I must say I've never seen this in my entire life, so
I wanted to get some input. This is sent to both the
linux-admin and linux-raid lists.
I have a customer/friend (yes, the two-in-one combo that
is often noted for being dangerous =) who recently upgraded
an old SCO setup they had to SCO Openserver 5.0.5. This
included selling him new hardware and the works. This meant
that his old HP Netserver LXe Pro (which is a gorgeous machine)
became a spare server for us to utilize.
Being the avid Linux geek I am, I immediately dumped Linux
on there. The Mylex DAC960 was not supported (the card
is the old 2.x firmware variety, and HP wanted money to upgrade
us to the 3.x series) so I could not utilize the hardware
RAID. Instead, I went with software.
It is a dual-processored machine (capable of 4, only utilizing
two PPro 200's at this time, 512k cache each), so I have
SMP compiled it. So at the current time, it basically is:
linux-2.2.11-SMP with raid0145-990824 with four 2gb drives
in a RAID-5. The SCSI bus is onboard Adaptec 78xx. The
network card is an Intel Etherexpress PRO.
The problem? It locks up. Solid. I've never in my life
seen a Linux box just lock up, with no hints anywhere in
logfiles. On the other hand, I've never gotten my hands on
hardware this "big" (This sucker was $31k retail when they
bought it). The machine is currently virutally unused (other
then qpopper for POP mail); SAMBA is setup on it, but is
currently completely unutilized at this time.
It seems to lock hard every few days. Maybe 3 or 4? I see
no coorelation of activity (ie, users doing something) and
lockups, but am willing to dig a little deeper if someone has
an idea.
I was wondering two things:
A. Are there any known incompatibilities with any of this
hardware? I've seen some mentions of aix78xx, SMP, and
raid causing problem. Is this what I'm bumping in to?
B. Is there anything I can do to figure out WHAT is causing
the hard lockups? Again, no hints in /var/log/messages or
anywhere else. Possibly a serial cable to a dumb terminal
constantly dumping system information?
Any information or clues would be more then appreciated. Replies
directly to the list are more then fine; I subscribe to both.
Thanks.
::: Jason A. Diegmueller
::: Microsoft Certified Systems Engineer
::: 513/542-1500 WORK // [EMAIL PROTECTED]
::: Systems Administrator, Bertke Systems Innovations
