[09:22] Hi all. Is there any command in Upstart to reset job hanging in "stop/killed" status? [09:22] I'm trying to write my own job file for my service, and it don't work well yet. So, I need to debug it, but now I've stuck with this "stop/killed" state. [09:31] Upstart’s current fork-tracking code is preliminary and gets confused easily. You probably used the ‘expect’ stanza claimin a certain style of daemonizing and the behavior of your main process was different. Can you reboot? [09:32] http://heh.fi/tmp/workaround-upstart-snafu if not. [09:33] ‘# workaround-upstart-snafu n’, where n is the pid ‘status jobname’ says, given that no such process actually exists. [09:33] Yes, I can. But I'm not very happy to do it every time during development. [09:35] If you’re not absolute sure of the forking behavior of the main process, it’s better to make it not fork and not use the ‘expect’ stanza. [09:35] ly [09:36] WOW! I've just started workaround-upstart-snafu without arguments... and it start output a lot of numbers (pids?) and not interruptable by ctrl+c [09:37] Not quite good behaviour.. [09:38] pkill -f workaround [09:38] I never said it’s a good program. :-) As the name says, it’s a workaround. [09:38] Ok. It's died. ) [09:39] But... ‘# workaround-upstart-snafu n' not helped. [09:39] Same numbers. [09:41] Updated the file. Now it shouldn’t behave incorrectly when given no parameters. [09:41] Exactly what does ‘status jobname’ print? [09:42] root@dahari:~# status stargazer [09:42] stargazer stop/killed, process 12148 [09:43] What does ‘ps -p 12148’ print? [09:45] Oh. It's over. I used workaround-upstart-snafu 12148 and wait a bit longer, letting it proceed with numbers. :) Now status is stop/waiting. Thank you. :) [20:03] when I run an upstart job on hary, initctl start doesn't return to the shell and just leave me hanging. if I ctrl-c, the process I exec'ed in my upstart job is running fine though