Docstoc

Cascade 500 CSRN 1.00.05.00

Document Sample
Cascade 500 CSRN 1.00.05.00 Powered By Docstoc
					    Software Release Notice

B-STDX 8000/9000 Switch Software
      Release 04.02.05.00


        November 21, 1997

Ascend Communications Corporation
          Alameda, CA




                       Part Number 104-00410-05 Rev. 00
Release Abstract


TABLE OF CONTENTS
1.      RELEASE ABSTRACT ......................................................................................................................................5
     1.1     SUPPORTED SOFTWARE ..................................................................................................................................5
     1.2     SUPPORTED HARDWARE .................................................................................................................................5
        1.2.1    Identifying a CP Classic ........................................................................................................................6
        1.2.2    Identifying IOP Types ............................................................................................................................7
     1.3     PRODUCT DOCUMENTATION ...........................................................................................................................8
2.      MINIMUM REQUIREMENTS .........................................................................................................................8
     2.1        HARDWARE ....................................................................................................................................................8
     2.2        SOFTWARE......................................................................................................................................................8
3.      NEW FEATURES AND ENHANCEMENTS ...................................................................................................8
     3.1        NEW FEATURES AND ENHANCEMENTS IN RELEASE 04.02.03.00 ....................................................................8
     3.2        NEW FEATURES AND ENHANCEMENTS IN RELEASE 04.02.02.00 ....................................................................9
4.      PROBLEMS CORRECTED ............................................................................................................................ 13
     4.1     CORRECTED IN VERSION 04.02.05.00 ........................................................................................................... 13
     4.2     CORRECTED IN VERSION 04.02.04.00 ........................................................................................................... 16
        4.2.1   General ................................................................................................................................................ 16
        4.2.2   PPP ...................................................................................................................................................... 17
     4.3     CORRECTED IN VERSION 04.02.03.00 ........................................................................................................... 17
     4.4     CORRECTED IN VERSION 04.02.02.00 ........................................................................................................... 19
        4.4.1   General ................................................................................................................................................ 19
        4.4.2   Loopback ............................................................................................................................................. 20
        4.4.3   Logical Ports ....................................................................................................................................... 20
        4.4.4   Frame Relay......................................................................................................................................... 20
        4.4.5   ATM Cards........................................................................................................................................... 21
        4.4.6   HSSI Card ............................................................................................................................................ 21
        4.4.7   DSX-10 Card ....................................................................................................................................... 21
        4.4.8   CS DS3 Card........................................................................................................................................ 21
        4.4.9   E1 Card................................................................................................................................................ 21
        4.4.10  CP 30/40/50 Cards .............................................................................................................................. 22
        4.4.11  Bulk Statistics....................................................................................................................................... 22
     4.5     CORRECTED IN RELEASE 04.02.00.00 ........................................................................................................... 22
        4.5.1   General ................................................................................................................................................ 22
        4.5.2   CP 30/40/50 ......................................................................................................................................... 25
        4.5.3   IOP....................................................................................................................................................... 25
        4.5.4   Channelized DS3 ................................................................................................................................. 25
        4.5.5   HSSI ..................................................................................................................................................... 25
        4.5.6   ATM ..................................................................................................................................................... 26
        4.5.7   SMDS ................................................................................................................................................... 26
        4.5.8   Logical Ports ....................................................................................................................................... 28
        4.5.9   PVCs .................................................................................................................................................... 28
        4.5.10  UIO ...................................................................................................................................................... 29
        4.5.11  ATM DS3 ............................................................................................................................................. 29
        4.5.12  DS1 and DS0 Loopback ....................................................................................................................... 30
        4.5.13  BERT Testing ....................................................................................................................................... 30
        4.5.14  Bulk Statistics....................................................................................................................................... 30
        4.5.15  ISDN .................................................................................................................................................... 30
5.      RESTRICTIONS AND SPECIAL CONSIDERATIONS .............................................................................. 32
     5.1        NETWORKS WITH 800 OR MORE OSPF LSAS ............................................................................................... 32
     5.2        LOGIN TO CONSOLE ...................................................................................................................................... 32
     5.3        SCRIPT DOWNLOAD ...................................................................................................................................... 32


Ascend Communications Inc.                                                                                                                                        Page 2
Release Abstract


     5.4     HARDWARE .................................................................................................................................................. 32
     5.5     IOM ............................................................................................................................................................. 32
     5.6     RIP ............................................................................................................................................................... 32
     5.7     STDX INTEROPERABILITY............................................................................................................................ 32
     5.8     SOFTWARE DOWNLOAD ............................................................................................................................... 33
     5.9     REAL-TIME CLOCK ....................................................................................................................................... 33
     5.10 CONNECTING TO B-STDX THROUGH A MODEM CONNECTION .................................................................... 33
     5.11 SWITCH MANAGEMENT ................................................................................................................................ 33
     5.12 IOP TYPE A .................................................................................................................................................. 34
     5.13 LOGICAL PORT ............................................................................................................................................. 34
     5.14 T1/E1 ........................................................................................................................................................... 34
     5.15 HSSI ............................................................................................................................................................ 34
     5.16 ATM CS/IWU.............................................................................................................................................. 35
     5.17 POWER CONSUMPTION ................................................................................................................................. 35
     5.18 QUALITY OF SERVICE (QOS) ........................................................................................................................ 35
     5.19 SMDS .......................................................................................................................................................... 35
     5.20 MANAGEMENT TRUNK ................................................................................................................................. 36
     5.21 ATM ............................................................................................................................................................ 36
     5.22 ISDN ............................................................................................................................................................ 36
     5.23 NEW FRAME RELAY RATE ENFORCEMENT ................................................................................................... 36
     5.24 FRAME RELAY SVC MAXIMUMS.................................................................................................................. 36
        5.24.1    Cards that Support Frame Relay SVCs................................................................................................ 37
        5.24.2    Card Limits .......................................................................................................................................... 37
        5.24.3    Node and Port Prefix and Address Limits ........................................................................................... 37
     5.25 CHANGING THE DEFAULT PASSWORD .......................................................................................................... 38
     5.26 IPX TRAFFIC ................................................................................................................................................ 38
     5.27 UNDOCUMENTED NEW FEATURES ................................................................................................................ 38
        5.27.1    B-STDX 9000 ATM CS and IWU I/O Module Direct Cell Trunk Shaper Priority Assignment .......... 38
        5.27.2    Shaper Guidelines ................................................................................................................................ 39
        5.27.3    Configuring the Shaper Settings .......................................................................................................... 41
        5.27.4    G.826 Statistics .................................................................................................................................... 42
        5.27.5    Displaying G.826 Statistics .................................................................................................................. 42
6.      USAGE BASED BILLING ............................................................................................................................... 44
     6.1        BRINGING UP A NEW SWITCH....................................................................................................................... 44
     6.2        INSTALLING A NEW CP CARD........................................................................................................................ 44
     6.3        SETTING THE CLOCK .................................................................................................................................... 44
7.      KNOWN ANOMALIES .................................................................................................................................... 45
     7.1     ANOMALIES IN 04.02.05.00 .......................................................................................................................... 45
     7.2     ANOMALIES IN 04.02.03.00 .......................................................................................................................... 45
        7.2.1   General ................................................................................................................................................ 45
     7.3     ANOMALIES IN 04.02.02.00 .......................................................................................................................... 46
        7.3.1   General ................................................................................................................................................ 46
        7.3.2   Loopbacks ............................................................................................................................................ 46
        7.3.3   Channelized DS3 ................................................................................................................................. 46
        7.3.4   T1/E1 ................................................................................................................................................... 47
     7.4     ANOMALIES IN 04.02.00.00 .......................................................................................................................... 47
        7.4.1   General ................................................................................................................................................ 47
        7.4.2   CP ........................................................................................................................................................ 47
        7.4.3   T1/E1 ................................................................................................................................................... 47
        7.4.4   DSX-10................................................................................................................................................. 48
        7.4.5   HSSI ..................................................................................................................................................... 48
        7.4.6   ATM Features ...................................................................................................................................... 48
        7.4.7   ATM DS3/E3 Cards ............................................................................................................................. 48
        7.4.8   Channelized DS3 Cards ....................................................................................................................... 48


Ascend Communications Inc.                                                                                                                                           Page 3
Release Abstract


        7.4.9          SMDS ................................................................................................................................................... 49
        7.4.10         Console ................................................................................................................................................ 49
        7.4.11         PVCs .................................................................................................................................................... 49
        7.4.12         DS0 Loopback ...................................................................................................................................... 50
        7.4.13         ISDN .................................................................................................................................................... 50
        7.4.14         OSPF ................................................................................................................................................... 50
8.      SWITCH SOFTWARE UPGRADE INSTRUCTIONS ................................................................................. 51
     8.1     PREREQUISITES ............................................................................................................................................. 51
     8.2     UPGRADE PROCEDURES................................................................................................................................ 51
        8.2.1    Installing the Software into a Switch with a Single CP ....................................................................... 52
        8.2.2    Installing the Software into a Switch with Redundant CPs .................................................................. 53
        8.2.3    Installing the software into a Switch with a Single CP Using Multifile Download ............................. 54
        8.2.4    Installing the software into a Switch with Redundant CPs Using Multifile Download ....................... 55




Ascend Communications Inc.                                                                                                                                           Page 4
Release Abstract




1. Release Abstract
   This document serves as the Software Release Notice (SRN) for this release of B-STDX software. The
   following sections describe the switch software revision for this release and the hardware that it supports. It
   also provides a list of current Ascend documentation you may need to reference for complete information.

1.1 Supported Software
   Release 4.1 introduced a new version numbering scheme. In addition to the four-part version ID (ex. 4.1.3.0)
   there is an additional “Software ID” portion, that consists of a letter code followed by a spin number (ex. R301,
   where R represents released software). To view the complete software tag, issue the show software
   command from a switch console to display the level of software currently running on each card. You can also
   issue the show software flash command to see the software that resides in the CP FLASH area.
   This release is bundled as CP version 04.02.05.00, IOPA version 04.02.05.00, IOPB version 04.02.05.00, and
   IOPC version 04.02.05.00. The following list describes specific switch software and build IDs contained on the
   software installation disk:
      cp1.rom version 04.02.05.00 (Software ID: EC-R000009F9) supports the Control Processor cards. This file
       is the CP application code (version 04.02.05.00). This file consists of two parts, the application code
       portion (04.02.05.00) and a CP1 boot flash code portion (version 04.02.04.00).
      ioptypea.rom version 04.02.05.00 (Software ID: EC-R00000836) supports the UIO/T1/E1/DSX-1 IOP
       cards. This file consists of two parts internally: an IOPA application code portion (version 04.02.05.00) and
       an IOPA boot flash code portion (version 04.02.00.03).
      ioptypeb.rom version 04.02.05.00 (Software ID: EC-R0000058B) supports the HSSI/ATM DS3/ATM E3
       IOP cards. This file consists of two parts internally: an IOPB application code portion (version 04.02.05.00)
       and an IOPB boot flash code portion (version 04.02.06.00).
      ioptypec.rom version 04.02.05.00 (Software ID: EC-R0000004BC) supports the ATM UNI CS IOP and
       ATM IWU OC3/STM-1 cards. This file consists of two parts internally: an IOP3 application code portion
       (version 04.02.05.00) and an IOPC boot flash code portion (04.02.02.01).

1.2 Supported Hardware
        This release supports the following B-STDX hardware:
            Control Processor Card (CP Basic, CP+, CP-30, CP-40, CP-50), and CP Adapter (CPA)

       NOTE: This release does not support CP Classic cards or non-IOP+ cards. To identify CP Classic cards
             and non-IOP+ cards, refer to the next two sections, Identifying a CP Classic and Identifying
             IOPs.
       NOTE: Networks with 800 or more OSPF LSAs require a CP+, CP40, or CP50 control processor card.

            8-port Universal Input/Output (UIO) card with V.35 or X.21 adapter (single and redundant)
            10-port DSX-1 card with RJ-45 adapter (single and redundant)
            4-port Channelized DSX card
            4-port Channelized T1 card with DB-15/RJ-45 adapter (single and redundant)
            4-port Channelized E1 card with DB-15/Coaxial adapter (single and redundant)
            4-port 24-Channel PRI T1 card for ISDN (single and redundant)
            4-port 24-Channel PRI E1 card for ISDN (single and redundant)
            4-port Unchannelized T1 card with DB-15/RJ-45 adapter (single and redundant)



Ascend Communications Inc.                                                                                   Page 5
Release Abstract


            4-port Unchannelized E1 card with DB-15/RJ-45 adapter (single and redundant)
            2-port HSSI card with adapter (single and redundant)
            1-port ATM DS3/E3 (single and redundant)
            ATM UNI CS (cell switching) card (single and redundant)
            ATM IWU OC3/STM-1 (cell switching) card (single and redundant)
            1-port Channelized DS3 card
            Redundant power supplies and fan units

       1.2.1 Identifying a CP Classic
       You can identify a CP Classic by any one of the following three methods:

           Console - Issue the following command through the console:

                   > show card

                       where

                       CP = CP-CLASSIC

                       CP Basic = CP-BASIC

                       CP Plus = CP-PLUS

           NMS – From the Monitor menu select MIB-Values  Browse MIB.
             Select the following MIB object:
             private.enterprises.cascade.cascfr.card.cardTable.cardEntry.cardHwRev
             The system displays the hardware revision of all cards for a given node. The values are:
             1.00 = CP-CLASSIC
             3.00 = CP-PLUS
             10.00 = CP-BASIC

           Visual - Visually identify the CP cards as follows:

                          CP Type               DIP Switch Color on Front Panel      Board Part Number
                         CP Classic                         Black                      710-00001-00
                          CP Basic                          Black                      710-00069-00
                          CP Plus                            Red                       710-00034-01




Ascend Communications Inc.                                                                               Page 6
Release Abstract


       1.2.2 Identifying IOP Types
       The following cards, including all high-speed cards, have only IOP Plus and IOP 16 versions. Therefore,
       they do not need to be verified.

          4 Port Unchannelized T1
          4 Port ISDN T1
          4 Port Channelized E1
          4 Port Unchannelized E1
          4 Port ISDN E1
          4 Port Channelized DSX-1
          2 Port HSSI
          1 Port ATM UNI DS3 Interworking
          1 Port ATM UNI E3 Interworking
          1 Port Channelized DS3

       The following low-speed cards have IOP, IOP Plus, and IOP 16 versions:
        8 Port UIO
        4 Port Channelized T1
        10 Port DSX-1

       You can identify IOPs by any one of the following three methods:

          Console - Issue the following command through the console:

                   > show card
            Find the h/w revision field. The revisions listed below are the minimum revisions you can use for this
            release (i.e. these were the first IOP Plus). If you have earlier revisions, you must upgrade the IOPs.
            UIO = 04.00.00
            DSX-1 = 02.00.00
            4-port Ch. T1 = 06.00.00

          NMS – Select the switch (double click) that contains the IOP you want to identify. Select the IOP you
           want to identify.
            Find the h/w revision field. The revisions listed below are the minimum revisions you can use for this
            release (i.e. these were the first IOP Plus). If you have earlier revisions, you must upgrade the IOPs.
            UIO = 04.00.00
            DSX-1 = 02.00.00
            4-port Ch. T1 = 06.00.00

          Visual - Visually identify the IOPs as follows:

                        IOP Type                DIP Switch Color on Front Panel
                          IOP                               Black
                        IOP Plus                             Red




Ascend Communications Inc.                                                                                Page 7
Minimum Requirements


1.3 Product Documentation
   This document covers last minute changes and restrictions that your user documentation does not include. For
   complete information about this product, refer to the following documents:
         Software Release Notice for CascadeView/UX Version: 02.05.02.00 or 02.04.03.10
   To successfully install and administer the B-STDX products, refer to one of the following manuals:
         Network Management Station Installation Guide (Product Code 80014)
         Network Configuration Guide for B-STDX/STDX (Product Code 80017)
         Diagnostic and Troubleshooting Guide for B-STDX/STDX (Product Code 80018)
         SYBASE 11 Upgrade Guide (Product Code 80040)
         CascadeView/UX Documentation Set Addendum (Part Number 104-00102-00)


2. Minimum Requirements
2.1 Hardware
           This release supports Version 04.02.05.00 switch software.

2.2 Software
           You must install the latest UNIX NMS Release software (minimum 02.04.03.00 or greater) before you
           install this B-STDX switch software. Refer to one of the following Software Release Notices (depending
           on the NMS version that you are using), for more information about this NMS release.
              Software Release Notice for CascadeView/UX Version: 02.04.03.10

          NOTE: In order to use the latest fixes and new features in this release, Ascend recommends using UNIX
                NMS Version 02.05.02.00 or 02.04.03.10.




3. New Features and Enhancements
   This section lists the new features and enhancements in previous releases. Release 04.02.05.00 does not have
   any new features or enhancements.

3.1 New Features and Enhancements in Release 04.02.03.00

       Feature                 Description
       SNMP Access to          SNMP access is supported for 5-minute peak values for trunks and Frame Relay
       Trunk and PVC           virtual circuits (VC). The MIB objects to access the 5-minute peak values for
       5-Minute Peak           trunks are defined under the logical port table (as lportEntry). The MIB
       Statistics              objects to access the 5-minute peak values for PVCs are defined under the circuit
                               table (as cktEntry).
       Idle/Unassigned cell    You can now send idle cells in either of the following formats:
       type for ATM cards       ATM Forum style idle cells have CLP=0 and Payload=0x6A in the idle cells
       on the 9000                  transmitted.
                                ITU style idle cells have CLP=1 and Payload=0x6A.




Ascend Communications Inc.                                                                                  Page 8
New Features and Enhancements




    Feature                 Description
    4K circuit support      New versions of the ATM DS3 CS (product code 59094A-16) and ATM
                            OC3/STM-1 (product code 59087A-16) cards have been released with additional
                            memory to support 4K circuits over trunks. Version 4.2.3.x is required to take
                            advantage of the additional memory.
    G.826 Performance       Display performance-monitoring statistics for digital transmission paths on the 4
    Monitoring Statistics   port E1 and channelized E1 modules.
    Shaper Priority         Used to prioritize circuit traffic traversing a direct cell trunk on ATM CS and
                            IWU modules.
    Management              Management VPI/VCI is now supported on the ATM CS/IWU modules.
    VPI/VCI

3.2 New Features and Enhancements in Release 04.02.02.00

    Feature                 Description
    General Features
    Class B IP              Increase networks to include up to 400 switches and 1000 trunks.
    Addressing
    Console                 Authenticate a user‟s login authority by verifying the user‟s password.
    Authentication
    RIP Support             Run the IP Routing Information Protocol (RIP) over the B-STDX Ethernet port.
                            This enables you to use redundant gateway switches. You configure RIP on the
                            gateway switch only. This allows the switch software to send and receive RIP
                            packets.
    Tunable Keep Alive      Configure a Keep Alive (KA) timer on the trunk. The timer enables you to set the
    Timer                   number of retries that the trunk protocol tries before bringing the trunk down.
                            The retry interval is represented in seconds.
    Configurable            Configure discard thresholds that determine under what network condition traffic
    Congestion              will be discarded. You configure the mild, severe, and absolute congestion
    Threshold               thresholds.
    Closed Loop             Prevent excess data from entering the network during congested states.
    Congestion
    Management
    Trunk Backup            Configure a module to signal to an external ISDN terminal adapter (T/A) that a
                            circuit needs to be established. You can designate a backup trunk for any trunk in
                            the network. If a primary trunk fails, all PVCs associated with that trunk are
                            rerouted to the backup trunk. You can reroute the trunks either automatically or
                            manually.
    Application Specific    Dedicate trunks to specific end-users and, if desired, allow customers to monitor
    Routes/Customer         their own networks. However, control and configuration of the switches stays
    Specific Routes         with you as the network provider. This feature is not supported for SMDS
    (Virtual Private        services.
    Networks)
    Single PVC              In a combined B-STDX and CBX 500 network, you can now establish PVCs
    Establishment           across paths with numerous switch hops, simply by specifying the endpoints.
                            This enables you to provision circuits more efficiently and results in an overall
                            lower cost of operation.




Ascend Communications Inc.                                                                                  Page 9
New Features and Enhancements




    Feature              Description
    Distributed OSPF     The Open Shortest Path First (OSPF) function is no longer centralized at the CP.
                         It has been distributed so that each IOP can now process OSPF functions.
                         Distributed OSPF enables the IOPs to make routing decisions. With this new
                         architecture, each IOP has full knowledge of the network topology and is
                         therefore able to make independent routing decisions.
                         Distributed OSPF reduces the load on the CP in intermediate nodes because the
                         IOPs can forward packets without using the CP. This does not increase network
                         background traffic, as the CP continues to aggregate and propagate the OSPF
                         Link State Advertisements (LSAs).
    Priority Rerouting   Priority Rerouting allows a PVC configured with a higher priority to select a
                         more efficient path (one that satisfies its QoS) than those PVCs configured with a
                         lower priority. As it selects this path, the higher priority PVC can choose to
                         redirect lower priority traffic to a less efficient path. This guarantees that the
                         network will always give preference to higher priority PVCs, even during periods
                         of congestion or link failure. This preference provides a higher percentage of up-
                         time, an increase capacity to burst past the guaranteed QoS, and shorter delays.
    PRAM Kermit          Enables a customer to kermit a PRAM image directly to the switch via the switch
                         console. This feature can simplify the initial switch commissioning.
    Multilink PPP        Multilink PPP allows two or more PPP links to be combined into one logical link
                         to provide greater effective bandwidth on the multilink PPP link than on any of
                         the individual links. There can be up to 32 links in each multilink PPP bundle.
                         Multilink PPP is a method for splitting, recombining, and sequencing datagrams
                         across multiple logical data links. See the Cascade Documentation Set
                         Addendum for instructions on configuring a multilink PPP connection.

                         All IOP Type A cards support the multilink PPP protocol (MLP-1490) and
                         include the following modules: E1-PRI (ISDN), T1-PRI (ISDN), Unchannelized
                         T1/E1, Channelized T1/E1, 4-port DSX-1, 10-port DSX-1, Universal IO (V.35,
                         X.21).
    ATM Features
    ATM Redundancy       This release supports redundancy on the ATM IWU OC3/STM-1 cards.
    Single OSPF Domain   This release provides Cascade OSPF routing between B-STDX and CBX 500
                         switches. A single OSPF routing domain can support a network that contains
                         both B-STDX and CBX 500 switches.
    OSPF Delay Metric    The OSPF delay metric allows for PVC circuits (ATM circuits only) to be routed
                         based on an end-to-end delay routing metric.
    ATM Direct Cell      A Direct Cell Trunk enables a single OSPF routing domain in a network that
    Trunk                includes both B-STDX and CBX 500 switches. You can establish VCs between
                         any B-STDX and CBX 500 switch, using either B-STDX frame- or cell-based
                         trunks or CBX 500 cell trunks. Routing decisions allow frame-based traffic
                         coming from the B-STDX to transit either frame or cell-based trunks. Cell-based
                         traffic coming from the CBX 500 is restricted to routes, which transit Direct Cell
                         Trunks.
    Multiservice PVCs    Multiservice PVCs refer to a virtual circuit in a mixed network of both B-STDX
                         and CBX 500 switches that can transit either frame or cell-based direct trunks in
                         any combination. The full integration of the B-STDX into the CBX 500 network
                         using a common trunking protocol enables this function. Multiservice PVCs
                         allow end-to-end connectivity between any pair of access interfaces which are
                         either compatible or support interworking.
    CLP Bit              Sets the cell loss priority on FR-ATM Service Interworking circuits.



Ascend Communications Inc.                                                                            Page 10
New Features and Enhancements



    Feature               Description
    SMDS Features
    SMDS Selective        Broadcasts group address frames to specific addresses rather than a general
    Group Address         broadcast.
    Country Code Prefix   Extends SMDS networks across international borders.
    SMDS Address Mask     Enables migration from another vendor‟s SMDS network to a Cascade SMDS
                          network.
    Alien Address         Increases network scaleability by increasing the Alien Address capacity from
                          1024 to 2048 per switch.
    Frame Relay
    Features
    Frame Relay SVCs      This feature is available to users of the CascadeView/UX NMS Version
                          02.04.00.00 or later. If you are using Version 02.04.00.00 or later of the NMS,
                          you can now set and monitor switch-wide SVC configuration parameters on each
                          of your B-STDX 8000/9000 switches. See the Cascade Documentation Set
                          Addendum (part number 104-00102-00) for more information about Frame Relay
                          SVCs.
    Zero CIR Frame        To provide a best effort data service, you can now configure PVCs with CIR
    Relay PVC             values of zero. End users subscribing to zero CIR services can burst to the port
                          speed if there is network bandwidth available to deliver frames, however the
                          network does not guarantee to deliver these frames. The discard eligible bit of all
                          frames entering the network on a zero CIR PVC is set to one (eligible for
                          discard).
    Frame Relay           Asymmetric CIR allows you to specify different QoS parameters in each
    Asymmetric CIR        direction of a Frame Relay PVC.
    ISDN Features
    E1 ISDN I/O Module    This release adds support for the European E1 ISDN interface that supports 30 B
                          channels and one D channel.
    European ISDN         The ISDN access card supports European ISDN signaling standards (CTR4) as
    Signaling             well as switch-specific extensions as required.
    Australian ISDN       The ISDN access card supports Australian signaling standards (TS014) as well as
    Signaling             switch-specific extensions as required.
    PAP                   Password Authentication Protocol (PAP) provides a simple authentication
                          method using a two-way handshake. This is done only upon initial link
                          establishment. With PAP, passwords are sent over the circuit without encryption
                          and there is no protection from repeated trial and error attacks.
    CHAP                  Challenge-Handshake Authentication Protocol (CHAP) is used periodically to
                          verify the identity of a client using a three-way handshake. This is done only
                          upon initial link establishment. With PAP, passwords are sent over the circuit
                          without encryption and there is no protection from repeated trial and error
                          attacks.
    RADIUS                Remote Authentication Dial-In User Service (RADIUS) provides authentication,
                          authorization, and configuration information between a Network Access Server
                          and a shared Authentication Server. RADIUS is used to ensure security when
                          establishing a Point-to-Point (PPP) connection to a Cascade B-STDX switch over
                          an ISDN remote access link. RADIUS supports PPP, PAP, and CHAP.




Ascend Communications Inc.                                                                              Page 11
New Features and Enhancements




    Feature             Description
    Hardware Features
    ATM IWU             The ATM IWU OC3/STM-1 (cell switching) module is a one port OC3c/STM1
    OC3/STM-1           card that supports the same frame-to-cell capabilities of our current ATM
                        DS3/E3 module including ATM Network Interworking and ATM Service
                        Interworking.
                        In addition, the ATM IWU OC3/STM-1 module performs pure cell switching.
                        Segmentation and Reassembly are performed using AAL-5. You can provision a
                        flexible output-buffering scheme for CBR, VBR-RT, VBR-NRT, and UBR
                        classes of service. These functions enable the B-STDX to perform cell switching
                        in addition to Frame Relay/ATM interworking functions.
                        The ATM IWU OC3/STM-1 module supports ATM UNI DCE, UNI DTE, ATM
                        OPTimum Trunk, ATM OPTimum Cell Trunk, and Cascade Direct Cell Trunk
                        logical port types.
    IOP 16 Meg Module   The NMS provides support for the IOP 16 Meg module. This new module also
                        provides the architecture to support future new services. This release only
                        recognizes the IOP 16 Meg module as an 8 Meg module.
    CP-30, CP-40, and   This release supports three new B-STDX Control Processor modules. The new
    CP-50 Modules       Control Processor modules provide the architecture to support future services -
                        such as IP Navigator and new Billing functions, as well as value-added features,
                        which will enhance upgradeability and scalability. You can install these new
                        modules in any switch or network that uses the current CP Modules (CP Basic
                        and CP Plus only).
    ATM-UNI CS          The ATM CS (cell switching) Module is a one port DS3 card that supports the
    Module (DS3)        same frame-to-cell capabilities of our current ATM DS3 module including ATM
                        Network Interworking and ATM Service interworking.
                        In addition, the ATM CS module performs pure cell switching. Segmentation
                        and Reassembly are performed using AAL-5. You can provision a flexible
                        output-buffering scheme for CBR, VBR-RT, VBR-NRT, and UBR classes of
                        service. These functions enable the B-STDX to perform cell switching in
                        addition to Frame Relay/ATM interworking functions.
                        The ATM CS module supports ATM UNI DCE, UNI DTE, ATM OPTimum
                        Trunk, ATM OPTimum Cell Trunk, and Cascade Direct Cell Trunk logical port
                        types.




Ascend Communications Inc.                                                                         Page 12
Problems Corrected


4. Problems Corrected
This section lists the problems corrected in this and previous releases.

4.1 Corrected in Version 04.02.05.00

            Issue                                               Resolution
            Circuit DLCI                                        The LMI Full Status message now reports all circuit
                                                                DLCIs.
            Exchange state condition                            The problem that caused cards/trunks to enter into a
                                                                continuous exchange state due to a resource shortage
                                                                (memory or bandwidth) has been corrected.
            PRAM Upload/Compare                                 Compare discrepancies no longer occur following a
                                                                PRAM Sync for the following features: CLLM,
                                                                Pport data rate on CH DS3, Tx and Rx clock on the
                                                                IWU OC-3.
            Telnet login reset                                  Following a warmboot, the telnet login is no longer
                                                                reset to 2.
            Telnet login trap                                   The switch now sends a trap to the NMS to notify
                                                                when a telnet login occurs on the switch.
            STM-1 ATM IWU does not raise pport traps            The STM-1 card now raises pport traps for “pport
            (C97-13284)                                         down with red alarm” and “pport down with yellow
                                                                alarm”.
            Frame Buffer Exhaustion                             A problem was corrected whereby control traffic
                                                                could drive the buffer count below the user
                                                                threshold, exhausting frame buffers and potentially
                                                                causing the reuse of a previously allocated buffer.
            Multicast Over Cell Trunks                          Multicast circuits no longer pass traffic to circuit
                                                                endpoints which are not members of the Multicast
                                                                circuit following a cold boot of a switch. This
                                                                problem was previously observed when the circuits
                                                                were established over cell trunks.
            Upgrading the B-STDX switch causes the              The problem whereby upgrading a B-STDX switch
            redundant CP to restart (C97-15909)                 from Release 4.2.2.1 to 4.2.2.3 caused the redundant
                                                                CP to restart has been corrected.
            Abnormal cell discarding at 9000 PVC (end           The problem has been corrected and now DS3 ports
            to end DS3 ports) (C97-16091)                       correctly transmit and receive cells.
            After a coldboot, a trap states the PVC is          The problem whereby PVCs were restored but the
            down (C97-16650)                                    system displayed a trap stating they were inactive
                                                                has been corrected.
            PPP negotiation traps flood the event               PPP negotiation traps are no longer continuously
            browser (C97-16652)                                 generated.
            ATM DS3 receive counter resets to zero              The receive counter no longer resets to zero.
            (C97-16726)




Ascend Communications Inc.                                                                                 Page 13
Problems Corrected



        Issue                                        Resolution
        No alarms on PPP circuit when up-stream      The problem whereby an Lport/Pport failure on
        equipment goes down (C97-16731)              C500 did not generate an alarm on PPP circuits in a
                                                     B-STDX has been corrected.
        PPP circuits can’t come back after switch    PPP circuits now come back after rebooting a B-
        reboot (OAM disabled) (C97-16845)            STDX switch with ATM ports (OAM disabled).
        Pram sync causes circuits to go down on a    Performing a PRAM Sync no longer causes circuits
        Ch DS3 (C97-17237)                           to go down on a channelized DS3 card.
        Translate OAM alarms to PPP alarms (C97-     The system now generates a trap stating the circuit is
        17585)                                       down due to OAM.
        Crash address not cleared (C97-13576)        Crash addresses associated with Fatal Error
                                                     messages are now cleared when subsequent Fatal
                                                     events are encountered that contain a crash address
                                                     of zero.
        After coldbooting B-STDX with Ch DS3         Ch DS3 circuits now correctly come up after you
        cards, circuits do not come up (C97-17257)   coldboot the switch.
        IfType and IfNum descriptions are            The system now displays the correct interface
        inaccurate (C97-10709)                       descriptions.
        No alarm when OAM F4 is Received (C97-       The alarm is now displayed in the event window
        16847)                                       when a Cisco 7500 and C500 is down and the C500
                                                     sends an OAM F4 alarm to the B-STDX.
        Issue console prompt (C97-16796)             On a switch that is booted with no PRAM on the
                                                     CP, the console prompt is now displayed as “>>”
                                                     (instead of “>”).
        T1/E1 Transmit Driver                        A problem with the transmit driver of the 10-port
                                                     DSX-1, 4-port T1, and 4-port E1 IOP cards has been
                                                     corrected. This problem previously caused transmit
                                                     buffer overruns, which in turn resulted in a garbled
                                                     frame. This problem occurred very infrequently and
                                                     only with very large frame sizes.
        QoS                                          The problem whereby the QoS circuit parameters for
                                                     lost frames and octets intermittently would report
                                                     extremely large and unrealistic numbers following
                                                     circuit reroutes has been corrected.
        Bulk Statistics                              Bulk statistics now calculates the correct QoS lost
                                                     frames and octets. A peak value is now maintained
                                                     for QoS statistics.
        Multicast Traffic over ATM Cards             Multicast traffic over ATM UNI CS or ATM IWU
                                                     OC3/STM-1 cards now works properly.
        Circuits Down (C97-17588)                    The source endpoint of the circuit is no longer
                                                     dropping the “call confirm” message from the
                                                     destination endpoint of the circuit.



Ascend Communications Inc.                                                                      Page 14
Problems Corrected



        Issue                                        Resolution
        IPX Traffic over ATM Cards (C97-16748)       The problem whereby IPX traffic received on an
                                                     Ascend PPP Lport and routed over an ATM UNI CS
                                                     or ATM IWU OC3/STM-1 card acting as a trunk
                                                     card, caused the trunk card to crash, has been
                                                     corrected. IPX traffic now correctly passes over the
                                                     trunk card, if the trunk is defined as a direct cell or
                                                     OPTimum cell trunk. If the trunk is defined as an
                                                     OPTimum trunk, the IPX data is discarded, but the
                                                     trunk card will not crash.
        PPP                                          The PPP link now recovers when a PPP link on a
                                                     channelized DS3 card connected to a router is
                                                     physically disconnected and then reconnected.
                                                     (C97-15910, C97-16430, and C97-16552)
                                                     The problem whereby a 9000 with PPP Lports were
                                                     rebooted causing the PPP links not to come back
                                                     after the switch had recovered has now been
                                                     corrected. PPP links have now been changed to
                                                     recover in this case. (Cases C97-16484 and
                                                     C97-16651)
                                                     When a new PPP connection is added to the switch,
                                                     the Lport now correctly reports its operational
                                                     status.
                                                     If a PPP link on a HSSI IOP goes down and up, the
                                                     card no longer needs to be rebooted to recover.
        Missing Transmit Cells from PVC Statistics   If a virtual circuit is set up between two ATM UNI
        (C97-16091)                                  CS cards in the same switch, the lower-numbered
                                                     slot now correctly reports cells transmitted on the
                                                     PVC statistics screen.
        PVC Operational Status                       The code has been corrected to resolve a condition
                                                     where the PVC operational status would be incorrect
                                                     after modifying the PVC’s parameters or
                                                     administrative status. (Case C97-16387)
                                                     Intermittently PVCs would fail to come up after
                                                     being added to a channelized DS3 card. This
                                                     problem has been resolved. (C97-16414)
                                                     If a channelized DS3 or HSSI card were latched
                                                     down and up again, the PVCs terminating on that
                                                     card would not recover. This problem has been
                                                     corrected.




Ascend Communications Inc.                                                                       Page 15
Problems Corrected


4.2 Corrected in Version 04.02.04.00

4.2.1 General
        Issue                                              Resolution
        FR-ATM ARP Packets not compatible with             The code now correctly sets address fields in the Frame
        routers. (C97-13300)                               Relay packets to work with routers that have ARP or
                                                           Inverse ARP enabled when FR-ATM Service
                                                           Interworking is selected on the B-STDX switch.
        If a B-STDX switch was rebooted with a new         The software was modified to correct this problem.
        4.2 version of code, some IOP type A cards
        would fail to come up.
        IOP cards crash after deleting circuits from the   The problem was fixed as follows:
        card. Running BulkStats on the card                - A pointer to a BulkStats data structure now clears
        exacerbates the crash frequency.                        the data after the structure is freed.
        (C97-16574)                                        - Deleting a PVC now causes the “back” pointer in a
                                                                doubly linked list to be properly handled.
        Internal diagnostics on channelized and            A software delay was introduced in the diagnostic
        unchannelized E1 cards would occasionally          initialization routines to allow additional time for the
        fail.                                              E1 framer to transition into a framed state. The delay
        (C97-12200)                                        guarantees that the framer will be in frame if all
                                                           hardware components are functioning correctly.
        If the “any path” option is used in a defined      A local pointer variable which was uninitialized at the
        path circuit and a trunk on the path goes down,    start of the routine that calculates a defined path route
        the IOP which contains the caller endpoint of      is now initialized correctly.
        the circuit will crash. (C97-16539)
        Displaying trunk statistics for an OPTimum         OSPF and traffic delivery problems no longer exist on
        trunk on an ATM CS or IWU card displays            ATM frame optimum trunks that are configured on
        question marks for the available bandwidth.        ATM CS/IWU cards. Circuits on a negative bandwidth
        (SQA-2299)                                         trunk load balance off of the trunk and new circuits
                                                           will not be routed over this trunk.
        Bulk statistics were incorrect after a circuit     The software was modified to correct this problem.
        reroute. (C97-13386)



        Circuit Receive counters may be incorrect after    The software was modified and the counters were
        a PVC is rerouted. (C97-12150)                     corrected.




Ascend Communications Inc.                                                                              Page 16
Problems Corrected


4.2.2 PPP
        Issue                                           Resolution
        If intervening equipment between the switch     The Link Control Protocol (LCP) now sends configure
        and the CPE experience an outage which          requests continuously, instead of stopping after 10 retries.
        neither switch nor CPE detect at the physical   The CPE device can now recover through a restart after
        level, PPP will not come up if the outage       service is restored.
        exceeds the Config Request retry count.
        (C97-16007)
        NCP would initiate negotiations of Bridging     NCP will no longer attempt to initiate negotiations for
        at link up. This may be a problem for certain   Bridging. However, the switch will negotiate if initiated
        CPE‟s. (C97-16007)                              by CPE.

        CHAP Challenge would fail using certain         The name field in the CHAP challenge is now
        Routers that depend on Name (Peer ID) to be     “CascadeXXX.YYY where XXX.YYY is the switch
        known at the Router. (C97-11602)                number as displayed on Cascadeview ->Administer ->
                                                        Set Parameters ->Get Sw Attributes.



4.3 Corrected in Version 04.02.03.00

        Issue                         Resolution
        Inverse ARP Failure on        The problem whereby inverse ARP failed on PPP-1490 translation FRAD
        PPP-1490 (C97-15500)          circuits which traversed a trunk which existed on the same card as the
                                      ingress PPP-1490 Lport has been corrected.
        OSPF Calculates Wrong         On an E3 trunk, OSPF now calculates the correct value for the
        Value (C97-13015)             oversubscription factor.
        SMDS Upgrades                 The problem whereby upgrading the last switch in your network to version
                                      4.2.3 from a release version lower than 4.2.0 results in SMDS paths to all
                                      switches being lost, has been corrected.
        Low Memory on IOP             The problem whereby low memory on IOPs caused the trunk state in OSPF
        (C97-14758)                   to trash limits for adding new PVCs has been corrected.
        LMI Asynch Delay              Problems with LMI asynch delay and full status updates on NNI ports
        (C97-13409)                   causing circuits to delay in coming up has been resolved.
        OPTimum Trunk on              You can now configure and OPTimum Trunk on an unchannelized E1 NNI
        Unchannelized E1              port.
        (C97-13519)
        Changing CP Capability        Changing the CP capability no longer causes both CPs to go down at the
        (C97-10173)                   same time.
        ISDN Channel Alignment        There are no longer any problems with ISDN channel alignment above
        (C97-14451)                   TS16 (Australian ISDN).
        UPC Parameters                For very high MBS values, UPC parameters correctly round off to 1
        (C97-13103)                   (instead of zero).
        Sleepy PPP (C97-14897)        PPP link now comes up without resetting the port.
        4.0 to 4.2 PVCs               The problem whereby 4.0 to 4.2 PVCs would not pass traffic has been
        (C97-14849)                   corrected.
        SMDS Lport                    The problem whereby SMDS Lports can‟t be modified once internal
        (C96-9507)                    diagnostics are executed and the pport is still down has been corrected.
        IOP Cards not Receiving       IOP cards are now receiving updates on SMDS paths.
        Updates (C97-14434)
        Priority Routing              Priority routing now functions correctly in this release.
        OAM on B-STDX                 OAM is now supported on B-STDX ATM cards.
        MIB-II Variables              The MIB-II variables ifType, ifMTU, and ifDesc are now correctly
        (C97-10709)                   implemented.



Ascend Communications Inc.                                                                              Page 17
Problems Corrected



        Issue                       Resolution
        CP Rebooting                The problem whereby downloading the CP boot image caused the CP to
        (C97-15670)                 reboot whether requested or not, has been corrected.
        PPP Causing Out of          PPP no longer caused “out of memory” condition on channelized DS3 I/O
        Memory Condition            modules.
        C97-15719)
        OSPF Default Route          With large network topology and major topology shifts, the switch no
        Calculations                longer loses memory or crashes. The problem was caused by OSPF route
                                    calculations which were not aged out the tables.
        Direct Cell Trunks          The problem whereby direct cell trunks configured on the CS DS3 and
                                    IWU OC3/STM-1 card may go down and fail to come back up when put
                                    under severe stress (over driving the trunk with small frames (less than 100
                                    bytes) and/or continuously bouncing the trunk) has been corrected. This
                                    problem may have also caused the card to get shot; background diagnostics
                                    displays the error message Warm Boot NEC_CHIP.
        DSX card reboots 30.x       This problem no longer occurs after the addition of a logical port with
        (C97-12658)                 congestion control set to on.
        Overloaded trunk            An overloaded trunk no longer causes the CP to become CPU-bound at
        (C97-13552)                 100% utilization.
        ATM CS/DS3 card cables      The problem whereby long cables do not work with ATM CS cards on the
        (C97-14716)                 B-STDX or ATM DS3 cards on the CBX has been corrected. See Section
                                    5.5 on page 32 for more information on pport parameters.
        Coldbooting a redundant     Coldbooting a redundant HSSI card no longer causes a switchover to the
        HSSI                        standby card.
        PRAM conflicts              The problem whereby the switch does not reject SNMP sets for cards in a
                                    PRAM conflict state has been corrected.
                                    If you move an IOP from one switch to another switch or one slot to
                                    another slot and leave it there in a PRAM conflict state for more than 24
                                    hours, the node no longer goes unreachable.
        Management VPI/VCI for      Management VPI/VCI endpoints are now supported on the CS/IWU card.
        CS/IWU card
        STM-1 card continuously     If there is a multiple hop SMDS virtual path established and the switching
        crashing with error 20.0    points in a single switch include the following trunks:
        (C97-14300)                  An SMDS OPTimum trunk
                                     An STM-1 ATM direct trunk
                                    The STM-1 card no longer continuously crashes with error 20.0.
        ATM circuit fails to come   The problem whereby an ATM circuit may fail to come up due to lack of
        up                          bandwidth has been corrected.
        Redundant IOP cards         A redundant IOP card that is heavily loaded (at a CPU utilization rate that
                                    is greater than 85%) no longer fails when a new card is inserted as the
                                    standby card, or any event that would require PRAM to be copied to the
                                    standby card.
        LMI                         The problem whereby a full status message interval must elapse before a
                                    circuit will become active after a reroute if that circuit is running over an
                                    NNI has been corrected. This is now only necessary if the outage exceeds
                                    the async delay threshold. Additionally, following a reroute, the switch
                                    will request a full status inquiry on the LMI poll immediately following the
                                    reroute. This minimizes the recovery to a single LMI poll interval. This is
                                    implemented where LMI async messages are disabled.
        OSPF Memory Leak            Memory leak due to OSPF default path pre-calculation has been identified
        (C97- 15155)                and fixed.
        Management VPI/VCI          Management VPI/VCI can now be properly configured and executed.
        (C97-14901)                 NOTE: CascadeView/UX Version 02.03.01.02 or higher is required to
                                    configure and support Management VPI/VCI.


Ascend Communications Inc.                                                                            Page 18
Problems Corrected


4.4 Corrected in Version 04.02.02.00

4.4.1 General
        Issue                        Resolution
        Large OSPF route and         It may take up to 30 minutes for trunks to come up following a switch reset
        database table               if the OSPF route and database table become too large (more than 450
                                     entries).
        56.0 CP Failure              The gateway switch's CP no longer fails with a 56.0 error when there is a
                                     broadcast storm of RIP updates and the CP has RIP enabled.
        Console modem won‟t          You can now dial into an uninitialized switch.
        answer on uninitialized
        switch
        (C96-7522)
        10-port DSX with             The problem with a 10-port DSX card displaying an “Outbound heap low”
        “Outbound heap low” error    error and the reduction of available memory until the card failed has been
        (C97-11479)                  resolved.
        ATM OPTimum cell trunk       In a previous release, an ATM OPTimum cell trunk would not pass traffic
        with different VPIs          when the VPIs differed along the trunk. This problem has been fixed.
        (C97-13833)
        Define Path Circuit          You can now use Define Path to force a circuit to traverse a management
        (C97-14401)                  trunk.
        Upgrade                      An upgrade from release 4.1 to 4.2 is no longer traffic affecting. The CP
        (C97-12825)                  code has now been divided into two images: cp1.rom (the application code)
                                     and cp1boot.rom (the boot flash image).
        Multicast PVCs               An operator can no longer delete individual PVCs without first deleting
        (C95-2154)                   them from the Multicast group.
        Channelized DS3 deletion     Pseudo logical ports that are created as part of a Channelized DS3
                                     configuration are now removed if the Channelized DS3 card is removed.
        E1 Transmit Clock Source     The NMS no longer incorrectly changes the parameter value
        (C96-7396)                   “pportDs1LineBuildOut”, resulting in a Card Out of Sync condition, during
                                     modification of the E1 transmit clock source.
        “mootman” password           The “mootman” password is now disabled when the debug password is
        (C97-12720)                  changed. Use the set debug password console command to change the
                                     debug password.
                                     NOTE: If the default password is changed and the new password is lost,
                                     there is no way to access the switch through the console.
        Management DLCI              The problem with management DLCI reporting as a multicast DLCI when
        reporting                    using LMI revision 1 no longer occurs.
        (C96-6513)
        Install routine with         The installation routine that is used when installing redundant CPs no
        redundant CPs                longer allows the redundant slot of CP2 to become active.
        (C95-54)
        Direct trunk bounces         The problem with direct trunk bounces on heavy load no longer occurs.
        (C97-12810)
        Circuit endpoint returning   The problem of creating a circuit endpoint on a logical port and receiving a
        noSuchname                   message of noSuchname has been resolved.
        SNMP Errored Looped          The SNMP errored looped condition which occurred on SNMP GetNext
        Condition                    requests no longer occurs.
        (C96-9240)




Ascend Communications Inc.                                                                            Page 19
Problems Corrected



        Issue                        Resolution
        CP continually rebooting     The problem of long frames on an Ethernet connection that caused the CP
        (C96-9623)                   to continually reboot no longer occurs.
        Authentication Domain        The problem of the switch PRAM having the authentication domain ID
                                     after the authentication domain deletion from the NMS has been fixed.

        ARP translation on an        The problem of ARP translation on an interworking circuit which caused
        interworking circuit         invalid protocol address mapping in the router has been resolved.
        (C97-13300)
        ICMP packets                 In a previous release ICMP packets were issued when the destination was
        (C96-9511)                   the node‟s LAN IP address. This problem has been resolved.
        Switch not responding to     In a previous release the switch did not appear to be responding to ARPs.
        ARPs                         This problem has been fixed.
        Multiple PVCs going down     The problem of multiple PVCs going down for no apparent reason when
        for no apparent reason       the Admin Status and LMI were both set to up and graceful discard was set
        (C97-13371)                  to off has been fixed.
        Lost PRAM in CP card         Problem number C96-9925 which involved losing PRAM on CP Basic
        (C96-9925)                   cards has been fixed.
        VC Buffers                   The problem of some PVCs staying inactive and listing a “Not Enough VC
        (C97-12543)                  Buffers” message even though there is adequate bandwidth has been fixed.
        Cell bypass circuits         Traffic no longer continues to transmit out of a physical port that is
        (C97-12953)                  admin‟d down on an ATM cell bypass circuit configured between two end
                                     points on a CS DS3 or IWU OC3/STM-1 card.

4.4.2 Loopback

        Issue                        Resolution
        Line loopback                The line loopback function now operates correctly when testing an OC3
                                     card.
        ATM-CS card loopbacks        The loopback function now operates correctly when testing an ATM-CS
                                     card.
        DS0 Loopback with            The problem with a DS0 loopback option failing when the repeater option
        repeater option fails        was greater than 0 has been fixed.
        (C97-10577)

4.4.3 Logical Ports

        Issue                        Resolution
        Logical ports greater than   Logical ports that are greater than ifnum 24 no longer fail configuration for
        ifnum 24                     Q.922 signaling.


4.4.4 Frame Relay

        Issue                        Resolution
        Frame Relay NNI failing      The problem of Frame Relay NNI failing several Bellcore compliance tests
        Bellcore compliance tests    has been resolved.
        (C96-5568)




Ascend Communications Inc.                                                                             Page 20
Problems Corrected


4.4.5 ATM Cards
        Issue                       Resolution
        Lost traffic with ATM-CS    The problem of losing traffic between two ATM-CS cards and the
        card                        associated statistics not capturing the lost traffic has been fixed.
        (C97-12828)
        STM-1                       The NMS now properly sends the correct value to the switch to execute the
        (C97-13546)                 SDH or Sonet code in the switch. It also now correctly updates the Pport
                                    interface in the database. This fix corrects the “Yellow Alarm” condition
                                    that occurred on the STM-1 card.
        Congestion management       The problem with FECN and BECN not getting set during congestion of a
        on ATM IWU OC3/STM-1        cell trunk has been resolved.
        and ATM UNI CS IOP
        cards
        (C97-12835)

4.4.6 HSSI Card

        Issue                       Resolution
        HSSI performance            The problem which occurred when a frame relay circuit between two HSSI
                                    cards could not sustain 500 byte frame traffic at 32 Mbs when the circuit
                                    was established over an IWU OC3/STM-1 cell trunk. This problem has
                                    been resolved.


