Seagate SATA RAID 1 problems

jameso · 05-30-2005, 11:00 AM

Hi all,

I have a debian linux box with kernel 2.6.11.8 from kernel.org.

I have 2 new 160GB Seagate SATA drives (ST3160827AS), and have been trying to set this up as a software raid1 after installing the OS on one of the drives (/dev/sda).

I have followed the instructions at xtronics.com/reference/SATA-RAID-debian-for-2.6.html, with mixed success.

Basically everything works as expected, with two of the 3 partitions raiding together successfuly.
mdadm /dev/md0 -a /dev/sdb1
mdadm /dev/md1 -a /dev/sdb2

My problem occurs whendoing an mdadm -a command for the remaining md device:

mdadm /dev/md2 -a /dev/sdb3

This syncs data from /dev/sda3 usually up to 99.9%, but never completes.

dmesg output:

Code:

SCSI error : <1 0 0 0> return code = 0x8000002
sdb: Current: sense key: Aborted Command
    Additional sense: Scsi parity error
end_request: I/O error, dev sdb, sector 124101858
ATA: abnormal status 0xD9 on port 0xE08020C7
ATA: abnormal status 0xD9 on port 0xE08020C7
ATA: abnormal status 0xD9 on port 0xE08020C7
ata2: command 0x35 timeout, stat 0xd9 host_stat 0x61
ata2: status=0xd9 { Busy }
SCSI error : <1 0 0 0> return code = 0x8000002

Eventually it causes the box to lock up, forcing me to reset it.

I thought this might indicate bad sectors or something, so I ran a full scan using Seatools, but this did not detect any problems.

Code:

# mdadm -D /dev/md2
/dev/md2:
        Version : 00.90.01
  Creation Time : Fri May 20 13:28:09 2005
     Raid Level : raid1
     Array Size : 155115520 (147.93 GiB 158.84 GB)
    Device Size : 155115520 (147.93 GiB 158.84 GB)
   Raid Devices : 2
  Total Devices : 2
Preferred Minor : 2
    Persistence : Superblock is persistent

    Update Time : Tue May 31 00:00:33 2005
          State : clean, degraded, recovering
 Active Devices : 1
Working Devices : 2
 Failed Devices : 0
  Spare Devices : 1

 Rebuild Status : 1% complete

           UUID : 92f52439:c5f8b6d4:b4214681:8c9a0db2
         Events : 0.347937

    Number   Major   Minor   RaidDevice State
       0       8        3        0      active sync   /dev/.static/dev/sda3
       1       0        0        -      removed

       2       8       19        1      spare rebuilding   /dev/.static/dev/sdb3

Any help would be greatly appreciated.

Cheers

jameso · 06-02-2005, 08:11 PM

So no one has encountered a similar problem to this?

I still haven't found a solution to the problem, so any help would be appreciated.

Cheers,

James

darkseer · 06-06-2005, 08:39 AM

I'm getting the same problem although my setup is a little different. I have 6 250 GB Western Digital SATA drives in a raid 5 group. I cannot get them out of degraded mode and get the 4th drive back in the array. When I try I get the same problem on shutdown you do. here is my madam output.

/dev/md0:
Version : 00.90.01
Creation Time : Sun Apr 24 08:42:22 2005
Raid Level : raid5
Array Size : 1220979520 (1164.42 GiB 1250.28 GB)
Device Size : 244195904 (232.88 GiB 250.06 GB)
Raid Devices : 6
Total Devices : 5
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Mon Jun 6 09:34:21 2005
State : clean, degraded
Active Devices : 5
Working Devices : 5
Failed Devices : 0
Spare Devices : 0

Layout : left-symmetric
Chunk Size : 32K

UUID : eca8bdf3:4bc13e24:2de80b9c:bbf5ccb1
Events : 0.329953

Number Major Minor RaidDevice State
0 8 17 0 active sync /dev/sdb1
1 8 33 1 active sync /dev/sdc1
2 8 49 2 active sync /dev/sdd1
3 0 0 - removed
4 8 81 4 active sync /dev/sdf1
5 8 1 5 active sync /dev/sda1

/dev/sde1 is the device that got kickout and the array wont let it back in(doesn't appear here). Physically the device is fine. Since it is out of the array I reformatted it and mounted it as ext3 just to check. If I try to hot insert it back into the array I get the exact same problems you do. Did you ever find a solution?

