Hazelcast supports companies for real-time data processing

Originally an expert in database acceleration, Hazelcast has expanded its field of action to support the emerging IoT market and provide applications with faster access to real-time data.

Many analytics and machine learning tools connect to data stored in data warehouses or data lakes, run algorithms across datasets or a subset of data, and compute results in cloud architectures. This approach works well when the data does not change as often. But what if the data changes frequently? Today, more and more companies need to process data and calculate analyzes in real time. The IoT is driving much of this paradigm shift, as streaming data from sensors requires immediate processing and analysis to control downstream systems. Real-time analytics is also important in many industries, including healthcare, financial services, manufacturing and advertising, where small changes to data can have financial, health, safety and other impacts. . important to business.

Rivals, for some specific functionality, of GridGain or FlashGrid, the start-up Hazelcast now focuses on real-time exploitation of data flows with its Platform 5.1. The most common use cases today are data filtering and aggregation. With the large-scale deployment of IoT sensors, businesses are faced with the need to apply real-time analytics, machine learning, and artificial intelligence algorithms that require optimized resources. In addition to storing data in memory so that it can be viewed, processed and analyzed faster, Hazelcast offers a platform that can continuously mine data as it is created, yielding insights that are more significant. The idea is more than just batch processing. Our goal is for the Hazelcast platform to be the last mile data storage and processing layer for real-time applications, whether transactional, batch or streaming. », said David Brimley, vice president of product at Hazelcast.

To ensure seamless data processing as it is created, Hazelcast has added more functionality to its in-memory platform. (Credit SL)

Version 5.1 of the Hazelcast platform includes additional stream integration and SQL query support, and better JOIN support. It also adds SQL support for JSON so that businesses can store and query using this data format to add real-time processing capabilities. By integrating the latter, the Hazelcast platform offers to enrich the data collected with information relevant to their context. This 5.1 adds the ability to create views and indexes, and run recovery plans in SQL. Hazelcast also ensures that its latest platform brings higher uptime by requiring less downtime for maintenance. In showing her ambition, Hazelcast CEO Kelly Herrell, intends to replace some large systems: “Mainframes work well, but they’re a thing of the past. And for customers who don’t want to change those existing system, the use of Hazelcast can improve its overall. ”Thus, at BNP Paribas, it took six months to deploy the Hazelcast platform and expedite the bank card fraud detection solution.

Adding tiered storage-in beta right now-to the Hazelcast platform eliminates the need to use a third-party database. Tiered storage also allows customers to easily enrich real-time data using larger historical reference sets stored on HDDs or SSDs to create the necessary context. Production is expected in version v5.2 of the platform. The Hazelcast 5.1 platform is available in the vendor’s cloud, as software to deploy in locations (a VM in fact) or in the cloud environment from AWS, GCP or Microsoft Azure. Hazelcast sells its solution in direct mode with little indirectly through its partner IBM. “We’ve been working with IBM for two years, because their data streaming solution isn’t very cloud friendly,” the CEO concludes. Other partnerships are expected in the coming months, for example with HPE.

Leave a Comment