Dear users of linux-raid mailing list.
I'm looking for some help about errors that I did when trying to add 2
new disks to my existing 4drive *320gb array. I added the disks to the
array and used mdadm to grow em. After it was done rebuilding I "forgot"
to check the status of the disks. 1 of the disks added itself as a hot
spare (never had that yet before, but was planned for next upgrade since
I was planning more disks/controllers ETA 2 weeks later), and the other
disk was added as removed-state. I managed then to use resize2fs on the
/dev/md1 to increase it too full since. I think it showed some errors at
the end, but since I was doing this in school time I was of course as
lucky to manage to close the terminal instead of logging out of screen
since I was running to class.
Simple case is that I managed to resize the file system to a bigger size
then it isn't. One of the disks died too, spin up&spin down&click
sounds. I'am afraid that it could be me who killed the drive but
software shouldn't do that as far as I know. (2 drives actually died, so
their in for RMA. I bought 3 new hdds, but didn't have sata-slot for the
last hdd, so it was only powered on in the hot swap bay without even
being connected to the motherboard - and that drive died..).
This is the log from after I've resized and activated the array and it
failed:
http://pastebin.ca/357261
These are the disks that should be in the array: (mdadm --examine disk)
http://pastebin.ca/357284
First all 4 originally drives was in sync and the disk one was in hot
spare and disk two removed-state.
I though I had to stop the array to be able to make them active fully in
my array - this was quite a mistake I think.
I now ended up with all drives in removed-state, without the hot spare
one which was now synced if I forced the array to run with -R option. (
-R, --run
Insist that mdadm run the array, even if some of the
components
appear to be active in another array or file system.
Normally
mdadm will ask for confirmation before including such
components
in an array. This option causes that question to be
suppressed.
)
This is /proc/mdstat after I was using -R option:
sinope:/etc/mdadm# cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5]
[raid4] [multipath] [faulty]
md1 : inactive sdf[6]
312571136 blocks super 1.0
/dev/sda (old disk)
/dev/sdb (old disk)
/dev/sdc (old disk)
/dev/sdd (old disk)
/dev/sde (new removed-state disk)
/dev/sdf (new hot spare which at the end activated itself as synced
after using -R)
I know I've managed to do quite odd things, but is it still possible to
rescue my raid5 array?
I'm afraid that if I try again I might be able to crash more hdds - but
I highly doubt it was the system's fault since one drive that wasn't
even connected died too.
What should I do now to try rescue the array?
---
Roy Sindre Norangshol
-
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
