| [2010/07/18 08:09:16] @ Log started by gepetto | ||
| [2010/07/18 08:09:16] @ rcrowley joined channel #puppet-dev | ||
| [2010/07/18 08:58:49] @ Quit: markus: Quit: markus | ||
| [2010/07/18 09:06:40] @ markus joined channel #puppet-dev | ||
| [2010/07/18 09:17:17] @ Quit: markus: Quit: markus | ||
| [2010/07/18 09:57:46] @ markus joined channel #puppet-dev | ||
| [2010/07/18 11:18:14] @ jmccune joined channel #puppet-dev | ||
| [2010/07/18 11:53:10] <masterzen> markus: just sent an answer to your e-mail in puppet-dev. I'm concerned about the compilation time that is way greater under 2.6 than it was on 0.25.5. | ||
| [2010/07/18 11:53:37] <masterzen> markus: I didn't pay attention when testing this week, but I re-ran 0.25.5 and clearly see the difference... | ||
| [2010/07/18 12:02:37] <duritong> masterzen: do you have some ideas, where I could put debug statements to debug the issue | ||
| [2010/07/18 12:03:26] <masterzen> duritong: I put some benchmark(:notice, "debug msg") { ... } in Puppet::Parser::Compiler#compile | ||
| [2010/07/18 12:03:40] <masterzen> duritong: my issue is in evaluate_definitions | ||
| [2010/07/18 12:03:45] <duritong> masterzen: ok | ||
| [2010/07/18 12:03:50] <masterzen> duritong: but I don't see the import issue you have | ||
| [2010/07/18 12:03:55] <masterzen> duritong: only the slowdown | ||
| [2010/07/18 12:05:57] <duritong> I can reproduce the import issue for every run | ||
| [2010/07/18 12:07:23] <duritong> and I don't see what I could probably do wrong | ||
| [2010/07/18 12:09:05] <duritong> masterzen: I'm looking for the point where puppet tries to import things or not | ||
| [2010/07/18 12:09:33] <masterzen> duritong: ok. I'm going offline for a couple of hours | ||
| [2010/07/18 12:15:30] <duritong> masterzen: ok | ||
| [2010/07/18 12:29:26] <duritong> markus: if you have an idea where I could put some debug statements to see why I'm autoimporting class? | ||
| [2010/07/18 12:37:50] <markus> masterzen (if you're still here) do you see 10x slowdown? | ||
| [2010/07/18 12:39:32] <markus> duritong: I'll look for strategic points to put checks. | ||
| [2010/07/18 12:48:16] @ lak joined channel #puppet-dev | ||
| [2010/07/18 12:51:15] @ Quit: jmccune: Ping timeout: 246 seconds | ||
| [2010/07/18 12:56:36] <markus> masterzen: thank you for some excellent leg work. I'm going to try to concoct a fix. | ||
| [2010/07/18 13:18:28] @ jmccune joined channel #puppet-dev | ||
| [2010/07/18 13:22:08] <markus> duritong? | ||
| [2010/07/18 13:25:44] <duritong> markus: ? | ||
| [2010/07/18 13:26:10] <markus> Hi. | ||
| [2010/07/18 13:27:15] <markus> I've got a branch up on github (MarkusQ:ticket/master/4278) that addresses what I suspect is causing some of the slowdown Brice reported (and which may ultimately lead us to the autoloading problem). | ||
| [2010/07/18 13:27:30] <duritong> cool | ||
| [2010/07/18 13:27:42] <duritong> I'm going to test it | ||
| [2010/07/18 13:27:52] <markus> It's based off master (without the additional patches). Would it be possible for you to try it and see if it has any speed impact? | ||
| [2010/07/18 13:28:05] <duritong> I can | ||
| [2010/07/18 13:28:14] <markus> If needed, I can make a version off of testing (with all the patches) | ||
| [2010/07/18 13:28:37] <duritong> I think it would be good to have a version off of testing | ||
| [2010/07/18 13:28:45] <markus> Coming up | ||
| [2010/07/18 13:28:51] <duritong> because without the patches in testing my manifests won't compile | ||
| [2010/07/18 13:30:16] <markus> It's at MarkusQ:ticket/testing/4278 | ||
| [2010/07/18 13:30:20] <duritong> ok | ||
| [2010/07/18 13:31:33] * duritong building rpm | ||
| [2010/07/18 13:31:48] <duritong> it's strange that you can't reproduce my problem | ||
| [2010/07/18 13:33:50] <markus> Yeah. On the other hand, if they aren't strange either I never see them or they resolve quickly, so there's some sample bias. | ||
| [2010/07/18 13:34:20] <markus> I tend to spend most my days scratching my head over strange ones. | ||
| [2010/07/18 13:42:41] <duritong> as long as it doesn't get boring... | ||
| [2010/07/18 13:42:49] * duritong has kinda flacky connection | ||
| [2010/07/18 13:45:14] <duritong> markus: first run 1.17s, second run 0.29 seconds | ||
| [2010/07/18 13:45:30] <duritong> markus: but I still get the info: Automatically imported lighttpd from lighttpd | ||
| [2010/07/18 13:45:57] <markus> But this mostly resolves the speed issue? | ||
| [2010/07/18 13:46:04] <markus> Or at least partially? | ||
| [2010/07/18 13:46:55] <duritong> I'm testing it now with a big manifest | ||
| [2010/07/18 13:47:30] <markus> Ok. I'm looking to see where to add debug statements to ferret out the autoloading part. | ||
| [2010/07/18 13:47:57] <jamesturnbull> #4278 | ||
| [2010/07/18 13:47:58] <gepetto> jamesturnbull: #4278 is http://projects.puppetlabs.com/issues/show/4278 "Puppet - Bug #4278: Performance hole in 2.6.0 - Puppet Labs" | ||
| [2010/07/18 13:48:31] <jamesturnbull> #4268 | ||
| [2010/07/18 13:48:32] <gepetto> jamesturnbull: #4268 is http://projects.puppetlabs.com/issues/show/4268 "Puppet - Bug #4268: puppet master 2.6 always imports manifests newly - Puppet Labs" | ||
| [2010/07/18 13:48:47] <markus> jamesturnbull yes, what about them? | ||
| [2010/07/18 13:48:59] <duritong> markus: before my first compile was 110s now it is 44 | ||
| [2010/07/18 13:49:11] <duritong> let's see how long the second one will take | ||
| [2010/07/18 13:49:31] <jamesturnbull> markus: I was being lazy | ||
| [2010/07/18 13:50:12] <markus> Ah, using gepetto to look things up for you? | ||
| [2010/07/18 13:50:16] <jamesturnbull> markus: yeah | ||
| [2010/07/18 13:55:41] <masterzen> markus: your patch on top of master is better but still not on par with 0.25.5. I got a speedup of 2 compared as straight master | ||
| [2010/07/18 13:55:55] <jamesturnbull> yeah that feels like what I got | ||
| [2010/07/18 13:56:22] <markus> masterzen: yeah, I'm working on the other instances of it. | ||
| [2010/07/18 13:57:06] <markus> I put it out for testing as soon as I had part of it to see if I was on the right track. | ||
| [2010/07/18 13:57:37] <masterzen> markus: you're definitely on the right track. | ||
| [2010/07/18 13:58:17] <markus> masterzen: *smile* Thanks to you. | ||
| [2010/07/18 13:58:39] <masterzen> markus: it still seems that evaluating definitions is taking most of the time | ||
| [2010/07/18 13:58:49] <markus> Yes. | ||
| [2010/07/18 13:59:00] <duritong> markus: yeah, the second one still 41s, as it still re-imports all the modules | ||
| [2010/07/18 13:59:22] <markus> We've effectively got an O(n^2) problem with defined resources. | ||
| [2010/07/18 13:59:39] <markus> Twice as many takes 4x as long, roughly. | ||
| [2010/07/18 13:59:43] <jamesturnbull> duritong: so you're running the client and then running it again? | ||
| [2010/07/18 14:00:15] <duritong> jamesturnbull: yes | ||
| [2010/07/18 14:00:32] <duritong> jamesturnbull: and I did not change anything between | ||
| [2010/07/18 14:00:40] <duritong> and no other client ran against it | ||
| [2010/07/18 14:03:00] <jamesturnbull> markus: testing HEAD is broken for me BTW | ||
| [2010/07/18 14:03:16] <markus> Broken how? | ||
| [2010/07/18 14:03:27] <jamesturnbull> markus: http://pastie.org/1049684 | ||
| [2010/07/18 14:04:16] <markus> jamesturnbull: "debug: Puppet::Type::File::ProviderPosix: feature posix is missing" | ||
| [2010/07/18 14:04:20] <markus> Ouch! | ||
| [2010/07/18 14:04:25] <markus> What OS are you on? | ||
| [2010/07/18 14:05:29] <jamesturnbull> Fedora | ||
| [2010/07/18 14:05:45] <jamesturnbull> though now after re-installing puppet I get weirdness | ||
| [2010/07/18 14:06:55] <markus> So last I heard Fedora was posix. Any idea why it isn't recognizing that fact? | ||
| [2010/07/18 14:07:45] <jamesturnbull> markus: uninstalling and reinstalling fixed that issue | ||
| [2010/07/18 14:07:46] <jamesturnbull> but now | ||
| [2010/07/18 14:07:47] <masterzen> markus: yes, we certainly have an O(n2) issue. sqrt(16) is roughly what I have on 0.25.x | ||
| [2010/07/18 14:07:49] <jamesturnbull> http://pastie.org/1049693 | ||
| [2010/07/18 14:08:10] <jamesturnbull> it's dying and not going into the foreground | ||
| [2010/07/18 14:08:53] <markus> masterzen: extended fix in the works. | ||
| [2010/07/18 14:09:11] <masterzen> markus: late here, might be able to test only tomorrow evening :( | ||
| [2010/07/18 14:09:14] <markus> jamesturnbull: the posix test seems to be" | ||
| [2010/07/18 14:09:15] <markus> 38 # We can use POSIX user functions | ||
| [2010/07/18 14:09:15] <markus> 39 Puppet.features.add(:posix) do | ||
| [2010/07/18 14:09:16] <markus> 40 require 'etc' | ||
| [2010/07/18 14:09:16] <markus> 41 Etc.getpwuid(0) != nil && Puppet.features.syslog? | ||
| [2010/07/18 14:09:16] <markus> 42 end | ||
| [2010/07/18 14:09:37] <jamesturnbull> markus: I think it's a null issue | ||
| [2010/07/18 14:09:38] <markus> masterzen: I understand. You've been a huge help, and thanks yet again. | ||
| [2010/07/18 14:09:51] <markus> No syslog? | ||
| [2010/07/18 14:09:57] <masterzen> markus: what do you think is the issue? | ||
| [2010/07/18 14:10:10] <jamesturnbull> no I think I had some 0.25.5 code in the RUBYLIB | ||
| [2010/07/18 14:11:08] <markus> masterzen: when we go through the resources to see which are evaluated, we effectively go through all the defined resources for each defined resource, though this isn't obvious | ||
| [2010/07/18 14:11:49] <markus> We need to do some targeted caching / short circuiting and should get it down to ~O(n) | ||
| [2010/07/18 14:12:11] <markus> One pass through to see what we need to do, and one to do it, both O(n) | ||
| [2010/07/18 14:12:32] <masterzen> markus: ok, I don't really understand the compiler... I was thinking there could be some issues in Puppet::Parser::Resource::Type#evaluate_parent_type | ||
| [2010/07/18 14:13:04] <masterzen> markus: I don't think the eval def code changed that much between 0.25 and 2.6 | ||
| [2010/07/18 14:13:35] <markus> No, but I'm wanting to cut out the send overhead while I'm at it. | ||
| [2010/07/18 14:14:00] <markus> Send is slow in MRI, and even slower in JRuby (kills branch prediction caching). | ||
| [2010/07/18 14:14:19] <markus> And there's no need for it, since it's always doing the same thing. | ||
| [2010/07/18 14:14:25] <masterzen> markus: I think the problem is when we evaluate the code of said definition. that's a wild-guess though :) | ||
| [2010/07/18 14:14:36] <masterzen> markus: sure, you're right about send. | ||
| [2010/07/18 14:15:39] <jamesturnbull> markus: so one error to another | ||
| [2010/07/18 14:15:49] <jamesturnbull> markus: the master HEAD single binary works | ||
| [2010/07/18 14:15:51] <markus> I don't think it's when we evaluate so much as when we check to see if we need to evaluate (though any sort of property check seems to be at risk). | ||
| [2010/07/18 14:15:57] <jamesturnbull> markus: the testing HEAD one does not | ||
| [2010/07/18 14:16:27] <duritong> jamesturnbull: mine does | ||
| [2010/07/18 14:16:31] <markus> jamesturnbull: can you bisect to see which of the ~5 patches broke it? | ||
| [2010/07/18 14:16:47] <jamesturnbull> duritong: on testing? | ||
| [2010/07/18 14:16:54] <markus> Or email me enough info that I could? | ||
| [2010/07/18 14:17:12] <jamesturnbull> oh wait | ||
| [2010/07/18 14:17:18] <jamesturnbull> f*cking hell | ||
| [2010/07/18 14:17:57] <duritong> jamesturnbull: on testing + markus wip patch | ||
| [2010/07/18 14:18:18] <jamesturnbull> duritong: markus yeah I'm going to blow away this VM - it's borked seemingly | ||
| [2010/07/18 14:18:29] @ Quit: jmccune: Ping timeout: 265 seconds | ||
| [2010/07/18 14:18:34] <markus> new speed patch coming up. | ||
| [2010/07/18 14:20:05] @ jmccune joined channel #puppet-dev | ||
| [2010/07/18 14:20:42] <masterzen> markus: I'm wondering the following (still following my idea). when we evaluate definition code, we try to find the parent resource (either a class or a node). Each time we end up in the catalog where we create a new resource ref to canonicalize thing (which we weren't doing in 0.25). Don't you think creating those tons of objects could be the issue? | ||
| [2010/07/18 14:20:50] <markus> duritong: I've updated MarkusQ:ticket/testing/4278 with an extension of the speed fix. | ||
| [2010/07/18 14:21:20] <markus> masterzen: Yeah, that has the same shape as the other, but on depth of the parent tree. | ||
| [2010/07/18 14:21:45] @ Quit: jmccune: Read error: Connection reset by peer | ||
| [2010/07/18 14:21:52] <markus> Caching them could be a bug plus as well. | ||
| [2010/07/18 14:21:57] <duritong> markus: rebuilding packages | ||
| [2010/07/18 14:23:26] <markus> s/bug/big/ | ||
| [2010/07/18 14:25:22] <duritong> markus: yeah | ||
| [2010/07/18 14:25:43] <markus> yeah? | ||
| [2010/07/18 14:25:53] <duritong> caching a big plus | ||
| [2010/07/18 14:25:56] <duritong> sorry | ||
| [2010/07/18 14:26:05] <markus> Ah, yeah. | ||
| [2010/07/18 14:26:20] <masterzen> markus: your latest patch is a little bit better: 12s instead of 16s (previous patch) | ||
| [2010/07/18 14:27:25] <markus> masterzen: So parent_resources... | ||
| [2010/07/18 14:27:54] <masterzen> markus: especially the resource ref we create on each catalog resource lookup :) | ||
| [2010/07/18 14:28:06] <markus> Yep. | ||
| [2010/07/18 14:28:56] <markus> Though I'm not sure the object-creation / GC is as bad as the call tree aspects. | ||
| [2010/07/18 14:29:35] <masterzen> markus: yes, I'm certainly wrong, we're also doing this on 0.25.5. | ||
| [2010/07/18 14:29:43] <masterzen> markus: I | ||
| [2010/07/18 14:29:55] <masterzen> markus: I guess, I'm too tired :) | ||
| [2010/07/18 14:30:16] <markus> No worries. It also looks like parent_type is already cached. | ||
| [2010/07/18 14:30:30] <markus> I'll keep looking. | ||
| [2010/07/18 14:34:45] <jamesturnbull> hmmm can replicate #4255 now whereas I couldn't before | ||
| [2010/07/18 14:34:46] <gepetto> jamesturnbull: #4255 is http://projects.puppetlabs.com/issues/show/4255 "Puppet - Bug #4255: Puppet 2.6 'keeps setting content' on files - Puppet Labs" | ||
| [2010/07/18 14:35:28] <masterzen> guys, I'm going to sleep. See you tomorrow. Good luck for the bug hunt! | ||
| [2010/07/18 14:35:39] <duritong> masterzen: good night | ||
| [2010/07/18 14:35:46] <jamesturnbull> masterzen: thanks mate - much appreciated | ||
| [2010/07/18 14:36:59] @ jmccune joined channel #puppet-dev | ||
| [2010/07/18 14:36:59] <markus> duritong: Any word on the speed of the revised branch? (masterzen was 25% faster) | ||
| [2010/07/18 14:37:58] <duritong> markus: I didn't get any speedup with the second improvement | ||
| [2010/07/18 14:38:11] <markus> Ok, so that's a dry hole. | ||
| [2010/07/18 14:38:12] <duritong> markus: but I'm verifing things | ||
| [2010/07/18 14:38:16] <markus> Ok. | ||
| [2010/07/18 14:38:31] <duritong> markus: but that's a bit difficult with the flacky connection here | ||
| [2010/07/18 14:38:56] <markus> Yeah, I was remote last week. You have my sympathy, | ||
| [2010/07/18 14:42:04] @ Quit: lak: Quit: lak | ||
| [2010/07/18 14:42:50] <duritong> markus I just verified that I really have your latest changes installed and they didn't help for me | ||
| [2010/07/18 14:43:09] <duritong> the first change improved speed by factor 3 | ||
| [2010/07/18 14:43:24] <duritong> so only ~40s against ~110s | ||
| [2010/07/18 14:43:28] <markus> Ok. I just sent you a patch that will produce additional feedback on the automating loading issue. | ||
| [2010/07/18 14:43:43] <duritong> ok | ||
| [2010/07/18 14:43:52] <markus> I'm going to take a brief break (~15 min) and then start looking for other instances of the speed issue. | ||
| [2010/07/18 14:44:06] <duritong> ok trying to get it running | ||
| [2010/07/18 14:44:46] <markus> If you could run the test case again & email me the server output for when it handles the first & second client runs. | ||
| [2010/07/18 14:45:26] <duritong> I'll do | ||
| [2010/07/18 14:45:31] <duritong> with --debug ? | ||
| [2010/07/18 14:45:53] <duritong> ah it is a simple puts | ||
| [2010/07/18 14:45:54] <duritong> ok | ||
| [2010/07/18 14:58:54] <markus> duritong: back | ||
| [2010/07/18 14:59:13] <duritong> markus: fetching the data | ||
