LTO Tape Drive
 
Sun StorageTek SL3000 Modular Library System
 
Oracle Database - Enterprise Edition
 

PLA-Support>Sun Systems>TAPE>Tape Hardware>SN-TP: SL3000-8500 Library
 

In this Document

Symptoms

Changes

Cause

Solution

Created from <SR 3-7095043881>

Applies to:

Symptoms

For past 3 Fridays, LTO drives will hang with tape in the drive.

Below is customer's update and observation working with NBU support.

"No backups that were out of the ordinary could be identified in Networker as a culprit.  From what I can tell RMAN doesn't play a role either, though will ask the RMAN expert on Monday.
Time line of all three failures start around 19:15 on Friday.  All backups were left running in each circumstance, which for the most part completed without intervention from me after the library was put back into service.
From the networker logs I pulled the following time lines around all three events: The only time the library was taken off line was the first event on April 05, 2013.
First indicators first event --
04/05/13 19:20:22  nsrd media warning: rd=boxmgr:/boxmgr_bay03 writing: I/O error, at file 393 record 2267
04/05/13 19:20:23  nsrmmgd lcpd 1 at host devnull reported error 'Jukebox:SL3000 access:/dev/scsi/changer/c5t500104F000B09137d0 failed:MOVE MEDIUM key:4 status:CHECK CONDITION Hardware Error, Diagnostic

Failure on Component 0x02
04/05/13 19:20:40  nsrmmd failed to open ndmp tape device, no device is at the specified target
04/05/13 19:20:42  nsrd media warning: rd=boxmgr:/boxmgr_bay02 writing: I/O error, at file 160 record 1916
04/05/13 19:31:52  rd=wastebin:/wastebin_bay04 eject operation failed: read open error: a system error occurred - see errno for more information (drive status is DRIVE_STATUS_UNSET)
04/05/13 19:36:13  /devnull_bay08 eject operation failed: read open error: there was not a device file name passed in to CDI (drive status is DRIVE_STATUS_UNSET)

Secondary indicators first event --
04/05/13 19:32:40  rd=bitbucket:/bitbucket_bay09 <NULL> operation failed: read open error: a system error occurred - see errno for more information (drive status is Drive reports no error - but state is

unknown)
04/05/13 19:33:53  /devnull_bay07 eject operation failed: read open error: a system error occurred - see errno for more information (drive status is Drive reports no error - but state is unknown)
04/05/13 19:46:47  rd=boxmgr:/boxmgr_bay03 eject operation failed: read open error: a system error occurred - see errno for more information (drive status is Drive reports no error - but state is unknown)
04/05/13 19:57:55  rd=bitbucket:/bitbucket_bay10 eject operation failed: read open error: a system error occurred - see errno for more information (drive status is Drive reports no error - but state is

unknown)
04/05/13 19:57:55  rd=boxmgr:/boxmgr_bay02 eject operation failed: read open error: a system error occurred - see errno for more information (drive status is Drive reports no error - but state is unknown)
04/05/13 20:11:41  rd=wastebin:/wastebin_bay04 eject operation failed: read open error: a system error occurred - see errno for more information (drive status is DRIVE_STATUS_UNSET)
04/05/13 21:15:39  /devnull_bay07 eject operation failed: read open error: a system error occurred - see errno for more information (drive status is Drive reports no error - but state is unknown)"

Changes

 Customer implemented NEXPOSE - Vulnerability Management Software.

Cause

 NEXPOSE - Scans customer IP network for any vulnerability. This software is intrusive to tape device ethernet interface.

Solution

Customer will need to partition NEXPOSE to not scan any tape drives or to isolate tape drives IP subnet as a private network. 

This solution has no attachment