tests/hermes_cli/test_gateway_service.py

@@ -64,6 +64,10 @@ class TestSystemdServiceRefresh:
 
         monkeypatch.setattr(gateway_cli, "get_systemd_unit_path", lambda system=False: unit_path)
         monkeypatch.setattr(gateway_cli, "generate_systemd_unit", lambda system=False, run_as_user=None: "new unit\n")
+        # Test fixtures don't have a real user D-Bus session — stub the preflight
+        # so the test exercises the unit-refresh + systemctl call sequence rather
+        # than the linger / loginctl probe (covered by TestPreflightUserSystemd).
+        monkeypatch.setattr(gateway_cli, "_preflight_user_systemd", lambda *a, **kw: None)
 
         calls = []
 
@@ -87,6 +91,9 @@ class TestSystemdServiceRefresh:
 
         monkeypatch.setattr(gateway_cli, "get_systemd_unit_path", lambda system=False: unit_path)
         monkeypatch.setattr(gateway_cli, "generate_systemd_unit", lambda system=False, run_as_user=None: "new unit\n")
+        # Same rationale as test_systemd_start_refreshes_outdated_unit: this test
+        # validates the unit-refresh + systemctl call shape, not the D-Bus probe.
+        monkeypatch.setattr(gateway_cli, "_preflight_user_systemd", lambda *a, **kw: None)
 
         calls = []
 
@@ -115,10 +122,20 @@ class TestGeneratedSystemdUnits:
         assert "ExecStop=" not in unit
         assert "ExecReload=/bin/kill -USR1 $MAINPID" in unit
         assert f"RestartForceExitStatus={GATEWAY_SERVICE_RESTART_EXIT_CODE}" in unit
-        # TimeoutStopSec must exceed the default drain_timeout (60s) so
-        # systemd doesn't SIGKILL the cgroup before post-interrupt cleanup
-        # (tool subprocess kill, adapter disconnect) runs — issue #8202.
-        assert "TimeoutStopSec=90" in unit
+        # TimeoutStopSec must exceed the default drain_timeout so systemd
+        # doesn't SIGKILL the cgroup before post-interrupt cleanup (tool
+        # subprocess kill, adapter disconnect) runs — issue #8202.
+        # gateway.py: restart_timeout = max(60, drain_timeout) + 30s headroom.
+        # PR #18761 (2026-05-02) raised the default drain_timeout from 60s
+        # to 180s after a /restart force-interrupted three mid-API-call
+        # agents inside the old 60s budget, so the expected TimeoutStopSec
+        # is now 180 + 30 = 210s.
+        expected_timeout = max(60, int(DEFAULT_GATEWAY_RESTART_DRAIN_TIMEOUT)) + 30
+        assert f"TimeoutStopSec={expected_timeout}" in unit
+        assert expected_timeout == 210, (
+            "drain_timeout default changed; update DEFAULT_CONFIG and this "
+            "test in lockstep — see PR #18761 commit message for rationale."
+        )
 
     def test_user_unit_includes_resolved_node_directory_in_path(self, monkeypatch):
         monkeypatch.setattr(gateway_cli.shutil, "which", lambda cmd: "/home/test/.nvm/versions/node/v24.14.0/bin/node" if cmd == "node" else None)
@@ -134,10 +151,16 @@ class TestGeneratedSystemdUnits:
         assert "ExecStop=" not in unit
         assert "ExecReload=/bin/kill -USR1 $MAINPID" in unit
         assert f"RestartForceExitStatus={GATEWAY_SERVICE_RESTART_EXIT_CODE}" in unit
-        # TimeoutStopSec must exceed the default drain_timeout (60s) so
-        # systemd doesn't SIGKILL the cgroup before post-interrupt cleanup
-        # (tool subprocess kill, adapter disconnect) runs — issue #8202.
-        assert "TimeoutStopSec=90" in unit
+        # See test_user_unit_avoids_recursive_execstop_and_uses_extended_stop_timeout
+        # for the TimeoutStopSec rationale; both unit generators share the same
+        # max(60, drain_timeout) + 30s formula so the user and system units stay
+        # in lockstep on this contract.
+        expected_timeout = max(60, int(DEFAULT_GATEWAY_RESTART_DRAIN_TIMEOUT)) + 30
+        assert f"TimeoutStopSec={expected_timeout}" in unit
+        assert expected_timeout == 210, (
+            "drain_timeout default changed; update DEFAULT_CONFIG and this "
+            "test in lockstep — see PR #18761 commit message for rationale."
+        )
         assert "WantedBy=multi-user.target" in unit
 
