auto-sync: 2026-05-26 11:21:57
This commit is contained in:
@@ -1,334 +1,173 @@
|
||||
# v2.7.2 Counter 生命周期修复
|
||||
# v2.7.2 防重复调用 & 防无限续杯 — 完整方案
|
||||
|
||||
**版本**: v1.1
|
||||
**版本**: v2.0
|
||||
**日期**: 2026-05-26
|
||||
**作者**: 庞统
|
||||
**状态**: 评审修订中(v1.1 采纳部分评审意见)
|
||||
**状态**: 已实现 + 待最终评审
|
||||
|
||||
---
|
||||
|
||||
## 1. 问题背景
|
||||
## 1. 核心原则
|
||||
|
||||
### 1.1 核心原则
|
||||
|
||||
> 每次 agent 调用都是独占的。openclaw 无论成功失败都会返回,最差情况是 timeout 返回。谁占用谁持有,进程退出就 release。
|
||||
|
||||
### 1.2 当前偏差
|
||||
|
||||
| 偏差 | 当前行为 | 核心原则要求 |
|
||||
|------|---------|-------------|
|
||||
| counter 生命周期 | 任务级(贯穿 retry 链) | 调用级(spawn acquire,退出 release) |
|
||||
| retry 绕过 dispatcher | `_do_retry` 直接调 `spawn_full_agent`,不检查 counter | spawn 只有 `spawn_full_agent` 一个入口,内部统一检查 |
|
||||
| A5/A6 fallback 也 retry | fallback 出现说明 agent 被占用时 spawn 了,根因错误 | 不应出现 fallback;出现了说明流程有 bug |
|
||||
| A10/A12 也 retry | compact 失败/未知错误都走 retry | 进程退出 = agent 空闲,走正常 dispatch |
|
||||
| 429 不推回 pending | release counter 但任务 still working → 30 分钟超时后才标 failed | 应推回 pending,让 ticker 重新调度 |
|
||||
| 429 无冷却 | ticker 下次(30秒后)又 dispatch → 又 429 → 循环 | per-agent 冷却期 |
|
||||
| 进程意外退出 counter 泄漏 | `_monitor_process` 异常/PM2 重启时 on_complete 不调用 | ticker 层兜底:检测进程存活性 |
|
||||
|
||||
### 1.3 司马懿事件复盘(2026-05-25 20:58-21:00)
|
||||
|
||||
```
|
||||
20:58:12 dispatch mail → counter acquire → spawn 司马懿 (pid=50512)
|
||||
20:59:50 进程退出 (gateway_timeout, exit=0) → release_counter=False → _do_retry
|
||||
→ 直接 spawn (pid=50580),不检查 counter → 司马懿被 spawn 第二次
|
||||
21:00:08 进程退出 → _do_retry → 直接 spawn (pid=50613) → 第三次
|
||||
21:00:13 API 429 → exit=1 → api_error → release_counter=True → 结束
|
||||
|
||||
问题:counter 占用但 retry 不检查,3 次 spawn 在 ~2 分钟内完成
|
||||
每次 retry 没有延迟,叠加 API 调用触发 zhipu 429
|
||||
```
|
||||
> **每次 agent 调用都是独占的。** openclaw 无论成功失败都会返回,最差情况 timeout 返回。谁占用谁持有,进程退出就 release。
|
||||
|
||||
---
|
||||
|
||||
## 2. 设计方案
|
||||
## 2. 根因分析
|
||||
|
||||
### 2.1 核心改动:counter 下沉到 spawn_full_agent
|
||||
### 2.1 2026-05-25/26 事件复盘
|
||||
|
||||
**原则:spawn 只有 `spawn_full_agent` 一个入口,acquire/release 统一在内部。**
|
||||
司马懿 + 庞统被 moziplus-v2 连续 spawn,叠加 API 调用触发 zhipu 429,双模型不可用导致 Gateway 假死。
|
||||
|
||||
```python
|
||||
async def spawn_full_agent(self, agent_id, ...):
|
||||
# 1. 检查 counter
|
||||
if self.counter and not await self.counter.can_acquire(agent_id):
|
||||
raise AgentBusyError(agent_id)
|
||||
|
||||
# 2. acquire
|
||||
if self.counter:
|
||||
await self.counter.acquire(agent_id)
|
||||
|
||||
# 3. 构建 on_complete(release counter)
|
||||
original_on_complete = on_complete
|
||||
async def _wrapped_on_complete(aid, outcome):
|
||||
if self.counter:
|
||||
self.counter.release(aid)
|
||||
if original_on_complete:
|
||||
await original_on_complete(aid, outcome)
|
||||
|
||||
# 4. spawn 进程
|
||||
proc = await asyncio.create_subprocess_exec(...)
|
||||
|
||||
# 5. schedule monitor(传 wrapped_on_complete)
|
||||
asyncio.create_task(self._monitor_process(..., on_complete=_wrapped_on_complete))
|
||||
### 2.2 发现的三个根因
|
||||
|
||||
| 根因 | 影响 | 严重度 |
|
||||
|------|------|--------|
|
||||
| **P0:`_parse_stdout_json` 解析路径错误** | `data.get("meta")` 应为 `data["response"]["meta"]`。68% 的 spawn 结果 transport=null(62/91 次)。A 场景分类全部失效 | **致命** |
|
||||
| counter 生命周期是任务级 | retry 绕过 dispatcher 直接 spawn,不检查 counter | 高 |
|
||||
| spawn 前缺 session state 检查 | webchat 占用 main session 时仍然 spawn,注定失败 | 高 |
|
||||
|
||||
### 2.3 P0 详细说明
|
||||
|
||||
openclaw agent `--json` 输出格式:
|
||||
```json
|
||||
{ "kind": "agent-response", "response": { "meta": { "transport": "gateway", ... } } }
|
||||
```
|
||||
|
||||
**变化**:
|
||||
- dispatcher 不再 acquire/release counter,只负责路由和构建业务回调
|
||||
- spawn_full_agent 内部 acquire + 注册 wrapped on_complete(保证 release)
|
||||
- on_complete 不再包含 counter.release,只做业务逻辑(幻觉门控、状态标记等)
|
||||
|
||||
### 2.2 _classify_outcome 简化
|
||||
|
||||
去掉 `release_counter` 字段。**进程退出 = release counter**(由 wrapped_on_complete 保证)。
|
||||
|
||||
只保留两个维度:
|
||||
- `should_retry`:是否触发续杯(只有 A2/A3 gateway_timeout)
|
||||
- `retry_field`:计入哪个计数器
|
||||
|
||||
| 情况 | outcome | should_retry | 说明 |
|
||||
|------|---------|-------------|------|
|
||||
| A1 正常完成 | completed | False | 任务终态 |
|
||||
| A4 Agent failed | agent_failed | False | 尊重 Agent 判断 |
|
||||
| A2/A3 Gateway timeout | gateway_timeout | True | **唯一续杯场景** |
|
||||
| A5/A6 Fallback | fallback_timeout | False | **不应出现**,记录 warning,不 retry |
|
||||
| A7 认证失败 | auth_failed | False | 不 retry |
|
||||
| A8 Gateway 不可达 | gateway_unreachable | False | 不改任务状态,等 ticker |
|
||||
| A9 API 错误/429 | api_error | False | **推回 pending**(见 2.4) |
|
||||
| A10 Compact 失败 | compact_failed | False | **不 retry**,推回 pending 等 ticker |
|
||||
| A11 Lock 冲突 | lock_conflict | False | 不改任务状态,等 ticker |
|
||||
| A12 其他 | agent_error | False | **不 retry**,推回 pending 等 ticker |
|
||||
|
||||
**A5/A6 不应出现**:如果出现了,说明 spawn 时 agent 被占用(counter 检查失效)。记录 ERROR 级日志,标 failed + escalate,不 retry。
|
||||
|
||||
**A10/A12 不 retry**:进程退出了 = agent 空闲。release counter → 推回 pending → ticker 重新调度。
|
||||
|
||||
### 2.3 _do_retry 重构
|
||||
|
||||
```python
|
||||
async def _do_retry(self, session_id, agent_id, task_id, on_complete, db_path, retry_field="retry_count"):
|
||||
# 1. 检查任务终态
|
||||
if self._is_terminal(db_path, task_id):
|
||||
return # wrapped_on_complete 已 release counter
|
||||
|
||||
# 2. 检查 retry 上限
|
||||
count = self._increment_retry(db_path, task_id, retry_field)
|
||||
if count >= self.max_retries:
|
||||
self._mark_task(db_path, task_id, "failed", {...})
|
||||
return # wrapped_on_complete 已 release counter
|
||||
|
||||
# 3. 构建续杯 message
|
||||
message = self._build_retry_message(...)
|
||||
|
||||
# 4. 通过 spawn_full_agent 重新 spawn(内部会 can_acquire)
|
||||
# 此时 counter 已由 wrapped_on_complete release
|
||||
try:
|
||||
await self.spawn_full_agent(
|
||||
agent_id=agent_id,
|
||||
message=message,
|
||||
task_id=task_id,
|
||||
on_complete=on_complete, # 业务回调(不含 counter)
|
||||
use_main_session=(session_id is None),
|
||||
reuse_session_id=session_id if session_id else None,
|
||||
task_db_path=db_path,
|
||||
)
|
||||
except AgentBusyError:
|
||||
# agent 被其他任务占用,release counter,等 ticker
|
||||
logger.warning("Retry spawn skipped: %s busy", agent_id)
|
||||
# counter 已被 wrapped_on_complete release
|
||||
# 任务保持 working,等 ticker 检查
|
||||
```
|
||||
|
||||
**关键变化**:
|
||||
- counter 在进程退出时由 wrapped_on_complete release
|
||||
- `_do_retry` 调 `spawn_full_agent`,内部 can_acquire 检查
|
||||
- 如果 agent 忙(不应该发生,但防御)→ 等 ticker
|
||||
|
||||
### 2.4 429 处理:推回 pending + 冷却机制
|
||||
|
||||
#### 推回 pending
|
||||
|
||||
```python
|
||||
# _handle_exit 中 A9 处理
|
||||
if outcome == "api_error":
|
||||
# wrapped_on_complete 已 release counter
|
||||
# 推回 pending,让 ticker 重新调度
|
||||
self._transition_task_status(db_path, task_id, "pending", {
|
||||
"reason": "api_error_retry",
|
||||
"api_retry_count": count,
|
||||
})
|
||||
```
|
||||
|
||||
#### 冷却机制
|
||||
|
||||
```python
|
||||
class ActiveAgentCounter:
|
||||
def __init__(self):
|
||||
self._active: dict[str, int] = {} # agent_id → count
|
||||
self._cooldown_until: dict[str, float] = {} # agent_id → timestamp
|
||||
|
||||
def set_cooldown(self, agent_id: str, seconds: float = 120.0):
|
||||
"""设置冷却期"""
|
||||
self._cooldown_until[agent_id] = time.time() + seconds
|
||||
|
||||
def is_cooling_down(self, agent_id: str) -> bool:
|
||||
"""检查是否在冷却期"""
|
||||
until = self._cooldown_until.get(agent_id)
|
||||
if until and time.time() < until:
|
||||
return True
|
||||
# 冷却期已过,清理
|
||||
self._cooldown_until.pop(agent_id, None)
|
||||
return False
|
||||
|
||||
async def can_acquire(self, agent_id: str) -> bool:
|
||||
if self.is_cooling_down(agent_id):
|
||||
return False
|
||||
return self._active.get(agent_id, 0) == 0
|
||||
```
|
||||
|
||||
**冷却时间**:120 秒(zhipu 瞬时限流一般 30-60 秒恢复,留余量)。
|
||||
|
||||
**冷却触发点**:
|
||||
- A9(api_error/429):`counter.set_cooldown(agent_id, 120)`
|
||||
- 冷却只阻止新 dispatch,不影响 retry 的 can_acquire(retry 时冷却已过或不存在)
|
||||
|
||||
### 2.5 进程意外退出兜底
|
||||
|
||||
在 ticker 的 `_check_timeouts` 中增加**进程存活性检查**:
|
||||
|
||||
```python
|
||||
def _check_timeouts(self, db_path: Path) -> List[str]:
|
||||
# ... 现有超时检查 ...
|
||||
|
||||
# 新增:检查 counter 占用但进程已死的情况
|
||||
if self.counter and self.spawner:
|
||||
for agent_id in self.counter.active_agents:
|
||||
session_info = self.spawner.get_session_by_agent(agent_id)
|
||||
if not session_info:
|
||||
continue
|
||||
pid = session_info.get("pid")
|
||||
if pid and not self._is_pid_alive(pid):
|
||||
# 进程已死但 counter 还占着
|
||||
crash_count = self._get_crash_count(db_path, task_id) + 1
|
||||
self._set_crash_count(db_path, task_id, crash_count)
|
||||
|
||||
if crash_count >= 3:
|
||||
# 3 次连续崩溃,标 failed
|
||||
logger.error("Agent %s crashed %d times, marking failed", agent_id, crash_count)
|
||||
self._mark_task(db_path, task_id, "failed", {
|
||||
"reason": "process_crash",
|
||||
"crash_count": crash_count,
|
||||
})
|
||||
self.counter.release(agent_id)
|
||||
else:
|
||||
# 释放 counter,让 ticker 下次重新 dispatch
|
||||
logger.warning("Agent %s process dead (crash %d/3), releasing counter",
|
||||
agent_id, crash_count)
|
||||
self.counter.release(agent_id)
|
||||
# 任务保持 working,下次 tick 会看到 counter 空闲 → 重新 dispatch
|
||||
```
|
||||
|
||||
**crash_count 存储**:`task_attempts.metadata` 的 `crash_count` 字段。
|
||||
|
||||
**重置时机**:agent 成功完成一次任务后重置为 0。
|
||||
|
||||
### 2.6 dispatcher 简化
|
||||
|
||||
dispatcher 不再管 counter:
|
||||
|
||||
```python
|
||||
async def dispatch(self, task, ...):
|
||||
# 不再 acquire counter(spawn_full_agent 内部处理)
|
||||
# 不再构建含 counter.release 的 on_complete
|
||||
|
||||
# 只做路由和业务回调
|
||||
on_complete = None
|
||||
if is_mail:
|
||||
on_complete = lambda aid, outcome: self._mail_auto_complete(task_id, aid, db_path, must_haves)
|
||||
# Task 不需要业务回调(counter release 由 spawn_full_agent 的 wrapped_on_complete 处理)
|
||||
|
||||
try:
|
||||
session_id = await self.spawner.spawn_full_agent(
|
||||
agent_id=agent_id,
|
||||
message=message,
|
||||
task_id=task.id,
|
||||
on_complete=on_complete, # 只含业务逻辑
|
||||
use_main_session=is_mail,
|
||||
task_db_path=db_path,
|
||||
)
|
||||
except AgentBusyError:
|
||||
return {"status": "skipped", "reason": "Agent busy"}
|
||||
```
|
||||
代码取的是 `data.get("meta")` 而不是 `data["response"]["meta"]`,导致:
|
||||
- 所有 transport=null → A1/A5/A6 分类失效
|
||||
- session lock 阻塞退出被误判为 gateway_timeout(A2/A3)→ 续杯循环
|
||||
- 续杯循环叠加 API 调用 → 429 → Gateway 假死
|
||||
|
||||
---
|
||||
|
||||
## 3. 改动范围
|
||||
## 3. 完整场景清单
|
||||
|
||||
| 文件 | 改动 | 行数 |
|
||||
### 3.1 Spawn 前检查(拦截无效 spawn)
|
||||
|
||||
| # | 场景 | 检测方法 | 检测到后方案 |
|
||||
|---|------|---------|-------------|
|
||||
| L1 | moziplus 内部并发 | counter.can_acquire() | AgentBusyError → 等 ticker |
|
||||
| L2 | API 429 冷却期 | counter.is_cooling_down() | AgentBusyError → 等 ticker |
|
||||
| L3a | main session 被外部占用 | _check_session_state → lock_pid_alive | AgentBusyError → 等 ticker |
|
||||
| L3b | main session 正在执行 | _check_session_state → status=processing | AgentBusyError → 等 ticker |
