Refreshing Cubes

The operation of the REFRESH CUBE statement varies depending on the data source to which PivotTable® Service is connected.

• If PivotTable Service is connected to an OLAP server running Microsoft® SQL Server™ OLAP Services, the REFRESH CUBE statement causes the local memory cache to be synchronized with the server.
• If PivotTable Service is connected to a relational database, executing the REFRESH CUBE statement causes the local memory and disk cache to be cleared and the dimensions repopulated.
• If PivotTable Service is connected to a local cube, executing the REFRESH CUBE statement rebuilds the local cube file with CREATE CUBE and INSERT INTO statements that reference the original source database. These statements and a reference to the original source database are stored in the local cube file.

For more information, see REFRESH CUBE.

Refreshing a cube can also be initiated by the server when PivotTable Service is connected to an OLAP server running OLAP Services. In such a case, OLAP Services updates the data and notifies PivotTable Service on the client to refresh the client’s cache. This notification is done automatically during the next available communication initiated by the client. At that time, all of the client caches are erased, and open queries (OLE DB for OLAP datasets) are refreshed. The LAST_UPDATED column of the CUBES schema rowset changes the time stamp.

It is the client application’s responsibility to track refresh notifications. The refreshing of a cube can result in situations where queries are no longer valid, but this situation can be avoided by setting the DBPROP_MSMD_DEFAULT_ISOLATION_MODE property to T (true). It is the client application’s responsibility to revalidate queries after refresh.