Many times user complain of queries taking a lot of time to execute and we are clueless how to proceed on this. RSRT transaction can provide us a lot of help in exactly analyzing where the issue is and then we can work on the improvements.
Query runtime mainly consist of two components.
How to find out where exactly the query has taken time.
Goto RSRT and enter the query name. Click on Execute + Debug
After that we would get the following screen
Click on Display statistics data as we would like to see where exactly query is taking time.
Please also select Do not use cache to get the exact time otherwse , data would be fetched from the cache and you would net get the correct time. In case you are using Accelerator please also select the option DO not use BIA index option. Click on the tick option at the bottom.
You would get the selection screen. Enter the input parameter and run the query.
After you get the result click on the back button in green at the top. It would take you to the statistics of query
We would see in the frontend/Calculation Layer time taken by the query at various stages.
We can identify the time taken at various layers with the help of event ID.
Data Manager time: We can identify the time taken by the query to read the data from database layer by event ID’s starting from 9000 like (9000,9001,9002,9010,9011,etc) . We can sum up the time for all the event ID's in the 9000 to 9100 range to get the database time.
OLAP time: This is the frontend time taken by the query to calculate CKF, RFK, formula variable, text variable, exception aggregation,etc.We can identify the OPLAP time with event ID’s like 3000, 3010, 3100, 3110, etc.They would be in the 3000 series. We can sum up the time for all the event ID's in the 3000 range to get the OLAP time.
There would be some other smaller component also but the majority of time would be taken either at the database level or OLAP level.
Once you have identified where the problem is, we could further propose solution to improve the query time.
If query is taking a lot of time at database level, possible steps can be.
If query is taking a lot of time at OLAP layer then possible steps can be.
Event ID and there relation can be find out from this link.
http://help.sap.com/saphelp_erp60_sp/helpdata/en/45/f0488a1aa03115e10000000a1553f7/content.htm
Awaiting your inputs for further improving the document
Thanks,
Shyam
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
37 | |
10 | |
7 | |
5 | |
5 | |
5 | |
4 | |
4 | |
3 | |
3 |