I think I may be overreacting to a monitoring issue. The CPU utilization
stays around 100% and the load jumps up for around 30min and then drops
back down. Obviously this amount of time will vary, but as I roll this out
to the rest of the environment, I want to minimize the alerts coming out.
I wa
The "nice" utility is your very best friend. It yields CPU time to other
operations but will run like crazy of nothing else is a higher priority. Clam is
a disk IO heavy process for obvious reasons, and can drive disk waits up quite
high. It is also CPU intensive but should occupy a single core.
What's "high cpu" in this instance ... you should expect system resources
to be consumed when the on-demand scans run, are you seeing high load
averages, what are you using to diagnose high cpu, is it simply a per core
spike?
On Jan 26, 2016 13:27, "Jeff Johnson" wrote:
> I have rolled out clamd
I have rolled out clamd to a handful of Red Hat systems and they all seem
to have high CPU usage when clamdscan runs at 2am. The rest of the day,
clamd drops down to minimal usage. There isn't a lot of change on these
systems so I can't imagine it's finding much, but it really pegs the CPU's.
What
