OCPNODE-4529: Migrate test case 44493 - configurable terminationGracePeriodSeconds for probes

test/extended/node/README.md

@@ -20,6 +20,7 @@ This directory contains OpenShift end-to-end tests for node-related features.
 - **image_volume.go** - Tests mounting container images as volumes in pods, including subPath and error handling
 - **node_swap.go** - Tests default kubelet swap settings (failSwapOn and swapBehavior) and rejection of user overrides
 - **zstd_chunked.go** - Tests building and running images with zstd:chunked compression format
+- **node_e2e/probe_termination.go** - Probe-level terminationGracePeriodSeconds (OCP-44493) - Tests configurable termination grace period for liveness and startup probes. Includes 3 test cases: probe-level config for liveness probe, probe-level config for startup probe, and fallback to pod-level config when probe-level is not set [Lifecycle:informing]
 
 ## Directory Structure
 

test/extended/node/node_e2e/probe_termination.go

@@ -0,0 +1,305 @@
+package node
+
+import (
+	"context"
+	"fmt"
+	"strings"
+	"time"
+
+	g "github.com/onsi/ginkgo/v2"
+	o "github.com/onsi/gomega"
+	ote "github.com/openshift-eng/openshift-tests-extension/pkg/ginkgo"
+
+	corev1 "k8s.io/api/core/v1"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/util/intstr"
+	"k8s.io/apimachinery/pkg/util/wait"
+	e2e "k8s.io/kubernetes/test/e2e/framework"
+	"k8s.io/utils/ptr"
+
+	nodeutils "github.com/openshift/origin/test/extended/node"
+	exutil "github.com/openshift/origin/test/extended/util"
+)
+
+var _ = g.Describe("[sig-node] Probe configuration", func() {
+	var (
+		oc = exutil.NewCLIWithoutNamespace("probe-termination")
+	)
+
+	//author: bgudi@redhat.com
+	g.It("[OTP] Liveness probe should respect probe-level terminationGracePeriodSeconds [OCP-44493]", ote.Informing(), func() {
+		ctx := context.Background()
+
+		oc.SetupProject()
+		namespace := oc.Namespace()
+
+		g.By("Create pod with liveness probe having probe-level terminationGracePeriodSeconds=10s")
+		pod := &corev1.Pod{
+			ObjectMeta: metav1.ObjectMeta{
+				Name:      "liveness-probe-level",
+				Namespace: namespace,
+			},
+			Spec: corev1.PodSpec{
+				TerminationGracePeriodSeconds: ptr.To[int64](60),
+				SecurityContext: &corev1.PodSecurityContext{
+					RunAsNonRoot: ptr.To(true),
+					SeccompProfile: &corev1.SeccompProfile{
+						Type: corev1.SeccompProfileTypeRuntimeDefault,
+					},
+				},
+				Containers: []corev1.Container{
+					{
+						Name:  "test",
+						Image: "quay.io/openshifttest/nginx-alpine@sha256:04f316442d48ba60e3ea0b5a67eb89b0b667abf1c198a3d0056ca748736336a0",
+						SecurityContext: &corev1.SecurityContext{
+							AllowPrivilegeEscalation: ptr.To(false),
+							Capabilities: &corev1.Capabilities{
+								Drop: []corev1.Capability{"ALL"},
+							},
+						},
+						Command: []string{"sh", "-c", "sleep 100000000"},
+						Ports: []corev1.ContainerPort{
+							{ContainerPort: 8080},
+						},
+						LivenessProbe: &corev1.Probe{
+							ProbeHandler: corev1.ProbeHandler{
+								HTTPGet: &corev1.HTTPGetAction{
+									Path: "/healthz",
+									Port: intstr.FromInt(8080),
+								},
+							},
+							InitialDelaySeconds:           5,
+							FailureThreshold:              1,
+							PeriodSeconds:                 60,
+							TerminationGracePeriodSeconds: ptr.To[int64](10),
+						},
+					},
+				},
+			},
+		}
+
+		_, err := oc.KubeClient().CoreV1().Pods(namespace).Create(ctx, pod, metav1.CreateOptions{})
