Partitions enable you to access data that is shared between databases that may span applications or servers. When you create a partition, you specify which cells are part of the partition and how the source and target cells map to each other. You also specify the type of partition to use and the connection information for the source and target databases. When you create and save a partition, the partition definition is saved to two different .ddb files. One .ddb file is stored in the source database directory (ARBORPATH\app\appname\sourcedbname) and the other .ddb file is stored in the target database directory (ARBORPATH\app\appname\targetdbname).
How you access the partitioned data depends on the type of partition you use.
Replicated partitions enable you to copy data inside the partitioned area from one database to another.
Transparent partitions enable you to view data that is stored in two different databases as if it is all stored in one database. For example, when you drill down in a spreadsheet to view data for different member combinations, some of the data is stored locally, but the data inside of the partitioned area is stored remotely.
Linked partitions enable you to specify one or more member combinations in one database to use as a link to the same combinations in another database. From a spreadsheet, you can drill down on one of the partitioned member combinations to launch the Linked Objects Browser. From there, you can open another spreadsheet; that spreadsheet is connected to the other database and displays the data cells for the corresponding member combination.
Two sample partitioning applications are provided with the Essbase Server software: Samppart and Sampeast. If these sample applications have been installed on Essbase Server, they appear in Enterprise View in Administration Services Console.
“Designing Partitioned Applications” in the Oracle Essbase Database Administrator's Guide
“Creating and Maintaining Partitions” in the Oracle Essbase Database Administrator's Guide