feat(app): add ignore-healthcheck annotation

Signed-off-by: cef <[email protected]>

common/common.go

@@ -185,6 +185,10 @@ const (
 	// AnnotationCompareOptions is a comma-separated list of options for comparison
 	AnnotationCompareOptions = "argocd.argoproj.io/compare-options"
 
+	// AnnotationIgnoreHealthCheck when set on an Application's immediate child indicates that its health check
+	// can be disregarded.
+	AnnotationIgnoreHealthCheck = "argocd.argoproj.io/ignore-healthcheck"
+
 	// AnnotationKeyManagedBy is annotation name which indicates that k8s resource is managed by an application.
 	AnnotationKeyManagedBy = "managed-by"
 	// AnnotationValueManagedByArgoCD is a 'managed-by' annotation value for resources managed by Argo CD

controller/health.go

@@ -10,6 +10,7 @@ import (
 	log "github.com/sirupsen/logrus"
 	"k8s.io/apimachinery/pkg/runtime/schema"
 
+	"github.com/argoproj/argo-cd/v2/common"
 	"github.com/argoproj/argo-cd/v2/pkg/apis/application"
 	appv1 "github.com/argoproj/argo-cd/v2/pkg/apis/application/v1alpha1"
 	"github.com/argoproj/argo-cd/v2/util/lua"
@@ -24,6 +25,9 @@ func setApplicationHealth(resources []managedResource, statuses []appv1.Resource
 		if res.Target != nil && hookutil.Skip(res.Target) {
 			continue
 		}
+		if res.Target != nil && res.Target.GetAnnotations() != nil && res.Target.GetAnnotations()[common.AnnotationIgnoreHealthCheck] == "true" {
+			continue
+		}
 
 		if res.Live != nil && (hookutil.IsHook(res.Live) || ignore.Ignore(res.Live)) {
 			continue

controller/health_test.go

@@ -65,6 +65,15 @@ func TestSetApplicationHealth(t *testing.T) {
 	healthStatus, err = setApplicationHealth(resources, resourceStatuses, nil, app, true)
 	require.NoError(t, err)
 	assert.Equal(t, health.HealthStatusHealthy, healthStatus.Status)
+
+	// now we set the `argocd.argoproj.io/ignore-healthcheck: "true"` annotation on the job's target.
+	// The app is considered healthy
+	failedJob.SetAnnotations(nil)
+	failedJobIgnoreHealthcheck := resourceFromFile("./testdata/job-failed-ignore-healthcheck.yaml")
+	resources[1].Target = &failedJobIgnoreHealthcheck
+	healthStatus, err = setApplicationHealth(resources, resourceStatuses, nil, app, true)
+	require.NoError(t, err)
+	assert.Equal(t, health.HealthStatusHealthy, healthStatus.Status)
 }
 
 func TestSetApplicationHealth_ResourceHealthNotPersisted(t *testing.T) {

controller/testdata/job-failed-ignore-healthcheck.yaml

@@ -0,0 +1,36 @@
+apiVersion: batch/v1
+kind: Job
+metadata:
+  annotations:
+    argocd.argoproj.io/ignore-healthcheck: "true"
+  labels:
+    job-name: fail
+  name: fail
+  namespace: argoci-workflows
+  selfLink: /apis/batch/v1/namespaces/argoci-workflows/jobs/fail
+spec:
+  backoffLimit: 0
+  completions: 1
+  parallelism: 1
+  template:
+    metadata:
+      creationTimestamp: null
+      labels:
+        job-name: fail
+    spec:
+      containers:
+      - command:
+        - sh
+        - -c
+        - exit 1
+        image: alpine:latest
+        imagePullPolicy: Always
+        name: fail
+        resources: {}
+        terminationMessagePath: /dev/termination-log
+        terminationMessagePolicy: File
+      dnsPolicy: ClusterFirst
+      restartPolicy: Never
+      schedulerName: default-scheduler
+      securityContext: {}
+      terminationGracePeriodSeconds: 30

docs/operator-manual/health.md

@@ -229,3 +229,16 @@ App (healthy)
 └── CustomResource (healthy) <- This resource's health check needs to be fixed to mark the App as unhealthy
     └── CustomChildResource (unhealthy)
 ```
+## Ignoring Child Resource Health Check in Applications
+
+To ignore the health check of an immediate child resource within an Application, set the annotation `argocd.argoproj.io/ignore-healthcheck` to `true`. For example:
+
+```yaml
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  annotations:
+    argocd.argoproj.io/ignore-healthcheck: "true"
+```
+
+By doing this, the health status of the Deployment will not affect the health of its parent Application.