hermes-agent/tests/gateway/test_planned_stop_watcher.py
Teknium 597615ade4
fix(ci): make tests, workflows, and attribution reliable under load (#66373)
* feat(attribution): conflict-free contributor mappings via contributors/emails/ directory

The AUTHOR_MAP dict in scripts/release.py was a merge-conflict magnet:
every concurrent salvage PR appended entries to the same lines of the
same file, so parallel PRs re-conflicted on every merge to main.

New system: one file per email under contributors/emails/ — filename is
the commit-author email, first non-comment line is the GitHub login.
File additions never conflict, so any number of PRs can add mappings
concurrently.

- scripts/release.py: AUTHOR_MAP is now LEGACY_AUTHOR_MAP (frozen)
  merged with the directory at import time (directory wins). All
  existing consumers (resolve_author, contributor_audit.py) unchanged.
- scripts/add_contributor.py: idempotent CLI to add a mapping; refuses
  conflicting reassignments (incl. against the legacy map), validates
  email/login shapes.
- contributor-check.yml: attribution gate now accepts a mapping file OR
  a legacy entry; failure message prints the exact add_contributor
  command. Also auto-resolves bare <login>@users.noreply.github.com
  emails is intentionally NOT added (kept id+login form only, matching
  previous behavior).
- contributor_audit.py: guidance now points at add_contributor.py.
- tests/scripts/test_contributor_map.py: 12 tests covering loader,
  merge precedence, CLI idempotency/conflict/validation, subprocess E2E.

* feat(ci): one-shot per-file flake retry in the parallel test runner

A failing test FILE is re-run once in a fresh subprocess. Pass-on-retry
counts as green but is loudly reported in a '⚠ FLAKY' summary section
(with both attempts' output preserved) so the flake gets fixed instead
of eating a full-run rerun. Deterministic failures fail both attempts —
regressions cannot be laundered green.

- --file-retries N / HERMES_TEST_FILE_RETRIES (default 1, 0 disables)
- E2E verified: simulated first-run-fail flake goes green with banner;
  deterministic failure still exits 1; retries=0 restores old behavior.

This converts the dominant CI failure mode (one timing-sensitive test
flaking a 4600-test shard, requiring a manual 10-minute rerun and an
agent triage loop) into a self-healing retry that costs one file's
runtime.

* test(approval): loosen wall-clock perf bounds 0.15s -> 2.0s

These guard against catastrophic regex backtracking (seconds-to-minutes
class), but 0.15s is within scheduler-stall noise on loaded shared CI
runners — test_max_accepted_separator_free_input_is_fast failed a CI
shard this week on runner load alone. 2.0s still catches the regression
class with zero flake surface.

* fix(ci): job timeouts everywhere + retries on all network installs

Reliability pass over every workflow:
- timeout-minutes on all 21 jobs that lacked one (a hung job previously
  burned the 6-hour default runner budget)
- ./.github/actions/retry wrapped around every network-fetching install
  that lacked it: pip installs (deploy-site, skills-index), npm ci
  (deploy-site website, upload_to_pypi web + ui-tui), uv sync (docker
  test deps). Deterministic build steps (npm run build) deliberately
  NOT retried — split into separate steps so a real build failure fails
  fast instead of retrying 3x.

* docs(agents): document the file-retry flake policy

* fix(ci): curl retries on deploy hook + skills-index probe

* fix(ci): kill the remaining transient-failure classes in workflows + Dockerfile

