling in SETUP message and fast start support is disabled in H323 peer.

 

Note: The CQ has been raised to address ISSUE ONE of RTD 8094. Original CQ is SONUS00138400.

Severity 3: H323 to SIP calls could fail

Signaling

SONUS00144651: GSX dropped the call due to PCI (though the parameter was not in the IAM).

The code has been modified to fix this issue.

Severity 3: If an IAM includes a PCI with a parameter that is not actually in the message, (and other parameters in the message) the call may not complete.

Logging

SONUS00145378: Ingress Protocol Variant Spec Data field was missing in CDR. As part of optimization of PAM messages, a defect SONUS00134966 arose, where in certain cases Egress Protocol Variant Specific Data field of CDR records was being overwritten.

This was resolved, but introduced another defect SONUS00145348 where the Ingress Protocol Variant Specific Data Field was not being populated.

This issue has now been corrected so that all Protocol Variant Specific Data Fields are populated properly and are according to the configured features.

Severity 3: Ingress Protocol Variant Spec Data field could be missing in the CDR.

Regression

SONUS00145947: Junk value was printed in DBG logs for SAM message.

The junk log was observed during Overlap dialing scenario.

The code has been modified to remove the event log notification that does not carry any information required to be displayed. The DBG logs are now displayed correctly after receiving SAM message

Severity 3: Junk value can be printed in DBG logs.

Signaling

 

SONUS00142555: In SIP-SIP call scenario, on expiry of the Timer B in Overlap Dialing, GSX sent 503 response instead of 504. Also Crankback/HPC call will not work on Timer B expiry from the first route.

The code has been modified to fix this issue.

Severity 3: On expiry of Timer B, GSX can send 503 response instead of 504. Crankback/HPC call may not work in such situation.

DSP

SONUS00142732: CNS/SPS card coredumped with error code 0x00001040.

The following changes have been made to fix this issue:

  • Added enhanced error checking before attempting to activate CNS/SPS.
  • Added additional logging to help identify the call that is attempting to activate a CNS/SPS with an out-of-range resource ID.
  • Removed logic that cores a server if it repeatedly gets bad frames from a CNS/SPS.

Severity 1: CNS/SPS card could coredump.

 

Overload

SONUS00129586: The cable for port 2 was pulled and reseated. Calls began to fail MC2 reported. CPU utilization was 100%.

The code has been modified to address the faulty adapter issue.

Severity 1: MNS switchover will happen.

Call Processing

SONUS00128009: Calls failed after the MNS switchover.

Added new mechanism in XRM to do NIF audit between MNS and PNS when there is a MNS/PNS switchover event. XRM on MNS will mark the NIF once it is audited with PNS. Then XRM on MNS will audit the NIF with NRS.

Severity 1: Calls can fail after the MNS switchover.

Call Processing

SONUS00124561: Under certain configurations, unauthorized calls were allowed to complete.

The code has been modified to properly deny unauthorized calls.

Severity 1: Under certain configurations, unauthorized calls may be allowed to complete.

Call Processing

SONUS00123707: 486 response to REFER initiated INVITE caused call termination in an on-hold call scenario.

The code has been modified such that the call does not get terminated when the 3rd party sends 486 Busy response to the INVITE (refer).

Severity 2: An on-hold call can get terminated (during a 3rd party SIP REFER call).

Call Processing

SONUS00131272: Video GW-GW transparency - re-Invite new offer did not work.

The code has been modified to fix this issue.

Severity 2: Call may teardown or revert to audio only.

H323

SONUS00123240: H.323 to SIP calls got disconnected after blind transfer from Cisco Call Manager (CCM). H.323 blind Xfer got disconnected after four rings or received no audio if answered.

The code has been modified to fix this issue.

Severity 2: Calls could fail.

Signaling

SONUS00128681: SIP (video) GW-GW transparency need dynamic memory for cache message.

The code has been modified to rework cache message for video GW-GW.

Severity 2: Calls may fail.

Call Processing

SONUS00128976: SIPS 491 with late media re-Invite in call revert. SIP-to-SIP calls did not complete.

The code has been modified to fix this issue.

Severity 2: Calls may fail.

Signaling

SONUS00128784/SONUS00114896: 911 operator calls were not complete over CAS. When a call was placed on CAS trunk group, the switch did not relay the hardware answer from the network to the ingress side.

The code has been modified such that the CAS logic issue is fixed so that the network answer will be relayed to the ingress side. Also, the core dumps that were seen with CQ 112681 are addressed in the CQ 114896.

Severity 2: CAS trunk group does not relay the hardware answer from the network.

LSWU

SONUS00127931: During LSWU, the standby CNS card while transitioned to active coredumped.

LSWU conversion routine was modified to fix this issue.

Severity 2: CNS card during LSWU could coredump while transitioning to active.

Signaling

SONUS00129499: DTMF digits were not being sent in a P2C call.

The code has been modified such that the DTMF digits are detected.

Severity 2: DTMF digits may not be sent in a P2C call.

Call Processing

SONUS00120773: GSX failed to transit National CPC in Russia-ISUP IAM to Russia SIP-I encapsulated data.

Note: This CQ has been fixed as part of PCR 1223.

The code has been modified to correctly map the Russia CPC into the MIME content of Egress SIPI.

Severity 2: GSX can fail to transit National CPC in Russia-ISUP IAM to Russia SIP-I encapsulated data.

Signaling

SONUS00130330: GSX failed to send all configured AMR-WB mode-sets. However, this leads to inability to negotiate proper codec or setup a call, depending upon remote peer configuration.

The code has been modified to read the complete mode set for AMR-WB, when formatting the SDP.

Severity 2: GSX fails to send all configured AMR-WB mode-sets leading to inability to negotiate proper codec, or even setup the call at all, depending upon remote peer configuration.

Protocol Conformance

SONUS00130334: The GSX should only include “SCP Carrier Information” of the CIT parameter in the ACM when the PSX modifies it.

The code has been modified so that if the “SCP Carrier Information” returned from the PSX matches the one in the IAM, the ACM will not include it.

Severity 2: If the IAM includes “SCP Carrier Information” the ACM will also include it, even when not changed by the PSX.

Signaling

SONUS00127974: INVALID_ROUTE_ID on bregsx12 after 7.2.10F1 upgrade. CICs have failed to recover with both local and remote bounces.

Updated MTP3 stack to populate route ID properly if the route and link are both accessible during the switchover.

Severity 2: INVALID_ROUTE_ID on bregsx12 after 7.2.10F1 upgrade.

Signaling – MTP2 link

SONUS00128673: MTP2 link did not come up when set to “BASIC”.

Currently the ECM change in MTP2 profile does not get to apply to the MTP2 link until MTP2 link state has been bounced because MTP2 stack doesn't support it. The ECM is only being passed in when MTP2 link is initialized.

Updated MTP2 COM to only allow user to change ECM configuration in MTP2 link profile when there is no active link using it.

Severity 2: MTP2 link may not come up when set to “BASIC”.

Signaling

SONUS00121185: Optical interface sent RFI on path level, causing wrong remote status.

The code has been modified to fix this issue.

Severity 2: Optical interface could send RFI on path level, causing wrong remote status.

Resource Manager

SONUS00115630: PNS card switchover occurred due to internal software failure.

Debug logic has been added to help identify the root cause of this issue.

Severity 2: PNS card could switchover due to software failure.

Coredump

SONUS00130540: Multiple PNS cards in each of the 3 redundancy groups coredumped with a 51 error code within a couple of minutes of each other. Username in the Diversion header was null, which caused the crash.

Added defensive check to fix this issue.

Severity 2: PNS cards can coredump.

CLI debug command

SONUS00119103: PNS MAC counters could not be displayed with R-build because the CLI debug command "seeq" was not included in the R-build.

Access level for "seeq" debug command is changed such that "seeq" debug command will be available for R/F/S builds.

Severity 2: PNS MAC counters could not be displayed with R-build.

Call Processing

SONUS00128121: REFER method failed when brackets were missing from the FROM and TO headers.

The function to parse the message is changed to move and process the semicolon (;) which is used to indicate the end of the parameter. Thus, the parser will not throw an error when the angle brackets are missing in the FROM and TO headers.

Severity 2: REFER method could fail when brackets are missing from the FROM and TO headers.

Coredump

SONUS00116049: Receipt of RTP packets with size less than RTP_HEADER_SIZE (12-byte) led to DSP crash.

The code has been modified such that the RTP packets having size less than RTP_HEADER_SIZE (12-bytes) are handled.

Severity 2: RTP packets with size less than RTP_HEADER_SIZE could not be handled.

Coredump

SONUS00129077: SIP- SIP video call caused core.

The code has been modified to avoid the coredump.

Severity 2: SIP- SIP video call could coredump.

Signaling – SIP

SONUS00098354: Memory corruption when copying transparency content.

The code was modified to detect this and clean up call. Extra debug messages were also added.

Severity 2: The CNS86 slot8 switchover may fail.

DSP

SONUS00097931: The crash was around StartBSPPipeLine() with high number of interrupts occurred. The interrupts were enabled, while the BSP/DMA was reset and reconfigured. The TimerSlot and the ISR was also SET, while the bit was monitored in IFR.

The code has been modified to save/set/reset Timerslot and IsrBSP bits in proper manner.

Severity 2: The crash can occur around StartBSPPipeLine() with high number of interrupts occurred.

Call Processing

SONUS00123411: SIP-GW-GW-SIP could cause major error “SipsMsgReleaseCmd: Invoked with invalid message”.

The code has been modified to fix this issue.

Severity 2: There can be high MNS CPU and GSX may not respond to the calls.

Call Processing

SONUS00126922: Loss of ISUP to SIP calls when CNS card is reset. When CNS card is reset, calls on another PNS and CNS card fail for small duration (maximum 1 second). This issue is caused due to NIF statistics reset to maximum on other active cards causing interface selection failure, when CNS card becomes active after reset.

The code has been modified to prevent the NIF statistics reset on other cards, when CNS card becomes active after reset, which will avoid call failures.

Severity 2: There could be loss of calls.

XRM

SONUS00110991: There were XRM issues and marlin errors after GSX was upgraded to 7.3.4R003.

The code has been modified such that specific checks are added to filter out the instance where the standby route has already bound to the requested active route and changed SYS ERR to MINOR DBG log for the real mismatch cases.

Severity 2: Upgrade can rollback.

Billing

SONUS00092154: If a comma is present in the following sub fields: To (subfield 3), Contact Header (subfield 13) and Userinfo/Hostname (subfield 9) of the SIP Protocol Variant Specific Data it could affect billing systems from parsing the accounting record properly.

The code has been modified to parse these fields and replace any comma with a semicolon.

Severity 3: Affect on

proper Billing.

Core dump

SONUS00119595: PNS10 had failed with error code 0x0000030a L3 reported slot 4 switched over with a core dump.

The code has been modified to save the bad packet buffer from shared memory in ICM debug memory for debugging if it happens again in the future.

Severity 3: An error can occur in the PNS 10.

Signaling

SONUS00130350: After upgrade, SIPFE sends an error message. Calls were not completed through primary route due to SIPFE errors after upgraded. The error was due to PNA nifgrpTbl does not have the correct NIF group ID for the NIF.

The code has been modified to correctly populating the nifGrpIndex in NIF and SIF to insure NRS/NRA to update PNA NIF group table properly.

Severity 3: Calls can fail after PNS switchover.

Signaling

SONUS00129592: For SIP-ISUP calls which use INFO message for DTMF, the GSX did not release VPAD Swea Channel even when the call is released.

The code has been modified to release VPAD Swea during call release and to clear all Swea Channels which are in use during the call cleanup.

Severity 3: Calls can fail due to un-availability of VPAD Swea Channel resource.

Call Processing

SONUS00124591: There were call failures with CC42 server equipment congestion.

The code has been modified to update OSPF host route with proper ID after PNS switchover.

Severity 3: Calls can fail.

Billing

SONUS00126176: CDR field 20 in Attempt record (Number of Called num Translations) was incorrectly set to “1” if the translation occurred on a prior switch. This field must only reflect the number of translations that were performed by the PSX. In addition to the issue above, the usage of the following four fields were documented incorrectly in the GSX Operations Guide:

1. Called Number Before Translation #1

2. Translation Type 1

3. Called Number Before Translation #2

4. Translation Type 2

These fields can contain values from translations performed by a prior switch. Note: The documentation will be modified in the next major release. The code has been modified to address the first part of this issue.

Severity 3: CDR field 20 in Attempt record can be populated with incorrect LNP information.

Fault management

SONUS00122371: CNS card has multiple switchovers and active and standby CNS get into an oscillation mode. The problem is FM declaring a DSP failure wrongly. Additional counter was added to detect DSP failure correctly.

The code has been modified to read the online cards on the Fault Management (FM) and the Maker/IXP counters to report as an event to MNS FM. EVENT_MAKER_BAD_SDU_LEN is added to notify the error counters to the FM on MNS card.

Severity 3: CNS cards may switchover repeatedly and can cause to coredumps.

Coredump

SONUS00129708: DSP coredumped on running iLBC load of 10cpc on the CNS card with call-hold time of 20 seconds.

The code has been modified to avoid the coredump.

Severity 3: DSP could coredump.

Call Processing

SONUS00107300: GSX was unable to handle

the DFC mid call action from PSX while

GSX was waiting for the trigger response, which resulted in the call getting released.

The GSX code has been modified to handle the reception of mid call action pertaining to the DFC from PSX while GSX is waiting for Trigger Response during the CAMEL_ETC script execution. The SRF leg will be released while processing the CAMEL_CONNECT script.

Severity 3: GSX cannot handle the DFC mid call action from PSX while GSX is waiting for the trigger response. This can result in GSX releasing the call.

Signaling

SONUS00116668: H323 calls are not released immediately when we receive a SETUP with an incomplete phone number and with canOverlapSend option set to FALSE.

The code has been modified to immediately release an H323 call if the SETUP contains an incomplete phone number and the canOverlapSend option is set to FALSE.

Severity 3: H323 calls may not release immediately when we receive a SETUP with an incomplete phone number.

Coredump

SONUS00128310: GSX crashed doing direct replace of the PNS cards in redundant slot. The PNS when partially inserted continuously generated Control Bus interrupts, making the MNS busy processing the false CB interrupts, eventually leading to watchdog reboots.

Fix is to detect the false CB interrupt, disable the interrupt line, and enter polling mode on MNS. Once stabilized as a result of the card properly plugged in, the CB interrupts are enabled again.

Severity 3: If the card is improperly or partially inserted, the MNS could reboot as a result of continuous CB interrupts generated by the half inserted card.

Call Control

SONUS00112072: CAMEL Apply Charging Report (ACR) was not reflecting correct total amount of usage on Prepaid calls.

The code has been modified such that ACR has correct time reported after the call is disconnected.

Severity 3: Incorrect billing on Prepaid calls which encounter the coding error.

Signaling

SONUS00126569: TOS field was not setup correctly in XRM flow change.

The code has been modified to restore the original TOS and COS.

Severity 3: Incorrect TOS field could be applied to RTP stream.

DRM

SONUS00126026: When a DSP channel is deallocated without deactivate first, the data fields in DRM's channelRes are cleared before channel status event sent from DDH and then the wrong packet type was used to check packet outage and caused invalid events generated.

The code has been modified to reserve the Channel Packet Type in the DrmResFree(),when resetting DSP to Channelres.

Severity 3: Invalid events can occur during upgrade.

Coredump

SONUS00115477: MNS coredumped while attempting to register to the SGX when ISUP manager was down.

The code has been modified to fix this issue.

Severity 3: MNS could coredump.

Signaling

SONUS00128920: GSX tried to register all CICs from all nodes to SGX4K. When SS7 node was migrated, all CICs for a slot were registered. This leads to a problem when the GSX has multiple nodes configured.

The code has been fixed to register only those CICs belonging to the nodes being migrated.

Severity 3: On migrating the SS7 node, overlapping error can occur while GSX registers the CICs slot.

Call Processing

SONUS00120533: H323SG received RequestMode for G711 when already doing G711 and did not respond.

The problem was that the code was not sending a response to a request mode message when there was no change to media information.

Added a check for request mode received with a network offer. If so, to process (break and reestablish media) even if nothing has changed with the media.

Note: When processing a request mode, make sure that PSP changes are handled for direct media.

Severity 3: PCR4777 TS 17.2 fax call failed, if Cisco Fax protocol pass-through G711u.

Coredump

SONUS00129433: Need to add defensive code to avoid accessing or freeing NULL pointer in association with CQ120600.

Defensive code has been added to fix this issue.

Severity 3: Potential MNS coredump could occur.

Traffic management

SONUS00129682: Alarm appears to indicate 90% of our capacity on the SIF. But, the SIF displayed high water mark alarms on PNS 40, where load was shown 3616 with "show nif sub all stat" and speedCur was on negative value with "nif slot 1". This occurs due to load balancing traffic. But, on reset manually, the problem was not noticed.

The code has been modified to update the bandwidth speed and the SDU values only when there is no error returned.

Severity 3: SIF may display high water mark alarms on PNS 40.

Call Processing

SONUS00115533: SMM on the SBC 5200 fails to trigger on response with specific statusCode. Introduced a new type SIP_TYPE_REQUEST_ALL=0 to indicate that the request type is wildcarded.

Severity 3: SMM on the SBC 5200 fails.

Coredump

SONUS00126635: SPS/PNS cored multiple times on different slots. The Hash linked list was in infinite loop situation in SIPS sub-system. One of the next pointers in list was pointing to one of the previous pointer. So search in the linked list did not work and caused memory corruption, and coredumps.

The code has been modified to fix this issue.

Severity 3: SPS/PNS cards could coredump multiple times on different slots.

Signaling

SONUS00128986: The debug command getcicstatus considered the node name as an argument and displayed CICs with matching RPCs instead of the node index.

The code has been modified to verify both the node and the RPC. This ensures that the debug command does not show CICs from the other nodes.

Severity 3: The debug command getcicstatus can show CICs from other nodes.

Signaling

SONUS00128680: The first 183 response that the egress GW sent to ingress GW was missing the Transparency headers.

The code has been modified to pick up the latest response cache message for sending out the response.

Severity 3: The first 183 response received from egress may not have the Transparency headers.

CLI

SONUS00122923: Users were not able to delete the routes from the route table, nor configure the sub-interface OOS and state disabled since the GW was in OOS status (although it was enabled and in-service on the Admin view).

After NIF/SIF resource being deallocated, the admnState was at ASTATE_IS to prevent users from issuing multiple OOS requests while NIF/SIF is waiting for XRM reply.

Modified FSM to update the admnState properly so that users are allowed to disable the NIF/SIF due to network issue.

Severity 3: Users may not be able to delete the routes from the route table, nor configure the sub-interface.

Coredump

SONUS00125114: Prefetch or data was aborted due to a hardware failure, which caused the IXP based cards to crash.

The code has been added to identify and log the nature of prefetch/data abort. If it is detected as a recoverable prefetch/data abort, every attempt is made to recover from it, thus preventing the card crash. However, if the abort is unrecoverable and imprecise, the system crashes with enough statistics to nail down the root cause.

Severity 4: IXP cards could coredump due to a hardware failure.

Call Processing

SONUS00114225: The pes.log does not show that a PSP with G711SS was selected and sent to the GSX. But, the GSX shows that the call tried G711U_SilenceSupression. For PASSTHRU calls, if the offer contains multiple codecs, then the NRMA wrongly adds SS to codec for filling CDRs.

The code has been modified to fix this issue.

Severity 4: Show call detail may wrongly display SilenceSupression even though it was not used.

MTP2

SONUS00131194: MTP2 trap was not being cleared. The trap sonusSs7Mtp2LinkInServiceNotification was not being called when the link came into service due to an MSU from the network.

The code has been modified to fix this issue.

Severity 4: Without this fix the trap may not reflect the state of the link.

Signaling

SONUS00132104: GSX was not included “To header tag” in the SIP INFO message.

When PRACK disable, and GSX sends INFO request for early dialog, GSX may not include tag in To header.

The code has been modified to save the remote tag after receiving 18x. So that GSX can send out INFO with correct peer tag in To header.

Severity2: GSX may not include To header tag in the SIP INFO message.

Call Processing

SONUS00117328: There was a call failure on a specific slot on the CNS card. The actual problem lies in the failure of the “DSP leak detection” logic. CQ78526 addressed this problem when the DSP was stuck in the activation mode, but not in the deactivation mode, as is the case here. The code has been modified to fix this issue.

Severity 1: 90% of the traffic may be dropped, which can result in call failures.

Call Processing

SONUS00122914: Calls were getting rejected after an MNS switchover, followed by the error message: Max System Call Limit Set The issue with this was that a standby MNS had a bad sonusCallPegCountsActiveCalls.

Added a defensive code when a standby MNS transitions to active to make sure the sonusCallPegCountsActiveCalls is correct only when becoming active.

Severity 1: Calls can fail after the MNS switchover.

Call Processing

SONUS00126116: NIF group containing SIFs was getting synchronized before synchronization of SIF on the standby card. This resulted in SIF lookup failure during nifgroup synchronization.

The order of synchronization has been changed accordingly.

Severity 1: Calls can fail during and after an upgrade.

Coredump

SONUS00124403: After an upgrade CNS card core dumped continuously, resulting in call failures.

The code has been modified to correct the error that caused the coredump.

Severity 1: CNS card could coredump continuously, which could result in call failures.

Call Processing

SONUS00117328: There was a call failure on a specific slot on the CNS card. The actual problem lies in the failure of the “DSP leak detection” logic. CQ78526 addressed this problem when the DSP was stuck in the activation mode, but not in the deactivation mode, as is the case here.

The code has been modified to fix this issue.

Severity 1: 90% of the traffic may be dropped, which can result in call failures.

Call Processing

SONUS00122914: Calls were getting rejected after an MNS switchover, followed by the error message: Max System Call Limit Set The issue with this was that a standby MNS had a bad sonusCallPegCountsActiveCalls. Added a defensive code when a standby MNS transitions to active to make sure the sonusCallPegCountsActiveCalls is correct only when becoming active.

Severity 1: Calls can fail after the MNS switchover.

Call Processing

SONUS00126116: NIF group containing SIFs was getting synchronized before synchronization of SIF on the standby card. This resulted in SIF lookup failure during nifgroup synchronization.

The order of synchronization has been changed accordingly.

Severity 1: Calls can fail during and after an upgrade.

Coredump

SONUS00124403: After an upgrade CNS card core dumped continuously, resulting in call failures.

The code has been modified to correct the error that caused the coredump.

Severity 1: CNS card could coredump continuously, which could result in call failures.

Call Processing

SONUS00123318: GSX9k rebooted during shelf congestion. GSX9k started to experience high call volume. Logs indicated shelf CPU level 3 congestion. Link Detection started to fail and PNS cards start to fail over. Slot 1 eventually switched over to Slot 2 but no core was written. The SPS 80s in the box all core dumped and then aborted the core. The box rebooted 52 minutes into this event.

The code has been modified to fix this issue.

Severity 1: GSX9k can reboot during shelf congestion.

Signaling

SONUS00122850: After an upgrade, PSTN-IP (H.323) calls failed as the VLAN was not configured appropriately on the far end. While resolving this issue, a bug in NRS was noticed which caused watchdog timeout and crash.

The code has been modified to correct the endless loop in NRS.

Severity 1: MNS can crash due to watchdog timeout.

DSP

SONUS00122623: This defect affects ECM enabled T.38 call when GSX or SBC is acting as a T.38 emitting gateway. SPS70 is the card affected. GSX/SBC failed to encode T.4 image data correctly, and thus caused receiving gateway to repeatedly send Partial Page Request (PPR) messages, which in turn caused GSX/SBC to continuously retrain and train down. As part of the resolution, the T.38 fax functionality is temporarily rolled back to the previous working version. The following CQs are re-introduced by rolling back the T.38 stack:

• CQ118124

• CQ110620

• CQ107121/103111

Severity 1: T.38 ECM fax call can fail with no actual page data transmitted.

Call Processing

SONUS00123479: Transcoding calls failed after an upgrade from R7.1. The logs indicated the DRM handles as “NULL” when NRMA allocates XPADs for the call after the upgrade.

The code has been modified to fix this issue.

Severity 1: Transcoding calls could fail after an upgrade from R7.1.

Signaling

SONUS00125575: The GSX (configured network side INS1500) received a SETUP with an illegal IE. The GSX accepted the call, but sent the status call as State 0. This caused the other end to release the call.

The code has been modified such that the GSX responds to the call with State 1.

Severity 2: Call may not be complete.

Coredump

SONUS00122842: During PNS switchover, the call was disconnected and coredump occurred.

The code has been modified to remove all the messages in the msgList with matching GCID.

Severity 2: MNS core dump can occur during PNS switchover.

Call Control

SONUS00124069: PCR3222 - SIPSG disconnect treatment was broken. As part of the code change in SipSgProcessNrmaUpdateNfy() some structures are always freed, as a result the SIPSG disconnect treatment logic is broken.

The code has been modified to revert the Cpc deletion only in case of fresh creation.

Severity 2: SIPSG disconnect treatment can be broken.

Coredump

SONUS00124792: The “SamProcess” coredumped with a Health Check System error due to a corrupted entry in the Trunk Group Name hash list.

The software has been modified to prevent freeing (and re-use) of the data structure which is on the hash list.

Severity 2: The “SamProcess” can coredump due to a corrupted entry in the hash list.

Coredump

SONUS00114777: GWSG task received an unrecognizable event that got the task to enter the SHUTDOWN state.

The code has been modified to core the card immediately upon receiving an invalid shutdown event for GWSG task.

Severity 2: An invalid shutdown event for GWSG task could be received, which could get the task to enter the SHUTDOWN state.

Signaling

SONUS00119518: When NFS failover occurred, some Announcement files were not sent from CNS to CNS cards due to “Segment Load Failure”.

The code has been modified to fix this issue.

Severity 2: Announcement file cannot be played.

Signaling

SONUS00122652: “Sip Parser” needs to have a strict rule for all type of attributes so that it can ignore all other unknown attributes. Note: This is a follow-up CQ for 115528 and 122115.

The grammar for attributes has been changed to exactly match the attribute. If it does not match, it will fall into “unknown” attributes.

Severity 2: Calls can fail.

Signaling

SONUS00124112: GSX sent incorrect Gratuitous ARP (G-ARP) when a PNS slot was reset. Note: This problem happens only under very specific circumstances:

• Two NIFS in the NIF GROUP used by the GW Sig Port

• ****The link is down on the standby PNS card when the GSXs attempt to switchover a PNS****

This problem does not happen if:

• The PNS Redundancy group is DISABLED.

• If the PNS card successfully switches over to a redundant card.

The code has been modified to put the correct address in the Source IP address field when sending a G-ARP.

Severity 2: Calls can fail.

Call Processing

SONUS00123646: Calls failed when SIPSG received a 200OK when an UPDATE request was still outstanding. (The previous UPDATE from the peer was still processed by SG/NRMA.) Also, when 200OK was received immediately after the UPDATE, NRMA did not change the data path mode and there was a transient time when the local data path and peer modes were inconsistent at SG/NRMA and at SIPSG.

