@rkazak helped me figure out some problems I was having with persisting data from a Postgres Docker container. I wanted to save the notes, so I’m posting them here in the forum.
Scenario: I uninstalled Postgres on my laptop and have been running it in Docker containers instead. Everything has been working fine, except that I’ve been losing the data when the container disappears. This post explains how to persist it in a Docker volume.
This page assumes that you know at least the very basics of Docker already.
Create a Docker volume
The following command creates a docker volume where your data will be stored. You can change the name of the volume, but be sure to also change it in the later commands.
$ docker volume create my_app_data
Then start a Docker container with a command like this. You can change the settings if you want, but this will work fine.
$ docker run --rm \ -p 5432:5432 \ -v my_app_data:/var/lib/postgresql/data \ -e POSTGRES_USER=postgres \ -e POSTGRES_PASSWORD=postgres \ --name my-postgres-container postgres
This table explains what the flags mean:
||removes the container after it shuts down|
||binds the port in the Docker container to a port on your computer|
||tells Docker to persist your volume at the given location on your computer|
||sets an environment variable|
||sets the name of the Docker container|
Finding the Postgres command line
To enter the Docker container, type this:
$ docker container exec -it my-postgres-container bash
Change to the
postgres user (or whatever username you gave it).
# su -u postgres
The next command will list the databases.
You can now connect your application to Postgres with the username and password that you set with the
docker container command. When you stop the container (destroying it, because of the
--rm flag), the data will still be there when you start a new temporary Docker container.
Additional tip: you can inspect the Docker volume with this command:
$ docker volume inspect my_app_data
If anyone is interested in Docker and Kubernetes, this course looks good, and I’m going to watch the videos soon. (Don’t pay more than $10-12 on Udemy. The courses go on sale regularly.)