diff --git a/pkg/kubelet/prober/worker.go b/pkg/kubelet/prober/worker.go
index 30aa113d38e..cdefc1da2c6 100644
--- a/pkg/kubelet/prober/worker.go
+++ b/pkg/kubelet/prober/worker.go
@@ -240,7 +240,7 @@ func (w *worker) doProbe() (keepGoing bool) {
 		// chance of hitting #21751, where running `docker exec` when a
 		// container is being stopped may lead to corrupted container state.
 		w.onHold = true
-		w.resultRun = 1
+		w.resultRun = 0
 	}
 
 	return true
diff --git a/pkg/kubelet/prober/worker_test.go b/pkg/kubelet/prober/worker_test.go
index a36c0afde60..d7473a35bc6 100644
--- a/pkg/kubelet/prober/worker_test.go
+++ b/pkg/kubelet/prober/worker_test.go
@@ -352,7 +352,7 @@ func TestResultRunOnLivenessCheckFailure(t *testing.T) {
 	expectContinue(t, w, w.doProbe(), msg)
 	expectResult(t, w, results.Success, msg)
 	if w.resultRun != 1 {
-		t.Errorf("Prober resultRun should 1")
+		t.Errorf("Prober resultRun should be 1")
 	}
 
 	m.prober.exec = fakeExecProber{probe.Failure, nil}
@@ -360,7 +360,7 @@ func TestResultRunOnLivenessCheckFailure(t *testing.T) {
 	expectContinue(t, w, w.doProbe(), msg)
 	expectResult(t, w, results.Success, msg)
 	if w.resultRun != 1 {
-		t.Errorf("Prober resultRun should 1")
+		t.Errorf("Prober resultRun should be 1")
 	}
 
 	m.prober.exec = fakeExecProber{probe.Failure, nil}
@@ -372,13 +372,13 @@ func TestResultRunOnLivenessCheckFailure(t *testing.T) {
 	}
 
 	// Exceeding FailureThreshold should cause resultRun to
-	// reset to 1 so that the probe on the restarted pod
+	// reset to 0 so that the probe on the restarted pod
 	// also gets FailureThreshold attempts to succeed.
 	m.prober.exec = fakeExecProber{probe.Failure, nil}
 	msg = "3rd probe failure, result failure"
 	expectContinue(t, w, w.doProbe(), msg)
 	expectResult(t, w, results.Failure, msg)
-	if w.resultRun != 1 {
-		t.Errorf("Prober resultRun should be reset to 1")
+	if w.resultRun != 0 {
+		t.Errorf("Prober resultRun should be reset to 0")
 	}
 }