DataCore CEO: MayaData Container Tech ‘Ready For Prime Time’

‘When there’s a generational foundational change like what we’re seeing with containers, a lot of innovations come from the smaller companies, but the channel generally is owned by the larger companies. ... And that’s why you see companies like Veeam, Pure Storage, and DataCore go out and try to acquire some of this innovative technology to basically run through their channels,’ says DataCore CEO David Zabrowski.

Bringing Storage And Containers Together: DataCore Acquires MayaData

When software-defined storage pioneer DataCore acquired MayaData, leading independent developer of container-attached storage and the original developer of the OpenEBS open-source Kubernetes storage technology, it was the third major storage vendor to do so following Veeam’s October 2020 acquisition of Kasten and Pure Storage’s September 2020 acquisition of Portworx.

DataCore early last year invested $26 million in MayaData to build storage-focused container technology, a move inspired by Google’s move a few years ago to develop the Kubernetes open-source framework, said David Zabrowski, CEO of the Fort Lauderdale, Fla.-based company.

“Containers was going to be a once-in-a-generation type of transformation,” Zabrowski told CRN. “And we had some particular expertise because of our performance and our availability expertise from the block storage side, so we were pretty excited about containers. So I actually started a internal ‘skunkworks’ project that ran for about a year, and it became apparent at that time that we thought we had a better approach.”

DataCore is already seeing the fruit of its initial investment in MayaData and in its acquisition of the company as it prepares to release both an open source and a licensed enterprise-hardened version of the OpenEBS storage container technology that was developed by MayaData.

Zabrowski talked with CRN at length about DataCore’s acquisition of MayaData and its prior Caringo acquisition, and about his company’s plans going into 2022. For a look at where DataCore is going, click through the slideshow.

First of all, how would you describe DataCore and what’s the latest in the company?

We’ve spent the last three years at DataCore building out a platform company. We are the largest independent provider of software-defined storage solutions, and we have a vision called DataCore ONE which was the work that I had done with Insight Venture Partners, which is our largest investor and one of the most successful investors in the industry today. And also, when we get to MayaData, [Insight is] also the largest investor in container companies, ironically.

We did an assessment of the storage industry, and it’s very siloed. It’s mostly built on proprietary hardware and proprietary homogeneous solutions. Most of the other technologies in the data center have gone into a software-defined kind of heterogeneous migration, and because of innovator’s dilemma storage hadn’t. So our goal at DataCore through the DataCore ONE vision was to essentially break down those silos, deliver an entire storage solution under the software-defined storage umbrella. DataCore started on the block side. So the higher the performance and the more available that your data is needed, the better we do. And then we identified the container space as one of the most strategic and eventually the highest growth parts of the storage market. We also identified object and file as two markets that we wanted to expand in and start really elevating up the industry from these siloed kind of homogeneous solutions into a software-defined stack. So that’s what we’ve done the last three years.

[DataCore’s Caringo acquisition] brought to the company what we think is the industry’s best hybrid object store. And we’re in the process of ramping that business up pretty significantly. Caringo had had a lot of success in certain verticals, like M&E (media and entertainment) and government, and we’re adding to those verticals and bringing that through our entire channel. And you’ll remember DataCore is very, very channel focused. We think we’re the only provider that is exclusively channel-focused in this space.

You just said DataCore is one of the few in ‘this space’ to be channel-focused. When you say ‘this space,’ how do you describe the space that you’re in? A lot of storage hardware vendors are also focused on software development, not hardware.

Well, we broadly just define it as storage, if you look at the analysts’ view of storage. But most of those software investments are really software investments in homogeneous solutions, whereas DataCore is known for being very heterogeneous, and we work across anybody’s hardware, anybody’s software installed base, as well as new stuff coming out. So that’s really part of our secret sauce. It’s one of the reasons that DataCore has been successful. And it’s quite unusual in this space because most people are still trying to peddle their stuff on their own goods. If you look at any of the competitors out there, even if they say they work with others, they really struggle because they’re not architected to do that. So that’s how we define the space: It’s just all-up storage.

Why did DataCore acquire MayaData?

[Three years ago,] it was really obvious that containers was going to be a once in a generation type of transformation. And we had some particular expertise because of our performance and our availability expertise from the block storage side, so we were pretty excited about containers. So I actually started a internal ‘skunkworks’ project that ran for about a year, and it became apparent at that time that we thought we had a better approach. ...

