March 27, 2003
Products Affected
|
Product |
Comments |
|---|---|
|
AXSM-16-T3E3-E |
With 2.1.7x software and set up for redundancy |
|
AXSM-2-622-E= |
With 2.1.7x software and set up for redundancy |
|
AXSM-32-T1E1-E= |
With 2.1.7x software and set up for redundancy |
|
AXSM-8-155-E= |
With 2.1.7x software and set up for redundancy |
Problem Description
The minimum cell rate (MCR) of available bit rate (ABR) connections is not correctly programmed on the standby ATM Service Switching Service Module-Enhanced (AXSM-E) card in some cases. MCR is set to the initial cell rate (ICR) on the standby card. The default ICR is the peak cell rate (PCR).
If ABR connections are not present, or AXSM-E cards are not used, there is no issue.
Background
This anomaly occurs only with ABR connections on AXSM-E cards that have Automatic Protection Switching (APS) or Y-redundancy enabled. In addition, the following conditions must apply before this anomaly is observed. This anomaly may present itself at the card level (see "I") or at the port level (see "II") so both conditions must be checked.
The problem will occur if the following conditions exist:
I) Card level:
-
The sum PCR (over all CBR connections on the card) * BookingFactor of CBR + the sum SCR (over all VBR connections on the card) * BookingFactor of VBR + sum ICR (over all ABR connections on the card) * BookingFactor of ABR exceeds OC12.
-
The standby card becomes active.
II) Port level:
-
Sum PCR (over all CBR connections on this port) * CBR_bookfactor + sum SCR (over all VBR connections on this port) * VCR_bookfactor + sum ICR (over all ABR connections on this port) * ABR_bookfactor exceeds MAX RATE of the port
-
The standby card becomes active
For I)
After a switchover, the following effects will be seen
-
It appears as though the card is over-booked and Quality Of Service (QoS) will be affected, if traffic into the card exceeds OC12.
-
Connection rebuilds "could" fail. Connections will be OK until rebuilds are attempted. What this implies is: Connection re-routes, line failures, dnport/upport, dnpnport/uppnport could cause connections that were earlier in a good state to fail. "Could" is used since: If the ABR conns are re-built the problem is actually being fixed. If a connection is being re-committed and condition 1. exists, the connections will fail.
-
No more connections will be allowed while condition I) exists.
For II)
-
No more connection additions will be allowed on this port, while the condition II) exists
-
Connection rebuild on this port could fail. Other ports will be OK.
Problem Symptoms
Connections that were previously working correctly will fail.
Workaround/Solution
Workaround: Set ICR of ABR conns to MCR (using cnfabr from the AXSME command line). OR Set BookingFactor of ABR connections to a low value.
Solution: Upgrade software to 2.1.80 or later release. Also, if upgrading to 3.0.00 or later, the network must be upgraded to 2.1.80 first.
Note:?This problem caused Release 2.1.79 to be removed from the Order Tool on Cisco.com.
DDTS
To follow the bug ID link below and see detailed bug information, you must be a registered user and you must be logged in.
|
DDTS |
Description |
|---|---|
|
CSCdx83285 (registered customers only) |
AXSM-E UNI port stuck in building vc state after upgrade. |
For More Information
If you require further assistance, or if you have any further questions regarding this field notice, please contact the Cisco Systems Technical Assistance Center (TAC) by one of the following methods:
Receive Email Notification For New Field Notices
Product Alert Tool - Set up a profile to receive email updates about reliability, safety, network security, and end-of-sale issues for the Cisco products you specify.