DataStax shifts focus back towards the Cassandra community

Towards the end of last year, NoSQL database provider DataStax announced that its long-standing CEO Billy Bosworth would be standing down so that the company could pursue a “more ambitious, bold approach”. 

Former Google VP Chet Kapoor was recruited to take the top job, heading up a fresh leadership team to shift the company’s focus. But, to what end? DataStax is the leading commercial entity behind the Apache Cassandra project, an open source database that has proven popular with organisations that deploy complex, multi-region applications built for the internet. However, there was clearly a sense that more could be done to fuel its popularity. 

This week we got some insight into what that future direction might be. It was announced that DataStax would be acquiring The Last Pickle, a leading Apache Cassandra consulting and services company (for an undisclosed amount), which counts Spotify, T-Moble and AT&T amongst its customer base. 

But the acquisition, whilst interesting in itself, reflects a broader wave of change happening internally at DataStax. 

We got the chance to speak with DataStax’s new Chief Strategy Officer, Sam Ramji, and the company’s VP of Server Software Engineering, Joshua McKenzie, to find out why acquiring The Last Pickle indicates how DataStax is shifting its focus back towards the Cassandra community. 

Ramji explained: 

So, the new strategic direction probably won’t surprise anyone that’s a student of the explosive open source businesses that you see in data. You’ve got great examples like DataBricks, Cloudera, Confluence, etc. 

The model is pretty straightforward. You work with the community to generate an increasingly capable project, solving a really technical problem, making sure that open source solution works and makes the impossible possible. 

By creating that abundance of something that really seemed shockingly hard, and doing it in open source and making it free, you create a new scarcity - now that you’ve made that thing really trivial. People are then curious about how to solve higher order problems.

Ramji said that for Cassandra, in particular, there are three core areas of focus: 

1. Solving for multi-region clusters - creating a high performing database that is serving one or more coherent applications around the world, on a common dataset. 

2. Integration - people want to use it with other tools, such as Kafka or Spark. 

3. As-a-Service - Operating at scale is a challenge and companies are looking for service providers to manage that for them. 

Referencing the acquisition of The Last Pickle, Ramji said that the organisation has experience of these three areas, working directly with the open source Cassandra community. 

The shift that’s signalled by The Last Pickle is to take on a tonne of open source community orientation, because they are an outstanding citizen of the Cassandra project. They know Apache Cassandra well. Also, they really know these thorny problems of management, integration, as well as serving Cassandra at scale. 

We’ve seen a lot of excitement as we start to turn our attention back towards solving our community problems and contributing to open source.

Making the lives of developers easier

McKenzie added that DataStax will be investing in tools that The Last Pickle has fostered, such as Medusa (which helps with back-up and restores) and Reaper (focuses on cluster repair). He said: 

What the The Last Pickle has done specifically is they’ve worked with some of the biggest Cassandra installations in the world. For example, what Spotify has tended to do is solve the problems in-house in a way that is generically usable by others, as well as contribute that back to the community. The Last Pickle has been the facilitator to bring that forward.

But one of the key things to note is that DataStax’s renewed emphasis on the community will be about making the lives of developers easier. This is something that DataStax competitors - such as MongoDB - have been investing in for a while, recognising that if they can simplify the work of the developer, traction and support will come from the ground up in the enterprise. 

McKenzie said: 

One of the big things we are looking at right now is, what is the ecosystem of usability and tooling for developers with Cassandra? And how can we help make that more robust so that it’s an easier ecosystem to get into and it’s easier to run your modern applications on? That’s been a big shift in our thinking and our focus, so that we can become a force multiplier for what the rest of the community is investing in and doing.

My take

As alluded to above, DataStax has seen a lot of success in recent years, but its focus has largely been about leading from the front with its own products, rather than necessarily supporting the community from the ground up and taking its lead. That appears to be changing and I think it’s a smart move. We’ve seen success elsewhere in the market from this approach. However, it’s a fine balance between carving out your own commercial value and priorities, alongside getting stuck in with the open source community. As Ramji noted to me on our call, the biggest risk to DataStax is execution, not necessarily the NoSQL market, which is growing at a rapid rate. It’s still early days as the new leadership team gets stuck in, but we will be watching this closely to see how it develops.