I've been able to successfully spin up the necessary infrastructure with terraform (EKS cluster, autoscaling groups, etc.), as well as test that jobs can be dispatched to the cluster correctly. However, currently, the jobs always fail, as I've been dispatching them locally, and they can't access my local deployment to access the database, minio/s3, etc. Once we have a public heroku deployment, I can test it against that.
TODO
[x] Add cluster name output as DJANGO_K8S_CLUSTER_NAME env var in heroku terraform (depends on heroku terraform being added)
[x] Test cluster against prod API run in heroku
[x] Configure triggers for building AMI with packer
[ ] Fix K8s code to easily work with both local and prod cluster deployments
[x] Add README.md to the infrastructure folder that explains how everything interacts
I've been able to successfully spin up the necessary infrastructure with terraform (EKS cluster, autoscaling groups, etc.), as well as test that jobs can be dispatched to the cluster correctly. However, currently, the jobs always fail, as I've been dispatching them locally, and they can't access my local deployment to access the database, minio/s3, etc. Once we have a public heroku deployment, I can test it against that.
TODO
DJANGO_K8S_CLUSTER_NAMEenv var in heroku terraform (depends on heroku terraform being added)README.mdto theinfrastructurefolder that explains how everything interacts