PostgreSQL Gets a Parallel Processing Boost
Database vendors are rolling out frameworks for accelerating analytics on streaming data, including attempts to address growing pains associated with the scaling Postgres-based analytics.
For example, Swarm64 is aiming its database acceleration software at open source PostgreSQL deployments that, despite their popularity, fall short when it comes to analytics and data warehousing. The Norwegian vendor claims the latest version of its platform boosts PostgreSQL performance at least 20-fold, making the open source platform more attractive than proprietary solutions for analytics and data warehousing upgrades.
As the company name suggests, Swarm64 leverages parallel processing to accelerate and scale PostgreSQL performance, spinning up 64 parallel threads to execute queries. The 4.0 version released last week installs as an extension to open source PostgreSQL, thereby eliminating the need to change a user’s application or SQL code.
The extension is designed to expand parallel query planning and execution, the company said, along with columnar indexing, data compression and reduction I/O. Along with faster queries, the latest version also accelerate data ingestion while promising consistent response times.
The new version also speeds the process of joining data from different tables, a widely used step for SQL queries that can also slow them down. Swarm64 said its latest release accelerates a variety of queries that previously slowed databases.
Along with software, Swarm64 also offers hardware accelerators in the form of optional FPGA support designed to add SQL read/write processes running in parallel on a co-processor. The result, the company said, is the ability to handle larger databases and real-time analytics.
Other cloud-based approaches to accelerating PostgreSQL performance have emerged over the last year. For example, Pivotal Software also unveiled an open-source Postgres platform targeting applications such as transactional workloads and prototyping. The company also announced collaboration with Crunchy Data to offer its PostgreSQL framework for running both application and container services on Pivotal Cloud Foundry.
Along with data warehouse modernization, Swarm64 is positioning its data acceleration platform upgrade as a way to scale emerging software services that incorporate PostgreSQL analytics while reducing cloud database costs. The upgrade also is geared to scaling machine learning and Internet of Things platforms used to analyze time-series data.
Swarm64 Data Accelerator 4.0, which also runs on any cloud, is available now via the company’s web site or from the Amazon AWS Marketplace. Swarm64 said its software is licensed on an annual subscription basis.