Currently, the ProductionTransit ECS service is configured to use the default healthcheck path, /. That path only returns static files, and isn't a good indicator about whether or not the backend service is up. This PR updates the ECS service health_check_path to match the one in Panopta, which actually tests tests the backend.
Additionally, encapsulate CloudWatch's log retention period in terraform. It's currently set to 30 days.
Finally, add docker-compose.ci.yml for containerized deployments, and make sure to clean up CI after deployments.
Testing
This change has already been deployed. The service hasn't been cycled out, so I think it's safe to say that the health check is working.
Overview
Currently, the
ProductionTransitECS service is configured to use the default healthcheck path,/. That path only returns static files, and isn't a good indicator about whether or not the backend service is up. This PR updates the ECS servicehealth_check_pathto match the one in Panopta, which actually tests tests the backend.Additionally, encapsulate CloudWatch's log retention period in terraform. It's currently set to 30 days.
Finally, add
docker-compose.ci.ymlfor containerized deployments, and make sure to clean up CI after deployments.Testing
This change has already been deployed. The service hasn't been cycled out, so I think it's safe to say that the health check is working.
Closes #91