February 17, 2015

Written by Nate A.

Hyper-converged infrastructure has picked up a lot of press lately.  Some tech industry analysts have even predicted 2015 to be a big year for hyper-convergence.   The VMware Partner Exchange conference saw a steady stream of debuts from vendors announcing new systems.  The past few weeks saw entries from EMC, HP and NetApp join offerings from Dell, and Nutanix.  Hyper-converged systems combine compute, storage and networking components into one package.  Where this differs from a traditional server is that the package consists of enough parts to create a highly redundant system that provides high availability – think “cluster-in-a-box”.  In addition, the systems often use simplified management tools which enable rapid deployment.  Moreover, the systems usually work like building blocks, or Legos.  When more capacity is needed, customers purchase another unit, and tie them together via the management interface.  Virtualization software such as VMware’s hypervisor often provides the virtualization layer although some vendors offer Linux based alternatives.   Many of the systems leverage software defined capabilities to customize aspects of the stack to the meet the customer needs.  Hyper-converged systems disrupt the traditional datacenter offerings by providing simplified “bricks” which take away the complexities in ordering, implementing, and operating new datacenter equipment.