Monitor Failing Switch?

[opel70]

Hello,

We have a 3560g switch that seems to fail about every two weeks. First it slows down, and then we can't access anything connected to it. Finally it reboots itself. All this happens over the course of about three hours. We have MRTG configured, but are currently only monitoring bandwidth on a few of the ports.

What should we try monitoring that might show what is going wrong? I am assuming there is something that keeps increasing until it gets too high and the switch has the problem, just not sure what that thing might be.

Thanks,

Tim

 

[lerdalt]

Do you have a syslog server? You might want to get something setup to capture the log file on the local switch to see what if any error messages are coming up.

 

[opel70]

Oh yeah, I forgot to mention that we do have logging to a syslog server enabled and there were no messages sent prior to the reboot. The most recent message was 12 hours before. We have logging set to "informational". I believe the only other option is to set it to "debug". Which I suppose we should do for this stretch.

Tim

 

[VinceWhirlwind]

You should be logging CPU and MEMORY.

 

[opel70]

You wouldn't happen to know the SNMP OID for those would you?

I'll search when I get to work tomorrow, but if you have them handy, that would be great.

Thanks,

Tim

 

[vipergg]

Also check software version on it for bugs. The latest version is like 12.2.53SG or soemthing like that.