🕷️ Crawler Inspector

Toggle navigation [mozillazg's Blog](https://mozillazg.com/) - [Feed](https://mozillazg.com/feeds/all.atom.xml) - [About](https://mozillazg.com/2014/10/pages/about-me.html) - [Archives](https://mozillazg.com/archives.html) # [Python: 僵尸进程的产生和清除方法](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html "Permalink to Python: 僵尸进程的产生和清除方法") 2017-07-20 [python](https://mozillazg.com/category/python.html) Tags [process](https://mozillazg.com/tag/process.html) / [zombie](https://mozillazg.com/tag/zombie.html) ## 僵尸进程产生的原因[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-1 "Permalink to this headline") 在 unix 或 unix-like 的系统中，当一个子进程退出后，它就会变成一个僵尸进程，如果父进程没有通过 wait 系统调用来读取这个子进程的退出状态的话，这个子进程就会一直维持僵尸进程状态。 [Zombie process - Wikipedia](https://en.wikipedia.org/wiki/Zombie_process) 中是这样描述的： > On Unix and Unix-like computer operating systems, a zombie process or defunct process is a process that has completed execution (via the exit system call) but still has an entry in the process table: it is a process in the "Terminated state". This occurs for child processes, where the entry is still needed to allow the parent process to read its child's exit status: once the exit status is read via the wait system call, the zombie's entry is removed from the process table and it is said to be "reaped". A child process always first becomes a zombie before being removed from the resource table. In most cases, under normal system operation zombies are immediately waited on by their parent and then reaped by the system – processes that stay zombies for a long time are generally an error and cause a resource leak. 并且僵尸进程无法通过 kill 命令来清除。 本文将探讨如何手动制造一个僵尸进程以及清除僵尸进程的办法。 ## 手动制造一个僵尸进程[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-2 "Permalink to this headline") 为了便于后面讲解清除僵尸进程的方法，我们使用日常开发中经常使用的 multiprocessing 模块来制造僵尸进程（准确的来说是制造一个长时间维持僵尸进程状态的子进程）： ``` $ cat test_a.py from multiprocessing import Process, current_process import logging import os import time logging.basicConfig( level=logging.DEBUG, format='%(asctime)-15s - %(levelname)s - %(message)s' ) def run(): logging.info('exit child process %s', current_process().pid) os._exit(3) p = Process(target=run) p.start() time.sleep(100) ``` 测试： ``` $ python test_a.py & [1] 10091 $ 2017-07-20 21:28:14,792 - INFO - exit child process 10106 $ ps aux |grep 10106 mozillazg 10126 0.0 0.0 2434836 740 s006 R+ 0:00.00 grep 10106 mozillazg 10106 0.0 0.0 0 0 s006 Z 0:00.00 (Python) ``` 可以看到，子进程 10106 变成了僵尸进程（状态是 Z ）。 既然已经可以控制僵尸进程的产生了，那我们就可以进入下一步如何清除僵尸进程了。 ## 清除僵尸进程[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-3 "Permalink to this headline") 清除僵尸进程至少有三种方法： - 第一种方法就是结束父进程（一般是主进程）。当父进程退出的时候僵尸进程随后也会被清除。 当然这个是个暴力的手段，因为我们一般肯定是希望父进程继续运行的。 - 第二种方法就是通过 wait 调用来读取子进程退出状态。比如通过 multiprocessing.Process 产出的进程可以通过子进程的 join() 方法来 wait，也可以在父进程中处理 SIGCHLD 信号，在处理程序中调用 wait 系统调用或者直接设置为 SIG\_IGN 来清除僵尸进程。 - 第三种办法就说把进程变成孤儿进程，这样进程就会自动交由 init 进程（pid 为 1 的进程）来处理，一般 init 进程都包含对僵尸进程进行处理的逻辑。（这里有个坑，那就是 docker 容器中一般 pid 为 1 进程就是主程序的进程，而不是我们预期的 init 进程。