CPU spike / Linode Crash
This is the second time this has happened to my server in the past month and I'm starting to wonder if its something to do with my server config.
Server is a 720 running a medium loaded php/mysql website / Nagios / Teamspeak 2 server.
From what little logs there are and user reports there was only about 30 people on TS during the time, but that uses next to no memory as is.
Memory:
~~
CPU:
~~
VMstats:
~~
Mysql Thread:
~~
It looks like my mysql server went through the roof causing all active memory to be sucked up causing swap to fill up until the OOM killer diced the server. But I can't find or don't know where to look to verify that.
Suggestions?~~~~
2 Replies
@MrRx7:
It looks like my mysql server went through the roof causing all active memory to be sucked up causing swap to fill up until the OOM killer diced the server. But I can't find or don't know where to look to verify that.
One would imagine that mysql was being called by another process, ie apache/php.
Check the logs of your website to see if it started getting a heap of traffic coinciding with mysql going through the roof.
You might like to ensure that mysql has some sensible limits placed on it so it can't use all the memory of your Linode.
@josephb:
@MrRx7:It looks like my mysql server went through the roof causing all active memory to be sucked up causing swap to fill up until the OOM killer diced the server. But I can't find or don't know where to look to verify that.
One would imagine that mysql was being called by another process, ie apache/php.
Check the logs of your website to see if it started getting a heap of traffic coinciding with mysql going through the roof.
You might like to ensure that mysql has some sensible limits placed on it so it can't use all the memory of your Linode.
I looked at yesterday access logs and they looked normal, no odd spikes or anything
I have mysql configured to use a max of 370mb with apache limited to 150. I suppose I will keep looking.