Description

The article is about the PLX switch on the PTX devices.A major alarm is generated when the PLX switch fails.
PLX Swithc

Symptoms

root@XX-lax31> show chassis alarms no-forwarding

2 alarms currently active

Alarm time        Class Description

2023-12-07 04:44:11 UTC Major SPMB 0 Stalled (See Logs for Recovery Options)

2023-12-07 04:44:11 UTC Major SPMB 0 CB_PLX switch failed

 

Jan  4 14:14:54.946  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)
Jan  4 14:14:55.447  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)
Jan  4 14:14:55.949  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)
Jan  4 14:14:56.449  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)
Jan  4 14:14:56.950  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)
Jan  4 14:14:57.452  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)
Jan  4 14:14:57.951  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)
Jan  4 14:14:58.449  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)
Jan  4 14:14:58.949  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)
Jan  4 14:14:59.450  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)
Jan  4 14:14:59.949  lab123 chassisd[21646]: SPMB 0: Has been stalled due to Hardware Fault, RecoveryOption: Replace hardware (CB_PLX switch)


 

Solution

SPMB0, "Switch Processor Mezzanine Board " runs as a process on routing engine to manage SIBs nad PCIe boards. SPMB 0 is the master SMPB that manages SIB and fabric cards.
Backup SPMB1 is on backup RE1.
As per the alarm, we need to replace the RCB0 because the internal PLX switch hardware has failed.
The routing engine has PCIe switch PLX86xx with it's downstream ports connecting to SIBs in the different slots and it's upstream port connected to the CPU root complex.

As of now, there should not be any impact as all SIBs are online. If any SIB card reboots, SPMB0 / master SPMB might not be able to bring it up as the PLX switch has failed which connect SPMB with SIB cards.

Thus, requested customer to switch over the routing engine  to make RE1 master so that SPMB1 also becomes master, until RE0 is  replaced.

 
 

Modification History

2024-02-26 : Article Created