Create helm tests for worker/server charts

[jamiezieziula]

as defined here

[parkedwards]

oh @jamiezieziula - we can kubectl logs by deployment name, eg.

kubectl logs deployment/shoveler | grep something-interesting

[ialejandro]

@jamiezieziula can help with this example? helm lint-test, the main config ct.yaml put on root dir, and you can define CI values per chart (example: ci-values.yaml)

More info.: https://github.com/helm/chart-testing

[mitchnielsen]

Thanks for the pointer @ialejandro 🤝 We have since implemented chart-testing to some degree (see https://github.com/PrefectHQ/prefect-helm/blob/db-secret-config/.github/linters/server-ct.yaml, for example). Looks like it's been helpful for validating that it can install the chart properly.

I see it can lint as well, but on that note: as part of #365, I tried helm-unittest and have had a great experience with it so far. It's pretty easy to make a set of tests and confirm the output is what we expect. This can also serve as example configuration values for end users with varying use cases.

If the feedback in that PR is positive, then once it's merged I'll follow up and further implement tests for the worker and server charts.

[mitchnielsen]

We have tests now for each chart under tests/*_test.yaml. We can expand on these over time.