| [2010/07/18 15:11:18] <markus> duritong: Also, a dumb question, but I should probably ask: are you sure there isn't anything touching your module files (e.g. changing the time stamps) between client runs? | ||
| [2010/07/18 15:12:24] <duritong> markus: I also thought about that, but I can't come up with anything, as I don't check in any code | ||
| [2010/07/18 15:12:38] <markus> ? | ||
| [2010/07/18 15:12:52] <duritong> which would change time stamps | ||
| [2010/07/18 15:12:58] <markus> You could just look and see if the timestamps are changing, no? | ||
| [2010/07/18 15:14:05] <duritong> and also the manifest setup I sent you was running in a dedicated folder which shouldn't have been touched at all | ||
| [2010/07/18 15:14:11] <duritong> but yeah we can check it | ||
| [2010/07/18 15:14:54] @ Quit: ricky: Read error: Connection reset by peer | ||
| [2010/07/18 15:16:04] <duritong> markus: I pasted the url to the output in a query to you | ||
| [2010/07/18 15:16:13] @ ricky joined channel #puppet-dev | ||
| [2010/07/18 15:16:23] <duritong> markus would ls -l be sufficient to check timestamps checkings? | ||
| [2010/07/18 15:16:28] <markus> duritong: the reason I ask, the only way I could duplicate anything that looked like the problem was touching the files in a bash for-loop. | ||
| [2010/07/18 15:16:34] <markus> Looking at the output now... | ||
| [2010/07/18 15:17:49] @ lak joined channel #puppet-dev | ||
| [2010/07/18 15:19:35] @ Quit: nasrat: Quit: ... | ||
| [2010/07/18 15:23:50] <duritong> markus: ls -l on a file that is imported shows a date in Sept. 2009 | ||
| [2010/07/18 15:24:07] <markus> So no joy there. | ||
| [2010/07/18 15:24:15] <duritong> I don't think so | ||
| [2010/07/18 15:25:50] <duritong> and it would also be surprising as I don't have that behavior with 0.2.5x | ||
| [2010/07/18 15:32:35] <duritong> markus: the conference is ending and I'll moving around shortly. I will be online later. | ||
| [2010/07/18 15:32:42] <duritong> thanks for looking into things | ||
| [2010/07/18 15:32:45] <markus> Ok. | ||
| [2010/07/18 15:32:55] <duritong> and just let me know if I should test further things | ||
| [2010/07/18 15:32:59] <markus> I'm not seeing anything too helpful, but one odd bit... | ||
| [2010/07/18 15:33:18] <markus> I don't see anything like these lines: | ||
| [2010/07/18 15:33:19] <markus> 6094 info: Could not find filesystem info for file 'modules/common/modules_dir' in environment production | ||
| [2010/07/18 15:33:32] <markus> And you seem to have a fair number of them | ||
| [2010/07/18 15:34:07] <markus> Any idea what they are? | ||
| [2010/07/18 15:34:27] <markus> I'll look into it. | ||
| [2010/07/18 15:34:29] <markus> ttyl | ||
| [2010/07/18 15:34:31] <duritong> non existing directories in the modules | ||
| [2010/07/18 15:34:36] <duritong> I'll fix them | ||
| [2010/07/18 15:34:39] <markus> Ok. | ||
| [2010/07/18 15:34:45] <markus> Probably not it then. | ||
| [2010/07/18 15:35:03] <duritong> but they are actually only a problem when the client applies the catalog | ||
| [2010/07/18 15:35:17] <duritong> they don't appear earlier | ||
| [2010/07/18 15:36:09] <markus> Another observation: I think the remaining speed issue is the import replication. | ||
| [2010/07/18 15:36:33] <duritong> I think I agree with that | ||
