| [2010/08/25 00:00:48] @ Log started by gepetto | ||
| [2010/08/25 00:00:48] @ rawler_ joined channel #puppet | ||
| [2010/08/25 00:03:21] @ Quit: nexx: Quit: quit | ||
| [2010/08/25 00:05:20] @ flooose joined channel #puppet | ||
| [2010/08/25 00:07:39] @ Niek-work joined channel #puppet | ||
| [2010/08/25 00:08:56] @ kenneho joined channel #puppet | ||
| [2010/08/25 00:11:05] @ Quit: eternaltyro: Ping timeout: 252 seconds | ||
| [2010/08/25 00:13:53] @ \ask joined channel #puppet | ||
| [2010/08/25 00:14:56] @ Quit: lak: Quit: lak | ||
| [2010/08/25 00:24:33] @ ohookins joined channel #puppet | ||
| [2010/08/25 00:30:48] @ auslmo joined channel #puppet | ||
| [2010/08/25 00:30:57] <auslmo> hi all there | ||
| [2010/08/25 00:31:14] <auslmo> I couldn't find a list of global variables, like operatingsystem | ||
| [2010/08/25 00:31:32] <Volcane> type facter | ||
| [2010/08/25 00:32:07] <auslmo> thanks Volcane | ||
| [2010/08/25 00:32:08] <auslmo> :) | ||
| [2010/08/25 00:34:45] <Niek-work> :) | ||
| [2010/08/25 00:36:10] @ Ramonster joined channel #puppet | ||
| [2010/08/25 00:39:28] @ keyur joined channel #puppet | ||
| [2010/08/25 00:39:37] <keyur> #expect | ||
| [2010/08/25 00:39:54] <keyur> oops wrong window | ||
| [2010/08/25 00:40:53] @ Quit: goldfish: Ping timeout: 255 seconds | ||
| [2010/08/25 00:42:53] @ anderbubble joined channel #puppet | ||
| [2010/08/25 00:44:20] @ Quit: jfreeman: Remote host closed the connection | ||
| [2010/08/25 00:45:10] @ MattyM joined channel #puppet | ||
| [2010/08/25 00:47:23] @ Quit: auslmo: Ping timeout: 252 seconds | ||
| [2010/08/25 00:49:01] @ littleidea joined channel #puppet | ||
| [2010/08/25 00:49:45] <Kaelten> err: Could not retrieve catalog from remote server: Could not intern from pson: Syntax error at '/' at /etc/services:13 | ||
| [2010/08/25 00:49:50] <Kaelten> I keep getting errors like that | ||
| [2010/08/25 00:49:56] <Kaelten> but they don't seem to be making anysense | ||
| [2010/08/25 00:50:00] <Kaelten> anyone have any insight? | ||
| [2010/08/25 00:50:04] @ Quit: ceren: Quit: ceren | ||
| [2010/08/25 00:50:22] <anderbubble> Kaelten, is PWD/CWD /etc? | ||
| [2010/08/25 00:50:28] <anderbubble> That is, are you running puppet from /etc? | ||
| [2010/08/25 00:50:48] <Kaelten> from /etc | ||
| [2010/08/25 00:50:49] <Kaelten> ? | ||
| [2010/08/25 00:51:30] <anderbubble> I've seen a problem where puppet tries to use files in the current working directory as modules. So if you have a "services" module, but also happen to be in a directory that has a "services" file, it will try to import that as a module. | ||
| [2010/08/25 00:51:35] <Kaelten> damn yes I was in /etc when I was running it | ||
| [2010/08/25 00:52:21] <anderbubble> I should have probably filed a bug report about that! ~_^ | ||
| [2010/08/25 00:53:01] @ Quit: yannL: Remote host closed the connection | ||
| [2010/08/25 00:54:02] @ yannL joined channel #puppet | ||
| [2010/08/25 00:56:23] @ Quit: ecolitan: Read error: Operation timed out | ||
| [2010/08/25 01:04:27] @ macfly joined channel #puppet | ||
| [2010/08/25 01:13:28] @ Quit: keyur: Quit: Leaving | ||
| [2010/08/25 01:16:51] @ goldfish joined channel #puppet | ||
| [2010/08/25 01:18:05] @ Welsh_Dwarf joined channel #puppet | ||
| [2010/08/25 01:18:26] @ e1mer joined channel #puppet | ||
| [2010/08/25 01:21:20] <Hallaj> hey guys, if my puppet/facter is too old to have $fqdn.filename, is there another way I could get around this? | ||
| [2010/08/25 01:23:13] <zipkid> Hallaj: what version do you have? | ||
| [2010/08/25 01:24:27] @ TREllis joined channel #puppet | ||
| [2010/08/25 01:24:49] @ nul0op joined channel #puppet | ||
| [2010/08/25 01:24:58] <Hallaj> zipkid: 0.18.4-1 | ||
| [2010/08/25 01:25:24] <zipkid> ..... Very old.... Upgrade is the best option.... | ||
| [2010/08/25 01:31:11] <Hallaj> errr, I don't have that option at the moment. Those machines are at client site. No other way to get around it | ||
| [2010/08/25 01:31:12] <Hallaj> ? | ||
| [2010/08/25 01:31:34] <zipkid> Hallaj: make a custom fact..... | ||
| [2010/08/25 01:31:37] @ GioGio joined channel #puppet | ||
| [2010/08/25 01:31:43] <GioGio> hi everyone | ||
| [2010/08/25 01:31:48] <zipkid> and... you can update puppet remotely.... | ||
| [2010/08/25 01:32:41] <Hallaj> zipkid: It's installed via the debian deb for edgy which is old. Not quite sure if it'll continue working if I update it? | ||
| [2010/08/25 01:33:05] <zipkid> Hallaj: me niether... no debian here.... | ||
| [2010/08/25 01:35:36] @ ecolitan joined channel #puppet | ||
| [2010/08/25 01:36:07] @ anderbubble left channel #puppet () | ||
| [2010/08/25 01:45:52] <Hallaj> so, other than using the $fqdn.filename to separate files from hosts, there's no way to get around it? | ||
| [2010/08/25 01:53:40] @ Quit: ke4qqq: Ping timeout: 252 seconds | ||
| [2010/08/25 01:54:40] @ zambaboo joined channel #puppet | ||
| [2010/08/25 01:55:41] @ zambaboo_ joined channel #puppet | ||
| [2010/08/25 01:55:41] @ Quit: zambaboo: Read error: Connection reset by peer | ||
| [2010/08/25 01:55:42] @ zambaboo_ is now known as zambaboo | ||
| [2010/08/25 01:56:02] @ Quit: Welsh_Dwarf: Read error: Connection reset by peer | ||
| [2010/08/25 01:56:07] <zipkid> Hallaj: is $hostname not sufficient? | ||
| [2010/08/25 01:59:58] @ itguru joined channel #puppet | ||
| [2010/08/25 02:00:01] @ ke4qqq joined channel #puppet | ||
| [2010/08/25 02:03:15] @ ceren joined channel #puppet | ||
| [2010/08/25 02:03:49] <dcarley> Hallaj: by "$fqdn.filename" you mean it just doesn't have a fact for fqdn? | ||
| [2010/08/25 02:03:57] @ Quit: kenneho: Quit: Ex-Chat | ||
| [2010/08/25 02:04:11] @ kenneho joined channel #puppet | ||
| [2010/08/25 02:04:21] @ Quit: ]oestewart: Ping timeout: 240 seconds | ||
| [2010/08/25 02:05:09] <Hallaj> zipkid: yeah, I need the full fqdn for some reason | ||
| [2010/08/25 02:05:23] <Hallaj> dcarley: on the client side, facter | grep fqdn doesn't return anything. | ||
| [2010/08/25 02:05:25] <zipkid> Hallaj: DEFINE "for some reason" | ||
| [2010/08/25 02:05:45] @ yure joined channel #puppet | ||
| [2010/08/25 02:05:54] <dcarley> Hallaj: combine "${hostname}.${domain}" ? | ||
| [2010/08/25 02:06:07] @ Quit: SpacePope: Ping timeout: 272 seconds | ||
| [2010/08/25 02:06:38] @ SpacePope joined channel #puppet | ||
| [2010/08/25 02:06:38] <Hallaj> zipkid: zipkid.domain1 vs zipkid.domain2 -- hostname would just return zipkid | ||
| [2010/08/25 02:06:41] @ Quit: ceren: Client Quit | ||
| [2010/08/25 02:07:09] <zipkid> Hallaj: what nitwit decided on the naming scheme? | ||
| [2010/08/25 02:08:16] @ ]oestewart joined channel #puppet | ||
| [2010/08/25 02:12:46] <dcarley> zipkid: i'd guess it's just an example. but it's not that uncommon to have machines named after their purpose and qualified by their domain/location. | ||
| [2010/08/25 02:13:09] <zipkid> dcarley: ... true... | ||
| [2010/08/25 02:15:02] <dcarley> i certainly do :) | ||
| [2010/08/25 02:15:18] <Hallaj> :) | ||
| [2010/08/25 02:16:38] <ohookins> anyone here using cucumber-puppet and successfully overriding the confdir and/or manifest? | ||
| [2010/08/25 02:23:11] @ Quit: jubei__: Ping timeout: 276 seconds | ||
| [2010/08/25 02:27:19] @ andrewh joined channel #puppet | ||
| [2010/08/25 02:29:49] @ rcrowley joined channel #puppet | ||
| [2010/08/25 02:42:20] @ Quit: zambaboo: Quit: zambaboo | ||
| [2010/08/25 02:42:27] @ Quit: rmiller4pi8: Ping timeout: 265 seconds | ||
| [2010/08/25 02:45:13] @ Quit: flooose: Remote host closed the connection | ||
| [2010/08/25 02:46:12] @ vad3R joined channel #puppet | ||
| [2010/08/25 02:48:50] <daff> what does "err: /File[/var/lib/puppet/reports]: Could not evaluate: Could not find group puppet" mean? this is on a 2.6.0 client talking to a 0.25.5 master | ||
| [2010/08/25 02:50:39] <masterzen> daff: it is not recommended to have clients with version > master | ||
| [2010/08/25 02:52:11] @ Quit: vad3R: Read error: Connection reset by peer | ||
| [2010/08/25 02:52:34] @ vad3R joined channel #puppet | ||
| [2010/08/25 02:56:33] @ coleman79 left channel #puppet () | ||
| [2010/08/25 03:00:02] <daff> masterzen: perhaps, but that is very probably not the issue | ||
| [2010/08/25 03:00:07] <daff> I found http://projects.reductivelabs.com/issues/4336 | ||
| [2010/08/25 03:00:34] <daff> is there a workaround for this? can't really tell from the bug report | ||
| [2010/08/25 03:01:23] <masterzen> daff: yes the bug is fixed and the fix will appear in 2.6.1 | ||
| [2010/08/25 03:02:18] @ Quit: hephaestus: Ping timeout: 240 seconds | ||
| [2010/08/25 03:02:38] @ Quit: vad3R: Remote host closed the connection | ||
| [2010/08/25 03:03:02] @ hephaestus joined channel #puppet | ||
| [2010/08/25 03:06:55] @ Quit: rcrowley: Quit: rcrowley | ||
| [2010/08/25 03:17:18] @ tholl583xx joined channel #puppet | ||
| [2010/08/25 03:17:38] <tholl583xx> what tool is best for checking that the puppet clients are checking in, and whether they had any problems? | ||
| [2010/08/25 03:18:34] @ Quit: Hallaj: Quit: Leaving | ||
| [2010/08/25 03:20:46] <telmich> I'm trying to nest if/else in puppet 0.24.5, with this manifest http://git.sans.ethz.ch/?p=puppet-modules/ns_server;a=blob;f=manifests/website.pp;h=2993bd077a13bfd925253b8b08aa6f280e7f25bf;hb=HEAD, but getting an error that "'if'; expected '}' at ... :25" | ||
| [2010/08/25 03:20:54] <telmich> anyone an hint on how to solve that correctly? | ||
| [2010/08/25 03:22:19] <masterzen> telmich: you cannot use an if in the middle of a resource. Use a selector. | ||
| [2010/08/25 03:24:59] <matti> Hi masterzen | ||
| [2010/08/25 03:25:22] <masterzen> matti: hi! | ||
| [2010/08/25 03:25:31] <matti> :) | ||
| [2010/08/25 03:25:39] <matti> You good folks? :) | ||
| [2010/08/25 03:29:58] <zipkid> telmich: we use a script that queries our Dashboard db used from nagios. | ||
| [2010/08/25 03:30:06] <zipkid> oops | ||
| [2010/08/25 03:30:20] <zipkid> tholl583xx: that answer was for you | ||
| [2010/08/25 03:34:21] @ Quit: pmorillo: Ping timeout: 240 seconds | ||
| [2010/08/25 03:34:54] <tholl583xx> zipkid: my puppet dashboard appears to be doing not a lot at the moment, there is an update from the repo so I will update and see if that fixes | ||
| [2010/08/25 03:35:08] <tholl583xx> does the puppet client log errors to the puppet dashboard? | ||
| [2010/08/25 03:35:40] <ohadlevy> tholl583xx: i use foreman | ||
| [2010/08/25 03:39:27] @ Quit: jmeeuwen: Ping timeout: 240 seconds | ||
| [2010/08/25 03:43:54] <tholl583xx> ohadlevy: cool, and who might be the author of that fine software? | ||
| [2010/08/25 03:45:20] <dcarley> subtle :) | ||
| [2010/08/25 03:46:38] <zipkid> tholl583xx: ohadlevy LOOOOL ! | ||
| [2010/08/25 03:48:03] <dcarley> ohadlevy: is it possible to use foreman with storeconfigs whilst keeping a bit of separation between the schema and tables of the two? | ||
| [2010/08/25 03:48:09] @ Quit: alexine_dsouza: Read error: Connection reset by peer | ||
| [2010/08/25 03:49:05] @ redpepper is now known as RedPepper | ||
| [2010/08/25 03:51:53] @ Quit: TREllis: Quit: brb | ||
| [2010/08/25 03:52:06] @ jmeeuwen joined channel #puppet | ||
| [2010/08/25 03:53:41] <ohadlevy> dcarley: you can use another db | ||
| [2010/08/25 03:53:52] <ohadlevy> dcarley: but why would you like to keep them separated? | ||
| [2010/08/25 03:54:09] @ TREllis joined channel #puppet | ||
| [2010/08/25 03:54:39] <dcarley> ohadlevy: will it still pull in the storeconfig data? | ||
| [2010/08/25 03:54:54] <ohadlevy> dcarley: no, but it can pull it from other sources (e.g. the yaml files on the master) | ||
| [2010/08/25 03:55:18] <dcarley> ohadlevy: i feel uncomfortable about one project modifying another's schema or mixing tables (without prefixes) in the same db. | ||
| [2010/08/25 03:56:00] <ohadlevy> dcarley: i spend a great deal of time making sure I'm not affecting puppet's schema (in general, I just afew more tables and add a few columns to the db) | ||
| [2010/08/25 03:56:23] <ohadlevy> dcarley: but foreman does not require storeconfigs | ||
| [2010/08/25 03:57:34] <dcarley> ohadlevy: i don't doubt the quality of your work. it's just a best practice that i have engrained :) | ||
| [2010/08/25 03:58:04] <dcarley> ohadlevy: sure, i figured that since i have the storeconfig poison anyway. but there isn't any significant disadvantage to pulling it from yaml? | ||
| [2010/08/25 03:59:14] @ vad3R joined channel #puppet | ||
| [2010/08/25 03:59:28] <ohadlevy> dcarley: not that much, it will import only new data, which in case of facts, is not so often | ||
| [2010/08/25 04:00:34] <dcarley> ohadlevy: okay, will re-investigate then. thanks | ||
| [2010/08/25 04:03:12] <ohadlevy> dcarley: you might want to look at the puppet module for managing foreman, it supports both options (with and without storeconfigs) | ||
| [2010/08/25 04:06:01] @ Quit: ke4qqq: Ping timeout: 240 seconds | ||
| [2010/08/25 04:06:31] @ Quit: VAlu: Ping timeout: 258 seconds | ||
| [2010/08/25 04:08:16] @ VAlu joined channel #puppet | ||
| [2010/08/25 04:08:58] @ Quit: kenneho: Ping timeout: 265 seconds | ||
| [2010/08/25 04:12:38] @ ke4qqq joined channel #puppet | ||
| [2010/08/25 04:13:14] @ Quit: VAlu: Ping timeout: 264 seconds | ||
| [2010/08/25 04:14:00] @ Quit: e1mer: Quit: Leaving | ||
| [2010/08/25 04:14:16] @ VAlu joined channel #puppet | ||
| [2010/08/25 04:18:28] @ kenneho joined channel #puppet | ||
| [2010/08/25 04:18:33] <kenneho> Hi all, I've set up my puppet clients to send reports to the Puppet Dashboard. The reports shows up in the dashboard, but for some reason the reports don't include resources that have failed. To trigger an error I've included a non-existing class for the puppet client, but the report that shows up in the dashboard says that everything is ok. | ||
| [2010/08/25 04:18:53] <kenneho> Do I manually have to specify what is going to be reported or something? | ||
| [2010/08/25 04:19:36] @ coleman79 joined channel #puppet | ||
| [2010/08/25 04:25:16] @ MPSimmons joined channel #puppet | ||
| [2010/08/25 04:27:00] @ ickymettle joined channel #puppet | ||
| [2010/08/25 04:28:39] @ Feffe joined channel #puppet | ||
| [2010/08/25 04:30:06] @ mvn071 joined channel #puppet | ||
| [2010/08/25 04:30:33] @ Quit: jmeeuwen: Ping timeout: 245 seconds | ||
| [2010/08/25 04:30:59] <Feffe> hello, I've been having problems that puppet agent randomly hangs when applying the configuration | ||
| [2010/08/25 04:31:17] <Feffe> anyone knows anything about it, or could help me troubleshoot? | ||
| [2010/08/25 04:31:27] <ohookins> Feffe: it is more likely to be something that puppet is calling out to, like yum | ||
| [2010/08/25 04:31:39] <ohookins> Feffe: check your process listing | ||
| [2010/08/25 04:32:00] <Volcane> or that you're running webrick and have too many nodes | ||
| [2010/08/25 04:32:24] <Feffe> so far I only test puppet on a single node | ||
| [2010/08/25 04:32:31] <Feffe> so that's probably not a problem | ||
| [2010/08/25 04:32:44] <Feffe> the last debug output when it hanged just now was | ||
| [2010/08/25 04:32:49] <Feffe> debug: Executing 'grep -xqe '^keymap[[:space:]]*=[[:space:]]*"swedish.iso"' -- /etc/rc.conf' | ||
| [2010/08/25 04:32:57] <Feffe> grep | ||
| [2010/08/25 04:33:01] <Feffe> and test | ||
| [2010/08/25 04:33:14] <Feffe> has been the latest commands the last times it hang | ||
| [2010/08/25 04:33:23] <ohookins> how big is swedish.iso? it could just be taking a long time | ||
| [2010/08/25 04:33:43] <Feffe> no, swedish.iso is just a string | ||
| [2010/08/25 04:33:50] <Feffe> it changes in rc.conf | ||
| [2010/08/25 04:34:15] <Feffe> or rather, tests if keymap is set to swedish.iso | ||
| [2010/08/25 04:34:29] <ohookins> ah, right... ignore me ;) | ||
| [2010/08/25 04:34:35] <Feffe> :) | ||
| [2010/08/25 04:35:04] <ohookins> does that command work if you run it interactively? | ||
| [2010/08/25 04:35:16] @ brothers_ joined channel #puppet | ||
| [2010/08/25 04:35:17] <ohookins> perhaps try to attach strace to it when puppet it trying to run it | ||
| [2010/08/25 04:35:37] <Feffe> puppet agent --test --noop --debug | ||
| [2010/08/25 04:35:42] <Feffe> was what I ran | ||
| [2010/08/25 04:35:52] <ohookins> no, the grep | ||
| [2010/08/25 04:36:15] <Feffe> yes | ||
| [2010/08/25 04:36:20] <Feffe> the thing is | ||
| [2010/08/25 04:36:28] <Feffe> it does not always hang on the same commands | ||
| [2010/08/25 04:37:04] <ptman> what machine are you running it on? does it run out of memory? does the oom killer kick in? | ||
| [2010/08/25 04:37:11] <kenneho> I took a look at the report itself, and could not find any indications of the error I triggered (i.e. that I tried to include a non-existing class). Are the default reports not supposed to report this kind of error? | ||
| [2010/08/25 04:37:17] <zipkid> Feffe: i suspect the log shows completed stuff.... so it hangs on the next... | ||
| [2010/08/25 04:37:38] <zipkid> and as puppet does not guarantee order of execution... | ||
| [2010/08/25 04:38:12] * zipkid may be wrong about the log.... | ||
| [2010/08/25 04:38:48] <Feffe> well, I thought the debug output showed what happend next | ||
| [2010/08/25 04:39:11] @ Quit: brothers: Ping timeout: 276 seconds | ||
| [2010/08/25 04:39:13] <zipkid> Feffe: we'll have to ask here ... Anyone? | ||
| [2010/08/25 04:39:26] @ Quit: brothers_: Ping timeout: 240 seconds | ||
| [2010/08/25 04:39:51] <Feffe> ptman: the machine has 14G memory and only about 1G used | ||
| [2010/08/25 04:39:57] <Feffe> so that's not a problem either | ||
| [2010/08/25 04:40:18] <ptman> ok | ||
| [2010/08/25 04:40:39] <ptman> strace was suggested, I'd follow that advice | ||
| [2010/08/25 04:41:14] <Feffe> mm I'm currently reading up on truss, as this is a bsd-machine | ||
| [2010/08/25 04:41:28] <ptman> dtrace? | ||
| [2010/08/25 04:41:41] <ptman> or is it open/netbsd? | ||
| [2010/08/25 04:42:05] <Feffe> freebsd | ||
| [2010/08/25 04:42:21] <Feffe> ... | ||
| [2010/08/25 04:42:30] <Feffe> I started truss on the pid | ||
| [2010/08/25 04:42:42] <Feffe> and then puppet started again and finished the run :/ | ||
| [2010/08/25 04:43:36] <ohookins> while it is hanging, look at the process listing | ||
| [2010/08/25 04:43:51] <ohookins> if your ps supports -f i'd suggest using that so you can see what puppet has forked | ||
| [2010/08/25 04:44:07] <ohookins> then you'll see the command that it is waiting on | ||
| [2010/08/25 04:44:37] @ jmeeuwen joined channel #puppet | ||
| [2010/08/25 04:45:07] <Feffe> pstree reports puppet as a leaf process... | ||
| [2010/08/25 04:45:09] <Feffe> no children | ||
| [2010/08/25 04:47:04] @ Quit: ghaskins: Quit: Leaving | ||
| [2010/08/25 04:48:02] <Feffe> can't see that it waits on any process | ||
| [2010/08/25 04:51:23] <ohadlevy> Feffe: lsof? | ||
| [2010/08/25 04:54:04] @ someonexh joined channel #puppet | ||
| [2010/08/25 04:54:43] <someonexh> hello.. anyone available to help with a few questions? | ||
| [2010/08/25 04:55:04] <Volcane> you'll never know unless you ask them | ||
| [2010/08/25 04:55:12] <ssm> someonexh: don't ask to ask, just ask | ||
| [2010/08/25 04:55:20] <someonexh> sorry :/ .. :) | ||
| [2010/08/25 04:55:35] <someonexh> I am running redhat 5.5 | ||
| [2010/08/25 04:55:54] <someonexh> just installed the puppet and facter rpm's | ||
| [2010/08/25 04:55:58] <someonexh> via epel | ||
| [2010/08/25 04:56:13] <Feffe> ohadlevy: mostly ruby libraries and two pipes | ||
| [2010/08/25 04:56:15] <someonexh> I notice in the wiki instructions it lists a /usr/sbin/puppetmasterd file | ||
| [2010/08/25 04:56:16] <Feffe> can't see much | ||
| [2010/08/25 04:56:30] <someonexh> to run.. yet I don't see that on my systems | ||
| [2010/08/25 04:56:49] <Volcane> someonexh: on the master server you need the puppet-server rpm - only the master gets that | ||
| [2010/08/25 04:56:59] <someonexh> and tailing my "messages" log when I start (service puppet start) on the server machine it says "could not request certificates connection refuesed" | ||
| [2010/08/25 04:57:08] <someonexh> ok | ||
| [2010/08/25 04:57:23] <Volcane> and whatever host runs that, should be in dns as 'puppet' (a cname or something) | ||
| [2010/08/25 04:58:11] <someonexh> correct | ||
| [2010/08/25 04:58:17] <someonexh> I added puppet in the /etc/hosts file | ||
| [2010/08/25 04:59:11] @ bug joined channel #puppet | ||
| [2010/08/25 05:03:50] <someonexh> awesome.. that file is there now :) | ||
| [2010/08/25 05:04:48] @ Quit: bug: Quit: bug | ||
| [2010/08/25 05:06:11] @ gebi joined channel #puppet | ||
| [2010/08/25 05:07:32] @ Quit: pinoyskull: Quit: Leaving | ||
| [2010/08/25 05:07:37] @ Quit: AZed: Remote host closed the connection | ||
| [2010/08/25 05:09:02] @ Quit: VAlu: Ping timeout: 264 seconds | ||
| [2010/08/25 05:10:22] @ VAlu joined channel #puppet | ||
| [2010/08/25 05:12:18] @ Quit: ReinH: Ping timeout: 240 seconds | ||
| [2010/08/25 05:13:26] @ ReinH joined channel #puppet | ||
| [2010/08/25 05:15:35] @ uphillian joined channel #puppet | ||
| [2010/08/25 05:17:53] @ Spacefish joined channel #puppet | ||
| [2010/08/25 05:18:47] <Spacefish> hello i have fixed a typo in the documentation but can´t create a redmine account because the site seems somehow down (not responding after the first bytes are sent). this is my forked repo git@github.com:Spacefish/puppet-docs.git | ||
| [2010/08/25 05:21:56] <barn> so that upgrade to unicorn didn't go so well /: | ||
| [2010/08/25 05:22:12] @ Quit: ickymettle: Quit: leaving | ||
| [2010/08/25 05:22:31] @ Quit: FoXMaN: Quit: leaving | ||
| [2010/08/25 05:23:46] <Spacefish> well unicorn?! | ||
| [2010/08/25 05:25:57] <barn> unicorn, the ruby rack web application container | ||
| [2010/08/25 05:27:06] <barn> it's there, but very very slow | ||
| [2010/08/25 05:29:41] <Spacefish> yes i see server seems pretty overloaded i have a very high packetloss (76%) mtr shows gige-g1-1.core2.fmt1.he.net as the loosing point. | ||
| [2010/08/25 05:30:30] <Spacefish> maybe the router is dropping packets? | ||
| [2010/08/25 05:30:31] @ [GuS] joined channel #puppet | ||
| [2010/08/25 05:30:31] @ Quit: bkero: Ping timeout: 265 seconds | ||
| [2010/08/25 05:30:43] @ ahasenack joined channel #puppet | ||
| [2010/08/25 05:30:56] @ carlasouza joined channel #puppet | ||
| [2010/08/25 05:31:57] @ shug joined channel #puppet | ||
| [2010/08/25 05:33:03] @ Quit: uphillian: Read error: Connection reset by peer | ||
| [2010/08/25 05:35:02] @ Quit: someonexh: Quit: Page closed | ||
| [2010/08/25 05:36:46] <Naresh> i used the multiple-mongrels + nginx setup for puppetmaster2.6.0 a few days ago and i came across a weird issue: | ||
| [2010/08/25 05:37:32] <Naresh> i use an array of puppet:// paths for 'source' in my modules. the first path is $name.conf.$hostname, second: $name.conf | ||
| [2010/08/25 05:37:56] @ coleman79 left channel #puppet () | ||
| [2010/08/25 05:38:19] <Naresh> and when i ran the agent on the client for a particular host for which I didn't have a $name.conf.$hostname (so that it'd source the 'default' $name.conf instead) it kept getting an empy file | ||
| [2010/08/25 05:38:51] <Naresh> emtpy*. a quick revert back to the default webrick server didn't show this problem anymore. | ||
| [2010/08/25 05:38:53] @ Quit: jmeeuwen: Ping timeout: 245 seconds | ||
| [2010/08/25 05:39:00] <ohadlevy> Naresh: its a know bug.. search the list | ||
| [2010/08/25 05:39:09] <Naresh> oh | ||
| [2010/08/25 05:39:11] * Naresh searches | ||
| [2010/08/25 05:39:57] @ Quit: stahnma: Quit: Disconnecting from stoned server. | ||
| [2010/08/25 05:40:54] @ stahnma joined channel #puppet | ||
| [2010/08/25 05:42:36] <Spacefish> downloading the documentation pdf with this high packetloss sucks.. feels like 14.4 baud on a 100Mbit Uplink.. | ||
| [2010/08/25 05:46:11] @ Quit: ]oestewart: Quit: This computer has gone to sleep | ||
| [2010/08/25 05:46:39] @ Quit: realist_: Remote host closed the connection | ||
| [2010/08/25 05:48:35] <Naresh> http://groups.google.com/group/puppet-users/browse_thread/thread/ffc5e86addb7cc47/6914473c66a120fb <- relevant thread. thanks ohadlevy | ||
| [2010/08/25 05:53:54] @ bkero joined channel #puppet | ||
| [2010/08/25 05:55:37] @ realist joined channel #puppet | ||
| [2010/08/25 05:55:52] @ jmeeuwen joined channel #puppet | ||
| [2010/08/25 05:56:41] @ Bass10 joined channel #puppet | ||
| [2010/08/25 06:02:21] <Naresh> i had already applied the two one-liner patches from http://projects.puppetlabs.com/issues/4319 | ||
| [2010/08/25 06:04:22] @ Welsh_Dwarf joined channel #puppet | ||
| [2010/08/25 06:05:29] @ Quit: Hunner: Ping timeout: 260 seconds | ||
| [2010/08/25 06:13:03] @ kolla joined channel #puppet | ||
| [2010/08/25 06:13:38] <Spacefish> is it advisable to use the debian lenny 0.24 puppet or better to user the 2.6 stable? because using stable means a lot of work to be done after install.rb is called. | ||
| [2010/08/25 06:13:53] @ Hunner joined channel #puppet | ||
| [2010/08/25 06:18:19] @ Quit: itguru: Read error: Connection reset by peer | ||
| [2010/08/25 06:18:55] @ gmason joined channel #puppet | ||
| [2010/08/25 06:22:16] <unixdaemon> Spacefish: try and grab one from backports - 0.24 is very out of date. | ||
| [2010/08/25 06:24:00] @ Quit: Spacefish: Quit: Page closed | ||
| [2010/08/25 06:24:10] @ bug joined channel #puppet | ||
| [2010/08/25 06:27:15] @ ber_Toulouse joined channel #puppet | ||
| [2010/08/25 06:27:56] <ber_Toulouse> I want to know some basics how puppet dashboard works as an external node tool | ||
| [2010/08/25 06:29:01] <ber_Toulouse> As i understood it i can define my node with the hostnames in the tool and the external node code is converting the data into correct node.pp code | ||
| [2010/08/25 06:29:05] <ber_Toulouse> is that right? | ||
| [2010/08/25 06:29:35] @ Demosthenes joined channel #puppet | ||
| [2010/08/25 06:29:45] <zipkid> ber_Toulouse: no | ||
| [2010/08/25 06:30:00] <zipkid> the tool returns the yaml needed | ||
| [2010/08/25 06:30:26] <ber_Toulouse> And which process uses the yaml file? | ||
| [2010/08/25 06:30:32] <schimmi_> Hi! I am looking for a hint for the following design problem: class xorg { ... xorg.conf.erb ... } class xorg::nvidia { ... $driver = "nv" ... package nv-driver ... } class xorg::ati { ...} node { include xorg; include xorg::nvidia } | ||
| [2010/08/25 06:31:00] <nico> schimmi_: A custom fact ? | ||
| [2010/08/25 06:31:02] <schimmi_> this does not work because xorg does not see the variable assignments of xorg::nvidia | ||
| [2010/08/25 06:31:23] @ ]oestewart joined channel #puppet | ||
| [2010/08/25 06:31:27] <schimmi_> nico: a fact giving which value? | ||
| [2010/08/25 06:31:46] <nico> the one you want | ||
| [2010/08/25 06:31:47] <joe-mac> the driver to use | ||
| [2010/08/25 06:31:53] <joe-mac> easily done via lspci | ||
| [2010/08/25 06:31:57] <schimmi_> I want to put the decision into the nodes whether ati or nvidia should be used | ||
| [2010/08/25 06:31:59] <nico> toy with lspci | ||
| [2010/08/25 06:32:05] * nico ^5 joe-mac | ||
| [2010/08/25 06:32:54] <schimmi_> that would work, yes. but the point here is the design problem, i.e. how to solve that _inside_ the puppet language | ||
| [2010/08/25 06:33:24] <schimmi_> in C++ I would write a virtual method for the driver which is then overwritten in the ati or nvidia instantiation | ||
| [2010/08/25 06:33:38] <nico> please pastie your manifest | ||
| [2010/08/25 06:34:07] <schimmi_> i.e. the base class (xorg) has to get some information of a derived class (xorg::nvidia) | ||
| [2010/08/25 06:34:29] @ Quit: kolla: Remote host closed the connection | ||
| [2010/08/25 06:35:04] @ Quit: notbrien: Quit: notbrien | ||
| [2010/08/25 06:36:36] @ Quit: mvn071: Quit: Leaving | ||
| [2010/08/25 06:37:24] <Volcane> set the desired card in the node | ||
| [2010/08/25 06:37:30] <Volcane> $driver = nv at the node {} level | ||
| [2010/08/25 06:37:35] <Volcane> then in class xorg | ||
| [2010/08/25 06:37:39] <Volcane> include the right driver based on that | ||
| [2010/08/25 06:37:51] <schimmi_> well, yet | ||
| [2010/08/25 06:38:00] <schimmi_> but this exposed details which should be hidden | ||
| [2010/08/25 06:38:10] <Volcane> you're already typeing 'include xorg::nv' | ||
| [2010/08/25 06:38:10] <schimmi_> a node should not "know" the driver name | ||
| [2010/08/25 06:38:14] <Volcane> that exposes the sam | ||
| [2010/08/25 06:38:14] <Volcane> e | ||
| [2010/08/25 06:38:26] <Volcane> so how does your node know to include that line? | ||
| [2010/08/25 06:39:00] <Volcane> the right way is to just have a fact that detects the card type | ||
| [2010/08/25 06:39:03] <schimmi_> in this example it's just one line, but it can be more than that | ||
| [2010/08/25 06:39:05] <Volcane> sets that from the node automagically | ||
| [2010/08/25 06:39:17] <Volcane> then your xorg class inlucdes the right driver based on that fact | ||
| [2010/08/25 06:39:54] <schimmi_> let's make it more abstract: class base { ... $x ... } class a inherits base { $x=1 } class b inherits base { $x=2 } | ||
| [2010/08/25 06:40:30] <Volcane> doesnt work, as per the faq | ||
| [2010/08/25 06:41:07] <Volcane> you're asking what the puppet model is, the model is the node needs to either set a fact or the classifier - the human, the external node classifier etc - needs to inform the manifest what a node is | ||
| [2010/08/25 06:41:20] <Volcane> so either the node needs to make a fact, or something needs to tell the node what it is | ||
| [2010/08/25 06:41:30] @ radkin joined channel #puppet | ||
| [2010/08/25 06:42:10] <Volcane> and with the trend being toward external node classifiers, you should think in node parameters and included classes at node level and not so much in node inheritance | ||
| [2010/08/25 06:42:48] @ Quit: wilmoore: Remote host closed the connection | ||
| [2010/08/25 06:42:56] <nico> for hardware consideration a fact is at 99% of chance the best choice | ||
| [2010/08/25 06:43:04] <Volcane> yup | ||
| [2010/08/25 06:47:55] <schimmi_> ic. so the only place to make decision is in the node. classes are meant to be static. they can only be changed via node variables or facts. | ||
| [2010/08/25 06:48:26] <Volcane> mostly | ||
| [2010/08/25 06:48:38] <Volcane> classes should be building blocks to build nodes, nodes need to decide what they are | ||
| [2010/08/25 06:48:47] <Volcane> one way or the other, based on facts or external info | ||
| [2010/08/25 06:49:10] <schimmi_> yes, but a class inheritence in the node should not influence the semantics of another inherited class | ||
| [2010/08/25 06:49:17] <schimmi_> that's basically what my example above had in mind | ||
| [2010/08/25 06:49:26] <Volcane> node inheritance mostly doesnt work how people think | ||
| [2010/08/25 06:49:28] <Volcane> its best avoided | ||
| [2010/08/25 06:49:40] <FiXion> anyone can explain what the pass option to mount is? I'm not understanding the typeref. docu | ||
| [2010/08/25 06:49:54] @ Quit: Naresh: Ping timeout: 240 seconds | ||
| [2010/08/25 06:50:05] <Volcane> FiXion: options => "your options" | ||
| [2010/08/25 06:50:15] <FiXion> not the options | ||
| [2010/08/25 06:50:17] <FiXion> the pass | ||
| [2010/08/25 06:50:20] <FiXion> pass => | ||
| [2010/08/25 06:50:26] <FiXion> I'm wondering what that is | ||
| [2010/08/25 06:50:28] <Volcane> man fstab | ||
| [2010/08/25 06:50:52] <FiXion> it sounds more like a puppet thing. the description says: the pass in which the mount is checked | ||
| [2010/08/25 06:51:06] <Volcane> yes, its a fstab option | ||
| [2010/08/25 06:51:53] <FiXion> ohh it's fs_passno | ||
| [2010/08/25 06:51:57] <FiXion> wasn't really clear to me :) | ||
| [2010/08/25 06:52:00] <FiXion> thanks | ||
| [2010/08/25 06:52:25] <Volcane> the man page is rubbish | ||
| [2010/08/25 06:52:25] <FiXion> anyone have a good idea of how best to make puppet ensure that ONLY nfs mounts defined in puppet are present in fstab? | ||
| [2010/08/25 06:52:31] <Volcane> fstab man page has more info :) | ||
| [2010/08/25 06:53:00] @ gmason_ joined channel #puppet | ||
| [2010/08/25 06:53:01] <FiXion> I was thinking of having an exec, required by mount - which cleared fstab for nfs mountpoints | ||
| [2010/08/25 06:53:28] <FiXion> and it being required by all mount references - it would be executed before them. and they would then be added to the fstab - now only containing local mounts | ||
| [2010/08/25 06:53:49] <FiXion> not sure if it would work - or if it's actually guaranteed to work in future versions :) | ||
| [2010/08/25 06:54:08] <Volcane> that would just keep editing the file over and over again and leave you open to corruption if puppet dies at just the wrong time | ||
| [2010/08/25 06:54:10] <FiXion> it could hurt rather badly on upgrade if something screws up there :) | ||
| [2010/08/25 06:54:36] <FiXion> I would like to avoid having to create an fstab file for each type of server. | ||
| [2010/08/25 06:54:49] <FiXion> currently I have mountpoints defined with the modules that need them | ||
| [2010/08/25 06:55:11] <FiXion> but I need a way to ensure the puppet defined "external" mountpoints - are the only ones actually in fstab | ||
| [2010/08/25 06:55:11] @ Quit: gmason: Ping timeout: 258 seconds | ||
| [2010/08/25 06:55:11] @ gmason_ is now known as gmason | ||
| [2010/08/25 06:55:29] <FiXion> sometimes people add a mount point to 1 server in a cluster - and suddenly a service ONLY works on that server :( | ||
| [2010/08/25 06:56:00] @ Quit: ohookins: Quit: Leaving. | ||
| [2010/08/25 06:56:17] @ ohookins joined channel #puppet | ||
| [2010/08/25 06:56:17] <FiXion> I do this with vhosts files.. but those are files.. it works like a charm :) | ||
| [2010/08/25 06:56:54] <Volcane> if you could arrange to build fstab from snippets it could do what you want, but mostly you should address whatever the problem is thats causing stale mounts | ||
| [2010/08/25 06:56:59] <Volcane> cos thats the problem | ||
| [2010/08/25 06:57:36] @ Quit: p3rror: Ping timeout: 272 seconds | ||
| [2010/08/25 06:58:33] @ kaptk2 joined channel #puppet | ||
| [2010/08/25 07:00:17] <FiXion> I don't have stale mounts. I have people who forget to add a mountpoint via puppet - but just adds it manually to fstab | ||
| [2010/08/25 07:00:29] <FiXion> and I'd like to forcefully fix that with puppet :) | ||
| [2010/08/25 07:00:33] <Volcane> thats what i mean with 'stale' ie. ones that shouldnt be there | ||
| [2010/08/25 07:00:59] <FiXion> well - some human problems can't easily be fixed :) | ||
| [2010/08/25 07:01:12] <FiXion> they are best detected/fixed by automation :) | ||
| [2010/08/25 07:01:24] @ AZed joined channel #puppet | ||
| [2010/08/25 07:02:05] <Volcane> technical solutions that tries to work around uncooperative coworkers just gives off a cold war between systems | ||
| [2010/08/25 07:02:18] <Volcane> and mostly you'll always loose cos the human can actually think unlike automations :) | ||
| [2010/08/25 07:02:45] <FiXion> it actually works rather well | ||
| [2010/08/25 07:03:01] <FiXion> all files puppet manages are sometimes modified by someone who didn't think puppet was managing it | ||
| [2010/08/25 07:03:13] <FiXion> but since puppet changes it back - they notice and get the change into puppet | ||
| [2010/08/25 07:03:28] @ Quit: jmeeuwen: Ping timeout: 245 seconds | ||
| [2010/08/25 07:03:40] @ notbrien joined channel #puppet | ||
| [2010/08/25 07:03:40] <FiXion> so I don't see it being any different from the rest of the things puppet manages today | ||
| [2010/08/25 07:03:55] <FiXion> only thing with fstab - is I can't control the entire file | ||
| [2010/08/25 07:04:14] @ thekad-afk is now known as thekad | ||
| [2010/08/25 07:04:17] <Volcane> nods, then none of the usual purge methods will work either really | ||
| [2010/08/25 07:04:26] <FiXion> :( | ||
| [2010/08/25 07:05:35] @ Quit: Ramonster: Quit: So long, thanx for all the fish | ||
| [2010/08/25 07:05:35] <Volcane> in that situation I'd probably do something with snippets, it can purge a file and still allow local edits etc | ||
| [2010/08/25 07:05:35] * FiXion thinks that perhaps mount could use an "ensuremode" - 0=just ensure this mount, 1=ensure only puppet defined and local mounts are there, 2=ensure only puppet defined mounts | ||
| [2010/08/25 07:05:36] <Volcane> you'd need to somehow preserve the original fstab into another location | ||
| [2010/08/25 07:05:43] <FiXion> that's easy | ||
| [2010/08/25 07:06:02] <Volcane> and probably stop using mount{} :P | ||
| [2010/08/25 07:06:07] <FiXion> in my case. cat /etc/fstab | grep -v nfs | /tmp/fstab.local | ||
| [2010/08/25 07:06:11] <Volcane> some kind of define to replace it | ||
| [2010/08/25 07:06:20] <Volcane> tmp? seriously? | ||
| [2010/08/25 07:06:22] <FiXion> it's actually already a define today | ||
| [2010/08/25 07:06:30] <FiXion> which ever | ||
| [2010/08/25 07:06:34] <Volcane> heh | ||
| [2010/08/25 07:06:58] @ kolla joined channel #puppet | ||
