Sun Fire[TM] V440 server: SCSI errors ASC: 0x29 ASCQ: 0x3

Asset ID:	1-72-1012440.1
Update Date:	2018-02-14
Keywords:

Solution Type Problem Resolution Sure

Solution 1012440.1 : Sun Fire[TM] V440 server: SCSI errors ASC: 0x29 ASCQ: 0x3

Applies to:

Sun Netra 440 Server - Version Not Applicable and later
Sun Fire V440 Server - Version Not Applicable and later
All Platforms

Symptoms

The following errors from the internal hard-disks are sometimes seen on Sun Fire[TM] V440 and Netra 440 servers.

scsi: [ID 107833 kern.warning] WARNING: /pci@1f,700000/scsi@2/sd@2,0 (sd3):
    Error for Command: write             Error Level: Informational  
scsi: [ID 107833 kern.notice]         Requested Block: 16 Error Block: 16
scsi: [ID 107833 kern.notice]         Vendor: FUJITSU Serial Number: 0323Q02MDH
scsi: [ID 107833 kern.notice]         Sense Key: Unit Attention
scsi: [ID 107833 kern.notice]         ASC: 0x29 (<vendor unique code 0x29>), 
    ASCQ: 0x3, FRU: 0x0

Most of the ASC would be 0x29 and the ASCQ be 0x3.

Cause

These warning messages are aften caused by Bug 4886938. Sun CR: 4886938 ==> Oracle BugID: 15166622

Solution

To discuss this information further with Oracle experts and industry peers, we encourage you to review, join or start a discussion in the My Oracle Support Community - SPARC Legacy Servers

Bug 4886938 can be resolved by applying one of the following patches as applicable:

Solaris[TM] 8:
Patch 115275-02 or later (SunOS 5.8: mpt driver patch)

Solaris{TM} 9:
Patch 115665-01 or later (SunOS 5.9: Chalupa platform support patch)
OR
Patch 115663-02 or later ( WITHDRAWN PATCH Obsoleted by: 115665-03 SunOS 5.9: kernel/drv/mpt patch)

It is strongly suggested that the most current version of the applicable patch be installed, not merely the minimum version to fix the issue.

Under no circumstances should the disk(s) be replaced as a result of this error unless one of the above patches has first been installed.

If the applicable patch HAS been installed, then further investigation is required in order to determine the appropriate corrective action.

Sun CR: 4886938 ==> Oracle BugID: 15166622

The device reset is coming from the mpt driver which has determined that a disconnected command timeout has occurred.

The command is an inquiry command coming from cfgadm. PICL issues cfgadm command every time it runs to figure out the disks state to determine if the OK-To-Remove LED should be lit. There is a small timing window in which the cfgadm inquiry command could be incorrectly timed out. This is due to the fact that the inquiry command has a pkt_time (timeout) set to 5 seconds and the MPT's scsi_watchdog_tick is set to 10 seconds. If the watchdog thread fired later than the SCSI inquiry command, the command could be incorrectly timed out.

Additional Reference: (Further investigation)

Other than BugID 4886938 which is addressed in this document, please take note of FIN I1040-1 for BugID 549558 which is related.

ASC 0x29, ASCQ 0x3, 0x29, V440, Netra 440, SCSI, SCSI Error, unit attention, internal, disk
Previously Published As
79291

Attachments

This solution has no attachment