| [2010/07/18 15:36:39] <markus> And another: it looks like explicit imports (e.g. common.pp) are repeating as well | ||
| [2010/07/18 15:37:00] <markus> At least, from the debug lines. | ||
| [2010/07/18 15:37:04] <markus> I don't see that, | ||
| [2010/07/18 15:37:10] <markus> when I run it. | ||
| [2010/07/18 15:37:27] <duritong> we'll also look at that | ||
| [2010/07/18 15:37:28] <markus> I'm going to take a walk-around-and-clear-the-cobwebs break myself, and then try to concoct another debug patch. | ||
| [2010/07/18 15:37:44] <markus> Any idea when you'll be on next? | ||
| [2010/07/18 15:37:51] <duritong> cool, thanks! | ||
| [2010/07/18 15:37:54] <duritong> have fun | ||
| [2010/07/18 15:41:08] <duritong> hmm, no I can't really tell | ||
| [2010/07/18 15:41:25] <duritong> but I will be on the whole day tomorrow and work on different things | ||
| [2010/07/18 15:41:38] <duritong> but I will also have time to do things you come up | ||
| [2010/07/18 15:42:05] <duritong> and you can also send me always a list of ideas I whould try out and I can report back by mail | ||
| [2010/07/18 15:42:32] <duritong> cu later | ||
| [2010/07/18 15:43:20] @ Quit: lak: Quit: lak | ||
| [2010/07/18 16:04:20] @ Quit: jmccune: Ping timeout: 265 seconds | ||
| [2010/07/18 16:06:02] @ jmccune joined channel #puppet-dev | ||
| [2010/07/18 18:43:36] @ Quit: rcrowley: Quit: rcrowley | ||
| [2010/07/18 18:54:31] @ Quit: jmccune: Ping timeout: 258 seconds | ||
| [2010/07/18 19:17:55] @ lak joined channel #puppet-dev | ||
| [2010/07/18 19:23:57] @ jmccune joined channel #puppet-dev | ||
| [2010/07/18 19:31:13] @ Quit: lak: Quit: lak | ||
| [2010/07/18 19:31:55] @ Quit: jmccune: Ping timeout: 260 seconds | ||
| [2010/07/18 19:33:45] @ jmccune joined channel #puppet-dev | ||
| [2010/07/18 19:47:32] @ Quit: jmccune: Ping timeout: 248 seconds | ||
| [2010/07/18 19:49:37] @ jmccune joined channel #puppet-dev | ||
| [2010/07/18 19:58:40] @ github joined channel #puppet-dev | ||
| [2010/07/18 19:58:40] <github> puppet: master Brice Figureau * dcf094a (2 files in 2 dirs): Fix #4262 - Puppetmaster used to log compilation time ... | ||
| [2010/07/18 19:58:41] <gepetto> github: puppet: #4262 is http://projects.puppetlabs.com/issues/show/4262 "Puppet - Bug #4262: Puppetmaster doesn't log the compilation time anymore - Puppet Labs" | ||
| [2010/07/18 19:58:41] <github> puppet: master Markus Roberts * 6bf9b8c (1 files in 1 dirs): Fix for #4257 -- problems resolving ::-prefixed classes ... | ||
| [2010/07/18 19:58:41] <gepetto> github: puppet: #4257 is http://projects.puppetlabs.com/issues/show/4257 "Puppet - Bug #4257: include ::foo is not working anymore - Puppet Labs" | ||
| [2010/07/18 19:58:41] <github> puppet: master Matt Robinson * da15bde (2 files in 2 dirs): [#4256] External nodes parameters can now be assigned to nodes ... | ||
| [2010/07/18 19:58:41] <github> puppet: master Jesse Wolfe * 0e1ccf4 (4 files in 4 dirs): [#4247] storeconfigs was calling Puppet::Parser::Resource.new with the wrong arguments ... | ||
| [2010/07/18 19:58:42] <github> puppet: master Markus Roberts * afb1322 (3 files in 2 dirs): Minimal fix for #4243 -- import isn't thread safe ... | ||
| [2010/07/18 19:58:42] <gepetto> github: puppet: #4243 is http://projects.puppetlabs.com/issues/show/4243 "Puppet - Bug #4243: Multithread issue with manual manifest import - Puppet Labs" | ||
