GridGain Systems Announces New Integrations For In-Memory Data Analytics
GridGain Systems announced GridGain Enterprise Edition version 7.1 and GridGain open-source Community Edition version 1.1. The announcements were made at the 2015 In-Memory Computing Summit and offer new features that allow for improved integration between Apache Ignite clusters and Apache Spark and Hadoop when deployed within Mesos, or YARN-based data infrastructures.
“The explosive growth of data consumption and the rapidly falling costs of DRAM have been aggressive drivers for new data access and processing techniques handling Big Data, and more specifically Fast Data, at the speed of RAM,” said Abe Kleinfeld, President and CEO of GridGain.
GridGain In-Memory Data Fabric is designed to deliver an ultra high-speed, scalable data solution that offers real-time streaming, as well as high speed transactions and analysis using commodity hardware. GridGain provides a unified API that allows disparate applications like Java, .Net, C++, SQL and Hive query engines to connect with multiple data sources that contain structured and unstructured data stored in SQL, NoSQL, or Hadoop data stores.
GridGain In-Memory Data Fabric offers the following key features:
- Clustering and Compute Grid – high performance, integrated, distributed memory systems with computational and transactional analytics to analyze large-scale data sets in real-time.
- Database Agnostic Data Grid – supports real-time import of data nodes with scalability to hundreds of nodes supporting local, replicated, and partitioned data sets. Data can be freely queried between data sets using standard SQL with full ACID-compatible transaction support.
- Real-Time Streaming Engine – supports both event workflow management and complex event processing (CEP) data querying.
- Hadoop Acceleration – has the ability to work as either a standalone primary file system within a Hadoop cluster or as an intelligent caching layer with HDFS as the main file system.
The New GridGain In-Memory Data Fabric Enterprise and Community editions add a number of new features, including the new Spark Shared Memory feature that allows Spark users to share data between previously isolated Spark jobs using native Spark Resilient Distributed Datasets (RDD) and DataFrame APIs. Now, users of either the Apache Mesos and YARN resource managers have the ability to manage Ignite resources based on demand. Interactive data analysis including graph-visualization using Apache Zeppelin can now be done using the Ignite SQL and Ignite APIs.
The New GridGain In-Memory Data Fabric Enterprise and Community editions are available for download on the GridGain website.