Specifying the InfoProvider Grouping for Delta Caching

You can optimize the delta cache process by using the InfoProvider Grouping setting. If a query is based on a MultiProvider, you can specify that the data that each of the InfoProviders for the MultiProvider supplies is to be treated separately and stored separately in the cache.

The advantage of this grouping is that when data for just one InfoProvider in the MultiProvider changes, you do not have to read all data for all InfoProviders again.

The disadvantage of this grouping, however, is that it requires more memory space because the data cannot be aggregated using the various InfoProviders. The data is not aggregated until the query is run.

The grouping is only taken into account if the delta cache process is activated for the query.

If you use the delta process and the query is based on a MultiProvider, choose the required setting for the option InfoProvider Grouping. The following modes are supported:

• No Provider Partitioning (0)

  There is no partitioning. All results for the InfoProvider are merged into one result and cached. If the data for an InfoProvider changes, the cache has to be rebuilt.

  This is the default setting.

• Grouping depending on InfoProvider Types (1)

  With grouping based on the InfoProvider types, all InfoCubes are saved in a group. The remaining InfoProviders are either stored in one group or stored in multiple groups according to their properties. For example, VirtualProviders, DataStore objects, or InfoProviders that can supply hierarchy nodes are grouped. Open requests for a real-time InfoCube are also stored separately from standard InfoCubes.

• Partitioning in Groups, InfoCubes Separate (2)

  The difference between this option and Partitioning in Groupsis that with this option, the InfoCubes are also stored separately.

• Every Provider Separate (3)

  With this setting, all InfoProviders are stored separately.