Re: ADMIN: Wonky server

Top Page
Attachments:
Message as email
+ (text/plain)
Delete this message
Reply to this message
Author: Kevin Brown
Date:  
To: plug-discuss
Subject: Re: ADMIN: Wonky server
> This guy has been having debilitating problems with this server for weeks if
> not months.
>
> I think it's a mission critical server not just for the fan list that I know
> him from but in a for-profit business. The fact that he's trying to solve it
> himself makes me strongly suspect that the business is mortally
> undercaptialized, but if anyone has useful advice, he no doubt would
> appreciate it. (Especially if it isn't "find $5,000 and buy a new server,
> then find $10,000 and hire the services a certified, experienced Red Hat
> admin ... namely me.)
>
> http://plug.phoenix.az.us/index.php
>
> =================================================================
> On Wednesday 2004-04-28 19:38, Nick Arnett wrote:
>
>>We're running on a somewhat crippled server because a few of its
>>essential binaries are seg faulting, suggesting that we're building up
>>to another disk failure, perhaps. It's been up and down today as I've
>>been trying to fix the broken parts.
>>
>>If anybody here has any insight or suggestions on how to deal with a
>>system in which grep, chmod and ln are all seg faulting, I'd love to
>>hear them. I never realized how essential grep is to Linux. And
>>without chmod, I can copy grep from elsewhere, but then can't make it
>>executable!
>>
>>Sadly, there's no "reinstall everything" that I can come up with, and
>>individual re-installs seem to be failing. This system is running RH 9,
>>a/k/a Fedora.
>>
>>BTW, it's running now because I can boot from a system CD, then chroot
>>to our normal system and start everything manually. Ugh.


I don't see a contact address, so I'll just send this to the list.

First, I would utilize rpm to verify that the binaries on the drive are what rpm
thinks it installed

a basic way (but will take some time) is:
rpm -Va

you could also just try:
rpm -Vf /bin/grep

if nothing is spit out, then it means the file matched what is expected, else it
will spit out a line like:
S.5....T c /etc/pam.d/system-auth

the order or those items is:
File Size
Mode
MD5 sum
major number/minor number
symlink string
owner
group
mod time

so, in my case, that file has a mismatch for Size (S), MD5 sum (5) and Mod time
(T). The little c tells me it is a config file.

You can also verify them agains rpm packages (in case you suspect corruption or
tampering of the rpm database itself) with the -p flag.

Hope this helps to begin to see what is up with your system. For me, the system
doesn't segfault so much as just have its load skyrocket and become unresponsive
when a system drive is dying.
---------------------------------------------------
PLUG-discuss mailing list -
To subscribe, unsubscribe, or to change you mail settings:
http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss