| [2009/03/30 00:08:12] @ Log started by gepetto | ||
| [2009/03/30 00:08:12] @ lak joined channel #puppet | ||
| [2009/03/30 00:25:22] @ Quit: nwp: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 00:41:48] @ Quit: justdave: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 00:44:36] @ justdave joined channel #puppet | ||
| [2009/03/30 00:46:28] <Djelibeybi> Man, I love Puppet. Fixing testing defects across 100s of servers in a few minutes = EPIC WIN. | ||
| [2009/03/30 00:51:59] <plathrop> gepetto: seen andrewcshafer | ||
| [2009/03/30 00:51:59] <gepetto> plathrop: andrewcshafer was last seen 2 hours, 12 minutes and 26 seconds ago, saying "might burrow" | ||
| [2009/03/30 00:54:19] <PaulWay> Djelibeybi: are you using puppet-run or something to push the updates out? | ||
| [2009/03/30 00:54:46] <Djelibeybi> PaulWay: no, I wasn't including the 30-min cycle to update the actual servers, just the few minutes of work it takes me to update the manifests. | ||
| [2009/03/30 00:56:00] <PaulWay> Ah, fair enough - yeah, that has the property of rocking. | ||
| [2009/03/30 01:03:10] @ Quit: Djelibeybi: "Leaving" | ||
| [2009/03/30 01:38:15] @ Quit: lak: | ||
| [2009/03/30 01:40:43] @ zombiefx_ joined channel #puppet | ||
| [2009/03/30 01:40:48] @ zombiefx_ left channel #puppet () | ||
| [2009/03/30 01:45:19] @ Quit: tuf8_: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 01:46:20] @ Quit: kolla: Remote closed the connection | ||
| [2009/03/30 01:46:24] @ tuf8 joined channel #puppet | ||
| [2009/03/30 01:49:02] @ fujin joined channel #puppet | ||
| [2009/03/30 01:53:40] @ toi joined channel #puppet | ||
| [2009/03/30 02:00:53] @ PaulWay left channel #puppet () | ||
| [2009/03/30 02:02:52] @ joe-mac1 left channel #puppet () | ||
| [2009/03/30 02:11:05] @ benblack is now known as benblack|away | ||
| [2009/03/30 02:15:27] @ hugh joined channel #puppet | ||
| [2009/03/30 02:17:20] @ Quit: hugh: Client Quit | ||
| [2009/03/30 02:17:51] @ mr_jackson joined channel #puppet | ||
| [2009/03/30 02:23:48] @ joe-mac1 joined channel #puppet | ||
| [2009/03/30 02:24:52] @ raphink joined channel #puppet | ||
| [2009/03/30 02:25:20] @ Quit: stijnbe: "Leaving..." | ||
| [2009/03/30 02:45:54] @ Quit: ivan: "Coyote finally caught me" | ||
| [2009/03/30 02:46:06] @ ivan joined channel #puppet | ||
| [2009/03/30 02:49:14] @ Quit: clayning: | ||
| [2009/03/30 02:55:01] @ Djelibeybi joined channel #puppet | ||
| [2009/03/30 02:57:22] @ Quit: mr_jackson: Remote closed the connection | ||
| [2009/03/30 03:14:29] @ johnw joined channel #puppet | ||
| [2009/03/30 03:15:16] @ madrescher joined channel #puppet | ||
| [2009/03/30 03:15:48] @ kibahop joined channel #puppet | ||
| [2009/03/30 03:17:22] @ Quit: Djelibeybi: "Leaving" | ||
| [2009/03/30 03:18:41] @ aymerick joined channel #puppet | ||
| [2009/03/30 03:25:13] @ Quit: joe-mac1: Read error: 110 (Connection timed out) | ||
| [2009/03/30 03:25:59] @ DerekW joined channel #puppet | ||
| [2009/03/30 03:26:07] @ thegcat joined channel #puppet | ||
| [2009/03/30 03:28:03] <eythian> In case anyone is using puppet to configure EC2 servers, I threw this together: https://code.launchpad.net/~eythian/+junk/ec2facts | ||
| [2009/03/30 03:37:37] @ tuf8_ joined channel #puppet | ||
| [2009/03/30 03:45:21] @ mvn071 joined channel #puppet | ||
| [2009/03/30 03:45:32] @ Quit: mvn071: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 03:45:33] @ madrescher1 joined channel #puppet | ||
| [2009/03/30 03:46:17] @ Quit: madrescher: Read error: 110 (Connection timed out) | ||
| [2009/03/30 03:47:12] <johnw> can I lengthen the timeout for an exec rule? | ||
| [2009/03/30 03:47:24] <johnw> i'm trying to run ntpdate, and it's not waiting long enough | ||
| [2009/03/30 03:51:31] @ MattyM joined channel #puppet | ||
| [2009/03/30 03:56:21] @ Quit: tuf8: Read error: 110 (Connection timed out) | ||
| [2009/03/30 03:56:26] @ Quit: madrescher1: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 03:59:53] @ Welsh_Dwarf joined channel #puppet | ||
| [2009/03/30 04:02:02] @ fbe__ joined channel #puppet | ||
| [2009/03/30 04:09:35] @ glaw joined channel #puppet | ||
| [2009/03/30 04:11:19] @ madrescher joined channel #puppet | ||
| [2009/03/30 04:18:24] @ gaveen joined channel #puppet | ||
| [2009/03/30 04:21:03] @ friendly12345 joined channel #puppet | ||
| [2009/03/30 04:25:38] @ stijnbe joined channel #puppet | ||
| [2009/03/30 04:25:48] <SyTonnerre> netnoc.eu?! | ||
| [2009/03/30 04:26:12] <stijnbe> SyTonnerre: Netlog network :) | ||
| [2009/03/30 04:26:46] <SyTonnerre> I only discovered | ||
| [2009/03/30 04:26:48] <SyTonnerre> How odd | ||
| [2009/03/30 04:27:11] <SyTonnerre> I thought it was something like The Internet Company | ||
| [2009/03/30 04:27:23] <SyTonnerre> stijnbe, you don't happen to be connected to the SwissIX? ;) | ||
| [2009/03/30 04:28:11] <stijnbe> I don't know :p | ||
| [2009/03/30 04:28:25] <SyTonnerre> So you're in the noc but don't know anything about the network? | ||
| [2009/03/30 04:28:44] <SyTonnerre> BTW, we're using puppet to manage the bgpd configuration of our routers here :) | ||
| [2009/03/30 04:29:17] @ yarihm joined channel #puppet | ||
| [2009/03/30 04:29:27] <SyTonnerre> And a finecom customer! Hello :) | ||
| [2009/03/30 04:29:47] <f3ew> heh | ||
| [2009/03/30 04:30:00] <SyTonnerre> f3ew, heh? | ||
| [2009/03/30 04:30:21] <f3ew> heh! | ||
| [2009/03/30 04:30:46] <f3ew> johnw timeout | ||
| [2009/03/30 04:31:21] @ Quit: f3ew: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 04:37:38] @ briandquinn joined channel #puppet | ||
| [2009/03/30 04:37:45] <jamesturnbull> fucking RST | ||
| [2009/03/30 04:38:23] * kjetilho <3 reST | ||
| [2009/03/30 04:40:02] @ fdz joined channel #puppet | ||
| [2009/03/30 04:42:39] <Volcane> RST? | ||
| [2009/03/30 04:44:17] <gepetto> ::trac:: test edited by jamtur01 @ http://reductivelabs.com/trac/puppet/wiki/test (by james@lovedthanlost.net) | ||
| [2009/03/30 04:44:34] <Volcane> oh the trac wiki thing RST | ||
| [2009/03/30 04:44:35] <Volcane> right | ||
| [2009/03/30 04:49:36] @ f3ew joined channel #puppet | ||
| [2009/03/30 04:53:25] @ alban2 joined channel #puppet | ||
| [2009/03/30 04:55:01] @ Quit: andrewcshafer: Read error: 110 (Connection timed out) | ||
| [2009/03/30 04:56:11] @ Quit: thegcat: Connection timed out | ||
| [2009/03/30 04:56:23] @ int joined channel #puppet | ||
| [2009/03/30 04:57:05] @ thegcat joined channel #puppet | ||
| [2009/03/30 04:57:17] @ Quit: fdz: "Leaving." | ||
| [2009/03/30 04:57:24] @ fdz joined channel #puppet | ||
| [2009/03/30 04:58:35] <tim|imac> a sysadmin question, unrelated to puppet: is there some way to have apache tell the loadbalancer in front of it "hey, I'm going to shut down/be unavailable, remove me from the pool, please" without custom scripting? is there a standard for such things? | ||
| [2009/03/30 04:59:00] <tim|imac> the loadbalancer is currently lvs (ipvsadm/ldirectord), but this can change, of course | ||
| [2009/03/30 04:59:10] @ Quit: fdz: Read error: 54 (Connection reset by peer) | ||
| [2009/03/30 04:59:26] @ Quit: yarihm: "This computer has gone to sleep" | ||
| [2009/03/30 05:00:15] <jenza> I don't think I'd have puppet controlling load balancers | ||
| [2009/03/30 05:00:34] <jenza> The way we do it is that our application is in an RPM which has the functionality to do that kind of thing via API calls | ||
| [2009/03/30 05:00:43] <jenza> with puppet you don't know when exactly it will going | ||
| [2009/03/30 05:00:46] <jenza> go on | ||
| [2009/03/30 05:01:35] <jenza> And if you're in an environment where you can't handle any downtime, ie the time between when you take the web box down and the load balancer yanks it from the pool you probably want more control over the process. | ||
| [2009/03/30 05:02:40] @ ssm_ joined channel #puppet | ||
| [2009/03/30 05:02:44] <SyTonnerre> Then again there are HTTP load balancers like Pound which take care of their active host list themselves | ||
| [2009/03/30 05:03:10] <SyTonnerre> You only feed them a list of all nodes and if one of them goes down, Pound removes it from the list internally until it comes back | ||
| [2009/03/30 05:03:22] <SyTonnerre> I think there's a 5 second convergence time though | ||
| [2009/03/30 05:03:23] <jenza> All load balancers do that | ||
| [2009/03/30 05:03:33] <SyTonnerre> jenza, no, unfortunately not ;) | ||
| [2009/03/30 05:03:43] <jenza> Well, all proper load balancers do | ||
| [2009/03/30 05:03:47] <SyTonnerre> True. | ||
| [2009/03/30 05:04:38] <jenza> For my environment I'd rather build it into the RPM deployment | ||
| [2009/03/30 05:05:12] <jenza> really depends on what your load balancer is, what your app is, what you want to achieve and what kind of outage is ok. | ||
| [2009/03/30 05:06:27] <SyTonnerre> And how well the load balancer handles an outage ;) | ||
| [2009/03/30 05:08:43] @ MarlondB joined channel #puppet | ||
| [2009/03/30 05:11:37] <jenza> Well, how well they work, sure. | ||
| [2009/03/30 05:11:54] <jenza> Handling an outage is Load balancer 101 | ||
| [2009/03/30 05:13:50] <gepetto> ::trac:: Puppet Manifest Documentation edited by lazzurs @ http://reductivelabs.com/trac/puppet/wiki/PuppetManifestDocumentation (by rob@lazzurs.net) | ||
| [2009/03/30 05:17:50] @ kolla joined channel #puppet | ||
| [2009/03/30 05:26:45] @ Quit: ohadlevy: Remote closed the connection | ||
| [2009/03/30 05:26:46] @ verwilst joined channel #puppet | ||
| [2009/03/30 05:42:42] @ Quit: f3ew: "You can /quit any time you like, but you can never leave" | ||
| [2009/03/30 05:43:39] @ Quit: johnw: | ||
| [2009/03/30 05:43:56] @ f3ew joined channel #puppet | ||
| [2009/03/30 05:51:04] <eythian> when I create a custom fact provider, and put it in modulename/facter - what must the .rb file be called? | ||
| [2009/03/30 05:51:28] <eythian> modulename/plugins/favter I mean | ||
| [2009/03/30 05:54:56] @ nakano_ is now known as nakano | ||
| [2009/03/30 05:55:04] <jamesturnbull> eythian: usually the name of the fact.rb | ||
| [2009/03/30 05:56:50] <eythian> jamesturnbull: is that a requirement? because mine doesn't know beforehand. | ||
| [2009/03/30 05:57:03] <jamesturnbull> eythian: nope - best practise | ||
| [2009/03/30 05:58:29] <eythian> jamesturnbull: OK cool. And 'facter -p' should show them up, or will they only ever show up when puppet is actually running it? | ||
| [2009/03/30 05:59:23] <jamesturnbull> eythian: should do but some people have reported issues | ||
| [2009/03/30 05:59:38] <jamesturnbull> eythian: with older versions | ||
| [2009/03/30 05:59:57] <eythian> jamesturnbull: hmm. cheers. I think I'll look into it more tomorrow. | ||
| [2009/03/30 06:00:13] <jamesturnbull> eythian: testing is your friend :P | ||
| [2009/03/30 06:02:11] <eythian> jamesturnbull: well they aren't showing up for me, hence the questions :) | ||
| [2009/03/30 06:02:21] <jamesturnbull> eythian: version? | ||
| [2009/03/30 06:02:33] <eythian> 0.24.8 | ||
| [2009/03/30 06:02:50] <jamesturnbull> eythian: crap | ||
| [2009/03/30 06:03:01] <jamesturnbull> eythian: as in "annoying" - not "you're lying" | ||
| [2009/03/30 06:03:02] <eythian> jamesturnbull: no really :) | ||
| [2009/03/30 06:03:43] <eythian> jamesturnbull: yeah, but I've been stabbing at servers for 11 hours straight today, I'm reaching the point of making silly mistakes, so I'll wait 'til I'm more awake tomorrow. | ||
| [2009/03/30 06:04:05] <jamesturnbull> eythian: you've got the fatc in modulename/plugins/facter? | ||
| [2009/03/30 06:04:27] <jamesturnbull> eythian: and factpath defined? | ||
| [2009/03/30 06:04:28] <eythian> (for example "why won't puppet move the file?" and discovering that I had ensure=>absent) | ||
| [2009/03/30 06:04:36] <eythian> jamesturnbull: yep, I saw it get copied to the client | ||
| [2009/03/30 06:05:09] <eythian> oh wait | ||
| [2009/03/30 06:05:12] <jamesturnbull> eythian: hmmm well try it tomorrow - if still fails could you log a big please? | ||
| [2009/03/30 06:05:18] <eythian> I think maybe I was doing something wrong. | ||
| [2009/03/30 06:05:18] <jamesturnbull> s/big/bug/ | ||
| [2009/03/30 06:05:28] <eythian> I tried it just then and it worked. | ||
| [2009/03/30 06:05:36] <jamesturnbull> woot! | ||
| [2009/03/30 06:05:38] <eythian> I don't know what changed in between now and the last time. | ||
| [2009/03/30 06:05:39] <jamesturnbull> I am a genius | ||
| [2009/03/30 06:05:42] <jamesturnbull> :P | ||
| [2009/03/30 06:05:42] <eythian> :) | ||
| [2009/03/30 06:06:01] <eythian> so I'm going to leave the computer thing alone for a wee while now. | ||
| [2009/03/30 06:06:06] <jamesturnbull> eythian: good move | ||
| [2009/03/30 06:06:14] <eythian> jamesturnbull: cheers, cya | ||
| [2009/03/30 06:06:27] <jamesturnbull> later | ||
| [2009/03/30 06:08:39] @ kibaho1 joined channel #puppet | ||
| [2009/03/30 06:09:39] @ kibaho1 left channel #puppet () | ||
| [2009/03/30 06:17:03] @ giles_ joined channel #puppet | ||
| [2009/03/30 06:17:48] <giles_> heya puppetiers | ||
| [2009/03/30 06:18:05] <verwilst> i would like to create a variable that would do like [{a => bleh, b => bleh}, ... ] so i can loop through them | ||
| [2009/03/30 06:18:06] <giles_> i've still got a weird problem on some hosts that I'm trying to upgrade to 0.24 | ||
| [2009/03/30 06:18:17] <verwilst> is that possible? | ||
| [2009/03/30 06:18:17] <giles_> notice: Did not receive certificate | ||
