Skip to content

fix(tasks): distinguish blocked process cleanup from failures - #411

Merged
Ed1s0nZ merged 1 commit into
AIPentest:mainfrom
Sycun:codex/issue-408-task-cleanup
Nov 19, 2026
Merged

Ed1s0nZ merged 1 commit into
AIPentest:mainfrom
Sycun:codex/issue-408-task-cleanup

Conversation

@Sycun

@Sycun Sycun commented Nov 19, 2026

Copy link
Copy Markdown
Contributor

关联 #408。

改动

  • Linux 上,直接子进程回收后,只剩僵尸进程的组视为已退出。watchdog 使用同样的进程组判定。
  • 使用进程组 watchdog 时,清理超时且剩余进程均处于 D 态、已有挂起的 SIGKILL,等本地 worker 退出后将任务记为 cleanup_unconfirmed,并在服务日志里记录 run ID 和 PGID。其他清理失败继续走原有重试流程。
  • CS_CLEANUP_GRACE、CS_CLEANUP_KILL 和 CS_CLEANUP_GUARD 分别配置退出宽限期、SIGKILL 后等待和隔离后端等待,默认各 3 秒。

验证

  • macOS 上运行 go test ./...,通过。
  • Linux 上运行进程清理相关回归和 -race 检测,覆盖真实的仅僵尸进程组、D 态 /proc 测试数据和任务状态通知。
  • 报告中的 WSL2/9p 阻塞场景仍待在原环境复测。

@Ed1s0nZ
Ed1s0nZ merged commit 85a81be into AIPentest:main Nov 19, 2026
10 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants