Improve Extraction performance for 0BBP_TD_PO_1

[Former Member]

Hi,

I am using the datasource 0BBP_TD_PO_1 to bring in EBP PO's into BW. During the init/full loads from this datasource, the extraction time for 25,000 records is around 45mins, which is unacceptable.

Has anybody come across this situation, any advices will be appreciated.

Thank You

Ann

[Former Member · Accepted solution]

Hi,

First determine where the time is being taken, i.e during extraction itself at the source system side or while posting the data on the BW side.

Once you determine that, if the problem is on the BW side, you can check the indexes and drop and recreate them, or if there are missing indexes you can take help from the DBAs to have them created. Also check to see if the MD tables need to be reorged on the BW side.

If the problem is on the source system side, then you can take help from the DBAs and see if you can create indexes on the table to speed up the extraction. Also check to see if there are any enhancements done to the datasource that might be causing the delays.

Cheers,

Kedar

[Former Member]

SAP delivered OSS note to improve performance

[Former Member]

Hi

try dropping indices of your data target before loading, and build them after load completes.

Also, some times, due to some problems(indixes) with master data tables in BI, those are used in transformation rules, consume much time for sequential read.

Check out SM66 while loading for process that is at Sequential read. goto DB02 for missing indexes and build them if necessary!

Hope this is helpful

Regards

GSK

• Assign pts if useful**

[Former Member]

I'd run a trace in R/3 to find missing indexes.