Add Helm v4 E2E tests.

Signed-off-by: Josh Baird <jbaird@galileo.io>

Author: joshuabaird

Makefile

@@ -97,6 +97,14 @@ cleanup-test-e2e:
 .PHONY: test-e2e
 test-e2e: setup-test-e2e fluentd-e2e ## Run the e2e tests against a Kind k8s instance that is spun up.
 
+.PHONY: helm-v4-e2e
+helm-v4-e2e: setup-test-e2e ## Run Helm v4 chart e2e tests against a Kind k8s instance.
+	@export KIND_CLUSTER=$(KIND_CLUSTER); export PATH="$$PWD/bin:$$PATH"; tests/scripts/helm_v4_e2e.sh
+
+.PHONY: helm-upgrade-e2e
+helm-upgrade-e2e: setup-test-e2e ## Run Helm v4 upgrade e2e tests (v3 → v4) against a Kind k8s instance.
+	@export KIND_CLUSTER=$(KIND_CLUSTER); export PATH="$$PWD/bin:$$PATH"; tests/scripts/helm_v4_upgrade_e2e.sh
+
 %-e2e: ginkgo
 	@export KIND_CLUSTER=$(KIND_CLUSTER); export PATH="$$PWD/bin:$$PATH"; tests/scripts/$*_e2e.sh
 

charts/fluent-operator/MIGRATION-v4.md

@@ -263,8 +263,13 @@ When upgrading, manually apply CRD updates before upgrading the chart:
 helm repo update
 
 # Extract and apply CRD updates
+# Note: --server-side --force-conflicts is required:
+# - --server-side: the v4 CRDs exceed the 262144-byte annotation limit of
+#   client-side apply.
+# - --force-conflicts: Helm owns the existing CRD fields; kubectl must take
+#   over management to apply the v4 schema changes.
 helm pull fluent/fluent-operator --version 4.0.0 --untar
-kubectl apply -f fluent-operator/crds/
+kubectl apply --server-side --force-conflicts -f fluent-operator/crds/
 
 # Then upgrade the chart
 helm upgrade fluent-operator fluent/fluent-operator --version 4.0.0
@@ -341,15 +346,30 @@ helm upgrade fluent-operator fluent/fluent-operator --version 4.0.0
 
 ### Migrating to Helm-Managed CRDs
 
-If you want to switch to full Helm management of CRDs:
+If you want to switch to full Helm lifecycle management of CRDs after upgrading to v4.0:
 
 ```bash
-# Step 1: Install the new CRDs chart (existing CRDs will be adopted)
+# Step 1: Apply v4 CRDs manually (see Upgrading Standard Installation above)
+kubectl apply --server-side --force-conflicts -f fluent-operator/crds/
+
+# Step 2: Add Helm ownership metadata to all existing fluent.io CRDs.
+# Helm cannot adopt pre-existing cluster resources without this metadata.
+NAMESPACE=fluent  # set to your release namespace
+kubectl get crds -o name | grep '\.fluent\.io' | while read crd; do
+  kubectl label "$crd" app.kubernetes.io/managed-by=Helm --overwrite
+  kubectl annotate "$crd" \
+    meta.helm.sh/release-name=fluent-operator-crds \
+    meta.helm.sh/release-namespace="${NAMESPACE}" --overwrite
+done
+
+# Step 3: Install the CRDs chart — it will adopt the annotated CRDs
 helm install fluent-operator-crds fluent/fluent-operator-crds \
+  --namespace "${NAMESPACE}" \
   --set additionalAnnotations."helm\.sh/resource-policy"=keep
 
-# Step 2: Upgrade operator to v4.0 with --skip-crds
-helm upgrade fluent-operator fluent/fluent-operator --version 4.0.0 --skip-crds
+# Step 4: Upgrade operator to v4.0 with --skip-crds
+helm upgrade fluent-operator fluent/fluent-operator --version 4.0.0 \
+  --namespace "${NAMESPACE}" --skip-crds
 ```
 
 ### Fresh v4.0 Installation

charts/fluent-operator/_docs.md

@@ -23,12 +23,12 @@ When upgrading, manually apply CRD updates before upgrading the chart. You can o
 ```bash
 # Option 1: Extract from the chart
 helm pull fluent/fluent-operator --untar
-kubectl apply -f fluent-operator/crds/
+kubectl apply --server-side --force-conflicts -f fluent-operator/crds/
 
 # Option 2: Clone the repository
 git clone https://github.com/fluent/fluent-operator.git
 cd fluent-operator
-kubectl apply -f charts/fluent-operator/crds/
+kubectl apply --server-side --force-conflicts -f charts/fluent-operator/crds/
 
 # Then upgrade the chart
 helm upgrade fluent-operator fluent/fluent-operator

tests/fixtures/fluentbit-cr.yaml

@@ -0,0 +1,17 @@
+apiVersion: fluentbit.fluent.io/v1alpha2
+kind: FluentBit
+metadata:
+  name: fluent-bit
+  namespace: fluent
+spec:
+  image: ghcr.io/fluent/fluent-operator/fluent-bit:v4.2.3
+  positionDB:
+    hostPath:
+      path: /var/lib/fluent-bit/
+  resources:
+    requests:
+      cpu: 10m
+      memory: 25Mi
+    limits:
+      cpu: 500m
+      memory: 200Mi