Divya Dadlani

Removed everything from EC2 and RDS, which was the majority of our cost. Left the S3 buckets as per @vito. Monthly cost should now be around $50-60.

Relevant context from @pivotal-jamie-klassen: Branch with spike to add metrics: https://github.com/concourse/concourse/tree/hack/2874 Wiki page: https://github.com/concourse/concourse/wiki/Adding-a-new-instance-to-the-metrics-site

For now our thoughts are revolving around keeping worker_artifacts without any managed states, and leaving the 'Creating' and 'Created' states in the volume. We can just leave around artifacts that...

Artifact lifecycle flow: - an artifact can exist without an underlying volume. The assumption is that a volume will eventually be created that points to the artifact. - an artifact...

Quick update: So far, our work has been to expand the `worker_artifact` abstraction to include *all* volumes rather than just the ones uploaded by `fly execute`. We ran into a...

After some discussion with the team, we have decided to only implement `artifacts` without any GC lifecycle. Artifacts will have a UUID which will be attached by core logic to...

We are currently working on #3810 which will enable using artifacts.

Changes required for this issue: - Workers currently specify their base resource types and paths at registration, and these are stored in the `base_resource_types` and `worker_base_resource_types` table. This can no...

Steps: - [ ] Allow volumes to be created for artifacts rather than for containers - [ ] Create artifacts at the start of every step and register outputs before...

We went over this one more time and realized that we don't actually need artifacts for this. Instead, we can have the existing `worker_base_resource_types` "own" the volume created for each...