fix: root-cause Flux Setters silent no-op - namespace mismatch

ImageUpdateAutomation only lists ImagePolicy objects in its own
namespace (image-automation-controller's getPolicies() scopes the
List() to obj.Namespace, confirmed by reading v1.2.4 source) - the
$imagepolicy marker's namespace:name is only used to match against
that pre-filtered list, not to broaden the search. Our
ImageUpdateAutomation lived in flux-system while its ImagePolicy
lived in hello-app, so the policy was invisible and Setters always
found zero markers to update ("repository up-to-date" forever),
regardless of correct marker syntax/RBAC/policy resolution.

Moved ImageUpdateAutomation into the hello-app namespace (alongside
its ImagePolicy), keeping a cross-namespace sourceRef back to the
flux-system GitRepository. Also fixed the commit messageTemplate,
which used the removed .Updated field (v1.2.4 requires .Changed).

Verified live: Flux pushed commit 0273f15 updating deployment.yaml's
tag on its own, and the cluster rolled out that image without any CI
involvement. Removed the update-deployment-tag CI workaround job
accordingly - it's redundant now and would otherwise race with
Flux's own commits.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GJNvvV3RrvX6TRZGAKEUeY
This commit is contained in:
k3s kubernetes 2026-09-04 12:05:43 +01:00
parent 0273f15f06
commit 10846d1082
2 changed files with 17 additions and 34 deletions

View file

@ -15,8 +15,6 @@ jobs:
container:
image: gcr.io/kaniko-project/executor:debug
options: --entrypoint ""
outputs:
tag: ${{ steps.build.outputs.tag }}
steps:
# actions/checkout@v4 needs a Node.js runtime, which this minimal
# kaniko image doesn't have (no package manager to add it either) -
@ -45,40 +43,10 @@ jobs:
sed -i "s/__GIT_SHA__/$SHORT_SHA/; s/__BUILD_TIME__/$BUILD_TIME/" apps/hello-app/src/index.html
- name: Build and push
id: build
run: |
TAG="main-${GITHUB_SHA::7}-$(date +%s)"
echo "tag=$TAG" >> "$GITHUB_OUTPUT"
/kaniko/executor \
--context="${{ github.workspace }}/apps/hello-app/src" \
--dockerfile="${{ github.workspace }}/apps/hello-app/src/Dockerfile" \
--destination="git.boglabob.com/${{ vars.FORGEJO_ORG }}/hello-app:$TAG" \
--destination="git.boglabob.com/${{ vars.FORGEJO_ORG }}/hello-app:latest"
# Flux's ImageUpdateAutomation (image-automation.yaml) should be doing this
# commit-back step instead - its $imagepolicy Setters marker in
# deployment.yaml is correctly formatted and verified against Flux's own
# docs, ImagePolicy correctly resolves each new tag, and every other part
# of the chain works, but it never actually commits (always reports
# "repository up-to-date" with nothing to change) for reasons that
# resisted a lengthy investigation - even a from-scratch Flux bootstrap
# against a fresh controller instance didn't change the behavior.
# ImageRepository/ImagePolicy are kept for visibility into what tag is
# available; this job does the actual commit as a working substitute.
update-deployment-tag:
needs: build-and-push
runs-on: docker
container:
image: alpine/git
steps:
- name: Commit new image tag to deployment.yaml
run: |
git config --global user.email "flux@boglabob.com"
git config --global user.name "fluxcdbot"
git clone "https://${{ vars.FORGEJO_USER }}:${{ secrets.FORGEJO_TOKEN }}@git.boglabob.com/${{ github.repository }}.git" repo
cd repo
TAG="${{ needs.build-and-push.outputs.tag }}"
sed -i "s|\(image: git.boglabob.com/${{ vars.FORGEJO_ORG }}/hello-app:\)[^ ]*|\1$TAG|" apps/hello-app/deployment.yaml
git add apps/hello-app/deployment.yaml
git diff --cached --quiet || git commit -m "chore(hello-app): auto-update image to $TAG"
git push