[Gllug] Identifying what processes are waiting on IO

[damion.yates at gmail.com]

On Sun, 7 Dec 2008, Stephen Nelson-Smith wrote:

> I have a system whose load is very high for several hours .  It's a
> quad core machine, and I'm seeing load of around 20.
 
[snipped]
 
> I would like to be able to find out what processes are using the
> disks, and waiting on IO
> 
> I'm using CentOS 5, so I don't have a kernel with IO accounting, nor
> do I have Python 2.5 or 2.6, so I can't use iotop.

Not sure if vm.block_dump requires IO accounting (I suspect that means
compiled with BSD process accounting?).  But I've used the following as
a poor man's iotop:

sudo sysctl -w vm.block_dump=1
for i in `seq 0 30 `; do
sudo tail -100 /var/log/kern.log |awk '{print $7}' |sort |uniq -c |sort -rn |tail -22
sleep 1
clear
done
sudo sysctl -w vm.block_dump=0

You can probably refine this somewhat.

Damion
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug