Date: Wed, 14 Aug 2013 15:50:22 +0200 From: Willem Jan Withagen <wjw@digiware.nl> To: fs@freebsd.org Subject: Fwd: Disk scheduling activity... Message-ID: <520B8B1E.7060002@digiware.nl>
next in thread | raw e-mail | index | archive | help
Hi,
Just a point of information or curiosity, and I don't think/know if it
is any problem...
I have the raidz array with 8 disks, which I'm using to backup to.
It is configured
4 disks on a mvs controller
4 disks on an Areca controller (JBODs with battery)
Both controllers are on a PCI-E slot
dT: 1.366s w: 1.000s
L(q) ops/s r/s kBps ms/r w/s kBps ms/w %busy Name
6 127 4 2 0.1 122 6481 18.8 56.5| da0
4 118 0 0 0.0 116 6646 19.6 56.5| da1
0 124 2 1 0.1 120 6650 19.0 56.6| da2
5 117 1 1 0.1 114 6726 20.0 56.5| da3
4 121 1 1 0.2 118 6895 25.7 70.6| ada0
2 115 1 0 30.0 113 7051 22.1 67.7| ada1
0 113 3 1 11.6 108 7057 26.3 73.7| ada2
3 105 1 0 10.6 103 6728 24.1 66.9| ada3
Most of the time the source just fully loads the pipe and sends 1Gbit/s.
When that happens I see this alternating pattern of writing either to
the 4 mvs disks, or writing to the Areca disks.
But almost never are all disk accesses at the same time.
And really never, never is there a mix of writing between the controller
sets.
I even made a little movie of it. :)
http://www.tegenbosch28.nl./FreeBSD/disk-activity.mov
What is in the disk scheduling that makes it this alternating pattern?
And why are not both controllers writing (flushing?) at the same time.
Changing the 4 ada?/mvs ports to ports on the motherboard behind a
port-multiplier starts increasing the write time per disk, and thus
increases the percentage that the disks are busy. So it is better to
hang them off the mvs controller.
--WjW
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?520B8B1E.7060002>