Hello, linode community,

this is my first post to this forum. here is my problem:

I'm running centos on linode 360. I have several sites which doesn't get more thatn 70GB traffic each month. The "biggest" site I host is an web rss reader. Time to time my centos halts, and I have to login in linode panel to reboot it. I did several things I found in the wiki: apache linode optimization, then I increased the value ot max open files, even stopped the planetplanet python deamon (which I doubt is the problem).

So my question is.. how to debug this halts? After reboot I can't find anything in the logs..

Any ideas will be very welcomed because now I get nightmares that my server might be down

Greetings,
hash h8r

If the server freezes again, try and look at the console through LISH. You will probably get some sort of error message on the console if the kernel panics.

I am having exactly the same problem.

Was you able to fix it?

I am looking for a solution to this problem in the last two weeks... not sure why this started to happen very often, I need to reset the server from Linode account panel.

I am using CentOS 5.3 (final)

Thanks.

I have a Linode 560 box, running CentOS and only need Apache, PHP and MySQL in that server.

Julian

You may need to lower the MaxClients directive for Apache. The correct number depends on what exactly you're running, but one Linode 360 I manage running CentOS, MySQL, and Apache/MediaWiki works well at MaxClients 25. The default for the prefork MPM is 256, which is too large for most Linodes, which tend to be memory-limited.

There is an occasional linode problem. Just the other day my (CentOS) instance froze. Not an application issue, not a memory issue. Nothing on the console. Won't respond to pings. Nothing. Just a freeze. Hard reboot required.

What causes it, I don't know. There's never anything logged. Activity before the freeze appears nominal. Definitely no OOM issue or death-swap.

Which makes me think it's a kernel or Xen issue.

It's infrequent, but still non-zero.

_________________
Rgds
Stephen
(Linux user since kernel version 0.11)

Yeah, and what's even weirder is that the past two days' /var/log/messages was missing. It was only locked up for 3 hours, but missing two days' worth of logs. Something in the disk must have had problems.