email iconemail phone iconcall



Serving Customers, Serving the Community

By Jonathan Ellis, CTO and Co-founderNovember 2, 2016

Focusing on DSE

For six years, DataStax has taken a modular approach to building DataStax Enterprise (DSE): Analytics, Search, Security, and now Graph have all been added to the open source core of Apache Cassandra™.

With DSE 5.0, we have reached the limits of what we can do with this approach.  The feedback from our customers is clear: they need us to focus on providing a completely unified platform to drive the next generation of cloud applications.

My team of full-time Apache Cassandra™ developers is thus joining the DSE team and will be prioritizing their efforts towards improving and enhancing that product at every level, including the core.  Our goal is for DSE to continue to be the best version of Apache Cassandra™ in every sense.

What does this mean for Apache Cassandra™?

After I stepped down as project chair in August, the Apache Cassandra™ Project Management Committee nominated Nate McCall to step into that role. Apache Cassandra™ is healthier than ever, with eight new PMC members and four new committers added this Summer.  It is actually not good for an Apache project to be dominated by a single company’s contributions, and I look forward to the innovations resulting from this more diverse leadership and development.

A strong Apache Cassandra™ remains a high priority for DataStax and we are extremely committed to helping the community advance.  DataStax’s ongoing involvement with Apache Cassandra™ includes:

  1. Contributing selected features like range-based compaction, time-series support, materialized view enhancements, and improved backpressure.
  2. Providing testing, build, and continuous integration resources.
  3. Continuing to drive adoption with conference sponsorship and Meetup support.
  4. Donating resources like the dtest suite and Planet Cassandra to Apache.
  5. Making available free, self-paced training on Apache Cassandra™ through our DataStax Academy.

Where we’re going next

Apache Cassandra™ is closing in on its 4.0 release.  With tons of new features released in the tick-tock process since 3.0, there’s a lot to learn, test, and build on.  Apache needs your help!

While you’re checking out the latest releases, take advantage of the new development documentation and get involved.  Don’t be afraid to ask questions on the mailing list and irc; especially with a crop of new committers looking to learn together, there’s never been a better time to participate!

Meanwhile, DataStax is hard at work on exciting new features in DSE 5.1.  Stay tuned!



  1. Thomas says:

    This post makes it sound very much like your team will be moving away from focus on the OSS project, which doesn’t sound good. In fact, there’s been a fair amount of discussion on Twitter on exactly what that means for the future of Cassandra in addition to less public channels. Can you clarify what’s happening here and what non-DSE users should expect the impact of this change to be?

  2. Jonathan Ellis says:

    Thanks for the question, Thomas.

    I want to make it crystal clear that DataStax is *not* abandoning Cassandra. As I wrote above, a strong Apache Cassandra remains a high priority for DataStax and we are extremely committed to helping our customers and the community advance. This includes all the activities I listed above, as well as extensive bug fix contributions (which is less exciting and didn’t make my original list but in terms of sheer volume is going to be the largest piece).

    Remember that for years, DataStax Enterprise has been delivering a stable, enterprise-ready version Apache Cassandra, plus a whole lot more. We have always had enterprise features not available in the open-source version, including Analytics, Search, and most recently Graph.

    So all that is really changing is that with Cassandra now a mainstream success, we can turn further attention and effort towards ensuring our customers can leverage its amazing power, with the additional capabilities in DSE, for their own application development.

  3. Christian says:

    One of the big reasons we decided to move to DSE from open-source Cassandra was the possibility to easily move back from DSE to open-source Cassandra. (As long as we are not using any DSE-only features)
    That was true for DSE 4.8.

    So will that also be possible with DSE 5+? Especially as you plan to contribute only selected features.

  4. Jonathan Ellis says:

    As with current DSE vs Cassandra, that depends on how you decide to build and architect your application, and which features you choose to use. (This is even true even when migrating between different relational databases!)

    What we will make very clear to everyone with each release is exactly what is in DSE and what is in C* alone. From there, the architects can make decisions on what and how they will leverage in DSE.

  5. Carsten Ziegler says:

    So either it takes the node.js route with a temp clone (io.js) that later gets merged into a new main branch owned by a new broad foundation…
    … or Cassandra’s open source version just vanishes slowely. Datastax may hope for a lot of conversions into DSE. However, as a decision maker I would rather pull out completely. And the taken approach may also harm other open source technologies which are mostly driven by one company which is true for many. Not a nice outlook.

  6. George says:

    You even stopped making RPM builds for the community. Was it so much trouble?


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