Cisco MGX 8800 Series Switches

Table Of Contents

Release Notes for Cisco MGX Route Processor Module (RPM-PR) IOS Release 12.3(7)T3 for MGX Releases 1.3.10 and 5.0.10

Contents

About These Release Notes (IOS Release 12.3(7)T3 for MGX Release 1.3.10)

New Features

Features Introduced with MGX 8800 Release 1.2.21

Previously Released Features

Automatic Cell Bus Clocking

VISM-PR to RPM-PR Connectivity

Configuring the Cell Bus Clock (CBC) Rate

LDP on RPM Running MGX Release 1.2.02 and Cisco IOS Release 12.2(8)T1

Multi-LVC on RPM Running MGX Release 1.2.02 and Cisco IOS Release 12.2(8)T1

Bypass Feature for RPM in Cisco IOS Release 12.2(4)T

Features Not Supported in This Release

RPM Redundancy Support

SNMP MIB

Notes and Cautions

Special Upgrade Procedure for Cisco IOS Release 12.1(5.3)T_XT

UPC Connection Parameters

Booting the RPM-PR

RPM-PR Bootflash Precautions

Solving the RPM-PR Bandwidth Issue When Adding a 12th VISM Card

Limitations and Restrictions

CWM Recognition of RPM-PR and MGX-RPM-128M/B Back Cards

RPM/B and RPM-PR Front Card Resets on the Back Card Removal

RPM/B and RPM-PR Back Ethernet Card Support

MGX-RPM-128M/B Ethernet Back Card Support

RPM/B and RPM-PR Limitations and Restrictions (MGX Release 1.3.10)

Resolved Caveats—Release 12.3(7)T3 for MGX 1.3.10

Resolved Caveats—Release 12.3(2)T6 for MGX 1.3.00

Resolved Caveats—Release 12.3(2)T5 for MGX 1.3.00

Open Caveats—Release 12.3(7)T3 for MGX 1.3.10

Compatibility Notes

RPM Boot File and Firmware File Names and Sizes

RPM Compatibility Matrix

MGX RPM/B and RPM-PR Hardware

Special Installation and Upgrade Requirements

Cisco IOS Release Compatibility Information

About the Cisco IOS 12.2(15)T5 Release

About the Cisco IOS 12.2(11)T1 Release

About the Cisco IOS 12.2(8)T4 Release

About the Cisco IOS 12.2(8)T1 Release

About the Cisco IOS 12.2(4)T3 Release

About the Cisco IOS 12.2(4)T1 Release

About the Cisco IOS 12.2(4)T Release

About the Cisco IOS 12.2(2)T2 and 12.2(2)T3 Release

About the Cisco IOS 12.1(5.3)T_XT Release

Problems Fixed with Cisco IOS 12.1(5.3)T_XT

Upgrading from an MGX-RPM-128M/B Card to an RPM-PR Card

Upgrade Procedures for RPM Cards in MGX Release 1 (PXM1) Switches

Upgrading RPM Boot Software

Upgrading RPM Runtime Software

Upgrade Procedure for Boot Software and Runtime Software for Non-Redundant Cards

Upgrading RPM Boot Software and Runtime Software for 1:N Redundancy

Using XModem to Download Flash to RPM Cards

About These Release Notes (MGX 5.0.10)

New Features

Features Introduced with MGX 8800 Release 4.0.10

Previously Released Features

RPM Image Directory Change From E:RPM to C:/FW

Automatic Cell Bus Clocking

New Fields Added to dspcd

Using the switchredcd Command with RPM-PR Cards to Switch from Active to Standby Card

VISM-PR to RPM-PR Connectivity

Configuring the Cell Bus Clock (CBC) Rate

LDP on RPM-PR in MGX 8850 and MGX 8950

Multi-LVC on RPM in MGX 8850 and MGX 8950 Release 2.1.76 Running Cisco IOS Release 12.2(8)T1

Multiprotocol Label Switching (MPLS) over ATM using VC Merge in MGX 8850 and MGX 8950 Release 2.1.76 Running Cisco IOS Release 12.2(8)T

Bypass Feature for RPM in Cisco IOS Release 12.2(4)T

Features Not Supported in This Release

RPM Redundancy Support

SNMP MIB

Notes and Cautions

UPC Connection Parameters

Booting the RPM-PR

RPM-PR Bootflash Precautions

CLI Modifications in MGX Release 5.0.10

Limitations and Restrictions

CWM Recognition of RPM-PR Back Card

RPM Front Card Resets on the Back Card Removal

RPM-PR Back Ethernet Card Support

RPM-PR Limitations and Restrictions (PXM45 and PXM1E)

Resolved Caveats—Release 12.3(7)T3 for MGX 5.0.10

Resolved Caveats—Release 12.3(2)T6 for MGX 5.0.00

Resolved Caveats—Release 12.3(2)T5 for MGX 5.0.00

Open Caveats—Release 12.3(7)T3 for MGX 5.0.10

Compatibility Notes

RPM Boot File and Firmware File Names and Sizes

RPM Compatibility Matrix

MGX RPM-PR Hardware

Special Installation and Upgrade Requirements

Cisco IOS Release Compatibility Information

About the Cisco IOS 12.2(11)T1 Release

About the Cisco IOS 12.2(8)T4 Release

About the Cisco IOS 12.2(8)T1 Release

About the Cisco IOS 12.2(4)T3 Release

About the Cisco IOS 12.2(4)T1 Release

About the Cisco IOS 12.2(4)T Release

Upgrade Procedures for RPM-PR Cards in MGX 8000 Release 2.1 and Release 3 (PXM45 and PXM1E) Switches

Upgrading RPM Boot Software

Upgrading RPM Runtime Software

Upgrade Procedure for Boot Software and Runtime Software for Non-Redundant Cards

Upgrade Procedure for RPM-PR cards in MGX8000 Release 4 (PXM45 and PXM1E) Switches

Using XModem to Download Flash to RPM Cards

Related Documentation

Obtaining Documentation

Documentation Feedback

Obtaining Technical Assistance

Cisco.com

Technical Assistance Center

Cisco TAC Web Site

Cisco TAC Escalation Center

Release Notes for Cisco MGX Route Processor Module (RPM-PR) IOS Release 12.3(7)T3 for MGX Releases 1.3.10 and 5.0.10

---

These release notes are part OL-6309-01 Rev. A0, August 18, 2004

Contents

About These Release Notes (IOS Release 12.3(7)T3 for MGX Release 1.3.10)

Note that for MGX Release 1.3.10, the user documentation (command reference, overview, and installation and configuration guides) were not updated. Use the Release 1.1.3 and 1.2.10 documents in addition to this release note.

