| [2009/04/03 00:01:11] @ Log started by gepetto | ||
| [2009/04/03 00:01:11] @ tuf8_ joined channel #puppet | ||
| [2009/04/03 00:06:00] <PhabX> is it possible to pass multiple values from nodes.pp (hostname,ip,ns,gw) to a template (.erb) ? | ||
| [2009/04/03 00:06:43] <dmosemp> PhabX: a function can take multiple args sure | ||
| [2009/04/03 00:07:59] <PhabX> dmosemp: would that function go into nodes.pp or into the .erb file? | ||
| [2009/04/03 00:08:25] @ Quit: fujin: | ||
| [2009/04/03 00:08:27] <dmosemp> eh I've only used functions in manifests not sure how you do them in templates | ||
| [2009/04/03 00:09:41] <PhabX> oh ok. so if the function is in the manifest and its able to receive args. where do i sent it the args from? from nodes.pp? | ||
| [2009/04/03 00:09:55] @ Deesl joined channel #puppet | ||
| [2009/04/03 00:11:32] <dmosemp> PhabX: qood question, sorry my template experience is limited | ||
| [2009/04/03 00:11:50] <PhabX> maybe i'm asking the wrong questions. thanks for helping | ||
| [2009/04/03 00:12:50] <dmosemp> PhabX: I think once a variable is set you have access to it from a template, like any Facts | ||
| [2009/04/03 00:13:11] <dmosemp> so you don't need to "pass" it variables | ||
| [2009/04/03 00:13:14] @ andrewcshafer joined channel #puppet | ||
| [2009/04/03 00:15:25] <Djelibeybi> PhabX: yes, it is | ||
| [2009/04/03 00:15:41] <Djelibeybi> PhabX: Any variable in scope is available to the .erb file | ||
| [2009/04/03 00:15:46] @ ^authentic joined channel #puppet | ||
| [2009/04/03 00:17:57] @ Quit: lak: | ||
| [2009/04/03 00:18:24] @ Quit: Deesl: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 00:18:53] <PhabX> that actually clarifies a few things thanks | ||
| [2009/04/03 00:18:57] @ Quit: tuf8: Read error: 113 (No route to host) | ||
| [2009/04/03 00:19:01] @ Deesl joined channel #puppet | ||
| [2009/04/03 00:22:49] @ Quit: authentic: Read error: 145 (Connection timed out) | ||
| [2009/04/03 00:22:54] @ ^authentic is now known as authentic | ||
| [2009/04/03 00:26:20] @ Quit: d3vilb0x: | ||
| [2009/04/03 00:31:58] @ Quit: eythian: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 00:32:01] @ eythian_ joined channel #puppet | ||
| [2009/04/03 00:36:52] @ shake-n-bake joined channel #puppet | ||
| [2009/04/03 00:37:00] @ Deesl_ joined channel #puppet | ||
| [2009/04/03 00:37:43] @ Quit: Deesl: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 00:40:52] @ Quit: chiggsy: | ||
| [2009/04/03 00:41:07] @ gstratton joined channel #puppet | ||
| [2009/04/03 00:41:53] @ Quit: Deesl_: Client Quit | ||
| [2009/04/03 00:50:17] @ Quit: PhabX: "Leaving..." | ||
| [2009/04/03 00:58:24] @ Quit: andrewcshafer: | ||
| [2009/04/03 00:59:14] @ andrewcshafer joined channel #puppet | ||
| [2009/04/03 01:07:02] @ Quit: shake-n-bake: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 01:07:04] @ shake-n-bake_ joined channel #puppet | ||
| [2009/04/03 01:07:20] @ Quit: sebykrueger: "http://www.mibbit.com ajax IRC Client" | ||
| [2009/04/03 01:24:35] @ fujin joined channel #puppet | ||
| [2009/04/03 01:31:11] @ Quit: cwebber: | ||
| [2009/04/03 01:31:26] @ jizquierdo joined channel #puppet | ||
| [2009/04/03 01:34:08] @ shake-n-bake joined channel #puppet | ||
| [2009/04/03 01:34:08] @ Quit: shake-n-bake_: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 01:36:49] @ shake-n-bake_ joined channel #puppet | ||
| [2009/04/03 01:36:50] @ Quit: shake-n-bake: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 01:37:56] @ shake-n-bake joined channel #puppet | ||
| [2009/04/03 01:37:56] @ Quit: shake-n-bake_: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 01:39:33] @ shake-n-bake_ joined channel #puppet | ||
| [2009/04/03 01:48:53] <justdave> hmm, is there a way to use a template by default for a config file, but use a static file instead if there's one present for that machine? | ||
| [2009/04/03 01:49:57] <justdave> suppose that'd be a conditional thing of some sort | ||
| [2009/04/03 01:52:37] @ pleemans joined channel #puppet | ||
| [2009/04/03 01:53:29] @ Quit: raphink: Remote closed the connection | ||
| [2009/04/03 01:54:12] @ iTonnerre joined channel #puppet | ||
| [2009/04/03 01:54:18] @ Quit: SyTonnerre: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 01:54:18] @ iTonnerre is now known as SyTonnerre | ||
| [2009/04/03 01:55:25] @ raphink joined channel #puppet | ||
| [2009/04/03 01:55:31] @ ido joined channel #puppet | ||
| [2009/04/03 01:55:45] @ ido left channel #puppet () | ||
| [2009/04/03 01:55:50] @ Quit: shake-n-bake: Read error: 110 (Connection timed out) | ||
| [2009/04/03 02:03:52] @ kibaho1 joined channel #puppet | ||
| [2009/04/03 02:04:10] <gstratton> I've been having problems with getting execs to work. Puppet keeps reporting errors like: | ||
| [2009/04/03 02:04:11] <gstratton> err: //nagios/Exec[set_command_file]/returns: change from notrun to 0 failed: | ||
| [2009/04/03 02:04:24] <gstratton> But with no further output. | ||
| [2009/04/03 02:04:42] <plathrop> gstratton: put logoutput => true in the resource and see if that helps | ||
| [2009/04/03 02:04:53] <plathrop> but likely the command isn't outputting anything | ||
| [2009/04/03 02:04:56] <gstratton> It turned out the command was no on my path, but I'd have expected a better error message... | ||
| [2009/04/03 02:05:02] <gstratton> plathrop: Ah, thanks | ||
| [2009/04/03 02:09:39] <gstratton> Hmm, that makes no difference in my case | ||
| [2009/04/03 02:28:38] @ Quit: mikepea: | ||
| [2009/04/03 02:32:15] @ PsychoSid joined channel #puppet | ||
| [2009/04/03 02:41:12] <gstratton> Is it possible to specify that exported resources depend on other resources? | ||
| [2009/04/03 02:41:41] <gstratton> In this case, nagios services shouldn't be realised until Nagios is installed | ||
| [2009/04/03 02:47:09] @ stijnbe joined channel #puppet | ||
| [2009/04/03 02:48:18] @ kolla joined channel #puppet | ||
| [2009/04/03 02:54:33] @ Quit: fbe: Read error: 110 (Connection timed out) | ||
| [2009/04/03 02:57:07] @ madrescher joined channel #puppet | ||
| [2009/04/03 02:57:45] @ PaulWay left channel #puppet () | ||
| [2009/04/03 02:57:56] @ Quit: punkcut: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 03:00:21] @ Quit: kolla: Remote closed the connection | ||
| [2009/04/03 03:09:04] @ f3ew joined channel #puppet | ||
| [2009/04/03 03:09:45] @ Quit: alban3: Read error: 113 (No route to host) | ||
| [2009/04/03 03:11:41] @ aymerick joined channel #puppet | ||
| [2009/04/03 03:12:54] @ Quit: Djelibeybi: "Leaving" | ||
| [2009/04/03 03:20:20] @ Quit: gstratton: | ||
| [2009/04/03 03:23:28] @ DerekW joined channel #puppet | ||
| [2009/04/03 03:23:39] @ Quit: maxagaz: "Leaving" | ||
| [2009/04/03 03:27:46] @ maxagaz joined channel #puppet | ||
| [2009/04/03 03:33:29] @ Quit: MattyM: "ta ta" | ||
| [2009/04/03 03:36:25] @ ^authentic joined channel #puppet | ||
| [2009/04/03 03:41:22] @ Quit: authentic: Read error: 145 (Connection timed out) | ||
| [2009/04/03 03:41:31] @ ^authentic is now known as authentic | ||
| [2009/04/03 03:41:57] @ Innocenti joined channel #puppet | ||
| [2009/04/03 03:43:07] @ Quit: fujin: | ||
| [2009/04/03 03:56:48] @ stasheck joined channel #puppet | ||
| [2009/04/03 03:58:14] <stasheck> Hi everyone | ||
| [2009/04/03 03:58:17] @ MattyM joined channel #puppet | ||
| [2009/04/03 03:58:25] <stasheck> got a problem on one of my machines | ||
| [2009/04/03 03:58:34] <stasheck> puppetd starts and stalls | ||
| [2009/04/03 03:58:55] <stasheck> trying to do puppetrun on this host results with error 3 "already running" | ||
| [2009/04/03 03:59:30] <stasheck> when running puppetd --test everything is ok, puppet d/ls it's config and applies | ||
| [2009/04/03 03:59:50] <stasheck> but running puppetd --verbose --debug --no-daemonize causes it stall | ||
| [2009/04/03 04:00:32] <stasheck> thats the output: http://pastie.org/435732 | ||
| [2009/04/03 04:03:05] @ Quit: raphink: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 04:04:37] @ fbe joined channel #puppet | ||
| [2009/04/03 04:07:27] @ fujin joined channel #puppet | ||
| [2009/04/03 04:12:58] <score> stasheck: is it really..really slow, or does it actually stop? | ||
| [2009/04/03 04:14:24] <stasheck> it stops | ||
| [2009/04/03 04:14:31] @ Quit: madrescher: "Leaving." | ||
| [2009/04/03 04:14:40] <stasheck> It didn't do nothing for the last month | ||
| [2009/04/03 04:14:41] @ verwilst joined channel #puppet | ||
| [2009/04/03 04:15:19] <stasheck> (or "it did nothing"? What's the proper way of saying this in English? :D) | ||
| [2009/04/03 04:16:00] <stasheck> I spotted this one accidentaly, because all tests seemed to confirm that it's up and running | ||
| [2009/04/03 04:16:08] <kjetilho> stasheck: "didn't do anything" | ||
| [2009/04/03 04:16:14] <stasheck> time to switch to monitoring logfile.... | ||
| [2009/04/03 04:16:42] <stasheck> kjetilho: right, thanks (time for some morning coffee ;-) ) | ||
| [2009/04/03 04:24:07] @ kolla joined channel #puppet | ||
| [2009/04/03 04:24:25] @ mikepea joined channel #puppet | ||
| [2009/04/03 04:25:00] @ keithlard joined channel #puppet | ||
| [2009/04/03 04:27:59] <stasheck> so, no ideas... | ||
| [2009/04/03 04:35:24] @ tim|mb joined channel #puppet | ||
| [2009/04/03 04:36:23] @ Quit: jizquierdo: Read error: 110 (Connection timed out) | ||
| [2009/04/03 04:39:13] @ Quit: andrewcshafer: | ||
| [2009/04/03 04:39:25] @ Quit: mikepea: | ||
| [2009/04/03 04:40:02] @ Quit: SyTonnerre: Success | ||
| [2009/04/03 04:40:04] @ iTonnerre joined channel #puppet | ||
| [2009/04/03 04:40:17] @ iTonnerre is now known as SyTonnerre | ||
| [2009/04/03 04:40:24] @ Quit: fujin: | ||
