Recently, I can click on any post that says it has as few as 1 comment, and I see the skeleton loading state on the comments for like 30 seconds before any of them load.
This happens on my 4G data, or at home on tested 400 Mbps fiber.
This is greatly disrupting my Reddit addiction.
Edit: holy wow! It appears to be fixed now?
Has anyone else had issues with branch reset on persistent branches?
I had a persistent development branch tied to my gh dev branch. I hit "Reset" in the dashboard multiple times. Row counts are identical before and after. Even data I inserted through MCP execute_sql (not part of any migration) survived the reset. The branch status showed ACTIVE_HEALTHY like everything was fine. I looked in branch management logs, all good.
All of my migrations were only related to schema, yet data stayed! I saw the branch logs show happy rebuild. Still no change. The branch seemed to replay cached copies of the original migrations regardless of what was in git.
Eventually I deleted the branch entirely and created a new one (not persistent this time). The new branch came up with clean schema and zero data, exactly as expected.
So it seems like reset is just broken on persistent branches? Or am I missing something about how persistent branches handle resets differently?
This has happened to me multiple times now. Each time I end up deleting the branch and creating a new one, which means a new project ref and updating API keys, env files, and configs everywhere.
Pro plan, linked to a github development branch.
I am on us-east-2, and when that went down, it took me down. It truly sucked. Thanks to pure luck, it did not destroy my fledgling company.
I am about to have a truly huge contract sign, where something like that is just not acceptable.
Does anyone know the timeline for Supabase multi-region failover? Is it very soon, or do I need to bite the bullet and migrate to AWS/CF/some auth for everything?
Any guidance would be truly appreciated!