Advanced Computing in the Age of AI | Friday, March 29, 2024

Hazelcast Platform Targets Cloud, App Container Rollouts 

In-memory computing specialist Hazelcast Inc. rolled out the latest version of its "data grid" that includes support for new cloud management and application container deployments along with a higher density memory store for web-scale application deployments.

Hazelcast, Palo Alto, Calif., said Thursday (Jan. 21) its version 3.6 release upgrades the performance and functionality of its in-memory data grid designed to allow users to share and partition application data across their installed clusters and servers. "With this release we have become a platform," Hazelcast CEO Greg Luck declared in a statement announcing the release.

The company's platform is intended to help customers manage their data and distribute processing using in-memory storage and parallel execution. Those features are used to scale and speed the delivery of enterprise applications.

Along with new cloud management and container deployment features designed to ease configuration and management of cloud and hybrid on-premise rollouts, Hazelcast said it has added a high-end persistence feature called "Hot Restart Store." The new capability targets enterprise and e-commerce applications, allowing apps to start up "hot," or with their data. The feature is said to eliminate the need to reload data.

The result, Hazelcast claims, is a startup time of 1.3 Gb/sec. per node on an SSD, resulting in application restarts within one minute for a 1-Tb data size. The reboot capability also in intended as a backup in case all nodes in a cluster crash while enabling software updates without having to reload data. Hot Restart Store also supports a range of data structures, including JCache and Web Sessions, the company said.

Version 3.6 also includes high-density memory for supported data structures. The company said the memory expansion would enable memory caches for large enterprise scale-out and scale-up application rollouts. The memory store also boosts the amount of data held by each node—in the range of hundreds of gigabytes per node, the company claimed.

Along with sheer horsepower upgrades, the compute platform reflects the growing shift toward "lightweight" micro-services. Hazelcast said its version 3.6 release includes new deployment options for industry-leading Docker containers along with support for the Kubernetes container management software released to the open source community by Google (NASDAQ: GOOG) in 2014.

Other cloud enhancements include support for Apache jclouds along with the CloudFoundry and OpenShift platforms-as-a-service, Hazelcast added. Meanwhile, a new cloud discovery interface allows auto-discovery of cloud or on-premise nodes with reduced configuration and management requirements. The feature also includes an Amazon Web Services (NASDAQ:AMZN) cloud module that helps Hazelcast cluster members discover each other on AWS.

Version 3.6 also includes an open client protocol designed to support development of new clients. CEO Luck noted that the latest version of the compute platform now supports clients for eight different programming languages. The company stresses its support for developers looking to upgrade existing applications or building new ones.

Hazelcast customers included Capital One (NYSE: COF), the Chicago Board Option Exchange, Deutsche Bank (NYSE: DB) and Ellie Mae, the mortgage compliance specialist.

About the author: George Leopold

George Leopold has written about science and technology for more than 30 years, focusing on electronics and aerospace technology. He previously served as executive editor of Electronic Engineering Times. Leopold is the author of "Calculated Risk: The Supersonic Life and Times of Gus Grissom" (Purdue University Press, 2016).

EnterpriseAI