| [2010/08/25 07:06:59] <Volcane> well my concat can pull in a local file into a concat allowing you to edit the local file so its contents show up | ||
| [2010/08/25 07:07:06] <Volcane> and then purge the rest with only what you're putting in | ||
| [2010/08/25 07:07:15] <FiXion> the define could perhaps just add fstab formatted lines to a file - and then I'd concat them in the end | ||
| [2010/08/25 07:07:19] <Volcane> so if you can convince people not to edit /path/to/copy/of/fstab | ||
| [2010/08/25 07:07:22] <Volcane> but /etc/fstab | ||
| [2010/08/25 07:07:26] <Volcane> then their changes will just go away | ||
| [2010/08/25 07:07:42] <FiXion> I'd rebuild the /path/to/copy/of/fstab-localonly every time | ||
| [2010/08/25 07:08:15] <Volcane> no, not what i mean | ||
| [2010/08/25 07:08:19] <FiXion> I "just" need to ensure the final concat is handled in the end though | ||
| [2010/08/25 07:08:23] <Volcane> u need somewhere where the machines base fstab is | ||
| [2010/08/25 07:08:30] <Volcane> ie. whatever your installer put in | ||
| [2010/08/25 07:08:41] <Volcane> you then pull that into /etc/fstab plus your own mounts | ||
| [2010/08/25 07:08:41] <FiXion> yes - my cat above (into tmp though) generates it | ||
| [2010/08/25 07:08:51] @ Quit: Maliuta: Remote host closed the connection | ||
| [2010/08/25 07:08:52] <Volcane> if you do that with my concat tool, it will purge it of any hand edits | ||
| [2010/08/25 07:08:57] <FiXion> and since fstab starts out fine - I'd ofcourse have to do this first. | ||
| [2010/08/25 07:09:01] <Volcane> without needing to know the contents of the lcoal per machine fstab | ||
| [2010/08/25 07:09:08] <FiXion> Volcane: link to this concat tool? | ||
| [2010/08/25 07:09:19] <Volcane> http://github.com/ripienaar/puppet-concat | ||
| [2010/08/25 07:09:37] @ reyjrar joined channel #puppet | ||
| [2010/08/25 07:09:52] @ Maliuta joined channel #puppet | ||
| [2010/08/25 07:10:16] <FiXion> thanks - I'll see if can figure it out, using that | ||
| [2010/08/25 07:11:21] @ Quit: bronto: Ping timeout: 240 seconds | ||
| [2010/08/25 07:11:59] @ p3rror joined channel #puppet | ||
| [2010/08/25 07:14:05] @ bronto joined channel #puppet | ||
| [2010/08/25 07:16:31] @ Quit: AZed: Remote host closed the connection | ||
| [2010/08/25 07:16:52] @ AZed joined channel #puppet | ||
| [2010/08/25 07:17:48] @ jmeeuwen joined channel #puppet | ||
| [2010/08/25 07:21:00] @ gniks joined channel #puppet | ||
| [2010/08/25 07:27:36] <FiXion> Volcane: that looks like it'll do nicely | ||
| [2010/08/25 07:27:53] <FiXion> and I can just make nagios check the motd.local file for none local mounts | ||
| [2010/08/25 07:28:08] <FiXion> so I'm sure no one takes a shortcut, unwillingly or otherwise :) | ||
| [2010/08/25 07:28:26] <FiXion> I'm going to check for active mounts, not in fstab anyways | ||
| [2010/08/25 07:28:46] <FiXion> one might forget such at a mid-night job :) | ||
| [2010/08/25 07:42:49] @ morpheus joined channel #puppet | ||
| [2010/08/25 07:43:28] @ Quit: [GuS]: Read error: Connection reset by peer | ||
| [2010/08/25 07:43:30] @ |GuS| joined channel #puppet | ||
| [2010/08/25 07:44:25] <morpheus> Hi@all | ||
| [2010/08/25 07:44:36] <morpheus> do you build custom types on a regular basis? | ||
| [2010/08/25 07:44:59] @ Quit: Niek-work: Read error: Connection reset by peer | ||
| [2010/08/25 07:45:17] <morpheus> I mean do you prefer to build new types whenever you wish to manage a ressource or is this kind of the last thing you would do? | ||
| [2010/08/25 07:45:43] <morpheus> hope you do it a lot, cause I have problems ;-) | ||
| [2010/08/25 07:45:56] <FiXion> I haven't need to yet - and since I'm not really good at ruby - it would probably take a lot longer :) | ||
| [2010/08/25 07:46:10] <FiXion> what is it, you need to handle by adding a new type? | ||
| [2010/08/25 07:46:16] <morpheus> I'm totally new to ruby too | ||
| [2010/08/25 07:46:26] <Volcane> morpheus: never had the need. defines tend to do everything i need | ||
| [2010/08/25 07:46:47] <morpheus> wow I'm trying to manage /etc/project on Solaris for the last three days now | ||
| [2010/08/25 07:46:53] <FiXion> Volcane: how does the concat module handle the fstab.local file isn't there? I'd like it to fail miserably - not touching /etc/fstab :) | ||
| [2010/08/25 07:47:27] <FiXion> morpheus: manage - as in ensure only files added by puppet to that folder is actually there? | ||
| [2010/08/25 07:47:40] <morpheus> Is the appropiate line there? no-> use projadd to add it. Its there but wrong -> use projmod to correct it | ||
| [2010/08/25 07:47:48] @ Niek-work joined channel #puppet | ||
| [2010/08/25 07:48:28] <Volcane> FiXion: best to do some testing | ||
| [2010/08/25 07:49:05] <morpheus> and different application on one hosts needs an entry in the /etc/project | ||
| [2010/08/25 07:49:29] <FiXion> morpheus: seems this guy thinks like you: http://github.com/plathrop/puppet-module-apache | ||
| [2010/08/25 07:49:30] <morpheus> I havent looked at auegias because I thought I would be smart | ||
| [2010/08/25 07:49:41] <FiXion> you could probably modify that to do your bidding | ||
| [2010/08/25 07:50:09] <FiXion> my apache module uses a2enmod - but exec and unless/onlyif shell commands to do the same. | ||
| [2010/08/25 07:50:36] <FiXion> but that's because a2enmod etc. only handles symlinks | ||
| [2010/08/25 07:50:57] <morpheus> I unfortunalty I dont have just create and destroy I also want to modify things | ||
| [2010/08/25 07:51:10] * FiXion has to go home now.. | ||
| [2010/08/25 07:51:25] <morpheus> one entry could be: | ||
| [2010/08/25 07:51:28] <FiXion> morpheus: but if you know how it's suppose to look in it's entirety - you could just destroy and create afterwards | ||
| [2010/08/25 07:51:46] <morpheus> myproj:12334:user1,user2:group1,group2:key1=value1;key2=value2 | ||
| [2010/08/25 07:51:56] @ wajasu joined channel #puppet | ||
| [2010/08/25 07:52:45] <morpheus> so I use exists? to parse the line and look for myproj than extract projectID, users (as an array), group (as an array), attributes (as an array) | ||
| [2010/08/25 07:53:07] <morpheus> maybe I should provide some pasties | ||
| [2010/08/25 07:53:26] <FiXion> to do it as a provider - I don't know. would be happy to see the result though. nice to have examples to use, if I ever find myself in that situation | ||
| [2010/08/25 07:53:31] <barn> morpheus: http://en.wikipedia.org/wiki/Pasties those?! | ||
| [2010/08/25 07:53:38] @ Quit: rawler_: Remote host closed the connection | ||
| [2010/08/25 07:54:03] <morpheus> ok I learned a new english word ;-) | ||
| [2010/08/25 07:54:09] <morpheus> I meant some code snippet | ||
| [2010/08/25 07:54:44] <morpheus> but I'm at home now and no solaris at hand | ||
| [2010/08/25 07:54:45] <barn> </helpful> | ||
| [2010/08/25 07:55:30] <morpheus> My problem was: If the line isnt there (exists?=false) puppet called create of the provider which worked fine | ||
| [2010/08/25 07:56:00] <morpheus> and than I created getters for all the attributes which returned the parsed stuff from exists | ||
| [2010/08/25 07:56:17] @ pinoyskull joined channel #puppet | ||
| [2010/08/25 07:56:29] @ ohookins left channel #puppet ("Leaving.") | ||
| [2010/08/25 07:56:38] <morpheus> And even though puppet used the getter to get the current value it always that its wrong | ||
| [2010/08/25 07:56:48] <morpheus> hard to explain though | ||
| [2010/08/25 07:57:10] <morpheus> I think it was an array problem. Lets say the users in /etc/project was user1,user2 | ||
| [2010/08/25 07:57:29] <morpheus> in exists? I convert that to an array [ user1, user2 ] with split | ||
| [2010/08/25 07:57:29] @ |GuS| is now known as [GuS] | ||
| [2010/08/25 07:57:46] <morpheus> and in my manifest I also use user => [ user1, user2 ] | ||
| [2010/08/25 07:58:33] <morpheus> And IIRC puppet always that user is "user1,user2" should be "user1 user" | ||
| [2010/08/25 07:59:47] <ber_Toulouse> Again a puppet user beginner question ... what is normally define as the nodes Class and what the group... Are there best practices how to setup nodes? For example i created a client for my Configuration Control server and want to define it in puppet-dashboard as a node... what would be good values for class and group in that case? | ||
| [2010/08/25 08:00:07] <morpheus> So I dont know how puppet compares arrays. Because resource[:user] IS an array. And the @user i build during exists? is an array too | ||
| [2010/08/25 08:00:46] <morpheus> I did some Puppet.notice @user.join('|') and resource[:user].join('|') and there are identical | ||
| [2010/08/25 08:02:05] <morpheus> So thought there would be anyone saying: Yeah when you use arrays dont forget to ... | ||
| [2010/08/25 08:03:32] @ Tas_ joined channel #puppet | ||
| [2010/08/25 08:03:45] @ rcrowley joined channel #puppet | ||
| [2010/08/25 08:04:52] @ thekad is now known as thekad-afk | ||
| [2010/08/25 08:09:06] @ benkevan joined channel #puppet | ||
| [2010/08/25 08:09:08] @ Quit: benkevan: Changing host | ||
| [2010/08/25 08:09:08] @ benkevan joined channel #puppet | ||
| [2010/08/25 08:09:18] @ gutseb_ joined channel #puppet | ||
| [2010/08/25 08:11:34] @ RussS joined channel #puppet | ||
| [2010/08/25 08:12:40] @ rasputnik joined channel #puppet | ||
| [2010/08/25 08:15:59] @ Quit: rasputnik: Client Quit | ||
| [2010/08/25 08:17:59] @ Quit: gmason: Quit: gmason | ||
| [2010/08/25 08:22:45] @ gmason joined channel #puppet | ||
| [2010/08/25 08:23:36] @ thekad-afk is now known as thekad | ||
| [2010/08/25 08:23:45] @ Quit: Niek-work: Remote host closed the connection | ||
| [2010/08/25 08:24:41] @ Quit: gutseb_: Quit: gutseb_ | ||
| [2010/08/25 08:25:33] @ Quit: lkoranda: Quit: Ex-Chat | ||
| [2010/08/25 08:25:42] @ apb joined channel #puppet | ||
| [2010/08/25 08:26:56] <apb> can an inline_template return a list? my simple test with $var = inline_template("<%= ['a','b'] %>") seems to return an empty string. | ||
| [2010/08/25 08:27:27] <Volcane> erb's make strings | ||
| [2010/08/25 08:27:41] <apb> ok, thanks | ||
| [2010/08/25 08:28:48] @ Quit: kenneho: Ping timeout: 240 seconds | ||
| [2010/08/25 08:28:55] @ Quit: mikey_p: Quit: mikey_p | ||
| [2010/08/25 08:30:13] @ Quit: spacex: Quit: Leaving. | ||
| [2010/08/25 08:30:50] @ kenneho joined channel #puppet | ||
| [2010/08/25 08:34:09] <apb> My real problem is, I have a template that makes file { ... source => [a list whose contents depend on several conditions] }. can't use a bunch of if <condition> { $list += ["append another element"] } because of variable scoping issues | ||
| [2010/08/25 08:36:41] @ christophermlee joined channel #puppet | ||
| [2010/08/25 08:37:54] @ cowboycoder joined channel #puppet | ||
| [2010/08/25 08:38:56] <cowboycoder> is there a way to use the successful execution of a command as a prerequisite of a class? | ||
| [2010/08/25 08:39:20] @ nexx joined channel #puppet | ||
| [2010/08/25 08:39:24] @ Quit: apb: Quit: Leaving. | ||
| [2010/08/25 08:39:46] <morpheus> require => Exec['important_command'] ? | ||
| [2010/08/25 08:40:10] <cowboycoder> morpheus: if the command exits with an error status, will the dependency not apply? | ||
| [2010/08/25 08:40:31] <morpheus> I'm not sure but I would guess not | ||
| [2010/08/25 08:41:00] <morpheus> at least it would make a lot of sense | ||
| [2010/08/25 08:41:31] <cowboycoder> and if the command goes from working to not, and then back to working will it re-apply the class? | ||
| [2010/08/25 08:42:41] @ [ack] joined channel #puppet | ||
| [2010/08/25 08:43:20] @ Quit: kenneho: Quit: Ex-Chat | ||
| [2010/08/25 08:44:47] <jbooth> Deps don't go if something they require failed | ||
| [2010/08/25 08:44:59] @ danj joined channel #puppet | ||
| [2010/08/25 08:45:47] <jbooth> apb: why are you implementing a varying source list? It already skips non-existent files, so just list it all and let puppet omit them? | ||
| [2010/08/25 08:46:00] <danj> i've been trying to figure out how stages work in coordination with an external node classifier and the Main class | ||
| [2010/08/25 08:47:37] <danj> do classes that come from the node classifier get put into the main stage by default? | ||
| [2010/08/25 08:47:38] * Volcane is considering the boat | ||
| [2010/08/25 08:47:42] <Volcane> oops, wrong # | ||
| [2010/08/25 08:49:56] @ alcy joined channel #puppet | ||
| [2010/08/25 08:56:25] @ spacex joined channel #puppet | ||
| [2010/08/25 08:57:47] <[ack]> is puppet safe to use to provide recipes to users in a multi-tenant cloud? | ||
| [2010/08/25 08:58:08] <[ack]> for example, if i want to provide a web server recipe that people can use. | ||
| [2010/08/25 08:58:21] <Diranged> [ack]: depends on how well written it is.. :P | ||
| [2010/08/25 08:59:06] <[ack]> sure, but i'm asking if this is an intended use case for puppet | ||
| [2010/08/25 08:59:24] <jbooth> Do you mean leatting your users actually write puppet configs? | ||
| [2010/08/25 08:59:28] <[ack]> no | ||
| [2010/08/25 08:59:29] <Volcane> why would it be unsafe? its only unsafe if you let your tenants also edit the modules contents that then go out to machines that arent theirs | ||
| [2010/08/25 08:59:39] <[ack]> just to pull recipes from the puppet server | ||
| [2010/08/25 09:00:30] <[ack]> the puppet server would be administered by me, and the clients would run puppet on their machines and be clients for the puppet server. | ||
| [2010/08/25 09:00:34] <geekatcmu> apb: if you need a list, write a function. | ||
| [2010/08/25 09:03:39] <Volcane> or use split | ||
| [2010/08/25 09:03:55] @ toi joined channel #puppet | ||
| [2010/08/25 09:04:46] <jbooth> [ack]: That should be fine as long as you don't turn on auto-sign and don't screw up your manifests. | ||
| [2010/08/25 09:06:34] <GioGio> I have a probably stupid question, but I didn't have a configured environment to test that. So what happened if there is two modules who try to activate/desactivate the same service ? (the same question could be answered with other resources like files...) | ||
| [2010/08/25 09:07:29] <Volcane> GioGio: on the same machine, resources have to be unique | ||
| [2010/08/25 09:07:53] @ jdcasey joined channel #puppet | ||
| [2010/08/25 09:07:58] @ nathanr joined channel #puppet | ||
| [2010/08/25 09:08:02] <GioGio> what do you mean with on the same machine ? on the same catalog ? | ||
| [2010/08/25 09:08:08] <Volcane> yes | ||
| [2010/08/25 09:08:27] <Volcane> you can declare service{"httpd": } in 100 classes, you can only include 1 a time on a node though | ||
| [2010/08/25 09:08:36] <Volcane> you should make a http module | ||
| [2010/08/25 09:08:43] <GioGio> ok thanks | ||
| [2010/08/25 09:08:45] <Volcane> and whenever you need a webserver in any of your other modules | ||
| [2010/08/25 09:08:47] <Volcane> you include that module | ||
| [2010/08/25 09:08:54] <Volcane> you can include the same thing many times | ||
| [2010/08/25 09:09:05] <GioGio> what happened if i include twice service httpd ? | ||
| [2010/08/25 09:09:17] <Volcane> you can include the same class many times | ||
| [2010/08/25 09:09:29] <GioGio> the client will not be configured ? ot it will configure the rest but not that ? | ||
| [2010/08/25 09:09:47] <Volcane> since its the same class, it'll do what the class says | ||
| [2010/08/25 09:09:52] @ fugit13 joined channel #puppet | ||
| [2010/08/25 09:09:54] @ Quit: Welsh_Dwarf: Read error: Connection reset by peer | ||
| [2010/08/25 09:10:56] <GioGio> Volcane: ok, thank you for your answer | ||
| [2010/08/25 09:11:02] <morpheus> ok ready for some action about my provider | ||
| [2010/08/25 09:11:15] <morpheus> thats a sample of the manifest: http://pastie.org/1115409 | ||
| [2010/08/25 09:11:24] <morpheus> so this is kind of the dream | ||
| [2010/08/25 09:11:42] <nathanr> is there a way to have puppet use the atboot attribute of mount for mac? | ||
| [2010/08/25 09:11:54] <morpheus> this is my type: http://pastie.org/1115413 | ||
| [2010/08/25 09:12:08] <morpheus> nothing special so far I think | ||
| [2010/08/25 09:12:23] <Volcane> best to ask the dev list about types | ||
| [2010/08/25 09:12:48] <morpheus> and my provider: http://pastie.org/1115415 | ||
| [2010/08/25 09:13:25] <morpheus> its not really functional at the moment because I cannot run projadd, projmod etc so I just fake it with /bin/true | ||
| [2010/08/25 09:14:14] <morpheus> if i start with an empty /etc/project puppet recognises it (exists? does return false) | ||
| [2010/08/25 09:14:32] <morpheus> thats working: debug: Puppet::Type::Project::ProviderProjadd: Executing '/bin/true -p 1234 -c my comment -U user1,user2,user3 -G group1 -K key1=value1 -K key2=value2 proj1 | ||
| [2010/08/25 09:15:43] <morpheus> because /bin/true doesnt do anything I insert the right line in /etc/project by myself | ||
| [2010/08/25 09:15:54] <morpheus> proj1:1234:my comment:user1,user2,user3:group1:key1=value1;key2=value2 | ||
| [2010/08/25 09:17:08] <morpheus> but puppet always says that attributes had changed notice: /Stage[main]/Project/Project[proj1]/attributes: attributes changed 'key1=value1key2=value2' to 'key1=value1 key2=value2' | ||
| [2010/08/25 09:17:29] @ gutseb_ joined channel #puppet | ||
| [2010/08/25 09:19:37] <jbooth> morpheus: Your exists? detection isn't setting @attributes in a way that matches what's coming in from the type. | ||
| [2010/08/25 09:19:42] <jbooth> If I'm reading it right, which I might not be. | ||
| [2010/08/25 09:20:29] @ rmiller4pi8 joined channel #puppet | ||
| [2010/08/25 09:20:32] <morpheus> jbooth: Yeah but I dont know where the error is | ||
| [2010/08/25 09:20:58] <morpheus> because in the setter I have these to outputs: http://pastie.org/1115415 | ||
| [2010/08/25 09:21:04] <morpheus> they are identical | ||
| [2010/08/25 09:21:10] <jbooth> morpheus: Start debugging. Print out via puppet.debug what you set @attributes to and what the newprop is | ||
| [2010/08/25 09:21:20] <jbooth> oh, while you're at it, most of your newprops should be newparams | ||
| [2010/08/25 09:21:32] <jbooth> at least with how you have your provider written | ||
| [2010/08/25 09:21:59] @ Reepicheep joined channel #puppet | ||
| [2010/08/25 09:22:05] <morpheus> I tought a prop is something you want to test against and a param is just a param for the provider (like manage_home) | ||
| [2010/08/25 09:22:08] <jbooth> File.open("puppet-setup").each do |line| is also going to be a lot clearer to read | ||
| [2010/08/25 09:22:18] <jbooth> a prop is something that "does work" | ||
