Hi all,
I have some trouble in recreating a likely undamaged RAID5-Array with 4 disks.
Initially I created the array one year ago:
mdadm --create /dev/md0 --level=5 --chunk=128 --raid-devices=4 /dev/sdb1 /dev/sdc1 /dev/sdd1 /dev/sde1
/dev/md0 is base for a luks container which is used as physical volume for LVM.
Recently I moved the array to a new machine, all working fine. Then I have
experimented with hdparm to put drives into sleep-mode. After a few
sleep/wake-cycles all still working fine. Sunday an event occured that 2 of
the 4 disks dind't wake up properly. I waked up the array by issuing this
command:
dd if=/dev/zero bs=512 count=1 > mnt/tmp/test.img
(mnt/tmp resides in the array/luks/lvm)
I cannot remember if an error message occured, but I see that 2 drives aren't
accessible. I decided to restart the machine to get drives working. After
restart all 4 drives are working, but the array didn't came up because of 2
missing drives. After some googling I found
https://raid.wiki.kernel.org/index.php/RAID_Recovery
This guide seems to deal with my problem. I tried starting recreating my array
with this steps:
1. mdadm --examine /dev/sd[acde]1 > raid.status
raid.status contains this:
/dev/sda1:
Magic : a92b4efc
Version : 1.2
Feature Map : 0x0
Array UUID : 65089aca:483766d3:0db55271:27c73384
Name : oldman:0
Creation Time : Sun Jun 26 14:35:51 2011
Raid Level : raid5
Raid Devices : 4
Avail Dev Size : 3907024896 (1863.01 GiB 2000.40 GB)
Array Size : 5860536576 (5589.04 GiB 6001.19 GB)
Used Dev Size : 3907024384 (1863.01 GiB 2000.40 GB)
Data Offset : 2048 sectors
Super Offset : 8 sectors
State : clean
Device UUID : 3d7a888c:7488a859:f53a0fe5:5bbc1bda
Update Time : Sat Jul 14 22:08:08 2012
Checksum : eeb6d1f1 - correct
Events : 266
Layout : left-symmetric
Chunk Size : 128K
Device Role : Active device 3
Array State : AAAA ('A' == active, '.' == missing)
/dev/sdc1:
Magic : a92b4efc
Version : 1.2
Feature Map : 0x0
Array UUID : 65089aca:483766d3:0db55271:27c73384
Name : oldman:0
Creation Time : Sun Jun 26 14:35:51 2011
Raid Level : raid5
Raid Devices : 4
Avail Dev Size : 3907024896 (1863.01 GiB 2000.40 GB)
Array Size : 5860536576 (5589.04 GiB 6001.19 GB)
Used Dev Size : 3907024384 (1863.01 GiB 2000.40 GB)
Data Offset : 2048 sectors
Super Offset : 8 sectors
State : clean
Device UUID : 578cb781:387a3537:e42e9414:3d6e25e7
Update Time : Sat Jul 14 22:08:08 2012
Checksum : fe017b9c - correct
Events : 266
Layout : left-symmetric
Chunk Size : 128K
Device Role : Active device 1
Array State : AAAA ('A' == active, '.' == missing)
/dev/sdd1:
Magic : a92b4efc
Version : 1.2
Feature Map : 0x0
Array UUID : 65089aca:483766d3:0db55271:27c73384
Name : oldman:0
Creation Time : Sun Jun 26 14:35:51 2011
Raid Level : raid5
Raid Devices : 4
Avail Dev Size : 3907024896 (1863.01 GiB 2000.40 GB)
Array Size : 5860536576 (5589.04 GiB 6001.19 GB)
Used Dev Size : 3907024384 (1863.01 GiB 2000.40 GB)
Data Offset : 2048 sectors
Super Offset : 8 sectors
State : clean
Device UUID : 3dc3e9e2:ab9a2199:5618c6cd:ca0631fa
Update Time : Sun Jul 15 11:25:20 2012
Checksum : 8d5c0fd0 - correct
Events : 271
Layout : left-symmetric
Chunk Size : 128K
Device Role : Active device 2
Array State : A.A. ('A' == active, '.' == missing)
/dev/sde1:
Magic : a92b4efc
Version : 1.2
Feature Map : 0x0
Array UUID : 65089aca:483766d3:0db55271:27c73384
Name : oldman:0
Creation Time : Sun Jun 26 14:35:51 2011
Raid Level : raid5
Raid Devices : 4
Avail Dev Size : 3907024896 (1863.01 GiB 2000.40 GB)
Array Size : 5860536576 (5589.04 GiB 6001.19 GB)
Used Dev Size : 3907024384 (1863.01 GiB 2000.40 GB)
Data Offset : 2048 sectors
Super Offset : 8 sectors
State : clean
Device UUID : 22ecbee2:a26678a3:c2aeca18:007edd48
Update Time : Sun Jul 15 11:25:20 2012
Checksum : 3139124b - correct
Events : 271
Layout : left-symmetric
Chunk Size : 128K
Device Role : Active device 0
Array State : A.A. ('A' == active, '.' == missing)
2. grep Role raid.status
gives:
Device Role : Active device 3
Device Role : Active device 1
Device Role : Active device 2
Device Role : Active device 0
3. mdadm --create --assume-clean --level=5 --chunk=128 --raid-devices=4 /dev/md0 /dev/sde1 /dev/sdc1 /dev/sdd1 /dev/sda1
THis gives the message that all drives apears to be part of a raid array.
I confirmed "Continue creating array" and /dev/md0 was created. But data
in /dev/md0 is corrupted. I cannot luksOpen the device. The first bytes
of /dev/md0 did not contain the luks header. Something went wrong in
recreating the array. I've build a simple script that tries to create the
array and luksOpen it in all possible 24 permutations in disk-arrangement,
but none of them brings me up my data.
I'm pretty sure that I didn't write something to the array, except the 512
bytes to wake up it from sleep (see above). The kernel messages around the
time drive didn't respond are:
Jul 15 11:21:12 EBENE kernel: [30046.863193] sd 2:0:0:0: [sdc] Unhandled error code
Jul 15 11:21:12 EBENE kernel: [30046.863197] sd 2:0:0:0: [sdc] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
Jul 15 11:21:12 EBENE kernel: [30046.863201] sd 2:0:0:0: [sdc] CDB: Read(10): 28 00 2f 91 75 00 00 00 a0 00
Jul 15 11:21:12 EBENE kernel: [30046.863374] sd 0:0:0:0: [sda] Unhandled error code
Jul 15 11:21:12 EBENE kernel: [30046.863376] sd 0:0:0:0: [sda] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
Jul 15 11:21:12 EBENE kernel: [30046.863378] sd 0:0:0:0: [sda] CDB: Read(10): 28 00 2f 91 75 00 00 00 a0 00
Jul 15 11:21:12 EBENE kernel: [30046.864442] sd 0:0:0:0: [sda] Unhandled error code
Jul 15 11:21:12 EBENE kernel: [30046.864445] sd 0:0:0:0: [sda] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
Jul 15 11:21:12 EBENE kernel: [30046.864449] sd 0:0:0:0: [sda] CDB: Read(10): 28 00 2f 91 76 00 00 00 a0 00
Jul 15 11:21:12 EBENE kernel: [30046.864508] sd 2:0:0:0: [sdc] Unhandled error code
Jul 15 11:21:12 EBENE kernel: [30046.864510] sd 2:0:0:0: [sdc] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
Jul 15 11:21:12 EBENE kernel: [30046.864513] sd 2:0:0:0: [sdc] CDB: Read(10): 28 00 2f 91 75 a0 00 00 60 00
Jul 15 11:21:12 EBENE kernel: [30046.864638] sd 0:0:0:0: [sda] Unhandled error code
Jul 15 11:21:12 EBENE kernel: [30046.864640] sd 0:0:0:0: [sda] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
Jul 15 11:21:12 EBENE kernel: [30046.864642] sd 0:0:0:0: [sda] CDB: Read(10): 28 00 2f 91 75 a0 00 00 60 00
Jul 15 11:21:12 EBENE kernel: [30046.864680] sd 2:0:0:0: [sdc] Unhandled error code
Jul 15 11:21:12 EBENE kernel: [30046.864681] sd 2:0:0:0: [sdc] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
Jul 15 11:21:12 EBENE kernel: [30046.864683] sd 2:0:0:0: [sdc] CDB: Read(10): 28 00 2f 91 76 00 00 00 a0 00
Jul 15 11:23:38 EBENE kernel: [30192.746530] sd 0:0:0:0: [sda] Unhandled error code
Jul 15 11:23:38 EBENE kernel: [30192.746534] sd 0:0:0:0: [sda] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
Jul 15 11:23:38 EBENE kernel: [30192.746539] sd 0:0:0:0: [sda] CDB: Read(10): 28 00 00 00 00 00 00 00 20 00
Jul 15 11:23:38 EBENE kernel: [30192.746810] sd 0:0:0:0: [sda] Unhandled error code
Jul 15 11:23:38 EBENE kernel: [30192.746812] sd 0:0:0:0: [sda] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
Jul 15 11:23:38 EBENE kernel: [30192.746815] sd 0:0:0:0: [sda] CDB: Read(10): 28 00 00 00 00 00 00 00 08 00
I have modified the Permute_array.pl on the wiki page to not mount /dev/md0
but print out the first 4 bytes of /dev/md0. These must be "LUKS", if array
is recreated in a succesful way. But I didn't ge this result after all
permutations.
I'm afraid of having killed my array by not using "missing" on some drive in
the first try (see above Point 3). Will may array remain corrupted?
Any suggestions?
Thanks in advance
André
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html