ftrack-service-example

This is an example of how ftrack actions or event listeners can be built as a python module when they should run as a service. Multiple actions can be built into the same module if they need to share logic or are part of the same solution.

The python module can be packaged as a docker image to simplify execution and distribution. The module has a built in health check which is used to test that the module stay in a healthy state.

Note

General knowledge on python, how ftrack actions are built and how to build and run docker containers is expected. To use the commands below the current shell is expected to have the following environment variables exported FTRACK_SERVER, FTRACK_API_USER, FTRACK_API_KEY and have a python virtual environment activated.

This module does not handle scaling out actions. It would however be possible add a message to a queue instead of doing work async in the same process. Separate workers could be setup to consume those messages in parallel.

Examples

Below is a list of useful commands that can be used to work with this module.

Install for development

pip install -e .

Run the module with python

python -m ftrack_service_example

Run the health check. It will exit with code 0 if the service is running and hang forever if not

python -m ftrack_service_example --healthcheck

Build as a docker image

docker build -t ftrack-service-example:latest --target final .

Test the image by running it as a container in the foreground

docker run --name ftrack-service-example -it --rm --env FTRACK_SERVER --env FTRACK_API_USER --env FTRACK_API_KEY ftrack-service-example:latest

See that the container is working and is healthy

docker ps

Running the container using docker will not restart it automatically if it enters an unhealthy state. To have the container automatically restarted when it becomes unhealthy a container orchestrator must be used, such docker swarm or kubernetes.

To run the container as a docker service (swarm) and allow it to heal after failure run the following

docker swarm init
docker service create --name ftrack-service-example --env FTRACK_SERVER --env FTRACK_API_USER --env FTRACK_API_KEY ftrack-service-example:latest

Check that the service is running

docker service ls

Look at the logs of the service

docker service logs ftrack-service-example