Deployed vs HEAD
Each workload in the clusters, with the image tag that runs and the newest commit of its repo. It tells you if a deploy is behind.
How to read it
- The badge at the top: green "all at HEAD", or yellow "N not at HEAD".
- Workload: namespace/name, and under it the cluster and the repo.
- Running: the image tag. HEAD: the first 7 characters of the newest commit.
| State | It means |
|---|---|
| ✓ current | The HEAD commit starts with the image tag |
| N behind | The image is N commits behind HEAD. Some of them can be commits that do not change the image. |
| ≠ differs | The image is not at HEAD, and the count is not known (the tag is not a commit id) |
| ? no repo | fleet found no repo for this image |
Problems come first.
Turn it on
This panel is part of Projects and shows with it. Set dash.clusters (a list of
{label, context}), dash.profile.registry (only images from this registry count) and
dash.profile.image_repo (image name to repo). Only admins see it.
When it looks wrong
| You see | It means | Do this |
|---|---|---|
| Green "all at HEAD" and an empty table | No clusters are set, or no image matches registry |
Set dash.clusters and dash.profile.registry |
| "⚠ kubectl: …" and no rows | One cluster failed. Then the panel drops every row. | Fix that context, or remove it from clusters |
| "? no repo" | The image has no entry in image_repo |
Add it |
| "N behind" after a deploy | The new image is not out, or the commits do not change the image | Check the deploy. See a release. |
Code: fleet/internal/dash/projects.go.