Unexpected switch reboots

[Dave]

I asked Stephen to get back to you on this.

[Stephen]

OK I have been running a WS-12-250-AC with this config for some time. I haven't seen any restart's or crashing so far. Can you let me know a bit more about how it's connected on your network? It doesn't seem to replicate in my environment Also, as Ludvick suggest's check to see what the memory utilization on the failing switch is on the Device->Status tab to see if it is growing slowly until the switch crash's. Also a full log of the switch currently just to see if's normal behavior would also be helpful.

[craig.moscardini]

Unfortunately this issue is affecting ANY of our WS12s. We have hundreds of them. This isn't a case of constant reboots so it's difficult to monitor and I don't think it can be put down to a config issue. Most switches probably have only rebooted once or twice in the past few months. If we only had a couple of switches it wouldn't ever get noticed, but because of the quantity we see a few a week. Another switch rebooted last night. Memory usage after reboot is under 50% and CPU constant around 25-40%. Not sure how it was before

[Dave]

How are you powering the WS-12-250-DC?

[craig.moscardini]

They are all DC running either 12v or 24v. Some are from batteries and some are from various types of inverter direct from the mains

[craig.moscardini]

Stephen can you use a DC switch as I think we are only seeing reboots on those. We have very few AC switches, and most of them are bigger ones.

[Stephen]

Yes I can but the config you sent me seemed to be for an AC switch, can you send me one of the DC ones?
I have one of every model of our switches active at all times in my lab. Under normal conditions (at least in my lab) I have not seen this behavior.
Something else that might help would be to configure a syslog server and have some of your DC switches log there that way the information leading up to the crash isn't lost - maybe it can give us some insight.

[craig.moscardini]

We have very few 12 port AC switches so I am surprised it's an AC config. It does confirm then this happens on AC and DC switches. We have specifically set up a syslog server to capture logs from these reboots, but no error is logged prior to the watchdog reboot. Is there are way to increase the logged information?

[Stephen]

I think the best way I can help right now is if you posted (or PM'd if you prefer) the logs from syslog to see if I can find a pattern.
Also if you could send me a config from one of the known afflicted DC switches it will give me a larger range of test's to try.

[craig.moscardini]

I have just sent you a PM with all logs and config files from switches that have rebooted in the past week