-Darkseer

jameso · 06-06-2005, 08:53 AM

No, not yet.

Currently the degraded array is running, and I'm hoping that the other hard drive doesn't die.

I'm still desperate for a solutions.

Anybody?

James

darkseer · 06-06-2005, 09:16 AM

been trolling other fourms:

http://www.ussg.iu.edu/hypermail/lin...01.3/1029.html

I'm still in the early stages of looking at this but my instincts are telling me this is a problem with the ata-scsi driver somehow. If it is not that then I think it may be some wierdness with the mdadm tools.

-Darkseer

jameso · 06-07-2005, 03:12 AM

Yeh my instincts tell me that my hard drives are fine (they are only a few weeks old), and that its either a driver or mdadm issue.

Unfortunately though, I'm far from a linux expert and have no ideas for fixing the problem.

Any other ideas?

James

darkseer · 06-07-2005, 08:06 AM

My intuition says that the way the spare is reconstructed is perturbing a flaw in teh ata-scsi driver. I just need to keep searching for a patch or a fix. I'll be trolling the kernel dev fourms next to see if there is an instance of this happening somewhere else. I don't see a solution in the near future unless we're really lucky.

-Darkseer

jameso · 06-07-2005, 09:07 AM

Yeh.

If i can't get this to work I might be forced to try Fedora instead or something. Would fedora use the same driver or a different one?

Cheers

James

jameso · 06-16-2005, 10:25 PM

Ok i did an apt-get update and an apt-get dist-upgrade this morning, and noticed that mdadm was one of the tools that was to be updated. I proceeded with the update and let it finish, rebooted, then tried to add my second disk into the degraded array.

Everything went well! I now have an active raid1 array:

Quote:

mdadm -D /dev/md2
/dev/md2:
Version : 00.90.01
Creation Time : Fri May 20 13:28:09 2005
Raid Level : raid1
Array Size : 155115520 (147.93 GiB 158.84 GB)
Device Size : 155115520 (147.93 GiB 158.84 GB)
Raid Devices : 2
Total Devices : 2
Preferred Minor : 2
Persistence : Superblock is persistent

Update Time : Fri Jun 17 11:27:40 2005
State : clean
Active Devices : 2
Working Devices : 2
Failed Devices : 0
Spare Devices : 0

UUID : 92f52439:c5f8b6d4:b4214681:8c9a0db2
Events : 0.1008286

Number Major Minor RaidDevice State
0 8 3 0 active sync /dev/.static/dev/sda3
1 8 19 1 active sync /dev/.static/dev/sdb3
inferno:~#

It seems that mdadm was the problem, and an update to a newer version (mdadm - v1.9.0 - 04 February 2005) fixed my problem!

James

darkseer · 06-20-2005, 07:56 AM

Turned out I had a bad drive. Once I replaced it with a fresh one it was all OK. Looks like my problem was PEBCAK.

-Darkseer

jameso · 06-20-2005, 07:59 AM

Did you run a seatools diagnostic on it?

I did that and there were no errors.

darkseer · 06-20-2005, 08:04 AM

Well, I got 2 new drives. The idea is to have 2 hot spares, a couple of friends depend on this box being up. I droped one of the new drives in and it rebuilt without a hitch. Also, there was no I/O error during shutdown on the replacement drive. I'm just gonna RMA the bad one and call it a day. I really don't think its worth looking into to much when it is so easy to replace the drive with a new one.

-Darkseer