feat: serialize autonomous release runs (#335)
CI / lint (pull_request) Successful in 35s
CI / build-frontend (pull_request) Successful in 5s

This commit is contained in:
timmy committed 2026-08-08 19:54:20 +00:00
1 parent bf96560831
commit 4969cadd82
4 files changed
+176 -5

No files matched your search

+73 -3
View File
@@ -2,12 +2,16 @@
from __future__ import annotations
import argparse
import errno
import fcntl
import json
import os
import re
import signal
import shlex
import subprocess
import tempfile
import time
from dataclasses import asdict, dataclass
from pathlib import Path
from typing import Any, Iterable, Protocol
@@ -97,9 +101,71 @@ class StateStore:
def save(self, state: RunState) -> None:
self.path.parent.mkdir(parents=True, exist_ok=True)
tmp = self.path.with_name(self.path.name + ".tmp")
tmp.write_text(json.dumps(asdict(state), indent=2) + "\n")
tmp.replace(self.path)
temporary_path = None
try:
with tempfile.NamedTemporaryFile(
mode="w",
encoding="utf-8",
dir=self.path.parent,
prefix=self.path.name + ".",
suffix=".tmp",
delete=False,
) as temporary:
temporary_path = Path(temporary.name)
temporary.write(json.dumps(asdict(state), indent=2) + "\n")
temporary.flush()
os.fsync(temporary.fileno())
os.replace(temporary_path, self.path)
directory = os.open(self.path.parent, os.O_RDONLY | os.O_DIRECTORY)
try:
os.fsync(directory)
finally:
os.close(directory)
finally:
if temporary_path is not None:
temporary_path.unlink(missing_ok=True)
class RepositoryLease:
"""Non-blocking, process-backed single-writer lease for one worktree."""
def __init__(self, repo_path: Path):
self.path = Path(repo_path).resolve() / ".release-engine" / "run.lock"
self._file = None
def __enter__(self):
self.path.parent.mkdir(parents=True, exist_ok=True)
lock_file = self.path.open("a+", encoding="utf-8")
try:
fcntl.flock(lock_file.fileno(), fcntl.LOCK_EX | fcntl.LOCK_NB)
except OSError as exc:
if exc.errno in (errno.EACCES, errno.EAGAIN):
detail = ""
try:
lock_file.seek(0)
metadata = json.loads(lock_file.read(512))
owner_pid = int(metadata["pid"])
acquired_at = float(metadata["acquired_at"])
detail = f" (owner pid={owner_pid}, acquired_at={acquired_at:.3f})"
except (KeyError, TypeError, ValueError, json.JSONDecodeError):
pass
lock_file.close()
raise RuntimeError(f"release run already active{detail}") from exc
lock_file.close()
raise
self._file = lock_file
lock_file.seek(0)
lock_file.truncate()
lock_file.write(json.dumps({"pid": os.getpid(), "acquired_at": time.time()}) + "\n")
lock_file.flush()
os.fsync(lock_file.fileno())
return self
def __exit__(self, exc_type, exc, traceback):
if self._file is not None:
fcntl.flock(self._file.fileno(), fcntl.LOCK_UN)
self._file.close()
self._file = None
def _label_names(labels: Iterable[Any]) -> list[str]:
@@ -267,6 +333,10 @@ class ReleaseEngine:
self.store = StateStore(config.state_path)
def execute(self, issue_number: int | None = None, dry_run: bool = False) -> dict[str, Any]:
with RepositoryLease(self.config.repo_path):
return self._execute_locked(issue_number=issue_number, dry_run=dry_run)
def _execute_locked(self, issue_number: int | None = None, dry_run: bool = False) -> dict[str, Any]:
deadlines = {
"agent timeout seconds": self.config.agent_timeout_seconds,
"test timeout seconds": self.config.test_timeout_seconds,