Running TPA in a Docker container v23

If you're using a system for which there are no TPA packages available, and it's difficult to run TPA after installing from source (for example, because it's not easy to obtain a working Python 3.9+ interpreter), your last resort might be to build a Docker image and run TPA inside a Docker container.

You don't need to run TPA in a Docker container to deploy to Docker containers. It's always preferable to run TPA directly if you can, even on MacOS X.

Quick start

Make sure you have Docker installed and working on your system.

To clone the tpaexec source repository from Github and build a new Docker image named tpa/tpaexec, run the following commands:

$ git clone ssh://
$ cd tpa
$ docker build -t tpa/tpaexec .

Double-check the created image:

$ docker image ls tpa/tpaexec
tpa/tpaexec   latest    e145cf8276fb   8 seconds ago   1.73GB
$ docker run --platform=linux/amd64 --rm tpa/tpaexec tpaexec info
# TPAexec v20.11-59-g85a62fe3 (branch: master)
PYTHON=/opt/EDB/TPA/tpa-venv/bin/python3 (v3.7.3, venv)
ANSIBLE=/opt/EDB/TPA/tpa-venv/bin/ansible (v2.8.15)

Create a TPA container and make your cluster configuration directories available inside the container:

$ docker run --platform=linux/amd64 --rm -v ~/clusters:/clusters \
    -it tpa/tpaexec:latest

You can now run commands like tpaexec provision /clusters/speedy at the container prompt. (When you exit the shell, the container is removed.)

If you want to provision Docker containers using TPA, you must also allow the container to access the Docker control socket on the host:

$ docker run --platform=linux/amd64 --rm -v ~/clusters:/clusters \
    -v /var/run/docker.sock:/var/run/docker.sock \
    -it tpa/tpaexec:latest

Run docker ps in the container to make sure that your connection to the host Docker daemon is working.

Installing Docker

See the Docker documentation if you need help installing Docker and getting started with it.

On MacOS X, you can install Docker Desktop for Mac and launch Docker from the application menu.