[10:27] <tstone> hi, i have just added upstart to ptxdist. The problem is that dbus is running but not responing to any connection any more?! I only change the init method...
[14:55] <pgmcgee> hey guys, im having issues with writing an upstart script for nginx, nginx starts up fine, but the pid upstart gets for nginx is wrong and so i can't stop or restart it
[15:08] <soren> Did you try "expect daemon"?
[15:08] <soren> pgmcgee: ^
[15:09] <soren> pgmcgee: See the init(5) man page for details.
[15:10] <soren> pgmcgee: Ah, make that "expect fork", apparantly:
[15:10] <soren> http://cyruslopez.net/writing/upstart.html
[15:14] <pgmcgee> soren: i did try expect fork... hmm
[15:15] <soren> Try expect daemon then.
[15:15]  * soren wanders off
[15:16] <pgmcgee> soren: yep, trying that now
[15:19] <pgmcgee> hmm.. on expect daemon, the start and stop commands just hang (don't give any output and don't exit) and nginx keeps running even on stop
[16:05] <halfline> sadmac: if i have one thing that does "stop on starting runlevel" and one thing that does "start on runlevel"
[16:05] <halfline> sadmac: the stop on starting runlevel job should be synchronously stopped before the start on runlevel job is run right?
[16:05] <sadmac> halfline: runlevel is an event, not a job (in fedora's stack anyway) so there is no starting runlevel
[16:06] <halfline> sadmac: oh, there's no way to trigger on the edge before the level? 
[16:07] <sadmac> halfline: try working relative to the rc job rather than the runlevel event
[16:07] <halfline> sadmac: basically i need prefdm.conf to finish completely before runlevel 5's rc scripts start shutting down
[16:07] <sadmac> halfline: start on starting rc
[16:07] <halfline> so you saying "stop on starting rc" ?
[16:07] <sadmac> yeah
[16:08] <halfline> sadmac: so if i do that
[16:08] <halfline> sadmac: and then do "start on stopped prefdm" for a task
[16:09] <halfline> will that task get run synchronously before rc is run?
[16:09] <sadmac> halfline: don't think so. start on stopping prefdm would do that though, but you'd run before prefdm
[16:10] <sadmac> before prefdm stopped
[16:10] <halfline> is there a way to hook in after prefdm is stopped before rc is started?
[16:11] <halfline> can i put in my stopping script some sort of barrier?
[16:11] <halfline> initctl wait on prefdm stopped  or something?
[16:12] <sadmac> halfline: start on starting rc and stopped prefdm
[16:12] <sadmac> halfline: you are tapping into overwhelming evil with that line but it should technically work
[16:13] <halfline> hmm i guess i need to filter that a bit though
[16:13] <halfline> start on starting rc RUNLEVEL=0 or RUNLEVEL=6 and stopped prefdm ?
[16:13] <sadmac> right
[16:14] <sadmac> well, wrong
[16:14] <halfline> i can just put an "or" like that?
[16:14] <sadmac> start on starting rc RUNLEVEL=[06] and...
[16:14]  * halfline tries
[16:18] <halfline> hmm seems to just be sitting there instead of rebooting
[16:19] <sadmac> hmm.
[16:19] <halfline> http://fpaste.org/mmVo/
[16:19] <halfline> so prefdm seems to be caught up on stop on startng rc RUNLEVEL=5
[16:20] <halfline> ohh
[16:20] <halfline> that should be stop on starting rc RUNLEVEL=[!5]
[16:20] <sadmac> bingo
[16:22] <halfline> seems to work
[16:22] <halfline> thanks
[16:44] <halfline> sadmac: hmm
[16:45] <halfline> sadmac: so if i boot into runlevel 3 and type reboot things go awry
[16:45] <halfline> it's just sitting there
[16:45] <halfline> http://fpaste.org/SfYh/
[16:47] <halfline> sadmac: could it be stop on starting rc RUNLEVEL=[!5] is cuasing problems if prefdm is already stopped?
[16:47] <halfline> (that line is in prefdm.conf)
[16:48] <sadmac> halfline: no, should be fine
[16:48] <halfline> hmm if i do "start prefdm"
[16:48] <halfline> and then type reboot again
[16:49] <sadmac> its the and line thats creating the problem when prefdm is already stopped
[16:49] <halfline> it reboots
[16:49] <sadmac> halfline: I have to run out though. bbiab
[16:49] <halfline> k
[21:48] <JayT> Is hwclock startup script necessary for an Ubuntu system to boot? if my RTC is screwed (perhaps because of a dead CMOS battery), can I do 'update-rc.d -f hwclock remove' (and the same for hwclockfirst), install ntp, set it up, and reboot?
[21:49] <JayT> on hardy that is
[22:18] <JanC> JayT: that sounds like an Ubuntu question, and not an upstart question  ;)