Around a year ago, I spent some time talking to an interesting new startup that, I have to say, I was a little dubious about. The idea of Wasabi was to bring together the knowledge that founders Jeff Flowers and David Friend had in the cloud storage space (they were the founders of both Carbonite and Pilot Software). That knowledge, alongside what they saw as a huge market opportunity, would enable them to put a dent in the storage market share that Amazon Web Services enjoys.

Wasabi had, as I said, great founder credibility, and had raised a bunch of cash from some pretty credible backers. As I suggested at the time, part of the reason (I surmise) that some of those investors jumped on board was because of the next-level nature of the companies hyperbole. Wasabi wasn’t about taking market share or building a business it was, in the words of Friend:

…launching cloud storage that is so fast, so cheap and so reliable that it will mark the beginning of cloud storage as a commodity.

Oh, my. No pressure, then.

What Wasabi was doing is pretty simple. They provide cloud storage that is 100 percent compatible with the Amazon Simple Storage Service (S3) API. At the time, Friend claimed Wasabi was six times the speed and one-fifth the price of Amazon S3, and even cheaper than Glacier, Amazon Web Services’ cold storage offering.

wasabi1

 

wasabi2

On top of that, it was also touted as being far faster than AWS.

As I said at the time;

…the three big public cloud vendors—Amazon, Microsoft and Google—spend billions of dollars on both capital expenditure and R&D. They have some of the world’s best engineers and developers building their products. That fact alone suggests that while incremental improvements might be made by smart people focusing on a particular problem set or use case, massive improvements in one step are unlikely to be made. So, color me skeptical. I’m pretty sure Andy Jassy, CEO of AWS, isn’t shaking in his shoes at the Wasabi launch. But we’ll have to wait and see what eventuates.

A year on

I caught up with Wasabi recently, a year on from their May 2017 launch. The company was pleased to report that customers had stored over one billion objects with Wasabi and that the company had hit over 1,500 paying subscribers. On top of that, there are a claimed 4,000+ customers in a trial. Wasabi suggests a mix of enterprise, SMB, and consumers as customers, but doesn’t give any reference customers.

Today’s announcement

Which takes us to today’s announcement from Wasabi who are rolling out what they’re calling the Wasabi Ball Data Transfer Appliance, a shameless reference (albeit a wordier one) to Amazon’s own Snowball Appliance. What the Ball (from either vendor) is about, is allowing large organizations to more rapidly upload data into the respective clouds. It’s a pretty simple concept: grab a ball, plug it into your data center, slurp up all that mass of data you have, ship it to the vendor and – voila – they rapidly ingress it into the cloud.

In Wasabi’s case, their Ball is actually the product of a partnership with NETGEAR. Wasabi is using NETGEAR’s RN628X Network Attached Storage (NAS) system as the guts of the ball and, as Wasabi configures it, the ball can transport up to 87 terabytes of data. On top of that, multiple appliances can be used in parallel to transfer petabytes of data into, or out of, Wasabi Hot Cloud Storage.

As one would expect for a device that will be out in the public arena, security is key here. Wasabi ensures customers that their Ball is a highly secure, tamper-resistant appliance that automatically encrypts the stored data using AES 256-bit volume-based encryption, designed to ensure data security and chain-of-custody throughout the transfer process.

MyPOV

If you were to ask me a question as to whether a physical device that is designed to make it faster and easier to move data from on-premises data centers to the cloud is a good idea, I’d have to say yes. After all, AWS seems to be seeing some real success with its snowball offering. So, yeah, it makes sense for Wasabi to also offer a ball.

But if you were to ask me whether, twelve months on, I have any more of a sense that Andy Jassy is worrying about Wasabi and its threat to his company’s dominance, I’d have to say that I don’t see it. Price is a difficult differentiator to perpetuate, especially when you’re a small startup up against some massive global vendors. Performance is also a little bit temporary – despite claims from pretty much every startup I know to the contrary, no one really has the secret technology sauce that will suddenly render their competitors obsolete. Technological progress is far more incremental than that.

I’m certainly interested in keeping in contact with Wasabi and seeing how they progress, but won’t be selling down my AWS shares anytime soon (actually, just to be clear, I don’t own, and never have any Amazon shares)

Ben Kepes

Ben Kepes is a technology evangelist, an investor, a commentator and a business adviser. Ben covers the convergence of technology, mobile, ubiquity and agility, all enabled by the Cloud. His areas of interest extend to enterprise software, software integration, financial/accounting software, platforms and infrastructure as well as articulating technology simply for everyday users.

3 Comments
  • Mike O’Dell |

    Wasabi is hardly the first serious challenger for bulk storage. Likewise, everyone doing it has appliances for massive data transfer. The even come in different sizes – from small box like Wasabi all the way to 20-foot cargo containers full of multi-terabyte disks that can slurp data at a spectacular rate. In this last case, the container may be where the data stays. The container is parked on a pad supplied with “plumbing, power, and ping” and the contents joins “the cloud”.

    -mo

  • Well, neither David Friend nor Jeff Flowers has said exactly how they get 6 times the performance over S3 storage. They do mention that the performance gains come from writing directly to the disk drives without using a general-purpose file system.

    Here is an interesting quote from David Friend.

    “If you’re relying on Linux or Windows or some operating system to control how the bits are physically written to disk, you’re never going to get to our price or our performance,” he said. “This is why it took three years to build this product, after having spent 10 years building a previous version at Carbonite. You have to go in and take control of the heads on the disk drives. And you have to write the data to disks the way you want to write it to disk, not the way Microsoft or Linux wants to write it to disks. Those operating systems were not specifically designed to do the kind of data storage that we do. They were designed for a variety of things that go on inside of a PC.

    “You’d be hard-pressed to find a developer who’s ever touched any part of that stack. It’s a very arcane end of software. Outside of the team we’ve put together at Carbonite, I’ve never met anyone who’s been told they have to figure out how to write the bits to the disk. You start thinking about sectors and tracks, the movement of the heads on the drives – things like that. Nobody ever has to worry about that. We do. That skill and that ability to really control those disk drives in a way that’s optimized for our particular application is what gives us the ability to do what we do.”

    Wasabi recently filed for a U.S. Patent on how they do it, but I could not find anything about it by searching the USPTO. I suspect the USPTO has not placed the Wasabi application in the “patent pending” category yet. So if there is a “secret sauce” in the Wasabi recipe you might eventually find it described in their patent application. That said, a measure of skepticism is healthy considering the claims that Wasabi is making.

  • Well, thinking about what David Friend said about how the Wasabi “application” takes control over how data is written to HDDs, what are they going to do when flash storage becomes the new normal for data storage? Dealing with caged electrons isn’t the same thing as controlling how read/write heads manipulate magnetic domains on HDD platters. Curious about whether the Wasabi “application” will be able to improve writing and reading operations on flash which is already incredibly fast compared to HDDs.

Leave a Reply