Product documentation for MGX 8850 is available at the following URL:
http://www.cisco.com/univercd/cc/td/doc/product/wanbu/mgx8850/12/index.htm

Product documentation for MGX 8250 is available at the following URL: http://www.cisco.com/univercd/cc/td/doc/product/wanbu/mgx8250/12/index.htm

Product documentation for MGX 8230 is available at the following URL:
http://www.cisco.com/univercd/cc/td/doc/product/wanbu/mgx8230/12/index.htm

Product documentation for VISM 3.0(0) is available at the following URLs:
http://www.cisco.com/univercd/cc/td/doc/product/wanbu/mgx8850/

Product documentation for RPM-PR 2.1 and these release notes are available at the following URLs:
http://www.cisco.com/univercd/cc/td/doc/product/wanbu/mgx8850/12/rpm/index.htm
http://www.cisco.com/univercd/cc/td/doc/product/wanbu/mgx8250/12/rpm/index.htm
http://www.cisco.com/univercd/cc/td/doc/product/wanbu/mgx8230/12/rpm/index.htm

If you are reading Cisco product documentation on the World Wide Web, you can submit comments electronically. Click Feedback in the toolbar, select Documentation, and click Enter the feedback form. After you complete the form, click Submit to send it to Cisco. We appreciate your comments.

New Features

No new features are introduced with this release.

Features Introduced with MGX 8800 Release 1.2.21

RPM in MGX 8800 Release 1.2.21 supports all new and existing features introduced in the Release 1.2.x baseline. There were four new features introduced for RPM implementations using IOS Release 12.2(15)T5:

•MPLS CoS Transparency

This feature allows the service provider to set the MPLS experimental field instead of overwriting the value in the customer's IP precedence field. The IP header remains available for the customer's use; the IP packet's CoS is not changed as the packet travels through the multiprotocol label switching (MPLS) network.

Configuration information can be found at http://www.cisco.com/univercd/cc/td/doc/product/software/ios122/122newft/122t/122t13/ftdtmode.htm

•cRTP with MQC

The MQC acronym refers to the Modular Quality of Service (QoS) Command-Line Interface (CLI). RPM in MGX 8800 Release 1.2.21 supports using the MQC to configure the Compressed Real-Time Protocol (CRTP) header.

Below are the CLI commands introduced to support this feature:

–ip rtp header-compression - enables RTP header compression for a particular interface.

–no ip rtp header-compression - disables RTP header compression for a particular interface.

–clear ip rtp header-compression <interface> - resets all statistics for the interface to 0.

–show ip rtp header-compression <interface> [detail] - shows all statistics for an interface.

–show policy-map int sw1.x -shows the number of packets which are compressed because of match in policy map.

Configuration information can be found at http://www.cisco.com/univercd/cc/td/doc/product/software/ios122/122newft/122t/122t13/fthdrcmp.htm

•LSC Redundancy

Configuration information can be found at
http://www.cisco.com/univercd/cc/td/doc/product/wanbu/8850px1e/rel4/scg/rpm.htm

•MVPN

The frame-based Multicast VPN (MVPN) feature enables the RPM in MGX 8800 Release 1.2.21 to pass frame-based multicast traffic to VPNs across the ATM core.

Configuration information can be found at:

http://www.cisco.com/univercd/cc/td/doc/product/software/ios122s/122snwft/release/122s14/fs_mvpn.htm

Previously Released Features

Automatic Cell Bus Clocking

To implement automatic cell bus clocking, a new -autoClkMode option has been added to the xcnfcbclk command. The default is disabled for backward compatibility. When the feature is enabled by entering the xcnfcbclk -autoClkMode enable command, the PXM scans the whole shelf to see whether there are any two RPMs residing on the same cell bus and changes that cell bus to be running at 42MHz clock rate. The clock rate for the rest of the cell buses are not changed. The active PXM updates the disk DB and sends the update to the standby PXM.

When the feature is enabled, the user will not be able to configure the cell bus clock rate manually for any of the cell buses. When disabled with xcnfcbclk -autoClkMode disable command, the PXM will not change the clock rate for any of the cell bus, but will still update the disk DB and send the update to the standby PXM.

---

Note The command to enable or disable the feature is on a per shelf basis.

---

---

Note The clock will be automatically changed to 21 MHz if one of the two RPM-PRs residing on the same cell bus is removed from the shelf.

---

---

Note After disabling the automatic cell bus clocking, you can manually configure the cell bus clock.

---

The output of dspcbclk command changes to reflect this new feature. A new column is added to show whether the feature is enabled or disabled on the cell buses.When the feature is enabled and an RPM card is inserted, the PXM checks whether the card that resides next to it on the same cell bus is also an RPM card. If both cards are RPM cards, and neither of them is in failed, reserved, unknown, self-test-fail, or no-card state, the cell bus clock rate is automatically set to 42MHz.

Conversely, when the feature is enabled, and an RPM card with a cell bus clock rate of 42MHz is removed or fails, the PXM sets the cell bus to 21MHz, as shown in the following example.

mgx574.1.7.PXM.a > dspcbclk

     CellBus    Rate (MHz)     Slot     AutoClkMode

    --------------------------------------------------

       CB1         21           1, 2       disable

       CB2         21           3, 4       disable

       CB3         21           5, 6       disable

       CB4         21        17 - 22       disable

       CB5         21          9, 10       disable

       CB6         21         11, 12       disable

       CB7         21         13, 14       disable

       CB8         21        25 - 30       disable

mgx574.1.7.PXM.a > cnfcbclk 1 42

WARNING: Certain Service Modules will not operate at the clock rate you specified.

