Hello,
Is there anyway to force a resync of RAID1 arrays that need to be
resynced? I have 26 RAID1 arrays and when the system shuts down
uncleanly, or the machine crashes hard, sometimes I need to hotadd
devices back into the array. For example if I have the following in
/etc/raidtab:
raiddev /dev/md23
raid-level 1
nr-raid-disks 2
persistent-superblock 1
chunk-size 64
device /dev/sdao1
raid-disk 0
device /dev/sdbc1
raid-disk 1
/proc/mdstat for md23 will show:
md23 : active raid1 sdbc1[1]
17783808 blocks [2/2] [U_]
And I will need to use mdadm or raidhotadd to add /dev/sdao1 back into
the array. When I add a device back into the array the reconstruction
doesn't start immediately and/or if another reconstruction is underway,
the new recontruction will not occur until the first one has completed.
Is this behavior expected/by design? Waiting for 26 arrays to
reconstruct one at a time is painfully slow. When I created the arrays
they all started to sync at the same time so I know the RAID code will
sync more than one at a time, and when my box came up yesterday after a
hard power cycle, about 5 arrays started syncing, but the rest waited
until those were finished and then synced one at a time.
I looked through md.c and raid1.c and I see that hot_add_disk() calls
md_recover_arrays() which wakes up the md recovery thread, but does the
thread only allow one synchronization to occur at a time?
Here are some of my system details:
Red Hat 7.3 with updates
kernel-bigmem-2.4.20-30 with megaraid2 patch
If someone could point me in the direction in the code to understand the
way this process works I would appreciate it. Let me know if you need
any more information.
Regards,
Andrew Rechenberg
Infrastructure Team, Sherman Financial Group
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