4.4.7 DSX-10 Card

       Issue                        Resolution
       DSX-10 Cards Disable         The problem of DSX-10 cards interrupting the T1 framing pattern during a
       Framing                      configuration state change has been fixed.
       (C96-7142)

4.4.8 CS DS3 Card

       Issue                        Resolution
       Coldbooting a redundant CS   Coldbooting a redundant CS DS3 card no longer results in a switchover.
       DS3 card

4.4.9 E1 Card

       Issue                        Resolution
       E1 Trunk with CRC4           The problem with E1 trunks with CRC4 disabled has been fixed.
       disabled                     In a previous release, E1 trunks, with CRC4 disabled, might not become
       (C97-12682)                  active. The STDX port was in yellow alarm and the B-STDX port in red-
                                    alarm, which caused the trunk between them to be down.




Ascend Communications Inc.                                                                            Page 21
Problems Corrected


4.4.10 CP 30/40/50 Cards

        Issue                        Resolution
        CP 30/40/50 and              The problem with redundant CP 30/40/50 cards has been fixed.
        redundancy                   Previously, in a redundant configuration (CP 30/40/50 with CP 30/40/50),
                                     the IOP cards would reboot when you inserted the second CP 30/40/50
                                     card into the switch. When this problem occurred, the IOP cards would
                                     indicate parity errors and/or init errors. Some cards came up successfully,
                                     while others enter the „loadFailed‟ state, and did not become operational
                                     thereafter.



4.4.11 Bulk Statistics

        Issue                        Resolution
        SMDS Total Discard           The problem with the SMDS Total Discard counter being incorrect has
        Counter                      been resolved. See the Bulk Statistics Collector for B-STDX/STDX
                                     Software Release Notice for further details.

        Invalid Circuit “in”         The problem whereby circuit “in” statistics (receive statistics) are
        Statistics (C97-12150)       inaccurate following short trunk outages has been resolved. Outages of
                                     very short durations are now detected and the correct “in” statistics
                                     maintained on the trunk port are now reported.

4.5 Corrected in Release 04.02.00.00

4.5.1 General

        Issue                        Resolution
        Invalid power supply traps   The NMS no longer reports a false power supply trap indicating that the
        (C97-11069)                  power supply is down.
        Manually defined paths       You can now use manually defined paths for paths with up to 16 hops.
        CP trap                      A trap now appears to indicate a successful boot up of a standby CP card.
        Redundant CP cards           The following redundant CP card problems were resolved:
                                      The problem of CP Basic and CP+ redundant pair combinations
                                          resulting in CP crashes or switchovers has been corrected.
                                      The standby CP has been modified to ignore information directed to it
                                          by the console UART. Previously the standby CP would attempt to
                                          process data coming in on the console UART, resulting in symptoms
                                          such as the standby CP reporting 100% CPU utilization, or being shot
                                          by the active CP.
                                      Warmbooting a redundant CP no longer causes the CPs to switchover.
        CP problems during trunk     In previous releases, trunk initialization on a switch would result in
        initialization               intermittent memory corruption on the CP. This corruption could cause CP
                                     crashes, background diagnostic errors, or loss of CP PRAM. This problem
                                     has been fixed.
        CP constantly shooting       The problem of CPs constantly shooting has been corrected in this release.




Ascend Communications Inc.                                                                            Page 22
Problems Corrected



       Issue                       Resolution
       CP Background Diagnostics   The problem of temporarily pausing output from the console during a MIB
       Stall                       walk would result in the information queuing up and stalling the
                                   background diagnostics. For example, if the operator issued a <ctrl>S in
                                   the middle of a “next card” operation, the CP background diagnostics
                                   would stall. This problem would cause the CP to enter into an infinite loop
                                   attempting to process the data. This problem has been fixed.
       CP fatal error              When multiple telnet sessions are displaying IOP memory simultaneously,
       (C96-9372)                  the Control Processor would fail with a fatal error (error 30.1). This
                                   problem has been corrected.
       System                      If a Control Processor with uninitialized PRAM is inserted into a switch,
       CP with uninitialized       IOP cards in slots 3 through 16 of the switch may determine their
       PRAM                        administrative capability incorrectly and report failure during their boot
       (C96-9834)                  sequence. This problem has been corrected.
       pSOS                        A new pSOS kernel has been introduced to correct stack corruption
       (Solution: S96-7485)        problems that occurred with a high interrupt rate and a high context-switch
       (Solution: S96-1011, S96-   rate environment on the switch. During stack switching within the pSOS
       266)                        kernel, there was a chance of corrupting the stack, leading to unpredictable
                                   behavior of the system, such as a CP reboot. The new pSOS kernel fixes
                                   this problem. The new kernel also fixes an Intel error, which requires three
                                   or more bus cycles to be executed during interrupt stack switching.
       Telnet                      The switch no longer crashes when you telnet into the switch during a load
                                   balance reroute of a circuit.
       Frame errors                Frame errors no longer occur over a trunk when you provision a DSX
                                   card.
       Erase Standby               IOPs no longer become unreachable following an erase standby of a
                                   redundant CP.
       Card Admin State            In a redundant configuration where the active card is in the higher
                                   numbered slot, if you modify the Admin status of this card, the NMS no
                                   longer inadvertently changes the standby card instead.
       Trunks                      Trunks no longer fail to come up immediately after you provision or reset
                                   them.
       E1 yellow alarm states      Rebooting an E1 card no longer results in the physical ports coming up in
                                   a yellow alarm state.
       RIP                         A configured static entry now replaces a timed-out RIP table entry; you do
                                   not have to warm boot the CP card.
       Multicast DLCI              The V.35 UIO trunk card no longer hangs and crashes with error 30.3
                                   “heap out of memory” when it is oversubscribed with multicast data.
       Invalid UDP frames          If a UDP frame was received with an odd byte length count in the header,
       (Solution: S96-1011)        the switch would add a 0 byte to pad the frame to an even length, thereby
                                   corrupting other data regions. This problem has been corrected.
       UDP Performance             To enhance UDP performance, UDP input processing now performs a
                                   validation of the UDP advertised header length vs. the actual amount of
                                   data. If an incoming frame fails this check, such as there is less actual data
                                   than advertised, the frame is discarded before its checksum is verified.
       TCP                         The problem whereby if a TFTP “Write” request timed out, the retry
       (Solution: S96-1010)        would be sent to UDP port 0, has been corrected.
       Software Download           The problem whereby the B-STDX was inadvertently accepting and
       (Solution: S96-768)         committing Cascade 500 software images as well as B-STDX corrupted
                                   images into FLASH memory has been corrected.




Ascend Communications Inc.                                                                            Page 23
Problems Corrected



       Issue                        Resolution
       MIB II                       The following responses to SNMP MIB-2 queries for ifSpeed and ifType
                                    have been corrected.
                                     ifSpeed query response from the UIO card was incorrect. The logical
                                        port bandwidth, not the physical port rate, is the correct response.
                                     ifSpeed query response from the HSSI card showed an incorrect
                                        number.
                                        ifSpeed query response from the ATM-DS3 card showed an incorrect
                                        number.
                                     ifType query response from the ATM-DS3 card has been changed
                                        from lapb to DS3.
       MIB variables                 The Channelized DS3 MIB values now have correct default values.
                                     The MIB values cardDiagNonFatalTime and cardDiagFatalTime no
                                        longer report the wrong values.
       Module Identification        MIM is supported via a console command (show hardware).
       Memory (MIM)
       Console commands                  When executing show Lport attributes console command, a logical
                                          port that is added when its physical port was down was reported as
                                          physical port 0 instead of the actual physical port id. This problem has
                                          been corrected.
                                    The show card console command has been corrected to indicate that the
                                    memory statistic it reports is Memory available and not Memory utilized,
                                    as it had reported previously.
       System                        Corrected a problem with the IOPA parity detector.
                                     Modified node number field to correctly report node numbers greater
                                         than 127 for Bulk Statistics file names.
                                     Enhanced error checking to prevent unknown or corrupted images
                                         from downloading into the switch.
       Statistics                   Added internal counters to transmit statistics for frame sizes over 10000
                                    bytes and for TX interrupts with no frames on queue.
       36.0 IOP Type A and B card   The problem that caused IOP Type A and IOP Type B cards to crash with
       crash                        Error 36.0 has been corrected.
       SNMP                          Corrected a problem where if multiple IP addresses are assigned to a
                                          community name, only the first address would be recognized for SET
                                          actions.
                                     When PRAM synching two cards in one switch from two NMS
                                          workstations simultaneously, the SNMP pram sync complete trap
                                          returned for each event indicated the same card number. This problem
                                          has been corrected.
                                     A problem has been fixed where Community Names greater than 31
                                          bytes in length were being accepted by the B-STDX.
                                     On a UIO card, SNMP traps were not being generated when physical
                                          port type did not match with the type defined by the user. This
                                          problem has been corrected.
       SNMP                         SNMP get requests for lportDTEFailReason and lportDCEFailReason
       (C96-9535)                   were returning improper values. This problem has been fixed.
       Ethernet                     The Ethernet link remains alive if any traffic is received during the time-
                                    out interval (60 seconds). Previously it would remain alive only if it saw
                                    directed IP or ARP traffic.
       OSPF                         Relaxed a restriction which prevented release 4.01 from functioning in a
                                    network of class „B‟ IP addresses.




Ascend Communications Inc.                                                                             Page 24
Problems Corrected



       Issue                        Resolution
       LMI                          Corrected a problem where NNI and UNI-DCE logical ports would
                                    initialize and report PVC‟s active before receiving an active bit from a
                                    remote logical port of a PVC segment.
       TFTP                         The problem that caused the file transfer to become suspended, if the
                                    directory in which either the card memory dump file or the PRAM upload
                                    file was placed was write protected, has been fixed.
       Illegal base IP Address      The switch no longer allows the illegal base IP address 255.255.255.255 to
                                    be set.


4.5.2 CP 30/40/50

       Issue                        Resolution
       CP 30/40/50 cold             In a switch with a CP 30/40/50 card, IOP cards no longer fail with 'Bus
       boot/power up                Initialization' errors when you coldboot or powered up the switch. Once
                                    the cards get shot (reloaded), they become operational.

4.5.3 IOP

       Issue                        Resolution
       Redundant IOP Warmboot       When you warmboot or PRAM Synch redundant IOPs, the active IOP no
                                    longer fails or is shot (reloaded) by the standby, resulting in a switchover.

4.5.4 Channelized DS3

       Issue                        Resolution
       Admin Down Physical Port     When you set the physical port Admin Status to Down, the card no longer
       on Channelized DS3 Card      performs a cold boot.
       ISDN, SMDS and ATM           The Channelized DS3 card now supports Frame Relay , Frame Relay to
       DXI/FUNI support             ATM Interworking, ISDN, and ATM DXI/FUNI.
       Trunks on channelized DS3    This release supports trunks on a channelized DS3 card.

4.5.5 HSSI

       Issue                        Resolution
       Intermittent failures        The problem of a HSSI card intermittently failing when you continuously
                                    bounce a trunk that is configured on it has been fixed.
       HSSI driver oversubscribed   The HSSI driver has been fixed to prevent a deadlock condition between
       (Solution: S96-940)          the driver and the DMA controller when the HSSI is being oversubscribed
                                    at low bandwidth.
       HSSI driver and circuits     The HSSI driver has been fixed to prevent circuits from bouncing when
       (Solution: S96-892)          small packets (less than 160 bytes) are transmitted at a high throughput.
       ATM UNI and HSSI             The problem that caused a HSSI card to report a bad function in the trace
                                    area when an SNMP get of a FEAC status was submitted to a switch
                                    containing both an ATM-UNI and a HSSI card has been corrected. The
                                    problem occurred because the HSSI card was inadvertently picking up the
                                    SNMP get of a FEAC status that was actually being sent to the ATM-UNI
                                    card.
       Trunk Protocol               On a HSSI trunk card, if a trunk line was connected to one of the ports that
                                    had a noisy line with numerous frame bit errors, the card would crash with
                                    a fatal diagnostic error of 41.0. This problem has been corrected.


Ascend Communications Inc.                                                                             Page 25
Problems Corrected


4.5.6 ATM

       Issue                         Resolution
       ATM CS/IWU card in a 4.1      An ATM CS/IWU card no longer loads the wrong boot flash when
       switch                        inserted in a 4.1 switch. When the switch is later upgraded to 4.2, the
                                     CS/IWU card will load the correct code and boot without failure.
       9000 cell trunk not showing   You can now display statistics for the B-STDX side of a CBX to B-STDX
       statistics (C97-11515)        cell trunk.
       ATM UNI CS cards              CP boot operations no longer cause the ATM CS cards to stop responding
                                     with the NMS message “response too big”.
       768 byte frames               The ATM IWU card no longer has problems receiving 768 byte frames on
                                     the ATM interface.
       ATM service interworking      The problem of ATM service interworking circuits not passing ATM
                                     traffic that has the CPCS bit set has been corrected.
       NNI ports for CS/IWU          ATM CS/IWU cards now support NNI logical ports types.
       cards
       Logical port bandwidth        The logical port bandwidth between 9000 CS and 500 DS3 ports is now
                                     equal.
       ATM UNI CS cards              The ATM UNI CS card now supports traffic shaping features.
       Defined paths on ATM          The problem of using defined paths for interworking circuits (ATM cards)
       cards                         has been corrected.
       Circuit statistics wrong      Interworking circuit throughput statistics are no longer doubled on the CS
                                     or IWU endpoint of the circuit.
       Hot inserting IOPs            The ATM IWU OC3/STM-1 card no longer goes into red alarm or
                                     experiences any cell errors/discards when hot swapping another IOP.
       OPTimum cell trunks           ATM cards now support more than one ATM OPTimum cell trunk.
       ILMI support                  The problem with logical ports configured on the DS3 CS card and IWU
                                     OC3/STM-1 card with ILMI enabled not coming up has been fixed.
       Physical port statistics      The physical port statistics screens now provide an accurate number of
       wrong                         frames, bits per second, and octet counts for CS and IWU physical ports.
       Physical port Summary         Summary transmit statistics for the physical port of a trunk are now
       statistics                    correctly reported.
       LMI                           In order to allow Frame Interworking circuits to be established quickly, the
                                     LMI handler has been modified to get an LMI exchange 3 seconds after
                                     “link up” rather than waiting until the T391 timer has expired.
                                     The ATM NNI port now defaults to LMI “link up” rather than “link down”
                                     until the first successful LMI poll.
       ATM Network Interworking      ATM Network Interworking circuits now become active without a long
       PVCs                          delay.


4.5.7 SMDS
       Issue                         Resolution
       SMDS PDU violations           SMDS over ISDN no longer shows SMDS PDU violations with SA not
                                     found. The SIP3 Header area for the SMDS PDU Violation Dialog screen
                                     displays the ISDN E.164 address in ASCII format.
       Modify IA area mask           Modification of the IA area mask no longer time out if traffic is running.
                                     The set completes prior to NMS the timeout period.
       SMDS areas missing            The area table is no longer missing SMDS areas following an upgrade to
                                     04.02.00.00.




Ascend Communications Inc.                                                                            Page 26
Problems Corrected



       Issue                         Resolution
       SMDS Optimum Trunk/GA         GA traffic arriving on an SSI with an optimum trunk path to a node with
       traffic                       the GA area no longer causes the card to become CPU bound. In a
                                     previous release this issue occurred under the following conditions:
                                      GA traffic entering node 1 on a DXI that is multiplexed to the SSI at
                                          node 2 (and carries the optimum trunk)
                                      Group Address exists at node 2 and the GA area is known at node 1
       SMDS GA traffic on            When DSX and UIO cards receive SMDS GA traffic at a very high data
       DSX and UIO cards             rate, this causes the cards to crash with memory corruption. This problem
       (C96-8795)                    has been corrected.
       SMDS header                   Traffic from an SMDS DXI-DTE port through an SSI-DTE port failed at
                                     the SMDS DSU because the direction bit in the DXI2 header was
                                     incorrect. This problem affected IA frames only. The problem has been
                                     corrected.
       SMDS vector table             The problem whereby a 4-port T1/E1 card would fail due to vector table
       corruption                    corruption upon receiving excessive amounts of group addressed traffic
                                     has been corrected.
       SMDS GA on IOP-B              SMDS GA handling on the IOP-B platform has been fixed to prevent a
       (Solution: S96-808)           frame buffer leak where a trunk (Direct or Optimum) exists on the same
                                     card as the SSI receiving the GA traffic.
       SMDS DXI on HSSI card         An SMDS problem has been fixed where if a DXI port is configured in the
       (Solution: S96-714)           same HSSI card as a Direct Trunk then AIA frames input into the DXI
                                     cannot be sent over the Direct Trunk to a remote HSSI. This problem only
                                     occurred if the area of the AIA was not known to the switching system.
       SMDS Path setup               During SMDS virtual path setup, the SMDS component allowed a virtual
                                     path to be setup even on a frame-relay only card. This problem has been
                                     fixed.
       SMDS Trunk between            When a direct trunk carrying traffic between two SMDS clouds is broken,
       clouds                        traffic reverts to an SSI which has a “management only” trunk. Previously,
                                     data would be erroneously discarded.
       PRAM Synch problems           After a card reboots from a PRAM sync, a logical port which is
       with SMDS logical ports       operational and passing traffic will incorrectly send an Lport Down trap if
                                     it receives no responses to its heartbeat polls.
       SMDS logical and physical     The switch no longer reports invalid values in the SMDS logical and
       port statistics               physical port statistics.
       SMDS Protocol Error           Protocol error checking now correctly occurs on SMDS traffic.
       Checking
       SMDS traffic                  Oversubscribing SMDS traffic on a 10-port DSX or 8-port UIO card will
       oversubscription              no longer cause the card to crash with a 36.0 diagnostic error code.
       Corrupt IA Screen Member      The problem that caused the PRAM upload to inadvertently indicate
       map                           differences in the IA screen map on a DXI of an E1 card that had many
                                     members and was PRAM synched, followed by PRAM upload, has been
                                     fixed.
       SMDS SS                       The problem that caused DXIs muxed to a remote SSI with particular slot
                                     number associations to fail to switch traffic has been fixed.
       CP with large configuration   Control Processors with a large SMDS configuration may get shot by a
       (C96-9153)                    standby CP when warm booted. This problem has been corrected.
       SNMP                          If a Group Address is modified (added to or removed from), while the
                                     card(s) with the Individual Address(es) is down, the SNMP set will fail.