         Please check the Service Modules in the slots where the Cell Bus clock rate is 

effected by this command.

mgx574.1.7.PXM.a > cnfcbclk 5 42

WARNING: Certain Service Modules will not operate at the clock rate you specified.

         Please check the Service Modules in the slots where the Cell Bus clock rate is 

effected by this command.

mgx574.1.7.PXM.a > dspcbclk

     CellBus    Rate (MHz)     Slot     AutoClkMode

    --------------------------------------------------

       CB1         42           1, 2       disable

       CB2         21           3, 4       disable

       CB3         21           5, 6       disable

       CB4         21        17 - 22       disable

       CB5         42          9, 10       disable

       CB6         21         11, 12       disable

       CB7         21         13, 14       disable

       CB8         21        25 - 30       disable

mgx574.1.7.PXM.a > xcnfcbclk

Not enough arguments (?)

xcnfcbclk "-cb <cellBus> -rate <clockRate> -autoClkMode <autoClkEnable>"

    -cb <cellBus>, where cellBus is a string CB1..CB8

    -rate <clockRate>, where clockRate is 21 or 42 (MHz)

    -autoClkMode <autoClkEnable>, where autoClkEnable is enable or disable

mgx574.1.7.PXM.a > xcnfcbclk -autoClkMode enable

mgx574.1.7.PXM.a > dspcbclk

     CellBus    Rate (MHz)     Slot     AutoClkMode

    --------------------------------------------------

       CB1         42           1, 2       enable 

       CB2         21           3, 4       enable 

       CB3         21           5, 6       enable 

       CB4         21        17 - 22       enable 

       CB5         21          9, 10       enable 

       CB6         42         11, 12       enable 

       CB7         21         13, 14       enable 

       CB8         21        25 - 30       enable

VISM-PR to RPM-PR Connectivity

VISM Release 3.0 introduced the new VISM-PR front cards for Cisco MGX Release 1.2.21 and Cisco IOS Release 12.2(15)T. The new VISM-PR-8E1 and VISM-PR-8T1 cards work in the MGX 8230, MGX 8250, and MGX 8850 PXM1-based switches, in combination with the PXM1 Processor Module card. The VISM-PR card supports 144 channels when used with the G.723.1 codec, whereas the current VISM card supports 64 channels with the G.723.1 codec.

Setting connections between a VISM-PR card and a RPM-PR card in your MGX 8230, MGX 8250, or MGX 8850 PXM1-based switch chassis requires that you use the new VBR (NRT)3 connection type.

For more information, refer to the Cisco VISM Installation and Configuration Guide.

Configuring the Cell Bus Clock (CBC) Rate

As of Cisco MGX Release 1.2.10 and Cisco IOS Release 12.2(8)T4), when two RPM-PR cards are on the same cell bus, that is, they occupy adjacent slots (for example, slots 1 and 2 or slots 3 and 4), the cell bus clock (CBC) rate should be manually set to 42MHz. Correspondingly, if there is only one RPM on the cell bus, the clock should be at the default value of 21 MHz.

If, for any reason, one of the adjacent RPM-PRs goes to Failed or Empty state, the CBC for that cell bus must be reconfigured for the Traffic Shaping to work correctly on the active RPM. On MGX 1 switches with Release 1.2.10, the 42MHz to 21 MHz change must be explicitly performed using the cnfcbclk command. Use the dspcbclk command from the PXM1 to confirm the cell bus clock rate.

The following screen output displays the use of the cnfcbclk and dspcbclk commands used to change the clock on cell bus 1 (for slots 1 and 2) from 21 MHz to 42 MHz and confirm the change.

PXM> dspcbclk

     CellBus    Rate (MHz)    Slot

    -------------------------------

       CB1         21        1, 2

       CB2         21        3, 4

       CB3         21        5, 6

       CB4         21        17 - 22

       CB5         21        9, 10

       CB6         21        11, 12

       CB7         21        13, 14

       CB8         21        25 - 30

PXM> cnfcbclk CB1 42

WARNING: Certain Service Modules will not operate at the clock rate you specified.

         Please check the Service Modules in the slots where the Cell Bus clock rate is 
effected by this command

mgx3.1.7.PXM.a > dspcbclk

     CellBus    Rate (MHz)    Slot

    -------------------------------

       CB1         42        1, 2

       CB2         21        3, 4

       CB3         21        5, 6

       CB4         21        17 - 22

       CB5         21        9, 10

       CB6         21        11, 12

       CB7         21        13, 14

       CB8         21        25 - 30

RPM makes use of idle cells for Traffic Shaping and Scheduling. If there are two RPMs in adjacent slots on the same cell bus and one of the RPMs is put into a Failed state by the PXM, while that card is actually alive, then the "Failed" RPM must stop sending idle cells to avoid impacting the Traffic Shaping on the adjacent functional RPM. The command that implements the RPM support for this feature is rpm-auto-cbclk-change.

rpm-auto-cbclk-change enables the RPM to stop sending idle cells in the event of being put into a "FAILED" state by the PXM and thus prevent an impact on the Traffic Shaping on an adjacent functional RPM.

no rpm-auto-cbclk-change which disables the feature to stop sending of idle cells if the RPM is put into a FAILED state. This command may be used if Traffic Shaping is not required.

The following screen output displays an example of the rpm-auto-cbclk-change command.

RPM-11#config  terminal

  Enter configuration commands, one per line.  End with CNTL/Z.

  RPM-11(config)#int sw1

  RPM-11(config-if)#rpm-auto-cbclk-change

  RPM-11(config-if)#end

  RPM-11#write mem

  Building configuration...

  [OK]

  RPM-11#show run int sw1

Building configuration...

Current configuration :142 bytes

!

interface Switch1

 no ip address

 no atm ilmi-keepalive

 rpm-auto-cbclk-change

 switch autoSynch off

end

! rpm_tag_id Apr 04 2002 02:49:04

If Traffic Shaping is not a requirement, enter the no rpm-cbclk-change command, either manually or during card configuration. The following screen output displays an example of the no rpm-auto-cbclk-change command.

