| === bradleyayers_ is now known as bradleyayers | ||
| j4m3s | SpamapS, you around? | 18:11 |
|---|---|---|
| j4m3s | is it possible to specify how many times a program will fork for the expect stanza? | 19:16 |
| SpamapS | j4m3s: I am now, sorry, whats up? | 19:19 |
| SpamapS | j4m3s: no, you can do 1 (expect fork) or 2 (expect daemon) | 19:19 |
| SpamapS | j4m3s: any reasonable program should only ever fork that many times to become the eventual master process anyway | 19:19 |
| j4m3s | bigcouch forks 33 times before | 19:19 |
| SpamapS | I doubt that | 19:19 |
| SpamapS | j4m3s: those are probably threads | 19:19 |
| SpamapS | the advice in the cookbook isn't entirely accurate.. I'm convinced the clone bits need to be re-evaluated | 19:20 |
| j4m3s | well i used the cookbook to determine the fork count | 19:21 |
| j4m3s | SpamapS, this is what i have so far | 19:22 |
| j4m3s | http://pastie.org/private/okibfp83s2oi5lb6emag | 19:22 |
| SpamapS | j4m3s: erts should have a "don't daemonize" flag somewhere.. | 19:23 |
| SpamapS | err, not erts, but erl | 19:23 |
| j4m3s | SpamapS, so if i start a process in pre-start and i want to stop it in post-stop, but it's not being killed. what am i doing wrong? http://pastie.org/private/apk4xedro2xtrbinygsg | 20:25 |
| SpamapS | j4m3s: that looks like it should, in theory, work | 20:26 |
| j4m3s | because it tells me it stop the job, and the main process is stopped but the other one still exists | 20:27 |
| SpamapS | j4m3s: btw, your start on is crap. "runlevel [2345]" is what you want (Unless you're on a release of Ubuntu before 11.10, then you want 'net-device-up IFACE=whateveryourethXis' | 20:27 |
| j4m3s | lol | 20:27 |
| SpamapS | j4m3s: pgrep isn't exactly an exact science | 20:27 |
| SpamapS | j4m3s: perhaps log the pid you try to kill | 20:28 |
| j4m3s | SpamapS, can you tell me if this variable usage is not allowed in upstart 0.6.5? http://pastie.org/private/wk1pteb9vmdkbmepkicq | 21:12 |
| j4m3s | specifically line 21 & 22 | 21:12 |
| j4m3s | ie. it works with line 22 and doesn't work with line 21 | 21:13 |
| SpamapS | j4m3s: there is no expansion in env statements, even in the latest upstart | 21:16 |
| j4m3s | SpamapS, that answers that! thx again! | 21:16 |
| j4m3s | SpamapS, does that mean i couldn't do var=$foo:/usr/bin ; env PATH=$var ? | 21:18 |
| SpamapS | j4m3s: correct, that will not work | 21:27 |
| j4m3s | SpamapS, will variable expansion work in an exec stanza? | 21:39 |
| SpamapS | j4m3s: yes | 21:43 |
| SpamapS | exec and script are fine | 21:43 |
| SpamapS | and start/stop on too | 21:43 |
| SpamapS | and instance | 21:43 |
| j4m3s | SpamapS, so i am trying to use $EMU in the exec and it's not being set properly any ideas? http://pastie.org/private/ti12eytjhg23h07evyohww | 21:47 |
| SpamapS | j4m3s: in a script statement, you don't want to say 'env XXX...' | 22:01 |
| SpamapS | j4m3s: just XXX=... | 22:01 |
| SpamapS | j4m3s: and you don't need to export it | 22:01 |
| j4m3s | SpamapS, ok gotcha | 22:01 |
| SpamapS | j4m3s: as the shell that you're setting it in, is the one evaluating the exec line | 22:01 |
| j4m3s | SpamapS, can you have two execs in a script statement? | 22:03 |
| SpamapS | j4m3s: sure, but only one will happen because exec replaces the current process | 22:04 |
| SpamapS | j4m3s: you can background stuff tho | 22:04 |
| SpamapS | foo1 & exec foo2 | 22:04 |
| j4m3s | SpamapS, is it possible to do something like TMP=`env VAR=$foo` within a script statement? | 23:09 |
| j4m3s | SpamapS, is there another way i can set an environment variable in that shell that can handle variable expansion? | 23:11 |
| j4m3s | SpamapS, http://pastie.org/private/df0a1urpr2ajq1dslzjaq | 23:12 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!