Ascend Communications Inc.                                                                            Page 27
Problems Corrected


4.5.8 Logical Ports

       Issue                          Resolution
       Logical port conflict traps    When the user deletes a logical port on a switch and then adds a logical
                                      port on a different card that uses the same interface number, the switch
                                      would sometimes report an 'Lport Conflict' trap. This was due to the
                                      original logical port being marked as deleted, but left in the interface table.
                                      The deleted logical port will now be correctly removed from the table.

4.5.9 PVCs

       Issue                          Resolution
       PVC re-routes after            While OSPF is reinitializing, if an IOPs PVC manager executes the
        CP reset                      periodic load balance re-route, then the evaluated PVC(s) no longer
                                      experience the effects of a load balance reroute if a better route does not
                                      exist. OSPF is re-initialized upon a CP reset (either cold-start, warm boot,
                                      switch-to-redundant, or PRAM sync).
       Resilient NNI/UNI (fault       PVCs now function properly following an upgrade.
       tolerant PVCs)
       PVCs fail to come up           The problem of circuits failing to come up under extreme conditions has
                                      been corrected.
       Switches with multiple         Circuits no longer fail to come up because they are trying to use a trunk
       trunks                         which has reached the maximum number of allowed circuits on a single
                                      trunk. The OSPF cost of using this trunk is correctly calculated and
                                      circuits are correctly routed to a lower cost trunk.
       Asymmetric circuits            The problem of reverse circuit parameters being set to the same values as
                                      the forward circuit parameters has been fixed.
       LMI                            When a PVC down event is received during PVC re-route process, any full
       (C96-9437)                     status inquiry message will get a PVC state active response until the LMI
                                      asynchronous update delay timer is expired.
       PVC                            Corrected the problem that caused valid data to be discarded when the
                                      Admin Status value of a circuit is set to up before the logical port Admin
                                      Status value is set to up. This problem would occur following an external
                                      diagnostic test on a logical port.
       Manually Defined Paths         PVC routes configured with defined paths requiring greater than 8 hops
                                      would occasionally fall off the defined path and not recover without an
                                      admin down/up of the PVC. The fix supports the documented limit of 16
                                      hops.
       Management DLCI                The problem that limited Management DLCI on FR-UNI-DTE and FR-
                                      NNI has been corrected.
       PPP-1490 Translation           Corrected a problem where circuits on a PPP-1490 Translation logical
       Logical ports                  ports on 10 port DSX-1 cards on ports 5 through 10 would cause a system
                                      reboot when activated.
       PVC routes                     Circuit routes are no longer torn down when the circuit is on a negative
                                      virtual bandwidth trunk. However, the route may momentarily fail and
                                      then recover.
       PVC Manager                    The Circuit Failure Reason is no longer incorrect when a PVC is
                                      established for the first time and the port at the destination IOP has no
                                      modem signals.
       PVC re-routes after CP reset   PVC no longer re-routes after CP reset.
       Generation of FECNs and        The problem where a card configured with a PVC with no congestion
       BECNs                          generates FECNs and BECNs has been corrected. Performance has
                                      improved to full T1/E1 speed on 21 channels with 128 byte frames.




Ascend Communications Inc.                                                                                Page 28
Problems Corrected



       Issue                        Resolution
       PVC Manager                  Problem with re-routes not being performed correctly when older revisions
       (C96-8116)                   of switch software exist in the network has been fixed.
                                    If a “switch to redundant” or a warmboot was done to a CP, spurious re-
       (C96-8689)                   routes of PVCs would occur. The algorithm to determine if re-routes are
                                    necessary has been corrected for this situation. Note that leaving the
                                    switch tuning parameters at default levels may minimize the occurrence of
                                    this problem.


4.5.10 UIO

       Issue                        Resolution
       UIO physical port type       If a UIO card's physical port type does not match the configured type, a
       mismatch                     trap is generated to the NMS.


4.5.11 ATM DS3

       Issue                        Resolution
       DS3 Loopback                 If you run a Far-End Loopback test followed by an External Test, the test
                                    no longer fails with a “Bad Data Compare” or “No Data Received” error.
       ATM DS3 Cards                Problem with executing „Internal Diagnostic‟ where test reported
       (C96-5919)                   “FAILURE: Bad Data Compare” or “FAILURE: Time out - No data
                                    Received” on good cards has been fixed.
       ATM DS3 Loopbacks            When initiating a DS3 far-end loopback, if the remote port was already in
       (C96-7897)                   loopback, the line would become looped at both ends preventing access to
                                    the remote port. This problem has been corrected by ensuring that a looped
                                    back signal may not set the local port in far-end loopback.
       ATM DS3 Loopbacks            A new status, Far-end loopback Initiated, may be reported at the NMS.
       Far-end Loopback Initiated   This status is reported when a far-end loopback is initiated from a physical
       status                       port on an ATM DS3 card. NMS 2.0.2.14 or greater is required.
       (C96-7897)
       ATM DS3 Loopbacks            Initiating a far-end or line loopback from an ATM DS3/E3 logical port
       Logical port remains down    would cause the port to remain down after attempts to clear the loopback.
       (C96-7897)                   This has been corrected by requiring all DS3 loopbacks to be initiated at
                                    the physical port level on ATM DS3 cards.
       ATM E3 OPTimum Trunk         The problem that caused the ATM E3 OPTimum trunk redundant cards to
       Redundant cards              crash with internal error 36.0 and switch over after a system reboot has
                                    been fixed.
       DS3 Line Loopback            The loopback status no longer displays incorrect mode or invalid values
                                    for the Received FEAC Status on a DS3 card that has frame chip PM7321.
       DS3 Line Loopback to         The problem that caused the loopback status and the received FEAC status
       Payload Loopback or vice     to incorrectly reflect the change when running a line loopback test
       versa                        followed by a payload loopback test, without running the stop test in
                                    between tests, or vice versa, has been corrected.
       Data corruption              Data corruption no longer occurs after you change the DS3 physical port
                                    Admin Status to Down.




Ascend Communications Inc.                                                                           Page 29
Problems Corrected


4.5.12 DS1 and DS0 Loopback


        Issue                        Resolution
        DS0 Far End Loopback         The following DS0 far-end loopback issues have been corrected:
                                        When you send the test bit pattern to a DS0 that is in far-end
                                         loopback, it no longer goes out of sync.
                                        The NMS prevents you from selecting multiple DS0s for far-end
                                         loopback testing. This prevents the port from getting stuck in far-end
                                         loopback mode.
                                        When deactivating a DS0 far-end loopback an appropriate trap is now
                                         sent to the NMS
        Loopback                     A DS0 or DS1, which is already in loopback, will no longer accept another
        (Solution: S96-924)          loopback activation from being set until the first is stopped/disabled. Only
                                     one loopback is allowed at a time.
        DS1 Loopback                 Nx56 Far-end out-of-band is now supported.

4.5.13 BERT Testing

        Issue                        Resolution
        BERT Pattern                 The problem where the pattern “3 of 24” generates an incorrect pattern has
                                     been corrected.
        BERT Bit Error Rate          You can inject an error rate at a single bit rate.
        Bert Out of Frame            Lower byte user-defined BERT patterns no longer result in Bert Out of
                                     Frame.


4.5.14 Bulk Statistics

        Issue                        Resolution
        Negative Values in Reports   The problem whereby a Node ID greater than 128 bytes resulted in
                                     negative values in the Bulk Statistics reports has been corrected.

        Memory Corruption            The memory corruption problem in Bulk Statistics collection, which
                                     resulted in Timer Chain Corrupted error, has been corrected.


4.5.15 ISDN

        Issue                        Resolution
        Card crash                   DS3 card no longer crashes when a cable is removed from a 4-port ISDN
                                     card.
        IPX over PPP                 IPX over PPP traffic traversed anywhere along the circuit path would crash
        (C96-9445)                   a high speed IOP card.
        PPP Echo request Timeouts    If CPE equipment over a PPP link passing through a DSU have the Admin
        cause physical port traps    value set to down, PPP echo request time-outs recognize this condition and
        (C96-7997)                   report circuit down with UNI/NNI down as the reason code. However,
                                     physical port down and up traps were also inadvertently reported. This has
                                     been corrected.
        Inverse ARP                  Inverse ARP responses and gratuitous ARP requests from a PPP-1490
        (C96-9280)                   logical port through a switch to a frame relay CPE would not succeed if
                                     passing through a trunk.



Ascend Communications Inc.                                                                             Page 30
Problems Corrected



       Issue                       Resolution
       IPCP configuration          IPCP configuration request commands issued without the IP address
       (C96-9412)                  option are now ACK‟d with no IP address provided rather than NAK‟d
                                   with IP address 0.0.0.0 returned.
                                   NOTE: The above is utilized on an IP unnumbered serial port. IP
                                   unnumbered serial ports on routers need another serial port or Ethernet
                                   port on the same router with an IP address configured and operational
                                   before the unnumbered serial port will pass data.
       ISDN PPP - Gratuitous ARP   In this release, an individual dial-in user is mapped to a nailed up PVC.
       feature                     Cascade maintains a local ARP cache, mapping the DLCI to the client IP
                                   address. As dial-in users connect and disconnect, the ARP cache is
                                   updated and the PVC state changes are communicated to the egress routers
                                   via LMI asynchronous status updates. The egress router then transmits
                                   inverse ARPs and the ARPs attempt to resolve changes in the DLC I state.
                                   The B-STDX software intercepts the inverse ARP in its 1490-to-PPP data
                                   stream and responds with the appropriate information. This configuration
                                   works well except in instances where a particular dial-in client (or specific
                                   IP address) or DLCI is reused before the egress router has updated its ARP
                                   cache.
                                   Reusing the same DLCI before the egress router has updated the ARP
                                   cache, causes a scenario where if a second call is placed on the same DLCI
                                   with a different IP address, it is not recognized by the router, and data will
                                   not pass. Gratuitous ARP ensures that when a new call is accepted by the
                                   B-STDX, the new IP address associated with the DLCI is updated in the
                                   router‟s cache.
       PPP                          A problem has been fixed in PPP where an Echo Request that was not
                                        8 bytes in length was ignored. PPP now sends an ARP when IPCP
                                        goes into open state.
                                    The BCHAN_MSG that is passed from the ISDN Call Control Task to
                                        the LCP in the PPP stack was not being freed. This condition was
                                        causing a memory leak when calls were being placed and torn down.

       ISDN                        Corrected the following issues:
                                    ISDN Timer Expires were causing the Telenetworks routines to be re-
                                       entered. This caused the globals that stored the PKT addresses to be
                                       corrupted which in turn caused unpredictable results.
                                    The ISDN main routines for each of the layers made a copy on the
                                       stack of all of the ISDN messages that go through them. This was not
                                       necessary because a common memory pool is used by all of the ISDN
                                       tasks and the block cannot be reused until it is freed. This became an
                                       issue when the Timer Expires began to be posted to the ISDN Tasks
                                       queues as the amount of traffic through these routines increased
                                       significantly.
                                    Timers that were being restarted were not being stopped then started
                                       so they would not expire when expected.
                                    When the D-channel drops for any reason and Layer 2 has
                                       information frames that are in the information queue waiting for
                                       acknowledgment, the queue entry was not being properly removed.
                                       As a result of the queue entry not being removed properly, the block
                                       that was holding the information frame on the queue was being freed,
                                       but the address of the block was not being removed from the queue.
                                       This condition would cause the block to be freed twice which in our
                                       system causes a fatal error.



Ascend Communications Inc.                                                                            Page 31
Restrictions and Special Considerations


5. Restrictions and Special Considerations
5.1 Networks with 800 or More OSPF LSAs
        Networks with 800 or more OSPF LSAs require a CP+, CP40, or CP50 control processor card.

5.2 Login to Console
        When trying to login to the console running Release 4.2.3 and 4.2.4, be sure that your console cable has
        pin 8 (DCD) on the switch side “high.” This is accomplished by connecting pin 8 to pins 20 or 4. The
        cable that is currently shipped with the switch has the correct pin settings and will work properly. Cables
        missing pin 8 that worked in previous releases will no longer work with Release 4.2.3 and 4.2.4 or higher.
        Crossing pin 8 to pins 20 or 4 to make DCD high enables modems to disconnect properly when exiting
        from a switch.
        The pinout is not required in Release 4.2.5, however, you may have problems accessing uninitialized
        switches.

5.3 Script Download
        When you send an initialization script to a B-STDX switch via the console interface, a 2/10 second delay
        is required between each set command.

5.4 Hardware
        Do not pull down the reset latch on the active CP at any time. If you want to remove the current active CP
        card, switch the card to Standby mode, then remove it. For a single CP system, the switch should be
        powered off before the CP is removed.
        The DS0 loopback feature is only supported on channelized T1s which are Rev C19 or greater. If you
        have a card prior to Rev C19, the loopback feature on the NMS will be disabled (grayed out) for that card.

5.5 IOM
        If you make pport parameter changes when the pport is admin‟d down, these changes may be ignored by
        the switch software. Therefor, if you want to make changes to the physical layer parameters such as line
        length or PLCP mode, you should make them when the pport admin state is up.

5.6 RIP
        If you are using SunOS 5.5.1 (System V Release 4) and want to use the RIP protocol to update the routing
        table entries, you must run the in.routed process on the host system. Once the process is started, any
        manual additions and deletions of the routing entries cause the table to freeze. The table no longer accepts
        updates from RIP. To resolve this problem, you must restart the in.routed process. In addition, you may
        also need to remove and/or flush the existing entries before restarting.
        If you use RIP with multiple gateways on a flat network (all gateway nodes and NMS server are on the
        same LAN segment), you may end up managing a gateway node through another gateway node located
        multiple hops from this node. This also means that nodes located close to gateway nodes may not be
        managed through that gateway node.
        Before you enable RIP on a gateway node, verify that an appropriate firewall is in place to prevent the
        router network and Cascade network from learning routes that are not meant for each other.

5.7 STDX Interoperability
        This release of switch software supports interoperability with STDX switches running switch code
        versions 2.4.0.0 and above.




Ascend Communications Inc.                                                                                Page 32
Restrictions and Special Considerations


5.8 Software Download
        The NMS is temporarily unable to reach the switch while the software is being committed to flash
        memory during a B-STDX system software download.

5.9 Real-time Clock
        An NMS script sets the real-time clock for all switches that have a community name of cascade. If a
        switch does not have a community name of cascade, the real-time clock is not set.

5.10 Connecting to B-STDX Through a Modem Connection
        If you are connecting to a B-STDX switch through a modem connection you must check and set the
        modem as follows:
              Enable Auto Answer
              Override Data Terminal Ready

5.11 Switch Management

            Issue                   Description
            Trunk limitation        Although a Release 04.02.02.00 or greater network can support up to 50 trunks
                                    per node, in a mixed software network (switches with 04.02.02.00 code and 4.1
                                    code), any node that has over 40 trunks configured will fail to come up when
                                    added and/or fail to recover when bounced. The trunk appears to remain in a
                                    loading state. To resolve this problem, delete the extra trunks (over the 40
                                    limitation) on the node or nodes which have them. Until you upgrade the entire
                                    network to Release 04.02.02.00 or greater, the trunk limitation remains at 40.
            Multicast DLCI          The following multicast DLCI restrictions exist:
                                     Service is temporarily disrupted if you delete a PVC from a Multicast
                                          DLCI.
                                     Do not use Multicast DLCIs with SMDS OPTimum trunks.
            Install console         The install console command does not support the initial configuration of
            command                 SMDS and ATM services.
            Traps Generation        Dynamically disabling the Traps Generation field of a switch‟s NMS entry has
                                    no effect until you PRAM Synch the switch.
            Management Interface    You cannot use multiple management interfaces simultaneously on the same
                                    switch.
            Console interface       The console interface requires a user name and password. You can use any user
            password                name, when the CP has a PRAM image; the password is any of the defined
                                    read/write community names. Note that login and logout traps are sent to the
                                    NMS.
            Synchronizing PRAM      Immediately after you PRAM Sync a CP, there is a 30-second window of time
                                    during which any NMS user activity may cause undesirable results. Therefore,
                                    it is recommended that you do not perform any switch configuration, IOP
                                    PRAM Synch, or any other NMS operations during this 30-second period of
                                    time.
            SNMP set                When you add a multicast DLCI with 200 members, the SNMP set may be
                                    rejected and result in a “bad value” error for MIB object
                                    cktMcastMemberList.12.101. The set succeeds when you reduce the number of
                                    members to 100.




Ascend Communications Inc.                                                                             Page 33
Restrictions and Special Considerations


5.12 IOP Type A

         Issue                       Description
         Maximum IOP card            Maximum IOP card performance has been tested at 12 to 14 Mbps,
         performance                 depending on the network configuration.
         IOP traps                   In situations where the IOPs are generating a large number of traps, the CP
                                     may become overloaded and drop traps.
         Redundant IOP Coldboot      When you coldboot redundant IOPs, there is no way to determine which
                                     IOP will become the active IOP.

5.13 Logical Port

         Issue                       Description
         PPP-IPX encapsulation       The Cascade switch supports only PPP to IPXCP RFC 1552. To get
         (Point-to-Point) to         IPXWAN support, you must attach a router between the Cascade switch
         (Internet Packet Exchange   and the Novell network that supports both IPXWAN and IPXCP. Cascade
         Protocol)                   supports PPP encapsulation of IP with no option support (except IP
                                     address assignment for ISDN), IPX with no option support (additionally,
                                     no IPXWAN support) and Bridging with no option support.

5.14 T1/E1

         Issue                       Description
         T1 physical port            The following T1 physical port configuration is not recommended due to
         configuration               the possibility of a false alarm condition:
                                     Link Framing Type: D4,       DS0s selected: 1-24
                                     Zero Encoding Method: B8ZS Logical Port Bit Stuffing: OFF
         Internal Foreground         Executing T1/E1 Internal Foreground Diagnostics on a single logical port
         Diagnostics                 affects all logical ports on the physical port.
         Cable restriction           When you use a cable that is longer than 100 feet, the T1 card does not
                                     work unless it is connected directly to the network.

5.15 HSSI

         Issue                       Description
         Frame errors                The switch reports statistics that include frame errors on a HSSI physical
                                     port that is being overdriven on the receive. The count reflects BTUs
                                     which are being dropped. When this condition occurs the frame error
                                     counts may exceed the rate of the frames coming in. In previous releases
                                     the BTUs dropped were not reported in the physical port statistics.




Ascend Communications Inc.                                                                             Page 34
Restrictions and Special Considerations


5.16 ATM CS/IWU

         Issue                         Description
         ATM OPTimum Frame             The EFCI and CLP bits are set to zero on all cells sent on a B-STDX 9000
         Trunks                        ATM OPTimum Frame Trunk from an ATM CS/IWU card. Data
                                       traversing an OPTimum Frame Trunk on an ATM CS/IWU card will not
                                       map the DE and FECN bits to the CLP and EFCI bits. The DE and FECN
                                       information is carried within the trunk header of the frame and will be
                                       reconstituted at the end of the trunk. The DE to CLP and FECN to EFCI
                                       mapping is done on all Direct and OPTimum cell trunk lports as well as on
                                       NNI lports.

5.17 Power Consumption
       The -48 VDC and 120 VAC power supplies for the B-STDX switch have been tested to deliver 700 watts
       of +5V power. When you calculate configurations, use 40 watts for IOP Type C cards (ATM UNI CS/ATM
       IWU), use 45 watts for IOP Type B card (HSSI), use 40 watts for CP cards and IOP Type A cards (T1, E1,
       DSX, and UIO), and use 35 watts for the Channelized DS3 card.

5.18 Quality of Service (QoS)
        QoS functions only if the entire PVC path supports QoS. This means that if any intermediate node does
        not support QoS, then QoS will not report statistics for that circuit.
        The QoS Min/Max/Ave round trip delay is now calculated over a 15-minute interval instead of over the
        life of the circuit.

5.19 SMDS

        Issue                          Description
        Trunk card failure             After extensive trunk bouncing, it is possible to crash the IOP Module.
                                       This condition exists if there is a single trunk from a switch to the
                                       network. The workaround for this problem is to create multiple trunks
                                       between switches.
        SMDS IA traffic over ATM       If the SMDS frame BA size is > 8132, SMDS IA traffic over an ATM
        OPTimum trunk                  OPTimum trunk is discarded at the trunk port.
        SMDS Group Members             You cannot delete a group address without first deleting all of the
                                       members in the group.
        SMDS Switching System          After you disable the SMDS Switching System (SS) on a currently
                                       functioning SS node, you must synchronize the CP‟s PRAM and then
                                       coldboot the CP card before the disable will take affect. Group Address
                                       traffic from a higher numbered node may still pass over this node‟s trunks
                                       until you coldboot the remote node‟s CP.
        CP Warmboot                    The B-STDX is temporarily unable to switch SMDS user data when you
                                       warmboot an active CP.
        SMDS Trunks                    On any system where SMDS devices are configured, make sure you set
                                       the card service capability for all trunk cards to Multiservice. These cards
                                       must physically be IOP Plus cards (this applies for SMDS device cards as
                                       well). SMDS data cannot pass over a direct trunk if the trunk is on a card
                                       that has its service capability defined as Frame-Relay Only. Due to the
                                       nature of the SMDS switching system software (which routes SMDS
                                       traffic locally without having to access the SMDS cloud), if you have any
                                       trunk card configured as Frame-Relay Only, it is possible that no SMDS
                                       data will be allowed to exit from the switch. This situation can occur even
                                       if other SMDS trunk cards or SSI cards exist.



Ascend Communications Inc.                                                                               Page 35
Restrictions and Special Considerations



        Issue                          Description
        Multiservice                   When you change an IOP card‟s capability from Frame-Relay Only to
                                       Multiservice (or vice-versa), the PRAM on the card is configured with a
                                       different size layout. This condition causes the card to fail itself and reboot
                                       so that the new PRAM layout can take affect. Background diagnostics for
                                       the card displays a Service Capability Changed diagnostic message. Note
                                       that when you replace an IOP card, if the replacement PRAM layout is
                                       different from the configured layout, the card will fail and reboot with the
                                       new layout.
        SMDS SS on 10-port DSX         On a 10-port DSX card, if the network passes bi-directional SMDS IA
                                       traffic at 9.2 Mbps, ports may begin to experience frame errors.
        SMDS SS on 2-port HSSI         A 2-port HSSI card configured for SMDS DXI may experience discarded
                                       IA frames on the opposite DXI at 24.5 Mbps.
        IA/GA area mask changes        When changing an IA/GA area mask, it will take approximately 12
                                       seconds for the switch to complete any updates of these area routing
                                       tables. Do not change the IA/GA area masks until you upgrade the entire
                                       network to 04.02.02.00 or greater. If you change these masks before you
                                       upgrade to 04.02.02.00 or greater, traffic delivery will not be reliable.
        Changing SSI muxing            The NMS now prevents you from changing DXI/SSI muxing on switches
                                       running pre-04.02.02.00 switch code. Release 04.02.02.00 changes the
                                       implementation of this operation to make it easier to perform, however,
                                       these changes are not backward compatible.
        Class B addressing             Do not introduce Class B addressing into an SMDS network until you
                                       upgrade all nodes in the network to 04.02.02.00 or greater. If you use
                                       Class B addresses before you upgrade, traffic delivery will not be reliable.

5.20 Management Trunk
        Management Trunks are supported only on networks in which every node is running the following
        minimum release versions of the switch software:
           B-STDX, Release 04.01.00.00
           STDX, Release 02.04.00.00
        Do not enable management trunks on any node in a network that is running a version of the software that
        does not support management trunks.

5.21 ATM
        The maximum ATM frame size is 8160.

5.22 ISDN
        ISDN enables you to dynamically configure the IP address for dialing in, however, you must manually
        configure the subnet mask from the PC.

5.23 New Frame Relay Rate Enforcement
        To use jump rate enforcement, circuits must be configured with a CIR equal to the Bc.

5.24 Frame Relay SVC Maximums
        Frame Relay SVCs are available to users of the CascadeView/UX NMS Version 02.04.00.00 or later. If
        you are using Version 02.04.00.00 or later of the NMS, you can now set and monitor switch-wide SVC
        configuration parameters on each of your B-STDX 8000/9000 switches. See the Cascade Documentation
        Set Addendum for more information about Frame Relay SVCs.


Ascend Communications Inc.                                                                                 Page 36
Restrictions and Special Considerations


5.24.1 Cards that Support Frame Relay SVCs
       All low-speed cards (IOPA) cards support Frame Relay SVCs. In addition, the following high-speed cards
       (IOPB) support Frame Relay SVCs:
        HSSI
        Channelized DS3

5.24.2 Card Limits
       The following limits indicate the maximum numbers for Frame Relay SVC circuits for the IOPA and IOPB
       images. These limits are based upon one address defined as the Calling Party, and one address defined as
       the Called Party.
       IOPA (Low-speed) Card Limits
       Maximum number of Frame Relay SVCs per port: 991
       Maximum number of Frame Relay SVCs per card: 2214
       IOPB (High-speed) HSSI Card Limits
       Maximum number of Frame Relay SVCs per port: 981
       Maximum number of Frame Relay SVCs per card: 1338
       IOPB (High-speed) Channelized DS3 Card Limits
       Maximum number of Frame Relay SVCs per port: 991
       Maximum number of Frame Relay SVCs per card: 4262

5.24.3 Node and Port Prefix and Address Limits
       The following node and port prefix and address limits exist for Frame Relay SVCs:
            64 node prefixes per switch
            The total number of SVC port prefixes + SVC port addresses must be less than or equal to a total of
             512 for each card
       There is also a limit of 8 SVC E.164 addresses for any IOPB or IOPA that is configured for Multi Service
       mode. For example, you could potentially configure a channelized DS3 card with 4262 SVC circuits and
       have all 4262 SVC circuits use only 8 different destination addresses.
       You should also note that some cards (such as the T1) can be configured for Frame Relay or Multi Service
       and the limits vary sharply based on the configuration. For example, a T1 card that is configured for Frame
       Relay service can support a total of 512 E.164 addresses. However, if the same card is configured for Multi
       Service, the limit is 8.
       Use the following steps to check on a card‟s configuration:
        1.    Select the switch.
        2.    Choose Monitor Cascade ObjectsShow Detail. The Switch Back Panel displays.
        3.    Select the card from the Switch Back Panel.
        4.    Choose the Get Attr… option button. The View Card Attributes dialog box displays.
       If you are not sure how the card has been configured, select the card from the Switch Back Panel and
       choose the Get Attr… option button.




Ascend Communications Inc.                                                                               Page 37
Restrictions and Special Considerations


5.25 Changing the Default Password
        You can use the set debug password console command to change the debug password. The default
        “mootman” password is disabled when the debug password is changed.

       NOTE: If the default password is changed and the new password is lost, there is no way to access the
             switch through the console.

5.26 IPX Traffic
   If IPX data traffic is received on an Ascend PPP Lport, and forwarded over an OPTimum trunk on an ATM
   UNI CS or ATM IWU OC3/STM-1 card, the IPX traffic will be discarded.

5.27 Undocumented New Features
   The following new features are not included in the Network Configuration Guide for B-STDX/STDX or the
   Diagnostic and Troubleshooting Guide for B-STDX/STDX.

5.27.1 B-STDX 9000 ATM CS and IWU I/O Module Direct Cell Trunk
Shaper Priority Assignment
       The section describes the mechanism used in this release to prioritize circuit traffic traversing a direct cell
       trunk on the B-STDX 9000 ATM CS and IWU cards.
       In the initial release of the cell trunk, all traffic on the trunk was assigned to a single shaper for
       transmission from the card. Frames were segmented into cells and the cells transmitted at the rate
       configured. This mechanism gave all circuit data the same transmission priority, so that VBR rt, VBR nrt
       and UBR circuits all would send data equally.
       This release uses the CS/IWU card‟s NEC chip‟s shaper priority to provide a means of prioritizing circuit
       data for transmission. The NEC SAR chip performs the segmentation and reassembly function on the
       CS/IWU cards. In addition it provides 16 shapers used to limit the speed at which data is transmitted from
       the ATM port. The shapers are configured with a peak cell transmission rate, sustainable cell transmission
       rate, maximum burst size (MBS), and a shaper priority. The NEC chip services circuits associated with the
       shapers depending on the priority of the shaper. The lower the priority value the higher the priority of the
       shaper. Shaper 0 has the highest priority. Shaper 15 has the lowest priority. The CS/IWU cards reserve the
       first shaper (shaper 0) of the 16 to be used as a line rate shaper (PCR=SCR=line rate). You can configure
       the other 15 shapers on the Set ATM Physical Port Attributes dialog box. An example of the default Set
       ATM Physical Port Attributes dialog box for an IWU card is shown below.




Ascend Communications Inc.                                                                                   Page 38
Restrictions and Special Considerations




5.27.2 Shaper Guidelines
       To allow for the prioritization of circuits by the circuit‟s QoS class, the switch code creates three groups of
       shapers that are used for VBR rt, VBR nrt and UBR circuits. There are five shapers assigned to each group.
       When you modify the shaper settings, the following guidelines apply:
          Shaper priority set for shaper n+1 must be >= the priority set for shaper n
          VBR rt circuits are assigned to shapers 1 - 5
          VBR nrt circuits are assigned to shapers 6-10
          UBR circuits are assigned to shapers 11 - 15
       Shaper 0 is set to PCR=SCR and can not be modified. This shaper carries control circuits.
       Assignment of a circuit to an individual shaper within a QoS shaper class is based on the SCR (CIR) of the
       circuit. At connection establishment the QoS value is used to select a shaper group, then the circuit is
       assigned to the shaper whose SCR most closely approximates the SCR (CIR) of the circuit. The conversion
       formula is the same as that used by the CAC when determining bandwidth allocation for circuits. For
       example:
       SCR0       = CIR * IOH1 / 8,
       where:
       IOH1 is the Interworking Overhead =
                                  ((N_Avg + AALHS) / 48)/N_Avg,
                     For an average frame size of 256, IOH1 is approximately equal to 0.02148




Ascend Communications Inc.                                                                                  Page 39
Restrictions and Special Considerations


       This enables you to set the shapers to whatever PCR, SCR and MBS desired within each of the three QoS
       groups. The shaper entry priority field determines the transmit priority of the traffic associated with the
       shaper. For example, using the following shaper parameters, a circuit with QoS VBR nrt with an SCR =
       50000 would be assigned to shaper 9 (second QoS group for nrt and 60000 is the next highest SCR). A
       circuit with QoS of UBR, SCR=80000 would be assigned to shaper 11. A circuit with QoS VBR rt , SCR=
       49000 would be assigned to shaper 3.

           Shaper Id        Sust. Cell Rate      Peak Cell Rate      Burst Tolerance                 QoS

                1               353207               353207                  2
                2                89285               353207                  2
                3                60000               353207                 100                     VBR rt
                4                48000                48000                  2
                5                24000                24000                  2

               6                353207               353207                  2
               7                 80000               353207                 100
               8                 72000                72000                  2                     VBR nrt
               9                 60000                60000                  2
               10                44642               353207                  2

               11               353207               353207                  2
               12                72000                72000                  2
               13                50000                50000                  2                      UBR
               14                35000                35000                  2
               15                22321               353207                  2


       The NEC chip schedules the shaper about every 70 ns. When this scheduling takes place a cell is
       segmented from the frame associated with the circuit assigned to the highest priority shaper. The following
       example shows how the priority works when multiple circuits assigned to different priority shapers all
       present data to be transmitted simultaneously.
       Using the shapers set in the example above, 7 circuits are active on a cell trunk. Circuits VC a and VC b
       have QoS VBR rt. Circuits VC c, VC d and VC e are of QoS VBR nrt. Circuits VC f and VC g are UBR
       circuits.

        Circuit Name         QoS class            SCR

        VC a                 VBR rt               72000
        VC b                 VBR rt               68000

        VC c                 VBR nrt              44000
        VC d                 VBR nrt              44000
        VC e                 VBR nrt              44000

        VC f                 UBR                  20000
        VC g                 UBR                  20000




Ascend Communications Inc.                                                                               Page 40
Restrictions and Special Considerations




       The selected shapers are configured with:

                     Shaper #2          Shaper #10           Shaper #15
        PCR           353207              353207               353207
        SCR            89285               44642                22321
        MBS              2                   2                    2
       priority          1                   2                    3
       For shaper #2 the I/M ratio is 1 cell in 4 cell times (SCR/PCR = ¼)
       For shaper #10 the I/M ratio is 1 cell in 8 cell times (SCR/PCR = 1/8)
       For shaper #15 the I/M ratio is 1 cell in 16 cell times (SCR/PCR = 1/16)


       Each of the circuits are assigned to the following shapers:
                  Shaper #2         VC a           VC b

                   Shaper #10       VC c           VC d            VC e

                   Shaper #15       VC f           VC g

      The cell stream emitted for this configuration would appear as follows:
             Slot 1 2        3     4    5    6    7   8    9   10     11   12         13    14    15    16
     Shaper #2      a b                 a    b             a   b                      a     b
     Shaper #10              c     d              e                   c    d                      e
     Shaper #15                                       f                                                 g

        Cell stream vc a and b have 1 cell sent every 4 cell times, then vc c and d send their cells, at this point the
        scheduler recognizes that shaper #2 has higher priority and jumps back and segments a cell from a and b
        again. Then the scheduler returns to shaper #10 and sends a cell for vc e. The scheduler finds no work to
        do for shaper #2 or shaper #10 so it sends a cell from shaper #15 vc f. The cycle then repeats.

