31a7f69c-e284-469d-bfac-e7c9ca208610

!t

The user has mentioned that this hasn't been working for the past 3 days https://discord.com/channels/713503345364697088/1067670962276945961/1389253623065743360.

@bledar - What environment should I be looking into?

Production and beta

Would you mind if I tried to deploy a database into beta? just so I can grab the debug info.

Yes please, you can do it

You might be able to see more details about this error

I have reported the issue to the team.

this may not be a database specific issue actually. im seeing the same thing on my project, it says creating volume-name then after a bit unknown error like above. potentially to do with volumes? im in us-west

oh

i think i know what it is actually

prior to the metal migration i noticed when i create a service with a volume, even if the location is in metal, it seems to create the volume on gcp - because when the service is created i see "we migrated this volume for you, looks good?" even if its completely new

my guess is after the full metal migration it fails to create the volume now and that results in the service unable to deploy

will have to see if it works if i delete the volume

hey @bledar and @sixfalls - we have pushed a fix, can you let me know if you are still experiencing issues?

It works now.

Take more time than normal, but it is creating the DB instance successfully

I’m trying to create a new PostgreSQL instance, but the process is extremely slow. When the instance is finally created, it’s placed in a legacy region by default. Additionally, attempting to change the PostgreSQL version from v16 to v17 results in errors. I also tried wiping the storage to resolve the issue, but that doesn’t seem to be working either.

Legacy regions should be completely deprecated and shouldn't be deployable, can you show me where you see that you are deploying to them?

Postgres itself does not support upgrading major versions, a volume wipe would be needed.

By default, all new PostgreSQL instances are created on a legacy region.

You can change it later, and a migration will be run automatically.

Wiping volume takes a lot of time and is failing

Deploys on Legacy regions will fail as Legacy regions have been disabled, can you show me where you see that you are deploying to them?

Please provide errors for the volume wipe.