| [2009/03/30 06:18:19] <giles_> notice: Set to run 'one time'; exiting with no certificate | ||
| [2009/03/30 06:18:30] <giles_> I've cleaned the cert from the puppetmaster | ||
| [2009/03/30 06:18:40] <giles_> cleared out the client too and I still get it | ||
| [2009/03/30 06:18:49] <giles_> ran the client with debug | ||
| [2009/03/30 06:18:54] <giles_> stumped.. | ||
| [2009/03/30 06:18:59] <giles_> it's only a couple of hosts | ||
| [2009/03/30 06:19:29] <giles_> ah crap just worked it out | ||
| [2009/03/30 06:19:45] <giles_> they have differerent fqdn's so aren't autosigning | ||
| [2009/03/30 06:19:46] <giles_> doh | ||
| [2009/03/30 06:19:52] <giles_> l8trz | ||
| [2009/03/30 06:19:55] @ Quit: giles_: Client Quit | ||
| [2009/03/30 06:20:28] <tim|imac> SyTonnerre: yeah, but i was hoping to keep the convergence time as low as possible... it's currently 3 seconds max in our current setup (sub 1 second if you're lucky) | ||
| [2009/03/30 06:21:23] @ fdz joined channel #puppet | ||
| [2009/03/30 06:23:22] <tim|imac> and 3 seconds on a real node failure is acceptable, however, if you do a manual restart or something (which happens quite often, or at least more often than crashes), there should be a method in which you can reduce the convergence to 0 seconds | ||
| [2009/03/30 06:23:38] * tim|imac was hoping for some sort of established protocol that he didn't know about yet | ||
| [2009/03/30 06:30:47] <verwilst> foo = {'a' => 'dddd', 'b' => 'ccccc'} doesn't work | ||
| [2009/03/30 06:30:56] <verwilst> in the external_nodes script | ||
| [2009/03/30 06:30:58] <verwilst> any ideas? | ||
| [2009/03/30 06:31:14] <verwilst> "foo: " that is | ||
| [2009/03/30 06:31:17] <huggie> It would be neat if puppet had a "do this first" which could be used to take a machine out of a load balancer and only allowed the rest of the config to run if it was successful. I suppose it might be possible in config. | ||
| [2009/03/30 06:32:55] @ webx joined channel #puppet | ||
| [2009/03/30 06:33:17] @ Quit: webx: Client Quit | ||
| [2009/03/30 06:33:42] <Volcane> yeah, would be nice :( could be as simple as a kind of meta resources that you could just use in before or require | ||
| [2009/03/30 06:33:57] <Volcane> one for first and one for last, donno how feasable that would be to code though :) | ||
| [2009/03/30 06:34:19] <Volcane> anyway, you wouldnt want puppet to remove from the LB on each run would you? | ||
| [2009/03/30 06:34:29] @ webx joined channel #puppet | ||
| [2009/03/30 06:35:07] <webx> I'm running into an interesting problem with iptables | ||
| [2009/03/30 06:35:17] <tim|imac> actually, i wasn't really thinking about a puppet solution in this ;-) just a general inquiry, since there are a lot of sysadmins here ;-) | ||
| [2009/03/30 06:35:46] <tim|imac> verwilst: I'd do that in an "if" block | ||
| [2009/03/30 06:35:49] <tim|imac> or case | ||
| [2009/03/30 06:35:57] <tim|imac> yeah, case is better | ||
| [2009/03/30 06:35:57] <webx> I have a series of rules that I send out to a set of nodes, and if the file changes I ask puppet to notify iptables.. yadda yadda yadda | ||
| [2009/03/30 06:36:14] <webx> I use service iptables... ensure running.. has restart, etc etc | ||
| [2009/03/30 06:36:26] @ m0zzzy joined channel #puppet | ||
| [2009/03/30 06:36:28] <m0zzzy> hello | ||
| [2009/03/30 06:36:35] <webx> but I think puppet is looking for something that service iptables status doesn't return | ||
| [2009/03/30 06:36:50] <webx> because every time puppet runs (30 minutes) iptables is restarted | ||
| [2009/03/30 06:37:06] <webx> which is a problem for production, obviously | ||
| [2009/03/30 06:37:11] <Volcane> webx: which distro/ | ||
| [2009/03/30 06:37:12] <tim|imac> webx: iptables isn't a running process and unless your init script doesn't return an LSB compliant value for "status", puppet thinks the service is down | ||
| [2009/03/30 06:37:33] <m0zzzy> got a question about virtual resource. I am realizing users from virtual class. Can I create group automatically using only user type? | ||
| [2009/03/30 06:37:39] <tim|imac> unless your init script returns an LSB blabla, i meant to say | ||
| [2009/03/30 06:38:04] <tim|imac> m0zzzy: you can use a user defined resource for that | ||
| [2009/03/30 06:38:05] <m0zzzy> usually, if running useradd in linux, group with the same name is created, however puppet requires it before | ||
| [2009/03/30 06:38:15] <webx> ok | ||
| [2009/03/30 06:38:18] <webx> makes sense | ||
| [2009/03/30 06:38:35] <webx> so, if I want iptables to restart when that file changes... | ||
| [2009/03/30 06:38:43] <m0zzzy> tim|imac: how can I do that? It is assumed group is created before, isn't it? | ||
| [2009/03/30 06:38:43] <webx> how can I do the equivalent ? | ||
| [2009/03/30 06:39:10] <tim|imac> m0zzzy: uh yeah, you're right... um... i'd create the group seperately, I guess... that way you can require on it | ||
| [2009/03/30 06:39:10] <Volcane> webx: what distro? | ||
| [2009/03/30 06:39:16] <webx> Volcane: centos | ||
| [2009/03/30 06:40:07] <tim|imac> webx: exec { "restart iptables": cmd => "/etc/init.d/puppet restart", subscribe => "/path/to/rules/file", } is a good start... not sure if it's "cmd" or "command", though... probably the latter | ||
| [2009/03/30 06:40:18] <tim|imac> argh | ||
| [2009/03/30 06:40:21] <tim|imac> and not puppet restart | ||
| [2009/03/30 06:40:26] <tim|imac> but iptables restart :) | ||
| [2009/03/30 06:40:29] <webx> :P | ||
| [2009/03/30 06:40:38] <tim|imac> pardon my caffeine-lacking brain | ||
| [2009/03/30 06:40:44] * Volcane 's iptables script on centos return sane values | ||
| [2009/03/30 06:41:03] <webx> right, it returns whatever iptables has loaded | ||
| [2009/03/30 06:41:05] <webx> right? | ||
| [2009/03/30 06:41:07] <tim|imac> webx: listen to Volcane, I'm a debian user, he knows centos better than me :) | ||
| [2009/03/30 06:41:40] <Volcane> webx: the return code is whats important, not the STDOUT spew | ||
| [2009/03/30 06:41:53] @ Quit: tuf8_: Read error: 110 (Connection timed out) | ||
| [2009/03/30 06:42:06] <Volcane> webx: so what does 'service iptables status;echo $?' say at the end? | ||
| [2009/03/30 06:42:41] <webx> 0, which it should obviously | ||
| [2009/03/30 06:42:54] <webx> http://pastie.org/431249 | ||
| [2009/03/30 06:43:03] <webx> that's what I'm doing.. | ||
| [2009/03/30 06:43:11] <webx> but the service iptables ensure => seems to be the issue | ||
| [2009/03/30 06:43:15] @ Quit: ssm_: "WeeChat 0.2.6" | ||
| [2009/03/30 06:43:25] <Volcane> show us a output from your run | ||
| [2009/03/30 06:43:31] <webx> iptables rules? | ||
| [2009/03/30 06:43:36] <webx> or? | ||
| [2009/03/30 06:43:52] <webx> it's all disabled now because I can't have it restart iptables | ||
| [2009/03/30 06:44:02] <webx> there are thousands of users playing poker that go byebye when I do that | ||
| [2009/03/30 06:44:13] <tim|imac> webx: try adding hasstatus => true to your service | ||
| [2009/03/30 06:44:19] <Volcane> what puppet shows when you do puppetd --test | ||
| [2009/03/30 06:44:27] <tim|imac> Volcane: at least, i assume the init script does the same as the service command, right? | ||
| [2009/03/30 06:44:38] <Volcane> yes, hasstatus is the key | ||
| [2009/03/30 06:44:50] <Volcane> foo, donno how i didnt see that, was blinded by the hasrestart :P | ||
| [2009/03/30 06:45:02] <Volcane> webx: just add hasstatus => true to the service | ||
| [2009/03/30 06:45:05] <tim|imac> webx: if you do not add hasstatus, it checks the ps list for a process with $name (iptables, in this case) | ||
| [2009/03/30 06:45:23] <webx> ah | ||
| [2009/03/30 06:45:31] <webx> just add "hasstatus" and don't remove anything? | ||
| [2009/03/30 06:46:04] <Volcane> just add | ||
| [2009/03/30 06:46:07] <tim|imac> webx: http://pastie.org/431252 | ||
| [2009/03/30 06:46:39] <webx> awesome | ||
| [2009/03/30 06:46:41] <webx> thanks :) | ||
| [2009/03/30 06:46:49] <webx> I was going to wait until the puppet class tomorrow to ask this question | ||
| [2009/03/30 06:46:52] <webx> but I'm impatient | ||
| [2009/03/30 06:50:48] @ Quit: friendly12345: "Leaving." | ||
| [2009/03/30 06:52:25] @ Quit: madrescher: Read error: 113 (No route to host) | ||
| [2009/03/30 06:55:39] @ Quit: ssm: "leaving" | ||
| [2009/03/30 07:03:15] @ ssm joined channel #puppet | ||
| [2009/03/30 07:06:20] @ Quit: glaw: Remote closed the connection | ||
| [2009/03/30 07:18:27] @ FiXion joined channel #puppet | ||
| [2009/03/30 07:19:53] @ Quit: FiXion: "leaving" | ||
| [2009/03/30 07:19:53] @ Quit: FiXion_: "leaving" | ||
| [2009/03/30 07:21:01] @ FiXion joined channel #puppet | ||
| [2009/03/30 07:22:12] <FiXion> anyone using puppetshow? | ||
| [2009/03/30 07:22:19] <FiXion> I'm trying to install it.. | ||
| [2009/03/30 07:22:37] <FiXion> rake production db:migrate | ||
| [2009/03/30 07:22:41] <FiXion> fails miserably | ||
| [2009/03/30 07:22:51] <FiXion> with "uninitialized constant ActiveSupport::Dependencies" | ||
| [2009/03/30 07:22:58] <FiXion> on ubuntu 8.04 with rails 2.02 | ||
| [2009/03/30 07:23:13] <FiXion> and I am not proficient in rubyspeak :/ | ||
| [2009/03/30 07:23:41] <FiXion> I've worked a lot with cfengine and used alist at that time. | ||
| [2009/03/30 07:24:04] <FiXion> I'm really looking forward to having the client info and puppet integracted via the facts system | ||
| [2009/03/30 07:24:30] <FiXion> and puppetshow looks like a nice way to get an overview of different things, such as client mac's etc. | ||
| [2009/03/30 07:26:19] <FiXion> hmm - seems I like a newer rails version.. according to the few results I could google | ||
| [2009/03/30 07:26:30] * Volcane doesnt think puppetshow is working at all atm | ||
| [2009/03/30 07:27:01] <jenza> you might want to look at puppetview | ||
| [2009/03/30 07:29:03] <FiXion> what's the pro/con's vs. puppetshow? | ||
| [2009/03/30 07:29:10] <FiXion> couldn't find much online docu. | ||
| [2009/03/30 07:29:16] <FiXion> puppetview only showed some graphs :) | ||
| [2009/03/30 07:30:09] <FiXion> but if puppetview also gives you webaccess to the client facts then it could be just as fine I guess. | ||
| [2009/03/30 07:30:43] <Volcane> it shows the facts | ||
| [2009/03/30 07:31:28] <Volcane> and logs | ||
| [2009/03/30 07:31:42] <FiXion> doesn't puppetview do the same? | ||
| [2009/03/30 07:31:45] <FiXion> puppetshow | ||
| [2009/03/30 07:32:02] <FiXion> but if puppetshow is broken.. :( | ||
| [2009/03/30 07:32:06] <huggie> Volcane: Well the way we run it currently, yes I'd want it out of the LB every run. We run with --no-client and use puppetrun to trigger with another script that controls how many are allowed out of the LB at any one time and smoke tests them before they go back in etc. Kinda works for now but feels a bit wrong ultimately. | ||
| [2009/03/30 07:32:12] <huggie> </lag> | ||
| [2009/03/30 07:32:14] <FiXion> puppetshow author also seems to not believe in releases :) | ||
| [2009/03/30 07:32:24] <jenza> huggie: We do the same here with no client | ||
| [2009/03/30 07:32:26] <Volcane> huggie: ah nice | ||
| [2009/03/30 07:32:33] <jenza> except we manually do the load balacner stuff | ||
| [2009/03/30 07:32:56] <verwilst> tim|imac: case? | ||
| [2009/03/30 07:33:01] <verwilst> ( sorry for late answer :) ) | ||
| [2009/03/30 07:34:12] <tim|imac> verwilst: hm... I kinda assumed you meant, "if $foo = 'a', set it to 'dddd'", but rereading it, I'm not sure what you're trying to achieve, actually... an array? | ||
| [2009/03/30 07:34:23] <verwilst> yeah | ||
| [2009/03/30 07:34:35] <verwilst> an array of hashes | ||
| [2009/03/30 07:34:51] <verwilst> then loop through the array | ||
| [2009/03/30 07:34:59] <verwilst> and use the hash keys/values in each | ||
| [2009/03/30 07:35:36] <tim|imac> I don't think puppet knows about dictionaries, only plain array (ie. no key, or rather, position is the key) | ||
| [2009/03/30 07:35:52] <verwilst> hm.. | ||
| [2009/03/30 07:36:02] <verwilst> i would like to fill in my /etc/network/interfaces | ||
| [2009/03/30 07:36:09] <verwilst> but the devices are not pre-defined | ||
| [2009/03/30 07:36:55] <verwilst> i've done sth like this before... but forgot how :( | ||
| [2009/03/30 07:37:29] <tim|imac> I don't really have a solution for that, I'm afraid | ||
| [2009/03/30 07:37:38] <tim|imac> at least, not something that pops into mind | ||
| [2009/03/30 07:37:53] <verwilst> sucky, something like this must be possible :) | ||
| [2009/03/30 07:37:55] <tim|imac> maybe if i think about it a little harder (but chances are, someone else will have an answer for you by then) | ||
| [2009/03/30 07:42:58] @ Quit: fdz: Read error: 113 (No route to host) | ||
| [2009/03/30 07:43:28] @ fdz joined channel #puppet | ||
| [2009/03/30 07:43:41] @ Quit: fdz: Client Quit | ||
| [2009/03/30 07:46:05] @ glaw joined channel #puppet | ||
| [2009/03/30 07:46:56] @ martha joined channel #puppet | ||
| [2009/03/30 07:54:41] @ yarihm joined channel #puppet | ||
| [2009/03/30 08:01:57] @ keithlard joined channel #puppet | ||
| [2009/03/30 08:04:19] @ Innocenti joined channel #puppet | ||
| [2009/03/30 08:08:35] @ yure_ joined channel #puppet | ||
| [2009/03/30 08:08:44] @ madrescher joined channel #puppet | ||
| [2009/03/30 08:10:03] @ thegcat_ joined channel #puppet | ||
| [2009/03/30 08:12:54] @ Quit: fujin: Read error: 110 (Connection timed out) | ||
| [2009/03/30 08:17:19] <FiXion> I've enabled reports = true in puppet.conf - but I see no reports. | ||
| [2009/03/30 08:17:24] <FiXion> I've run puppetd with puppetd --test --debug --trace | ||
| [2009/03/30 08:17:31] <FiXion> and there's no mention of reports in the output :( | ||
| [2009/03/30 08:17:47] <FiXion> I can't find any other hints on the wiki to help me debug | ||
| [2009/03/30 08:18:34] @ ^authentic joined channel #puppet | ||
| [2009/03/30 08:19:53] <jenza> reports go into /etc/puppet/reports | ||
| [2009/03/30 08:20:03] <jenza> assuming that's your puppethome | ||
| [2009/03/30 08:20:13] <FiXion> but shouldn't the client log that it was doing reports or something | ||
| [2009/03/30 08:20:23] <jenza> it'll say sending report to server | ||
| [2009/03/30 08:20:25] <FiXion> I've setup reports on the server - according the the puppetview config | ||
| [2009/03/30 08:20:30] <FiXion> jenza: but it doesn't say that :( | ||
| [2009/03/30 08:20:45] <jenza> you also need reports = store | ||
| [2009/03/30 08:20:58] <jenza> reports = store,rrdgraph if you want the rrdgraphs stored too | ||
| [2009/03/30 08:21:14] <jenza> also under [puppetd] you want report = true | ||
| [2009/03/30 08:21:20] <jenza> I think | ||
| [2009/03/30 08:21:30] <jenza> yeah | ||
| [2009/03/30 08:21:42] <jenza> last one is for the client | ||
| [2009/03/30 08:24:05] @ rsquared joined channel #puppet | ||
| [2009/03/30 08:24:17] <FiXion> I have report = true on client | ||
| [2009/03/30 08:24:45] <FiXion> and reports = store,rrdgraph on master | ||
| [2009/03/30 08:24:54] <FiXion> and reportdir etc. | ||
| [2009/03/30 08:25:17] <FiXion> reports = true | ||
| [2009/03/30 08:25:23] <jenza> yup | ||
| [2009/03/30 08:25:29] <FiXion> is it report or reports ? | ||
| [2009/03/30 08:25:31] <FiXion> on client | ||
| [2009/03/30 08:25:35] <jenza> report | ||
| [2009/03/30 08:25:41] <FiXion> I'll change it.. | ||
| [2009/03/30 08:26:03] <jenza> also add catalog_format = marshal | ||
| [2009/03/30 08:26:24] <jenza> nuder [puppetd] for the client | ||
| [2009/03/30 08:26:32] <jenza> also under puppetmasterd | ||
| [2009/03/30 08:27:51] <FiXion> jenza: thank you - that fixed it :) | ||
| [2009/03/30 08:27:58] <FiXion> marshal | ||
| [2009/03/30 08:28:04] <FiXion> what other formats are there available? | ||
| [2009/03/30 08:28:13] <FiXion> and why m arshal ? :) | ||
| [2009/03/30 08:28:15] * FiXion is curious | ||
| [2009/03/30 08:28:33] @ Quit: maxagaz: "Leaving" | ||
| [2009/03/30 08:29:08] <jenza> I believe it's a more efficient reporting method | ||
| [2009/03/30 08:29:21] <jenza> or it speeds up catalog processing | ||
| [2009/03/30 08:29:26] <jenza> it's been a while since I put it in :) | ||
| [2009/03/30 08:29:41] <FiXion> fair enough - thanks :) | ||
| [2009/03/30 08:30:28] <jenza> no worries | ||
| [2009/03/30 08:30:42] @ Quit: thegcat: Read error: 113 (No route to host) | ||
| [2009/03/30 08:34:06] @ Quit: authentic: Read error: 110 (Connection timed out) | ||
| [2009/03/30 08:35:45] @ Quit: DerekW: Read error: 110 (Connection timed out) | ||
| [2009/03/30 08:35:54] @ DerekW joined channel #puppet | ||
| [2009/03/30 08:37:56] @ ^authentic is now known as authentic | ||
| [2009/03/30 08:39:25] @ kibahop left channel #puppet () | ||
| [2009/03/30 08:45:45] <FiXion> anyone installed ruby rrd bindings on ubuntu and remember what path they choose? | ||
| [2009/03/30 08:45:57] * FiXion is hoping backports has an rrdtool version build with ruby bindings | ||
| [2009/03/30 08:46:13] <FiXion> or perhaps I should simply rebuild the rrdtool package with a small change to the debian/rules | ||
| [2009/03/30 08:46:21] @ Quit: madrescher: Read error: 110 (Connection timed out) | ||
| [2009/03/30 08:46:28] @ bajan joined channel #puppet | ||
| [2009/03/30 08:51:06] @ madrescher joined channel #puppet | ||
| [2009/03/30 08:53:09] @ mattock joined channel #puppet | ||
| [2009/03/30 08:54:11] @ kambiz_away is now known as kambiz | ||
| [2009/03/30 08:59:51] @ d3vilb0x joined channel #puppet | ||
| [2009/03/30 09:01:21] <FiXion> I'll rebuild the rrdtools package so it includes the ruby rrd bindings.. | ||
| [2009/03/30 09:04:02] @ Quit: webx: "..(cyp): bleh" | ||
| [2009/03/30 09:19:01] <ashp> oh man i get to work to find 6500 emails from majordomot | ||
| [2009/03/30 09:19:07] <ashp> this won't be a good week :/ | ||
| [2009/03/30 09:19:44] <jenza> you're not filtering enough to /dev/null | ||
| [2009/03/30 09:21:16] <FiXion> ashp: procmail's your friend :) | ||
| [2009/03/30 09:21:36] <ashp> someone managed to trigger a loop, i don't handle majordomo, but as nobody else was in it became my issue | ||
| [2009/03/30 09:21:47] <ashp> i fixed it by commenting out the mailing list :D | ||
| [2009/03/30 09:24:57] <ashp> also i really hate our backup solution, deploying updates is a shell script that you have to manually type 'y' at various points | ||
| [2009/03/30 09:25:00] <ashp> hardly easy to puppetise :/ | ||
| [2009/03/30 09:25:35] <realist> yes | ./install | ||
| [2009/03/30 09:25:59] <realist> Or use expect :-) | ||
| [2009/03/30 09:26:04] <ashp> i'm not sure that's going to work, as it's horrible, but i'm either going to do that or hack up the .sh to remove all the prompts and assume yes | ||
| [2009/03/30 09:26:25] <realist> I was cursing interactive installers the other day | ||
| [2009/03/30 09:26:32] * f3ew agrees | ||
| [2009/03/30 09:26:39] <joe-mac> god expect | ||
| [2009/03/30 09:26:41] <f3ew> Learn to take flags | ||
| [2009/03/30 09:26:44] <ashp> they really need to grasp that people aren't running 1 server anymore | ||
| [2009/03/30 09:26:47] <realist> At least this one had a "response file" | ||
| [2009/03/30 09:26:50] <joe-mac> omfg /me loads his revolver | ||
| [2009/03/30 09:26:53] <ashp> if you can't do some kind of --yes-to-all it's a terrible script | ||
| [2009/03/30 09:27:27] <fsweetser> that's actually one of my favorite features about rpm packages | ||
| [2009/03/30 09:27:47] <joe-mac> ashp: in the past i've used amanda if that's the kind of backup you're talking about | ||
| [2009/03/30 09:27:49] <joe-mac> it was really nice | ||
| [2009/03/30 09:27:49] <realist> joe-mac: you would cringe at looping ssh + expect scripts then :-) | ||
| [2009/03/30 09:27:50] <fsweetser> it's not possible for the install process to get user input, so the packager has no choice but to make it work non-interactively | ||
| [2009/03/30 09:27:53] <ashp> sadly it's 'commvault' | ||
| [2009/03/30 09:27:53] <joe-mac> imo | ||
| [2009/03/30 09:28:04] @ Quit: nmalik: "leaving" | ||
| [2009/03/30 09:28:17] @ nmalik joined channel #puppet | ||
| [2009/03/30 09:28:27] <realist> ashp: netbackup has the same issues installing/updating (y/n prompts) | ||
| [2009/03/30 09:29:30] @ Quit: nmalik: Client Quit | ||
| [2009/03/30 09:29:40] @ nmalik joined channel #puppet | ||
| [2009/03/30 09:30:01] @ Quit: madrescher: Read error: 110 (Connection timed out) | ||
| [2009/03/30 09:34:15] @ teratoma joined channel #puppet | ||
| [2009/03/30 09:34:20] <FiXion> annoying. failed to load library RRDtool for feature rrd | ||
| [2009/03/30 09:34:25] <FiXion> I'ts using ruby 1.8 | ||
| [2009/03/30 09:34:30] <FiXion> and it has a RRD.so file | ||
| [2009/03/30 09:34:44] <FiXion> all hints are welcome :) | ||
| [2009/03/30 09:37:07] @ Quit: Agrajag`: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 09:37:32] @ Agrajag` joined channel #puppet | ||
| [2009/03/30 09:37:35] <FiXion> I rebuild the ubuntu 8.10 package on 8.04 to get the bindings btw. | ||
| [2009/03/30 09:44:12] <omry|work> what is the 'api' for a service status command? | ||
| [2009/03/30 09:44:34] @ Quit: Agrajag`: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 09:45:02] @ Agrajag` joined channel #puppet | ||
| [2009/03/30 09:48:40] <Volcane> omry|work: LSB compliance | ||
| [2009/03/30 09:48:54] <Volcane> http://refspecs.freestandards.org/LSB_3.1.0/LSB-Core-generic/LSB-Core-generic/iniscrptact.html | ||
| [2009/03/30 09:49:08] <omry|work> Volcane, thanks. looking at it | ||
| [2009/03/30 09:50:12] @ lak joined channel #puppet | ||
| [2009/03/30 09:52:42] <FiXion> solved. I had to manually install RubyRRDTool-0.6.gem | ||
| [2009/03/30 09:53:08] <omry|work> Volcane, how can I easily verify what puppet is thinking about a particular service (specifically - if it detects a status command) | ||
| [2009/03/30 09:53:59] <Volcane> omry|work: it doesnt 'detect' status commands, u need to set hasstatus => true | ||
| [2009/03/30 09:54:24] <omry|work> Volcane, how does the detection works? | ||
| [2009/03/30 09:54:30] <Volcane> it doesnt detect | ||
| [2009/03/30 09:54:39] <omry|work> ah | ||
| [2009/03/30 09:54:45] <Volcane> if you dont tell it hasstatus, it ps -auxw|grep ${name} | ||
| [2009/03/30 09:54:49] <omry|work> thought you said IF it doesnt | ||
| [2009/03/30 09:55:06] <omry|work> right. and this fails because it's a java program | ||
| [2009/03/30 09:55:17] <omry|work> and the process name is java | ||
| [2009/03/30 09:55:21] <Volcane> yeah | ||
| [2009/03/30 09:55:24] @ cwebber joined channel #puppet | ||
| [2009/03/30 09:55:33] @ markl__ joined channel #puppet | ||
| [2009/03/30 09:55:39] <Volcane> so if your status command works, use hasstatus, else i belive you can pass it a pattern or something | ||
| [2009/03/30 09:55:55] <omry|work> the pattern for the process name is useless in this case | ||
| [2009/03/30 09:56:01] <omry|work> there are other java programs running around | ||
| [2009/03/30 09:56:06] <omry|work> I`ll try hasstatus | ||
| [2009/03/30 09:56:20] <omry|work> looking at the script, it should work. although it's not LSB complient | ||
| [2009/03/30 09:58:47] @ d3vilb0x_ joined channel #puppet | ||
| [2009/03/30 09:58:54] @ mattock left channel #puppet () | ||
| [2009/03/30 10:00:42] @ clayning joined channel #puppet | ||
| [2009/03/30 10:01:49] <FiXion> is it normal that puppet client constantly says: ensure changed stopped to running ? | ||
| [2009/03/30 10:01:53] <FiXion> even on services that are running? | ||
| [2009/03/30 10:02:12] @ Quit: kolla: Remote closed the connection | ||
| [2009/03/30 10:02:19] <Volcane> FiXion: what service? | ||
| [2009/03/30 10:03:20] <FiXion> Volcane: apache2 | ||
| [2009/03/30 10:03:23] <FiXion> and 2 others. | ||
| [2009/03/30 10:03:26] <Volcane> FiXion: read the exec type reference, esp about pattern, hasstatus etc | ||
| [2009/03/30 10:03:50] <FiXion> apache2 init.d file doesn't have a status - so I'll go read about patterns :) | ||
| [2009/03/30 10:03:54] <FiXion> Volcane: thanks | ||
| [2009/03/30 10:03:58] <Volcane> it basically just ps -auxw|grep name | ||
| [2009/03/30 10:04:07] <Volcane> so if you have service{"apache2": ensure => running} | ||
| [2009/03/30 10:04:13] <Volcane> it'll ps -auxw|grep apache2 | ||
| [2009/03/30 10:04:21] <FiXion> that's a match | ||
| [2009/03/30 10:04:22] @ ethan_rowe joined channel #puppet | ||
| [2009/03/30 10:04:45] <Volcane> yup, so apache2 should be fine, things like nfs, iptables etc, needs help | ||
| [2009/03/30 10:04:45] <FiXion> 0 /usr/sbin/apache2 -k start | ||
| [2009/03/30 10:04:49] <FiXion> 4 lines of these | ||
| [2009/03/30 10:05:31] <Volcane> well it probably doenst use -auxw but whatever, it greps the ps table | ||
| [2009/03/30 10:05:33] <FiXion> then why does it keep saying it's stopped | ||
| [2009/03/30 10:05:44] <Volcane> show us your whole puppetd --test output? | ||
| [2009/03/30 10:05:51] <FiXion> 2 secs | ||
| [2009/03/30 10:05:56] <Volcane> and the service definition | ||
| [2009/03/30 10:06:36] <FiXion> http://pastebin.com/m6ba8d313 | ||
| [2009/03/30 10:07:35] <FiXion> class (incl. service) def. http://pastebin.com/m6db7bd2b | ||
| [2009/03/30 10:09:24] <Volcane> if you tell it hasstatus => true, it'll try status on the service | ||
| [2009/03/30 10:09:36] <Volcane> which on debian* fails ofcourse, so it thinks the service is shut | ||
| [2009/03/30 10:09:51] <omry|work> how can I make all Package{} actions to depend on a particular action? | ||
| [2009/03/30 10:09:59] <omry|work> (without adding it to each an every one of them) | ||
| [2009/03/30 10:10:10] <Volcane> omry|work: site.pp Package{ require => ... } | ||
| [2009/03/30 10:10:31] <omry|work> Volcane, cool, I`ll try that. | ||
| [2009/03/30 10:10:38] <Volcane> omry|work: but ofcourse if you then set specific requires later on on a package, it'll override afaik | ||
| [2009/03/30 10:10:42] <omry|work> I already had Package { require .. } but it was not in site.pp | ||
| [2009/03/30 10:10:50] <FiXion> done.. embarrassing.. :( | ||
| [2009/03/30 10:10:58] <Volcane> yeah just need to set it out of a class/node scope and high up in the parsing | ||
| [2009/03/30 10:11:01] <omry|work> Volcane, okay - I`ll remember that. | ||
| [2009/03/30 10:11:04] <Volcane> FiXion:) | ||
| [2009/03/30 10:13:02] <duritong> lak: just got the confirmation. will be in nuernberg. :) | ||
| [2009/03/30 10:13:10] <FiXion> now I'm only wondering about one line.. http://pastebin.com/m6ba8d313 - line 3. why does it say that every time? | ||
| [2009/03/30 10:15:15] @ Quit: d3vilb0x: Read error: 110 (Connection timed out) | ||
| [2009/03/30 10:19:10] @ gaveen_ joined channel #puppet | ||
| [2009/03/30 10:19:20] @ Quit: gaveen: No route to host | ||
| [2009/03/30 10:19:21] <omry|work> Volcane, sounds logical to me that require => would normally add up to the exiting class requirements and not replace them | ||
