IN-DATABASE EXECUTION

Traditionally, analytics have been developed and used within independent departments across an organization using a variety of specialized, and isolated, tools from various vendors. Different departments typically manage their own data marts and tools. Use of analytic tools is often restricted to highly-skilled analytically-oriented business users, who have to collaborate closely with database administrators and application programmers. Development of analytical applications becomes time-consuming, costly and the analytics are often inflexible and not easily re-used.

Relational database vendors provide a wide range of best-of-breed analytical functions via SQL and SQL extensions that operate directly within the database, providing security, scalability and integrity. Although, in theory, these analytics can be accessed via graphical user-interfaces, their use beyond performing simple analysis over isolated data sets is not straightforward and requires the use of advanced SQL or programming APIs.  Database users are thus faced with a choice:

• Building their own customized applications using database analytics, which require access to database and application programming expertise to develop end-user applications,

Or

• Resorting to third party analytical tools that offer a simplified environment for developing and using analytics, but typically require data to be pulled out of the database for analysis. Thus compromising efficiency, scalability, security and data integrity.

In-database analytics brings analysis to the data rather than the other way around.  Eliminating data movement enhances performance with efficient execution of the compute-intensive analytics over large data sets while effectively addressing data security and integrity issues.  Executing analytics directly in the database also reduces the administrative costs of learning, implementing and maintaining an environment separate from the database.