Fixing a stale endpoint issue that blocks ingress network cleanup

fixes #24400

Signed-off-by: Madhu Venugopal <madhu@docker.com>

daemon/network.go

@@ -127,6 +127,14 @@ func (daemon *Daemon) SetupIngress(create clustertypes.NetworkCreateRequest, nod
 				return
 			}
 
+			// Cleanup any stale endpoints that might be left over during previous iterations
+			epList := n.Endpoints()
+			for _, ep := range epList {
+				if err := ep.Delete(true); err != nil {
+					logrus.Errorf("Failed to delete endpoint %s (%s): %v", ep.Name(), ep.ID(), err)
+				}
+			}
+
 			if err := n.Delete(); err != nil {
 				logrus.Errorf("Failed to delete stale ingress network %s: %v", n.ID(), err)
 				return