Files
hermes-agent/tests/gateway/test_shutdown_flush.py
T
kshitijk4poor db339f0051 fix(state): consolidate gateway SessionDB writers via process-wide shared registry
A gateway process opened state.db from ~12 call sites, each minting its
own writer connection, self._lock, close-time WAL checkpoint, and
token-writer thread. With N independent writers on one WAL file, one
connection's close-time checkpoint could race another's growth — the
lost/reordered-page-write signature across 11+ incidents (#90837).

Adds hermes_state_registry.py: a process-wide, per-path, refcounted
shared registry owning the writer boundary.

- acquire(path): same resolved path returns the same instance (one
  writer connection, one lock, one token-writer thread) for every
  long-lived in-process caller (gateway runner, SessionStore, per-agent
  lazy recall, cron per-job, mirror, channel_directory, slash_commands,
  shutdown_flush, session_search, react_to_message, delegate, mcp_serve,
  auto_archive, tui_gateway).
- close() on a shared instance is a NO-OP — the registry owns the
  lifecycle, so one caller's close can never tear down a writer other
  callers still hold.
- Generation-aware retirement on inode change: a replaced state.db
  RETIRES the live generation (never lent again) but keeps it alive for
  existing holders; release is object-keyed so holders of the old
  generation drain it independently of the new one. The old
  generation's own write path still fails with the typed
  StateDbReplacedError (existing protection, unchanged).
- Replacement-open failure leaves NO registry entry for the path —
  the next acquire retries fresh, never hands out a closed stale object.
- All teardown runs OUTSIDE the registry lock: a final release's WAL
  checkpoint can never stall acquisition for every state.db.
- close_shared_session_dbs() at gateway shutdown drains every
  generation (live + retired) as the final safety net.

CLI one-shots, recovery flows, and read-only cross-profile opens keep
using SessionDB() directly with their own close() — only long-lived
in-process sites route through the registry.

References #90837 (root-cause tracker stays open: the #10 EOF signature
and the WAL-lifecycle A/B verdict remain under investigation there).
2026-09-01 20:55:35 +05:30

171 lines
5.0 KiB
Python

"""Tests for gateway/shutdown_flush.py — pending message durability (#72680)."""
import json
import os
import stat
import time
from pathlib import Path
from unittest.mock import MagicMock
import pytest
from gateway.shutdown_flush import (
_serialise_value,
flush_pending_to_file,
recover_pending_to_db,
)
def _make_flush_dir(tmp_path: Path) -> Path:
"""Create a temp flush dir and monkeypatch _get_flush_dir to use it."""
flush_dir = tmp_path / "pending_messages"
flush_dir.mkdir(parents=True, exist_ok=True)
return flush_dir
def test_flush_writes_string_pending_to_file(tmp_path, monkeypatch):
flush_dir = _make_flush_dir(tmp_path)
monkeypatch.setattr(
"gateway.shutdown_flush._get_flush_dir", lambda: flush_dir
)
pending = {"agent:main:telegram:supergroup:123": "hello world"}
count = flush_pending_to_file(pending, reason="shutdown")
assert count == 1
files = list(flush_dir.glob("*.json"))
assert len(files) == 1
payload = json.loads(files[0].read_text(encoding="utf-8"))
assert payload["session_key"] == "agent:main:telegram:supergroup:123"
assert payload["reason"] == "shutdown"
assert payload["data"]["text"] == "hello world"
assert ":" not in files[0].name
assert "telegram" not in files[0].name
def test_flush_writes_message_event_to_file(tmp_path, monkeypatch):
flush_dir = _make_flush_dir(tmp_path)
monkeypatch.setattr(
"gateway.shutdown_flush._get_flush_dir", lambda: flush_dir
)
event = MagicMock()
event.text = "user message"
event.session_id = "20260728_120000_abc"
event.platform = "telegram"
event.sender_id = "456"
event.sender_name = "Alice"
event.reply_to = None
event.media = None
event.raw_event = None
count = flush_pending_to_file({"session_key_1": event}, reason="adapter_shutdown")
assert count == 1
files = list(flush_dir.glob("*.json"))
assert len(files) == 1
payload = json.loads(files[0].read_text(encoding="utf-8"))
assert payload["data"]["text"] == "user message"
assert payload["data"]["session_id"] == "20260728_120000_abc"
def test_recover_inserts_via_append_message_and_deletes_file(tmp_path, monkeypatch):
flush_dir = _make_flush_dir(tmp_path)
monkeypatch.setattr(
"gateway.shutdown_flush._get_flush_dir", lambda: flush_dir
)
ts = int(time.time())
# Write a flush file with session_id
payload = {
"session_key": "agent:main:telegram:supergroup:123",
"reason": "shutdown",
"ts": ts,
"data": {
"text": "lost message",
"session_id": "20260728_120000_abc",
},
}
flush_file = flush_dir / "test_session_123.json"
flush_file.write_text(json.dumps(payload), encoding="utf-8")
mock_db = MagicMock()
count = recover_pending_to_db(mock_db)
assert count == 1
mock_db.append_message.assert_called_once_with(
session_id="20260728_120000_abc",
role="user",
content="lost message",
timestamp=ts,
)
assert not flush_file.exists()
def test_recover_closes_owned_db_when_unexpected_exception_escapes(
tmp_path, monkeypatch
):
"""Owned SessionDB must close even when recovery is interrupted."""
flush_dir = _make_flush_dir(tmp_path)
monkeypatch.setattr(
"gateway.shutdown_flush._get_flush_dir", lambda: flush_dir
)
(flush_dir / "pending.json").write_text(
json.dumps(
{
"session_key": "agent:main:telegram:123",
"data": {"text": "message", "session_id": "sid"},
}
),
encoding="utf-8",
)
class InterruptingDB:
released = False
def append_message(self, **_kwargs):
raise KeyboardInterrupt
db = InterruptingDB()
monkeypatch.setattr("hermes_state.get_shared_session_db", lambda: db)
monkeypatch.setattr(
"hermes_state.release_or_close", lambda _: setattr(db, "released", True)
)
with pytest.raises(KeyboardInterrupt):
recover_pending_to_db()
assert db.released is True
def test_serialise_object_with_text():
obj = MagicMock()
obj.text = "msg"
obj.session_id = "sid"
obj.platform = None
obj.sender_id = None
obj.sender_name = None
obj.reply_to = None
obj.media = None
obj.raw_event = None
result = _serialise_value(obj)
assert result is not None
assert result["text"] == "msg"
assert result["session_id"] == "sid"
def test_get_flush_dir_uses_get_hermes_home(tmp_path, monkeypatch):
"""Flush dir must use get_hermes_home(), not hardcoded Path.home()."""
import gateway.shutdown_flush as mod
captured = {}
def fake_get_hermes_home():
from pathlib import Path
captured["called"] = True
return tmp_path
monkeypatch.setattr(
"hermes_constants.get_hermes_home", fake_get_hermes_home
)
result = mod._get_flush_dir()
assert captured.get("called") is True
assert result == tmp_path / "pending_messages"