Table Of Contents
1.1.22 Version Software Release Notes
Cisco WAN MGX 8850 SoftwareFeatures Introduced in Release 1.1.22
Features Introduced in Release 1.1.21
Features Introduced in Release 1.1.12
Features Introduced in Release 1.1.11
Features Introduced in Release 1.1.10
Release 1.1.22 MGX 8850 Hardware
MGX 8220 Hardware not supported on Release 1.1.22 of the MGX 8850
MGX 8220 Hardware that has been superseded on the MGX 8850 by MGX 8850-specific Hardware
MGX 8220 Hardware that will not be supported on the MGX 8850
Features not Supported in this Release:
Major Network Management Features
Problems Fixed in Release 1.1.22
Problems Fixed in Release 1.1.21
Problems Fixed in Release 1.1.20
Problems Fixed in Release 1.1.12
Problems Fixed in Release 1.1.11
Problems Fixed in Release 1.1.10
Problems Fixed in Release 1.1.01
Problems Fixed in Release 1.1.00
Problems Fixed for RPM in 12.0.5T1
Problems Fixed for RPM in 12.0.4T
Problems Fixed for VISM in 1.1.20 release of PXM
Special Installation and Upgrade Requirements
Single PXM Installation Procedure
Installation Procedure For Redundant PXMs:
Service Module Firmware Download Procedure
Service Module Installation/Upgrade and Flashdownload Requirements.
Known Anomalies for Platform Software and Service Module Firmware
RPM Configuration Examples for MPLS-based Virtual Private Networks
One PE - Two CE Configuration - OSPF & IBPG Between PEs & EBGP between PE-CE
One PE - Two CE Configuration - OSPF & IBPG Between PEs & RIP between PE-CE
One PE - Two CE Configuration - OSPF & IBPG Between PEs & STATIC ROUTES between PE-CE
1.1.22 Version Software Release Notes
Cisco WAN MGX 8850 Software
About These Release Notes
Cisco documentation and additional literature are available in a CD-ROM package, which ships with your product. The Documentation CD-ROM, a member of the Cisco Connection Family, is updated monthly. Therefore, it might be more current than printed documentation. To order additional copies of the Documentation CD-ROM, contact your local sales representative or call customer service. The CD-ROM package is available as a single package or as an annual subscription. You can also access Cisco documentation on the World Wide Web at http://www.cisco.com, http://www-china.cisco.com, or http://www-europe.cisco.com.
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.
About the 1.1.22 Release
Release 1.1.22 of the MGX 8850 supports the same network scenarios as Release 1.1.12 and 1.1.21.
1.
Feeder concentration to the BPX 8600 and all other endpoints (no BPX 8600 BNI trunk connections). IGX endpoints are supported in this release using Switch Software 9.2.
The MGX 8850 provides multiservice, high density ATM, Circuit Emulation and Frame Relay feeder concentration to the BPX 8600. The MGX 8850 connects to the BPX 8600 using the feeder trunk protocol over a PXM port. On the BPX 8600 side the feeder connection trunk to the MGX 8850 is supported on the BXM card only. Interoperability support is limited to (a) MGX 8850 to MGX 8850, (b) MGX 8850 to MGX 8220, c) MGX 8850 to BPX 8600 (FR to ATM service interworking) and (d) MGX 8850 to IGX.
2.
Stand-alone capability allows the MGX 8850 to act as an edge concentrator to any vendor ATM network which implies service interoperability with other vendor's equipment. All connections for stand-alone are local switching connections.
Features
This section lists the features supported in each of the maintenance releases that comprise MGX 8850 Release 1.1.22. Except where noted, all features are generally available.
Features Introduced in Release 1.1.22
MGX 8850 Release 1.1.22 is a maintenance release of Release 1.1.21. In addition, it introduces the following new features:
•
•
•
Features Introduced in Release 1.1.21
MGX 8850 Release 1.1.21 includes all the features in Release 1.1.12 and earlier. In addition, it introduces the following new features:
•
•
Support for the FRSM-HS1B dual-personality back-card that can be configured for either V.35 interfaces or X.21 interfaces.