| [2009/04/03 04:55:53] <f3ew> I am getting an "Invalid parameter provider" error when trying to create a group | ||
| [2009/04/03 04:55:59] @ mikepea joined channel #puppet | ||
| [2009/04/03 04:56:03] <f3ew> Suggestions? | ||
| [2009/04/03 04:59:05] <f3ew> Never mind | ||
| [2009/04/03 05:01:24] @ Quit: rellis_: Read error: 113 (No route to host) | ||
| [2009/04/03 05:10:04] @ madrescher joined channel #puppet | ||
| [2009/04/03 05:12:05] @ rraphink joined channel #puppet | ||
| [2009/04/03 05:13:18] @ rraphink is now known as raphink | ||
| [2009/04/03 05:13:51] @ Quit: keithlard: | ||
| [2009/04/03 05:15:05] <f3ew> I am not at a problem which looks like #948 | ||
| [2009/04/03 05:15:07] <gepetto> f3ew: #948 is http://projects.reductivelabs.com/issues/show/948 "Puppet - Bug #948: err: Could not create root: undefined method `groups' for #<Puppet::Type::User:0xb77e3ef4> - ReductiveLabs.com" | ||
| [2009/04/03 05:15:08] <f3ew> now* | ||
| [2009/04/03 05:15:16] <f3ew> err: Got an uncaught exception of type NoMethodError: undefined method `groups' for #<Puppet::Type::User:0x2aaaab1f3440> | ||
| [2009/04/03 05:15:59] <Baum> sometimes im having the problem that if i remove includes from a node it will be cached somehow even after a restart of the master and client. the include classses are still being processed | ||
| [2009/04/03 05:21:25] @ Quit: sunoano: "Leaving." | ||
| [2009/04/03 05:37:26] @ Quit: grim_radical: Read error: 110 (Connection timed out) | ||
| [2009/04/03 05:54:37] @ sunoano joined channel #puppet | ||
| [2009/04/03 05:54:51] @ keithlard joined channel #puppet | ||
| [2009/04/03 06:16:31] @ d3vilb0x joined channel #puppet | ||
| [2009/04/03 06:23:11] @ ethan_rowe joined channel #puppet | ||
| [2009/04/03 06:33:59] @ Shamgar_ joined channel #puppet | ||
| [2009/04/03 06:33:59] @ Quit: Shamgar: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 06:38:28] <justdave> ok, so here's a funny situation... we had puppet deployed on a bunch of boxes without the initscript enabled, and we've been manually going in and running "puppetd --test" on boxes to pull config changes | ||
| [2009/04/03 06:38:43] <justdave> so now I'm confortable with how it's working and wanting to have the daemon running on all the boxes... | ||
| [2009/04/03 06:39:06] <justdave> figured the easy way is to just put the puppet service in our base class and ensure it's enabled and running | ||
| [2009/04/03 06:39:19] <justdave> next time we run it manually it would fire up the daemon, or so I thought... | ||
| [2009/04/03 06:39:42] <justdave> turns out, /etc/init.d/puppet status says it's running even if it got run from the command line in one-shot mode | ||
| [2009/04/03 06:40:09] <justdave> so the "ensure => running" is essentially a no-op because it *was* running when it tested, but it stopped right after. | ||
| [2009/04/03 06:40:36] <justdave> any ideas how to get around that? :) | ||
| [2009/04/03 06:41:56] <justdave> maybe there's a way I can trigger on the 'enable' state change and exec it? | ||
| [2009/04/03 06:42:03] <Volcane> justdave: on what distro? | ||
| [2009/04/03 06:42:10] <justdave> rhel | ||
| [2009/04/03 06:42:26] <Volcane> justdave: ok, by default, it doesnt call service puppet status, or the init script for status, it does a grep of the ps | ||
| [2009/04/03 06:42:40] <Volcane> justdave: to make it use the status, you need to set hasstatus => true in the service{} | ||
| [2009/04/03 06:42:42] <justdave> ah. hasstatus=>true? | ||
| [2009/04/03 06:43:18] <justdave> and hopefully the status looks for pid file... | ||
| [2009/04/03 06:43:30] <justdave> hmm, nope, no change. | ||
| [2009/04/03 06:43:45] <Volcane> the init script's status should work right | ||
| [2009/04/03 06:44:04] <Volcane> heh, maybe puppetd --test creates the pid too? not sure | ||
| [2009/04/03 06:44:44] <justdave> # service puppet status | ||
| [2009/04/03 06:44:44] <justdave> puppetd is stopped | ||
| [2009/04/03 06:44:50] <justdave> # puppetd --test | ||
| [2009/04/03 06:44:51] <justdave> info: Caching catalog at /var/lib/puppet/localconfig.yaml | ||
| [2009/04/03 06:44:56] <justdave> (meanwhile, from another window) | ||
| [2009/04/03 06:45:04] <justdave> # service puppet status | ||
| [2009/04/03 06:45:04] <justdave> puppetd (pid 10508) is running... | ||
| [2009/04/03 06:45:12] <Volcane> hehe, puppetd must write the pid | ||
| [2009/04/03 06:45:14] <Volcane> that sux | ||
| [2009/04/03 06:46:03] <Volcane> well instead of running it manually 'next time' just start the service? :) | ||
| [2009/04/03 06:46:14] <justdave> yep, guess so. | ||
| [2009/04/03 06:46:32] <Volcane> you could do something weird like have the service subscribe to the config file, and just add a comment to the config | ||
| [2009/04/03 06:46:33] <justdave> or just leave it with the initscript enabled (it does do that part) | ||
| [2009/04/03 06:46:37] <Volcane> then it'll go and restart the service | ||
| [2009/04/03 06:46:43] <justdave> eventually they'll get rebooted for a kernel security patch :) | ||
| [2009/04/03 06:46:51] <Volcane> kewl | ||
| [2009/04/03 06:48:03] <justdave> hmm, there's a way to override the command used to check status I assume... | ||
| [2009/04/03 06:48:24] <Volcane> yeah theres a option | ||
| [2009/04/03 06:48:33] <justdave> "ps ax | grep puppet | grep -v test | grep -v one-shot" | ||
| [2009/04/03 06:48:35] <justdave> or some such | ||
| [2009/04/03 06:48:40] @ Quit: kolla: Read error: 110 (Connection timed out) | ||
| [2009/04/03 06:49:30] <justdave> yeah, 'status => "command to use"' | ||
| [2009/04/03 06:49:41] <justdave> except it doesn't say what kind of output it's expecting | ||
| [2009/04/03 06:49:56] <Volcane> grep's return code should be fine | ||
| [2009/04/03 06:55:19] <justdave> doh, have to grep -v grep, too :) | ||
| [2009/04/03 06:55:26] <justdave> notice: //Node[default]/base/Service[puppet]/ensure: ensure changed 'stopped' to 'running' | ||
| [2009/04/03 06:55:29] <justdave> bingo, there we go | ||
| [2009/04/03 06:55:41] @ gebi_ is now known as gebi | ||
| [2009/04/03 06:55:47] <kjetilho> justdave: use pgrep | ||
| [2009/04/03 06:56:08] <Volcane> wont work for his scenario | ||
| [2009/04/03 06:56:08] <kjetilho> (if you're on Linux, FreeBSD or Solaris) | ||
| [2009/04/03 06:56:20] <Volcane> wants to find all puppetd's not running with --test | ||
| [2009/04/03 06:57:17] <justdave> status => "ps ax | grep puppetd | grep -v grep | grep -v test | grep -v onetime", | ||
| [2009/04/03 06:57:25] <kjetilho> well, OK. I would search for the specific daemon command instead, though | ||
| [2009/04/03 06:58:05] <justdave> yeah, '/usr/bin/ruby /usr/sbin/puppetd' | ||
| [2009/04/03 06:58:13] <Volcane> kjetilho: puppet writes a pid, so /etc/init.d/puppet status thinks the daemon is running even when using --test, so service ensure=> running doesnt start it | ||
| [2009/04/03 06:58:40] @ Quit: d3vilb0x: | ||
| [2009/04/03 07:06:19] @ Quit: keithlard: | ||
| [2009/04/03 07:09:51] @ Deesl joined channel #puppet | ||
| [2009/04/03 07:19:26] <duritong> once somebody had a passenger module for puppet. unfortunately I lost the link for it and I don't remember the person | ||
| [2009/04/03 07:19:35] <duritong> anybody remembers? | ||
| [2009/04/03 07:21:58] <duritong> gepetto: seen DavidS | ||
| [2009/04/03 07:21:58] <gepetto> duritong: DavidS was last seen 4 days, 22 hours, 32 minutes and 6 seconds ago, quitting IRC (Read error: 104 (Connection reset by peer)) | ||
| [2009/04/03 07:33:26] @ rmiller4pi8 joined channel #puppet | ||
| [2009/04/03 07:36:26] @ Quit: omry: Remote closed the connection | ||
| [2009/04/03 07:48:37] @ kolla joined channel #puppet | ||
| [2009/04/03 07:51:58] <gepetto> ::trac:: Recipes/Debian Apache2Recipe edited by ScottTFrazer @ http://reductivelabs.com/trac/puppet/wiki/Recipes/DebianApache2Recipe (by sfrazer@gmail.com) | ||
| [2009/04/03 07:52:15] @ Quit: rmiller4pi81: Read error: 110 (Connection timed out) | ||
| [2009/04/03 07:52:23] <duritong> ah just found andrew's module | ||
| [2009/04/03 07:54:39] @ Quit: mrrx: Read error: 110 (Connection timed out) | ||
| [2009/04/03 07:56:11] @ lludwig joined channel #puppet | ||
| [2009/04/03 07:56:24] @ Quit: lludwig: Client Quit | ||
| [2009/04/03 07:58:36] @ kibaho1 left channel #puppet () | ||
| [2009/04/03 08:00:49] @ mrrx joined channel #puppet | ||
| [2009/04/03 08:04:25] @ fujin joined channel #puppet | ||
| [2009/04/03 08:19:20] @ avleen joined channel #puppet | ||
| [2009/04/03 08:20:20] <avleen> afternoon :) I'm trying to install two packages using puppet (php-cli and php-curl), but puppet insists on installing php-curl before php-cli. as a return mod_php, apache, and a bunch of other stuff gets installed which shouldn't | ||
| [2009/04/03 08:20:27] <avleen> anyone know how i could work around this? | ||
| [2009/04/03 08:23:11] <f3ew> avleen make an explicit dependency with Require | ||
| [2009/04/03 08:23:48] <avleen> ahh! ok. i'll do that, ty :) | ||
| [2009/04/03 08:24:39] @ Quit: mrrx: Read error: 110 (Connection timed out) | ||
| [2009/04/03 08:29:04] <avleen> cheers that worked great | ||
| [2009/04/03 08:37:19] @ kibahop joined channel #puppet | ||
| [2009/04/03 08:37:19] @ thegcat joined channel #puppet | ||
| [2009/04/03 08:48:11] @ chiggsy joined channel #puppet | ||
| [2009/04/03 08:51:40] <gepetto> ::trac:: Installation Guide edited by timrosenblatt @ http://reductivelabs.com/trac/puppet/wiki/InstallationGuide (by tim@timrosenblatt.com) | ||
| [2009/04/03 09:00:33] @ kibahop left channel #puppet () | ||
| [2009/04/03 09:08:59] @ kambiz_away is now known as kambiz | ||
| [2009/04/03 09:09:00] @ johnw joined channel #puppet | ||
| [2009/04/03 09:10:05] @ Quit: chiggsy: | ||
