garyaustin

You need to contact Supabase thru their support contact form if you have an account or support@supabase.com. There are sales and enterprise contacts also somewhere on the main website. You likely won't get much info from users here in Discord unless someone who has done HIPAA happens to see your post. HIPAA info... https://supabase.com/docs/guides/security/hipaa-compliance https://supabase.com/docs/guides/deployment/shared-responsibility-model#managing-healthcare-data

Nothing we can do without ticket numbers. Even then it is an informal bump path one employee checks when he can. Direct support is the way.

Have you reached out to the two support teams? Vercel kicked of Supabase is a small subset of users and chances of another user having advice is slim. There is no resolution to this being stuck other than Supabase (I don't know the Vercel relationship with project ownership by them).

There is no back up email feature for the account login. You could create a 2nd account and invite that account to the org for your projects. I don't think they would consider this a violation of terms especially if you share the org without extra free plans.

Is only one project missing?

You said you were invited to the org that the project was in. This means you were part of a team. Can other teammates see the project? Did someone remove you from the team? Can you still see the organization itself and just the project is gone?

You did not mention anything about projects showing up or not. So you get into a supabase account with Github or that email? Then you are in there and you don't see an Organization you are part of a team for and you accepted an invite earlier and it worked at one time? Do you have two accounts one with that email and another with github?

What happens when you try to login the various ways?

Not really clear. You can't use email login in now even though you created the account with email/password? You can't login with Github with the same email? You can't login with Github with different email?

If you don't get an automated message within 6 hours saying you have been moved to another team (like infra) drop your ticket # here and we have an informal way to make sure it gets found in general support queue.

Did you leave in RETURNS TEXT?

Then you have to wait for support.

I added a note to the repository about that. I need to find some time to go in and get it updated with pg_net bulk delete and this.

And I'm guessing as that was not needed before with the JWT style service_role key.

Basically same as Authorization but without `bearer `

You may need to add the apikey header with the new SB_ style secrets. I've not tested with those.

Are you sending both headers? Like in your orginal code clip?

I'm just telling what the info says that error means.

Your error is that your apikey/authorization headers are wrong format or wrong key.

You are using a supabase hosted project. It runs on their servers.

You need to contact support from top right of dashboard if it has been more than a few minutes. Make sure you try on a different device or at least another browser.

There are dashboards for all three. Does your url look like this: `https://ugcxxxxxxxxxxxxxx.supabase.co` If so you are hosted.

So you are running CLI local supabase project or self-hosted?

If you are using the SB_ format keys you need to use that for both apikey and Authorization header I'm pretty sure.

Double check the keys match from the API Keys settings. Is this a hosted Supabase cloud project?

Check the auth and postgres logs. Check API settings to see if data API is on. (should not impact auth though). Check usage in settings for any overage. Otherwise probably contact support.

Have you restarted the instance/project from general settings?

I deleted your post as you showed what I think was your real service key.

What is the status of your project from the home page of the dashboard? Where are you located (there is a DNS for one provider in Brazil area)? No other reports so seems project specific.

Postgres logs showing password errors?

Did you already have a support request in? Sure you are going to the same instance you are changing the password on?

I did bump it to the informal queue, but the one employee that monitors it has not marked any processsed since Fri/Sat.

Add some logging and log out the url you build. Check Postgres log for those or errors. Check the Gateway API log to see if your request is getting to the server.

If you need to debug you can use `raise log 'thing = %', var;` and check postgres logs. Also in the net schema there is a table with the errors (if any) that come back from the request. Delete normally does not error though if I recall for bad file names.

This shows the URL formats for storage. https://supabase.github.io/storage/

That is normally used to read a public bucket only.

You also have the public in the URL. That probably should not be there.

And what is in the policy? That is just a description Also is that on storage.objects?

We have no access to status within Supabase other than that status page.

You have to meet both select and delete policy on storage.objects OR use the secret key.

I don't know anything about next.js environment, but they would be in your dashboard under API keys.

Unless your RLS lets anyone delete then you should use the secret key.

That is the public key.

Besides needing to change to the new API keys not sure why it would not work. Most of the dated stuff is just ability to do things better now with storage meta and pg_net doing bulk delete.

https://github.com/GaryAustin1/supa-file-helper Is dated but gives some examples of pg_net and http with cron.

And I assume this was for creating an account with them, versus on your app trying to signup with the API. I believe so based on the message wording.

Have you used that email to signup before with possibly a 3rd party integration like Vercel?

Yes.

You need to contact them thru that email you show.

Also if you drop your project id in DM (no keys needed) I can confirm it works or not from the US. But using a VPN confirms it is something to do with your office network or DNS provider they are using. Not likely a Supabase issue.