05-20-2021 01:21 PM
We are having an issue with a 2960X switch stack consisting of 7 switches. Software version is Version 15.2(7)E4, RELEASE SOFTWARE (fc2). Whenever we hook up the stacking modules in a complete ring, we get strange issues resulting in switches dropping out of the stack ring and the duplicates counter in misc counters for the stack incrementing. If we remove the cable going from the last switch's stacking module to the first switch's stacking module, the stack stabilizes and runs fine.
We suspect a bad stacking module, but are not quite sure how to go about isolating which one might be causing the issues in order to get it replaced. Any help is appreciated. Thanks!
05-20-2021 01:55 PM
Hi,
Have you replaced the stacking cable that goes between the first and last switch?
Does the output of "show stack" show the correct speed on the ring?
HTH
05-21-2021 08:33 AM
Reza,
We have not had a chance to locate another stacking cable to try swapping this out. Show Stack does not display ring speed, just an FYI, it is actually show switch stack-ring speed. Now that the stack is stabilized, the only chance I have to play with it for troubleshooting is late night when users aren't present in the area.
I have collected some info and attached the text file for review. The first part of the file shows things I was seeing with the full stack ring enabled that jumped out at me as being goofy. Everything after that is switch info collected after stack stabilization.
Thanks,
Marvin
05-21-2021 09:16 AM - edited 05-21-2021 09:18 AM
Hi,
Appears to be some sort of buffer/hardware issue with the Asics when you connect that last cable. Almost at the same time, the stack members are being removed.
Do you have a ticket with Cisco on this?
Also, have a look at this link for a possible bug in the software
https://community.cisco.com/t5/switching/cisco-2960-x-error-supq-4-port-queue-stuck/td-p/2717665
May 20 03:46:45.565: Error disabling queue 1 for asic 1 port 14
May 20 03:46:45.565: Error disabling queue 5 for asic 1 port 14
May 20 03:46:46.047: %STACKMGR-4-SWITCH_REMOVED: Switch 3 has been REMOVED from the stack
May 20 03:46:49.294: %STACKMGR-4-SWITCH_REMOVED: Switch 7 has been REMOVED from the stack
05-21-2021 09:46 AM
Reza,
Tried to put a TAC ticket in with Cisco, but they said our switches weren't covered by any contract and we would need to purchase Smartnet for them before they would do anything. Will be running that possibilty by management for possible approval of purchase.
As far as the bug report you listed, that is for a much older version of firmware and supposedly was fixed in later versions. That is not to say that Cisco might not have reintroduced the same or similar bug in later versions. My config does have the MLS QOS command, so I guess I could try removing that and try to complete the stack ring again.
Thanks,
Marvin
05-21-2021 10:06 AM
Yes, if you don't have support, Cisco will not help.
Also, just for testing, can you remove the MLS QOS command for a period of time and see if there is a difference. As you said correctly, Cisco may have reintroduced the same bug in the version you are running. If things stay stable for a while, then an upgrade may fix the issue but you would need to put the switches under support in order to be able to download the IOS.
Thanks,
05-21-2021 10:31 AM
Reza,
I can try to find a time after hours to try removing that command. Unfortunately, some of our software systems are eventually wanting us to implement end-to-end QOS and, if I am not mistaken, we would need that command available to accomplish that. As far as upgrading if that fixes the issue, I don't see that happening as 15.2.7 is the newest version available for this switch model.
Our best bet still might be to purchase a Smartnet for these switches and get Cisco to fix the problem.
Thanks,
Marvin
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide