Fix OCP-44493 test event matching and add err variable

- Add missing err variable declaration after removing featureSet skip
- Fix event matching to look for Killing and Started events in proper format
- Update to match actual oc describe pod event output format
- Increase timeout to 5 minutes for probe failure detection

Author: BhargaviGudi

test/extended/node/node_e2e/node.go

@@ -174,14 +174,7 @@ var _ = g.Describe("[sig-node] [Jira:Node/Kubelet] Kubelet, CRI-O, CPU manager",
 	//automates: https://issues.redhat.com/browse/OCPBUGS-44493
 	g.It("[OTP] add configurable terminationGracePeriod to liveness and startup probes [OCP-44493]", ote.Informing(), func() {
 		ctx := context.Background()
-
-		g.By("Check if featureSet is empty in cluster")
-		featureSet, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("featuregate", "cluster", "-o=jsonpath={.spec.featureSet}").Output()
-		o.Expect(err).NotTo(o.HaveOccurred(), "failed to get featuregate")
-		e2e.Logf("featureSet is: %s", featureSet)
-		if featureSet != "" {
-			g.Skip("featureSet is not empty, skipping test")
-		}
+		var err error
 
 		oc.SetupProject()
 		namespace := oc.Namespace()
@@ -218,58 +211,78 @@ var _ = g.Describe("[sig-node] [Jira:Node/Kubelet] Kubelet, CRI-O, CPU manager",
 		}
 
 		// Helper to verify probe termination period
-		verifyProbeTermination := func(podName string, expectedTerminationSec int) error {
-			return wait.PollUntilContextTimeout(ctx, 10*time.Second, 4*time.Minute, true, func(ctx context.Context) (bool, error) {
+		verifyProbeTermination := func(podName, containerName string, expectedTerminationSec int) error {
+			return wait.PollUntilContextTimeout(ctx, 10*time.Second, 5*time.Minute, true, func(ctx context.Context) (bool, error) {
 				podDesc, err := oc.AsAdmin().WithoutNamespace().Run("describe").Args("pod", podName, "-n", namespace).Output()
 				if err != nil {
 					e2e.Logf("Error describing pod: %v", err)
 					return false, nil
 				}
 
-				// Look for probe failure and container start events
-				probeFailLine := ""
-				containerStartLine := ""
+				// Look for probe failure (killing) and container restart events
+				// Event format: "Normal  Killing  <time>  kubelet  Container <name> failed <probe> probe, will be restarted"
+				// Event format: "Normal  Started  <time>  kubelet  Container started"
+				killingLine := ""
+				restartLine := ""
+
+				inEvents := false
 				for _, line := range strings.Split(podDesc, "\n") {
-					if strings.Contains(line, "Container") && strings.Contains(line, "failed") && strings.Contains(line, "probe") && strings.Contains(line, "will be restarted") {
-						probeFailLine = line
+					if strings.Contains(line, "Events:") {
+						inEvents = true
+						continue
+					}
+					if !inEvents {
+						continue
+					}
+
+					// Look for killing event with container name
+					if strings.Contains(line, "Killing") && strings.Contains(line, containerName) &&
+					   strings.Contains(line, "failed") && strings.Contains(line, "probe") &&
+					   strings.Contains(line, "will be restarted") {
+						killingLine = line
 					}
-					if strings.Contains(line, "Started container") {
-						containerStartLine = line
+					// Look for Started event after Killing
+					if killingLine != "" && strings.Contains(line, "Started") && strings.Contains(line, "Container started") {
+						restartLine = line
+						break
 					}
 				}
 
-				if probeFailLine == "" || containerStartLine == "" {
-					e2e.Logf("Waiting for probe failure and container start events")
+				if killingLine == "" || restartLine == "" {
+					e2e.Logf("Waiting for probe failure (killing) and container restart events")
 					return false, nil
 				}
 
-				e2e.Logf("Probe failure event: %s", probeFailLine)
-				e2e.Logf("Container start event: %s", containerStartLine)
+				e2e.Logf("Killing event: %s", killingLine)
+				e2e.Logf("Restart event: %s", restartLine)
 
 				// Extract timestamps (format: "1m30s" or "45s")
-				probeFailFields := strings.Fields(probeFailLine)
-				containerStartFields := strings.Fields(containerStartLine)
-				if len(probeFailFields) < 3 || len(containerStartFields) < 3 {
+				// Event format: "Normal  Killing  2m30s  kubelet  Container..."
+				killingFields := strings.Fields(killingLine)
+				restartFields := strings.Fields(restartLine)
+				if len(killingFields) < 3 || len(restartFields) < 3 {
 					e2e.Logf("Unable to parse event timestamps")
 					return false, nil
 				}
 
-				probeFailTime := probeFailFields[2]
-				containerStartTime := containerStartFields[2]
+				killingTime := killingFields[2]
+				restartTime := restartFields[2]
 
-				probeFailSec, err := parseDurationToSeconds(probeFailTime)
+				killingSec, err := parseDurationToSeconds(killingTime)
 				if err != nil {
-					e2e.Logf("Error parsing probe fail time: %v", err)
+					e2e.Logf("Error parsing killing time: %v", err)
 					return false, nil
 				}
 
-				containerStartSec, err := parseDurationToSeconds(containerStartTime)
+				restartSec, err := parseDurationToSeconds(restartTime)
 				if err != nil {
-					e2e.Logf("Error parsing container start time: %v", err)
+					e2e.Logf("Error parsing restart time: %v", err)
 					return false, nil
 				}
 
-				timeDiff := probeFailSec - containerStartSec
+				// Time difference: killing happened earlier, restart happened later
+				// So we need to calculate how long between them
+				timeDiff := killingSec - restartSec
 				e2e.Logf("Time difference: %d seconds (expected: %d ±10 seconds)", timeDiff, expectedTerminationSec)
 
 				// Allow range: [expectedTerminationSec-3, expectedTerminationSec+10]
@@ -331,7 +344,7 @@ var _ = g.Describe("[sig-node] [Jira:Node/Kubelet] Kubelet, CRI-O, CPU manager",
 		o.Expect(err).NotTo(o.HaveOccurred(), "failed to create liveness probe pod")
 		g.DeferCleanup(oc.KubeClient().CoreV1().Pods(namespace).Delete, ctx, "liveness-probe", metav1.DeleteOptions{})
 
-		err = verifyProbeTermination("liveness-probe", 10)
+		err = verifyProbeTermination("liveness-probe", "test", 10)
 		o.Expect(err).NotTo(o.HaveOccurred(), "liveness probe termination grace period not honored")
 
 		g.By("Test startup probe with probe-level terminationGracePeriodSeconds")
@@ -382,7 +395,7 @@ var _ = g.Describe("[sig-node] [Jira:Node/Kubelet] Kubelet, CRI-O, CPU manager",
 		o.Expect(err).NotTo(o.HaveOccurred(), "failed to create startup probe pod")
 		g.DeferCleanup(oc.KubeClient().CoreV1().Pods(namespace).Delete, ctx, "startup-probe", metav1.DeleteOptions{})
 
-		err = verifyProbeTermination("startup-probe", 10)
+		err = verifyProbeTermination("startup-probe", "teststartup", 10)
 		o.Expect(err).NotTo(o.HaveOccurred(), "startup probe termination grace period not honored")
 
 		g.By("Test liveness probe without probe-level terminationGracePeriodSeconds (should use pod-level)")
@@ -433,7 +446,7 @@ var _ = g.Describe("[sig-node] [Jira:Node/Kubelet] Kubelet, CRI-O, CPU manager",
 		o.Expect(err).NotTo(o.HaveOccurred(), "failed to create liveness probe pod without probe termination")
 		g.DeferCleanup(oc.KubeClient().CoreV1().Pods(namespace).Delete, ctx, "liveness-probe-no-term", metav1.DeleteOptions{})
 
-		err = verifyProbeTermination("liveness-probe-no-term", 60)
+		err = verifyProbeTermination("liveness-probe-no-term", "test", 60)
 		o.Expect(err).NotTo(o.HaveOccurred(), "liveness probe should use pod-level termination grace period when probe-level not set")
 	})
 })