Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-05-20

| Channels | #salt index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:02 whiteinge andrej: slow reply. glad you got it :)
00:04 Nexus_x1 joined #salt
00:13 n8n joined #salt
00:21 faldridge joined #salt
00:22 Luke_ joined #salt
00:30 bhosmer joined #salt
00:32 simonmcc jcockhren: v0.0.16 allows you to set a collection_name (for clarity) https://github.com/simonmcc/kitchen-salt/commit/2c647ed4dfa0e7104894975b74edc46bdb9f6110
00:34 tyler-baker left #salt
00:37 XenophonF joined #salt
00:42 happytux joined #salt
00:47 ajolo joined #salt
00:47 hipster joined #salt
00:47 n8n joined #salt
00:57 travisfischer joined #salt
01:03 malinoff joined #salt
01:18 jcockhren simonmcc: thanks!
01:19 Zuru joined #salt
01:20 pdayton joined #salt
01:20 andrej whiteinge : all good; not having people answer all my questions is actually good for me ;D ... takes WAY longer, but the result is more *sticky*, if you know what I mean :}
01:20 XenophonF heh
01:21 mgw joined #salt
01:24 HeadAIX joined #salt
01:25 Joseph @UtahDave: considering the early conversation about defining the mean of state.
01:25 Joseph i have a write up that i'd like you to review
01:25 whiteinge andrej: haha
01:25 Joseph would you be willing?
01:25 fragamus joined #salt
01:27 whiteinge Joseph: i think he's gone home for the day. he's often back online in the evenings or tomorrow, of course.
01:27 Joseph gotcha
01:28 Joseph its quite a significant rewrite of how state is used in the documentation
01:28 whiteinge i'm a poor substitute for UtahDave but i'd be happy to give it a look if you'd rather not wait
01:28 Joseph oh sure!!
01:28 Joseph i am not sure how this will be received by the community
01:28 Joseph but i think it really helps clarify that state means one and only one thing
01:28 bhosmer joined #salt
01:28 whiteinge the docs play pretty fast and loose with the term
01:28 Joseph yes it drives me absolutely crazy
01:28 Joseph and thus my rewrite
01:29 bhosmer joined #salt
01:29 malinoff Joseph, I would be interested too
01:31 bhosmer_ joined #salt
01:32 bhosmer_ joined #salt
01:34 Joseph okay
01:34 Joseph here you guys go: https://gist.githubusercontent.com/jaloren/30f87d6f0b68b5137fd1/raw/2f0ce05f0dc99a3bf05446a17eb3857249e79897/gistfile1.txt
01:34 Joseph i apologize of the length
01:35 whiteinge Joseph: is this intended to modify an existing doc? or is this a new addition?
01:36 Joseph it would modify an existing doc
01:36 Joseph either http://docs.saltstack.com/en/latest/topics/tutorials/starting_states.html
01:36 Joseph or the highstate data structure section
01:36 Joseph if this was accepted it would require many changes across the doc set to unify terms
01:36 * whiteinge nods
01:37 whiteinge reading...
01:38 AdamSewell joined #salt
01:39 malinoff Joseph, just my 2 cents, people don't want to read the wall of text. Since salt state files syntax are quite difficult for a newbie, it is not THAT difficult. I'd like to see many iterative examples, from 'easy' to 'hard'
01:39 jcockhren malinoff likes flowcharts
01:39 jcockhren s/flow/gant/
01:39 jcockhren ;p
01:40 * whiteinge would like more flow and gantt charts in the docs
01:40 malinoff jcockhren, yep :)
01:41 whiteinge what good is sphinx support for embedded graphviz if we never use it!
01:41 jcockhren heh. yep
01:42 malinoff One of the best tutorials I've ever seen is rabbitmq tutorial - probably, you can take something from it: http://www.rabbitmq.com/getstarted.html
01:45 XenophonF i think the existing tutorial covers this same material pretty well
01:45 malinoff from "The simplest thing that does something" to "Remote procedure call implementation"
01:45 jcockhren not is you're asing, "how do the events flow in salt"
01:45 jcockhren asking*
01:45 Joseph malinoff: I understand. I could axe the section on the requisites completely but the rest of it was the only way i know to differentiate between terms and get away from using state to mean differnet things.
01:45 Joseph Speaking as a recent newbie myself, i was incredibly confused and struggled for over a week trying to figure out what "state" mean in context A vs Context B
01:45 Joseph trying to figure out how requisites worked was really challenging because of this
01:45 Joseph when in reality its quite smiple
01:46 malinoff Joseph, so the problem is not about the syntax of a state file, but about salt terms
01:46 Joseph yes!!
01:46 Joseph no functional changes needed
01:46 malinoff So why don't you write a doc describing "terms", not state file syntax?
01:46 Joseph better metaphors and one word always mean a single thing
01:46 Joseph state has multiople meanings
01:46 Joseph hmmm
01:47 Joseph but how helpful is a keyword terms if you don't have a concrete example?
01:47 malinoff Joseph, salt already has examples, you can refer to them in your doc
01:47 malinoff Or, you can write your own examples
01:47 hipster joined #salt
01:47 malinoff But right now you're describing a state file syntax
01:47 malinoff Not terms
01:48 Joseph while i guess my issue is the terms we use for that file syntax
01:48 Joseph for example what is this thing called a "highstate component" or this thing called a "state declaration"
01:49 Joseph i just found those terms utterly meaningless
01:49 Joseph in some cases kinda tautological
01:49 Joseph http://docs.saltstack.com/en/latest/ref/states/highstate.html
01:49 Joseph take this entire page
01:50 Joseph for ID declaration, all it says is Defines an individual highstate component.
01:50 Joseph what does that mean!?
01:50 Joseph i have no idea
01:50 Joseph but it seems important
01:50 cpowell joined #salt
01:51 malinoff I'd like to see something like "Execution module is a sequence of actions required to execute a single task in imperative style, e.g. 'install a package'. State module, internally, is a sequence of execution modules calls in declarative style, e.g. 'ensure that a package is installed'"
01:51 malinoff Start from the beginning, describe each significant term
01:52 Joseph model it off the highstate link i just pasted avoe?
01:52 Joseph above
01:52 malinoff Because if I don't understand the difference between state module and execution module, I can't understand nothing
01:52 Joseph completely agree
01:52 jcockhren Joseph, malinoff throw some of the terms defined in an issue. I like this discussion. Needs to happen. Just need to write down some key things from it so all can comment. ;)
01:52 Joseph and this is the current definition !! "State Modules are the components that map to actual enforcement and management of Salt states"
01:53 whiteinge sorry, got pulled afk. reading now...
01:53 redondos joined #salt
01:53 redondos joined #salt
01:53 whiteinge Joseph: lol. what page has that definition of state modules?
01:53 Joseph http://salt.readthedocs.org/en/latest/ref/states/writing.html
01:53 whiteinge yikes
01:53 Joseph which is as far as i can tell the only place a state module is actually defined!!
01:54 ckao joined #salt
01:54 Joseph also, state tree and SLS modules.....so murky
01:55 Joseph okay jcockren i 'll work key terms and definitions issue
01:55 Joseph and link it here when i have something based
01:55 malinoff So it don't make sense for me to describe high-level terms (like requisites) until low-level terms are described well
01:55 jcockhren awesome!
01:55 Joseph yes malinoff you are right
01:55 Joseph because they build on one another
01:55 malinoff Yep
01:55 Joseph i also agree about the wall of text
01:55 Joseph that's going to turn novices off
01:56 Joseph we need a single goto page for a list of key terms, where if a user says "what does this mean" then that's where they go to look
01:56 whiteinge the glossary page was a recent addition (~3 months?). we need to cross-reference glossary entries everywhere so when readers hit an unfamiliar term they can click
01:56 Joseph that would probably have solved most of my confusion
01:56 Joseph yes yes yes!!!
01:56 whiteinge what Joseph said :)
01:56 whiteinge i think there's a ticket for that for the sprint tomorrow
01:56 Joseph but whiteinge....the elephant in the room is that once everyone agrees on what a "state" means
01:56 Joseph that's going to have to mean some major clean up
01:57 Joseph because there is no way the documentation uses state is coherent right now
01:57 Joseph the context switching makes my head spin sometimes
01:57 malinoff Also, I see the lack of a birds-eye view of salt architecture. I work with it almost for a year, and I understand how it works, but for a newbie salt is like magic
01:57 Joseph yes and that's the shame
01:57 whiteinge bunch of work: yes. problem: no. we've got a number of "comb through the docs..." tickets that will be addressed in (hopefully) short order
01:57 whiteinge https://github.com/saltstack/salt/issues/12446
01:57 Joseph the architecture is beautifully pythonic
01:57 Joseph i really appreciate it
01:59 whiteinge hm. i can't find the glossary ticket
01:59 Joseph http://docs.saltstack.com/en/latest/glossary.html
01:59 malinoff Well, I'm not talking about python actually (since there are many dirty hacks and awkward decisions in the source code), but about how actually salt works. I've described salt authentication recently on the mailing list, so why there is nothing about it in the docs?
01:59 Joseph oh my that's quite sparse
02:00 Joseph uh malinoff, that's a loaded question on that one
02:00 fllr joined #salt
02:00 Joseph because if people start digging into how the auth works some very nasty questions aer going to start getting asked about the security implementation
02:00 Joseph they already area but you have to know where to look
02:01 whiteinge Joseph: https://github.com/saltstack/salt/issues/12906
02:01 Joseph got it whiteing....should i add my list of terms as a commetn there?
02:02 whiteinge please do!
02:02 Joseph okay should be excited....imaging the passionate arguments about the meaning of is etc :)
02:04 * whiteinge hurries to add his definition of is before anyone else does
02:04 Joseph ha!!
02:05 malinoff Also, it would be awesome to have a development guide (as a documentation, or pdf, or something). http://tinypic.com/view.php?pic=fxxi5j&s=8 - this were on my whiteboard for 2 or 3 months because I couldn't remember where should I look :)
02:05 dccc joined #salt
02:06 malinoff It's a bit outdated, of course
02:06 Joseph and oh dear lord don't get me started on the overstate and reactor system documentation
02:07 Joseph i know there are tickets opened for that
02:07 Joseph but i feel like i am given a small little glimmer of something awesome but no info on how to make the awesome thing happen
02:07 whiteinge Joseph: (about half-way through) your opening for states is fantastic. this is good stuff and we need it committed :)
02:08 Joseph but where? malinoff has a fair point that its a wall of text
02:08 Joseph i can see that initimidating a novice from learning salt
02:09 googolhash joined #salt
02:09 whiteinge offhand, i'd say break it up between a couple different pages (high-level vs. advanced when you get into the data-structure-level stuffs). itersperse both with example and links
02:09 Joseph ahhh
02:09 Joseph so embarrassing question
02:09 Joseph is there a tutorial on how to write long form documentation for saltstack
02:09 yongfu joined #salt
02:09 Joseph i keep on screwing it up
02:09 whiteinge new users don't need to care that it's a dictionary that maps to a function call but intermediate users definitley will
02:09 Joseph i know i do!!
02:10 whiteinge Joseph: https://github.com/saltstack/salt/issues/12792  :-P
02:11 Joseph hahaha
02:11 Joseph yes that's what i want to know hwo to do
02:12 whiteinge Joseph: i don't suppose you're available to join us in tomorrow's doc sprint? (or perhaps the one in june?)
02:12 Joseph well
02:12 Joseph funny you should mention that
02:12 Joseph i sent an email today asking to join
02:12 Joseph haven't heard anything  back as of yet
02:12 whiteinge it's incredibly helpful to have fresh-user eyes on that
02:13 whiteinge oh, dang. that may be my bad
02:13 sashka_u1 guys, can someone tell ig he is missing state.test. module in his distribution?
02:13 sashka_u1 I run centos6 and found no test state in lib dir
02:13 jalbretsen joined #salt
02:14 whiteinge Joseph: i've been adding people to the google calendar but haven't emailed anyone directly. (someone forwards me the email, then i add 'em to the hangout list.)
02:15 malinoff sashka_ua, I think this state is actually new and will be available in the next release
02:15 Joseph i just got it
02:15 Joseph woohoo
02:15 whiteinge gah. thanks for mentioning that. i didn't realize the person checking that mailbox hadn't been replying
02:16 Joseph oh haha no problem
02:16 sashka_ua malinoff: thank you
02:17 Joseph i'll make sure to get my glossary suggestions done before that
02:17 sashka_ua can then someone tell when Helium is coming out?
02:17 malinoff sashka_ua, but can you please say why do you need this module? :)
02:18 malinoff I mean, you can pass test=True to the cli
02:18 sashka_ua malinoff: I have kind of api and app. app checks if job failed. I want to send it failed job with a comment
02:19 sashka_ua with "send job" I mean run job on a minion that will fail with a message
02:19 whiteinge sashka_ua: the helium RC is likely to be cut in a couple weeks time
02:20 sashka_ua whiteinge: ty. I just didnt find release roadmap, only milestones which is marked as completed
02:20 whiteinge sashka_ua: the test state module doesn't contain any helium-isms. you can just grab the .py file off github and put it in your /srv/salt/_states folder to use it now
02:20 sashka_ua whiteinge: yeah, thats what I was thinking about
02:20 sashka_ua whiteinge: nice catch about :/srv/salt/_states
02:21 sashka_ua whiteinge: we are releasing our system soon, so you will see saltstack running on ~1K hosts :) I think you have enough infra that size in customers
02:22 rgbkrk joined #salt
02:22 malinoff I don't know, are those connection issues solved?
02:23 saurabhs left #salt
02:24 ajprog_laptop joined #salt
02:26 ajw0100 joined #salt
02:30 whiteinge sashka_ua: nice!
02:32 mgw joined #salt
02:38 snuffeluffegus joined #salt
02:39 hipster joined #salt
02:41 krow joined #salt
02:42 lionel joined #salt
02:43 whiteinge malinoff: that whiteboarded chart is very interesting
02:44 malinoff whiteinge, that's how I tried to track salt calls
02:44 whiteinge i'll bet that took a bit of effort to put together :)
02:44 malinoff yeah
02:45 icebourg joined #salt
02:46 malinoff I really missed a doc that could say "hey, if you want to see how the payload is formed, look `here`, and if you want to see how do we handle cli options, look `there`"
02:46 hipster_ joined #salt
02:47 * whiteinge nods
02:47 pentabular joined #salt
02:47 whiteinge that would be a good addition
02:48 malinoff whiteinge, in other words, it would be awesome to have two documentations - for users (the current docs), and for developers
02:49 whiteinge makes sense
02:49 dsolsona joined #salt
02:50 malinoff But please, don't do that like eucalyptus team - pdf with 800 pages is not something that can attract to develop :)
02:51 malinoff without any navigation
02:51 whiteinge ha
02:55 cruatta_ joined #salt
02:56 ashw7n_ joined #salt
02:57 mateoconfeugo joined #salt
02:59 Chris___ joined #salt
02:59 Chris___ hi there
02:59 Chris___ anyone here?
02:59 malinoff Chris___, yep
03:01 zooz joined #salt
03:01 Chris___ i was reading thru the salt table of contents.. but it cant do like change a specific rule? example i wanner change just a small portion of a config in centos.. but others remains the same can it be done?
03:03 malinoff Chris___, since you use a configuration management tool, you have to manage configuration files with it. But, if you really don't want to, you can always use file.replace: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.file.html#salt.modules.file.replace
03:06 Chris___ meaning for example i have 3 pc's. pc1 = server host(192.168.1.1), pc2 = wifi client(192.168.1.2), pc3 = wifi client(192.168.1.3). So with this 3 pc's installed centos, and i need to change the netmask of the both pc2 and pc3. So without touching the ip's change in the config in the netwok i just wanner change the netmask but not the IP thats what i want to do..
03:06 stanchan joined #salt
03:07 Chris___ pc1 saltserver , the 2 are the clients
03:08 malinoff Chris___, have you read the walkthrough? http://docs.saltstack.com/en/latest/topics/tutorials/walkthrough.html
03:08 Chris___ i dont want to replace the files .. if i replace 1 the second pc will follow.. that means the ip also change which follows the saltserver 192.168.1.1?
03:08 malinoff file.replace won't replace any files, it replaces the content by pattern
03:10 ipalreadytaken joined #salt
03:11 whiteinge Chris___: salt can perform any action that you can perform yourself. in many cases salt has a wrapper module to make that nicer/easier. but as malinoff said, you should start with the tutorials.
03:12 Furao joined #salt
03:12 whiteinge !modules.rh_ip.apply_network_settings | Chris___
03:12 wm-bot4 Chris___: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.rh_ip.html#salt.modules.rh_ip.apply_network_settings
03:12 whiteinge ^^ may be helpful
03:12 Chris___ i know before i want to test it .. i need to know does it has this function to do or not..
03:12 Chris___ ok
03:16 zooz joined #salt
03:16 schimmy joined #salt
03:18 krow joined #salt
03:19 bhosmer joined #salt
03:19 bhosmer_ joined #salt
03:21 schimmy joined #salt
03:24 pdayton joined #salt
03:26 pdayton joined #salt
03:32 Chris___ ok just a question..
03:32 hipster joined #salt
03:34 Chris___ i have 100 pc's. IP ranging from 192.168.1.1 ~ 192.168.1.100. i need to change the config netmask to 192.168.1.200 for all of the pc's of 100. Can it be done? centos
03:34 mpanetta joined #salt
03:35 Chris___ using saltstack software to change all
03:36 malinoff Chris___, yes. If you can change something on a single node, you can change something on bunch of nodes
03:36 mpanetta joined #salt
03:37 jalaziz joined #salt
03:39 ajolo_ joined #salt
03:39 whiteinge Chris___: salt -S 192.168.1.1/32 cmd.run 'ifconfig eth0 netmask 255.255.255.0'
03:39 whiteinge ^^ something like that?
03:40 Chris___ yes but if all ? 1-100?
03:40 whiteinge the -S flag targets hosts via CIDR notation. there are many, many other ways to target hosts
03:42 whiteinge the above command will cause all hosts, 1-100, to execute that command and will return the result back to you within a second or two.
03:43 Chris___ if i want to change a line in a config? just that line
03:43 Chris___ for all pc's..
03:43 whiteinge same thing but file.replace instead of cmd.run
03:43 whiteinge (unless you want to cmd.run and call out to sed)
03:44 Chris___ file.replace it can be replace the whole file ?
03:45 whiteinge Chris___: no. file.replace is a Python implementation of the sed workflow
03:45 zain joined #salt
03:47 whiteinge in general the answer to "how do you do XX using Salt?" is "the same way you do XX without Salt". salt just lets you do it with granularity and with great speed, and usually with a simple syntax.  :)
03:48 catpigger joined #salt
03:49 ejo joined #salt
03:56 schimmy1 joined #salt
03:58 jnials joined #salt
04:01 ipalreadytaken joined #salt
04:01 schimmy joined #salt
04:06 schimmy joined #salt
04:08 ashw7n joined #salt
04:19 fragamus joined #salt
04:27 hipster joined #salt
04:29 hipster_ joined #salt
04:40 cnelsonsic joined #salt
04:40 XenophonF left #salt
04:41 ajw0100 joined #salt
04:50 kermit joined #salt
04:53 cpowell joined #salt
04:55 mpanetta hmm
04:59 ajolo joined #salt
05:01 malinoff mpanetta, ?
05:01 mpanetta I should be asleep, and I am not :P
05:01 malinoff mpanetta, that's really strange
05:02 mpanetta Not when you consider that I got a call from work to fix something :P
05:07 hipster joined #salt
05:07 bhosmer joined #salt
05:08 cruatta joined #salt
05:08 wincus joined #salt
05:10 jnials joined #salt
05:21 Ryan_Lane joined #salt
05:26 anuvrat joined #salt
05:27 flebel joined #salt
05:36 Tekni joined #salt
05:44 krow joined #salt
05:44 ejo I'm starting to use salt as a provisioner with vagrant... anyone know how to get vagrant to tell salt to be more verbose?
05:44 ejo (have asked on #vagrant and they appear to be asleep)
05:46 ejo I tried salt.log_level = "debug" or "all" etc. but still all I'm getting as output is that salt.highstate is being called.
05:46 __number5__ ejo: vagrant ssh; sudo -i; salt-call -l debug state.highstate
05:46 ejo mm, good idea... rather than worry about adjusting my Vagrantfile
05:46 ejo __number5__: thanks
05:47 __number5__ you won't get the debug output when doing vagrant provision
05:47 __number5__ ejo: np
05:47 ejo so vagrant is just not set up to pass that through no matter what?
05:47 hipster joined #salt
05:48 __number5__ I don't think so
05:48 ejo cool, this is really feeding my curiosity from the past day, now that i can see what was going wrong in the highstate call
05:48 ejo it's so strange that vagrant doesn't even have the decency to tell me that the highstate call was failing :/
05:49 ejo kind of fits with my general feeling of oddity about vagrant.  It often seems to me as though it just kind of throws stuff at a machine to see which part sticks.
05:49 Ryan_Lane joined #salt
05:55 pentabular joined #salt
06:02 n8n joined #salt
06:08 anuvrat joined #salt
06:09 saravanans joined #salt
06:13 krow joined #salt
06:16 ipalreadytaken joined #salt
06:23 ndrei joined #salt
06:26 roolo joined #salt
06:30 picker joined #salt
06:33 TyrfingMjolnir joined #salt
06:47 hipster joined #salt
06:48 n8n joined #salt
06:52 Ryan_Lane joined #salt
06:56 bhosmer joined #salt
07:04 ml_1 joined #salt
07:07 Outlander left #salt
07:07 TyrfingMjolnir joined #salt
07:09 cruatta joined #salt
07:10 Kenzor joined #salt
07:12 TyrfingMjolnir joined #salt
07:14 zooz joined #salt
07:16 TyrfingMjolnir joined #salt
07:17 n8n joined #salt
07:18 harobed joined #salt
07:19 stanchan joined #salt
07:29 happytux_ joined #salt
07:30 marnom joined #salt
07:30 TyrfingMjolnir joined #salt
07:30 esogas_` joined #salt
07:32 crop joined #salt
07:32 miqui__ joined #salt
07:33 it_dude joined #salt
07:34 sulky_ joined #salt
07:37 ggoZ joined #salt
07:39 troyready joined #salt
07:41 alunduil joined #salt
07:43 crashmag joined #salt
07:43 ilako joined #salt
07:53 Ryan_Lane joined #salt
07:57 mortis anyone know if its possible / will be possible (in the future) to use libvirt as a provider for salt-cloud?
07:59 defunctzombie_zz joined #salt
08:00 nkuttler mortis: well there already is the virt module
08:01 mortis nkuttler: yeah
08:02 mortis but thats a module :)
08:05 darkelda joined #salt
08:13 jcockhren mortis: yeah. salt-cloud was merged into salt
08:13 jcockhren you mean like a backport?
08:14 mortis jcockhren: yeah, thats really nice, and im thinking ...if you could use salt-cloud with libvirt as a provider, you could manage vms in the cloud or on your local libvirt setup without having to go for something like openstack
08:17 jcockhren mortis: hm. I'm not seeing a libvirt module. I see a runner
08:17 jcockhren I don't know/see of any reason why that couldn't be added
08:18 mortis hm yeah
08:18 jcockhren ah. I see the module
08:18 mortis wouldnt it be simpler to relate to only salt-cloud and not salt-cloud and salt-virt together?
08:18 mortis just an idea
08:19 stevednd joined #salt
08:20 jcockhren well. the underlying functionality is already persent in modules.virt
08:20 jcockhren it'll make sense to build the salt-cloud extras on top of that
08:21 jcockhren in fact, that basically how eveything works: built on top of modules
08:21 jcockhren as a provider, have salt-cloud make the right calls to module.virt
08:23 mortis that would be really cool
08:23 mortis i wonder if that is something i should make a github feature request from..
08:23 jcockhren yep
08:23 mortis ill try, and see what response it gets
08:27 anuvrat joined #salt
08:31 ipmb joined #salt
08:38 giantlock joined #salt
08:38 babilen Hi, are there any examples of how to use the pydsl renderer to generate pillars or in file.managed ? I'm not even sure that's possible and http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.pydsl.html doesn't mention it.
08:39 marnom anyone willing/able to test vagrant salt windows guest provisioning?  I've created a pull request here https://github.com/mitchellh/vagrant/pull/3825
08:44 bhosmer joined #salt
08:44 bhosmer_ joined #salt
08:52 zooz joined #salt
08:54 Ryan_Lane joined #salt
08:54 Ryan_Lane joined #salt
09:00 CeBe joined #salt
09:03 TyrfingMjolnir joined #salt
09:09 wigit I was trying to do a recursive watch on a directory as described here: https://github.com/saltstack/salt/issues/663 - but it fails. State and error here: https://gist.github.com/sollie/213eb77c8f20b0d2d87f
09:09 wigit Does anyone know if this feature was changed or scrapped?
09:10 cruatta joined #salt
09:10 andabata /win 19
09:10 andabata whoops
09:12 jab416171 joined #salt
09:21 sgviking joined #salt
09:32 fivethreeo mortis: you can with openstack
09:42 mortis fivethreeo: yeah, but we dont really want to set up such a "complicated" environment for our rather small environment :)
09:43 fivethreeo i almost got it set up ;)
09:43 mortis nice
09:43 fivethreeo https://github.com/fivethreeo/openstack-automation
09:45 mortis sweet, we should prolly start playing with it
09:46 fivethreeo and if you decide to use that one, please share progress :)
09:47 mortis sure
09:47 mortis what distro are you running it on atm?
09:47 fivethreeo got som issues getting the vm's up.. ubuntu 14.04
09:47 mortis what kinda issues?
09:47 hipster joined #salt
09:48 fivethreeo i get a neutron resource does not exist
09:50 mortis we are having a company hackathon, so we might just give it a go now :)
09:50 mortis we found a test-server \o/
09:50 mortis hehe
09:50 fivethreeo :)
09:53 fivethreeo hrm, my salt s not returning :(
09:53 mortis we could prolly change that to have it all run on the same test-server i guess
09:53 mortis :\
09:54 fivethreeo i try do that  :)
09:54 fivethreeo just list the same host
09:54 mortis yeah :)
09:54 fivethreeo but that might be where the issue lies
09:54 Ryan_Lane joined #salt
09:55 che-arne joined #salt
09:55 fivethreeo https://github.com/fivethreeo/openstack-automation/blob/master/Openstack_Icehouse/file/nova/compute_kvm.sls overrides neutrin conf in init.sls
09:55 mortis did you see this? http://saltstarters.org/example/EntropyWorks/salt-openstack
09:56 xt joined #salt
09:56 mortis ah
09:57 fivethreeo yeah, but less documented
09:58 joehoyle joined #salt
10:00 giantlock joined #salt
10:12 dsolsona joined #salt
10:16 N-Mi joined #salt
10:21 shwaiil joined #salt
10:21 shwaiil hi
10:22 fivethreeo mortis: if you are still on openstack i fixed a bug in my formula now
10:22 giantlock joined #salt
10:27 shwaiil Q: Anyone using the https://github.com/saltstack-formulas/apache-formula/ ? I'm having the same issue reported here https://github.com/saltstack-formulas/apache-formula/issues/8
10:27 joehoyle joined #salt
10:32 bhosmer joined #salt
10:39 shwaiil Q: without changing the pillar https://raw.githubusercontent.com/saltstack-formulas/apache-formula/master/pillar.example I'm getting http://pastie.org/private/jfeitftm9zo4merisgfq
10:47 mortis fivethreeo: ait, cool :) we havent quite started yet
10:47 hipster joined #salt
10:48 jcsp joined #salt
10:49 bhosmer joined #salt
10:52 elfixit joined #salt
10:55 Ryan_Lane joined #salt
10:58 ml_1 joined #salt
10:59 xt sigh, gitfs fileserver stopped working after latest update
11:00 kossy joined #salt
11:03 ml_1 joined #salt
11:03 bhosmer_ joined #salt
11:04 kossy joined #salt
11:10 masterkorp hello fellow salters
11:10 oc- hi
11:11 cruatta joined #salt
11:14 rofl____ anyone know if its possible to call a runner, ex virt.init through salt-api?
11:17 oc- There's a bug in loading salt.utils from useradd.py: https://gist.github.com/oc/db8a564fb6eb6789e952
11:17 diegows joined #salt
11:18 rofl____ oc-: filed an issue? my experience is that issues on github are quickly responded to
11:20 Tekni joined #salt
11:21 felskrone joined #salt
11:21 felskrone hey
11:22 shwaiil Q: Is there any working formula for apache and vhost that I could use to learn ? * tried the apache-formula doesn't work out of the box, etc.
11:23 bhosmer_ joined #salt
11:23 picker joined #salt
11:23 shwaiil also found this one https://github.com/bechtoldt/httpd-formula
11:23 tristianc joined #salt
11:23 shwaiil but getting an error
11:24 shwaiil http://pastie.org/private/7xvpszjjdvtir8ibhb2gq
11:27 fivethreeo shwaiil: i suspect the issue is here https://github.com/bechtoldt/httpd-formula/blob/master/httpd/init.sls at the top
11:28 shwaiil fivethreeo: thanks for looking. It's been a few days now, I decided to try to learn saltstack. I'm finding it a bit hard. I now tried to test this formulas etc but they never work
11:30 harobed joined #salt
11:31 fivethreeo Formulas usually has some issues, not generic enough, to generic, hard coded  paths, requires some wonky pillar and wonky yaml ;)
11:32 shwaiil fivethreeo: how can I learn how to create a vhost state by myself ?
11:32 joehoyle joined #salt
11:32 dsolsona joined #salt
11:32 shwaiil fivethreeo: is puppet easier then salt ? it's been a few days I feel I'm wasting time
11:32 jrdx joined #salt
11:33 fivethreeo shwaiil: pernonally i dont understand puppet at all, salt is easier
11:33 jrdx joined #salt
11:33 bhosmer_ joined #salt
11:33 fivethreeo a vhost state should be doable
11:34 fivethreeo shwaiil: https://github.com/fivethreeo/openstack-automation/blob/master/Openstack_Icehouse/file/horizon/init.sls simple example
11:35 oc- bug in useradd.py :~ anybody familiar with this? https://gist.github.com/oc/db8a564fb6eb6789e952
11:36 ml_1 joined #salt
11:36 fivethreeo shwaiil:  or just https://gist.github.com/fivethreeo/de4f0117227a77fde82e
11:42 N-Mi joined #salt
11:43 harobed_ joined #salt
11:46 ml_1 joined #salt
11:49 shwaiil fivethreeo: thanks. but that's an example for apache2, I'm looking for vhosts
11:49 shwaiil http://httpd.apache.org/docs/2.2/vhosts/examples.html
11:50 oc- seems like /usr/lib/python2.7/dist-packages/salt/utils/__init__.py:1968 isn't loaded properly (get_groups_list). Suggestions?
11:51 malinoff joined #salt
11:51 shwaiil how can I check the current salt version I'm working with ?
11:52 marnom shwaiil: salt --version?
11:52 Sypher Tricky question: Is there a salt method available to parse event keys like "file_|-/srv_|-/srv_|-directory" ? Not sure what those parts are exactly and why there are two of the same
11:52 marnom or salt --versions to also report version for certain dependancies like zmq
11:53 shwaiil Thanks :)
11:56 CeBe left #salt
11:56 Ryan_Lane joined #salt
11:56 pdayton joined #salt
12:03 anuvrat joined #salt
12:06 joehoyle joined #salt
12:09 pressureman joined #salt
12:23 JordanRinke joined #salt
12:24 jas-_ joined #salt
12:26 sverrest joined #salt
12:26 TyrfingMjolnir joined #salt
12:30 AdamSewell joined #salt
12:32 sgviking joined #salt
12:34 Furao left #salt
12:39 xmj shalom
12:40 xmj how do I override an included state and add - watch arguments to it?
12:41 xmj I'm trying to generalize a wiki/init.sls state so that the apache2 parts of it are to be included from apache/init.sls, and I want to have two additional files watched (the /etc/httpd/conf.d/* wiki-specific files)
12:42 xmj or can i just do something like
12:42 jslatts joined #salt
12:42 xmj file.managed:\n  - watch_in: service: httpd ?
12:42 jslatts joined #salt
12:42 CeBe2 joined #salt
12:43 rocket joined #salt
12:43 rocket has anyone run into a case where you need to install a file conditionally
12:43 CeBe1 joined #salt
12:43 rocket eg I need to restore the my.cnf only after the databases are restored
12:44 rocket does anyone have any tips on how to tell if the database is restored?
12:49 vexati0n joined #salt
12:55 cpowell joined #salt
12:55 ccase joined #salt
12:56 XenophonF joined #salt
12:57 jeddi rocket: send over a bash script that can return a value when run?  (remember - you can always refactor the ugliness out of your solutions later :)
12:58 rocket thanks jeddi .. I keep forgetting my fallback tricks .. :p
13:09 XenophonF i'm loving salt formulas, but i've got a problem now that i'm adding operating systems not previously supported by my state config
13:09 XenophonF I get errors like "Rendering SLS "base:utils" failed: Jinja variable 'None' has no attribute 'pkgs'".
13:10 XenophonF So what's the best way to go about writing a formula that, if nothing matches, the corresponding state turns into one big no-op?
13:10 resmike joined #salt
13:11 XenophonF should I add a {% if grains['filter_by'] != 'None' %} to the top of every state?
13:11 jaimed joined #salt
13:11 XenophonF i don't think that will DTRT
13:12 cruatta joined #salt
13:13 racooper joined #salt
13:15 jeddi XenophonF: i've taken to wrapping large tracts of state files with similar 'if this grain doesn't exist' or 'if this pillar value isn't there' ... then bypass the section.  it's not truly elegant, but it'll get you working again for OS's that you need fully functional.
13:17 XenophonF jeddi: thanks that's the route i'm going to take
13:17 XenophonF some states in my config make no sense for the target O/S
13:17 XenophonF e.g., configuring NIS on Windows
13:18 XenophonF man, i'd love to see someone's real config
13:18 XenophonF something that manages on the order of 100 computers across BSD, Unix, Linux, and Windows
13:19 SEJeff_work joined #salt
13:19 XenophonF i have no idea whether what i'm writing (a) is idiomatic, (b) makes sense, or (c) can be maintained easily
13:19 nkuttler XenophonF: you could ask the google group for input
13:19 nkuttler i'd rather find my own style tbh, and i'm still working on that
13:20 ajprog_laptop joined #salt
13:22 vejdmn joined #salt
13:25 musinsky left #salt
13:26 quickdry21 joined #salt
13:26 dude051 joined #salt
13:28 jaycedars joined #salt
13:29 hipster joined #salt
13:30 HeadAIX joined #salt
13:32 wincus joined #salt
13:32 MTecknology Wow... I can't figure out why I'm getting this. http://dpaste.com/01E9YY9/  Any thoughts? I see the key get accepted on the syndic box
13:32 XenophonF you know, i'm just going to support Windows/FreeBSD/CentOS
13:32 vejdmn joined #salt
13:33 XenophonF if i need to add support for Debian/Gentoo/Suse/MacOS, I'll add that in later
13:35 MTecknology I support windows/debian/centos
13:35 shadowsun ZMQError: Operation cannot be accomplished in current state
13:35 shadowsun What version of ZMQ are you running?
13:35 XenophonF you know what the salt docs are missing? a nice, top-down, putting-all-the-pieces-together summary at the end of the tutorial
13:36 MTecknology woah...
13:36 XenophonF like, for-real how to support different operating systems in different roles, moderately simple to moderately complex targeting in base.sls, etc.
