Saturday, July 7, 2012

HA GROUP ERROR: DISK/SHELF COUNT MISMATCH ERROR

When a disk enters maintenance center on one node (There will be lot of reason behind this (One reason is it’s not a valid disk to use it)), the partner may not be aware of this. This may result in one node excluding the disk in its disk inventory, while the partner including the same disk in its inventory. This will result in generating "cf.disk.inventory.mismatch" and "CLUSTER ERROR: DISK/SHELF COUNT MISMATCH" auto support.

If the disk in the maintenance center is repaired after testing, the disk inventory mismatch will be automatically resolved. If the disk in the maintenance center is failed after testing, the disk should be removed.


filer1> disk show 1b.57
DISK OWNER POOL SERIAL NUMBER HOME
------------ ------------- ----- ------------- -------------
1b.57 filer2 (151704668) Pool0 3SK0AFTX0000902778L9 filer2 (151704668)
whmn3849*>
filer2> disk show 1b.57
DISK OWNER POOL SERIAL NUMBER HOME
------------ ------------- ----- ------------- -------------
1b.57 filer2 (151704668) FAILED 3SK0AFTX0000902778L9 filer2 (151704668)
Filer2>

From syslog (/etc/messages)

Fri Jul  6 23:02:24 EDT [filer1: raid.disk.missing:info]: Disk 0b.57 Shelf 3 Bay 9 [NETAPP   X291_S15K7420F15 NA00] S/N [3SK0AFTX0000902778L9] is missing from the system

Fri Jul  6 18:52:10 EDT [filer1: cf.disk.inventory.mismatch:CRITICAL]: Status of the disk 1b.57 (20000024:B6564A61:00000000:00000000:00000000:00000000:00000000:00000000:00000000:00000000) has recently changed or the node (filer2) is missing the disk.

But after sometime it will completely go to failed pool if ONTAB failed to rectify the soft error on that disk if not it will back to inventory stating that soft error has been fixed. See below message from syslog.

Fri Jul  6 20:41:11 EDT [filer1: cf.disk.inventory.mismatchOK:info]: The node (filer2) included the disk 0b.57 (20000024:B6564A61:00000000:00000000:00000000:00000000:00000000:00000000:00000000:00000000) in its inventory.

No comments:

Post a Comment