 
@@ -437,6 +460,11 @@ class TestGatewayServiceDetection:
         monkeypatch.setattr(gateway_cli, "is_linux", lambda: True)
         monkeypatch.setattr(gateway_cli, "is_termux", lambda: False)
         monkeypatch.setattr(gateway_cli, "is_wsl", lambda: False)
+        # supports_systemd_services() also branches on is_container() — without
+        # this stub the test runs the container fallback which calls real
+        # systemctl, returning False on a CI runner that has /.dockerenv but
+        # no working systemd. The test models a native-Linux host, so pin it.
+        monkeypatch.setattr(gateway_cli, "is_container", lambda: False)
         monkeypatch.setattr(gateway_cli.shutil, "which", lambda name: "/usr/bin/systemctl")
 
         assert gateway_cli.supports_systemd_services() is True
@@ -487,6 +515,10 @@ class TestGatewaySystemServiceRouting:
         calls = []
 
         monkeypatch.setattr(gateway_cli, "_select_systemd_scope", lambda system=False: False)
+        # systemd_restart() calls _preflight_user_systemd() before refreshing
+        # the unit; the test fixture has no real D-Bus session so stub it. The
+        # preflight contract itself is covered by TestPreflightUserSystemd below.
+        monkeypatch.setattr(gateway_cli, "_preflight_user_systemd", lambda *a, **kw: None)
         monkeypatch.setattr(gateway_cli, "refresh_systemd_unit_if_needed", lambda system=False: calls.append(("refresh", system)))
         monkeypatch.setattr(
             "gateway.status.get_running_pid",
@@ -541,6 +573,8 @@ class TestGatewaySystemServiceRouting:
 
     def test_systemd_restart_recovers_failed_planned_restart(self, monkeypatch, capsys):
         monkeypatch.setattr(gateway_cli, "_select_systemd_scope", lambda system=False: False)
+        # See sibling test for preflight stub rationale.
+        monkeypatch.setattr(gateway_cli, "_preflight_user_systemd", lambda *a, **kw: None)
         monkeypatch.setattr(gateway_cli, "refresh_systemd_unit_if_needed", lambda system=False: None)
         monkeypatch.setattr(
             "gateway.status.read_runtime_status",

tests/hermes_cli/test_gateway_wsl.py

@@ -145,6 +145,13 @@ class TestSupportsSystemdServicesWSL:
         monkeypatch.setattr(gateway, "is_linux", lambda: True)
         monkeypatch.setattr(gateway, "is_termux", lambda: False)
         monkeypatch.setattr(gateway, "is_wsl", lambda: False)
+        # Pin is_container() too: CI runners often have /.dockerenv which would
+        # send supports_systemd_services() down the _container_systemd_operational()
+        # branch (calls real systemctl) and return False here.
+        monkeypatch.setattr(gateway, "is_container", lambda: False)
+        # Make sure a systemctl binary appears present so we don't trip the
+        # which() short-circuit in supports_systemd_services().
+        monkeypatch.setattr(gateway.shutil, "which", lambda name: "/usr/bin/systemctl")
         assert gateway.supports_systemd_services() is True
 
     def test_termux_still_excluded(self, monkeypatch):