•
–
–
–
•
–
–
–
•
–
–
–
This feature reduces switchover time for the above service modules.
•
–
–
–
–
•
This feature supports running the T3 or E3 lines on the FRSM at sub rates. The interfaced DSU/CSU supported are the Digital Link DL3100 for T3 and the Digital Link DL3100E for E3.
Features Introduced in Release 1.1.12
MGX 8850 Release 1.1.12 was a maintenance release of Release 1.1.11. In addition, it introduced the following new features:
•
–
–
Note
•
•
Features Introduced in Release 1.1.11
MGX 8850 Release 1.1.11 was a maintenance release of Release 1.1.10. In addition, it introduced the following new features:
•
–
–
–
–
–
–
–
–
–
Note
•
The RPM has a custom ASIC on it called the ATMizer. This chip is being replaced by the ATMizer II+, AKA "G10" version of the chip. The RPM-B contains this new chip, along with a 16Mbyte Flash SIMM instead of 4Mbyte. The RPM-B can operate at 21Mhz and 42Mhz Cellbus clock rate with 1.1.11 and 12.0.5T1, while the RPM-A can only operate at 21 Mhz.
•
•
Note
•
•
•
•
•
•
Features Introduced in Release 1.1.10
MGX 8850 Release 1.1.10 provided the following features in addition to the ones provided in Release 1.1.01 and earlier:
•
•
•
Release 1.1.22 MGX 8850 Hardware
MGX 8850 is a 45 Gbps backplane with 1.2 Gbps switching fabric for Release 1.1.22. The same backplane is used with different switching fabric cards (1.2, 45 Gbps) to achieve scalability. MGX 8850 Release 1.1.22 hardware components and their revisions that are supported are as follows:
Support for embedded Cisco IOS router (Router Processor Module - RPM)
•
–
–
–
MGX 8220 Hardware not supported on Release 1.1.22 of the MGX 8850
The following cards are not supported in Release 1.1.22:
•
•
•
•
•
•
MGX 8220 Hardware that has been superseded on the MGX 8850 by MGX 8850-specific Hardware
•
The MGX-SRM-3T3-B front card replaces the original AX-SRM-3T3-A front card and the MGX-BNC-3T3 back card replaces the original AX-BNC-3T3 back card. This change allows the use of slots 9, 10, 25, and 26 for 1:n redundancy and BERT in the MGX 8850 chassis. Both the AX-SRM-3T3-A/AX-BNC-3T3 card set and the MGX-SRM-3T3-B/MGX-BNC-3T3 card set are supported on the MGX 8220.
New card should have enabled use of bulk distribution in slots 9 and 10. 1:N redunancy should have been supported in those slots with the model A card.
•
Superseded by the MGX-SCSCI2-2HSSI/B, which works with the MGX-FRSM-HS2 front card. A V.35 interface is supported on the MGX-FRSM-HS1/B in this release.
•
Superseded by MGX-AUSM-8T1/B and MGX-AUSM-8E1/B
•
Superseded by MGX-AUSM-8T1/B and MGX-AUSM-8E1/B
MGX 8220 Hardware that will not be supported on the MGX 8850
•
•
•
•
Software Platform Features
MGX 8850 provides high speed native ATM interfaces which can be configured as ATM UNI ports or trunks
Support for 1:N and 1:1 Service Module Redundancy, as indicated in the table below:
Support for Bulk Distribution using SRM-3T3B card.
Service module and PXM upgrades
Features not Supported in this Release:
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Note
Major Network Management Features
•
•
•
•
•
•
•
For more details refer to the CWM Release 9.2.07 release notes part number 78-6659-07.
Connection Limits
•
•
•
•
SNMP MIB:
The SNMP MGX 8850 MIB is being provided with the delivery of Release 1.1.22 of the MGX 8850 software on CCO. The MIB is in standard ASN.1 format and is located in the ASCII text files MGX8800Mib.my file which is included in the same directory within CCO. These files may be compiled with most standards-based MIB compilers. For changes in this MIB from release 1.1.21 please refer to the MIB release notes on CCO.