RPM-11#config  terminal

  Enter configuration commands, one per line.  End with CNTL/Z.

  RPM-11(config)#int sw1

  RPM-11(config-if)#no rpm-auto-cbclk-change

  RPM-11(config-if)#end

  RPM-11#write mem

  Building configuration...

  [OK]

  RPM-11#show run int sw1

Building configuration...

Current configuration :145 bytes

!

interface Switch1

 no ip address

 no atm ilmi-keepalive

 no rpm-auto-cbclk-change

 switch autoSynch off

end

! rpm_tag_id Apr 04 2002 02:49:57

---

Note By default on the RPM this feature is enabled.

---

LDP on RPM Running MGX Release 1.2.02 and Cisco IOS Release 12.2(8)T1

The MPLS label distribution protocol (LDP), as standardized by the Internet Engineering Task Force (IETF) and as enabled by Cisco IOS software, allows the construction of highly scalable and flexible IP Virtual Private Networks (VPNs) that support multiple levels of services.

LDP provides a standard methodology for hop-by-hop or dynamic label distribution in an MPLS network by assigning labels to routes that have been chosen by the underlying Interior Gateway Protocol (IGP) routing protocols. The resulting labeled paths, called label switch paths (LSPs), forward label traffic across an MPLS backbone to particular destinations. These capabilities enable service providers to implement Cisco's MPLS-based IP VPNs and IP+ATM services across multivendor MPLS networks.

From a historical and functional standpoint, LDP is a superset of Cisco's pre-standard Tag Distribution Protocol (TDP), which also supports MPLS forwarding along normally routed paths. For those features that LDP and TDP share in common, the pattern of protocol exchanges between network routing platforms is identical. The differences between LDP and TDP for those features supported by both protocols are largely embedded in their respective implementation details, such as the encoding of protocol messages, for example.

This software release of LDP provides the means for transitioning an existing network from a TDP operating environment to an LDP operating environment. Thus, you can run LDP and TDP simultaneously on any given router platform. The routing protocol that you select can be configured on a per-interface basis for directly-connected neighbors and on a per-session basis for non-directly-connected (targeted) neighbors. In addition, a label switch path (LSP) across an MPLS network can be supported by LDP on some hops and by TDP on other hops.

MPLS LDP offers the following features:

•IETF Standards-based label distribution protocol

•Multi-vendor interoperability

•TDP to LDP migration and interoperability

Multi-LVC on RPM Running MGX Release 1.2.02 and Cisco IOS Release 12.2(8)T1

This feature enables support for initiation of multiple label switched paths (LSPs) per destination on the RPM. Different label switched paths are established for different class of services. This feature enables interface level queueing rather than per-vc level on the RPM based on MPLS class of service policy. With Multi-LVC support, customers can deploy IP VPN services with Class of Service SLAs.

Bypass Feature for RPM in Cisco IOS Release 12.2(4)T

---

Note Information about the bypass feature and the IOS commands used to support it was not available at the time of the printing of the RPM documents; therefore, it is included in the these release notes.

---

RPM cards have a maximum storage of 128 KB for the NVRAM. This size limitation creates a problem for customers with large configurations, who find it impossible to store the complete configuration in the NVRAM, even with compression enabled.

In order to support storage of large configuration files, a new bypass feature is now available in the 12.2(4)T IOS Release. With the bypass feature enabled, the enhanced "write memory" is used to bypass the NVRAM and save the configuration on:

•For MGX Release 1, the file auto_config_slot## located in the C:/RPM directory on the PXM1.

Where ## represents the zero-padded slot number in which the RPM card is seated in the MGX chassis.

To enable the bypass feature, issue the command rpmnvbypass from the IOS run time image—not in the IOS boot image.

To disable the bypass feature, enter the command no rpmnvbypass.

To verify that the bypass feature is either enabled or disabled, enter the show running-configuration command. If the bypass feature is enabled, rpmnvbypass is seen on the display. If it is not seen, the feature is not enabled.

---

Note Because the bypass feature bypasses NVRAM, it is not necessary to compress the configuration file using the command service compress-config.

---

Table 1 contains cautions important to the successful usage of the bypass feature.

Table 1 Boot Cautions 

Caution	Why is This Important?
When using the bypass feature, you can load the run time IOS image from the PXM hard-drive or from the boot flash.	In the case of an RPM module, the IOS image can be loaded in 3 ways: 1. From the PXM hard-drive. 2. From the boot flash. 3. From the network (for example, via TFTP) from the RPM backcard (Ethernet or Fast Ethernet). When the bypass feature is enabled, the boot config statement: c:auto_config_slot## is automatically generated. The NVRAM configuration is cleared upon entering a write memory command. In order to load from the network, the RPM must have an IP address for its backcard. This information is part of the NVRAM configuration, which was just cleared by enabling the bypass feature. Hence, it is not possible to load the IOS image from the network upon a reload of the RPM after the rpmnvbypass and write memory commands have been executed.
Do not execute the command no boot config. Doing so may prevent the bypass feature from working properly.	When the bypass feature is enabled, the boot config statement: c:auto_config_slot## is automatically generated, and the NVRAM configuration is cleared. Any writes are subsequently directed to the boot config file. This is essential, as a write memory command expects the boot config statement to be present. If the boot config statement is not present, entering the write memory command would write the configuration into the NVRAM, which is not desirable when the objective is to save a complete configuration when the configuration is large and requires more space.
If the command write memory is issued with the bypass feature enabled, and is consequently followed by an RPM card reset, previous versions of the boot image will trigger the RPM card to go into boot mode (unable to load run-time IOS).	For safety purposes, the location of the system image is stored in a special area (called the ROMMON area) in the NVRAM. The ROMMON is always intact. The 12.2(4)T boot image accesses and reads ROMMON in order to load the IOS image. Boot images prior to 12.2(4)T do not read the ROMMON area. Generally, the IOS boot and run-time images are of the same versions. However, if you change the boot image to one prior to 12.2(4)T, on a reload, the boot image would see that the NVRAM configuration is empty, which is normal when the bypass feature is enabled. But because boot images prior to 12.2(4)T cannot access the ROMMON area, it cannot read the location of the IOS image. Unable to see the IOS image, it instead loads itself.

