As the NoSQL database market continues to evolve and mature, the biggest players involved are beginning to not just think about the technology capabilities, but also the delivery models that would set enterprise buyers.
The latest to broaden its offering is DataStax, which offers the leading distribution of Apache Cassandra via its DataStax Enterprise data platform. It has now announced that the same platform capabilities will be made available as a managed service, running in the public cloud and with a ‘white glove service’, following demand from customers.
It’s not hard to understand why customers may wish for DataStax to take control of the operational aspect of their deployment. Firstly, Cassandra operates as a distributed system and complexities with managing sprawl can be seen. Secondly, skills are always hard to come by, particularly in a relatively immature market. Thirdly, and most importantly, many companies will likely see the operational aspect of their NoSQL deployments as not value-add to their business.
DataStax has said that the newly announced managed service is purely for enterprise customers looking to run production environments in the cloud - it isn’t for test/dev.
Billy Bosworth, CEO at DataStax, said:
At DataStax, we understand the stakes are high for organizations to deliver engaging experiences in today’s right-now economy and we’re committed to helping our customers quickly get the best business returns from their data infrastructures.
DataStax Managed Cloud provides flexibility for our customers, which allows them to focus on higher-level pieces of their business. DataStax Managed Cloud makes our portfolio the most comprehensive in the data management industry and the most laser-focused on cloud applications that define computing today. I’m excited about what this means for DataStax and I’m even more excited about how this translates into real value for our customers.
The managed service is immediately available on Amazon Web Services, and will soon be made available on Microsoft Azure and Google Cloud Platform.
I got the opportunity to speak with Martin van Ryswyk, EVP of product and engineering at DataStax, to hear a bit more about the launch and to try and understand a bit more about why customers are demanding this service, as well as why DataStax sees it important.
The biggest problem for existing DataStax customers is finding the right skills to manage the often large systems. Van Ryswyk said:
What we’ve seen is that as business has picked up and as we have grown and our customer base has grown, like any new technology, finding skills starts to become a challenge. Particularly when it’s not 30 years old, something like Oracle.
We’ve had a number of customers that have come to us and said that they love the end result, the application developers get it, but it’s a challenge to actually operate it day by day. It’s a distributed system, it’s not just one box. They sometimes have to monitor hundreds or thousands of machines, all as part of this one system - and they want to know if we can do that for them you. And the simple answer is yes, we can.
Van Ryswyk said that it was important for customers that the DataStax Enterprise platform that they have on premise would be exactly the same in the cloud, with the same capabilities, so that applications are portable. DataStax has ensured this is the case.
He added that the managed service will be mean working very closely with customers.
We are calling it ‘White Glove’. And by that, we mean it’s very hands on. We will work with them before we even spin up an instance. What are your performance needs? What are your application demands? What is the schema you were using? We need to know everything about it so that we can craft and provision an environment for you. We are watching it on a day to day basis, with monitoring and management.
We have been doing this a lot of years. We have learned a lot about what it takes to manage and operate a production DataStax Enterprise cluster. Because we do it every day, we see the things that they might never run into.
Prior to launching the service, DataStax carried out some research on the potential appetite for it from customers. What it found was that approximately half of its customers are already running their DataStax platforms in some sort of public cloud environment (except are running it themselves) - and as such, are perceived to be very cloud friendly.
The research also found that approximately half of DataStax’s current customers would be interested in letting it run their operations for them. However, van Ryswyk was cautious about what speed the company will see uptake in the service - but did estimate that at some point in the future it could be as much as 50%. He said:
We’re serving enterprises. And they’re conservative. So for them to change anything can be a long process. I think we will get new customers. We will get existing customers putting projects on, maybe not converting older projects. But based on everything we’ve heard, there’s huge demand. Our job is to move as fast as they want to move. I think we will approach the 50/50 mark eventually, based on what we’ve seen.
Van Ryswyk was also cautious when I asked him if this meant a higher margin opportunity for DataStax. He said that the economics of it are yet to be fully established, but added that DataStax is doing what it can do introduce automation into the managed service, so that it can scale and focus resources more effectively.
I don’t know how the economics like that will work out. There are certainly some thoughts that by managing it for customers, we will get fewer support calls, because they will have the A-team in the world managing it.
At the same time, there will be extra costs, because we will have people managing it. We’ve been working on the automation to make it efficient for us, so that we can handle many customers doing the routine things in an automated way. Then keep our experts for the things that aren’t so routine.