| [2009/03/30 10:19:43] @ bajan left channel #puppet () | ||
| [2009/03/30 10:23:09] @ Quit: aZaFred: | ||
| [2009/03/30 10:29:00] @ lludwig joined channel #puppet | ||
| [2009/03/30 10:29:34] <lludwig> anyone know where I can get 1.5.4 of facter RPM? | ||
| [2009/03/30 10:30:27] @ mfoster1 joined channel #puppet | ||
| [2009/03/30 10:31:09] <Volcane> lludwig: theres a spec file in the tarball | ||
| [2009/03/30 10:31:38] <lludwig> yea I understand that, I would prefer if it was pre-built already. Lazy sysadmin :-) | ||
| [2009/03/30 10:31:42] <Volcane> lludwig: so it should take like a minute to build one :) | ||
| [2009/03/30 10:31:50] <Volcane> infact less time than it took to ask for it here :P | ||
| [2009/03/30 10:32:02] <omry|work> I found some .vim files for puppet files on the site. any idea where I drop them on debian? | ||
| [2009/03/30 10:32:06] <lludwig> not when you have 3, 4, 5 and i386 and x86_64 | ||
| [2009/03/30 10:33:14] <Volcane> :) | ||
| [2009/03/30 10:33:20] <Volcane> its a noarch on 5 at least :P | ||
| [2009/03/30 10:33:41] @ unxfrek joined channel #puppet | ||
| [2009/03/30 10:34:05] <lludwig> it's bad enough I have to support 3 manually | ||
| [2009/03/30 10:35:30] <lludwig> I HATE becoming my own distro | ||
| [2009/03/30 10:35:39] <Volcane> http://nephilim.ml.org/~rip/puppet/ | ||
| [2009/03/30 10:35:43] <Volcane> thats for 5 | ||
| [2009/03/30 10:35:47] <Volcane> and the srpm | ||
| [2009/03/30 10:35:49] @ Quit: thegcat_: | ||
| [2009/03/30 10:35:55] @ Quit: toi: Read error: 60 (Operation timed out) | ||
| [2009/03/30 10:36:02] <lludwig> thanks | ||
| [2009/03/30 10:36:14] <Volcane> i think local repos are unavoidable, esp since you cant rely on people like EPEL not to inject bullshit dependencies on beta software :( | ||
| [2009/03/30 10:39:11] @ jonmills joined channel #puppet | ||
| [2009/03/30 10:39:35] <jonmills> how does facter calculate the value of the FQDN? | ||
| [2009/03/30 10:39:45] <jonmills> I do not believe it uses resolv.conf | ||
| [2009/03/30 10:39:45] <lludwig> uname -a I believe | ||
| [2009/03/30 10:39:55] <Volcane> /usr/lib/ruby/site_ruby/1.8/facter/fqdn.rb | ||
| [2009/03/30 10:40:10] <Volcane> ah sorry, lame file to look at :) | ||
| [2009/03/30 10:40:13] <joe-mac> uname -a doesn't contain fqdn | ||
| [2009/03/30 10:40:22] <joe-mac> my guess is etc/hosts which is how hostname -f gets it | ||
| [2009/03/30 10:40:29] <kjetilho> uname -n is first choice, but it will look for domain/search in resolv.conf as one workaround | ||
| [2009/03/30 10:40:45] <Volcane> it concats /usr/lib/ruby/site_ruby/1.8/facter/hostname.rb and /usr/lib/ruby/site_ruby/1.8/facter/domain.rb | ||
| [2009/03/30 10:40:52] <kjetilho> if /etc/hosts doesn't have the fqdn as the first entry, either | ||
| [2009/03/30 10:42:55] @ andrewcshafer joined channel #puppet | ||
| [2009/03/30 10:47:41] <jonmills> okay, i'm reading domain.rb | ||
| [2009/03/30 10:47:51] <jonmills> it looks like facter looks at resolv.conf third | ||
| [2009/03/30 10:47:57] <jonmills> before that, it looks at | ||
| [2009/03/30 10:48:06] <jonmills> domain = Facter::Util::Resolution.exec('dnsdomainname') | ||
| [2009/03/30 10:48:15] <jonmills> and then | ||
| [2009/03/30 10:48:15] <jonmills> domain = Facter::Util::Resolution.exec('domainname') | ||
| [2009/03/30 10:50:14] <Volcane> where those are the comands /bin/dnsdomainname et | ||
| [2009/03/30 10:50:15] <Volcane> c | ||
| [2009/03/30 10:50:46] @ madrescher joined channel #puppet | ||
| [2009/03/30 10:50:48] <jonmills> i guess so, yeah. i don't know ruby, but I suppose the '.exec' means "run the shell command" | ||
| [2009/03/30 10:52:25] @ benblack|away is now known as benblack | ||
| [2009/03/30 10:52:34] <jonmills> my dire problem is that when it runs 'domainname' on the command line, the value returned is the NIS domain my company uses. This however is NOT the domain in DNS | ||
| [2009/03/30 10:52:49] <jonmills> so all my clients get their certname with the wrong FQDN | ||
| [2009/03/30 10:52:59] <jenza> yeah you might want to fix that. | ||
| [2009/03/30 10:53:05] <jonmills> no doubt | ||
| [2009/03/30 10:53:07] <jonmills> but how? | ||
| [2009/03/30 10:53:32] <jenza> add entries for the NIS domains into DNS | ||
| [2009/03/30 10:53:34] <jonmills> is there a way to override that, in puppet.conf for example? | ||
| [2009/03/30 10:53:46] <Volcane> i think there was a discussion on the list just recently titled overriding hostname | ||
| [2009/03/30 10:53:48] <jonmills> @ jenza: ha | ||
| [2009/03/30 10:53:49] <Volcane> something like that | ||
| [2009/03/30 10:53:53] <lak> duritong: good to hear, see you there | ||
| [2009/03/30 10:54:21] <jonmills> @ volcane: i'll search my list archives, thx | ||
| [2009/03/30 10:55:19] <jonmills> IMHO, they should modify domain.rb to force it to use resolv.conf first. that'd make a lot more sense | ||
| [2009/03/30 10:56:26] <kjetilho> yeah, it makes little or no sense to use the NIS domain name. | ||
| [2009/03/30 10:56:48] @ yure__ joined channel #puppet | ||
| [2009/03/30 10:56:57] <kjetilho> so dnsdomainname first, resolv.conf second | ||
| [2009/03/30 10:58:57] <jonmills> @ kjetilho: nope, resolv.conf is 3rd. it does `dnsdomainname` first, then `domainname` | ||
| [2009/03/30 10:59:33] <joe-mac> it's annoying that nagios counts SCHEDULED downtime as downtime, because when the $SUITS want the reports i have to go in and manually subtract any outages during the scheduled maintenance window, anybody here notice that? or i am just a dumbass | ||
| [2009/03/30 11:00:25] @ hessml|away joined channel #puppet | ||
| [2009/03/30 11:01:00] <jonmills> so, /bin/dnsdomainname is there in linux, and returns correct value for me. but it doesn't seem to be there on my endless numbers of Solaris 10 clients | ||
| [2009/03/30 11:02:36] @ Bass10 joined channel #puppet | ||
| [2009/03/30 11:02:45] <kjetilho> jonmills: so use a proper FQDN as the nodename | ||
| [2009/03/30 11:03:39] @ Quit: Bass10: SendQ exceeded | ||
| [2009/03/30 11:03:59] @ mib_w3fp7z97 joined channel #puppet | ||
| [2009/03/30 11:05:57] @ Quit: authentic: "[IRSSI]" | ||
| [2009/03/30 11:06:19] @ authentic joined channel #puppet | ||
| [2009/03/30 11:06:25] @ Quit: mib_w3fp7z97: Client Quit | ||
| [2009/03/30 11:06:42] @ Quit: grim_radical: Remote closed the connection | ||
| [2009/03/30 11:06:43] @ mib_vt8tc348 joined channel #puppet | ||
| [2009/03/30 11:09:34] @ Quit: verwilst: "Ex-Chat" | ||
| [2009/03/30 11:09:35] <mib_vt8tc348> I'm having a problem with using the nagios_service type. It's creating a _naginator_name (which I assume is added by the naginator back end). Unfortunately, nagios won't start because it doesn't recognize that option. | ||
| [2009/03/30 11:11:03] <ashp> this script had a hidden -silent option that's not in the docs, that was handy | ||
| [2009/03/30 11:11:47] <Volcane> they were obviously running -silent while writing the docs | ||
| [2009/03/30 11:15:07] @ hessml|away is now known as hessml|away|away | ||
| [2009/03/30 11:16:25] <SyTonnerre> Hm. If I'm rolling a custom Puppet type which requires an additional Ruby package, how would I declare that? | ||
| [2009/03/30 11:16:49] <SyTonnerre> Because I would think that there's a way to make the provider only includable if the Ruby module is installed | ||
| [2009/03/30 11:17:15] <SyTonnerre> But if I just require the module, puppetd will probably just fail? | ||
| [2009/03/30 11:17:25] @ benblack is now known as benblack|away | ||
| [2009/03/30 11:17:44] <Volcane> SyTonnerre: begin/trap | ||
| [2009/03/30 11:18:37] <SyTonnerre> Hm, I guess that's a way | ||
| [2009/03/30 11:18:41] @ Bass10 joined channel #puppet | ||
| [2009/03/30 11:19:21] <mib_vt8tc348> I just noticed the latest puppet has nagios fixes. I'll try that first and get back | ||
| [2009/03/30 11:19:24] <Volcane> http://pastie.org/431480 | ||
| [2009/03/30 11:19:30] @ Quit: Bass10: SendQ exceeded | ||
| [2009/03/30 11:19:31] <Volcane> SyTonnerre: like that, see that kind of thing often | ||
| [2009/03/30 11:19:57] @ Bass10 joined channel #puppet | ||
| [2009/03/30 11:20:14] <SyTonnerre> So the entire Puppet::Type declaration would be inside the begin? | ||
| [2009/03/30 11:20:49] * Volcane doesnt know about writing types, just saying if you want to do something thats gonna fail and screw up puppet, using exceptions is the way to go :) | ||
| [2009/03/30 11:21:14] @ benblack|away is now known as benblack | ||
| [2009/03/30 11:21:59] <mib_vt8tc348> Nevermind, latest nagios seems to fix my issue | ||
| [2009/03/30 11:22:04] <mib_vt8tc348> *latest puppet | ||
| [2009/03/30 11:22:31] @ Quit: benblack: "Leaving..." | ||
| [2009/03/30 11:29:02] <jonmills> @ Volcane: the only way to override my FQDN problem, I've found, is to create the puppet.conf file and echo in a correct value for 'certname = ' BEFORE installing puppet, or at least before allowing puppetd to launch (but many packages launch it automatically) | ||
| [2009/03/30 11:29:07] @ axisys joined channel #puppet | ||
| [2009/03/30 11:29:20] @ aZaFred joined channel #puppet | ||
| [2009/03/30 11:36:52] @ jonmills left channel #puppet () | ||
| [2009/03/30 11:38:05] @ alfism joined channel #puppet | ||
| [2009/03/30 11:40:56] @ Quit: MarlondB: | ||
| [2009/03/30 11:42:13] <lludwig> is there an env variable within puppet to get the path of your modules? | ||
| [2009/03/30 11:43:37] @ Quit: stijnbe: "Leaving..." | ||
| [2009/03/30 11:44:08] @ Quit: Innocenti: Client Quit | ||
| [2009/03/30 11:44:24] @ Quit: mib_vt8tc348: "http://www.mibbit.com ajax IRC Client" | ||
| [2009/03/30 11:50:44] @ jeanpierre joined channel #puppet | ||
| [2009/03/30 11:56:30] @ Llama joined channel #puppet | ||
| [2009/03/30 11:58:57] @ Quit: jeanpierre: "ChatZilla 0.9.84 [Iceweasel 3.0.6/2009020409]" | ||
| [2009/03/30 11:59:57] @ jeanpierre joined channel #puppet | ||
| [2009/03/30 12:00:17] @ kali67 joined channel #puppet | ||
| [2009/03/30 12:01:01] @ Quit: d3vilb0x_: Remote closed the connection | ||
| [2009/03/30 12:01:13] @ d3vilb0x joined channel #puppet | ||
| [2009/03/30 12:04:52] @ stijnbe joined channel #puppet | ||
| [2009/03/30 12:05:18] @ Quit: MattyM: "ta ta" | ||
| [2009/03/30 12:08:14] @ Quit: jeanpierre: "ChatZilla 0.9.84 [Iceweasel 3.0.6/2009020409]" | ||
| [2009/03/30 12:12:32] @ kolla joined channel #puppet | ||
| [2009/03/30 12:15:33] @ Quit: Welsh_Dwarf: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 12:15:48] @ benblack joined channel #puppet | ||
| [2009/03/30 12:18:55] @ jeanpierre joined channel #puppet | ||
| [2009/03/30 12:22:47] @ Quit: alfism: "Connection reset by beer" | ||
| [2009/03/30 12:27:07] @ Quit: alban2: Read error: 110 (Connection timed out) | ||
| [2009/03/30 12:28:34] @ Quit: DerekW: "Leaving" | ||
| [2009/03/30 12:29:31] @ nevele joined channel #puppet | ||
| [2009/03/30 12:35:50] @ erm_ joined channel #puppet | ||
| [2009/03/30 12:38:02] @ Quit: fbe__: Read error: 60 (Operation timed out) | ||
| [2009/03/30 12:42:24] <omry|work> does puppetmaster and puppetd leaks memory? | ||
| [2009/03/30 12:43:23] @ benblack is now known as benblack|away | ||
| [2009/03/30 12:46:00] @ benblack|away is now known as benblack | ||
| [2009/03/30 12:53:15] @ Quit: axisys: "leaving" | ||
| [2009/03/30 12:53:45] <plathrop> omry|work: There have been a couple memory leaks in the past but most (all?) have been plugged. It does, however, *use* a lot of memory, depending on what you are doing. | ||
| [2009/03/30 12:53:56] <plathrop> omry|work: A lot of fileserving will eat memory like crazy. | ||
| [2009/03/30 12:54:47] <omry|work> plathrop, I mostly serve small files (I use apt-get to deliver big files) | ||
| [2009/03/30 12:55:21] <plathrop> omry|work: what about recursive directory management? That's another big hog. | ||
| [2009/03/30 12:55:22] <omry|work> I noticed that puppetmaster used around 250mb, after I restarted it it went down to 40mb. | ||
| [2009/03/30 12:55:35] <plathrop> omry|work: Oh man, that's nothing. | ||
| [2009/03/30 12:55:39] <plathrop> :-P | ||
| [2009/03/30 12:55:42] <omry|work> yeah? | ||
| [2009/03/30 12:56:08] <omry|work> it's just that the machine running puppetmaster is already short on memory | ||
| [2009/03/30 12:56:12] <plathrop> omry|work: My puppetmasters hang out at around 500mb most of the time | ||
| [2009/03/30 12:56:38] <omry|work> what about puppetd? | ||
| [2009/03/30 12:57:20] <plathrop> I don't have easy stats on what my typical puppetd uses, but I know a lot of people run it from cron instead of as a daemon because of memory concerns. | ||
| [2009/03/30 13:00:19] @ Quit: MrHeavy: Remote closed the connection | ||
| [2009/03/30 13:05:24] @ toi joined channel #puppet | ||
| [2009/03/30 13:12:00] @ benblack is now known as benblack|away | ||
| [2009/03/30 13:12:48] @ alfism joined channel #puppet | ||
| [2009/03/30 13:14:10] <Volcane> http://pastie.org/431605 | ||
| [2009/03/30 13:14:14] <Volcane> theres some of my puppetd sizes | ||
| [2009/03/30 13:14:40] @ benblack|away is now known as benblack | ||
| [2009/03/30 13:28:05] <score> does anyone have any idea what all that memory is being used for? | ||
| [2009/03/30 13:28:49] @ jief- joined channel #puppet | ||
| [2009/03/30 13:29:29] <Volcane> the ruby vm on its own probably takes a load | ||
| [2009/03/30 13:29:38] <score> no way | ||
