High Server Load Logs

Hi.

I receive an e-mail from NodeWorx somtimes warning me about high server load. I’ve noticed that bandwidth activity has been going iffy (it was raised to 66GB in less than a day), so I know there’s something going on, and I’d like to investigate. Problem is, I don’t see this from within NodeWorx > System Logs.

Where do I start searching for things related to this scenario? I’m not very familiar with how InterWorx logs things.

Thanks.

Hi Liam

Hope your well

I have always found the most useful is to login ssh when notified, and run top

This usually shows the culprit, service etc…

There is also the Apache server status, which is useful for identifying where bandwidth maybe used the most. In general, we have found some host scripts appear to keep open and serving pages, and once we know host siteworx name, have used the siteworx stats to see which files it relates too, ended them and notified client.

I have based this on assumptions as you stated bandwidth.

Usually though, high loading we have seen in the past, is result of backup been run, or restore of course.

We use r1soft now for backups, no server loading on backup or restore, and added benefit of backups been held in another country.

Also, do you run BFD, meldect and rkhunter. If not, I would advise you do, and in particular for word press sites, which are constantly under login attack, which BFD will stop and your loading/bandwidth goes down a lot.

I hope that helps

Many thanks

John

Hi d2d4j.

I have no WordPress sites running (most of the sites running on my server are a couple of installations of Ghost blog and a MyBB forum). I don’t think this is related to backup, seeing the sudden spike in bandwidth usage.

I would’ve run top, but I usually get the e-mail at night time so I log in to the server in the morning to see what’s happening. Every time I log in it seems to be fine.

Any idea where these logs are at, specifically which ones do I look at in /var/log ?

Thanks.

Hi Liam

I would certainly look at dmessage, and Apache logs, to see if anything is obvious.

I read somewhere about sar, but never used it myself and if not installed, I’d yum install and read man page

Also, have you looked at monit or external monitoring like nagios perhaps, we have WUG, what’s up gold

I’ll have a little think though if you don’t mind

Many thanks

John

Thanks. I’ll look into Nagios and see what I can get from there.