The code has been modified to fix this issue.

Severity 2: Calls could fail.

Configuration

SONUS00125631: A static route existed for a SIF that was no longer on the GSX. Although the card was deleted and removed, the static route still remained configured.

The code has been added to allow users to delete the admin route if the corresponding interface does not exist anymore. Also, modified the routine to cleaning up SIF’s admin route(s) when the PNS40 server is being deleted.

Severity 2: Cannot delete the static route that is no longer on the GSX.

Signaling

SONUS00119591: Users could not enable the state of maximum PDU SIP Signaling controls from the GSX. The issue was that the call blocks were not de-allocated correctly in case when the negative response is generated for call loop-backed on the GSX.

The code has been modified to fix this issue.

Severity 2: Cannot enable the state of maximum PDU SIP Signaling controls from the GSX.

Memory Management - Coredump

SONUS00122196: CNS cards coredumped during a switchover due to memory corruption.

The code has been modified such that after the switchover the standby card comes up and any module on this card may allocate this memory. This ensures that there is no coredump. Note: If this memory is not allocated by any module, it resides as garbage and when the pointer points to this garbage value, coredump occurs.

Severity 2: CNS cards can coredump during a switchover.

Signaling

SONUS00123768: The COLP/COLR feature was not working for ISDN for the following reasons:

1. Even after disabling the Trusted for COL flag on PSX, the Connected Number is still sent to the egress.

2. Even after enabling different flags like “Don't send Connected Line Identity”, “Don't send Restricted Connected Line Identity flag enabled”, “Connected Line Identity flag is enabled”, the Connected Number is forwarded instead of getting dropped.

3. Even with the IGW unsupported with the COLP COLR on the GSX for ISDN, the Connected Number is sent.

The code has been changed to resolve the issue by modifying certain functions.

Severity 2: Connected Number can get dropped.

Coredump

SONUS00124677: There was a software failure during PNS revert. This was a NRMA context data corruption issue as an incorrect context data was being freed up by standby.

The code has been modified to pass along the standby context ID to be destroyed.

Severity 2: Coredump can occur.

Call Processing - Media

SONUS00125066: There was dead AIR/poor audio quality on some calls using DSP resources.

The code has been modified to fix this issue.

Severity 2: Dead AIR/poor audio quality on some calls using DSP resources.

Coredump

SONUS00122643: DSP coredumped during a single PXPAD call. This bug was exposed after DSP hardening. The memory will not be allocated for SideB. If this issue is not fixed, it will eventually throw a BUS error.

Severity 2: DSP can coredump after single PXPAD call.

DSP Coredump

SONUS00124430: DSP crashed after continuous execution of approximately 1000 test cases for universal XPAD. Note: This crash is not reproducible on single or a small group of tests. Decoder object was set to zero.

The incorrect size for the decoder object has been fixed to resolve this issue.

Severity 2: DSP can crash.

Call Processing

SONUS00125113: In certain circumstances in IUP->ISUP interworking the ACM should not be sent on the IUP side until the criteria is met.

The code has been modified such that till certain criteria are met, the ACM is not sent. For example: If all the BCI (in given conditions) is false and Optional parameter is excluded or false, then ACM is not sent.

Severity 2: Even if the criteria are not met, the ACM is sent.

FAX

SONUS00124351: Fax call failed as GSX was sending unnecessary re-INVITE to G.729 (audio), after successful switch to FAX passthru.

The code has been modified to resolve the issue.

Severity 2: Fax call might fail.

Call Processing

SONUS00124830: Fax calls failed when re-INVITE collision w/ 491 response.

The code has been modified to fix this issue.

Severity 2: Fax calls can fail.

Coredump

SONUS00125577: The MGMT IF’s logical IP address was no longer pingable after MNS switchover.

The code has been modified to ensure the route for logical MGMT IF IP address is added properly in pNA.

Severity 2: GSX can become unavailable when the MNS switchover occurs, resulting in unnecessary alerting/trouble reporting.

Signaling

SONUS00125579: GSX failed to send 4xx transparently. GSX after receiving 400 for the 2nd redirect Invite, the call teardown w/o transparency information.

The code has been modified to fix the problem so that 4xx-6xx can relay correctly.

Severity 2: GSX can fail to send 4xx transparently.

Protocol Conformance

SONUS00120190: In an IUP transit case, the GSX did not build the BT IUP Line Identity Address signal properly. GSX added an extra HexB in the LineID.

The code has been modified to fix this issue.

Severity 2: GSX could add an extra HexB in the LineID.

Signaling

SONUS00122693: ICMP routing priority changed after a software upgrade. Because of this, users had to manually add a static route to restore the behavior expectancy as handled in the older software version (before the upgrade).

The ICMP code has been modified to fix this issue.

Severity 2: ICMP traffic can be directed to the MGMT NIFs. This could require manual intervention to add a static route to restore the behavior expectancy as handled in the older software version.

Signaling

SONUS00118339: This is a follow-up CQ for CQ115717. During a Direct Media call, the response to a Refresh INVITE sent by the GSX (after a PNS switchover) to the call leg performing X-dmi signaling caused a new offer to be sent on the other call leg to restart the media flow. However, some IADs were not capable of accepting the new offer.

The code has been modified to fix this issue and required to turn ON “SDP O-line Only Compares” in IP Signaling Profile.

Severity 2: If an IAD does not have the capability to accept a “new offer”, the call could tear down.

Call Processing - SIPE

SONUS00120261: SIPS V07.03.04F007 did not respond to 200OK message. When SIPS proxy received 200OK message with Session-Expires header value greater than 65535, SIPS proxy did not forward the 200OK and the call was released.

The code has been modified to forward the 200OK message with Session-Expires header value of 65535 when the Session-Expires header value received is greater than 65535.

Severity 2: If SIPS proxy receives 200OK with Session-Expires header value greater than 65535, the call could fail.

IP Input Filter

SONUS00125069: IP Input Filters affected NIF indexes outside the configured range.

A new flag has been introduced, which is initialized to FALSE and it will be set to TRUE if NRS has successfully sent the NOTIFY message when enabling the IP input filter. When applying the action, this flag will be checked. If the flag is FALSE, then NRS will log a MAJOR level message indicating the action is not applied.

Severity 2: IP Input Filters can affect NIF indexes outside the configured range.

Protocol Conformance

SONUS00124695: ISUP National specific parameters were not always deleted, whereas when interworking between ISUP variants national specific parameters should be deleted to avoid confusion in the target network.

The code has been modified to fix this issue.

Severity 2: ISUP National specific parameters may not always get deleted whereas they must be deleted to avoid confusion in the target network.

Coredump

SONUS00125930: MNS coredumped when loading TLS certificates. The caCertificate has been correctly initialized thereby avoiding the access of invalid memory later on.

Severity 2: MNS can coredump when loading TLS certificates.

Coredump

SONUS00126284: MNS card coredumped with “FmMonitorCpuUsage” error mentioning that task Id 00000000 is continuously using CPU extensively for long time.

The logic in CPU utilization monitoring module has been corrected to fix this issue. Task Id 00000000 is not the active task.

Severity 2: MNS can coredump.

- CAMEL

SONUS00126075: There was no voice path on call to the media resource (first leg) in a CAMEL scenario. The IAM from the B-party (second call leg) transits through the GSX and it responds with an ACM containing BCI Interworking indicator and OBCI INBAND indicator. Because the GSX did not perform a full duplex cut through (the OVERRIDE INBAND control was enabled on the egress side), the tones from the A-party did not make it to the B-party. Code has been modified in ISUPSG and CC as follows:

• The ISUPSG code is changed to remove the check on the OVERRIDE_INBAND control while sending cut through request to CC.

• CC Progress Event Manager code is modified to ignore the INBAND indicator while GSX has the media leg up as part of CAMEL call flow.

Severity 2: No voice path on call to the media resource (first leg) in a CAMEL scenario.

Coredump

SONUS00126400: PNS card coredumped during SIP registration.

Null check has been added to fix this issue.

Severity 2: PNS card can coredump during SIP registration.

Coredump

SONUS00122695: PNS card coredumped after an upgrade. There is a race condition, which may cause memory corruption if an H323 is in the process of being set up during an MNS or PNS switchover. This memory corruption may cause the card to core.

The code has been modified to prevent the memory corruption.

Severity 2: PNS card may coredump if H323 calls are in the process of being set up during an MNS or PNS switchover. This is the result of a race condition and will therefore not always happen when H323 calls are arriving during a switchover.

Call Processing

SONUS00123089: SIP re-INVITE call was processed as a Fax Relay call.

The code has been modified to check if the NRMA service mode is fax and if T.38 is offered in the PSP as the first offered codec. In case when the mode is not fax and T.38 is offered in index 0, SIPSG skips index 0 and uses the next codec index for sending a Re-INVITE.

Severity 2: SIP re-INVITE call can be processed as a Fax Relay call.

Call Processing

SONUS00120660/120644/119194: When ISUP MIME relay flag was enabled on SIP Service Group, GSX was not relaying MIME body for SIP-I GW-GW calls; due to which SIP-I to SIP-I calls over GW failed

The code has been modified to fix this issue.

Severity 2: SIP-I to SIP-I calls over GW can fail.

Signaling

SONUS00123331: SMM rule caused Parse error. Parse error is due to the removal of the space between the SIP-URIs multiple Contact header instances.

The code has been modified to fix the problem.

Severity 2: SMM rule can cause Parse error.

Coredump

SONUS00112339: There was a software failure, which resulted in PNS failure. There was an invalid pointer reference when logging a DBG message.

The code has been modified to ensure that there is a valid pointer reference when logging a DBG message.

Severity 2: Software failure can result in the failure of the PNS card. Also, there can be loss of redundancy.

Call Processing - H323

SONUS00123027: Stable calls were getting dropped in PNS switchover for H323 calls. H323 Mirror to sby was broken by CQ106828, which changed the behavior of RTM_MSG_CLASS_LATE_CALL_DATA.

The fix is to change H323 Mirror to RTM_MSG_CLASS_CALL_DATA instead.

Severity 2: Stable calls can get dropped in PNS switchover for H323 calls.

Fault Management

SONUS00124774: Under rare conditions, an invalid GCID reported during the call audit would cause stable calls to be cleaned up/terminated.

The code has been modified such that it prevents the stable call terminations and reports the condition in the debug log file.

Severity 2: Stable calls could be lost during call audit.

GSX-PSX Registration

SONUS00104315: GSX to PSX registration did not automatically reconnect after MNS switchover. During certain unknown network conditions and during an MNS switchover, the GSX will not automatically retry its registration sequence to reconnect to the PSX.

The code has been modified to always retry until the connection is successful as this was the intended behavior.

Severity 2: The affected PSX can remain in the “Down” state until its state is manually toggled.

Signaling

SONUS00125073: SIPS does not format the display name within double quotes of the embedded header (PAI) correctly.

The code has been modified to check if the display name is within double quotes and formatting accordingly.

Severity 2: The display name of embedded PAI part of the Contact header may not have quotes.

SIPFE - Redundancy

SONUS00119756: The Maximum SIP PDU Size value was lost on PNS switchover/ revert when both the Distributed SIPFE and Maximum SIP PDU features were used. Note: This problem does not occur when the Distributed SIPFE feature is not used. Standby D-SIPFE did not initialize buffer size properly.

The code has been modified to fix the issue.

Severity 2: The Maximum SIP PDU Size value can be lost on PNS switchover/revert, when Distributed SIPFE feature is used.

Signaling

SONUS00110714: User Service Information (USI) coding was set too strict.

The code has been modified to treat layer 1 protocol of G.711 u-law in the same way as layer 1 protocol of V.110 while decoding and encoding the mandatory variable USI parameter in ANSI. This prevents the USI from being discarded.

Severity 2: The USI parameter cannot be decoded and the call could be released.

Call Control

SONUS00113631: During SIP to CAS call flow, the DTMF was received through SIP INFO messages (using dynamic VPAD), but the same did not appear to be sent on the CAS leg. The Send digit function code has been modified to resolve this issue.

The changes ensure the playout curTs are incremented suitably and subsequently the DTMF digits are dequeued accordingly during complete absence of voice packets.

Severity 2: There can be errors with respect to DTMF tones.

Signaling

SONUS00123415: There was a problem in the routine that searched for the signaling address with the given NIF group index.

Modified the lookup logic to resolve the problem.

Severity 2: There can be multiple entries per address in the logical signaling binding status, impacting the services.

Fault Management

SONUS00116204: There were huge number of loose routing failures and major faults in DBG logs for SIP NOTIFY.

The alert level has been changed from Major to Minor.

Severity 2: There could be huge number of loose routing failures and major faults in DBG logs for SIP NOTIFY.

CIC Registration

SONUS00120048: CICs did not operate properly after an upgrade (or MNS switchover) until they were deleted and then re-grown.

The code has been modified to fix this issue.

Severity 2: There could be MTP errors after the MNS switchover.

DSP

SONUS00120746: There was a server module software failure with error code 0x00001040. This resulted in CNS to core dump.

The invalid pointer code has been modified to resolve the problem.

Severity 2: TPAD can crash resulting in core dump.

Configuration

SONUS00123848: SIP port IP address was stuck on PNS and even got deleted from the GSX. When an attempt was made to create another SIP port with same IP (different ID), the GSX did not allow and returned an error.

Added logic to delete SIP Signaling on PNS.

Severity 2: Unable to create different SIP Signaling port with the same IP address.

Coredump

SONUS00121473: The SIP call entry (address) in the hash table was corrupted and the memory was marked as free.

The code has been modified to the following:

1. Remove the entry directly from the hash table even though it cannot find the match entry and log a major log.

2. Improve the memory pool validation.

Severity 2: xNS card can coredump.