From the workflow reliability audit:
- tests.yml: duration-cache restore had NO restore-keys while saves use
  run_id-suffixed keys — the cache never matched once, so LPT slicing
  always ran blind and unbalanced slices pushed heavy files toward the
  per-file timeout. One-line restore-keys fixes slice balancing.
- Label gates (lint ci-reviewed, supply-chain mcp-catalog-reviewed):
  'gh pr view || true' turned an API blip into 'label absent' → false
  BLOCKING failure. Now 3x retry, and API failure is reported as an API
  failure instead of a missing label.
- detect-changes action: compare API retried before failing open (was
  silently running all lanes on any blip).
- uv-lockfile-check: 'uv lock --check' resolves against PyPI — retried
  so registry blips don't read as 'lockfile stale'.
- docker.yml merge job: imagetools create retried (Docker Hub eventual
  consistency on just-pushed digests).
- Dockerfile: apt-get Acquire::Retries=3; s6-overlay ADDs converted to
  curl --retry 3 (ADD cannot retry; checksums still enforced); npm
  --fetch-retries=5; playwright chromium fetch retried 3x.
- Advisory artifact uploads (per-slice durations, ci-timings report)
  get continue-on-error so an artifact-service blip can't fail a green
  test slice.

* fix(tests): kill the two root-cause flakes — leaking pre-warm timer + env-dependent provider list

- test_tui_gateway_server.py: session.create / non-eager session.resume
  arm a 50ms threading.Timer (_schedule_agent_build) that outlives its
  test and fires into the NEXT test's _make_agent mock, racily
  corrupting captured state (the recurring session_resume shard
  failures). Replaced the per-test whack-a-mole stub with a module-wide
  autouse fixture; the 3 worker-lifecycle tests that genuinely need the
  deferred build opt back in via @pytest.mark.real_agent_prewarm (new
  marker in pyproject).
- test_api_key_providers.py: PROVIDER_ENV_VARS is now derived from the
  live PROVIDER_REGISTRY instead of a hand-list that had drifted
  (missing HF_TOKEN / DEEPINFRA_API_KEY) — resolve_provider('auto')
  tests failed on any machine with HF_TOKEN exported. E2E-verified with
  HF_TOKEN/DEEPINFRA_API_KEY set: 42/42 pass.

* test: de-flake 30 timing-sensitive test files for loaded CI runners

Root-cause fixes from the flake audit (session-DB mining + repo sweep):

Event-based sync instead of sleep-sync:
- title_generator: mock sets threading.Event, wait(10) replaces
  sleep(0.3) hoping the daemon thread got scheduled
- docker zombie_reaping / profile_gateway: poll-for-state helpers
  replace fixed 1-3s sleeps (s6 transitions + SIGCHLD reaping are async)
- process_registry tree test: select()-bounded readline replaces an
  unbounded blocking read (parent wedge now fails THIS test with a clear
  message instead of an opaque rc=124 file kill); SIGTERM grace 1s->2s
  (the 1s partition window mid-interpreter-startup is how a child PID
  escaped the live-system guard in CI)

Timeout raises (loaded 8-way-sliced runners see ~5s scheduling floors;
all of these complete in ms-to-1s when healthy so the raises cost
nothing on green runs):
- subprocess/thread waits <= 2s raised to 10-15s across mcp_tool,
  mcp_circuit_breaker, mcp_reconnect_retry_reset, mcp_parked_self_probe,
  mcp_cancelled_error_propagation, registry, clarify_gateway, interrupt,
  voice_cli_integration, docker_environment, session_store_lock_io,
  planned_stop_watcher, cli_interrupt_subagent, thread_scoped_output
  (joins now also assert not is_alive() so stragglers fail loudly)
- wall-clock discrimination ceilings loosened where the guarded hang is
  10x larger: local_background_child_hang 4s->10s, interrupt_cleanup
  setup 5s->20s + pgid-exit 30s->60s, mcp_stability grandchild spinup
  5s->15s, protocol/gil-starvation fast-handler 0.5s->2s,
  iso_certify_seam 1.5s->5s, wait_for_mcp_discovery 0.1s->1s
- narrow assertion windows widened: honcho first-turn wait 0.4..0.65 ->
  0.25..2.0 (property is bounded-not-hung, not an exact wall-clock);
  compression fork-lock TTL 1s->3s (12 refresh chances per lease);
  compression-lock expiry margins symmetric (ttl 0.05->0.5, sleep 1.0)
- telegram hung-DNS bound 1.0->1.4 (fake hang is 1.5s — must stay under)

* fix(tests): repair indentation from de-flake batch edit

* fix(tests): harden env isolation and replace remaining sleep-sync races

The full 42k-test run and complete npm check surfaced three more classes:

