Welcome to Essbase Partitioning 101. In this post, we will introduce the basic Essbase Partition functions, and how each partitioning scheme can help ease data pain in enterprise applications.
Essbase Partitioning Functionality
Essbase partitioning is used to simplify user access to data across multiple Essbase cubes. The goal of partitioning is to simplify the data management process, and avoid many of the data synchronization and network bandwidth headaches often associated with replicating data between applications. Essbase partitions enable administrators to select specific subsets of data and metadata (Essbase outline) to synchronize between Essbase cubes, using replicated, transparent, or linked partitioning schemes. Each partition type has advantages specific to particular needs and application requirements. We’ve outlined the general functionality and advantages of each partitioning type below.
Types of Partitions
A replicated Essbase partition creates a copy of the selected portion of the source cube in the target cube. The primary considerations with replicated partitions are synchronization timing and data submission requirements. Because data is being copied to the target database, the timing of source to target synchronization may have an impact on end users. Data replication can be automated, but data in the target application is not “up to the minute” and any changes to the source will not reflect in the target until after replication has completed. The primary limitation to replicated partition is the inability to submit data. If data submission is a requirement, transparent or linked partitions will be more suitable.
The primary advantages of a replicated partition are reduced network traffic to source database, improved calculation and query speeds, and ease of recovery. Replicating data to a target database closer to the users has the impact of reducing network access to the source database. Databases can be more focused by only coping relevant data where needed. These smaller databases improve user query and calculation times. When comparing to other data transfer approaches, replicated partitions perform data transfer operations more quickly than XREF, and offer improved simplicity over database export/import scripting. Finally, because the data exists in multiple locations, the data remains accessible in the case one of the databases fails.
Transparent Essbase partitions operate by retrieving data from the source database in real time. Primary considerations with transparent partitions are database design, as well as available network bandwidth in each database location. Transparent partitions can have significantly negative performance implications when partitioning dense or base dimensions, limiting retrievals to summary-level, full blocks specified by sparse dimensions will provide the best performance. The careful design, testing, and refinement of transparent partitions is critical to their success in production environments.
Transparent partitions partitions are the most invisible to end users, and show users the latest up-to-the-minute data. Because data is not being copied, only one master data source exists, requiring less overall disk space. Data can be loaded and accessed from both source and target databases, though performance will be best at the source database. Transparent partitions will often be the lowest maintenance solution, at the cost of network bandwidth.
Linked partitions connect two databases with a linked data cell. Clicking a linked cell causes drill across to the source database in a new worksheet. The primary consideration with a linked database is user access to each database. A linked partition effectively makes hyperlinks between two separate databases, so any user accessing data in the partitioned portion of the data must have access to both databases.
The most effective uses of linked partitions include use cases where multiple separate data sources need consolidated into a central data target. This enables spreadsheet users to access unified data at a high level, while individual data sources can be uniquely customized as needed.
Next Steps | Essbase Partitioning Step-by-step
We hope you’ve found this Essbase Partitioning 101 post useful. Next time we will walk step-by-step through creating an Essbase Partition, and overview some of the common obstacles and errors.
For more information, Oracle’s technical documentation can be found here: https://docs.oracle.com/cd/E12825_01/epm.111/esb_dbag/frameset.htm?dotprtde.htm
Need Essbase support? We can help! 1-800-EPM-PROS