For the last several days-- I guess ever since I rebooted into the latest kernel (running Centos 5.3 32-bit) -- I have had a bizarre CPU usage plot (attached).
Disk I/O is normal, network traffic shows slightly elevated but nothing is showing up in any logs to indicate any malicious activity, and the system is generally pretty responsive. I do have a "heartbeat" script that sends me a text message every 30 minutes with the output of "uptime" and the load average peaks at around 2.5.
I checked Cron for anything that runs at 8-hour intervals (or at the times shown on the plot) but I can't find anything that would account for this.
I'm wondering if anyone else is seeing anything similar, or might have any idea where to look for answers? It's not impacting performance (that I've noticed-- and since VoIP runs through this box it's pretty load-sensitive) but I'd like to figure out a potential cause.
I am on atlanta51.
