On Tue, 15 Jan 2002 19:27:54 -0500
"Matt Sales" <[EMAIL PROTECTED]> implied:

> Hello,
> I have recently have put RH7.2 on two machines.  I upgraded an Intel
> Celeron 400MHz that was working great running 6.0.  I also put 7.2 on
> a Pentium II-MMX 266MHz machine that has a brand new hard drive. 
> Everything seemed to be fine; the install went smoothly, etc., but now
> both machines are crashing frequently.  Neither have crashed while I
> work on them, but if they're left alone anywhere from 1/2 hour to 3
> hours, they crash.  The monitor goes dark and I cannot get any
> responses from them--no pings, nothing.  I have to pull the power cord
> to restart them.  I have been monitoring this list on and off, and I
> have not seen other people with this problem... It's frustrating
> because I've never had this problem with RedHat before (5.2 and 6.0) .
>  Now I have a backup DNS server that won't stay up.  If anyone could
>  point me toward an answer it would be much appreciated.

I've uninstalled 7.2 and went back to 7.1 due to problems such as your
DNS stuff dying. There are ways to keep it working, but it became too
much for me to deal with after finding a new one every day (was over 10
before I dumped 7.2).

One method is to crontab a restart of services that die. Since I
couldn't figure out how often it was happening (and it looks like you
have the same problem) I ran it every hour. Cute if someone is in the
middle of a lookup when it gets restarted.

As for lockups, is there a desktop involved? Any NFS exports from or to
the machine? any processes that go into overload and suck up all ram or
cpu resources?

Is there any clue in the logs about something failing or something
getting killed or anything else about the time the lockups start?

A desktop using KDE or gnome would be suspect IMHO since I've had
problems with both, particularly gnome. Screensavers can also cause
similar problems.

How much ram is installed? It might be too little, or even a bad module
(I had that problem once with a ram module causing lockups; it became
more evident as time went by and I started getting lockups in the middle
of doing things).

One way to try to track it down is telnet into the machine (if possible)
and see if it's really locked up or if it's just bogged down. That might
not be helpfull because you say you aren't getting pings, but if
something is really screwy it might prevent answering pings because it's
using all it has for running whatever is going nuts. If you get an
immediate refusal, chances are it's dead. If it takes a long time before
timeout or letting you in, it's likely something eating all it can get
of the machine's resources. If it times out, you might be able to get in
setting a much longer timeout then use top to see if you can find what's
running rampant.

-- 
Intel: where Quality is job number 0.9998782345!



_______________________________________________
Redhat-list mailing list
[EMAIL PROTECTED]
https://listman.redhat.com/mailman/listinfo/redhat-list

Reply via email to