Cisco 6500 soft and hard error question

[andrewswanson]

Hi

I had a 6503 (supervisor WS-SUP720-3B) switch suddenly crash/restart. When I logged onto the switch after it reloaded I found the reason for the reload:

System image file is "sup-bootdisk:s72033-ipservicesk9_wan-mz.122-33.SXI11.bin"
Last reload reason: processor memory parity error at PC 0x42AE9120, address 0x0

I had a look at the the crashinfo and found:

Real cache error detected.  System will be halted.

From the Cisco documentation:

http://www.cisco.com/c/en/us/support/docs/switches/catalyst-6500-series-switches/116135-trouble-6500-parity-00.html

Explanation

This is the result of a parity error within the Level 2 (L2) cache (static random-access memory, or SRAM) used by route processor (RP) or switch processor (SP) CPU of the Multilayer Switch Feature Card 3 (MSFC3).

Recommendation

Monitor the system regularly for reoccurrence. If no further events are observed, it is a soft error. If the error occurs frequently, request a Return Material Authorization (RMA) in order to replace the Supervisor Engine, and mark the module for equipment failure analysis (EFA).

I haven't come across this issue before and was looking to find out whether this type of incident is more likely to be "soft" rather than "hard" based on other's experiences.

Thanks
Andy

[Reza Sharifi · Accepted solution]

Hi,

Based on what happened to the switch (crash and restart) this is a hard error. If this switch is not a critical, you can upgrade the IOS, monitor and wait for recurrence. If the switch is critical to your environment, I would open a ticket with TAC, RMA the sup and get a replacement.

BTW, 122-33.SXI is al least 8 years old.

HTH

[andrewswanson]

Thanks for the quick reply. The switch has come recently come under my administration and I was planning an ios upgrade soon. The switch is critical as it does BGP routing with an ISP. I shelf the IOS upgrade and bring forward a plan to upgrade the sup or at the very least have it replaced.

Thanks

Andy

[Reza Sharifi]

Hi Andy,

Since the device is running BGP with the provider than it is really critical to your environment as every crash/restart will disable access to Internet. I would open a P1 ticket with TAC right a way, send them the info you have and prepare for upgrade and replacement.

Good Luck!