On Wed, Jun 15, 2011 at 3:13 PM, <m.roth@xxxxxxxxx> wrote:
> Lamar Owen wrote:
>> On Wednesday, June 15, 2011 09:12:44 AM Brunner, Brian T. wrote:
>>> Damage to circuitry is not all "instant-or-never"; damaged junctions can
>>> take their own time (sometimes zero) to degenerate from
>>> damaged-but-perfectly-functional to occasional errors to persistent
>>> failure.
> <snip>
>> Dust isn't as bad of an issue as you might think, but stains are. And I
>> do the vacuuming of the data center spaces myself, with a dedicated vac
>> with HEPA filtration. Takes less than half an hour for the critical
>> spaces, and gives me a good reason to inspect everything.
> <snip>
> Ever heard the old, old m'frame (I think) story, of the guy who needed to
> do a backup, and the tape failed, and they had to go to an older one. The
> next few days, he was trying to find out why, and discovered all of the
> tapes on the bottom row of the tape (reel) rack were bad. Stayed late one
> day, working on it... and watched as the cleaner came in, and ran the
> floor cleaner right up to the rack....
And the apocryphal mystery why the servers rebooted at 6:30 every evening.
Having said that we once had a cleaner unplug a router to plug a kettle in,
but that's rather more likely because it was tucked away in a corner than
unplugging a whole computer. Was back in '85 as well so your average
cleaner had no clue what a computer was.
On the other hand I managed to prank the cleaner in our computer lab. We
had 20 or so computers with touch screens and speech synth back in '85
when they were both magic. I left a program running one every one of them
so say 'thank you' when she finished cleaning the screen. We had the
cleanest screens ever.
_______________________________________________
CentOS mailing list
CentOS@xxxxxxxxxx
http://lists.centos.org/mailman/listinfo/centos