如果要使用这种方法的话，需要注意一下类似的场景） ### 处理 SIGCHLD 信号[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsigchld "Permalink to this headline") 子进程退出时系统会向父进程发送 SIGCHLD 信号，父进程可以通过注册 SIGCHLD 信号处理程序，在信号处理程序中调用 wait 系统调用来清理僵尸进程。 ``` $ cat test_b.py import errno from multiprocessing import Process, current_process import logging import os import signal import time logging.basicConfig( level=logging.DEBUG, format='%(asctime)-15s - %(levelname)s - %(message)s' ) def run(): exitcode = 3 logging.info('exit child process %s with exitcode %s', current_process().pid, exitcode) os._exit(exitcode) def wait_child(signum, frame): logging.info('receive SIGCHLD') try: while True: # -1 表示任意子进程 # os.WNOHANG 表示如果没有可用的需要 wait 退出状态的子进程，立即返回不阻塞 cpid, status = os.waitpid(-1, os.WNOHANG) if cpid == 0: logging.info('no child process was immediately available') break exitcode = status >> 8 logging.info('child process %s exit with exitcode %s', cpid, exitcode) except OSError as e: if e.errno == errno.ECHILD: logging.warning('current process has no existing unwaited-for child processes.') else: raise logging.info('handle SIGCHLD end') signal.signal(signal.SIGCHLD, wait_child) p = Process(target=run) p.start() while True: time.sleep(100) ``` 效果： ``` $ python test_b.py & [1] 10159 $ 2017-07-20 21:28:56,085 - INFO - exit child process 10174 with exitcode 3 2017-07-20 21:28:56,088 - INFO - receive SIGCHLD 2017-07-20 21:28:56,089 - INFO - child process 10174 exit with exitcode 3 2017-07-20 21:28:56,090 - WARNING - current process has no existing unwaited-for child processes. 2017-07-20 21:28:56,090 - INFO - handle SIGCHLD end $ ps aux |grep 10174 mozillazg 10194 0.0 0.0 2432788 556 s006 R+ 0:00.00 grep 10174 ``` 可以看到，子进程退出变成僵尸进程后，系统给父进程发送了 SIGCHLD 信号，我们在 SIGCHLD 信号的处理程序中通过 os.waitpid 调用 wait 系统调用后阻止了子进程一直处于僵尸进程状态，从而实现了清除僵尸进程的效果。 也可以把 SIGCHLD 设置为 SIG\_IGN ### 把进程变成孤儿进程[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-4 "Permalink to this headline") 什么是孤儿进程：当父进程已经退出但是子进程仍旧在运行时，这个子进程就变成了孤儿进程。 系统会把孤儿进程的父进程设置为 init 进程，将由 init 进程来管理这个孤儿进程。 我们修改一下前面的程序，改成由子进程的子进程来执行具体逻辑： ``` $ cat test_c.py from multiprocessing import Process, current_process import logging import os import time logging.basicConfig( level=logging.DEBUG, format='%(asctime)-15s - %(levelname)s - %(message)s' ) def run(): time.sleep(30) logging.info('exit grandchild process %s', current_process().pid) os._exit(3) def worker(): p = Process(target=run) p.start() logging.info('exit worker process %s, grandchild is %s', current_process().pid, p.pid) os._exit(1) p = Process(target=worker) p.start() p.join() time.sleep(100) ``` 效果: ``` $ python test_c.py & [1] 79565 2017-07-30 18:18:27,680 - INFO - exit worker process 79585, grandchild is 79586 $ ps -f |grep test_c.py mozillazg 79565 62003 0 6:18PM ttys009 0:00.06 python test_c.py mozillazg 79586 1 0 6:18PM ttys009 0:00.00 python test_c.py mozillazg 79611 62003 0 6:18PM ttys009 0:00.00 grep test_c.py $ 2017-07-30 18:18:57,681 - INFO - exit grandchild process 79586 $ ps -f |grep 79586 mozillazg 79697 62003 0 6:19PM ttys009 0:00.00 grep 79586 $ ps -f |grep test_c.py mozillazg 79565 62003 0 6:18PM ttys009 0:00.06 python test_c.py mozillazg 79741 62003 0 6:19PM ttys009 0:00.00 grep test_c.py ``` 可以看到当结束了进程 79585 之后，它的子进程 79586 的父进程的 pid 就变成了 1 ， 随后退出 79586 进程后，进程 79586 并没有出现一直维持僵尸进程状态的情况。 ## 结语[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-5 "Permalink to this headline") 希望本文能对你有所帮助。 ## 参考资料[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-6 "Permalink to this headline") - [16\.1. os — Miscellaneous operating system interfaces — Python 3.6.2 documentation](https://docs.python.org/3/library/os.html#os._exit) - [16\.1. os — Miscellaneous operating system interfaces — Python 3.6.2 documentation](https://docs.python.org/3/library/os.html#os.waitpid) - [waitpid(3) - Linux man page](https://linux.die.net/man/3/waitpid) - [IBM Knowledgecenter - waitpid()--Wait for Specific Child Process](https://www.ibm.com/support/knowledgecenter/ssw_i5_54/apis/waitpid.htm) - [gunicorn/arbiter.py at 19.7.1 · benoitc/gunicorn](https://github.com/benoitc/gunicorn/blob/19.7.1/gunicorn/arbiter.py#L506) - [cpython/forkserver.py at 13e96cc596d158b98996db3fa291086ea4afecd9 · python/cpython](https://github.com/python/cpython/blob/13e96cc596d158b98996db3fa291086ea4afecd9/Lib/multiprocessing/forkserver.py#L198-L223) - [深入浅出---unix多进程编程之wait()和waitpid()函数 - wintree的专栏 - CSDN博客](http://blog.csdn.net/wallwind/article/details/6998602) - [Zombie process - Wikipedia](https://en.wikipedia.org/wiki/Zombie_process) - [Orphan process - Wikipedia](https://en.wikipedia.org/wiki/Orphan_process) 有任何建议和想法欢迎在下方评论区留言或者加我[微信](https://mozillazg.com/static/wechat.png)交流 *** ## Comments *** © 2025 mozillazg · Powered by [pelican-bootstrap3](https://github.com/DandyDev/pelican-bootstrap3), [Pelican](http://docs.getpelican.com/), [Bootstrap](http://getbootstrap.com/) · [Privacy](https://mozillazg.com/privacy-policy.html) [](http://creativecommons.org/licenses/by-sa/4.0/) "mozillazg's Blog" by [mozillazg](https://mozillazg.com/) is licensed under a [Creative Commons Attribution-ShareAlike 4.0 International License](http://creativecommons.org/licenses/by-sa/4.0/), except where indicated otherwise. [Back to top](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html)

2017-07-20 ## 僵尸进程产生的原因[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-1 "Permalink to this headline") 在 unix 或 unix-like 的系统中，当一个子进程退出后，它就会变成一个僵尸进程，如果父进程没有通过 wait 系统调用来读取这个子进程的退出状态的话，这个子进程就会一直维持僵尸进程状态。 [Zombie process - Wikipedia](https://en.wikipedia.org/wiki/Zombie_process) 中是这样描述的： > On Unix and Unix-like computer operating systems, a zombie process or defunct process is a process that has completed execution (via the exit system call) but still has an entry in the process table: it is a process in the "Terminated state". This occurs for child processes, where the entry is still needed to allow the parent process to read its child's exit status: once the exit status is read via the wait system call, the zombie's entry is removed from the process table and it is said to be "reaped". A child process always first becomes a zombie before being removed from the resource table. In most cases, under normal system operation zombies are immediately waited on by their parent and then reaped by the system – processes that stay zombies for a long time are generally an error and cause a resource leak. 并且僵尸进程无法通过 kill 命令来清除。 本文将探讨如何手动制造一个僵尸进程以及清除僵尸进程的办法。 ## 手动制造一个僵尸进程[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-2 "Permalink to this headline") 为了便于后面讲解清除僵尸进程的方法，我们使用日常开发中经常使用的 multiprocessing 模块来制造僵尸进程（准确的来说是制造一个长时间维持僵尸进程状态的子进程）： ``` $ cat test_a.py from multiprocessing import Process, current_process import logging import os import time logging.basicConfig( level=logging.DEBUG, format='%(asctime)-15s - %(levelname)s - %(message)s' ) def run(): logging.info('exit child process %s', current_process().pid) os._exit(3) p = Process(target=run) p.start() time.sleep(100) ``` 测试： ``` $ python test_a.py & [1] 10091 $ 2017-07-20 21:28:14,792 - INFO - exit child process 10106 $ ps aux |grep 10106 mozillazg 10126 0.0 0.0 2434836 740 s006 R+ 0:00.00 grep 10106 mozillazg 10106 0.0 0.0 0 0 s006 Z 0:00.00 (Python) ``` 可以看到，子进程 10106 变成了僵尸进程（状态是 Z ）。 既然已经可以控制僵尸进程的产生了，那我们就可以进入下一步如何清除僵尸进程了。 ## 清除僵尸进程[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-3 "Permalink to this headline") 清除僵尸进程至少有三种方法： - 第一种方法就是结束父进程（一般是主进程）。当父进程退出的时候僵尸进程随后也会被清除。 当然这个是个暴力的手段，因为我们一般肯定是希望父进程继续运行的。 - 第二种方法就是通过 wait 调用来读取子进程退出状态。比如通过 multiprocessing.Process 产出的进程可以通过子进程的 join() 方法来 wait，也可以在父进程中处理 SIGCHLD 信号，在处理程序中调用 wait 系统调用或者直接设置为 SIG\_IGN 来清除僵尸进程。 - 第三种办法就说把进程变成孤儿进程，这样进程就会自动交由 init 进程（pid 为 1 的进程）来处理，一般 init 进程都包含对僵尸进程进行处理的逻辑。（这里有个坑，那就是 docker 容器中一般 pid 为 1 进程就是主程序的进程，而不是我们预期的 init 进程。如果要使用这种方法的话，需要注意一下类似的场景） ### 处理 SIGCHLD 信号[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsigchld "Permalink to this headline") 子进程退出时系统会向父进程发送 SIGCHLD 信号，父进程可以通过注册 SIGCHLD 信号处理程序，在信号处理程序中调用 wait 系统调用来清理僵尸进程。 ``` $ cat test_b.py import errno from multiprocessing import Process, current_process import logging import os import signal import time logging.basicConfig( level=logging.DEBUG, format='%(asctime)-15s - %(levelname)s - %(message)s' ) def run(): exitcode = 3 logging.info('exit child process %s with exitcode %s', current_process().pid, exitcode) os._exit(exitcode) def wait_child(signum, frame): logging.info('receive SIGCHLD') try: while True: # -1 表示任意子进程 # os.WNOHANG 表示如果没有可用的需要 wait 退出状态的子进程，立即返回不阻塞 cpid, status = os.waitpid(-1, os.WNOHANG) if cpid == 0: logging.info('no child process was immediately available') break exitcode = status >> 8 logging.info('child process %s exit with exitcode %s', cpid, exitcode) except OSError as e: if e.errno == errno.ECHILD: logging.warning('current process has no existing unwaited-for child processes.') else: raise logging.info('handle SIGCHLD end') signal.signal(signal.SIGCHLD, wait_child) p = Process(target=run) p.start() while True: time.sleep(100) ``` 效果： ``` $ python test_b.py & [1] 10159 $ 2017-07-20 21:28:56,085 - INFO - exit child process 10174 with exitcode 3 2017-07-20 21:28:56,088 - INFO - receive SIGCHLD 2017-07-20 21:28:56,089 - INFO - child process 10174 exit with exitcode 3 2017-07-20 21:28:56,090 - WARNING - current process has no existing unwaited-for child processes. 