UPSTREAM: 100378: Tag Multi-AZ scheduling tests as serial

As per mdame, we can't ensure that the cluster is actually balanced
if other tests are adding or deleting pods in parallel.

test/e2e/scheduling/ubernetes_lite.go

@@ -69,11 +69,11 @@ var _ = SIGDescribe("Multi-AZ Clusters", func() {
 			cleanUp()
 		}
 	})
-	ginkgo.It("should spread the pods of a service across zones", func() {
+	ginkgo.It("should spread the pods of a service across zones [Serial]", func() {
 		SpreadServiceOrFail(f, 5*zoneCount, imageutils.GetPauseImageName())
 	})
 
-	ginkgo.It("should spread the pods of a replication controller across zones", func() {
+	ginkgo.It("should spread the pods of a replication controller across zones [Serial]", func() {
 		SpreadRCOrFail(f, int32(5*zoneCount), framework.ServeHostnameImage, []string{"serve-hostname"})
 	})
 })