This is Terraform and Python code to help you track AWS Batch Jobs without having to poll Batch's APIs.

It sets up this integration

AWS Batch --(job state change)--> EventBridge --(message)--> SQS

• A Service is a homogenous set of Task Definitions. Think of it as a fleet definition. It's responsible for making sure there are enough Tasks (servers or background workers) running at all times. It includes
  • Which Task Definition to run - name + version (e.g. foo-service:30)
  • How many instances of the Task Definition to run
  • Uptime strategies (minimum/maximum healthiness of the fleet)
• A Task Definition is a definition of how to run a set of docker containers. This includes what volumes to mount, what environment variables to pass into the container, what its CPU/Memory requirements are, port mappings - think everything available as part of the docker run command. Generally we only run one at a time, so we can update services one-type-of-process-at-a-time.
• A Task is an instance of a running Task Definition. Usually this will be a single docker container. We won't generally run Tasks manually - we let a Service run them for us, aiming

From the API examples, in Chrome console:

snippets = $('.syntaxhighlighter.js table')
bits = []; snippets.each((i, snippet) => bits.push(snippet.innerText));
console.log(bits.join('\n'));

After that, I ran ts-node against squel-test.ts until it compiled... and I gave up when I started hititng all the class bits.