We want to optimize the Material Stock infocube (0IC_C03) in order to have better query runtime, it uses non-cumulative key figures. Before the compression of the request of the infocube, the query runtime is around 10 min. Doing the compression of all the requests with marker update the query runtime is around 50 min.
The model we are using has a DSO before the infocube because we want to save the history at granular level.
The DSO has just one request of 20 million records updated as delta. The history goes from NOV 2008 to FEB 2010.
The infocube has several requests that are divided by periods (selection of different posting date) all the requests are updated as full:
The model doesnu2019t use 2LIS_03_BX, it uses just 2LIS_03_BF and 2LIS_03_UM, therefore we are compressing the infocube requests with marker update
The expected result would be that the query runtime be less than 10 min.