Signaling

SONUS00126208: GSX unnecessarily utilizes XPADs, whereas for an ISUP-GW2GW-ISUP call there is no need for XPADs when dynamics are available.

The code is modified not to use XPADs when dynamics are available on GSX.

Severity 2: XPADs can get utilized even when Dynamic VPADs are available.

Signaling

SONUS00127999: 200OK was being sent with SDP when it was configured to NOT include. When the Minimize flag is enabled, “Send SDP in 200 OK when 18x reliable” flag becomes ineffective. Currently SendNtwkUnchanged() does not control sending of duplicate SDP in next message.

Fix has been added to control it.

Severity 3: 200OK could be sent with SDP although it is configured to NOT include.

Signaling

SONUS00123705: After an LSWU, mask in the routing data (IP Routing table) was changed to 255.0.0.0. User’s attempt to change it to 0.0.0.0 failed. Note: This function is only called for MNS1x. So, this problem did not show on MNS2xs.

The mask error has been corrected when adding the MNS port 2 in pNA.

Severity 3: After LSWU, mask in the routing data can be changed.

Signaling

SONUS00125042: SIP-I Terminating side hold/resume flow with switchover failed to resume.

The code has been modified to fix this issue.

Severity 3: “Call hold” from the terminating party cannot be resumed if a card switchover occurs before the resume happens.

Call Processing – PDD timer

SONUS00126241: Post Dial Delay (PDD) timer failed to cancel for a SIP to ISUP call when response was received from the egress and GSX dropped the call.

The code has been modified to stop the PDD timer if the response is received from the egress.

Severity 3: Calls can fail.

Call Processing

SONUS00126776: After applying multiple media profiles to the servers, calls can fail when one resource pool is full with error “out of compression resources”. GSX product accidently marked both pools as “empty”.

The code has been modified to mark only the correct resource pools as “empty” when necessary.

Severity 3: Calls can fail.

Call Processing -SIP

SONUS00125817: SIP receiving side Filtering did not work after the switchover. The standby card did not update the Filtering profile properly. The code has been modified such that GSX does not generate the syntax error.

Severity 3: Calls can fail.

Signaling

SONUS00123480: There was a bug in ISUP Signaling Profile to configure UPV Exchange Roles with the correct values. For MIB object sonusIsupsgSignalingProfileUPVExchangeRoles, the valid options are “int” representing International and other is “intwk” representing Interworking. As part of the change, “International” is represented by “intnl”. This avoids one option from becoming the substring of another.

Severity 3: Cannot configure UPV Exchange Roles with the correct values.

Call Processing

SONUS00119808: Hong Kong Ring Feature (PCR 2431) - Users could not enable the feature after configuring the PSX. When FLRBT is set on PSX, the Ingress GSX must disregard any in-band tone indications and/or cutthru received from Egress GSX and continue to provide LRBT. But in this case, the LRBT was aborted. Following changes have been made as part of this fix:

• To handle FLRBT for alert/call progress response.

• To not CUTTHRU if FLRBT is enabled and local ring back tone is being played before Network Connect (Answer response) event is received.

Note: The fix has a limitation, which does not fully address the call transfer case.

Severity 3: Cannot enable the feature (Hong Kong Ring Feature) after configuring the PSX.

Call Processing

SONUS00120072: The CAS casfsmstat was broken on GSX4000.

The code has been modified to fix this issue.

Severity 3: CAS debug could be difficult.

Signaling

SONUS00123484: The CDR field “Call Disconnect Reason Sent to Egress” was populated when it should have been empty.

The code has been modified such that the CDR call disc reason sent to egress field is empty.

Severity 3: CDRs can lead to wrong statistics.

Coredump

SONUS00123654: CNS card switched over because channel deactivation was attempted on a DSP channel that was already inactive. From DSP core, there is nothing wrong.

The code has been modified to :

1. Change the log level from info to major when GWFE receives multi-force.

2. Increase the threshold from 5 to 10 and for error timeout case only.

Severity 3: CNS card can switchover.

Call Processing - SIP

SONUS00125825: TCP/RST resulted in the timeout of Timer B. This change affects SIP calls over TCP. When a RST is received for a SYN packet from the remote side while sending out egress INVITE, GSX performs a crank back (if configured) or fails the call based on the TCP retransmit interval. This CQ fixes the case when TCP retransmit interval is set to 0.

The code changes are made in the SIP transaction layer to change this time to 2*Timer T1 value. This will ensure that crank backs happen faster for the call.

Severity 3: Crank back/failure of the call can occur in 64 times the Timer T1 value (in about 32 seconds).

Alarms

SONUS00123611: There was a bug in the alarm trap due to which a wrong event sonusIsupCircuitUnblockedTrapsSuppressed was associated with the alarm notification and the trap.

The code has been modified to associate the correct event sonusIsupCircuitResetTrapsSuppressed.

Severity 3: Due to the bug, a wrong event could be associated with the alarm notification and the trap.

Call Processing

SONUS00125922: During a SIP- SIP re-directed call, GSX generated a “182 Queued” message intermittently”.

The code has been modified to initialize the Event Indicator structure when using for redirection.

Severity 3: During a SIP re-directed call, GSX can generate a “182 Queued” response intermittently”.

Signaling

SONUS00102938: GSX sent media using the same source and destination MAC address.

The code has been modified to force an interface-specific route lookup if the NIF groups of the source and destination are different. This will prevent mis-directed packets to overlap addresses within the source’s NIF group and force them to reach the correct destination (using the correct MAC address).

Severity 3: GSX can send media using the same source and destination MAC address.

Signaling

SONUS00124177: GSX did not send re-INVITE in early media scenario.

The code has been modified such that GSX sends a re-INVITE with previous SDP just after receiving ACK for 200OK.

Severity 3: GSX may not send re-INVITE in early media scenario.

CLI

SONUS00122435: The SHOW ALARMS ISUP SERVICE command showed an incorrect point code format.

The CLI has been modified to fix this issue.

Severity 3: Incorrect point code format can be displayed.

Signaling

SONUS00122575: SS7 gateway status that was displayed after reconnecting a cable for primary SGX4K was incorrect. The SCTP code declares a “peer address change” when it receives data from a new target address. If the heartbeat processing code gets a heartbeat from the “preferred” address, it updates the target address without declaring a peer address change. Later when data is received, the target address change is not detected.

The fix is to modify the heartbeat processing code to declare a peer address change when it modifies the target address.

Severity 3: Incorrect SS7 gateway status can be displayed.

Signaling

SONUS00116107: ANM was received after ACM at the ingress side, instead of a CPG. When a CON (CONNECT) message is received from ETSI side, based on the Charge Indicator value, GSX must send either CPG or ANM. Instead of picking the Charge Indicator value present in CPC BCI information, it was picked from MDA table.

The code has been modified to retrieve the charge indicator information from CPC.

Severity 3: Instead of a CPG, ANM can be received after ACM at the ingress side.

Signaling

SONUS00123742: ISDN D-Channel bounced when other ISDN services bounced. GSX dropped ISDN D-channels for customers not related to actual customers having ISDN issues.

The code has been modified to fix this issue.

Severity 3: ISDN D-Channel can bounce when other ISDN services bounce.

CLI

SONUS00123966: The commands SHOW ALARMS ISDN SERVICE and SHOW ALARMS ISDN BCHANNEL did not display the accurate data. D-channels and Bearer channels that were actually down frequently were not displayed in these reports. Additionally, D-channels that were up and had active calls showed up on these reports.

The CLI script has been modified to fix this issue.

Severity 3: Manual intervention can be required.

Memory Management

SONUS00125566: There was high memory utilization on the SPS card.

The code has been modified to prevent the memory leak.

Severity 3: Memory leak can occur.

Coredump

SONUS00117457: MNS can core when checking the status of the Gateway SIG link if the task is swapped out during the processing of the command. This bug was introduced with 7.3.4.

The code has been modified to avoid the coredump.

Severity 3: MNS can reset when checking the status of the gateway signaling link.

Coredump

SONUS00119451: MNS card coredumped due to a corrupted ICM message that tried to update the announcement file.

Defensive code has been added to avoid the corrupted ICM message.

Severity 3: MNS card can coredump due to a corrupted ICM messages.

Coredump

SONUS00128206/120600: show gateway signaling port all status command resulted in MNS coredump. The size of GWFE_SOCKET_CTRL_STR had grown beyond MEM_BLK_SIZE_LAST (10200) and hence the system was not able to allocate memory.

The size of MEM_BLK_SIZE_LAST has been increased to 10752.

Severity 3: MNS could coredump.

Coredump

SONUS00122189: MNS coredumped on changing the SS7 gateway to M3UA type.

The code has been modified to fix this issue.

Severity 3: MNS switchovers and calls may be dropped.

Signaling

SONUS00125044: While changing SIP signaling slot, there was an error in the SBC.

Modified the routines to check the interfaces in the specified NIFGROUP when running distributed SIPFE.

Severity 3: SBC can give an error while changing SIP signaling slot.

Media

SONUS00125257: CNS86 XPAD DTMF was not working on out-dials after an upgrade to 7.3.6.

The code has been modified such that when a change in SSRC is detected, the packets are dropped and probation is set. The next packet is processed only when it is in sequence and with the same SSRC.

Severity 3: Non-compliant handling of RTP probation can occur. DTMF may not work on out-dials after an upgrade to 7.3.6.

Call Processing

SONUS00123174: NOTIFY message was not routed by the AOR. It was being rejected with a “404 Not Found” message. The issue is that the Registrar did not include the reg-info in the NOTIFY message, due to which the GSX could not find the registration.

The code has modified to add another userinfo lookup from R-URI.

Severity 3: NOTIFY message could be rejected.

Coredump

SONUS00126565: PNS coredumped during an upgrade after executing the RTP Port base change command. The code has been modified to free the XRESs before changing the base.

Severity 3: PNS could coredump, which can cause a delay in the upgrade.

Signaling

SONUS00125101: SIP-I with G722 was being rejected.

The code has been modified to fix this issue.

Severity 3: SIP-I with G722 may get rejected.

Coredump

SONUS00125309: There was a memory leak and SPS card coredumped with error code 0x00000074. The issue was that the SIP Prack entry did not get cleaned up properly when a call was deleted. The code has been modified to skip the check for Prack service when releasing the call.

Severity 3: SPS card can coredump.

Call Processing

SONUS00117486: SPS slot did not host any calls. The bug was related to there being no calls on the card and NRMA thinking there were many. The slot’s counter became inaccurate and had a high watermark value well above all the other servers. This removed itself from being viewed as an available resource by NRM because all of the other servers were least loaded.

The code has been modified to fix this issue.

Severity 3: SPS slot may not host any calls.

Call Processing

SONUS00120118: When receiving a BT IUP CNA message with cause value 7 (Congestion - Reroute Not Permitted), the Disconnect Initiator in the CDR was incorrectly pegged as Internal. This happened because the Internal Disconnect value (Congestion Reroute Not Permitted) is not in the default Q850 values.

The code has been modified to fix this issue.

Severity 3: The Disconnect Initiator for BTIUP cause value 7 could be incorrect.

Redundancy

SONUS00108968: The Redundancy group showed ACTIVESYNCEDHOLDOFF/STANDBYHOLDOFF abnormally. Added a new flag in RTM_REDUND_HOLDOFF_NFY_MSG_STR indicating whether the state must be HOLDOFF (TRUE) or not HOLDOFF (FALSE).

Severity 3: The Redundancy group can show abnormal values.

Accounting

SONUS00118052: CDRs showed inconsistent values for the “Packetization Time” field for pass through calls in a GW-GW call.

For pass through calls, the logic is improved to better determine what packetization time is used for all legs of the call before writing it to the accounting record.

Severity 3: The RTP packetization times recorded in CDR cannot be accurate.

CLI

SONUS00120113: The show command SHOW CAS CHANNEL SERVICE CAS PORT E1-1 CHANNEL 1-3 FSM STATUS did not show the signal bits when the channel was idle.

The code has been modified to fix this issue.

Severity 3: The show command may not show the signal bit when the channel is down.

Call Control

SONUS00103971: The way of counting GHG message is wrong. In Japan egress protocol variant, subfield 62, number of CHG Messages is slotted “before receiving ANM” but it should be “after receiving ANM”.

Modified the restrictions to ignore CHG message in the call processing state before receiving ACM or in the call processing state in between ACM and ANM.

Severity 3: The way of counting CHG message will be wrong.

Call Processing

SONUS00120316: There was unnecessary major log in sipsgValidateMediaDirection, and sipsgSavePsp. It should log major only when there is a real error.

The code has been modified to log only at level INFO or calltrace.

Severity 3: There can be unnecessary major logs.

Media

SONUS00110207: While testing the Teletypewriter Signaling (TTY) functionality for the 7.1.8 release, there were missing characters at term end of the call. A TTY letter is 22ms long, but there was a period of no audio for 5ms. This was a unidirectional issue, as the TTY calls in the opposite direction had no missing/incorrect characters at the term end. The ~5ms discontinuities in TTY letters were introduced due to aggressive suppression by in-band DTMF remove algorithm from D2 technologies.

The “in-band DTMF removal” has been disabled whenever modem tones (BA1 and BA2) are encountered at the beginning of the TTY session establishment.

Severity 3: TTY tones can get dropped out of the RTP stream.

Accounting

SONUS00102154: For calls that ingress using the ISDN protocol the Disconnect initiator field in STOP CDR field 64 will show a ‘0’ when the disconnect comes from the calling party. This issue is only seen if we are configured for “IMMEDIATE DISCONNECT WITH INBAND TONES” and we receive a disconnect with inband info.

The code has been modified to fix this issue.

