SSH and Web-Management breaks down on 2920-24g-poe+ switches

[MG2018]

Hi,

in our company we have several 2920-24g-poe+ switches, which have the problem that all of them loose ssh after the same period of time. We recognize this in our monitoring. Then, the switch isn´t manageable any more, even via the web-interface. It is not possible to view the event log any more because when I try to connect via the console port, the switch reboots automatically and the log is gone. All I can do is power off and on the device, which isn´t really acceptabe in working time. Has anyone any Idea what can be the problem?

[parnassus]

Pretty strange behaviours (especially considering that all of your switches are suffering of the same issue at the very same time).

Can you share a sanitized running configuration and explain what "monitoring" (system) you have in place (and how it is configured to monitor your switches)?

Have you ever considered the idea that the way your monitoring system is actually controlling your switches could be the ultimate cuplrit (it was seen before)? what will happen if you stop to monitor all of them (or just a part of them)?

[MG2018 · Accepted solution]

Hi there,

i guess the problem is solved.

I remembered that a few months ago, when the problem occured the last time, I updated the firmware in two of the switches from version 16.05.0007 to 16.06.0006 to test if this is the problem. These two switches didn´t loose ssh this time, although they were supposed do. So,

everybody having the same issue should consider a firmware update.