Anatomy of a High-Performance Hive ODBC 3.52 Driver

Contributed by

2 min read

ODBC has been the flagship API for SQL ever since it was first developed by Microsoft and Simba Technologies in 1992. An acronym for Open DataBase Connectivity, ODBC is the standard API used by popular applications like Excel, Crystal Reports, MicroStrategy and Tableau to connect to SQL databases.

ODBC is also the API used to directly connect the same popular analytics and reporting applications to NoSQL Big Data sources, such as Hadoop/Hive. But, there’s a catch. Not all ODBC drivers for Hadoop/Hive are equally capable. Some offer universal, direct connectivity while others offer barely any connectivity at all. The difference between capable versus non-capable depends on the version of ODBC supported and the degree of “special sauce” built into the driver to map SQL queries to HiveQL.

While it’s proven that MapR’s Hive ODBC 3.52 driver is certainly capable, what may not be known is that it’s actually the most advanced ODBC driver for Hadoop/Hive available. Developed with Simba Technologies—the ODBC pioneer—it provides universal ODBC access to Hadoop-based data.

You may ask, what makes it so special?

  • It’s compliant with the latest ODBC 3.52 specification. Other drivers only support ODBC 2.X with limited functionality.
  • It supports both 32- and 64-bit platforms. Other drivers only support 32-bit platforms.
  • It translates SQL-92 queries to HiveQL on the fly, opening up full SQL-based analysis and reporting on data. No other drivers support SQL-92.
  • It connects popular BI tools such as Excel, Crystal Reports, MicroStrategy and Tableau directly to the MapR Hadoop distribution – data stays in MapR, no extraction needed. Many other drivers hardcode connectivity to a singular application and only support certain functionality.

SQL-based BI has come to Hadoop; you need to make sure your ODBC driver can provide open access for analysis of your data.

This blog post was published August 08, 2012.