Merge pull request #121007 from borg-land/npd-host-exec-rewrite

Configure COS to use NPD in daemonset mode and align kubeup NPD manifests with the manifests in the NPD repo

build/dependencies.yaml

@@ -97,6 +97,8 @@ dependencies:
       match: registry.k8s.io/node-problem-detector/node-problem-detector
     - path: cluster/addons/node-problem-detector/npd.yaml
       match: registry.k8s.io/node-problem-detector/node-problem-detector
+    - path: cluster/addons/node-problem-detector/npd.yaml
+      match: app.kubernetes.io/version
     # TODO(dims): Ensure newer versions get uploaded to
     # - https://console.cloud.google.com/storage/browser/gke-release/winnode/node-problem-detector
     # - https://gcsweb.k8s.io/gcs/kubernetes-release/node-problem-detector/

cluster/addons/node-problem-detector/npd.yaml

@@ -26,24 +26,22 @@ subjects:
 apiVersion: apps/v1
 kind: DaemonSet
 metadata:
-  name: npd-v0.8.9
+  name: node-problem-detector
   namespace: kube-system
   labels:
-    k8s-app: node-problem-detector
+    app.kubernetes.io/name: node-problem-detector
-    version: v0.8.9
+    app.kubernetes.io/version: v0.8.13
-    kubernetes.io/cluster-service: "true"
     addonmanager.kubernetes.io/mode: Reconcile
 spec:
   selector:
     matchLabels:
-      k8s-app: node-problem-detector
+      app.kubernetes.io/name: node-problem-detector
-      version: v0.8.9
+      app.kubernetes.io/version: v0.8.13
   template:
     metadata:
       labels:
-        k8s-app: node-problem-detector
+        app.kubernetes.io/name: node-problem-detector
-        version: v0.8.9
+        app.kubernetes.io/version: v0.8.13
-        kubernetes.io/cluster-service: "true"
     spec:
       containers:
       - name: node-problem-detector
@@ -69,6 +67,9 @@ spec:
         volumeMounts:
         - name: log
           mountPath: /var/log
+        - name: kmsg
+          mountPath: /dev/kmsg
+          readOnly: true
         - name: localtime
           mountPath: /etc/localtime
           readOnly: true
@@ -76,6 +77,9 @@ spec:
       - name: log
         hostPath:
           path: /var/log/
+      - name: kmsg
+        hostPath:
+          path: /dev/kmsg
       - name: localtime
         hostPath:
           path: /etc/localtime
@@ -84,5 +88,7 @@ spec:
       tolerations:
       - operator: "Exists"
         effect: "NoExecute"
+      - operator: "Exists"
+        effect: "NoSchedule"
       - key: "CriticalAddonsOnly"
         operator: "Exists"

cluster/gce/config-default.sh

@@ -287,12 +287,7 @@ export ENABLE_DNS_HORIZONTAL_AUTOSCALER="${KUBE_ENABLE_DNS_HORIZONTAL_AUTOSCALER
 #   none           - Not run node problem detector.
 #   daemonset      - Run node problem detector as daemonset.
 #   standalone     - Run node problem detector as standalone system daemon.
-if [[ "${NODE_OS_DISTRIBUTION}" == "gci" ]]; then
+export ENABLE_NODE_PROBLEM_DETECTOR="${KUBE_ENABLE_NODE_PROBLEM_DETECTOR:-daemonset}"
-  # Enable standalone mode by default for gci.
-  ENABLE_NODE_PROBLEM_DETECTOR="${KUBE_ENABLE_NODE_PROBLEM_DETECTOR:-standalone}"
-else
-  export ENABLE_NODE_PROBLEM_DETECTOR="${KUBE_ENABLE_NODE_PROBLEM_DETECTOR:-daemonset}"
-fi
 NODE_PROBLEM_DETECTOR_VERSION="${NODE_PROBLEM_DETECTOR_VERSION:-}"
 NODE_PROBLEM_DETECTOR_TAR_HASH="${NODE_PROBLEM_DETECTOR_TAR_HASH:-}"
 NODE_PROBLEM_DETECTOR_RELEASE_PATH="${NODE_PROBLEM_DETECTOR_RELEASE_PATH:-}"

cluster/gce/config-test.sh

@@ -329,13 +329,7 @@ export ENABLE_DNS_HORIZONTAL_AUTOSCALER=${KUBE_ENABLE_DNS_HORIZONTAL_AUTOSCALER:
 #   none           - Not run node problem detector.
 #   daemonset      - Run node problem detector as daemonset.
 #   standalone     - Run node problem detector as standalone system daemon.
-if [[ "${NODE_OS_DISTRIBUTION}" = 'gci' ]]; then
+export ENABLE_NODE_PROBLEM_DETECTOR=${KUBE_ENABLE_NODE_PROBLEM_DETECTOR:-daemonset}
-  # Enable standalone mode by default for gci.
-  ENABLE_NODE_PROBLEM_DETECTOR=${KUBE_ENABLE_NODE_PROBLEM_DETECTOR:-standalone}
-else
-  ENABLE_NODE_PROBLEM_DETECTOR=${KUBE_ENABLE_NODE_PROBLEM_DETECTOR:-daemonset}
-fi
-export ENABLE_NODE_PROBLEM_DETECTOR
 NODE_PROBLEM_DETECTOR_VERSION=${NODE_PROBLEM_DETECTOR_VERSION:-}
 NODE_PROBLEM_DETECTOR_TAR_HASH=${NODE_PROBLEM_DETECTOR_TAR_HASH:-}
 NODE_PROBLEM_DETECTOR_RELEASE_PATH=${NODE_PROBLEM_DETECTOR_RELEASE_PATH:-}

test/e2e/node/node_problem_detector.go

@@ -380,7 +380,7 @@ func getNpdPodStat(ctx context.Context, f *framework.Framework, nodeName string)
 
 	hasNpdPod := false
 	for _, pod := range summary.Pods {
-		if !strings.HasPrefix(pod.PodRef.Name, "npd") {
+		if !strings.HasPrefix(pod.PodRef.Name, "node-problem-detector") {
 			continue
 		}
 		cpuUsage = float64(*pod.CPU.UsageNanoCores) * 1e-9