| [2008/07/14 00:00:08] @ Quit: a-priori: | ||
| [2008/07/14 00:06:42] @ inigo_montoya is now known as mmcgrath | ||
| [2008/07/14 00:22:09] @ asa_ joined channel #puppet | ||
| [2008/07/14 00:29:54] @ Quit: asa_: | ||
| [2008/07/14 00:57:39] @ ski98033 joined channel #puppet | ||
| [2008/07/14 00:57:57] @ f3ew_ joined channel #puppet | ||
| [2008/07/14 00:59:59] @ f3ew_ is now known as f3ew | ||
| [2008/07/14 01:20:09] @ Quit: ski98033: "Leaving." | ||
| [2008/07/14 01:22:23] @ Quit: mikearr: Read error: 110 (Connection timed out) | ||
| [2008/07/14 01:26:01] @ mmcgrath is now known as inigo_montoya | ||
| [2008/07/14 01:27:39] @ mikearr joined channel #puppet | ||
| [2008/07/14 01:42:00] @ DavidS joined channel #puppet | ||
| [2008/07/14 01:48:32] @ Quit: \ask_: Remote closed the connection | ||
| [2008/07/14 01:49:05] @ \ask_ joined channel #puppet | ||
| [2008/07/14 01:53:10] @ pleemans joined channel #puppet | ||
| [2008/07/14 02:17:56] @ lak joined channel #puppet | ||
| [2008/07/14 02:20:44] @ Quit: kolla: Remote closed the connection | ||
| [2008/07/14 02:22:17] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 02:25:06] @ mikearr_ joined channel #puppet | ||
| [2008/07/14 02:29:00] @ Quit: mikearr: Read error: 110 (Connection timed out) | ||
| [2008/07/14 02:53:25] @ Quit: andrewcshafer: | ||
| [2008/07/14 03:09:37] @ f--z joined channel #puppet | ||
| [2008/07/14 03:10:21] @ _NiC_ is now known as _NiC | ||
| [2008/07/14 03:10:22] @ Quit: lak: | ||
| [2008/07/14 03:22:15] @ Quit: Maliuta: Remote closed the connection | ||
| [2008/07/14 03:31:28] @ Maliuta joined channel #puppet | ||
| [2008/07/14 03:33:28] @ kolla joined channel #puppet | ||
| [2008/07/14 03:41:38] @ Quit: thegcat_: | ||
| [2008/07/14 03:42:00] @ MaliutaLap joined channel #puppet | ||
| [2008/07/14 03:52:09] @ aymerick joined channel #puppet | ||
| [2008/07/14 03:56:05] @ Quit: DavidS: "Leaving." | ||
| [2008/07/14 03:56:24] @ Quit: aymerick: Client Quit | ||
| [2008/07/14 04:33:58] @ Quit: f--z: "KVIrc 3.2.5 Anomalies http://www.kvirc.net/" | ||
| [2008/07/14 05:16:51] @ roald joined channel #puppet | ||
| [2008/07/14 05:25:05] @ aymerick joined channel #puppet | ||
| [2008/07/14 05:36:56] @ meandtheshell joined channel #puppet | ||
| [2008/07/14 05:56:08] <silent-ion> Hi. | ||
| [2008/07/14 05:56:09] <silent-ion> Can I order puppet, that one module should configure before others? | ||
| [2008/07/14 05:58:54] <tim|imac> only by requiring the parts of the other module | ||
| [2008/07/14 06:00:53] @ DerekW joined channel #puppet | ||
| [2008/07/14 06:16:34] @ Joined channel #puppet | ||
| [2008/07/14 06:16:34] @ Topic is "Welcome to Puppet on Redmine (http://tinyurl.com/48ek6g) | Please see http://snurl.com/1udr1 for channel guidelines | See http://snurl.com/1udr3 and http://snurl.com/2901u about two SSL issues" | ||
| [2008/07/14 06:16:34] @ Topic set by jamesturnbull on Tue Jun 03 01:57:42 -0400 2008 | ||
| [2008/07/14 06:22:45] @ snerdly joined channel #puppet | ||
| [2008/07/14 06:34:49] @ Quit: \ask_: | ||
| [2008/07/14 06:37:53] @ sparanjape joined channel #puppet | ||
| [2008/07/14 07:01:03] <sparanjape> Starting puppetmaster: Could not match ' | ||
| [2008/07/14 07:01:03] <sparanjape> ' at /etc/puppet/manifests/common/nodes/skenzo_base.pp:1 | ||
| [2008/07/14 07:01:09] <sparanjape> any idea why this error | ||
| [2008/07/14 07:01:46] <msf> maybe you could paste line1 of your manifest ? | ||
| [2008/07/14 07:02:08] <sparanjape> node skenzo_base { | ||
| [2008/07/14 07:03:18] @ tazz joined channel #puppet | ||
| [2008/07/14 07:04:04] <fujin> put it in quotes | ||
| [2008/07/14 07:04:11] <fujin> _ will throw the lexer | ||
| [2008/07/14 07:04:37] @ Quit: DerekW: Remote closed the connection | ||
| [2008/07/14 07:05:02] <msf> _ isn't a valid character in a hostname anyways | ||
| [2008/07/14 07:07:25] <sparanjape> that definately isnt the reason | ||
| [2008/07/14 07:07:37] <sparanjape> i have a similar manifest with an _ which is working fine | ||
| [2008/07/14 07:07:50] <sparanjape> i renamed the problematic 1 & remove3d the _ | ||
| [2008/07/14 07:07:55] <sparanjape> the error still persists | ||
| [2008/07/14 07:08:22] <jamesturnbull> sparanjape: can you pastie the whole thing? | ||
| [2008/07/14 07:09:37] <f3ew> msf, that isn't a hostname, it's a generic base type | ||
| [2008/07/14 07:09:43] <f3ew> gepetto pastie | ||
| [2008/07/14 07:09:50] <sparanjape> http://pastie.org/232927 | ||
| [2008/07/14 07:10:14] <f3ew> sparanjape what's in the file being included before that? | ||
| [2008/07/14 07:11:20] <sparanjape> f3ew those are my classes.. i'm configuring a node which would include all those classes.. a host would then inherit that node | ||
| [2008/07/14 07:49:29] @ M- joined channel #puppet | ||
| [2008/07/14 08:18:01] <gepetto> ::puppet:: Recipes/Hostgroup Fact edited by mwr @ http://reductivelabs.com/trac/puppet/wiki/Recipes/HostgroupFact (by renfro@tntech.edu) | ||
| [2008/07/14 08:18:58] <Volcane> hmm, with facter 1.5.0 is facter -p supposed to show facts that were distribued via puppet? | ||
| [2008/07/14 08:19:05] <Volcane> seems so, but not having joy from it | ||
| [2008/07/14 08:22:15] @ a-priori joined channel #puppet | ||
| [2008/07/14 08:25:46] <silent-ion> gepetto pastie | ||
| [2008/07/14 08:35:23] @ Quit: a-priori: | ||
| [2008/07/14 08:37:56] <lazzurs> Hello, if I define a file in a class however in "sub-class" (class that inherits from that class) I don't want to touch that file what do I change in the File["/path/to/file"] { } line? | ||
| [2008/07/14 08:42:45] <lazzurs> it would seem that ignore is what I am looking for | ||
| [2008/07/14 08:43:26] @ riddley joined channel #puppet | ||
| [2008/07/14 08:44:02] <lazzurs> this is what I am thinking of http://pastebin.com/d11219c49 | ||
| [2008/07/14 08:50:51] @ kambiz joined channel #puppet | ||
| [2008/07/14 08:51:33] @ brothers joined channel #puppet | ||
| [2008/07/14 08:51:48] @ brothers left channel #puppet () | ||
| [2008/07/14 08:53:48] <jamesturnbull> silent-ion: gepetto is one bot and pastie is another | ||
| [2008/07/14 08:53:49] <lazzurs> I have replaced the ignore with ensure => present and that appears to work | ||
| [2008/07/14 08:54:20] <jamesturnbull> Volcane: try facter --puppet | ||
| [2008/07/14 08:55:29] @ gileswork joined channel #puppet | ||
| [2008/07/14 08:55:38] <silent-ion> ok.. Can anybody help me with my module puppetd: | ||
| [2008/07/14 08:55:40] <silent-ion> http://pastie.org/232959 | ||
| [2008/07/14 08:55:41] <gileswork> elo puppetiers | ||
| [2008/07/14 08:55:41] <silent-ion> Maybe it's a bug??? | ||
| [2008/07/14 08:55:55] <gileswork> we have a custom fact for the default gateway | ||
| [2008/07/14 08:56:29] <gileswork> i'm wanting to create one to get a cidr address for the primary network | ||
| [2008/07/14 08:56:40] <gileswork> that a server is on | ||
| [2008/07/14 08:56:50] <gileswork> anyone got ideas about how best to get this? | ||
| [2008/07/14 08:58:10] <jamesturnbull> gileswork: platform? | ||
| [2008/07/14 08:58:20] <gileswork> linux | ||
| [2008/07/14 08:58:22] <gileswork> rhel and deb | ||
| [2008/07/14 08:58:36] <gileswork> i think I use netstat for the dft gw | ||
| [2008/07/14 08:58:56] <jamesturnbull> gileswork: I think there is already a fact for default gateway in the Recipes namespace | ||
| [2008/07/14 08:59:09] <gileswork> I just want to make sure that I get the right cidr range | ||
| [2008/07/14 08:59:22] <gileswork> yea well I've done that | ||
| [2008/07/14 08:59:46] <jamesturnbull> gileswork: http://jodies.de/ipcalc | ||
| [2008/07/14 08:59:52] <jamesturnbull> gileswork: parse the output of that | ||
| [2008/07/14 08:59:54] <Volcane> jamesturnbull: no diff | ||
| [2008/07/14 09:00:52] <kjetilho> gileswork: the netmask is pretty clearly available in netstat -rn output? | ||
| [2008/07/14 09:01:43] <gileswork> k thats what I thought I'd use | ||
| [2008/07/14 09:01:46] <jamesturnbull> gileswork: or there are some ruby snippets here - http://snippets.dzone.com/tag/ip/2 | ||
| [2008/07/14 09:02:01] <gileswork> presumably there's a ruby lib to do cidr conversion | ||
| [2008/07/14 09:02:40] <jamesturnbull> gileswork: yeah on that link I just sent you | ||
| [2008/07/14 09:03:39] <jamesturnbull> Volcane: it was working for me | ||
| [2008/07/14 09:04:03] <gileswork> cool I shall give it a whirl | ||
| [2008/07/14 09:04:51] @ Quit: M-: "Leaving" | ||
| [2008/07/14 09:05:02] <kjetilho> 32 - log2((1<<32) - netmask) :-) | ||
| [2008/07/14 09:05:54] <gileswork> that looks nicer heh | ||
| [2008/07/14 09:06:10] <gileswork> I just need a way if finding out the primary nic | ||
| [2008/07/14 09:06:14] <gileswork> of rather | ||
| [2008/07/14 09:07:29] <kjetilho> there isn't an objective definition of that, but I think most people will consider the interface with the default route | ||
| [2008/07/14 09:08:34] <gileswork> ya that was what I was going to try | ||
| [2008/07/14 09:13:51] <gileswork> actually on a side note we struggle a bit with testing our custom facts | ||
| [2008/07/14 09:14:16] @ pjw joined channel #puppet | ||
| [2008/07/14 09:14:21] <gileswork> require 'facter' gives you access to the built in facts | ||
| [2008/07/14 09:14:28] <gileswork> but not the custom ones | ||
| [2008/07/14 09:18:56] @ Quit: tmz: Read error: 110 (Connection timed out) | ||
| [2008/07/14 09:19:35] @ tmz joined channel #puppet | ||
| [2008/07/14 09:22:34] <jamesturnbull> gileswork: facter 1.5 has this capability - though Volcane seems to be having issues with it | ||
| [2008/07/14 09:23:16] <gileswork> k I'm making my facts self contained for the mo | ||
| [2008/07/14 09:23:33] <gileswork> might use ip to get the cidr of the primary nic | ||
| [2008/07/14 09:25:20] @ ski98033 joined channel #puppet | ||
| [2008/07/14 09:39:34] @ tmz_ joined channel #puppet | ||
| [2008/07/14 09:42:47] @ Quit: tmz: Nick collision from services. | ||
| [2008/07/14 09:43:02] @ tmz_ is now known as tmz | ||
| [2008/07/14 09:46:10] @ MNK2 joined channel #puppet | ||
| [2008/07/14 09:49:15] <MNK2> Hi all, is there any policy regarding being able to mirror the puppet website and documentation? | ||
| [2008/07/14 09:52:51] <duritong> where is the official link to get facter 1.5 ? | ||
| [2008/07/14 10:02:25] @ Quit: pjw: | ||
| [2008/07/14 10:03:29] @ shenson_not_here is now known as shenson | ||
| [2008/07/14 10:07:09] <thom> duritong: http://reductivelabs.com/downloads/facter/facter-1.5.0.tgz | ||
| [2008/07/14 10:07:30] <duritong> thom: thanks | ||
| [2008/07/14 10:07:35] <duritong> and how did you find that? | ||
| [2008/07/14 10:08:02] <jamesturnbull> MNK2: why do you ask? | ||
| [2008/07/14 10:09:06] @ a-priori joined channel #puppet | ||
| [2008/07/14 10:09:42] <jamesturnbull> duritong: http://reductivelabs.com/trac/facter | ||
| [2008/07/14 10:09:52] @ efaistos_ joined channel #puppet | ||
| [2008/07/14 10:09:55] <efaistos_> hi | ||
| [2008/07/14 10:09:59] <jamesturnbull> duritong: I believe andrewshafer is slowly updating the website | ||
| [2008/07/14 10:10:32] <duritong> jamesturnbull: I didn't see -latest as 1.5 and then didn't think about to browse the directory | ||
| [2008/07/14 10:10:51] <jamesturnbull> duritong: it's always the latest I think - let me check | ||
| [2008/07/14 10:11:12] <jamesturnbull> duritong: yep it is | ||
| [2008/07/14 10:12:53] <duritong> hmm Timed out seeking value for lsbdistcodename | ||
| [2008/07/14 10:14:01] <duritong> hmm maybe the host is just a bit overloaded | ||
| [2008/07/14 10:14:10] <jamesturnbull> duritong: hmmm there was a bug | ||
| [2008/07/14 10:14:10] <duritong> jamesturnbull: thanks for checking | ||
| [2008/07/14 10:14:21] <jamesturnbull> duritong: are you sure you've deleted all old files? | ||
| [2008/07/14 10:14:47] <duritong> jamesturnbull: I just made a new rpm | ||
| [2008/07/14 10:15:46] <duritong> jamesturnbull: it looks like the timeout is too short if the machine is overloaded (which currently means a load > 2) | ||
| [2008/07/14 10:15:57] <jamesturnbull> duritong: crap - I knew hat'd backfire on us | ||
| [2008/07/14 10:16:03] <jamesturnbull> duritong: you can adjust the timeout | ||
| [2008/07/14 10:16:08] <duritong> where? | ||
| [2008/07/14 10:16:12] <jamesturnbull> duritong: in the code | ||
| [2008/07/14 10:16:17] <jamesturnbull> duritong: but log a ticket please | ||
| [2008/07/14 10:16:21] <duritong> ok | ||
| [2008/07/14 10:16:53] <duritong> the funny thing is that it still shows up, also when the error is shown | ||
| [2008/07/14 10:22:12] <MNK2> O | ||
| [2008/07/14 10:22:14] <MNK2> oops | ||
| [2008/07/14 10:23:16] <MNK2> I am located in japan and hitting the main website is quite slow for me :], it'd be nice if I could have a cached copy to reference. However, if this is frowned on, I'll just keep using the main website as is. | ||
| [2008/07/14 10:24:40] <jamesturnbull> MNK2: email Luke | ||
| [2008/07/14 10:24:50] <MNK2> Thanks, will do. | ||
| [2008/07/14 10:25:14] <jamesturnbull> MNK2: I don't think he'll have an issue with the wiki but the other content is ReductiveLabs | ||
| [2008/07/14 10:25:48] <MNK2> I mainly just want the wiki content regarding Puppet stuff. The company stuff he can keep | ||
| [2008/07/14 10:27:31] <kjetilho> MNK2: as long as you don't publish your copy for the rest of the world (or for money), he really has no say | ||
| [2008/07/14 10:28:17] <MNK2> Well, I consider this more of being polite and respecting wishes than anything else. | ||
| [2008/07/14 10:28:50] <jamesturnbull> kjetilho: it's always good manners to ask | ||
| [2008/07/14 10:29:56] @ Quit: pleemans: "Ex-Chat" | ||
| [2008/07/14 10:31:09] <jamesturnbull> night all | ||
| [2008/07/14 10:31:17] <kjetilho> sure, honour robots.txt | ||
| [2008/07/14 10:33:04] <duritong> jamesturnbull: #1422 | ||
| [2008/07/14 10:33:05] <gepetto> duritong: jamesturnbull: #1422 is http://reductivelabs.com/redmine/issues/show/1422 | ||
| [2008/07/14 10:34:43] @ Quit: kolla: Remote closed the connection | ||
| [2008/07/14 10:49:13] <efaistos_> im doing some investigation about puppet but a colleague said that there is a lot of vulnerabilities with Ruby and wants cfengine which is a pain to configure | ||
| [2008/07/14 10:49:27] <efaistos_> any arguments to oppose to that ? | ||
| [2008/07/14 10:52:49] <kjetilho> it's hard to oppose FUD | ||
| [2008/07/14 10:53:31] <Volcane> efaistos_: upgrade to a fixed ruby? | ||
| [2008/07/14 10:54:04] <Volcane> efaistos_: there are old and already-fixed vulnerabilities in almost every bit of software, ruby too | ||
| [2008/07/14 10:54:51] <efaistos_> Volcane: i know that | ||
| [2008/07/14 10:54:54] <efaistos_> ;) | ||
| [2008/07/14 10:55:15] <Volcane> efaistos_: so its identical | ||
| [2008/07/14 10:55:39] <Volcane> tell him there are many vulnerable programs written in C, so you dont want software writting in that language :P | ||
| [2008/07/14 10:55:47] <aelizondo> efaistos: i'm in exactly the same position. the timing of the ruby vulnerbability was bad. :( | ||
| [2008/07/14 10:56:07] <Volcane> its bad that redhat has not released patched versions yet :( | ||
| [2008/07/14 10:56:20] <MNK2> Hmm that's not good | ||
| [2008/07/14 10:56:33] <MNK2> I thought there were some Redhat guys that were getting into Puppet, too? | ||
| [2008/07/14 10:56:36] @ Quit: aymerick: | ||
| [2008/07/14 10:56:58] <aelizondo> MNK2 yes http://cft.et.redhat.com/ | ||
| [2008/07/14 10:59:06] <duritong> there are a lot of people @ redhat in puppet | ||
| [2008/07/14 11:00:24] <efaistos_> i think they are using puppet for the dep,oyment of their servers ... | ||
| [2008/07/14 11:03:37] <bda> http://www.redhat.com/spacewalk/ | ||
| [2008/07/14 11:03:55] <MNK2> Well, I don't have to do much selling to get Puppet in the door at work :D | ||
| [2008/07/14 11:04:03] <MNK2> More like 'when the hell are you going to have it running?' | ||
| [2008/07/14 11:05:40] <magnachef> where do I assign devices to a node? So if I have a pool of webservers, how do I associate those servers with that node definition? | ||
| [2008/07/14 11:08:41] <magnachef> I've checked through the docs, but seem to be missing where that association is | ||
| [2008/07/14 11:12:13] @ pjw joined channel #puppet | ||
| [2008/07/14 11:14:05] @ martha joined channel #puppet | ||
| [2008/07/14 11:17:32] <Disconnect> magnachef: you either use an external classifier or you do "include webserver" in the node definition | ||
| [2008/07/14 11:18:43] <magnachef> does the include webserver get defined on the client machine (aka, the webserver)? | ||
| [2008/07/14 11:19:47] <magnachef> I've also looked at adding facts to each machine as to what their "role" is, but I'm not sure if that's the right way to go about it | ||
| [2008/07/14 11:20:16] <duritong> anybody knows if I can run several mongrel instances and a webrick instance on the same source, but on different ports? | ||
| [2008/07/14 11:20:18] <Disconnect> external classifiers are good for that, but without those you generally put it into node.pp when you define it. | ||
| [2008/07/14 11:20:24] <duritong> or will I have a locking problem? | ||
| [2008/07/14 11:21:03] <Disconnect> so you'd say (not syntax checked..) "node { web1: include webserver .. include log-client .. etc.. }" | ||
| [2008/07/14 11:21:24] <magnachef> the node.pp on the client? | ||
| [2008/07/14 11:22:13] <Disconnect> no everyting goes on the server | ||
| [2008/07/14 11:22:21] <Disconnect> (did you go through the docs at all?) | ||
| [2008/07/14 11:22:32] <magnachef> yeah I did | ||
| [2008/07/14 11:22:53] <Disconnect> all the client ever gets is the puppet client and a hostname (and a puppet server name, etc from /etc/puppet/puppet.conf) | ||
| [2008/07/14 11:22:57] <Disconnect> everything else comes from the erver | ||
| [2008/07/14 11:23:05] <magnachef> I'm just not understanding the node definitions fully | ||
| [2008/07/14 11:24:06] <Disconnect> they are basically just a magic class that only applies to that node | ||
| [2008/07/14 11:24:38] <magnachef> right, but where I'm confused is how a server knows whether it's a webserver, or a database server, etc | ||
| [2008/07/14 11:24:53] <magnachef> can you point me to a link for the node.pp syntax, etc | ||
| [2008/07/14 11:25:31] @ DerekW joined channel #puppet | ||
| [2008/07/14 11:28:15] <Disconnect> the server knows because you tell it.. http://reductivelabs.com/trac/puppet/wiki/SimplestPuppetInstallRecipe has the node definition | ||
| [2008/07/14 11:28:18] @ kolla joined channel #puppet | ||
| [2008/07/14 11:28:23] @ Quit: meandtheshell: "Leaving." | ||
| [2008/07/14 11:30:10] * Disconnect isn't trying to be flippant - you can 'tell it' with facts or directly, but somewhere along the way you say "this server gets these classes" .. the most basic way is in the node definition | ||
| [2008/07/14 11:30:56] @ Quit: MNK2: | ||
| [2008/07/14 11:31:03] @ muerr joined channel #puppet | ||
| [2008/07/14 11:33:05] <magnachef> yeah, I understand what you're saying, my issue was not as much the theory, but the mechanics and syntax of making the association of node and device | ||
| [2008/07/14 11:34:12] <duritong> magnachef: just go step by step, the longer you will be into it, the better you'll understand and the more you think the puppet way | ||