| [2009/03/30 13:29:46] <plathrop> score: in fileserving, currently (this will change with 0.25), puppetmaster has to store essentially 3 copies of the file in memory due to the XMLRPC encoding | ||
| [2009/03/30 13:29:59] <Volcane> plathrop: those figures are puppetd though | ||
| [2009/03/30 13:30:36] <plathrop> Volcane: Well, puppetd gets to do the same while decoding the files it receives, but it is typically lower than the master | ||
| [2009/03/30 13:30:48] <score> sure, but the numbers don't add up. 3x the size of the files i'm loading are almost nothing.. not 500mb | ||
| [2009/03/30 13:30:52] <Volcane> yeah but so why hang onto it? | ||
| [2009/03/30 13:31:05] <score> heh, for speed! | ||
| [2009/03/30 13:31:23] <Volcane> the daemons dont fluctuate that much afaik between running time and idling time | ||
| [2009/03/30 13:31:39] <plathrop> I'm not sure of the details, but a lot of work has been done to fix this in 0.25 | ||
| [2009/03/30 13:43:45] @ fujin joined channel #puppet | ||
| [2009/03/30 13:44:37] @ Quit: fujin: Client Quit | ||
| [2009/03/30 13:45:27] @ Quit: briandquinn: | ||
| [2009/03/30 13:46:10] @ benblack is now known as benblack|away | ||
| [2009/03/30 13:48:22] <jief-> hello. im using this code: http://pastebin.ca/1377087. but when i define $enabledmodules as an array, my config doesnt work anymore. it would try to enable for example headersstatus | ||
| [2009/03/30 13:48:32] <jief-> im really lost here, not sure how to fix it | ||
| [2009/03/30 13:49:52] @ WALoeIII joined channel #puppet | ||
| [2009/03/30 13:50:07] <jief-> this is based on the wiki page DebianApache2 | ||
| [2009/03/30 13:50:14] @ Tuf8 joined channel #puppet | ||
| [2009/03/30 13:51:55] <Volcane> 'doesnt work anymore' doesnt really help us a whole lot | ||
| [2009/03/30 13:52:10] <jief-> ok lemme develop this | ||
| [2009/03/30 13:52:28] <jief-> if i set $enabledmodules = [ "headers" ] | ||
| [2009/03/30 13:52:31] <jief-> in my node | ||
| [2009/03/30 13:52:59] <jief-> and call i t with apache2::modules::enabled { $$enabledmodules: ensure => present} | ||
| [2009/03/30 13:53:09] <Volcane> yes, i get that, but when you say 'doesnt work' I can only assume the server blows up, take out 5 street blocks with it etc? | ||
| [2009/03/30 13:53:09] <jief-> then the recipe works as expected, the module is enabled | ||
| [2009/03/30 13:53:12] @ Quit: unxfrek: Read error: 110 (Connection timed out) | ||
| [2009/03/30 13:53:25] <jief-> if i set $enabledmodules to an array, it expends the array into one argument | ||
| [2009/03/30 13:53:28] <jief-> instead of several | ||
| [2009/03/30 13:53:31] <jief-> that's where im lost | ||
| [2009/03/30 13:53:50] <Volcane> dont do "$array" for one | ||
| [2009/03/30 13:53:56] <Volcane> just $array is fine | ||
| [2009/03/30 13:54:13] @ benblack|away is now known as benblack | ||
| [2009/03/30 13:54:45] @ unxfrek joined channel #puppet | ||
| [2009/03/30 13:54:57] <Volcane> and in the define, say around line 6, please put notice("The module it: ${name}") | ||
| [2009/03/30 13:55:05] <Volcane> lets see what it says when you run puppetd --test | ||
| [2009/03/30 13:55:32] <jief-> weird, by removing the " " around the $array, it seems to work now | ||
| [2009/03/30 13:55:38] <jief-> lemme add more modules to test | ||
| [2009/03/30 13:55:41] <Volcane> yup | ||
| [2009/03/30 13:55:50] <Volcane> "$array" says make a string out of this array | ||
| [2009/03/30 13:56:15] <Volcane> so it does the only sane thing it can do and flatten it | ||
| [2009/03/30 13:56:23] <jief-> im still very new to puppet | ||
| [2009/03/30 13:56:26] <jief-> and its language | ||
| [2009/03/30 13:56:33] <jief-> some of it doesnt make sense yet | ||
| [2009/03/30 13:56:46] <Volcane> yeah, its a bit odd, though the array thing would bite you in most languages :) | ||
| [2009/03/30 13:57:17] <jief-> i sysadmin, programming is something i do for fun in my free time hehe | ||
| [2009/03/30 13:57:28] <jief-> although ive been trying to learn more Python these days | ||
| [2009/03/30 13:57:50] <jief-> thanks for your help, my recipe is fine now | ||
| [2009/03/30 13:57:54] <Volcane> no probs | ||
| [2009/03/30 13:58:25] <jief-> so much to think about when you're putting configs in puppet | ||
| [2009/03/30 13:58:27] <jief-> its insane | ||
| [2009/03/30 13:58:39] <jief-> should this be a class? subclass? will i want to reuse this? | ||
| [2009/03/30 13:59:01] <jief-> reminds me of packaging .deb | ||
| [2009/03/30 14:00:06] @ shake-n-bake joined channel #puppet | ||
| [2009/03/30 14:00:12] @ Quit: jeanpierre: "ChatZilla 0.9.84 [Iceweasel 3.0.6/2009020409]" | ||
| [2009/03/30 14:02:22] <Volcane> hehe | ||
| [2009/03/30 14:05:12] @ Quit: unxfrek: "Leaving" | ||
| [2009/03/30 14:05:28] @ Quit: stijnbe: Read error: 60 (Operation timed out) | ||
| [2009/03/30 14:08:38] @ stijnbe joined channel #puppet | ||
| [2009/03/30 14:24:27] <jief-> arg. just found another debianity/ubuntunity .. | ||
| [2009/03/30 14:30:29] @ MrHeavy joined channel #puppet | ||
| [2009/03/30 14:33:39] @ MarlondB joined channel #puppet | ||
| [2009/03/30 14:35:53] @ Quit: jief-: "Lost terminal" | ||
| [2009/03/30 14:37:09] <MrHeavy> puppetdoc is producing no output for me on any of my documented manifests. --debug shows only stuff like this: | ||
| [2009/03/30 14:37:17] <MrHeavy> info: scanning: ["modules/java6/manifests/init.pp"] | ||
| [2009/03/30 14:37:18] <MrHeavy> Any ideas? | ||
| [2009/03/30 14:37:45] @ Quit: glaw: ""Remember, information is not knowledge, knowledge is not wisdom, wisdom is not truth, truth is not beauty, beauty is not lov | ||
| [2009/03/30 14:38:09] <ashp> i really really | ||
| [2009/03/30 14:38:11] <ashp> should abandon nagios | ||
| [2009/03/30 14:38:15] <ashp> for zenoss or zabbix or something | ||
| [2009/03/30 14:38:41] <MrHeavy> I tried a long time ago | ||
| [2009/03/30 14:38:44] <ashp> i am tired of having to manually delete nagios_* and rerun puppet as the nagios types can't handle removing, only adding :/ | ||
| [2009/03/30 14:38:54] <MrHeavy> My environment is just too heterogeneous for anything but Nagios | ||
| [2009/03/30 14:39:02] <MrHeavy> Filling in a few templates is easier than clicking 500 times | ||
| [2009/03/30 14:39:41] <ashp> well, it has to be automated like nagios on puppet, just.. not as annoying | ||
| [2009/03/30 14:40:04] <MrHeavy> I haven't even attempted to bridge the Puppet/Nagios gap | ||
| [2009/03/30 14:40:29] <MrHeavy> I have way overtemplated my Nagios 3.0 configs since they allowed you to do much cooler things with multiple inheritance and user-defined macros | ||
| [2009/03/30 14:43:24] <ashp> i think i just really dislike nagios, it seems clumsy as hell, maybe once you're an expert it's ok | ||
| [2009/03/30 14:43:28] <ashp> zenoss seemed... 'heavy' | ||
| [2009/03/30 14:43:31] <ashp> maybe zabbix is better | ||
| [2009/03/30 14:43:58] <ashp> all my requirements really come down to 'monitor and email if it breaks.. trigger ssh to remote host/restart certain things if monitoring fails | ||
| [2009/03/30 14:43:59] <joe-mac> lol, i think by the time anyone becomes a nagios 3.x "expert" 5.x will be rolling around | ||
| [2009/03/30 14:44:44] <MrHeavy> I became a Nagios 3.x "expert" in about two days of tooling around with the new features, there's really not that much to it. It's the ecosystem of addons that make Nagios useful, like pnp4nagios, that make Nagios actually tricky | ||
| [2009/03/30 14:45:16] <MrHeavy> NSCA is the really annoying part, but DNX has made it largely unnecessary | ||
| [2009/03/30 14:45:17] <joe-mac> i don't use the nagios types in puppet. fwiu they are almost the same as typing out the configs, so without puppet storedconfigs they are kinda useless | ||
| [2009/03/30 14:45:26] <joe-mac> what is DNX? | ||
| [2009/03/30 14:45:29] <joe-mac> i use nsca | ||
| [2009/03/30 14:45:33] <MrHeavy> http://dnx.sourceforge.net | ||
| [2009/03/30 14:45:35] <ashp> i am using storedconfigs | ||
| [2009/03/30 14:45:50] <joe-mac> it runs on my nagios server and accepts checks froma ll my nodes via a small shell script | ||
| [2009/03/30 14:46:16] <ashp> hmm, i'll check out dnx | ||
| [2009/03/30 14:46:28] <joe-mac> looks like a lot of work | ||
| [2009/03/30 14:46:57] <joe-mac> i install nagios plugins, push out my check_psv script and send_nsca script via puppet, and the results are sent in that way | ||
| [2009/03/30 14:46:59] <MrHeavy> DNX is still a tiny bit unstable at times and needs a little babysitting, but was way easier for me than dealing with NSCA, even with Puppet | ||
| [2009/03/30 14:47:06] <joe-mac> i don't run any actual daemon on the hosts | ||
| [2009/03/30 14:47:57] <MrHeavy> I had considered running that way, but I didn't want to mix and match, since we have a lot of Windows servers running NSClient++ in our environment | ||
| [2009/03/30 14:48:01] @ Quit: shake-n-bake: Read error: 110 (Connection timed out) | ||
| [2009/03/30 14:53:25] <ed209> we use zenoss at work and I really don't like it | ||
| [2009/03/30 14:53:46] <ashp> ed209: any major reason, or just feels clumsy? | ||
| [2009/03/30 14:53:59] <ed209> I find it a lot harder to tell what is actually going on in the network | ||
| [2009/03/30 14:54:12] <ashp> the only real reason I use nagios is because there's a puppet module I can crib into building a lot of my monitoring for me :) | ||
| [2009/03/30 14:54:37] <ashp> I guess I need to go back to deciding if I want to use snmp or nrpe or whatever to get my local info from hosts rather than playing with alternatives | ||
| [2009/03/30 14:54:55] <ed209> zenoss has its "discover" tool which I find is reliable only on our redhat boxes | ||
| [2009/03/30 14:55:13] <MrHeavy> I really wish Nagios had some better agent tools | ||
| [2009/03/30 14:55:32] <MrHeavy> Something that took the guesswork out of figuring out what to monitor on a host would be great | ||
| [2009/03/30 14:55:33] <ed209> windows hosts, solaris, ups's, switches, etc all seem to be a crap shoot when you hit "discover" | ||
| [2009/03/30 14:55:47] <ed209> use munin | ||
| [2009/03/30 14:55:58] <ashp> I thought about trying munin but isn't that really focused on graphs | ||
| [2009/03/30 14:56:01] <ashp> not host monitoring | ||
| [2009/03/30 14:56:02] <ed209> or ganglion | ||
| [2009/03/30 14:56:17] <ed209> use munin and have it drop alerts through nagios | ||
| [2009/03/30 14:56:25] <ashp> i do wonder if 'god' is any good | ||
| [2009/03/30 14:56:39] <ed209> never had any use for god | ||
| [2009/03/30 14:56:43] <ed209> ;) | ||
| [2009/03/30 14:56:47] <ashp> has to be easy to integrate.. both being ruby | ||
| [2009/03/30 14:56:57] <fsweetser> I just wish I could find a good monitoring tool that could handle a non-DAG dependency tree... | ||
| [2009/03/30 14:57:21] <ashp> it's frustrating that something that should be easy (monitoring) is so hard to get right and find a tool | ||
| [2009/03/30 14:57:35] <MrHeavy> Network monitoring is by no means something that should be easy :( | ||
| [2009/03/30 14:57:37] <ed209> who said monitoring should be easy? | ||
| [2009/03/30 14:57:42] <ed209> ha | ||
| [2009/03/30 14:57:43] <MrHeavy> High five. | ||
| [2009/03/30 14:58:03] <ashp> of course it _should_ be easy, everything should be :) | ||
| [2009/03/30 14:58:20] <fsweetser> monitoring is pretty easy, really | ||
| [2009/03/30 14:58:28] <fsweetser> it's correlation that's tricky | ||
| [2009/03/30 14:58:39] <MrHeavy> We made subprime lending, short-term corporate incentivization and toxic asset handling really easy and look where that got us | ||
| [2009/03/30 14:58:52] <ed209> simple event correlator works ... if you can figure it out :) | ||
| [2009/03/30 14:59:14] <fsweetser> sadly, "simple" stuff would look at my actual dependencies and die an instant death | ||
| [2009/03/30 14:59:36] <ed209> simple events, not a simple correlator | ||
| [2009/03/30 14:59:45] <MrHeavy> There's lots of products out there that aim to make it simple | ||
| [2009/03/30 14:59:45] <fsweetser> ah | ||
| [2009/03/30 14:59:48] <MrHeavy> Hyperic HQ is one | ||
| [2009/03/30 15:00:00] <fsweetser> yeah, I've been playing with a relatively new one from nortel | ||
| [2009/03/30 15:00:00] <MrHeavy> The problem is that they make it so simple that it never, ever does anything right in the real world | ||
| [2009/03/30 15:00:18] <ed209> simple event correlator is different | ||
| [2009/03/30 15:00:20] <fsweetser> sadly, most of the FOSS ones out there can't handle the kinds of loops that are standard fare in overengineered networks | ||
| [2009/03/30 15:00:48] <MrHeavy> I want an open-source monitoring app that actually acknowledges that sometimes people run remote sites | ||
| [2009/03/30 15:01:09] <ed209> you might find SEC would work for you, cuz you gotta write all the rules yourself | ||
| [2009/03/30 15:01:26] @ mvn071 joined channel #puppet | ||
| [2009/03/30 15:01:36] <fsweetser> ed209: got a url? | ||
| [2009/03/30 15:01:44] <ed209> just diggin it up now | ||
| [2009/03/30 15:01:58] <MrHeavy> http://kodu.neti.ee/~risto/sec/ | ||
| [2009/03/30 15:02:08] <ed209> ^ thats the one | ||
| [2009/03/30 15:02:10] <ed209> thanx | ||
| [2009/03/30 15:02:49] <ed209> its got its own language, so for puppet enthusiasts I assume you'd be cool with that | ||
| [2009/03/30 15:02:49] <MrHeavy> OSSEC is nice for system logs and FS changes out of the box | ||
| [2009/03/30 15:03:13] <ed209> I like ossec as well | ||
