This week was the Spark Summit in San Francisco. It was a great event with a lot of interesting announcements. Databricks is the company promoting Spark and Shark and they made some interesting announcements. One interesting piece of news is that they are ending development of Shark and instead focusing their efforts on Spark SQL. The claim is that the Hive codebase is legacy and they need to break free to improve SQL performance. This seems similar to what Cloudera did with Impala and MapR did with Drill. It will be interesting to see how this progresses. You can read more about this in a post from Reynold Xin entitled “Shark, Spark SQL, Hive on Spark, and the future of SQL on Spark“.
Another interesting read is an interview by Doug Henschen at InformationWeek. Doug interviews Databricks’ Ion Stoica and Arsalan Tavakoli. They talk about their plans to become the epicenter of Big Data analytics and how they plan to make deploying Hadoop and Analytics software so much easier. You can read about this at “Databricks Spark Plans: Big Data Q&A“.
At Simba we have been a leader in connecting BI and analytics tools to the various technologies and products in the Big Data space. We started in 1992 by co-developing the ODBC specification with Microsoft. We continue in this tradition of innovation by delivering ODBC and JDBC in the world of Big Data. Databricks selected Simba as the partner to deliver ODBC for Shark and Simba will continue to support these efforts with Spark SQL. As the world leader in connecting any data, anywhere, any time, Simba continues to deliver the connectivity technology that enables the world’s data to flow.
Many of the announcements this week at Spark Summit have been underpinned by the data connectivity from Simba. For example, when SAP HANA connects to Spark, the ODBC driver used to connect from HANA Smart Data Access (SDA) to Shark is the one developed by Simba.