| [2008/07/14 11:34:36] <duritong> and don't try to do one big thing to start | ||
| [2008/07/14 11:35:39] <magnachef> I'm starting slow | ||
| [2008/07/14 11:36:01] <magnachef> I have a default node currently managing a file on the webservers | ||
| [2008/07/14 11:36:29] <f3ew> magnachef, you basically can do something like this | ||
| [2008/07/14 11:36:56] <f3ew> node 'webserver' { include webserver_classes } | ||
| [2008/07/14 11:37:07] <f3ew> node 'foo.example.com' inherits webserver | ||
| [2008/07/14 11:37:08] <magnachef> but the database servers have a different version of the file, and that's where I'm running into a block as to how I associate these with different node definitions | ||
| [2008/07/14 11:37:16] <f3ew> where foo.example.com is your webserver | ||
| [2008/07/14 11:37:23] <duritong> you have to do this | ||
| [2008/07/14 11:37:38] <f3ew> node 'dbserver' { include dbserver_classes } | ||
| [2008/07/14 11:37:47] <f3ew> node 'db.example.com' inherits dbserver | ||
| [2008/07/14 11:37:59] <f3ew> that should get you started :) | ||
| [2008/07/14 11:38:01] <magnachef> ahhhh, gotcha | ||
| [2008/07/14 11:38:24] <f3ew> Do you have any sort of understanding of OO? | ||
| [2008/07/14 11:38:39] <magnachef> so the node name is the device which just inherits from the node definition | ||
| [2008/07/14 11:38:41] <magnachef> yep | ||
| [2008/07/14 11:39:35] <magnachef> and this is all within the site.pp | ||
| [2008/07/14 11:39:52] <magnachef> or defined in the node.pp file and included into the site.pp | ||
| [2008/07/14 11:39:55] <magnachef> correct? | ||
| [2008/07/14 11:40:39] * magnachef apologizes for his ignorance | ||
| [2008/07/14 11:42:29] <Disconnect> yah | ||
| [2008/07/14 11:43:07] @ f--z joined channel #puppet | ||
| [2008/07/14 11:43:08] @ lutter joined channel #puppet | ||
| [2008/07/14 11:46:10] <f3ew> ignorance is not to be apologised for, but refusal to fix the ignorance is | ||
| [2008/07/14 11:47:15] <magnachef> :-) | ||
| [2008/07/14 11:49:17] @ shake-n-bake_ joined channel #puppet | ||
| [2008/07/14 11:53:31] @ tazz_ joined channel #puppet | ||
| [2008/07/14 11:53:48] @ Quit: tazz: Read error: 104 (Connection reset by peer) | ||
| [2008/07/14 11:55:31] <magnachef> woohoo! success | ||
| [2008/07/14 11:55:58] <magnachef> thanks for all the help | ||
| [2008/07/14 11:56:10] <jfluhmann> magnachef, going to OSCON this year? If so, http://en.oreilly.com/oscon2008/public/schedule/detail/2607 | ||
| [2008/07/14 11:56:21] <magnachef> seems like I need to restart puppetmaster after each config change though? | ||
| [2008/07/14 11:56:24] <jfluhmann> I wish I were going. I'd like to see that talk | ||
| [2008/07/14 11:56:28] <magnachef> unfortunately not | ||
| [2008/07/14 11:56:32] <jfluhmann> :-( | ||
| [2008/07/14 11:56:42] <magnachef> hey jeremy! | ||
| [2008/07/14 11:56:52] <jfluhmann> hey Dan! | ||
| [2008/07/14 11:56:54] <f3ew> magnachef just wait for a few minutes | ||
| [2008/07/14 11:57:18] <magnachef> ok, cool...it was just impatience | ||
| [2008/07/14 12:01:55] @ tazz_ is now known as tazz | ||
| [2008/07/14 12:02:19] @ shake-n-bake joined channel #puppet | ||
| [2008/07/14 12:05:25] <magnachef> so is there a best way to handle the nodes in the config when you have, say 20 servers of a group type? | ||
| [2008/07/14 12:05:58] <ezekiel-v3> efaistos_: I just got redhat notifications about fixes for the Ruby vulnerabilities, if that helps your investigation | ||
| [2008/07/14 12:06:12] <f3ew> a custom function and a string in the hostname? | ||
| [2008/07/14 12:06:45] <efaistos_> ezekiel-v3: okay thanks ;) | ||
| [2008/07/14 12:08:56] @ Quit: shake-n-bake: Read error: 60 (Operation timed out) | ||
| [2008/07/14 12:09:02] @ Quit: f--z: Read error: 110 (Connection timed out) | ||
| [2008/07/14 12:09:44] @ Quit: shake-n-bake_: Read error: 110 (Connection timed out) | ||
| [2008/07/14 12:12:14] <Disconnect> magnachef: external classifier or maybe you can swing something with facts and the default node.. | ||
| [2008/07/14 12:14:39] * magnachef reads about 'How to use external nodes' | ||
| [2008/07/14 12:15:19] <magnachef> f3ew: what you're saying is just have the node name be a function that returns a strong of hostnames? | ||
| [2008/07/14 12:15:51] <magnachef> or did I totally miss your point? | ||
| [2008/07/14 12:16:00] <f3ew> magnachef Puppet uses the output of facter in getting node names | ||
| [2008/07/14 12:23:25] @ Quit: authentic: Read error: 104 (Connection reset by peer) | ||
| [2008/07/14 12:24:01] <magnachef> am I natively able to use a regexp to define a node: ex: node '/^web[\d+]\-domain.com/' | ||
| [2008/07/14 12:26:30] @ shake-n-bake_ joined channel #puppet | ||
| [2008/07/14 12:26:33] <refuseresisted> yawn | ||
| [2008/07/14 12:30:32] @ ezralini joined channel #puppet | ||
| [2008/07/14 12:31:47] <Volcane> magnachef: no | ||
| [2008/07/14 12:31:59] <Volcane> magnachef: external node tools for that kind of thing | ||
| [2008/07/14 12:32:20] @ Quit: tazz: Read error: 104 (Connection reset by peer) | ||
| [2008/07/14 12:33:06] @ tazz joined channel #puppet | ||
| [2008/07/14 12:35:11] @ Quit: kambiz: "Leaving" | ||
| [2008/07/14 12:35:21] @ kambiz joined channel #puppet | ||
| [2008/07/14 12:38:09] @ Quit: sigmonsays: Read error: 110 (Connection timed out) | ||
| [2008/07/14 12:38:41] @ sigmonsays joined channel #puppet | ||
| [2008/07/14 12:39:27] @ pleemans joined channel #puppet | ||
| [2008/07/14 12:40:30] @ Quit: roald: Remote closed the connection | ||
| [2008/07/14 12:42:11] @ refuseresisted_ joined channel #puppet | ||
| [2008/07/14 12:42:22] @ Quit: refuseresisted: "Leaving" | ||
| [2008/07/14 12:42:37] @ refuseresisted_ is now known as refuseresisted | ||
| [2008/07/14 12:47:53] <Volcane> centos ruby updates also coming through now | ||
| [2008/07/14 12:50:51] @ Quit: shake-n-bake_: Read error: 110 (Connection timed out) | ||
| [2008/07/14 12:52:56] @ Quit: tazz: Connection timed out | ||
| [2008/07/14 12:55:23] @ lak joined channel #puppet | ||
| [2008/07/14 12:56:42] @ asa_ joined channel #puppet | ||
| [2008/07/14 12:58:53] <asa_> hello all, I am having some issues with puppet + openssl, while writing a parser function to read encrypted text. | ||
| [2008/07/14 12:59:51] <asa_> my OpenSSL::PKey::RSA.new(File.read("/tmp/blah.pem")) will fail if permissions are less than 4 for others ie 704 for /tmp/blah.pem. | ||
| [2008/07/14 13:00:04] <asa_> even though I am running as root on the server and the client. | ||
| [2008/07/14 13:00:55] <asa_> is there a setuid thing going on with these plugins? I don't want my private key to have to be visible to others but this code fails if it is not. | ||
| [2008/07/14 13:02:43] <asa_> ok... nevermind. if owned by puppet then all is well. =) ignore me. | ||
| [2008/07/14 13:04:11] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 13:05:05] @ Quit: neh: Read error: 110 (Connection timed out) | ||
| [2008/07/14 13:07:24] <lazzurs> has anyone here manged to get yum to downgrade a package through puppet? | ||
| [2008/07/14 13:11:46] <ezralini> lazzurs: with Exec, anything is possible | ||
| [2008/07/14 13:12:22] <lazzurs> ezralini: indeed, just had the same chat in #yum but I would really have liked to just specify the package version and let puppet take care of it | ||
| [2008/07/14 13:13:06] <ezralini> lazzurs: I haven't checked too recently, but I've never seen an option like that in the Package type, and I wouldn't be surprised if there aren't plans for it | ||
| [2008/07/14 13:13:32] <Volcane> yum on its own isnt even too hot with downgrading | ||
| [2008/07/14 13:13:49] <lazzurs> Volcane: so I have heard | ||
| [2008/07/14 13:14:26] <Volcane> not really even just yum, its more rpm | ||
| [2008/07/14 13:15:20] <lazzurs> it is not great, the thing is what I want to do can be done. It can be done on the command line with great ease, yum remove autofs ; yum install autofs-new.ver.num | ||
| [2008/07/14 13:15:47] <Volcane> you'll need to write a custom define then | ||
| [2008/07/14 13:15:47] <lazzurs> but yum install autofs-new.ver.num will not work and because of that puppet will not do the downgrade either | ||
| [2008/07/14 13:15:53] <lazzurs> yea indeed I do :( | ||
| [2008/07/14 13:17:01] @ Quit: andrewcshafer: | ||
| [2008/07/14 13:17:15] <lazzurs> I guess it would be nice if puppet understood it was having to downgrade a package and tried to remove the package before installing the new version | ||
| [2008/07/14 13:17:29] * Volcane would hate that | ||
| [2008/07/14 13:17:31] <lazzurs> is that something people would like puppet to do or is it just me? | ||
| [2008/07/14 13:18:28] <Volcane> you never know what kind of package is badly behaved on uninstall, might nuke data/configs etc | ||
| [2008/07/14 13:18:31] @ Quit: sparanjape: "When the chips are down, well, the buffalo is empty" | ||
| [2008/07/14 13:19:05] <lazzurs> Volcane: I would suggest puppet should already be putting those configurations in place. there should be a clean seperation between programs, data and configurations | ||
| [2008/07/14 13:20:11] <Volcane> rpm just isnt designed for downgrades, expecting a one size fits all big hammer approach to be logcial in all possible scenarios is naive | ||
| [2008/07/14 13:20:25] <Volcane> suggesting that rpm upgrades will work, well thats fine cos things are designed for that | ||
| [2008/07/14 13:20:57] <f3ew> rpm --older-version IIRC | ||
| [2008/07/14 13:21:32] <lazzurs> I just have a problem with the idea that an rpm install or uninstall can break anything. if my configuration management tool understands my system state well enough then it should not matter as long as the package is able to provide the requirements | ||
| [2008/07/14 13:21:52] <f3ew> err, rpm --oldpackage autofs | ||
| [2008/07/14 13:22:01] @ authentic joined channel #puppet | ||
| [2008/07/14 13:22:11] <f3ew> lazzurs, it would be nice | ||
| [2008/07/14 13:22:25] <lazzurs> eg the rpm pre-post instal scripts should not be able to break anything that puppet is not going to fix on this/the next run | ||
| [2008/07/14 13:22:28] <Volcane> f3ew: yes, though thats on par with --force, package designers just do not generally code with that in mind and in so many cases it just wouldnt work | ||
| [2008/07/14 13:22:38] <Volcane> f3ew: like, i dont know, say cacti, you just cant downgrade it | ||
| [2008/07/14 13:23:07] <Volcane> and what do you do in that case with dependencies? | ||
| [2008/07/14 13:23:17] <lazzurs> Volcane: that is due to the data set not being compatiable, it has nothing to do with dependencies or configurations | ||
| [2008/07/14 13:23:20] <Volcane> foo-1.0 might require OLDER versions of things than foo-1.1 | ||
| [2008/07/14 13:23:50] <lazzurs> then those things should be downgraded if required as well, it is just the same dependency tracking logic, in reverse | ||
| [2008/07/14 13:24:10] <lazzurs> no, I am talking rubbish, it is just the same dependency tracking logic | ||
| [2008/07/14 13:26:05] <lazzurs> I have just done the package remove manually on the 3 systems requiring the change and the world did not end and the sky is not falling down...because I knew the effect of this because I had tested it before hand, as anyone should do before rolling out any type of new configuration | ||
| [2008/07/14 13:27:30] <Volcane> if it worked reliably for everyone, they wouldnt have gone through great effort to put the transactional rollback stuff in place (see --repackage and --rollback) | ||
| [2008/07/14 13:27:53] <Volcane> those exist because downgrading is just a nightmare in general | ||
| [2008/07/14 13:28:14] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 13:28:37] <lazzurs> Volcane: I would suggest downgrading is a nightmare because the tools do not support it and it is not something people are used to doing | ||
| [2008/07/14 13:29:20] <lazzurs> I have done this many, many times and it has never caused an issue in production | ||
| [2008/07/14 13:29:26] <Volcane> exactly, so it will be unreliable in many cases | ||
| [2008/07/14 13:29:34] <Volcane> so not something that the yum provider should just support | ||
| [2008/07/14 13:29:58] <Volcane> making a define that removes and then installs using package is trivial, if ppl want it they can do it, like 10 lines tops | ||
| [2008/07/14 13:30:21] <lazzurs> that is all just a matter people testing their configurations and data sets before going and messing with their packages on production systems | ||
| [2008/07/14 13:30:25] @ Quit: andrewcshafer: Client Quit | ||
| [2008/07/14 13:30:29] @ Quit: refuseresisted: "Lost terminal" | ||
| [2008/07/14 13:30:33] @ Quit: lak: | ||
| [2008/07/14 13:31:01] <lazzurs> Volcane: you are right, maybe I should write the patch and see if I can get it included. I do believe this is a useful feature and is something the configuration management tool should be able to do | ||
| [2008/07/14 13:31:19] * Volcane never suggested writing a patch :) | ||
| [2008/07/14 13:31:41] <lazzurs> when I give puppet a new manifest I want to be SURE that configuration is now live on a system, not oh it might be if a newer version of package x was not installed | ||
| [2008/07/14 13:32:59] <lazzurs> do you understand where I am coming from now? | ||
| [2008/07/14 13:33:55] @ Quit: kambiz: "Leaving" | ||
| [2008/07/14 13:34:23] @ kambiz joined channel #puppet | ||
| [2008/07/14 13:34:51] <Volcane> i can understand your desire, i dont think its the right place to impliment it | ||
| [2008/07/14 13:34:58] <Volcane> its the yum provider, does things with yum | ||
| [2008/07/14 13:35:03] <Volcane> knows about yum and does what yum can do | ||
| [2008/07/14 13:35:40] <Volcane> if yum cant do it, it shouldnt put in hacks that some people - not the yum maintainers/designers - think are good ideas | ||
| [2008/07/14 13:35:44] <Volcane> might as wlel use debian then | ||
| [2008/07/14 13:37:34] @ elshaa joined channel #puppet | ||
| [2008/07/14 13:37:36] <elshaa> hi | ||
| [2008/07/14 13:39:26] <lazzurs> Volcane: I kinda agree that maybe the fault is in yum | ||
| [2008/07/14 13:39:39] <lazzurs> but this is a configuration management issue, regardless of the tool underneath | ||
| [2008/07/14 13:40:10] <lazzurs> if I say I want configuration x,y,z ensured then it must happen to autonomous systems configuration and management will never work | ||
| [2008/07/14 13:40:51] <lazzurs> but yea, this is not the place for this debate, maybe the configuration management list is, either way home time :) thanks for the chat | ||
| [2008/07/14 13:41:54] @ stevenjenkins joined channel #puppet | ||
| [2008/07/14 13:42:37] <elshaa> Sorry to be brutal. I'm trying to remove lines from an .erb when a variable is empty. Could someone check that : http://rafb.net/p/AiPswt32.html | ||
| [2008/07/14 13:43:07] <elshaa> It's roughly what it does. | ||
| [2008/07/14 13:43:27] <elshaa> The problem is that the line is removed even when the variable is not empty. | ||
| [2008/07/14 13:44:41] <Volcane> that looks like it should work | ||
| [2008/07/14 13:45:07] <Volcane> oh | ||
| [2008/07/14 13:45:13] <Volcane> you're checking if its empty, put the line | ||
| [2008/07/14 13:45:23] <Volcane> if blah == "" | ||
| [2008/07/14 13:45:32] <Volcane> you want | ||
| [2008/07/14 13:45:35] <Volcane> if blah != "" | ||
| [2008/07/14 13:45:35] <Volcane> ? | ||
| [2008/07/14 13:47:19] <elshaa> Volcane: if the variable is empty, I want to remove the line | ||
| [2008/07/14 13:47:29] <elshaa> so I'm correct, right / | ||
| [2008/07/14 13:47:31] <elshaa> > | ||
| [2008/07/14 13:47:32] <elshaa> ? | ||
| [2008/07/14 13:47:33] <Volcane> if blah !== "" | ||
| [2008/07/14 13:47:37] <Volcane> put line | ||
| [2008/07/14 13:47:38] <Volcane> end | ||
| [2008/07/14 13:47:47] <elshaa> hum... let's try | ||
| [2008/07/14 13:47:48] <Volcane> if the the variable is not empty, put the line | ||
| [2008/07/14 13:47:55] <Volcane> which is the same as, if the variable is empty dont put the line | ||
| [2008/07/14 13:48:17] @ ^authentic joined channel #puppet | ||
| [2008/07/14 13:48:22] <elshaa> yes, that even makes more sense :) | ||
| [2008/07/14 13:49:48] @ teratoma_ joined channel #puppet | ||
| [2008/07/14 13:51:09] <elshaa> (erb):19: syntax error if uri !== "" ; _erbout.concat "\n" | ||
| [2008/07/14 13:51:32] <elshaa> end ; _erbout.concat "\n" | ||
| [2008/07/14 13:51:32] <elshaa> ^ in method puppetmaster.getconfig | ||
| [2008/07/14 13:52:11] @ Quit: kambiz: "Leaving" | ||
| [2008/07/14 13:52:12] <martha> !=, not !== | ||
| [2008/07/14 13:52:19] <Volcane> oh sorry | ||
| [2008/07/14 13:52:24] <Volcane> i had a stray = in there | ||
| [2008/07/14 13:52:38] <elshaa> weird, I think that worked though | ||
| [2008/07/14 13:52:41] <Volcane> thats what you get for up-arrowing :) | ||
| [2008/07/14 13:52:43] <elshaa> ok, I'll clean that | ||
| [2008/07/14 13:53:18] <elshaa> sweet, that works :) | ||
| [2008/07/14 13:53:22] <elshaa> Volcane: Thanks a lot ! | ||
| [2008/07/14 13:53:22] @ kambiz joined channel #puppet | ||
| [2008/07/14 13:53:49] <sigmonsays> dang it | ||
| [2008/07/14 13:54:01] <sigmonsays> keeping directories created in the right mode is proving difficult | ||
| [2008/07/14 13:54:11] <sigmonsays> because I have to write damn dependancies | ||
| [2008/07/14 13:54:37] <sigmonsays> is there a basename() function ? | ||
| [2008/07/14 13:54:50] <sigmonsays> dirname() ? | ||
| [2008/07/14 13:55:27] <Volcane> sigmonsays: how do you create them? with file? | ||
| [2008/07/14 13:55:36] <sigmonsays> yah | ||
| [2008/07/14 13:55:57] <Volcane> then I'm not following you :) | ||
| [2008/07/14 13:55:59] <sigmonsays> i'm thinking of using a define "smart_dir" which does require => File[dirname($name)] | ||
| [2008/07/14 13:56:01] @ refuseresisted joined channel #puppet | ||
| [2008/07/14 13:56:55] <sigmonsays> I have to make file { "/path/a/b" } require file { "/path/a" } | ||
| [2008/07/14 13:57:00] <sigmonsays> because it's *not smary* | ||
| [2008/07/14 13:57:19] <Volcane> hmm, it has autorequires | ||
| [2008/07/14 13:57:54] <Volcane> the trac for language refs is becoming unbearable | ||
| [2008/07/14 13:58:34] @ teratoma_ is now known as dryrot | ||
| [2008/07/14 14:00:18] <Volcane> it autorequires parents, owner and group | ||
| [2008/07/14 14:02:23] @ Quit: authentic: Connection timed out | ||
| [2008/07/14 14:02:30] @ ^authentic is now known as authentic | ||
| [2008/07/14 14:02:51] <sigmonsays> when did that start? | ||
| [2008/07/14 14:02:52] <sigmonsays> (what version) | ||
| [2008/07/14 14:04:10] <Volcane> not sure | ||
| [2008/07/14 14:04:21] <Volcane> http://www.pastie.org/233157 theres the code though | ||
| [2008/07/14 14:04:30] @ Quit: dryrot: "leaving" | ||
| [2008/07/14 14:04:45] <Volcane> from file.rb | ||
| [2008/07/14 14:05:09] <Disconnect> Volcane: not parents .. parent | ||
| [2008/07/14 14:05:42] <Volcane> and if you made the parent with file, it will require its parent surely? | ||
| [2008/07/14 14:05:46] <Volcane> and so on and so on? | ||
| [2008/07/14 14:06:45] <sigmonsays> well that's common | ||
| [2008/07/14 14:06:55] <sigmonsays> "parents parent and so no" | ||