Severity 3: When the calling side terminates the call, CDR STOP field 64 can be incorrectly set to ‘0’.

User Tool

SONUS00120418: ISDN unit test tool was broken on GSX4000.

The code has been modified to fix this issue.

Severity 5: Does not affect a real application.

NIFs/SIFs

SONUS00130195: The NIFs and SIFs on the GSX became unresponsive to all external input in one of the instances. The logs show that IXP erroring occurs due to nopktbuf increasing at a rate of about 8-14 packets every 5-6 seconds.

The code has been modified to delete the nilan entry if PNA fails to add NIF/SIF.

Severity 2: All packets sent to those NIFs/SIFs would be dropped.

 

 

Signaling

SONUS00130406: There were frequent DSP cores on SPS cards. The problem was that packets of 1-byte appeared in T.38 (fax) channels.

Added defensive code to drop such packets before it reaches fax channels Also, added a new counter to keep track of the count of such packets received.

Severity 2: Coredump could occur.

Regression

SONUS00125811: When H323 overlap to H323 enbloc is done using customized script, the egress SETUP is sent even when the egress is ENBLOC and ingress (overlap) dialing is not complete.

The code has been modified such that Overlap Addressing will work correctly while running certain scripts.

Severity 2: Overlap Addressing will not work correctly while running certain scripts.

Show Screens

SONUS00131386: GSX shows incorrect status if the port that is requested does not exist. It shows the next existing port.

The code has been modified such that there are checks to see if the signaling port exists, before displaying the port.

Severity 2: GSX will display a port that does not exist.

Signaling

SONUS00133312: The parameter "mode" goes in the SDP "a=ftmp" attribute by copying it directly from MIME media type string as "mode=value".

The code has been modified to fix this issue.

Severity 3: ILBC calls can fail if “Use_COLON in SDP Media Type Parameter” is enabled in the IP signaling profile.

Signaling

SONUS00133572: Event Header in the NOTIFY message is being dropped by the SBC.

The code has been modified such that the optional header count is changed to check for atleast SIP_MAX_HEADERS in function SipMsgSetSingleInstanceHeaderCmd. This will make sure that sip messages can populate at least 64 optional headers.

Severity 2: The peer might not accept the NOTIFY without the event header.

 

Signaling

SONUS00129940: Switchover on PNS41 generated MAJOR alarms, and deleted one or more default IP routes. The user-configured default routes (admin routes) for the NIF on Standby PNS were missing after the Standby PNS came online.

The code has been modified such that a debug command has been added to directly delete the route in PNA.

Severity 2: Call may fail due to the invalid route in the route table.

LSWU

SONUS00130495: Two media profiles were configured, both at 50%. When there was a Live Software Upgrade to 8.3.0, it was found that the first profile was set to 50% and the other was blank.

The code has been modified such that the LSWU code has been added in 8.3.x to initialize the profile percentage variables correctly, as there are 4 media profiles supported in 8.3.x and 2 in the 7.x.

Severity 2: The profile percentage variables will not be initialized correctly.

Signaling

SONUS00132996: Calls failed with MTP3_SM_OBJECT_ERR m3_srm.c 759 since MNS failure. For F-links, there was a bug in the way current routes were synced to the standby. The routes get converted from internal to user route IDs. Usually these IDs match. This bug is exposed when they do not match.

Severity 1: Over time on a GSX that has a good amount of F-link reconfiguration, if an MNS sync occurs (for instance as a part of LSWU), when the next MNS switchover occurs, F-links may have a variety of failures that require a full GSX reboot to recover.

Mirror Groups

SONUS00132427: GSX failed to reboot after PNS was deleted from GSX.

When a PNS is deleted, references to its deleted interfaces in mirror groups would be left.

The code has been fixed to modify the delete routines to search mirror groups and remove references to interfaces that have been deleted.

Severity 3: After a GSX reboot, MNS will continuously core and prevent GSX from booting successfully.

Signaling

SONUS00129680: The ITU SS7 F-link may not recover after E1 alarms. Multiple E1 alarms in a short period of time expose a bug in the handling of MTP3 ITU T21 (overall MTP restart timer at a signaling point adjacent to one whose MTP restarts).

The code has been modified to fix this issue.

Severity 2: After E1 alarms, the SS7 route may not recover.

Signaling

SONUS00123561: For SIP-ISUP calls, GSX would normally distribute calls among available PNS/CNS resource based on resource availability. However, after SIP sig port is moved and enabled on other slot, and then moved back to the original slot, GSX no longer distributes calls and only uses one slot for all calls.

The code has been modified such that whenever the SIP Signaling Port is disabled, the mask value is resumed.

Severity 3: Calls are not distributed, when SIP signaling port is reconfigured.

Memory

SONUS00131456: Standby MNS card showed high memory utilization.

The code has been modified such that the memory utilization is within reasonable range (as per the number of registrations).

Severity 2: Manual intervention could be required to clear the high memory.

H323 Call Processing

SONUS00130068: In H.323 Release Complete to ISUP REL interworking, cause value 22 (H.323) is mapped to 102 (ISUP). This needs to be adjusted.

Modified H323SgRelCauseIeTake() so that it only maps H225_CC_INVNMBFRM to CPC_DISC_RECOVERY_ON_TIMER_EXPIRY if the H225_CC_INVNMBFRM was generated internally (as a result of overlap timer expiration).

Severity 2: Without this fix, the GSX will incorrectly map H225_CC_INVNMBFRM to CPC_DISC_RECOVERY_ON_TIMER_EXPIRY when the H225_CC_INVNMBFRM error comes from the ingress peer.

Signaling

SONUS00132457: The PSX Signaling Profile screen box “Change Bearer Cap from 3.1 KHz to Speech” worked for an ISUP to ISUP call but did not work for an IUP to ISUP call.

The code has been modified to work for all interworking cases.

Severity 2: Calls could fail.

Signaling

SONUS00133497: On receiving improperly formed RFC2833/RFC4733 dual-tone multi-frequency (DTMF) digit events from a peer, the SBC 9000 was generating duplicate DTMF digits. The received improperly formed events had a consistent 200 millisecond (ms) gap between the first few packets in the event and the succeeding (final) packets in the event. The SBC 9000 interpreted the 200 ms gap between events packets as the end of the event, and finished generating the associated DTMF digit. The SBC 9000 subsequently interpreted the RFC2833/4733 DTMF event packets following the 200 ms gap as a second DTMF event, and generated a second DTMF digit.

The SBC 9000 code has been modified to tolerate an inter-packet gap of up to 250 ms without interpreting the inter-packet gap as the end of the DTMF event. Thus, the 200 ms inter-packet gap in the events received from the peer no longer results in double DTMF digit generation by the SBC 9000.

Severity 1: There can be a duplicate DTMF.

Call Processing

SONUS00133703: When the user made an outbound call, the call was set up correctly with two way speeches, but when the user (calling party) placed the call on hold, the called party did not hear the music/tone on hold.

The issue was that the GSX processed the first media index from the SDP in the ACK, when multiple m-lines (2 audio and 2 video) were present. This resulted in incorrect data path mode sent to NRMA.

The code has been modified such that the data path mode send to NRMA is “recvonly”.

Severity 2: The called party could not hear the music on hold.

Coredump

SONUS00132215: HKIGSX02 node restarted.

The ping -R command was causing the MNS to coredump.

The “-R” option has been removed to resolve this issue.

Severity 2: MNS could coredump.

Signaling

SONUS00133924: GSX responded with 423 message when it received re-REGISTER message instead of passing this message to the other end.

The code has been modified to fix this issue.

Severity 2: GSX could respond with 423 message when it receives re-REGISTER message.

Signaling

SONUS00132961: For ISDN TBCT (Two B-channel Transfer), if the call was transferred to SIP (or any IP) peer, it results in one-way audio. While bridging the call, NRMA was not associating proper leg context to the resource chain.

The code is modified to associate the proper leg context.

Severity 1: If the call may transfer to SIP (or any IP) peer, it can results in one-way audio.

Signaling

SONUS00133347: The “Force 503 Relay” IP Signaling Profile flag is not working properly in Gateway to Gateway scenarios.

The code has been modified in the SBX to handle SIP 5xx responses in the same way handled in GSX 7.3.x. The “Force 503 Relay” IP Signaling Profile relay flag has been renamed to “Force 503 to 500 Relay”.

The default value for the “Force 503 to 500 Relay” flag is DISABLED (unselected).

Handling of 5xx responses depends on the IP Signaling Profile relay flag settings:

1. When the "Force 503 to 500 Relay" relay flag (in the egress IP Signaling Profile) is DISABLED, and the "Status Code 4XX-6XX" flag is:

  • ENABLED (in the egress IP Signaling Profile), the 503 response to INVITE is relayed to the ingress call leg
  • DISABLED, the Cause Code Mapping Profiles (SIP-to-CPC / CPC-to-SIP) are used to generate the response that is sent on the ingress call leg (for the 503 response to INVITE).

2. When the "Force 503 to 500 Relay" flag is ENABLED, a 500 Server Internal Error is sent on the ingress call leg (in response to the 503 response to INVITE on egress call leg), regardless of how the "Status Code 4XX-6XX" relay flag is set.

Severity 3: “Force 503 Relay” IP Signaling Profile flag may not work properly in Gateway to Gateway scenarios.

 

 

Signaling

SONUS00133735: When the ingress side of a SIP to SIP call sent a re-INVITE with G711U and silenceSupp was off, GSX sends T38 re-INVITE on the egress leg. Due to change of Codec to T38, the audio call failed.

The code has been modified not to send T38 RE-INVITE when G711 SS:Off RE-INVITE is received.

Severity 2: Audio Calls will fail due to T38 RE-INVITE towards egress leg.

Fault management

SONUS00122371: CNS card has multiple switchovers and active and standby CNS get into an oscillation mode. The problem is FM declaring a DSP failure wrongly. Additional counter was added to detect DSP failure correctly.

The code has been modified to read the online cards on the Fault Management (FM) and the Maker/IXP counters to report as an event to MNS FM. EVENT_MAKER_BAD_SDU_LEN is added to notify the error counters to the FM on MNS card.

Severity 3: CNS cards may switchover repeatedly and can cause to coredumps.

IP Routing

SONUS00133290: There are two signaling addresses on a PNS slot. Invite is sent from the Ingress to the Egress side. The Egress side fails to receive the invite.

The code has been modified to do lookup in local address hash along with interface hash.

Severity 3: Distributed SIPFE addresses cannot communicate with each other.

Signalin

g

SONUS00088751: When the GSX was connected to a PSX in mode TEST, the GSX was not using the PSX congestion level information from the PSX in mode ACTIVE and state ACTIVE as input into the machine congestion level calculations. This occurred because the GSX uses the lowest congestion level from all the state ACTIVE PSX’s as the input value for machine congestion and the PSX in mode TEST would never be in congestion so PSX congestion level 0 was always being used in the machine congestion level calculations.

The GSX code has been modified so that it no longer considers the PSX in mode TEST when calculating the lowest congestion level from all the PSX’s in state ACTIVE.

Severity 2: GSX

congestion control does

not work correctly when

one of the PSX is in

TEST mode.

CLI

 

SONUS00101104: The CLI command SHOW SIP SIGNALING REGISTRATION ADMIN must be deprecated

The code has been modified to deprecate REGISTRAR related CLI commands.

Severity 3: CLI interface will exist, but has no impact.

RTCP

SONUS00111677: The latency field in the ingress/egress RTCP statistics are populated as 0 and 65535 regardless of any latency introduced in the RTP stream.

The code has been modified to resolve the issue. The calculation for round trip time is done and the same is saved for RTCP session. The SHOW CALL DETAIL FULL returns latency using round trip time divided by two, similar to what CDR has.

Severity 3: The Latency fields can have either 0 or 65535 instead of CDR fields 67 and 151.

Signaling

SONUS00133831: Unable to enable NIF. The reason for rejecting the NIF enable command was that NIF's admnState = 5 (OOS_DRYUP). NRS COM only allows NIF/SIF to be enabled or re-enabled when it is admnState = 0 (NULL) or 7 (DISABLED).

The code has been modified to fix this issue.

Severity 3: Unable to enable NIF.

Signaling

 

SONUS00122870: GSX sets different Backward Call Indicator (BCI) values in ACM and ANM when receiving 180 message without SDP. When the originating GW gets the ANM, it allocates resources before forwarding the ANM to CC. While waiting, the cut through event comes in and the message information is lost. The default BCI is then used for the ANM.

The code has been modified to free the existing CPC message information, only if the event being processed includes CPC message information.

Severity 3: The BCI should be the same in the ACM and ANM.

 

Signaling

SONUS00134761: MNS11 crashes during a Live Software Upgrade from 7.2.9R0 to 8.3.2R0.

The code has been modified to fix this issue.

Severity 2: MNS11 crashes.

SBX

SONUS00127254: SBC looks pass through all option tags on supported header. Even after “100rel” is disabled at egress side (TG51_112),  “100rel” option tag on Supported header is visible. SBC must not pass through the option-tags in Supported header unconditionally.

The code has been modified to fix this issue.

Severity 3:

Signaling

SONUS00133500: GSX sent UPDATE message when received subsequent 18x prior to sending PRACK. Added extra logic to prevent sending UPDATE when received subsequent 18x with PRACK support.

Severity 3: Calls may not get completed.

Signaling

SONUS00134075: GNS tried to send an ACK for the received error response from the egress leg. The error response had remote Tag. Hence, the ACK msg pointing to the remote tag (garbage) caused the crash.

The code has been modified to correctly initiate remote tag properly.

Severity 1: May cause crash.

Signaling

SONUS00131344: When SBC 5200 received INVITE in TLS with Secure RTP SDP, Parse error occurred and it sent back 400 Bad message to the ingress peer. The issue was caused by crypto parser error. Reworked the crypto parser.

Severity 3: Parse error could occur on SBC 5200.

Call Processing

SONUS00133170: GSX failed to respond to late media re-INVITE relay. After the call transfer, the SG type was not updated properly.

The code has been modified to update the SG type correctly.

Severity 2: GSX could fail to respond to the re-INVITE.

SSH

SONUS00130209: GSX went into Machine Congestion level 2 (MC2) when SHOW TRUNK GROUP ALL ADMIN command was issued from EMS CLI. If the same command was issued from GSX CLI there was no congestion observed.

Defensive code has been implemented to prevent task that uses up the CPU time and causes machine CPU congestion that may cause call drop.

Severity 3: GSX can go into MC2 congestion.

LSWU

SONUS00130495: Two media profiles were configured, both at 50%. When there was a Live Software Upgrade to 8.3.0, it was found that the first profile was set to 50% and the other was blank.

The code has been modified such that the LSWU code has been added in 8.3.x to initialize the profile percentage variables correctly, as there are 4 media profiles supported in 8.3.x and 2 in the 7.x.

Severity 2: The profile percentage variables will not be initialized correctly.

Call Control

SONUS00134841: There was a problem that could cause the ‘From’ field in an egress SIP INVITE to be incorrect in a SIP->GW->GW->SIP call under certain circumstances. The problem affects calls, where the calling number sent back from the PSX is not the same as the original calling number received in the ingress SIP INVITE due to application of DM rules, globalization, and so on. The code has been modified such that the correct calling number is always sent in the egress SIP INVITE.

Severity 1: The ‘From’ field in an egress SIP INVITE will be incorrect in a SIP->GW->GW->SIP call under certain circumstances.

Coredump

SONUS00134229: There were high numbers of crashes on SPS cards.

The original fix (CQ 119393) used a hardware feature provided by 5561 DSP chips, where any invalid memory access would be caught and an interrupt would be raised. This was done to prevent the cases where such invalid memory accesses went un-noticed and caused some other problems (not necessarily crashes).

To prevent these crashes, the feature where an interrupt is raised with every invalid memory access is disabled. Essentially, the fixes made under CQ 119393 have been removed.

Severity 2: SPS cards could coredump.

LSWU

SONUS00134636: LSWU from V7.03.06R004 to R005 failed due to error in ISDNSG.

Fixes were made to how calls transition to active and how calls are synchronized to the standby card.

Severity 3: After a CNS switchover the standby card may fail to sync to an active card.

Coredump

SONUS00135827: PNS card coredumped because of accessing invalid memory.

The code has been modified to remove a semicolon at the end of an “if” statement that made an undesired check fell through.

Severity 3: PNS card could coredump.

Regression

SONUS00135885: The message “DRM: *DrmDspDidGet() DID already assigned, will be retained” is repeatedly received.

The code has been modified to fix this issue.

Severity 2: There will be RTCP DID related errors for TNA Pads.

Coredump

SONUS00135611: PNS cored twice with error 0x00000022.

The code has been modified to correct the memory double free issue.

Severity 3: PNS could coredump.

Regression

SONUS00132695: GSX allows UDP checksum to be configured for softswitches when the softswitches are ENABLED. If UDP Checksum is changed to ENABLED without setting the Softswitch state to DISABLED first, the CLI indicates that the feature is ENABLED but the feature will not work.

The code has been modified such that if any of the configured softswitches are not in DISABLED state when the administrator attempts to ENABLE UDP checksums, an error will be displayed to indicate the same.

Severity 3: If UDP Checksum is changed to ENABLED without setting the Softswitch state to DISABLED first, the CLI with indicate that the feature is ENABLED but the feature will not work.

 

Call Control

SONUS00132430: The Ringback tone is not heard when MCT is enabled. In this case, it is the SBC that is providing local RBT. When MCT is disabled, RBT is heard.

The code has been modified to fix this issue.

Severity 2: Ringback tone is not heard when MCT is enabled.

Signaling

SONUS00132806: CDR field 52 (for GSX accounting) was missing in H323 ingress call flow.

The code has been modified to resolve the problem.

Severity 2: There can be billing issues.

Memory management

SONUS00134623: GW-GW transparency memory leak.

This CQ resolves a memory leak resulting from a GW-GW transparency failure indicated in the DBG logs as “Unable to copy Transparency content” messages.

Severity 2: A memory leak will occur for every “Unable to copy Transparency content” GW-GW transparency error.

CDR

SONUS00133746: CDR STOP field 169 “Remote GSX Billing Indicator” may be incorrectly populated after a redundancy switchover if GW link is down when call terminates. This is only an issue when using the accounting feature” Populate RemoteGateway Info”.

The code has been modified such that the software preserves the value of “Remote GSX Billing indicator” field if it has already been set to TRUE. That is we do not need to clear this field once it has been set to TRUE even if we do not receive any more Billing information.

Severity 2: CDR STOP field 169 might not get populated correctly if a call terminates after a redundancy switchover.

 

CDR

SONUS00134342: CDR STOP fields 56, 138, 139 and 140 are not populated after a redundancy switchover. The data used to populate STOP fields: 56, 138, 139 and 140 were not mirrored to the redundant server.

The code has been modified to add redundancy logic to mirror these data fields.

Severity 3: CDR STOP fields 56, 138, 139 and 140 will not get populated in CDR if call terminates after a redundancy switchover.

Signaling

SONUS00135425: GSX responds to re-INVITE for changing media port with 100 but nothing follows. This is for REFER with Replaces call bridging only.

The code has been modified to fix this issue.

Severity 2: The OA-FSM Timer Expiry will be triggered and the call will tear down.

Coredump

SONUS00136539: All PNS cards coredumped on receiving INVITE message without out m-line.

The code has been modified to prevent the coredump.

Severity 1: PNS cards could coredump.

Signaling

SONUS00135297: SPS cards failed health checks when PXPADS and DSPs were enabled on it. This issue was tracked to a compiler bug.

The code has been modified to compensate for the compiler bug.

Severity 1: SPS cards can fail health checks when it is enabled with PXPADS and DSPs.

Coredump

SONUS00130868: Numerous marlin sys errors occurred followed by redundant MNS coredump with error 0x112d (Watchdog timer failure).

The issue was XRM task in infinite loop and some logical error.

The code has been modified to add defensive check to avoid infinite loop and correction has been made in the illogical code.

Severity 3: Redundant MNS could coredump.

DSP Coredump

SONUS00136524: After the upgrade to 8.3.2R0, CNS81 coredumps due to activation and deactivation failures.

The core dumps were due to DSP health check time outs.

The code has been modified to fix this issue.

Severity 2: Core dumps due to activation or deactivation failures will continue to happen.

Call Processing - SIP

SONUS00129249: In a GW-GW call, GSX ignored the unknown attributes and the FMTP line. Since the unknown attributes were ignored, the video call was not established.

The code has been modified such that the GW-GW video call is established by passing transparently the Unknown attributes, FMTP and Bandwidth under media line VIDEO.

Note: This is applicable for P2P Video Call also.

Severity 2: GW-GW video call cannot be established.

Call Control

SONUS00134388: CPU Utilization on PNS card was running unbalanced on slots. The issue was that there was no SIF data on CNS/SPS servers due the hardware limitation.

The code has been enhanced to populate SIF’s slot information to CNS/SPS servers so that it can pick proper PNS slot for the call.

Severity 2: Calls could fail.

Call Processing

SONUS00136635: After MNS switchover, IP to PSTN calls failed when GSX assigned them to in-busy trunkgroup. It was found that if you have a PSTN to IP attempt that gets cranked back, both the PSTN and IP legs get deleted from the standby TRM (Trunk Group Resource Manager). If the next route is IP, only the IP leg gets mirrored. If you then have an MNS switchover, TRM and the ISUPSG are out of sync and TRM will select the PSTN leg that it considers available, but is actually in use.

Note: If the next route is PSTN, the problem clears itself because both PSTN legs get mirrored.

This TRM redundancy bug has been fixed.

Severity 3: TRM selects channels that are not actually available and this leads to call failures.

Signaling

SONUS00129718: If the GSX receives a SIP invite with 32 or more called digits, it truncates the digits to 30, but sipsgCpcUtils.c compares the request URI with the “To header” when only 1 number has been truncated. This leads to the GSX 'creating' a redirecting number even though the original request URI and TO header were identical.

The code has been modified to fix this issue.

Severity 3: This will cause an unnecessary redirecting number.

CLI

SONUS00128590: GSX4000 only has its Oscillator as a valid timing source but when a "SHOW ST STATUS" command is sent, it shows N/A rather than the Oscillator as its Active Source.

The code has been modified such that the MTA card need not be verified since object is not supported on GSX4000 series switches.

Severity 2: GSX4000 does not show valid timing source.

 

SIP Signaling

SONUS00136017: When a call is made from a SIP terminal connected to the untrust side of SBC 9000 for more than 6 minutes, if the originator releases the call, SBC 9000 sends back 481 response for CANCEL message, but it does not send CANCEL to the trust side. Thus, the call on trust side cannot be released. Also, if the terminator answers the call, it becomes in talking state. However, SBC 9000 sends back 500 response for re-invite (refresh) coming from untrust side. Thus, the call is released once the session timer expires.

The code has been modified such that CANCEL is sent to Egress and 408 is sent to Ingress when Guard timer expires.

Severity 2: Ingress rings longer than 6 minutes and then hangs up. The egress will get hung if Answer Supervision timer is disabled.

Call Processing

SONUS00136940: When the recorder call switched over from standby to active, GSX accidentally inserted the recorder call into the GCID hash table. After the recorder call tore down, GSX was not able to remove its entry from the GCID hash table. The memory was later used for other purposes and caused the GCID hash table corruption. Note: This issue is introduced by NICE feature only.

The code has been modified to not insert the recorder call into the GCID hash table when the call switches over from standby to active. The recorder call is not intended to be inserted into the GCID hash table.

Severity 2: Coredump could occur.

Signaling

SONUS00135575: REFER based call transfer fails. The policy query contains previous General Digit.

The code has been modified to fix this issue.

Severity 2: Wrong data is sent to PSX. Call may fail due to routing to the wrong destination.

Coredump

SONUS00136580: PNS40 coredumps with 0x51 when memory allocation fails.

The code has been modified to fix this issue.

Severity 1: Coredump may occur on memory allocation failure.

SBX 2.2

SONUS00127254: SBC looks pass through all option tags on Supported header even though it is disabled on “100rel” at the egress side (TG51_112).

The code has been modified in 2.2 to fix the transparently passing of Unknown option tags in supported header. So, only the supported tags get transparently passed. No flags were added to control the behavior:

  • No Supported option-tags in the supported header will be transparently passed by default. There are specific flags to transparently pass the supported option-tags like History info, Target dialog, preconditions,100 rel, timer. Only these will be transparently passed when the flags are enabled.
  • Unknown extensions will not be passed through for Invite message and responses.

Severity 3: The option-tags in Supported header will pass unconditionally.

ISUP Signaling

 

SONUS00133697: For double Hop Service Node scenarios, there is no Early-ACM with SCP Carrier Info in the CIT.

The code has been modified to send Early ACM, which also includes SCP Carrier Info in CIT. Also, the second INVITE w/IAM to SN will contain CIT parameter with SCP Carrier Info.

Severity 2: No Early-ACM towards Ingress. Second INVITE w/IAM will not include SCP carrier Info in CIT parameter.

 

Statistics

 

SONUS00125849: During data collection of the FTP polling feature on the EMS, it is found that the EMS FTP login at the NFS goes to passive mode, finds the files, enters binary mode, transfers and then deletes the files from the NFS.

The code has been modified such that maximum number of configured trunks are handled.

Severity 3: The trunk group statistics are logged for a maximum of 29 trunk groups. Even if there are more, only 29 trunk groups are configured in the system.

 

Call Flow

 

SONUS00131448: Ring back tone is heard at a different DSP channel.

The code has been modified to prevent ECAN memory corruption by including the data memory holding ECAN data pointers in the run time checksum mechanism.

Severity 3: There will be a voice leak at other DSP channel.

 

Enhancement

 

SONUS00132508: PCR 4949 (ARS Crankback requirements) is an enhancement on ARS so that it optionally can use a new cause code 169 rather than 151(INVITE request retransmission Timeout). This would enable crankback for blacklisted peers even if cause code 151 is not in the crankback profile.

The code has been modified such that new CPC disc code 169 is added to handle ARS separately.

Severity 2: It is not possible to configure specific cause code to enable crankback for ARS black listed entries.

 

ISUP Service

 

SONUS00134479: The “Call Cleanup Calls Per Segment” of ISUP Service Profile has a value out of the range sets on GSX.

The code has been modified to set the correct default values.

Severity2: It will display values that are out of the range sets of GSX.

 

Regression- IP

SONUS00133483: Dual stack NIF (IPV4 and IPV6) has multiple SIF's on it. Upon node restart, an IPV6-only SIF is coming up in OOS state. This prevents a SIP signaling port that is created on this SIF to become active and in service.

The code has been modified to integrate the ND with ARP infrastructure (similar to IPv4), which sends the Neighbor solicitation messages at an interval of 2 seconds regularly until the MAC address is resolved.

Severity 1: SIFs will remain in OOS state. This prevents a SIP signaling port that is created on the SIF to become active and in service.

Signaling

SONUS00133698: SBC sends Re-Invite without getting ACK for 200OK-INVITE in LM Call.

The code has been modified such that the Re-Invite is not generated before getting the Acknowledgment.

Severity 2: Call may not get connected.

Overload

SONUS00129586: The cable for port 2 was pulled and reseated. Calls began to fail MC2 reported. CPU utilization was 100%.

