topic RE: X460 and X440 stack locking up in ExtremeSwitching (EXOS/Switch Engine)

X460 and X440 stack locking up

Tim_Senaldi — Wed, 17 Feb 2016 05:32:00 GMT

I have a customer with 6 x440's and 1 x460 switches in a stack. The 460 is the master in slot 1. The problem is that the switches on occasion lockup. When this happens all the lights go solid and no traffic will pass. The fix is to reboot the stack. When I have looked in the logs there were no signs of what caused the issue. They were running 15.3.1.4 patch 1-33. I have just upgraded to 15.6.3.1 patch 1-9. I am looking for anyone that has the switch stacks lock up. I have made sure it was not a loop. I have checked that is not an Intel NIC multicast issue.

RE: X460 and X440 stack locking up

David_Rahn — Wed, 17 Feb 2016 05:47:00 GMT

look here to see if this helps

I had same issues...

https://gtacknowledge.extremenetworks.com/articles/Solution/CRC-errors-are-seen-on-Summit-x460-stack...

is fixed in some software, or you set some setting on the stacking, if it is this issue.

RE: X460 and X440 stack locking up

Tim_Senaldi — Thu, 18 Feb 2016 03:09:00 GMT

This is helpful. It is hard to explain the issue to GTAC after the customer reboots the switches and the logs buffer out. Thanks.

RE: X460 and X440 stack locking up

BrandonC — Thu, 18 Feb 2016 03:09:00 GMT

Hi Tim,

If this happens again, please collect the output of the commands below from both the master and backup (if you can) in the stack before rebooting.

show ports stack-ports rxerrors no-refresh show ports stack-ports txerrors no-refresh show stacking stack-ports show stacking detail show stacking show slot show slot detail show tech all show log messages nvram
You should be able to connect to the backup in the stack by telnetting across the backplane with the command 'telnet slot '.

This information will help us once you do open up a case with GTAC after rebooting the stack to recover.

-Brandon

RE: X460 and X440 stack locking up

Tripathy__Priya — Thu, 18 Feb 2016 09:04:00 GMT

Hello Tim,
Going with Brandon and David suggestion would like to add few more points here to address if there is any Rx crc errors in the stacking ports when the issue reappears again on customer switch.

For this as this is X460 stack switch hence
To identify the presence of RX errors in the stack-port, use the command : "sh port stack-ports rxerrors "

The article provided by David earlier in the post should help you troubleshoot this!

CRC errors are seen on Summit x460 stack links and mixed Summit x440 - x460 stack links

Apart from the outputs requested by Brandon, please also collect the following outputs from the stack nodes and contact GTAC with these outputs.

debug ems show trace devmgr all

debug ems show trace nodemgr all

debug ems show trace hal all

Hope this helps!

RE: X460 and X440 stack locking up

Tim_Senaldi — Fri, 19 Feb 2016 04:09:00 GMT

Thanks. I will add these to the others.