| [2008/07/14 14:07:08] <sigmonsays> especially since you can't control the order of the resources | ||
| [2008/07/14 14:07:29] <bgupta> So when I install fastthread with gem directly, it prompts to ask which version we want (mswin32/ruby/i386-mswin32). (The ruby version.) Puppet however is thrown by the prompt. Anone know how to get it to work? | ||
| [2008/07/14 14:08:08] <Volcane> sigmonsays: I'm not following you, theres only one logical thing to do when you make /1/2/3/4/5 and that is that 5 would require 4, 4 would require 3 all the way down to 1 and it would be made in the correct order | ||
| [2008/07/14 14:08:27] <Volcane> anyway, gotta go home | ||
| [2008/07/14 14:13:10] @ Quit: msf: Read error: 110 (Connection timed out) | ||
| [2008/07/14 14:13:10] @ Quit: mcbride: Read error: 110 (Connection timed out) | ||
| [2008/07/14 14:15:14] <martha> can functions be used directly in types, or do I have to assign a variable first? | ||
| [2008/07/14 14:15:56] <martha> ie, can I do cron { "aptitude update": hour => hour_from_ipaddress() } ? | ||
| [2008/07/14 14:17:03] @ shake-n-bake joined channel #puppet | ||
| [2008/07/14 14:18:03] @ meandtheshell joined channel #puppet | ||
| [2008/07/14 14:19:29] @ meandtheshell1 joined channel #puppet | ||
| [2008/07/14 14:23:59] <benp-> wow massive power failure in vancouver. the building where all my servers is on has no power. which is surprising because they supposedly have lots of UPSes and backup generators.. sigh | ||
| [2008/07/14 14:24:16] <martha> benp-: peer1? | ||
| [2008/07/14 14:24:20] <benp-> yes | ||
| [2008/07/14 14:24:34] <martha> ah, I wondered where all our machines went | ||
| [2008/07/14 14:24:54] <refuseresisted> pwned | ||
| [2008/07/14 14:25:10] @ ^authentic joined channel #puppet | ||
| [2008/07/14 14:26:08] <refuseresisted> you guys have an SLA? | ||
| [2008/07/14 14:26:26] <benp-> http://www.canada.com/vancouversun/news/story.html?id=92275acb-1e0b-4ff8-b2b6-f0731881009a | ||
| [2008/07/14 14:26:50] <benp-> heh, even their phone system is down | ||
| [2008/07/14 14:26:53] <benp-> yeah | ||
| [2008/07/14 14:32:04] @ Quit: muerr: "Time wasted on IRC: 3 hours 1 minute 24 seconds" | ||
| [2008/07/14 14:33:59] @ Quit: meandtheshell: Connection timed out | ||
| [2008/07/14 14:34:28] @ Quit: shake-n-bake: Read error: 110 (Connection timed out) | ||
| [2008/07/14 14:39:29] <refuseresisted> d'oh! -> "Page LanguageReference not found" | ||
| [2008/07/14 14:40:04] <refuseresisted> hmm, maybe my bookmarks got screwed up. ../LanguageTutorial works | ||
| [2008/07/14 14:41:06] @ Quit: authentic: Read error: 110 (Connection timed out) | ||
| [2008/07/14 14:41:12] @ ^authentic is now known as authentic | ||
| [2008/07/14 14:43:21] @ f--z joined channel #puppet | ||
| [2008/07/14 14:48:18] @ shake-n-bake_ joined channel #puppet | ||
| [2008/07/14 14:59:05] @ Quit: f--z: Remote closed the connection | ||
| [2008/07/14 14:59:10] @ Quit: stevenjenkins: "Leaving." | ||
| [2008/07/14 15:00:24] @ dysinger joined channel #puppet | ||
| [2008/07/14 15:01:22] @ f--z joined channel #puppet | ||
| [2008/07/14 15:05:44] @ lak joined channel #puppet | ||
| [2008/07/14 15:06:40] @ hessml joined channel #puppet | ||
| [2008/07/14 15:06:55] <hessml> morning all | ||
| [2008/07/14 15:07:35] <hessml> anybody mind if I go off topic and bring up network switches again | ||
| [2008/07/14 15:08:05] @ Quit: f--z: Remote closed the connection | ||
| [2008/07/14 15:08:36] <fastjay> hessml: i dont mind :) | ||
| [2008/07/14 15:11:30] <hessml> i've noticed that dell and netgear have more or less a very similar command set -- are they all stealing the same command language or are they oeming the same software? | ||
| [2008/07/14 15:11:52] <refuseresisted> dell probably OEMed netgear hardware | ||
| [2008/07/14 15:12:20] @ plathrop joined channel #puppet | ||
| [2008/07/14 15:12:48] <refuseresisted> those things suck anyway | ||
| [2008/07/14 15:13:15] <hessml> netgear has less features then dell so that seems unlikely | ||
| [2008/07/14 15:13:29] <hessml> but netgear does suck | ||
| [2008/07/14 15:14:07] <hessml> why do you say that dell sucks? (I'm trying to find something to upgrade from -- I'm on netgear now) | ||
| [2008/07/14 15:14:19] <refuseresisted> i prefer cisco hardware | ||
| [2008/07/14 15:15:02] <hessml> ah... i'm trying to not pay $600/gbit port | ||
| [2008/07/14 15:15:07] <refuseresisted> how many ports | ||
| [2008/07/14 15:15:23] <refuseresisted> 2960 is only a few k new | ||
| [2008/07/14 15:15:27] <refuseresisted> cheaper on ebay | ||
| [2008/07/14 15:15:37] <hessml> short term i need 6 switches each with 24 ports | ||
| [2008/07/14 15:15:45] <hessml> gbit | ||
| [2008/07/14 15:16:13] <refuseresisted> http://preview.tinyurl.com/66xabh | ||
| [2008/07/14 15:16:43] <refuseresisted> i run those in my racks | ||
| [2008/07/14 15:16:51] <refuseresisted> 48 port version though | ||
| [2008/07/14 15:17:34] <refuseresisted> as long as you arent pushing full line rate on each port you'll be fine. but then again you'd surely be oversubscribed with a netgear as well | ||
| [2008/07/14 15:17:38] <hessml> $1471 not too bad for not going insane and wasting my friking life away | ||
| [2008/07/14 15:18:11] <refuseresisted> :) | ||
| [2008/07/14 15:18:40] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 15:19:31] <hessml> netgears are bricking on me --- just sitting in the rack humming away and then brick | ||
| [2008/07/14 15:19:59] <refuseresisted> high ambient temp? | ||
| [2008/07/14 15:20:10] <hessml> mid 70s | ||
| [2008/07/14 15:20:44] @ f--z joined channel #puppet | ||
| [2008/07/14 15:20:49] <score> hessml: i think i heard it was really close to smc firmware | ||
| [2008/07/14 15:23:16] @ Quit: andrewcshafer: Client Quit | ||
| [2008/07/14 15:23:17] @ Quit: lak: | ||
| [2008/07/14 15:23:48] <hessml> do people put all their switches behind firewalls? | ||
| [2008/07/14 15:24:28] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 15:24:32] <refuseresisted> depends. the firewall has to have connectivity as well | ||
| [2008/07/14 15:24:34] <hessml> I secure mine and stick them right on the internet with a puplic IP and everything, but I'm told I'm insane.... | ||
| [2008/07/14 15:24:51] <hessml> public | ||
| [2008/07/14 15:25:03] <refuseresisted> your router has to be connected to a switch to give your internal hosts access through it | ||
| [2008/07/14 15:25:40] <refuseresisted> you should have two seperate subnets, at least logically - internal and external | ||
| [2008/07/14 15:26:29] <refuseresisted> it's a bad idea to give it a public ip address though | ||
| [2008/07/14 15:32:04] @ Quit: f--z: Remote closed the connection | ||
| [2008/07/14 15:34:22] @ f--z joined channel #puppet | ||
| [2008/07/14 15:34:45] <refuseresisted> anyone know if it's possible to specify a tag as a variable in a collection? | ||
| [2008/07/14 15:36:33] @ ^authentic joined channel #puppet | ||
| [2008/07/14 15:47:05] @ Quit: andrewcshafer: Read error: 110 (Connection timed out) | ||
| [2008/07/14 15:48:21] @ Quit: authentic: Read error: 110 (Connection timed out) | ||
| [2008/07/14 15:48:26] @ ^authentic is now known as authentic | ||
| [2008/07/14 15:50:04] @ Quit: pleemans: "Ex-Chat" | ||
| [2008/07/14 15:50:37] @ hessml is now known as hessml|away | ||
| [2008/07/14 15:50:39] @ hessml|away is now known as hessml | ||
| [2008/07/14 16:04:59] @ Quit: f--z: "KVIrc 3.2.5 Anomalies http://www.kvirc.net/" | ||
| [2008/07/14 16:06:13] <hessml> ok. i'm back with more off topic questions | ||
| [2008/07/14 16:06:28] <hessml> does anyone here use ipmi? | ||
| [2008/07/14 16:08:44] @ Quit: meandtheshell1: "Leaving." | ||
| [2008/07/14 16:10:10] <Volcane> for monitoring only | ||
| [2008/07/14 16:10:34] @ Quit: asa_: | ||
| [2008/07/14 16:10:47] <hessml> whose servers are you using? | ||
| [2008/07/14 16:10:56] <Volcane> ibm, hp | ||
| [2008/07/14 16:11:00] <Volcane> some generic junk | ||
| [2008/07/14 16:11:28] <hessml> are you happy with the ipmi? does it work as advertised? | ||
| [2008/07/14 16:12:12] <Volcane> yip, I've not tried it for management etc though i guess the server management software that came with my kit probably use it, donno | ||
| [2008/07/14 16:12:30] <Volcane> was looking at it for some STONITH use with heartbeat but not got so far yet | ||
| [2008/07/14 16:13:05] <hessml> have you ever tried to use the serial console stuff on it? | ||
| [2008/07/14 16:13:21] <Volcane> i used machines that had it setup, but not setup myself | ||
| [2008/07/14 16:14:09] <hessml> i'm just trying to figure out which companies implement ipmi well and which don't | ||
| [2008/07/14 16:14:41] <hessml> i'm on supermicro right now, and they do a mostly ok job.... | ||
| [2008/07/14 16:14:41] <Volcane> they all have little hangups | ||
| [2008/07/14 16:15:00] <Volcane> like my old ibm blades, they are not supported under redhat for ipmi from inside the os | ||
| [2008/07/14 16:15:04] <hessml> thats what I was afraid of | ||
| [2008/07/14 16:15:06] <Volcane> newer ones are sweet | ||
| [2008/07/14 16:15:45] <Volcane> i have some cheap junk hp machines, literally paid 120 pounds for them brand new, and they seem great | ||
| [2008/07/14 16:15:53] <Volcane> supported out the box with redhat 5 | ||
| [2008/07/14 16:16:15] <Volcane> http://nephilim.ml.org/~rip/ipmi.txt | ||
| [2008/07/14 16:16:17] <Volcane> thats from the hp box | ||
| [2008/07/14 16:16:39] <hessml> nice | ||
| [2008/07/14 16:17:52] <Volcane> do you have super micro blades? | ||
| [2008/07/14 16:21:01] <hessml> no, just 1u servers | ||
| [2008/07/14 16:22:02] <Volcane> their blade systems looks quite nice | ||
| [2008/07/14 16:22:41] @ snerdling joined channel #puppet | ||
| [2008/07/14 16:22:47] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 16:23:03] <hessml> I haven't really found blades to be more economical than 1u systems. Maybe I've been doing the math wrong, but 1u systems keep coming up cheaper for me. | ||
| [2008/07/14 16:23:25] <Volcane> they're much more convenient | ||
| [2008/07/14 16:23:26] @ Quit: snerdly: Read error: 110 (Connection timed out) | ||
| [2008/07/14 16:23:28] <refuseresisted> i'm paying ~1800 for dual proc quad core low energy dells, 8gb ram | ||
| [2008/07/14 16:23:51] <Volcane> refuseresisted: $? | ||
| [2008/07/14 16:23:56] <refuseresisted> 1800 usd | ||
| [2008/07/14 16:24:11] <Volcane> refuseresisted: thats very cheap, how many do you buy? | ||
| [2008/07/14 16:24:12] <refuseresisted> they pull like, 1amp under load | ||
| [2008/07/14 16:24:16] <refuseresisted> just bought 32 | ||
| [2008/07/14 16:24:27] <refuseresisted> first order | ||
| [2008/07/14 16:24:29] <hessml> wow, thats good | ||
| [2008/07/14 16:24:35] <Volcane> that insanely cheap then | ||
| [2008/07/14 16:24:40] <refuseresisted> :) | ||
| [2008/07/14 16:24:42] <hessml> does that come with ipmi? | ||
| [2008/07/14 16:24:50] <refuseresisted> drac included | ||
| [2008/07/14 16:25:08] <hessml> drac is ipmi? | ||
| [2008/07/14 16:25:09] <Volcane> anyway, a blade, 14 servers = 4 x power, 1 x kvm (2nd if you want), 2 to 8 network cables, thats it | ||
| [2008/07/14 16:25:24] <refuseresisted> it's their bmc, supports full console access over the web | ||
| [2008/07/14 16:25:26] <Volcane> compared to 28 power, 28 nic, 14 kvm | ||
| [2008/07/14 16:26:03] <Volcane> that alone is a huge win to me | ||
| [2008/07/14 16:26:30] @ lak joined channel #puppet | ||
| [2008/07/14 16:26:36] <hessml> what model number is the dell? | ||
| [2008/07/14 16:26:49] <refuseresisted> energy smart 1950 | ||
| [2008/07/14 16:27:21] <hessml> volcane: what model is your blade? | ||
| [2008/07/14 16:27:34] <refuseresisted> one of the biggest pros for me is drac, becausr you just need a second ethernet drop - no kvm necessary | ||
| [2008/07/14 16:27:51] <Volcane> hessml: hs21 | ||
| [2008/07/14 16:27:53] @ Quit: lak: Read error: 104 (Connection reset by peer) | ||
| [2008/07/14 16:28:11] <hessml> is drac the same thing as ipmi? | ||
| [2008/07/14 16:28:13] @ lak joined channel #puppet | ||
| [2008/07/14 16:28:31] <refuseresisted> no, it's proprietary | ||
| [2008/07/14 16:29:01] <refuseresisted> http://www.dell.com/content/topics/global.aspx/power/en/ps2q02_bell | ||
| [2008/07/14 16:29:17] @ f--z joined channel #puppet | ||
| [2008/07/14 16:29:39] <refuseresisted> we were using penguin servers when i started. stay away from them | ||
| [2008/07/14 16:29:49] <hessml> they claim they are ipmi 2.0 | ||
| [2008/07/14 16:30:07] <refuseresisted> it does support ipmi as well yes | ||
| [2008/07/14 16:32:13] <Volcane> refuseresisted: those 1U's are exactly the same price as i pay for ibm blades with the same spec (at least via dells website) | ||
| [2008/07/14 16:32:51] <refuseresisted> Volcane: you mean list price? | ||
| [2008/07/14 16:33:08] <Volcane> nods, i know you'll get htem for quite a bit less if you buy 32 | ||
| [2008/07/14 16:33:12] <Volcane> ditto for ibm | ||
| [2008/07/14 16:33:32] <refuseresisted> yea.. i dunno what ibm's discounts are like but dell can offer some pretty steep ones since they are so in bed with intekl | ||
| [2008/07/14 16:34:14] <Volcane> had really bad experience with dell in the past though, very unhappy | ||
| [2008/07/14 16:34:22] <refuseresisted> that sucks, what happened? | ||
| [2008/07/14 16:34:30] <hessml> what makes it "Energy Smart" | ||
| [2008/07/14 16:34:31] <Volcane> litterally just tossed about 100 dells out 6 months before their warrenty were out | ||
| [2008/07/14 16:34:45] <riddley> yea dell is bad news | ||
| [2008/07/14 16:34:45] <refuseresisted> hessml: it's a low power server | ||
| [2008/07/14 16:34:47] <riddley> esp for linux | ||
| [2008/07/14 16:34:53] <Volcane> refuseresisted: we had 4 of the biggest dell yhou could buy at the time running sql server | ||
| [2008/07/14 16:34:57] <refuseresisted> riddley: they're great for me | ||
| [2008/07/14 16:35:00] <riddley> and their support model is completely _retarded_ | ||
| [2008/07/14 16:35:11] <Volcane> refuseresisted: they died pretty soon from each other and even though we had top level of support it took dell more than a month to fix each | ||
| [2008/07/14 16:35:22] <riddley> 5 years maximum support from the time _the machine is made_ (not when you bought it) | ||
| [2008/07/14 16:35:29] <riddley> that's complllletely unacceptable | ||
| [2008/07/14 16:35:36] <Volcane> riddley: they swapped *everything* in my servers back then | ||
| [2008/07/14 16:35:46] <Volcane> riddley: so when i asked them wtf is going on why do i keep getting shit | ||
| [2008/07/14 16:35:56] <Volcane> riddley: they owned up to the fact that they were replacing it with refurbs | ||
| [2008/07/14 16:36:04] <riddley> we use HP's. open-source drivers that work great for all of it. | ||
| [2008/07/14 16:36:12] <riddley> Volcane, unbelievable | ||
| [2008/07/14 16:36:13] <refuseresisted> hessml: a 1u es 1950 (8 core intel) is drawing like 1amp under load for me, where as a 1u generic server with only 4 total cores (amd) is drawing almost 2 amps | ||
| [2008/07/14 16:36:14] <Volcane> riddley: and they didnt actually have new stock, so i had to wait 4 weeks for a new baseboard to come etc | ||
| [2008/07/14 16:37:01] <riddley> we were using HP 48xx and 58xx series, but their blades are so head-and-shoulders above the rest, we've switched to buying them exclusively | ||
| [2008/07/14 16:37:07] <riddley> (for linux) | ||
| [2008/07/14 16:37:23] <refuseresisted> riddley: how is HP's bmc? | ||
| [2008/07/14 16:37:29] <riddley> bmc? | ||
| [2008/07/14 16:37:34] <riddley> oh managment? | ||
| [2008/07/14 16:37:36] <refuseresisted> i've heard it's nice but havent seen it | ||
| [2008/07/14 16:37:36] <refuseresisted> yea | ||
| [2008/07/14 16:37:43] <riddley> that's the H&S above the rest stuff | ||
| [2008/07/14 16:37:48] <riddley> it's _amazing_ | ||
| [2008/07/14 16:37:54] <refuseresisted> cool | ||
| [2008/07/14 16:38:04] <riddley> it's so good I never would have believed anyone if they'd tried to convince me hehe | ||
| [2008/07/14 16:38:09] <Volcane> sun's intel/amd blades are quite kewl too | ||
| [2008/07/14 16:38:15] <refuseresisted> haha | ||
| [2008/07/14 16:38:26] <hessml> i need low power, lots or memory, 1 drive, 2 nics, 1 ipmi. need it work for 3 years before it burns out | ||
| [2008/07/14 16:38:26] <riddley> sun's blades are crap in terms of management | ||
| [2008/07/14 16:38:30] <riddley> I haven't seen intels | ||
| [2008/07/14 16:38:55] <riddley> the HP stuff is nice because it's all ajax-y and doesn't require plugins or windows | ||
| [2008/07/14 16:39:00] <riddley> and can all be done CLI if you like | ||
| [2008/07/14 16:39:11] <Volcane> riddley: have u seen the brand new sun kit? | ||
| [2008/07/14 16:39:25] <riddley> we demoed the sun blades and couldn't see the win over 1u suns | ||
| [2008/07/14 16:39:29] <riddley> yea | ||
| [2008/07/14 16:39:37] <riddley> well brand new as of 2mo ago | ||
| [2008/07/14 16:39:47] <riddley> or 1mo ... whenever that sun guy was here last | ||
| [2008/07/14 16:39:50] <Volcane> ah | ||
| [2008/07/14 16:40:11] <riddley> no kidding I'd say that the hp blades are 10,000x better | ||
| [2008/07/14 16:40:24] <riddley> I work for HP in the blade dept... | ||
| [2008/07/14 16:40:26] <riddley> jk | ||
| [2008/07/14 16:40:26] <riddley> :) | ||
| [2008/07/14 16:40:46] <refuseresisted> haha | ||
| [2008/07/14 16:41:02] <Volcane> yeah well i was at a big client | ||
| [2008/07/14 16:41:07] <Volcane> we were replacing about 60 sun blades | ||
| [2008/07/14 16:41:11] <riddley> we got in 106 blades at one of our sites and within 48hrs had them all delivered to our apps group: thanks to their awesome management and puppet | ||
| [2008/07/14 16:41:11] <Volcane> and got ibm, sun etc to quote | ||
| [2008/07/14 16:41:13] <Volcane> hp too | ||
| [2008/07/14 16:41:16] <Volcane> they didnt even bother quoting | ||
| [2008/07/14 16:41:27] <Volcane> quote came WEEKS late | ||
| [2008/07/14 16:41:46] <riddley> we have a great HP var, but yea, it can take a while with HP | ||
| [2008/07/14 16:46:54] @ [newbie] joined channel #puppet | ||
| [2008/07/14 16:49:02] @ Quit: ezralini: | ||
| [2008/07/14 16:49:43] <hessml> refuseresisted: I got $2896 when I configured one of the dell systems the way you described. I need 18. I wonder if they will give me them at your price of $1800 if I buy that many. | ||
| [2008/07/14 16:50:59] <refuseresisted> hessml: probably fairly close | ||
| [2008/07/14 16:56:26] @ jbrothers joined channel #puppet | ||
| [2008/07/14 16:57:30] @ Quit: f--z: Read error: 110 (Connection timed out) | ||
| [2008/07/14 16:58:10] <jbrothers> would someone clue me in on getting manufacturer / product info from facter? i'd like to install hardware specific packages if appropriate | ||
| [2008/07/14 16:58:23] <Disconnect> install dmidecode | ||
| [2008/07/14 16:58:30] <jbrothers> it is | ||
| [2008/07/14 16:58:40] <Disconnect> factor should pick up manuf and such from there | ||
| [2008/07/14 16:58:50] <Volcane> when run as root | ||
| [2008/07/14 16:58:56] <Disconnect> er, facter | ||
| [2008/07/14 16:58:58] <Disconnect> yah as root | ||
| [2008/07/14 16:59:33] <Disconnect> manufacturer => Dell Inc. | ||
