Sometimes you need to quickly dump and restore a PostgreSQL database, but what's the easiest way to do this when your database is in a Docker container?

Check out this quickie here if you're using MariaDB or MySQL.

I ran into this just today, and thought I'd share one method that I felt was easy, fast and served my purpose. Depending on why you need to dump/restore a database, this might help for you, too.

This quickie assumes you have nothing directly installed on your development machine, so everything is run straight from and to the Docker PostgreSQL container you're running.

Dump using pg_dump

docker exec -i pg_container_name pg_dump --username pg_username [--password pg_password] database_name > /desired/path/on/your/machine/dump.sql

Restore using psql

docker exec -i pg_container_name psql --username pg_username [--password pg_password] database_name < /path/on/your/machine/dump.sql

Note: If you are attempting to restore data from a custom format dump, you should instead use pg_restore as I described in my How to set up and use Postgres locally article.

Note: By default PostgreSQL keeps importing even when errors occur. If you would instead prefer to stop the import completely upon error, be sure to add --set ON_ERROR_STOP=on to your above command.

Dump and restore in one command

If you, for example, are moving data from one (e.g. manually created) container to another, you could use pipes to do this in one command, like so:

docker exec -i pg_old_container_name pg_dump --username pg_username [--password pg_password] database_name | docker exec -i pg_new_container_name psql --username pg_username [--password pg_password] database_name


There are certainly other ways to achieve something similar, but this method will work in a pinch. Please keep in mind that you should always ensure your production databases are properly backed up, and ideally automatically so. If you use any of these manual steps as a means to create backups, you're probably doing something not entirely correct. These steps are mostly for moving development data around or pulling (partial) production data locally for debugging, or something along those lines.

That's it, really. I hope this note-to-self quickie will be of some help to you as-well.

Thank you.