[00:08] I finally got it working with the following https://gist.github.com/anonymous/d8e61294931052f96f76 [08:42] joshu: you can do env HOME=/home/test \n exec /usr/bin/brscan-skey, thus no need for script / end-script. [08:43] joshu: but it looks ok / good none-the-less. [08:44] hi xnox thanks for looking at it. I can make those changes. It took a long time to figure out that chdir wasn't working. I still don't understand why, but with the export or env in your case it works [13:13] I have a system daemon that takes a long time to shutdown, I want rsyslog to stay up until that daemon is completely down but rsyslog.conf states stop on runlevel[06]. I'm not seeing how to accomplish this without altering rsyslog (and really others). Suggestions? [13:33] scampbell: echo "stop on stopped your-job" |sudo tee /etc/init/rsyslog.override [13:34] ah, thank you. [13:37] scampbell: note that if your daemon takes a really long time to stop, you may need to meddle with /etc/init.d/sendsigs as that is what currently handles "final cleanup" before shutdown. [13:49] thx, this points me in the right direction to read up :) [15:28] hrm [15:29] * SpamapS gets annoyed by the default respawn limit [17:10] SpamapS: too much, too little, not often enough? =) [17:10] xnox: Actually I took a moment to reason about what was happening and it is good. [17:10] xnox: my program does a few http fetches and fails.. and then gets respawned.. infinitely [17:10] SpamapS: a cup of tea solves everything ! [17:10] but.. that's probably a good default [17:11] even though the fail is not related to the fetches.. [17:11] kind of makes sense to just keep trying something that fails 1 time per second forever in the thought that it may very well just be failing temporarily because of network [17:11] It's turning out to be real fun to arrange for large application to be stopped on shutdown/reboot before the other services are stopped.