diff --git a/pkg/resourceinterpreter/defaultinterpreter/aggregatestatus.go b/pkg/resourceinterpreter/defaultinterpreter/aggregatestatus.go
index 6a714f67e..4bb5989fe 100644
--- a/pkg/resourceinterpreter/defaultinterpreter/aggregatestatus.go
+++ b/pkg/resourceinterpreter/defaultinterpreter/aggregatestatus.go
@@ -160,6 +160,11 @@ func aggregateJobStatus(object *unstructured.Unstructured, aggregatedStatusItems
 		return nil, err
 	}
 
+	// If a job is finished, we should never update status again.
+	if finished, _ := helper.GetJobFinishedStatus(&job.Status); finished {
+		return object, nil
+	}
+
 	newStatus, err := helper.ParsingJobStatus(job, aggregatedStatusItems)
 	if err != nil {
 		return nil, err
diff --git a/pkg/resourceinterpreter/interpreter.go b/pkg/resourceinterpreter/interpreter.go
index 06b7aefcb..ed5ea2259 100644
--- a/pkg/resourceinterpreter/interpreter.go
+++ b/pkg/resourceinterpreter/interpreter.go
@@ -146,6 +146,11 @@ func (i *customResourceInterpreterImpl) Retain(desired *unstructured.Unstructure
 func (i *customResourceInterpreterImpl) AggregateStatus(object *unstructured.Unstructured, aggregatedStatusItems []workv1alpha2.AggregatedStatusItem) (*unstructured.Unstructured, error) {
 	klog.V(4).Infof("Begin to aggregate status for object: %v %s/%s.", object.GroupVersionKind(), object.GetNamespace(), object.GetName())
 
+	// If status has not been collected, there is no need to aggregate.
+	if len(aggregatedStatusItems) == 0 {
+		return object, nil
+	}
+
 	obj, hookEnabled, err := i.customizedInterpreter.Patch(context.TODO(), &webhook.RequestAttributes{
 		Operation:        configv1alpha1.InterpreterOperationAggregateStatus,
 		Object:           object.DeepCopy(),
diff --git a/pkg/util/helper/job.go b/pkg/util/helper/job.go
index 7ad423c52..37cefbfca 100644
--- a/pkg/util/helper/job.go
+++ b/pkg/util/helper/job.go
@@ -36,7 +36,7 @@ func ParsingJobStatus(obj *batchv1.Job, status []workv1alpha2.AggregatedStatusIt
 		newStatus.Succeeded += temp.Succeeded
 		newStatus.Failed += temp.Failed
 
-		isFinished, finishedStatus := getJobFinishedStatus(temp)
+		isFinished, finishedStatus := GetJobFinishedStatus(temp)
 		if isFinished && finishedStatus == batchv1.JobComplete {
 			successfulJobs++
 		} else if isFinished && finishedStatus == batchv1.JobFailed {
@@ -88,9 +88,9 @@ func ParsingJobStatus(obj *batchv1.Job, status []workv1alpha2.AggregatedStatusIt
 	return newStatus, nil
 }
 
-// getJobFinishedStatus checks whether the given Job has finished execution.
+// GetJobFinishedStatus checks whether the given Job has finished execution.
 // It does not discriminate between successful and failed terminations.
-func getJobFinishedStatus(jobStatus *batchv1.JobStatus) (bool, batchv1.JobConditionType) {
+func GetJobFinishedStatus(jobStatus *batchv1.JobStatus) (bool, batchv1.JobConditionType) {
 	for _, c := range jobStatus.Conditions {
 		if (c.Type == batchv1.JobComplete || c.Type == batchv1.JobFailed) && c.Status == corev1.ConditionTrue {
 			return true, c.Type
diff --git a/test/e2e/resource_test.go b/test/e2e/resource_test.go
index 783a7f364..045b122fa 100644
--- a/test/e2e/resource_test.go
+++ b/test/e2e/resource_test.go
@@ -331,7 +331,6 @@ var _ = ginkgo.Describe("[resource-status collection] resource status collection
 	ginkgo.Context("JobStatus collection testing", func() {
 		var jobNamespace, jobName string
 		var job *batchv1.Job
-		var patch []map[string]interface{}
 
 		ginkgo.BeforeEach(func() {
 			policyNamespace = testNamespace
@@ -352,13 +351,6 @@ var _ = ginkgo.Describe("[resource-status collection] resource status collection
 				},
 			})
 
-			patch = []map[string]interface{}{
-				{
-					"op":    "replace",
-					"path":  "/spec/placement/clusterAffinity/clusterNames",
-					"value": framework.ClusterNames()[0 : len(framework.ClusterNames())-1],
-				},
-			}
 		})
 
 		ginkgo.BeforeEach(func() {
@@ -386,25 +378,6 @@ var _ = ginkgo.Describe("[resource-status collection] resource status collection
 				})
 				gomega.Expect(err).ShouldNot(gomega.HaveOccurred())
 			})
-
-			framework.PatchPropagationPolicy(karmadaClient, policy.Namespace, policyName, patch, types.JSONPatchType)
-
-			ginkgo.By("check if job status has been update with new collection", func() {
-				wantedSucceedPods := int32(len(framework.Clusters()) - 1)
-
-				klog.Infof("Waiting for job(%s/%s) collecting correctly status", jobNamespace, jobName)
-				err := wait.PollImmediate(pollInterval, pollTimeout, func() (done bool, err error) {
-					currentJob, err := kubeClient.BatchV1().Jobs(jobNamespace).Get(context.TODO(), jobName, metav1.GetOptions{})
-					gomega.Expect(err).ShouldNot(gomega.HaveOccurred())
-
-					if currentJob.Status.Succeeded == wantedSucceedPods {
-						return true, nil
-					}
-
-					return false, nil
-				})
-				gomega.Expect(err).ShouldNot(gomega.HaveOccurred())
-			})
 		})
 	})