Example 1 through Example 5 illustrate how the bypass feature is enabled and disabled, and how to validate each of these actions from the configuration display.

Example 1 Running configuration without the bypass feature enabled

rpm_slot02#show running-config

Building configuration...

Current configuration : 470 bytes

!

version 12.2

service timestamps debug uptime

service timestamps log uptime

no service password-encryption

!

hostname rpm_slot02

!

boot system c:rpm-js-mz.<new_rel>

enable password cisco

!

ip subnet-zero

!

!

!

!

interface Switch1

 no ip address

 no atm ilmi-keepalive

 switch autoSynch off

!

ip classless

no ip http server

ip pim bidir-enable

!

!

snmp-server community public RO

snmp-server community private RW

!

!

line con 0

line aux 0

line vty 0 4

 no login

!

end

Example 2 Enable the bypass feature (rpmnvbypass)

rpm_slot02#

rpm_slot02#configure terminal

Enter configuration commands, one per line.  End with CNTL/Z.

rpm_slot02(config)#rpmnvbypass

The "boot config" statement has been (re)added to your

running configuration. Do not remove it else risk not

using the nvbypass feature

rpm_slot02(config)#end

rpm_slot02#

Example 3 Running configuration with bypass feature enabled (note rpmnvbypass at end of output)

rpm_slot02#show running-config

Building configuration...

Current configuration : 515 bytes

!

version 12.2

service timestamps debug uptime

service timestamps log uptime

no service password-encryption

!

hostname rpm_slot02

!

boot system c:rpm-js-mz.<new_rel>

boot config c:auto_config_slot02    <==== Line added as per output above

enable password cisco

!

ip subnet-zero

!

!

!

interface Switch1

 no ip address

 no atm ilmi-keepalive

 switch autoSynch off

!

ip classless

no ip http server

ip pim bidir-enable

!

!

snmp-server community public RO

snmp-server community private RW

!

!

line con 0

line aux 0

line vty 0 4

 no login

!

rpmnvbypass

end

Example 4 Disable the bypass feature (no rpmnvbypass)

rpm_slot02#configure terminal

Enter configuration commands, one per line.  End with CNTL/Z.

rpm_slot02(config)#no rpmnvbypass

rpm_slot02(config)#end

rpm_slot02#

Example 5 Running configuration after the bypass feature is disabled

rpm_slot02#show running-config

Building configuration...

Current configuration : 503 bytes

!

version 12.2

service timestamps debug uptime

service timestamps log uptime

no service password-encryption

!

hostname rpm_slot02

!

boot system c:rpm-js-mz.<new_rel>

boot config c:auto_config_slot02

enable password cisco

!

ip subnet-zero

!

!

!

!

interface Switch1

 no ip address

 no atm ilmi-keepalive

 switch autoSynch off

!

ip classless

no ip http server

ip pim bidir-enable

!

!

snmp-server community public RO

snmp-server community private RW

!

!

line con 0

line aux 0

line vty 0 4

 no login

!

end

rpm_slot02#

Features Not Supported in This Release

The following features are not supported on RPM:

•MPLS inter AS

•MPLS TE

RPM Redundancy Support

RPM 1:N redundancy is used to switch configuration and traffic from one RPM card to another. The main benefits are:

•Route processing continues even if an RPM fails and there is no operator or direct access to swap the failed card or fix the problem.

•An RPM card with hardware problems can be fixed while the redundant standby card takes over its functionality.

•Software upgrades are easier and can be done with less downtime.

SNMP MIB

SNMP MGX Release 1 MIB are provided with the delivery of this release. The MIB is in standard ASN.1 format and is located in the same directory within the release bundle on CCO. These files may be compiled with most standards-based MIB compilers. The tar file for MIB contains the file release notes that contains the MIB release notes.

For changes in this MIB from the previous release, please refer to the MIB release notes.

---

Note The old_mib_Format is discontinued as of this release.

---

Notes and Cautions

The following notes and cautions should be reviewed before using this release.

Special Upgrade Procedure for Cisco IOS Release 12.1(5.3)T_XT

Use the following procedure when upgrading from your current RPM/IOS runtime image 12.1(5.3)T_XT and MGX version for MGX Release 1.2.02 and 1.2.10:

---

Step 1 RPM IOS boot and runtime images should be upgraded before MGX images are upgraded. Please follow the RPM/IOS image upgrade procedure as specified in the "Upgrade Procedures for RPM Cards in MGX Release 1 (PXM-1) Switches" as described in later sections of this Release Notes Document.

Step 2 MGX software should be upgraded next as illustrated in the following steps.

a. install <image-name>

b. newrev <image-name>

c. commit <image-name>

---

For more detail on the MGX upgrade procedures, refer to the Release Notes for Cisco MGX 8230, MGX 8250, and MGX 8850 (PXM1), Software Version 1.2.10.

UPC Connection Parameters

In Release 1.1.40 and higher, the default PCR is 50 cps, and the default for policing is "enabled." These settings are insufficient for running RPM ISIS protocol over the connection, and with such settings, the ISIS protocol will fail. The PCR value needs to be increased, depending upon the number of interfaces configured for ISIS on the RPM.

Depending upon your connection type, you can use the following CLIs to modify the PCR parameter.

•cnfupccbr

•cnfupcvbr

•cnfupcabr

•cnfupcubr

Booting the RPM-PR

Refer to Chapter 5, "Configuring the MGX RPM" in the Cisco MGX Route Processor Module Installation and Configuration Guide, Release 1.1, (DOC-7812278=) and for complete details on configuring the RPM-PR cards. (See the "Obtaining Documentation" section for information on how to order a printed copy of this manual or locate the manual online.) A summary of the booting and upgrading procedures is presented here for your convenience.

When the RPM-PR is booted, the boot image must be the first file in the bootflash. If the bootflash does not have a valid boot image as a first file, the card may not be able to boot.

