[c-nsp] Mysterious 2924 reboots

[Tuc at T-B-O-H.NET]

> 
> > 	Still happening... We had a power outage at the site, 
> > so everything
> > was off for about 5 minutes. Then it came back on :
> > 
> > c2924-1 uptime is 2 days, 14 hours, 48 minutes
> > c2924-2 uptime is 18 hours, 40 minutes
> > 
> > 	so still wondering why one unit took about 1 day and 20 
> > hours AFTER
> > power was applied before it booted up. I did upgrade that one 
> > to the latest
> > IOS I had, c2900xl-c3h2s-mz.120-5.WC8.bin . The c2924-1 is 
> > still running
> > c2900XL-c3h2s-mz.120-5.WC2.bin . 
> 
> 
> Do you have an NMS or monitoring system tracking these devices ?  Are
> the switches sending traps or data to a syslog server with all detail ?
> Is there anything in the log data ?  Any crash info, etc ?  Does the NMS
> report anything ? 
> 
> Although you may interpret the above results as one switch booting 44
> hours later than the other [which would be one hell of a delay and
> mostly unimaginable], one could also think they were both originally up
> and one has since rebooted... 
> 
> Collecting the data and monitoring them live vs random show ver checks
> should help with the reasons why and how
> 
	I do have syslog server set on them, and combination of NMS/monitoring
happening. So I know when they go down and then come back up. They don't
send traps, but do syslog anything that happens. Nothing in the logs about
crashes, all claim power on.

	Nope, if the other came up and then re-crashed, I'd know. I'd get an
up event in the NMS, and then another down event (Or multiples). So its 
not like it comes back and re-crashes later. As soon as I get the down
notifications I sit there and watch the one unit come back and an up notification,
and I ping the other unit for up to an hour afterwards with no response.

	I am collecting it live, I have all the data. I only showed the 
"show vers" to show my point more.

	But in general, yes, there are NMS and monitoring running at 2
locations against the same equipment, and neither shows the c2924-2 
coming back up at all during that 44 hour gap.

			Tuc