Severalnines: Mixing DevOps with database management and going its own way

James is editor in chief of TechForge Media, with a passion for how technologies influence business and several Mobile World Congress events under his belt. James has interviewed a variety of leading figures in his career, from former Mafia boss Michael Franzese, to Steve Wozniak, and Jean Michel Jarre. James can be found tweeting at @James_T_Bourne.


Sweden-based database management provider Severalnines is an interesting company. Its name would suggest SLAs but the firm, with approximately 7000 users, supports MySQL, MariaDB and MongoDB among others and enables businesses to support both SQL and NoSQL open source databases in the cloud or on-premise.

Yet the way Severalnines has built its success is of most interest. Instead of going cap in hand to VCs and building up cash without going near making a profit, Severalnines has done it the hard way, relying predominantly on the success of its ClusterControl product to expand, with customers such as BT, Orange and Cisco on board. And with a 100% growth in the past 12 months, things are certainly on track.

Vinay Joosery is CEO and co-founder of Severalnines. He argues startups today are focusing at their goals from the wrong angle.

He tells CloudTech: “Five or 10 years ago, I didn’t see as many young people really dreaming about starting their own company, but now it seems like today, when you go to some of these networking events, the dream is [to] have a good idea, build the prototype with your friends, and then go and contact the VCs.

“Basically you get excited about the wrong things. Solve a real problem and find real customers who are willing to pay for that problem – I think that’s probably where the focus should be.”

Severalnines has done that. The company’s focus is on architecting for failover – a big problem when managing databases and the majority of a firm’s staff doesn’t know how to set things up. Add cloud failures and multiple data centres to the mix and it all gets very complicated.

Covering both SQL and NoSQL use cases, Joosery argues Severalnines doesn’t push its customers in any direction – “they are the experts in what they want” – but adds: “What they don’t know though is – does this database work for me? We think that the technology is not enough. It needs to be maintained, it needs to be scaled, if there are issues you need to know how to fix it. These are the kinds of things that we address with the product.”

Plenty of complexity gets thrown in with the working environments in large enterprises, who will typically have an army of Oracle database admins in one room, an army of sysadmins in another, and armies of developers elsewhere. The growing DevOps movement is changing this landscape, and for Severalnines – which includes support for Puppet and Chef – it’s a key trend to play into.

“Our product has been addressing that angle from day one in a way,” Joosery says. “You have people managing the database, they’re doing some programming, they’re building web services, they know a bit of everything. What you will see in ClusterControl today, it’s doing a number of things addressing mainly on the sysadmin point of view, all the database aspects. The stuff it does is a combination of sysadmin tasks that you need to do on a database, as well as pure database administration tasks.”

Joosery adds another benefit of DevOps is writing infrastructure as code. In the future, you won’t need to put boxes together, screw in cables, or do anything manually. You write a program to spin up VMs in the cloud and set up connections instead.

Watch this space for further developments on that front. Yet there is a common theme between pursuing DevOps architecture and building distributed systems to architect against failover.

“We are helping people with their databases, but we are not doing it as the other tools vendors are,” Joosery explains. “They are taking an approach where they are looking at single instances. What we do is focus from day one to build tools for systems, and the system for us today is a distributed system, which can be distributed across multiple machines.

“In a distributed system you assume there will be failures. The shit will hit the fan, that’s just by default. Our software is built from the ground up to make sure that these things are easy.

“Let’s say you have a bunch of services, you’re running it in a couple of Amazon regions and you want to have the same data in your own private data centre as well. If for example the Irish region of Amazon gets knocked out, [you] failover to Singapore or failover to the US or failover to your own data centre.

“This is why we are called Severalnines – people need more availability.”

View Comments
Leave a comment

Leave a Reply

Your email address will not be published. Required fields are marked *