Log all pods on a node when cordoning the node (#259)

Author: wanwenli

cmd/node-termination-handler.go

@@ -221,6 +221,10 @@ func drainOrCordonIfNecessary(interruptionEventStore *interruptioneventstore.Sto
 				os.Exit(1)
 			}
 			log.Log().Str("node_name", nodeName).Msg("Node successfully cordoned")
+			err = node.LogPods(nodeName)
+			if err != nil {
+				log.Log().Err(err).Msg("There was a problem while trying to log all pod names on the node")
+			}
 			metrics.NodeActionsInc("cordon", nodeName, err)
 		} else {
 			err := node.CordonAndDrain(nodeName)

pkg/node/node.go

@@ -22,6 +22,7 @@ import (
 
 	"github.com/aws/aws-node-termination-handler/pkg/config"
 	"github.com/aws/aws-node-termination-handler/pkg/uptime"
+	"github.com/rs/zerolog"
 	"github.com/rs/zerolog/log"
 	corev1 "k8s.io/api/core/v1"
 	"k8s.io/apimachinery/pkg/api/errors"
@@ -286,6 +287,21 @@ func (n Node) TaintSpotItn(nodeName string, eventID string) error {
 	return addTaint(k8sNode, n, SpotInterruptionTaint, eventID, corev1.TaintEffectNoSchedule)
 }
 
+// LogPods logs all the pod names on a node
+func (n Node) LogPods(nodeName string) error {
+	podList, err := n.fetchAllPods(nodeName)
+	if err != nil {
+		return fmt.Errorf("Unable to fetch all pods from API: %w", err)
+	}
+	podNamesArr := zerolog.Arr()
+	for _, pod := range podList.Items {
+		podNamesArr = podNamesArr.Str(pod.Name)
+	}
+	log.Log().Array("pod_names", podNamesArr).Str("node_name", nodeName).Msg("Pods on node")
+
+	return nil
+}
+
 // TaintScheduledMaintenance adds the scheduled maintenance taint onto a node
 func (n Node) TaintScheduledMaintenance(nodeName string, eventID string) error {
 	if !n.nthConfig.TaintNode {
@@ -399,6 +415,12 @@ func (n Node) fetchKubernetesNode(nodeName string) (*corev1.Node, error) {
 	return n.drainHelper.Client.CoreV1().Nodes().Get(nodeName, metav1.GetOptions{})
 }
 
+func (n Node) fetchAllPods(nodeName string) (*corev1.PodList, error) {
+	return n.drainHelper.Client.CoreV1().Pods("").List(metav1.ListOptions{
+		FieldSelector: "spec.nodeName=" + nodeName,
+	})
+}
+
 func getDrainHelper(nthConfig config.Config) (*drain.Helper, error) {
 	drainHelper := &drain.Helper{
 		Client:              &kubernetes.Clientset{},

pkg/node/node_test.go

@@ -240,6 +240,30 @@ func TestMarkForUncordonAfterRebootAddActionLabelFailure(t *testing.T) {
 	h.Assert(t, err != nil, "Failed to return error on MarkForUncordonAfterReboot failing to add action Label")
 }
 
+func TestLogPods(t *testing.T) {
+	resetFlagsForTest()
+
+	client := fake.NewSimpleClientset(
+		&v1.Pod{
+			ObjectMeta: metav1.ObjectMeta{
+				Name: "myPod",
+				Labels: map[string]string{
+					"spec.nodeName": nodeName,
+				},
+			},
+		},
+		&v1.Node{
+			ObjectMeta: metav1.ObjectMeta{
+				Name: nodeName,
+			},
+		},
+	)
+
+	tNode := getNode(t, getDrainHelper(client))
+	err := tNode.LogPods(nodeName)
+	h.Ok(t, err)
+}
+
 func TestIsLableledWithActionFailure(t *testing.T) {
 	resetFlagsForTest()