|
||||
| L3c | main session 正在 compact | _check_session_state → recent_compact | AgentBusyError → 等 ticker |
|
||||
|
||||
L1-L3 统一结果:不 spawn,任务保持 working,ticker 30 秒后重新调度。
|
||||
|
||||
L1+L3 互补:counter 防 moziplus 内部并发,session state 防外部占用(webchat/Control UI/cron)。
|
||||
|
||||
### 3.2 Spawn 后 — 进程退出(情况 A,0-630 秒内)
|
||||
|
||||
| # | 场景 | 检测方法 | 条件 | 检测到后方案 |
|
||||
|---|------|---------|------|-------------|
|
||||
| A1 | 正常完成 | stdout transport + 任务状态 | exit=0 + transport≠embedded + 终态 | release counter → 结束 |
|
||||
| A2/A3 | Gateway timeout | stdout transport + 任务状态 | exit=0 + 非终态 + transport 正常 | release counter → 续杯 |
|
||||
| A4 | Agent 自标 failed | 任务状态 | 任务=failed | release counter → 结束 |
|
||||
| A5/A6 | Gateway fallback | stdout transport=embedded + fallbackReason | exit=0 + transport=embedded | release counter → 标 failed + escalate |
|
||||
| A7 | 认证失败 | exit≠0 + stderr | exit≠0 + stderr 含 401/403 | release counter → 标 failed + escalate |
|
||||
| A8 | Gateway 不可达 | exit≠0 + stderr | exit≠0 + stderr 含 ECONNREFUSED/ETIMEDOUT | release counter → 等 ticker |
|
||||
| A9 | API 429 | exit≠0 + stderr | exit≠0 + stderr 含 rate_limit/500/503 | release counter → 推回 pending + 冷却 120s |
|
||||
| A10 | Compact 失败 | exit≠0 + stderr | exit≠0 + stderr 含 compaction-diag | release counter → 等 ticker |
|
||||
| A11 | Session lock 冲突 | exit≠0 + stderr | exit≠0 + stderr 含 lock/busy/concurrent | release counter → 等 ticker |
|
||||
| A12 | 兜底未知错误 | 兜底 | exit≠0 不匹配以上 | release counter → 等 ticker |
|
||||
| A2+P2 | transport=null 兜底 | stderr 辅助判断 | exit=0 + transport=null + 非终态 | 检查 stderr → lock/compact/api_error 或走 gateway_timeout |
|
||||
|
||||
**续杯:只有 A2/A3 才触发续杯。其他都 release counter,等 ticker 或推回 pending。**
|
||||
|
||||
### 3.3 Spawn 后 — 进程不退出(情况 B,630 秒后)
|
||||
|
||||
| # | 场景 | 检测方法 | 条件 | 检测到后方案 |
|
||||
|---|------|---------|------|-------------|
|
||||
| B1 | 假死(lock PID 死了) | _check_session_state | lock_pid 存在但 PID 不存活 | 标 failed + escalate + release counter |
|
||||
| B2 | Compact 进行中 | _check_session_state + stderr | lock_pid 存活 + stderr 有 compact | 继续等(不递增计数,独立上限) |
|
||||
| B3 | 进程不退出 | _check_session_state | lock_pid 存活 + 无 compact | 继续等(递增计数,≥3 次 → failed) |
|
||||
| B4 | Session 状态不匹配 | _check_session_state | sessions.json status≠running | 等 60s → 按 B3 处理 |
|
||||
|
||||
### 3.4 续杯限制
|
||||
|
||||
| # | 计数器 | 上限 | 超限后方案 |
|
||||
|---|--------|------|-----------|
|
||||
| R1 | retry_count | 3 次(每次 ~10 分钟) | 标 failed + escalate |
|
||||
| R2 | connect_retry_count | 3 次 | 标 failed |
|
||||
| R3 | api_retry_count | 3 次 | 标 failed |
|
||||
| R4 | lock_retry_count | 3 次 | 标 failed |
|
||||
| R5 | monitor_timeout_count | 3 次 | 标 failed |
|
||||
|
||||
### 3.5 Ticker 兜底(每 30 秒检查)
|
||||
|
||||
| # | 检查 | 方案 |
|
||||
|---|------|------|
|
||||
| T1 | 进程存活性:counter 占用但 PID 不存活 | release counter + 推回 pending |