- Environment isolation: local ~/.honcho defaultHost and SSH_* variables
  leaked into Python/TUI tests. Pin the default Honcho host in the
  hermetic fixture, isolate the one fallback test from ~/.honcho, and
  blank SSH_* around terminalSetup tests. This flipped 20 false failures
  back to deterministic behavior on developer machines.
- Background-thread sleep-sync: Honcho async writer tests patched
  time.sleep globally, then busy-polled with that same mocked sleep. Under
  full-suite load the poller could starve the writer. Each test now waits
  on an Event emitted by the exact flush/retry transition; 30/30 passed
  under 15-way contention.
- Desktop streaming: the test slept 80ms and assumed a 500ms timer could
  not fire before its assertion. A loaded runner descheduled the test for
  >500ms and both chunks arrived. Producer controls now gate second-chunk
  and completion transitions explicitly.

Also make file-retry observability complete: a self-healed flaky file now
prints BOTH attempts' full output in the FLAKY summary. Two behavioral
runner tests prove pass-on-retry is green+loud+traceback-preserving, while
a deterministic failure remains red.

* refactor(ci): use gh bot pat, better retries

refactor(ci): use retry action for PR label fetch
the retry action now captures stdout as a step output, so it can serve
double duty: retry + output capture for commands like 'gh pr view' whose
result must be consumed by later steps.

Retry action gains:
- 'stdout' output (heredoc-delimited to preserve newlines)
- tee to temp file so stdout still streams to the job log
- step id 'retry' for output reference

Both lint.yml and supply-chain-audit.yml now use the retry action
directly with 'command: gh pr view ...' and read
steps.<id>.outputs.stdout.

ci: use AUTOFIX_BOT_PAT for all gh CLI / GitHub API auth

Replace secrets.GITHUB_TOKEN and github.token with
secrets.AUTOFIX_BOT_PAT across all workflows and composite actions
that use the gh CLI or GitHub API. The PAT has consistent permissions
across fork PRs (where GITHUB_TOKEN is read-only), avoids API rate
limit sharing with the default token, and is already used by
js-autofix.yml for the same reasons.

19 sites swapped across 9 files:
- lint.yml (3): label fetch, comment post/edit, comment update
- supply-chain-audit.yml (5): scan, critical comment, unbounded dep
  comment, label fetch, mcp-catalog comment
- lockfile-diff.yml (1): PR comment post/update
- skills-index-freshness.yml (1): issue creation on degraded probe
- skills-index.yml (2): index build, trigger deploy workflow
- upload_to_pypi.yml (2): release view poll, release upload
- ci.yml (1): timings report
- deploy-site.yml (2): skills index crawl
- detect-changes/action.yml (1): compare API call

---------

Co-authored-by: ethernet <arilotter@gmail.com>
2026-07-17 20:55:24 +00:00

393 lines
14 KiB
Python

"""Tests for the planned-stop marker watcher thread (gateway/run.py).
The watcher is the Windows-fallback path for the v0.13.0 session-resume
feature — on Windows ``asyncio.add_signal_handler`` raises
NotImplementedError, so the SIGTERM signal handler never runs and the
shutdown drain (which writes ``resume_pending=True``) is skipped. The
watcher closes this gap by polling for the planned-stop marker file
and translating its existence into the same shutdown-handler call a
real SIGTERM would have produced.
See issue #33778 for the original Windows session-loss bug report.
"""
import asyncio
import json
import os
import threading
import time
from unittest.mock import MagicMock
from gateway.run import _run_planned_stop_watcher
from gateway import status as status_mod
def _write_self_marker(marker, *, stale: bool = False):
"""Write a planned-stop marker that targets the CURRENT process.
The watcher only fires for markers naming our PID + start_time (the
fix for issue #34597), so tests that expect a fire must write a
self-targeting marker. Pass ``stale=True`` to backdate ``written_at``
past the TTL.
"""
written_at = "2000-01-01T00:00:00+00:00" if stale else status_mod._utc_now_iso()
record = {
"target_pid": os.getpid(),
"target_start_time": status_mod._get_process_start_time(os.getpid()),
"stopper_pid": os.getpid(),
"written_at": written_at,
}
marker.write_text(json.dumps(record), encoding="utf-8")
class _FakeRunner:
"""Stand-in for GatewayRunner — only exposes the two flags the watcher reads."""
def __init__(self, *, running: bool = True, draining: bool = False):
self._running = running
self._draining = draining
def _make_loop_capturing_calls():
"""Build a fake asyncio loop whose call_soon_threadsafe records its args."""
loop = MagicMock(spec=asyncio.AbstractEventLoop)
loop._captured = []
def fake_call_soon_threadsafe(fn, *args):
loop._captured.append((fn, args))
loop.call_soon_threadsafe = fake_call_soon_threadsafe
return loop
def test_watcher_fires_shutdown_when_marker_appears(tmp_path, monkeypatch):
"""When a marker targeting THIS process exists, fire the shutdown handler."""
marker = tmp_path / ".gateway-planned-stop.json"
# Patch the marker-path resolver so the watcher polls our temp location.
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", lambda: marker)
runner = _FakeRunner(running=True, draining=False)
loop = _make_loop_capturing_calls()
shutdown_handler = MagicMock(name="shutdown_signal_handler")
stop_event = threading.Event()
# Drop a self-targeting marker before the thread starts.
_write_self_marker(marker)
watcher = threading.Thread(
target=_run_planned_stop_watcher,
args=(stop_event, runner, loop, shutdown_handler),
kwargs={"poll_interval": 0.05},
daemon=True,
)
watcher.start()
watcher.join(timeout=10.0)
assert not watcher.is_alive(), "Watcher should exit after firing"
assert len(loop._captured) == 1, (
f"Expected exactly one shutdown invocation, got {loop._captured}"
)
fn, args = loop._captured[0]
assert fn is shutdown_handler
# The handler must be called with signal=None (planned stop sentinel).
assert args == (None,)
def test_watcher_does_not_fire_when_marker_absent(tmp_path, monkeypatch):
"""No marker = no shutdown call. Watcher just spins until stop_event."""
marker = tmp_path / ".gateway-planned-stop.json"
# Deliberately do NOT create the marker.
from gateway import status as status_mod
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", lambda: marker)
runner = _FakeRunner(running=True, draining=False)
loop = _make_loop_capturing_calls()
shutdown_handler = MagicMock()
stop_event = threading.Event()
watcher = threading.Thread(
target=_run_planned_stop_watcher,
args=(stop_event, runner, loop, shutdown_handler),
kwargs={"poll_interval": 0.05},
daemon=True,
)
watcher.start()
time.sleep(0.3) # let it poll a few times
stop_event.set()
watcher.join(timeout=10.0)
assert not watcher.is_alive()
assert loop._captured == [], (
f"No marker present, but watcher fired shutdown: {loop._captured}"
)
shutdown_handler.assert_not_called()
def test_watcher_skips_when_runner_already_draining(tmp_path, monkeypatch):
"""If shutdown is already in progress, don't re-fire the handler.
This prevents a race where the SIGTERM handler is mid-drain and the
watcher would double-tap the shutdown path. We check ``_draining``
