[<prev] [next>] [day] [month] [year] [list]
Message-ID: <hdo5u6d3rqfo7v349262i3854gkde4re10@4ax.com>
Date: Mon, 30 May 2011 10:17:50 +1000
From: Grant Coady <gcoady.lk@...il.com>
To: linux-kernel@...r.kernel.org
Subject: No /proc/diskstats accounting for loopN or md...p. disks??
I built a RAID box, 6 x 1TB SATA drives, RAID 6, RAID10, runs well, but
I notice that /proc/diskstats is not accumulating numbers for loop[0-7]
devices, nor does it grab I/O numbers for partitioned mdadm RAIDs.
For that second, I partition the drives, and several RAIDs using different
partitions are fine, but if I make a RAID6, then partition that device,
then the eg. md125p1 device will not get stats collected, they all go to
the parent device. My 'cure' for this one is 'don't do that'.
Loop devices simply show all zeroes on /proc/diskstats. No cure?
So what's the spell to get full /proc/diskstats accounting? I looked in
/sys/... and the same dev gets all zeroes in there too, so it looks as
if the numbers are not simply getting calculated.
I'm running 2.6.38.7 on Slackware64-13.1 but this happens on earlier
version kernel. Be upgrading to Slackware64-13.37 soon, but I doubt
it's user space...
Happy to do directed testing for you. Also if you tell me it doesn't
work like that, that's okay too, but then loopX should disappear from
/proc/diskstats, no?
Thanks heaps for info, ideas,
Grant.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists