You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Problem: We are using GCP Cloud Deploy to deploy the chart as a dependency to a GKE cluster. Cloud Deploy calls skaffold render, which, in our case, uses Helm to render the manifest. The rendered manifest then gets applied, including the test-db-interaction deployment. Although the test runs successfully and is marked as completed/successful, Cloud Deploy still marks the release as failed since the test pod is not marked as ready.
Potential Suggested Solution: Introduce a conditional render around the test deployment.
Why did I create a new issue? This is essentially the same request as described in issue #77 . That issue was marked as resolved because the test passed, and the deployment tool no longer flagged the release as failed. However, the underlying problem of preventing the test deployment from being rendered still persists.
I'm happy to create a PR for the suggested solution myself—just wanted to confirm that this mechanism would be an acceptable approach before proceeding.
The text was updated successfully, but these errors were encountered:
SlowSnowFox
changed the title
allow test hooks to be turned off (again)
allow preventing test hooks from being rendered (again)
Sep 23, 2024
Problem: We are using GCP Cloud Deploy to deploy the chart as a dependency to a GKE cluster. Cloud Deploy calls skaffold render, which, in our case, uses Helm to render the manifest. The rendered manifest then gets applied, including the test-db-interaction deployment. Although the test runs successfully and is marked as completed/successful, Cloud Deploy still marks the release as failed since the test pod is not marked as ready.
Potential Suggested Solution: Introduce a conditional render around the test deployment.
Why did I create a new issue? This is essentially the same request as described in issue #77 . That issue was marked as resolved because the test passed, and the deployment tool no longer flagged the release as failed. However, the underlying problem of preventing the test deployment from being rendered still persists.
I'm happy to create a PR for the suggested solution myself—just wanted to confirm that this mechanism would be an acceptable approach before proceeding.
The text was updated successfully, but these errors were encountered: