[lug] More Server Problems

[George Sexton]

> -----Original Message-----
> From: lug-bounces at lug.boulder.co.us 
> [mailto:lug-bounces at lug.boulder.co.us] On Behalf Of David L. Anselmi
> Sent: Wednesday, February 15, 2006 7:38 PM
> To: Boulder (Colorado) Linux Users Group -- General Mailing List
> Subject: Re: [lug] More Server Problems
> 

> Really, check the disks before you reinstall (the machine 
> will be down 
> anyway).

Unfortunately, this is my web server and I can't have it down. Taking the
system offline will mean buying a new machine and installing it.

I've seen disk problems on this machine actually. One of the mirrored set
was bad when I first installed it. The messages were totally different, and
the software raid system dropped the faulty drive out of the array. I put in
a new drive (that I did test thoroughly). This doesn't look anything like
that drive failure problem. Also, since I'm using software raid, the raid
control software SHOULD just drop the faulty drive (as it did before).


> 
> Did you burn this in well before you sent it to the colo?  Maybe a 
> lesson for next time...
> 
> Dave

Well, I really am that sophisticated. Its just the vendor was two week late
delivering the equipment and he ate up my burn in time.

The real lesson for next time is to not buy new state of the art technology.
Every time I do I get burnt. I remember 5 years ago the kernel just puked if
you had an SMP machine. Over time the kernel issues got sorted out and I
have the machines that used to puke daily in production now with uptimes in
the hundreds of days. The other machine that I bought at the same time is
the same vendor, but a normal P4 CPU. No problems rock solid, up over a
hundred days now.

The linux kernel just doesn't stablize new technology like EMT-64 or SMP
very quickly.

George Sexton
MH Software, Inc.
http://www.mhsoftware.com/
Voice: 303 438 9585