so the watcher backs off once any shutdown is in flight.
"""
marker = tmp_path / ".gateway-planned-stop.json"
_write_self_marker(marker)
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", lambda: marker)
# Already draining — watcher should be a no-op.
runner = _FakeRunner(running=False, draining=True)
loop = _make_loop_capturing_calls()
shutdown_handler = MagicMock()
stop_event = threading.Event()
watcher = threading.Thread(
target=_run_planned_stop_watcher,
args=(stop_event, runner, loop, shutdown_handler),
kwargs={"poll_interval": 0.05},
daemon=True,
)
watcher.start()
time.sleep(0.2)
stop_event.set()
watcher.join(timeout=10.0)
assert loop._captured == [], "Watcher fired while runner was already draining"
def test_watcher_skips_when_runner_not_started(tmp_path, monkeypatch):
"""If the runner hasn't started, the marker is for a previous instance —
we shouldn't shutdown a not-yet-running gateway.
"""
marker = tmp_path / ".gateway-planned-stop.json"
marker.write_text('{"target_pid": 9999}', encoding="utf-8")
from gateway import status as status_mod
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", lambda: marker)
runner = _FakeRunner(running=False, draining=False)
loop = _make_loop_capturing_calls()
shutdown_handler = MagicMock()
stop_event = threading.Event()
watcher = threading.Thread(
target=_run_planned_stop_watcher,
args=(stop_event, runner, loop, shutdown_handler),
kwargs={"poll_interval": 0.05},
daemon=True,
)
watcher.start()
time.sleep(0.2)
stop_event.set()
watcher.join(timeout=10.0)
assert loop._captured == [], "Watcher fired before runner was running"
def test_watcher_responds_to_stop_event_promptly(tmp_path, monkeypatch):
"""Setting stop_event must exit the watcher within ~poll_interval seconds."""
marker = tmp_path / ".gateway-planned-stop.json"
from gateway import status as status_mod
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", lambda: marker)
runner = _FakeRunner(running=True, draining=False)
loop = _make_loop_capturing_calls()
stop_event = threading.Event()
watcher = threading.Thread(
target=_run_planned_stop_watcher,
args=(stop_event, runner, loop, MagicMock()),
kwargs={"poll_interval": 0.1},
daemon=True,
)
watcher.start()
time.sleep(0.05)
started_stop = time.monotonic()
stop_event.set()
watcher.join(timeout=10.0)
elapsed = time.monotonic() - started_stop
assert not watcher.is_alive()
assert elapsed < 2.0 # 0.05s-poll thread; loose bound for scheduler stalls, f"Watcher took {elapsed:.2f}s to honour stop_event"
def test_watcher_fires_only_once_when_marker_persists(tmp_path, monkeypatch):
"""Marker file existing for multiple polls must NOT spam the handler.
The watcher fires once and exits its loop (the shutdown handler is
responsible for consuming the marker on its own thread). If we
re-fired on every tick, the handler would be invoked dozens of
times before the gateway actually shuts down.
"""
marker = tmp_path / ".gateway-planned-stop.json"
_write_self_marker(marker)
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", lambda: marker)
runner = _FakeRunner(running=True, draining=False)
loop = _make_loop_capturing_calls()
stop_event = threading.Event()
watcher = threading.Thread(
target=_run_planned_stop_watcher,
args=(stop_event, runner, loop, MagicMock()),
kwargs={"poll_interval": 0.05},
daemon=True,
)
watcher.start()
# Let the watcher tick several times — but it should exit after the first fire.
watcher.join(timeout=10.0)
assert not watcher.is_alive()
assert len(loop._captured) == 1, (
f"Watcher fired {len(loop._captured)} times; should fire once "
f"and exit (events={loop._captured})"
)
def test_watcher_tolerates_marker_path_resolution_errors(tmp_path, monkeypatch, caplog):
"""If _get_planned_stop_marker_path() raises, the watcher logs and continues."""
from gateway import status as status_mod
call_count = [0]
def explode():
call_count[0] += 1
# First call (the one outside the loop, at thread start) is fine —
# but subsequent .exists() calls on a corrupt Path could explode.
if call_count[0] == 1:
return tmp_path / "nonexistent"
raise OSError("filesystem failed")
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", explode)
runner = _FakeRunner(running=True, draining=False)
loop = _make_loop_capturing_calls()
stop_event = threading.Event()
watcher = threading.Thread(
target=_run_planned_stop_watcher,
args=(stop_event, runner, loop, MagicMock()),
