On Sun, Nov 24, 2013 at 02:49:20PM -0600, Michael Lustfield wrote:
This has been corrected and pushed. It will be fixed in the next upload. Thanks for taking the time and effort for this!
I think we should keep type=forking, copying from `man systemd.service`: type=simple (default) If set to simple (the default value if BusName= is not specified) **it is expected that the process configured with ExecStart= is the main process of the service.** In this mode, if the process offers functionality to other processes on the system its communication channels should be installed before the daemon is started up (e.g. sockets set up by systemd, via socket activation), as **systemd will immediately proceed starting follow-up units** type=forking If set to forking it is expected that the process configured with ExecStart= **will call fork() as part of its start-up. The parent process is expected to exit when start-up is complete and all communication channels set up.** The child continues to run as the main daemon process. This is the behavior of traditional UNIX daemons. If this setting is used, it is recommended to also use the PIDFile= option, so that systemd can identify the main process of the daemon. **systemd will proceed starting follow-up units as soon as the parent process exits.** nginx with master_process & daemon fits exactly the forking type, it calls fork() and spawn a process that writes /run/nginx.pid file. When the main process exits, systemd starts all follow-up units, this is not the case with type=simple, in that case nginx immediately starts follow up units. So type=simple works by coinsidence: We have no follow up units that depend on nginx (if we had they would be started before nginx had setup it's sockets). This is because nginx service jumps to state "running" before the actual process is up, with type=forking it first changes to state "start" and when the control process dies, correctly jumps to "running" logs with type=simple: About to execute: /usr/sbin/nginx -g 'daemon on; master_process on;' Forked /usr/sbin/nginx as 23570 nginx.service changed start-pre -> running Job nginx.service/start finished, result=done Started A high performance web server and a reverse proxy server. Closed jobs progress timerfd. Executing: /usr/sbin/nginx -g 'daemon on; master_process on;' Received SIGCHLD from PID 23570 (nginx). Got SIGCHLD for process 23570 (nginx) Child 23570 died (code=exited, status=0/SUCCESS) Child 23570 belongs to nginx.service Main PID changing: 23570 -> 23573 logs with type=forking: About to execute: /usr/sbin/nginx -g 'daemon on; master_process on;' Forked /usr/sbin/nginx as 23462 nginx.service changed start-pre -> start Executing: /usr/sbin/nginx -g 'daemon on; master_process on;' Accepted connection on private bus. Received SIGCHLD from PID 23462 (nginx). Got SIGCHLD for process 23462 (nginx) Child 23462 died (code=exited, status=0/SUCCESS) Child 23462 belongs to nginx.service nginx.service: control process exited, code=exited status=0 nginx.service got final SIGCHLD for state start Main PID loaded: 23463 nginx.service changed start -> running Job nginx.service/start finished, result=done Notice the "Job nginx.service/start" line, and the "control process exited" line, systemd understands the notion of a control process and waits for it to die. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
