Dharmesh Talaviya

Gaining An Insight Into The BI Data Structure Of OLAP

OLAP or otherwise known as Online Analytic Processing is a data structure under Business Intelligence software that enables one to quickly analysis the data. It can also be characterized as the means of studying and manipulating data from different perspectives.

This BI tool provides the building blocks to facilitate analysis (e.g. rich functions, multi-dimensional models, analysis types). These business intelligence tools are inclined towards slicing and dicing of the data.  OLAP enables the possibility of business intelligence solutions generally through the transformation of data into multi-dimensional cubes, summarization of pre-aggregated and derived data, management of strong query and usage of multitude of calculation and modeling functions.

In Business Intelligence, OLAP stands between the Data Warehouse and end-user BI tools.  Data warehouse software is a database utilized mainly for BI reporting. Data is uploaded from the operational systems and is later on stored in the warehouse. Before the data can be used for reporting, it passes first through an operational data for additional operations. To be able to provide overall data-access for the end-user tools, OLAP and Data warehouse must work in collaboration. Cubes within a BI data warehouse are stored in different forms known as relational storage mode (ROLAP) and a multidimensional storage mode (MOLAP).

MOLAP

MOLAP stands for Multidimensional Online Analytical Processing. It is designed to enable the analysis of data through using a multidimensional data model. In this type of OLAP, a cube is collected from the relational data source. This particular BI tool can generate a report quickly when a user generates a report request because the data is pre-aggregated within the cube.

ROLAP

ROLAP stands for Relational Online Analytical Processing. This BI tool is also designed to enable the analysis of data through using a multidimensional data model. However, it differs significantly from MOLAP since it does not necessitate the pre-computation and storage of information into a cube. The ROLAP engine fundamentally operates as a smart SQL generator. It ideally includes a 'Designer' piece, where the data warehouse administrator is responsible for indicating the relationship between different relational tables as well as figuring out the attributes, dimensions and specific hierarchy maps in regard to the underlying BI database tables.

HOLAP

HOLAP or Hybrid Online Analytical Processing is ideally a blend of MOLAP and ROLAP and combines the strengths of these two BI tools. HOLAP stores a certain part of the data in MOLAP and another part of the data in ROLAP, thereby offering the advantage of both. The extent of the control that the cube designer would have over this partitioning can differ from one particular BI product to another.

Comparison in the usage of MOLAP and ROLAP

When it comes to comparing the different methods of these BI tools what matters is the kind of effect this type of storage can have on the processing, storage and browsing of the cube. For instance when the three Business intelligence data structure methods are compared, cube browsing is the best in MOLAP. Since the data is piled up in a format, which is compressed and multidimensional therefore it can be accessed at a faster rate than in relational database. On the other hand, browsing is not a strong point in ROLAP because the time it takes to process is much slower especially where levels of high aggregation are concerned.

In terms of storage, the BI tool of MOLAP requires more space that ROLAP because the duplication of the data takes place at low levels of aggregation. Meanwhile, ROLAP hardly requires any space for storage as data doesn't need to be copied. However, ROLAP aggregations need more space for storage than HOLAP or MOLAP aggregations.

In terms of viewing the data through these different BI solutions, all the data that is stored in MOLAP can be observed even when the original data source is not available, whereas, in ROLAP data cannot be analyzed unless it is attached to the data source.

Article Tags: Online Analytical Processing, Data Structure, Business Intelligence, Data Warehouse, Online Analytical, Analytical Processing, Data Through, Data Source

Source: Free Articles from ArticlesFactory.com

ABOUT THE AUTHOR