mirror of
https://github.com/headroomlabs-ai/headroom.git
synced 2026-08-27 14:17:10 -04:00
Unit 3 of the Codex proxy resilience plan. Eliminates the "aged process
has leaked relay tasks" hypothesis by making every WS session explicitly
tracked and both relay tasks deterministically cancelled when either
exits.
- New headroom/proxy/ws_session_registry.py: dict-backed
WebSocketSessionRegistry + WSSessionHandle with register /
deregister / attach_tasks / snapshot. Deregister is idempotent and
clears relay-task references so coroutine frames are not retained
past session end.
- HeadroomProxy exposes proxy.ws_sessions so /debug/ws-sessions
(Unit 5) can read the live snapshot.
- handle_openai_responses_ws now registers on websocket.accept()
success and deregisters in the outermost finally so no leak can
survive handshake-phase, mid-stream, or upstream-error paths. The
session_id / termination_cause is threaded through both relay
halves and both sides raise asyncio.CancelledError cleanly.
- Replaced asyncio.gather(_client_to_upstream(), _upstream_to_client(),
return_exceptions=True) with explicit asyncio.create_task(...)
(named codex-ws-c2u-<sid> / codex-ws-u2c-<sid>) +
asyncio.wait(FIRST_COMPLETED) + cancel-and-await on the survivor.
Termination cause is classified as client_disconnect /
client_error / upstream_disconnect / upstream_error /
response_completed from which task completed first plus inline
error captures from the halves.
- Prometheus metrics: new active_ws_sessions and active_relay_tasks
gauges plus ws_session_duration_ms_{sum,count,max} histogram
bucketed by termination cause. Mirrors the Unit 2 stage_timing_*
shape.
Preserved: upstream WS retry loop, WS→HTTP fallback, memory-context
timeout, compression pipeline, Unit 2 stage timings. Memory-tool
execution inside _upstream_to_client still runs when the client task
exits first; however, if the client disconnects *before* the upstream
emits response.completed, pending memory writes in `pending_fcs` are
dropped (unchanged from prior behavior — a crashing upstream has the
same effect). Note: handle_openai_responses (HTTP, line ~800) is a
single-shot HTTP request; lifecycle tracking isn't added there
(scope boundary).
Tests:
- tests/test_ws_session_registry.py: 8 registry unit tests
(register/deregister idempotency, snapshot shape, attach merging,
reference release).
- tests/test_openai_codex_ws_lifecycle.py: 6 integration tests
using real relay tasks (only upstream WS endpoint mocked):
happy-path, failing-test-first "client disconnect cancels upstream
relay within 100 ms", upstream-closes-first, upstream-error mid-
stream, handshake-failure deregister, 50 concurrent sessions.
- Regression: test_openai_codex_ws_timings, test_openai_codex_routing,
test_proxy_codex_route_aliases, test_ws_memory_relay all pass.
- Tests pass under python -W error::RuntimeWarning (no "coroutine
was never awaited").
177 lines
5 KiB
Python
177 lines
5 KiB
Python
"""Unit tests for :class:`WebSocketSessionRegistry`.
|
|
|
|
These tests exercise the in-memory registry in isolation — no network,
|
|
no WS server. They pin down register/deregister semantics, ``snapshot``
|
|
shape, and the task-attachment accounting that feeds the
|
|
``active_relay_tasks`` gauge.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import asyncio
|
|
|
|
import pytest
|
|
|
|
from headroom.proxy.ws_session_registry import (
|
|
WebSocketSessionRegistry,
|
|
WSSessionHandle,
|
|
)
|
|
|
|
|
|
def _make_handle(session_id: str = "sess-1") -> WSSessionHandle:
|
|
return WSSessionHandle(
|
|
session_id=session_id,
|
|
request_id="req-1",
|
|
client_addr="127.0.0.1:12345",
|
|
upstream_url="wss://upstream/test",
|
|
)
|
|
|
|
|
|
def test_register_and_deregister_round_trip():
|
|
reg = WebSocketSessionRegistry()
|
|
handle = _make_handle()
|
|
|
|
reg.register(handle)
|
|
assert reg.active_count() == 1
|
|
assert reg.get("sess-1") is handle
|
|
|
|
reg.deregister("sess-1", cause="response_completed")
|
|
assert reg.active_count() == 0
|
|
assert reg.get("sess-1") is None
|
|
assert handle.termination_cause == "response_completed"
|
|
|
|
|
|
def test_deregister_unknown_session_is_idempotent():
|
|
reg = WebSocketSessionRegistry()
