14 Dec What is the potential scale of hyperconverged infrastructure?
At inception, hyperconverged infrastructure had some limited use cases, simply due to immaturity. This is expected with any new technology; when x86 virtualization first became mainstream, it had a limited use case until it could be fine-tuned. But as the architecture matures and different vendors perfect little pieces of the puzzle, HCI begins to be applicable to most use cases and most organizations. A common question a potential HCI candidate in the SMB space will ask is “Can this be small enough to be affordable to me?” On the opposite end of the spectrum, enterprise customers might ask, “Can this scale high enough to meet my needs?”
Developing a new technology – and launching a business in general – is all about developing a product that will appeal to the largest segment of your target market first. Once that is developed and stabilized, it may make sense to extend outwards to the ends of the spectrum. In this regard, hyperconvergence is no exception. The first iterations of the architecture were the best fit for the small to medium business that needed anywhere from 3 to 8 nodes in a cluster. That covers a substantial segment of the addressable market.
More recently, solutions have been announced like VSAN 6.1 offering a two-node deployment (with witness appliance for quorum) or SimpliVity’s OmniStack 3.0 doubling the number of sites that can be managed in a given federation. HCI has now reached the maturity that offerings are starting to reach the edges of the spectrum where the smallest and largest customers are also served well.
The question of scale doesn’t apply only to technological limits, however. “Does it work?” is an important question; equally important is, “How easy is it to use at [insert scale]?” This is the beauty of the HCI paradigm – no matter the scale, great attention will be given to making sure that simplicity doesn’t begin to fade. Except for in the smallest of small environments and the largest of large environments, hyperconverged infrastructure can likely serve the needs of an environment.