5.27.3 Configuring the Shaper Settings
       If you do not change the default Pport shaper settings, the cell trunk operation will remain the same as it
       was in the previous software release. In this case all shapers are set to line rate and all priorities are equal.
       If you use the default shaper settings, the switch will still assign the circuits based on QoS to the first
       shaper in each QoS shaper group, but all circuits will be serviced with equal priority since the shapers
       priorities are equal.

       NOTE: You must use caution when setting the shapers for circuit prioritization when circuits traversing
             the cell trunk have the Graceful Discard option ON. With graceful discard ON, the frame relay
             circuit can burst data up to the access rate of the ingress card. The potential problem is that SCR
             is calculated using CIR which does not take into account the bursts allowed by Graceful Discard.
             If the SCR set in the Pport shaper field closely matches the SCR, calculate from the CIR value then
             when the circuit exceeds the Burst Tolerance value data will accumulate in the VC transmit
             queues. If the source of the data continues to exceed the CIR then discard is inevitable due to
             shaper congestion.




Ascend Communications Inc.                                                                                     Page 41
Restrictions and Special Considerations


5.27.4 G.826 Statistics
       The G.826 statistics display performance monitoring statistics for digital transmission paths.
       G.826 statistics are supported on the following B-STDX I/O modules:
           4-port E1 card
           Channelized E1 card

5.27.5 Displaying G.826 Statistics
       To display G.826 statistics:
       1. On the network map, select the switch object that contains a 4-port E1 or channelized E1 I/O module.
       2. From the Administer menu, select Cascade Parameters -> Set Parameters. The switch back panel
           dialog box appears.
       3. Select (double-click) the 4-port E1 or channelized E1 physical port for which you want to display
           G.826 statistics. The following dialog box appears. (For information about the Set Physical Port
           Attributes dialog box, refer to the Network Configuration Guide for B-STDX/STDX.)




       4.   Select G.826 Statistics. The following dialog box appears.




Ascend Communications Inc.                                                                              Page 42
Restrictions and Special Considerations




        The following table describes the E1 Statistics fields.


                  Field                                                  Description
 Switch Name                                      Displays the name of the switch
 IP Address                                       Displays the IP address of the selected switch.
 Pport ID                                         Displays the physical port ID.
 Refresh Time                                     Displays the time of the last refresh.
           E1 Current Statistics
 Errored Blocks                                   A count of detected block errors (frames with CRC or
                                                  framing errors) within a 15-minute interval.
 Errored Seconds                                  A count of one-second intervals with one or more
                                                  errored blocks or at least one defect within a 15-minute
                                                  interval.
 Severely Errored Seconds                         A count of one-second periods containing ≥ 30% errored
                                                  blocks or at least one defect within a 15-minute interval.
 Background Block Errors                          A count of block errors detected but not occurring as part
                                                  of a severely errored second within a 15-minute interval.
 Errored Seconds Ratio                            A ratio of accumulated errored seconds to total seconds
                                                  in the available time.
 Severely Errored Seconds Ratio                   A ratio of accumulated severely errored seconds to total
                                                  seconds in the available time.
 Background Block Error Ratio                     A ratio of background block errors to total received
                                                  blocks (E1 frames) in the available time.




Ascend Communications Inc.                                                                                Page 43
Usage Based Billing


6. Usage Based Billing
   Usage-based billing for SMDS requires the use of a real-time clock that is present on all CP/CP+ cards. This
   clock is used to schedule billing aggregation periods and mark the beginning and ending of billing periods
   within usage records.
   Whenever you insert a new CP card into a switch, use CascadeView/UX to set the value of its real-time clock.
   This ensures that the real-time clock on the switch remains synchronized between a redundant CP pair and the
   clock on the CascadeView/UX station.

6.1 Bringing Up a New Switch
        You can configure Billing on a switch at either of the following times:
            When the switch is off-line (the switch is unmanaged)
            When the switch is on-line and actively being managed
        The clock is automatically set when billing parameters are changed while the switch is reachable and
        being actively managed. Thus, CascadeView/UX sets the clock when you enable billing after you perform
        the initial switch configuration (i.e., after the switch becomes reachable).
        If you enable billing while the switch is off-line and unreachable, the clock may not be set properly. Thus,
        you must set the clock manually immediately after the switch becomes reachable and before the CP cards
        are PRAM-synched. If you set the clock after the CP cards are PRAM synched, you should reset the CPs
        so that the NMS can correctly schedule the billing software's aggregation period.

6.2 Installing a new CP card
        You must manually set the real-time clock whenever you install a new CP card into an existing switch (for
        example, when a failed CP card is swapped-out or when a redundant CP is added). To avoid affecting the
        billing operation, set the clock immediately after the installed CP card becomes operational.

6.3 Setting the Clock
        To set the clock from CascadeView/UX:
        1.   Select the desired switch on the network map. A selected switch appears highlighted.
        2.   From the Administer menu, choose Cascade Parameters  Set Parameters.
        3.   Choose the Set Sw Attr command button.
        4.   Select the Billing button.
        5.   Choose Apply. The real-time clock is set when the Apply button is activated on the CascadeView/UX
             Billing screen.

       NOTE: Note that CascadeView/UX automatically sets the clock once per day. For more information about
             this, refer to the Cascade SMDS Billing System Administrator's Guide.




Ascend Communications Inc.                                                                                Page 44
Known Anomalies


7. Known Anomalies
7.1 Anomalies in 04.02.05.00

       Issue                           Description
       Upgrade from 4.2.2.6            When upgrading from version 4.2.2.6, the primary CP can no longer
                                       see the standby CP when issuing a show card or show sys
                                       command. To workaround this problem, issue the show software
                                       flash command.
       PRAM differences                When you perform a PRAM Compare, PRAM differences will be
                                       noted for cktAtmQoS and cktAtmRQoS on Management DLCIs. This
                                       discrepancy has no operational impact on the switch.
       Defined Paths                   When a circuit on a defined path is deleted, the defined path object
                                       remains in PRAM. This may result in circuit set failures due to the
                                       maximum number of defined path circuits. To workaround this
                                       problem, delete the defined path prior to deleting the circuit.
       B-STDX makes the router         The B-STDX IW circuit will continuously bounce while receiving
       go up and down every            OAM cells.
       second (C97-16722)
       ATM Management                  You must reset the port on ATM CS/IWU cards after the configuration in
       VPI/CVI on ATM CS/IWU           order for the ATM Management VPI/CVI to function.
       ports
       Trunk on CS/IWU cards           Trunks configured on cell cards (ATM CS/IWU) sometimes fail to come
       sometimes fail to come up       up following a cold boot. To recover the trunk, you must reset the far-end
                                       trunk endpoint on the card.
       Bulk Statistics collection      When the Bulk Statistics collection interval is set to '0' on the switch, a
       occurs at the top of the hour   TFTP transfer occurs for that switch at the top of the hour. Expected
       when the collection interval    functionality is that Bulk Statistics TFTP transfer is disabled when the
       is set to '0'.                  NMS interval is set to 0.
       Circuit lost counts are not     Circuit lost counts (cktOutLostFrames, cktOutLostOctets,
       correct.                        cktOutLostDEFrames, and cktOutLostODEFrames) are inaccurate.
                                       Instead, they are instantaneous snapshots that are reset frequently by QOS
                                       processing.

7.2 Anomalies in 04.02.03.00

7.2.1 General
       Issue                           Description
       Switch Authentication           On switches which have been configured for authentication, if it is
                                       disabled, you will not be able to enable debug mode on the console.
       G.826 Performance               The accuracy of the statistics have been observed to be off by
       Statistics                      approximately 2%.
       Invalid Interface Numbers       When performing a query on the MIB-II object, the if (interface) numbers
                                       associated with OPTimum trunks are reported as “invalid”. The
                                       configured feeder ports associated with these OPTimum trunks report as
                                       usual.
       PRAM Conflict When              Moving an IOP card from one switch to another will cause the IOP to
       Moving an IOP Card              report a “PRAM Conflict” condition. After this condition, a PRAM sync
                                       operation may cause the IOP card to crash. The CP will shoot the card and
                                       the IOP will come up successfully without any intervention.



Ascend Communications Inc.                                                                              Page 45
Known Anomalies


7.3 Anomalies in 04.02.02.00

7.3.1 General

       Issue                         Description
       Mismatched subnet masking     If there is a router on the same LAN as the gateway switch, and:
       with RIP enabled
                                        Both the router and gateway switch have RIP enabled
                                        The router has a Class C subnet mask (255.255.255.0) and the
                                         gateway switch has a Class B subnet mask (255.255.0.0)

                                     Then the switch may respond to some RIP updates with "frame too small"
                                     packets. The background diagnostics for the CP may also show a non-
                                     fatal error of 2.2 "outbound heap low".
       ATM Frame Optimum             OSPF and traffic delivery problems exist on ATM frame optimum trunks
       trunks                        that are configured on frame-based cards other than ATM UNI. When this
                                     condition occurs you will see large negative bandwidth values on the
                                     trunk. Circuits on this trunk will not load balance off of the trunk and new
                                     circuits may be routed over this trunk.
       Incorrect red frame           Red frames are calculated incorrectly when a card is busy. This condition
       calculation                   occurs on circuits that are configured with a rate enforcement value of
                                     simple.
       Transmit Discards             Transmit discards and congestion can occur on a looped V.35 port
       (C96-6332)                    operating at 6 M.
       Direct trunk to a 6000        A direct trunk connection to an STDX 6000 may not come up until the
                                     9000 card is rebooted.
       Loopback status inaccurate    A failed far-end loopback activation may be reported as successful.
       T1 direct trunk               A T1 direct trunk may not recover after a cold boot of the switch.
       QoS Statistics                QoS Statistics may be inaccurate by 5 percent.
       (C95-3359)
       Circuits backed up to a DTE   Circuits backed up to a DTE backup port do not always come up.
       backup port
       Standby CP getting shot       While downloading switch software to a switch configured with a
                                     redundant CP, the standby CP will sometimes get shot.


7.3.2 Loopbacks

       Issue                         Description
       DS1 closed loop               When running an unframed CSU Line loopback test on a DS1 port of a
                                     channelized DS3 card, it is possible for a closed loop condition to occur on
                                     the DS1 port. When the far-end device loops back the loopback code, the
                                     DS1 also goes into loopback.
       Loopback state of OC3 and     Traps that indicate whether a port is in or out of a loopback state are not
       Channelized DS3               generated for all loopbacks on an OC3 card or for DS1 loopbacks on a
                                     channelized DS3 card.

7.3.3 Channelized DS3

       Issue                         Description
       Adding DS3 card               The addition of a DS3 card occasionally fails when doing multiple adds
                                     and/or deletes on the same DS3 card..



Ascend Communications Inc.                                                                              Page 46
Known Anomalies


7.3.4 T1/E1

       Issue                        Description
       E1 configured for 56K        If a logical port is configured for 56 K on an E1 ISDN card, the card
                                    crashes when a call comes in.


7.4 Anomalies in 04.02.00.00

7.4.1 General
       Issue                        Description
       MIB walk                     You cannot perform a complete MIB walk with this release.
       Multicast DLCI frame         Cards discard data when the multicast circuit contains many members.
       discards
       TFTP Server                  If you PRAM synch a card when the TFTP server is not running, the card
                                    becomes unreachable from the NMS, until the PRAM synch operation
                                    times out.
       Erase PRAM                   If you use the NMS Erase PRAM for an IOP card in maintenance mode,
                                    the switch rejects the SNMP request, and PRAM is not erased. If a card is
                                    in maintenance mode, use the switch console to erase PRAM.
       Switch unreachable at the    If an IOP is extremely busy (100% CPU utilization) card is busy and
       time of statistics request   physical or logical port statistics are requested for the associated port, the
       (C95-2080)                   switch becomes unreachable until the statistics window is closed.
       Provisioning PRAM            Refer to the Software Release Notice for CascadeView/UX Version
       compare differences          02.04.03.00 for more information about PRAM compare/upload.
       Upgrade PRAM compare          A switch configured with a non-redundant CP displays a PRAM
       differences                  difference following a software upgrade.
       Foreground diagnostics       Foreground diagnostics fail on full channelized T1/E1 cards.
       OSPF - Ethernet port down    A switch with a configured gateway Ethernet address will not stop
                                    advertising the route if the Ethernet port goes down. To avoid this
                                    problem, any switch that is not intended as a gateway should not have an
                                    address entered on the CP. If a switch was configured as a gateway, and
                                    the Ethernet is unplugged, the Ethernet address must then be deleted and
                                    the CP warmbooted to delete the advertised OSPF entry.
       Rev 1 LMI                    Rev 1 LMI does not work on HSSI, CS DS3, or IWU OC3/STM-1 cards.


7.4.2 CP

       Issue                        Description
       Redundant CP cards           Continuously toggling the active/redundant CP cards may cause IOPs to
                                    reset.

7.4.3 T1/E1
       Issue                        Description
       Frame errors                 Frames of 8190 to 8192 in length on T1/E1 cards are dropped as a “frame
                                    error” when received by a physical port at the switch.
       Frame discards on E1         E1 statistics report frame discard errors. These errors are reported when
       interface (C97-12000)        the port receives CBX 500 signaling.




Ascend Communications Inc.                                                                              Page 47
Known Anomalies


7.4.4 DSX-10
       Issue                         Description
       Background diagnostic error   DSX cards may report a background diagnostic error of 27.2 (bus
       27.2 (bus transmits full)     transmits full) when the card is heavily stressed. Such conditions include:
                                     a card that is fully configured with trunks and PVCs, and the maximum
                                     utilization is defined for all trunks.

7.4.5 HSSI

       Issue                         Description
       Hot swapping with UIO         Hot swapping a UIO card that is switching SMDS frames to a HSSI trunk
       cards                         may cause the HSSI IOP to lose its connection to the CP.
       Overclocking bandwidth        Overclocking the HSSI card beyond 45 Mbps may cause it to fail.
       Foreground diagnostics on     Foreground diagnostics executed on a HSSI logical port causes the
       HSSI logical ports            operational status of the physical port to remain down. Bounce (admin
                                     down/up) the physical port to correct this problem.


7.4.6 ATM Features

       Issue                         Description
       VPC circuits not supported    VPC circuits will not pass traffic over a cell trunk configured on any of the
                                     ATM cards supported on the 9000.
       OPTimum cell to direct cell   When you reconfigure a trunk from a direct cell trunk to an optimum cell
       trunk                         trunk, you must perform a PRAM sync for the trunk to come up.


7.4.7 ATM DS3/E3 Cards

       Issue                         Description
       ATM PVC                       ATM FUNI to FUNI PVC allows only 2M of data to pass through when
                                     the CIR = 20M.
       ATM PVC                       You cannot create an ATM to ATM PVC where VPI x 256 + VCI > 1023.
                                     The switch rejects the set with “Invalid length or type”.
       ATM DS3/E3 trunk              If you perform a coldboot of an ATM DS3/E3 trunk with the endpoints in
                                     an Admin Down state, then Admin Up the endpoints, the trunk may not
                                     become operational. To work around this problem, you must warmboot the
                                     card.
       Link status down              The link status displays down if LMI Rev 1 is used for an ATM NNI
                                     logical port.
       ATM E3 card                   A fully configured E3 card may crash with a 30.3 error “Heap out of
                                     memory”.

7.4.8 Channelized DS3 Cards

       Issue                         Description
       PRAM sync required            If you add a channelized DS3 card to a switch configuration for the first
                                     time, you must PRAM sync the card after you use the NMS to configure
                                     the card.




Ascend Communications Inc.                                                                             Page 48
Known Anomalies


7.4.9 SMDS

       Issue                         Description
       SMDS traffic on non-          The switching system does not prevent SMDS traffic from using paths that
       switching nodes               contain non-switching nodes. This may result in SMDS traffic not being
                                     delivered.
       Physical port summary         Summary statistics for the physical port of a trunk will not account for
       statistics                    SMDS Group Address traffic when the direction of the traffic is from a
                                     lower-numbered node to a higher-numbered node.
                                     The following error condition frames are not detected:
                                     - invalid link ID (set to 1)
                                     - L2 header subfield (set to any invalid value)
                                     - BA size too small
                                     - missing or invalid version field
       SMDS Billing                  If the Adjunct Processor is serviced by NIS, the AP installation changes
                                     that you make to /etc/services have no affect.
       SMDS paths                    SMDS virtual paths will never get established to a switch if the best path
                                     to the switch goes over a trunk which already has the maximum number of
                                     circuits supported on it.

7.4.10 Console

       Issue                         Description
       show card command             The show card command times out for 15 seconds on any card that is in
                                     the process of booting up.

