5

Currently I have 5 Catalyst 3750 V2-48TS as slaves and 1 Catalyst 3750g 24TS-1U as the Stack Master.

I was wondering if there was any way to change the Stack master to one of the slaves without devices plugged into the ports on the master loosing connection. Since these devices cant lose connection the master cannot reboot to cause a new master election. I've looked at the documentation for the Catalyst 3750's Switch Stacking features (https://www.cisco.com/c/en/us/support/docs/switches/catalyst-3750-series-switches/71925-cat3750-create-switch-stks.html) but there seems to be no way to change the master without rebooting the Catalyst 3750g 24TS-1U.

This is the output to show platform switch all (due to internal policies i had to censor the MAC addresses): https://pastebin.com/6PVpHjns

Improve this question
10
  • Why are you trying to change the master or why do you not want it to be master anymore?
    – user36472
    Jan 25, 2018 at 15:17
  • For some reason the Switch that is currently the master (The Catalyst 3750g 24TS-1U)will periodically stop accepting any sort of connection to manage it. This includes attempts with Cisco Network Assistant, Telnet, and even physically connecting with a console cable. The only way we have been able to fix this is to restart the switch which messes with the devices connected to it that need to stay up. A coworker has talked to Cisco support about this and they where unable to provide an answer to why it was happening.
    – Quitehatty
    Jan 25, 2018 at 15:25
  • Ok, then change the priority of the switches. (config)# switch 1 priority 0 and then give one of the others the priority of master switch 3 priority 15. The users will still be disconnected when the switch is shutdown, but switch 3 will take the master role and run the others.
    – user36472
    Jan 25, 2018 at 15:27
  • on a side note, you mention telnet, which is very unsecure (password are transmitted in clear text as well as anything you type) you should disable telnet and use SSH - preferably with RSA key.
    – JFL
    Jan 25, 2018 at 15:30
  • @JFL That is something we are also working on fixing.
    – Quitehatty
    Jan 25, 2018 at 15:33

2 Answers 2

Reset to default
1

You will either need to shutdown the current master, or you will need to restart the stack. Once the stack is up and running, there is no election unless the current master stops.

You should schedule a time to do this. There is no such thing as never being able to (at least, briefly) disconnect a device (stuff happens), but you can minimize the impact.

If you have devices that cannot go down, you should have dual connections, preferably using a port channel (nearly instantaneous failover) to two different switches in the stack. Stopping the current master would then not disconnect the devices.

Improve this answer
2
  • Scheduling a time to do this is what we are tying to avoid but it is an option we have considered. I was just looking for an alternative solution. The suggestion about using a port channel for redundancy may be an option for us to look into.
    – Quitehatty
    Jan 25, 2018 at 15:29
  • He's eventually going to have to reboot it anyway -- at the next lock-out. So there is a window to change the stack master.
    – Ricky
    Jan 25, 2018 at 21:10
-1

Run "sh process CPU history" from CLI. Look back when switch had no console or vty conenctions...

There are versions that report fine when being monitored but are realy running at 70-90 pct CPU utilz

The only time I have seen a Cisco switch refuse console is over memory/ cpu issue.

Check mem/ CPU and your firmware for bug.

Improve this answer
1
  • I failed to see the relation between the question and your answer.
    – JFL
    Feb 8, 2018 at 20:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.