2017-07-20 21:28:56,090 - INFO - handle SIGCHLD end $ ps aux |grep 10174 mozillazg 10194 0.0 0.0 2432788 556 s006 R+ 0:00.00 grep 10174 ``` 可以看到，子进程退出变成僵尸进程后，系统给父进程发送了 SIGCHLD 信号，我们在 SIGCHLD 信号的处理程序中通过 os.waitpid 调用 wait 系统调用后阻止了子进程一直处于僵尸进程状态，从而实现了清除僵尸进程的效果。 也可以把 SIGCHLD 设置为 SIG\_IGN ### 把进程变成孤儿进程[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-4 "Permalink to this headline") 什么是孤儿进程：当父进程已经退出但是子进程仍旧在运行时，这个子进程就变成了孤儿进程。 系统会把孤儿进程的父进程设置为 init 进程，将由 init 进程来管理这个孤儿进程。 我们修改一下前面的程序，改成由子进程的子进程来执行具体逻辑： ``` $ cat test_c.py from multiprocessing import Process, current_process import logging import os import time logging.basicConfig( level=logging.DEBUG, format='%(asctime)-15s - %(levelname)s - %(message)s' ) def run(): time.sleep(30) logging.info('exit grandchild process %s', current_process().pid) os._exit(3) def worker(): p = Process(target=run) p.start() logging.info('exit worker process %s, grandchild is %s', current_process().pid, p.pid) os._exit(1) p = Process(target=worker) p.start() p.join() time.sleep(100) ``` 效果: ``` $ python test_c.py & [1] 79565 2017-07-30 18:18:27,680 - INFO - exit worker process 79585, grandchild is 79586 $ ps -f |grep test_c.py mozillazg 79565 62003 0 6:18PM ttys009 0:00.06 python test_c.py mozillazg 79586 1 0 6:18PM ttys009 0:00.00 python test_c.py mozillazg 79611 62003 0 6:18PM ttys009 0:00.00 grep test_c.py $ 2017-07-30 18:18:57,681 - INFO - exit grandchild process 79586 $ ps -f |grep 79586 mozillazg 79697 62003 0 6:19PM ttys009 0:00.00 grep 79586 $ ps -f |grep test_c.py mozillazg 79565 62003 0 6:18PM ttys009 0:00.06 python test_c.py mozillazg 79741 62003 0 6:19PM ttys009 0:00.00 grep test_c.py ``` 可以看到当结束了进程 79585 之后，它的子进程 79586 的父进程的 pid 就变成了 1 ， 随后退出 79586 进程后，进程 79586 并没有出现一直维持僵尸进程状态的情况。 ## 参考资料[¶](https://mozillazg.com/2017/07/python-how-to-generate-kill-clean-zombie-process.html#hidsection-6 "Permalink to this headline") - [16\.1. os — Miscellaneous operating system interfaces — Python 3.6.2 documentation](https://docs.python.org/3/library/os.html#os._exit) - [16\.1. os — Miscellaneous operating system interfaces — Python 3.6.2 documentation](https://docs.python.org/3/library/os.html#os.waitpid) - [waitpid(3) - Linux man page](https://linux.die.net/man/3/waitpid) - [IBM Knowledgecenter - waitpid()--Wait for Specific Child Process](https://www.ibm.com/support/knowledgecenter/ssw_i5_54/apis/waitpid.htm) - [gunicorn/arbiter.py at 19.7.1 · benoitc/gunicorn](https://github.com/benoitc/gunicorn/blob/19.7.1/gunicorn/arbiter.py#L506) - [cpython/forkserver.py at 13e96cc596d158b98996db3fa291086ea4afecd9 · python/cpython](https://github.com/python/cpython/blob/13e96cc596d158b98996db3fa291086ea4afecd9/Lib/multiprocessing/forkserver.py#L198-L223) - [深入浅出---unix多进程编程之wait()和waitpid()函数 - wintree的专栏 - CSDN博客](http://blog.csdn.net/wallwind/article/details/6998602) - [Zombie process - Wikipedia](https://en.wikipedia.org/wiki/Zombie_process) - [Orphan process - Wikipedia](https://en.wikipedia.org/wiki/Orphan_process) ***