7.4.11 PVCs

       Issue                         Description
       Zero CIR circuits             You may see some statistics which report green frames on zero CIR
                                     circuits even when all frames are red.
       PVCs fail to establish        Under extreme conditions (over 9000 PVCs and continuous trunk
                                     bouncing), PVCs may not be established due to low memory. This is a
                                     non-fatal error.
       Manually defined paths        When you attempt to define a path over an active circuit with Admin
                                     Status Up and an NNI endpoint, the circuit may fail with either “PVC
                                     segment down beyond node 2” or “Admin Status Down.” The workaround
                                     for this condition is to bounce the circuit. The circuit should then come up
                                     using the defined path.
       Resilient NNI reroute traps   Reroute traps are not generated when circuits are rerouted using resilient
                                     NNI services.
       Circuit fails to come up      Occasionally a circuit will fail to come up with a fail reason of not enough
                                     bandwidth. This usually occurs when the bandwidth of the circuit is equal
                                     to (or slightly under) the available bandwidth on a trunk or logical port.
       Circuits not load balance     Occasionally some circuits will not load balance reroute to a better path.
       rerouting                     This usually occurs when the best path already has almost the maximum
                                     number of circuits on it and there is an alternative path available for the
                                     circuit.




Ascend Communications Inc.                                                                             Page 49
Known Anomalies


7.4.12 DS0 Loopback

       Issue                         Description
       DS0 Far End Loopback          DS0 far end loopback does not get activated when a DS0 is not selected.
                                     The NMS indicates that the loopback will be activated.
       Multiple DS0 Far-end          If you enable and disable DS0 far-end loopback tests more than one time
       Loopback enable/disable       on a DS0 that is on a physical port in a red alarm state, the following error
                                     occurs when you disable the test a second time: “value given has the
                                     wrong type or length”.

7.4.13 ISDN

       Issue                         Description
       ISDN T1 card                  The ISDN T1 card supports short haul applications from 0 to 110 feet.
       SMDS frames lost              Frames may be lost in an SMDS over ISDN configuration.
       Card crash                    Under extreme conditions, while continually toggling the Admin status of
                                     a physical port on the ISDN card up and down, the 4-port ISDN card may
                                     fail.
       PRI failure,                  The PRI card may fail with an internal error 30.x while dialing many calls
       Cable Disconnect/             at once.
       Reconnect
       New circuit not activated     When an MP call is established, the circuit is created for 1 DS0 of the MP
                                     bundle. If the DS0 the circuit is activated on is disconnected, the circuit
                                     goes inactive but the other DS0s in the bundle remain connected. A new
                                     circuit is not established for the next DS0 in the bundle.
       DS0 not disconnected          If a DS0 of an MP bundle (other than the DS0 the circuit is connected on)
                                     is admin‟d down, the call is not disconnected and the DS0 is not released
                                     from the MP bundle.
       Base IP address               Base IP Address assignment requires a warmboot in order to take affect.
       PPP Authentication failures   PPP Authentication failures on Non-ISDN logical ports generates a trap
                                     referring to ISDN authentication.
       Dynamic IP address            There are issues with dynamic IP address assignment. For example, after
       Assignment                    getting the client to connect successfully 22 times, at the 23rd call, the
                                     Client reports a BAD-IP error message. The switch sends a trap that states
                                     the IP Address is in use. Since the client is dialing in, the switch should
                                     assign a unique address to the client. If the address assigned is in use, the
                                     switch should assign the next available address.
                                     You must warmboot the IOP to recover from this error. If the client dials
                                     in with a specific IP address, then PPP converges.

7.4.14 OSPF

       Issue                         Description
       Duplicate NMS paths           If there are switches with duplicate NMS paths configured, it is possible
                                     for OSPF paths to be calculated incorrectly if the switch‟s CPs are
                                     rebooted simultaneously. This may make the switches unreachable from
                                     the NMS. To recover, warmboot the CP of each switch separately.




Ascend Communications Inc.                                                                              Page 50
Switch Software Upgrade Instructions




8. Switch Software Upgrade Instructions
   This section describes the procedure for upgrading a B-STDX switch to the current 04.02.05.00 version of
   software. For information on installing a new switch into your network, refer to the Network Management
   Station Installation Guide.

8.1 Prerequisites
        The Network Management Station must be running CascadeView 02.04.03.00 (for UNIX environment)
        before you can install the B-STDX software supplied with this release in any switch. If you have not yet
        upgraded CascadeView/UX, refer to one of the following Software Release Notices (depending on the
        NMS version that you are using) for more information about this NMS release:
           Software Release Notice for CascadeView/UX Version: 02.04.03.00

       NOTE: Use the following upgrade path:

                1. Install the NMS.

                2. Upgrade Bulk Statistics Version to Version 2.5.4.0 (if applicable).

                 Bulk Statistics Version 2.2.0.0 cannot collect statistics from the upgraded B-STDX switches.
                However, Bulk Statistics Version 2.5.2.0 can collect statistics from the following switches:
                - B-STDX switches running Switch Software Release 4.0.18 and higher
                - STDX 3000/60000 switches running Switch Software Version 2.3.7 and higher

                3. Upgrade the STDX switch software. Note that if you have STDX 6000 switches in your
                network, you must upgrade to release 2.4.0.0 STDX software (at a minimum) before upgrading
                your B-STDX switch to 4.2.5.0.

                 A B-STDX must be running release 04.00.xx.xx or greater before you can upgrade to this version.
                If you are running a release prior to release 04.00.xx.xx, you must first upgrade to release 04.00.x.
                Refer to the appropriate Software Release Notice for upgrade instructions.

8.2 Upgrade Procedures
        There are slight differences in the upgrade process depending on the hardware present in the B-STDX
        switch.

       NOTE: Follow the upgrade procedures exactly as specified. PRAM upload and PRAM compare may
             show differences if you do not perform a PRAM sync.

           If you are upgrading a switch with a single CP, refer to Installing the Software into a Switch with a
            Single CP on page 52.
           If you are upgrading a switch with redundant CPs, refer to Installing the Software into a Switch with
            Redundant CPs on page 53.




Ascend Communications Inc.                                                                                Page 51
Switch Software Upgrade Instructions


        Using Multifile Download
        If you are using UNIX NMS Version 02.04.02.00 or greater, you can use multifile download to install the
        switch software. Depending on the type of hardware installed in the switch, use one of the following
        upgrade paths:
            If you are upgrading a switch with a single CP, refer to Installing the software into a Switch with a
             Single CP Using Multifile Download on page 54.
            If you are upgrading a switch with redundant CPs, refer to Installing the software into a Switch with
             Redundant CPs Using Multifile Download on page 55.

       NOTE: Follow the upgrade procedures exactly as specified.

8.2.1 Installing the Software into a Switch with a Single CP
        This section provides installation procedures for a switch with a single CP. Make sure that copies of the
        software images (.rom files) are located in the /tftpboot/cv_switchSoftware directory before you
        begin this installation.
        1.   Verify that the TFTP daemon is running. If it is not, refer to the Network Management Station
             Installation Guide for instructions on how to configure the TFTP daemon to start automatically.
        2.   From the CascadeView map, highlight the switch. Then, from the Administer menu, select Cascade
             Switches  Download Switch Software.
        3.   Select the file cp1.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify that
             the “File Description” box contains version 4.2.5.0 of the CP Application Image before initiating the
             download.
        4.   After the file TFTP transfer completes, the switch is temporarily unreachable and may appear red on
             the map. This should last only a short while, as the CP is burning the new software image
             permanently into its Flash storage area.
        5.   From the CascadeView map, highlight the switch. Then, from the Administer menu, select Cascade
             Switches  Download Switch Software.
        6.   Select the file ioptypea.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify
             that the “File Description” box contains version 4.2.5.0 of the IOPA Application Image before
             initiating the download.
        7.   After the file TFTP transfer completes, the switch is temporarily unreachable and may appear red on
             the map. This should last only a short while as the CP is burning the new software image permanently
             into its Flash storage area.
        8.   From the CascadeView map, highlight the switch. Then, from the Administer menu, select Cascade
             Switches  Download Switch Software.
        9.   Select the file ioptypeb.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify
             that the “File Description” box contains version 4.2.5.0 of the IOPB Application Image before
             initiating the download.
        10. From the CascadeView map, highlight the switch. Then, from the Administer menu, select Cascade
            Switches  Download Switch Software.
        11. Select the file ioptypec.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify
            that the “File Description” box contains version 4.2.5.0 of the IOPC Application Image before
            initiating the download.
        12. Once you successfully download all five images, perform a “show software flash” command from the
            console to verify that the version numbers and build IDs match those provided in this document.




Ascend Communications Inc.                                                                                 Page 52
Switch Software Upgrade Instructions


        13. Once verification of the images is complete, cold boot the switch from the NMS or use the “reset
            system” console command.
        14. The switch takes longer to reboot than normal because each IOP is burning in a new version of the
            boot flash before loading and adopting the new software. This is a one-time only burn in.

8.2.2 Installing the Software into a Switch with Redundant CPs
        This provides the installation procedures for switches with redundant CPs. Make sure that copies of the
        software images (.rom files) are located in the /tftpboot/cv_switchSoftware directory before you
        begin this installation.
        1.   Verify that the TFTP daemon is running. If it is not, refer to the Network Management Station
             Installation Guide for instructions on how to configure the TFTP daemon to start automatically.
        2.   From the CascadeView map, highlight the switch. Then, from the Administer menu, select Cascade
             Switches  Download Switch Software.
        3.   Select the file cp1.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify that
             the “File Description” box contains version 4.2.5.0 of the CP Application Image before initiating the
             download.
        4.   After the file TFTP transfer completes, the switch is temporarily unreachable and may appear red on
             the map. This should last only a short while, as the CP is burning the new software image
             permanently into its Flash storage area.
        5.   From the CascadeView map, highlight the switch. Then, from the Administer menu, select Cascade
             Switches  Download Switch Software.
        6.   Select the file ioptypea.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify
             that the “File Description” box contains version 4.2.5.0 of the IOPA Application Image before
             initiating the download.
        7.   After the file TFTP transfer has completed, the switch will be temporarily unreachable and will turn
             red on the map. This should last only a short while as the active CP is burning the new software image
             permanently into its Flash storage area.
        8.   From the CascadeView map, highlight the switch. Then, from the Administer menu, select Cascade
             Switches  Download Switch Software.
        9.   Select the file ioptypeb.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify
             that the “File Description” box contains version 4.2.5.0 of the IOPB Application Image before
             initiating the download.
        10. After the file TFTP transfer has completed, the switch will be temporarily unreachable and will turn
            red on the map. This should last only a short while as the active CP is burning the new software image
            permanently into its Flash storage area.
        11. From the CascadeView map, highlight the switch. Then, from the Administer menu, select Cascade
            Switches  Download Switch Software.
        12. Select the file ioptypec.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify
            that the “File Description” box contains version 4.2.5.0 of the IOPC Application Image before
            initiating the download.
        13. After the file TFTP transfer has completed, the switch will be temporarily unreachable and will turn
            red on the map. This should last only a short while as the active CP is burning the new software image
            permanently into its Flash storage area.
        14. From the CascadeView Switch Back Panel screen, select “Switch to Redundant Unit” to switch to the
            standby CP.




Ascend Communications Inc.                                                                               Page 53
Switch Software Upgrade Instructions


        15. Return to the CascadeView map, and verify that the switch is highlighted. Then, from the menu bar,
            select Administer  Cascade Switches  Download Switch Software.
        16. Select the file cp1.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify that
            the “File Description” box contains version 4.2.5.0 of the CP Application Image before initiating the
            download.
        17. After the file TFTP transfer has completed, the switch will be temporarily unreachable and will turn
            red on the map. This should last only a short while, as the active CP is burning the new software
            image permanently into its Flash storage area.
        18. Return to the CascadeView map, and verify that the switch is highlighted. Then, from the menu bar,
            select Administer  Cascade Switches  Download Switch Software.
        19. Select the file ioptypea.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify
            that the “File Description” box contains version 4.2.5.0 of the IOPA Application Image before
            initiating the download.
        20. After the file TFTP transfer has completed, the switch will be temporarily unreachable and will turn
            red on the map. This should last only a short while as the active CP is burning the new software image
            permanently into its Flash storage area.
        21. Return to the CascadeView map, and verify that the switch is highlighted. Then, from the menu bar,
            select Administer  Cascade Switches  Download Switch Software.
        22. Select the file ioptypeb.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify
            that the “File Description” box contains version 4.2.5.0 of the IOPB Application Image before
            initiating the download.
        23. After the file TFTP transfer has completed, the switch will be temporarily unreachable and will turn
            red on the map. This should last only a short while as the active CP is burning the new software image
            permanently into its Flash storage area.
        24. Return to the CascadeView map, and verify that the switch is highlighted. Then, from the menu bar,
            select Administer  Cascade Switches  Download Switch Software.
        25. Select the file ioptypec.rom from the list box, then select „No‟ for “Reboot After Download?”. Verify
            that the “File Description” box contains version 4.2.5.0 of the IOPC Application Image before
            initiating the download.
        26. Once you successfully download all images, perform a “show software flash” command from the
            console to verify that the version numbers and build IDs match those provided in this document.
        27. Once verification of the images is complete, cold boot the switch from the NMS or use the “reset
            system” console command.
        28. The switch takes longer to reboot than normal because each IOP is burning in a new version of the
            boot flash before loading and adopting the new software. This is a one-time only burn in.

8.2.3 Installing the software into a Switch with a Single CP Using Multifile Download
       This section describes the multifile download installation instructions for a switch with a single CP. Verify
       that copies of the software images (.rom files) are located in the /tftpboot/cv_switchSoftware
       directory before you continue with this installation.
       1.   Verify that the TFTP daemon is running. If it is not running, refer to the Network Management Station
            Installation Guide for instructions on how to configure the TFTP daemon to start automatically.
       2.   From the CascadeView/UX map, highlight the switch. Then, from the Administer menu, choose
            Cascade Switches  Download Switch Software.




Ascend Communications Inc.                                                                                Page 54
Switch Software Upgrade Instructions


         3.   Select the following software image files from the list box.
                 cp1.rom
                 ioptypea.rom
                 ioptypeb.rom
                 ioptypec.rom
4.   Verify that the “File Description” box contains Version 04.02.05.00 of the corresponding software image file
     and that the “Reboot After Download” option is set to “No” before you initiate the download.
5.   A trap message appears to inform you the download is complete. Use the console command “show software
     flash” to verify that the version numbers and build IDs match those provided in Section 2 on page 8 of this
     SRN.
6.   After you verify that the software image download is complete, cold boot the switch from the NMS or use the
     “reset system” console command.

8.2.4 Installing the software into a Switch with Redundant CPs Using Multifile Download
         This section describes the multifile download installation instructions for a switch with redundant CPs.
         When the download from the NMS takes place, the images are only written to the active CP. This means
         that once the first active CP is downloaded, you must perform a switch-over to download the other CP.
         Verify that copies of the software images (.rom files) are located in the
         /tftpboot/cv_switchSoftware directory before you continue with this installation.

         NOTE: Because the automatic PRAM copy process between the active and standby CPs does not take
               place when the two CPs have different operating software, do not perform any node configuration
               changes between the time you download the first CP and the time you download the second. This
               prevents configuration mismatches between the two CPs.
         1.   Verify that the TFTP daemon is running. If it is not running, refer to the Network Management Station
              Installation Guide for instructions on how to configure the TFTP daemon to start automatically.
         2.   From the CascadeView/UX map, highlight the switch. Then, from the Administer menu, choose
              Cascade Switches  Download Switch Software.
         3.   Select the following software image files from the list box.
                 cp1.rom
                 ioptypea.rom
                 ioptypeb.rom
                 ioptypec.rom
         4.   Verify that the “File Description” box contains Version 04.02.05.00 of the corresponding software
              image file and that the “Reboot After Download” option is set to “No” before you initiate the
              download.
         5.   A trap message appears to inform you the download is complete. Use the console command “show
              software flash” to verify that the version numbers and build IDs match those provided in Section 2 on
              page 8 of this SRN.
         6.   After you verify that the software image download is complete, perform a CP switch-over. To do this,
              access the Set Switch Back Panel dialog box, select the CP, and choose the “Switch to Redundant”
              command. (You can also use the “toggle card 1” console command.




Ascend Communications Inc.                                                                                Page 55
Switch Software Upgrade Instructions


       NOTE: After the switch-over is complete and the download of the new active CP is progressing, the NMS
             will report a marginal condition on the CP for “software mismatch detected.” This is a normal
             occurrence as part of the upgrade process.
       7.   After the switch-over is complete and the NMS can reach the switch, highlight the switch on the
            CascadeView/UX map. Then, from the Administer menu, choose Cascade Switches Download Switch
            Software.
       8.   Select the following software image files from the list box:
               cp1.rom
               ioptypea.rom
               ioptypeb.rom
               ioptypec.rom
       9.   Verify that the “File Description” box contains Version 04.02.05.00 of the corresponding software
            image file and that the “Reboot After Download” option is set to “No” before you initiate the
            download.
       10. Once the download is complete a trap message appears informing you that the download is complete.
           Use the console command “show software flash” to verify that the version numbers and build IDs
           match those provided in Section 2 on page 8 of this SRN.
       11. After you verify that the software image download is complete, cold boot the switch from NMS or use
           the “reset system” console command.




Ascend Communications Inc.                                                                              Page 56

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:16
posted:7/13/2011
language:English
pages:56