[nycbug-talk] Help!!! (server problems...)
George at sddi.net
Thu May 27 10:11:37 EDT 2004
A wise person once said. . . Kit Halsted
> At 7:21 PM -0400 5/26/04, Jesse Callaway wrote:
>>On May 24, 2004, at 10:43 AM, Kit Halsted wrote:
>>>Thanks, Jim. Heat was one of the first things to occur to me, & one
>>>thing I did last time I was there was cleaning the filter on the
>>>server room AC. (Dropped ambient temp by about 15 degrees! That
>>>filter was nasty...) Still, it never made sense to me that heat
>>>problems would happen when the box was close to idle; I would
>>>expect crashes at peak times rather than at 7:00 AM when the office
>>>is empty & no one's hitting the web db...
>>hmm... maybe it wants coffee.
> No, that I'm sure of! Same client, a year or 2 back, I get a call
> about a PowerBook not working: turns out it took about half a grande
> latte in the keyboard. Nice piece of design, that keyboard: it kept
> the liquid away from the important bits.
Ike has some insight about Dos Equis and PowerBooks. . .
>>Regarding what Jim suggests, check the BIOS and see if it has a
>>thermal panic feature.
> I have to look all through the new BIOS when I can get back there, but
> that could be it.
I think it's easy to forget the importance of cooling and computers.
That is one of the threshholds of speed, right?
I've also spent a lot of time dealing with heat issues that weren't
apparent. Some years back, I was at a firm with a number of Florida
offices. I was working on a Cisco router remotely as the WAN connection
was acting very flaky. One thought was that the router had been
compromised. When I had a consultant go onsite he was hit by a blast of
heat when he entered the server room. I had lost an argument with upper
management to spend $6k on decent HVAC in the room not too long before
that. Needless to say, the hardware and labor costs hit that in one
More information about the talk