Another Word For It Patrick Durusau on Topic Maps and Semantic Diversity

March 3, 2013

Project Panthera…

Filed under: Hadoop,SQL — Patrick Durusau @ 1:38 pm

Project Panthera: Better Analytics with SQL and Hadoop

Another Hintel project focused on Hadoop.

From the project page:

We have worked closely with many enterprise users over the past few years to enhance their new data analytics platforms using the Hadoop stack. Increasingly, these platforms have evolved from a batch-style, custom-built system for unstructured data, to become an integral component of the enterprise application framework. While the Hadoop stack provides a solid foundation for these platforms, gaps remain; in particular, enterprises are looking for full SQL support to seamlessly integrate these new platforms into their existing enterprise data analytics infrastructure. Project Panthera is our open source efforts to provide efficient support of standard SQL features on Hadoop, so as to enable many important, advanced use cases not supported by Hadoop today, including:

  • Exploring data with complex and sophisticated SQL queries (such as nested subqueries with aggregation functions) – for instance, about half of the queries in TPC-H (a standard decision support benchmark) use subqueries
  • Efficient storage engine for high-update rate SQL query workloads – while HBase is often used to support such workloads, query processing (e.g., Hive) on HBase can incur significant overheads as the storage engine completely ignores the SQL relational model
  • Utilizations of new hardware platform technologies (e.g., new flash technologies and large RAM capacities available in modern servers) for efficient SQL query processing

The objective of Project Panthera is to collaborate with the larger Hadoop community in enhancing the SQL support of the platform for a broader set of use cases. We are building these new capabilities on top of the Hadoop stack, and contributing necessary improvements of the underlying stack back to the existing Apache Hadoop projects. Our initial goals are:

SQL is still alive! Who knew? 😉

A good example of new technologies not replacing old ones, but being grafted onto them.

With that grafting, semantic impedance between the systems remains.

You can remap over that impedance on an ad hoc and varying basis.

Or, you can create mapping today that can be re-used tomorrow.

Which sounds like a better option to you?

No Comments

No comments yet.

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

Powered by WordPress