DataStax reconnects with Apache Cassandra
Breaking up is hard to do. A couple years ago, DataStax vacated leadership of the Apache Cassandra project that it created as the community demanded a bigger voice. New players had to move into new roles, while the annual community event was allowed to lapse. Members of the community voiced their angst.
As the Apache Cassandra community ventures down the homestretch of readying release 4.0 of the open source platform for release, there is light at the end of the tunnel as DataStax has started reconnecting with the community. It’s releasing commercial support of the current open source version, reinstating the Cassandra community event, and taking a higher profile on its contributions, which continue to include documentation and free training, plus some stability enhancements for the upcoming Apache 4.0 release.
This is the company that, in its white paper The Untold Story of Apache Cassandra, billed itself as “the driving force” behind the platform.
DataStax’s re-upping with the community reflects several developments. First, that time heals all, and secondly, that both sides need their space, but also need each other. But there’s another important lesson: both sides need to step up to the plate.
While DataStax is not the only commercial player supporting Apache Cassandra — Instaclustr offers a managed Cassandra database service in the cloud — DataStax is practically the only game in town if you want a supported version on premises.
As an Apache project, Cassandra needs to thrive on a diversity of authors. In the interim, Netflix, Uber, Facebook, and Apple have grown more vocal. But with users taking the gavel, it doesn’t make sense for them to part ways with the company that started it all and the talent that got the database developed to the point where it is the 11th most popular database today. Cassandra needs the big features, like a management plane, but it also needs attention to the details, like documentation and training. And it needs some entity with the resource to hold an annual conference to get the wider community together to bring in fresh blood and ideas.
But with diversity comes responsibility. With DataStax giving up leadership, it took some time for project roles to get adequately filled. With the 4.0 version having been in the works for a couple years, the community has finally gotten to the point of freezing code back in September. It now needs to get all the new features tested.
Whoever said open source was pretty?
The wider dilemma is the dynamics of open source projects that are driven by individual vendors. As we’ve discussed in the past, there are many types of open source models, and many types of open source licenses. It’s a challenge that DataStax, along with players like MongoDB and Redis Labs are facing, but for different reasons. It’s a midlife crisis for the community for which there is no shortage of opinions and solutions.
MongoDB and Redis databases are vendor-led open source projects that they control, and where the lion’s share of development is by and from the vendor. The existential threat there is that of third parties profiteering from their work.
For much of its first decade, DataStax also dominated Apache Cassandra development, but as an Apache project, it was a community effort allowing third parties to commercialize the technology, thanks to the vendor-friendly Apache license. While third parties could profit, the original challenge to the project was community legitimacy; after the changing of the guard, the challenge becomes who makes sure that the trains run on time.
The good news is that Apache Cassandra 4.0 is getting closer to reality, and that the community is on speaking terms. But the challenge to both sides is to continue to suck it in, so to speak. The time for DataStax to lead the Apache Cassandra project has passed, but the time for it to have a seat at the leadership table should open once more.