How to use an existing postgres database

[Tarang]

How can I set up edgedb so it uses a postgres database that already exists? Is there any documentation that explains how this works?

As far as I can tell the current tutorials and configuration handles & manages a postgres instance behind the scenes for you.

[elprans]

One can use the --postgres-dsn=<libpq-compatible-uri> command-line argument when starting edgedb-server or specify it via the EDGEDB_POSTGRES_DSN if using the Docker image. We'll be adding proper documentation for split-Postgres production deploys soon.

[Tarang]

Thats right. I was wondering if we can use Edgedb and Postgres. If we use Edgedb does it mean we cant use the postgres part directly anymore?

[elprans]

EdgeDB can run on an existing PostgreSQL cluster (via --postgres-dsn), but you are not supposed to modify the actual PostgreSQL databases managed by EdgeDB directly. You can continue to use the non-EdgeDB databases in the cluster as regular Postgres databases, of course.

[korjavin]

I try using podman run --name=edgedb --pod edge -e EDGEDB_POSTGRES_DSN="username=postgres host=pg db=edgedb" -e EDGEDB_SERVER_SECURITY=insecure_dev_mode edgedb

but it creates another instances of pg instead of using one I provided

[elprans]

Since the comment above was posted, the option was renamed to EDGEDB_SERVER_POSTGRES_DSN. Also, it should be in the URI format: postgres://user@host/db

[korjavin]

Thank you! It works this way

[RashiqAzhan]

So, currently, it is impossible to switch to edgedb with an existing postgres instance with production data?

Is there any way to have existing infrastructure to use the current pg instance while newer services use the edgedb layer to access and modify the same pg instance?