Hi,
I'm trying to understand the way SSAS Tabular pulls data and computes the query results. I'm probably asking very basic questions. My questions are the following:
1. If I use the 'Cached'/'In-memory' mode, during processing of data, will all the base data be stored locally by the SSAS tabular engine as well as loaded in memory? The size shown by the cube is much smaller (about 1/100th of the underlying DB) - which leads me to assume that after processing and before any queries are fired, it really does not stock all the data locally. Or is my understanding wrong?
2. Does it make any DB calls (assuming 'Cached'/'In-memory' mode) at the time queries are fired on the db?
3. I noted that when queries were fired, the memory utilization was shooting up - is this because although it had all the base data in the memory, it needed to compute the joins / aggregates etc while responding to the queries and since all that happened in memory, the memory utilization was high?