| [2009/03/30 15:03:20] <MrHeavy> The problem I find with these things is that nothing makes it really easy to correlate data that describes different types of things | ||
| [2009/03/30 15:03:25] <MrHeavy> So like | ||
| [2009/03/30 15:03:37] <MrHeavy> There's lots of stuff out there like OSSEC and Splunk that correlates logs with other logs | ||
| [2009/03/30 15:03:49] <MrHeavy> There's almost nothing correlating logs with performance data with process tables with logged-in users | ||
| [2009/03/30 15:04:05] <fsweetser> interesting, thanks | ||
| [2009/03/30 15:04:24] <ed209> did you check out SEC, MrHeavy? | ||
| [2009/03/30 15:04:27] <MrHeavy> Yeah | ||
| [2009/03/30 15:04:44] <MrHeavy> The problem is that not everything you want to correlate is necessarily an "event" | ||
| [2009/03/30 15:04:53] <MrHeavy> Something can be relevant data without being an event | ||
| [2009/03/30 15:05:00] <ed209> oh I see what you're saying | ||
| [2009/03/30 15:05:16] <ed209> ya thats why you get paid the big bucks tho :) | ||
| [2009/03/30 15:05:17] <MrHeavy> There's events and there's states | ||
| [2009/03/30 15:05:22] <MrHeavy> Nothing correlates one to the other | ||
| [2009/03/30 15:05:44] <MrHeavy> It seems like it would be a great way to monetize the entire IT industry if you could figure out a good way to do that :) | ||
| [2009/03/30 15:05:53] <MrHeavy> Anomaly detection in states is ridiculously tough to do though | ||
| [2009/03/30 15:06:08] @ Quit: yarihm: "This computer has gone to sleep" | ||
| [2009/03/30 15:06:49] <ed209> you'd need to have a snapshot of the states at the timestamp of the events | ||
| [2009/03/30 15:07:15] <ed209> just super imposed events over your state graphs | ||
| [2009/03/30 15:08:04] <ed209> that would take some doin, and it would make distributed monitoring tougher | ||
| [2009/03/30 15:08:14] <ed209> but I don't think its an intractable problem | ||
| [2009/03/30 15:15:46] @ Quit: nmalik: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:47] @ Quit: int: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:47] @ Quit: FUF: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:48] @ Quit: kjetilho: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:48] @ Quit: tlockney: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:48] @ Quit: lutter: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:49] @ Quit: omry|work: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:49] @ Quit: Tuf8: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:49] @ Quit: d3vilb0x: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:50] @ Quit: jenza: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:50] @ Quit: goozbach: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:51] @ Quit: Superfly_: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:51] @ Quit: glut: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:52] @ Quit: stick: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:52] @ Quit: Bosse: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:53] @ Quit: Ned_: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:53] @ Quit: justdave: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:53] @ Quit: fox2mike: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:55] @ Quit: cheflog__: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:55] @ Quit: rlpowell: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:55] @ Quit: philip__: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:55] @ Quit: markl___: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:56] @ Quit: zirpu: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:56] @ Quit: josb: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:56] @ Quit: Disconnect: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:57] @ Quit: dmosemp: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:57] @ Quit: ralfgro_: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:57] @ Quit: lurbs: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:58] @ Quit: jason^: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:58] @ Quit: tsb_: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:15:58] @ Quit: ricky: kubrick.freenode.net irc.freenode.net | ||
| [2009/03/30 15:16:16] @ Tuf8 joined channel #puppet | ||
| [2009/03/30 15:16:16] @ d3vilb0x joined channel #puppet | ||
| [2009/03/30 15:16:16] @ nmalik joined channel #puppet | ||
| [2009/03/30 15:16:16] @ int joined channel #puppet | ||
| [2009/03/30 15:16:16] @ justdave joined channel #puppet | ||
| [2009/03/30 15:16:17] @ jenza joined channel #puppet | ||
| [2009/03/30 15:16:17] @ tlockney joined channel #puppet | ||
| [2009/03/30 15:16:17] @ FUF joined channel #puppet | ||
| [2009/03/30 15:16:17] @ kjetilho joined channel #puppet | ||
| [2009/03/30 15:16:17] @ lutter joined channel #puppet | ||
| [2009/03/30 15:16:17] @ omry|work joined channel #puppet | ||
| [2009/03/30 15:16:17] @ goozbach joined channel #puppet | ||
| [2009/03/30 15:16:17] @ cheflog__ joined channel #puppet | ||
| [2009/03/30 15:16:17] @ stick joined channel #puppet | ||
| [2009/03/30 15:16:17] @ Bosse joined channel #puppet | ||
| [2009/03/30 15:16:17] @ glut joined channel #puppet | ||
| [2009/03/30 15:16:17] @ Ned_ joined channel #puppet | ||
| [2009/03/30 15:16:17] @ Superfly_ joined channel #puppet | ||
| [2009/03/30 15:16:17] @ rlpowell joined channel #puppet | ||
| [2009/03/30 15:16:18] @ philip__ joined channel #puppet | ||
| [2009/03/30 15:16:18] @ markl___ joined channel #puppet | ||
| [2009/03/30 15:16:18] @ zirpu joined channel #puppet | ||
| [2009/03/30 15:16:18] @ fox2mike joined channel #puppet | ||
| [2009/03/30 15:16:18] @ lurbs joined channel #puppet | ||
| [2009/03/30 15:16:18] @ ralfgro_ joined channel #puppet | ||
| [2009/03/30 15:16:18] @ dmosemp joined channel #puppet | ||
| [2009/03/30 15:16:18] @ jason^ joined channel #puppet | ||
| [2009/03/30 15:16:18] @ ricky joined channel #puppet | ||
| [2009/03/30 15:16:18] @ Disconnect joined channel #puppet | ||
| [2009/03/30 15:16:18] @ tsb_ joined channel #puppet | ||
| [2009/03/30 15:16:18] @ josb joined channel #puppet | ||
| [2009/03/30 15:21:17] @ Quit: lak: | ||
| [2009/03/30 15:22:22] @ shake-n-bake_ joined channel #puppet | ||
| [2009/03/30 15:27:04] @ silassewell joined channel #puppet | ||
| [2009/03/30 15:27:49] <silassewell> Anyone know if its possible to do a multi-line string in a manifest? | ||
| [2009/03/30 15:29:24] <silassewell> Searching google for puppet + string is fail. | ||
| [2009/03/30 15:32:59] @ shake-n-bake joined channel #puppet | ||
| [2009/03/30 15:33:38] @ Quit: silassewell: "leaving" | ||
| [2009/03/30 15:34:58] @ Quit: d3vilb0x: Read error: 110 (Connection timed out) | ||
| [2009/03/30 15:41:35] @ alban2 joined channel #puppet | ||
| [2009/03/30 15:42:54] @ Quit: shake-n-bake_: Read error: 110 (Connection timed out) | ||
| [2009/03/30 15:45:26] @ omry joined channel #puppet | ||
| [2009/03/30 15:45:47] @ lak joined channel #puppet | ||
| [2009/03/30 15:47:49] @ Quit: yure__: "Odhajam" | ||
| [2009/03/30 15:49:45] @ fbe joined channel #puppet | ||
| [2009/03/30 15:49:46] @ Quit: shake-n-bake: Read error: 110 (Connection timed out) | ||
| [2009/03/30 15:50:06] <joe-mac> hey is anybody here really slick with what chars need escaping in puppet manifests? | ||
| [2009/03/30 15:50:28] <joe-mac> i am assuming = signs need tyo be escaped | ||
| [2009/03/30 15:59:01] @ shake-n-bake joined channel #puppet | ||
| [2009/03/30 16:01:54] @ Quit: gaveen_: "Enough lurking for now" | ||
| [2009/03/30 16:03:00] @ Quit: Bass10: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 16:04:24] @ Bass10 joined channel #puppet | ||
| [2009/03/30 16:04:48] @ Quit: markl_: Read error: 60 (Operation timed out) | ||
| [2009/03/30 16:07:00] @ markl_ joined channel #puppet | ||
| [2009/03/30 16:07:13] @ Innocenti joined channel #puppet | ||
| [2009/03/30 16:09:33] <joe-mac> hasving a hell ogf a time with a one line sed replacement | ||
| [2009/03/30 16:09:38] <joe-mac> driving me to suicide | ||
| [2009/03/30 16:12:23] <andrewcshafer> joe-mac: what are you trying to do? | ||
| [2009/03/30 16:13:26] <joe-mac> andrewcshafer: replace the include line in php.ini | ||
| [2009/03/30 16:13:35] <joe-mac> i can do it from the command line but just can't get it to work in puppet | ||
| [2009/03/30 16:14:24] <andrewcshafer> you are just trying to exec the sed? what is the command? | ||
| [2009/03/30 16:15:55] <joe-mac> in the shell it would be sed -i 's/^;include_path = ".:\/usr\/share\/php"$/include_path = "\/usr\/share\/php:\/usr\/share\/pear:\/var\/www\/pci\/includes"/' /etc/php5/apache2/php.ini | ||
| [2009/03/30 16:17:24] <joe-mac> i thought i got somewhere when puppet --parseonly finally stopped erroring but now i just can't get it to exec, maybe it's my onlyif | ||
| [2009/03/30 16:18:21] <joe-mac> o man hold on i am going to punc my own face i think | ||
| [2009/03/30 16:20:33] @ ijcd_ joined channel #puppet | ||
| [2009/03/30 16:20:37] <andrewcshafer> is it not running the exec at all? what is the output from --debug? | ||
| [2009/03/30 16:22:53] @ Quit: aymerick: "kit mais sage" | ||
| [2009/03/30 16:23:08] <joe-mac> idk i am at the point where i think i just made things a lot worse | ||
| [2009/03/30 16:23:34] <joe-mac> i'm so freaking tired, i hate when sed pisses me off wehen i am tired | ||
| [2009/03/30 16:23:46] <joe-mac> i am doing just parseonlyu | ||
| [2009/03/30 16:25:07] <joe-mac> remember kids, for every " you need a matching "! | ||
| [2009/03/30 16:25:14] * joe-mac cocks .45 | ||
| [2009/03/30 16:28:00] <joe-mac> still can't get this, andrewcshafer is there a way for me to take a single exec block and like parse it to see what puppet would exec? | ||
| [2009/03/30 16:28:32] <kjetilho> joe-mac: you should use that recipe on the wiki | ||
| [2009/03/30 16:28:43] <Volcane> joe-mac: check out ralsh | ||
| [2009/03/30 16:29:08] <andrewcshafer> why not just run it on a VM with --debug? How do you test your manifests? | ||
| [2009/03/30 16:29:41] <joe-mac> i just tested it, it's included the backslashes as literaals | ||
| [2009/03/30 16:30:00] <joe-mac> ralsh can build exec statements???? | ||
| [2009/03/30 16:30:36] @ Quit: madrescher: Read error: 110 (Connection timed out) | ||
| [2009/03/30 16:31:40] <Volcane> ralsh exec /bin/date | ||
| [2009/03/30 16:31:51] <joe-mac> thanks Volcane let me check that out | ||
| [2009/03/30 16:32:11] @ madrescher joined channel #puppet | ||
| [2009/03/30 16:41:28] @ martha left channel #puppet () | ||
| [2009/03/30 16:42:48] @ Quit: lak: | ||
| [2009/03/30 16:43:15] <JD> I'm guessing there's no way to allow a define to be called twice with the same value | ||
| [2009/03/30 16:43:22] <JD> and just have puppet do it once | ||
| [2009/03/30 16:43:53] @ Demosthenes joined channel #puppet | ||
| [2009/03/30 16:44:36] <fsweetser> JD: check out virtual resource and the realize function | ||
| [2009/03/30 16:46:31] @ Quit: Bass10: Connection timed out | ||
| [2009/03/30 16:46:36] <JD> fsweetser: ah. yeah, I forgot about that | ||
| [2009/03/30 16:46:49] <JD> I'd only seen it used for built in types | ||
| [2009/03/30 16:47:01] <JD> didn't think about using it for user-defined defines | ||
| [2009/03/30 16:51:24] <plathrop> So, here's an annoying problem | ||
| [2009/03/30 16:51:35] <plathrop> I want to install lighttpd with a manifest | ||
| [2009/03/30 16:51:46] <plathrop> I want to install it on a server that isalready running apache | ||
| [2009/03/30 16:52:08] <plathrop> But, because of Debian's brain-dead packaging policies, lighttpd tries to start as part of its install | ||
| [2009/03/30 16:52:19] <plathrop> of course, it can't start, because apache is running, and I'm okay with that. | ||
| [2009/03/30 16:52:41] <plathrop> But aptitude still considers this a failed install and thus puppet reads it as an error | ||
| [2009/03/30 16:52:46] <plathrop> How do I work around this? | ||
| [2009/03/30 16:53:18] <joe-mac> who wants to win a cookie and tell me why this doesn't work http://www.pastie.org/431904 | ||
| [2009/03/30 16:53:28] @ Quit: yure_: Connection timed out | ||
| [2009/03/30 16:53:43] <fsweetser> '/bin/sed 's | ||
| [2009/03/30 16:53:44] <joe-mac> plathrop: do debconf-get-selections on the package | ||
| [2009/03/30 16:53:45] @ Quit: yure: Read error: 110 (Connection timed out) | ||
| [2009/03/30 16:53:50] <fsweetser> you need to escape the second ' | ||
| [2009/03/30 16:54:00] <joe-mac> fsweetser: let me give it a try | ||
| [2009/03/30 16:54:16] @ yure_ joined channel #puppet | ||
| [2009/03/30 16:54:18] <plathrop> joe-mac: can you either give me a bit more detail or point me at the relevant docs? | ||
| [2009/03/30 16:54:26] @ yure joined channel #puppet | ||
| [2009/03/30 16:54:38] <fsweetser> you also need to escape the inner ' pair on the onlyif line | ||
| [2009/03/30 16:54:45] <joe-mac> sorry plathropi couldn't finish | ||
| [2009/03/30 16:54:57] <plathrop> joe-mac: no worries :-) | ||
| [2009/03/30 16:55:01] <joe-mac> after you run debconf-get-selections on lighthttpd you should see a bunch of preseed values | ||
| [2009/03/30 16:55:41] <joe-mac> hopefully, and i mean really HOPE (with debian you need a lot of hope and macgyverisms) that there is a variable to start automatically, and you can disable that. then yopu can save the preseed file and feed it to your package definition | ||
| [2009/03/30 16:55:52] <joe-mac> i use a preseed like that to configure postfix | ||
