NEC Express5800 Series User Manual page 305

"I/O error" is recorded in the syslog by the cable disconnection during the access to
the Fibre Channel disk:
When the Fibre Channel connected to the primary PCI module Cable is
disconnected during the access to the Fibre Channel disk, logs may be recorded,
which seem that "SCSI disk (/dev/sdd**)I/O error"(** represent numbers)
occurred on the devices which are not running because the internal status
managed by the OS and the actual status on the server side are different. The
operations are not affected.
HASIPMI-related logs are recorded in the log:
The following event logs may be recorded during the system operation. These
logs are recorded when the software that access to BMC could not receive
response from BMC within a given period of time. There is no problem to the
system by this warning.
A timeout occurred for waiting BMC response. Process = (Although Primary
BMC is ready)
A timeout occurred for waiting BMC response. Process = (Secondary BMC is still
running)
SCSI-related logs are recorded in the system event log:
SCSI driver records mounting and unmounting of a module in the system event
log, but these events do not affect the system. However, if they are recorded for
over 5 minutes, contact your sales agent.
<When mounting a module > ("*" is different for each system.)
scsi0: Someone reset channel A
[aic79xx/scsi*] occur Hardware error interrupt.
scsi*: hwerrint, Discard Timer has timed out
scsi*: Recovery Initiated - Card was paused
scsi*: Recovery Initiated - Card was not pausedA timeout occurred for waiting
<When unmounting a module > ("*" is different for each system.)
scsi*: Recovery Initiated - Card was paused
scsi*: Recovery Initiated - Card was not paused
scsi*: Recovery Initiated - Card was paused, not connected.
scsi*: Recovery Initiated - Card was not paused, not connected.
(scsi*:A:0:0): SCB 0x0 - timed out
(scsi*:A:0:0): Other SCB Timeout
(scsi*:A:0:0): No other SCB worth waiting for...
[aic79xx/scsi*] HBA seems to be removed.
Troubleshooting 7-33