I've been reading the threads about kernel panic here, and I'm not really sure what it means yet. I've never had one before. Let me know what you think of this:

I had a RAM upgrade on one of my servers tonight, and within 30 minutes or so, the server crashed. I had it rebooted and the tech said there was a kernel panic. Do you think it was caused by the RAM? If so, is it going to continue crashing the server until I replace it?

I looked through the logs and don't see anything pointing to it... Traffic on the server was VERY light at the time.

Thanks for your thoughts...

AtoZ,

From the way you have put it, it sounds like the RAM could be fault, ask for a replacement and see how it goes from there.

However, it could be just a coinsidence that the day you get RAM put in your server gets a kernel panic.

Kernel panics in my opinion is where you either your kernel is unable to access something at boot time, therefore is unable to perform an operation needed to boot the system, or faulty hardware and your kernel doesn't want to disk mounting hard drives etc.

I would suggest the change in memory modules, see how that goes.

DislexiK

Since I posted this message, I had numerous server errors, and eventually a segmentation fault and the server crashed again. This server never had any problems before except that it would occassionally lock up due to inadequate memory.

Server is down now 90 min. since the segmentation fault. I just got a ticket update that the RAM was replaced, but it's not back up yet, so I don't know if it has resolved the problem.

Maybe this issue just now caused a problem because it's the first time the server has been taken down completely since it went up almost a year ago?

Originally posted by AtoZ
Since I posted this message, I had numerous server errors, and eventually a segmentation fault and the server crashed again. This server never had any problems before except that it would occassionally lock up due to inadequate memory.

Server is down now 90 min. since the segmentation fault. I just got a ticket update that the RAM was replaced, but it's not back up yet, so I don't know if it has resolved the problem.

Maybe this issue just now caused a problem because it's the first time the server has been taken down completely since it went up almost a year ago?

Sounds like you were rooted if this is the first time you took it down. Lots of kernel upgrades and such. (guessing its linux)

I am supporting thelinuxguy . What the version of kernel you are using ? I think it is 2.4.20-2.4.22 , and it is LKM infected. Try installing new kernel and then boot on 2nd new kernel.

I doubt its just a kernel issue. His system is probably mass infected. Critical binarys are bad it sounds like.

Thanks for all the replies. So what I'm hearing is that the server may have been hacked at one time, and I'm just now seeing the effects due to bringing it down? It has been rebooted about 5 times or so since Sept. 20... Zero times from when the server went up in Dec 03 through September.

When the server was finally brought up again after my last message, I upgraded the kernel. The server was running just fine for several hours, and since then it has completely crashed again twice. Currently it's completely down because it didn't come back up after a reboot. I'm waiting to hear results of another reboot.

Thanks for any insight. I have never had a problem like this before, so I'm clueless.

by the way, I'm using redhat 9... I upgraded the kernel with up2date. I don't even know what version it is. (I really don't know anything about this!)

Have the extra memory taken out entirely and see if it fixes the problem. I've had similar problems in the past that were caused by faulty memory, but it could just as easily be the board in the machine. If the problems remain then I'd entertain the idea that the server is compromised as suggested.

Brian

I had the initial RAM removed, and they supposedly put in a different stick this morning. The server stayed up a lot longer with the new stick, but I'm still having similar problems.

How much ram was initially on the server ? how much ram was added ? Have you had the DC run memtest on the sticks ? Are they the correct type or ram ?

I had 512 MB RAM before and 512 was added. I have no idea what type of RAM it is. I would assume that they would know what the hell they are doing, but it's hard to come by that these days without having to second guess it.

run chkrootkit or rkhunter on the box

Okay, they got the server back up again, so I left apache down and ran the tar command on one of my data directories so I could move it to another server just in case. I encountered a segmentation fault when doing this.

I didn't see your post until after I made mine linuxguy... I have rkhunter on one of my other servers, so I'll install that here if I can ever get it working long enough.

I am going to have them remove the "new" memory as well. What a waste of time.

Just an update for anyone with a similar problem. There are so many possibilities. I don't really have a definite solution. It ended up that the new memory didn't change anything so they installed a new kernel. The situation improved, but the server crashed again so they put my hard drives in a new chassis and now all is well. :) (rkhunter also showed everything was clean)