MSL5052 tape library NSR reboot and error issue

[Jason Antes [2]]

Last night our MSL rebooted unexpectedly with the following error:

2012. 09/11/2006 21:04:40 38d09h14m16.42s EXCEPTION: TLB Load Exception Type : 02
2013. 09/11/2006 21:04:40 38d09h14m16.42s Address : 801337A8 (inside fcpTrns_sendSpoofResponse)
2014. 09/11/2006 21:04:40 38d09h14m16.42s Return Address : 8013377C (inside fcpTrns_sendSpoofResponse)
2015. 09/11/2006 21:04:40 38d09h14m16.43s Virtual Address: 00000018 Task : psOutQ
2016. 09/11/2006 21:05:46 0d00h00m42.50s New device is added to location 0/0/0
2017. 09/11/2006 21:05:46 0d00h00m42.50s New device is added to location 0/1/0
2018. 09/11/2006 21:05:46 0d00h00m42.50s New device is added to location 1/2/0
2019. 09/11/2006 21:05:46 0d00h00m42.62s FC Port 0 Link is UP.
2020. 09/11/2006 21:05:46 0d00h00m42.62s Unit restart and initialization, Firmware Version: 5.6 Build Level: 5.6.78

Previous and following events include:

2002. 09/09/2006 09:13:03 35d21h22m40.15s SCSI UDC Error. Port 0 Target 0, S_ID 0xFFFFFF, CDB[0] 0x00, Device Type 31
2003. 09/09/2006 09:23:34 35d21h33m11.14s SCSI UDC Error. Port 0 Target 0, S_ID 0xFFFFFF, CDB[0] 0x00, Device Type 31
2004. 09/11/2006 07:05:10 37d19h14m46.66s CHK COND with Sense Key=0x3, ASC=0x11, ASCQ=0x0 from SCSI LUN 0/1/0

and

2021. 09/12/2006 01:07:19 0d04h02m15.60s CHK COND with Sense Key=0x3, ASC=0x11, ASCQ=0x0 from SCSI LUN 0/1/0

We are running Windows 2003 with DP 5.5. The MSL is fiber connected and zoned properly. I've looked around and seen that the CHK COND messages could be attributed to a firmware issue or the RsetPTRLO option on a connected servers' HBA. We have several servers zoned into the tape library but only 1 of them reported the reboot via an Insight Manager event. The Insight Manager trap flags are 16 for the drives and controller.

[Marino Meloni_1]

2004. 09/11/2006 07:05:10 37d19h14m46.66s CHK COND with Sense Key=0x3, ASC=0x11, ASCQ=0x0 from SCSI LUN 0/1/0

this is indicating a medium error read not recovered error


2002. 09/09/2006 09:13:03 35d21h22m40.15s SCSI UDC Error. Port 0 Target 0, S_ID 0xFFFFFF, CDB[0] 0x00, Device Type 31

This is indicating a Unexpected Disconnect on the device in SCSI Parallel port O

I think that you should troubleshoot your drive, or scsi connection between NSR and drive to see if errors on the scsi path cause trouble to the NSR and causing the reboot.

The cause may also be a bad cartidge.

You can use LTT and run acceptance test to identify the cause of the problem

Marino