RomanNum3ral
/
hermes-agent
mirror of https://github.com/NousResearch/hermes-agent.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

refactor(gateway): drop dead degraded token field; de-churn salvage diff 

Follow-up to the #80376 salvage:

- TurnLeaseToken.degraded is dead code since acquire() started raising
  TurnLeaseTimeoutError: the only constructor site passes the default and
  repo-wide there are zero external readers. Remove the field, its ctor
  param, the repr segment, and the always-False guards in rebind()/release();
  the class docstring's 'retained for compatibility' claim described
  consumers that do not exist.
- Revert pure black-rewrap churn in test_config_env_bridge_authority.py and
  test_turn_lease.py (hunks on functions this change does not touch), keeping
  the functional Windows-env/encoding additions.
- Turn the default-value test into an invariant: config default must equal
  gateway.turn_lease.DEFAULT_LEASE_WAIT instead of pinning the 1800 literal.
- Rewrap the dangling 'Released' comment line in gateway/run.py.

Verified: 20/20 targeted gateway tests, ruff clean; mutation check — with
gateway/turn_lease.py reverted to pre-fix main the module fails, restored it
passes.
This commit is contained in:
kshitij 2026-08-07 18:17:37 +05:30
parent 3a3aed3c1f
commit 2a0d0bc698
4 changed files with 39 additions and 54 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
gateway/run.py
View File

@ -16663,10 +16663,9 @@ class GatewayRunner(GatewayAuthorizationMixin, GatewayKanbanWatchersMixin, Gatew
# Fail-closed on timeout: never enter the transcript region without a
# lease. Outer dispatch returns a bounded rejection/resend notice rather
# than recreating the exact concurrent-turn corruption this lease exists
# to prevent. Released
# in _handle_message's finally via _release_turn_lease — granted per
# (routing key, run generation) so a stale unwind can't release a
# newer turn's lease.
# to prevent. Released in _handle_message's finally via
# _release_turn_lease — granted per (routing key, run generation) so a
# stale unwind can't release a newer turn's lease.
_lease_registry = getattr(self, "_turn_leases", None)
if _lease_registry is not None:
try:

22
gateway/turn_lease.py
View File

@ -97,32 +97,29 @@ class TurnLeaseTimeoutError(TimeoutError):
class TurnLeaseToken:
"""Handle returned by :meth:`SessionTurnLeaseRegistry.acquire`.
``degraded`` is retained for compatibility with older callers and test
doubles, but :meth:`acquire` no longer returns degraded tokens: a timeout
raises :class:`TurnLeaseTimeoutError` instead. ``released`` makes release
idempotent.
A timeout raises :class:`TurnLeaseTimeoutError` instead of returning a
token, so every token handed out is a held lease. ``released`` makes
release idempotent.
"""
__slots__ = ("session_id", "owner_key", "generation", "degraded", "released")
__slots__ = ("session_id", "owner_key", "generation", "released")
def __init__(
self,
session_id: str,
owner_key: str,
generation: int,
degraded: bool = False,
) -> None:
self.session_id = session_id
self.owner_key = owner_key
self.generation = generation
self.degraded = degraded
self.released = False
def __repr__(self) -> str: # pragma: no cover - debug aid
return (
f"TurnLeaseToken(session_id={self.session_id!r}, "
f"owner_key={self.owner_key!r}, generation={self.generation}, "
f"degraded={self.degraded}, released={self.released})"
f"released={self.released})"
)
@ -291,7 +288,6 @@ class SessionTurnLeaseRegistry:
"""
if (
token is None
or token.degraded
or token.released
or not new_session_id
or new_session_id == token.session_id
@ -329,11 +325,11 @@ class SessionTurnLeaseRegistry:
"""Release ``token``'s lease. Idempotent; ownership-checked.
Returns True only when this exact token was the current holder and
the lock was freed. A degraded token, a re-release, or a stale token
whose slot has since been granted to a newer turn are all safe
no-ops — a stale unwind can never release a newer turn's lease.
the lock was freed. A re-release or a stale token whose slot has
since been granted to a newer turn are both safe no-ops — a stale
unwind can never release a newer turn's lease.
"""
if token is None or token.degraded or token.released:
if token is None or token.released:
return False
token.released = True
lease = self._leases.get(token.session_id)

56
tests/gateway/test_config_env_bridge_authority.py
View File

@ -21,9 +21,7 @@ import pytest
PROJECT_ROOT = Path(__file__).resolve().parents[2]
def _run_gateway_import(
hermes_home: Path, initial_env: dict[str, str]
) -> dict[str, str]:
def _run_gateway_import(hermes_home: Path, initial_env: dict[str, str]) -> dict[str, str]:
"""Import gateway.run in a clean subprocess and return the post-import env.
The bridge runs at module-import time, so simply importing is enough
@ -100,15 +98,9 @@ def _run_gateway_import(
return out
def _write_config(
home: Path,
agent_cfg: dict | None = None,
display_cfg: dict | None = None,
timezone: str | None = None,
gateway_cfg: dict | None = None,
) -> None:
def _write_config(home: Path, agent_cfg: dict | None = None, display_cfg: dict | None = None,
timezone: str | None = None, gateway_cfg: dict | None = None) -> None:
import yaml
cfg: dict = {}
if agent_cfg:
cfg["agent"] = agent_cfg
@ -135,22 +127,16 @@ def hermes_home(tmp_path: Path) -> Path:
def test_config_gateway_timeout_wins_over_stale_env(hermes_home: Path) -> None:
"""Every agent.* bridge key must be config-authoritative, not .env-authoritative."""
_write_config(
hermes_home,
agent_cfg={
"gateway_timeout": 1800,
"gateway_timeout_warning": 900,
"session_stall_timeout": 300,
},
)
_write_env(
hermes_home,
{
"HERMES_AGENT_TIMEOUT": "60",
"HERMES_AGENT_TIMEOUT_WARNING": "30",
"HERMES_SESSION_STALL_TIMEOUT": "15",
},
)
_write_config(hermes_home, agent_cfg={
"gateway_timeout": 1800,
"gateway_timeout_warning": 900,
"session_stall_timeout": 300,
})
_write_env(hermes_home, {
"HERMES_AGENT_TIMEOUT": "60",
"HERMES_AGENT_TIMEOUT_WARNING": "30",
"HERMES_SESSION_STALL_TIMEOUT": "15",
})
env = _run_gateway_import(hermes_home, initial_env={})
@ -190,15 +176,21 @@ def test_default_turn_lease_timeout_overrides_stale_env_when_key_is_omitted(
def test_default_turn_lease_timeout_matches_the_runtime_fallback() -> None:
"""The advertised config default must match the fail-closed runtime."""
"""The advertised config default must match the fail-closed runtime fallback.
Invariant, not a snapshot: whatever the default becomes, config and the
lease registry's DEFAULT_LEASE_WAIT must move together.
"""
from gateway.turn_lease import DEFAULT_LEASE_WAIT
from hermes_cli.config import DEFAULT_CONFIG
assert DEFAULT_CONFIG["agent"]["gateway_turn_lease_timeout"] == 1800
assert (
float(DEFAULT_CONFIG["agent"]["gateway_turn_lease_timeout"])
== DEFAULT_LEASE_WAIT
)
def test_config_platform_connect_timeout_supplies_env_when_unset(
hermes_home: Path,
) -> None:
def test_config_platform_connect_timeout_supplies_env_when_unset(hermes_home: Path) -> None:
"""config.yaml:gateway.platform_connect_timeout supplies the env var when
it isn't already set (#19776 — config surface for the Discord connect
timeout, replacing the undocumented env-var-only workaround)."""

8
tests/gateway/test_turn_lease.py
View File

@ -48,7 +48,7 @@ def test_alias_key_turn_waits_and_order_is_preserved():
token = await registry.acquire(
"sess-1", owner_key=owner_key, generation=generation, timeout=5
)
assert token is not None and not token.degraded
assert token is not None
events.append(f"load:{owner_key}")
await asyncio.sleep(hold) # simulate run + flush
events.append(f"flush:{owner_key}")
@ -394,9 +394,7 @@ def test_rebind_moves_serialization_to_new_session_id():
async def scenario():
registry = SessionTurnLeaseRegistry()
token = await registry.acquire(
"parent", owner_key="key-a", generation=1, timeout=5
)
token = await registry.acquire("parent", owner_key="key-a", generation=1, timeout=5)
assert registry.rebind(token, "child") is True
assert token is not None and token.session_id == "child"
@ -409,7 +407,7 @@ def test_rebind_moves_serialization_to_new_session_id():
assert registry.release(token) is True
t2 = await waiter
assert t2 is not None and not t2.degraded
assert t2 is not None
registry.release(t2)
_run(scenario())