| [2009/04/03 09:10:55] <johnw> ergh, my puppetmaster is taking 443M of virtual memory | ||
| [2009/04/03 09:11:07] <johnw> what in the world could it even be doing with all that mem? virtsz is 1.2G | ||
| [2009/04/03 09:11:43] @ Quit: maxagaz: Read error: 113 (No route to host) | ||
| [2009/04/03 09:13:04] @ iTonnerre joined channel #puppet | ||
| [2009/04/03 09:14:35] @ Quit: SyTonnerre: Read error: 60 (Operation timed out) | ||
| [2009/04/03 09:14:35] @ iTonnerre is now known as SyTonnerre | ||
| [2009/04/03 09:18:28] @ masterzen joined channel #puppet | ||
| [2009/04/03 09:22:34] @ nakano_ is now known as nakano | ||
| [2009/04/03 09:24:53] @ Quit: stasheck: Remote closed the connection | ||
| [2009/04/03 09:26:01] <johnw> ok, so I have a manifest that will consume memory without bounds | ||
| [2009/04/03 09:27:13] <jenza> With great power comes high memory utilisation | ||
| [2009/04/03 09:27:27] <johnw> well, until the OOM comes along | ||
| [2009/04/03 09:29:29] <johnw> i'm trying to reduce it down to a test cas enow | ||
| [2009/04/03 09:29:35] <johnw> because the manifest that causes it is very simple I thought | ||
| [2009/04/03 09:30:41] @ d3vilb0x joined channel #puppet | ||
| [2009/04/03 09:32:36] <johnw> so, am I right in assuming that transferring large files via puppet file repositories is just a bad idea all around? | ||
| [2009/04/03 09:33:25] <johnw> my puppetmaster is at 90% of total memory now, all because I tried to serve the JDK | ||
| [2009/04/03 09:33:37] <f3ew> johnw yes | ||
| [2009/04/03 09:33:52] <johnw> and it's not giving the memory back either | ||
| [2009/04/03 09:33:58] <johnw> ok | ||
| [2009/04/03 09:34:10] <f3ew> johnw convert it into a regular package, or pull via HTTP and an exec | ||
| [2009/04/03 09:34:21] <johnw> a regular package? | ||
| [2009/04/03 09:34:47] <int> hello! after upgrade to 0.24.8 puppet starts to install ssh keys on every run, anyone saw that too? | ||
| [2009/04/03 09:34:56] <johnw> also, the binary is 64M, any reason why puppet needs 500M of memory to try and transfer that? | ||
| [2009/04/03 09:37:28] <int> johnw: i guess it's not send it in optimal way | ||
| [2009/04/03 09:37:33] <johnw> haha | ||
| [2009/04/03 09:37:42] <f3ew> johnw XML-RPC | ||
| [2009/04/03 09:37:54] <int> johnw: make package (.deb or .rpm) put in local repository and install via puppet | ||
| [2009/04/03 09:37:55] <f3ew> Lots of base64 encoded small packets | ||
| [2009/04/03 09:38:01] <johnw> int: ah, thx | ||
| [2009/04/03 09:38:04] <int> # HEADER: This file was autogenerated at Fri Apr 03 17:36:04 +0400 2009 | ||
| [2009/04/03 09:38:22] <int> ^^^ wtf? it updates it on every run | ||
| [2009/04/03 09:43:56] @ mrrx joined channel #puppet | ||
| [2009/04/03 09:45:53] @ Quit: aymerick: "kit mais sage" | ||
| [2009/04/03 09:46:22] @ aymerick joined channel #puppet | ||
| [2009/04/03 09:52:45] @ cwebber joined channel #puppet | ||
| [2009/04/03 09:57:40] @ briandquinn joined channel #puppet | ||
| [2009/04/03 09:59:40] <Dyresen> is there a timeout value in facter? | ||
| [2009/04/03 10:00:11] <Dyresen> Or to pitch it different, how quick do I need to return custom facts? | ||
| [2009/04/03 10:00:12] <joe-mac> int: puppet' | ||
| [2009/04/03 10:00:17] <joe-mac> i mean | ||
| [2009/04/03 10:00:19] <joe-mac> johnw: | ||
| [2009/04/03 10:00:46] <joe-mac> puppet's file serving capabilities suck ATM, but afaik lak and crew area ddressing it | ||
| [2009/04/03 10:00:52] <johnw> good to know | ||
| [2009/04/03 10:00:56] <johnw> at the moment i'll use vsftpd +curl | ||
| [2009/04/03 10:00:57] <joe-mac> by suck i mean really slow and eat tons of ram | ||
| [2009/04/03 10:01:44] <joe-mac> i made a definition for a source package and it worked great i could push out source and do upgrades easily to a couple of apckages we use custom compuiles of and it works great.... except the manifest took 10 minutes to complete serving out all the .c files lol | ||
| [2009/04/03 10:02:02] @ alban2 joined channel #puppet | ||
| [2009/04/03 10:04:29] @ PsychoSid1 joined channel #puppet | ||
| [2009/04/03 10:04:31] <cwebber> does this error make sense to any1 when doing a puppet run? Certificates were not trusted: SSL_connect returned=1 errno=0 state=SSLv3 read finished A: sslv3 alert bad record mac | ||
| [2009/04/03 10:05:24] * int had to patch puppet to remove Time.now from header of autogenerated files :( | ||
| [2009/04/03 10:05:28] @ Quit: rmiller4pi8: "Leaving." | ||
| [2009/04/03 10:05:55] @ Deesl left channel #puppet () | ||
| [2009/04/03 10:07:03] @ MrHeavy joined channel #puppet | ||
| [2009/04/03 10:10:11] @ Quit: PsychoSid1: "Leaving." | ||
| [2009/04/03 10:14:49] @ markl___ joined channel #puppet | ||
| [2009/04/03 10:15:41] @ Quit: PsychoSid: Read error: 113 (No route to host) | ||
| [2009/04/03 10:16:46] @ d3vilb0x_ joined channel #puppet | ||
| [2009/04/03 10:17:16] @ Quit: rsquared: "Leaving" | ||
| [2009/04/03 10:25:58] @ andrewcshafer joined channel #puppet | ||
| [2009/04/03 10:25:59] @ flakrat joined channel #puppet | ||
| [2009/04/03 10:32:15] @ Quit: d3vilb0x: Read error: 110 (Connection timed out) | ||
| [2009/04/03 10:36:24] <tim|mb> hm... how does the virtual fact work? shouldn't it tell me if a machine is a domU instead of a dom0? | ||
| [2009/04/03 10:36:42] <Volcane> use the source luke | ||
| [2009/04/03 10:36:43] <tim|mb> I'm getting virtual => physical in both the dom0 and the domU... | ||
| [2009/04/03 10:37:05] <Volcane> all kinds of tests, see /usr/lib/ruby/site_ruby/1.8/facter/virtual.rb | ||
| [2009/04/03 10:37:39] @ d3vilb0x joined channel #puppet | ||
| [2009/04/03 10:37:48] <tim|mb> ah got it, only root can access the stuff it needs to check :) | ||
| [2009/04/03 10:39:32] @ rmiller4pi8 joined channel #puppet | ||
| [2009/04/03 10:39:32] <int> hmm, it still replaces authorized_keys on every run | ||
| [2009/04/03 10:42:26] <joe-mac> int: are you using the authorized_keys type or serving out a file? what version of puppet? | ||
| [2009/04/03 10:43:07] <int> joe-mac: ssh_authorized_key 0.24.8 (it worked ok with 0.24.7) | ||
| [2009/04/03 10:44:07] @ jief- joined channel #puppet | ||
| [2009/04/03 10:44:38] <int> joe-mac: key => template("keys.erb") (if that's important) | ||
| [2009/04/03 10:44:43] <jief-> hello. there's not much info on the features of puppetshow. what can i expect with it? | ||
| [2009/04/03 10:45:06] <Volcane> headaches mostly seems to be the concensus :P | ||
| [2009/04/03 10:45:15] <jief-> lol | ||
| [2009/04/03 10:45:35] <jief-> so what's a good webGUI to install alongside Puppet then? | ||
| [2009/04/03 10:45:39] <jief-> if there's any? | ||
| [2009/04/03 10:45:49] <jief-> my storeconfigs are in psql | ||
| [2009/04/03 10:45:51] <int> joe-mac: template is <% if name == "user1" -%>123<% elsif name == "user12" -%>23535<% end -%> | ||
| [2009/04/03 10:47:12] @ nickv joined channel #puppet | ||
| [2009/04/03 10:47:42] <nickv> Hey all, puppet appears to be working on my boxes, but i'm having a firewall issue i think with some boxes......i know that the clients talk to the puppet master on TCP port 8140 | ||
| [2009/04/03 10:48:03] <nickv> what other additional ports need to be opened for proper communication? I am getting ssl_connect errors | ||
| [2009/04/03 10:48:09] <nickv> on a few servers | ||
| [2009/04/03 10:48:35] <Volcane> only 8140 | ||
| [2009/04/03 10:48:47] <Volcane> unless u set special things, u can split the ca from the server etc | ||
| [2009/04/03 10:49:11] @ fbe_ joined channel #puppet | ||
| [2009/04/03 10:49:23] <Volcane> ssl errors though generally suggest tcp comms are working already | ||
| [2009/04/03 10:49:25] <nickv> hmm...i got the cert on my puppetmaster server and signed it......but I am getting a bunch of ssl_connect untrusted cert errors in log, i cleaned out the certs on the puppetmaster, uninstalled puppet and deleted the ssl dir and started form scratch, same thing | ||
| [2009/04/03 10:49:35] <jief-> oh wth, its friday, lets give puppetshow a spin | ||
| [2009/04/03 10:49:36] <Volcane> but that things werent validated/authorized properly | ||
| [2009/04/03 10:49:51] @ Quit: fbe: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 10:50:05] <Volcane> nickv: not fireall then | ||
| [2009/04/03 10:50:29] <nickv> Volcane: How else can i troubleshoot this? why isn't my cert trusted even if i signed it? | ||
| [2009/04/03 10:50:35] <thegcat> anyone knows the naming conventions and restrictions for facter facts? | ||
| [2009/04/03 10:50:44] <Volcane> nickv: show the error pls | ||
| [2009/04/03 10:51:04] <Volcane> thegcat: a-z_ is a safe bet | ||
| [2009/04/03 10:51:11] <thegcat> ok, thanks | ||
| [2009/04/03 10:52:04] @ Quit: madrescher: "Leaving." | ||
| [2009/04/03 10:52:06] <Volcane> nickv: generally, if you specify 'server' or anything like that to point puppetd to a specific box, u need specify a matching 'certname' on the master | ||
| [2009/04/03 10:52:10] <nickv> http://www.mibbit.com/pb/SgVDXw | ||
| [2009/04/03 10:52:28] <Volcane> nickv: which is why the suggestion is to just leave things as default and add cnames for 'puppet' to dns | ||
| [2009/04/03 10:52:51] <Volcane> I'd say you are using a specific hostname in your puppet:// style file retrievals | ||
| [2009/04/03 10:52:59] <Volcane> instead of what the master's ssl is configured to use | ||
| [2009/04/03 10:53:44] <nickv> Volcane: I have puppet.domain.com in my puppet.conf file.....but i have puppet.domain.com in ALL my servers hosts files regardless | ||
| [2009/04/03 10:53:50] <nickv> nscd is disabled also btw | ||
| [2009/04/03 10:53:55] @ wassy121 joined channel #puppet | ||
| [2009/04/03 10:54:05] <Volcane> do you have certname to puppet.domain.com ? | ||