| [2010/07/18 19:58:42] <github> puppet: master Brice Figureau * d1b984c (4 files in 2 dirs): Fix #4238 - if should match undef as '' ... | ||
| [2010/07/18 19:58:42] <gepetto> github: puppet: #4238 is http://projects.puppetlabs.com/issues/show/4238 "Puppet - Bug #4238: an undef variable is not anymore an empty string - Puppet Labs" | ||
| [2010/07/18 19:58:42] <github> puppet: master Markus Roberts * daa0f57 (2 files in 2 dirs): Fix for #4236 -- Only interpolate $ if followed by a variable ... | ||
| [2010/07/18 19:58:43] <gepetto> github: puppet: #4236 is http://projects.puppetlabs.com/issues/show/4236 "Puppet - Bug #4236: $ without a variable name is interpolated to undef - Puppet Labs" | ||
| [2010/07/18 19:58:43] <github> puppet: master Markus Roberts * 7717d03 (1 files in 1 dirs): Fix for #4234 -- ruby DSL fails on second resource ... | ||
| [2010/07/18 19:58:43] <gepetto> github: puppet: #4234 is http://projects.puppetlabs.com/issues/show/4234 "Puppet - Bug #4234: Ruby DSL fails if you specify more than one resource - Puppet Labs" | ||
| [2010/07/18 19:58:43] <github> puppet: master Jesse Wolfe * 61d2f83 (6 files in 6 dirs): [#4233] Ruby regexps are not multiline by default, but Resource titles can be multiline ... | ||
| [2010/07/18 19:58:44] <github> puppet: master Markus Roberts * dd03ac9 (5 files in 3 dirs): Partial fix for #4278 -- the performance aspects ... | ||
| [2010/07/18 19:58:44] <gepetto> github: puppet: #4278 is http://projects.puppetlabs.com/issues/show/4278 "Puppet - Bug #4278: Performance hole in 2.6.0 - Puppet Labs" | ||
| [2010/07/18 19:58:44] <github> puppet: master Markus Roberts * 5b68afe (1 files in 1 dirs): Fix for #4255 -- misleading diagnostic message ... | ||
| [2010/07/18 19:58:44] <gepetto> github: puppet: #4255 is http://projects.puppetlabs.com/issues/show/4255 "Puppet - Bug #4255: Puppet 2.6 'keeps setting content' on files - Puppet Labs" | ||
| [2010/07/18 19:58:44] <github> puppet: master Brice Figureau * 6e07a19 (2 files in 2 dirs): Fix #4262 - Puppetmaster used to log compilation time ... | ||
| [2010/07/18 19:58:45] <gepetto> github: puppet: #4262 is http://projects.puppetlabs.com/issues/show/4262 "Puppet - Bug #4262: Puppetmaster doesn't log the compilation time anymore - Puppet Labs" | ||
| [2010/07/18 19:58:45] <github> puppet: master Markus Roberts * 680dd1a (1 files in 1 dirs): Fix for #4257 -- problems resolving ::-prefixed classes ... | ||
| [2010/07/18 19:58:45] <gepetto> github: puppet: #4257 is http://projects.puppetlabs.com/issues/show/4257 "Puppet - Bug #4257: include ::foo is not working anymore - Puppet Labs" | ||
| [2010/07/18 19:58:45] <github> puppet: master Matt Robinson * 9f91540 (2 files in 2 dirs): [#4256] External nodes parameters can now be assigned to nodes ... | ||
| [2010/07/18 19:58:49] @ github left channel #puppet-dev () | ||
| [2010/07/18 20:38:36] @ lak joined channel #puppet-dev | ||
| [2010/07/18 20:50:57] @ Quit: beata-: Ping timeout: 240 seconds | ||
| [2010/07/18 22:11:16] @ Quit: lak: Quit: lak | ||
| [2010/07/18 22:29:20] @ rcrowley joined channel #puppet-dev | ||
| [2010/07/18 22:34:45] @ Quit: rcrowley: Quit: rcrowley |
Generated by irclog2html.py 2.6 by Marius Gedminas - find it at mg.pov.lt!