Menlo Park & Strata+Hadoop World – Galactic Exchange, Inc. officially came out of stealth mode this week to announce initial beta availability of ClusterGX, an open source clustering solution which provides unprecedented simplicity of deployment and management of Spark/Hadoop clusters.
Galactic Exchange has developed a unique cloud-managed clustering technology (ClusterGX) which can be deployed on-premise or within the cloud-compute platform of a customer’s choice. Designed to work on any common operating system — including Windows, Mac OS, Linux, or bare metal — ClusterGX installs in minutes with zero previous experience required in clustering, Hadoop or Spark.
«We take simplicity and ease of use to a whole new level,» said Rob Mustarde, co-founder and CEO of Galactic Exchange. «If you can install a smartphone app, you can install ClusterGX™.» The big data industry is weighed down by complexity which prevents all but the very largest of companies from achieving rapid success and return on investment from their data analysis endeavors. «With ClusterGX™ we are democratizing the playing field for all businesses by bridging this skills gap,» said Mustarde.
Deploying ClusterGX is as simple as setting up a free Galactic Exchange cloud account and downloading the ClusterGX software onto each node that will make up the cluster. Within minutes, the smart technology within the Galactic Exchange cloud automatically assembles the compute nodes into a single, Docker container-enabled cluster. Apache Hadoop and Apache Spark are pre-loaded and ready to run with the Hadoop/Spark master nodes running from the cloud. With no defined limit on cluster scaling, there is zero additional downloads or pre-configuration required to operate ClusterGX™ — everything has been automated into the system. The administrator can easily invite other users — local or remote — to access the cluster and provide them with a secure sandbox environment to take advantage of the cluster compute power and/or data associated with it.
«The Big Data market has grown rapidly to be a multi-billion dollar industry but it is generally recognized that continued complexity is the greatest barrier to broad adoption,» said Nik Rouda, Senior Analyst for Big Data at research firm Enterprise Strategy Group. «Anything which helps reduce this complexity and accelerate the time to value for an organization must be a considered a very positive thing.»
Containers and the ClusterGX Path to Hyperconvergence
As well as simplifying Hadoop/Spark delivery, ClusterGX™ is natively built around a container architecture and utilizes technology from market leader Docker, Inc. Containers create a virtual sandbox compute environment within a single node or cluster and provide greatly increased flexibility over traditional Virtual Machine (VM) architectures.
Containers deliver significantly greater speed of set-up, better sharing of cluster compute resources across all containers and all with virtually zero compute performance overhead normally associated with VM-centric platforms.
Within ClusterGX, any type of application — not just Big Data related — could run inside its own container, allowing an enterprise to consolidate all of its applications and services onto a single Galactic Exchange cluster.
«Having ClusterGX designed natively around containers provides us with an incredible growth path for the future,» says Mustarde. «We can deliver a hyperconvergence platform to provide enhanced application performance & redundancy/back-up, reduced complexity, centralized cloud based management as well as huge cost savings from increased server utilization efficiencies and reduced server sprawl.»
«Ease of use aside, Galactic Exchange is pioneering advances in two distinct areas,» said Mike Matchett, Senior Analyst at technology consultants Taneja Group. «They are not encumbered by legacy VM constraints, so they are catching the Container wave and creating this new ground-breaking management approach by both using and enhancing the flexibility and agility of a Docker-centric architecture. Second, by moving the complexity of cluster and Hadoop/Spark management into their cloud service while preserving all big data and computing nodes on-premise, they are really simplifying and accelerating customer adoption. They are essentially unlocking the opportunity for many more businesses to take easy advantage of new analysis solutions, data lakes, and even IoT opportunities. We view this smart cloud-centered management approach as a very positive trend which we are starting to also see in a number of other areas.»