[SunHELP] scsi warnings FC4700

[Markham, Richard]

On boot I'm getting scsi warnings for each LUN which is assigned to a
particular host. An example is below the only thing is see changing is the
requested block and LUN device path.  I can't get an answer from EMC on how
to diagnose this issue. From the Clarrion standpoint everything checks out
and there is no error in the event log that matches up with this warning.
Since the error is retryable I am assuming it completes what it needs to do
and moves on.  I haven't seen any issue with data integrity which should be
immediately obvious since an Oracle database resides on these LUN's.
Perhaps this is a timing issue, but I have seen these errors thrown when
initially setting up powerpath during the "powermt config" command.



Dec  7 12:43:20 oracledb scsi: WARNING:
/ssm at 0,0/pci at 19,600000/fibre-channel at 1/sd at 0,3 (sd122): Dec  7 12:43:20
oracledb        Error for Command: write                   Error
 Level: Retryable
Dec  7 12:43:20 oracledb scsi:  Requested Block: 98513
Error
 Block: 98513
Dec  7 12:43:20 oracledb scsi:  Vendor: DGC
Seria
l Number: 010599333ACL
Dec  7 12:43:20 oracledb scsi:  Sense Key: Unit Attention
Dec  7 12:43:20 oracledb scsi:  ASC: 0x2a (<vendor unique code 0x2a>), ASCQ:
0x8
0, FRU: 0x0

Please advise on  what I should do to diagnose this issue.

Clarrion FC4700
Powerpath 4.0.2
Solaris 8 (Dec 02'03 Recommended cluster)