+		o.Expect(err).NotTo(o.HaveOccurred(), "failed to create liveness probe pod")
+
+		g.By("Verify probe-level terminationGracePeriodSeconds is honored (10s)")
+		expectedSec := 10
+		minSec := expectedSec - 3
+		maxSec := expectedSec + 10
+		timeDiff, err := verifyProbeTermination(ctx, oc, namespace, "liveness-probe-level", "test", expectedSec)
+		o.Expect(err).NotTo(o.HaveOccurred(), "failed to get probe termination events")
+		o.Expect(timeDiff).To(o.BeNumerically(">=", minSec), fmt.Sprintf("time difference %ds is less than expected minimum %ds", timeDiff, minSec))
+		o.Expect(timeDiff).To(o.BeNumerically("<=", maxSec), fmt.Sprintf("time difference %ds is greater than expected maximum %ds", timeDiff, maxSec))
+	})
+
+	//author: bgudi@redhat.com
+	g.It("[OTP] Startup probe should respect probe-level terminationGracePeriodSeconds [OCP-44493]", ote.Informing(), func() {
+		ctx := context.Background()
+
+		oc.SetupProject()
+		namespace := oc.Namespace()
+
+		g.By("Create pod with startup probe having probe-level terminationGracePeriodSeconds=10s")
+		pod := &corev1.Pod{
+			ObjectMeta: metav1.ObjectMeta{
+				Name:      "startup-probe-level",
+				Namespace: namespace,
+			},
+			Spec: corev1.PodSpec{
+				TerminationGracePeriodSeconds: ptr.To[int64](60),
+				SecurityContext: &corev1.PodSecurityContext{
+					RunAsNonRoot: ptr.To(true),
+					SeccompProfile: &corev1.SeccompProfile{
+						Type: corev1.SeccompProfileTypeRuntimeDefault,
+					},
+				},
+				Containers: []corev1.Container{
+					{
+						Name:  "teststartup",
+						Image: "quay.io/openshifttest/nginx-alpine@sha256:04f316442d48ba60e3ea0b5a67eb89b0b667abf1c198a3d0056ca748736336a0",
+						SecurityContext: &corev1.SecurityContext{
+							AllowPrivilegeEscalation: ptr.To(false),
+							Capabilities: &corev1.Capabilities{
+								Drop: []corev1.Capability{"ALL"},
+							},
+						},
+						Command: []string{"sh", "-c", "sleep 100000000"},
+						Ports: []corev1.ContainerPort{
+							{ContainerPort: 8080},
+						},
+						StartupProbe: &corev1.Probe{
+							ProbeHandler: corev1.ProbeHandler{
+								HTTPGet: &corev1.HTTPGetAction{
+									Path: "/healthz",
+									Port: intstr.FromInt(8080),
+								},
+							},
+							InitialDelaySeconds:           5,
+							FailureThreshold:              1,
+							PeriodSeconds:                 60,
+							TerminationGracePeriodSeconds: ptr.To[int64](10),
+						},
+					},
+				},
+			},
+		}
+
+		_, err := oc.KubeClient().CoreV1().Pods(namespace).Create(ctx, pod, metav1.CreateOptions{})
+		o.Expect(err).NotTo(o.HaveOccurred(), "failed to create startup probe pod")
+
+		g.By("Verify probe-level terminationGracePeriodSeconds is honored (10s)")
+		expectedSec := 10
+		minSec := expectedSec - 3
+		maxSec := expectedSec + 10
+		timeDiff, err := verifyProbeTermination(ctx, oc, namespace, "startup-probe-level", "teststartup", expectedSec)
+		o.Expect(err).NotTo(o.HaveOccurred(), "failed to get probe termination events")
+		o.Expect(timeDiff).To(o.BeNumerically(">=", minSec), fmt.Sprintf("time difference %ds is less than expected minimum %ds", timeDiff, minSec))
+		o.Expect(timeDiff).To(o.BeNumerically("<=", maxSec), fmt.Sprintf("time difference %ds is greater than expected maximum %ds", timeDiff, maxSec))
+	})
+
+	//author: bgudi@redhat.com
+	g.It("[OTP] Liveness probe should fall back to pod-level terminationGracePeriodSeconds when probe-level is not set [OCP-44493]", ote.Informing(), func() {
