Fragmetation issue on TSM

[Rajatsm]

Hi All,

Need a support to resolve below error:

Date/Time Message
-------------------- ----------------------------------------------------------
2017-05-03 15:58:27 ANR0984I Process 1814 for MOVE DATA started in the
BACKGROUND at 15:58:27. (SESSION: 396129, PROCESS: 1814)
2017-05-03 15:58:27 ANR1140I Move data process started for volume H06161L3
(process ID 1814). (SESSION: 396129, PROCESS: 1814)
2017-05-03 15:58:27 ANR3247W Process 1814 skipped 2 files on volume H06161L3
because of pending fragments. (SESSION: 396129, PROCESS:
1814)
2017-05-03 15:58:27 ANR1141I Move data process ended for volume H06161L3.
(SESSION: 396129, PROCESS: 1814)
2017-05-03 15:58:27 ANR0985I Process 1814 for MOVE DATA running in the
BACKGROUND completed with completion state SUCCESS at
15:58:27. (SESSION: 396129, PROCESS: 1814)

Due to Fragmentation issue, Backup copy pool, Move data are getting effected. PMR suggested to upgrade TSM to 7.1.7.0 to 7.1.7.1, which was performed 2 weeks back, but still same issue.

Now need a support to fix this. Please help.

 

[marclant]

PMR suggested to upgrade TSM to 7.1.7.0 to 7.1.7.1, which was performed 2 weeks back, but still same issue.

My suggestion would be to follow-up on that PMR.

 

[opeth]

Just did a quick google of the warning message, ANR3247W and there are several ways that could happen with a few solutions.
Also, should go IBM to ask for "support to fix this" probably.

 

[Rajatsm]

Just did a quick google of the warning message, ANR3247W and there are several ways that could happen with a few solutions.
Also, should go IBM to ask for "support to fix this" probably.

IBM PMR suggested for upgrade, that is already done. But still no result.
What are few solution? share let me try those if that can help.

 

[Rajatsm]

Guys, any solution for Fragmentation. I am still on same page.

 

[[email protected]]

Hallo Rajatsm

I think, that MARCLANT is absolutely right: - you have to insist and follow-up on that PMR.

Additional Remarks:

The two following SWg - Support Infos

1.) http://www-01.ibm.com/support/docview.wss?uid=swg1IT13236 and
2.) http://www-01.ibm.com/support/docview.wss?uid=swg1IT03650

describe also "pending fragments" related issues which should have been fixed.........but.........

There exist also an undocumented option for the FRAGMENT SIZE:

if you do a q node xxx f=d you will see the value "SPLITLARGEObjects yes".
This is related to the FRAGMENT SIZE a node can have.

There is an undocumented option: MAXFRAGMENTSIZE which describe and can influence the value:

q opt MAXFRAGMENTSIZE (with SETOPT you can change this value)

May be you can ask IBM, if it makes sense to reduce this value...

Not a final solution, but that's all what comes into my mind...

rgds Michael