Deciding between ROLAP mode (relational online analytical processing) and MOLAP mode (multidimensional online analytical processing) when using OLAP with Microsoft‘s SQL Server Analysis Services (SSAS) has always been a decision of choosing between high performance and data validity.
The problem with MOLAP
In standard MOLAP mode, you precompute your multidimensional data cubes first which allows you to query the pre-aggregated data with high performance. But this approach has clear disadvantages:
- A lengthy pre-computation is often needed
- You operate on potentially outdated information as your queries are being executed on the precomputed data cube – not on the live data
That means that you have to sacrifice the validity of data in favour of performance.
The problem with ROLAP
If you use ROLAP mode on a standard relational database, you can overcome these disadvantages as you access the relational data directly. However, now you have to sacrifice fast query performance levels as you are no longer using pre-computation, like you do in MOLAP mode.
There is another way: EXASOL.
With EXASOL and the OLAP Connector for SSAS you can accelerate you OLAP-based BI solution without sacrificing data validity. In other words, your data is always up-to-date.
The combination of SSAS in ROLAP mode, combined with EXASOL’s leading in-memory analytic technology, delivers groundbreaking performance without the need for pre-computations.
Are you interested in learning more about EXASOL’s MS SQL Server Analysis Services Accelerator?