performance bottleneck in Linux MD RAID-1

James Ralston qralston+ml.dell-poweredge at andrew.cmu.edu
Thu Jul 15 13:10:00 CDT 2010


On 2010-07-14 at 20:32-07 Bond Masuda <bond.masuda at jlbond.com> wrote:
> We're seeing a performance bottleneck of about 200MBytes/sec
> sequential writes when testing with iozone.  We were expecting with
> 7x effective spindles on the RAID-10, to get about ~350MBytes/sec
> sustained writes for sequential access.

200MiB/s is the default value for /proc/sys/dev/raid/speed_limit_max,
so I'm wondering if the md driver is counting non-recovery RAID1
syncing against speed_limit_max.

Do you see better performance if you increase speed_limit_max?

(If you increase speed_limit_max, make sure that the instance-specific
value in sysfs isn't being set separately--see Documentation/md.txt
from the Linux kernel.)



More information about the Linux-PowerEdge mailing list