OLAP : On-line Analytical Processing
What is OLAP?
We know that there are two kinds of database applications available, OLAP and OLTP. Before we go for OLAP, I would like to tell you something about OLTP.
OLTP is On-line Transaction Processing. By transaction, we mean any operation including INSERT, UPDATE, and DELETE statements. And, just like any other transaction system, the main concern here is ACID rules.
Let’s take an example of banking transaction systems, where a large number of users simultaneously perform transactions. In this case, Online Transaction processing (OLTP) applications provide optimal environment for concurrency and atomicity of transactions.
OLAP is On-line Analytical Processing and its main concern is analysis. This is how OLAP is different from OLTP, where the main concern is concurrency and atomicity.
Now, With OLTP we always have a consistent state of data but how this data can be useful to business?
OLAP has an answer to it. In OLAP, the current data is analyzed for any trends or summarize through many filters. The resultant data is not changed or tampered with anytime.
“Because of OLAP, data speaks for itself.”
For example: With the help of OLAP component; Sales, Marketing, Budgeting and Forecasting can provide multi-dimensional analysis and reporting.
What are the advantages of OLAP?
Multi-dimensional: Multi-dimensional views of various kinds of business activities.
Decision Support: Analytical capabilities promote decisions.
Complex Calculations Support: When it comes to determining future trends then complex calculations are necessary like averages and percentage growth.
Access from individual’s point of view: Access to fast, consistent, interactive access to a wide variety of possible views of information.
What are the types of OLAP?
There are three basic types of OLAP :
MOLAP relies on a multidimensional database.
ROLAP relies on a relational database.
HOLAP is a hybrid model.
More Details are given below:
MOLAP: Multidimensional Online Analytical Processing uses a multidimensional database stored in various cubes. This information is pre-summarized and optimized, due to which, fast data retrieval is possible. Also, a range of complex calculations are already performed.
ROLAP: Relational Online Analytical Processing uses a relational database. ROLAP do not use pre-calculated data cubes like MOLAP. ROLAP uses regular SQL statement to query the standard relational database. Although, it gives the appearance of traditional OLAP’s slicing and dicing functionality.
HOLAP: Hybrid Online Analytical Processing is a hybrid approach which combines the advantages of MOLAP and ROLAP. Here, database divides data between relational and specialized storage like the small aggregated totals are stored in a multidimensional database while large quantities of detail data is stored in the relational database. So we have advantage of both pre-summarized/optimized cubes and relational database.