OLAP memory management
The OLAP database requires memory in order to store data, run
calculations, and answer requests. Depending on the kind of requests and calculations, the
amount of memory required to perform such an operation varies greatly.
OLAP checks the available memory on a regular basis. This is done to avoid a memory allocation that overwhelms the system. If a request requires more memory than configured for OLAP, the request is canceled with an out-of-memory exception.
These memory checks are in the MDX Engine and the OLAP Kernel.
These are just a few reasons for large memory consumption:
- Very large cross-joins
- Sorting very large sets
- Returning a very large amount of data
- Value-generating rules.
- Operations that create a large number of cell values.