fix(api): keep heartbeat storage successful if check restart fails

Log restart failures without rejecting an already stored heartbeat.
Continue processing online status after a restart failure.
Move the empty queue marker check into HeartbeatMonitor.RequiresCheck.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Copilot-Session: f27d8f52-9620-4ee2-a564-ec4870a5f840

Author: AchoArnold

api/pkg/entities/heartbeat_monitor.go

@@ -20,7 +20,7 @@ type HeartbeatMonitor struct {
 
 // RequiresCheck returns true if the heartbeat monitor requires a check
 func (h *HeartbeatMonitor) RequiresCheck() bool {
-	return h.UpdatedAt.Add(2 * time.Hour).Before(time.Now())
+	return h.QueueID == "" || h.UpdatedAt.Add(2*time.Hour).Before(time.Now())
 }
 
 // PhoneIsOffline returns true if the phone is offline

api/pkg/services/heartbeat_service.go

@@ -124,15 +124,15 @@ func (service *HeartbeatService) Store(ctx context.Context, params HeartbeatStor
 		return heartbeat, nil
 	}
 
-	if monitor.QueueID == "" || monitor.RequiresCheck() {
+	if monitor.RequiresCheck() {
 		if err = service.scheduleHeartbeatCheck(ctx, heartbeat.Timestamp, &HeartbeatMonitorParams{
 			Owner:     monitor.Owner,
 			PhoneID:   monitor.PhoneID,
 			UserID:    monitor.UserID,
 			MonitorID: monitor.ID,
 			Source:    params.Source,
 		}); err != nil {
-			return nil, service.tracer.WrapErrorSpan(span, stacktrace.Propagatef(err, "cannot restart heartbeat check for monitor [%s]", monitor.ID))
+			ctxLogger.Error(service.tracer.WrapErrorSpan(span, stacktrace.Propagatef(err, "cannot restart heartbeat check for monitor [%s]", monitor.ID)))
 		}
 	}
 

api/pkg/services/heartbeat_service_test.go

@@ -54,9 +54,13 @@ func (repository *heartbeatServiceMonitorRepository) UpdatePhoneOnline(_ context
 
 type heartbeatServiceQueue struct {
 	phoneNotificationEventQueue
+	err error
 }
 
 func (queue *heartbeatServiceQueue) Enqueue(ctx context.Context, task *PushQueueTask, timeout time.Duration) (string, error) {
+	if queue.err != nil {
+		return "", queue.err
+	}
 	if _, err := queue.phoneNotificationEventQueue.Enqueue(ctx, task, timeout); err != nil {
 		return "", err
 	}
@@ -160,3 +164,37 @@ func TestHeartbeatServiceMonitor_StopsCheckingDeletedMonitor(t *testing.T) {
 	}))
 	assert.Empty(t, queue.events)
 }
+
+func TestHeartbeatServiceStore_ReturnsHeartbeatWhenRestartFails(t *testing.T) {
+	for _, test := range []struct {
+		name   string
+		online bool
+	}{
+		{name: "online phone", online: true},
+		{name: "offline phone"},
+	} {
+		t.Run(test.name, func(t *testing.T) {
+			logger := &phoneNotificationLogger{}
+			tracer := telemetry.NewOtelLogger("test", logger)
+			queue := &heartbeatServiceQueue{err: stacktrace.NewError("queue unavailable")}
+			repository := &heartbeatServiceRepository{}
+			monitor := &entities.HeartbeatMonitor{
+				ID: uuid.New(), PhoneID: uuid.New(), UserID: entities.UserID("user-1"),
+				Owner: "+18005550199", UpdatedAt: time.Now().UTC(), PhoneOnline: test.online,
+			}
+			service := NewHeartbeatService(logger, tracer, repository,
+				&heartbeatServiceMonitorRepository{monitor: monitor},
+				NewEventDispatcher(logger, tracer, nil, queue, PushQueueConfig{}))
+
+			heartbeat, err := service.Store(context.Background(), HeartbeatStoreParams{
+				UserID: monitor.UserID, Owner: monitor.Owner, Source: "https://example.com", Timestamp: time.Now().UTC(),
+			})
+
+			require.NoError(t, err)
+			require.NotNil(t, heartbeat)
+			assert.Same(t, repository.heartbeat, heartbeat)
+			assert.Empty(t, monitor.QueueID)
+			assert.True(t, monitor.PhoneOnline)
+		})
+	}
+}