From ce96f5abf4ca280cbb9d620feb8865d40f2e2e2a Mon Sep 17 00:00:00 2001
From: Dan Notestein <dan@syncad.com>
Date: Mon, 29 Dec 2025 02:31:58 -0500
Subject: [PATCH] Add TCP port wait to benchmark job template to fix service
 race condition

- Add before_script that waits for app service using nc -z
- Add SERVICE_WAIT_TIMEOUT variable (default: 120s)
- Progress logging every 2 seconds
- Prevents flaky CI failures due to DNS/service startup timing
---
 templates/test_jobs.gitlab-ci.yml | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/templates/test_jobs.gitlab-ci.yml b/templates/test_jobs.gitlab-ci.yml
index 3a24d9c..f2f471e 100644
--- a/templates/test_jobs.gitlab-ci.yml
+++ b/templates/test_jobs.gitlab-ci.yml
@@ -131,8 +131,27 @@ include:
   variables:
     HAF_COMMIT: $HAF_COMMIT # dotenv artifacts can be passed to service as long as they appear in section variables
     FF_NETWORK_PER_BUILD: 1
+    # Maximum time to wait for service to be ready (seconds)
+    SERVICE_WAIT_TIMEOUT: 120
   services:
     - !reference [.haf_app_pattern_tests_template, services]
+  before_script:
+    # Wait for the app service to be ready before running tests
+    # This eliminates race conditions where DNS or the service isn't ready yet
+    - |
+      echo "Waiting for app service on port ${HAF_APP_PORT}..."
+      start_time=$(date +%s)
+      while ! nc -z app ${HAF_APP_PORT} 2>/dev/null; do
+        current_time=$(date +%s)
+        elapsed=$((current_time - start_time))
+        if [ $elapsed -ge ${SERVICE_WAIT_TIMEOUT} ]; then
+          echo "ERROR: Timeout waiting for app:${HAF_APP_PORT} after ${SERVICE_WAIT_TIMEOUT}s"
+          exit 1
+        fi
+        echo "Waiting for app:${HAF_APP_PORT}... (${elapsed}s elapsed)"
+        sleep 2
+      done
+      echo "Service app:${HAF_APP_PORT} is ready"
 
 .tox_test_job:
   extends: .job-defaults
-- 
GitLab