I occasionally have seen similar behavior with my RAID 6
array, it
almost always has been a problem with an individual disk.
If one disk
is misbehaving and the IDE layer has to wait for timeouts to
be hit when
writing to the disk it can kill performance. The way I
diagnose
problems like this is with iostat, which on my Debian system
is part of
the sysstat package. If you run something like
# iostat -x -k 5
and watch the await column, you should quickly be able to
see if one of
the disks is causing the wait states.
_______________________________________________
mythtv-users mailing list
mythtv-users mythtv.org
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
a>
|