VirtualminGPL Odd Server Lockup

Hi folks.
I’m not sure if this is the right place to post this thread, so apologize if I’ve mistakenly posted in teh wrong place.
I’m a new user of Virtualmin GPL addition, using CentOS6 on a dedicated server, with 2GB of RAM (maxed out as far as what it can hold RAm wise) with a 500GB hard-drive.
I’m trying ot track down an issue as to why the server suddenly grined to a halt, so much so that when the datacenter staff member tried to access the console of the server, only control-alt-del responded apparently for a reboot.
At the time, I had spam server scanning and mail filter scanning going, so the first thing I did was to make Spam assassin and virus scanning standalone programs.
I tuned Apache to run a max of 5 servers, the primary amount of servers being 3, and set MySQL to use about 512MBusing the wizard.
I have Apache as the primary core, with Postfix, Mailman, etc running, (without any other modifications.)
I also run DNS with named.
So I’m trying to figure out what might have used up so much RAM?
My guess is that is what made the system crash.
I only have squirrelMail installed on one virtual host at the moment, as well.
I do not wish to leave my spam filtering or virus scanning to a third party such as GMail, however.
I’m trying to track down what might be causing the issue, and right now I have no idea.
Another interesting thing I’m running into is that due to the default bind configuration, wich I haven’t touched either, I’m getting lots of DNS resolving attempts and “network unreachable” messages in /var/log/messages.
Do you folks have any suggestions as to what I could look into or tune values wise to try and reduce RAM useage?


The DNS issues I wouldn’t worry about too much – those are likely just lookups being done of spam that you’re receiving. If it contains a forged “from” address, the DNS lookup could generate an error or warning.

And for “Spam and Virus Scanning” – I’d recommend using “spamc (Client for SpamAssassin filter server spamd)” and “Server scanner (clamdscan)” – those are the fastest and least resource intensive, for folks with as much RAM as yourself. If you had 512MB of RAM or less, you’d need to use the Standalone scanner, which requires less RAM.

But as far as the lockups go – the key is going to be to figure out what’s going on when those problems are occurring.

Do you see any errors/warnings in your logs at around the time these lockups occur? That could be in /var/log/messages, or maybe /var/log/secure.

Also, what do the commands “free -m” and “mailq | tail -1” show?


Erick, hear is what you asked for command-wise.
Command free -m (due to some of the modifications I’ve done temporarily):
total used free shared buffers cached
Mem: 2013 660 1353 0 36 312
-/+ buffers/cache: 311 1702
Swap: 2047 0 2047
Now for mailq | tail -1:
Mail queue is empty
Hope that helps.
Feel free to let me know if you need any other information.