[c-nsp] Mysterious 2924 reboots

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

> 
> On Sun, 5 Aug 2007, Tuc at T-B-O-H.NET wrote:
> 
> > 	And both units do have "no ip http server". Should it be
> > an issue?
> 
> There are some well-documented bugs in some older spins of IOS where 
> people can do fun things like crash/reload the switch if the built-in web 
> interface is enabled.  The way to guard against many of those bugs is 
> to disable the web server, which you have done.
> 
	Its just odd that they both do it at the same exact time.
One 2924 is the one connected to all the end devices, and thats the one
(FORTUNATELY) that comes back quickly. The other is supposed to be
a warm spare. It only has a X-connect over to the 3640 and nothing else.
THAT one is the one thats taken up to 3 days to "reappear". 

	If I'm having power issues, I'd think my Soekris would be
pretty prone to it too, but thats been fine. I have a UPS at the site
for other equipment (LONG story) and I've checked twice and while 
there was a power "issue" earlier this morning that it logged,
none of the devices other than the 2924's seemed to have any qualms
about it. 

	If they were just REALLY power sensitive, I could deal with it.
But the fact that one of them has taken DAYS to come back is even
more mysterious.

			Tuc