Severity 2:

NIFs/SIFs

 

SONUS00130195: The NIFs and SIFs on the GSX became unresponsive to all external input in one of the instances. The logs show that IXP erroring occurs due to nopktbuf increasing at a rate of about 8-14 packets every 5-6 seconds.

The code has been modified to delete the nilan entry if PNA fails to add NIF/SIF.

Severity 2: All the packets sent to those NIFs/SIFs would be dropped.

Circuit Signaling

SONUS00135328: The SHOW SS7 NODE STATUS command does not show SGX4K during migration from SGX2K to SGX 4K.

The code has been modified for the SHOW SS7 NODE <> STATUS command to display the SGX4K information (i.e. configured as Alternate Gateway during migration).

Severity 2: The SHOW SS7 NODE STATUS command does not show SGX4K during migration.

Overlap Scenarios

SONUS00135761: PCR 5442 seems to be supported for ISUP and SIP ingress SG only. Support for H323 or ISDN or CAS is missing.

The code has been modified to update the new values of minDigit and maxDigit to Egress SG as well so that the comparison will happen with updated values during timer expiry.     CcStoreOverlapSendingData function is modified to manipulate both mindDigits and maxDigits for valid phone and the calledDigitsOffset points to the difference of dialedDigits to the actual called Number digits length.

Severity 2: Irrespective of the protocol, scenarios where egress leg is enbloc and Total  calledNum digits is less than minDigits configured in PSX, they will get cleared without this fix.

CLI

SONUS00130096: There is a delay in printing the sh nifgr all ad or sh nifgr SIF-1-5-2-2079-OUTSIDE ad command or any other NIF group name and the prompt appears after a delay.

Also, the output of show ipv6 routes command is being displayed only after couple of seconds.

Note: The following commands must not be executed during the Live Software Upgrade:

SHOW IP ROUTE ALL and sub commands.

SHOW IPV6 ROUTE ALL and sub commands.

The code has been modified to traverse and get only relevant records from the COM and display them, instead of fetching all records and displaying when required.

Severity 2: Delay in printing the SHOW output.

NIFs/SIFs

SONUS00136480: Overlap LIFs must be part of disjoint Nifgroups. Overlap LIF is only for NIF or SIF-type LIF.

The code has been modified to fix this issue.

Severity 2: Overlap LIF checking will happen for Mgmt type NIF also, whereas Overlap LIF checking is supposed to be for PNS NIF or SIF only.

SIP Signaling

SONUS00135356: Case 1: A re-Invite is triggered towards Ingress once the resources for LRBT is deallocated.The LRBT is not really played as there is no support for 100rel.

Case 2:A re-Invite is triggered towards theIngress peer on the receipt of re-invite from Egress, though the OA is not complete in case of Late Media Call

For Case 1, the code has been modified such that the actual issue reported due to LRBT is solved in CQ 133698.

For case 2, the code has been modified such that the re-Invite is triggered towards the peer when the OA is still incomplete. As part of this CQ, the new offer from the Egress is rejected and care has been taken to overcome the video only and direct media call by checking the flag noAudio to TRUE.

Severity 2: A reInvite will be seen towards ingress. The call may get released.

H323 Signaling

SONUS00136613: Announcement is not played to the H323 ingress leg for the No Routes Found case.

 

The code has been modified such that mediaSignalingInfo CPC in H323SgSetCpcMediaSignalingInfo() sets the Late Media flags only when the call is really Slow start or when the fast start fails/refuses, which makes the h323capNegReqd flag TRUE.

Severity 2: No tones/announcement is played to H323 ingress leg when PSX returns “No Routes Found” error with script that plays tone and announcement.

Media

SONUS00137024: SBC is configured to send RFC 2833 digits on egress when a user input event occurs in the ingress leg. When H323 to SIP call is in progress, sending User input event from H245 end, SBC sends RFC2833 RTP event packets. First RTP event packet is sent with correct digit info but subsequent RTP Event packet with E-bit set will always be sent as a DTMF Zero event.

The code has been modified to fix this issue.

Severity 2: SBC sends wrong DTMF event ID, while marking the end of the event.

 

Signaling

SONUS00137224: SIP-GW-GW-SIP and the transparency AUTHCODE ON.  ReInvite fax call fails with internal error at the ingress GSX.

The code has been modified to fix this issue.

Severity 3: Calls may fail.

Call Processing

SONUS00130251: Emergency calls were not allowed when Registration mode Required-NonPriority.

The code has been modified to allow emergency calls without registration.

Severity 2: Calls could get rejected.

Signaling

SONUS00132195: SIPFE may find the wrong registration handle when both A and B are registered, A calls B, and loopback to GSX.

The code has been modified such that a defensive check is added to ensure that the call is coming from the same registration signaling port.

Severity 2: SIPFE may find the wrong registration handle.

Signaling

SONUS00132682: After switchover, when keep alive receive/response triggers the other leg to remove Video CODECS.

The code has been modified to fix this issue. Also, GSX mirroring the peer SDP checksum to avoid unnecessary re-Invite to the other leg.

Severity 2: Video CODECS can be missing in Re-Invite sent to the other leg after PNS switchover.

Signaling

SONUS00136529: ISDN debug logging was broken.

An ISDN event was added, but the ISDN event string array was not updated as well. Thus the string array did not match the actual events.

Modified the event string array to be consistent with the event enum.

Severity 3: ISDN logging is misleading.

Call Processing

SONUS00136237: REFER/REPLACE landed on the wrong slot when CPU got congested. NRM selected a different slot causing the call to tear down.

The code has been modified such that NRM is updated to support the preferred slot only.

Severity 3: Calls could tear down.

 

 

 

Known Problems

The following problems are known to exist in GSX V09.00.xx:

Issue ID

Severity

Category

Problem Description

Workaround

SONUS00164394

3

Coredump

RTCP may fail when xNS cards reach their defined socket limit. SYS ERR in log will occur.

None

SONUS00153503

2

Configuration

Invalid prefix length configuration for IPv6 on MGMT NIFs results in a crash.

None

Known Limitations

The section describes limitations that are known to exist for the GSX4000 and GSX9000 in GSX release V09.00.xx.

Limitations for GSX9000 and GSX4000

The section describes platform, signaling, and call processing limitations for both the GSX9000 and the GSX4000.

Platform Limitations

The section describes the platform limitations for the GSX9000 and the GSX4000.

MNS and Management

Due to security changes to protect the GSX, the GSX will no longer respond when attempting a traceroute to the GSX.
Management NIFs no longer have 'active/standby' mode. Both MGMT NIFs are 'active' from operational point of view. Until release 6.0, route tables are modified such a way that, routes configured on "active" MGMT NIF have higher precedence over the standby MGMT NIF. Starting with release 6.0, both MGMT NIF ports on a given slot have equal preference, and routes are added to the routing table in the order they are configured/loaded. It is likely that an equivalent route on second management port might end up on top of the routing table because of the order in which ports are brought up.
The twenty-four hour cumulative (total) T1, T3, and E1 statistics cannot be reset.

SGX SS7 Gateway

Zero calls and loss of Primary OMNI CE may take 20 seconds to connect CE2.

Trunk Group Configuration Considerations

IP Trunk Groups – H.323 and SIP calls using G.711 with silence suppression and RTCP should disable the SID heartbeat to avoid premature Packet Outage Detection (POD) detection and the associated RTCP events. POD is an IP Trunk Groups feature that identifies significant network outages and can be used to reduce the capacity of an IP Trunk Group automatically. Silence Suppression has a silence insertion descriptor (SID) component and a silence insertion descriptor heartbeat; it is the latter that can trigger the false packet outage state. RTCP provisioning allows the GSX to detect when silence occurs and possibly disconnect a call, generate an SNMP trap and report statistics. If the SID heartbeat is not disabled in this scenario the RTCP-generated events will not be accurate. GW-GW calls are not affected by this limitation (SONUS00041846.)
Trunk Groups and Performance Statistics – Performance statistics are saved in a number of "intervals" (default = 4, maximum = 12) and performance statistics are run per trunk group per interval. Increasing the number of trunk groups, and to a greater extent increasing the number of intervals, increases the memory requirements for the interval table. So changing the number of intervals can have a dramatic effect on how much memory is used. Therefore we recommend that customers do not increase the number of intervals when using a large number of trunk groups.
Older CPL Scripts May Cause System Errors – Older CPL scripts that were created to work with a different GSX script architecture may not work properly and can even cause system errors if used with newer GSX versions (SONUS00047034.)
Service Profile Editor loop coding – Although SPE 6.x (e.g., SPE 06.03.05 used for GSX 7.2.3) does not allow any pointers to point back to the Begin SIBB, ensure that the script being built with the Sonus SPE tool does not have an infinite loop. An unintended consequence of using a loop-programming construct could be a forever loop, and such a script when executing on the GSX card could cause high CPU utilization.
Disabling Default IP Trunk Groups – You cannot directly disable the defaultIPTG or its service groups; trying to do so through the CLI will generate errors such as "The defaultiptg cannot be set out of service" or "Default IP service groups can never be modified or deleted" (SONUS00053613). However, you can configure the defaultIPTG trunk to not accept calls by setting the defaultIPTG call limit and bandwidth limit = 0 through the CLI or EMS. Via the CLI, use:

CONFIGURE TRUNK GROUP defaultiptg IP CALL LIMIT 0
CONFIGURE TRUNK GROUP defaultiptg IP BANDWIDTH LIMIT 0 

On the EMS IP Trunk and defaultIPTG Trunk screens, use CONFIGURE SELECTED IP TRUNK GROUP - NO DISRUPTION to change the IP call and bandwidth limit without taking the trunk out of service.

Signaling and Call Processing Limitations

Different DTMF relay with conditional transcoding

Different DTMF Relay works best if PSPs and endpoints are partitioned. For example, Ingress is all 2833 (only) and Egress is all OOB (only). PSPs should be set in the same way and endpoints should be enabled in the same manner.

G729r8 Support in H.323 Signaling

The GSX can now process an incoming H.323 call request with G.729 payload in the offer (PCR1528, SONUS00074701). On the egress side, when configured appropriately with G.729A in the PSP-Ri and PSP-Re, the GSX sends both G.729 and G.729A in the offer to the peer. Upon receiving the answer from the terminating endpoint, the GSX sends G.729/G.729A in the answer to the originating side depending on what it received from the egress endpoint.
Previously the H.323 GW supported only G.729A, not G.729. Although G.729 was supported on the DSP it was not supported in the H.323 implementation, so no common codec could be negotiated for an incoming offer with G.729 payload and as a result calls.

Lawful Intercept Solution

Lawful Intercept (LI) functionality has the following current limitations for the LI solution:

  • Billing Number intercept criterion is not supported.
  • SIP URI intercept criterion is not supported.
  • VPN intercept criterion is not supported.
  • IP Loopback only works on GSX PNS40 cards (GSX 9000 only).
  • Inband DTMF digits are reported by Sonus Equipment to the LI Server.  However, final reporting to the Law Enforcement Agency (LEA) will depend on the LI Server vendor.

Contact Sonus regarding implementation of Lawful Intercept, including details of interfacing with an Intercept Server and any current functional limitations.

SIP

If the Mandatory Header fields have an illegal value and cannot be parsed, the GSX will not respond but rather will ignore the SIP packet (SONUS00031454).
For SIP calls, the GSX rejects a modified SDP offer in a PRACK.
RURI transparency is not consistent for relayed SUBSCRIBE/NOTIFY messages.

SIP Replaces

SIP Replaces is supported as an independent feature, it is not supported in conjunction with SIP REFER on the same call (i.e. you cannot perform a SIP Replaces operation on a call which has already performed SIP REFER, and vice versa).
Multiple SIP Replaces operations on the same call are not supported; you can only perform one SIP Replaces operation on any given call.
SIP Replaces is only supported in scenarios where the new Replaces call results in a pass-through call. This type of call is supported: A call from A  B is passed-through or transcoded and the SIP Replaces call from A  C is passed-through. This type of call is not supported: A call from A  B is passed-through or transcoded and the SIP Replaces call from A  C is transcoded.

Interoperability Constraints with the SIP Application Server

When using packet cards other than PNS40s, SIP-based applications that instruct the GSX to bridge the media streams of two independent calls may result in the following types of issues:

  • One way media flow
  • No media flow
  • Call failures

Call flow validation is required when the GSX is interworking with SIP Application Servers that provide call transfer scenarios (including those with SIP Refer), prepaid calling card applications, call forwarding applications, click-to-dial applications, Class 5 applications, or any application involving the bridging of two or more calls through the GSX.
These types of applications are typically implemented by an application server swapping the SDP of two independent calls. This action, by the application, results in the GSX sending media from one call back to itself for the other call. When the calls land on the same NIF, the GSX must recognize this situation and internally bridge the media between the two calls. This is referred to as an IP-loopback call. When the calls land on different NIFs, the media flows out of the GSX on one NIF through the IP network to the other NIF and back into the GSX; this scenario does not utilize internal bridging.
Why Bridge? Media streams for IP loopback calls must be bridged internally by the GSX and not send packets on to the IP network; packets of looped streams if sent out on the IP network are dropped by the routers/switches due to the fact that they have the same source and destination addresses.
IP loopback is not restricted to the scenarios where the two calls are PSTN to IP calls; loopback can also occur when A and B are IP endpoints or a mix of PSTN and IP endpoints.

ISDN to BT IUP Interworking

The table below indicates the interworking functionality that is provided between these protocols.

