The Evolution of Hyperconvergence
Hyperconvergence is a type of infrastructure system that integrates servers, networking, and storage in a single box, supported by a single vendor. Hyperconverged systems are managed as a single system and have evolved rapidly, with each iteration further simplifying operations and reducing costs.
Traditional Systems: The Dinosaur of Server Infrastructure
Traditional infrastructure setups required expensive hardware, and lots of it. The amount of manual involvement was also immense. The servers had to be managed by server professionals, networking had to be managed by networking professionals, and the storage by storage experts. Each of these groups had their own tools that added to the expense of operating on traditional infrastructure setups.
Not only were these traditional setups complicated, but they were difficult to scale. As a company grew, adding more storage became complex and expensive. All three groups—servers, networking, and storage—had to work together to build out additional VMs. This level of complexity was a barrier for growth and stalled a company’s ability to tackle new initiatives.
Cisco Delivers Converged Technology
Roughly five years ago, Cisco was the first to market with converged technology. They saw how complex traditional systems had become, and offered a new solution. Their idea was to provide a single system where everything the client needed to build a data center (servers, storage, networking) were supplied in one rack and managed by one tool.
While converged technology took the market by storm, it was not without its problems. In this first iteration of converged technology, clients purchased the solution from Cisco, but “under the covers” it was still several different products, albeit tightly integrated ones. Since the solution comprised various components from different vendors, it was still somewhat complicated to set up. Hundreds of pieces of hardware would arrive on-site followed by a group of techs to put it all together. Then, another group of techs would have to come to configure the system. The whole process could take up to six weeks.
Once the solution was in place, however, a single console was used to manage the entire system. Though converged technology was somewhat complicated to deploy, once it was up and running, it was much easier to manage and much less expensive than traditional systems.
Nutanix Introduces the World to Hyperconvergence
Since converged technology was not without its limitations, it was only a matter of time before someone developed an even better system. A company called Nutanix took a different approach to Cisco’s hyperconverged solution. With Cisco’s converged technology, clients still had to purchase a rack that included severs, switches, trays of storage, etc. At the end of the day, it was still a huge piece of equipment made up of components from different vendors.
Nutanix’s first iteration of hyperconvergence took a web-scale architecture mindset. Everything an enterprise needed—storage, networking, and servers—was built right into a box. In order to increase capacity, an enterprise simply added more boxes. A layer of software consolidated everything and provided easy console management.
While Nutanix helped to simplify things, its first attempt at hyperconvergence was not perfect. There was no data management layer and multiple sites typically required LAN acceleration and outside backup solutions. Of course, this meant that there was room in the market for a newer solution.
SimpliVity Delivers True Hyperconvergence
A small startup called SimpliVity created some very disruptive hyperconverged technology. So much so that Cisco, HP, and Dell all announced their own plans to launch similar versions of the solution.
What made SimpliVity so disruptive? They were the first to add a layer of data management to their hyperconverged technology. The solution was so robust that the biggest downside was job loss. Because everything—including backups and data deduplication— was managed in one system, it eliminated the need to have separate teams of networking, server, and storage professionals.
The business case for hyperconvergence is tremendous. For example, HP has a hyperconverged solution called Moonshot. This solution uses web-scale architecture and the solution itself is nothing more than a processor, memory, and storage. HP was able to replace the 11 server racks that ran hp.com with just three Mooonshots.Hyperconvergence is poised to change the way companies do business. Providing simplicity though a single solution, companies can not only save money, but they are able to scale up on demand, and small enterprises now have the same accessibility to advanced infrastructure solutions as their larger counterparts, helping to level the playing field like never before.