SUMMARY: Es40 "Path to device has been reduced"

From: Bugs (bb1@humboldt.edu)
Date: Fri May 03 2002 - 11:56:29 EDT


Hello friends,

By using the command as suggested by Robert, "hwmgr -show scsi -full",
I was able to come up with what I need to show compaq.
as in:
# hwmgr -show scsi -full | egrep 'stale|valid'
      0 0 0 valid
      0 1 0 valid
      0 3 0 valid
      0 2 0 valid
      0 4 0 valid
      1 0 0 stale
      2 0 0 stale
      1 0 0 stale
      2 0 0 valid
      1 0 1 stale
      2 0 1 valid
      1 0 2 stale
      2 0 2 valid
      1 0 3 stale
      2 0 3 valid
      0 6 0 stale
      1 6 0 valid
      3 0 0 valid
      1 4 0 valid

Thanks to all who responded.
Vincent.D'Antonio, Matt.Morris, Jay Nash, Lecia Lynn,
and Robert Mulley

My original problem:

On Thu, 2 May 2002, Bugs wrote:

->
->Hello,
->Our es40, Tru64 5.1 pk2, with hsg80 just had its first hickup
->after over a year of operation. The Oracle database stopped
->working, and could not be restarted so I rebooted and this
->showed up:
->
->May 2 08:08:22 willow vmunix: cam_logger: SCSI event packet
->May 2 08:08:22 willow vmunix: cam_logger: bus 0 target 6 lun 0
->May 2 08:08:22 willow vmunix: ccmn_path_ping3
->May 2 08:08:22 willow vmunix: Path to device has been reduced
->May 2 08:08:22 willow vmunix: Hard Error Detected
->
->However, the machine came back up, and Oracle is running. Is
->this an indication of a failing part? Also getting a few
->CAM SCSI errors.
->
->Thanks...

=========================================================

>From robert.mulley

Hello,

This would seem to me to indicate that one of the channels/KZPBA/HSG80 ports
is no longer working. The OS has maked this path as being stale/bad and is
now using another channel to the disks, which is what it should do. However
the path probably failed whilst Oracle was waiting for some info causing it
to lock up. Using the hwmgr command show as "hwmgr -show scsi -full" you
should be able to get a list of paths to the disks and one of them will be
marked stale. So get a hardware engineer in to investigate. Could also be
that somebody unplugged the wrong cable :)

Robert Mulley
Unix Admin

Bugs Brouillard Unix system administrator
Humboldt State Univ. Information Technology Services
Arcata, Calif.

email bb1@humboldt.edu



This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:48:40 EDT