| [2008/07/14 16:59:33] <Disconnect> productname => PowerEdge SC1435 | ||
| [2008/07/14 16:59:42] <jbrothers> still no manufacturer as root | ||
| [2008/07/14 16:59:53] <Volcane> jbrothers: do you see the info in dmidecode on its own? | ||
| [2008/07/14 17:00:03] <jbrothers> facter -v 1.3.8 | ||
| [2008/07/14 17:00:30] <refuseresisted> i see it with 1.3.8 and my dell h/w | ||
| [2008/07/14 17:01:09] <jbrothers> just dmidecode says HP, ProLiant DL380, etc | ||
| [2008/07/14 17:01:11] @ f--z joined channel #puppet | ||
| [2008/07/14 17:02:23] @ Quit: f--z: Client Quit | ||
| [2008/07/14 17:02:57] @ WesleyS joined channel #puppet | ||
| [2008/07/14 17:03:07] <jbrothers> facter --debug | ||
| [2008/07/14 17:03:07] <Volcane> what os? | ||
| [2008/07/14 17:03:13] <jbrothers> value for productname is still nil | ||
| [2008/07/14 17:03:33] <jbrothers> this is centos 4.4 | ||
| [2008/07/14 17:04:12] <Volcane> # dmidecode|grep "Product Name" | ||
| [2008/07/14 17:04:14] <Volcane> says what? | ||
| [2008/07/14 17:04:39] @ WesleyS left channel #puppet () | ||
| [2008/07/14 17:04:53] <jbrothers> Product Name: ProLiant DL380 G4 | ||
| [2008/07/14 17:05:08] <Volcane> what does 'facter kernel' say? | ||
| [2008/07/14 17:05:31] <jbrothers> Linux | ||
| [2008/07/14 17:05:32] * Volcane will be back in a bit | ||
| [2008/07/14 17:05:38] <Volcane> path to dmidecode? | ||
| [2008/07/14 17:05:52] <jbrothers> /usr/sbin/dmidecode | ||
| [2008/07/14 17:06:38] @ Quit: [newbie]: Read error: 110 (Connection timed out) | ||
| [2008/07/14 17:11:21] @ shake-n-bake joined channel #puppet | ||
| [2008/07/14 17:16:31] <Volcane> weird, someone else had the same problem on centos 4.x the other day but i were unable to reproduce | ||
| [2008/07/14 17:16:46] <Volcane> oh actually he couldnt get the lsb stuff | ||
| [2008/07/14 17:17:33] <Volcane> yeah defo was lsb stuff not dmi | ||
| [2008/07/14 17:17:40] <Volcane> my centos 4 has the same probs as u | ||
| [2008/07/14 17:19:06] <Volcane> yes, on centos 4.x it wont work at all | ||
| [2008/07/14 17:19:13] <Volcane> facter 1.5 will fix it for u | ||
| [2008/07/14 17:19:47] <Volcane> on 1.3.8 its searching for "System Information" | ||
| [2008/07/14 17:19:52] <Volcane> in dmidecode output | ||
| [2008/07/14 17:19:56] @ Quit: shake-n-bake_: Read error: 110 (Connection timed out) | ||
| [2008/07/14 17:20:15] @ Quit: andrewcshafer: | ||
| [2008/07/14 17:20:28] @ Quit: lak: | ||
| [2008/07/14 17:20:47] <fastjay> anyone here using puppetrun? | ||
| [2008/07/14 17:21:11] <fastjay> my puppetd is returning a 502 when i try it | ||
| [2008/07/14 17:27:17] @ Quit: a-priori: | ||
| [2008/07/14 17:29:03] <Volcane> yes, 1.5 fixes it | ||
| [2008/07/14 17:29:44] <sigmonsays> Anyone ever experiement using binary diffs to restore large chunks of data? | ||
| [2008/07/14 17:29:51] <Volcane> productname => PowerEdge SC1425 | ||
| [2008/07/14 17:34:03] <jbrothers> thanks for looking into it volcane | ||
| [2008/07/14 17:34:22] <refuseresisted> anyone familiar with using checksums to monitor directory changes? it seems to only notify on the next catalog run | ||
| [2008/07/14 17:34:42] <jason^_> how can i get puppet to put a file on one specific machine but for the rest it gets a generic file? | ||
| [2008/07/14 17:35:03] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 17:35:45] <Volcane> jason^_: source => [ "puppet://puppet/path/to/file.$fqdn", "puppet://puppet/path/to/file"] | ||
| [2008/07/14 17:36:02] <Volcane> jason^_: if file.host.you.com exist it will get that, else it will get file | ||
| [2008/07/14 17:36:08] <jason^_> oh nice | ||
| [2008/07/14 17:36:12] <jason^_> i didn't realize that worked | ||
| [2008/07/14 17:36:20] @ Quit: andrewcshafer: Client Quit | ||
| [2008/07/14 17:39:58] @ Quit: fujin: "leaving" | ||
| [2008/07/14 17:42:30] @ punkcut joined channel #puppet | ||
| [2008/07/14 17:43:00] <punkcut> How can I get the User type to automatically create the home directory if needed? | ||
| [2008/07/14 17:43:09] <refuseresisted> managehome => true | ||
| [2008/07/14 17:44:01] <punkcut> Thanks refuseresisted | ||
| [2008/07/14 17:44:15] <punkcut> what is the full scope of the "management" of a homedir? | ||
| [2008/07/14 17:46:18] <refuseresisted> uses -m for useradd in centos/rh, etc | ||
| [2008/07/14 17:46:32] <punkcut> lol. ok. | ||
| [2008/07/14 17:46:39] <punkcut> thanks refuseresisted | ||
| [2008/07/14 17:46:41] <refuseresisted> np | ||
| [2008/07/14 17:56:41] @ Wakko666 joined channel #puppet | ||
| [2008/07/14 17:59:43] @ Quit: ski98033: "Leaving." | ||
| [2008/07/14 18:00:36] @ randybias joined channel #puppet | ||
| [2008/07/14 18:00:39] @ Quit: Aaton: | ||
| [2008/07/14 18:00:59] @ Aaton joined channel #puppet | ||
| [2008/07/14 18:03:42] @ Aaton left channel #puppet ("Leaving...") | ||
| [2008/07/14 18:04:48] @ aymerick joined channel #puppet | ||
| [2008/07/14 18:07:34] @ fujin joined channel #puppet | ||
| [2008/07/14 18:09:44] @ Quit: aymerick: Client Quit | ||
| [2008/07/14 18:09:54] @ inigo_montoya is now known as mmcgrath | ||
| [2008/07/14 18:10:32] @ Quit: pjw: | ||
| [2008/07/14 18:13:05] @ jbrothers left channel #puppet () | ||
| [2008/07/14 18:15:19] @ lak joined channel #puppet | ||
| [2008/07/14 18:16:26] <Volcane> lak: is this known behaviour: facter serialnumber shows nothing but facter|grep serialnumber is fine? | ||
| [2008/07/14 18:16:29] <Volcane> 1.5 | ||
| [2008/07/14 18:21:22] @ a-priori joined channel #puppet | ||
| [2008/07/14 18:21:35] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 18:22:04] @ Wakko666 left channel #puppet () | ||
| [2008/07/14 18:22:07] @ Wakko666 joined channel #puppet | ||
| [2008/07/14 18:25:53] @ punkcut left channel #puppet () | ||
| [2008/07/14 18:31:23] @ \ask_ joined channel #puppet | ||
| [2008/07/14 18:31:40] <Volcane> seems to be same as #1365 i guess | ||
| [2008/07/14 18:31:41] <gepetto> Volcane: #1365 is http://reductivelabs.com/redmine/issues/show/1365 | ||
| [2008/07/14 18:33:08] <jamesturnbull> Volcane: is serialnumber a custom fact? | ||
| [2008/07/14 18:34:19] <jamesturnbull> Volcane: ah ignore me | ||
| [2008/07/14 18:34:59] <jamesturnbull> Volcane: and yes you are right - that's the issue | ||
| [2008/07/14 18:36:05] <Volcane> sux | ||
| [2008/07/14 18:36:10] <Volcane> but not really a problem | ||
| [2008/07/14 18:36:20] <Volcane> just screwed me around earlier trying to ehlp someone :) | ||
| [2008/07/14 18:36:30] <Volcane> built rpms and upgraded my facter | ||
| [2008/07/14 18:36:42] <Volcane> and still couldnt see the facts, but the code seems to suggest it would work etc | ||
| [2008/07/14 18:36:49] <Volcane> turns out that bug got me :) | ||
| [2008/07/14 18:40:09] @ shenson is now known as shenson_not_here | ||
| [2008/07/14 18:49:12] @ punkcut joined channel #puppet | ||
| [2008/07/14 18:53:26] @ Quit: lak: | ||
| [2008/07/14 18:53:37] @ Quit: andrewcshafer: | ||
| [2008/07/14 18:54:26] <fujin> bloody dynamic facts eh jamesturnbull | ||
| [2008/07/14 18:55:13] <fujin> I've just locally modified bin/puppet to load *all* facts and then |'d the facts array to the user-supplied arguments | ||
| [2008/07/14 18:55:18] <fujin> err, bin/facter | ||
| [2008/07/14 18:55:24] <fujin> so lazy.. I win :D | ||
| [2008/07/14 18:58:04] <jamesturnbull> Volcane: it's not a bug as such - more of a feature and a way to do other things | ||
| [2008/07/14 18:58:28] <jamesturnbull> Volcane: but it needs some further thought - will be seeing LUke next week so we'll chat | ||
| [2008/07/14 18:58:38] <punkcut> Volcane : I wanted to ask you about the YumProvider error you had seen before. You said there was some issue with architecture hints? | ||
| [2008/07/14 18:58:38] <Volcane> nods, cos it invalidates the --help output | ||
| [2008/07/14 18:58:56] <Volcane> punkcut: yeah i dont really remember now, read that log from the other day that u found :) | ||
| [2008/07/14 18:59:06] <Volcane> punkcut: its prolly more useful than me at midnight | ||
| [2008/07/14 18:59:38] @ MrProper__ joined channel #puppet | ||
| [2008/07/14 18:59:40] <punkcut> yah I scanned through it but didnt understand a lot of the references to puppets underlying code. | ||
| [2008/07/14 19:01:00] <jamesturnbull> Volcane: is this loading rm's of differing architectures? | ||
| [2008/07/14 19:01:08] <jamesturnbull> s/rm/rpms/ | ||
| [2008/07/14 19:01:26] <MrProper__> mornin all | ||
| [2008/07/14 19:01:29] <Volcane> jamesturnbull: i thnk so, i very vaguely remember the problem, punkcut what was the problem you're having again? | ||
| [2008/07/14 19:05:05] @ Quit: \ask_: | ||
| [2008/07/14 19:06:04] @ \ask_ joined channel #puppet | ||
| [2008/07/14 19:13:16] * Volcane goes to bed | ||
| [2008/07/14 19:13:28] <fastjay> i would love to be in bed right now | ||
| [2008/07/14 19:14:41] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 19:14:56] @ andrewcshafer left channel #puppet () | ||
| [2008/07/14 19:15:09] @ hessml is now known as hessml|away | ||
| [2008/07/14 19:16:20] @ Quit: hessml|away: "Leaving..." | ||
| [2008/07/14 19:27:12] <refuseresisted> does exec {} still support before => ? | ||
| [2008/07/14 19:27:48] <refuseresisted> nm, finally found it | ||
| [2008/07/14 19:34:10] @ pjw joined channel #puppet | ||
| [2008/07/14 19:34:18] @ hessml joined channel #puppet | ||
| [2008/07/14 19:34:58] @ Quit: hessml: Client Quit | ||
| [2008/07/14 19:42:21] @ Quit: Wakko666: "Leaving." | ||
| [2008/07/14 19:54:35] @ hessml joined channel #puppet | ||
| [2008/07/14 19:56:37] @ Quit: snerdling: Remote closed the connection | ||
| [2008/07/14 19:57:59] <hessml> refuseresisted: they quoted me $1530 | ||
