Back to threads
Loading user profile...
Sorry I don't self-host. One of the new self-host devs Supabase recently hired comes over to discord quite a bit you might make a post there and tag it self-host.
You did not get an auto responder email with a ticket? Or no follow up after that? How long ago did you submit the ticket? Free support can be days. Drop your ticket # here and over on Discord I have an informal path to an employee to make sure tickets for infra issues are not "lost" in free.
Is this self-hosted? If so the right now you can't manage them thru studio. https://github.com/supabase/supabase/issues/10319
Sorry, just a mod/user so can't dig deeper for you. They probably are the only way to resolve it unless you have a browser state issue. Do you have multiple projects? Maybe a mix up on which one they are looking at?
You must disable the JWT verification for new API keys and JWT's. There are 3 spots that might be involved the .toml, the Edge UI has a switch, and then your deploy command (if you don't have .toml set correctly. https://supabase.com/docs/guides/functions/auth
The one thing when REST API appears unhealthy is if you disable it but the UI even though at first looks that way the description says disabled. But that would not impact auth. Any useful errors in the PostgREST or Auth logs? Read only warnings? But this is most likely a support request to resolve as you already tried restarting.
It is a totally independent thing. Your cron task uses either pg_net (which won't wait for an answer at all from the function or http extension that will wait for a response. Http is very bad to do if calling it from an external request to the db as it ties up a connection. I don't know the impact from cron. But say you use pg_net. Then your edge function gets kicked off. When it returns a response (before it goes in background mode) pg_net puts the response in an internal table in the net schema. Your function can then continue to run in background but it can't return any response to the original request, but could make its own DB/REST calls. Http you have to respond in 5 seconds I believe by default, but it has been awhile since I messed with it. The background part of the function is still independent of that.
Everyone on Discord is says it is back for them so far. And an employee says the issue was found.
No reports of issues on the Status page or Discord right now. You did not say how long. If less than a day or two wait and see if resolves (assume this is the dashboard) which you also did not clarify. Otherwise contact support as it does not appear to be a mass thing.
If this has been for less than a day and you are talking about the Dashboard there is a known issue that has not been root caused. It will usually start working again with the browser/device/network it failed on after a day or so. Otherwise what Region are YOU in (not your project)?
500MB/8GB for your database disk. 1GB/100GB storage for files. https://supabase.com/pricing
Still no when. I can't tell if I can use an informal path or not to get your ticket out of free support as it has to have been "awhile" depending on the issue. Although it sounds like it won't do any good if you don't have access, or now if it is for another free account because that implies you are using multiple emails to get multiple free accounts versus the 2 limit. Any way if you say when you filed the ticket and the ticket # I'll bump to my contact. But this is still a single employee that checks when he can, not an official queue as I'm not an employee.
If you can't prove you own the other email you won't be able to get in. Do you not know it or can't access it? You should have received emails while using it or if it was paused. If you can't get access to that old email anymore then my guess is you are out of luck as they can't just give you access to an account based on api codes or urls. But you did reply if you got a ticket # back and when.
It can vary quite a bit. How long has it been? Are you free or pro? What is your login issue? Did you get the default email with a ticket # back?
Sorry for not seeing this sooner. You need to contact support and title "Project stuck Pausing".
I believe this was resolved by the user on Discord and was logging with a different email.
This has always been you are not meeting select RLS for many different reasons (if you get deletes working and nothing else). You have to figure why you don't meet select RLS on the table as it could be no user session when you subscribe to realtime, lost session later, just bad RLS on the table. You provided no info so no one can even suggest where to look.
See the Gemini CLI section here (select dropdown). You don't add project info according to this and will get a log in prompt to sign in to supabase I believe. https://supabase.com/docs/guides/getting-started/mcp But I've not used it for Gemini but it worked for Claude Code and just brought me to a window to signup with my supabase account info if I recall.
I am not an employee. Security sort of seems like the place as you are trying to get access to an account you don't have access to. I can try and have my contact look if it is right place, but he his not a member of support or a group that can solve this issue. What does not respond mean? He says it can't be done? He asked for info and you have not heard back? If the latter the method I have is not really the right way as it is for getting "lost" tickets out of the main support queue not moving another team in supabase along.
Support is the main way. If you drop your ticket # (s) here I have an INFORMAL way in Discord to get an employee to get it out of the Free queue and to the right team IF that is where it is where it is stuck. This is dependent on when the employee checks a queue and he can only get it out of being stuck, not solve it. This has come up before but not sure how they resolve you being the real owner as they may not have your credit card details directly (I think they use Stripe) or a way to confirm you own the card. Certainly they can't just take your word for it.
Sorry. You did not state your region. So less likely. More info would be needed probably for users checking here to help (not regularly monitored by SB staff) or check the Supabase Discord as busy on the issue I linked. Otherwise contact support.
https://status.supabase.com/incidents/9t3fvctpj1hq <img width="803" height="385" alt="image" src="https://github.com/user-attachments/assets/edd314de-e485-4ff4-a861-8d0251e166b4" />
What is your ticket #? I have an Informal way to escalate after that long when the employee checks the queue in discord.
You will need to have support unstick it. Did you already file a support request and get a ticket # and when?