| [2009/03/30 16:56:17] <joe-mac> plathrop: (//Node[scan-stag.osdc.bos.rapid7.com]/pci_portal/Exec[/bin/sed -i s/^;include_path \= .\$:\/usr\/share\/php\$/include_path \= \/usr\/share\/php:\/usr\/share\/pear:\/var\/www\/pci\/includes/ /etc/php5/apache2/php.ini]/returns) change from notrun to 0 failed: /bin/sed: -e expression #1, char 3: unterminated `s' command sh: include_path: not found | ||
| [2009/03/30 16:56:18] <gepetto> joe-mac: plathrop: #1 is http://projects.reductivelabs.com/issues/show/1 "Puppet - Feature #1: Differentiate classes from definitions - ReductiveLabs.com" | ||
| [2009/03/30 16:56:19] @ d3vilb0x joined channel #puppet | ||
| [2009/03/30 16:56:42] <plathrop> joe-mac: whoah, that's a lot of output (debconf-get-selections lighttpd) | ||
| [2009/03/30 16:57:02] <joe-mac> yea you'll have to sift through it, that last error message was for fsweetser though | ||
| [2009/03/30 16:57:09] <joe-mac> i can't freaking get this to worrkkkkkkkkkkkkkkkkk | ||
| [2009/03/30 16:57:24] <joe-mac> :-( it works in the shell fine, i am li ke two seconds away from just serving out a one line bash script to do it | ||
| [2009/03/30 16:58:28] @ Quit: toi: Read error: 110 (Connection timed out) | ||
| [2009/03/30 16:58:42] <plathrop> joe-mac: do I have to do this on a server where I *can* install lighttpd successfully? | ||
| [2009/03/30 16:58:50] <plathrop> joe-mac: I don't see anything about lighttpd in here | ||
| [2009/03/30 16:59:00] <joe-mac> the package has to be present | ||
| [2009/03/30 16:59:21] <joe-mac> then you have to dump the debconf-get-selections out to a file i usually name $package.preseed | ||
| [2009/03/30 16:59:32] <joe-mac> there is actually a decent little recipe for preseed packages on the wiki too which i also use | ||
| [2009/03/30 17:00:54] <fsweetser> joe-mac: can you pastebin that again? | ||
| [2009/03/30 17:02:33] <joe-mac> fsweetser: i'm ging to paste what works in a shell first then the exec def | ||
| [2009/03/30 17:02:39] <fsweetser> ok | ||
| [2009/03/30 17:03:52] @ hessml|away|away is now known as hessml|away | ||
| [2009/03/30 17:04:19] @ oboxodo joined channel #puppet | ||
| [2009/03/30 17:07:14] <joe-mac> well fsweetser somehow i have f'd up the exec resource beyond all fixing but here is what works in bash | ||
| [2009/03/30 17:07:30] <joe-mac> http://pastie.org/431926 | ||
| [2009/03/30 17:08:55] @ kambiz is now known as kambiz_away | ||
| [2009/03/30 17:11:07] <kjetilho> joe-mac: again, I recommend you just use ensure_key_value from http://reductivelabs.com/trac/puppet/wiki/Recipes/PasswordManagement | ||
| [2009/03/30 17:11:13] <plathrop> joe-mac: hrm. I can't find anything that looks like it might keep lighttpd from starting. Any pointers on what kind of thing I'd be looking for? | ||
| [2009/03/30 17:11:48] <kjetilho> joe-mac: what distro are you using btw? | ||
| [2009/03/30 17:11:50] <joe-mac> plathrop: the debconf stuff is poorly documented unfortunately. | ||
| [2009/03/30 17:12:04] <joe-mac> kjetilho: ubuntu (don't laugh and point, i was forced to stick with status quo" | ||
| [2009/03/30 17:12:05] <joe-mac> ) | ||
| [2009/03/30 17:12:37] <kjetilho> joe-mac: doh. Ubuntu specifically disable grep -P for some idiotic reason | ||
| [2009/03/30 17:12:52] <joe-mac> lol, ubuntu specifically disables my sanity for some reason | ||
| [2009/03/30 17:12:58] <fsweetser> joe-mac: give this a try http://www.pastie.org/431934 | ||
| [2009/03/30 17:15:03] @ bolt joined channel #puppet | ||
| [2009/03/30 17:15:21] <joe-mac> fsweetser: err: Could not parse for environment development: Syntax error at 'php5'; expected '}' at /etc/puppet/manifests/definitions/pci_portal.pp:60 | ||
| [2009/03/30 17:15:46] <fsweetser> ah, yep, I see the problem | ||
| [2009/03/30 17:15:59] <fsweetser> make the onlyif line end with php.ini'", | ||
| [2009/03/30 17:16:15] <joe-mac> OH CHRIST | ||
| [2009/03/30 17:16:30] <oboxodo> Hi. I've inherited puppet running on a bunch of RHEL 5 servers and am having a problem when puppet tries to create a new user telling me "Could not find a default provider for user". I guess it's not finding useradd because RHEL doesn'รพ have /usr/sbin in the path. Anybody has a clue on how to fix this? | ||
| [2009/03/30 17:16:51] <fsweetser> oboxodo: check if you have the ruby-shadow package installed | ||
| [2009/03/30 17:17:07] <fsweetser> joe-mac: is that good or bad? | ||
| [2009/03/30 17:17:16] <oboxodo> fsweetser: in the puppet master or the client? master, right? | ||
| [2009/03/30 17:17:30] <fsweetser> client | ||
| [2009/03/30 17:17:43] <fsweetser> all of the useradd/etc stuff is actually executed client side | ||
| [2009/03/30 17:20:02] <oboxodo> fsweetser: no I don't. :) | ||
| [2009/03/30 17:20:31] <fsweetser> that's probably your problem, then! | ||
| [2009/03/30 17:20:54] <oboxodo> thanks. I'll keep going. | ||
| [2009/03/30 17:21:22] <Volcane> oboxodo: check the path | ||
| [2009/03/30 17:21:32] <joe-mac> fsweetser: step in the rigght direcvtyon for sure | ||
| [2009/03/30 17:21:36] <joe-mac> bvut still not right | ||
| [2009/03/30 17:21:55] <fsweetser> out of curiosity, is removing the original line that starts with ';' actually necessary? | ||
| [2009/03/30 17:22:20] <Volcane> oboxodo: try running 'puppetd --test --path=/usr/bin:/usr/sbin:/bin:/sbin | ||
| [2009/03/30 17:22:40] <fsweetser> or can you just append the new line you want? | ||
| [2009/03/30 17:22:45] <joe-mac> fsweetser: i suppose not | ||
| [2009/03/30 17:22:51] @ hessml|away is now known as hessml|away|away | ||
| [2009/03/30 17:23:00] <fsweetser> in that case, you could just use the append_line definition | ||
| [2009/03/30 17:23:02] <fsweetser> much simpler | ||
| [2009/03/30 17:23:02] <joe-mac> i guess i will cause i think that semicolon is causing some kind of issue | ||
| [2009/03/30 17:23:13] <joe-mac> whoa whoa what append_line definition? | ||
| [2009/03/30 17:23:43] <fsweetser> hm, can't find it on the wiki now | ||
| [2009/03/30 17:23:44] <joe-mac> i gotta head home fsweetseri will look into it, thanks | ||
| [2009/03/30 17:23:50] <fsweetser> hold on, let me dig... | ||
| [2009/03/30 17:23:51] @ duncanmv joined channel #puppet | ||
| [2009/03/30 17:23:56] <duncanmv> hi guys | ||
| [2009/03/30 17:24:06] <joe-mac> if you find something and feel so obnliged send it to me joseph.e.mcdonagh@gmail.com, thanks | ||
| [2009/03/30 17:24:17] @ joe-mac left channel #puppet () | ||
| [2009/03/30 17:24:38] <fsweetser> figures, I just found it... | ||
| [2009/03/30 17:25:16] @ Quit: raphink: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 17:25:28] @ Quit: erm_: "leaving" | ||
| [2009/03/30 17:26:12] <duncanmv> guys, I want to have a list of packages that should be ensure installed or ensure latest for every node, this is dynamic (not per class) am reading the external_nodes documentation and I see one can only say which class a node is. Is possible to specify the list of packages per node? | ||
| [2009/03/30 17:27:10] <Volcane> duncanmv: make an array per host, use the array in package{} | ||
| [2009/03/30 17:27:36] <duncanmv> I have evidence against me: http://markmail.org/message/vuet5klz4hxvks7g#query:puppet%20external_nodes%20ensure%20latest+page:1+mid:4zy6qc5uvxfl4ld2+state:results | ||
| [2009/03/30 17:27:53] <duncanmv> Volcane: any pointer to an example with array per host? | ||
| [2009/03/30 17:28:54] <JD> $packages = ["foo", "bar"] package{$packages: ensure => latest} | ||
| [2009/03/30 17:28:56] <Volcane> duncanmv: inside the node statement, define an array | ||
| [2009/03/30 17:28:57] @ Quit: mfoster1: Read error: 54 (Connection reset by peer) | ||
| [2009/03/30 17:28:57] <JD> potentially | ||
| [2009/03/30 17:29:05] @ mfoster joined channel #puppet | ||
| [2009/03/30 17:29:38] @ shake-n-bake_ joined channel #puppet | ||
| [2009/03/30 17:29:40] <duncanmv> Volcane: sure, but that only works if I generate the complete sites.pp with each node | ||
| [2009/03/30 17:30:01] @ nakano is now known as nakano_ | ||
| [2009/03/30 17:30:16] <JD> duncanmv: I suppose at this point we should ask what exactly you're trying to do | ||
| [2009/03/30 17:30:29] <Volcane> duncanmv: nope, node statements just generally have a list of includes and maybe some variables | ||
| [2009/03/30 17:30:45] <duncanmv> Volcane: but for external_nodes, I can only specify classes, therefore I still need to generate classes | ||
| [2009/03/30 17:30:54] <Volcane> duncanmv: you can also specify variables | ||
| [2009/03/30 17:31:07] <duncanmv> Volcane: I want to have the list of packages per host, individually. Not per class. Therefore I generate it from another system. | ||
| [2009/03/30 17:31:26] <andrewcshafer> duncanmv: Is there some reason not to break things up semantically? What problem are you actually trying to solve? | ||
| [2009/03/30 17:32:06] <Volcane> duncanmv: in a node{} file you can just go $packages = ["apache", "bash"] etc | ||
| [2009/03/30 17:32:13] <duncanmv> andrewcshafer: I have a db which maps hosts to a list of packages that need to be installed or ensured recent. I want to generate puppetmasterd config from that. | ||
| [2009/03/30 17:32:15] <Volcane> duncanmv: in external nodes, u can do similar with parameters afaik | ||
| [2009/03/30 17:32:52] <duncanmv> Volcane: uhm, AFAIK external_nodes YAML only allow to specify classes | ||
| [2009/03/30 17:32:56] <andrewcshafer> you can add them as variables using external nodes as other people mentioned | ||
| [2009/03/30 17:33:10] <duncanmv> andrewcshafer: ahh now I see some light | ||
| [2009/03/30 17:33:18] <Volcane> duncanmv: http://reductivelabs.com/trac/puppet/wiki/ExternalNodes | ||
| [2009/03/30 17:33:33] <duncanmv> Volcane: you mean in the parameters section? | ||
| [2009/03/30 17:34:16] <Volcane> those are varaibles yes, i think you can do arrays too | ||
| [2009/03/30 17:34:33] <andrewcshafer> yes, then just reference that parameter/variable in the manifest | ||
| [2009/03/30 17:34:38] <Volcane> duncanmv: otherwise u can write tiny fnctions that extend the puppet language: http://reductivelabs.com/trac/puppet/wiki/WritingYourOwnFunctions | ||
| [2009/03/30 17:34:50] <andrewcshafer> duncanmv: do you already have this database? how many nodes do you have? | ||
| [2009/03/30 17:35:09] <andrewcshafer> you can kind of think of building up classes as normalizing your database | ||
| [2009/03/30 17:36:34] @ Quit: shake-n-bake: Read error: 110 (Connection timed out) | ||
| [2009/03/30 17:37:23] <duncanmv> andrewcshafer: the database can have thousands of nodes, and every node has a list of packages (most of them are empty) | ||
| [2009/03/30 17:38:04] <Volcane> duncanmv: sounds like you'll need a external node script anyway | ||
| [2009/03/30 17:39:12] <andrewcshafer> duncanmv: do you already have this database, or you are building it? Long term you would be happier with the semantics of classes IMHO. | ||
| [2009/03/30 17:39:38] @ joe-mac joined channel #puppet | ||
| [2009/03/30 17:39:51] <Volcane> andrewcshafer: erm, you'd rather hand write thousands of classes than have a nice UI to manage the packages that goes on these many may machines? | ||
| [2009/03/30 17:39:52] <joe-mac> hey fsweetser, that looks freaking phenomenal | ||
| [2009/03/30 17:40:05] <joe-mac> thanks for that def i'm going to be using that in other places too | ||
| [2009/03/30 17:40:16] <joe-mac> see ya | ||
| [2009/03/30 17:40:20] @ joe-mac left channel #puppet () | ||
| [2009/03/30 17:40:33] <andrewcshafer> Volcane: you won't have thousands of classes, unless every machine is different | ||
| [2009/03/30 17:40:38] @ yarihm joined channel #puppet | ||
| [2009/03/30 17:40:45] <andrewcshafer> and you still have to configure packages | ||
| [2009/03/30 17:41:07] <plathrop> well, joe-mac's idea didn't help | ||
| [2009/03/30 17:41:10] <plathrop> anyone else? | ||
| [2009/03/30 17:41:12] <plathrop> I'm at a los | ||
| [2009/03/30 17:41:13] <plathrop> loss | ||
| [2009/03/30 17:41:18] <Volcane> andrewcshafer: easily handled with defines :) | ||
| [2009/03/30 17:41:51] <andrewcshafer> Volcane: Which are in? your thousand classes? | ||
| [2009/03/30 17:41:54] @ hessml|away|away is now known as hessml|away | ||
| [2009/03/30 17:42:05] <andrewcshafer> plathrop: what are you trying to solve? | ||
| [2009/03/30 17:42:42] <Volcane> andrewcshafer: you pass the array to a define, which creates the resources, unique array per host, or the same one 100s of times, doesnt make a difference | ||
| [2009/03/30 17:43:02] <andrewcshafer> it's easy to install the packages | ||
| [2009/03/30 17:43:10] <plathrop> andrewcshafer: I want to install lighttpd (with puppet) on a machine that already has apache running. But due to Debian policies, if a daemon can't start at install, that's a failed install | ||
| [2009/03/30 17:43:11] <andrewcshafer> that's not a problem | ||
| [2009/03/30 17:43:19] <plathrop> andrewcshafer: but I want to run lighttpd on a different port | ||
| [2009/03/30 17:43:27] <plathrop> andrewcshafer: so, I'm trying to figure out a workaround | ||
| [2009/03/30 17:43:50] <Volcane> plathrop: yeah that pisses me off too, debian is retarded on so many levels | ||
| [2009/03/30 17:44:15] <plathrop> Volcane: I usually at least *understand* the policy, but this one is completely retarted and contrary | ||
| [2009/03/30 17:44:18] <Volcane> plathrop: exactly the same, dev machines need both apache and lighty | ||
