Kent,
I have just recently installed RAID-5 on my 2.2.10 w/ MOSIX machine. I
did a similar test and really all you have to do in case of that kind of
failure is (assuming you unplugged sdd):
1.) Shutdown the machine
2.) Replace (or re-plug-in the drive) the failed drive with a new one
3.) Power up
4.) Parition the new drive (in the case of just un-plug/re-plug, you
don't need to, the partition table stays intact)
5.) Do a "raidhotadd /dev/sdd1 /dev/md0"
The RAID partition gets marked as degraded when the failure occurs and
it stays degraded until the reconstruction completes. If you do a "cat
/proc/mdstats" before you raidhotadd the drive, while it's still in
degraded mode, you should see something like this (assuming you
unplugged sdd) :
md0 : active raid1 sdb1[0] XXX blocks level 1, 4k chunk, algorithm 2
[1/2] [U_]
Note that only the live, working sdb appears there, and that the last
two fields show 1 of 2 disks working and U_, which says one is fine(U),
the other isn't(_). When you do the "raidhotadd", you will get the same
output as above, but only with additional info showing that sdd1 now in
the list of partitions, and info telling you the progress of the
re-construction and estimated time to completion. When it's all done,
you should see something like:
md0 : active raid1 sdd1[1] sdb1[0] XXX blocks level 1, 4k chunk,
algorithm 2 [2/2] [UU]
And that should be all you need to do. You don't need to do mkraid,
because that is just for the initial creation of the raid partitions.
The raid drivers in the kernel and the raid recovery processes will
handle the recontruction once you do "raidhotadd", you don't have to
copy data around by hand.
Hope this helps,
Tom
--
Tom Kunz Tool Developer Software Consulting Services
PGP Key http://www.users.fast.net/~tkunz/pgp.html
1452 1F99 E2BB 632E 6EAE 2DF0 EF11 4DFC
DB62 7EBC 3BA0 6C40 88C0 C509 DA85 91B4 D5E9 EFD3
