[00:19] magicrobotmonkey, if you want to restart it without rebooting, you can rename the job file and start that job name. === marrusl is now known as marrusl_afk === Md_ is now known as Md === phroddon_ is now known as phroddon === marrusl_afk is now known as marrusl [17:22] lets say you've got this daemon that, according to strace, forks 5 times [17:23] magicrobotmonkey: "this daemon" is doing it wrong [17:23] i know [17:25] magicrobotmonkey: do you have any control over the daemon? [17:25] magicrobotmonkey: like, can you change its code or pass options to run it in the foreground? [17:25] yea i can probably get in there and hack around a bit [17:26] https://gist.github.com/acdha/1506392 [17:26] im also not sure that it actually is forking 5 times [17:27] its a shell script? [17:27] unless the shell script uses 'exec', thats one fork [17:29] no its a shell script that calls a python script that does it's own daemonizing [17:29] ah [17:29] graphite is made by pretty sane people, I bet it can run in the foreground [17:30] yea i agree, i'll keep hacking around, thanks [20:29] hey what happens if your script has a prestart but no script/exec, like this: https://github.com/gosquared/graphite-cookbook/blob/master/templates/default/carbon-cache.upstart.erb [20:33] is there anyway to fix the hang on start/stop without a reboot or rename the conf in /etc/init [20:37] jY: yes, you can loop the pid space to re-create the dead pid [20:37] SpamapS: but i need to do that for all services [20:37] or just run it once? [20:37] all? [20:38] jY: you have one stuck job right? [20:38] yes [20:38] jY: just once [20:38] whats a command to do that [20:38] jY: there's a ruby script that does it linked in the bug [20:38] thanks [20:40] there's a bash script on it too [20:40] ok i'll try to find the bug [20:41] https://bugs.launchpad.net/upstart/+bug/406397 [20:44] i take it i use the ruby script to hit the pid of the hung start/stop? [20:45] yup [20:47] thanks guys