| [2010/08/25 09:22:33] <jbooth> ie: ensure needs to "do work" to make it exist or not | ||
| [2010/08/25 09:24:27] <morpheus> ok but when I look at puppet/type/mount for example | ||
| [2010/08/25 09:24:31] <jbooth> morpheus: What you might be running into is symbol-vs-string conflicts. | ||
| [2010/08/25 09:24:42] <morpheus> fstype, pass, options are all properties | ||
| [2010/08/25 09:26:22] <jbooth> I didn't say it didn't work, but what it was explained to me as | ||
| [2010/08/25 09:26:36] @ plathrop-away is now known as plathrop | ||
| [2010/08/25 09:28:11] @ Quit: GioGio: Quit: Page closed | ||
| [2010/08/25 09:28:18] <jbooth> morpheus: If you're sure about that attributes changed line being pasted correctly, I'm pretty sure that's why it won't work for you. | ||
| [2010/08/25 09:31:06] @ Quit: nul0op: Ping timeout: 240 seconds | ||
| [2010/08/25 09:32:11] @ Quit: Dantman: Ping timeout: 265 seconds | ||
| [2010/08/25 09:33:14] <morpheus> arg! | ||
| [2010/08/25 09:34:05] <morpheus> in the exists? function where it iterates over the lines... | ||
| [2010/08/25 09:34:19] <morpheus> I added a simple line.chomp! and now it seems to work... | ||
| [2010/08/25 09:34:41] @ Quit: reyjrar: Quit: Leaving. | ||
| [2010/08/25 09:34:52] <jbooth> Alternately, .each do |line| probably would have also fixed it | ||
| [2010/08/25 09:34:56] @ tonyskapunk joined channel #puppet | ||
| [2010/08/25 09:35:06] <jbooth> (it'd split on the newlines, which is most likely what your chomp is eating) | ||
| [2010/08/25 09:36:09] <morpheus> you mean File.open("/etc/project",'r').each do |line| | ||
| [2010/08/25 09:36:14] <jbooth> in some ways, you'd be better being more advanced. Have just the one type, leave things as properties, and let the out-of-sync properties fix themselves | ||
| [2010/08/25 09:36:17] <jbooth> yeah | ||
| [2010/08/25 09:36:18] @ Quit: bkero: Changing host | ||
| [2010/08/25 09:36:18] @ bkero joined channel #puppet | ||
| [2010/08/25 09:36:33] <jbooth> but to do that you have to get into loading providers for what exists already on the system | ||
| [2010/08/25 09:36:37] <jbooth> which is a mess when I tried it. | ||
| [2010/08/25 09:36:48] <morpheus> doesnt work without the chomp | ||
| [2010/08/25 09:37:06] <jbooth> huh | ||
| [2010/08/25 09:37:08] <jbooth> shrug | ||
| [2010/08/25 09:37:50] @ Quit: toi: Ping timeout: 264 seconds | ||
| [2010/08/25 09:39:30] <joe-mac> yea open with .each has newlines i think | ||
| [2010/08/25 09:39:33] <joe-mac> maybe .readlines doesn't? | ||
| [2010/08/25 09:39:41] <joe-mac> easily tested with irb, buty i am lazy :-P | ||
| [2010/08/25 09:41:03] @ Diranged1 joined channel #puppet | ||
| [2010/08/25 09:41:15] @ Quit: MattyM: Remote host closed the connection | ||
| [2010/08/25 09:41:31] @ ecapriolo joined channel #puppet | ||
| [2010/08/25 09:42:27] <morpheus> I think that writing custom types is easy if once understand what hooks puppet calls and what helperfunctions (like parsefile) exist | ||
| [2010/08/25 09:42:48] <morpheus> unfortunately i don't ;-) | ||
| [2010/08/25 09:43:01] @ Quit: andrewh: Read error: Connection reset by peer | ||
| [2010/08/25 09:44:02] <morpheus> but the last 4 hours shoot to hell because one simple line is probably just because of my lack of ruby skills | ||
| [2010/08/25 09:44:10] @ Quit: Diranged: Ping timeout: 272 seconds | ||
| [2010/08/25 09:45:07] <joe-mac> even readlines takes the \n with it | ||
| [2010/08/25 09:45:20] <joe-mac> i guess just chomping is the way to go, i thought there was a method that didn't read the new lines though | ||
| [2010/08/25 09:45:38] @ zambaboo joined channel #puppet | ||
| [2010/08/25 09:46:20] <morpheus> jbooth: is this the approach you meant: http://projects.puppetlabs.com/projects/1/wiki/Development_Practical_Types | ||
| [2010/08/25 09:47:01] @ Quit: alcy: Quit: KVIrc Insomnia 4.0.0, revision: 3900, sources date: 20100125, built on: 2010-02-15 08:27:55 UTC http://www.kvirc.net/ | ||
| [2010/08/25 09:47:12] <morpheus> with retrieve reporting insync or not? | ||
| [2010/08/25 09:47:15] @ mikey_p joined channel #puppet | ||
| [2010/08/25 09:47:29] @ Quit: ber_Toulouse: Ping timeout: 252 seconds | ||
| [2010/08/25 09:47:56] <zambaboo> hi all | ||
| [2010/08/25 09:48:33] <morpheus> but I dont wouldnt know if this would help me if there is just one part (say user or just the comment) out of sync | ||
| [2010/08/25 09:48:48] <zambaboo> whats the recommended method of dealing with different ips for test and prod environments? | ||
| [2010/08/25 09:48:51] <morpheus> zamboo: hi | ||
| [2010/08/25 09:49:36] <morpheus> I hope your servers have different IPs | ||
| [2010/08/25 09:49:52] <zambaboo> morpheus: yes they do | ||
| [2010/08/25 09:50:08] <zambaboo> im wondering how to deal with that from a configuration standpoint | ||
| [2010/08/25 09:50:19] @ mclarke joined channel #puppet | ||
| [2010/08/25 09:50:51] <morpheus> do you want to configure IPs with puppet? | ||
| [2010/08/25 09:51:08] <zambaboo> yes please. | ||
| [2010/08/25 09:51:29] <nathanr> any puppet+mac users in here? have a question about the best way to ensure mounts on bootup | ||
| [2010/08/25 09:51:30] <morpheus> or do you want to decide which environment the server is in by looking at his IP? | ||
| [2010/08/25 09:52:00] <zambaboo> morpheus: id like decide on the ip based on the FACTER_ENV setting | ||
| [2010/08/25 09:52:12] @ Quit: shug: Quit: Leaving | ||
| [2010/08/25 09:53:13] @ Quit: mikey_p: Quit: mikey_p | ||
| [2010/08/25 09:53:34] @ mikey_p joined channel #puppet | ||
| [2010/08/25 09:53:47] <morpheus> I'm no expert so I have to ask: what do you mean by the "FACTER_ENV" setting | ||
| [2010/08/25 09:54:23] <zambaboo> er :) | ||
| [2010/08/25 09:55:15] <zambaboo> facter returns a bunch of facts about the client, environment variables that start with 'FACTER_whatever' are considered to be useful for puppet. | ||
| [2010/08/25 09:58:04] <nathanr> I don't get what you mean there | ||
| [2010/08/25 09:58:24] <hajile> anyone know if you can use if statements within a user {} block.. like this: http://pastie.org/1115523 | ||
| [2010/08/25 09:58:26] <nathanr> I can use everything from facter in puppet, without any FACTER_ | ||
| [2010/08/25 09:59:33] <zambaboo> nathanr: you can set an env variable, for example $FACTER_test="123", and facter will show it. | ||
| [2010/08/25 09:59:51] <zambaboo> at least thats what the doc says | ||
| [2010/08/25 09:59:53] <nathanr> instead of using custom facts? | ||
| [2010/08/25 10:00:27] <zambaboo> i.. suppose? | ||
| [2010/08/25 10:00:37] <zambaboo> i jsut read the docs, i havent even implemented anything yet | ||
| [2010/08/25 10:00:39] <nathanr> If you're talking about environment variables, I don't see any difference in facter from doing FACTER_blahh=something; facter | ||
| [2010/08/25 10:00:46] <nathanr> what are you trying to do exactly | ||
| [2010/08/25 10:01:00] @ Quit: yannL: Read error: Connection reset by peer | ||
| [2010/08/25 10:01:07] <morpheus> hajile: use a selector groups => $host ? { 'web' => "apache", default => undef}, | ||
| [2010/08/25 10:01:18] @ Quit: littleidea: Ping timeout: 265 seconds | ||
| [2010/08/25 10:01:52] <hajile> ok, and 'groups' can be 'undef'? | ||
| [2010/08/25 10:01:52] <zambaboo> i have a dev, test, and prod environments. i want to decide on the ip address of the host based on the environment in which it resides. | ||
| [2010/08/25 10:02:04] <hajile> it just can't be '', right? | ||
| [2010/08/25 10:02:27] <nathanr> where are you setting the environment? cron task? | ||
| [2010/08/25 10:02:29] <morpheus> nathanr: export FACTER_blahh and you will see it | ||
| [2010/08/25 10:02:55] <nathanr> morpheus: ahh yes there it is, interesting | ||
| [2010/08/25 10:03:56] <nathanr> zambaboo: without knowing more about your set up I would say that at the least you can have a custom fact | ||
| [2010/08/25 10:03:59] <zambaboo> nathanr: well, i can set it just it in bashrc for the user that runs facter | ||
| [2010/08/25 10:04:08] <zambaboo> or can i? | ||
| [2010/08/25 10:04:24] @ lak joined channel #puppet | ||
| [2010/08/25 10:04:36] <nathanr> you want puppet to make changes based on it right? so how are you scheduling puppet? | ||
| [2010/08/25 10:04:38] @ Quit: vad3R: Ping timeout: 276 seconds | ||
| [2010/08/25 10:05:25] <zambaboo> er.. why does this matter? | ||
| [2010/08/25 10:05:32] <morpheus> if the same user runs puppet than I guess you can make decissions in you manifests based on that fact | ||
| [2010/08/25 10:05:46] <nathanr> ask someone else if you don't want to share the most basic info | ||
| [2010/08/25 10:05:54] <nathanr> rediculous | ||
| [2010/08/25 10:06:01] @ Quit: Bass10: Ping timeout: 265 seconds | ||
| [2010/08/25 10:06:07] <morpheus> so you have an environment like FACTER_myip=10.0.0.1 | ||
| [2010/08/25 10:06:18] <morpheus> and want to change your IP based on that fact? | ||
| [2010/08/25 10:06:52] <nathanr> export FACTER_group=$(command to find out which environment its in which I can't show you since you wont say how it's scheduled) | ||
| [2010/08/25 10:07:25] @ Quit: TREllis: Quit: ukuleleeeeeeeeeee | ||
| [2010/08/25 10:08:16] <morpheus> cause puppet doesnt provide a native provider for IP-Adresses this can be pretty hard. But you can use exec and ifconfig or appropiate calls | ||
| [2010/08/25 10:08:26] <morpheus> But thats maybe not the best way to do things | ||
| [2010/08/25 10:09:26] <morpheus> so how is FACTER_foo set and what do you want to do with it? | ||
| [2010/08/25 10:09:53] <nathanr> believe he wants to have puppet adjust his ip based on what environment the machine is in | ||
| [2010/08/25 10:10:16] <morpheus> yeah but you dont want every machine share the same ip right? | ||
| [2010/08/25 10:10:37] <nathanr> well he can just use fqdn_rand for the ip I'm sure | ||
| [2010/08/25 10:10:44] <nathanr> or rather | ||
| [2010/08/25 10:10:56] <nathanr> another sort of method | ||
| [2010/08/25 10:11:02] <nathanr> that wouldn't work since it could conflict | ||
| [2010/08/25 10:11:07] @ aroth joined channel #puppet | ||
| [2010/08/25 10:11:15] <nathanr> who knows, he's not talking | ||
| [2010/08/25 10:12:18] <morpheus> I'm just curious about the goal | ||
| [2010/08/25 10:12:33] <nathanr> it sounds like he doesn't know how to use dhcp | ||
| [2010/08/25 10:12:35] <nathanr> =] | ||
| [2010/08/25 10:13:13] <morpheus> because in my imagination puppet is configuring IP addresses and after that nothing is working and you cannot reach your servers anymore | ||
| [2010/08/25 10:13:45] <morpheus> but maybe puppet should just report "Hey your server has an IP he doesnt supposed to have because he belongs to ..." | ||
| [2010/08/25 10:13:49] <hajile> morpheus: it doesn't look like the undef stuff is working for groups | ||
| [2010/08/25 10:13:51] <nathanr> maybe it's already on the same subnet and he has logical splitting | ||
| [2010/08/25 10:14:05] <hajile> i'll have to look at it in a bit though.. are you going to be around in about an hour? | ||
| [2010/08/25 10:14:07] <morpheus> ok then try '[undef]' or '[]' | ||
| [2010/08/25 10:14:11] <hajile> ok | ||
| [2010/08/25 10:14:12] <hajile> cool | ||
| [2010/08/25 10:14:15] <hajile> thank you! | ||
| [2010/08/25 10:14:43] @ littleidea joined channel #puppet | ||
| [2010/08/25 10:14:47] <morpheus> i think I'm not the only want who can help you here ;-) | ||
| [2010/08/25 10:14:50] <hajile> true | ||
| [2010/08/25 10:14:57] <hajile> but you're helping and not giving me shit ;) | ||
| [2010/08/25 10:15:19] <nathanr> the problem with 'easy' operating systems is that no one in chat rooms knows a damn thing about them | ||
| [2010/08/25 10:15:21] <hajile> not that everyone does this, but still :) | ||
| [2010/08/25 10:15:22] <nathanr> annoying | ||
| [2010/08/25 10:16:37] <hajile> weird.. it's not picking up the change | ||
| [2010/08/25 10:17:26] <hajile> ok bbiaf | ||
| [2010/08/25 10:17:31] <hajile> thank you for your help so far! | ||
| [2010/08/25 10:18:04] <morpheus> nathanr: I cant follow you | ||
| [2010/08/25 10:18:06] @ danj left channel #puppet () | ||
| [2010/08/25 10:18:30] @ grape joined channel #puppet | ||
| [2010/08/25 10:19:15] @ Quit: littleidea: Remote host closed the connection | ||
| [2010/08/25 10:19:20] @ littleidea joined channel #puppet | ||
| [2010/08/25 10:19:25] @ toi joined channel #puppet | ||
| [2010/08/25 10:20:13] <zambaboo> sorry had a standup | ||
| [2010/08/25 10:20:44] <zambaboo> ok you guys so here's an example im trying to work with | ||
| [2010/08/25 10:21:21] <zambaboo> you have a prod env, and a dev env. in dev you test your manifests, and if they work you commit to source control. | ||
| [2010/08/25 10:21:41] <zambaboo> now. you do NOT want to modify the manifests before you apply them to production. | ||
| [2010/08/25 10:21:57] <nathanr> just blathering. channels like #ubuntu and #mac are practically impossible to actually get help in | ||
| [2010/08/25 10:22:01] @ Quit: littleidea: Client Quit | ||
| [2010/08/25 10:22:22] <zambaboo> yeah #ubuntu has a zillion people in it, all asking about running photoshop in wine. | ||
| [2010/08/25 10:22:34] <nathanr> haha | ||
| [2010/08/25 10:22:41] @ Quit: aroth: Ping timeout: 240 seconds | ||
| [2010/08/25 10:22:42] @ Quit: madduck_: Read error: Operation timed out | ||
| [2010/08/25 10:23:28] <zambaboo> SO your manifests set the IP addresses for your servers. what is the recommended way to decide on an IP address based on the environment in which the system resides? | ||
| [2010/08/25 10:23:33] @ mrwacky42 joined channel #puppet | ||
| [2010/08/25 10:23:59] @ madduck joined channel #puppet | ||
| [2010/08/25 10:24:08] @ Quit: madduck: Changing host | ||
| [2010/08/25 10:24:08] @ madduck joined channel #puppet | ||
| [2010/08/25 10:25:01] <nathanr> well like morpheus was hinting at | ||
| [2010/08/25 10:25:10] <morpheus> I dont know if managing IP adresses should be done by puppet | ||
| [2010/08/25 10:25:15] <nathanr> you need to have more information than just the environment | ||
| [2010/08/25 10:25:30] <nathanr> otherwise you have 40 machines trying for 2 ips | ||
| [2010/08/25 10:25:40] <zambaboo> morpheus: oh? an interesting notion. | ||
| [2010/08/25 10:25:43] <nathanr> also, if the ip is wrong in the first place, how are they talking to puppet | ||
| [2010/08/25 10:26:00] <morpheus> i dont know if puppet agent needs a stable connection to the master from start to end and changing the address in the middle of the run can be a bad idea | ||
| [2010/08/25 10:26:00] <zambaboo> good point indeed, nathanr | ||
| [2010/08/25 10:26:34] <zambaboo> most likely not, we're working with https here | ||
| [2010/08/25 10:28:33] @ bitmand joined channel #puppet | ||
| [2010/08/25 10:29:44] @ Quit: wajasu: Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100724065753] | ||
| [2010/08/25 10:30:53] <bitmand> I just got puppet up and running - 1 master, 3 agents, _very_ small and simple setup atm. But I am wondering, can I somehow tell the agent to checke for changes with the master? I tried kill -HUP - but nothing happens. And restarting the client seems a bit overkill | ||
| [2010/08/25 10:31:23] @ Quit: goldfish: Ping timeout: 258 seconds | ||
| [2010/08/25 10:32:19] <joe-mac> bitmand: you can enable 'listen = true' in the config file and use puppetrun, now called 'puppet kick' | ||
| [2010/08/25 10:32:46] <fsweetser> bitmand: also, sending it -USR1 will cause it to do an immediate run | ||
| [2010/08/25 10:33:40] @ littleidea joined channel #puppet | ||
| [2010/08/25 10:33:50] @ shadoi joined channel #puppet | ||
| [2010/08/25 10:33:50] <bitmand> thank you both :) | ||
| [2010/08/25 10:36:10] @ aroth joined channel #puppet | ||
| [2010/08/25 10:41:34] @ Quit: geekatcmu: Quit: Time to go home | ||
| [2010/08/25 10:42:01] @ geekatcmu joined channel #puppet | ||
| [2010/08/25 10:43:44] <threescoops> Is all puppetd <-> puppetmaster traffic over port 8140? | ||
| [2010/08/25 10:44:22] <threescoops> Having some issues with a remote machine and just want to rule out the obvious firewall issues | ||
| [2010/08/25 10:44:23] @ Dantman joined channel #puppet | ||
| [2010/08/25 10:46:05] <threescoops> Ignore me, that's not the problem :) | ||
| [2010/08/25 10:46:08] @ Bass10 joined channel #puppet | ||
| [2010/08/25 10:48:00] @ reyjrar joined channel #puppet | ||
| [2010/08/25 10:48:01] <nathanr> done | ||
| [2010/08/25 10:50:09] @ Quit: jdcasey: Remote host closed the connection | ||
| [2010/08/25 10:51:34] @ Quit: benkevan: Quit: leaving | ||
| [2010/08/25 10:52:36] @ WALoeIII joined channel #puppet | ||
| [2010/08/25 10:52:45] @ Quit: yure: Ping timeout: 246 seconds | ||
| [2010/08/25 10:53:39] <Diranged1> ok.. my puppet-master build definition works.. but ive got an issue. a bunch of links it uses to populate files point to puppet:///. if i use "puppet --modulepath=$SVNDEST/modules $SVNDEST/manifests/site.pp —pluginsync" to cause the server to run its puppet config on itself (which self-configures all the puppetmasterd stuff, and a bunch of other crap), it fails | ||
| [2010/08/25 10:54:09] <Diranged1> err: /Stage[main]/Idc-basenode/Auth::Ldap[default_ldap_config]/File[/etc/nsswitch.conf]: Could not evaluate: getaddrinfo: Name or service not known Could not retrieve file metadata for puppet:///idcadmin/auth/CentOS/nsswitch.conf: getaddrinfo: Name or service not known at /etc/puppet//modules/auth/manifests/init.pp:111 | ||
| [2010/08/25 10:54:15] <Diranged1> that kind of error.. | ||
| [2010/08/25 10:55:28] <zambaboo> Diranged1: not knowing ANYTHING about this, i could point at the getaddrinfo and say that something with your name resolution is messed up. | ||
| [2010/08/25 10:56:19] <Diranged1> whats it trying to look up though? notice the url is just puppet:/// … shouldnt that, when running against a local puppet source, simply get the data from the local tree? | ||
| [2010/08/25 10:56:40] <zambaboo> thaaaaaaaaaats a question for the docs | ||
| [2010/08/25 10:57:22] <Diranged1> ok | ||
| [2010/08/25 10:57:25] @ jab_doa joined channel #puppet | ||
| [2010/08/25 10:58:02] <Diranged1> another one.. can i make module a require module b? | ||
| [2010/08/25 10:58:18] <zambaboo> docs bro | ||
| [2010/08/25 10:58:53] <nathanr> I'm wondering, if adding a fact is as easy as exporting a variable, what is the point of doing all the steps on the adding facts page | ||
| [2010/08/25 10:59:50] <zambaboo> nathanr: vars might not be as reliable? | ||
