| [2010/08/27 00:36:26] @ Log started by gepetto | ||
| [2010/08/27 00:36:26] @ Quit: jhelwig: Excess Flood | ||
| [2010/08/27 00:37:46] @ jhelwig joined channel #puppet-dev | ||
| [2010/08/27 04:26:44] @ Quit: sejo: Read error: Operation timed out | ||
| [2010/08/27 04:28:50] @ sejo joined channel #puppet-dev | ||
| [2010/08/27 04:54:51] @ Quit: ohadlevy: Read error: Operation timed out | ||
| [2010/08/27 04:58:48] @ ohadlevy joined channel #puppet-dev | ||
| [2010/08/27 06:44:48] @ notbrien joined channel #puppet-dev | ||
| [2010/08/27 07:01:41] @ blkperl_ joined channel #puppet-dev | ||
| [2010/08/27 07:04:13] @ Quit: blkperl: Ping timeout: 252 seconds | ||
| [2010/08/27 07:12:39] @ rcrowley joined channel #puppet-dev | ||
| [2010/08/27 08:12:58] @ ohookins left channel #puppet-dev ("Leaving.") | ||
| [2010/08/27 08:22:03] @ Quit: notbrien: Quit: notbrien | ||
| [2010/08/27 08:23:25] @ notbrien joined channel #puppet-dev | ||
| [2010/08/27 08:24:21] @ Quit: notbrien: Client Quit | ||
| [2010/08/27 08:47:19] @ lak joined channel #puppet-dev | ||
| [2010/08/27 08:47:23] @ Quit: lak: Client Quit | ||
| [2010/08/27 10:17:13] @ lak joined channel #puppet-dev | ||
| [2010/08/27 10:29:38] @ Quit: lak: Quit: lak | ||
| [2010/08/27 11:27:17] @ lak joined channel #puppet-dev | ||
| [2010/08/27 11:50:58] @ Quit: lak: Quit: lak | ||
| [2010/08/27 12:24:00] <eric0> tmz: you there? i have a little tweak to the puppet init script in the srpm i'd like to run by you | ||
| [2010/08/27 12:25:42] <tmz> eric0: Yeah, what do you have? | ||
| [2010/08/27 12:26:43] <eric0> IMO the 'once' argument should stop any running puppetd, run with --no-splay as a hardcoded argument, then restart puppetd once its done -- that way you can use it to trigger an immediate execution | ||
| [2010/08/27 12:31:13] <tmz> So it only gets screwed up if you have splay = true and there's a run "in progress" I take it? | ||
| [2010/08/27 12:31:46] <tmz> How do we tell if there is a run that's simply waiting for the splay timeout versus midstream? | ||
| [2010/08/27 12:32:04] <eric0> tmz: not even in progress; it exits on the existence of the pid | ||
| [2010/08/27 12:32:24] <eric0> midstream it'll also have $statedir/puppetdlock | ||
| [2010/08/27 12:38:48] <tmz> eric0: So it just calls '$puppetd -o ${PUPPET_OPTS} ${PUPPET_EXTRA_OPTS} $@' | ||
| [2010/08/27 12:39:24] <tmz> If puppet is exiting there, it seems like forcing it to stop the running puppetd might not be so good. | ||
| [2010/08/27 12:39:56] <tmz> As in, this should be done in puppetd directly if it's worth doing, as it would then benefit other distros. | ||
| [2010/08/27 12:40:28] <tmz> I run puppetd -tv a lot on my systems while puppetd is running and unless it's midstream, it works fine for me. | ||
| [2010/08/27 12:40:39] <tmz> the difference is that you have splay = true, right? | ||
| [2010/08/27 12:40:45] <eric0> yeah | ||
| [2010/08/27 12:41:27] <tmz> I think it puppetd is set to exit when it finds that, it's something we don't want to override in the initscript. | ||
| [2010/08/27 12:42:08] <tmz> But it may be worth filing a ticket to see if it can be made smarter, to recognize that it was just waiting and to pause while the --onetime run proceeds. | ||
| [2010/08/27 12:45:02] <eric0> hm, so --onetime is supposed to ignore a running-but-idle (i.e. between runinterval, splayed or not) puppetd? | ||
| [2010/08/27 13:05:41] @ Quit: ricky: Ping timeout: 276 seconds | ||
| [2010/08/27 13:07:33] @ ricky joined channel #puppet-dev | ||
| [2010/08/27 13:50:57] <tmz> eric0: In my experience, using --onetime (well, via -t, which implies it), my running puppetd doesn't interfere. Except when it's mid-run. | ||
| [2010/08/27 13:51:27] <tmz> If splay is locking it longer, it might be considered a bug worthy of fixing. But I don't know how difficult that is to do properly. | ||
| [2010/08/27 13:51:45] <tmz> More important is not having two puppetd's stepping on each other, of course. | ||
| [2010/08/27 13:53:14] <Volcane> eric0: sadly only if its in the foreground, not --one-time in the background :( | ||
| [2010/08/27 13:53:46] <Volcane> to go background it needs to be able to make a pid | ||
