Splice Machine boosts hybrid capabilities of its Big Data platform


Splice Machine Inc. has beefed up its relational database management system’s hybrid workload capabilities with the release of version 2.5 of its platform, which is powered by Apache Hadoop and Apache Spark.

Specifically, Splice Machine’s platform is a dual-engine RDBMS that specializes in something called hybrid transactional and analytical processing. What this means is that it can perform both complex online analytical processing queries and time-sensitive online transaction processing queries simultaneously, without any performance degradation.

With the update, Splice Machine’s platform now uses resource isolation – separating processes and resource management for its Hadoop and Spark components, to ensure that OLAP queries don’t overwhelm OLTP queries and cause workloads to slow down. Basically, the idea is to eliminate the delay involved in querying a relatively high-latency secondary storage medium such as Amazon S3. With Splice Machine’s new hybrid architecture, customers can perform analytical and transactional workloads concurrently, which should be a massive benefit for use cases such as ETL acceleration, operational data lakes, data warehouse offloads, Internet of Things applications, web, mobile and social applications and operational applications.

Besides boosting its hybrid capabilities, Splice Machine has also added support for columnar storage, cost-optimized storage and in-memory caching for Amazon Web Services’ users to its platform. Splice Machine was on hand at AWS re:Invent to demonstrate how users can take advantage of the new capabilities on AWS to integrate compute and storage engines into an elastically scalable database, which serves as a combined data warehouse and relational database.

“The new capabilities further emphasize the benefits of Splice Machine’s hybrid architecture,” Monte Zweben, cofounder and chief executive of Splice Machine, said in a statement. “For modern applications that need to combine fast data ingestion, web-scale transactional and analytical workloads and continuous machine learning, one storage model does not fit all. The Splice Machine SQL RDBMS tightly integrates multiple compute engines, with in-memory and persistent storage in both row-based and columnar formats.”

Another interesting new feature in Splice Machine 2.5 is support for Data Sketches, a library created by Yahoo Inc. to speed the analysis of high-volume, streaming information such as website metrics. It avoids the overhead of individually scanning every new item by using statistical estimation methods that sacrifice some accuracy for a big performance boost. During a test that Yahoo conducted last year, the framework managed to blaze through a sample set of 100 million values in just under three seconds.

Splice Machine SQL RDBMS 2.5 is available now.

Additional reporting by Maria Deutscher.