| [2010/08/25 11:00:04] <zambaboo> zambaboo: your question about the setting time really made me think about it. | ||
| [2010/08/25 11:00:08] @ Quit: geekatcmu: Quit: Time to go home | ||
| [2010/08/25 11:00:20] @ adrian_broher joined channel #puppet | ||
| [2010/08/25 11:00:42] <nathanr> did you just call yourself in the second person | ||
| [2010/08/25 11:01:12] <zambaboo> er. | ||
| [2010/08/25 11:01:18] <zambaboo> yes, i have. | ||
| [2010/08/25 11:01:30] * zambaboo moves away from the coffee | ||
| [2010/08/25 11:02:40] <nathanr> did you just past particible a past verb | ||
| [2010/08/25 11:02:59] <zambaboo> indeed i have. | ||
| [2010/08/25 11:03:13] <zambaboo> you're strict, nathanr. | ||
| [2010/08/25 11:03:14] @ geekatcmu joined channel #puppet | ||
| [2010/08/25 11:03:55] <morpheus> You may want to add a fact because you dont have to export variables on ALL your nodes | ||
| [2010/08/25 11:04:19] <morpheus> and a variable is static and a fact can be dynamic | ||
| [2010/08/25 11:04:45] <nathanr> did you just say a variable is static | ||
| [2010/08/25 11:04:54] <nathanr> that kind of goes against the definition | ||
| [2010/08/25 11:05:02] <morpheus> I think puppet:/// will imply that your puppetmaster is called puppet | ||
| [2010/08/25 11:05:10] <nathanr> anyways, in order for your nodes to use custom facts, they need to have access to the file | ||
| [2010/08/25 11:05:25] <nathanr> which in my case means an nfs mount or more recently just pushing the files down | ||
| [2010/08/25 11:05:32] <morpheus> I think its the same when you dont set server in your config. Puppet will suppose your server is called puppet | ||
| [2010/08/25 11:05:37] <nathanr> seems like less work to just have bashrc for example set the variable | ||
| [2010/08/25 11:05:48] @ pting joined channel #puppet | ||
| [2010/08/25 11:06:22] @ yure joined channel #puppet | ||
| [2010/08/25 11:06:34] <jbooth> morpheus: Yes, the puppet type dev stuff. There's like, 3 pages on it. Practical, the a couple on types and providers. | ||
| [2010/08/25 11:06:34] <nathanr> for example if I have all vlans numbered in line with their subnets, for example vlan 101 machines will be 192.168.101.x, and I want a custom fact for vlan | ||
| [2010/08/25 11:07:01] <nathanr> I could have something like FACTER_vlan=$(facter ipaddress | cut -d. -f3) | ||
| [2010/08/25 11:07:05] <jbooth> also, puppet:/// will talk to whoever you're talking to -- not puppet persay. It is akin to puppet://$servername/ | ||
| [2010/08/25 11:08:11] <zambaboo> nathanr: yeah man thats exactly the stuff im talking about. | ||
| [2010/08/25 11:08:11] <nathanr> seems a lot easier than making a file, having it be a custom fact, and having to set the RUBYLIB (or whatever) environment variable _anyways_ | ||
| [2010/08/25 11:08:23] <zambaboo> precisely. | ||
| [2010/08/25 11:08:37] <nathanr> zambaboo: I'm not helping you because I asked you for the slightest information which I still don't have | ||
| [2010/08/25 11:08:57] * zambaboo scrolls back | ||
| [2010/08/25 11:08:58] <nathanr> I don't help people who try to make it difficult for me to do so, at least not for free =] | ||
| [2010/08/25 11:09:05] <zambaboo> im sorry what are you talking about | ||
| [2010/08/25 11:10:02] <zambaboo> oh the scheduling bit | ||
| [2010/08/25 11:10:25] <zambaboo> im not sure about that yet, i was under the impression that puppet magically does its thing every thirty minutes by dafult. | ||
| [2010/08/25 11:10:27] <zambaboo> default. | ||
| [2010/08/25 11:11:00] <jbooth> Meh. You dump the fact in modules/local/lib/facter/<fact>.rb and let pluginsync distribute it. Done. | ||
| [2010/08/25 11:12:03] @ Quit: VAlu: Ping timeout: 272 seconds | ||
| [2010/08/25 11:13:17] <morpheus> if you put in in .bashrc and than run puppet in daemon mode you have no chance to change its environment afterwards | ||
| [2010/08/25 11:13:23] <morpheus> thats what I meant by static | ||
| [2010/08/25 11:13:33] @ VAlu joined channel #puppet | ||
| [2010/08/25 11:13:48] <morpheus> try realize the uptime fact as an environment variable | ||
| [2010/08/25 11:17:21] <nathanr> zambaboo: there is no magic in nix | ||
| [2010/08/25 11:17:39] <nathanr> you can run it as a service which would do the default check ins | ||
| [2010/08/25 11:17:41] <zambaboo> nathanr: by magic i mean an internal schedular. | ||
| [2010/08/25 11:17:49] <zambaboo> scheduler | ||
| [2010/08/25 11:17:51] <nathanr> or you can decide the frequency | ||
| [2010/08/25 11:18:02] <zambaboo> (what is _up_ with my spelling today) | ||
| [2010/08/25 11:18:17] <nathanr> for example all my nodes are once an hour, spread out over the hour | ||
| [2010/08/25 11:18:41] <nathanr> how do you know which machine is in which environment | ||
| [2010/08/25 11:18:48] <nathanr> also | ||
| [2010/08/25 11:19:05] <nathanr> there are environments in puppet. are you talking about actual environments or puppet's --environment | ||
| [2010/08/25 11:19:22] <zambaboo> actual environments. | ||
| [2010/08/25 11:19:27] <nathanr> ok. so first of all | ||
| [2010/08/25 11:19:33] <Diranged1> I dont think my issue with puppet:/// not resolving is a dns issue.. I think it may be a syntax issue or something. When using puppet:/// as a reference to a file, does it not even check whether "puppet:///" refers to a locally run puppet command against a local checkout of the code? | ||
| [2010/08/25 11:19:34] <zambaboo> i have vm farm that mimics production | ||
| [2010/08/25 11:19:38] <nathanr> I would recommend setting up a puppet environment for each actual environment | ||
| [2010/08/25 11:19:51] <nathanr> for example here I have a dev and prod environment | ||
| [2010/08/25 11:20:09] <nathanr> I can change any node to the development branch of puppet manifests, modules and files by just changing | ||
| [2010/08/25 11:20:12] @ jubei__ joined channel #puppet | ||
| [2010/08/25 11:20:23] <nathanr> --environment=production to --environment=development | ||
| [2010/08/25 11:20:31] <zambaboo> right, right | ||
| [2010/08/25 11:20:32] <nathanr> don't have to mess with the server or anything else | ||
| [2010/08/25 11:20:44] <nathanr> ok so here's where it's all going to come together for you | ||
| [2010/08/25 11:20:55] <nathanr> WHERE I DO THIS IS WHERE IT IS SCHEDULED | ||
| [2010/08/25 11:21:10] <nathanr> hence, you can determine quite easily what environment it's in by just checking that location | ||
| [2010/08/25 11:21:24] <zambaboo> ok so you schedule this stuff in your crontab? | ||
| [2010/08/25 11:21:27] <nathanr> my personal recommendation is to always use cron for your puppet check ins | ||
| [2010/08/25 11:21:31] <nathanr> yes like so: | ||
| [2010/08/25 11:21:55] <zambaboo> (mind you, i have NOT implemented anything yet, i just have the vms installed and the networking stuff sorted out.) | ||
| [2010/08/25 11:22:02] <Diranged1> ugh: http://projects.puppetlabs.com/issues/4489 | ||
| [2010/08/25 11:23:03] <nathanr> http://pastebin.com/5eqKPtPL | ||
| [2010/08/25 11:23:47] <nathanr> what that says is that this node will check in with production environment once an hour | ||
| [2010/08/25 11:23:48] <zambaboo> nathanr: i see! | ||
| [2010/08/25 11:24:04] <nathanr> if I wanted to see programmatically what environment this node was on I could have a fact | ||
| [2010/08/25 11:24:08] @ Quit: ayanich: Ping timeout: 276 seconds | ||
| [2010/08/25 11:25:04] <nathanr> export FACTER_enviro=$(grep -o production /etc/cron.d/puppet-checkin || echo development) | ||
| [2010/08/25 11:25:12] <nathanr> now I can say facter enviro | ||
| [2010/08/25 11:25:18] @ Quit: kaptk2: Ping timeout: 240 seconds | ||
| [2010/08/25 11:25:20] <nathanr> and it will either say production or development | ||
| [2010/08/25 11:25:23] @ ayanich joined channel #puppet | ||
| [2010/08/25 11:25:30] <zambaboo> i see | ||
| [2010/08/25 11:25:36] <zambaboo> makes sense indeed | ||
| [2010/08/25 11:25:48] @ Quit: shog: Ping timeout: 240 seconds | ||
| [2010/08/25 11:25:52] <nathanr> that's not foolproof, it will also say development if that file doesn't exist for example | ||
| [2010/08/25 11:25:55] <nathanr> but anyways you get the point | ||
| [2010/08/25 11:26:39] <zambaboo> i do indeed | ||
| [2010/08/25 11:26:42] @ kaptk2 joined channel #puppet | ||
| [2010/08/25 11:26:47] <FiXion> morpheus: you could probably use the mount provider - it's handling of the fstab is very similar to your needs | ||
| [2010/08/25 11:26:58] <FiXion> morpheus: as a code example is what I mean :) | ||
| [2010/08/25 11:33:10] <hajile> morpheus: it worked, thanks again for your help. I had some bad logic left elsewhere in the manifest. | ||
| [2010/08/25 11:33:51] @ Ramonster joined channel #puppet | ||
| [2010/08/25 11:34:28] @ Quit: bug: Ping timeout: 252 seconds | ||
| [2010/08/25 11:34:28] @ thekad is now known as thekad-afk | ||
| [2010/08/25 11:34:45] @ bug joined channel #puppet | ||
| [2010/08/25 11:34:51] @ nakano is now known as nakano_ | ||
| [2010/08/25 11:35:10] <Diranged1> I guess due to bug 4489, i have to first install puppet .25 from the epel library.. then do my puppet run against my local source, and have that puppet run self-update after its got the apache server up and running | ||
| [2010/08/25 11:35:41] @ thekad-afk is now known as thekad | ||
| [2010/08/25 11:38:47] @ shog joined channel #puppet | ||
| [2010/08/25 11:38:51] @ Quit: lak: Quit: lak | ||
| [2010/08/25 11:39:41] @ Quit: Bass10: Ping timeout: 240 seconds | ||
| [2010/08/25 11:40:08] @ ryep joined channel #puppet | ||
| [2010/08/25 11:41:09] <morpheus> nathanr: why not using $environment? | ||
| [2010/08/25 11:41:18] <morpheus> this works in puppet 2.6.0 | ||
| [2010/08/25 11:42:27] <morpheus> FiXion: I looked at ssh_auhorized_key, mount and host and I always come up with the conclusion: There is more than one way to do it | ||
| [2010/08/25 11:43:30] @ Quit: aroth: Quit: Leaving | ||
| [2010/08/25 11:44:55] <morpheus> the problem with realizing with parsedfile I have to join and split the arrays back and forth and I just couldnt get it to work | ||
| [2010/08/25 11:45:52] @ nakano_ is now known as nakano | ||
| [2010/08/25 11:45:52] @ bitmand left channel #puppet () | ||
| [2010/08/25 11:46:50] @ Quit: gebi: Ping timeout: 264 seconds | ||
| [2010/08/25 11:47:21] <morpheus> I tried with record_line :post_parse and :pre_gen but I didnt understand it | ||
| [2010/08/25 11:48:49] <morpheus> and host provider for example has a self.to_line function in it and I guess thats a hook for record_line but its really hard to figure all this things out | ||
| [2010/08/25 11:49:18] <morpheus> especially if you dont know any ruby. | ||
| [2010/08/25 11:49:40] @ bmahe joined channel #puppet | ||
| [2010/08/25 11:49:50] @ benkevan joined channel #puppet | ||
| [2010/08/25 11:51:58] <morpheus> and when I browse through the different providers I sometimes see resource[:option], @resource[:option], resource.value(:option) and resource.should(:option) | ||
| [2010/08/25 11:52:29] <morpheus> at least for me this is just confusing because I dont know the difference | ||
| [2010/08/25 11:56:25] <tholl583xx> hi | ||
| [2010/08/25 11:57:06] <tholl583xx> im getting nothing in my puppet-dashboard, i have report=true in [puppet] and reports = http, store | ||
| [2010/08/25 11:57:15] <tholl583xx> any suggestions on troubleshooting? | ||
| [2010/08/25 12:00:27] <morpheus> think the resports will not be stored in the db of puppet-dashboard per default | ||
| [2010/08/25 12:00:43] <morpheus> do you have something in /var/lib/puppet/reports? | ||
| [2010/08/25 12:01:42] <morpheus> when i last tested the dashboard there was some rake target for importing reports I think I read somewhere that you can integrate the reporting | ||
| [2010/08/25 12:04:07] <morpheus> just google around. maybe heres an answer: http://zcentric.com/tag/puppet-dashboard/ | ||
| [2010/08/25 12:05:28] <mrwacky42> is there a puppet module to set up an HA messaging server so I can easily enable stored configs ? | ||
| [2010/08/25 12:05:45] <jbooth> tholl583xx: Did you do a smart thing and put dashboard on https or similar, expecting to secure it behind a login mechanism? | ||
| [2010/08/25 12:12:21] <tholl583xx> jbooth: nope | ||
| [2010/08/25 12:14:17] @ Bass10 joined channel #puppet | ||
| [2010/08/25 12:15:41] @ Quit: mikey_p: Ping timeout: 240 seconds | ||
| [2010/08/25 12:16:05] @ bug_ joined channel #puppet | ||
| [2010/08/25 12:16:16] <jbooth> tholl583xx: Did you distribute the dashboard reporting module to the nodes? | ||
| [2010/08/25 12:16:26] <jbooth> er, maybe just to the puppetmaster | ||
| [2010/08/25 12:16:31] <jbooth> I never quite got clear on that | ||
| [2010/08/25 12:18:53] @ Quit: bug: Ping timeout: 245 seconds | ||
| [2010/08/25 12:18:53] @ bug_ is now known as bug | ||
| [2010/08/25 12:19:14] <telmich> if I | ||
| [2010/08/25 12:19:30] @ jdcasey joined channel #puppet | ||
| [2010/08/25 12:20:57] <telmich> if I've a basedir, let's say www/nico, which requires the user nico and a define that creates subdirectories below www/nico/$name/, at which point would I realise the user nico? | ||
| [2010/08/25 12:21:35] @ shtylman joined channel #puppet | ||
| [2010/08/25 12:21:49] <telmich> at the end, I would like to call ns_server::website_dir { "mydomain": owner => "nico" }, to create a new web instance directory for the user nico | ||
| [2010/08/25 12:21:55] <jbooth> If you haven't defined the user, never. | ||
| [2010/08/25 12:22:09] @ vad3R joined channel #puppet | ||
| [2010/08/25 12:22:17] <shtylman> the puppet docs say that if I have listen=true in my config, then puppet won't update every 30 seconds... I don't seem to be seeing that behavior tho | ||
| [2010/08/25 12:22:18] @ Quit: vad3R: Remote host closed the connection | ||
| [2010/08/25 12:22:23] <shtylman> is there something else I overlooked? | ||
| [2010/08/25 12:22:27] <tholl583xx> im not doing any configuratino on the nodes other than bootstrapping the /etc/hosts so the nodes know where the puppet server is | ||
| [2010/08/25 12:22:42] <shtylman> I would basically like to update puppet through fixed intervals (in cron) or remotely | ||
| [2010/08/25 12:22:45] <tholl583xx> i would presume that the puppetmaster sends the reports | ||
| [2010/08/25 12:22:46] <telmich> jbooth: I want to ensure that the user exists if calling ns_server::website_dir | ||
| [2010/08/25 12:22:53] @ rgsteele joined channel #puppet | ||
| [2010/08/25 12:23:36] <jbooth> telmich: Then you'd better define the user somewhere too. If you've done it as a virtual resource, you can realize it anywhere (or multiple times) | ||
| [2010/08/25 12:24:16] <jbooth> tholl583xx: Nodes don't have report=true in puppet.conf? | ||
| [2010/08/25 12:24:19] @ Quit: littleidea: Quit: littleidea | ||
| [2010/08/25 12:25:14] @ Quit: shog: Ping timeout: 276 seconds | ||
| [2010/08/25 12:26:19] <Diranged1> ok.. need some help. i want to migrate my puppet nodes to a new system where they connect to an arbitrary IP based on their network locatino, woith the hostname 'puppet' and an ssl cert signed for the hostname 'puppet'. but when i try to take an existing puppet client and run it against this new hostname, it actually seg faults | ||
| [2010/08/25 12:26:36] <telmich> hmm, it seems the creation of a user and the directory together are not even handled perfectly by virtual ressources | ||
| [2010/08/25 12:29:16] @ lak joined channel #puppet | ||
| [2010/08/25 12:29:52] <jbooth> telmich: eh? | ||
| [2010/08/25 12:30:45] <zambaboo> Diranged1: stab in the dark - is the arch the same on source and destination machines./ | ||
| [2010/08/25 12:31:26] <telmich> jbooth: at which point would I realize the virtual users? | ||
| [2010/08/25 12:31:33] <Diranged1> zam: yes | ||
| [2010/08/25 12:31:55] <Diranged1> even if i remove /var/lib/puppet from the client and start it over, it errors out: "Could not prepare for execution: Retrieved certificate does not match private key; please remove certificate from server and regenerate it with the current key" | ||
| [2010/08/25 12:31:58] <telmich> if you've a look at http://git.sans.ethz.ch/?p=puppet-modules/ns_server;a=blob;f=manifests/website_dir.pp;h=b4d95ef4a68f1a7a7f7a8ebdda9337ad6bbb5918;hb=HEAD, you see the use case | ||
| [2010/08/25 12:32:15] @ gebi joined channel #puppet | ||
| [2010/08/25 12:33:19] @ mikey_p joined channel #puppet | ||
| [2010/08/25 12:34:33] <christophermlee> anyone here familur with cobbler, I am having an issue and the cobbler channel does not respond | ||
| [2010/08/25 12:35:46] @ Quit: MPSimmons: Quit: Leaving. | ||
| [2010/08/25 12:36:09] <christophermlee> i am getting a tftpd error saying it does not accept options | ||
| [2010/08/25 12:36:25] <zambaboo> what does google say? | ||
| [2010/08/25 12:37:36] @ shog joined channel #puppet | ||
| [2010/08/25 12:38:55] <christophermlee> so far only finding people complaining about it but no resolutions... still searching | ||
| [2010/08/25 12:39:46] @ Quit: \ask: Remote host closed the connection | ||
| [2010/08/25 12:40:09] @ Quit: mikey_p: Quit: mikey_p | ||
| [2010/08/25 12:41:01] <Tas_> hmm, does anyone know how to shut down puppet-dashboard on Debian? | ||
| [2010/08/25 12:41:07] <Tas_> I can't seem to kill the pid | ||
| [2010/08/25 12:41:34] <Tas_> it shows up as 'ruby script/server -d', I find no .pid files (like the rest of puppet seems to have) etc | ||
| [2010/08/25 12:41:45] <Tas_> this makes it difficult to make an init.d script. | ||
| [2010/08/25 12:42:17] @ Quit: ezekiel: Quit: what | ||
| [2010/08/25 12:42:30] @ ezekiel joined channel #puppet | ||
| [2010/08/25 12:45:59] <christophermlee> figured it out, the pxelinux file was bad | ||
| [2010/08/25 12:48:38] @ Quit: shog: Ping timeout: 264 seconds | ||
| [2010/08/25 12:49:27] <zambaboo> christophermlee: post an answer then :) | ||
| [2010/08/25 12:49:55] <jbooth> telmich: Before you need them | ||
| [2010/08/25 12:50:24] @ RedPepper is now known as redpepper | ||
| [2010/08/25 12:50:26] <jbooth> telmich: Since puppet likes to cache users/groups it may take 2 runs before it realizes the user exists when you go to create a file and assign ownership | ||
| [2010/08/25 12:50:38] @ yannL joined channel #puppet | ||
| [2010/08/25 12:51:05] <jbooth> telmich: But neither of those are virtual, so realizing them would be meaningless | ||
| [2010/08/25 12:53:07] <jbooth> Tas_: Dashboard generally runs as a passenger app in apache, so killing it will likely just have apache restart it | ||
| [2010/08/25 12:53:40] <husimon> does puppet have "hashes | ||