| [2009/04/03 10:54:17] <wassy121> Morning all. | ||
| [2009/04/03 10:54:17] <nickv> Volcane: on the client? | ||
| [2009/04/03 10:54:22] <Volcane> master | ||
| [2009/04/03 10:54:24] @ Quit: d3vilb0x_: Read error: 110 (Connection timed out) | ||
| [2009/04/03 10:54:35] <Volcane> without it, the master will have a cert called 'puppet' | ||
| [2009/04/03 10:54:40] <int> it says: target changed '/home/user/.ssh/authorized_keys' to '/home/user/.ssh/authorized_keys' - but it doesn't changed! | ||
| [2009/04/03 10:54:42] <Volcane> if you then try to speak to 'puppet.domain.com' | ||
| [2009/04/03 10:54:51] <Volcane> puppet != puppet.domain.com | ||
| [2009/04/03 10:55:09] <nickv> Volcane: where on the paster master should I see if I have that puppet.domain.com cert? | ||
| [2009/04/03 10:55:26] <nickv> it works fine for like 70 other systems just these 3 are having issues, ill look now for it | ||
| [2009/04/03 10:55:42] <Volcane> ah, and you're certain they all have the same configs? | ||
| [2009/04/03 10:55:51] <Volcane> then it could be down to things like clocks not being synced | ||
| [2009/04/03 10:56:06] <nickv> yes...where can i check for the cert on the master? | ||
| [2009/04/03 10:56:12] @ lak joined channel #puppet | ||
| [2009/04/03 10:56:23] <Volcane> if it works for the rest i think you're ok on the certname front | ||
| [2009/04/03 10:56:30] <Volcane> check your clocks | ||
| [2009/04/03 10:56:52] <nickv> wow | ||
| [2009/04/03 10:56:52] <nickv> Wed Apr 3 10:56:39 EDT 2030 | ||
| [2009/04/03 10:57:00] <Volcane> nice | ||
| [2009/04/03 10:57:13] <nickv> haha 10:56am up 7836 day(s), 34 min(s), 2 users, load average: 0.02, 0.02, 0.02 | ||
| [2009/04/03 10:57:14] <Volcane> so you'll need to fix that, clear the old ssl certs and make new ones for these machines | ||
| [2009/04/03 10:57:41] <nickv> ok got it, lemme take a look at this | ||
| [2009/04/03 10:57:44] <nickv> thanks for all your help! | ||
| [2009/04/03 10:57:49] <Volcane> np | ||
| [2009/04/03 10:58:24] <wassy121> heh. 2030. No wonder the certs weren't passing verification. | ||
| [2009/04/03 10:58:55] @ PsychoSid joined channel #puppet | ||
| [2009/04/03 11:00:21] <wassy121> Is there a way to make the "yum" package source use "--enablerepo=BLAH" ? | ||
| [2009/04/03 11:00:31] @ Quit: PsychoSid: Client Quit | ||
| [2009/04/03 11:00:32] <jief-> nickv: you don't patch often | ||
| [2009/04/03 11:00:56] <nickv> OpenSolaris you dont patch, you install new builds | ||
| [2009/04/03 11:01:38] <jief-> no kernel change in close to 8000 days? | ||
| [2009/04/03 11:02:30] <nickv> hahaha ntp is off :-) | ||
| [2009/04/03 11:02:36] <nickv> which is my problem with puppet certs | ||
| [2009/04/03 11:02:44] <nickv> Wed Apr 3 10:56:39 EDT 2030 | ||
| [2009/04/03 11:02:47] <nickv> its not 2030 :-) | ||
| [2009/04/03 11:02:50] <nickv> few years off | ||
| [2009/04/03 11:02:57] <nickv> not sure how that happened | ||
| [2009/04/03 11:03:16] <tim|mb> smurfs! | ||
| [2009/04/03 11:03:16] <jief-> it is in my parallel universe | ||
| [2009/04/03 11:03:45] <jief-> git://reductivelabs.com/puppetshow | ||
| [2009/04/03 11:06:12] <wassy121> bah - enablerepo options - http://projects.reductivelabs.com/issues/915 | ||
| [2009/04/03 11:06:21] <wassy121> 0% complete :-/ | ||
| [2009/04/03 11:08:03] <tim|mb> wassy121: create a patch and submit it? :) | ||
| [2009/04/03 11:08:19] <wassy121> tim|mb: Ah, but if I could . . . | ||
| [2009/04/03 11:08:35] <wassy121> I know enough python to debug stuff, writing new patches is something totally different ;-) | ||
| [2009/04/03 11:08:48] <Volcane> and this isnt even python :P | ||
| [2009/04/03 11:09:46] <wassy121> ./fail | ||
| [2009/04/03 11:10:13] <wassy121> Ruby, python, perl. I know its one of those "interpreted" languages. | ||
| [2009/04/03 11:11:27] @ mccune joined channel #puppet | ||
| [2009/04/03 11:12:32] <int> joe-mac: it was b/c http://projects.reductivelabs.com/issues/2124 | ||
| [2009/04/03 11:13:45] @ freak4u joined channel #puppet | ||
| [2009/04/03 11:14:33] <jief-> python's the best ;) | ||
| [2009/04/03 11:17:36] @ Quit: tim|mb: | ||
| [2009/04/03 11:21:30] * int wishes to have puppet in python | ||
| [2009/04/03 11:22:36] <jief-> indeed | ||
| [2009/04/03 11:22:43] <jief-> although, ive been writing a few custom facts | ||
| [2009/04/03 11:22:46] <jief-> ruby's not that bad | ||
| [2009/04/03 11:23:31] <nickv> anyone familiar with solaris ntp? having issues forcing the server to get the time from our ntp box | ||
| [2009/04/03 11:23:51] <jief-> not familiar enough with it to help troubleshoot something | ||
| [2009/04/03 11:24:22] <nickv> hmm...some people say it takes 5min | ||
| [2009/04/03 11:26:05] <kjetilho> yes. check with ntpq peers | ||
| [2009/04/03 11:26:30] <kjetilho> it won't get a lock before 5-10 minutes, then ntptrace will give a proper answer too | ||
| [2009/04/03 11:26:37] <Volcane> if its that big a difference ntp wont fix it | ||
| [2009/04/03 11:26:46] <Volcane> use ntpdate manually once to fix it then ntp to keep it fixed | ||
| [2009/04/03 11:28:18] @ sigmonsays joined channel #puppet | ||
| [2009/04/03 11:28:25] <jief-> Volcane: it will, but over a very long delta | ||
| [2009/04/03 11:28:35] <jief-> and with this much difference, its gonna take years to get back in sync | ||
| [2009/04/03 11:28:38] <kjetilho> no, it won't | ||
| [2009/04/03 11:28:39] <Volcane> yes, fixing a 25 year time skew will take som time :P | ||
| [2009/04/03 11:28:56] <Volcane> you'd probably find the hardware will die first though | ||
| [2009/04/03 11:29:02] <sigmonsays> hehe | ||
| [2009/04/03 11:29:07] <jief-> that'd be a nice experiment | ||
| [2009/04/03 11:29:53] <joe-mac> puppet is good, but i find that ruby is a notch above | ||
| [2009/04/03 11:30:08] @ Quit: verwilst: "Ex-Chat" | ||
| [2009/04/03 11:30:18] <joe-mac> granted, that's from an SA perspective, i don't do any 'heavy-lifting' in the programming area with those languages | ||
| [2009/04/03 11:31:00] <sigmonsays> joe-mac, SA ? | ||
| [2009/04/03 11:31:08] <bda> bah. Time to refactor my classes. Such a mess. | ||
| [2009/04/03 11:31:26] <sigmonsays> joe-mac, systems architect? | ||
| [2009/04/03 11:31:59] @ Quit: d3vilb0x: Remote closed the connection | ||
| [2009/04/03 11:32:12] @ d3vilb0x joined channel #puppet | ||
| [2009/04/03 11:32:12] <jief-> sysadmin probably | ||
| [2009/04/03 11:32:34] <jief-> http://reductivelabs.com/trac/puppet/wiki/PuppetShow | ||
| [2009/04/03 11:32:37] @ alfism joined channel #puppet | ||
| [2009/04/03 11:32:38] <jief-> that's all the doc available for it? | ||
| [2009/04/03 11:37:11] @ Quit: Innocenti: Client Quit | ||
| [2009/04/03 11:38:17] <johnw> what the biggest reason to use stored configs, other than running something like puppet show? | ||
| [2009/04/03 11:38:38] @ jmarki joined channel #puppet | ||
| [2009/04/03 11:38:52] <Volcane> johnw: http://reductivelabs.com/trac/puppet/wiki/UsingStoredConfiguration | ||
| [2009/04/03 11:39:07] <Volcane> erm, sorry | ||
| [2009/04/03 11:39:09] <Volcane> bad link | ||
| [2009/04/03 11:39:20] @ unxfrek joined channel #puppet | ||
| [2009/04/03 11:39:26] <Volcane> http://reductivelabs.com/trac/puppet/wiki/ExportedResources thats what i was looking for | ||
| [2009/04/03 11:39:32] <johnw> thx | ||
| [2009/04/03 11:39:41] <Volcane> thats only possible with stored confs | ||
| [2009/04/03 11:41:31] <jief-> hmm puppetshow seems to assume it's running where puppetmasterd is running | ||
| [2009/04/03 11:41:33] @ Quit: stijnbe: "Leaving..." | ||
| [2009/04/03 11:42:35] <sigmonsays> is it safe to kill -TERM puppetd ? | ||
| [2009/04/03 11:42:40] <sigmonsays> i'm seeing it hang on yumhelper | ||
| [2009/04/03 11:42:50] <sigmonsays> and I don't know why! | ||
| [2009/04/03 11:43:05] <Volcane> happens if theres problems with your rpm locks | ||
| [2009/04/03 11:43:11] <Volcane> but yeah kill em no harm | ||
| [2009/04/03 11:43:20] <sigmonsays> well it's doing a call recvfrom on some network socket | ||
| [2009/04/03 11:44:29] <Volcane> i had it happen all the time when i had shitty repos like epel and rpforge enabled | ||
| [2009/04/03 11:44:33] <Volcane> cos they're just not reliable | ||
| [2009/04/03 11:44:43] <Volcane> and when things like mirror.centos.org dns thing goes titsup | ||
| [2009/04/03 11:44:53] <sigmonsays> yah, that is probably it, even though all my repos should eb local mirrors I bet someone is adding hteir own :) | ||
| [2009/04/03 11:44:55] <rmiller4pi8> Volcane: just a heads up, epel has stabilized nicely, but rpmforge is still down a lot so i mirror that myself | ||
| [2009/04/03 11:45:20] <Volcane> yeah its not bad, still, best to just yumdownload whatever and pop in local | ||
| [2009/04/03 11:47:39] <Volcane> plus I'd rather carefully check what they include | ||
| [2009/04/03 11:47:51] <Volcane> since i dont want random stuff to just arrive, they seem to be a bit loose and fast with requires | ||
| [2009/04/03 11:50:31] @ Quit: jmarki: "Leaving" | ||
| [2009/04/03 11:50:58] @ jmarki joined channel #puppet | ||
| [2009/04/03 11:51:23] @ n0ts___ joined channel #puppet | ||
| [2009/04/03 11:53:42] <z00dax> under what circumstances might puppet's run end without it outputting the ususal 'notice: Finished catalog run in <blah> seconds' ? | ||
| [2009/04/03 11:54:38] @ webx joined channel #puppet | ||
| [2009/04/03 11:55:20] @ Quit: thegcat: | ||
| [2009/04/03 11:55:25] @ Quit: mrrx: Read error: 110 (Connection timed out) | ||
| [2009/04/03 11:55:30] <webx> is there a way to install the ruby-shadow package and change a password string of a user without requiring 2 puppet runs? | ||
| [2009/04/03 11:55:43] <Volcane> z00dax: when run in the foreground or the daemon? | ||
