GridGain Systems Inc. is trying to fix a need for speed with a new update to its in-memory data fabric platform that promises a 2X performance boost.

GridGain offers a high-performance, integrated and distributed in-memory data fabric based on Apache Ignite that’s designed for computing and transacting on large-scale data sets in real-time. GridGain Enterprise Edition 7.5 is designed to deliver programmatic real-time analytics, machine-to-machine communication and high-performance transactional processing using clustered x86 server nodes in a grid defined by a distributed, massively parallel architecture. The end result is that applications to run far more rapidly than they would if running on SSD-based systems, and “thousands of times faster” than they would on disk.

The startup donated Ignite to the Apache Software Foundation in 2014, and its first major release came out last year. The GridGain platform, which sits on top of databases such as Oracle, DB2, Postgres, MySql, and Microsoft SQL, integrates compute, data, service and streaming grids into a single system that also comes with Hadoop and Spark integration. GridGain’s unified API encompasses the C++, Java and .Net programming languages. GridGain’s speed derives from the fact that data is cached in-memory, with Ignite’s compute grid performing parallel in-memory processing. It also provides fault tolerance and continuous availability of deployed services via a service grid that lets users control the number of instances that can be deployed on each cluster.

The new features in GridGain Enterprise Edition 7.5 deliver an enormous speed boost, with most cache operations and SQL queries now up to two times faster than the previous version, according to the company. GridGain has also introduced something it calls “deadlock-free transactions” that eliminate issues like deadlocks and lock ordering, meaning data pools can be accessed by multiple teams at the same time. In addition, GridGain Enterprise 7.5 now supports OSGI containers, enabling automated deployments and re-deployments within a container environment. Finally, the platform now supports open-source projects like Apache Camel and Apache Flume, support for automatic data ingest from Twitter or MQTT, and support for compact protocol, enabling dynamic schema changes and SQL indexing on the fly.

“With the latest release of the GridGain In-Memory Data Fabric, we have focused on increasing speed, accessibility and reliability of data to ensure our customers can gain the competitive edge they need,” said Abe Kleinfeld, President and CEO of GridGain.

Those who’re unfamiliar with Apache Ignite might think GridGain’s platform is something of a rival to the more popular Apache Spark engine. But although both Ignite and Spark use in-memory as the main storage paradigm, the two platforms are actually quite complimentary to one another.

“Apache Ignite fills a gap in the ecosystem by providing a solution for high-performance transactional processing and real-time machine analytics,” GridGain’s CTO and founder Nikita Ivanov told SiliconANGLE last year.​ “That’s something that neither Hadoop nor Spark were designed to do.”

Ivanov explained that one of Spark’s major weaknesses is that it doesn’t have an integrated storage layer, and relies on external storage layers like Cassandra, HDFS and HBase. But when Spark is integrated with GridGain, Ignite acts like an in-memory database that allows Spark to process data in-memory alone, with Ignite helping to choose which data stays in-memory and which can be stored persistently.

image credit: alan9187 via pixabay.com