Improving Cube Performance

The application enables you to schedule jobs that improve cube performance. These jobs include:

  • Restructure Cube—Performs a full restructure of a block storage cube to eliminate or reduce fragmentation. This will also remove empty blocks. Running this job won't push any changes from Planning to Essbase.

  • Compact Outline—Compacts the outline file of an aggregate storage cube. Compaction helps keep the outline file at an optimal size. Compacting the outline doesn't clear the data. Running this job won't push any changes from Planning to Essbase.

  • Merge Data Slices—Merges incremental data slices of an aggregate storage cube. Fewer slices improve a cube’s performance. You can merge all incremental data slices into the main database slice or merge all incremental data slices into a single data slice without changing the main database slice. You can optionally remove cells that have a value of zero.

  • Optimize Aggregation—Generates optimized views based on collected query tracking information in an aggregate storage cube. This job has two actions:

    • Enable Query Tracking—Records a meaningful set of queries, and then uses the recorded query data to select the most appropriate set of aggregate views to materialize for that database. You must enable query tracking and allow it sufficient time to collect data before executing the aggregation process based on query data.

      Once enabled, query tracking continues until query tracking is disabled, the cube is reloaded or refreshed, the outline is compacted, the application is shut down, or additional aggregate views are materialized for the database. Query tracking does not resume automatically after these actions occur.

      Note:

      The following actions clear the stored query tracking information from the aggregate storage cube, but they do not disable query tracking:

      • Merging incremental data slices

      • Clearing data

      • Clearing existing aggregations

      • Clearing partial data

      • Loading data into the aggregate storage cube

    • Execute Aggregation Process—Calculates aggregations for aggregate storage databases that contain data and to which you are granted Calculation permission. To perform an aggregation, you use system recommended views. The selection of views and aggregation processes are combined into one, non-configurable operation performed by the server. You can optionally specify the maximum disk space for the resulting files, base the view selection on user querying patterns, and include rollup hierarchies in the view selection. Select one or more of the following options:

      • Based on query data?—Aggregate the views the server selects, based on collected user querying patterns. This option is only available if query tracking is turned on.

      • Include rollup option?—Include secondary hierarchies (with default level usage) in the view selection process.

      • Include growth size option?—Aggregate the views the server selects, until the maximum growth of the aggregated database exceeds limits you specify. Enter the size (in ratio) beyond which the server should stop the aggregation.

You can start the jobs right away or schedule them to run later.

Note:

Users can stay logged in during these actions.

To create a job that improves cube performance, see Scheduling Jobs.