I'm having a real weird problem all of a sudden on my redhat machine.


I have some old backups of +/- 700MB per file that i removed and then after about 5 minutes the machine gets real slow and eventually freezes up, this might take 5 to 15 minutes and then the machine responds again.

I've looked at 'top' and i see that the RM command stays listed in there, consuming 5-15% of my cpu. If i removed more than one file there are more RM's in there.

The top process it kswapd taking 20-50% cpu usually.

I've googles a lot already but i haven't found a satisfying answer so far.

Could this be a disc problem? Or can i configure something to fix this?

I've looked at /var/log/messages and there is nothing unusual there.

I'd be real suspicious of the hard disk. I'd also make certain that I didn't have any valuable data on it.

Can you do an strace on the rm process and see what it's doing?

What does the command strace do then?

Can i do a HD test?

The HD's are in a RAID 5 system, so how can i see which disc is bad?