WebScaleSQL Moving Forward

December 2016

WebScaleSQL began as a collaboration among engineers that had similar challenges in running MySQL at scale. Members of the MySQL community — Facebook, Google, LinkedIn, Twitter, and eventually Alibaba — began working together to create a database technology tailored for their needs.

The collective goal of the companies was to enable the scale-oriented members of the MySQL community to work more closely together in order to prioritize the aspects that are most important to us. We learned over the course of this experiment, however, that sometimes each company need may not be similar enough to make sustained collaboration possible. Everyone was at different stages of MySQL lifecycle timelines; it wasn't productive to try to work on the same one. That's why, as you may have noticed, the WebScaleSQL collaborators are no longer contributing to the WebScaleSQL branch.

Many of the WebScaleSQL companies continue to make MySQL better through other projects: Alibaba with many changes for InnoDB performance, Facebook with MyRocks, Google with encrypted InnoDB. We'll continue a meet-up of WebscaleSQL participants at Percona Live once a year. It is always great to exchange ideas.

Please welcome Alibaba to WebScaleSQL!

January 2015

We're excited to announce that the Alibaba MySQL team has joined WebScaleSQL. In fact, they've already pushed their first few commits. We know they'll be a great fit for this project based on the work they've previously done in this space, from general bug fixes and specific performance improvements to enhanced large-scale parallelism and micro-optimizations in compression.

Things have been moving along well with WebScaleSQL. We've begun to push the planned replication and GTID server-side fixes. In recent weeks, we've pushed several other improvements including bug fixes and testing fixes, some contributions from the wider community, and the foundation for the async client feature. Major enhancements are in the works as well: the async client protocol, logical read-ahead, query throttling, improved server level thread pooling, large InnoDB page support, and a major reduction in memory usage are all on the horizon. With Alibaba joining the group, we anticipate even more progress to come.

Things are ramping up with this collaboration, and soon we'll be seeing more of their code and code reviews getting into WebScaleSQL.

Welcome to the project!