Disk failure during grow, what is the current state.

Disk failure during grow, what is the current state.

[Steve Fairbairn]

Hi All,

I was wondering if someone might be willing to confirm what the current
state of my RAID array is, given the following sequence of events (sorry
it's pretty long)....

I had a clean, running /dev/md0 using 5 disks in RAID 5 (sda1, sdb1,
sdc1, sdd1, hdd1).  It had been clean like that for a while.  So last
night I decided it was safe to grow the array into a sixth disk....

[root@space ~]# mdadm /dev/md0 --add /dev/hdi1
mdadm: added /dev/hdi1
[root@space ~]# mdadm -D /dev/md0
/dev/md0:
        Version : 00.90.03
  Creation Time : Wed Jan  9 18:57:53 2008
     Raid Level : raid5
     Array Size : 1953535744 (1863.04 GiB 2000.42 GB)
  Used Dev Size : 488383936 (465.76 GiB 500.11 GB)
   Raid Devices : 5
  Total Devices : 6
Preferred Minor : 0
    Persistence : Superblock is persistent

    Update Time : Tue Feb  5 23:55:59 2008
          State : clean
 Active Devices : 5
Working Devices : 6
 Failed Devices : 0
  Spare Devices : 1

         Layout : left-symmetric
     Chunk Size : 64K

           UUID : 382c157a:405e0640:c30f9e9e:888a5e63
         Events : 0.429616

    Number   Major   Minor   RaidDevice State
       0       8        1        0      active sync   /dev/sda1
       1       8       17        1      active sync   /dev/sdb1
       2       8       33        2      active sync   /dev/sdc1
       3      22       65        3      active sync   /dev/hdd1
       4       8       49        4      active sync   /dev/sdd1

       5      56        1        -      spare   /dev/hdi1
[root@space ~]# mdadm --grow /dev/md0 --raid-devices=6
mdadm: Need to backup 1280K of critical section..
mdadm: ... critical section passed.
[root@space ~]# cat /proc/mdstat 
Personalities : [raid6] [raid5] [raid4] 
md0 : active raid5 hdi1[5] sdd1[4] sdc1[2] sdb1[1] sda1[0] hdd1[3]
      1953535744 blocks super 0.91 level 5, 64k chunk, algorithm 2 [6/6]
[UUUUUU]
      [>....................]  reshape =  0.0% (29184/488383936)
finish=2787.4min speed=2918K/sec
      
unused devices: <none>
[root@space ~]# 

OK, so that would take nearly 2 days to complete, so I went to bed happy
about 10 hours ago.

I come to the machine this morning, and I have the following....

[root@space ~]# cat /proc/mdstat 
Personalities : [raid6] [raid5] [raid4] 
md0 : active raid5 hdi1[5] sdd1[6](F) sdc1[2] sdb1[1] sda1[0] hdd1[3]
      1953535744 blocks super 0.91 level 5, 64k chunk, algorithm 2 [6/5]
[UUUU_U]
      
unused devices: <none>
You have new mail in /var/spool/mail/root
[root@space ~]# mdadm -D /dev/md0
/dev/md0:
        Version : 00.91.03
  Creation Time : Wed Jan  9 18:57:53 2008
     Raid Level : raid5
     Array Size : 1953535744 (1863.04 GiB 2000.42 GB)
  Used Dev Size : 488383936 (465.76 GiB 500.11 GB)
   Raid Devices : 6
  Total Devices : 6
Preferred Minor : 0
    Persistence : Superblock is persistent

    Update Time : Wed Feb  6 05:28:09 2008
          State : clean, degraded
 Active Devices : 5
Working Devices : 5
 Failed Devices : 1
  Spare Devices : 0

         Layout : left-symmetric
     Chunk Size : 64K

  Delta Devices : 1, (5->6)

           UUID : 382c157a:405e0640:c30f9e9e:888a5e63
         Events : 0.470964

    Number   Major   Minor   RaidDevice State
       0       8        1        0      active sync   /dev/sda1
       1       8       17        1      active sync   /dev/sdb1
       2       8       33        2      active sync   /dev/sdc1
       3      22       65        3      active sync   /dev/hdd1
       4       0        0        4      removed
       5      56        1        5      active sync   /dev/hdi1

