[SOLVED] Software RAID1, need some help

Hi,
We run OMV for some months. I recycled an old computer, put 2 * 2 TB disks, and used OMV to create a software RAID1. This server is connected to our old Windows Server 2003 Active Directory. This has worked without any trouble since the beginning.

Some days ago, first alert, my colleagues couldn't access the Samba shares anymore by using their regular AD credentials. Not a big deal, I rebooted the server and everything came back as usual.

But few days ago (last friday, let's be accurate!), something weird and worrying happened, it was like if the server RAID switched back some weeks in the past (as if we switched back to an old snapshot for a VM). We lost some data, and see old data back which are normally automatically cleaned after 10 days.

Here I am, I don't want to let things like this, and would like to diagnose a bit to prevent, as much as possible, the same thing to happen again. I need some help, because, for instance, the syslog doesn't contain traces older than today. Any help would be greatly appreciated, thanks.

Hi, and thanks for helping.

I went through almost all logs and found nothing relevant. But I'm not an expert. I wonder if this can come from a disk failure. But as these disks are mirrored in the software RAID, there is no way to get a SMART assessment.

No, I don't run any rsync script nor cron jobs, except default OMV tasks. The OMV version is the latest available (now 0.5.48), the system is regularly updated. Any idea how I can diagnose this issue?

Hard drive, seems clean based on SMART data.

I think I've found someting. Normally, I setup a software RAID1, made of 2 WD 2 TB HDDs. But on OMV, I see this:

<machine>:raid - /dev/md0 - clean,degraded - mirror - 1,82TiB - /dev/sdc
<machine>:raid - /dev/md127 - clean,degraded - mirror - 1,82TiB - /dev/sdd

That's pretty disturbing don't you think? So I tried to deep into mdadm analysis tools.

If I run: mdadm --query --detail /dev/md0, the output is:

/dev/md0:
        Version : 1.2
  Creation Time : Mon Oct 21 11:14:26 2013
     Raid Level : raid1
     Array Size : 1953513424 (1863.02 GiB 2000.40 GB)
  Used Dev Size : 1953513424 (1863.02 GiB 2000.40 GB)
   Raid Devices : 2
  Total Devices : 1
    Persistence : Superblock is persistent


    Update Time : Tue May 27 16:11:20 2014
          State : clean, degraded
 Active Devices : 1
Working Devices : 1
 Failed Devices : 0
  Spare Devices : 0


           Name : <machine>:raid  (local to host <machine>)
           UUID : 9452f898:58290c79:1691d389:fbb1d13b
         Events : 4242


    Number   Major   Minor   RaidDevice State
       0       8       32        0      active sync   /dev/sdc
       1       0        0        1      removed

If I run: mdadm --query --detail /dev/md127, the output is:

/dev/md127:
        Version : 1.2
  Creation Time : Mon Oct 21 11:14:26 2013
     Raid Level : raid1
     Array Size : 1953513424 (1863.02 GiB 2000.40 GB)
  Used Dev Size : 1953513424 (1863.02 GiB 2000.40 GB)
   Raid Devices : 2
  Total Devices : 1
    Persistence : Superblock is persistent


    Update Time : Thu May 22 16:01:33 2014
          State : clean, degraded
 Active Devices : 1
Working Devices : 1
 Failed Devices : 0
  Spare Devices : 0


           Name : <machine>:raid  (local to host <machine>)
           UUID : 9452f898:58290c79:1691d389:fbb1d13b
         Events : 38014


    Number   Major   Minor   RaidDevice State
       0       0        0        0      removed
       1       8       48        1      active sync   /dev/sdd

My newbie first thought would be: "my raid doesn't work at all". What do you think?

/dev/md0 on /media/f605a663-0455-4580-9139-fe9350690b51 type ext4 (rw,noexec,_netdev,acl,user_xattr,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0)

No /md127 in the output. I guess I've to rebuild my raid, don't you think?

Hi,
I finally fixed the issue by removing the RAID totally. And re-created from scratch. Bonus, I didn't lose any data, OMV (mdadm indeed) synced everything right.

/dev/md127 must not be used, it's a sign that your raid is in a bad shape. So be aware, if you see /dev/md127, your raid needs some attention.