Sun Fire X2200 M2 Server
 
Sun Ultra 20 Workstation
 
Sun Ultra 40 M2 Workstation
 
Sun Fire X4540 Server
 
Sun Fire X4440 Server
 
Sun Java Workstation W1100z
 
Sun Fire X4200 Server
 
Sun Ultra 40 Workstation
 
Sun Fire X4200 M2 Server
 
Sun Fire X4600 M2 Server
 
Sun Ultra 20 M2 Workstation
 
Sun Fire X4100 M2 Server
 
Sun Fire X2100 M2 Server
 
Sun Fire X4100 Server
 
Sun Fire X4600 Server
 
Sun Fire X2100 Server
 
Sun Fire X4500 Server
 
Sun Java Workstation W2100z
 

PLA-Support>Sun Systems>x64>Server>SN-x64: MISC-SERVER
 
.Old GCS Categories>Sun Microsystems>Servers>x64 Servers
 

Applies to:

Goal

Symptoms

• disk errors.

Purpose/Scope:

This document describes how to identify if a Solaris[TM] x64 operating environments disk is online/offline or has reported errors. This document does not detail how to recover data or replace storage devices.

Fix

Steps to Follow

Always perform disk management commands as a root or UID0 user.

CHECKING DRIVE STATUS:

Execute the command "/usr/sbin/format" while logged in as root. Check for the following "drive type unknown" message:

#/usr/sbin/format

AVAILABLE DISK SELECTIONS:

1. c2t0d0 SEAGATE cyl 3817 alt 2 hd 128 sec 32

/pci@0,0/pci108e,5351@1f/disk@1,0

2. c2t1d0 <drive type unknown>

/pci@0,0/pci108e,5351@1f/disk@1,1

Line numbers and controller numbers (cXtXdX) may vary depending on failure condition, highlighted in BOLD.

This message appears when a disk was originally available to the operating system but now does not contain a valid disk label (VTOC) or the operating system is unable to read the disk label. This is an indication of a disk being corrupt.
If your pre-configured disk is currently showing as "drive type unknown" then the disk has become unavailable or the VTOC is corrupt therefore the disk should be replaced as a precaution. If however all disks show without this message then further investigation is required.

FURTHER INVESTIGATION:

Investigate further by running the following command:

# /usr/sbin/cfgadm

The resulting output will be similar to below depending on platform type and configuration:

#/usr/sbin/cfgadm

Ap_Id Type Receptacle Occupant Condition

c0 scsi-bus connected configured ok

c0::dsk/c0t0d0 disk connected configured ok

c0::dsk/c0t1d0 disk connected configured unknown

c1 scsi-bus connected configured ok

c1::dsk/c1t2d0 CD-ROM connected configured ok

The above output which may vary depending on configuration refers to a platform that has 2 controllers installed. The first controller is connected to an 2 disk backplane. The second controller is connected to a CD-ROM or DVD device. Possible failures are highlighted in BOLD.

• "Ap Id" refers to the controller and target location information.

• "Type" refers to the media type i.e. Disk, Tape, CD-ROM etc.

• "Receptacle" refers to the physical interface connection i.e. SAS PHY or SCSI bus connection.

• "Occupant" refers to the media devices usage being configured or unconfigured in the operating system.

• "Condition" refers to the reported failure analysis of the disk. This field is useful but may not be supported by older platforms. If ALL disks show as the condition "unknown" then please ignore this column.

Confirm that a disk is not disconnected or unconfigured which may explain why the disk is unavailable to Solaris[TM].
If a disk is unconfigured or disconnected, use the cfgadm command to configure the disk back online. See /usr/bin/man cfgadm(1M) for more details.

Finally, you may wish to confirm error messages have not been logged in the system messages file for a failed disk.
Execute the following command to extract disk related errors from the system messages file:

/usr/bin/grep -i SCSI /var/adm/messages*

We use the keyword SCSI because all storage devices in Solaris including IDE/PATA, SAS, SATA, SCSI, and USB emulate SCSI to be represented as a storage device. Therefore, most error messages reported in the system messages file are prefixed with the word SCSI.

Replace as a precaution any disks that show a siginficantly high number of read or write SCSI ASC/ASQ media errors if those errors are local to a particular disk only. It is normal to see a few errors logged against various targets but high numbers on any one particular target are a sign of a pending failure. Occasional read or write media errors are normal, although they may cause disk management software to 'fail' RAID volume components, this does not necessarily indicate a disk HW failure. However, repeated media errors on several different blocks may indicate imminent disk failure requiring preventive replacement and should be further investigated.

You sould also search for and consider platform bugs when experiencing a large number of SCSI ASC/ASQ errors. Search My Oracle Support and contact Oracle to verify if a bug is applicable to your platform type / disk type before replacing a disk drive.

Previously Published As 91499

This solution has no attachment