ci(regression): extend daily endurance to thirty minutes

.github/workflows/daily-endurance.yml

@@ -15,9 +15,9 @@ concurrency:
 
 jobs:
   endurance:
-    name: Daily endurance (10m)
+    name: Daily endurance (30m)
     runs-on: ubuntu-latest
-    timeout-minutes: 25
+    timeout-minutes: 45
     steps:
       - uses: actions/checkout@v4
         with:
@@ -77,7 +77,7 @@ jobs:
           with open(os.environ['GITHUB_STEP_SUMMARY'], 'a') as out:
               out.write(f"Endurance: {'PASS' if gate.get('passed') is True else 'FAIL / incomplete'}\n\n")
               out.write(f"Core: `{os.environ['GITHUB_SHA']}`; plugins: `{os.environ['PLUGIN_REVISION']}`.\n\n")
-              out.write('540s offered load + drain + 60s cooldown. Real Core/Runtime/plugins/Box; mock generation and embedding.\n\n')
+              out.write('1740s offered load + drain + 60s cooldown. Real Core/Runtime/plugins/Box; mock generation and embedding.\n\n')
               out.write('Observational run, not a required PR check. Budgets are initial QA limits, not product SLOs. Hosted runners are not a controlled performance baseline.\n\n')
               for phase in suite.get('phases', []):
                   out.write(f"Successful throughput: {phase['achieved_rps']:.3f} RPS.\n\n")
@@ -91,7 +91,7 @@ jobs:
                           out.write(f"| {target} | {metric} | {values['growth']} | {values['allowance']} |\n")
               errors = list(suite.get('errors', [])) + list(gate.get('resource_errors', []))
               if gate.get('timed_out'):
-                  errors.append('Suite exceeded its 900-second execution budget.')
+                  errors.append('Suite exceeded its 2100-second execution budget.')
               if gate.get('infrastructure_error'):
                   errors.append(gate['infrastructure_error'])
               if not gate or not suite:

tools/README.md

@@ -412,9 +412,9 @@ merge check.
 
 The workflow reuses the disposable smoke stack, locked SDK, pinned official
 plugins and lbctl v0.3.0. Generation and embedding are deterministic mocks.
-It runs the endurance profile below, with a 15-minute execution budget for
+It runs the endurance profile below, with a 35-minute execution budget for
 resource preparation, load, drain, cooldown and API cleanup. Service preparation
-is outside that budget; the whole job has a 25-minute limit. Service processes
+is outside that budget; the whole job has a 45-minute limit. Service processes
 and owned Box containers are stopped afterward, including after a failed suite.
 
 Actions summaries show throughput, per-case P95 and retained resource growth.
@@ -433,7 +433,7 @@ uv run --no-sync python tools/regression_ci.py --suite endurance \
 ```
 
 [`regression_profiles/endurance.json`](regression_profiles/endurance.json) runs
-nine minutes of mixed chat/Box/RAG traffic at 1 request/second across four stable
+29 minutes of mixed chat/Box/RAG traffic at 1 request/second across four stable
 groups, followed by one minute without traffic to check recovery. Configure its
 CLI context, plugin inputs, model endpoints, Core SQLite path and Core/Runtime/Box
 PID files as for smoke. Set `plugins: []` to reuse initialized plugins. Configure

tools/regression_ci.py

@@ -193,7 +193,7 @@ def run(args):
                 str(root / 'gate/suite'),
             ],
             root / 'gate',
-            60 if args.suite == 'smoke' else 900,
+            60 if args.suite == 'smoke' else 2100,
             pids,
             env,
         )

tools/regression_profiles/endurance.json

@@ -444,7 +444,7 @@
   "phases": [
     {
       "name": "sustained",
-      "duration_seconds": 540,
+      "duration_seconds": 1740,
       "rps": 1,
       "workers": 4,
       "cooldown_seconds": 60,

tools/test_regression_ci.py

@@ -32,7 +32,7 @@ class DailySummaryTest(unittest.TestCase):
         for gate, expected in [
             ({'passed': True}, 'Endurance: PASS'),
             ({'passed': False, 'infrastructure_error': 'teardown failed'}, 'teardown failed'),
-            ({'passed': False, 'timed_out': True}, '900-second execution budget'),
+            ({'passed': False, 'timed_out': True}, '2100-second execution budget'),
             (None, 'Missing verdict'),
         ]:
             with self.subTest(gate=gate), tempfile.TemporaryDirectory() as directory: