Hi

Thanks a lot, this helped me to get the raid up again. Now I am trying to copy 
anything to another place.
ATM I am still not sure about what caused the crash, maybe you are right and 
there is a hw problem. I assume it is the  memory.

thx
thomas


Zitat von Neil Brown [off-list ref]:

On Wednesday January 25, mdadm@lr-netcon.de wrote:

quoted

Hi,

I am using a 8 disc scsi raid5 array on a fedora 3 system.

after a system crash, while the array was rebuilding a missing hd, 
i am now unable to get this array running again.

thats all i get (/dev/sdr1 is not the one which was rebuild):
# mdadm -R /dev/md4
mdadm: failed to run array /dev/md4: Invalid argument   

# mdadm -A /dev/md4 -
v /dev/sdk1 /dev/sdl1 /dev/sdm1 /dev/sdn1 /dev/sdo1 /dev/sdp1 /dev/sdq1

/dev/sdr

quoted

1
looking for devices for /dev/md4
/dev/sdk1 is identified as a member of /dev/md4, slot 0.
/dev/sdl1 is identified as a member of /dev/md4, slot 1.
/dev/sdm1 is identified as a member of /dev/md4, slot 2.
/dev/sdn1 is identified as a member of /dev/md4, slot 8.
/dev/sdo1 is identified as a member of /dev/md4, slot 4.
/dev/sdp1 is identified as a member of /dev/md4, slot 5.
/dev/sdq1 is identified as a member of /dev/md4, slot 6.
mdadm: superblock on /dev/sdr1 doesn't match others - assembly aborted    


# mdadm --examine /dev/sdk1
/dev/sdk1:
          Magic : a92b4efc
        Version : 00.90.01
           UUID : 232635af:1afe6aa5:e06f8828:99966325

snip

quoted

# mdadm --examine /dev/sdr1
/dev/sdr1:
          Magic : a92b4efc
        Version : 00.90.01
           UUID : 232605af:1afe6aa5:e06f8828:99966325

Putting those UUIDs right next to each other:

quoted

           UUID : 232635af:1afe6aa5:e06f8828:99966325
           UUID : 232605af:1afe6aa5:e06f8828:99966325

                        ^
                        |

We have 2 bits of difference.  This suggests a bad drive, bad RAM, bad
cabling or.... something.

You could be able to recreate the array:

 mdadm -C /dev/md4 -l5 -n8 -c4 -pls /dev/sdk1 /dev/sdl1 /dev/sdm1 \
         missing /dev/sdo1 /dev/sdp1 /dev/sdq1 /dev/sdr1

(double check all of that before you actually run it, I think I have
the order right...)
but be warned that you have some dodgey hardware somewhere, and data
corruption is a real possibility.... What caused the system crash?

NeilBrown

----------------------------------------------------------------------
This mail sent through L&R NetCon webmail Client