Commvault Says Game On With New Hyperscale Appliance

Commvault has partnered with RedHat to build a scale-out hyper-converged appliance running Commvault's software, sold as a subscription. Called HyperScale, it's a clear validation of the Rubrik and Cohesity approach to backups and secondary data.
Both Rubrik and Cohesity have enjoyed strong growth with their approach to secondary data management, and Rubrik is boasting of valuations of greater than a billion dollars. That kind of market action demands attention.

The thing is, Commvault does have a good story to tell here, one that can work for its existing customers and to help it steal share from its major competitors like Dell EMC and HPE. It's not a mere copy, but rather Commvault's way of providing what customers clearly like about the Rubrik/Cohesity offering.

“Scale-up architectures are outdated,” said Commvault's Senior Director of Worldwide Solutions, Don Foster. “Customers need something that can handle the size and scale of secondary data in the modern enterprise, with all the data services Commvault can provide.”

Commvault uses RedHat's Gluster file-system to do the scale-out storage, and uses a 4+2 erasure coding approach for data resilience. Commvault is also selling the product on a pure subscription basis, priced based on usable target storage. If you purchase the Commvault appliance hardware, you get it refreshed for free every three years, but Commvault are also providing reference architectures for hardware vendors commonly found in the enterprise: Dell EMC, HPE, Lenovo, Super Micro Computer, Cisco, Fujitsu and Huawei.

This is smart, given Commvault's enterprise customer base. They often have whole-of-enterprise agreements with vendors for server and storage hardware, and this lets them use their existing relationships to acquire hardware they know and like (and have the tools and processes established to manage). This is a low-change option, which makes adoption easier. Just layer the Commvault software on top.

Commvault also supports end systems that startups choose not to, like Unix. Maybe the new folks will decide the effort is worth it, maybe they won't, but for an enterprise with substantial investments to protect, it's easier to choose an extension to a known and trusted brand than to deal with sprawling point solutions. Heterogeneity and brownfields is the norm, and dealing with that complexity is one reason these deals are so large.

There's also the not insubstantial investment in tools and processes that already exist. If you can get at least some of the benefits of the new approach without a wholesale re-tooling of your entire organisation, that's an attractive and lower risk option. The added value doesn't need to be as high as for a completely new approach. Change is not risk free, and how many companies do you know that handle large-scale organisational change well?

You can argue that this is a reaction by Commvault, but really, it is doing just what large enterprises should do. They don't have to invent all the new things themselves, because that leads to Not Invented Here syndrome. Equally, they can't ignore all new developments in the marketplace, because that leads to stagnation and death. If it's what their customers want, then they need to be providing it. There's no shame in being a fast follower, and there are plenty of examples of it being an extremely profitable approach.

The clear winner here is customers, and I say that without any sarcasm (a shock, I know). This is an example of customers winning when a new, fresh approach shakes up a staid and complacent market. There is clear value from the Rubrik/Cohesity approach, and the secondary data market had been more-or-less stagnant since Data Domain was acquired. Since then, hyper-converged technology has matured, storage densities have increased markedly, and flash has become vastly more affordable. Combine that with investments in engineering to make very complex systems easy to use, and you have a lovely confluence of technologies that make a new approach commercially viable.

What we see now is wonderful, because the timing is rather great. Rubrik/Cohesity have enough momentum to make them serious players, and Commvault has entered this market at quite a good time, I think. It's been proven as a market that's worth being in, but it's growing fast enough that the competition isn't too intense. That gives everyone time to make a bunch of money while their offerings improve, meaning we as customers end up with better products, rather than simply ones that are optimised to take as big a piece of a fixed pie as possible. There are incentives for all of the players to grow the pie as well as their own relative share.

Now look for Dell EMC, HPE, Veritas, and all the other heritage enterprise backup and recovery vendors to pile in with similar offerings. The fear of missing out on what is now very clearly a real thing means will ensure that.

Expect to see some minimum viable reworkings of other vendors' existing products as a knee-jerk reaction from those who didn't pick up on this market early enough. We should also see one or two quite well engineered offerings from those organisations who have reasonable long-term marketing and strategy capabilities. We'll see some entrants who come in too late, and are forced to exit after lackluster sales, and possible one truly great piece of engineering that will have sadly missed the wave, which will get acquired by one of the middle strugglers as an attempt to reinvigorate its mediocre offering. The wonderful tech will then sink deep into the belly of the acquiring beast where it will die from lack of attention, to the chagrin of technology lovers everywhere.

The startups have had time to establish a foothold, but now we're seeing competition really star

As Foster told me: “Game on.”

This article first appeared in Forbes.com here.