From c093283b1b27008078d77ac8bf8ffb2a3814dcd8 Mon Sep 17 00:00:00 2001
From: sashaodessa <140454972+sashaodessa@users.noreply.github.com>
Date: Fri, 21 Nov 2025 22:39:24 +0100
Subject: [PATCH] Replace fixed sleep delays with active polling in prometheus
 service test (#15828)

## **Description:**

**What type of PR is this?**

> Bug fix

**What does this PR do? Why is it needed?**

Replaces fixed `time.Sleep(time.Second)` delays in `TestLifecycle` with
active polling to wait for service readiness/shutdown. This improves
test reliability and reduces execution time by eliminating unnecessary
waits when services start/stop faster than expected.

**Which issues(s) does this PR fix?**

N/A - Minor test improvement

**Other notes for review**

- Uses 50ms polling interval with 3s timeout for both startup and
shutdown checks
- Maintains same test logic while making it more efficient and less
flaky
- No functional changes to the service itself

**Acknowledgements**

- [x] I have read
[CONTRIBUTING.md](https://github.com/prysmaticlabs/prysm/blob/develop/CONTRIBUTING.md).
- [ ] I have included a uniquely named [changelog fragment
file](https://github.com/prysmaticlabs/prysm/blob/develop/CONTRIBUTING.md#maintaining-changelogmd).
- [x] I have added a description to this PR with sufficient context for
reviewers to understand this PR.
---
 changelog/sashaodessa_patch-1.md      |  3 +++
 monitoring/prometheus/service_test.go | 31 +++++++++++++++++++++++----
 2 files changed, 30 insertions(+), 4 deletions(-)
 create mode 100644 changelog/sashaodessa_patch-1.md

diff --git a/changelog/sashaodessa_patch-1.md b/changelog/sashaodessa_patch-1.md
new file mode 100644
index 0000000000..75533ae012
--- /dev/null
+++ b/changelog/sashaodessa_patch-1.md
@@ -0,0 +1,3 @@
+### Ignored
+
+- Replace fixed sleep delays with active polling in prometheus service test to improve test reliability.
diff --git a/monitoring/prometheus/service_test.go b/monitoring/prometheus/service_test.go
index 00f85b4750..da2a37c8f1 100644
--- a/monitoring/prometheus/service_test.go
+++ b/monitoring/prometheus/service_test.go
@@ -26,8 +26,21 @@ func TestLifecycle(t *testing.T) {
 	port := 1000 + rand.Intn(1000)
 	prometheusService := NewService(t.Context(), fmt.Sprintf(":%d", port), nil)
 	prometheusService.Start()
-	// Give service time to start.
-	time.Sleep(time.Second)
+    // Actively wait until the service responds on /metrics (faster and less flaky than a fixed sleep)
+    deadline := time.Now().Add(3 * time.Second)
+    for {
+        if time.Now().After(deadline) {
+            t.Fatalf("metrics endpoint not ready within timeout")
+        }
+        resp, err := http.Get(fmt.Sprintf("http://localhost:%d/metrics", port))
+        if err == nil {
+            _ = resp.Body.Close()
+            if resp.StatusCode == http.StatusOK {
+                break
+            }
+        }
+        time.Sleep(50 * time.Millisecond)
+    }
 
 	// Query the service to ensure it really started.
 	resp, err := http.Get(fmt.Sprintf("http://localhost:%d/metrics", port))
@@ -36,8 +49,18 @@ func TestLifecycle(t *testing.T) {
 
 	err = prometheusService.Stop()
 	require.NoError(t, err)
-	// Give service time to stop.
-	time.Sleep(time.Second)
+    // Actively wait until the service stops responding on /metrics
+    deadline = time.Now().Add(3 * time.Second)
+    for {
+        if time.Now().After(deadline) {
+            t.Fatalf("metrics endpoint still reachable after timeout")
+        }
+        _, err = http.Get(fmt.Sprintf("http://localhost:%d/metrics", port))
+        if err != nil {
+            break
+        }
+        time.Sleep(50 * time.Millisecond)
+    }
 
 	// Query the service to ensure it really stopped.
 	_, err = http.Get(fmt.Sprintf("http://localhost:%d/metrics", port))