Online analytical processing (OLAP) Interview Preparation Guide

OLAP is known as online analytical processing which provides answers to queries which are multi dimensional in nature. It composes relational reporting and data mining for providing solutions to business intelligence. This term OLAP is created from the term OLTP.

Hyper cube or multidimensional cube forms the core of OLAP system. This consists of measures which are arranged according to dimensions. Hyper cube Meta data is created by star or snow flake schema of tables in RDBMS. Dimensions are extracted from dimension table and measures from the fact table.

Classic form of OLAP is known as MOLAP and it is often called as OLAP. Simple database structures such as time period, product, location, etc are used. Functioning of each and every dimension or data structure is defined by one or more hierarchies.

Functioning of ROLAP occurs simultaneously with relational databases. Data and tables are stored as relational tables. To hold new information or data new tables are created. Functioning of ROLAP depends upon specialized schema design.

OLAP can process complex queries and give the output in less than 0.1 seconds, for it to achieve such a performance OLAP uses aggregations. Aggregations are built by aggregating and changing the data along the dimensions. Possible combination of aggregations can be determined by the combination possibilities of dimension granularities.

Often calculating all the data is not possible by aggregations for this reason some of the complex data problems are solved. In order to determine which data should be solved and calculated, developers use View selection application. This solution is often used to reduce calculation problem.

Bitmaps are very useful in start schema to join large databases to small databases. Answer queries and bit arrays are used to perform logical operations on the databases. Bit map indexes are very efficient in handling Gender differentiation; also repetitive tasks are performed with much larger efficiency.

Bitmaps commonly use one bitmap for every single distinct value. Number of bitmaps used can be reduced by opting for a different type of encoding. Space can be optimized but when a query is generated bitmaps have to be accessed.

Binning process is very useful to save space. Performance may vary depending upon the query generated sometimes solution to a query can come within few seconds and sometimes it may take longer time. Binning process holds multiple values in the same bin.

The process which is underlined during the check of base data is known as candidate check. When performing candidate check performance varies either towards the positive side or to the negative side. Performance of candidate check depends upon the user query and also they examine the base data.