+		ctx := context.Background()
+
+		oc.SetupProject()
+		namespace := oc.Namespace()
+
+		g.By("Create pod with liveness probe without probe-level terminationGracePeriodSeconds")
+		pod := &corev1.Pod{
+			ObjectMeta: metav1.ObjectMeta{
+				Name:      "liveness-pod-level",
+				Namespace: namespace,
+			},
+			Spec: corev1.PodSpec{
+				TerminationGracePeriodSeconds: ptr.To[int64](60),
+				SecurityContext: &corev1.PodSecurityContext{
+					RunAsNonRoot: ptr.To(true),
+					SeccompProfile: &corev1.SeccompProfile{
+						Type: corev1.SeccompProfileTypeRuntimeDefault,
+					},
+				},
+				Containers: []corev1.Container{
+					{
+						Name:  "test",
+						Image: "quay.io/openshifttest/nginx-alpine@sha256:04f316442d48ba60e3ea0b5a67eb89b0b667abf1c198a3d0056ca748736336a0",
+						SecurityContext: &corev1.SecurityContext{
+							AllowPrivilegeEscalation: ptr.To(false),
+							Capabilities: &corev1.Capabilities{
+								Drop: []corev1.Capability{"ALL"},
+							},
+						},
+						Command: []string{"sh", "-c", "sleep 100000000"},
+						Ports: []corev1.ContainerPort{
+							{ContainerPort: 8080},
+						},
+						LivenessProbe: &corev1.Probe{
+							ProbeHandler: corev1.ProbeHandler{
+								HTTPGet: &corev1.HTTPGetAction{
+									Path: "/healthz",
+									Port: intstr.FromInt(8080),
+								},
+							},
+							InitialDelaySeconds: 5,
+							FailureThreshold:    1,
+							PeriodSeconds:       60,
+							// No TerminationGracePeriodSeconds - should use pod-level (60s)
+						},
+					},
+				},
+			},
+		}
+
+		_, err := oc.KubeClient().CoreV1().Pods(namespace).Create(ctx, pod, metav1.CreateOptions{})
+		o.Expect(err).NotTo(o.HaveOccurred(), "failed to create liveness probe pod without probe-level termination")
+
+		g.By("Verify pod-level terminationGracePeriodSeconds is used (60s)")
+		expectedSec := 60
+		minSec := expectedSec - 3
+		maxSec := expectedSec + 10
+		timeDiff, err := verifyProbeTermination(ctx, oc, namespace, "liveness-pod-level", "test", expectedSec)
+		o.Expect(err).NotTo(o.HaveOccurred(), "failed to get probe termination events")
+		o.Expect(timeDiff).To(o.BeNumerically(">=", minSec), fmt.Sprintf("time difference %ds is less than expected minimum %ds", timeDiff, minSec))
+		o.Expect(timeDiff).To(o.BeNumerically("<=", maxSec), fmt.Sprintf("time difference %ds is greater than expected maximum %ds", timeDiff, maxSec))
+	})
+})
+
+// findLatestEventByReason finds the latest event matching the given reason and message filter
+func findLatestEventByReason(events *corev1.EventList, reason string, msgFilter func(string) bool) *corev1.Event {
+	var latestEvent *corev1.Event
+	for i := range events.Items {
+		event := &events.Items[i]
+		if event.Reason == reason && msgFilter(event.Message) {
+			if latestEvent == nil || event.LastTimestamp.Time.After(latestEvent.LastTimestamp.Time) {
+				latestEvent = event
+			}
+		}
+	}
+	return latestEvent
+}
+
+// findEarliestEventAfter finds the earliest event matching the reason and filter that occurred after the given time
+func findEarliestEventAfter(events *corev1.EventList, reason string, msgFilter func(string) bool, afterTime time.Time) *corev1.Event {
+	var earliestEvent *corev1.Event
+	for i := range events.Items {
+		event := &events.Items[i]
+		if event.Reason == reason && msgFilter(event.Message) && event.FirstTimestamp.Time.After(afterTime) {
+			if earliestEvent == nil || event.FirstTimestamp.Time.Before(earliestEvent.FirstTimestamp.Time) {
+				earliestEvent = event
+			}
+		}
+	}
+	return earliestEvent
+}
+
+// verifyProbeTermination verifies that the probe termination grace period is honored
+// by checking the time difference between probe failure (Killing) and container restart (Started) events
+// Returns the time difference in seconds, or an error if events are not found
+func verifyProbeTermination(ctx context.Context, oc *exutil.CLI, namespace, podName, containerName string, expectedTerminationSec int) (int, error) {
+	var timeDiff int
+	// Timeout needs to account for: pod start (~30s) + probe period (60s) + termination (up to 60s) + restart (~30s) = ~3 minutes minimum
+	// Use 6 minutes to be safe for tests with 60s termination grace period
+	err := wait.PollUntilContextTimeout(ctx, 10*time.Second, 6*time.Minute, true, func(ctx context.Context) (bool, error) {
+		// Get events using the Events API
+		events, err := oc.KubeClient().CoreV1().Events(namespace).List(ctx, metav1.ListOptions{
+			FieldSelector: fmt.Sprintf("involvedObject.name=%s,involvedObject.kind=Pod", podName),
+		})
+		if err != nil {
+			e2e.Logf("Error getting events: %v", err)
+			return false, nil
+		}
+
+		// Find probe failure (Killing) event for the container
+		killingEvent := findLatestEventByReason(events, "Killing", func(msg string) bool {
+			return strings.Contains(msg, containerName) &&
+				strings.Contains(msg, "failed") &&
+				strings.Contains(msg, "probe")
+		})
+
+		if killingEvent == nil {
+			e2e.Logf("Waiting for probe failure (Killing) event")
+			return false, nil
+		}
+
+		// Find container restart (Started) event that occurred after the Killing event
+		startedEvent := findEarliestEventAfter(events, "Started", func(msg string) bool {
+			return strings.Contains(msg, "Container started")
+		}, killingEvent.LastTimestamp.Time)
+
+		if startedEvent == nil {
+			e2e.Logf("Waiting for container restart (Started) event after Killing event")
+			return false, nil
+		}
+
+		e2e.Logf("Killing event: %s at %v", killingEvent.Message, killingEvent.LastTimestamp)
+		e2e.Logf("Started event: %s at %v", startedEvent.Message, startedEvent.FirstTimestamp)
+
+		// Calculate time difference using the helper function
+		timeDiff = int(nodeutils.CalculateEventTimeDiff(killingEvent, startedEvent).Seconds())
+		e2e.Logf("Time difference: %d seconds (expected: %d ±10 seconds)", timeDiff, expectedTerminationSec)
+
+		return true, nil
+	})
+	if err != nil {
+		return 0, err
+	}
+	return timeDiff, nil
+}

test/extended/node/node_utils.go

@@ -764,3 +764,18 @@ func GetFirstReadyWorkerNode(oc *exutil.CLI) string {
 	o.Expect(false).To(o.BeTrue(), "no Ready worker node found among %v", workers)
 	return "" // unreachable; satisfies compiler
 }
+
+// CalculateEventTimeDiff calculates the time difference between two Kubernetes events.
+// It uses LastTimestamp for the start event and FirstTimestamp for the end event.
+// Falls back to FirstTimestamp/LastTimestamp respectively if the primary timestamp is zero.
+func CalculateEventTimeDiff(startEvent, endEvent *corev1.Event) time.Duration {
+	startTime := startEvent.LastTimestamp.Time
+	if startTime.IsZero() {
+		startTime = startEvent.FirstTimestamp.Time
+	}
+	endTime := endEvent.FirstTimestamp.Time
+	if endTime.IsZero() {
+		endTime = endEvent.LastTimestamp.Time
+	}
+	return endTime.Sub(startTime)
+}