Bandwidth/CPU Usage has me completely Stumped

[johnsee]

Hi,

I've used every tool and resource I can think of, but haven't managed to get to the bottom of this problem. I have WHM running on a decent sized linode. At irregular intervals normally lasting a few hours CPU is spiking to over 100% (which the linode allows) and bandwidth is going up to a sustained Mb/second.

The VPS manager reports that so far this month, it's 567GB (split roughly evenly between up and down). This number is also reflected in the number I see in bandmin (531GB). However, looking at "View Bandwidth Usage" and individual account, they barely even reach 80GB combined.

I'm in control of every account on the server, and don't offer web hosting, so it seems that someone could be doing something dodgy. The spikes in traffic are almost always equal in both incoming and outgoing, as though whatever is running is uploading and downloading an equal amount.

mod_userdir protection is enabled for all account
Almost all sites on the server are wordpress, are up to date, and almost all run supercache or similar

I'm completely out of ideas to find the source of the problem... help!

[techs4gnu]

Hello,

Do you have the exact time frame when the bandwidth spike occurs? Else install mrtg / cacti to monitor the traffic on the server and find it! When you find the time frame, login to the server and check the server traffic using a utility like iftop ( if not installed, install it using yum ) and see what is consuming all the bandwidth.

Thank you.