SL150 - Getting Error Message From Different Drives, HP Data Protector Showing in its Logs (No Such Device or Address)

Asset ID:	1-72-1929148.1
Update Date:	2017-05-18
Keywords:

Solution Type Problem Resolution Sure

Solution 1929148.1 : SL150 - Getting Error Message From Different Drives, HP Data Protector Showing in its Logs (No Such Device or Address)

Applies to:

StorageTek SL150 Modular Tape Library - Version All Versions and later
HP LTO6 Tape Drive - Version All Versions and later
Information in this document applies to any platform.

Symptoms

FCP ERROR

AIX error log was loads of errors by fcs1, which is one of the HBAs where the host has access to tapes.

HP Data Protector showing in its LOGs (no such device or address) our tape drives HP LTO6.

---------------------------------------------------------------------------------------------------------

[Major] From: BMA@bwdb.cosan.rede "TD6_DRV_02_R1" Time: 29/07/2014 16:23:39 [90:173] REQUEST SENSE call failed. Possible cleanme signal detection failure.
([6] No such device or address)

On debug.log:
07/29/14 16:23:39 BMA.36569266.0 ["/ma/spt/fc_aix.c $Rev: 31110 $ $Date:: 2012-05-17 15:44:31":1347] A.07.00 b105
FC_AIX_io: ioctl(fd=8, SCTL_IOCTL, cmd) failed. cdb_length=6, {errno=6}

07/29/14 16:23:39 BMA.36569266.0 ["/ma/dev/devseq.c $Rev: 35661 $ $Date:: 2013-01-23 17:24:19":7769] A.07.00 b105
AIX_SeqWrite: /dev/rmt18.1- write() fails: {6}

07/29/14 16:23:39 BMA.36569266.0 ["/ma/spt/fc_aix.c $Rev: 31110 $ $Date:: 2012-05-17 15:44:31":1347] A.07.00 b105
FC_AIX_io: ioctl(fd=8, SCTL_IOCTL, cmd) failed. cdb_length=6, {errno=6}

07/29/14 16:24:16 OB2BAR_SBT_CHANNEL.49741922.0 ["/integ/bar2/ob2common.c $Rev: 12815 $ $Date:: 2010-09-07 18:09:57":700] A.07.00 b105
[OB2_BmWriteData] RmWriteObjData(a00010000100138, 0, 261832, 0, 0, 0, 0) failed.

Function trace (currentLevel=0):
Level 0: sbtwrite2()

Cause

When using encryption without Dynamic Tracking, firmware upgrades, interop mode changes, or reboots, etc., can cause VI/VT PID change, and this could result in AIX or HPUX host losing connectivity. This is because the proxy virtual initiators and virtual targets may change their internal PID and therefore the host may have to rescan for the devices after a firmware upgrade or other encryption configuration change. Most of the operating systems do a dynamic rescan by default. Whenever there is a fabric change, host receives RSCN and rescans the fabric. In the case of AIX, rescan is done only when Dynamic Tracking is enabled. Reboot of host also does this rescan during OS boot up.

Solution

Dynamic Tracking was recommended to turn off on HBAs in the AIX server, no more hangs (freeze) have been seen in the library.

Disable dynamic tracking in both HBAs following one of the options below:

a.If you can not stop the I / O for all the tapes:

# chdev -l fscsi0 -a dyntrk=no -P
# chdev -l fscsi1 -a dyntrk=no -P

Note: After that you need to perform the reboot of AIX to validate the change.b. as this may interrupt I/O for all tapes as:

# rmdev -Rl fcs0
# rmdev -Rl fcs1
# chdev -l fscsi0 -a dyntrk=no
# chdev -l fscsi1 -a dyntrk=no
# cfgmgr

Note: Before running the chdev verify that the HBAs and Tapes are in the Defined status.

* Dynamic Tracking of Fibre Channel Devices
http://www-01.ibm.com/support/knowledgecenter/api/content/ssw_aix_71/com.ibm.aix.kernextc/dynamictracking.htm

References

<NOTE:1587435.1> - SL150 - How to change the designation of the Bridge Drive.

Attachments

This solution has no attachment