Sunday, January 2, 2011

Basic troubleshooting for library and tape drives in AIX and TSM

Basic troubleshooting for library and tape drives in AIX
Source:
IBM Technote:
http://www-01.ibm.com/support/docview.wss?uid=swg21406601

Many times issues with Tivoli Storage Manager Server in AIX connecting to tape libraries and/or drives can be resolved by checking some basics. The first things to do when addressing hardware issues are given in the steps below.

1) AIX: Does AIX have the correct device driver for the library and/or tape drives?
2) AIX: Does AIX list the library and/or tape drives as Available?
If 6.1, does the db2 instance owner account have read/write access?
3) TSM: Does TSM list the drives as online?
4) TSM: Does TSM list the paths as online and using the correct device name?
5) AIX: Can you test the library and/or drives from device utilities in AIX?
6) TSM: Does reconfiguring the library and/or drives in AIX and TSM resolve the issue?



1) Verify that AIX has the correct device driver for the library and/or tape drives

IBM tape drives and libraries typically use Atape (currently 11.7.7.0) or atldd (for IBM 3494 tape libraries, currently 6.8.4.0), though there are a few exceptions that use tsmscsi. Non-IBM tape drives and libraries typically use the tsmscsi device driver (at exactly the same version as the TSM Server version, to 4 digits) that comes packaged with TSM Server and Storage Agent, though there are a few exceptions that use other device drivers. A non-IBM library that uses IBM tape drives will typically use tsmscsi for the library and IBMtape for the drives.

Check the TSM supported device list for which device driver is needed for supported tape libraries and drives
http://www-01.ibm.com/software/sysmgmt/products/support/IBM_TSM_Supported_Devices_for_AIXHPSUNWIN.html

Check current versions installed with:
      lslpp -l Atape.*
      lslpp -l atldd.* (for IBM 3494 libraries)
      lslpp -l tivoli.tsm.devices.* (for non-IBM libraries)

Atape and atldd (for IBM 3494 libraries) are available for download at the following ftp site:
ftp://service.software.ibm.com/storage/devdrvr/AIX/

To install and commit the Atape driver, uninstall the older version:
      installp -u Atape.driver
Then install the new vesrion. For example, if you downloaded the file to /tmp/Atape.11.7.7.0.engr.bin:
      installp -acXd /tmp/Atape.11.7.7.0.engr.bin all

Configure the tape device:
      cfgmgr -v
(-v is not required but will show where it hangs if it does)

Verify the new devices are Available:
      lsdev -Cc tape

For details, see Technote # 1174910: Upgrading Atape device driver in AIX
http://www-01.ibm.com/support/docview.wss?uid=swg21174910

For details on configuring non-IBM library and/or tape drives to use tsmscsi in AIX, see TSM Server 5.5 for AIX Administrator's Guide, Ch. 5: "Attaching Devices to the Server System", Section: "Installing and Configuring Device Drivers", Subsections: "Configuring Tivoli Storage Manager Device Drivers for Autochangers" and "Configuring Tivoli Storage Manager Device Drivers for Tape or Optical Drives"
http://publib.boulder.ibm.com/infocenter/tivihelp/v1r1/topic/com.ibm.itsmaixn.doc/anragd55103.htm#crddl

2) Verify that AIX lists the library and/or tape drives
From AIX, run:
    * For tape devices using Atape or atldd
      # lsdev -Cc tape
    * For tape devices using tsmscsi
      # lsdev -Cc adsm

Tape devices should be listed as Available.
AIX device names:
    * Tape drives that use tsmsci*
      /dev/mtx (for example: /dev/mt0)
    * Tape libraries that use tsmsci*
      /dev/lbx (for example: /dev/lb0)
    * IBM 3575, 3581, 3583, 3584 libraries
      /dev/smcx (for example: /dev/smc0)
    * IBM® 349X libraries
      /dev/lmcpx (for example: dev/lmcp0)
* tsmsci: Tivoli Storage Manager device driver, packaged with TSM Server

Reference: TSM Server 5.5 Administrator's Guide, Ch 5. "Attaching Devices to the Server System", Section: "Installing and Configuring Device Drivers", Subsection: "Determining Device Special File Names"
http://publib.boulder.ibm.com/infocenter/tivihelp/v1r1/topic/com.ibm.itsmaixn.doc/anragd55100.htm#spclfl

For TSM Server 6.1, does the db2 instance owner account have read/write access?
      # ls -l /dev

For example:
      # ls -l rmt[0-9]
      crw-rw-rwT 1 root system 40,192 Mar 08 14:02 rmt1
      crw-rw-rwT 1 root system 40,256 Mar 08 14:02 rmt2

3) Check TSM Server drives are online:
From a TSM server command line run
      q drive

If drive(s) are not online, update them with:
      update drive online=yes

4) Check the TSM Server paths are online and using the correct device name
From a TSM server command line run
      q path f=d

If path(s) are not online or are using the wrong device device name (for example /dev/smc0 or /dev/rmt0), update them with the following (if device names are correct, you don't need to include the device parm):
Library (for example, using smc0):
      update path srct=server destt=libr device=/dev/smc0 online=yes

Tape drive (for example, using rmt0):
      update path srct=server destt=drive libr= device=/dev/rmt0 online=yes

5) Check for hardware issues from AIX
From an AIX command prompt window, use tapeutil (for devices using IBMtape) or lbtest or mttest (located in /usr/tivoli/tsm/devices/bin, for devices using tsmsci) to test the library or drives.

5a) lbtest, for libraries using tsmscsi
From AIX, run:
lbtest
Go into manual test and open the library
1: Manual test
1: set device special file
Enter device special file name:
For example: /dev/smc0 [enter]
6: open
Get library inventory
8: ioctl return element count
10: ioctl return library inventory
Move a tape from a tape slot to a drive
11: ioctl move medium
transport element address: XX
source address moving from: YY
destination address moving to: ZZ
Move the tape back to where it came from
11: ioctl move medium
transport element address: XX
source address moving from: ZZ
destination address moving to: XX
Close the device and exit lbtest
7: close
99: return to main menu
9: exit lbtest
Reference: Technote # 1326205: LBTEST - Practical Examples to Test Functionality
http://www-01.ibm.com/support/docview.wss?uid=swg21326205

5b) mttest, for tape drives using tsmscsi
From AIX, run:
mttest
Go into manual test and open the drive
1: Manual test
1: set device special file
Enter device special file name:
For example: /dev/rmt0 [enter]
20: open
Check basic functionality:
 22. read
 60. write block to "block.out"

Note: You will get an error stating that only 80 bytes could be read out of 1024. This is normal and can be ignored

Close the device and exit mttest
21: close
99: return to main menu
9: exit mttest

5c) tapeutil, for libraries and tape drives using IBMtape
tapeutil operates basically the same as lbtest and mttest, except that you access both the library and the drives from tapeutil.

A good basic test for tape drives in Base Mode is
6: test unit ready

6) Delete and redefine paths, drive, and library, and checkin scratch tapes, then private
Sometimes deleting and redefining everything works to reconnect TSM Server to the devices correctly. The following technote gives these steps in detail.

Redefining TSM Library and Drives for UNIX OS
http://radek-it.blogspot.com/2010/09/redefining-tsm-library-and-drives-for.html

No comments:

Post a Comment