Tell HN: Supabase database restore from backup corrupting projects
There are multiple reports that restoring a database from Supabase backups currently corrupts the entire project.
The Supabase dashboard shows the following error message:
“Something went wrong while restoring your project. Your project’s data is intact, but your project is inaccessible due to a restoration failure.”
However, it appears that at least some of the backups, if not all, are either corrupt or unavailable.
As a result, data is lost and users are unable to access their projects. There are numerous reports on the Supabase Discord server, including from paying (non-free) users, stating that they haven’t received a response for several days (up to 8, from what I’ve seen).
AFAIK, this issue is still ongoing: restoring a database can still corrupt your project, and I haven’t found any official communication from the Supabase team addressing the problem.
In my view, this represents a worst-case scenario for managed cloud services. Supabase should immediately disable the restore function and start communicating transparently about what happened to users’ data.
https://discord.com/channels/839993398554656828/1006358244786196510/threads/1431693614785892634
https://discord.com/channels/839993398554656828/1006358244786196510/threads/1431059097515720744
https://discord.com/channels/839993398554656828/1006358244786196510/threads/1430844413625634816
https://discord.com/channels/839993398554656828/1006358244786196510/threads/1430626295586226207
https://discord.com/channels/839993398554656828/1006358244786196510/threads/1429534157834420355
https://discord.com/channels/839993398554656828/1006358244786196510/threads/1429028717123862539
and more....
Comments
Strange they just got a huge investment cycle, maybe trying too hard too fast because of that? I love using them personally but their 'database is paused due to inactivity feature' is something I wouldn't really depend upon, If I let the database get that stale I basically see it as 'gone' in a sense. If it was important it would've been replicated locally, etc. Supabase serves as the front-end layer/proxy.
Looks like making a better Firebase is a bit hard. Superbase is neat, but hosting it yourself isn’t fun.
Hopefully they’ll be able to fix this data corruption, although a backup on the same host isn’t really a backup. The whole system can have issues
I'm only hosting a newly started side project there, but it does have paying users so I'm really unhappy at the moment.
Do you have any backups aside from what’s on Superbase ?
that's not good. i have 8 dbs with them.
This is a worst case scenario. Even worse is the no-communication and not turning off the restore function. This is having serious economic impact