| [2009/03/30 17:44:42] <plathrop> Volcane: how do you get around it? Or do you not use Debian? | ||
| [2009/03/30 17:44:58] <andrewcshafer> Volcane: but what about the packages that require configuration? are you going to make configuration packages and just manage everything purely with packages? Cause that is it's own style of fun. | ||
| [2009/03/30 17:45:25] <Volcane> plathrop: oh i get it, i mean all the stupid hacks and "fixes" hey make are there to just make things work out the box, which doesnt work most of the time :P | ||
| [2009/03/30 17:45:48] <andrewcshafer> just remembered I have an appt in 15 minutes... bbiab | ||
| [2009/03/30 17:46:33] <Volcane> andrewcshafer: i dont know, the guy said he needs to get a list of packages and install them, he specifically didnt want classes or per package configs. i suspect all he wants is to install packages, its not like things conjure config out of thin air, and if he didnt want to build configs for 100s of packages then Id say he'd be happy with just installing a list of packages | ||
| [2009/03/30 17:46:51] <Volcane> andrewcshafer: we can only make suggestions on how much data ppl give, with the data he give, thats the answer | ||
| [2009/03/30 17:47:34] <Volcane> plathrop: its a pain, i dont really get around it, we only have 1 or 2 devs who need both | ||
| [2009/03/30 17:47:45] <plathrop> Volcane: damn | ||
| [2009/03/30 17:47:48] <plathrop> *headdesk* | ||
| [2009/03/30 17:47:48] <Volcane> plathrop: so its not been worth it enough yet for me to make the effort to fix/work around | ||
| [2009/03/30 17:47:57] <plathrop> Yeah, this is for 1 server here, too | ||
| [2009/03/30 17:48:00] <Volcane> plathrop: is there no preseed trick you can make to disable the autostart? | ||
| [2009/03/30 17:48:06] <plathrop> I think I'm just going to separate it out. | ||
| [2009/03/30 17:48:21] <ch> plathrop: you may have luck by delivering configfiles and using the apt provider instead of the apitute provider | ||
| [2009/03/30 17:48:27] <ch> can't test your case now, though. | ||
| [2009/03/30 17:48:30] <Volcane> plathrop: else put the config down first then install it, hopefully with some luck debian wont overwrite the config? | ||
| [2009/03/30 17:48:48] <plathrop> ch: Hrm. I'll try that. | ||
| [2009/03/30 17:48:52] <duncanmv> Volcane: I think the idea of having a parameter in external_nodes is the best way to go, I define a class that only ensures packages installed and then each node changes this variable | ||
| [2009/03/30 17:48:57] <plathrop> Volcane: good idea, I'll try that too | ||
| [2009/03/30 17:48:58] <plathrop> thanks | ||
| [2009/03/30 17:48:58] <Volcane> plathrop: alternatively change both apache and lighty to different ports | ||
| [2009/03/30 17:49:19] <Volcane> duncanmv: and you wont care if some of these need configs and such? do you really just want to install packages? | ||
| [2009/03/30 17:49:32] <Volcane> plathrop: cos if u change *both* then that'll fix it | ||
| [2009/03/30 17:49:34] <kjetilho> can I iterate over an array in a define? | ||
| [2009/03/30 17:49:50] <Volcane> plathrop: puppet installs one, recofigs it, restarts it, installs the other, reconfigs it, restarts it | ||
| [2009/03/30 17:49:51] <kjetilho> e.g. foo { "bar": groups => ["zot", "zoo"] } | ||
| [2009/03/30 17:49:56] <Volcane> plathrop: tada, all happy | ||
| [2009/03/30 17:50:41] <kjetilho> hmm. group { $groups: } might work | ||
| [2009/03/30 17:50:56] @ ezmob joined channel #puppet | ||
| [2009/03/30 17:50:58] <plathrop> Volcane: Yeah, I can't do that in this case for complicated reasons. | ||
| [2009/03/30 17:51:13] @ benblack is now known as benblack|away | ||
| [2009/03/30 17:51:45] <Volcane> plathrop: lovely | ||
| [2009/03/30 17:51:57] <Volcane> plathrop: check the preseed thing, or the other package provider :( | ||
| [2009/03/30 17:52:14] <Volcane> plathrop: else it might even be less painful to make your own package :P | ||
| [2009/03/30 17:52:43] <plathrop> Volcane: the preseed thing was totally unhelpful because there's nothing that says anything about lighttpd in the output | ||
| [2009/03/30 17:52:50] <plathrop> Volcane: unless I'm running it wrong | ||
| [2009/03/30 17:53:19] <plathrop> Volcane: was going to roll my own package, but I can't figure out what about the package tells it to start | ||
| [2009/03/30 17:53:36] <Volcane> yeah debs are as retarded as the rest of debian | ||
| [2009/03/30 17:53:59] <Volcane> making debs is like some special form of punishment you reserve for your most hated enemies | ||
| [2009/03/30 17:54:26] <zirpu> but once you learn to make your own debs, you feel "special". i know i do. :-) | ||
| [2009/03/30 17:55:02] <Volcane> I'd rather forget :P | ||
| [2009/03/30 17:55:09] <ch> if you don't find a single good thing in debian, you might better stop using it. | ||
| [2009/03/30 17:56:08] <Volcane> not my choice unfort, one big client has 1000s of the stuff, big++ client, cant tell them to piss off :P | ||
| [2009/03/30 17:56:48] <Volcane> but i know i wont ever work for another debian shop no matter who/what | ||
| [2009/03/30 17:57:52] <duncanmv> Volcane: if later we need to configure other stuff that is common, I can always include more classes on what external_nodes return | ||
| [2009/03/30 17:58:28] <Volcane> duncanmv: nods, its easy with an array and a define then | ||
| [2009/03/30 17:58:29] @ kibahop joined channel #puppet | ||
| [2009/03/30 17:58:38] @ kibahop left channel #puppet () | ||
| [2009/03/30 17:59:01] @ lak joined channel #puppet | ||
| [2009/03/30 17:59:21] <plathrop> Oh, how nice. | ||
| [2009/03/30 17:59:45] <plathrop> Ask the debian folks and get flamed "What's the point of installing 2 webservers? That's stupid!" | ||
| [2009/03/30 18:00:13] <Volcane> heh | ||
| [2009/03/30 18:00:46] @ Quit: shake-n-bake_: Read error: 110 (Connection timed out) | ||
| [2009/03/30 18:00:57] <plathrop> honestly, I *don't* need them both, but that doesn't matter. | ||
| [2009/03/30 18:01:04] <plathrop> I should be able to install them both | ||
| [2009/03/30 18:01:15] <plathrop> Ah well, I'm going to fix this another way | ||
| [2009/03/30 18:02:11] <Volcane> yeah cant see much about preseeeds for lighty on mine either | ||
| [2009/03/30 18:03:42] @ kibahop joined channel #puppet | ||
| [2009/03/30 18:03:47] @ kibahop left channel #puppet () | ||
| [2009/03/30 18:05:07] @ Quit: mvn071: "Leaving" | ||
| [2009/03/30 18:05:56] <huggie> plathrop: please do file a bug if you really can't install two webservers in Debian. The maintainers should care about stuff like that | ||
| [2009/03/30 18:06:35] <Volcane> the bug is not being able to stop unconfigured daemons from starting up | ||
| [2009/03/30 18:06:54] <plathrop> huggie: There is already a bug. It's tagged 'wontfix' | ||
| [2009/03/30 18:06:55] @ Quit: stijnbe: "Leaving..." | ||
| [2009/03/30 18:07:50] @ Quit: lak: | ||
| [2009/03/30 18:07:58] <huggie> plathrop: heh what number? I'll look when I'm not ona bus | ||
| [2009/03/30 18:08:11] @ lak joined channel #puppet | ||
| [2009/03/30 18:08:39] <plathrop> huggie: 451242 | ||
| [2009/03/30 18:09:03] <huggie> Volcane: I'd have thought the config on disk first would fix that tho | ||
| [2009/03/30 18:09:36] <Volcane> nods, its worth a try | ||
| [2009/03/30 18:09:46] @ shake-n-bake__ joined channel #puppet | ||
| [2009/03/30 18:09:49] <Volcane> depends how the deb is built | ||
| [2009/03/30 18:14:32] <jrojas> dis the sql for stored configs anywhere in the source tree? | ||
| [2009/03/30 18:14:35] <jrojas> -d | ||
| [2009/03/30 18:15:32] @ Quit: jsm: "Leaving." | ||
| [2009/03/30 18:15:44] <huggie> Volcane: It's an RC bug if the config files aren't conffiles though. | ||
| [2009/03/30 18:17:56] @ benblack|away is now known as benblack | ||
| [2009/03/30 18:23:42] <duncanmv> Volcane: thanks for your help! | ||
| [2009/03/30 18:23:51] @ Quit: duncanmv: Remote closed the connection | ||
| [2009/03/30 18:28:35] @ Quit: shake-n-bake__: Read error: 110 (Connection timed out) | ||
| [2009/03/30 18:29:36] @ ezmob is now known as ezmob|away | ||
| [2009/03/30 18:30:52] @ Quit: eythian: Read error: 110 (Connection timed out) | ||
| [2009/03/30 18:33:39] @ Quit: omry: Remote closed the connection | ||
| [2009/03/30 18:36:57] @ Quit: oboxodo: Remote closed the connection | ||
| [2009/03/30 18:42:03] @ Quit: Demosthenes: "leaving" | ||
| [2009/03/30 18:43:33] @ Quit: ezmob|away: "Leaving..." | ||
| [2009/03/30 18:46:36] @ ezmob joined channel #puppet | ||
| [2009/03/30 18:47:39] @ nakano_ is now known as nakano | ||
| [2009/03/30 18:55:25] @ hessml|away is now known as hessml|away|away | ||
| [2009/03/30 18:59:33] @ Quit: mfoster: "Leaving." | ||
| [2009/03/30 19:07:36] @ specsmu joined channel #puppet | ||
| [2009/03/30 19:07:44] <specsmu> hey... is there a changelog for 0.24.8? | ||
| [2009/03/30 19:07:52] <specsmu> it only goes as far as 0.24.7 on the website | ||
| [2009/03/30 19:10:47] @ Quit: madrescher: "Leaving." | ||
| [2009/03/30 19:13:10] <Volcane> http://github.com/jamtur01/puppet/blob/0.24.x/CHANGELOG | ||
| [2009/03/30 19:26:01] @ Quit: keithlard: | ||
| [2009/03/30 19:28:16] @ keithlard joined channel #puppet | ||
| [2009/03/30 19:28:35] @ clayning is now known as claymation | ||
| [2009/03/30 19:39:27] @ Quit: themurph: Read error: 110 (Connection timed out) | ||
| [2009/03/30 19:41:23] @ Quit: yarihm: "This computer has gone to sleep" | ||
| [2009/03/30 19:43:20] @ specsmu left channel #puppet () | ||
| [2009/03/30 19:43:40] @ Quit: cwebber: | ||
| [2009/03/30 19:48:06] @ nakano is now known as nakano_ | ||
| [2009/03/30 19:58:15] @ shake-n-bake_ joined channel #puppet | ||
| [2009/03/30 20:03:32] @ Quit: ezmob: "Leaving..." | ||
| [2009/03/30 20:03:34] @ Quit: Innocenti: Read error: 110 (Connection timed out) | ||
| [2009/03/30 20:04:58] @ ezmob joined channel #puppet | ||
| [2009/03/30 20:05:53] @ Quit: alfism: "Connection reset by beer" | ||
| [2009/03/30 20:10:03] @ Quit: hessml|away|away: "Leaving..." | ||
| [2009/03/30 20:18:54] @ Quit: aZaFred: | ||
| [2009/03/30 20:19:38] @ shake-n-bake joined channel #puppet | ||
| [2009/03/30 20:20:57] @ Quit: shake-n-bake_: Read error: 110 (Connection timed out) | ||
| [2009/03/30 20:26:08] @ Djelibeybi joined channel #puppet | ||
| [2009/03/30 20:29:51] @ Quit: Djelibeybi: Client Quit | ||
| [2009/03/30 20:30:44] @ Quit: wakko666: Remote closed the connection | ||
| [2009/03/30 20:34:02] @ wakko666 joined channel #puppet | ||
| [2009/03/30 20:35:40] @ Quit: lak: | ||
| [2009/03/30 20:37:47] @ Quit: Tuf8: | ||
| [2009/03/30 20:38:05] @ Djelibeybi joined channel #puppet | ||
| [2009/03/30 20:39:46] @ Quit: nevele: "Xaric: If you have a better quit message then submit a patch!" | ||
| [2009/03/30 20:42:09] @ Quit: shake-n-bake: Read error: 60 (Operation timed out) | ||
| [2009/03/30 20:45:33] @ Quit: kali67: | ||
| [2009/03/30 20:48:15] @ shake-n-bake joined channel #puppet | ||
| [2009/03/30 20:53:32] @ Quit: ezmob: "Bye!" | ||
| [2009/03/30 20:53:33] @ Quit: keithlard: | ||
| [2009/03/30 20:59:07] @ shake-n-bake_ joined channel #puppet | ||
| [2009/03/30 21:04:27] @ Quit: shake-n-bake: Read error: 110 (Connection timed out) | ||
| [2009/03/30 21:10:26] @ shake-n-bake joined channel #puppet | ||
| [2009/03/30 21:18:14] @ Quit: benblack: "Leaving..." | ||
| [2009/03/30 21:19:37] @ xmslack joined channel #puppet | ||
| [2009/03/30 21:21:06] @ claymation_ joined channel #puppet | ||
| [2009/03/30 21:24:35] @ Quit: shake-n-bake_: Read error: 110 (Connection timed out) | ||
| [2009/03/30 21:26:03] @ Quit: WALoeIII: Read error: 60 (Operation timed out) | ||
| [2009/03/30 21:28:00] @ aZaFred joined channel #puppet | ||
| [2009/03/30 21:31:34] @ Quit: shake-n-bake: Read error: 110 (Connection timed out) | ||
| [2009/03/30 21:39:33] @ Quit: punkcut: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 21:41:30] @ Quit: claymation: Read error: 110 (Connection timed out) | ||
| [2009/03/30 21:41:48] @ shake-n-bake joined channel #puppet | ||
| [2009/03/30 21:45:03] @ punkcut joined channel #puppet | ||
| [2009/03/30 21:46:34] @ WALoeIII joined channel #puppet | ||
| [2009/03/30 21:48:10] @ Quit: andrewcshafer: | ||
| [2009/03/30 21:49:53] @ Quit: WALoeIII: Client Quit | ||
| [2009/03/30 21:52:25] @ shake-n-bake_ joined channel #puppet | ||
| [2009/03/30 21:52:33] @ alfism joined channel #puppet | ||
| [2009/03/30 21:54:06] @ lak joined channel #puppet | ||
| [2009/03/30 21:58:21] @ Quit: shake-n-bake: Read error: 110 (Connection timed out) | ||
| [2009/03/30 22:28:20] @ Quit: shake-n-bake_: | ||
| [2009/03/30 22:33:44] @ Quit: punkcut: Read error: 104 (Connection reset by peer) | ||
| [2009/03/30 22:39:14] @ punkcut joined channel #puppet | ||
| [2009/03/30 23:00:42] @ joe-mac joined channel #puppet | ||
| [2009/03/30 23:11:21] @ maxagaz joined channel #puppet | ||
| [2009/03/30 23:11:32] @ joe-mac left channel #puppet () | ||
| [2009/03/30 23:29:39] @ Quit: d3vilb0x: | ||
| [2009/03/30 23:32:00] @ eythian joined channel #puppet | ||
| [2009/03/30 23:36:13] @ Quit: ethan_rowe: "Lack of interest wins out." | ||
| [2009/03/30 23:43:20] <eythian> I'm having trouble having custom facts work. On the client if I do 'RUBYLIB=/var/lib/puppet/lib/ facter' it works. If I do 'facter -p' it doesn't, but I think it should. What could be going on? | ||
| [2009/03/30 23:46:52] @ benblack joined channel #puppet |
Generated by irclog2html.py 2.6 by Marius Gedminas - find it at mg.pov.lt!