|
||||
| T2 | 任务超时:working 超时(默认 30 分钟) | 标 failed |
|
||||
|
||||
### 3.6 假死复活术(2026-05-04 经验)
|
||||
|
||||
**现象**:sessions.json 状态为 running 但 agent 长时间无响应。
|
||||
|
||||
**根因**:Gateway 认为 session 还活着(running)但实际连接已断开,无超时清理。
|
||||
|
||||
**复活步骤**:
|
||||
1. 修改 sessions.json,把对应 session 的 status 从 running 改为 idle
|
||||
2. 发心跳激活(把当前任务再发一遍给 agent)
|
||||
|
||||
**待设计**:将此复活术集成到 ticker `_check_timeouts` 中,检测到 sessions.json status=running 但 lock PID 不存活时自动执行。
|
||||
|
||||
---
|
||||
|
||||
## 4. 改动文件清单
|
||||
|
||||
| 文件 | 改动 | 版本 |
|
||||
|------|------|------|
|
||||
| `spawner.py` | spawn_full_agent 加 counter acquire/release + wrapped_on_complete | ~30 |
|
||||
| `spawner.py` | _classify_outcome 去掉 release_counter 字段 | ~10 |
|
||||
| `spawner.py` | _handle_exit 简化(A9 推回 pending,A10/A12 不 retry) | ~20 |
|
||||
| `spawner.py` | _do_retry 通过 spawn_full_agent 重试 | ~15 |
|
||||
| `spawner.py` | 新增 AgentBusyError | ~5 |
|
||||
| `dispatcher.py` | 去掉 counter acquire/release/on_complete 逻辑 | ~-40 |
|
||||
| `counter.py` | 新增 cooldown 机制 | ~25 |
|
||||
| `ticker.py` | _check_timeouts 加进程存活性检查 | ~30 |
|
||||
| **合计** | | ~95 行净增 |
|
||||
|
||||
## 4. 续杯语义变化
|
||||
|
||||
| | 当前 | 修复后 |
|
||||
|---|------|--------|
|
||||
| counter 生命周期 | 任务级(贯穿 retry 链) | 调用级(spawn acquire,退出 release) |
|
||||
| retry 路径 | 直接 spawn,不检查 counter | 通过 spawn_full_agent,内部检查 |
|
||||
| retry 失败 | 只看 retry_count 上限 | can_acquire 失败也会停止(agent 忙) |
|
||||
| A9/429 | release + 任务 working → 30 分钟超时 | release + 推回 pending + 冷却 120s |
|
||||
| A10 compact | retry(counter 不 release) | 不 retry,推回 pending |
|
||||
| 意外退出 | counter 泄漏 | ticker 检测 → crash_count → 最多 3 次 |
|
||||
|
||||
## 5. 续杯时间计算
|
||||
|
||||
Gateway timeout = 600s(10 分钟),3 次续杯 = ~30 分钟。
|
||||
|
||||
每次续杯的流程:
|
||||
```
|
||||
进程退出 → release counter → _do_retry → spawn_full_agent(can_acquire)
|
||||
→ acquire → spawn → 新进程执行(最多 10 分钟)
|
||||
```
|
||||
|
||||
30 分钟内 agent 无法完成任务 → 第 3 次续杯后标 failed + escalate。
|
||||
|
||||
## 6. 测试计划
|
||||
|
||||
| 用例 | 验证 |
|
||||
|------|------|
|
||||
| 正常完成 | counter acquire → release,任务 done |
|
||||
| Gateway timeout 续杯 | release → re-acquire → spawn,复用 session |
|
||||
| 续杯时 agent 被占用 | can_acquire 失败 → 不 spawn,等 ticker |
|
||||
| 429 → 冷却 | 推回 pending,120s 内不 dispatch |
|
||||
| 进程崩溃 | ticker 检测,crash < 3 → 重新 dispatch |
|
||||
| 3 次崩溃 | 标 failed + escalate |
|
||||
| PM2 重启 | ticker 检测进程死 → release counter |
|
||||
| Mail 幻觉门控 | on_complete 业务回调正常执行 |
|
||||
| `counter.py` | 新增 cooldown 机制(is_cooling_down / set_cooldown) | v1 |
|
||||
| `spawner.py` | 新增 AgentBusyError | v1 |
|
||||
| `spawner.py` | spawn_full_agent 内部 counter acquire/release + wrapped_on_complete(try/finally) | v1 |
|
||||
| `spawner.py` | spawn_full_agent 加 L3 session state 检查 | v2 |
|
||||
| `spawner.py` | _classify_outcome 去掉 release_counter,只有 A2/A3 触发 retry | v1 |