Notes & Cautions
CLI modification and changes:
•
–
–
–
–
–
–
–
–
–
–
–
–
–
Syntax:
cnfifip "Interface IPaddr [NetMask [BroadcastAddr]]"
or cnfifip "Interface Flag"
Interface -- 26/28/37 (26:Ethernet 28:SLIP 37:ATM)
or Ethernet/SLIP/ATM
IP_Addr -- <n>.<n>.<n>.<n> (<n>: integer 0..255)
Net_Mask -- <n>.<n>.<n>.<n> (<n>: integer 0..255)
BroadcastAddr -- <n>.<n>.<n>.<n>
(<n>: integer 0..255)
Flag -- a string "UP" or "DOWN"
Example:
> cnfifip atm 192.9.200.1 255.255.255.128
This configures the ATM interface and brings it UP.
> cnfifip atm up
This will bring up the ATM interface with current information in database.
> cnfifip atm down
This will bring down the ATM interface and preserve the information in the database.
–
Syntax:
delifip Interface
Interface -- 26/28/37 (26:Ethernet 28:SLIP 37:ATM) or Ethernet/SLIP/ATM
Example:
> delifip 37
This will bring down the ATM interface and delete the information in the database
–
Example:
> dspifipInterface Flag IP Address Subnetmask Broadcast Addr--------------- ---- --------------- --------------- ---------------Ethernet/lnPci0 UP 172.29.37.77 255.255.255.0 172.29.37.255SLIP/sl0 DOWN 172.29.36.253 255.255.255.252 (N/A)ATM/atm0 UP 192.9.200.1 255.255.255.128 0.0.0.0
This command shows the current condition of all 3 interfaces. The data shown for the SLIP interface will apply when it is turned UP with, say, "cnfifip slip on".
–
Syntax:
cnfenetgw IPAddr
Example:
> cnfenetgw 172.29.37.1This command will set the default gateway and add the appropriate routes necessary to it.
–
Example:
> dspenetgwEnet Gateway: 172.29.37.1–
Several lines are essential for the network to function:
- boot device : lnPci
(The only Ethernet interface)
- inet on ethernet (e) : 172.29.37.40:ffffff00
(IP address and subnetmask)
- gateway inet (g) : 172.29.37.1
(Default Ethernet gateway)
The PXM will try to correct bad entries when it boots up. This information will be copied to the standby card and if different than the shelf ip address it will up the interface on the standby with the bootChange ip address. The shellconn version of this command only updates the local bootline values and is not copied to the other card.
–
Used to bring up the Ethernet interface when CLI prompt is not there or in backup boot if it's not enabled
The following commands which are related to FRSM-2CT3 line level loopbacks.
•
This loopback can be configured in FRSM-2CT3 using the following commands.
addds3rmtloop <lineno>
xcnfln -ds3 <lineno> -e 3 -lpb 2
•
This loopback can be configured in FRSM-2CT3 using the following commands.
addds3loop <lineno>
xcnfln -ds3 <lineno> -e 3 -lpb 3
DS3 Loopback status will be displayed with following commands:
dspds3ln <lineno>
dspalm -ds3 <lineno>
dspalms -ds3
FEAC codes monitoring and Inband loopbacks for DS3 are not supported in FRSM-2CT3
•
This loopback can be configured in FRSM-2CT3 using the following commands:
cnfbert (from PXM)
addrmtloop <lineno>
xcnfln -ds1 <lineno> -e 3 -lpb 2
•
This loopback can be configured in FRSM-2CT3 using the following commands:
cnfbert (from PXM)
addlnloop <lineno>
xcnfln -ds1 <lineno> -e 3 -lpb 3
DS1 Loopback status will be displayed with following commands:
dspln <lineno>
dspalm -ds1 <lineno>
dspalms -ds1
•
xcnfln -ds1 <lineno> -e 3 -detect 2
•
•
•
–
–
–
Node Related
At most one BERT test can be performed per shelf at any point in time. BERT can only be activated through the CLI.
Do not execute the restoreallcnf command in the middle of the installation process. If you follow the following steps:
Step 1
Step 2
Step 3
Step 4
The dsplns command will display a line as disabled, but you cannot run an addln command. Do not execute the restoreallcnf command until the install and newrev commands have completed.
The correct order for the restore procedure is:
Step 1
Step 2
Step 3
Step 4
(for more information, refer to CSCdm57683)
Addln should be issued before issuing addapsln.
The following line and alarm related commands have been modified to allow slots 8, 16 and 32 as valid arguments if PXM at slot 8 is active:
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Full SRM redundancy requires redundant SRMs. There must be SRMs in BOTH slot 15 and 16 to ensure service module redundancy for the upper shelf AND SRMs in BOTH slot 31 and 32 to ensure service module redundancy for the lower shelf. Lack of the second SRM in either shelf may result in mismatch conditions.
For service module redundancy support, if the active service module is physically removed from the slot then a switchcc would cause the now active service module to be inaccessible. The workaround is to make sure that both the active and standby cards are physically present in their slots. If the active card indeed needs to be removed then at shellconn type: pmmStartScmPolling(slotnumber) after the switchcc.
If you are moving service modules from an existing MGX 8220 platform to the MGX 8850, the MGX 8220 service modules (AX-FRSM-8T1/E1, and AX-CESM-8T1/E1) need to have the boot flash upgraded to MGX 8220 Release 5.0.00 common boot code (1.0.01 version) before they can be plugged in to the MGX 8850 chassis. All MGX-8220 service module versions that use release 4.0.xx of boot code and earlier are not supported in the MGX 8850.
If loading of the correct common boot code image is required then it will have to be performed on an MGX 8220 chassis, and cannot be performed on an MGX 8850 chassis. Please refer to the procedure below, which is also outlined in the Cisco MGX 8850 Installation and Configuration publication on the documentation CD.
Step 1
Step 2
Step 3
tftp <ip address of the MGX 8220 shelf >binput <boot filename> AXIS_SM_1_<slot#>.BOOTInsure that tftp downloaded the appropriate boot code by verifying the flash checksums.
Step 1
cc <slot #>'Step 2
chkflash'If NOT, repeat the process until they are the same. If they are the same, then you can safely remove the card. At this point the service module can be used in the MGX 8850 shelf.
Caution
Whenever an MGX 8850 is added as a feeder to a BPX 8600, SWSW automatically programs a channel with a VPI.VCI of 3.8 for use as the IP Relay channel. IP Relay is used to send IP data between nodes via the network handler; allowing every node in the domain to be directly addressable via IP addressing and CWM workstations to communicate with every node (especially feeders) using TELNET, SNMP and CWM protocols. If the user tries to add a channel with a VPI.VCI of 3.8, the BPX 8600 does not prevent the user channel from being added, but the MGX 8850 rejects it. To delete the added channel on the BPX 8600, and to get IP relay working you need to reset the BXM card.
In addition to clearing all the configuration, clrallcnf clears the network IP addresses. IP addresses and netmasks stay the same (dspifip). However, it's recommended by engineering to reconfigure them using the cnfifip command. Network IP is gone (dspnwip), and must be reconfigured using the cnfifip command. Refer to the entry on cnfifip in the Cisco MGX 8850 Command Reference publication on the documentation CD for syntax.
–
A minimum of two and up to four IP addresses are needed to be configured for MGX 8850 (one or more of the following: ethernet, ATM, SLIP) and the boot IP address. The user should use "bootChange" to set up IP gateway when the PXM card is just installed. The IP default gateway should be on the same subnet as the PXM board. Use the bootChange command to set correct IP address, netmask, and default gateway.
Do not install a Y cable on the UIA CP port for PXMs. If you do both serial ports will be enabled and you will not be able to communicate at all with the shelf through the console ports. If after switchcc standby PXM loses the downlevel port then it is due to a downlevel Beta version of UIA backcard that were shipped during field-trial only. Upgrading the UIA back card to the latest version should fix this problem.
To configure the external clock source, use the interface label 7.35. Do not use 0.33 or 7.33
There are also routeShow/routeAdd/routeDelete commands for modifying routing tables.
You must reboot your PXM after each modification with "bootChange" for it to take effect. Also make sure the subnet mask is 255.255.0.2
. bootChange- Only enter the ethernet IP address, netmask and default gateway.- Type "." to erase incorrect entries.tigers.1.7.PXM.a > bootChange'.' = clear field; '-' = go to previous field; ^D = quitboot device :lnPciprocessor number :0host name :C <-- Please put "C".file name :inet on ethernet (e) :172.29.37.40:ffff00 <-- Ethernet IP Addr/Netmaskinet on backplane (b):host inet (h) :gateway inet (g) :172.29.37.1 <-- Default Gatewayuser (u) :ftp password (pw) (blank = use rsh):flags (f) :0x0target name (tn) :startup script (s) :other (o) :- Type in reboot, after this the command "ping" will work:tigers.1.7.PXM.a > ping 171.71.54.53 1171.71.54.53 is aliveConfiguration save and restore is only supported through the CLI (CWM does not support configuration save and restore).- Service module upgrades error handling is not provided. If the user skips any of the steps during upgrade or if a power failure happens in the middle of the upgrade, results will be unpredictable. See the Special Installation and Upgrade requirements section for service module upgrades. To recover from procedural errors contact your TAC support personnel.
The MGX 8850 supports 15 simultaneous telnet sessions and 10 tftp sessions.
You must use the following Y cables for FRSM-HS2 and FRSM-CT3 redundancy as specified in the Product Orderability Matrix (Straight Cable: 72-0710-01, Crossover Cable: 72-1265-01, Straight Y-cable: FRSM-HS2: CAB-SCSI2-Y, FRSM-CT3: CAB-T3E3-Y). Other cables are not supported.
Y cable redundancy for FRSM-HS2, FRSM-2CT3, FRSM-2T3, FRSM-2E3 is only supported for adjacent slots.
Statistics are not supported for the RPM.
There is no need to issue the syncdisk and shutdisk commands before removing the PXMs. The system quiesces the disk by detecting the removal of the PXM board and flushes the write buffers to the disk and puts the PXM in sleep mode. This disables any further hard disk access since it locks the acctuator. When the card is reinserted the PXM automatically comes out of sleep mode.
Syntax of "addlink" command has changed as follows:
New Syntax:
Syntax: addlink <T3LineNum> <T1Slot> <NumberOfT1s> <TargetSlotNum>
<TargetSlotLineNum>
<T3LineNum> where = Slot.Line
Slot = 15,31
Line = 1 - 3
<T1Slot> where T1Slot = 1 - 28
<NumberOfT1s> where NumberOfT1s = 1-8
<TargetSlotNum> where TargetSlotNum = 1-6|11-14|17-22|27-30
<TargetSlotLineNum> where TargetSlotLineNum = 1-8
PAR command "cnfnwip" has been disabled in this release, please use "cnfifip" instead.
If you lose power, or remove the on-line PXM you lose the broadcast address. Use the "cnfifip" command to configure the broadcast address. To re-define your ATM address and IP Address that are in the same subnet, you have to change the ATM address to a temporary address not in the same subnet, then add back your IP Address with the original Broadcast address, then go back and correct your ATM address.
Cooling and Power limitations: Customer should be aware of the need for extra power supplies and fans beyond certain limitations. A single fan tray will support all configurations that draw between 1200 and 1400 watts. For power requirements, the MGX 8850 requires a minimum of one power supply per line cord to support the power requirement for 5 cards.
This is based on an estimated worst case power requirement of 190W plus margin per card slot.
CONNECTION MANAGEMENT RELATED
The name of the node cannot be changed if there are PVCs. The node name must be changed from the default value before adding connections, since it cannot be changed later. Use the cnfname command to change the node name.
Only one feeder trunk can be configured. No BNI trunk to MGX 8850 as a feeder is supported.
The slave end of a connection must be added first.
The slave end cannot be deleted and re-added back by itself. If you delete the slave end, the entire connection must be completely torn down and re-added back. If the slave end of the connection is deleted and re-added back by itself, then unpredictable results will happen.
For user connections, VCI 3 and VCI 4 on every VPI are reserved for VPC OAMs.
The actual number of feeder connections you can provision on the PXM is always two less than you have configured. (the dsprscprtns command shows max connections as 32767, but you can only use 32767 - 2 = 32765). One connection is used for LMI and another one for IP relay.
There is no error handling detection while provisioning through the CLI. Invalid endpoints and unsupported connection types (such as connections between FRSM-CESM ports or connections between structured and unstructured connections) are permitted using the CLI. The user should not configure these connections.
The sum of CIR of all channels of a port can be greater than port speed as long as CAC is disabled. However, it is not acceptable for one channel's CIR to be greater then port speed even if CAC is disabled. Two channels added up can exceed port speed. This means you cannot oversubscribe a port if only one channel is configured.
When trying to add a port on DS0 slot 32 of a CESM-8E1 line using an SNMP set or the CiscoView Equipment Manager, The SNMP agent in CESM will time out, without adding the port. The SNMP libraries treat the 32 bit DS0 slotmap (cesPortDs0ConfigBitMap) as an integer. The value for the last DS0 is treated as the sign value. This causes a corruption in the packet coming to the agent. As the agent does not receive a complete SNMP packet, it does not respond and times out. Use the command line interface to add a port on DS0 slot 32 of a CESM-8E1 line.
The cnfport command does not allow VPI ranges to be reduced. The cnfport command only allows the VPI range to expand. The correct sequence is to delete all connections on the partitions, delete the partitions, delete the port and add the port with new VPI range.
On an FRSM-2CT3, one can add 128 ports on a group of 14 T1 lines as indicated below.
lines 1 to 14 -- 128 ports (A)
lines 15 to 28 -- 128 ports (B)
lines 29 to 42 -- 128 ports (C)
lines 43 to 56 -- 128 ports (D)
So, to add 256 ports on one T3 one should add 128 ports on the first 14 T1 lines and the remaining 128 on the next 14 T1 lines.
Note that (A) and (D) are connected to 1st FREEDM and (B) and (C) are connected to the 2nd FREEDM. Each FREEDM supports only 128 ports. If 128 ports are added on one T3 as in (A), then there cannot be any more ports as in (D). The 129th port should be on lines 15 to 42 (as in B or C).
If the user adds a connection between an RPM and a PXM and then deletes the connection the RPM shows no connection but the PXM still has the connection. The MGX was designed and implemented in such a way that only the connections that have the master end show up on PXM (by dspcons command). Consider these three connections:
c1 - has only slave end,
c2 - has only master end,
c3 - has both master and slave end.
When using the dspcons command, c2 & c3 will be displayed, NOT c1. The connection will not show up once the master end (PXM) is deleted. Recommendation: When adding a connection, if one end of the connection is PXM, always configure the PXM side to be the slave. Thus when deleting the RPM side, which is the master, the connection will not show up on the PXM. However keep in mind that the slave end (PXM) still exists. This also provides a side benefit. When a connection exists with only the slave side, no bandwidth is occupied. The bandwidth is reserved only if the master end exists (with or without the slave).
The MGX-FRSM-HS1/B is capable of supporting a total throughput (card-level) of 16 Mbps. However, it is possible to configure 4 lines each supporting up to 8 Mbps, thus oversubscribing the card. This has been raised in bug #CSCdm71476 and a restriction/warning will be added in a future release.
Addlnloop on an FRSM-HS1/B line works only when there is a (valid) cable plugged in to the backcard on that line. This is a hardware limitation on the backcard and has been mentioned in the Release-notes in bug# CSCdm44993
RPM Related
The RPM is a NPE-150 based router card capable of sustaining 150,000 pps. With RPM versions earlier than 12.O.7T1, some limitations in Inter-Process Communication between the can cause the PXM to declare that the RPM has Failed, when the RPM is at high loads. To avoid this, with RPM 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 700 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. Currently, the RPM supports 700 subinterfaces, and hence 700 networks with BGP, RIP or static routing.
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. Using such configurations, the RPM with IOS software release 12.0.7T1 supports 700 customer networks which use OSPF internally. (For more information on readvertisement, 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.
Recommendations for Booting:
The current implementation provides the following options:
From PXM Disk
NetBoot (TFTP server)
Booting from PXM Disk is faster than NetBoot.
----------------------------------------------------------------------------------------------------------------------
Recommendations for saving RPM configuration
The current implementation provides the following options:
a. Save on flash / boot-flash.
b. Save on PXM Disk.
c. Save on network (TFTP server)
d. Save on RPM NVRAM (comes up faster; only for limited configuration size)
It is recommended to save the configuration on flash and on the PXM Disk, as well as on the network server. This ensures that the configuration can be restored; even in the case of multiple failures.
For example if an RPM card has problems, one can copy the configuration from either the PXM disk or from the network to new RPM card. In case of multiple hardware failures (both RPM and PXM cards have problems) one can copy the configuration from the network server.
----------------------------------------------------------------------------------------------------------------------
Replacing the existing RPM with a new card or a card with old configuration in flash:
The existing configuration (of the old card) can be restored on the newly inserted card by following the instructions given below:
Step 1
Step 2
Step 3
Step 4
Step 5
Please note that in RPM context the "config save/restore" feature of the PXM only restores the PXM part of the RPM configuration/connections. The RPM part of the configuration should also be saved from RPM CLI through copy command (For example: "copy run c: <config-filename>" for saving to PXM Disk) for future restoration.
----------------------------------------------------------------------------------------------------------------------
RPM Connection Resynchronization:
The RPM Connection Re-sync process is supported in the 12.04T and higher releases. This feature checks for consistency between the RPM and PXM connection databases. ----------------------------------------------------------------------------------------------------------------------
Limitations
restoreallcnf
The MGX 8850 does not support CESM endpoints with MGX 8220 Release 4.1 through CWM, but is supported through the CLI (CSCdm11835).
Do not execute the restoreallcnf command in the middle of the installation process. If you do, the dsplns command will display a line as disabled, but you cannot run an addln command. Do not execute the restoreallcnf command until the install and newrev commands have completed.
The correct order for the restore procedure is:
Step 1
Step 2
Step 3
Step 4
.(for more information, refer to CSCdm57683)
The Service MIB does not support resource partitions.
LIP is supported on the maintenance port, but there is no PPP support on the maintenance port.
BIS messages are constantly being sent from BPX to various nodes. This affects the frequency of TFTP updates, which may affect CWM performance and/or CWM database consistency.
Unable to provision virtual trunks in SWSW 9.1.10.
clrsmcnf
As a speedy way to wipe out all configuration on an SM, you can use clrsmcnf. This command works in the following scenarios:
•
•
•
To be able to use an SM of a different type from the current one in a slot you can also use clrsmcnf. For example, if there is a FRSM8t1e1 in the slot with some configuration and the customer wants to use this slot for an AUSM8t1e1 card.
clrsmcnf cannot delete a port or channel due to corruption or error locally on the SM. It is able to delete the port/channel from within the PXM, but it cannot delete a port or channel due to corruption/error on the PXM itself. You can save an SM configuration and restore it back to the same slot on the same node. If the SM configuration is corrupted on disk, but the run-time image is okay and the file contented is corrupted, this is supported.
