Critical: server hangs because of scsi resets

[Mihails Nikitins]

Hi,

I see the following on my console.

SCSI host 1 abort (pid 1207191) timed out - resetting
SCSI bus is being reset for host 1 channel 0.
ncr53c8xx_reset: pid=1207191 reset_flags=2 serial_number=1207192
serial_number_at_timeout=1207192
SCSI host 1 abort (pid 1207201) timed out - resetting
SCSI bus is being reset for host 1 channel 0.

The system hangs because of the resets. The problem repeats time by time.

I run K210 under linux-2.4.23-pa5. The problem appears on different scsi
controllers, so I hope there are neither controller nor disk problems. This
are current drivers from /var/log/messages.

kernel: loop: loaded (max 8 devices)
kernel: Found i82596 at 0xffd07000, IRQ 87
kernel: eth0: 82596 at 0xffd07000, 08 00 09 DA 44 88 IRQ 87.
kernel: 82596.c $Revision: 1.32 $
kernel: SCSI subsystem driver Revision: 1.00
kernel: zalon_scsi_callback: Zalon vers field is 0x1, IRQ 35
kernel: ncr53c8xx: 53c720 detected
kernel: ncr53c720-0: ID 7, Fast-10, Parity Checking, Differential
kernel: zalon_scsi_callback: Zalon vers field is 0x1, IRQ 36
kernel: ncr53c8xx: 53c720 detected
kernel: ncr53c720-1: ID 7, Fast-10, Parity Checking, Differential
kernel: scsi0 : ncr53c8xx-3.4.3b-20010512
kernel: scsi1 : ncr53c8xx-3.4.3b-20010512

Can it be a bad driver problem? I use Linux Software RAID, is it possible to
tune some parameters there?

Many thanks in advance for any help!