Cloud native and distributed SQL database presentations from FOSDEM

Key takeaways from the cloud native databases devroom, featuring TiDB, Vitess, Yugabyte, and other open source database communities.

The first cloud native databases devroom at FOSDEM 2025 brought together open source database communities to discuss distributed SQL, cloud-scale technologies, and more.

For many years, there have been different FOSDEM devrooms for database communities. For example, MySQL and Postgres devrooms have been long-standing fixtures at FOSDEM. In the past decade, several open source projects got started in order to build databases to take advantage of Cloud Native technologies that can scale across cloud infrastructure. Distributed SQL is a commonly used term to describe technologies from communities such as TiDB, Vitess, Yugabyte, and others.

Last year, I had the good fortune of connecting with community and DevRel team members from these communities. Together with people from Percona, we started discussing a FOSDEM devroom for cloud native database technologies. The purpose was to bring these communities together instead of focusing only on one community or project.

Making a FOSDEM devroom for Cloud Native database technologies

As organizers, we banded together to submit a devroom proposal, and were absolutely delighted when our devroom was accepted. When we announced our devroom and Call For Papers (CFP), we weren’t sure how this new devroom would be received and what kind of interest there would be. We were delighted when 49 submissions were made in response to our CFP. The quality of submissions made things really difficult for us to say “no,” especially since we only had one half day for our devroom. 

Being a devroom manager was a new experience for me, and I was both anxious and excited about how things would go as we got closer to FOSDEM. When we kicked off the devroom, it was amazing to see a completely full room that had a capacity for 200+ people. 

Highlights from the devroom

As for our devroom sessions, we started with presentations from several communities, and it was fascinating to see both similarities and different approaches taken by each community in developing their database technologies. We had presentations from Sunny Bains on TiDB, Franck Pachot on Yugabyte, Evgenii Ivanov on YDS, Mattias Crauwels and Rohit Nayak on Vitess (including their experience with massive database migrations), and Josh Lee’s presentation on Tracing the internals of ClickHouse.

We then had a very interactive presentation by Daniel van Eeden and Frank Pachot to compare how key Distributed SQL technologies–e.g., SQL Layer, data replication, storage, clock timestamp, etc.–are implemented between TiDB and Yugabyte.

Finally, we had several end-users who talked about their experience with Distributed SQL databases in a lightning round portion of the devroom:

• Peter Zaitsev discussing trade-offs between distributed and non-distributed architectures
• Mary Grygleski discussing practical strategies for building fault-tolerant, geo-distributed systems
• Jesus Espino discussing Mattermost’s experience with Yugabyte
• Sorin Dumitrescu sharing Omniconvert’s experience of migrating from MongoDB to TiDB
• Leandro Morgado discussing data replication in TiDB & MySQL

I really need to thank all of our presenters for making great presentations. I heard from several people in the devroom mention that they were impressed with the quality of the talks. My only regret is that we only had half of a day, and we couldn’t accommodate other proposed talks. 

If you couldn’t join us at FOSDEM, I encourage you to check out the slides and video recordings by selecting the individual sessions from our devroom page. We hope to be back at next year’s FOSDEM with another devroom, and we look forward to seeing many of you there!

More from We Love Open Source

• How to export data in DBeaver
• How to import data with DBeaver
• Why AI won’t replace developers
• What is OpenTelemetry?