PE 1850 mptbase error

[kxiluri at email.arizona.edu]

Dear List,

I am receiving the following error from a Power Edge 1850
which is a member of a cluster, attached to a RAID, with an LSI
FC949X card.

kernel: scsi : aborting command due to
timeout : pid 29096354, scsi1, channel 0, id 0, lun 0 Write (10) 00 7e
27 a2 c8 00 00 10 00
mptscsih: ioc1: id=0 OldAbort: scheduling ABORT SCSI IO (sc=f72e9a00)

This happens when users run a highly intense I/O code and the
node hangs. Top CPU average load reaches 40 at least.

kernel 2.4.21-58.ELsmp
BIOS A07
cat /proc/scsi/mptscsih/1
ioc1:LSIFC949X, FwRev=01031300h, Ports=1, MaxQ=203

cat /proc/mpt/version
mptlinux-2.06.16.02
Fusion MPT base driver
Fusion SCSI host driver
Fusion MPT ioctl driver

Would anyone have a script i can run in a cron job perhaps to
kill the code that causes the system to hang.

The LSI driver rpm does not apply to the current kernel
Any thoughts would be much appreciated.

Many thanks
Kiriaki