       6       8       49        -      faulty spare
[root@space ~]# df -k
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
                      56086828  11219432  41972344  22% /
/dev/hda1               101086     18281     77586  20% /boot
/dev/md0             1922882096 1775670344  69070324  97% /Downloads
tmpfs                   513556         0    513556   0% /dev/shm
[root@space ~]# mdadm /dev/md0 --remove /dev/sdd1
mdadm: cannot find /dev/sdd1: No such file or directory [root@space ~]# 

As you can see, one of the original 5 devices has failed (sdd1) and
automatically removed.  The reshape has stopped, but the new disk seems
to be in and clean which is the bit I don't understand.  The new disk
hasn't been added to the size, so it would seem that md has switched it
to being used as a spare instead (possibly as the grow hadn't
completed?).

How come it seems to have recovered so nicely?
Is there something I can do to check it's integrity?
Was it just so much quicker than 2 days because it switched to only
having to sort out the 1 disk? Would it be safe to run an fsck to check
the integrity of the fs?  I don't want to inadvertently blat the raid
array by 'using' it when it's in a dodgy state.

I have unmounted the drive for the time being, so that it doesn't get
any writes until I know what state it is really in.

Any suggestions gratefully received,

Steve.

No virus found in this outgoing message.
Checked by AVG Free Edition. 
Version: 7.5.516 / Virus Database: 269.19.20/1261 - Release Date:
05/02/2008 20:57
 

Re: Disk failure during grow, what is the current state.

[Nagilum]

[-- Attachment #1: Type: text/plain, Size: 2485 bytes --]

----- Message from steve@fairbairn-family.com ---------
     Date: Wed, 6 Feb 2008 12:58:55 -0000
     From: Steve Fairbairn <steve@fairbairn-family.com>
Reply-To: Steve Fairbairn <steve@fairbairn-family.com>
  Subject: Disk failure during grow, what is the current state.
       To: linux-raid@vger.kernel.org


> As you can see, one of the original 5 devices has failed (sdd1) and
> automatically removed.  The reshape has stopped, but the new disk seems
> to be in and clean which is the bit I don't understand.  The new disk
> hasn't been added to the size, so it would seem that md has switched it
> to being used as a spare instead (possibly as the grow hadn't
> completed?).
>
> How come it seems to have recovered so nicely?
> Is there something I can do to check it's integrity?
> Was it just so much quicker than 2 days because it switched to only
> having to sort out the 1 disk? Would it be safe to run an fsck to check
> the integrity of the fs?  I don't want to inadvertently blat the raid
> array by 'using' it when it's in a dodgy state.
>
> I have unmounted the drive for the time being, so that it doesn't get
> any writes until I know what state it is really in.


----- End message from steve@fairbairn-family.com -----

If a drive failes during reshape the reshape will just continue.
The blocks which were on the failed drive are calculated from the the  
other disks and writes to the failed disk are simply omitted.
The result is a raid5 with a failed drive.
You should get a new drive asap to restore the redundancy.
Also it's kinda important that you don't run 2.6.23 because it has a  
nasty bug which would be triggered in this scenario.
The reshape probably increased in speed after the system was no longer  
actively used and io bandwidth freed up.
Kind regards,
Alex.


========================================================================
#    _  __          _ __     http://www.nagilum.org/ \n icq://69646724 #
#   / |/ /__ ____ _(_) /_ ____ _  nagilum@nagilum.org \n +491776461165 #
#  /    / _ `/ _ `/ / / // /  ' \  Amiga (68k/PPC): AOS/NetBSD/Linux   #
# /_/|_/\_,_/\_, /_/_/\_,_/_/_/_/   Mac (PPC): MacOS-X / NetBSD /Linux #
#           /___/     x86: FreeBSD/Linux/Solaris/Win2k  ARM9: EPOC EV6 #
========================================================================


----------------------------------------------------------------
cakebox.homeunix.net - all the machine one needs..


[-- Attachment #2: PGP Digital Signature --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

FW: Disk failure during grow, what is the current state.

[Steve Fairbairn]

I'm having a nightmare with emails today.  I can't get a single one
right first time.  Apologies to Alex for sending it directly to him and
not to the list on first attempt.

Steve

> -----Original Message-----
> From: Steve Fairbairn [mailto:steve@fairbairn-family.com] 
> Sent: 06 February 2008 15:02
> To: 'Nagilum'
> Subject: RE: Disk failure during grow, what is the current state.
> 
> 
> 
> 
> > -----Original Message-----
> > From: linux-raid-owner@vger.kernel.org
> > [mailto:linux-raid-owner@vger.kernel.org] On Behalf Of Nagilum
> > Sent: 06 February 2008 14:34
> > To: Steve Fairbairn
> > Cc: linux-raid@vger.kernel.org
> > Subject: Re: Disk failure during grow, what is the current state.
> > 
> > 
> > 
> > If a drive failes during reshape the reshape will just
> > continue. The blocks which were on the failed drive are 
> > calculated from the the  
> > other disks and writes to the failed disk are simply omitted. 
> > The result is a raid5 with a failed drive. You should get a 
> > new drive asap to restore the redundancy. Also it's kinda 
> > important that you don't run 2.6.23 because it has a  
> > nasty bug which would be triggered in this scenario.
> > The reshape probably increased in speed after the system was 
> > no longer  
> > actively used and io bandwidth freed up.
> > Kind regards,
> > Alex.
> > 
> 
> Thanks for the response Alex, but....
> 
>      Array Size : 1953535744 (1863.04 GiB 2000.42 GB)
>   Used Dev Size : 488383936 (465.76 GiB 500.11 GB)
> 
> Surely the added disk should now been added to the Array 
> Size?  5 * 500GB is 2500GB, not 2000GB.  This is why I don't 
> think the reshape has continued.  As for speeding up because 
> of no IO badwidth, this also doesn't actually hold very true, 
> because the system was at a point of not being used anyway 
> before I added the disk, and I didn't unmount the drive until 
> this morning after it claimed it had finished doing anything.
> 
> It's because the size doesn't match up to all 5 disks being 
> used that I still wonder at the state of the array.
> 
> Steve.
> 
> 
> 
> 
> No virus found in this outgoing message.
> Checked by AVG Free Edition. 
> Version: 7.5.516 / Virus Database: 269.19.20/1261 - Release 
> Date: 05/02/2008 20:57
>  
> 

No virus found in this outgoing message.
Checked by AVG Free Edition. 
Version: 7.5.516 / Virus Database: 269.19.20/1261 - Release Date:
05/02/2008 20:57
 

RE: Disk failure during grow, what is the current state.

[Steve Fairbairn]

> > -----Original Message-----
> > From: Steve Fairbairn [mailto:steve@fairbairn-family.com]
> > Sent: 06 February 2008 15:02
> > To: 'Nagilum'
> > Subject: RE: Disk failure during grow, what is the current state.
> > 
> > 
> >      Array Size : 1953535744 (1863.04 GiB 2000.42 GB)
> >   Used Dev Size : 488383936 (465.76 GiB 500.11 GB)
> > 
> > Surely the added disk should now been added to the Array
> > Size?  5 * 500GB is 2500GB, not 2000GB.  This is why I don't 
> > think the reshape has continued.  As for speeding up because 
> > of no IO badwidth, this also doesn't actually hold very true, 
> > because the system was at a point of not being used anyway 
> > before I added the disk, and I didn't unmount the drive until 
> > this morning after it claimed it had finished doing anything.
> > 

Thanks again to Alex for his comments.  I've just rebooted the box, and
the reshape has continued on the degraded array and an RMA has been
raised for the faulty disk.

Thanks,

Steve.

No virus found in this outgoing message.
Checked by AVG Free Edition. 
Version: 7.5.516 / Virus Database: 269.19.20/1261 - Release Date:
05/02/2008 20:57