Lost mdadm superblock but not drives - can I rebuild the software RAID5?

[Syncro]

I have an old Athlon XP 3000 machine that I keep around as a file server.

It's currently got three 1TB drives which I had setup as mdadm raid 5 on FC10. The machine's original drive held the superblock for the raid array and it just had a massive heart attack.

I've searched, my biggest source being http://man-wiki.net/index.php/8:mdadm

I can't tell if I can reassemble the superblock info lost with the original hard drive or if I've lost it all...

[coffee412]

perhaps this link will help you out.

[Syncro]

Thank you jebus!
All fixed

[Syncro]

More problems...

The primary 200gb drive which held the OS died, when I replaced it with a new 500g unit and tried to reconstruct my old array disks some typical linux BS inevitably came up

Status of the three 1tb raid drives on the new OS:

[root@syncro /]# fdisk -l

Disk /dev/sda: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xa02c3486

Disk /dev/sda doesn't contain a valid partition table

Disk /dev/sdb: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x000df562

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1               1      121601   976760001   fd  Linux raid autodetect

Disk /dev/sdc: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xa021c1e4

   Device Boot      Start         End      Blocks   Id  System
/dev/sdc1               1      121601   976760001   fd  Linux raid autodetect

Disk /dev/sdd: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x00089c30

   Device Boot      Start         End      Blocks   Id  System
/dev/sdd1   *           1          26      204800   83  Linux
Partition 1 does not end on cylinder boundary.
/dev/sdd2              26       60801   488179201   8e  Linux LVM

Disk /dev/dm-0: 497.8 GB, 497779998720 bytes
255 heads, 63 sectors/track, 60518 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x00000000

Disk /dev/dm-0 doesn't contain a valid partition table

Disk /dev/dm-1: 2113 MB, 2113929216 bytes
255 heads, 63 sectors/track, 257 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x00000000

Disk /dev/dm-1 doesn't contain a valid partition table

Disk /dev/md0: 2000.4 GB, 2000409591808 bytes
255 heads, 63 sectors/track, 243202 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x000df562

    Device Boot      Start         End      Blocks   Id  System
/dev/md0p1               1      121601   976760001   fd  Linux raid autodetect

I don't know why sda shows up as not having a partition table all of a sudden. I panicked and forced the original OS to boot up (cranky drive) and it had no problem viewing\writing the array. So back to the new\os drive and...

[root@syncro /]# mdadm --create /dev/md0 -v -l 5 -n 3 /dev/sdb /dev/sda /dev/sdc
mdadm: layout defaults to left-symmetric
mdadm: chunk size defaults to 64K
mdadm: layout defaults to left-symmetric
mdadm: /dev/sdb appears to be part of a raid array:
    level=raid5 devices=3 ctime=Tue Mar 23 16:21:40 2010
mdadm: layout defaults to left-symmetric
mdadm: /dev/sda appears to be part of a raid array:
    level=raid5 devices=3 ctime=Tue Mar 23 16:21:40 2010
mdadm: layout defaults to left-symmetric
mdadm: /dev/sdc appears to be part of a raid array:
    level=raid5 devices=3 ctime=Tue Mar 23 16:21:40 2010
mdadm: size set to 976762496K
Continue creating array? y
mdadm: array /dev/md0 started.
[root@syncro /]# cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4]
md0 : active raid5 sdc[3] sda[1] sdb[0]
      1953524992 blocks level 5, 64k chunk, algorithm 2 [3/2] [UU_]
      [>....................]  recovery =  0.3% (3073664/976762496) finish=531.5min speed=30530K/sec

unused devices: <none>

It takes a good 8 hours for the so called 'spare' drive sdc to 'sync' or whatever and then attempting to re-assemble the array stills fails. Message says whatever drive is listed first on the assemble command has no superblock (be is sda,sdb,sdc) and assembly is aborted. If I keep trying to mess with it, such as recreating the array to do something else the bloody thing just starts to sync sdc all over again for another 8 hour session

I don't know why sda is listed without a superblock on fdisk -l nor why it is sdc that gets 're-synced'.

Current md0 status

[root@syncro /]# mdadm -D /dev/md0
/dev/md0:
        Version : 0.90
  Creation Time : Wed Mar 24 13:34:19 2010
     Raid Level : raid5
     Array Size : 1953524992 (1863.03 GiB 2000.41 GB)
  Used Dev Size : 976762496 (931.51 GiB 1000.20 GB)
   Raid Devices : 3
  Total Devices : 3
Preferred Minor : 0
    Persistence : Superblock is persistent

    Update Time : Wed Mar 24 13:34:37 2010
          State : clean, degraded, recovering
 Active Devices : 2
Working Devices : 3
 Failed Devices : 0
  Spare Devices : 1

         Layout : left-symmetric
     Chunk Size : 64K

 Rebuild Status : 4% complete

           UUID : 68e25f81:291da775:502a3737:121dd8f0 (local to host syncro.redirectme.net)
         Events : 0.3

    Number   Major   Minor   RaidDevice State
       0       8       16        0      active sync   /dev/sdb
       1       8        0        1      active sync   /dev/sda
       3       8       32        2      spare rebuilding   /dev/sdc

When sdc finishes with whatever it's doing it then shows up as 'active sync' as well, and system state changes to clean. etc etc.

Can anyone give me a few ideas on how to proceed?

[BrianMicek]

I see at least one of your problems.

[root@syncro /]# fdisk -l

Disk /dev/sda: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xa02c3486

Disk /dev/sda doesn't contain a valid partition table
...

That means the first 512 bytes are corrupted. The reason is because if this command:

[root@syncro /]# mdadm --create /dev/md0 -v -l 5 -n 3 /dev/sdb /dev/sda /dev/sdc

I think you wanted to issue:

[root@syncro /]# mdadm --create /dev/md0 -v -l 5 -n 3 /dev/sdb1 /dev/sda1 /dev/sdc1

[Syncro]

I had tried your correct-er version of the create command first since I'd c\p from the web

Brought to my attention the faulty partition table on /dev/sda and aborted

I tried leaving out specific partitions and got a bit further, as it recognized all 3 drives as being part of a common array and began the sdc rebuild process

It's currently at 65% on it's second go around. I'll make sure to not re-create the array from scratch a third time but without being able to assemble it for mount without seemingly fixing sda's partition tables whose status does not seem to change regardless of what mdadm does as far as the data I see from fdisk -l

[BrianMicek]

I should have pointed out something and apologize. I'm pretty sure there is nothing wrong with using an entire disk as a portion of a RAID like you did. I'm pretty sure people just don't do it because you loose some flexibility at almost no cost. The main drawback is on boot, mdadm searches for all your RAIDs and assembles them by detecting partitions of type 0xfd "Linux raid autodetect" in conjunction with their superblock and the contents of mdadm.conf. I'm pretty sure you will have to manually assemble your RAIDs each boot but I never tried it.

Don't give up and stick with it! Linux RAIDs are great. I never lost a byte of data.

[raidrecovery01]

This is said to know that you loss your precious data. I will suggest you to take Raid 5 services and get rid of this problem quickly.

[DBelton]

I hate to say this, but with you only having 2 posts, and digging up dead threads just to post something useless but with a link to a site selling Raid recovery services, I would have to classify your post as spam advertising.

[geekwlover]

i found a free partition tool that can repair RAID5 on donwload.cnet
Check here: http://download.cnet.com/Easeus-Part...#editorsreview
It might can help u.