Business Intelligence: OLTP vs OLAP

The acronym OLAP, On Line Analytical Processing, refers to a class of products that allow for ‘navigation’ through data (data surfing), providing easy-to-use tools that enable the user to independently formulate queries. The semantic layer provided frees the user from the knowledge of the databases, both in terms of location and organization. Typical of this class of access products is the multidimensional view of data, which allows for analysis along different paths (dimensions) (time, geographical location, etc.). To download the full text and for more information, click HERE

The range of OLAP tools is very broad. It is appropriate to clarify the concepts of ROLAP and MOLAP. The term ROLAP (Relational OLAP) identifies tools that operate on data stored in relational databases. In many cases, data is stored in denormalized structures such as star schemas and in aggregated form. ROLAP tools can be divided into two categories, depending on whether the reference architecture is two-tier or three-tier. MOLAP tools (Multidimensional OLAP) are based on a proprietary database, fed from staging areas or the Enterprise Data Warehouse. To date, there is no standard for accessing multidimensional data as there is for relational databases (SQL). At the time of loading, which must take place periodically, data is inserted into multidimensional structures, offering excellent performance guarantees. On the downside, analysis dimensions not foreseen at the time of loading cannot be performed without a subsequent data feeding operation. MOLAP tools are built according to two different architectural principles, which we will call Client and Server MOLAP. The two types differ in that in Client MOLAP the multidimensional database resides on the client machine, while in Server MOLAP the multidimensional database is hosted on a server. Strengths of ROLAP tools include the standard interface (SQL), access to any database, analytical flexibility, and scalability; conversely, MOLAP products have traditionally been strong in performance. In recent years, specific decision support functionalities have been incorporated into relational databases, which has significantly reduced the performance gap between the two technologies. Many OLAP vendors are adapting their offerings to provide solutions where MOLAP and ROLAP products cooperate (HOLAP, Hybrid OLAP solutions). MOLAP solutions are proprietary, less flexible, and require more setup and maintenance effort compared to ROLAP solutions. Their use is not to be excluded, although it certainly must be carefully evaluated. OLAP tools are the most widespread in the heterogeneous landscape of data access tools, as they provide advanced data analysis functionalities, as well as excellent graphical and reporting features. These characteristics make such products of interest to a very broad class of users. Normally, the analyst user sets up the report, then allows its use to other users who, depending on the case, simply view it or perhaps request an update by customizing certain parameters of interest. The use of these tools generally makes users self-sufficient in their ability to locate, understand, and access data. ALL RIGHTS RESERVED

Pubblicato in Business