| [2010/08/25 12:53:41] <husimon> " | ||
| [2010/08/25 12:54:15] <Tas_> jbooth: hmm, how can I halt it via the terminal, then? | ||
| [2010/08/25 12:54:51] @ Quit: morpheus: Quit: Leaving | ||
| [2010/08/25 12:55:10] @ Quit: [GuS]: Read error: Connection reset by peer | ||
| [2010/08/25 12:56:34] <ohadlevy> Tas_: you should find a pid in its tmp dir | ||
| [2010/08/25 12:56:37] @ Quit: Bass10: Ping timeout: 260 seconds | ||
| [2010/08/25 12:56:42] @ Quit: gebi: Ping timeout: 272 seconds | ||
| [2010/08/25 12:57:06] @ Quit: masterzen: Ping timeout: 240 seconds | ||
| [2010/08/25 12:57:11] <ohadlevy> Tas_: i think it has a startup script already, if not, you can use foreman's (they should be nearly identical) | ||
| [2010/08/25 13:05:38] @ masterzen joined channel #puppet | ||
| [2010/08/25 13:06:25] @ Quit: mrwacky42: Quit: Leaving. | ||
| [2010/08/25 13:06:56] @ shog joined channel #puppet | ||
| [2010/08/25 13:07:56] @ \ask joined channel #puppet | ||
| [2010/08/25 13:09:39] @ littleidea joined channel #puppet | ||
| [2010/08/25 13:09:41] <Diranged1> ugh i think im getting closer.. had a rack/passenger version issue that was causing some issues. ok new puppet client connecting to new puppet server. new pupet server uses a self-signed .key and .crt file with a really generic hostname (puppet). when i connect, it gets "Exiting; no certificate found and waitforcert is disabled" on the client | ||
| [2010/08/25 13:10:04] <Diranged1> i see "Aug 25 13:08:35 vdc01ppt01 puppet-master[17086]: Could not find certificate for 'puppet-client'" on the master host.. | ||
| [2010/08/25 13:11:33] @ Quit: viewlogic: Ping timeout: 260 seconds | ||
| [2010/08/25 13:12:09] @ MarkN1 joined channel #puppet | ||
| [2010/08/25 13:12:14] @ Quit: geekatcmu: Quit: Leaving. | ||
| [2010/08/25 13:12:53] @ geekatcmu joined channel #puppet | ||
| [2010/08/25 13:13:49] @ gebi joined channel #puppet | ||
| [2010/08/25 13:14:17] @ vad3R joined channel #puppet | ||
| [2010/08/25 13:14:19] @ Quit: vad3R: Remote host closed the connection | ||
| [2010/08/25 13:19:26] <tholl583xx> my puppet clients are definitely sending reports, as I can see them in the /var/lib/puppet/reports foldefr | ||
| [2010/08/25 13:19:30] @ Quit: toi: Quit: Ex-Chat | ||
| [2010/08/25 13:19:30] <tholl583xx> folder | ||
| [2010/08/25 13:19:33] @ Quit: bug: Quit: bug | ||
| [2010/08/25 13:23:06] <[ack]> is there a way for me to clients to certain environments? | ||
| [2010/08/25 13:23:12] <tholl583xx> there looks to be no http.rb in the /usr/lib/ruby/site_ruby/1.8/puppet/reports folder though | ||
| [2010/08/25 13:23:36] <cowboycoder> is there a way to make one class dependant on another? | ||
| [2010/08/25 13:23:37] @ lkoranda joined channel #puppet | ||
| [2010/08/25 13:24:05] @ mocker left channel #puppet ("WeeChat 0.3.2") | ||
| [2010/08/25 13:26:43] <christophermlee> zambaboo: the answer was I don't understand how cobbler was working. I unchecked a box I should have left checked. None of the google posts were specific about this problem. the guys over in the cobbler group finally responded and got me straight | ||
| [2010/08/25 13:29:24] @ mikey_p joined channel #puppet | ||
| [2010/08/25 13:32:05] @ Quit: mikey_p: Client Quit | ||
| [2010/08/25 13:32:33] @ Quit: shog: Ping timeout: 260 seconds | ||
| [2010/08/25 13:35:39] @ mikey_p joined channel #puppet | ||
| [2010/08/25 13:38:58] @ shog joined channel #puppet | ||
| [2010/08/25 13:40:01] @ Quit: bmahe: Quit: Leaving | ||
| [2010/08/25 13:40:06] @ bug joined channel #puppet | ||
| [2010/08/25 13:40:13] @ KB1JWQ joined channel #puppet | ||
| [2010/08/25 13:40:21] <KB1JWQ> Where in the configuration does the address for the puppetmaster generally live? | ||
| [2010/08/25 13:41:05] @ Quit: mikey_p: Ping timeout: 276 seconds | ||
| [2010/08/25 13:43:11] <geekatcmu> Depends. | ||
| [2010/08/25 13:43:20] <geekatcmu> puppetd looks in [puppetd]. | ||
| [2010/08/25 13:43:44] <geekatcmu> Look for "server = my.server.name" | ||
| [2010/08/25 13:48:33] <KB1JWQ> Thanks, digging. | ||
| [2010/08/25 13:48:42] <zambaboo> nathanr: so there's another way to do what we were talking about. tags. | ||
| [2010/08/25 13:50:13] @ morpheus joined channel #puppet | ||
| [2010/08/25 13:52:18] @ nul0op joined channel #puppet | ||
| [2010/08/25 13:53:48] @ cwebber joined channel #puppet | ||
| [2010/08/25 13:55:22] @ garfias joined channel #puppet | ||
| [2010/08/25 13:56:32] <garfias> how does one raise the db connection pool size? | ||
| [2010/08/25 13:57:28] @ mikey_p joined channel #puppet | ||
| [2010/08/25 13:58:13] <mackn> what db pool are you talking about? | ||
| [2010/08/25 13:58:52] <garfias> well, thats what i'd like to know. Setup stored configs and now i'm getting an error saying i need to up my pool size | ||
| [2010/08/25 13:59:21] <mackn> it's probably the pool size of the mysql database you are using for the stored config | ||
| [2010/08/25 13:59:26] <mackn> or whatever db backend you are using | ||
| [2010/08/25 13:59:38] <garfias> mysql, but whatever. Where do I change this? | ||
| [2010/08/25 14:00:13] <garfias> i don't see anything in puppet.conf's man page about it | ||
| [2010/08/25 14:00:18] @ Quit: cwebber: Read error: Connection reset by peer | ||
| [2010/08/25 14:03:22] <mackn> which version of puppet are you running/ | ||
| [2010/08/25 14:03:35] <garfias> .25.4 | ||
| [2010/08/25 14:03:44] <garfias> from deb backports | ||
| [2010/08/25 14:04:16] <mackn> there's supposedly a dbconnections option | ||
| [2010/08/25 14:04:20] @ Quit: mikey_p: Quit: mikey_p | ||
| [2010/08/25 14:04:22] <mackn> not sure which version it appeared in | ||
| [2010/08/25 14:04:38] <garfias> i tried: dbconnections = 256 | ||
| [2010/08/25 14:04:41] <garfias> no luck there | ||
| [2010/08/25 14:05:06] <garfias> i found a bug report for a patch, but i wasn't sure if it'd got merged in | ||
| [2010/08/25 14:06:12] @ Quit: joe-mac: Quit: Leaving. | ||
| [2010/08/25 14:06:12] <mackn> looks like it's in 2.6 | ||
| [2010/08/25 14:06:18] <garfias> damn | ||
| [2010/08/25 14:06:21] @ Quit: nul0op: Ping timeout: 265 seconds | ||
| [2010/08/25 14:06:41] <mackn> how many clients do you have? | ||
| [2010/08/25 14:06:42] @ nul0op joined channel #puppet | ||
| [2010/08/25 14:06:46] <garfias> ~ 200 | ||
| [2010/08/25 14:07:04] <mackn> you can try spacing them apart a bit more | ||
| [2010/08/25 14:07:47] @ Quit: yannL: Remote host closed the connection | ||
| [2010/08/25 14:07:56] <garfias> i just hacked the default limit of 5 in active record to 256 | ||
| [2010/08/25 14:08:20] @ Quit: nexx: Quit: quit | ||
| [2010/08/25 14:08:38] <garfias> probably not one of my best ideas, but it should get me running | ||
| [2010/08/25 14:08:44] <mackn> yeah | ||
| [2010/08/25 14:11:53] <garfias> looks like it did work | ||
| [2010/08/25 14:11:55] <garfias> woo | ||
| [2010/08/25 14:13:43] @ viewlogic joined channel #puppet | ||
| [2010/08/25 14:14:45] @ xaoslaad joined channel #puppet | ||
| [2010/08/25 14:16:21] @ bmahe joined channel #puppet | ||
| [2010/08/25 14:17:38] @ Bass10 joined channel #puppet | ||
| [2010/08/25 14:19:19] @ littleidea_ joined channel #puppet | ||
| [2010/08/25 14:21:41] @ Quit: littleidea: Read error: Operation timed out | ||
| [2010/08/25 14:23:32] @ fredden joined channel #puppet | ||
| [2010/08/25 14:28:19] @ mikey_p joined channel #puppet | ||
| [2010/08/25 14:29:00] @ \ask left channel #puppet ("Leaving...") | ||
| [2010/08/25 14:30:10] @ mrwacky42 joined channel #puppet | ||
| [2010/08/25 14:32:06] @ Quit: bug: Quit: bug | ||
| [2010/08/25 14:32:29] @ Quit: lak: Quit: lak | ||
| [2010/08/25 14:33:00] @ littleidea_ is now known as littleidea | ||
| [2010/08/25 14:33:19] @ randyb joined channel #puppet | ||
| [2010/08/25 14:34:25] @ randyb is now known as randybias | ||
| [2010/08/25 14:35:23] @ Quit: adrian_broher: Quit: Verlassend | ||
| [2010/08/25 14:37:24] @ Quit: gebi: Ping timeout: 272 seconds | ||
| [2010/08/25 14:37:53] @ Quit: bmahe: Remote host closed the connection | ||
| [2010/08/25 14:41:32] @ Quit: mrwacky42: Quit: Leaving. | ||
| [2010/08/25 14:41:38] @ bmahe joined channel #puppet | ||
| [2010/08/25 14:48:28] <husimon> anyone here use extlookup? | ||
| [2010/08/25 14:48:35] <husimon> wondering what the best way to get values into an array? | ||
| [2010/08/25 14:48:57] <husimon> $val = inline_template(<%= temp.split(',')[0] %>) apparently works if temp has 1,2,3,4 in it. | ||
| [2010/08/25 14:49:31] <husimon> but can I call a split directly on a extlookup("temp") | ||
| [2010/08/25 14:50:05] <husimon> like $val = inline_template(<%= extlookup("temp").split(',')[0] %>) | ||
| [2010/08/25 14:50:26] @ thekad is now known as thekad-afk | ||
| [2010/08/25 14:50:28] @ Quit: proton: Remote host closed the connection | ||
| [2010/08/25 14:51:23] @ Quit: cowboycoder: Quit: leaving | ||
| [2010/08/25 14:51:49] @ Quit: garfias: Ping timeout: 252 seconds | ||
| [2010/08/25 14:55:10] @ tecto_ joined channel #puppet | ||
| [2010/08/25 14:56:26] @ KipMacy joined channel #puppet | ||
| [2010/08/25 14:56:45] @ christophermlee_ joined channel #puppet | ||
| [2010/08/25 14:56:59] @ Quit: christophermlee: Ping timeout: 265 seconds | ||
| [2010/08/25 14:56:59] @ christophermlee_ is now known as christophermlee | ||
| [2010/08/25 14:58:01] @ Quit: gniks: Quit: Leaving. | ||
| [2010/08/25 14:59:37] @ Quit: gmason: Quit: gmason | ||
| [2010/08/25 14:59:45] @ Quit: Ramonster: Quit: Computer has gone to sleep | ||
| [2010/08/25 14:59:56] @ Quit: ecapriolo: Ping timeout: 255 seconds | ||
| [2010/08/25 15:00:43] @ Quit: radkin: Remote host closed the connection | ||
| [2010/08/25 15:05:41] <zambaboo> so.. funny story | ||
| [2010/08/25 15:05:54] <zambaboo> im reading the book over here and following the examples.. | ||
| [2010/08/25 15:06:10] <zambaboo> and um things dont work too well. | ||
| [2010/08/25 15:06:26] @ Quit: jmeeuwen: Ping timeout: 276 seconds | ||
| [2010/08/25 15:07:32] <zambaboo> oh i see what the deal is. | ||
| [2010/08/25 15:07:43] <zambaboo> like it says the group has to exist. | ||
| [2010/08/25 15:08:59] @ Quit: randybias: Quit: Leaving... | ||
| [2010/08/25 15:11:40] @ limed joined channel #puppet | ||
| [2010/08/25 15:12:15] @ Quit: lkoranda: Quit: Ex-Chat | ||
| [2010/08/25 15:13:04] @ thekad-afk is now known as thekad | ||
| [2010/08/25 15:15:26] @ rodnet joined channel #puppet | ||
| [2010/08/25 15:16:18] @ runrig joined channel #puppet | ||
| [2010/08/25 15:16:58] <zambaboo> how can i get the user resource to create the homedir? im looking at docs: user but its not clear there.. | ||
| [2010/08/25 15:17:14] @ Quit: rodnet: Client Quit | ||
| [2010/08/25 15:17:25] <miah> setup pam correctly? | ||
| [2010/08/25 15:18:08] @ Quit: VAlu: Ping timeout: 276 seconds | ||
| [2010/08/25 15:18:09] <zambaboo> miah: please elaborate. im working with freshly installed centos vms. | ||
| [2010/08/25 15:18:18] <miah> pam_mkhomedir | ||
| [2010/08/25 15:18:42] <husimon> question: is there some directory under a module thats standard for custom data used by facts? | ||
| [2010/08/25 15:18:44] <husimon> like flat files | ||
| [2010/08/25 15:18:52] <husimon> i mean i can create whatever dir just curious | ||
| [2010/08/25 15:19:19] <zambaboo> miah: thanks | ||
| [2010/08/25 15:20:06] <husimon> miah, i think managehome also works | ||
| [2010/08/25 15:20:24] @ Quit: kaptk2: Quit: Leaving. | ||
| [2010/08/25 15:20:27] <zambaboo> actually useradd test creates /home/test | ||
| [2010/08/25 15:20:31] <miah> manages_homedir, sure | ||
| [2010/08/25 15:20:49] <husimon> i use managehome = "true" | ||
| [2010/08/25 15:20:50] @ jmeeuwen joined channel #puppet | ||
| [2010/08/25 15:20:54] <miah> ya | ||
| [2010/08/25 15:20:55] <zambaboo> how do i set that? manages_homedir=true? | ||
| [2010/08/25 15:21:00] <zambaboo> where? in the resource def? | ||
| [2010/08/25 15:21:04] <miah> that feature needs to be supported by the provider that you're using it on | ||
| [2010/08/25 15:21:07] <miah> its in the documentation | ||
| [2010/08/25 15:21:12] <miah> under the "user" feature | ||
| [2010/08/25 15:21:28] <husimon> same place home = "your/home/dir" goes | ||
| [2010/08/25 15:21:31] <husimon> in the user resource | ||
| [2010/08/25 15:21:54] <zambaboo> like user { "test": manages_homedir=true, home=>"/home/test"} ? | ||
| [2010/08/25 15:22:06] @ Quit: rgsteele: Remote host closed the connection | ||
| [2010/08/25 15:22:18] @ hephaestus is now known as DrHorrible | ||
| [2010/08/25 15:22:35] <miah> zambaboo> its managehome, provided by the manages_home feature | ||
| [2010/08/25 15:22:46] <miah> its int he docs, at http://docs.puppetlabs.com/references/latest/type.html | ||
| [2010/08/25 15:22:56] <miah> if you look under the "user" section | ||
| [2010/08/25 15:23:29] <zambaboo> miah yes i am looking at the docs. it is not clear how to use this directive. | ||
| [2010/08/25 15:25:32] <zambaboo> ok i see. | ||
| [2010/08/25 15:25:35] <zambaboo> thanks guys | ||
| [2010/08/25 15:26:23] <zambaboo> what about automatic group creation? | ||
| [2010/08/25 15:26:50] <miah> group creation? | ||
| [2010/08/25 15:27:01] @ DrHorrible is now known as hephaestus | ||
| [2010/08/25 15:27:07] <zambaboo> when you add a user, a group of the same name is autocreated | ||
| [2010/08/25 15:27:28] <miah> ok, but do you use those groups? they're generally pointless | ||
| [2010/08/25 15:27:42] <zambaboo> i wouldnt call them pointless, but ok | ||
| [2010/08/25 15:27:42] <miah> i can't say that i've ever used such groups for anything of value | ||
| [2010/08/25 15:28:00] <miah> i tend to create groups like 'admins' or 'devs' | ||
| [2010/08/25 15:28:04] <miah> and then put people in the right group | ||
| [2010/08/25 15:28:08] <zambaboo> so you're saying this is not supported at this point | ||
| [2010/08/25 15:28:09] <miah> so that the groups are actually useful | ||
| [2010/08/25 15:28:13] @ jnewland joined channel #puppet | ||
| [2010/08/25 15:28:16] <zambaboo> at least not automatically | ||
| [2010/08/25 15:28:18] <miah> i dont know, i am not a puppet developer | ||
| [2010/08/25 15:28:34] <miah> if you want such groups, either the feature exists which would be specified by the documentation, or the code | ||
| [2010/08/25 15:29:03] @ Quit: jab_doa: Quit: Verlassend | ||
| [2010/08/25 15:29:15] <miah> i cant say i've ever used those groups, so i haven't ever missed that features existence, or used it if it does | ||
| [2010/08/25 15:29:50] <zambaboo> oh it _is_ supported. great, | ||
| [2010/08/25 15:30:07] <husimon> i use groups | ||
| [2010/08/25 15:30:08] <zambaboo> well, redhat put this stuff in for a reason. | ||
| [2010/08/25 15:30:11] <husimon> but i explicitely create them | ||
| [2010/08/25 15:30:24] <husimon> and then reference them in the user creation | ||
| [2010/08/25 15:31:03] @ Quit: runrig: Quit: Leaving | ||
| [2010/08/25 15:31:20] <husimon> i don't give users custom groups | ||
| [2010/08/25 15:31:23] <husimon> they get the "users" gropu | ||
| [2010/08/25 15:31:24] <husimon> group | ||
| [2010/08/25 15:32:21] <zambaboo> right | ||
| [2010/08/25 15:32:40] @ Quit: jdcasey: Remote host closed the connection | ||
| [2010/08/25 15:33:02] @ gniks joined channel #puppet | ||
| [2010/08/25 15:33:37] @ Quit: rmiller4pi8: Ping timeout: 272 seconds | ||
| [2010/08/25 15:33:41] @ VAlu joined channel #puppet | ||
| [2010/08/25 15:34:33] @ bodepd joined channel #puppet | ||
| [2010/08/25 15:35:04] <husimon> miah, do you know of a good reference for how to write generic modules? like how to setup variables that are set in the inherting modulet that are applied to the parent module | ||
| [2010/08/25 15:35:10] <husimon> ls | ||
| [2010/08/25 15:36:19] @ rodnet joined channel #puppet | ||
| [2010/08/25 15:40:22] @ nDuff joined channel #puppet | ||
| [2010/08/25 15:40:32] @ Quit: pinoyskull: Read error: Connection reset by peer | ||
| [2010/08/25 15:42:17] @ Quit: VAlu: Ping timeout: 260 seconds | ||
| [2010/08/25 15:44:46] @ Quit: spacex: Quit: Leaving. | ||
| [2010/08/25 15:46:54] @ Quit: miah: Quit: leaving | ||
| [2010/08/25 15:47:26] @ Quit: Reepicheep: Quit: Ex-Chat | ||
| [2010/08/25 15:47:38] @ Quit: Tas_: Ping timeout: 245 seconds | ||
| [2010/08/25 15:47:53] @ scashin133 joined channel #puppet | ||
| [2010/08/25 15:48:38] @ miah joined channel #puppet | ||
| [2010/08/25 15:49:12] @ Quit: ]oestewart: Ping timeout: 252 seconds | ||
| [2010/08/25 15:50:24] @ VAlu joined channel #puppet | ||
| [2010/08/25 15:56:11] @ pinoyskull joined channel #puppet | ||
| [2010/08/25 15:56:48] @ Quit: VAlu: Ping timeout: 240 seconds | ||
| [2010/08/25 15:58:46] @ VAlu joined channel #puppet | ||
| [2010/08/25 15:59:16] @ notbrien_ joined channel #puppet | ||
| [2010/08/25 15:59:19] @ Quit: notbrien: Remote host closed the connection | ||
| [2010/08/25 15:59:19] @ notbrien_ is now known as notbrien | ||
| [2010/08/25 15:59:47] @ ahuman joined channel #puppet | ||
| [2010/08/25 16:01:02] @ jeffmoss joined channel #puppet | ||
| [2010/08/25 16:01:10] <jeffmoss> how do I bypass the "command exceeded timeout" error? | ||
| [2010/08/25 16:01:34] <ahuman> If I return a third hash member form external nodes along with "parameters" and "classes", how do I access it in puppet templates? | ||
| [2010/08/25 16:01:52] @ Quit: shtylman: Remote host closed the connection | ||
| [2010/08/25 16:02:58] <geekatcmu> Change the timeout? | ||
| [2010/08/25 16:03:58] <geekatcmu> 'puppetd —genconfig | grep -i timeout' may be instructive. | ||
| [2010/08/25 16:04:34] @ Quit: jnewland: Read error: Connection reset by peer | ||
| [2010/08/25 16:05:16] @ Quit: shadoi: Quit: Leaving. | ||
| [2010/08/25 16:05:33] @ jnewland joined channel #puppet | ||
| [2010/08/25 16:06:06] @ Quit: jnewland: Read error: Connection reset by peer | ||
| [2010/08/25 16:06:07] @ jnewland joined channel #puppet | ||
| [2010/08/25 16:12:12] @ ivanoats joined channel #puppet | ||
| [2010/08/25 16:12:55] @ jnewland_ joined channel #puppet | ||
| [2010/08/25 16:13:28] @ Quit: jnewland: Read error: Connection reset by peer | ||
| [2010/08/25 16:13:28] @ jnewland_ is now known as jnewland | ||
| [2010/08/25 16:16:15] @ Quit: bodepd: Quit: bodepd | ||
| [2010/08/25 16:17:45] @ nexusone joined channel #puppet | ||
| [2010/08/25 16:18:28] <nexusone> is it possible to ensure some process running without converting it to be a service? that is, when it failed, puppet would auto start it .. | ||
| [2010/08/25 16:18:53] <AngryParsley> you probably want to use something like runit or daemontool | ||
