Differences between Aggregate Storage, Block Storage, and Federated Cubes
If you already have an Essbase ASO or BSO application, you can make it into a federated partition application.
Use this page in planning stages, if you need to compare characteristics of Essbase block storage (BSO) and aggregate storage (ASO) cubes, side by side with federated partition cubes.
| Aggregate Storage (ASO) | Block Storage (BSO) | Federated Partition Cube | |
|---|---|---|---|
| Data storage model |
Data is stored in Essbase. |
Data is stored in Essbase. |
Data is stored in a relational table in Autonomous Data Warehouse. Elsewhere in the documentation, it is referred to as the fact table. |
| How it works |
Number of dimensions can be very high, containing millions of members, but the cube has relatively sparse data slices (many dimensional intersections contain no data). Data is input at level 0 only. Cubes are optimized for rapid aggregation. |
The number and scale of dimensions are typically smaller as compared with ASO. BSO accomodates dense data sets. Some of the dimensions are defined as dense, with data at most intersections, and others are defined as sparse. This helps Essbase store data efficiently and optimize dependency analysis (so as not to overcalculate). Data can be input at any level. |
The Essbase outline is mapped to the fact table, allowing data storage to remain in Autonomous Data Warehouse, while being accessible for analysis using the logic you build into your Essbase application. The analytical capabilities of your Essbase outline enable you to analyze the flat relational table as hierarchies, employing whatever complex procedural math you may need for your multidimensional analysis. Calculations and aggregations, when possible, are converted by Essbase into SQL and pushed to Autonomous Data Warehouse, so that processing occurs closer to where data is stored. You can find the SQL Essbase writes in the platform log, located in |
| Typical use cases |
ASO cubes are commonly used for highly aggregational analytics, custom calculations, and allocations. Data loads can be broken into slices for frequent, highly parallelized updates. |
BSO cubes are commonly used for financial and operational planning, and interactive reporting on aggregate data relative to the source. BSO cubes are designed for complex analytical requirements requiring formulas/ math, and frequent procedural calculations. |
Data does not leave Autonomous Data Warehouse, eliminating the need for refreshing and restructuring in Essbase. Since you create the federated partition over an existing ASO or BSO cube, you can use either of those Essbase options and benefit from its style of calculations and queries, without ever having to load the data into Essbase or restructure the outline. If your organization already has a fact table stored in Autonomous Data Warehouse, a federated partition enables you to use Essbase functionality such as:
If your organization already uses Essbase, a federated partition enables you to access these benefits of storing data in Autonomous Data Warehouse:
|