[TriLUG] Odd problem with RH8.0 box

[Brian Weaver]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Peter,

You might want to try badblocks in NON-DESTRUCTIVE (read) mode from a
rescue disk just to verify that the disk is not going bad. Just because
fsck doesn't show any errors doesn't mean there isn't a problem.

A successful babblocks read-mode test it not a guarentee either. I've
had disks that passed the read-mode, but failed the destructive write
mode test. However if your disk is having problems that are manifesting
as a system lock then the read-mode will probably catch it.

If you do use the badblocks program backup your data. One bad flag and
you data is toast!

- -Weave

P.S. badblocks is usually located in /sbin and is part of the e2fsprogs
RPM on Redhat 9.


Peter Long wrote:
| Yikes, I hope that is not it! I don't see any errors reported by fsck
| either. I have a spare PCI IDE controler I could try instead of the
| onboard one.
|
| I recently installed APCUPSD to monitor my APS Back-UPS ES. I wonder if
| it could be messing with the APC stuff causing the disks to be powered
| down. How would I check if that is that case...
|
| --
| Peter Long
|
| Jon Carnes wrote:
|
|> I second Jason's theory. I've run into that exact problem before with
|> boxes having disk problems.
|>
|> When booting, does the fsck show any bad sectors or problems that it
|> can't fix?
|>
|> Jon
|>
|> On Wed, 2004-01-28 at 21:51, Jason Tower wrote:
|>
|>> it sounds like a problem with either the hard drive or the disk
|>> controller.  anything that accesses the disk is hanging.  unlike
|>> logins, mail, ssh, etc, routing and iptables don't generally require
|>> disk access once they're loaded into memory so they aren't failing.
|>>
|>> of course, that's just my opinion, i could be wrong.
|>>
|>> jason
|>>
|>> On Wednesday 28 January 2004 21:38, Peter Long wrote:
|>>
|>>> Hi all,
|>>>
|>>> I have been experiencing an intermittent problem with my trusty Linux
|>>> box. Unfortunately I can't figure out the cause. The machine is
|>>> running RedHat 8.0 and is a PII 350 with 384MB of RAM. It also has
|>>> more then enough RAM. The box serves as my gateway to the Internet.
|>>> It has two NICs and I am running IP Tables on it. I also use it as my
|>>> mail server.
|>>>
|>>> The symptoms of the problem are that every now and then it seems to
|>>> "lock up". By that I mean that I cannot login at the console, neither
|>>> via X or tty. The funny thing is that it still works fine as a
|>>> router. My internal machines can still send traffic through it after
|>>> it has locked up. At first I thought it was a problem with PAM since
|>>> it locked up when I tried to login. Incidently IMAP from remote
|>>> machines would also stop working as well as new SSH connections.
|>>> Named also stops responding. I tried testing my theory that it was a
|>>> PAM problem by leaving one tty console logged in after rebooting the
|>>> system. However when the problem re-occurred I discovered that I
|>>> could not do an 'ps' at the bash prompt. So it seems that you cannot
|>>> launch new processes either.
|>>>
|>>> There are no funny errors in the logs to indicate what the problem
|>>> could be. Actually it seems to stop writing to the log entirely. I
|>>> usually get an entry in the 'messages' log every 30 minutes from
|>>> iptables for BOOTP traffic that is dropped. These stop occurring.
|>>>
|>>> It does not seem to be a disk space or ram problem. I have gig free
|>>> and 230MB of the 384MB is unused.
|>>>
|>>> Any clues or tips would be greatly appreciated.
|>>>
|>>> --
|>>> Peter Long
|>
|>
|>
|
|
|
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFAGVgyKiU2VlicfIwRAvswAKDMVlyIUkvEIHsuUFoQWw1bEeKgUACeKF/u
21PUKUIYuZBo6hO+UpnM96k=
=xU+B
-----END PGP SIGNATURE-----