BT IUP

ISDN

Interworking Status

CLIP/CLIR

CLIP/CLIR

Full Interworking1

Overlap

Overlap

Full Interworking

Fallback

Fallback

Partial interworking2 3

NAE

SUB(Called)

Full Interworking

DDI

DDI

Full Interworking (PSX Functionality)

CCBS

CCBS

Cannot be mapped across the IWF

CUG

CUG

Cannot be mapped across the IWF 4

Call Diversion

Call Diversion

Cannot be mapped across the IWF4

Suspend/Resume

Suspend/Resume

Cannot be mapped across the IWF4

RBWF

n/a

Cannot be mapped across the IWF4

SWAP

n/a

Cannot be mapped across the IWF3

ACI

n/a

Cannot be mapped across the IWF

SIC

LLC

Partial Interworking

n/a

HLC

Cannot be mapped across the IWF

n/a

HOLD

Cannot be mapped across the IWF

n/a

CW

Cannot be mapped across the IWF

n/a

3PTY

Cannot be mapped across the IWF

n/a

CONF

Cannot be mapped across the IWF

n/a

ECT

Cannot be mapped across the IWF

n/a

COLP/COLR

Cannot be mapped across the IWF

n/a

UUS

Cannot be mapped across the IWF

1: The GSX ISDN implementation only Supports Single Number Delivery, interworking Calling and Presentation Number to ISDN as PNO 8, Issue 3.1, Table 8.3.
2: BT IUP only supports fallback from 3.1 kHz to Speech.
3: No messages requesting a Bearer Change can be interworked to BT IUP.
4: The GSX Euro-ISDN implementation does not support this functionality.

G.711

Starting with GSX 6.0, silence suppression descriptor (SID) packets are compliant with G711 Annex II. However, its use is not generally available with all signaling protocols and is only supported between GSXs with GW-GW signaling. G711 Annex II silence suppression with comfort noise generation (CNG) is not supported to SIP or H.323 endpoints.

Limitations for GSX9000 Only

The section describes platform, signaling, and call processing limitations for the GSX9000 only.

Platform Limitations

This section describes the platform limitations for the GSX9000.

PNS and Packet Transport

This release only supports a 3:1 PNS40 redundancy configuration.

PNS card switchover during high call load in GSX

In a configuration where all the PNS cards in GSX chassis are fully loaded with calls and incoming call load is significant, a PNS card switchover can create call distribution imbalance in the system. This call distribution imbalance can cause other PNS cards to go into cycles of congestion. This condition will abate once the call load in the system decreases. The problem was detected in a 3+1 PNS card configuration but is not specific to that configuration. It is possible in 1+1 PNS card configuration at sufficiently high call load during switchover.

MNS and Management

A maximum of four simultaneous telnet and/or SSH sessions is supported.
An NTP server is required in both Sonus SoftSwitch and third party softswitch models.

CNS and TDM Transport

Certain combinations of circuit redundancy groups must not be configured in a GSX. The restrictions in the table below are enforced. The CNS servers are divided into classes for the purpose of defining the restrictions within a class as well as between classes of servers. The table below shows the restrictions and the valid combinations:

Redundancy Group Class

CNS60

CNS71

CNS40 /45

CNS80/ 85

CNS86/ 81

CNS60

Multiple

Yes

Multiple

Multiple

Yes

CNS71

Yes

Multiple

Yes

Yes

Multiple

CNS40/45

Multiple

Yes

Multiple

Multiple

Yes

CNS80/85

Multiple

Yes

Multiple

Multiple

Yes

CNS86/81

Yes

Multiple

Yes

Yes

Multiple

Where:

  • One: A single redundancy group composed of one of these server types is permitted.
  • No: Redundancy groups composed of servers from each of these classes are not permitted.
  • Yes: Redundancy groups composed of servers from each class are permitted.
  • Multiple: Multiple redundancy groups composed of servers from this class are permitted.

All other redundant CNS modules must be installed in slots 14, 15, or 16. Only one T3 redundancy group is allowed to switchover at one time because it is dependent on the same hardware resources as the other T3 redundancy group.

CNS Card Announcement Memory

If a CNS card's announcement memory is full and all the announcements are currently in use (none have been unused long enough to be eligible for overwriting), an attempt to download another announcement will generate a Major error "...Segment Memory is Full..." The announcement will successfully load once an existing announcement is no longer in use.

Trunk Group Configuration Considerations

Release GSX 6.3 and later allow you to configure a maximum of 2048 trunk groups on MNS20 card. The maximum number of service groups allowed is 2047. However, in general, the more trunk groups and service groups that you configure, the more likely that call rate performance will be reduced, as follows:

  • Configuring large numbers of trunk groups with similar names can cause a severe degradation of calls per second performance. Trunk group names received from the PSX are located using a named lookup. This search is most efficient when the trunk groups have dissimilar names and least efficient when they take similar names such as TrunkGroup01, TrunkGroup02, etc.
  • Configuring larger performance statistics interval sizes and saving fewer intervals can reduce the load on the management port. Sonus Insight communicates with SNMP over the management port to maintain trunk group performance statistics. Other management port functions, such as policy requests and responses, could be delayed or lost by these performance statistics causing calls to fail.
  • A larger number of trunk groups and service groups will require more MNS module processing resources to complete periodic saves of the parameter files. This could result in fewer cycles to process calls when parameters are saved due to configuration changes.
  • The time that is taken to successfully perform a CNS or MNS switchover will be lengthened as the number of trunk groups increases.
  • PSX processing is also subjected to increased load when large numbers of trunk groups are configured. When large numbers of trunk groups are configured on a single routing label, the route label should not have the "All" flag enabled.
  • Any trap that is generated per trunk group or per service group can potentially cause a "trap storm" if there is an event that causes a trap to be generated for a large number trunk groups or service groups. This may occur during a CNS or MNS switchover, when communication with the SGX is lost, or when a high capacity port (e.g. OC-3) goes down. Trap storms can bog down the management port on the GSX and EMS, and the GSX may drop some traps when its queues get full.
EMS Support for Expanded Numbers of Trunk Groups

Due to EMS processing limitations, it is recommended that the number of provisioned trunk groups on the GSX not exceed several hundred trunk groups. Any number beyond that may detrimentally affect the EMS's ability to collect trunk group data and to manage the GSX via the GSX Navigator.
It is also recommended that as the number of trunk groups provisioned on the GSX is increased that the performance data collection frequency be decreased.
Finally it is recommended that the number of intervals configured on the GSX for the trunk group interval table be reduced as the number of provisioned trunk groups increases.

T.38 Fax Transmission

When the T.38 fax transmission ceases, the audio session should be terminated. GSX does not allow the audio session to be re-established after the successful fax call ends.

Packet-to-Packet Call Services

In Packet-to-Packet call services, the GSX sends zero duration for RTP Event packet.

Direct Media

It is a Direct Media call, but the Ingress Codec Info and Egress Codec Info of the CDR fields has Codec type negotiated in the signaling.

Redundancy for PNS40/41 Cards

Because of the increased memory footprint of the software, Release V09.00.02R000 supports 3:1 redundancy instead of 4:1 redundancy for PNS 40/41 cards.

If CICs are added on the GSX when all the M3UA links between GSX and SGX are down, then the registration information could go out of sync between GSX and SGX when the links come back up.
This can be rectified by making the SGP links to the standby MNS card outOfService and back inService at the SGX.

Performance Limitations

This section describes the performance limitations for the GSX9000.

Standby PNS41 crash

Standby PNS41 crashes with an out of memory condition in a 4+1 redundancy configuration in an Access SBC role. PNS41 crashes with 240,000 subscribers spread across 4 active cards backed up by 1 standby card.
Under load at 160 cps, PNS41 crashes for SIP-SIP call, where signaling and media is on IPv6 for both legs.
The PNS41 cards must be reconfigured into a 2+1 redundancy configuration.

AMR Limitations

The GSX9000 does not support interleaving, CRC, and robust sorting in transcoded call. But these are supported for pass through calls. Supported interleaving is 4-19 frame-blocks.
Sonus AMR, WB AMR, and EFR implementation has VAD turned ON and there is no configurable option currently to disable the VAD.
Currently, silence suppression check box in PSX will be used to configure VAD for the three codecs. When silence suppression box is checked, it implies that the VAD is ON. 
AMR-NB supports VAD1 and VAD2, the VAD type should be made configurable. For Backward compatibility default VAD2 must be configured.
However, AME-WB codec AMR-WB and EFR have only one VAD type.

INVITE with Replaces

When GSX receives an INVITE with Replaces from an End Point which has received a REFER, 
GSX does not relay the INVITE to the intended Party and there is no Replaces header in the message.

Traps

The following SNMP Traps are not supported by the GSX9000:

MIB

SNMP Trap Not Supported by GSX9000

rfc1573.mib

linkDown

rfc1573.mib

linkUp

rfc1907.mib

coldStart

rfc1907.mib

warmStart

rfc1907.mib

authenticationFailure

rfc2320.mib

ipoaMtuExceeded

rfc2320.mib

ipoaDuplicateIpAddress

rfc2320.mib

ipoaLisCreate

rfc2320.mib

ipoaLisDelete

rfc2495.mib

dsx1LineStatusChange

rfc2496.mib

dsx3LineStatusChange

rfc3498.smib

apsEventSwitchover

rfc3498.smib

apsEventModeMismatch

rfc3498.smib

apsEventChannelMismatch

rfc3498.smib

apsEventPSBF

rfc3498.smib

apsEventFEPLF

sonusAps.mib

sonusApsTrapProtectionSwitchingByteFailure

sonusAps.mib

sonusApsTrapChannelMismatch

sonusAps.mib

sonusApsTrapModeMismatch

sonusAps.mib

sonusApsTrapFarEndProtectionLineFailure

sonusAps.mib

sonusApsTrapSwitchover

sonusIpdc.mib

sonusIpdcSessionConnectNotification

sonusIpdc.mib

sonusIpdcSessionDisconnectNotification

sonusIpdc.mib

sonusIpdcSessionChangeNotification

sonusIpdc.mib

sonusIpdcSessionEndNotification

Limitations for GSX4000 Only

The section describes platform, signaling, and call processing limitations for the GSX4000 only.

Platform Limitations

This section describes platform limitations for the GSX4000.

Management

In a scenario where there was one MTRM server/client on one GSX4000, one MTRM client on another GSX4000 (with GW calls on an IP selected trunk group between them), prior to taking MTRM server outofservice, the calls allocated were balanced at 500 and 500 for each GSX4000 client. (The MTRM Server bandwidth limit was UNLIM while the global call limit was 1000.) Upon taking it outofservice and once again inservice, calls allocated did not achieve this balanced state again; thus, resources were not symmetrically allocated.
In order to work around this issue, use a more granular number for the following configuration:

CONFIGURE TRUNK GROUP <trunk group name> CALLS PER REQUEST <number>


There is a performance tradeoff with this workaround. If you set this number too low with respect to the overall expected rate of arrival of calls, then there will be an increased number of messages. If too high, then there is a possibility of the anomaly reported here. A good number would be MIN{20, peak rate/3}.

DSP Re-Loading

The GSX4000 does not support re-loading of DSPs. This is due to the existence of three distinct DSP XPAD images (UXPAD, GSXPAD, and EXPAD). There are two workarounds to this limitation:

  • Configure the DS1 type (CONFIGURE DS1 ALL TYPE circuittype) after you configure the MEDIA PROFILE.
  • Dynamically change the MEDIA PROFILE while the system is up and running, turn on parameter loading, and re-boot the GSX4000.
Local Disk and Event Logs

This platform does not support event logging to the local disk.

Packet Transport

Support for signaling and media flows between communicating packet peers is supported with restrictions. Any combination of SIP, SIP-T, and Gateway may be used on both ingress and egress call legs. Media negotiation occurs between the peers on a call-by-call basis, resulting either in the mapping (transcoding) or in the pass through (RTP relay) of media. Negotiations may occur during the call, as well as at call setup.
RTCP is not supported for packet-to-packet calls that are setup with RTP relay. For these calls, RTP flows through the GSX without DSP processing.

Traps

The following SNMP Traps are not supported by the GSX4000:

MIB

SNMP Trap Not Supported by GSX

rfc1573.mib

linkDown

rfc1573.mib

linkup

rfc1907.mib

coldStart

rfc1907.mib

warmStart

rfc1907.mib

authenticationFailure

rfc2495.mib

dsx1LineStatusChange

Emergency Link activation is possible without the CIC configured (SONUS00046398.) In the current design, SS7 Links stay in OUTOFSERVICE state as long as ISUP CICs are not configured, and are brought into INSERVICE state when the CICs are configured. But whenever the CLI command to perform emergency Activation of a linkset is used, the MTP2 Links go into INSERVICE state. Therefore, availability of MTP2 Links after using above CLI command doesn't indicate that the ISUP CICs are configured on the chassis. You can bring the link into service with the CLI command even though CICs are not configured. When bringing up MTP2 links, you must also configure the CICs.
MTP3 can report MTP2 Link status as L2OOS, when the actual status is INSERVICE (SONUS00052079.) Whenever an MTP3 link is deactivated and mode Blocked, followed by re-activation by 'EMERGENCY or NORMAL' option, the MTP3 link status indicate the MTP2 link state as UNAVL|LBLK|L2OOS instead of only UNAVL|LBLK. This indicates that MTP2 link is OUTOFSERVICE, instead of INSERVICE. Later, when the MTP3 Link is mode unblocked, the MTP3 link state changes to AVAILABLE as expected.

During CNS switchover, MTP2 may glitch the line. However, calls are not impacted.

Signaling and Call Processing Limitations

This section describes the signaling and call processing limitations for the GSX4000.

SS7

The GSX4000 can only support one SS7 MTP3 MGMT node.

GSX Behavior Change

The GSX CLI may occasionally appear to halt when a large number of commands are executed via the CLI. This behavior is not indicative of a problem with the system.