[c-nsp] Cisco 6509 reboots on its own... again...

[j.vaningenschenau at utwente.nl]

Hi,

> ==> Are you implying that a slight temperature raise of 3 degrees was
> enough 
> to provoque the dilation of the line cards inside the chassis ?!?
> 
> Hmm.... Curious, I noticed a slight increase of our DC's temp between
> last 
> saturday night and sunday morning for a few hours (approx 10h) :
> 
> - the SUP' temp went from 22°C to 24°C
> 
> - the chassis' temp went from 26°C to 28°C
> 
> Hmm.... It's a possibility but the odds are fairly small compared to a
> software bug, don't you think ?

To chip in here: over the last 6 weeks we had two crashes on a Cat6k Sup720-3B. Device was stuck in rommon afterwards; "reset" followed by "boot" hung the box, so we had to resort to full power down / power up. Did a reseat of the sup after the first crash, but it crashed again within two weeks after that. About the same conditions as in your case: IOS SXI2a, "software fault", crashes happened while the temperature was changing a few degrees.

So it's either:
* a software bug that we didn't hit for several months and then two times in a short period of time (and only on one of our two parallel routers)
* temperature-sensitive hardware issue
* other hardware failure

Does that change your perception of the odds? ;)


Regards,

Jeroen van Ingen
ICT Service Centre
University of Twente, P.O.Box 217, 7500 AE Enschede, The Netherlands