From c56aa14f9ddcae0c90e08823d77bd11e0a20bbb6 Mon Sep 17 00:00:00 2001
From: nexxo <nexxo@clupilot.local>
Date: Sat, 25 Jul 2026 12:15:26 +0200
Subject: [PATCH] fix(engine): don't reset started_at on poll so step deadlines
 accumulate

RunRunner::onPoll no longer resets started_at, so a poll step's own deadline
(WaitForGuestAgent 270s, ConfigureDnsAndTls cert 840s) actually fires instead
of resetting every poll. maxDuration raised above each own-deadline so the
step's fail wins over the generic timeout. Shared core fix (A + B).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
---
 app/Provisioning/RunRunner.php                       |  7 +++++--
 .../Steps/Customer/ConfigureDnsAndTls.php            |  3 ++-
 .../Steps/Customer/WaitForGuestAgent.php             |  3 ++-
 tests/Feature/Provisioning/RunRunnerTest.php         | 12 ++++++++++++
 4 files changed, 21 insertions(+), 4 deletions(-)

diff --git a/app/Provisioning/RunRunner.php b/app/Provisioning/RunRunner.php
index 1886bad..1adb542 100644
--- a/app/Provisioning/RunRunner.php
+++ b/app/Provisioning/RunRunner.php
@@ -147,12 +147,15 @@ class RunRunner
         $this->record($run, $step->key(), 'retry', $result->reason);
     }
 
-    /** Poll: wait and re-run without consuming the retry budget (step owns its deadline). */
+    /**
+     * Poll: wait and re-run without consuming the retry budget. started_at is NOT
+     * reset here, so a step's own deadline (measured from started_at) actually
+     * accumulates across polls instead of resetting every time.
+     */
     private function onPoll(ProvisioningRun $run, ProvisioningStep $step, StepResult $result): void
     {
         $run->status = ProvisioningRun::STATUS_WAITING;
         $run->next_attempt_at = now()->addSeconds($result->afterSeconds);
-        $run->started_at = now();
         $run->save();
         $this->record($run, $step->key(), 'info', $result->reason);
     }
diff --git a/app/Provisioning/Steps/Customer/ConfigureDnsAndTls.php b/app/Provisioning/Steps/Customer/ConfigureDnsAndTls.php
index c29ab02..a63dbf3 100644
--- a/app/Provisioning/Steps/Customer/ConfigureDnsAndTls.php
+++ b/app/Provisioning/Steps/Customer/ConfigureDnsAndTls.php
@@ -21,7 +21,8 @@ class ConfigureDnsAndTls extends CustomerStep
 
     public function maxDuration(): int
     {
-        return 900;
+        // Above the 840s cert own-deadline below, so the step's own fail fires first.
+        return 960;
     }
 
     public function execute(ProvisioningRun $run): StepResult
diff --git a/app/Provisioning/Steps/Customer/WaitForGuestAgent.php b/app/Provisioning/Steps/Customer/WaitForGuestAgent.php
index 0cf1278..c9b01d9 100644
--- a/app/Provisioning/Steps/Customer/WaitForGuestAgent.php
+++ b/app/Provisioning/Steps/Customer/WaitForGuestAgent.php
@@ -17,7 +17,8 @@ class WaitForGuestAgent extends CustomerStep
 
     public function maxDuration(): int
     {
-        return 300;
+        // Above the 270s own-deadline below, so the step's own fail fires first.
+        return 360;
     }
 
     public function execute(ProvisioningRun $run): StepResult
diff --git a/tests/Feature/Provisioning/RunRunnerTest.php b/tests/Feature/Provisioning/RunRunnerTest.php
index 84c1494..f3efaeb 100644
--- a/tests/Feature/Provisioning/RunRunnerTest.php
+++ b/tests/Feature/Provisioning/RunRunnerTest.php
@@ -93,6 +93,18 @@ it('polls without consuming the retry budget', function () {
     expect($run->fresh()->status)->toBe('waiting')->and($run->fresh()->attempt)->toBe(0);
 });
 
+it('keeps started_at stable across polls so step deadlines can accumulate', function () {
+    bindPipeline(['test' => [FakePollStep::class]]);
+    $run = ProvisioningRun::factory()->create([
+        'pipeline' => 'test', 'status' => 'running', 'started_at' => now()->subMinutes(5),
+    ]);
+    $original = $run->started_at->timestamp;
+
+    app(RunRunner::class)->advance($run);
+
+    expect($run->fresh()->started_at->timestamp)->toBe($original);
+});
+
 it('marks a Host subject as error when the run fails', function () {
     bindPipeline(['test' => [FakeFailStep::class]]);
     $host = Host::factory()->create(['status' => 'onboarding']);