| [2008/07/14 19:59:35] <hessml> oh, you got 8 core. I'm getting 4 | ||
| [2008/07/14 19:59:44] <hessml> never mind | ||
| [2008/07/14 20:00:14] @ martha left channel #puppet () | ||
| [2008/07/14 20:02:34] @ Quit: hessml: "Leaving..." | ||
| [2008/07/14 20:13:39] @ pinguskahn joined channel #puppet | ||
| [2008/07/14 20:22:04] <sigmonsays> There is some "file replication" daemon out there. I'm hoping to experiment with it to synchronize a common set of directories from one machine to many others | ||
| [2008/07/14 20:22:18] <sigmonsays> the problem: I don't remember the name | ||
| [2008/07/14 20:27:52] @ neh joined channel #puppet | ||
| [2008/07/14 20:29:05] <chadh> sigmonsays: rsync? | ||
| [2008/07/14 20:29:32] <chadh> or something lower level like drdb? | ||
| [2008/07/14 20:30:37] @ Quit: Kindred: Read error: 104 (Connection reset by peer) | ||
| [2008/07/14 20:31:14] @ Kindred joined channel #puppet | ||
| [2008/07/14 20:31:32] @ Quit: a-priori: | ||
| [2008/07/14 20:32:31] @ Wakko666 joined channel #puppet | ||
| [2008/07/14 20:34:27] <fujin> rsync :D | ||
| [2008/07/14 20:39:21] <chadh> anyone around that uses puppet on solaris10? | ||
| [2008/07/14 20:58:01] @ groove joined channel #puppet | ||
| [2008/07/14 20:59:23] <groove> can anyone refer me to a book that shows you how to write a high-level configuration management specification? | ||
| [2008/07/14 21:00:05] @ Quit: plathrop: "ERC Version 5.2 (IRC client for Emacs)" | ||
| [2008/07/14 21:00:20] @ lak joined channel #puppet | ||
| [2008/07/14 21:01:15] <dan___t> How high level are you talking? | ||
| [2008/07/14 21:01:16] @ Quit: lak: Read error: 104 (Connection reset by peer) | ||
| [2008/07/14 21:01:20] @ lak joined channel #puppet | ||
| [2008/07/14 21:01:25] <dan___t> And what does Puppet lack in that regard? | ||
| [2008/07/14 21:01:51] <groove> like a very generalised specification | ||
| [2008/07/14 21:02:01] <groove> that could be implemented in any config management system | ||
| [2008/07/14 21:02:12] <dan___t> Suppose you could start out with UML or something... | ||
| [2008/07/14 21:02:18] <dan___t> Or YAML | ||
| [2008/07/14 21:02:24] <groove> it sounds like you don't think it is worth it | ||
| [2008/07/14 21:02:39] <dan___t> What are you talking about? | ||
| [2008/07/14 21:02:50] <groove> i'm trying to describe in implementation-free terms our current cfengine configuration | ||
| [2008/07/14 21:03:05] <groove> in theory it will then be easier to re-implement in puppet | ||
| [2008/07/14 21:03:12] <dan___t> I'm sure it would, yes. | ||
| [2008/07/14 21:03:17] <groove> without thinking in legacy cfengine terms | ||
| [2008/07/14 21:03:26] <dan___t> Take sections of your cfe config and compare it against what Puppet can do with similar parts. | ||
| [2008/07/14 21:04:04] <groove> if i'm going to do that i may as well just migrate it over piece by piece rather than bothering to write a high-level specification | ||
| [2008/07/14 21:04:20] <dan___t> Look at using UML. | ||
| [2008/07/14 21:04:37] <dan___t> You're going to be going through your cfe configs anyway to see how they work. | ||
| [2008/07/14 21:05:25] <groove> it's going to be a lot of work isn't it | ||
| [2008/07/14 21:05:58] <dan___t> Half the battle is trying to figure out what to make your stuff do. | ||
| [2008/07/14 21:06:07] <dan___t> Sounds like you have an existing solution, and you already know what to do. | ||
| [2008/07/14 21:06:15] <dan___t> If anything, it's going to be less work. | ||
| [2008/07/14 21:06:47] <groove> unless i do it wrong | ||
| [2008/07/14 21:06:57] <groove> that's my main concern, hence trying to find a good book on the subject | ||
| [2008/07/14 21:07:17] <dan___t> Did you write the cfe bits? | ||
| [2008/07/14 21:07:25] <groove> some of it | ||
| [2008/07/14 21:07:31] <dan___t> Then you already *know* how it works. | ||
| [2008/07/14 21:07:43] <groove> i understand our cfengine implementation entirely, but that's not the problem | ||
| [2008/07/14 21:07:58] <groove> i'd like the conversion process to be auditable also to some extent | ||
| [2008/07/14 21:08:19] <dan___t> Then map it out in Visio or UML or something. | ||
| [2008/07/14 21:08:20] <groove> if we can look at the puppet manifests and compare to the original specification we can easily see if we've converted it correctly | ||
| [2008/07/14 21:08:45] <groove> ok i'll have a think about how it would be done in UML | ||
| [2008/07/14 21:08:47] <groove> thanks | ||
| [2008/07/14 21:09:01] @ hessml joined channel #puppet | ||
| [2008/07/14 21:09:28] <dan___t> I don't know, sounds like you're just going to backtrack on a lot of work, spending time on what could have been implementation | ||
| [2008/07/14 21:09:36] <dan___t> That's what leaves me confused. | ||
| [2008/07/14 21:10:42] <groove> i've been reading a lot of books on software design recently so i'm petrified about leaving the "design" phase out | ||
| [2008/07/14 21:14:02] @ Quit: tim|imac: Read error: 110 (Connection timed out) | ||
| [2008/07/14 21:14:28] <fujin> it'd have to be a very high level spec | ||
| [2008/07/14 21:14:31] @ tim|imac joined channel #puppet | ||
| [2008/07/14 21:14:34] <fujin> like before the abstraction phase | ||
| [2008/07/14 21:14:38] <refuseresisted> puppet design is pretty much restricted to a module by module basis imo | ||
| [2008/07/14 21:15:41] <punkcut> somehow puppetd has started throwing "unknown protocol" errors... lame... | ||
| [2008/07/14 21:19:07] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 21:19:54] <groove> fujin: do you know of any such books? | ||
| [2008/07/14 21:20:15] <fujin> no, can't say I do. | ||
| [2008/07/14 21:20:24] <fujin> generally companies formalise on a design structure for manifests/modules | ||
| [2008/07/14 21:20:33] <fujin> and then doc based on that | ||
| [2008/07/14 21:20:40] <groove> a colleague of mine was saying that there aren't any such books, since it's a relatively new field | ||
| [2008/07/14 21:20:53] <fujin> mm, I can't say I'm familiar with any. | ||
| [2008/07/14 21:22:02] <jamesturnbull> groove: what sort of books are you after? | ||
| [2008/07/14 21:22:45] <groove> jamesturnbull: something that is a good reference for creating configuration management specifications, in a high level language | ||
| [2008/07/14 21:22:53] <groove> implementation-free | ||
| [2008/07/14 21:23:18] <jamesturnbull> groove: hmmm | ||
| [2008/07/14 21:23:19] <groove> kind of like the puppet best practice guide but without referring to any particular config management system | ||
| [2008/07/14 21:23:38] <jamesturnbull> groove: doubt there is anything and I know the field pretty well | ||
| [2008/07/14 21:23:47] <jamesturnbull> groove: I'd adapt something from software dev land | ||
| [2008/07/14 21:25:40] * MrProper__ unicomp model m keyboard arrived....sweeeet | ||
| [2008/07/14 21:26:12] <jamesturnbull> groove: UML - perhaps the OMF implementation | ||
| [2008/07/14 21:26:17] <fujin> rofl, seriously MrProper__ ? | ||
| [2008/07/14 21:26:21] <jamesturnbull> s/OMF/OMG/ | ||
| [2008/07/14 21:26:33] <groove> jamesturnbull: i'll look into it, thanks | ||
| [2008/07/14 21:26:35] <MrProper__> fujin, aye clicking awesomeness | ||
| [2008/07/14 21:26:44] <MrProper__> fujin, buckling spring rules | ||
| [2008/07/14 21:26:51] * machpo loves his avant prime kbd | ||
| [2008/07/14 21:27:33] <MrProper__> machpo, nice | ||
| [2008/07/14 21:30:52] <machpo> programmable and indestructable, i just wish they'd make a usb model | ||
| [2008/07/14 21:33:55] @ a-priori joined channel #puppet | ||
| [2008/07/14 21:34:30] @ Quit: sigmonsays: "Leaving" | ||
| [2008/07/14 21:37:33] @ bma joined channel #puppet | ||
| [2008/07/14 21:44:23] @ Quit: a-priori: | ||
| [2008/07/14 21:50:22] @ Quit: shake-n-bake: | ||
| [2008/07/14 21:52:28] @ punkcut left channel #puppet () | ||
| [2008/07/14 21:58:49] @ Quit: Kindred: Read error: 104 (Connection reset by peer) | ||
| [2008/07/14 21:59:22] @ Kindred joined channel #puppet | ||
| [2008/07/14 22:00:40] @ Quit: lak: | ||
| [2008/07/14 22:03:36] <MrProper__> machpo, thats why i went with the unicomp as they also come in black and ps2/usb options | ||
| [2008/07/14 22:04:13] @ a-priori joined channel #puppet | ||
| [2008/07/14 22:04:41] @ Quit: andrewcshafer: | ||
| [2008/07/14 22:05:11] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 22:05:21] @ Quit: andrewcshafer: Client Quit | ||
| [2008/07/14 22:06:06] @ msf joined channel #puppet | ||
| [2008/07/14 22:08:02] @ andrewcshafer joined channel #puppet | ||
| [2008/07/14 22:08:09] @ lak joined channel #puppet | ||
| [2008/07/14 22:08:58] @ Quit: andrewcshafer: Client Quit | ||
| [2008/07/14 22:12:05] @ Quit: lak: Client Quit | ||
| [2008/07/14 22:14:33] @ pdt joined channel #puppet | ||
| [2008/07/14 22:36:47] @ jdputsch joined channel #puppet | ||
| [2008/07/14 22:44:51] @ mcbride joined channel #puppet | ||
| [2008/07/14 22:44:57] @ Quit: mcbride: Client Quit | ||
| [2008/07/14 22:46:52] @ mcbride joined channel #puppet | ||
| [2008/07/14 22:53:09] @ shadoi joined channel #puppet | ||
| [2008/07/14 23:03:07] @ Quit: dysinger: | ||
| [2008/07/14 23:07:00] @ totalchaos joined channel #puppet | ||
| [2008/07/14 23:11:09] @ Demosthenes joined channel #puppet | ||
| [2008/07/14 23:33:24] @ Quit: a-priori: | ||
| [2008/07/14 23:38:31] @ dysinger joined channel #puppet | ||
| [2008/07/14 23:42:09] @ hessml is now known as hessml|away | ||
| [2008/07/14 23:46:32] @ Quit: pdt: | ||
| [2008/07/14 23:55:29] @ Quit: hessml|away: "Leaving..." |
Generated by irclog2html.py 2.6 by Marius Gedminas - find it at mg.pov.lt!