Linux: very slow mdadm resync on raid 1

mdadm

I'm trying to rsync a RAID 1 on a system with absolutely nothing running ( i've moved all services to another server ). So it's just me via ssh. For some reason, it's very slow:

cat /proc/mdstat

Personalities : [raid1]
md1 : active raid1 sdc1[1] sda1[0]
      104320 blocks [2/2] [UU]

md2 : active raid1 sdd1[1]
      1953511936 blocks [2/1] [_U]
      bitmap: 10/233 pages [40KB], 4096KB chunk

md0 : active raid1 sdc3[1] sda3[2]
      240814272 blocks [2/1] [_U]
      [>....................]  recovery =  1.0% (2488576/240814272) finish=8243.9min speed=479K/sec
      bitmap: 94/230 pages [376KB], 512KB chunk

unused devices:

I already tried:

  • Setting /proc/sys/dev/raid/speed_limit_min to a very high value
  • Setting –bitmap=internal

None of these worked.

I also run smartctl on all HDs, no errors were detected. They seem fine. Hdparm also reports an average speed of 4000MB/s for reads. Which seems acceptable.

Any ideas of what might be happening?

I'm running centos 5.5 and my kernel is 2.6.18-274.7.1.el5.028stab095.1.

Best Answer

Probably a silly question, but have you verified that the speed_limit_max is also sufficiently high? 4000MB/s would be blindingly fast, that must be the cached speed not the actual disk speed.

Related Topic