Yeah, I've done that in a non-RAID setup before. It worked really well, except for a bunch of stray files and directories that were just INODE addresses instead of names. Everything inside the directories was fine though, and that's all I cared about. I had pretty much a 100% recovery of that drive and I was quite happy I figured it all out. I was also lucky because the data wasn't on any other drive. It's not backed up daily of course.

I don't think there's a problem with the card. One of the drives keeps making a chirping sound (not a click) for a while and I've been expecting it to fail. I'm guessing since I still hear it that's not the drive that actually DID fail on me. I'm assuming what happened was that 1 drive failed for real (check the super-block seems to verify corruption) and another one just dropped out because mdadm messed up a little. The super-block on the other drive was still fine and I was able to force it all back online like I mentioned. Now, back to copying all this data over...... this is going to take a long, long time.

-Joe

Eric Shubert wrote:

I did this last summer (rebuilt w/ --assemble --force) and it worked ok.

For a drive that's really failed, you can dd as much of it as you can to 
a new drive, then run fsck on the new drive, then add it back in w/ 
--assemble --force. That worked for me as well IIRC. It did lose 
whatever wasn't able to be read w/ dd from the old drive though.

Given that 2 drives went at once, consider that the i/o card or MB (or 
PS) might be having issues. That was the case with the incident last 
summer (replaced i/o card, then MB was failing).

Joe Fleming wrote:

That's exactly what I want to do here; just pull up one of the drives 
long enough that I can get the data off it. I suspect one of the drives 
really did fail, I've been waiting for it to happen in fact. But since 
the other drive claims to have failed at the EXACT same time, I really 
don't think that it did.

I saw the --force option but there's no indication that it wasn't going 
to rebuild the array. The assemble option might simply imply that 
though.... it does say "This usage assembles one of more raid arrays 
from pre-existing components" which sounds promising enough.

I think you've described exactly what I was trying to do; assemble (NOT 
rebuild) and copy. Thanks!

-Joe

I've had luck in the past recovering from a multi-drive failure, where 
the other failed drive was not truly dead but rather was dropped 
because of an IO error caused by a thermal calibration or something 
similar.  The trick is to re-add the drive to the array and using the 
option to force it NOT to try to rebuild the array.  This used to be 
an require several options like --really-force and --really-dangerous 
but now I think its just something like --assemble --force /dev/md0. 
This forces the array to come back up to its degraded (still down 1 
disk) state.  If possible replace the degraded disk or copy your data 
off before the other flakey drive fails.
------------------------------------------------------------------------

---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.plug.phoenix.az.us
To subscribe, unsubscribe, or to change your mail settings:
http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss