After yesterday's outage they admitted that their elasticsearch index for issues/prs lost data.

They seem to have changed the primary source of data in the issues and pull requests tabs (w/o filters applied) from the underlying database to the elasticsearch search index, which has the side effect that there's a noticeable delay between state change of an issue/pr and an update in the UI. But as seen today, these can get out of sync, and apparently they even had data loss in the index.

I would really like to know their reasoning for making that change. I can totally imagine that they wanted to "simplify" so the UI uses only a single data source instead of two.

As a user it's incredibly annoying to have a delay between issue/pr state changes and the search index picking it up.

What? React has nothing to do with current state of affairs. In fact, React on GitHub currently exists in mere islands, i.e. in Projects and recently in Pull Requests. Most of the frontend is still Web Components[1] paired with Turbo[2] for hot reloading. GitHub is still as slow even with JavaScript disabled, try it yourself. Backend just serves stuff really slow. In fact, there is an alternative GitHub frontend (no affiliation) that feels snappier and is written in React.[3]

With that said, Mitchell complains about outages. These started directly after Microsoft acquisition[4] and are attributed to migration from AWS to Azure.

[1] https://github.blog/engineering/architecture-optimization/ho...

[2] see html source for tags

[3] https://my.githero.app/

[4] https://damrnelson.github.io/github-historical-uptime/