test(docker): fix svstat 'want up' assertion in profile-gateway lifecycle test

After the supervise-perms fix lands, the s6 lifecycle actually works
for the hermes user — hermes -p <profile> gateway start now genuinely
brings the supervised gateway up rather than silently no-op'ing on
EACCES. That exposes a latent bug in this test's assertion: it
expected 'want up' to appear literally in s6-svstat output, but
s6-svstat elides redundancies — when the slot is currently up AND
s6 wants it up, the output is just 'up (pid N pgid N) X seconds';
the explicit 'want up' token only appears when current ≠ wanted
(e.g. 'down (exitcode 1) … , want up' on a crash-loop).

Add a small helper _svstat_wants_up() that reads the want-state
correctly across both spellings:
  * 'up …'                       → wanted up (unless explicit 'want down')
  * 'down …, want up'            → wanted up explicitly
  * 'down …'                     → wanted down

Both stop and start assertions now use the helper. Also rewords
the module docstring to acknowledge that the supervised process
may succeed OR crash-loop depending on environment, but the want-
state contract holds either way.

(cherry picked from commit 02c933aedc)
This commit is contained in:
Ben 2026-05-25 11:21:47 +10:00
parent 7d54288d82
commit c524b8a4dc

View file

@ -8,11 +8,14 @@ with the pre-Phase-4 informational message.
These tests were marked ``xfail(strict=True)`` through Phase 03 and
flip to plain ``test_`` once Phase 4 lands (now).
NB: The harness profile created here has no model/auth configured,
so the gateway process itself will exit with code 1 on every start
attempt (s6 will keep restarting it). We assert against s6's
``want up`` / ``want down`` state which reflects the lifecycle
command's intent, not the supervised process's health.
NB: The harness profile has no model/auth configured. Depending on
how the gateway run script handles missing config, the supervised
process may either spin up successfully (and svstat reports ``up``)
or exit fast and get throttled by s6 (and svstat reports ``down ,
want up``). Both states are valid "user asked for gateway up" results
what we assert is the *want* intent the lifecycle command set, NOT
the supervised process's health. ``s6-svc -u`` records ``want up`` in
the supervise/status file regardless of the run-script outcome.
Every ``docker exec`` here runs as the unprivileged ``hermes`` user
(via :func:`docker_exec_sh` in conftest); see the conftest module
@ -42,6 +45,27 @@ def _svstat(container: str) -> str:
return r.stdout if r.returncode == 0 else ""
def _svstat_wants_up(container: str) -> bool:
"""Read the slot's want-state from s6-svstat output.
s6-svstat formats the output to elide redundancies when the
service is currently up AND s6 wants it up, the literal token
``want up`` doesn't appear (it's implicit from the leading ``up``).
When the service is down but s6 wants it back up, ``, want up``
appears explicitly. So a comprehensive "is the want-intent set to
up" check has to accept both spellings.
"""
state = _svstat(container)
if not state:
return False
head = state.split()[0] if state.split() else ""
if head == "up":
# Currently up implies wanted-up unless ``want down`` is set.
return "want down" not in state
# Currently down — ``want up`` only shows up when explicitly set.
return "want up" in state
def test_profile_create_then_gateway_start(
built_image: str, container_name: str,
) -> None:
@ -66,17 +90,23 @@ def test_profile_create_then_gateway_start(
# After start, s6's intent is "up" — even if the supervised gateway
# process spin-fails (no model/auth in the test profile), the
# supervision-state contract holds.
# supervision-state contract holds. See ``_svstat_wants_up`` for
# why we accept both ``up …`` (currently up) and ``down …, want
# up`` (down but s6 wants up).
time.sleep(2)
state = _svstat(container_name)
assert "want up" in state, f"want up not in svstat: {state!r}"
assert _svstat_wants_up(container_name), (
f"slot want-state is not up after gateway start: "
f"{_svstat(container_name)!r}"
)
r = _sh(container_name, f"hermes -p {PROFILE} gateway stop", timeout=30)
assert r.returncode == 0
time.sleep(2)
state = _svstat(container_name)
assert "want up" not in state, f"want up still in svstat: {state!r}"
assert not _svstat_wants_up(container_name), (
f"slot want-state still up after gateway stop: "
f"{_svstat(container_name)!r}"
)
def test_profile_delete_stops_gateway(