13:36 XenophonF the existing docs are actually pretty good about building a config from the bottom up
13:37 MTecknology I thought I updated this box, but it's on 0.17.4, the master is on 2014.1.4
13:37 MTecknology that'd break it
13:37 shadowsun MTecknology: yep
13:38 MTecknology How does something like this happen?  http://i.imgur.com/75ZPKCu.jpg
13:38 shadowsun meteal shaving gets between the traces or secondary side on the PSU blew or a serious surge usually
13:38 * shadowsun detaches this terminal
13:38 shadowsun work time.
13:39 ipmb joined #salt
13:40 MTecknology shadowsun: thanks for making me check the versions report
13:44 rojem joined #salt
13:49 Tekni joined #salt
13:54 kaptk2 joined #salt
13:54 joehoyle joined #salt
13:54 darkelda joined #salt
13:55 redo joined #salt
13:55 redo how to install mine.get.docker?
13:56 xmj "I'm not slacking of, my salt highstate is running."
13:56 plastik_ joined #salt
13:56 timoguin xmj: ah yes, the new "Compiling!"
13:57 xmj give me all ur retweetz https://twitter.com/xmjEE/status/468752426770915328
13:58 harobed joined #salt
13:59 harobed joined #salt
13:59 GradysGhost joined #salt
14:00 abe_music joined #salt
14:04 vejdmn joined #salt
14:04 rocket what does it mean if salt-call state.highstate runs and works properly but salt 'dev' state.highstate has Failures?
14:04 rocket why would there be a difference?
14:07 ajolo joined #salt
14:10 mgw joined #salt
14:10 wendall911 joined #salt
14:13 edsonyf joined #salt
14:14 mikeputnam left #salt
14:16 rgbkrk joined #salt
14:21 debaser joined #salt
14:21 wendall911 joined #salt
14:21 debaser hello
14:21 jeremyBass joined #salt
14:23 debaser is it possible to insert a line at the specific place in a text file using file state?
14:26 __hudson__ joined #salt
14:26 timoguin debaser: yes, there is a file.replace state
14:26 timoguin where you can specify blocks of text to replace
14:28 lionel joined #salt
14:28 oz_akan_ joined #salt
14:28 joehoyle joined #salt
14:29 debaser I was looking at file.replace but I'm not sure if it can do what I need
14:29 debaser I need to find a pattern in text and add a line after it, but not replace it
14:29 plastik_ left #salt
14:30 debaser and I dont know the exact line that I'm looking for, it may change
14:31 felskrone joined #salt
14:31 debaser can I somehow store the found pattern and prepend it to the line I'm replacing it with?
14:31 debaser xD
14:32 debaser I know i can do it with sed or pure python but I'd rather stick to salt modules if I can
14:32 tyler-baker joined #salt
14:32 svs joined #salt
14:33 timoguin it looks like file.replace and/or file.sed should be able to do that with regex, but i haven't done it
14:33 timoguin and the docs are a bit scarse
14:34 JesseCW joined #salt
14:34 mateoconfeugo joined #salt
14:35 Shish joined #salt
14:35 ashw7n joined #salt
14:35 taion809 joined #salt
14:36 dccc joined #salt
14:38 faldridge joined #salt
14:39 hipster joined #salt
14:39 ajacoutot joined #salt
14:40 ajacoutot hi
14:40 alunduil joined #salt
14:40 ajacoutot pretty new to salt; is there *any* way for pillar to ignore a missing sls ?
14:40 N-Mi joined #salt
14:40 N-Mi joined #salt
14:41 XenophonF debaser: use backreferences in your regex
14:41 XenophonF hang on let me find an example for you
14:43 saravanans joined #salt
14:43 wincus joined #salt
14:44 JeroenH_ joined #salt
14:45 opapo joined #salt
14:46 mpanetta joined #salt
14:46 elfixit joined #salt
14:48 joehoyle joined #salt
14:51 N-Mi joined #salt
14:51 N-Mi joined #salt
14:54 jalbretsen joined #salt
14:56 debaser thanks timoguin and XenophonF
14:56 debaser I've found an example with backreference
14:56 XenophonF ah good
14:56 XenophonF sorry got distracted by my boss :-/
14:57 MTecknology she cute?
14:57 XenophonF what?
14:57 timoguin hey that's racist
14:57 jcockhren XenophonF: bosses can be distracting
14:57 XenophonF no we have a network down
14:57 bhosmer joined #salt
14:58 AdamSewell joined #salt
14:58 XenophonF debaser: i uncomment pam_krb5 using file.replace
14:58 XenophonF the pattern looks like '#(.*pam_krb5.*)'
14:58 Ryan_Lane joined #salt
14:58 XenophonF and the repl is '\1'
14:58 bhosmer_ joined #salt
14:58 XenophonF you can do the same thing in sed/awk/vi/emacs/perl/etc.
15:00 pentabular joined #salt
15:00 mpanetta joined #salt
15:00 XenophonF debaser: here's the complete state, if you want a moderately complicated example - http://paste.debian.net/100825/
15:05 n8n joined #salt
15:05 debaser thanks XenophonF thats exactly what I was looking for
15:05 mpanetta joined #salt
15:06 tpaul joined #salt
15:08 frasergraham joined #salt
15:08 it_dude joined #salt
15:08 bhosmer joined #salt
15:11 bhosmer joined #salt
15:13 cruatta joined #salt
15:15 DerekRBN joined #salt
15:15 bhosmer_ joined #salt
15:16 DerekRBN Hey all, Quick Yaml question, I'm using pillar and i have a string that ends in {{}} but its getting evaluated to () quoting doesnt stop it
15:16 smcquay joined #salt
15:16 jaycedars joined #salt
15:16 DerekRBN I've tried excaping and using | any ideas?
15:16 pdayton joined #salt
15:17 ipalreadytaken joined #salt
15:17 jchen what if you do something like https://gist.github.com/fly/f301414963cff84307b1
15:18 masterkorp Can I set template data on a pillar ?
15:19 masterkorp like i have a parameter that takes an Ip, can I template the like take the IP of certain minion ?
15:19 DerekRBN @jchen thanks for the feed back, that still evaluated to ()
15:20 jchen DerekRBN: damn, sorry man. not sure that's a YAML thing anymore
15:21 DerekRBN It did as it would but still not taking it literally
15:21 happytux joined #salt
15:21 Ryan_Lane joined #salt
15:23 MTecknology salt-call works perfect and great if I have "master: salt.domain.tld" but if I have "master: [ syndic1.domain.tld, syndic2.domain.tld ]" it'll do nothing for a long time and then spit out zmq errors
15:24 manfred which is why you have a --master flag for salt-call
15:25 mateoconfeugo joined #salt
15:25 raizyr joined #salt
15:25 MTecknology oooh
15:26 manfred :)
15:26 meteorfox joined #salt
15:26 MTecknology Can that be set in the config?
15:26 manfred that i have no idea
15:26 MTecknology default_master: ..  If not, it should be
15:26 manfred i haven't used it yet
15:26 meteorfox joined #salt
15:27 XenophonF debaser: np glad that helps
15:27 manfred i only learned about it when i was writing the zshell completion the other night at 2 am... then forgot about it until now
15:27 MTecknology nice, I'm glad you saw it
15:27 MTecknology It seems to still give me the same headache
15:28 manfred MTecknology: you can set random_master: True
15:28 manfred http://paste.gtmanfred.com/mDifv/
15:28 redo if I was to write a module for salt, how should I go about?
15:29 AdamSewell joined #salt
15:29 manfred redo: the key is to use __salt__ as much as possible, and sanatize inputs
15:29 manfred redo: also needs a __virtual__() and document every function, even the private ones
15:29 manfred redo: https://github.com/saltstack/salt/blob/develop/salt/modules/glusterfs.py
15:29 manfred that was my first one
15:30 MTecknology heh... new error now
15:30 MTecknology manfred: any thoughts here? http://dpaste.com/11THF4S/
15:31 ZombieFeynman joined #salt
15:31 MTecknology this all works fine if I change it back to pointing at only the master of masters
15:31 manfred are you on 2014.1.x?
15:33 manfred it looks like that whole portion has been redone in develop, so this may have been fixed
15:33 MTecknology .4, ya
15:33 raizyr joined #salt
15:34 ipalreadytaken joined #salt
15:34 manfred nevermind https://github.com/saltstack/salt/blob/develop/salt/fileclient.py#L986
15:34 manfred still in develop
15:35 manfred i am not famiiliar with much about the fileclient.py yet... i would just plug that into the issue tracker
15:37 deares joined #salt
15:38 shadowsun MTecknology: welcome
15:39 JordanRinke joined #salt
15:41 ashw7n joined #salt
15:41 joehoyle joined #salt
15:42 UtahDave joined #salt
15:42 joehoyle joined #salt
15:43 tligda joined #salt
15:43 ashw7n joined #salt
15:43 stanchan_ joined #salt
15:48 fragamus joined #salt
15:51 faldridge joined #salt
15:53 MTecknology manfred: fixit!
15:53 MTecknology I made the issue, you're turn to fix it!
15:53 mgw joined #salt
15:54 conan_the_destro joined #salt
15:54 manfred if i get a minute to look into it, i will try :P
15:56 ajolo joined #salt
15:57 cruatta joined #salt
15:57 cruatta_ joined #salt
15:58 fragamus joined #salt
15:58 jhulten joined #salt
15:59 KennethWilke joined #salt
15:59 ipmb joined #salt
16:01 joehoyle joined #salt
16:01 MTecknology manfred: it's interesting... against the master, it all works perfect; against the synidcs it goes kerblam; I'm rather certain these syndics are set up just like the master
16:01 MTecknology even the ssh keys match what's on the master
16:03 AdamSewell joined #salt
16:03 AdamSewell joined #salt
16:04 whiteinge rofl____: you can call runners via salt-api by specifying client='runner' in the request body
16:05 repl1cant joined #salt
16:07 faldridg_ joined #salt
16:07 schimmy joined #salt
16:07 hipster_ joined #salt
16:08 KyleG joined #salt
16:08 KyleG joined #salt
16:08 ajw0100 joined #salt
16:09 schimmy1 joined #salt
16:13 faldridge joined #salt
16:15 stephas joined #salt
16:19 linjan joined #salt
16:19 forrest joined #salt
16:22 jcockhren so much win: https://github.com/saltstack/salt/issues/11147
16:22 jcockhren and the fact that we're only a couple months behind ansible is pretty amazing
16:22 dsolsona joined #salt
16:23 mpanetta jcockhren:  Oh sweet
16:24 DerekRBN jchen: I figured out my issue wasnt with yaml or salt but more so jinja, should have figured because what i was using was {{}}, I ended up just putting something in there and using sed to change it after the fact.
16:25 ashw7n joined #salt
16:25 DerekRBN Thanks for your suggestion!
16:26 cruatta joined #salt
16:26 cruatta_ joined #salt
16:28 ldlework joined #salt
16:28 ldlework how do I salt call specific states from my state tree?
16:28 timoguin state.sls foo
16:29 ldlework thanks
16:30 conan_the_destro joined #salt
16:31 _gothix_ joined #salt
16:31 dlindquist joined #salt
16:33 vejdmn joined #salt
16:33 dlindquist I'm seeing additional bug fixes going into https://github.com/saltstack/salt/commits/2014.1.  Is a 2014.1.5 release planned?
16:33 XenophonF we hope so :)
16:34 XenophonF was asking UtahDave that very question last night
16:35 forrest timoguin, are you participating in the sprint today or this evening?
16:35 timoguin yes, i should be able to do some things this evening.
16:35 forrest cool
16:36 Ahlee i need to pull out the lowercase of hte non-fqdn of the minion in this state
16:36 faldridge left #salt
16:36 cruatta joined #salt
16:36 cruatta_ joined #salt
16:37 Ahlee best to do it in jinja, I guess?  {% set shortname = grains['hostname'].tolower() or something?
16:37 timoguin forrest: you?
16:37 dlindquist XenophonF: Thanks, I see that beginning of June was mentioned in yesterday's chat for the "next release".  Not sure if that is the next major version or the planned date to do a 2014.1.5 release.
16:38 forrest timoguin, in the evening I plan to do some work yes
16:38 UtahDave XenophonF: dlindquist, We're hoping to have the first release candidate for the next feature release in the next couple of weeks.
16:38 XenophonF UtahDave, that's great!
16:39 UtahDave We MAY do a 2014.1.5 if it looks like it's needed.
16:39 XenophonF hey, would you all consider switching to an MSI installer for Windows if someone contributed one?
16:39 smcquay_ joined #salt
16:39 Ahlee I would be.
16:40 UtahDave XenophonF: Yeah, definitely.  In fact, if you'd like to work with me to make it part of our official build process, that would be great.
16:40 joehillen joined #salt
16:41 XenophonF yes
16:41 XenophonF very much
16:41 XenophonF UtahDave: would you want me to send diffs against -HEAD?
16:41 KyleG1 joined #salt
16:41 XenophonF I was doing a first pass using the 2014.1.4 branch.
16:41 UtahDave Right now we're using NSIS, but not because it's that great.
16:42 timoguin I'd really like to get the windows builds automated. Is there a gotcha reason that hasn't been done or just time?
16:42 XenophonF UtahDave, timoguin: Is the Windows build environment documented anywhere?
16:42 timoguin it is, on the windows installer page
16:42 XenophonF OK
16:43 UtahDave timoguin: we're not that far off from automating it completely. The main problem is that I do a lot of manual testing on windows before packaging.
16:43 UtahDave We don't have integration tests running on windows yet
16:43 timoguin That's what I figured
16:43 UtahDave I have a 32bit and a 64bit vm that have the environment built and i use those to create the installers.
16:43 XenophonF OK
16:44 XenophonF UtahDave: which version of Windows do you use to host those builds?
16:46 UtahDave They're VMs on EC2.  I think they're both Windows Server 2008.
16:46 XenophonF OK
16:46 XenophonF let me throw a similar build env together here
16:47 ramteid joined #salt
16:48 UtahDave cool.
16:48 CeBe3 joined #salt
16:49 Networkn3rd joined #salt
16:49 UtahDave XenophonF: What do you use for creating the MSIs?  I've used wix a little bit in the past.
16:49 XenophonF wix
16:49 XenophonF UtahDave: actually, wix + emacs :)
16:49 * goodwill bellybumps UForgotten
16:49 * goodwill bellybumps UtahDave
16:49 goodwill ooopsy
16:50 UtahDave :)  I was wondering where my bellybump was!
16:50 UtahDave XenophonF: cool.  I'm wondering if we can set that up on the same VMs as my current NSIS setup
16:51 XenophonF yes
16:51 XenophonF WiX doesn't have any special requirements IIRC
16:52 UtahDave so right now, we do an  esky/bbfreeze  frozen build.  Then we use NSIS to create an installer that puts salt in C:\Salt
16:52 TheRealBill_ joined #salt
16:52 UtahDave We have a bunch of open requests to make the install location a more standard location, and also to make it adjustable. Several things like that
16:53 XenophonF that should all be doable
16:54 ashw7n_ joined #salt
16:55 Ryan_Lane joined #salt
16:55 travisfischer joined #salt
16:56 ashw7n_ joined #salt
16:56 XenophonF UtahDave: i am spinning up a windows server 2008 r2 instance now to do the development; am also forking saltstack/salt
16:57 UtahDave awesome.
16:57 XenophonF UtahDave: let me see what i can put together quickly for you to test
16:57 UtahDave FYI, i found that setting up pip and easy_install was the best way to get all the deps, rather than trying to collect them from around the internet
16:58 XenophonF OK
16:58 UtahDave easy_install because most of them have compiled versions to install
17:01 frasergr_ joined #salt
17:02 UtahDave XenophonF: seems like I might have had to still install Visual Studio 2008 for something though.
17:04 fraserg__ joined #salt
17:04 ZombieFe_ joined #salt
17:04 ajacoutot left #salt
17:05 tyler-baker_ joined #salt
17:06 jnials joined #salt
17:06 rgbkrk joined #salt
17:07 raizyr joined #salt
17:07 ml_1 joined #salt
17:08 ndrei joined #salt
17:09 ghartz_ joined #salt
17:09 jgarr does the salt bootstrap support rhel systems registered with subscription manager instead of rhn classic?
17:09 jgarr getting an error that the optional repo is not found. Just wanted to see if it's supported before I move forward troubleshooting
17:09 forrest jgarr, Yes I wrote that error jgarr
17:10 forrest jgarr, some asshat decided that it would be a good idea to put jinja2 in the optional repo for rhel, so if you try to use the EPEL one it will error out
17:10 forrest jgarr, it has a different naming scheme
17:10 forrest so just enable the optional repo and that check should pass.
17:11 jaycedars joined #salt
17:11 jgarr forrest: I already did enable optional with yum-config-manager (it shows up under repolist)
17:11 forrest oh really?
17:11 forrest hmm
17:12 stanchan joined #salt
17:13 jgarr forrest: let me see if I can actually get something from it to doublecheck. It worked fine on my rhn classic servers
17:13 XenophonF UtahDave: I think I have a Visual Studio 2008 Pro license available.
17:13 kermit joined #salt
17:13 jgarr forrest: btw this is all rhel 6. This particular box is 6.3 if that makes any difference
17:13 vejdmn joined #salt
17:13 forrest jgarr, yea I did my testing against a rhel6 machine
17:13 UtahDave XenophonF: The free version works just fine.  although sometimes it's hard to find these days
17:14 forrest jgarr, when you run rhn-channel -l | grep optional, what does that return?
17:15 forrest jgarr, hmm, it looks like someone modified the existing check... https://github.com/saltstack/salt-bootstrap/commit/22a65c83005fa951a59a8cc2bdc5b14d603ca3e9
17:15 resmike joined #salt
17:15 frasergraham joined #salt
17:16 forrest jgarr, in fact there have been quite a few updates since I initially put that check in
17:17 forrest jgarr, if you are sure you have optional (and therefore jinja), just have https://github.com/saltstack/salt-bootstrap/blob/develop/bootstrap-salt.sh#L2502 and https://github.com/saltstack/salt-bootstrap/blob/develop/bootstrap-salt.sh#L2505 return 0
17:17 cvale joined #salt
17:17 forrest and if you want to troubleshoot it further please do and submit a PR with a fix, otherwise open an issue.
17:17 jgarr forrest: the check returned rhel-x86_64-server-optional-6
17:17 forrest but yea, this has changed quite a bit since I initially wrote the check
17:18 forrest jgarr, yea, there's a big chunk of new code in there
17:18 aw110f joined #salt
17:19 forrest if [ $DISTRO_MAJOR_VERSION -eq 6 ] && [ $(rhn-channel -l | grep optional) != "rhel-${OPTIONAL_ARCH}-server-optional-${DISTRO_MAJOR_VERSION}" ]; then
17:19 forrest that was the old check I initially wrote that worked
17:19 forrest but like I said, just modify those two lines to return 0s instead of 1s, and it will continue on
17:19 forrest Sorry about that :(
17:20 jgarr forrest: k thanks, I'll see what I can do. I have a lot of boxes to push this out to but I'm not sure I have time to find a proper fix
17:21 forrest no worries, if you don't just file an issue with the error and as much detail as possible
17:22 cvale salt-2014.1.0-1.el6.noarch on CentOS 6.5 does not seem to be honoring group.present - members.
17:22 cvale Thought I would check here before submittting a bug.
17:22 cruatta_ joined #salt
17:22 cruatta joined #salt
17:24 jgarr forrest: haha, I ran rhn-channel -l on the wrong window. The box with the problem returns "Unable to locate SystemId file. Is this system registered?" but the system shows up in subscription manager on rh's customer portal website
17:24 jgarr I'll try to clear up my reg error before hacking up the boostrap
17:25 tyler-baker joined #salt
17:26 frasergraham joined #salt
17:26 HeadAIX joined #salt
17:26 cvale When I try to add existing users to an existing group, they do not get added but salt returns True with comment No Change.
17:27 forrest jgarr, ok cool
17:28 ndrei joined #salt
17:30 babilen cvale: Could you paste your state and the output to http://paste.debian.net and share it with us?
17:31 gq45uaethdj26jw6 joined #salt
17:31 toastedpenguin joined #salt
17:32 gq45uaethdj26jw6 hey there, working with joyent and having problems with salt-cloud identifying the images. using their cli tool, i get the image names just fine. salt-cloud with --list-images, however, returns an empty result. other providers are working fine, but I can't get joyent to go. anyone have ideas?
17:33 liquid_tool joined #salt
17:34 _jslatts joined #salt
17:34 XenophonF cvale: you will need modules/groupadd.py and states/group.py from salt's current branch
17:34 XenophonF cvale: i ran into the same issue with salt 2014.1.4 on both centos 6 and freebsd 10
17:35 n8n joined #salt
17:35 cvale babilen, certainly.
17:35 cvale XenophonF, Oh?
17:36 cvale XenophonF: Do I just drop them in place of existing modules?
17:36 XenophonF cvale: create _modules/ and _states/ in the appropriate file_root location
17:36 krow joined #salt
17:37 XenophonF cvale: that's on the master; i'm running freebsd, so my file_root is /usr/local/etc/salt/states
17:38 XenophonF cvale: then copy modules/groupadd.py (again, from the current dev branch of salt) to _modules/, and copy states/group.py to _states/
17:39 XenophonF cvale: then do a state.highstate - that will copy just those files to your minions and re-run the various states
17:41 KyleG joined #salt
17:41 KyleG joined #salt
17:41 XenophonF cvale: I ran into the exact same issue as you yesterday.
17:41 babilen well done :)
17:41 zain joined #salt
17:42 gq45uaethdj26jw6 alternatively, is there a debug mode that lets me see that raw data returned from a salt-cloud api call?
17:42 XenophonF cvale: my new friends on #salt here walked me through something similar when i hit a bug in states/mount.py, so that should be in the IRC log some time in the last week or so
17:43 mpanetta gq45uaethdj26jw6: Have you tried enabling debug or trace level logging?
17:43 gq45uaethdj26jw6 mpanetta: I'm using -l all
17:43 gq45uaethdj26jw6 mpanetta: I can see the request being made, and that it returns 200, but I need to know if it's returning data or not, or an empty object
17:43 mpanetta Ah
17:43 mpanetta Hmm...
17:45 gq45uaethdj26jw6 mpanetta: here's a snippet http://pastebin.com/DJZdwLD9
17:45 gq45uaethdj26jw6 other commands work. --list-locations works, for example
17:46 gq45uaethdj26jw6 thoughts?
17:46 jgarr forrest: I decided just to not use the boostrap. Everything I need is available in rhel optional, and epel-testing right?
17:47 forrest jgarr, yep, you can just install the salt master, then the minion package and you should be good to go. There should be a package in epel stable as well, but it will probably be a slightly older release.
17:47 jgarr ya, there was a bug in 2014.1.0 with yum
17:48 jgarr .1 fixed it but testing has .4
17:48 forrest gotcha
17:48 jgarr thanks for the help
17:48 forrest yea np
17:49 mpanetta gq45uaethdj26jw6: Hmm, sorry I am stumped :(  Maybe use tcpdump to wtach the API traffic?
17:49 jgarr oh, one other question. Why do my new minions have their node name the fqdn instead of just the hostname? Where is that pulled from whet it's not explicit in /etc/salt/minion?
17:49 gq45uaethdj26jw6 mpanetta: i think it will be wrapped in ssl
17:50 jgarr it's annoying because the key changes name after I have highstate explicitly put the hostname in /etc/salt/minion and needs to be reaccepted
17:50 forrest jgarr, I believe it pulls that from a python call to network
17:50 forrest jgarr, there's a doc on it somewhere.. timoguin do you remember which doc has the order in which salt looks for the minion name?
17:51 jgarr so it's actually polling from dns? how does that work with dns aliases? that could become a mess for me
17:51 mpanetta gq45uaethdj26jw6: Technically if you have the certs tcpdump can decrypt the traffic, all though I have never done that before.
17:52 forrest jgarr, it's probably socket.getfqdn(), but I can't remember what the ordering is that it looks. I thought it was hostname first, then that...
17:53 jgarr forrest: I think I just found it http://docs.saltstack.com/en/latest/topics/tutorials/walkthrough.html#minion-id-generation
17:53 jgarr /etc/hostname is first
17:53 jgarr i mean second
17:54 jgarr I'll have my install playbook create that file
17:54 forrest cool
17:58 chrisjones joined #salt
17:58 rocket joined #salt
17:59 jnials joined #salt
17:59 cvale XenophonF: I'll give that a shot. Thanks.
18:00 ipalreadytaken joined #salt
18:00 tyler-baker left #salt
18:01 ashw7n joined #salt
18:01 kickerdog joined #salt
18:02 giannello joined #salt
18:02 linjan_ joined #salt
18:03 ZombieFeynman joined #salt
18:03 ipalreadytaken joined #salt
18:04 kermit joined #salt
18:04 kermit joined #salt
18:05 ZombieFeynman joined #salt
18:05 chow joined #salt
18:05 giannello heya all, is there a way to debug pillar distribution from the master to the minions?
18:05 giannello it seems that my /srv/pillar/top.sls is ignored, I cannot see pillars on any minion
18:06 joehoyle joined #salt
18:07 chow is my syntax correct? trying to retrieve ip address of eth0 from all minions "salt '*' grains.item ip_interfaces.eth0... because is not working...
18:07 chow salt '*' grains.item ip_interfaces without the eth0 portion works
18:07 chow but gives me all interfaces...
18:08 manfred chow: salt '*' grains.get 'ip_interfaces:eth0'
18:08 chow ah
18:08 chow I like
18:08 chow thanks
18:08 manfred np
18:09 rocket_ joined #salt
18:10 saravanans joined #salt
18:15 cruatt___ joined #salt
18:15 cruatta__ joined #salt
18:16 chow can i do "grains.get 'ip_interfaces:eth0'" inside a jinja file?
18:16 chow to replace with my ip address?
18:17 manfred yes
18:17 ZombieFeynman joined #salt
18:17 manfred {{ salt['grains.get']('ip_interfaces:etho') }}
18:17 chow amazing
18:18 chow thanks
18:18 manfred np
18:21 jnials joined #salt
18:22 Dewara joined #salt
18:22 Dewara Hello all....I have a few questions ....I have been working with the salt cron state and module .....
18:22 stevednd curious for some thoughts from you guys. I have a few web applications that I want to deploy. They all follow the same basic pattern. app specific user, nginx config file, deploy via pulling a git repo, etc... The only thing that changes in each one is the specific values for those activities(username, app name, repo location, etc...). I was hoping to have one basic app.base sls file that I could include in each app specific sls
18:22 stevednd file, but can't come up with a good way to do it
18:23 manfred chow: yeah, i am not sure that grains.item can pull anything farther down than the top level one
18:23 Dewara However I have not been able to remove any cron that was not put there by salt ....
18:23 Dewara how should i handle that ?
18:23 manfred stevednd: you can do that, just use pillars to assign the information for each one, then put the server in multiple highstates by changeing the environment, which will change the pillar data
18:24 stevednd a server can have more than one app, and each app needs to be independently deployable,  so I can't just have different pillar or grain information per server
18:24 ZombieFeynman joined #salt
18:24 chow yea pillar should do it. I have similar scenerio
18:24 manfred stevednd: use saltenv
18:24 manfred to set the pillar data differently
18:24 manfred then you just deploy the same sls/highstate
18:25 jalaziz joined #salt
18:26 JasonSwindle joined #salt
18:26 JasonSwindle UtahDave: Howdy!
18:27 bhosmer joined #salt
18:27 manfred pfft, this guy
18:27 JasonSwindle manfred: Hello to you too :P
18:27 manfred
18:27 XenophonF stevednd: I think the answer to your question is to combine the states with pillar.
18:28 manfred he will want to combine the environments too
18:28 rgbkrk joined #salt
18:28 manfred or
18:28 manfred i guess you could just have a pillar that assigns which things go on each server, and then do a for loop on each config thing for each site
18:29 Bandikoto joined #salt
18:29 JasonSwindle With the event bus, when you do fire_master, and the master is in syndic..... it should go to the master of masters.  Right?
18:30 stevednd manfred: so instead of just having my environments that I have now of base, dev, stage, prod I'm going to need all of those *n applications?
18:30 manfred nah
18:30 joehoyle joined #salt
18:30 manfred that was a weird idea
18:30 manfred stevednd: what you should do is just use pillars
18:30 chrisjones joined #salt
18:30 manfred stevednd: have one pillar that says which applications should be installed on each minion... then have a for loop in ever sls file that loops through each of those assigned applications, and changes the states in the file
18:31 Eugene if is your friend
18:31 stevednd manfred: is there any way to know the name of the file being being run by salt?
18:32 stevednd I'm still struggling to see how looping a pillar would allow me to deploy each app individually
18:32 manfred yes, but i don't remember it right this second
18:32 manfred stevednd: ones econd, I am marking up an example environment
18:32 stevednd okay, thanks
18:32 XenophonF manfred: I want to see your example, too. :)
18:32 ZombieFe_ joined #salt
18:33 z3uS joined #salt
18:33 jslatts joined #salt
18:34 chow manfred: this is what I put in my jinja file  "rpc_address: {{ salt['grains.get']('ip_interfaces:etho') }}"
18:34 manfred eth0
18:35 chow but when I run it the rpc_addresss comes back blank
18:35 manfred eth0
18:35 manfred not etho
18:35 chow right...
18:35 chow wow
18:35 manfred :P
18:35 chow sorry...
18:35 chow "slowly walking away..."
18:35 chow thanks
18:36 manfred np :P
18:37 joehoyle joined #salt
18:38 jnials joined #salt
18:38 chow after all that embarrassment....I don't think it will work....the ip comes out in brackets. I just need the mere ip
18:38 manfred just put a [0]
18:39 chow o
18:39 chow k
18:39 manfred XenophonF: stevednd this is really rough, and doesn't do a lot... but http://paste.gtmanfred.com/f9cq4T/
18:39 manfred that is how I would do it
18:39 chow :)
18:39 manfred all it does is make the document roots
18:39 manfred bah
18:39 cruatta joined #salt
18:39 cruatta_ joined #salt
18:39 manfred this http://paste.gtmanfred.com/k7y9n1/
18:39 manfred but you can see how you could expand the whole thing
18:39 XenophonF thanks manfred
18:41 manfred or better yet... now i can't rememebr if you can use include in a pillar file
18:41 manfred yup
18:41 manfred XenophonF: stevednd http://paste.gtmanfred.com/CfL1N/
18:41 manfred there
18:42 manfred now you just have one pillar file that would incldue the information about each appication, and then you just have that one pillar per server
18:42 manfred ifif I stopped finding mistakes stevednd XenophonF http://paste.gtmanfred.com/ZvD/
18:45 chrisjones joined #salt
18:46 oz_akan_ joined #salt
18:46 jnials joined #salt
18:47 XenophonF thanks manfred!
18:48 it_dude joined #salt
18:49 jnials joined #salt
18:50 saguilar joined #salt
18:51 joehoyle joined #salt
18:51 jcsp joined #salt
18:52 saguilar joined #salt
18:53 z3uS| joined #salt
18:54 jnials joined #salt
18:54 Dewara We are working with the cron module and state....we can add and remove cron entries using them but only if managed by salt ......when we try and remove a job from a crontab entry that is not controled by salt it wont let us do that ....any suggestions on how we go about removing the entry
18:55 giannello oh, wow, looks like setting "environment: development" in the minion config makes pillars unreadable. If I remove the "environment" from the minion config, it works.
18:55 KyleG1 joined #salt
18:57 giannello I wonder why it doesn't behave like with states - going up to the "base" environment if something is not matched
18:59 rgbkrk joined #salt
19:00 surge joined #salt
19:01 krow joined #salt
19:03 ZombieFeynman joined #salt
19:04 n8n joined #salt
19:05 surge hey all
19:05 surge i have a problem
19:05 jcockhren we all do
19:05 jcockhren :_
19:05 jcockhren ;)
19:05 surge lol
19:05 chow yup
19:05 surge salt '*' disk.blkid
19:05 surge 'disk.blkid' is not available.
19:05 XenophonF oh, you have a salt problem
19:05 surge on any of my minions
19:05 surge Centos 6.5
19:05 XenophonF well, we can help with that :)
19:05 surge why would this be
19:06 eagen joined #salt
19:06 XenophonF surge: what version of salt do you have?
19:07 manfred surge: https://github.com/saltstack/salt/blob/2014.1/salt/modules/disk.py
19:07 manfred surge: unless you are using salt directly from git/develop, no currently tagged version has disk.blkid
19:07 surge [root@kraken salt]# salt --version salt 2014.1.0
19:07 manfred only develop branch has that module
19:08 surge oh
19:08 surge i see
19:08 surge is it not ready for prime time..?
19:08 manfred it just wasnt in the module when 2014.1 was tagged
19:08 surge oic
19:08 surge thank u
19:08 manfred it will be in 2014.2 which is going to be in rc1 soonish
19:09 timoguin which will actually be 2014.5 or 2014.6, depending on when the first RC is cut
19:10 surge do they have like a giant pair of scissors
19:10 surge O_~
19:10 ashw7n joined #salt
19:11 cheus Does anyone know, when jinja+yaml parse a pillar, is the data stored in a sequenced structure (eg, odict) ?
19:11 ashw7n joined #salt
19:11 manfred it should be
19:11 cvale XenophonF: Thanks again for your group.present fix. That did the trick.
19:12 nkuttler cheus: probably, considering that order matters since a few releases
19:12 XenophonF cvale: Sweet!  Glad to hear it.
19:12 defunctzombie joined #salt
19:13 forrest Is anyone in here located in the Seattle area?
19:13 mpanetta I wish :P
19:14 cheus Ok. I would expect it to be so with states but wasn't sure if pillar was parsed along a similar means. Makes me very nervy relying on that in formulas.
19:17 Dewara is there a better search and replace/delete option other then file.sed for line removal in a file?
19:17 krow joined #salt
19:18 surge another problem :|
19:18 surge i have 4 accepted keys on this master
19:18 surge sudo salt "*" test.ping
19:19 surge i only get three responses from the minions
19:19 steveoliver joined #salt
19:19 surge hrmzzzz
19:19 nkuttler surge: start by running the missing minion with -l debug
19:20 chow anyone has additional good docs on mines...
19:20 chow having a hard time with them..
19:20 rgbkrk joined #salt
19:20 chow tried this...http://docs.saltstack.com/en/latest/topics/mine/ but is not working for me
19:20 chow ...
19:21 paletto joined #salt
19:21 paletto hi
19:22 stevedndm joined #salt
19:23 agronholm joined #salt
19:24 stevedndm manfred: thanks. Not quite what I was hoping for, but I'll play around with it. We're you able to find a way to identify the file salt is currently running?
19:24 agronholm hi, I'm a tad confused about salt states -- is the parent label in the yaml configuration given to the state as the first argument?
19:24 manfred stevedndm: yes, it is ... {{ state }} iirc
19:25 manfred we just talked about this the other day... lemme find it
19:25 Vye Does it make sense to modify virtualenv_mod to set the 'env' key when passing virtualenv commands to cmd.run_all in order to supply a custom LD_LIBRARY_PATH environment variable? Or is there an easier way?
19:25 ipmb joined #salt
19:25 manfred stevedndm: in the state file.. it is {{ __sls__ }}, and then in side a jinja file i think it is {{ state }} and {{ file }}
19:25 AndChat-482736 joined #salt
19:25 XenophonF agronholm: i think that by default, the state's ID is passed to the state as the name argument
19:26 cvale left #salt
19:26 AndChat-482736 joined #salt
19:26 XenophonF agronholm: so you can have /etc/issue: file.managed: ...
19:26 manfred stevedndm: https://github.com/saltstack/salt/pull/12832 this was just added to develop to help assist with figuring those out
19:26 XenophonF agronholm: or you can have set login banner: file.managed: name: /etc/issue
19:26 agronholm XenophonF: makes sense, thanks
19:26 AndChat-482736 It would also be great if you could pass variables to a state file that is being included in another sls
19:27 simonmcc joined #salt
19:28 Ryan_Lane whiteinge, forrest: howdy. the doc sprint has started, right?
19:28 whiteinge indeed
19:28 Ryan_Lane where's the info on joining in?
19:28 manfred stevedndm: {{ source }} is the file it is currently acting on as well iirc
19:29 manfred {{ sls }} is the path.name of the state file
19:29 manfred stevedndm: https://github.com/saltstack/salt/pull/12832 this was just added to develop to help assist with figuring those out
19:29 manfred stevedndm: {{ source }} is the file it is currently acting on as well iirc
19:29 manfred {{ sls }} is the path.name of the state file
19:30 timoguin Ryan_Lane: good question. i haven't seen anything about the hangout, and no talking in here
19:30 whiteinge https://plus.google.com/hangouts/_/saltstack.com/docs-sprint?authuser=1&hceid=c2FsdHN0YWNrLmNvbV9icGRrbGJwcDg3anR1aWh1bGFvZ3Y1Yjk3OEBncm91cC5jYWxlbmRhci5nb29nbGUuY29t.9j83dk3htaiq0mfidj7isddrn0
19:30 timoguin i can't do much now anyway though
19:30 whiteinge sprint hangout ^^
19:31 Ryan_Lane hm. I can't join it?
19:31 agronholm is there a shortcut to get jinja2 to put a string in quotes  while escaping it properly?
19:31 agronholm it'd be nice for passwords that may themselves contain quotes
19:32 nkuttler agronholm: not sure what you mean, but can't you just \" ?
19:32 whiteinge Ryan_Lane: might need to be invited. pm me your email?
19:33 agronholm nkuttler: password: "{{ pillar['password'] }}"
19:33 perniferuse joined #salt
19:33 hipster joined #salt
19:34 hipster joined #salt
19:34 stevedndm manfred: thanks
19:34 nkuttler agronholm: and pillar['password'] contains quotes?
19:34 agronholm yeah
19:34 nkuttler agronholm: i'd assume jinja has some filter for that
19:34 agronholm so would I, that's why I'm asking
19:34 nkuttler agronholm: well, #pocoo is the official jinja support
19:35 nosleep77 joined #salt
19:35 rgbkrk joined #salt
19:35 Vye agronholm: escaping can be turned off by default or use safe(value). I haven't used the safe function but I turn it off by default in other apps (haven't in salt yet ).
19:35 agronholm Vye: what escaping are you referring to
19:35 supplicant {{ pillar["password"] | yaml }}
19:35 agronholm supplicant: gotcha
19:35 supplicant or maybe {{ pillar["password"] | yaml(True) }}
19:35 agronholm is that better than | json
19:36 agronholm what does the (True) do
19:36 agronholm I'm a total jinja2 newbie
19:36 supplicant inline format, so if "password" is an object it'll be a flat representation
19:36 supplicant this is also specific to salt, IIRC
19:36 agronholm I see
19:36 agronholm thanks
19:36 AndChat-482736 joined #salt
19:37 AndChat-482736 joined #salt
19:37 twobitsprite joined #salt
19:38 twobitsprite ugh... this whole zmq2 bug is killing me and I can't find any simple/workable solutions to getting pyzmq >13.0 on RHEL5... anyone have any advice?
19:38 agronholm twobitsprite: will it not compile if you pip install it?
19:39 twobitsprite agronholm: I can pip install it, but then I can't install the salt-minion RPMs because they depend on the python-zmq-2.1.9 package
19:39 agronholm ah
19:39 twobitsprite I guess I could pip install salt-minion too, but I really like to stick with RPMs wherever possible... just easier to manage, etc
19:39 agronholm what's the exact dependency that they declare?
19:40 XenophonF twobitsprite: what about rpmbuild-ing your own zmq rpm?
19:40 XenophonF not that it helps much
19:40 XenophonF i had to do that on an EC2 instance with mailgraph
19:40 XenophonF and now i have to do yum --skip-broken all the time
19:41 agronholm how should I go about installing wheels from the master to a minion virtualenv?
19:41 twobitsprite the exact depend is python26-zmq, so I guess it doesn't specify a version
19:42 twobitsprite I suppose I could try to rpmbuild pyzmq in to a package, but I can't find any existing spec files or anything, so that's a pain
19:42 fragamus joined #salt
19:43 twobitsprite I can't believe someone hasn't already solved this... there are a number of bugs on salt's github about the zmq2 problems, and everyone seems to be having the same problem with rhel5
19:43 twobitsprite and/or centos5
19:44 Ryan_Lane fixing something I broke, then I'll be in the sprint :)
19:44 ml_1 joined #salt
19:44 whiteinge ok
19:44 manfred everyone should just move to python27 from the ius repository... if ius ever compiles python27 for rhel 5
19:44 twobitsprite manfred: what does that solv?
19:44 manfred it gets you to python27-zmq something newer
19:45 manfred well... it doesn't have it in there...
19:45 manfred nevermind
19:45 twobitsprite yeah, I was going to say, I don't see any python27-zmq packages...
19:46 quickdry21 anyone else have problems with salt failing silently all the fucking time? even on info level debug
19:46 liquid_tool twobitsprite, can you link the zmq2 bug?
19:47 quickdry21 i've been trying to run highstate on a brand new minion. took some magic combination of restarting minions, masters, new keys and it finally worked. nothing in the logs
19:49 Networkn3rd joined #salt
19:49 twobitsprite liquid_tool: https://github.com/saltstack/salt/issues/6231
19:49 liquid_tool twobitsprite, thanks. Just want to check what the exact problem is to see if I havne't solved it accidently ;P
19:51 twobitsprite liquid_tool: lol :)
19:52 eriko joined #salt
19:55 jeblair joined #salt
19:55 joehoyle joined #salt
19:55 scott_w joined #salt
19:56 ghartz joined #salt
19:58 raizyr joined #salt
20:02 oz_akan_ joined #salt
20:04 bhosmer joined #salt
20:08 chow what would be the syntax to collect my minions ip using mines
20:08 chow ?
20:09 rgbkrk_ joined #salt
20:10 TyrfingMjolnir joined #salt
20:13 jrdx joined #salt
20:13 harobed_ joined #salt
20:13 twobitsprite ahh ha, found it! http://docs.saltstack.com/downloads/cent5/ :)
20:15 jaycedars joined #salt
20:16 cruatta joined #salt
20:16 cruatta_ joined #salt
20:16 smcquay joined #salt
20:16 it_dude joined #salt
20:17 cruatta_ joined #salt
20:17 cruatta joined #salt
20:18 smcquay joined #salt
20:20 rojem joined #salt
20:20 n8n joined #salt
20:21 mgw joined #salt
20:21 ggoZ joined #salt
20:24 KyleG joined #salt
20:24 KyleG joined #salt
20:24 liquid_tool twobitsprite, good job :)
20:27 sgviking joined #salt
20:27 dlindquist joined #salt
20:29 raizyr joined #salt
20:31 agronholm how would I go about installing wheels from the master's fileserver on a minion's virtualenv?
20:32 agronholm should I copy them over to minion's /tmp first or is there a better way?
20:37 ashw7n joined #salt
20:39 UtahDave XenophonF: sorry, I've been at a bunch of meetings.  How far did you get with your windows vm?
20:40 XenophonF UtahDave: it's built, am installing wix, git, python, etc. now
20:40 jnials joined #salt
20:40 UtahDave ok, cool.  I just wanted to make sure you didn't think I had abandoned you.  :)
20:41 Jarus Is it safe to set root's password with the state user.present?
20:41 XenophonF hah thanks!
20:41 XenophonF i'll work on this tonight, maybe have something for you to review tomorrow
20:41 UtahDave I'll be leaving for another meeting in 20 minutes.   I shouldn't be gone for more than 30 or 40 minutes.
20:41 UtahDave ok, that's great XenophonF
20:44 forrest UtahDave, getting fancy on windows
20:44 UtahDave :)
20:45 jcockhren "shouldn't be gone for more than 30 or 40 minutes"
20:45 jcockhren you know how many times in life I've said this and it not be true
20:45 jcockhren ?
20:45 jcockhren haha
20:45 jcockhren meetings--
20:46 jcockhren 4 hours later
20:46 UtahDave lol, jcockhren     so true
20:51 joehoyle joined #salt
20:54 jnials joined #salt
20:55 rgbkrk joined #salt
21:05 simonmcc joined #salt
21:05 frozenice joined #salt
21:05 hellome joined #salt
21:05 happytux joined #salt
21:09 stevednd manfred: is {{source}} only in develop currently, or is it in release versions?
21:10 manfred should be in released...
21:10 manfred the debug tool is only in develop
21:11 krow joined #salt
21:12 stevednd I'm running 2014.1.3, and I get jinja variable 'source' is undefined
21:14 jnials joined #salt
21:17 untamo13 joined #salt
21:19 untamo13 hi everybody! i had a working salt setup with one salt master and one salt minion. i recently upgraded to 2014.1.4 on both machines (and have verified this) and now i'm getting a "No Top file or external nodes data matches found" error when trying to run state.highstate. any ideas?
21:20 untamo13 i can run the test.ping and test.version commands successfully
21:20 untamo13 and when i use salt-call to run state.highstate on the minion i get the same error
21:22 stevednd does anyone know where in the source "include" is handled in statefiles?
21:25 agronholm untamo13: did your salt root directory change by any chance?
21:25 agronholm seems like it's not finding your files...
21:25 JasonSwindle joined #salt
21:26 untamo13 no, agronholm -- it's still at /srv/salt
21:26 gq45uaethdj26jw6 left #salt
21:26 agronholm and your salt master is running as root, right?
21:27 untamo13 no, agronholm -- i run it as a different user with limited permissions specifically for salt tasks
21:27 untamo13 but i've changed the user entry in the config to reference that user
21:28 agronholm has your config changed at all since before the upgrade?
21:29 untamo13 yes, a little, but i *think* not in any way that matters, i.e. the changes i made were moving my custom config into a separate file and removing a couple of lines from top.sls
21:29 untamo13 i have verified that the custom config file is getting pulled in
21:29 agronholm well, I'm out of ideas then
21:30 agronholm I'm a salt newbie
21:30 untamo13 me too :( thank you for your help though
21:31 ajolo_ joined #salt
21:33 roolo joined #salt
21:36 MindDrive joined #salt
21:37 krow joined #salt
21:38 jnials joined #salt
21:41 untamo13 for the record, it looks like 1 of 2 things happened: 1) when i extracted out my file_roots config into the new file, i didn't copy it over correctly; or 2) the upgrade changed how salt looks for the top.sls file so my existing file_roots config no longer worked. i'm guessing it was #1.
21:43 lionel__ joined #salt
21:43 fivethreeo joined #salt
21:45 ml_1 joined #salt
21:48 surge can i add a minion
21:48 surge to a master?
21:48 agronholm surge: not sure what you're asking
21:48 fivethreeo Anyone want to help out on a openstack formula? https://github.com/fivethreeo/openstack-automation
21:48 agronholm the minion connects to the master, you accept its key on the master
21:48 surge so lets say i have a master controlling 10 minions
21:48 surge but i want to make a minion on the same machine as the master
21:48 surge so it would be 11
21:48 agronholm surge: of course
21:48 surge sweet
21:48 surge thank u
21:48 agronholm I have just such a setup
21:49 jergerber joined #salt
21:50 V3h3m3nt joined #salt
21:51 jcockhren fivethreeo: I think I recall abe_music saying they use openstack. https://github.com/fivethreeo/openstack-automation
21:51 fivethreeo ah good  to know :)
21:52 SEJeff_work Question
21:52 bhosmer joined #salt
21:52 bhosmer_ joined #salt
21:52 SEJeff_work I know salt states run in the order they are defined in the state file. Is this true for the top file as well? If something matches at the very bottom of the top file, does that mean that it is included last and is executed last?
21:52 felskrone joined #salt
21:53 fivethreeo jcockhren: specifically https://github.com/fivethreeo/openstack-automation or just openstack at all?
21:54 jcockhren fivethreeo: I could be wrong. but at a meetup here I thought he said openstack
21:54 whiteinge SEJeff_work: that is true for the top file, yes
21:54 SEJeff_work whiteinge, you rock sir, thanks!
21:54 jcockhren not sure if they use that one specifically,
21:54 jcockhren that one = formula
21:54 whiteinge SEJeff_work: if you run state.show_lowstate the __order__ entry in that data structure is the order that Salt slurped things up
21:54 SEJeff_work We're including the salt.minion state and want it to execute at the very end so newly bootstrapped minions configure themselves last
21:55 SEJeff_work perfect
21:55 jaimed joined #salt
21:55 fivethreeo jcockhren: then just openstack my repo is 4 days old ;)
21:55 whiteinge SEJeff_work: you may prefer to use   ``- order: last`` to make that explicit
22:00 SEJeff_work whiteinge, You can do that in the top file?
22:00 SEJeff_work or in the specific state file?
22:00 perniferuse left #salt
22:01 whiteinge SEJeff_work: that goes in an individual state. sorry just re-read you want to do that at the include-level
22:01 fivethreeo SEJeff_work: specific state
22:02 ashw7n_ joined #salt
22:03 chuffpdx joined #salt
22:04 agronholm how can I display the current salt-minion options?
22:04 agronholm I'm not sure that it's respecting my tcp_keepalive_idle setting
22:04 jhulten joined #salt
22:04 Nazca joined #salt
22:04 Nazca joined #salt
22:06 aw110f_ joined #salt
22:06 kermit joined #salt
22:08 rojem joined #salt
22:15 joehh joined #salt
22:16 bhosmer joined #salt
22:16 bhosmer_ joined #salt
22:17 dlindquist left #salt
22:17 taion809 joined #salt
22:18 alunduil joined #salt
22:18 N-Mi joined #salt
22:19 N-Mi joined #salt
22:19 forrest Has anyone seen this postgres error when trying to use postgres_user.present? https://gist.github.com/gravyboat/b5e534a11ce0c651a710
22:27 Dattas joined #salt
22:29 dsolsona joined #salt
22:32 raizyr joined #salt
22:39 fivethreeo forrest: salt --version is?
22:39 forrest it's 2014.1, I figured it out
22:39 forrest postgres fun
22:39 forrest but it's not a salt problem
22:40 MTecknology forrest: fix it!
22:40 forrest I did, by not using postgres :P
22:40 MTecknology I meant my stuff...
22:40 forrest sorry but no
22:40 forrest :\
22:41 MTecknology oh :(
22:41 forrest You can hire me hourly, for the low low price of 999,999.99
22:41 forrest compared to most contractors, a steal
22:41 MTecknology who's stealing?
22:41 forrest what
22:42 MTecknology what's the steal? who's stealing?
22:42 forrest it's just a saying
22:42 * goodwill slaps forrest
22:42 fivethreeo all contractors?
22:42 goodwill or you can pay him in slaps
22:42 goodwill he likes them
22:42 forrest payment in slaps not accepted
22:42 MTecknology I like slapping and being slapped
22:42 forrest I said most contractors
22:42 forrest MTecknology, you've been working in IT too long
22:43 ajolo__ joined #salt
22:43 forrest glutton for punishment
22:43 MTecknology HAHA!!!!... that's uhm...
22:43 MTecknology I hear that a lot...
22:43 MTecknology I have salt using LDAP to authenticate users and that's working great, and also it's amazing that I can do that. I'd like to also use groups, which seems like it should be easy, but in practice, not so much. I have a report I need to build for a higher up from a pile of mismatched data, but... once that's done I'm going to try to tackle ldap groups again.
22:45 meteorfo_ joined #salt
22:45 SEJeff_work MTecknology, Are there docs on that up somewhere?
22:45 MTecknology I think part of the problem is our AD structure...
22:45 SEJeff_work I'd like to take a stab at it, but haven't looked yet
22:45 MTecknology SEJeff_work: http://docs.saltstack.com/en/latest/topics/eauth/index.html
22:45 fivethreeo IT is, do something good, everything works noone sees you, do something bad **ck you what did you do ;))
22:46 SEJeff_work sounds about right
22:46 MTecknology SEJeff_work: external auth is a bit under-documented, but not bad
22:46 SEJeff_work Nice!
22:46 SEJeff_work I'd love to see something that allowed a "thin" salt command on admin workstations
22:46 SEJeff_work that hit the salt api
22:47 SEJeff_work and ran real commands
22:47 SEJeff_work so you don't have to have everyone login to the salt master
22:47 SEJeff_work we have a team of ~15 people in 4 offices
22:47 MTecknology SEJeff_work: halite
22:47 SEJeff_work yeah a bit rough around the edges
22:47 SEJeff_work you use it?
22:47 forrest MTecknology, doesn't ldap handle users and groups differently? I thought that was the issue...
22:48 MTecknology I'm trying to combine halite and ldap auth and use AD groups so that people can log in and execute things like just state.sls against just a set of servers.
22:48 SEJeff_work forrest, depends on the schema, but in general yes
22:48 SEJeff_work MTecknology, Would you blog about that if you get it working?
22:48 MTecknology SEJeff_work: sure, I need the time too, mostly, I need to update docs when I figure it out
22:49 MTecknology forrest: it's different, but not that much if you talk to AD via LDAP calls... AD does some really fucky things like putting users into groups intead of groups onto users which can cause issues like incredibly long lookups, but sssd and python-sss takes care of that.
22:50 KyleG joined #salt
22:50 KyleG joined #salt
22:50 MTecknology The patches my company paid dave_den to make are making use of python-sss in salt and fixing a few other things in that user/group system (they made it into 2014.1.4)
22:51 MTecknology that was a massive pile of changes...
22:53 MTecknology forrest: I actually wrote a bunch of python and php sites that use AD authentication and treat AD almost just like and LDAP server. It works pretty well if you don't need to do a bunch of user lookups.
22:53 forrest nice
22:54 whiteinge MTecknology: have you seen the group support that has been added for the Helium release?
22:54 rgbkrk joined #salt
22:54 MTecknology whiteinge: nope, but I just read something about it
22:54 pdayton joined #salt
22:54 MTecknology This is in! #11324. It will be in the next feature release (codename Helium)
22:54 agronholm can salt do pip installs directly from the master's fileserver or do I have to copy the files to the minion first?
22:54 MTecknology Oh, and #11417 is the LDAP support.
22:54 aw110f joined #salt
22:55 MTecknology agronholm: my guess (without reading or testing), if you try to install from salt://foo, it'll copy foo to cache and install from there
22:56 agronholm MTecknology: I'll see what happens, thanks
22:56 MTecknology whiteinge: isn't this it?.. https://github.com/saltstack/salt/pull/11324
22:57 whiteinge hm, not sure. i thought we had ldap group support too
22:58 MTecknology I thought ldap group support was already in 2014.1.4.
22:58 MTecknology If it's not, that would explain why it's not working...
22:59 whiteinge it's in develop and not in the 2014.1 release
22:59 MTecknology hah... it looks like it's not there yet and that explains exactly why I can't make it work
22:59 MTecknology well... apple fritters.
23:00 MTecknology whiteinge: THANKS!!! I would have kept fighting and fighting to make something work that just didn't exist.
23:00 Networkn3rd joined #salt
23:00 whiteinge doh. glad to help
23:00 MTecknology so, helium is which release?
23:00 forrest MTecknology is another vote for showing the current docs as opposed to develop :P
23:00 MTecknology ^ YES!
23:01 whiteinge MTecknology: if you end up giving it a go in develop we love to hear if you run into trouble
23:01 whiteinge the Helium RC will be cut in a couple weeks most likely
23:01 MTecknology so, 2014.1.5?
23:02 Outlander joined #salt
23:02 whiteinge no it will be 2014.X where X is the month it gets cut
23:02 whiteinge (probably .6)
23:02 MTecknology 2014.6.0?
23:03 MTecknology heh.. interesting
23:03 whiteinge yeah
23:03 MTecknology https://github.com/saltstack/salt/issues/12915 <-- this is another issue I'm having that's causing me headaches.
23:04 MTecknology I might end up giving develop a shot, but it's now a nightmare for me to upgrade that many minions and hope nothing breaks. I don't have a dev or test environment. :(
23:05 MTecknology I have a protest environment... production + test + protest bad changes
23:05 SEJeff_work joined #salt
23:06 nineteeneightd Hah, "protest"...stealing that
23:08 cewood joined #salt
23:09 MTecknology whiteinge: I don't suppose you guys have some super secret repository for daring admins that happen to have a lot of minions that try to avoid building their own packages that has pre-release builds?...
23:10 whiteinge i can neither confirm nor deny. if you see black helicopters sporting the saltstack logo do not be afraid, stay near the window
23:10 MTecknology I don't have windows near my cube :(
23:10 * whiteinge calls of the hit
23:10 whiteinge *off
23:11 * goodwill slaps whiteinge
23:11 whiteinge MTecknology: we've discussed exactly that but we don't have it in place yet
23:12 whiteinge MTecknology: what platform?
23:12 MTecknology debian, centos, and windows; debian is where I would test
23:13 MTecknology I could just build it, it's not really that big of a pain.
23:13 krow joined #salt
23:13 jhulten joined #salt
23:13 pentabular joined #salt
23:13 whiteinge more so for debian than the others  :-P
23:14 ecdhe joined #salt
23:14 MTecknology hm?
23:15 whiteinge i know someone here could whip together an RPM or an .exe quick enough. i'm not sure about a .deb though
23:15 MTecknology I'll give you commands tomorrow for how I build packages for internal deployment. You might enjoy them.
23:15 rjc joined #salt
23:17 MTecknology gotta run
23:18 MTecknology whiteinge: I'll build from develop tomorrow, give you the commands I used, and update on how it went.
23:18 MTecknology back in 2-3hr
23:18 jslatts joined #salt
23:19 bhosmer_ joined #salt
23:20 bhosmer joined #salt
23:21 agronholm if I want to set up my app to run as a daemon on ubuntu, should I set it up as an upstart service?
23:21 agronholm or is there another way
23:22 agronholm it requires postgresql to be running
23:25 dlindquist joined #salt
23:26 dlindquist does anyone know if there's a way to specify a pillar file on the master when using salt-call on the command line?
23:26 bhosmer joined #salt
23:26 dlindquist wondering if there's a way to test without changing pillar/top.sls
23:28 XenophonF left #salt
23:33 manfred you need to specify it in the top.sls
23:33 manfred otherwise any minion could get ahold of data in your pillars, just by knowing the name
23:33 UtahDave dlindquist: Hm. I don't think so.  By definition, the salt master determines the pillar data and sends down the dict to the minion
23:34 pydanny joined #salt
23:37 JasonSwindle joined #salt
23:37 ashw7n joined #salt
23:39 orbit_darren joined #salt
23:43 manfred Ryan_Lane: https://github.com/saltstack/salt/pull/12924
23:44 whiteinge SEJeff_work: ooc have you seen pepper? https://github.com/saltstack/pepper
23:45 whiteinge it's still rough (though less rough than the current README makes it out)
23:45 whiteinge but it describes that salt-external interface you were talking about
23:46 ml_1 joined #salt
23:47 dlindquist manfred and UtahDave: Thanks!
23:47 fragamus joined #salt
23:50 schimmy joined #salt
23:52 UtahDave dlindquist: you're welcome!
23:52 Ryan_Lane joined #salt
23:55 untamo13 joined #salt
23:56 schimmy joined #salt

| Channels | #salt index | Today | | Search | Google Search | Plain-Text | summary