The following are NOT supported on the MGX 8850:
•
•
•
If you have more than 500 connections on a service module, before issuing clrsmcnf you need to change the session timeout default value.
Use CLI command timeout 0 (no timeout)
clrsmcnf
after it is done
Use CLI command timeout 600 (to set the timeout value back to the default)
Note
Core Dump Mask
There are no system performance implication unless you take a core dump, currently the default error mask to take coredump is attached, you can change the mask or take it manually.
Set the core dump mask to its default value. If you enable core dumps with power on reset and shell reset core dumps enabled you will end up with a PXM that continuously dumps the core and resets. The only way out is to use a download boot that does not have the core dump feature.
Default setting 0x262eeOFF 0001 Power ON ResetON 0002 DRAM Parity ErrorON 0004 WatchDog Timeout ResetON 0008 Resource OverflowOFF 0010 Clear All ConfigurationON 0020 Missing TaskON 0040 Reset because of PXM Low VoltageON 0080 Reset By Event Log TaskOFF 0100 Reset from ShellON 0200 UnknownOFF 0400 Reset from PXMOFF 0800 Reset SystemOFF 1000 Switch Core CardON 2000 Secondary Cache ErrorON 4000 Software Error ResetOFF 8000 S/W reset due to upgradeOFF 10000 Restore All ConfigurationON 20000 Device Driver ErrorNODENAME.1.8.PXM.a > core hot-dumpDo you want to proceed (Yes/No)? yDumping PXM Core Image[0]:......................................................................................... .......................................Done.NODENAME.1.7.PXM.s > core save 2 core.zipCreating core.zip.................................Core dump basics.There are two steps to save a core dump.1. The system will store a raw core dump image on the disk (this image ison a portion of the disk that is not used for the filesystem).2. After the raw core dump image is saved on the disk, use the cli command"core" to zip the image and save it in a file on the disk.Use the "core mask" command to display and to set the mask which determineswhich conditions will cause an autmatic core dump.NODENAME.1.7.PXM.s > core maskAutomatic Core Dumping is enabled..The Current Core slot is 0The Current Core mask is 0x4004OFF 0001 Power ON ResetOFF 0002 DRAM Parity ErrorON 0004 WatchDog Timeout ResetOFF 0008 Resource OverflowOFF 0010 Clear All ConfigurationOFF 0020 Missing TaskOFF 0040 Reset because of PXM Low VoltageOFF 0080 Reset By Event Log TaskOFF 0100 Reset from ShellOFF 0200 UnknownOFF 0400 Reset from PXMOFF 0800 Reset SystemOFF 1000 Switch Core CardOFF 2000 Secondary Cache ErrorON 4000 Software Error ResetOFF 8000 S/W reset due to upgradeOFF 10000 Restore All ConfigurationOFF 20000 Device Driver ErrorNODENAME.1.7.PXM.s > core mask 0x2e2eeAutomatic Core Dumping is enabled..The Current Core slot is 0The Current Core mask is 0x2e2eeOFF 0001 Power ON ResetON 0002 DRAM Parity ErrorON 0004 WatchDog Timeout ResetON 0008 Resource OverflowOFF 0010 Clear All ConfigurationON 0020 Missing TaskON 0040 Reset because of PXM Low VoltageON 0080 Reset By Event Log TaskOFF 0100 Reset from ShellON 0200 UnknownOFF 0400 Reset from PXMOFF 0800 Reset SystemOFF 1000 Switch Core CardON 2000 Secondary Cache ErrorON 4000 Software Error ResetON 8000 S/W reset due to upgradeOFF 10000 Restore All ConfigurationON 20000 Device Driver ErrorNODENAME.1.7.PXM.s >Use the "core mask default" command to set the mask back to the default.NODENAME.1.7.PXM.s > core mask defaultAutomatic Core Dumping is enabled..The Current Core slot is 0The Current Core mask is 0x262eeOFF 0001 Power ON ResetON 0002 DRAM Parity ErrorON 0004 WatchDog Timeout ResetON 0008 Resource Overflow
Guest