You can reboot the RPM-PR from the PXM by entering the command resetcd <card_number> from the switch CLI, where card_number is the slot number of the RPM-PR that is being rebooted.

---

Note Omitting the card number resets the entire system.

---

Also, you can reboot the RPM-PR from the RPM-PR using the RPM-PR console port and entering the reload command.

Each time you turn on power to the RPM-PR, by inserting the RPM-PR into the MGX 8850, it goes through the following boot sequence:

1. The RPM-PR runs diagnostics on the CPU, memory, and interfaces.

2. The system boot software, which is the boot image, executes and searches for a valid Cisco IOS image, which is the RPM-PR runtime software.

The source of the Cisco IOS image is determined by the configuration register setting. To verify this setting, you can enter either the show version or show bootvar command. (Refer to the "Viewing the Hardware Configuration" section of the Cisco MGX Route Processor Module Installation and Configuration Guide, Release 1.1 (DOC-7812278=).

a. If the configuration register is set to the factory-default setting of 0x01, RPM-PR will come up and stay in boot mode.

b. If the configuration register is 0x2, the RPM-PR will look for the runtime image either in bootflash or on the PXM1 C:/RPM drive.

3. The search for runtime image is determined by which boot system command is entered.

a. Entering the boot system c:<runtime_image_name> command will result in a search for a runtime image in the C:/RPM directory on the PXM1 hard disk.

b. Entering the boot system bootflash:<runtime_image_name> will result in a search for a run time image in the bootflash.

c. If the boot system bootflash:<runtime_image_name> is not entered, it will result in loading of the first available IOS image from C:/RPM, if one such image is present.

4. If the runtime software is not found after three attempts, the RPM-PR reverts to the boot mode.

5. If a valid Cisco IOS image is found, then the RPM-PR searches for a valid configuration, which can reside in NVRAM or as a configuration file either on the PXM hard disk C:/RPM drive or in bootflash.

If you want to load from a specific configuration file, you should enter either the boot config bootflash:<config_file> command or the boot config c:<config_file> command.

6. For normal RPM-PR operation, there must be a valid Cisco IOS image on the PXM-1 C:/RPM drive or in bootflash, and a configuration in NVRAM or configuration file in bootflash or on the PXM disk.

The first time you boot the RPM-PR, configure the RPM-PR interfaces and save the configuration to a file in NVRAM. Then follow the procedure described in "Initializing the RPM-PR Card." For information on the Cisco IOS instructions, refer to Appendix C, "IOS and Configuration Basics"of the Cisco MGX Route Processor Module Installation and Configuration Guide, Release 2.1 (DOC-7812510=)

RPM-PR Bootflash Precautions

The RPM-PR bootflash is used to store boot image, configuration and "run time" files. The Flash stores and accesses data sequentially, and the RPM-PR boot image must be the first file stored to successfully boot the card. Erasing the boot image or moving it from the first position on the Flash will cause the card to not boot.

The RPM boot image, which comes loaded on the Flash, will work for all RPM IOS images. Therefore, there is no reason to ever delete or move the factory installed boot image.

---

Caution Erasing or moving the boot image can cause RPM-PR boot failure. When this happens, the RPM card must be returned to Cisco and reflashed.

---

In order to avoid this unnecessary failure, requiring card servicing, you should

•Never erase the boot file from the RPM Flash

•Never change the position of the boot file on the RPM Flash

•Use care when "squeezing" the Flash to clean it up.

As long as the boot file remains intact in the first position on the flash, the RPM will successfully boot.

Solving the RPM-PR Bandwidth Issue When Adding a 12th VISM Card

Adding more than 11 VISM cards to an MGX chassis with RPM-PR cards requires enabling the expanded memory option on PXM45/B. The command to enable this option is cnfndparms (option 4). This expanded memory option does not impact chassis performance while allowing more connections.

Limitations and Restrictions

CWM Recognition of RPM-PR and MGX-RPM-128M/B Back Cards

CWM does not distinguish between the Ethernet back card versions installed with the MGX-RPM-128M/B or RPM-PR. There is no functionality difference.

RPM/B and RPM-PR Front Card Resets on the Back Card Removal

The RPM front card may reset on an MGX 8250 and MGX8850 switches with PXM1 as controller card when the ethernet back card is removed or inserted.

This reset problem can be easily avoided if "shut" interface is executed before the removal of the back card.

RPM/B and RPM-PR Back Ethernet Card Support

For Ethernet connectivity with the RPM-PR, the model "/B" four-port Ethernet back card is required (order number: MGX-RJ45-4E/B).

MGX-RPM-128M/B Ethernet Back Card Support

The model "/B" four-port Ethernet back card can be used with the MGX-RPM-128M/B module only in combination with IOS 12.2(2)T2 or higher. The model "/B" back card will not work on the MGX-RPM-128M/B with earlier versions of the IOS.

The order number is order number: MGX-RJ45-4E/B.

Older back cards can be used with any version of the IOS.

4-port Ethernet back card used with MGX-RPM-128M/B	Required IOS
model "/B" back card	12.2(2)T2
earlier back card models	Min. IOS for MGX-RPM-128M/B on MGX 8250 is 12.0(7)T

RPM/B and RPM-PR Limitations and Restrictions (MGX Release 1.3.10)

The RPM/B and RPM-PR limitations and restrictions that apply to this release are as follows:

•The MGX-RPM-128M/B is a NPE-150 based router card capable of sustaining 150,000 pps. The RPM-PR is an NPE-400 based router capable of sustaining over 350,000 pps. The RPM-PR will only operate with IOS 12.1(5.3)T_XT or later. For the following section "RPM" refers to both the MGX-RPM-128M/B and the RPM-PR, (unless specifically called out). Some software versions and limitations are not applicable to the RPM-PR because it does not support IOS versions before 12.1(5.3)T_XT.

