Aggregate have High Record Summarised (>10) But NOT Used

[Former Member]

Hi

I Have found the Query/Infocube which is taking high DB time(>30 %) and High Select/transfer Ration (>10) . From Table RSDDSTAT I found out the STATUID of the High Response time Query and put the STATID in the table RSDDSTATAGGRDEF to get the IOBJNM .

Then Created the Aggregates Based on these IOBJNM , Query Definitions and General Rules for creating aggregates.

The Aggrgate Created have high Record Summarised(mean Value) >10 but these Aggregates are NOT used by the Query .

Please tell me how to Debug WHY the Query is NOT using the Aggregates.

I found out the Quesry is NOT using the Aggregates when I run the Query in DEBUG mode in Transaction RSRT.

I want to increase the Performance of the System. Please help me to find out where I am going wrong ? Why the Query is Not selecting the Aggregate.

Kind regards,

Yogesh

[Former Member]

Are all the characteristics that are used for filtering/restricting in the query also in the aggregate? They need to be in the aggregate as well as any characteristics in the rows.

RSRT in debug mode - check the <b>Display Aggregate Found</b> and <b>Select Aggregate</b> debug options, and that will tell you why an aggregate could not be used.

[Former Member]

Hi.

You can also use transaction RSRTRACE. Add your user to the list and configure the user for your need.

Then execute the query in a new session and check the log in RSRTRACE. Here you will find an entry for every session. Double click on the session you want to analyze. Then check on the button 'DB Debugging'. Check 'Select Aggregate' and then continue. You can now see why the aggregate is not used in the column 'Not Possible Because...'.

/Christian Frier

[Former Member]

Hi,

rsrt is the right place to go. It will also provide you the information about the aggregate the query will need.

Siggi