September 30, 2015

Teradata Readies Links to Push Presto to Enterprises

George Leopold

Teradata Corp. said this week it would accelerate the open-source rollout of Presto, the SQL-on-Hadoop framework, by offering connectivity drivers for free.

Teradata (TDC) said Monday (Sept. 28) it would provide at no cost ODBC (Open Database Connectivity) and JDBC (Java Database Connectivity) drivers with the goal of making the Presto SQL query engine ready for primetime in the enterprise. The drivers provide connection and implementation protocols for transferring queries and results between applications and databases.

Teradata announced in June it would make a major investment in the Presto framework originally developed at Facebook to power interactive queries against its massive data warehouse. The San Diego-based company said its accelerated development of the ODBC/JDBC drivers based on feedback from the open source Presto community.

Along with stepping up to provide technical services and support for the open source SQL engine, Teradata also pledged in June to commit engineering resources to improve Presto over time.

Presto is a key analytics component in the Netflix (NFLX) big data platform. “One big challenge has been the absence of enterprise grade ODBC and JDBC drivers,” Kurt Brown, director of the Netflix data platform, noted in a testimonial.

Teradata said the Presto drivers were originally slated for delivery in mid-2016. Based on feedback from open source users, the global launch for the ODBC driver has been moved up to the fourth quarter of this year. Meanwhile, the JDBC driver will be released in the first quarter of 2016.

Teradata said the new Presto drivers would allow the SQL-on-Hadoop framework to leverage existing analytic environments to scale interactive queries. It would also help leverage existing SQL business intelligence and visualization tools.

The familiar ODBC and JDBC interfaces also would help speed development of business intelligence applications in data lakes. “The goal is to increase adoption beyond the world-class Silicon Valley companies that are already using the platform,” Teradata said in a statement.

Adding new drivers to the Presto open source platform is also expected to help enterprises speed application development based on the in-memory SQL framework. The links are also intended to help business intelligence vendors certify their tools with Presto for use with interactive workloads in data lakes, the company said.

Presto is a Java-based SQL engine developed by Facebook and released as open source software in November 2013. Facebook was an early Hadoop adopter.

Teradata said in June it would commit 15 engineers to work on Presto and to help shape the future of the open source project. Those engineers largely come from Hadapt, the SQL-on-Hadoop software company it acquired last year.

Presto is seen as filling a specific niche, primarily running interactive queries against large datasets with low latency and many users. It also can push queries beyond Hadoop to other data repositories, including NoSQL, relational databases and proprietary data stores. Hence, Teradata is positioning Presto as an alternative when quick answers are needed to database queries.

Recent items:

Terradata Bets Big on Presto For Hadoop SQL

Facebook’s Super Hive-Killing Query Machine New Yours

Share This