From 5289f0f19556627aa09f88540dd90ba0482a0c6b Mon Sep 17 00:00:00 2001
From: Celes Renata <celes@frameshift.net>
Date: Mon, 20 Apr 2026 07:47:10 +0000
Subject: [PATCH] fix: use kubectl wait for job completion detection in inttest
 pipeline

The polling loop checked conditions[0].type which missed the Complete
condition when it wasn't at index 0. Switch to kubectl wait
--for=condition=complete which handles condition matching reliably.
---
 infra/inttest/run_pipeline.sh | 23 +++++++++++------------
 1 file changed, 11 insertions(+), 12 deletions(-)

diff --git a/infra/inttest/run_pipeline.sh b/infra/inttest/run_pipeline.sh
index f08a8fd..bcad48f 100755
--- a/infra/inttest/run_pipeline.sh
+++ b/infra/inttest/run_pipeline.sh
@@ -428,24 +428,23 @@ envsubst < "$REPO_ROOT/infra/inttest/runner.yaml" \
   | kubectl apply -n "$NAMESPACE" -f -
 
 log "Waiting for test runner to complete (timeout: 600s) ..."
-# Wait for either complete or failed — whichever comes first
+# Use kubectl wait which handles condition matching reliably.
+# Wait for Complete first; if that times out, check for Failed.
 JOB_DONE=false
-for i in $(seq 1 120); do
-  STATUS=$(kubectl get job inttest-runner -n "$NAMESPACE" -o jsonpath='{.status.conditions[0].type}' 2>/dev/null || true)
-  if [ "$STATUS" = "Complete" ]; then
-    log "Test runner completed successfully"
-    stage_end "integration_tests" "ok"
-    JOB_DONE=true
-    break
-  elif [ "$STATUS" = "Failed" ]; then
+if kubectl wait --for=condition=complete job/inttest-runner -n "$NAMESPACE" --timeout=600s 2>/dev/null; then
+  log "Test runner completed successfully"
+  stage_end "integration_tests" "ok"
+  JOB_DONE=true
+else
+  # Check if it failed (non-zero exit) vs timed out
+  FAILED=$(kubectl get job inttest-runner -n "$NAMESPACE" -o jsonpath='{.status.conditions[?(@.type=="Failed")].status}' 2>/dev/null || true)
+  if [ "$FAILED" = "True" ]; then
     log "Test runner job reported failure (tests failed)"
     stage_fail "integration_tests"
     PIPELINE_EXIT_CODE=1
     JOB_DONE=true
-    break
   fi
-  sleep 5
-done
+fi
 if [ "$JOB_DONE" = false ]; then
   log "Test runner timed out after 600s"
   debug_pod_failure "inttest-runner" "app=inttest-runner"