| hro | Hi! Can someone please help me with this question: http://askubuntu.com/questions/355223/limit-the-parallelisation-of-upstart | 06:54 |
|---|---|---|
| hro | I just realized that my question is silly. If the jobs are daemons that never stop, then there is no way for Upstart to decide that they are done. | 07:03 |
| hro | Instead, what I need is a way to limit by disk i/o. That is, when disk i/o is high, upstart shouldn't spawn new jobs. It should only spawn after disk activity cools down | 07:04 |
| sneak | hi | 09:09 |
| sneak | how do i log the output of my job run with upstart (it does not background) to syslog? | 09:09 |
| sneak | when i do exec myapp | logger -t myapp | 09:09 |
| sneak | it still ends up in /var/log/upstart/myapp | 09:09 |
| jodh | sneak: 'exec myapp 2>&1 | logger -t myapp' | 09:28 |
| sneak | jodh: i have exactly that | 10:41 |
| sneak | exec /usr/local/bin/nsqd -data-path=/var/lib/nsqd 2>&1 | logger -t nsqd | 10:42 |
| sneak | still getting output in /var/log/upstart/nsqd.log and nothing in syslog | 10:42 |
| sneak | it's very confusing | 10:43 |
| sneak | same with exec /bin/bash -c "/usr/local/bin/nsqd -data-path=/var/lib/nsqd 2>&1 | logger -t nsqd" | 10:44 |
| jodh | sneak: look at /proc/<nsqd_pid>/fd/. | 10:45 |
| sneak | 0 is /dev/null, 1 and 2 are /dev/pts/3 | 10:46 |
| sneak | thanks for your help, btw | 10:47 |
| sneak | echo "test" | logger -t nsqd shows up in my logs | 10:49 |
| sneak | hrm, it's not showing up in "initctl list" | 10:58 |
| sneak | i stopped doing "restart nsqd" and tried "stop nsqd" and "start nsqd" | 10:59 |
| sneak | i think i need to get upstart to notice my nsqd.conf | 10:59 |
| xnox | sneak: as sudo...? | 11:06 |
| xnox | sneak: sudo initctl list | 11:06 |
| xnox | sneak: sudo initctl reload-configuration | 11:06 |
| xnox | sneak: also try $ init-checkconf /etc/init/nsqd.conf | 11:07 |
| sneak | i got it to show up by removing a bunch of lines from the conf | 11:07 |
| sneak | i didn't know about init-checkconf | 11:07 |
| sneak | and now it's logging to syslog | 11:07 |
| sneak | i wonder which one did it | 11:07 |
| sneak | there really could be some better logging going on there. init-checkconf says syntax ok and upstart logs nothing | 11:10 |
| sneak | very confusing | 11:10 |
| xnox | sneak: hm, you didn't paste full job to us, so it's hard to tell what you had written up. | 11:11 |
| sneak | you're right, the program should have told me the error | 11:19 |
| sneak | that's why i didn't paste it | 11:19 |
| xnox | sneak: full contents of /etc/init/nsqd.conf that is | 11:43 |
| ggayan | hey guys, I' | 14:58 |
| ggayan | I'm defining a service in CentOS 6.4 with upstart 0.6.5 | 14:58 |
| ggayan | but when I display the processes using ps | 14:58 |
| ggayan | I see two processes instead of one, it seems that exec isn't working | 14:59 |
| ggayan | https://gist.github.com/anonymous/76f54edea219eaf26889 | 15:01 |
| === JanC_ is now known as JanC | ||
| jamescarr | heyo | 20:22 |
| jamescarr | can I use usual bash type commands in upstart? like "/usr/bin/foo -e $(baz -d)" | 20:24 |
| jamescarr | with bas -d returning something like "239je23d | 20:24 |
| jamescarr | so it expands to /usr/bin/foo -e 239je23d | 20:24 |
| xnox | jamescarr: yes. | 22:18 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!