[linux-lvm] LVM v0.9 on 2.4.0 kernel: raid5syncd halt

[Andreas Dilger]

Scott, you write:
> The 2.4.0 kernel (patched with XFS and the appropriate
> LVM patches from the 0.9-patches tree) produces a stop
> of any resync when issuing lvcreate (i.e., lvcreate
> -L 500M -n vol volgr0).

When you are talking about resync, is this the rebuilding of a RAID
array?  While I don't know much about how the RAID code works, there
doesn't appear to be anything in lvm_do_lv_create() which would do
anything bad.  It may be that elsewhere in lvcreate it calls something
to confuse/halt the RAID code (i.e. fsync_dev(), invalidate_buffers(),
etc).  Is there any way you can put debugging into the RAID resync code
to see why it stopped?

How often does a RAID resync happen?

Cheers, Andreas
-- 
Andreas Dilger  \ "If a man ate a pound of pasta and a pound of antipasto,
                 \  would they cancel out, leaving him still hungry?"
http://www-mddsp.enel.ucalgary.ca/People/adilger/               -- Dogbert