Eric, sorry for the late reply.
On Thu, 7 Sep 2000, Eric Youngdale wrote:
> The oddness is this. We were observing stalls in the processing of
>commands that was traced to the fact that the queue had remained plugged
>for an excessive amount of time. The stalls last for about 5 seconds or
>so.
If during the stalls the disk is writing to disk at full speed (not idle),
then I agree the right fix is to unplug the queue before putting kflushd
to sleep (after all it makes sense to do the I/O while the disk is
otherwise idle).
However I'm also worried by the size of the queues, maybe the high latency
is a sign that the queue is too big (5 seconds at 20mbyte/sec are
100mbytes of stuff into the I/O queue, are you really waiting 5 seconds
while the disk is writing?).
Andrea
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Sep 07 2000 - 21:00:32 EST