Catalyst 9200 stack troubleshooting

[janelle.provine]

I just installed a three cisco c9200 in a stack to replace an existing stack, and created uplinks/port-channels back to a core switch. The only thing plugged into the switch are the uplinks to the core. We are seeing a small 16port switch connected to the old stack freezing and locking up. The only way to get it to work again is to reboot the 16 port switch. When I disconnect the uplinks to the core from the main switch the 16 port switch works fine.  This 10 port switch is the only device being affected by the uplinks of the new switch being plugged in.

[marce1000]

 - Configure a (central) syslog server on all involved devices (the core , c9200 stack , old stack  , the 16 port switch) , scrutinize logs send ;  especially when the 16 port switch freezes and or around that timeframe.

 M.

[janelle.provine]

I have been looking at the logs on the switches but do not have the option to set up a syslog server. The 16 port switch is unmanaged I can't even login to i

[marce1000]

 - Well that's unfortunate but foremost very easy to do , you can for instance spin up a linux vm , and after a 'few googles' on what you need you will have it going.  The reason for that is that I suspect a spanning tree conflict of some sort , which the reply as outlined in my first answer can easily determine ; on the other hand is it is strongly advised in modern networks to no longer use unmanaged switches , for starters because of the 'wording' (...) and the resulting problem you are experiencing (...).

 M.

[janelle.provine]

All good points but I am a consultant hired to replace a stack of switches with a new stack and the customer wanted to slowly move the devices from the old switch to the new stack at his convenience. I was unaware of the unmanaged switch until it started to lockup on the introduction of the new switches even though nothing is plugged into the new stack but the uplinks back to the core switch. so creating syslog server are out of scope of my contract

[marce1000]

- Convince the customer to ditch the unmanaged switch.

M.

[balaji.bandi]

what is the Core Switch ? Do you have rough diagram how they connected ?

We are seeing a small 16port switch connected to the old stack freezing and locking up. ?  - what model of this switch ?

Where is this 16port switches connected  to Core ?

[janelle.provine]

no idea what model the unmanaged is, it is connected to the production switches as shown above

[balaji.bandi]

in the diagram what switch you rebooting the fix the issue ?

Do you have port configuration of the 10port switch connected ?

[janelle.provine]

The one in the yellow circle, its an unmanaged switch so no I have no information for it, I can't even get to it

[balaji.bandi]

in the diagram what switch you rebooting the fix the issue ?  - so you reboot all the time 16port switch to fix the issue

and you claiming all this happend soon you connect Cat 9200 switches uplink (and before it was ok ?)

[janelle.provine]

Yes rebooting the unmanaged switch fixes the issue temporarily, it finally goes into lock up again later. Yes only when the second or new set of 9200s are brought on line with the old set

[balaji.bandi]

technically i do not see a reason, as suggested Migrate the device from that to cat 9200 is moving forward.

since we have less tools from the switch to see what is wrong ?

do you have show interface output that switch connected along with port config ?