Unable to sense robotic device

NetBackup 6.5 on Solaris 9 (SPARC), with a fibre-attached robot.

Server lost power. When it came up, the robot was unavailable. /var/adm/messages complained

 May 18 10:08:15 thackeray tldcd[326]: [ID 985024 daemon.error] TLD(0) key = 0x5, asc = 0x24, ascq = 0x0, INVALID FIELD IN CDB
May 18 10:08:15 thackeray tldcd[326]: [ID 466614 daemon.error] TLD(0) Mode_sense error
May 18 10:08:15 thackeray tldd[306]: [ID 320639 daemon.error] TLD(0) unavailable: initialization failed: Unable to sense robotic device 

and "sgscan all" showed an unexpected entry for the robot. It used to look like this

/dev/sg/c0t0l0: Changer: "STK     L700"

but now it looks like this

 /dev/sg/c0t0l0: Array-controller: "HP      HSV210" 

and cfgadm -al reports  (I don't know what it used to look like)

 Ap_Id                          Type         Receptacle   Occupant     Condition
c0                             scsi-bus     connected    configured   unknown
c0::dsk/c0t0d0                 disk         connected    configured   unknown
c0::dsk/c0t1d0                 disk         connected    configured   unknown
c0::sg/c1t0l0                  unknown      connected    configured   unknown
c0::sg/c1t0l1                  unknown      connected    configured   unknown
c0::sg/c1t0l2                  unknown      connected    configured   unknown
c0::sg/c1t1l0                  unknown      connected    configured   unknown
c0::sg/c1t1l1                  unknown      connected    configured   unknown
c0::sg/c1t1l2                  unknown      connected    configured   unknown
c0::sg/c1t2l0                  unknown      connected    configured   unknown
c0::sg/c1t2l1                  unknown      connected    configured   unknown
c0::sg/c1t2l2                  unknown      connected    configured   unknown
c0::sg/c1t2l3                  unknown      connected    configured   unknown
c1                             scsi-bus     connected    configured   unknown
c1::sg/c2t0l0                  unknown      connected    configured   unknown
c1::sg/c2t0l1                  unknown      connected    configured   unknown
c1::sg/c2t0l2                  unknown      connected    configured   unknown
c1::sg/c2t1l0                  unknown      connected    configured   unknown
c1::sg/c2t1l1                  unknown      connected    configured   unknown
c1::sg/c2t1l2                  unknown      connected    configured   unknown
c1::sg/c2t2l0                  unknown      connected    configured   unknown
c1::sg/c2t2l1                  unknown      connected    configured   unknown
c1::sg/c2t2l2                  unknown      connected    configured   unknown
c1::sg/c2t2l3                  unknown      connected    configured   unknown
c2                             scsi-bus     connected    configured   unknown
c2::dsk/c2t0d0                 CD-ROM       connected    configured   unknown
 

Is cfgadm trying to tell me something, or should I just reinstall the sg driver?

Thanks,

Glo

1 Solution

Accepted Solutions
Highlighted
Accepted Solution!

Do you have Persistent

Do you have Persistent Binding in place for fibre-attached devices? If not, device names are assigned in the order that devices are detected at boot time.

Please post the section of /var/adm/messages where HBA and devices were detected at boot time.

View solution in original post

3 Replies
Highlighted
Accepted Solution!

Do you have Persistent

Do you have Persistent Binding in place for fibre-attached devices? If not, device names are assigned in the order that devices are detected at boot time.

Please post the section of /var/adm/messages where HBA and devices were detected at boot time.

View solution in original post

Highlighted

Agree with Marianne

The device tree has changed so the existing c0t0d0 now points to other device. The message "INVALID FIELD IN CDB"  basically means "I do not understand this command - I am a disk array and not tape robot".

You need to verify path and device drivers point to the right devices.

Highlighted

Oops

Oops, a few months ago we reconfigured the switch in an experiment to grant this server SAN storage, then we abandoned the experiment before changing anything on the server, but also without undoing that change on the switch and without considering its effect on this server. Indeed, I should have studied /var/adm/messages more closely. Thanks!