Avoid cableEngine cleanup when the gateway pod is restarted

In the current code, it was seen that when the submariner-gateway
pod gets restarted, there is a brief datapath disruption until the
new pod comes online. The cleanup code in submariner-gateway pod is
not required as its handled by the route-agent code when there is
any transition in the Gateway nodes.

Partially fixes: #2498
Signed-off-by: Sridhar Gaddam <sgaddam@redhat.com>

main.go

@@ -248,10 +248,6 @@ func main() {
 
 	<-stopCh
 
-	if err := cableEngine.Cleanup(); err != nil {
-		klog.Errorf("error cleaning up cableEngine resources before removing Gateway")
-	}
-
 	klog.Info("All controllers stopped or exited. Stopping main loop")
 
 	if err := httpServer.Shutdown(context.TODO()); err != nil {