software raid resyncing

software raid resyncing

[urgrue]

ive set up a software raid. md0 and md1 are RAID-0 (two identical SATA 
drives on each) and md2 is a RAID-1 consisting of md0 and md1. 
it works, but:
Personalities : [linear] [raid0] [raid1] [raid5] read_ahead 1024 sectors
md2 : active raid1 md1[1] md0[0]
       490223168 blocks [2/2] [UU]
       [========>............]  resync = 43.6% (213810348/490223168) 
finish=18185.0min speed=252K/sec
md1 : active raid0 sdd1[1] [dev 08:21][0]
       490223232 blocks 32k chunks
       md0 : active raid0 sdb1[1] [dev 08:01][0]
       490223232 blocks 32k chunks

and mdadm --detail /dev/md2  shows md2:
           State : dirty, resyncing
  Active Devices : 2
Working Devices : 2
  Failed Devices : 0
   Spare Devices : 0
  Rebuild Status : 43% complete

question 1: why is it dirty? i just set this up a few days ago. should 
i be concerned, is something faulty? the individual RAID-0 drives md0 
and md1 report nothing out of the ordinary.

question 2: why is it piddling along at 252K/sec? at this rate resync 
will take weeks.

Re: software raid resyncing

[andrew mathes]

From my experience this is caused by some (or all) drives not being in 
dma.

run hdparm /dev/hd* (sd*) and make sure using_dma is on, you can also 
check what kind of dma is enabled with hdparm -i <device>

if it's not, give hdparm -d1 -c1 <dev> (d for dma and c for 32-bit) and 
hopefully it will turn on.  you can run hdparm -tT <dev> before and after 
you do this to test the difference.

If dma doesn't turn on most likely you haven't compiled support for your 
board into your kernel, so if hdparm fails, give that a look

as for WHY it's dirty, either it has never synced in the first place, or 
it was unmounted unclean, happens to me most by power failure (or me 
kicking my power supply)

On Fri, 18 Feb 2005, urgrue wrote:

> ive set up a software raid. md0 and md1 are RAID-0 (two identical SATA drives 
> on each) and md2 is a RAID-1 consisting of md0 and md1. it works, but:
> Personalities : [linear] [raid0] [raid1] [raid5] read_ahead 1024 sectors
> md2 : active raid1 md1[1] md0[0]
> 490223168 blocks [2/2] [UU]
> [========>............]  resync = 43.6% (213810348/490223168) 
> finish=18185.0min speed=252K/sec
> md1 : active raid0 sdd1[1] [dev 08:21][0]
> 490223232 blocks 32k chunks
> md0 : active raid0 sdb1[1] [dev 08:01][0]
> 490223232 blocks 32k chunks
>
> and mdadm --detail /dev/md2  shows md2:
>         State : dirty, resyncing
> Active Devices : 2
> Working Devices : 2
> Failed Devices : 0
> Spare Devices : 0
> Rebuild Status : 43% complete
>
> question 1: why is it dirty? i just set this up a few days ago. should i be 
> concerned, is something faulty? the individual RAID-0 drives md0 and md1 
> report nothing out of the ordinary.
>
> question 2: why is it piddling along at 252K/sec? at this rate resync will 
> take weeks.
> -
> To unsubscribe from this list: send the line "unsubscribe linux-admin" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>