| [2009/04/03 11:55:56] <Volcane> webx: doubtit | ||
| [2009/04/03 11:56:12] <webx> seems like I can't because there is no provider for the encrypted password when the manifest is created | ||
| [2009/04/03 11:56:18] <webx> yea, same feeling I get | ||
| [2009/04/03 11:56:25] <webx> I tried require and before | ||
| [2009/04/03 11:56:26] <webx> no love | ||
| [2009/04/03 11:56:53] <Volcane> it tires to load the shadow library early on, if it cant find it, it wont try to set passwords that run | ||
| [2009/04/03 11:56:58] <webx> yea | ||
| [2009/04/03 11:57:00] <webx> fek | ||
| [2009/04/03 11:57:05] <webx> thanks | ||
| [2009/04/03 11:57:17] @ herlo joined channel #puppet | ||
| [2009/04/03 11:57:17] <z00dax> Volcane: foreground only runs here | ||
| [2009/04/03 11:57:58] <Volcane> so it just exits basically? | ||
| [2009/04/03 11:58:02] @ Quit: nickv: "http://www.mibbit.com ajax IRC Client" | ||
| [2009/04/03 12:00:05] <Volcane> I've seen it get stuck and never return, but not just exit like that, not for ages anyway, maybe in the 0.23.x days last | ||
| [2009/04/03 12:00:12] <Volcane> but running with --verbose --trace might help | ||
| [2009/04/03 12:03:01] @ Quit: silk: Remote closed the connection | ||
| [2009/04/03 12:03:22] @ Quit: n0ts__: Read error: 110 (Connection timed out) | ||
| [2009/04/03 12:05:40] @ chiggsy joined channel #puppet | ||
| [2009/04/03 12:05:45] @ Quit: shake-n-bake_: | ||
| [2009/04/03 12:07:48] <z00dax> Volcane: this is 0.24.*; and its run 5 times in a loop to get things right. randomly some of the runs complete withut that at the end | ||
| [2009/04/03 12:07:57] <z00dax> it does look like its run through whatever is needed | ||
| [2009/04/03 12:09:15] <wassy121> I am having problems with puppetbucket.addfile, it fails the cert validation check. However, everything else runs fine with that certificate. What's the deal? (0.24.7) | ||
| [2009/04/03 12:09:18] <Volcane> thats pretty odd | ||
| [2009/04/03 12:09:31] <Volcane> wassy121: how do you config the bucket/ | ||
| [2009/04/03 12:09:56] <Volcane> z00dax: worth running in verbose and trace and try to see if theres any hints thats about the best advice | ||
| [2009/04/03 12:10:04] @ Quit: andrewcshafer: Read error: 110 (Connection timed out) | ||
| [2009/04/03 12:14:10] @ herlo left channel #puppet () | ||
| [2009/04/03 12:14:53] <wassy121> http://pastebin.ca/1381012 - I just changed the "enabled" flag in a repo file from 0 to 1. | ||
| [2009/04/03 12:15:51] <Volcane> wassy121: but where do you configure the file bucket? | ||
| [2009/04/03 12:16:03] <Volcane> wassy121: do you use the same hostnames as you do for your puppet host? or some other name? | ||
| [2009/04/03 12:16:41] @ Quit: lak: | ||
| [2009/04/03 12:16:49] <wassy121> volcane: blast, you are right. The filebucket has an incorrect hostname. | ||
| [2009/04/03 12:17:00] <Volcane> :) | ||
| [2009/04/03 12:17:36] @ Quit: MattyM: "ta ta" | ||
| [2009/04/03 12:18:00] <wassy121> I migrated between puppet servers recently, missed that mentioning of the hostname ;-) | ||
| [2009/04/03 12:18:33] @ lak joined channel #puppet | ||
| [2009/04/03 12:19:07] <Volcane> wassy121: once again proving that just sticking to the default 'puppet' server/certname is the best idea :P | ||
| [2009/04/03 12:19:52] <wassy121> Volcane: then adding an /etc/hosts entry? | ||
| [2009/04/03 12:20:03] <Volcane> dns? | ||
| [2009/04/03 12:20:25] @ Quit: masterzen: "Leaving" | ||
| [2009/04/03 12:20:26] <Volcane> puppet.your.com IN CNAME whatever | ||
| [2009/04/03 12:20:27] <wassy121> Wouldn't that resolve it to puppet.$DOMAINNAME ? | ||
| [2009/04/03 12:20:28] <webx> without manually viewing the nodes.pp, is there a way to show which classes a particular node is loading ? | ||
| [2009/04/03 12:20:54] <sigmonsays> webx, /var/lib/puppet/classes.txt | ||
| [2009/04/03 12:21:01] <sigmonsays> on the node | ||
| [2009/04/03 12:21:23] <Volcane> wassy121: by default puppetd will try to talk to a host called 'puppet' however u make that resolve is fine, but it takes away all these problems of changing to a new/better/faster server and then having to redo all certs for example cos its hostnames changed | ||
| [2009/04/03 12:21:48] @ shake-n-bake joined channel #puppet | ||
| [2009/04/03 12:21:59] <wassy121> Volcane: Ah, I see the logic there. I have just never set it up like that. | ||
| [2009/04/03 12:22:28] <webx> sigmonsays: ah, /var/puppet/state/classes.txt it looks like | ||
| [2009/04/03 12:22:29] <webx> thanks | ||
| [2009/04/03 12:22:49] <sigmonsays> webx, cool | ||
| [2009/04/03 12:23:08] @ Quit: briandquinn: | ||
| [2009/04/03 12:23:30] @ kibahop joined channel #puppet | ||
| [2009/04/03 12:23:49] @ kibahop left channel #puppet () | ||
| [2009/04/03 12:24:58] @ chrismat_ joined channel #puppet | ||
| [2009/04/03 12:25:42] <Volcane> wassy121: when done that way you dont need to change any certnames or anything | ||
| [2009/04/03 12:26:33] <chrismat_> I am a bit confused by the number of virt projects from RH, like there are Genome, virt-factory, spacewalk and func | ||
| [2009/04/03 12:26:52] <chrismat_> or more system managment projects | ||
| [2009/04/03 12:26:57] <Volcane> at least 2 of those arent virt packages :P | ||
| [2009/04/03 12:26:58] @ jmeeuwen joined channel #puppet | ||
| [2009/04/03 12:27:12] <Volcane> actually 3 | ||
| [2009/04/03 12:27:26] <Volcane> they just happen to have functionally to also manage virt machines, since well, they should be just like other machines | ||
| [2009/04/03 12:27:35] <chrismat_> I do not have a infrastructure for this in place today | ||
| [2009/04/03 12:28:05] <Volcane> genome is like a big architecture thing, includes puppet and all sorts | ||
| [2009/04/03 12:28:06] <chrismat_> which should I bet on, I manage both physicall hardware and virtual machines and would like Puppet to mange them somehow | ||
| [2009/04/03 12:28:12] <Volcane> func is for running adhoc commands kind of | ||
| [2009/04/03 12:28:24] <Volcane> spacewalk manages updates and so forth, like RHN does | ||
| [2009/04/03 12:28:30] <Volcane> there's also ovirt | ||
| [2009/04/03 12:28:57] <Volcane> most of these are like emerging products, i dont think they really know what will be the choice to make it to enterprise customers :( | ||
| [2009/04/03 12:29:28] @ Quit: DerekW: "Leaving" | ||
| [2009/04/03 12:29:36] <chrismat_> erm that explains it | ||
| [2009/04/03 12:29:54] <Volcane> spacewalk is the old redhat satelite server | ||
| [2009/04/03 12:30:15] <chrismat_> Cause I would need something, but putting development effort into something which will not make it in the long run seems kinda like wasted time | ||
| [2009/04/03 12:30:23] <Volcane> and come RH 6 they're abandoning Xen in favour for kvm | ||
| [2009/04/03 12:30:25] <Volcane> its all in flux | ||
| [2009/04/03 12:31:09] @ sc0ttB joined channel #puppet | ||
| [2009/04/03 12:31:25] <Volcane> though they at least use libvirt and the libvirt tools to manage running vms | ||
| [2009/04/03 12:31:33] <Volcane> so a lot wont change when they rip out Xen for KVM | ||
| [2009/04/03 12:32:04] <Volcane> personally i dont use any of those, i just let puppet make me LVM volumes, configs, kickstarts and then use anaconda to install a VM based on puppet settings | ||
| [2009/04/03 12:32:13] <Volcane> from there on its like any other machine, virtual of physical, no diff | ||
| [2009/04/03 12:32:18] <Volcane> but i dont do like live migrates and stuff | ||
| [2009/04/03 12:33:06] <Volcane> so creating VMs and the configs for them is using the normal RHEL documented libvirt based tools | ||
| [2009/04/03 12:33:46] @ Quit: mikepea: | ||
| [2009/04/03 12:33:49] <chrismat_> hmm | ||
| [2009/04/03 12:34:22] <Volcane> guess it depends on your needs | ||
| [2009/04/03 12:34:50] <chrismat_> I need something which says a server is webserver or a calculation server | ||
| [2009/04/03 12:34:55] @ Quit: alfism: "Connection reset by beer" | ||
| [2009/04/03 12:35:06] <chrismat_> and it would be nice to have something which automates reinstalls | ||
| [2009/04/03 12:35:18] <Volcane> did you look at cobbler? | ||
| [2009/04/03 12:35:51] <Volcane> its a fancy kickstarter/repo manager/rebuilder | ||
| [2009/04/03 12:36:17] <Volcane> has some rudimentary puppet support to make the right manifests go to a machine based on what type was chosen | ||
| [2009/04/03 12:36:22] <Volcane> its fugly, but quite nice | ||
| [2009/04/03 12:36:31] <chrismat_> Thanks a lot Volcane! | ||
| [2009/04/03 12:36:47] <chrismat_> I think I will investigate the Cobbler+Puppet path | ||
| [2009/04/03 12:36:48] <Volcane> else its easy to code your own | ||
| [2009/04/03 12:36:58] <Volcane> i mean if you break the problem in 2 | ||
| [2009/04/03 12:37:04] * sc0ttB uses cobbler and hearts it | ||
| [2009/04/03 12:37:14] <sc0ttB> can't wait for better deb support | ||
| [2009/04/03 12:37:17] <Volcane> 1 i installing the machines, make them all the same to start with apart from differences forced due to hardware differences | ||
| [2009/04/03 12:37:37] <Volcane> and then the 2nd part of the problem is taking a "standard" built machine and turning it into a web/calculate/whatever node | ||
| [2009/04/03 12:37:40] <Volcane> thats where puppet comes in | ||
| [2009/04/03 12:37:56] <chrismat_> Great thanks for explaining that | ||
| [2009/04/03 12:38:27] <Volcane> so if all your machiens are the same to begin with, kickstarting is easy | ||
| [2009/04/03 12:38:43] <Volcane> and if you then use a external node tool which could be very simple, you can easily put machines into categories | ||
| [2009/04/03 12:38:59] <chrismat_> I have lots of PXE netbooted thinclients, I am a bit unsure how that would fit into cobbler rewriting dhcpd.conf | ||
| [2009/04/03 12:39:01] <Volcane> puppet can query external databases/ldap/whatevers to decide what to put on a machine | ||
| [2009/04/03 12:39:25] <chrismat_> can you exclude cobbler from doing that or mix | ||
| [2009/04/03 12:39:30] <chrismat_> clever puppet | ||
| [2009/04/03 12:39:32] <Volcane> could mix i think | ||
| [2009/04/03 12:39:43] <Volcane> http://reductivelabs.com/trac/puppet/wiki/ExternalNodes theres some info about the external node thing | ||
| [2009/04/03 12:40:47] <chrismat_> How do you install Puppet, when the OS installs, do you put that in a separate repo? | ||
| [2009/04/03 12:41:09] <chrismat_> Is there any official puppet.repo ? | ||
| [2009/04/03 12:41:46] <Volcane> yeah I've a local repo that I include in my kickstarts with the repo option | ||
| [2009/04/03 12:41:51] <Volcane> and then just put puppet into the package list | ||
| [2009/04/03 12:41:58] <Volcane> so its there the moment the box first boots | ||
| [2009/04/03 12:42:40] <Volcane> i create my desired first run puppet.conf in the pre section of kickstart | ||
| [2009/04/03 12:43:05] <chrismat_> Aha! :) | ||
| [2009/04/03 12:43:16] <Volcane> z00dax: pretty new artwork in 5.3, kudos! | ||
| [2009/04/03 12:47:17] @ Quit: kolla: Remote closed the connection | ||
| [2009/04/03 12:49:27] @ madrescher joined channel #puppet | ||
| [2009/04/03 12:55:24] @ ezmob joined channel #puppet | ||
| [2009/04/03 12:56:22] @ benblack joined channel #puppet | ||
| [2009/04/03 12:56:47] @ Quit: fbe_: Read error: 60 (Operation timed out) | ||
| [2009/04/03 12:58:23] <Volcane> yeah interesting question | ||
| [2009/04/03 12:58:24] <Volcane> oops | ||
| [2009/04/03 12:59:32] <jief-> chrismat_: its part of our OS images, preconfigured to connect to our server | ||
| [2009/04/03 13:00:23] <gwar9999> Not sure if I should ask here or in #ruby, but I have an output formatting question inside an erb template. If I want to have fixed length string output, how can I accomplish this? Something like: | ||
| [2009/04/03 13:00:26] <gwar9999> foo bar 4 | ||
| [2009/04/03 13:00:26] <gwar9999> testing 999999 5 | ||
| [2009/04/03 13:00:26] <gwar9999> more - 6 | ||
| [2009/04/03 13:02:22] <Superfly_> ruby does support (s)printf, doesnt it? | ||
| [2009/04/03 13:02:46] <Volcane> yes | ||
| [2009/04/03 13:03:49] <gwar9999> hmmm, so I would need to use sprintf within the erb template to accomplish this. I'll give it a shot although it's probably going to clutter up the template. Oh well. | ||
| [2009/04/03 13:04:18] <Volcane> you know puppet has its own sprintf function now? | ||
| [2009/04/03 13:05:55] <gwar9999> nope. | ||
| [2009/04/03 13:06:27] <Volcane> http://reductivelabs.com/trac/puppet/wiki/FunctionReference#sprintf | ||
| [2009/04/03 13:08:03] <gwar9999> thanks Volcane for the link. However, from a template I think I'd be calling the ruby sprintf since it's in erb. | ||
| [2009/04/03 13:08:29] @ Quit: shake-n-bake: | ||
| [2009/04/03 13:08:43] <Volcane> yeah if you're in a tmeplate sure, just thought it was worth mentioning in case you were just wanting to manipulate a val or something | ||
| [2009/04/03 13:08:58] <jief-> Volcane: you should teach classes or something | ||
| [2009/04/03 13:09:10] <gwar9999> maybe I don't need to do this at all... I suppose the data I send to the erb can be pre-formatted from the external script. | ||
| [2009/04/03 13:09:32] <gwar9999> I agree with jief- | ||
| [2009/04/03 13:09:38] <TedC> I'm having a facter problem. None of my facter variables are making it into puppet. | ||
| [2009/04/03 13:09:52] <Volcane> jief-:) | ||
| [2009/04/03 13:10:37] <TedC> Anything I should look at? facter is installed on both the master and client, but if I do a notice or notify {"$hostname"}, I just get blank. | ||
| [2009/04/03 13:10:55] <jief-> TedC: first, are you facts getting pushed to clients? | ||
| [2009/04/03 13:11:09] <TedC> not sure. How would I test that? | ||
| [2009/04/03 13:11:09] <Volcane> i dont think puppet would even run if facts werent working | ||
| [2009/04/03 13:11:28] <jief-> TedC: it should end up on the client at $vardir/lib/facter by default | ||
| [2009/04/03 13:11:31] <Volcane> TedC: just do notice("hostname is ${hostname}") and run it, check logs on client and server | ||
| [2009/04/03 13:11:32] <jief-> at least on debian-based systems | ||
| [2009/04/03 13:12:48] <Volcane> put that notice() in your node or even site.pp | ||
| [2009/04/03 13:12:52] @ stijnbe joined channel #puppet | ||
| [2009/04/03 13:12:53] <Volcane> you should see them on the master | ||
| [2009/04/03 13:13:02] <TedC> my config has a notice("hostname is ${hostname}"), I get "notice: My hostname is" | ||
| [2009/04/03 13:13:23] <Volcane> try ${fqdn} ? | ||
| [2009/04/03 13:13:47] <Volcane> or even ${ipaddress} something that doesnt require the black magic of host name resolving? | ||
| [2009/04/03 13:15:36] <TedC> nope | ||
| [2009/04/03 13:15:40] <TedC> (Scope(Class[basedebian])) ip: | ||
| [2009/04/03 13:15:55] <Volcane> thats very weird | ||
| [2009/04/03 13:15:58] <TedC> I'm not seeing a /var/lib/facter directory at all. (on debian) | ||
| [2009/04/03 13:15:59] <TedC> indeed | ||
| [2009/04/03 13:16:09] <Volcane> what version facter on there? | ||
| [2009/04/03 13:16:18] <TedC> 1.5.1-0.1 | ||
| [2009/04/03 13:16:28] <Volcane> and if you run just facter you see the output right/ | ||
| [2009/04/03 13:16:44] <TedC> right. facter ipaddress returns 10.0.192.1 | ||
| [2009/04/03 13:16:58] <Volcane> on the node, ignoring the master for the moment, just make a small .pp file with just the notice in it | ||
| [2009/04/03 13:17:02] <Volcane> and run puppet whatever.pp | ||
| [2009/04/03 13:17:08] <Volcane> does that work? | ||
| [2009/04/03 13:17:27] <TedC> yes | ||
| [2009/04/03 13:17:30] <TedC> notice: Scope(Class[main]): ip: 10.0.192.1 | ||
| [2009/04/03 13:17:40] <Volcane> wow, well then indeed somethings up with sending facts to master, boggle. | ||
| [2009/04/03 13:17:42] <Volcane> never seen that | ||
| [2009/04/03 13:18:18] <TedC> thoughts for how to debug it? | ||
| [2009/04/03 13:18:37] <Volcane> none, i didnt think the master would even compile a manifest if facts didnt work. | ||
| [2009/04/03 13:18:38] <Volcane> check this. | ||
| [2009/04/03 13:18:56] <TedC> Hmmm. | ||
| [2009/04/03 13:18:57] @ freak4u left channel #puppet ("Leaving") | ||
| [2009/04/03 13:19:06] <TedC> I may have a version mismatch between my node and server. | ||
| [2009/04/03 13:19:26] <Volcane> how big a missmatch? | ||
| [2009/04/03 13:19:27] <TedC> nope. | ||
| [2009/04/03 13:19:38] <TedC> 0.24.3 master 0.24.5 client | ||
| [2009/04/03 13:19:50] <Volcane> might be ok | ||
| [2009/04/03 13:19:52] <TedC> so not that big, actualy. | ||
| [2009/04/03 13:20:19] <Volcane> check /var/lib/puppet/aml/nodes/ | ||
| [2009/04/03 13:20:32] <Volcane> there should be many yaml files | ||
| [2009/04/03 13:20:34] <Volcane> with facts in them | ||
| [2009/04/03 13:21:49] <Volcane> sorry, /var/lib/puppet/yaml/nodes | ||
| [2009/04/03 13:21:53] <TedC> there are many yaml files in /var/puppet/yaml/node, but almost none of them have facts in them. | ||
| [2009/04/03 13:22:18] <Volcane> the ones that have facts, are they of the same version as your master? | ||
| [2009/04/03 13:22:41] <Volcane> there'll be puppetversion fact in there | ||
| [2009/04/03 13:23:31] <TedC> there's only one that's not the master itself, and it's a different version (much older, actually: 0.22.4) | ||
| [2009/04/03 13:23:49] <Volcane> might be old, well try updating one of your clients to the same version as the server and see if its the same | ||
| [2009/04/03 13:23:55] <Volcane> long shot, not sure what else to suggest | ||
| [2009/04/03 13:24:56] <TedC> the thing is, that's a dead client. It was a test machine, and has been decommissioned. All of the clients I currently have have yaml files, but no real data in them. The parameters field is {}. | ||
| [2009/04/03 13:25:10] <Volcane> nods, that file is probably old then | ||
| [2009/04/03 13:25:11] @ alfism joined channel #puppet | ||
| [2009/04/03 13:25:23] <Volcane> try upgrading one, or downgrading the server | ||
| [2009/04/03 13:25:27] <jief-> Volcane: think i could puppet 2000 hosts with a single puppetmaster? | ||
| [2009/04/03 13:25:53] <Volcane> jief-: depends on many things, but i'd probably think you'd need more than one box | ||
| [2009/04/03 13:26:03] <TedC> I'll upgrade the server to match the clients. Right now, the clients are newer than the server. | ||
| [2009/04/03 13:26:13] <Volcane> k | ||
| [2009/04/03 13:27:21] <Volcane> jief: and more than one host gets pretty complex, anyway i think theres a wiki page dealing with a multi box scaling puppet type setup | ||
| [2009/04/03 13:29:50] <jief-> Volcane: i do that daily scaling code that shouldn't ;) | ||
| [2009/04/03 13:30:00] <jief-> ill put it behind our load-balancers farm | ||
| [2009/04/03 13:30:12] <jief-> and share the configs over NFS | ||
| [2009/04/03 13:30:18] <jief-> that should do it | ||
| [2009/04/03 13:30:25] <Volcane> nod, certificates are a pain too | ||
| [2009/04/03 13:30:49] <jief-> hmm | ||
| [2009/04/03 13:30:57] <jief-> right, haven't thought about that | ||
| [2009/04/03 13:30:59] <Volcane> http://reductivelabs.com/trac/puppet/wiki/PuppetScalability | ||
| [2009/04/03 13:31:00] <jief-> clients won't like it | ||
| [2009/04/03 13:31:47] <jief-> ill experiment with our wildcard cert i guess | ||
| [2009/04/03 13:32:06] <Volcane> clients need to be signed with the same ca | ||
| [2009/04/03 13:32:18] <Volcane> you can share the ca over nfs probably, as long as the certname is sane it should be fine | ||
| [2009/04/03 13:33:43] <jief-> ill experiment and will most likely write on the wiki | ||
| [2009/04/03 13:34:24] <Volcane> 0.25 should perform much better too | ||
| [2009/04/03 13:34:54] <rmiller4pi8> or you can run the ca on a separate port and only send that one place | ||
| [2009/04/03 13:36:09] @ Quit: unxfrek: "Leaving" | ||
| [2009/04/03 13:37:19] @ PhabX joined channel #puppet | ||
| [2009/04/03 13:39:09] @ mrrx joined channel #puppet | ||
| [2009/04/03 13:39:31] <TedC> Volcane, It's working now. I upgraded facter and puppet to the latest versions, and it's happy. | ||
| [2009/04/03 13:40:36] @ Quit: lak: | ||
| [2009/04/03 13:47:38] <Volcane> TedC: nice | ||
| [2009/04/03 13:51:29] @ nakano is now known as nakano_ | ||
| [2009/04/03 13:52:42] @ claymation joined channel #puppet | ||
| [2009/04/03 13:54:49] @ shake-n-bake joined channel #puppet | ||
| [2009/04/03 13:56:07] @ Quit: jmeeuwen: Read error: 110 (Connection timed out) | ||
| [2009/04/03 14:02:28] @ Quit: alban2: Read error: 110 (Connection timed out) | ||
| [2009/04/03 14:05:48] @ Ju` joined channel #puppet | ||
| [2009/04/03 14:06:11] @ Ju` is now known as Ju | ||
| [2009/04/03 14:06:48] <Ju> Hi there ! | ||
| [2009/04/03 14:07:36] @ Quit: jief-: "Lost terminal" | ||
| [2009/04/03 14:07:37] <PhabX> yoh Ju! | ||
| [2009/04/03 14:09:50] @ Quit: shake-n-bake: | ||
| [2009/04/03 14:13:43] @ ezmob is now known as ezmob|away | ||
| [2009/04/03 14:16:59] @ Quit: fujin: | ||
| [2009/04/03 14:42:54] @ Quit: d3vilb0x: | ||
| [2009/04/03 14:45:37] @ andrewcshafer joined channel #puppet | ||
| [2009/04/03 14:46:16] @ grim_radical joined channel #puppet | ||
| [2009/04/03 14:48:10] <MrHeavy> Before I submit a bug, has anyone else run into problems trying to remove broken symlinks with Puppet? | ||
| [2009/04/03 14:52:29] <sigmonsays> Hey, does anyone know of a more advanced monitoring tool than nagios? i'm hoping to use somethign that can script better | ||
| [2009/04/03 14:52:45] <sigmonsays> would also love to see a hierarchy notification system | ||
| [2009/04/03 14:53:01] @ aymerick_ joined channel #puppet | ||
| [2009/04/03 14:55:07] @ jmeeuwen joined channel #puppet | ||
| [2009/04/03 14:56:14] @ nakano_ is now known as nakano | ||
| [2009/04/03 14:59:11] @ Quit: nakano: "Coyote finally caught me" | ||
| [2009/04/03 14:59:32] @ nakano joined channel #puppet | ||
| [2009/04/03 15:01:10] @ lak joined channel #puppet | ||
| [2009/04/03 15:05:55] <jrojas> sigmonsays: good luck, nagios is unfortunately the best thing out there | ||
| [2009/04/03 15:06:30] <jrojas> sigmonsays: it does have a heiarchy notification system, 3 pages to this person then escalate to this person etc... | ||
| [2009/04/03 15:06:47] <Telmo> evening, how can i ensure that a directory exists? also making sure all the directories in the path exist? I haven't seen a directory resource in the wiki unless I am missing something | ||
| [2009/04/03 15:07:09] <jrojas> Telmo: directories are files | ||
| [2009/04/03 15:07:20] <jrojas> file { "/etc/blah": ensure => directory } | ||
| [2009/04/03 15:07:42] <Telmo> would that ensure the whole path? /path/to/my/directory/ ? | ||
| [2009/04/03 15:07:51] <jrojas> recurse => true | ||
| [2009/04/03 15:07:58] <Telmo> awesome thanks | ||
| [2009/04/03 15:08:22] @ Quit: bolt: Remote closed the connection | ||
| [2009/04/03 15:08:25] <jrojas> at least i think it will...havent tried tho.. | ||
| [2009/04/03 15:08:36] @ ezmob|away is now known as ezmob | ||
| [2009/04/03 15:08:42] @ thegcat joined channel #puppet | ||
| [2009/04/03 15:09:05] <z00dax> Volcane: its perhaps the most comprehensive set we have had so far - thank Alain and Ralph for that though - they put in the legwork for almost all of it | ||
| [2009/04/03 15:12:29] @ bolt joined channel #puppet | ||
| [2009/04/03 15:13:40] @ Quit: aymerick: Read error: 110 (Connection timed out) | ||
| [2009/04/03 15:14:02] @ Quit: benblack: "Leaving..." | ||
| [2009/04/03 15:18:31] @ fbe_ joined channel #puppet | ||
| [2009/04/03 15:24:13] <tim|imac> argh... | ||
| [2009/04/03 15:24:17] <tim|imac> what am I doing wrong here? | ||
| [2009/04/03 15:24:19] <tim|imac> case $lsbdistcodename { | ||
| [2009/04/03 15:24:19] <tim|imac> lenny: { include xen::dom0::lenny } | ||
| [2009/04/03 15:24:19] <tim|imac> etch: { include xen::dom0::etch } | ||
| [2009/04/03 15:24:19] <tim|imac> } | ||
| [2009/04/03 15:28:34] <tim|imac> nvm, pebkac | ||
| [2009/04/03 15:34:02] @ cheflog__ is now known as magnachef | ||
| [2009/04/03 15:39:58] @ Tuf8 joined channel #puppet | ||
| [2009/04/03 15:50:35] <justdave> so when I'm using tagmail.conf to send email reports of changes to servers, is there some way to get those emails to include the diffs of the files it changes? (same as you get in the output if you run it manually with --test) | ||
| [2009/04/03 15:54:25] @ dsch04 joined channel #puppet | ||
| [2009/04/03 15:59:22] <lak> justdave: no, there isn't | ||
| [2009/04/03 15:59:30] <lak> the diffs can only print on the console, at this point | ||
| [2009/04/03 16:01:43] @ Quit: pleemans: Read error: 110 (Connection timed out) | ||
| [2009/04/03 16:19:49] @ Quit: jmarki: Remote closed the connection | ||
| [2009/04/03 16:25:36] @ ezmob is now known as ezmob|away | ||
| [2009/04/03 16:29:12] @ ezmob|away is now known as ezmob | ||
| [2009/04/03 16:31:19] @ Quit: stijnbe: "Leaving..." | ||
| [2009/04/03 16:37:22] @ Quit: tmz: "leaving" | ||
| [2009/04/03 16:42:08] @ Quit: aymerick_: | ||
| [2009/04/03 16:43:15] @ tmz joined channel #puppet | ||
| [2009/04/03 16:48:04] @ claymation_ joined channel #puppet | ||
| [2009/04/03 16:48:05] @ Quit: claymation: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 16:48:19] @ Quit: claymation_: Client Quit | ||
| [2009/04/03 16:50:42] @ ezmob is now known as ezmob|away | ||
| [2009/04/03 16:51:08] <Volcane> z00dax: yeah i dont use it as a desktop, but i installed a little vm in gui mode just to take a look at it, very pretty and nicely consistent, must say its a lot more pleasing than the garish red certain other distros have :P | ||
| [2009/04/03 16:52:21] <joe-mac> sigmonsays: script wrt what? | ||
| [2009/04/03 16:52:38] <joe-mac> nagios is actually extremely advancesd | ||
| [2009/04/03 16:53:03] <joe-mac> well, you can extend its simple core to be extremely advanced i should say | ||
| [2009/04/03 16:53:13] <Volcane> the notification system needs a good kicking | ||
| [2009/04/03 16:53:13] <joe-mac> i've spent a lot of time with it and still haven't touched a lot of things | ||
| [2009/04/03 16:53:44] <Volcane> its way too hard to have one contact for "john" with sms and email and configure it to use sms only after hours without resorting to external hacks | ||
| [2009/04/03 16:54:06] <Volcane> everyone ends up with 2 contacts john-email john-sms and different groups | ||
| [2009/04/03 16:54:07] <Volcane> pita | ||
| [2009/04/03 17:01:43] @ kambiz is now known as kambiz_away | ||
| [2009/04/03 17:05:16] @ Quit: egerlach: "leaving" | ||
| [2009/04/03 17:08:02] @ alban2 joined channel #puppet | ||
| [2009/04/03 17:10:46] @ ezmob|away is now known as ezmob | ||
| [2009/04/03 17:11:21] @ Quit: andrewcshafer: | ||
| [2009/04/03 17:28:22] @ Quit: PhabX: "Leaving..." | ||
| [2009/04/03 17:28:24] @ andrewcshafer joined channel #puppet | ||
| [2009/04/03 17:30:39] @ Quit: MrHeavy: "Ex-Chat" | ||
| [2009/04/03 17:31:29] @ nakano is now known as nakano_ | ||
| [2009/04/03 17:58:45] @ ezmob is now known as ezmob|away | ||
| [2009/04/03 18:00:54] @ Djelibeybi joined channel #puppet | ||
| [2009/04/03 18:01:48] @ Quit: mccune: Remote closed the connection | ||
| [2009/04/03 18:02:30] @ Quit: DarkFlibb: Remote closed the connection | ||
| [2009/04/03 18:05:57] @ DarkFlib joined channel #puppet | ||
| [2009/04/03 18:15:19] @ thegcat_ joined channel #puppet | ||
| [2009/04/03 18:15:30] @ Quit: RexMundi: "Ik ga weg" | ||
| [2009/04/03 18:18:44] @ collier_s joined channel #puppet | ||
| [2009/04/03 18:22:38] @ Quit: thegcat: Read error: 113 (No route to host) | ||
| [2009/04/03 18:25:47] @ Quit: lak: | ||
| [2009/04/03 18:34:47] <justdave> hmm, http://reductivelabs.com/trac/naginator is broken? | ||
| [2009/04/03 18:35:42] @ Quit: flakrat: "Leaving" | ||
| [2009/04/03 18:38:40] @ nakano_ is now known as nakano | ||
| [2009/04/03 18:47:31] @ ezmob|away is now known as ezmob | ||
| [2009/04/03 18:49:59] @ sinwave_ joined channel #puppet | ||
| [2009/04/03 18:51:45] <jrojas> looks that way | ||
| [2009/04/03 18:52:02] @ LapCat-1 joined channel #puppet | ||
| [2009/04/03 18:52:08] <jrojas> try projects.reductivetry: http://projects.reductivelabs.com/projects/naginator | ||
| [2009/04/03 18:52:22] @ keithlard joined channel #puppet | ||
| [2009/04/03 18:52:47] @ LapCat-1 left channel #puppet ("Leaving") | ||
| [2009/04/03 18:52:59] @ Irunongames joined channel #puppet | ||
| [2009/04/03 18:53:18] @ Irunongames left channel #puppet () | ||
| [2009/04/03 18:58:04] @ lak joined channel #puppet | ||
| [2009/04/03 19:06:55] @ Quit: markl___: "Lost terminal" | ||
| [2009/04/03 19:13:06] @ Quit: sinwave_: "Bad BitchX, bad BitchX, whatcha gonna do? Whatcha gonna do when they come for you?" | ||
| [2009/04/03 19:14:49] @ Quit: chiggsy: | ||
| [2009/04/03 19:21:03] @ Quit: madrescher: "Leaving." | ||
| [2009/04/03 19:24:09] @ Quit: thegcat_: | ||
| [2009/04/03 19:50:16] @ Quit: ezmob: "Bye!" | ||
| [2009/04/03 19:51:24] @ Quit: lak: | ||
| [2009/04/03 19:55:58] @ Quit: Tuf8: | ||
| [2009/04/03 19:58:23] @ Quit: alfism: "Connection reset by beer" | ||
| [2009/04/03 20:04:59] @ benblack joined channel #puppet | ||
| [2009/04/03 20:08:01] <gwar9999> I can't seem to find the function I need on the wiki.... hopefully someone can point me in the right direction. If I need to ensure that a file contains a specific line (eg. "foo x y z" how do I accomplish this in puppet? | ||
| [2009/04/03 20:09:22] <gwar9999> looks like I need to build a custom function, I guess? | ||
| [2009/04/03 20:18:25] <Djelibeybi> gwar9999: depends on the file. You could use Augeas if it supports the file you're checking. | ||
| [2009/04/03 20:18:45] <Djelibeybi> gwar9999: otherwise, I did find a custom function on someone's complete config that did this using exec calls to grep | ||
| [2009/04/03 20:22:54] @ Quit: cwebber: | ||
| [2009/04/03 20:25:24] @ Quit: Djelibeybi: Read error: 104 (Connection reset by peer) | ||
| [2009/04/03 20:32:30] @ Djelibeybi joined channel #puppet | ||
| [2009/04/03 20:40:58] @ nakano is now known as nakano_ | ||
| [2009/04/03 20:41:17] @ nakano_ is now known as nakano | ||
| [2009/04/03 20:41:37] <johnw> can the store-configs-for-postgresql use the new "pg" gem, or does it need the older "postgres" one? | ||
| [2009/04/03 20:47:19] <justdave> Djelibeybi: we were doing this: http://pastebin.mozilla.org/639754 | ||
| [2009/04/03 20:47:48] <justdave> the commented out part is generically what you're trying to accomplish. it got replaced with the uncommented part when we discovered puppet had a built-in for managing lines in that particular file in our example | ||
| [2009/04/03 20:48:40] <justdave> if you don't need to restart/recompile anything in your case, you wouldn't need the notify trigger | ||
| [2009/04/03 20:48:47] <Djelibeybi> justdave: I wasn't looking for it -- I think someone else was. I was just replying. | ||
| [2009/04/03 20:49:02] <justdave> oh, yep, I misread my scrollback, that was gwar9999 asking | ||
| [2009/04/03 20:49:07] <justdave> gwar9999: ^^^^ | ||
| [2009/04/03 20:55:51] @ fujin joined channel #puppet | ||
| [2009/04/03 21:05:05] @ Quit: keithlard: | ||
| [2009/04/03 21:07:38] @ keithlard joined channel #puppet | ||
| [2009/04/03 21:14:14] @ eythian_ is now known as eythian | ||
| [2009/04/03 21:14:59] <eythian> I'm looking into making puppet types to manage glassfish. Anyone done this to save me reinventing it? | ||
| [2009/04/03 21:23:49] <johnw> if I just want an exec to run once after a package is installed or updated, do I need to both put notify on the package, and subscribe on the exec? or is just one of them sufficient? | ||
| [2009/04/03 21:24:22] <eythian> johnw: just a subscribe, I think. You may also want refresholny=>true | ||
| [2009/04/03 21:24:27] <eythian> although, spell it right | ||
| [2009/04/03 21:24:37] <johnw> well, this is what I had, but the exec's weren't being run | ||
| [2009/04/03 21:24:45] <johnw> at least, not always | ||
| [2009/04/03 21:25:10] <eythian> I don't think I've ever done it with a package for that matter, I usually use it with files. | ||
| [2009/04/03 21:25:30] <johnw> i'm trying to create a dependency chain among non-file resources | ||
| [2009/04/03 21:25:49] <johnw> so that after stuff is installed on the system, certain post-install scripts will get run | ||
| [2009/04/03 21:25:55] <johnw> sometimes, many of them, each dependent on the other | ||
| [2009/04/03 21:26:56] <eythian> hmm. I would expect that to work, but I'm still learning at this stuff too. | ||
| [2009/04/03 21:29:09] @ Quit: andrewcshafer: | ||
| [2009/04/03 21:34:58] <johnw> oh, i found my probelm | ||
| [2009/04/03 21:35:04] <johnw> didn't subscribe to enough things | ||
| [2009/04/03 21:37:24] <eythian> johnw: ah right :) Done that before... | ||
| [2009/04/03 21:38:05] @ nakano is now known as nakano_ | ||
| [2009/04/03 21:44:22] @ Quit: benblack: "Leaving..." | ||
| [2009/04/03 21:46:18] @ punkcut joined channel #puppet | ||
| [2009/04/03 22:07:12] @ Quit: keithlard: | ||
| [2009/04/03 22:13:19] @ mfoster joined channel #puppet | ||
| [2009/04/03 22:15:48] <johnw> has anyone written a script to generate a DOT file from a manifest? | ||
| [2009/04/03 22:28:07] @ cwebber joined channel #puppet | ||
| [2009/04/03 22:28:23] <rmiller4pi8> ? | ||
| [2009/04/03 22:28:33] <johnw> to show dependencies among modules and resources | ||
| [2009/04/03 22:28:47] <rmiller4pi8> oh, that kind of dot | ||
| [2009/04/03 22:29:08] <rmiller4pi8> if you find one, let me know, that would be cool | ||
| [2009/04/03 22:30:31] @ Quit: Agrajag^: Read error: 60 (Operation timed out) | ||
| [2009/04/03 22:30:43] @ Agrajag^ joined channel #puppet | ||
| [2009/04/03 22:31:06] @ d3vilb0x joined channel #puppet | ||
| [2009/04/03 22:31:46] <rmiller4pi8> johnw: looks like this is a start: http://gist.github.com/74566 | ||
| [2009/04/03 22:32:38] <rmiller4pi8> johnw: and there is the builtin bit, but it has some issues http://reductivelabs.com/trac/puppet/wiki/FrequentlyAskedQuestions#how-do-i-use-puppet-s-graphing-support | ||
| [2009/04/03 22:34:43] @ collier_s left channel #puppet () | ||
| [2009/04/03 22:37:24] @ Quit: Telmo: Remote closed the connection | ||
| [2009/04/03 22:38:13] <eythian> hmm. on issue is the link in the FAQ on how to enable graphing support is a 404. | ||
| [2009/04/03 22:38:18] <webx> is there any way to show which of my nodes had or are currently having errors when puppet runs ? | ||
| [2009/04/03 22:38:46] <eythian> webx: I have a nagios task that sends me nasty emails when things are goign wrong. | ||
| [2009/04/03 22:38:52] <webx> haha | ||
| [2009/04/03 22:38:58] <webx> eythian: log scraping ? | ||
| [2009/04/03 22:39:20] <eythian> webx: it checks a) is the puppet process running and b) is the state file modified recently. | ||
| [2009/04/03 22:39:42] <eythian> webx: http://www.kallisti.net.nz/blog/2009/02/monitoring-puppet-with-nagios/ | ||
| [2009/04/03 22:41:38] <webx> interesting.. having a look :) | ||
| [2009/04/03 22:42:35] <webx> If, for example, a developer who was writing a class made a mistake and umm... named his class the same as an existing class, so each of the nodes who used that class were complaining | ||
| [2009/04/03 22:42:42] <webx> does state.yaml get updated in that case? | ||
| [2009/04/03 22:44:41] <eythian> webx: Dunno. test it and find out :) It's been pretty good about letting me know about errors. | ||
| [2009/04/03 22:44:57] <eythian> I don't recall seeing things get logged without this telling me. | ||
| [2009/04/03 22:49:42] @ Quit: cwebber: | ||
| [2009/04/03 22:51:17] <webx> thanks | ||
| [2009/04/03 22:55:08] <webx> in order to install Nagios::Plugins, I'll have to either write a bunch of rpms for all of it's pre-reqs or do some perl one liner I'd guess to -MCPAN -e install <something or other to follow deps> | ||
| [2009/04/03 22:55:18] <webx> how did you automate the installation, or did you? | ||
| [2009/04/03 22:57:46] <eythian> I use Debian. On that, I can do something like dh-make-perl --build --cpan Nagios::Plugins and it gives me a .deb that I then deploy to all the hosts. | ||
| [2009/04/03 22:58:18] <webx> ah | ||
| [2009/04/03 22:58:21] <eythian> Note in the comments that you can use the check_age plugin if you are happy skipping the 'is the process running' check. that would be easier. | ||
| [2009/04/03 22:58:41] <webx> I'm not happy skipping that though :) | ||
| [2009/04/03 22:58:42] <rmiller4pi8> webx: there's hardly a CPAN lib in the world that DAG doesn't build | ||
| [2009/04/03 22:58:57] <eythian> (of course, if there's an RPM equiv for dh-make-perl, it's quick and easy. | ||
| [2009/04/03 22:59:00] <eythian> ) | ||
| [2009/04/03 22:59:04] <rmiller4pi8> webx: rpm.pbone.net search perl(Class::Subclass) | ||
| [2009/04/03 22:59:17] <webx> yea, I'm a lover of dag | ||
| [2009/04/03 23:00:16] <webx> Nagios::Plugin is one of the rare that dag doesn't have apparently | ||
| [2009/04/03 23:00:36] <rmiller4pi8> webx: i guess i forgot about building that one | ||
| [2009/04/03 23:00:41] <rmiller4pi8> cpan2rpm is so transparent | ||
| [2009/04/03 23:01:33] <webx> hmm | ||
| [2009/04/03 23:01:38] <webx> that still work right? | ||
| [2009/04/03 23:01:46] <webx> I'd like to use it | ||
| [2009/04/03 23:01:46] <rmiller4pi8> i use it every week | ||
| [2009/04/03 23:02:34] <rmiller4pi8> the only catch is that you have to have any dependencies already installed on the build machine | ||
| [2009/04/03 23:02:51] <rmiller4pi8> so enjoy the bad old days of playing yum yourself and walking the dep trees | ||
| [2009/04/03 23:03:13] <webx> haha | ||
| [2009/04/03 23:03:22] <rmiller4pi8> though it's DAG, again, almost always has whatever deps you're looking for | ||
| [2009/04/03 23:03:31] <webx> so basically, install the deps on the build machine | ||
| [2009/04/03 23:03:37] <rmiller4pi8> right | ||
| [2009/04/03 23:03:39] <webx> and then build all of the dep modules to rpm | ||
| [2009/04/03 23:03:41] <webx> yea yea.. OK | ||
| [2009/04/03 23:03:45] <webx> I'll give that ago | ||
| [2009/04/03 23:03:50] <webx> s/ag/a g/ | ||
| [2009/04/03 23:03:53] <rmiller4pi8> heh | ||
| [2009/04/03 23:04:12] <rmiller4pi8> i don't think the nagios module has any non-DAG dependencies | ||
| [2009/04/03 23:04:25] <rmiller4pi8> also, cpan2rpm depends on rpm-build | ||
| [2009/04/03 23:07:45] <webx> is there any RPM trickery that I may not know about where I can wrap all 3 of these rpms into one install? | ||
| [2009/04/03 23:10:15] <eythian> There's nothing I know of, short of building them manually. | ||
| [2009/04/03 23:12:54] <webx> yea, I have it all built now | ||
| [2009/04/03 23:15:00] @ mfoster left channel #puppet () | ||
| [2009/04/03 23:20:12] @ andrewcshafer joined channel #puppet | ||
| [2009/04/03 23:24:19] @ cwebber joined channel #puppet | ||
| [2009/04/03 23:49:15] @ benblack joined channel #puppet | ||
| [2009/04/03 23:56:37] @ Quit: cwebber: |
Generated by irclog2html.py 2.6 by Marius Gedminas - find it at mg.pov.lt!