|
|
# Deregistering a never-registered session must not raise.
|
|
reg.deregister("never-seen", cause="unknown")
|
|
assert reg.active_count() == 0
|
|
|
|
|
|
def test_register_twice_same_session_does_not_double_count():
|
|
reg = WebSocketSessionRegistry()
|
|
handle = _make_handle()
|
|
|
|
reg.register(handle)
|
|
reg.register(handle) # idempotent — overwrite, do not double-increment
|
|
assert reg.active_count() == 1
|
|
|
|
reg.deregister("sess-1", cause="client_disconnect")
|
|
assert reg.active_count() == 0
|
|
|
|
|
|
def test_snapshot_shape_is_json_serializable():
|
|
reg = WebSocketSessionRegistry()
|
|
reg.register(_make_handle("a"))
|
|
reg.register(_make_handle("b"))
|
|
|
|
snapshot = reg.snapshot()
|
|
assert len(snapshot) == 2
|
|
ids = {entry["session_id"] for entry in snapshot}
|
|
assert ids == {"a", "b"}
|
|
sample = snapshot[0]
|
|
for key in (
|
|
"session_id",
|
|
"request_id",
|
|
"client_addr",
|
|
"upstream_url",
|
|
"age_seconds",
|
|
"relay_task_count",
|
|
"termination_cause",
|
|
):
|
|
assert key in sample
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_attach_tasks_merges_and_tracks_active_count():
|
|
reg = WebSocketSessionRegistry()
|
|
handle = _make_handle()
|
|
reg.register(handle)
|
|
|
|
async def _idle() -> None:
|
|
await asyncio.sleep(5)
|
|
|
|
t1 = asyncio.create_task(_idle(), name="codex-ws-test-1")
|
|
t2 = asyncio.create_task(_idle(), name="codex-ws-test-2")
|
|
|
|
reg.attach_tasks("sess-1", [t1])
|
|
assert reg.active_relay_task_count() == 1
|
|
|
|
reg.attach_tasks("sess-1", [t2]) # merges, does not replace
|
|
assert reg.active_relay_task_count() == 2
|
|
|
|
# Deregister cancels the in-registry accounting; tasks themselves
|
|
# are the caller's responsibility to cancel — the registry is a
|
|
# bookkeeper, not an owner.
|
|
reg.deregister("sess-1", cause="client_disconnect")
|
|
assert reg.active_relay_task_count() == 0
|
|
|
|
t1.cancel()
|
|
t2.cancel()
|
|
for t in (t1, t2):
|
|
try:
|
|
await t
|
|
except asyncio.CancelledError:
|
|
pass
|
|
|
|
|
|
def test_deregister_releases_task_references():
|
|
"""Once deregistered, the handle's ``relay_tasks`` list is cleared.
|
|
|
|
This keeps large references (coroutine frames) from being held by
|
|
the registry after the session ends.
|
|
"""
|
|
reg = WebSocketSessionRegistry()
|
|
handle = _make_handle()
|
|
reg.register(handle)
|
|
|
|
# Use sentinel non-task objects; ``attach_tasks`` does not inspect type
|
|
# beyond appending to the list and updating counts.
|
|
class _FakeTask:
|
|
def __init__(self, name: str) -> None:
|
|
self._name = name
|
|
self._done = False
|
|
|
|
def done(self) -> bool:
|
|
return self._done
|
|
|
|
def cancel(self) -> bool:
|
|
self._done = True
|
|
return True
|
|
|
|
def get_name(self) -> str:
|
|
return self._name
|
|
|
|
fake_tasks = [_FakeTask("t-a"), _FakeTask("t-b")]
|
|
reg.attach_tasks("sess-1", fake_tasks)
|
|
assert len(handle.relay_tasks) == 2
|
|
|
|
reg.deregister("sess-1", cause="upstream_disconnect")
|
|
# After deregister, the handle's tasks list is empty — the registry
|
|
# does not retain references.
|
|
assert handle.relay_tasks == []
|
|
assert reg.active_relay_task_count() == 0
|
|
|
|
|
|
def test_attach_tasks_on_missing_session_is_noop():
|
|
reg = WebSocketSessionRegistry()
|
|
|
|
async def _idle() -> None:
|
|
await asyncio.sleep(0)
|
|
|
|
loop = asyncio.new_event_loop()
|
|
try:
|
|
t = loop.create_task(_idle())
|
|
reg.attach_tasks("missing", [t])
|
|
assert reg.active_relay_task_count() == 0
|
|
t.cancel()
|
|
loop.run_until_complete(asyncio.gather(t, return_exceptions=True))
|
|
finally:
|
|
loop.close()
|
|
|
|
|
|
def test_snapshot_age_seconds_is_non_negative():
|
|
reg = WebSocketSessionRegistry()
|
|
reg.register(_make_handle())
|
|
snapshot = reg.snapshot()
|
|
assert snapshot[0]["age_seconds"] >= 0.0
|