If you were watching social media or blog/media coverage of VMworld two weeks ago, you’ve heard a lot about hyperconvergence. Not unlike “cloud” was 5 years ago (and to some extent still today), it’s a word that is applied loosely to many different technologies and platforms. Now that it’s achieved a buzzword status, the industry must narrow down the definition of the term.
Clearly the IT industry is going through a major change. It started with the integration of the entire stack, including all the same infrastructure pieces you’re used to, delivered and supported in a rack-level appliance (what SimpliVity refers to as Convergence 1.0). The next step was the convergence of compute and storage using local SSDs and HDDs to provide fault-tolerant shared storage, across multiple hypervisor host servers (Convergence 2.0). Now we’re at the step where we can truly hyperconverge the legacy IT stack and deliver the entire rack in a single device (Convergence 3.0 or Hyperconvergence). You can read our CEO’s more detailed take in his post last December.
There are several elements that differentiate convergence 2.0 from hyperconvergence (Convergence 3.0), and enable the true vision of hyperconverged infrastructure:
- The inclusion of products such as backup, disk backup appliances, WAN acceleration, global data management and cloud gateways
- The collapsing of all the legacy management layers into a single, familiar interface, thereby reducing the training and management overhead required by the legacy IT stack of independent hardware/software stacks
- Crazy simple, almost appliance-like deployment of new nodes to enable easy up-scaling of the environment
- An equally simple removal of nodes should be provided to enable easy down-scaling of the environment when capacity is no longer needed or upgraded capacity has been added to the environment
- Single-vendor support for all aspects of the hyperconverged stack
- A VM-centric management approach, again simplifying the management experience
- Storage efficiency baked into the product to reduce the capacity and performance requirements of the underlying storage devices
For additional information I’d recommend picking up your free copy of the eBook Hyperconverged Infrastructure for Dummies by industry veteran and analyst Scott D. Lowe. After that, I’d encourage you to check out http://www.hyperconverged.org, a website that Scott created with David Davis (another long-time industry watcher) to help the industry start to understand and narrow in on a definition for the term “hyperconverged.”