Run a Dash app with a docker image on AWS ECS.
This is a basic Dash app that pulls two images from http://picsum.photos, displays a graph, and echoes input to an output div.
It uses the very small python:3.7-alpine image which is less than 100MB compared to python:3.7 which is 700MB+.
More detail at this excellent tutorial: https://linuxacademy.com/blog/linux-academy/deploying-a-containerized-flask-application-with-aws-ecs-and-docker/
git clone https:/github.com/russellromney/docker-dash
cd docker-dash
Install Docker, Python 3, and AWS CLI if not already done.
If needed, configure AWS CLI with aws configure.
Build:
docker build -t docker-dash .
Test:
docker run -p 8050:8050 docker-dash
Create a repository in AWS ECR:
aws ecr create-repository --repository-name <my-repo-name>
Get login auth token:
aws ecr get-login --region us-east-1 --no-include-email
Copy and paste the output of the above command back into the terminal. You're authenticated locally for 12 hours.
Tag the image to the ECR repo with:
docker tag docker-dash:latest ACCT_ID.dkr.ecr.us-east-1.amazonaws.com/<my-repo-name>
Push the image to AWS ECR with:
docker push ACCT_ID.dkr.ecr.us-east-1.amazonaws.com/<my-repo-name>
Finally, you're done with the command line!
Follow the instructions in step 4 here: https://linuxacademy.com/blog/linux-academy/deploying-a-containerized-flask-application-with-aws-ecs-and-docker/
While doing this, I find it easier to change the names to match -service, -cluster, etc. to make it consistent to refer back to later.
If you need to make changes later, you need to build, test, login (if token expired), tag, and push to get the image to ECR. Then run the following command in the terminal:
aws ecs update-service --cluster <cluster name> --service <service name> --force-new-deployment
This will force the service to reload with the new code.