Description of problem:
Spare disk added to a raid1 array by mdadm command is dropped on next boot.

If you add, remove, re-add a spare and then reboot the spare is lost due to mismatched sb info.


Version-Release number of selected component (if applicable):
RHEL 5.3


How reproducible:
100%


Steps to Reproduce:
1. create a raid1 array with a spare. 
   + All three devices should have the same sb event count (seen 
     with mdadm --examine). 
2. Use mdadm to remove the spare. 
   + Verify with /proc/mdstat that it is removed. 
3. Use mdadm to add the spare back to the same array. 
   + Verify with /proc/mdstat that the spare is part of the array. 
     You can use --examine to show that the spare sb has an event 
     count that is two less than the md and primary/secondary disks.
4. Reboot RHEL and the array will be autoassembled,either via a "nash 
   raidautorun <dev>" command or command or "/sbin/mdadm -A -s". 
5. Spare is dropped because it is "non-fresh".


  
Actual results:
Added spare is dropped.


Expected results:
Results of mdadm commands such as --add should persist across boots, spare
not dropped.


Additional info:

Fixed in the following upstream commit in 2.6.19.3 to make sure superblocks are
properly refreshed.  This code is already within RHEL6 stream.

"
[PATCH] md: assorted md and raid1 one-liners
author  NeilBrown <neilb>
        Sun, 10 Dec 2006 10:20:52 +0000 (02:20 -0800)
committer       Linus Torvalds <torvalds.org>
        Sun, 10 Dec 2006 17:57:21 +0000 (09:57 -0800)
commit  1757128438d41670ded8bc3bc735325cc07dc8f9
tree    e85679cbe949e337616ac53ab3b3fd1a3fa14a63        tree
parent  c2b00852fbae4f8c45c2651530ded3bd01bde814        commit | diff
[PATCH] md: assorted md and raid1 one-liners

Fix few bugs that meant that:
  - superblocks weren't alway written at exactly the right time (this
    could show up if the array was not written to - writting to the array
    causes lots of superblock updates and so hides these errors).

  - restarting device recovery after a clean shutdown (version-1 metadata
    only) didn't work as intended (or at all).

1/ Ensure superblock is updated when a new device is added.  <<<<<<
2/ Remove an inappropriate test on MD_RECOVERY_SYNC in md_do_sync.
   The body of this if takes one of two branches depending on whether
   MD_RECOVERY_SYNC is set, so testing it in the clause of the if
   is wrong.
3/ Flag superblock for updating after a resync/recovery finishes.
4/ If we find the neeed to restart a recovery in the middle (version-1
   metadata only) make sure a full recovery (not just as guided by
   bitmaps) does get done.

"

Patch consists of 4 line changes: 3 additions and 1 removal.  Customer
has added just one line from the full patch and tested to show this 
addresses the immediate issue.  The patch/line added was in add_new_disk():

                if (err)
                        export_rdev(rdev);
+               md_update_sb(mddev);
 
                set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);