| [2010/08/25 16:18:54] <AngryParsley> s | ||
| [2010/08/25 16:18:58] @ jeffmoss left channel #puppet () | ||
| [2010/08/25 16:20:59] @ Quit: ivanoats: Read error: Connection reset by peer | ||
| [2010/08/25 16:21:23] @ Quit: mikey_p: Ping timeout: 245 seconds | ||
| [2010/08/25 16:21:47] @ Quit: rodnet: Quit: rodnet | ||
| [2010/08/25 16:24:00] @ bug joined channel #puppet | ||
| [2010/08/25 16:28:02] @ Quit: littleidea: Quit: littleidea | ||
| [2010/08/25 16:29:53] <Diranged1> Ok.. turns out a 2-yearold bug is still present in puppet 2.6.. trying to use puppet to make sure the puppet service is running, doesnt work. /etc/init.d/puppet status reports true — even if puppet is being run manually in -test mode. | ||
| [2010/08/25 16:30:12] @ Quit: benkevan: Quit: leaving | ||
| [2010/08/25 16:31:39] <geekatcmu> Uhm, I'm not sure it's a bug. | ||
| [2010/08/25 16:32:33] <geekatcmu> When I run 'puppet —test" manually (0.25.5) it will start up the puppetd server if it's stopped. It's a configuration issue. | ||
| [2010/08/25 16:32:47] @ Quit: morpheus: Quit: Leaving | ||
| [2010/08/25 16:33:19] <geekatcmu> Make sure there's an actual *service* in the OS, that service provides as "status" method, and that your manifests know it. | ||
| [2010/08/25 16:33:29] <Diranged1> What triggers it to start up? puppetd —test for me runs against your config, but does not start up puppet by default. A module I have written takes care of checking whether puppet is running/enabled during that -test. | ||
| [2010/08/25 16:33:39] <kjetilho> sounds like a packaging bug, really | ||
| [2010/08/25 16:33:49] <geekatcmu> Because if you are just looking for "puppetd" to be running, well, duh, running by hand looks just like running as a service. | ||
| [2010/08/25 16:33:56] <Diranged1> Yep.. | ||
| [2010/08/25 16:34:15] <Diranged1> but if /etc/init.d/puppet status returns true, even when the *service* isnt running (but a manual test-run is), then thats a bug. IMO. | ||
| [2010/08/25 16:34:22] * Volcane 's init script does the right hting | ||
| [2010/08/25 16:34:27] <miah> thats a initscript issue | ||
| [2010/08/25 16:34:29] <geekatcmu> Sure. In that case, the init script is broken. | ||
| [2010/08/25 16:34:33] <Diranged1> :) yep | ||
| [2010/08/25 16:34:35] <miah> is that initscript packages with puppet? | ||
| [2010/08/25 16:34:41] <miah> er, packaged | ||
| [2010/08/25 16:34:41] <Diranged1> yep | ||
| [2010/08/25 16:34:47] <miah> so submit a fix | ||
| [2010/08/25 16:34:55] <Diranged1> i would… where though? :) | ||
| [2010/08/25 16:34:59] <geekatcmu> I provide my own service manifest/init script for all my systems. | ||
| [2010/08/25 16:35:01] <miah> the bug tracker? | ||
| [2010/08/25 16:35:13] <Diranged1> agh just found the sign in/reg button | ||
| [2010/08/25 16:35:14] <Diranged1> didnt see that | ||
| [2010/08/25 16:35:15] <Volcane> i run puppetd --tst in one window, in another the puppet status says not running with code 3 | ||
| [2010/08/25 16:35:52] <Diranged1> volcane i dont think it happens all the time, or maybe even on every system. i need to figure out why its happening on 10 of my systems reliably though. | ||
| [2010/08/25 16:36:08] <Diranged1> my temporary fix was just doing a custom 'status' check in the puppet service definition | ||
| [2010/08/25 16:36:29] <Volcane> are you using hasstatus => true on your service def? | ||
| [2010/08/25 16:36:42] <Diranged1> with and without that, it still failed.. until i put a custom 'status' bit in there | ||
| [2010/08/25 16:36:43] @ Quit: notbrien: Read error: Operation timed out | ||
| [2010/08/25 16:37:19] <Volcane> 'custom status bit' doesnt exactly mean anything to anyone but you tbh | ||
| [2010/08/25 16:37:54] <zambaboo> silly question, but can the package resource ensure a version? | ||
| [2010/08/25 16:38:18] @ Quit: ricky: Ping timeout: 240 seconds | ||
| [2010/08/25 16:38:23] <miah> yes | ||
| [2010/08/25 16:38:31] <miah> its clearly documented in the docs | ||
| [2010/08/25 16:38:31] <Diranged1> http://www.pastie.org/1116644 | ||
| [2010/08/25 16:38:41] <zambaboo> im looking at the doc, i see versionable, but nothing else? | ||
| [2010/08/25 16:39:08] <zambaboo> miah: looking here: http://docs.puppetlabs.com/guides/types/package.html#features | ||
| [2010/08/25 16:39:29] <miah> http://docs.puppetlabs.com/references/latest/type.html | ||
| [2010/08/25 16:39:32] <miah> im just looking there | ||
| [2010/08/25 16:40:11] <Volcane> Diranged1: and how do you start these runs that makes the init script think the servie is running? | ||
| [2010/08/25 16:40:23] <miah> look at what it says for "ensure" under package | ||
| [2010/08/25 16:40:28] <miah> read it a few times if you must | ||
| [2010/08/25 16:40:43] <Diranged1> i have a crontab entry as a backup incase puppet dies.. :0 * * * * /sbin/service puppet status > /dev/null || /usr/sbin/puppetd --server puppet -t | ||
| [2010/08/25 16:41:29] @ MarkN1 left channel #puppet () | ||
| [2010/08/25 16:45:45] @ lak joined channel #puppet | ||
| [2010/08/25 16:46:32] @ Quit: ashp: Ping timeout: 276 seconds | ||
| [2010/08/25 16:47:18] @ ashp joined channel #puppet | ||
| [2010/08/25 16:47:38] <husimon> say i have a module in my common directory which inside of it has a few variables set, say A and B | ||
| [2010/08/25 16:48:02] <husimon> now I want to inherit this class from my a module in my /etc/puppet/modules directory | ||
| [2010/08/25 16:48:08] <husimon> and assign values to A and B | ||
| [2010/08/25 16:48:12] <husimon> how is this done? | ||
| [2010/08/25 16:48:17] <husimon> or what docs should I be looking at ? | ||
| [2010/08/25 16:48:26] <zambaboo> variable scope | ||
| [2010/08/25 16:50:53] @ Quit: alexs__: Ping timeout: 272 seconds | ||
| [2010/08/25 16:50:58] @ alexs_ joined channel #puppet | ||
| [2010/08/25 16:51:43] <husimon> so if I have a class myclass1 in /etc/puppet/modules/myclass1/manifests/init.pp and a class in /etc/puppet/common/myclass2/manifests/init.pp and in myclass1 i include myclass2, i would access variables in myclass 2 via myclass2::varname? | ||
| [2010/08/25 16:52:18] <husimon> can i go myclass2::myvar = "blah" ? | ||
| [2010/08/25 16:52:25] @ proton joined channel #puppet | ||
| [2010/08/25 16:52:31] @ ezmobius joined channel #puppet | ||
| [2010/08/25 16:54:24] @ Quit: jnewland: Read error: Connection reset by peer | ||
| [2010/08/25 16:58:05] @ ivanoats joined channel #puppet | ||
| [2010/08/25 16:58:48] @ magn0lia joined channel #puppet | ||
| [2010/08/25 16:58:51] <magn0lia> !ping | ||
| [2010/08/25 16:58:51] <gepetto> magn0lia: pong | ||
| [2010/08/25 16:59:06] <magn0lia> hmm. | ||
| [2010/08/25 16:59:09] @ magn0lia left channel #puppet () | ||
| [2010/08/25 17:01:20] @ Quit: WALoeIII: Quit: Bai. | ||
| [2010/08/25 17:01:25] @ WALoeIII joined channel #puppet | ||
| [2010/08/25 17:01:30] @ Quit: WALoeIII: Remote host closed the connection | ||
| [2010/08/25 17:11:40] @ Quit: gniks: Quit: Leaving. | ||
| [2010/08/25 17:12:20] @ Quit: lak: Quit: lak | ||
| [2010/08/25 17:13:27] @ spacex joined channel #puppet | ||
| [2010/08/25 17:13:28] @ notbrien joined channel #puppet | ||
| [2010/08/25 17:14:01] @ Quit: ahasenack: Quit: Leaving | ||
| [2010/08/25 17:17:40] @ Quit: ivanoats: Remote host closed the connection | ||
| [2010/08/25 17:18:04] @ littleidea joined channel #puppet | ||
| [2010/08/25 17:19:34] @ Quit: littleidea: Client Quit | ||
| [2010/08/25 17:19:41] @ Quit: joergherzinger: Ping timeout: 276 seconds | ||
| [2010/08/25 17:28:08] @ joergherzinger joined channel #puppet | ||
| [2010/08/25 17:29:17] @ Quit: christophermlee: Quit: christophermlee | ||
| [2010/08/25 17:29:43] @ christophermlee joined channel #puppet | ||
| [2010/08/25 17:30:01] @ Quit: christophermlee: Client Quit | ||
| [2010/08/25 17:31:15] @ plathrop is now known as plathrop-away | ||
| [2010/08/25 17:35:34] @ Nilgiri joined channel #puppet | ||
| [2010/08/25 17:35:44] @ Quit: spacex: Ping timeout: 276 seconds | ||
| [2010/08/25 17:35:46] <husimon> hey i have a custom function | ||
| [2010/08/25 17:36:02] <husimon> ive put it in a modules lib/puppet/parser/functions dir | ||
| [2010/08/25 17:36:15] <husimon> and the /var/lib/puppet/lib/puppet/parser/functions dir | ||
| [2010/08/25 17:36:18] <husimon> and puppet still wont' find the function | ||
| [2010/08/25 17:36:19] <husimon> ideas? | ||
| [2010/08/25 17:36:45] <Nilgiri> I'm new to puppet and ruby and I was wonder if there puppet versions matter much? I mean, does it matter if the master and puppets are different versions at all? | ||
| [2010/08/25 17:37:29] <Nilgiri> I wonder, because I'm having trouble getting 2.6 to install in a consistant easy way | ||
| [2010/08/25 17:37:55] <husimon> god | ||
| [2010/08/25 17:38:02] <husimon> i had weird html tags at the top cause is aved it from my web browser | ||
| [2010/08/25 17:38:03] <husimon> laugh | ||
| [2010/08/25 17:38:39] <matti> ;D | ||
| [2010/08/25 17:40:52] @ Quit: Diranged1: Quit: Leaving. | ||
| [2010/08/25 17:41:14] @ Diranged joined channel #puppet | ||
| [2010/08/25 17:41:30] @ Quit: SpacePope: Read error: Connection reset by peer | ||
| [2010/08/25 17:42:22] <Nilgiri> [root@xenhost1 ~]# gem install --remote --source http://www.puppetlabs.com/downloads puppet ERROR: http://www.puppetlabs.com/downloads/ does not appear to be a repository ERROR: could not find gem puppet locally or in a repository | ||
| [2010/08/25 17:42:29] <Nilgiri> that's were I got stuck | ||
| [2010/08/25 17:43:29] <Nilgiri> am I doing something wrong with gem? | ||
| [2010/08/25 17:43:47] <matti> Nilgiri: Why don't you download tar-ball? | ||
| [2010/08/25 17:43:49] <Nilgiri> that's what the Pulling Strings with Puppet suggests for the command | ||
| [2010/08/25 17:43:58] <matti> Nilgiri: Or fetch it from git repository? | ||
| [2010/08/25 17:44:14] <matti> Nilgiri: http://docs.puppetlabs.com/guides/installation.html | ||
| [2010/08/25 17:44:38] <whack> gem installation == win | ||
| [2010/08/25 17:44:41] <Nilgiri> I got no git skills, and tar-balls typically hard to automate updates | ||
| [2010/08/25 17:44:58] <matti> Nilgiri: Look at the bottom of this page then. | ||
| [2010/08/25 17:45:05] @ brothers joined channel #puppet | ||
| [2010/08/25 17:45:09] <matti> whack: Depends. | ||
| [2010/08/25 17:45:17] <matti> whack: They sometimes tend to be a pain in the ... | ||
| [2010/08/25 17:45:17] <whack> Nilgiri: you don't need to specify --source, just 'gem install puppet' | ||
| [2010/08/25 17:45:30] @ SpacePope joined channel #puppet | ||
| [2010/08/25 17:46:01] <Nilgiri> oh? That will get me 2.6 from some preconfibured repo? | ||
| [2010/08/25 17:46:10] <whack> % sudo gem install puppet | ||
| [2010/08/25 17:46:12] <whack> Successfully installed puppet-2.6.0 | ||
| [2010/08/25 17:46:32] <whack> http://rubygems.org/gems/puppet | ||
| [2010/08/25 17:47:05] <Nilgiri> nice | ||
| [2010/08/25 17:47:09] <Nilgiri> uh oh... | ||
| [2010/08/25 17:47:35] @ Quit: pinoyskull: Quit: Leaving | ||
| [2010/08/25 17:47:41] <Nilgiri> Said "Successfully installed facter-1.5.7" but I already have a factor RPM installed... conflicts abound? | ||
| [2010/08/25 17:47:58] <whack> Nilgiri: unlikely as they probably install to different locations | ||
| [2010/08/25 17:48:01] @ pinoyskull joined channel #puppet | ||
| [2010/08/25 17:48:09] <whack> but you can check with rpm -q facter --verify | ||
| [2010/08/25 17:48:22] <whack> gems only install to the gem path (see 'gem env' output) | ||
| [2010/08/25 17:48:55] <Nilgiri> gem sure is taking a while to run... | ||
| [2010/08/25 17:49:33] <Nilgiri> it complained about "Could not find main page README" 4 times =/ | ||
| [2010/08/25 17:49:43] <Nilgiri> still chugging though | ||
| [2010/08/25 17:49:48] @ gniks joined channel #puppet | ||
| [2010/08/25 17:49:57] <ktdreyer> is it common to store sensitive information in puppet, eg. Kerberos keytabs, or /etc/shadow hashes? | ||
| [2010/08/25 17:51:27] <whack> Nilgiri: happens sometimes, I ignore most warnings from gem | ||
| [2010/08/25 17:51:34] <Nilgiri> ok cool | ||
| [2010/08/25 17:51:40] <Nilgiri> looks like it is done | ||
| [2010/08/25 17:52:12] <Nilgiri> puppet is now in the path with a bunch of others puppet commands... including puppetmasterd. Anyway to just get the client installed? | ||
| [2010/08/25 17:52:42] <whack> not with gem | ||
| [2010/08/25 17:52:53] <whack> but that doesn't really matter too much, fwiw | ||
| [2010/08/25 17:53:09] <whack> you don't have to run puppetmasterd. | ||
| [2010/08/25 17:53:16] <whack> (if it's installed, I mean) | ||
| [2010/08/25 17:53:42] <Nilgiri> yeah, I just like to run my boxes lean because it keeps the update and security tracking efforts smaller | ||
| [2010/08/25 17:54:15] @ Quit: zambaboo: Quit: zambaboo | ||
| [2010/08/25 17:54:38] @ Quit: Bass10: Quit: Leaving | ||
| [2010/08/25 17:54:42] <whack> Nilgiri: install puppet from yum/rpm/whatever | ||
| [2010/08/25 17:54:55] <whack> I think centos/fedora have separate puppet and puppetmaster bits | ||
| [2010/08/25 17:55:21] <whack> however, in 2.6.0 'puppetmasterd' is just 'puppet master' - all commands start from the 'puppet' tool | ||
| [2010/08/25 17:55:49] <Nilgiri> yeah, that would be ideal, but I worry about the version differences. Is there any reasons to be concerned if the master and puppets are not the same versions? | ||
| [2010/08/25 17:55:52] <nevyn> so is it just me that hates gem/pear/cpan etc? | ||
| [2010/08/25 17:56:10] @ PaulWay[w] joined channel #puppet | ||
| [2010/08/25 17:56:12] <whack> Nilgiri: sometimes. Generally newer masters support older clients, but not vice-versa | ||
| [2010/08/25 17:56:18] <nevyn> well not hates but thinks they're a bad idea. from a software management pov. | ||
| [2010/08/25 17:56:22] <whack> nevyn: I like gem more than most packaging systems | ||
| [2010/08/25 17:56:35] <whack> though in production I have historically built rpms from gems | ||
| [2010/08/25 17:56:39] <nevyn> I hate that they're secondary package management systems | ||
| [2010/08/25 17:56:41] <whack> to keep packaging systems consistent | ||
| [2010/08/25 17:56:48] <nevyn> and they reduce visibility and consistancy | ||
| [2010/08/25 17:56:57] <whack> you can likely convert pear/egg/cpan stuff to your one true packager, too | ||
| [2010/08/25 17:57:07] <Nilgiri> Perfect! I have 2.6 installed on my master already and I was worried it might confuse older clients, but you're saying the concern is the otherway around, whack? | ||
| [2010/08/25 17:57:22] <whack> Nilgiri: yeah. 2.6 master + 2.5 client should be fine | ||
| [2010/08/25 17:57:29] <whack> but 2.5 master + 2.6 client == sadface, usually | ||
| [2010/08/25 17:57:40] <whack> that's the general support line from the puppet folks, anyway | ||
| [2010/08/25 17:57:45] <Nilgiri> cool! THAT I can handle | ||
| [2010/08/25 17:58:17] <matti> whack += 1 | ||
| [2010/08/25 17:58:21] <Nilgiri> a hand ful of manual updates to masters is not a concern, but I was trying to automate the puppets | ||
| [2010/08/25 17:58:47] <Nilgiri> whack = whack + 1 indeed! | ||
| [2010/08/25 17:59:22] @ spacex joined channel #puppet | ||
| [2010/08/25 17:59:25] <Nilgiri> whack: thanks a million | ||
| [2010/08/25 18:00:02] <Nilgiri> now to undo all this hacking I did just to get the client installed! =) | ||
| [2010/08/25 18:00:54] <Nilgiri> whack: Also, just to be clean, 2.5 = 0.25, right? | ||
| [2010/08/25 18:01:20] <Nilgiri> the versioning is throwing me for a bit, so I want to be sure we are talking about the same thing | ||
| [2010/08/25 18:01:22] <Nilgiri> =) | ||
| [2010/08/25 18:02:40] @ Quit: pinoyskull: Quit: Leaving | ||
| [2010/08/25 18:03:35] @ Quit: rcrowley: Quit: rcrowley | ||
| [2010/08/25 18:04:43] @ Bass10 joined channel #puppet | ||
| [2010/08/25 18:06:14] @ Quit: Bass10: Max SendQ exceeded | ||
| [2010/08/25 18:06:43] @ Bass10 joined channel #puppet | ||
| [2010/08/25 18:07:41] <odyi> Nilgiri: yes 0.25 and 2.5 are synonymous. | ||
| [2010/08/25 18:08:49] @ Quit: auxesis: Ping timeout: 260 seconds | ||
| [2010/08/25 18:09:20] @ ricky joined channel #puppet | ||
| [2010/08/25 18:11:33] @ Rint joined channel #puppet | ||
| [2010/08/25 18:16:37] <Nilgiri> thanks much all! heading to be now. Good night! | ||
| [2010/08/25 18:16:48] @ Nilgiri left channel #puppet () | ||
| [2010/08/25 18:19:43] @ Quit: scashin133: Quit: scashin133 | ||
| [2010/08/25 18:20:54] @ cyrus_mc joined channel #puppet | ||
| [2010/08/25 18:21:20] <cyrus_mc> using a user resource to ensure a user is present. However if the user isn't present I am getting an error saying can't set comment, yadaydada...and it shows puppet trying to run a usermod. | ||
| [2010/08/25 18:22:05] @ Quit: tonyskapunk: Quit: ERC Version 5.3 (IRC client for Emacs) | ||
| [2010/08/25 18:22:52] <odyi> What does your resource look like? | ||
| [2010/08/25 18:24:34] @ Quit: mclarke: Quit: mclarke | ||
| [2010/08/25 18:26:44] <cyrus_mc> simple User { "nagios": uid => 100, gid => 101, comment => "Nagios monitoring user", ensure => present; } | ||
| [2010/08/25 18:27:21] @ randybias joined channel #puppet | ||
| [2010/08/25 18:28:06] <odyi> Is the capital U a typo? | ||
| [2010/08/25 18:28:12] <cyrus_mc> sorry, ya..typo | ||
| [2010/08/25 18:29:21] @ Quit: Edgan: Read error: Connection reset by peer | ||
| [2010/08/25 18:29:23] <cyrus_mc> shouldn't that do what I want | ||
| [2010/08/25 18:29:28] <n0ts> 電話で起こされました。。。 | ||
| [2010/08/25 18:29:39] <n0ts> Sorry, I missed. | ||
| [2010/08/25 18:30:22] @ Edgan joined channel #puppet | ||
| [2010/08/25 18:31:03] @ Quit: RussS: Quit: Leaving | ||
| [2010/08/25 18:32:12] <cyrus_mc> odyi: anything? | ||
| [2010/08/25 18:33:59] <odyi> Your sure `id nagios` returns nothing and you have verified that the uid number of 100 isn't being used by anything else? | ||
| [2010/08/25 18:34:32] <cyrus_mc> yep | ||
| [2010/08/25 18:34:49] <cyrus_mc> the erorr messages shows it running usermod on nagios and it returns user nagios not found | ||
| [2010/08/25 18:35:14] <odyi> Is there a user on the puppetmaster with the uid 100? | ||
| [2010/08/25 18:36:23] <cyrus_mc> no | ||
| [2010/08/25 18:37:07] <cyrus_mc> another more pressing issue...I am runnign puppet and getting ERROR 400 on Server: undefined method acl_type for nil:NilClass | ||
| [2010/08/25 18:37:44] <PaulWay[w]> I'm not sure I understand how to subclass things. Is there a helpful wiki page on it? | ||
