Partition Storage Models and Processing

The storage mode of a partition affects the query and process of performance. The storage requirements and storage locations of the partition measure group and cube
A partition can use one of three basic storage modes

Multidimensional OLAP (MOLAP)
Relational OLAP (ROLAP)
Hybrid OLAP (HOLAP)

Microsoft SQL Server Analysis Services supports all three basic storage modes. It also supports proactive caching. The characteristics of ROLAP and MOLAP storage of data and query performance

MOLAP

The MOLAP storage mode causes the aggregations of the partition and a copy. It is source data to store in a multidimensional structure in Analysis Services. This MOLAP structure is optimized to maximize query performance. The storage location can be store computer. Where the partition defined or on another computer running Analysis Services. Because a copy of the source data resides in the multidimensional structure. Queries can resolve without accessing the partition's source data. Query response times can decrease by using aggregations. The data in the partition's MOLAP structure is the current processing of the partition

The Process updates the data in the MOLAP structure, either full. The time between one processing and the next to create OLAP objects not match the source data. The update objects in MOLAP storage without taking the partition or cube offline. But, take a cube offline to process certain structural changes to OLAP objects

ROLAP

The ROLAP storage mode aggregations of the partition to store in index views. The relational database that specified in the partition's data source. Unlike the MOLAP storage mode, ROLAP does not cause a copy of the source data. To stored in the Analysis Services data folders. Instead, when results cannot derive from the query cache. The indexed views in the data source access to answer queries. The query response is slower with ROLAP storage than MOLAP or HOLAP storage modes. Processing time is also slower with ROLAP. But, ROLAP enables users to view data in real time and can save storage space. When you are working with large datasets. Those are queried, such as historical data

The partition cannot contain measures that use the Min or Max adds functions.
Each table in the schema of the ROLAP partition used in only one time.
Each table must be a table, not a view.
All table names in the partition's schema must qualify with the owner name.
All tables in the partition's schema must have the same owner.
The source columns of the partition's measures must not be nullable.
All tables used in the view must have created with the following options set to ON:
- ANSI_NULLS
- QUOTED_IDENTIFIER

HOLAP

The HOLAP storage mode combines attributes of both MOLAP and ROLAP. The partition to store multidimensional structure in an SQL Server Analysis Services instance. HOLAP does not cause a copy of the source data to store. That access only summary data in the aggregations of a partition, HOLAP is the equal of MOLAP. Queries that access source data. If you want to drill down to an atomic cube cell for which there is no aggregation data.
Partitions stored as HOLAP are smaller than the equal MOLAP partitions. They do not contain source data and respond faster. Than ROLAP partitions for queries involving summary data. HOLAP storage mode is generally suited for partitions in cubes. That needs rapid query response for summaries based on a large amount of source data. Where users generate queries that must touch leaf level data. Such as for calculating median values, MOLAP is generally a better choice

MSBI Tutorials Point