•With MGX-RPM-128M/B versions earlier than 12.0.7T1, some limitations in Inter-Process Communication when the MGX-RPM-128M/B is at high loads can cause the PXM to declare that the MGX-RPM-128M/B has Failed. To avoid this with MGX-RPM-128M/B, software releases earlier than 12.0.7T1, throughput is limited to 62,000 pps, and it is recommended that MPLS configurations are limited to 100 interfaces. With RPM software releases from 12.0.7T1, those limitations are removed. In a separate limitation, the number of directly connected OSPF networks supported by an RPM is currently limited to 27. This means that any or all of the subinterfaces supported by the RPM can run OSPF, but the number of distinct OSPF networks supported is limited to 27. (A work around is available and is discussed below.) The limit of 27 arises because of the overheads of supporting separate link-state databases for separate networks.

•In an application where the RPM is a Provider Edge Router in an MPLS Virtual Private Network service, a much better solution in any case is to use a distance-vector routing protocol between the customer routers and the RPM. A distance-vector routing protocol provides exactly the information required for this application: reachability information, and not link-state information. The distance-vector routing protocols supported by the RPM are BGP, RIP v1 and RIP v2, as well as static routing. With RPM software releases from 12.0.7T1, distance-vector routing protocols can be used with as many different networks as subinterfaces.

•Note that if the RPM is acting as a Provider Edge Router in an MPLS Virtual Private Network service, and even if OSPF is running in a customer network, it is not necessary to run OSPF between the customer router and the RPM. If the customer edge devices run Cisco IOS, they can redistribute OSPF routing information into RIP using the IOS commands, redistribute RIP in the OSPF configuration, and redistribute OSPF in the RIP configuration. Similar configurations are possible for BGP. (For more information on re advertisement, see the "Configuring IP Routing Protocol-Independent Features" chapter in the Cisco IOS Release 12.0 Network Protocols Configuration Guide, Part 1). Redistribution is not unique to Cisco CPE, and other vendors' equipment also supports redistribution.

•Whenever there are 2 RPM cards on adjacent slots, driven by the same cell bus clock, the clock rate should be set to 42 MHz for traffic shaping, using the command cnfcbclk. This configuration will be lost if the node rebuilds due to resetsys or a power cycle. The user will have to manually re-configure the cell bus clock rate after the rebuild, using the cnfcbclk command.

•On an MGX 8850 node, when the chassis is loaded with 6 or more RPM-PR cards, and if every card is configured to download the IOS runtime image from the PXM-1 hard disk, occasionally, upon entering a resetsys command or after a power cycle, some of the RPM-PR cards may go into the failed state. To reset the failed RPM-PR cards, enter the resetcd <slot #> command for each failed card.

•A single RPM-PR can only function as either an Edge LSR or as an LSC, but not as both.

•Total of (OC12 minus T3) Mbps intrashelf traffic for Cell bus based modules are supported.

•To configure redundancy, the primary and secondary RPM-PR cards need to be in the Active state and the secondary card should not have any configuration.

•Removing a back card does not cause RPM-PR switchover.

•After establishing redundancy between two RPM-PR cards with the addred command, you must enter the copy run start command on the primary RPM-PR card to save the configuration change.

•If a secondary RPM-PR card is redundant to primary cards x and y, you cannot delete redundancy for only card x.

•If you need to enter the switchredcd and switchcc commands, Cisco Systems recommends that you wait at least 5 seconds after issuing the switchredcd command, and then enter the switchcc command.

•IOS software images on primary and secondary RPM-PR cards do not have to be compatible, but the IOS software on a secondary card should be at the same level as the primary card or higher.

•Whenever the RPM-PR configuration is changed and a user wants to store that configuration, the user must enter the copy run start command on the RPM-PR. If this is not done, the changed configuration will be lost on RPM-PR card reboot or RPM-PR switchover in case of redundancy.

•Even though RPM-PR can have 1999 sub interfaces, the usage of sub interfaces should be planned in such a way that it does not cross a safe limit of 1985. This is because each sub interface takes one IDB (interface descriptor block) and the number of IDBs available in the card is 2000. Further, a user might need some IDBs for the RPM-PR back card and its ports.

•For RPM/B and RPM-PR PVC dax connections, the slave end must be deleted before the master endpoint.

Resolved Caveats—Release 12.3(7)T3 for MGX 1.3.10

Table 2 lists the resolved caveats in Cisco IOS Release 12.3(7)T3 for MGX 1.3.10 as of August 18, 2004.

Table 2 Resolved Caveats in Cisco IOS Release 12.3(7)T3 for MGX 1.3.10 

Caveat Number	Description
CSCee54524	During SCR provisioning (changing the bandwidth) on the ePVCs, the RPM-PR (Route Processor Module) PE may reload.
CSCed41381	LSNT: Input drops on framed PVC i/f cause ldp session flap
CSCed74835	When copying big files from Cisco MGX8850 Route Processor Module (RPM-PR) to MGX8850 hard drive, it silently failed.
CSCea85395	BGP suppressed prefixes not reinstated after condition removed
CSCea62571	%UTIL-3-TREE: Data structure error with trace back logged on LSC

Resolved Caveats—Release 12.3(2)T6 for MGX 1.3.00

Table 3 lists the resolved caveats in Cisco IOS Release 12.3(2)T6 for MGX 1.3.00 as of June 4, 2004.

Table 3 Resolved Caveats in Cisco IOS Release 12.3(2)T6 for MGX 1.3.00 

Caveat Number	Description
CSCec76875	A provider edge (PE) router may not be able to ping other PE routers or a label switch controller (LSC), nor may other platforms be able to ping the PE router.
CSCed82475	A Label Switch Controller (LSC) may reload unexpectedly when there is an Address Error (load or instruction fetch) exception.
CSCed85438	A Fast Ethernet 100BASE-TX port adapter on an RPM-PR card may stop receiving burst traffic packets.
CSCee03384	When an LC-ATM switch subinterface is created and then deleted on an RPM-PR, the index for the current subinterface for the LVC stuck detection and recovery mechanism is changed in such a way that the "LVC stuck" information for an existing LC-ATM is overwritten when a new LC-ATM is added.

Resolved Caveats—Release 12.3(2)T5 for MGX 1.3.00

Table 4 lists the resolved caveats in Cisco IOS Release 12.3(2)T5 for MGX 1.3.00 as of April 14, 2004.