| [2010/08/27 13:53:53] <Volcane> which the already running daemon would have | ||
| [2010/08/27 13:53:55] <tmz> So it just needs --no-daemonize perhaps. | ||
| [2010/08/27 13:54:04] <tmz> Then it'd be like -t. | ||
| [2010/08/27 13:54:13] <Volcane> limited use | ||
| [2010/08/27 13:54:40] <tmz> Which seems more like what the intent of the once option in the initscript would be? Or do folks use that in other ways, like via cron? | ||
| [2010/08/27 13:55:01] <Volcane> cron and mcollective | ||
| [2010/08/27 13:55:09] * tmz never uses service puppet once, when puppetd -t is shorter. | ||
| [2010/08/27 13:55:38] <tmz> So numerous valid use cases. Not the sort of thing we want to break then. :) | ||
| [2010/08/27 13:57:01] <tmz> Thanks for clarifying Volcane. I missed that just --onetime differed from --test by the lack of --no-daemonize. | ||
| [2010/08/27 13:57:45] <Volcane> i am sure i filed some feature request around this | ||
| [2010/08/27 13:57:48] <Volcane> trying to find it | ||
| [2010/08/27 14:00:38] <Volcane> no i think i am thinking about #3757 | ||
| [2010/08/27 14:00:41] <gepetto> Volcane: #3757 is http://projects.puppetlabs.com/issues/show/3757 "Puppet - Feature #3757: --enable and --disable should be improved - Puppet Labs" | ||
| [2010/08/27 14:00:47] <Volcane> unrelated | ||
| [2010/08/27 14:01:56] <tmz> Either way, thanks. You pointed out what I was missing. | ||
| [2010/08/27 14:54:07] @ lak joined channel #puppet-dev | ||
| [2010/08/27 15:10:44] @ Quit: lak: Quit: lak | ||
| [2010/08/27 15:19:46] <eric0> my other take on this same problem was #4624 | ||
| [2010/08/27 15:19:47] <gepetto> eric0: #4624 is http://projects.puppetlabs.com/issues/show/4624 "Puppet - Bug #4624: SIGUSR1 should ignore splay - Puppet Labs" | ||
| [2010/08/27 15:29:08] @ Quit: blkperl_: Read error: Connection reset by peer | ||
| [2010/08/27 15:29:11] @ blkperl joined channel #puppet-dev | ||
| [2010/08/27 15:40:18] @ lak joined channel #puppet-dev | ||
| [2010/08/27 16:56:49] @ github joined channel #puppet-dev | ||
| [2010/08/27 16:56:49] <github> facter: master Rein Henrichs * 32c0cb0 (3 files in 1 dirs): [#4594] Revert "fixes #2573, #2085, #1291..." ... | ||
| [2010/08/27 16:56:49] <github> facter: master Rein Henrichs * 44e5332 (3 files in 1 dirs): Merge branch 'bug/master/4594' | ||
| [2010/08/27 16:56:49] <github> facter: master Rein Henrichs * fca8861 (1 files in 1 dirs): [#4594] Reintroduce fix for #1291 from original patch ... | ||
| [2010/08/27 16:56:50] <github> facter: master Rein Henrichs * 3671c9f (1 files in 1 dirs): [#4583] Refactor uptime to use Resolution.exec ... | ||
| [2010/08/27 16:56:50] <github> facter: master Rein Henrichs * 244745e (1 files in 1 dirs): Merge branch 'ticket/master/4583' | ||
| [2010/08/27 16:56:50] <github> facter: master James Turnbull * ca2da36 (1 files in 1 dirs): Updated install.rb and created man page | ||
| [2010/08/27 16:56:50] <github> facter: master commits e7df4c0...ca2da36 - http://bit.ly/axOLyN | ||
| [2010/08/27 16:56:51] @ github left channel #puppet-dev () | ||
| [2010/08/27 16:58:36] @ github joined channel #puppet-dev | ||
| [2010/08/27 16:58:36] <github> facter: master James Turnbull * 3d287ac (1 files in 1 dirs): Updated CHANGELOG for 1.5.8 - http://bit.ly/cjHBem | ||
| [2010/08/27 16:58:36] @ github left channel #puppet-dev () | ||
| [2010/08/27 17:00:48] @ github joined channel #puppet-dev | ||
| [2010/08/27 17:00:48] <github> facter: master James Turnbull * ff473ef (1 files in 1 dirs): Updated signing rake task - http://bit.ly/9aLcaq | ||
| [2010/08/27 17:00:48] @ github left channel #puppet-dev () | ||
| [2010/08/27 17:38:59] @ Quit: lak: Quit: lak | ||
| [2010/08/27 17:57:54] @ paradox12 joined channel #puppet-dev | ||
| [2010/08/27 20:38:34] @ Quit: eric0: Ping timeout: 258 seconds | ||
| [2010/08/27 20:56:53] @ eric0 joined channel #puppet-dev | ||
| [2010/08/27 21:57:55] @ Quit: eric0: Ping timeout: 258 seconds | ||
| [2010/08/27 22:06:31] @ eric0 joined channel #puppet-dev | ||
| [2010/08/27 23:38:28] @ Quit: rcrowley: Quit: rcrowley |
Generated by irclog2html.py 2.6 by Marius Gedminas - find it at mg.pov.lt!