By Anoj Willy, Vice President, Cloud Solutions, Trace3
I was speaking to a friend who is a real car-aficionado and he lamented on how he used to go looking at cars with his father and ask questions about the piston pressure, rotations-per-minute, and even fuel-injection stats. “Nowadays people don’t even ask those questions” he said, “they have no idea what goes into everything they are buying!”. “Halleluiah!”, I say. We haven’t talked in a little while.
Personally I’m ecstatic about this change that came about when I was apparently too young to understand cars or how to buy them. Now that I’m older, and still don’t understand anything about cars, the fact that I can go on a website and pick out a model, select a few options, and price-shop has made this process so much easier. I don’t need to know the intricacies, but I do need to know that I am making a solid investment.
When it comes to data-centers this notion of a solid investment is far more critical and it is a massive exercise for companies. If your car breaks down, one or two people are inconvenienced. If your data-center goes down, on the other hand, your whole company could shudder to a halt.
Due to the increasing demands on IT, data-centers have grown more complex and unwieldy. Most enterprises have multiple storage and compute platforms, disparate networks, and much more. If you take a look inside your datacenter and you might feel like Noah, carrying two of every animal in your HVAC controlled ark. What’s more the cost to the business and the burden on IT to just maintain these systems is massive. And this is just to stay in place. This is not progress.
Hyperconverged Infrastructure (HCI) is a shift toward consolidating and simplifying all the disparate components that make up data-center infrastructure onto a centralized platform. To date the three major areas of spend for data-center infrastructure are storage, compute, and networking. Early players to the market like Pivot3, Nutanix, and Simplivity have entered the market by combining the storage and compute layers onto a consolidated platform and showing that there is real market share in hyperconverged. Now that HCI is a proven billion-dollar market growing at 60% CAGR big-fish like Cisco are coming to the party.
Who will be victorious and who will fall by the way-side is a crystal-ball mystery that few can determine but many will bet on. There are two clear trends that players in this space have to keep a close-eye on if they don’t want to end up an also-ran. The first is favoring simplicity over functionality (making the complex simple), and second, to focus on de-risking sales execution above all else.
Making the complex simple
James Green, from virtadmin.com, asserts “Because of the assumption that high-performing must be complex, there’s an inherent implication that a simple solution must be low-performing”. He goes on to say “Hyperconvergence defies the industry’s notions about complexity and performance by being both simple and complex, yet still high performing”.
The software that makes a hyperconverged solution work at highly performing levels is quite complex AND an equal amount of work is done to ensure the simplicity isn’t compromised. The focus is less about the details of the infrastructure but rather the outcomes. Key attributes like scale, reliability, and efficiency are important but are just table-stakes. The real question of how does this support my workloads and what is the lifetime cost of this asset loom larger. IT leaders want to know “ will it become a staple in my data-center or just another one-off?”. The big factor we have seen driving internal adoption are the number of interfaces admins have to deal with on a day-to-day basis and how much effort is needed to manage it.
If the management functions are consolidated “under the hood” then the interface must reflect that. There’s no point having all the functionality available to me if I have to rummage through screens or wrangle with inadequate API’s to control the things that are important. With new players like Cisco entering the market with Hyperflex (HX) and Dell (formerly EMC formerly VCE) pushing VxRail the big players are certain to push the feature parity game up to the next level. Interestingly enough Cisco is poised to bring one major differentiator to bear that has not been talked about – the network layer. To date the network has been the most challenging to automate and bring under a common management pane. It still remains one of the most complex systems to manage in a data-center. Regardless of the reasons why, the network remains as one of the biggest pain-points and therefore biggest areas of opportunity for new entrants. If anyone is to bring this massive new functionality to bear they must take an equal amount of care to ensure the simplicity that was the initial promise of HCI shines through.
This type of differentiator however is still a long-term play. In the next 24 months Gartner and IDC have both predicted a massive consolidation in HCI vendors and new entrants into the market. For a leader to emerge at this stage they have to focus relentlessly on de-risking sales execution. HCI is no longer considered “hype” it is part of every enterprise infrastructure conversation. The technology has been proven and the market response has been extremely positive. Now expanding to new markets before the competition can get there is the next hurdle for all companies and the most dangerous for new businesses. VC’s call this “the cliff” and it’s where the majority of their investments fail. Asheem Chandna from Greylock often says “a technology is not a product, and a product is not a business”. It doesn’t matter that you can stuff more IOPs into a box or add features more rapidly, can you rapidly take market share when customers are willing to buy.
Join us for part-2 of this post in 2 weeks