RAID5 recovery after disk-change

RAID5 recovery after disk-change

[Simon Frettloeh]

Hi!

Well, I'm wondering if I did something wrong after replacing probably 
faulty disk by a fresh new one.

The new disk just hangs on the same controller on the same cable 
and is identical to the replaced one.

I switched on power. The raid still was working (in degraded mode).
 I used fdisk to partitionate the new disk (/dev/hdg) and create a 
partition (/dev/hdg1) with the fd-flag. After that I tried to add the new 
disk by this command:
mdadm /dev/md0 --add /dev/hdg1

That's how /proc/mdstat looks like now:
=====snip============================================================
thismaschine:~ # cat /proc/mdstat
Personalities : [raid5]
read_ahead 1024 sectors
md0 : active raid5 hdg1[4] hdh1[3] hdf1[1] hde1[0]
      480214656 blocks level 5, 128k chunk, algorithm 2 [4/3] [UU_U]
      [==================>..]  recovery = 90.8% (145491784/160071552) 
finish=21.4min speed=11336K/sec
unused devices: <none>
====snap==============================================================

Strange is, that the new disk didn't get place 2 in the array.

What went wrong?
Do I have to repair this?
How can I repair this?
How do I have to modify my raidtab (see below)?

I'd be very thankful for helping me, after I tried the howto, the faq and 
searched this mailinglist.

Ciao,
Simon Frettloeh


=====snip=============================================================
# autogenerated /etc/raidtab by YaST2

raiddev /dev/md0
   raid-level       5
   nr-raid-disks    4
   nr-spare-disks   0
   persistent-superblock 1
   parity-algorithm left-symmetric
   chunk-size       128
   device   /dev/hde1
   raid-disk 0
   device   /dev/hdf1
   raid-disk 1
   device   /dev/hdg1
   raid-disk 2
   device   /dev/hdh1
   raid-disk 3
=====snap=============================================================

Re: RAID5 recovery after disk-change

[Simon Frettloeh]

Hi again!

Strange things happen and I don't know why. After resynching the RAID I
get following output:
====snip=============================================================
thismachine:~ # cat /proc/mdstat
Personalities : [raid5]
read_ahead 1024 sectors
md0 : active raid5 hdg1[2] hdh1[3] hdf1[1] hde1[0]
      480214656 blocks level 5, 128k chunk, algorithm 2 [4/4] [UUUU]

unused devices: <none>
====snap=============================================================
So it seems, that hdg1 moved by itself from "4" to "2". What happened
(according to my previous email)?

Do I have to do any further things or is my RAID now working fine?

Ciao,
Simon

Re: RAID5 recovery after disk-change

[Paul Clements]

Simon Frettloeh wrote:

> So it seems, that hdg1 moved by itself from "4" to "2". What happened
> (according to my previous email)?
> 
> Do I have to do any further things or is my RAID now working fine?

This is normal behavior. When a spare disk is added, it is placed into
the array above the normal disk slots (in your case, slots 4 and up) and
is then resynced. Once resync is complete, the disk is moved into the
first open normal slot (in your case, slot 2) and is then used as part
of the array.

--
Paul