|
||||
| `spawner.py` | _classify_outcome 加 P2 transport=null 兜底 | v2 |
|
||||
| `spawner.py` | _handle_exit:A9 推回 pending + 冷却,A5/A6 标 failed + context 日志 | v1 |
|
||||
| `spawner.py` | _do_retry 手动 release counter + 通过 spawn_full_agent 重试 | v1 |
|
||||
| `spawner.py` | P0:_parse_stdout_json 改为 data["response"]["meta"] | v2 |
|
||||
| `spawner.py` | 新增 get_session_by_agent(进程存活性检查用) | v1 |
|
||||
| `dispatcher.py` | 去掉 counter acquire/release,on_complete 只含业务逻辑 | v1 |
|
||||
| `ticker.py` | _spawn_available_agents 不再管 counter | v1 |
|
||||
| `ticker.py` | _check_timeouts 加进程存活性检查 + 推回 pending | v1 |
|
||||
| `ticker.py` | 新增 _is_pid_alive | v1 |
|
||||
| `main.py` | counter 创建提前,传给 spawner | v1 |
|
||||
|
||||
---
|
||||
|
||||
## 7. 司马懿评审意见处理(mail-1779726169654)
|
||||
## 5. 司马懿评审记录
|
||||
|
||||
### 第一次评审(mail-1779726169654)
|
||||
|
||||
| 评审意见 | 结论 | 理由 |
|
||||
|---------|------|------|
|
||||
| wrapped_on_complete 加 try/finally | ✅ 采纳 | 防御性编程,确保 counter release 和业务回调都执行 |
|
||||
| wrapped_on_complete 加 try/finally | ✅ 采纳 | 防御性编程 |
|
||||
| A5/A6 加 context 日志 | ✅ 采纳 | 排查方便 |
|
||||
| per-provider 冷却 | ⏭ 延后 | 低优先级,先做 per-agent |
|
||||
| crash_count per-agent 累计,禁用 agent | ❌ 不采纳 | 崩溃可能是任务问题不是 agent 问题。保持 per-task 3 次标 failed,通过 escalate 通知用户自行判断 |
|
||||
| can_acquire 失败推回 claimed | ❌ 不采纳 | retry 路径下 can_acquire 不会失败(asyncio 单线程无竞态)。release → can_acquire → acquire 是内存同步操作,中间无 await |
|
||||
| release 和 acquire 之间有竞态窗口 | ❌ 不存在 | Python asyncio 单线程,三步都是内存同步操作,无竞态 |
|
||||
| per-provider 冷却 | ⏭ 延后 | 低优先级 |
|
||||
| crash_count per-agent 累计,禁用 agent | ❌ 不采纳 | 崩溃可能是任务问题不是 agent 问题 |
|
||||
| can_acquire 失败推回 claimed | ❌ 不采纳 | asyncio 单线程无竞态 |
|
||||
| release 和 acquire 之间有竞态窗口 | ❌ 不存在 | 内存同步操作 |
|
||||
|
||||
## 8. 实现检查清单
|
||||
### 第二次评审(_do_retry counter 时序)
|
||||
|
||||
- [ ] counter.py:新增 cooldown 机制
|
||||
- [ ] spawner.py:spawn_full_agent 加 counter acquire/release + wrapped_on_complete(try/finally)
|
||||
- [ ] spawner.py:_classify_outcome 去掉 release_counter,只有 A2/A3 触发 retry
|
||||
- [ ] spawner.py:_do_retry 通过 spawn_full_agent 重试
|
||||
- [ ] spawner.py:_handle_exit 简化(A9 推回 pending,A5/A6 标 failed + context 日志)
|
||||
- [ ] dispatcher.py:去掉 counter acquire/release 逻辑
|
||||
- [ ] ticker.py:_check_timeouts 加进程存活性检查(per-task crash_count)
|
||||
| 评审意见 | 结论 |
|
||||
|---------|------|
|
||||
| _do_retry 续杯退化为 AgentBusyError | ✅ 采纳:_do_retry 入口手动 release counter |
|
||||
|
||||
### 第三次评审(P0/P1/P2 补丁)
|
||||
|
||||
| 评审意见 | 结论 |
|
||||
|---------|------|
|
||||
| P0 stdout 解析路径修复 | 待评审 |
|
||||
| P1 spawn 前检查 | 待评审 |
|
||||
| P2 transport=null 兜底 | 待评审 |
|
||||
|
||||
---
|
||||
|
||||
## 6. 待办
|
||||
|
||||
- [ ] 假死复活术集成到 ticker
|
||||
- [ ] 历史数据清理(P0 修复前 62 条 transport=null 的 task_attempts)
|
||||
- [ ] per-provider 冷却(低优先级)
|
||||
|
||||
Reference in New Issue
Block a user