MSA2312sa: disk fails but not automatically removed

[Peter Kruse2]

Hello,

two such disk arrays are connected to two
Proliant DL380 G6 running Ubuntu 10.04 LTS.
The Firmware of the Management Controller
Code Version is W441R06.
Each disk is mapped to a logical Volume and
mapped to both hosts. This is a HA setup.
(Pacemaker makes sure the disks are only used
on one hosts at a time).
Two software raids level 10 are created,
one with the SAS disks and one with the SATA
disks.
At one time the kernel (2.6.32-23-server)
logs:

mptscsih: ioc0: attempting task abort! (sc=ffff8806122f8f00)
sd 2:0:0:6: [sdc] CDB: Write(10): 2a 00 05 25 85 00 00 01 00 00
mptbase: ioc0: LogInfo(0x31140000): Originator={PL}, Code={IO Executed}, SubCode(0x0000)
mptscsih: ioc0: task abort: SUCCESS (sc=ffff8806122f8f00)

and continues indefinitely, the disk obviously
is broken but never disabled which causes all
applications trying to access the storage to
block. Shouldn't the disk be disabled
automatically? Or is this rather a Linux
question in which case I should probably
ask the Linux forum.

[Peter Kruse2]

apparently a new firmware fixed this issue, thanks.