| [2010/08/25 18:38:09] <cyrus_mc> i am tryign to do an install for work here..this all worked in my testing..so not sure why all of a sudden I am getting this error. | ||
| [2010/08/25 18:39:42] <odyi> cyrus_mc: Sorry, never ran into either of your issues. Feels like you need more debugging. | ||
| [2010/08/25 18:40:52] <odyi> (since it worked previously) | ||
| [2010/08/25 18:43:31] <PaulWay[w]> What I'm looking is something like http://fpaste.org/0yCL/ | ||
| [2010/08/25 18:44:32] @ auxesis joined channel #puppet | ||
| [2010/08/25 18:45:18] <odyi> PaulWay[w]: The require in your defined resource is un-needed. | ||
| [2010/08/25 18:46:22] <odyi> require is a meta parameter and can easily just as a require atribute when you use the resource. | ||
| [2010/08/25 18:47:07] @ Quit: n0ts: Quit: Tiarra 0.1+svn-37177: SIGTERM received; exit | ||
| [2010/08/25 18:47:27] @ n0ts joined channel #puppet | ||
| [2010/08/25 18:47:54] <PaulWay[w]> odyi: OK, I think I'm with you. | ||
| [2010/08/25 18:47:54] @ Djelibeybi joined channel #puppet | ||
| [2010/08/25 18:48:09] <PaulWay[w]> In other words, require works on anything including defines. | ||
| [2010/08/25 18:48:18] <PaulWay[w]> Neat. That saves a bit of odd-hackery. | ||
| [2010/08/25 18:48:43] <PaulWay[w]> I'm getting a syntax error at line 55 in the paste, though. | ||
| [2010/08/25 18:49:01] <PaulWay[w]> (change sshd_xforward to sshd::xforward throughout) | ||
| [2010/08/25 18:50:28] @ bodepd joined channel #puppet | ||
| [2010/08/25 18:51:09] <cyrus_mc> anyone know what the error message ERROR 400 on Server: undefined method acl_type for nil:NilClass | ||
| [2010/08/25 18:51:34] <PaulWay[w]> I suspect that your definition of the class is wrong... | ||
| [2010/08/25 18:51:39] @ Quit: n0ts: Client Quit | ||
| [2010/08/25 18:51:47] <odyi> PaulWay[w]: you are using the wrong override syntax | ||
| [2010/08/25 18:51:54] @ n0ts joined channel #puppet | ||
| [2010/08/25 18:52:13] @ Quit: Dantman: Ping timeout: 260 seconds | ||
| [2010/08/25 18:53:39] <PaulWay[w]> Ah. I don't know what the correct syntax is... | ||
| [2010/08/25 18:53:50] <PaulWay[w]> A brief look at the manual didn't turn up anything helpful there. | ||
| [2010/08/25 18:53:56] <PaulWay[w]> Now checking the syntax reference. | ||
| [2010/08/25 18:56:39] @ Quit: n0ts: Client Quit | ||
| [2010/08/25 18:56:58] @ n0ts joined channel #puppet | ||
| [2010/08/25 18:59:22] @ Quit: cyrus_mc: Quit: Lost terminal | ||
| [2010/08/25 19:02:21] @ Quit: brothers: Quit: brothers | ||
| [2010/08/25 19:02:34] <husimon> how should i debug a function not loading? | ||
| [2010/08/25 19:02:42] <husimon> i just says no such function on the client | ||
| [2010/08/25 19:02:46] <husimon> the server has no errors | ||
| [2010/08/25 19:03:21] <husimon> also it hangs the server | ||
| [2010/08/25 19:17:53] @ andypea joined channel #puppet | ||
| [2010/08/25 19:19:15] @ alexine_dsouza joined channel #puppet | ||
| [2010/08/25 19:29:23] @ Quit: paradox12: Quit: off like a prom dress | ||
| [2010/08/25 19:36:09] @ hungnv joined channel #puppet | ||
| [2010/08/25 19:39:59] <hungnv> hello everyone. I'm sorry if I did not read the document carefully. My question is: if I place one file in server location: /file/sshd/sshd_config, how can I tell server to update that file to client side with location /etc/sshd_config | ||
| [2010/08/25 19:40:06] <hungnv> thank you very much | ||
| [2010/08/25 19:41:04] @ Quit: bmahe: Quit: Leaving | ||
| [2010/08/25 19:46:40] @ Quit: gutseb_: Quit: gutseb_ | ||
| [2010/08/25 19:51:11] @ Quit: tecto_: Quit: tecto_ | ||
| [2010/08/25 19:56:15] <Djelibeybi> hungv: file { "/etc/sshd_config": source => "puppet:///modules/sshd/sshd_config", } | ||
| [2010/08/25 19:57:08] <hungnv> Djelibeybi: ahh, I see it in the document | ||
| [2010/08/25 19:58:24] <hungnv> but, in this case, puppet://modules/sshd/sshd_config specify source on the server. Where does the server look for modules path? | ||
| [2010/08/25 20:00:11] <Djelibeybi> hungnv: it looks in MODULE_PATH/module_name/files | ||
| [2010/08/25 20:00:20] <Djelibeybi> So, by default, /etc/puppet/modules/sshd/files/ | ||
| [2010/08/25 20:00:29] <Djelibeybi> Assuming you have a module named sshd | ||
| [2010/08/25 20:01:06] <hungnv> Djelibeybi: thank you very much, that's what I want to know :) | ||
| [2010/08/25 20:02:09] @ Quit: jaredrhine: Ping timeout: 272 seconds | ||
| [2010/08/25 20:02:49] @ Quit: reyjrar: Remote host closed the connection | ||
| [2010/08/25 20:03:24] @ Quit: nb: Read error: Connection reset by peer | ||
| [2010/08/25 20:06:44] @ nb joined channel #puppet | ||
| [2010/08/25 20:09:07] <husimon> is extlookup broken for 0.25.5-3 ? | ||
| [2010/08/25 20:09:16] <husimon> i tried the ruby code from the website and its all fuxed | ||
| [2010/08/25 20:10:12] @ Dantman joined channel #puppet | ||
| [2010/08/25 20:16:33] @ zambaboo joined channel #puppet | ||
| [2010/08/25 20:22:16] @ Quit: bodepd: Quit: bodepd | ||
| [2010/08/25 20:23:42] @ Quit: ryep: Quit: ryep | ||
| [2010/08/25 20:27:52] <PaulWay[w]> I've got the correct syntax now - yay! | ||
| [2010/08/25 20:29:23] <hungnv> the require syntax likes: require => Package[subversion] , if puppet check there's no $require, do we have any option to install it? | ||
| [2010/08/25 20:30:07] @ Quit: tecto__: Ping timeout: 265 seconds | ||
| [2010/08/25 20:30:29] <Djelibeybi> hungnv: can you ask that question again? I don't understand what you're asking. | ||
| [2010/08/25 20:31:01] @ e1mer joined channel #puppet | ||
| [2010/08/25 20:31:24] @ Quit: e1mer: Remote host closed the connection | ||
| [2010/08/25 20:31:25] @ gutseb joined channel #puppet | ||
| [2010/08/25 20:31:57] @ e1mer joined channel #puppet | ||
| [2010/08/25 20:31:57] @ Quit: ahuman: Write error: Broken pipe | ||
| [2010/08/25 20:32:01] @ Quit: explody: Read error: Operation timed out | ||
| [2010/08/25 20:32:12] @ ahuman joined channel #puppet | ||
| [2010/08/25 20:32:30] @ explody joined channel #puppet | ||
| [2010/08/25 20:32:38] @ Quit: colyte: Ping timeout: 245 seconds | ||
| [2010/08/25 20:33:00] <hungnv> Djelibeybi: if puppet check and found that client does not have subversion installed, can we have any option to let client automatic install that package? | ||
| [2010/08/25 20:33:17] @ danblick joined channel #puppet | ||
| [2010/08/25 20:33:22] <Djelibeybi> hungnv: yes, using the package type: package { "subversion": ensure => installed, } | ||
| [2010/08/25 20:33:40] <Djelibeybi> You would need to have your package provider configured first, i.e. yum or apt or whatever your platform uses | ||
| [2010/08/25 20:33:52] <hungnv> Djelibeybi: :-) | ||
| [2010/08/25 20:33:55] <hungnv> ok thank you | ||
| [2010/08/25 20:34:17] @ colyte joined channel #puppet | ||
| [2010/08/25 20:34:49] @ tecto_ joined channel #puppet | ||
| [2010/08/25 20:34:50] @ Quit: tecto_: Changing host | ||
| [2010/08/25 20:34:50] @ tecto_ joined channel #puppet | ||
| [2010/08/25 20:35:44] <Djelibeybi> hungnv: there is a great tutorial on getting started with Puppet here: http://www.devco.net/pubwiki/Puppet/GettingStarted | ||
| [2010/08/25 20:35:57] <Djelibeybi> Though, that's a bit old. | ||
| [2010/08/25 20:36:04] <Djelibeybi> Wait, let me find the new one. | ||
| [2010/08/25 20:36:47] <hungnv> Djelibeybi: I'm reading puppet document, just ask if I don't understand something | ||
| [2010/08/25 20:41:24] @ Quit: xaoslaad: Quit: Leaving | ||
| [2010/08/25 20:51:56] @ Quit: Djelibeybi: Quit: Leaving | ||
| [2010/08/25 20:56:17] @ lak joined channel #puppet | ||
| [2010/08/25 21:02:38] @ Quit: ricky: Ping timeout: 255 seconds | ||
| [2010/08/25 21:03:53] @ cyrus_mc joined channel #puppet | ||
| [2010/08/25 21:04:07] <cyrus_mc> I have the following variable declaration inside my class ssh - http://pastebin.com/jpmtMRfm | ||
| [2010/08/25 21:04:14] @ Quit: kogent: Ping timeout: 264 seconds | ||
| [2010/08/25 21:04:43] <cyrus_mc> also within that class I define a fuction. To reference that variable I use $ssh::basedir | ||
| [2010/08/25 21:04:48] <cyrus_mc> is that the correct way to reference it | ||
| [2010/08/25 21:06:09] <PaulWay[w]> No, you need to say basedir => $operatingsystem ... | ||
| [2010/08/25 21:06:25] <PaulWay[w]> And I think you'll find the OSes are lower case. | ||
| [2010/08/25 21:06:38] <PaulWay[w]> Test it with a notify and see. | ||
| [2010/08/25 21:06:49] <PaulWay[w]> or just do 'facter' at the command line. | ||
| [2010/08/25 21:07:08] @ kogent joined channel #puppet | ||
| [2010/08/25 21:07:21] <cyrus_mc> PaulWay[w]: so i don't put the $ in front when declaring a variable? | ||
| [2010/08/25 21:07:46] <PaulWay[w]> No, not AFAIK. | ||
| [2010/08/25 21:08:08] <cyrus_mc> ok..thanks..I will try | ||
| [2010/08/25 21:10:04] <cyrus_mc> PaulWay[w]: just read the puppet docs and it says ti use $variable = $operatingsystem => ... | ||
| [2010/08/25 21:10:13] <cyrus_mc> $variable => $operationsystem { .. } | ||
| [2010/08/25 21:10:36] <PaulWay[w]> Well, I don't have that in my Package declaration. | ||
| [2010/08/25 21:11:06] @ Quit: pting: Quit: Ex-Chat | ||
| [2010/08/25 21:11:15] <cyrus_mc> I think I see my mistake..it is $variable = $operatingsystem ? not => | ||
| [2010/08/25 21:11:22] <cyrus_mc> http://docs.reductivelabs.com/guides/language_tutorial.html | ||
| [2010/08/25 21:11:29] <PaulWay[w]> *nods* | ||
| [2010/08/25 21:14:04] @ Quit: kogent: Remote host closed the connection | ||
| [2010/08/25 21:15:44] @ Quit: ezmobius: Quit: Leaving... | ||
| [2010/08/25 21:16:20] @ ezmobius joined channel #puppet | ||
| [2010/08/25 21:16:24] @ Quit: ezmobius: Remote host closed the connection | ||
| [2010/08/25 21:20:48] @ kogent joined channel #puppet | ||
| [2010/08/25 21:23:12] @ Quit: cyrus_mc: Quit: leaving | ||
| [2010/08/25 21:25:44] @ thekad is now known as thekad-afk | ||
| [2010/08/25 21:29:02] @ Quit: Bass10: Ping timeout: 240 seconds | ||
| [2010/08/25 21:32:23] @ Bass10 joined channel #puppet | ||
| [2010/08/25 21:37:05] @ Quit: jmslagle: Ping timeout: 276 seconds | ||
| [2010/08/25 21:40:47] @ plathrop-away is now known as plathrop | ||
| [2010/08/25 21:41:36] @ FOCer joined channel #puppet | ||
| [2010/08/25 21:43:38] @ Quit: Bass10: Ping timeout: 272 seconds | ||
| [2010/08/25 21:46:38] @ Quit: e1mer: Ping timeout: 276 seconds | ||
| [2010/08/25 21:52:48] @ Quit: randybias: Quit: Leaving... | ||
| [2010/08/25 21:54:34] <zambaboo> anyone around? | ||
| [2010/08/25 22:00:09] <jamesturnbull> zambaboo: sort of | ||
| [2010/08/25 22:00:17] <zambaboo> hello sir! | ||
| [2010/08/25 22:00:19] <PaulWay[w]> no, more sort of tall and thin. | ||
| [2010/08/25 22:00:25] <zambaboo> and hello sir to you too | ||
| [2010/08/25 22:00:37] <zambaboo> looking for more info on extlookup | ||
| [2010/08/25 22:00:42] <PaulWay[w]> we say hello too! | ||
| [2010/08/25 22:00:51] * PaulWay[w] has no idea about extlookup. | ||
| [2010/08/25 22:00:52] <jamesturnbull> zambaboo: what do you need to know? | ||
| [2010/08/25 22:01:14] <zambaboo> id like to know if i can point my puppetmaster directly to a vcs repository | ||
| [2010/08/25 22:02:55] <zambaboo> for example, is it possible to tell the puppetmaster to serve config files directly from a vcs, for example, file { "/etc/hosts": path => bzr://myvcs/conf/trunk/hosts } | ||
| [2010/08/25 22:04:59] <zambaboo> or, if this cannot be done, what is the recommended way of triggering a production instance of puppet to pull configuration out of vcs once, for a example, a new tag is created | ||
| [2010/08/25 22:05:56] <zambaboo> btw james, the book rocks. kindle sucks. :) | ||
| [2010/08/25 22:06:40] <hungnv> zambaboo: I love your sentence | ||
| [2010/08/25 22:06:50] @ tonyskapunk joined channel #puppet | ||
| [2010/08/25 22:07:11] <zambaboo> thank you? | ||
| [2010/08/25 22:07:25] @ Quit: tonyskapunk: Client Quit | ||
| [2010/08/25 22:09:38] <zambaboo> here: http://groups.google.com/group/puppet-users/browse_thread/thread/68dda0ce9e363026?pli=1 Dan Carley states that they use "`manifest`/`modulepath`/extlookup " - this looks like what i need to implement | ||
| [2010/08/25 22:10:25] @ ezmobius joined channel #puppet | ||
| [2010/08/25 22:10:36] <_Sarge_> zambaboo: I run an svn up every minute on my puppetmaster | ||
| [2010/08/25 22:10:40] <_Sarge_> works nicely for me | ||
| [2010/08/25 22:10:57] <zambaboo> _Sarge_: i thought of cron as well, yes | ||
| [2010/08/25 22:11:48] <zambaboo> this uyg http://blog.nistu.de/tags/puppet/ uses a CI server, i could make hudson do the work | ||
| [2010/08/25 22:13:28] @ Quit: Diranged: Quit: Leaving. | ||
| [2010/08/25 22:13:31] @ Quit: andypea: Remote host closed the connection | ||
| [2010/08/25 22:16:31] @ Quit: wolverian: Remote host closed the connection | ||
| [2010/08/25 22:18:07] @ e1mer joined channel #puppet | ||
| [2010/08/25 22:20:25] @ Quit: e1mer: Client Quit | ||
| [2010/08/25 22:20:29] @ e1mer joined channel #puppet | ||
| [2010/08/25 22:22:23] <jamesturnbull> zambaboo: you can't pull directly from an VCS ... yet | ||
| [2010/08/25 22:22:58] <zambaboo> jamesturnbull, got it. reading up on extlookup now. | ||
| [2010/08/25 22:23:01] <jamesturnbull> zambaboo: we have a demo Git back-end I think but nothing production yet | ||
| [2010/08/25 22:24:47] @ Quit: kogent: Quit: kogent | ||
| [2010/08/25 22:26:23] <zambaboo> so it looks like i can either rely on custom facts, or on extlookup | ||
| [2010/08/25 22:26:55] @ Quit: ezmobius: Quit: Leaving... | ||
| [2010/08/25 22:27:10] <zambaboo> ok thanks all. | ||
| [2010/08/25 22:27:13] @ ezmobius joined channel #puppet | ||
| [2010/08/25 22:27:27] @ Quit: ezmobius: Remote host closed the connection | ||
| [2010/08/25 22:27:57] @ xaoslaad joined channel #puppet | ||
| [2010/08/25 22:28:38] @ kogent joined channel #puppet | ||
| [2010/08/25 22:31:29] @ jaredrhine joined channel #puppet | ||
| [2010/08/25 22:32:09] @ pinoyskull joined channel #puppet | ||
| [2010/08/25 22:33:24] @ mikey_p joined channel #puppet | ||
| [2010/08/25 22:38:05] @ Quit: gniks: Quit: Leaving. | ||
| [2010/08/25 22:39:07] @ rcrowley joined channel #puppet | ||
| [2010/08/25 22:40:06] @ Quit: mikey_p: Quit: mikey_p | ||
| [2010/08/25 22:40:32] @ mikey_p joined channel #puppet | ||
| [2010/08/25 22:41:14] @ nexx joined channel #puppet | ||
| [2010/08/25 22:43:41] @ Quit: zambaboo: Quit: zambaboo | ||
| [2010/08/25 22:46:37] @ Quit: fredden: Quit: Leaving | ||
| [2010/08/25 22:47:40] @ lkoranda joined channel #puppet | ||
| [2010/08/25 22:49:50] @ Quit: lak: Quit: lak | ||
| [2010/08/25 22:51:14] @ kenneho joined channel #puppet | ||
| [2010/08/25 22:55:22] @ Quit: xaoslaad: Quit: Leaving | ||
| [2010/08/25 23:06:35] @ Quit: spacex: Ping timeout: 276 seconds | ||
| [2010/08/25 23:18:55] @ spacex joined channel #puppet | ||
| [2010/08/25 23:22:57] @ Hallaj joined channel #puppet | ||
| [2010/08/25 23:23:00] <Hallaj> Hi all | ||
| [2010/08/25 23:23:08] @ Quit: gutseb: Quit: gutseb | ||
| [2010/08/25 23:23:09] <PaulWay[w]> Hi | ||
| [2010/08/25 23:23:21] <Hallaj> can someone tell me why the following isn't working: source => [ "puppet://server.ip.addr/path/to/$hostname.htpasswd", "puppet://server.ip.addr/path/to/.htpasswd" ] | ||
| [2010/08/25 23:26:03] @ ricky joined channel #puppet | ||
| [2010/08/25 23:26:18] @ ohookins joined channel #puppet | ||
| [2010/08/25 23:27:17] <tholl583xx> by the looks of it puppet dashboard does not install the reports processor into the lib/puppet/reports when using the rpm | ||
| [2010/08/25 23:27:23] <tholl583xx> am i missing some install step? | ||
| [2010/08/25 23:28:55] @ Quit: alban2: Ping timeout: 252 seconds | ||
| [2010/08/25 23:35:41] <FiXion> Hallaj: source (in 0.25+) should be puppet:///modules/$modulename/filename | ||
| [2010/08/25 23:36:04] <FiXion> unless you don't want to fetch the source from whatever puppetmaster the client is configured to talk to, but from somewhere else. | ||
| [2010/08/25 23:36:49] <FiXion> Hallaj: also - an actual error message would be helpful :) | ||
| [2010/08/25 23:36:49] <Hallaj> FiXion, I don't quite follow? So I can't saperate files via their host anymore? | ||
| [2010/08/25 23:37:11] <FiXion> sure | ||
| [2010/08/25 23:37:20] <FiXion> you can replace filename with $hostname.htpasswd | ||
| [2010/08/25 23:37:34] <Hallaj> FiXion: There's no error from the client. When I restart puppet, it'll grab the default .htpasswd file instead of the one meant for it. In this case, site.htpasswd | ||
| [2010/08/25 23:37:37] <FiXion> it was the part before the filename | ||
| [2010/08/25 23:37:40] <PaulWay[w]> puppet:///modules/$modulename/${filename}.${hostname} | ||
| [2010/08/25 23:37:45] <PaulWay[w]> ? | ||
| [2010/08/25 23:38:21] <FiXion> I use the hostname part some places. it works fine for me | ||
| [2010/08/25 23:39:00] <FiXion> Hallaj: you can always try and run the client (and server) with --debug if need be. | ||
| [2010/08/25 23:39:00] @ zambaboo joined channel #puppet | ||
| [2010/08/25 23:39:16] <FiXion> I have a test setup (with a master and clients) - where it's easier to debug | ||
| [2010/08/25 23:39:24] <FiXion> I'd recommend you get one two | ||
| [2010/08/25 23:39:28] <FiXion> using virtual servers it's easy :) | ||
| [2010/08/25 23:39:37] <FiXion> s/two/too/ | ||
| [2010/08/25 23:39:37] * Hallaj nods | ||
| [2010/08/25 23:39:42] <Hallaj> I should try it :) | ||
| [2010/08/25 23:39:46] * FiXion is obviously not completely awake yet | ||
| [2010/08/25 23:40:12] <FiXion> also - to ensure you don't "crap" on your production environment :) | ||
| [2010/08/25 23:40:22] <FiXion> always test before putting into production etc. | ||
| [2010/08/25 23:42:54] @ Quit: pinoyskull: Remote host closed the connection | ||
| [2010/08/25 23:44:22] <ohookins> Hallaj: I'd recommend using cucumber-puppet so you can have tests on your manifests | ||
| [2010/08/25 23:44:40] <ohookins> then any time you make changes you can verify they will do what you expect | ||
| [2010/08/25 23:47:00] <Hallaj> cucumber-puppet? | ||
| [2010/08/25 23:47:05] <Hallaj> I'll look it up :) | ||
| [2010/08/25 23:48:04] @ pinoyskull joined channel #puppet | ||
| [2010/08/25 23:51:00] @ pmorillo joined channel #puppet | ||
| [2010/08/25 23:54:53] @ yannL joined channel #puppet | ||
| [2010/08/25 23:57:05] @ PaulWay[w] left channel #puppet ("I must go, there are too many tangrams on this channel.") | ||
| [2010/08/25 23:57:39] <husimon> how can I tell a puppet package to tell the provider to fuck off and do it | ||
| [2010/08/25 23:57:41] <husimon> e.g. force |
Generated by irclog2html.py 2.8 by Marius Gedminas - find it at mg.pov.lt!