Postgres is not a supauser in Supabase. `Drop powersync_role;` is not working?
https://status.supabase.com/incidents/79kw380212jb
From the MFA docs: <img width="1120" height="171" alt="image" src="https://github.com/user-attachments/assets/61d017ad-b94d-4f89-8361-ff78ee23eef2" /> This was from support to another user recently: `Thanks for reaching out to Supabase Support. Unfortunately, we are unable to recover your account/MFA if you no longer have access to any of the MFA devices. It is recommended to set up a backup factor as mentioned in our documentation (https://supabase.com/docs/guides/platform/multi-factor-authentication#enable-mfa) to prevent situations like this.`
JS client IS the REST API. So you must have RLS protection. Do you have RLS enabled on the table? If so you must have a delete/select policy to protect from anyone deleting. Check the dashboard REST API logs for Deletes. Was it 1 user's rows only? If so seems something coming thru for that user IF you have RLS enabled and a delete policy.
Do you use the REST API? Do you have delete RLS on the table to protect it? Do you have more than you with access to the dashboard? Someone with CLI access to your project? You also seem to be saying it is 1 customer's rows that disappeared. That seems like a targeted delete from somewhere. Just one table or cascades from one table?
The AI is correct on other possible ways than your code. All the other things are controlled by you including migrations.
Are they gone in the dashboard table UI as postgres role with no filters? Someone on your team could have deleted in the dashboard. Your RLS could be bad and someone deleted (even if you don't use the REST API you have to have RLS enabled or turn off the API). Could the customer delete and not be saying so? Is there a pattern to the deleted data. I suspect you will have to solve it. I know of and have never seen postgres or Supabase delete "most" rows. Certainly reach out to their support, but I'm skeptical they have ability to see what caused the deletes.
The op found his issue was AI deleted rows. Supabase/Postgres on its own does not delete from your tables and I've never seen a confirmed case of the cause not being something in your code/AI/user (you or someone with access) deleting the info. If you are using the REST API check the logs. You might have backups if Pro that can help (restore to another instance to pull data if needed). I am not an employee and this is not official support so no one can see your instance, but there is not postgres tracking of deletes if your are using DB connections.
Storage.objects is enabled by default and as of 1/2 year or so ago locked down so you can't alter tables in the schema (just add/remove policies). Contact support assuming this is hosted supabase.
This was discussed in Discord and the only is to contact support. Never delete rows from storage.objects. The OP has contacted support.
You could remove public from the API schemas and put your functions in another schema that is added to the API. I'm assuming you really mean rpc versus SQL database connections calling a postgres function. There is also a pre-function that postgres runs that can tell a call is a function and potential error on everything else. https://supabase.com/docs/guides/api/securing-your-api
https://status.supabase.com/incidents/4ydyhb0wrtkc
What region? . EU-West-1 was/is having issues with project creation. https://status.supabase.com/incidents/4ydyhb0wrtkc
You need to provide more info on what not working is. Did you signup with Github? Many times a login issue to the dashboard is a known state issue bug. Try a different browser, then device, then vpn.
OK that is more than just your auth email not working. The other area is just a logging error usually temporary not your instance. But with your database and auth down then your choices are to try a restart from general settings. If that does not work you have to contact support. This is not support here for specific instance/project issues.
@athuljayaram Your issue seems like an auth.users insert trigger or auth hook you added going wrong. Probably don't have public. in front of profiles if you insert to that. You need public.profiles for any sql you do in the trigger function.
Check the auth logs in the dashboard. There should be a detail error there. If you switch to built in does it still work?
upgrade to 2.91.1 or higher was there was bug in 2.91.0.
Not heard of this not working... One of my projects: <img width="882" height="387" alt="image" src="https://github.com/user-attachments/assets/347e007a-4f5b-4542-9928-ea6747a188c7" /> Here: <img width="864" height="640" alt="image" src="https://github.com/user-attachments/assets/5ee7e3ce-8ece-4173-8150-2852b4a08101" /> Try hard refreshing browser? Otherwise contact support or like you said create a new account...
Not at all sure but suapbase-js is at.... 2.91.1 And you do have a setting for it, but default is one hour... <img width="1416" height="553" alt="image" src="https://github.com/user-attachments/assets/a10c5a3c-d1c6-44e3-8454-23c94f99fbdd" />
Try general settings at bottom.
If just your dashboard is not working try another browser then device and network/VPN. If it is the known random state issue dashboard bug it will start working again on the original browser/device in a day or so. Otherwise unless you have more info your support request is the best avenue.
If you have not figured it out yet there was a bug in 2.91.0. 2.91.1 is now released.
In Discord the Mods have an informal path to bump project down type situations to make sure they get to the infra team out of general support. I've added your ticket # to that. No guarantee if that will be faster than support getting to it or not. You should get a follow up email with a message moved to the infra team when out of the general queue. Supabase staff is unlikely to see your message here.
That is a community library and maintained directly by Supabase. Also I'm not an employee in case you were implying I could do something about it. You two might post in that repository or generate a PR to fix it.
Publishable and Secret keys are the replacement keys for anon and service_role. You might try `npx supabase status -o env` and see if that shows the legacy keys.