Table 4 Resolved Caveats in Cisco IOS Release 12.3(2)T5 for MGX 1.3.00 

Caveat Number	Description
CSCeb63762	RPM-SAR RAS enhancement - Ability to reset SAR after crash
CSCec14559	PXM spontaneously fails over during 1:N RPM redundancy testing
CSCed26565	RPM-PR-512 reset multiple times due to SAR APU stall
CSCeb22233	RPM crash after FE backcard OIR
CSCeb61872	sw rev varbind not available when trap 60056 sent for RPM
CSCed40082	Duplicate ping atm command in RPM-PR
CSCed46492	Ethernet 2/1 on RPM-PR stays in shutdown state after switchredcd
CSCeb74637	During the RPM upgrade to 12.2(8)MC2d, startup config disappeared.
CSCed05581	The switch connection configuration in switch subinterfaces is lost after reload in RPM-PR.
CSCdz79827	SNMP loop at cipPrecedenceEntry on switch subinterface of rpm
CSCea26869	IPC message header cache below traceback kept coming
CSCea44854	Error Messages from RPM-PR module for SNMP requests are wrong
CSCea74335	On reloading an RPM-PR card using reload command it prompts for saving the config even if the configuration has not been modified.
CSCeb10082	No indication in the log that clear int sw1 was executed via CLI
CSCeb11734	Traceback messages on slot 14 after upgrade
CSCeb41501	traceback on RPM-PR with only VCC VPC switch partition configured
CSCeb65685	ip cef load-sharing algorithm tunnel command lost upon reset of RPM-PR
CSCeb78535	Multiple alignment errors adding switch connection
CSCeb78905	Incorrect configuration generation for RPM commands
CSCeb81588	Set PCI retry timer to non-zero value for RPM-PR platform
CSCed25513	watchdog timeout when malloc fails while resynching

Open Caveats—Release 12.3(7)T3 for MGX 1.3.10

Table 5 lists the open caveats in Cisco IOS Release 12.3(7)T3 for MGX 1.3.10 as of August 18, 2004.

Table 5 Open Caveats in Cisco IOS Release 12.3(7)T3 for MGX 1.3.10 

Caveat Number	Description
CSCec16481	A Cisco device running Internetwork Operating System (IOS) and enabled for the Open Shortest Path First (OSPF) Protocol is vulnerable to a Denial of Service (DoS) attack from a malformed OSPF packet. The OSPF protocol is not enabled by default. The vulnerability is only present in IOS release trains based on 12.0S, 12.2, and 12.3. Releases based on 12.0, 12.1 mainlines and all IOS images prior to 12.0 are not affected. Refer to the Security Advisory for a complete list of affected release trains. Further details and the workarounds to mitigate the effects are explained in the Security Advisory which is available at the following URL: http://www.cisco.com/warp/public/707/cisco-sa-20040818-ospf.shtml.
CSCdv24154	Notify new tag MTU value after the interface mtu change. Symptom: Setting MTU size in xtagatm interface in LSC defaults to 4470 Condition: Setting MTU size in xtagatm interface in LSC defaults to 4470 Workaround: None.
CSCea84387	Two simultaneous policy map displays cause problems Symptoms: A user session may pause indefinitely, causing a Cisco router to become unresponsive. Conditions: This symptom is observed when multiple simultaneous users enter modular QoS CLI (MQC) commands on the same router via separate vty sessions. Workaround: Allow only one user at a time to enter MQC commands.
CSCea85160	snmpwalk on ifDescr should not respond to aal5 mib variable Symptom: snmpwalk on ifDescr responds to aal5 mib variable. Conditions: This symptom is observed when we query for the ifDescr Workaround: None.
CSCec02155	SLT:All Vsi sessions down, xtag to down/down state Symptoms: All Virtual Switch Interface (VSI) sessions go down and XTag interfaces go to the down/down state on a Label Switch Controller (LSC). Conditions: This symptom is observed on a Cisco Route Processor Module (RPM-PR) that acts as an LSC. Workaround: There is no workaround.
CSCec26512	After switchredcd 22 LVCs went to Bwait state Symptom: After switchover, 22 LVCs went to Bwait state. Condition: After switchover, 22 LVCs went to Bwait state. Work-around: None.
CSCec68542	LSNT:SAR not reassemble properly for big pkts cause protocol down Symptom: In Large scale network, under stressfull condition, Segmentation and reassembly (SAR) may not reassemble properly for large packets, causing the protocol to fail. Conditions: This symptom is observed on a Cisco MGX Route Processor Module PRemium (RPM-PR running 12.2(15)T4B) that functions as a Label Edge Router (LER) that is configured with two label switch controllers (LSCs) under the following conditions: •With a script running disable and reenable the multi-virtual-circuit (Multi-VC) mode on the RPM-PR. •There are 1000 soft permanent virtual circuits (SPVCs) that have Virtual Private Network (VPN) routing/forwarding (VRF) enabled. •There are 500 external Border Gateway Protocol (eBGP) sessions. •There is 80-Mbps input traffic and 6-Mbps output IP traffic. Workaround: Reload the RPM-PR.
CSCed83799	Error message ATMPA-3-BADPARTICLE on LSC Symptom: The following error message is noticed on RPM-PR terminal followed by a traceback. *Feb 24 14:36:35.140: %ATMPA-3-BADPARTICLE: Switch1: bad rx particle 0x61DE1540 flags 0x00000000 index 19000 Conditions: This error message will be noticed on RPM-PR terminal when there is considerable amount of incoming and outgoing traffic through the RPM-PR card. Workaround: None.
CSCed85540	LSNT:LSC crash in LSNT Symptom: Cisco MGX8850 Route processor module RPM-PR acting as LSC resets due to "software forced reload". Condition: This problem might seen on the RPM-PR in the following conditions: •acts as LSC in cell based MPLS network. •Frequent multi-vc flaps back to back. Workaround: None.
CSCed94598	RPM over shaping SCR by 0.3 to 0.5% Symptoms: A Route