Data Transfer Phase - Data Deletion Start in the Receiver System

[Former Member]

Hi All,

Our migration package is now running at "Data Deletion Start in the Receiver System" step. However, it's running there for almost 3days now. We noticed that is running for table VBOX and FAGLFLEXA. It appears that the deletion is running for table VBOX for more than 17 hours now, while for table FAGLFLEXA, it seems calculating for 9 hours.

TD09P_FAGLFLEXA  @4A@ 0 @DH@ 03.06.2014 04:14:27   376.372.732 330.056.308 0 0 09h 12m 56s

TD09P_VBOX 1 @4A@ 33 @DH@ 02.06.2014 19:47:56   0 107.988.496 36.203.826 36.203.826 17h 39m 27s

TD09P_VBOX 2 @4A@ 34 @DH@ 02.06.2014 19:47:56   0 107.958.008 36.752.646 36.752.646 17h 39m 27s

TD09P_VBOX 3 @4A@ 33 @DH@ 02.06.2014 19:47:56   0 107.958.008 36.374.570 36.374.570 17h 39m 27s

TD09P_VBOX 4 @4A@ 32 @DH@ 02.06.2014 19:47:56   0 107.958.008 35.240.342 35.240.342 17h 39m 27s

TD09P_VBOX 5 @4A@ 33 @DH@ 02.06.2014 19:47:56   0 107.958.008 35.892.828 35.892.828 17h 39m 27s

TD09P_VBOX 6 @4A@ 33 @DH@ 02.06.2014 19:47:56   0 107.929.235 35.947.710 35.947.710 17h 39m 27s

These are large tables. Can anyone share how can we optimize the deletion and calculation activity on these tables, and also share how can we optimize the TDMS processes runtimes, in the System Analysis Phase and Data Transfer Phase, as these 2 phase is where we encountered the to slow runtime.

We are at below level.

DMIS 2011_1_700 0005 SAPK-11305INDMIS

DMIS_CNT 2011_1_700 0005 SAPK-11305INDMISCNT

We would appreciate your response for inputs. Thanks.

Regards,

Philip

[Former Member]

Hello Philip,

Refer to the link http://scn.sap.com/docs/DOC-53873 for improving performance of data deletion activity.

Refer to the link http://scn.sap.com/docs/DOC-49910 for improving performance of data selection and data transfer activities.

Thanks

Anita

[Former Member]

Thanks Anita,

We will review these link to improve our current systems.

However, how about currently runnning Data transfer? Do you know of any way we can improve, especially for processes running for large tables, particularly VBOX, FAGLFLEXA, and the likes.

Thanks,

Philip

[Former Member]

Check if there are active secondary indices on these tables, if yes then stop the activity drop the indices and then restart the activity. Make sure to stop the activity using troubleshooter only, don't cancel any job manually. troubleshooter may take som time to stop the current execution of activity, you can monitor the progress of troubleshooter in extended process monitor.

[Former Member]

Hi Anita,

Thanks. Checked and found no secondary indices on these tables. The only existing are those standard indices for both the tables.

Regards,

Philip

[Former Member]

hi Philip, in that case it won't be possible to analyse the performance issue without having a look in the system as performance issues are specific to system environment. I would suggest you to raise an oss message in component ca-tdm. Meanwhile can you update the screen shot of current running activities from process tree ? By the way where did you check indices ? Receiver system right ?

Thanks

anita

[Former Member]

Hi Anita,

Please find attached screenshot of the current running process tree.

Also, yes, I checked the indices in the receiver.

Regards,

Philip

[Former Member]

HI Philip,

do one thing, go to transaction DTLMON and input the mass transfer id and access plan for your package. There go to view absolute runtime distribution and sort the column run and then upload the screen shot here. Also test all Tdms rfc connections, the response time should be 10-15 ms.

thanks

anita

[amit_sharma36]

You can delete even standard index. You can use drop index in receiver system troubleshoter for teh same in TDMS process tree.