kwargs={"poll_interval": 0.05},
daemon=True,
)
watcher.start()
time.sleep(0.2)
stop_event.set()
watcher.join(timeout=10.0)
assert not watcher.is_alive(), "Watcher should still honour stop_event after errors"
# No shutdown fired because the marker never reported existence.
assert loop._captured == []
# ---------------------------------------------------------------------------
# Regression coverage for issue #34597:
# A marker left behind by a PREVIOUS gateway instance (different PID, or
# past its TTL) must NOT crash the freshly booted gateway. The watcher
# only fires when the marker targets the current process, and self-heals
# by cleaning up stale/malformed markers.
# ---------------------------------------------------------------------------
def test_watcher_does_not_fire_for_foreign_pid_marker(tmp_path, monkeypatch):
"""A marker naming a DIFFERENT process must not trigger our shutdown.
This is the core #34597 regression: a stale marker from a prior
gateway instance was firing the handler, driving the new gateway into
a false "Received UNKNOWN" shutdown and a watchdog crash loop.
"""
marker = tmp_path / ".gateway-planned-stop.json"
# Foreign PID + a start_time that cannot match ours, freshly written
# so the TTL does NOT remove it — the watcher must still decline.
record = {
"target_pid": os.getpid() + 1,
"target_start_time": -1,
"stopper_pid": os.getpid() + 1,
"written_at": status_mod._utc_now_iso(),
}
marker.write_text(json.dumps(record), encoding="utf-8")
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", lambda: marker)
runner = _FakeRunner(running=True, draining=False)
loop = _make_loop_capturing_calls()
shutdown_handler = MagicMock(name="shutdown_signal_handler")
stop_event = threading.Event()
watcher = threading.Thread(
target=_run_planned_stop_watcher,
args=(stop_event, runner, loop, shutdown_handler),
kwargs={"poll_interval": 0.05},
daemon=True,
)
watcher.start()
time.sleep(0.3) # several poll cycles
stop_event.set()
watcher.join(timeout=10.0)
assert not watcher.is_alive()
assert loop._captured == [], (
f"Watcher fired on a foreign-PID marker (#34597 regression): {loop._captured}"
)
shutdown_handler.assert_not_called()
# Foreign (but live) marker is left in place — it may still belong to
# the process it names.
assert marker.exists()
def test_watcher_cleans_up_stale_marker_and_keeps_running(tmp_path, monkeypatch):
"""A marker older than the TTL is unlinked and never fires shutdown."""
marker = tmp_path / ".gateway-planned-stop.json"
# Self-targeting but backdated past the TTL: must be treated as dead.
_write_self_marker(marker, stale=True)
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", lambda: marker)
runner = _FakeRunner(running=True, draining=False)
loop = _make_loop_capturing_calls()
shutdown_handler = MagicMock(name="shutdown_signal_handler")
stop_event = threading.Event()
watcher = threading.Thread(
target=_run_planned_stop_watcher,
args=(stop_event, runner, loop, shutdown_handler),
kwargs={"poll_interval": 0.05},
daemon=True,
)
watcher.start()
time.sleep(0.3)
stop_event.set()
watcher.join(timeout=10.0)
assert not watcher.is_alive()
assert loop._captured == [], "Stale marker must not fire shutdown"
shutdown_handler.assert_not_called()
assert not marker.exists(), "Stale marker should have been cleaned up"
def test_planned_stop_marker_targets_self_probe_is_non_destructive(tmp_path, monkeypatch):
"""The probe returns True for a self-marker WITHOUT unlinking it.
The shutdown handler performs the authoritative consume on its own
thread, so the watcher's probe must leave a matching marker intact.
"""
marker = tmp_path / ".gateway-planned-stop.json"
_write_self_marker(marker)
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", lambda: marker)
assert status_mod.planned_stop_marker_targets_self() is True
assert marker.exists(), "Probe must not consume a matching marker"
# Idempotent: still True on a second call.
assert status_mod.planned_stop_marker_targets_self() is True
def test_planned_stop_marker_targets_self_drops_malformed(tmp_path, monkeypatch):
"""A malformed marker reports False and is cleaned up."""
marker = tmp_path / ".gateway-planned-stop.json"
marker.write_text("{not valid json", encoding="utf-8")
monkeypatch.setattr(status_mod, "_get_planned_stop_marker_path", lambda: marker)
assert status_mod.planned_stop_marker_targets_self() is False