Rate-limit the LoadBalancer rolling update test

It was trying to reconnect to the LoadBalancer as fast as possible to
try to catch any transient problems, but "as fast as possible" ended
up meaning about 12,500 times a second, which is clearly excessive.
Limit it to 100 times a second.

test/e2e/network/loadbalancer.go

@@ -1331,12 +1331,13 @@ func testRollingUpdateLBConnectivityDisruption(ctx context.Context, f *framework
 	var totalRequests uint64 = 0
 	var networkErrors uint64 = 0
 	var httpErrors uint64 = 0
-	done := make(chan struct{})
-	defer close(done)
+
+	pollCtx, cancel := context.WithCancel(ctx)
+	defer cancel()
 	go func() {
 		defer ginkgo.GinkgoRecover()
 
-		wait.Until(func() {
+		_ = wait.PollUntilContextCancel(pollCtx, time.Second/100, true, func(ctx context.Context) (bool, error) {
 			atomic.AddUint64(&totalRequests, 1)
 			client := &http.Client{
 				Transport: utilnet.SetTransportDefaults(&http.Transport{
@@ -1351,26 +1352,27 @@ func testRollingUpdateLBConnectivityDisruption(ctx context.Context, f *framework
 			if err != nil {
 				framework.Logf("Got error testing for reachability of %s: %v", url, err)
 				atomic.AddUint64(&networkErrors, 1)
-				return
+				return false, nil
 			}
 			defer func() { _ = resp.Body.Close() }()
 			if resp.StatusCode != http.StatusOK {
 				framework.Logf("Got bad status code: %d", resp.StatusCode)
 				atomic.AddUint64(&httpErrors, 1)
-				return
+				return false, nil
 			}
 			body, err := io.ReadAll(resp.Body)
 			if err != nil {
 				framework.Logf("Got error reading HTTP body: %v", err)
 				atomic.AddUint64(&httpErrors, 1)
-				return
+				return false, nil
 			}
 			if string(body) != msg {
 				framework.Logf("The response body does not contain expected string %s", string(body))
 				atomic.AddUint64(&httpErrors, 1)
-				return
+				return false, nil
 			}
-		}, time.Duration(0), done)
+			return false, nil
+		})
 	}()
 
 	ginkgo.By("Triggering DaemonSet rolling update several times")