| keesj | hi | 08:56 |
|---|---|---|
| keesj | how is halt/shutown to be implemented? | 09:36 |
| brendan_ | keesj: here's what i do on rc0/shutdown: | 14:44 |
| brendan_ | # rc0 - runlevel 0 compatibility | 14:44 |
| brendan_ | # | 14:44 |
| brendan_ | # This task runs the old sysv-rc runlevel 0 ("halt/poweroff") scripts with | 14:44 |
| brendan_ | # the decision as to whether to halt or power off the system left up to the | 14:44 |
| brendan_ | # script (and thus /etc/default/poweroff). | 14:44 |
| brendan_ | start on runlevel 0 | 14:44 |
| brendan_ | console output | 14:44 |
| brendan_ | script | 14:44 |
| brendan_ | echo "Sending all processes the TERM signal" | 14:44 |
| brendan_ | killall5 -15 | 14:44 |
| brendan_ | sleep 5 | 14:44 |
| brendan_ | echo "Sending all processes the KILL signal" | 14:44 |
| brendan_ | killall5 -9 | 14:44 |
| brendan_ | sleep 1 | 14:44 |
| brendan_ | /sbin/halt -f -p | 14:44 |
| brendan_ | end script | 14:44 |
| brendan_ | you might want a little more in there than that, of course | 14:45 |
| AlexExtreme | if you have that in a script called 'shutdown', you can make other scripts stop before that ones gets to run by putting 'stop on starting shutdown' in them | 14:55 |
| jdong | keesj: shutdown has to be implemented by you | 15:09 |
| jdong | in short | 15:09 |
| jdong | I do similar to brendan_, hook the sysv shutdown/reboot handlers into shutdown or reboot jobs. | 15:10 |
| mdales | I'm doing some tests, and just wanted this confirmed if possible - when a job with respawn fails and is restarted, there's no stopped failed signal emitted is there? | 17:19 |
| mdales | it's only when it's failed a number of times and upstart gives up that the stopped failed is emitted? | 17:20 |
| keesj | jdong: thank , I was afraid of that | 19:23 |
| keesj | I don't have sysv installed at al , no shutdown etc :p | 19:26 |
| keesj | and the shutdown/halt are booth provided by the upstart sysv compatibility layer | 19:28 |
| brendan_ | mdales: i believe so. you could put a sleep in your event script then watch initctl status to see the event status as it's respawning | 19:45 |
| brendan_ | keesj: you could just have shutdown and reboot as shell scripts that would call "initctl start shutdown" "initctl start restart" and just write event scripts for those | 19:47 |
| keesj | but what command would perform the "real" shudown/reboot | 19:51 |
| keesj | I guess I really need to understand how that works. thanks for the help | 19:52 |
| brendan_ | oh right, you'd be missing halt | 20:05 |
| brendan_ | duh | 20:05 |
| brendan_ | doing too many things at once, sorry :) | 20:06 |
| === cdahlin is now known as sadmac2 | ||
| keesj | :p | 20:43 |
| === keesj is now known as k-way | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!