At the time we thought the best way to go to market was to partner with someone who has the open source expertise, who has a similar type of approach through a container-attached storage type of architecture. We looked at everybody in the space and found MayaData. And we created a joint venture of sorts with them [with a $26-million investment in early 2020]. We created a separate board of directors [for the joint venture]. We put money, we put intellectual property, we put cross-license rights, and we transferred our team into the MayaData team. So we basically took the best of both worlds.

MayaData had OpenEBS at the time, which was at the time and still is today the only CNCF (Cloud Native Computing Foundation) track project for open storage container-attached storage. We combined our two efforts. And so we basically acquired them because the product’s getting ready for prime time. OpenEBS just went version 3.0. So it’s a very mature product in the CNCF world. The downloads are pretty substantial. It’s a million per week for downloads of OpenEBS. And that’s the open source version. A product that we’re working on today is really an enterprise-hardened version of OpenEBS that will be coming out under the MayaData brand, and that will basically be a licensed product. There will be an open source version and a licensed version of that. And so it was a marriage that really started three years ago when we identified containers as the highest growth area for storage and made the subsequent investment. And now this is the culmination of that marriage.

So the new container technology was developed by MayaData before the acquisition, or by DataCore?

It was built by MayaData combined with the DataCore team that we put it over there. When we made the investment, we merged our two teams. The output of that effort is what will be announced and released in the early part of next year.

Prior to the acquisition of MayaData, had DataCore released any offerings specific to Kubernetes or to containers?

We have a CSI (container storage interface), which is basically a plug-in that allows the existing installed base to adopt the Kubernetes framework and then access the DataCore storage on the back end using a CSI driver.

So why didn’t DataCore just continue with its own efforts rather than invest in MayaData?

Because of the container storage architecture, which by design has to be architected from a blank sheet of paper. It is based on a microservice architecture, and it’s really designed to be the highest performance, most agile storage for Kubernetes containers. Just think about the very foundation of a container. Basically it gets spun up when it’s needed, and when it’s not needed, it gets spun down. That’s basically how it works. And when you have storage, you need to keep persistent tracking of that. You spin something up, let’s say an e-commerce site, there’s storage that’s created from that transaction that has to be tracked. So there has to be persistence associated with that. That persistence has to be done in a very specific way with the architecture for container-attached storage.

The storage industry has been busy with container acquisitions. DataCore is acquiring MayaData about a year after Veeam acquired Kastenand Pure Storage acquired Portworx. Is there a trend here?

Like a lot of markets, when there’s a generational foundational change like what we’re seeing with containers, a lot of innovations come from the smaller companies, but the channel generally is owned by the larger companies. We have a channel. We have over 10,000 customers. So we can take the MayaData innovation, which was assisted by DataCore because of the merger that I mentioned before, [and] build it out through a MayaData channel, but we also have 10,000 end-user customers that are ready-made. And by the way, we just did a survey of this. Fifty percent of our installed base already has some version of containers running in their enterprise. So we have a ready-made channel. And that’s why you see companies like Veeam, Pure Storage, and DataCore go out and try to acquire some of this innovative technology to basically run through their channels.

How many channel partners does DataCore have?

We have 3000. We’re a very channel centric company. We sell exclusively through the channel. ... We’re known to be very channel friendly and very, very much a channel company partner.

Did MayaData have any strategic relationships with other storage vendors?

They were pretty agnostic. They worked with many players in the space. And so for us, they will be leveraged in our Alliance partners, and that will be an acceleration for their business as well.

What’s next for DataCore. What are some things you’re working on as you start approaching 2022?

Basically, our SANsymphony product is still doing great. It’s been the product that DataCore is most known for, but it’s growing. We’re investing in it. We’re coming out with a new version here shortly. This quarter we’ll grow 20-percent-plus in that product. Second is the object side with the Caringo product, which is called Swarm. DataCore Swarm is the brand now. And that product’s doubling and tripling this year and next year. So we’re very excited about that. MayaData, once we get the company integrated particularly with the enterprise product that I mentioned earlier, that’s the third strategic area for us for 2022.

And for people who may not be familiar with it, can you describe DataCore SANsymphony?

SANsymphony is our block storage. So in the industry, we talk about hot, warm, cool, and cold on the storage spectrum. This is on the hot side. So where applications are most demanding, sometimes databases would be an example of that, that’s a target of SANsymphony. It’s for high performance, high availability block storage. Caringo is object storage, which is unstructured data. Whereas the database is generally more structured data, and high performance, object store would be things like hospital images, media and entertainment, high-performance computing, artificial intelligence, machine learning language. These applications generate unstructured data. And that’s stored on an object storage. It’s generally not as high performance as block, but lower cost.