Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-06-25

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

All times shown according to UTC.

Time Nick Message
00:00 forrest MoonSweep, he lives in Australia, so he should be around,
00:00 UtahDave MoonSweep: Actually, opening a bug report on github will help us track it and make sure it gets fixed. We're actually looking at releasing a bugfix release very soon
00:00 forrest UtahDave, I still keep forgetting that we need to look into a build system :(
00:00 forrest should use that gentoo one
00:00 TaiSHi Exact same error, gonna run it with debug
00:00 forrest that builds all the things
00:01 TaiSHi cc: manfred
00:01 manfred forrest: we could make one that kicks off the builds using salt...
00:01 manfred and the reactor
00:01 MoonSweep UtahDave: the problem doesn't lie in salt itself, but in the debian repository
00:01 forrest manfred, the gentoo build system is pretty good
00:01 UtahDave TaiSHi: can you pastebin your sanitized providers file
00:01 manfred forrest: indeed
00:01 forrest it could easily tie into Pedro's stuff on jenkins I imagine
00:02 forrest or, just a manual process to cut the release.
00:02 manfred yeah
00:02 TaiSHi UtahDave: http://dpaste.com/2SXT606
00:02 manfred there is also another one... that I can't remember right now
00:02 whitepaws for anyone googling their way into this chat log later,
00:02 forrest manfred, that builds for multiple OS's?
00:02 manfred forrest: http://openbuildservice.org/
00:03 manfred that one
00:03 whitepaws the way to escape the colon for cmd.run arguments is not to escape the colon. it's to escape the whitespace following the colon.
00:03 forrest lol
00:03 forrest whitepaws, can you create an issue to add an example where that is done?
00:03 manfred whitepaws: that is ... unexpected
00:03 forrest and paste your example in said issue?
00:03 forrest whitepaws, that would be a lot more effective than googling the IRC log I think
00:03 MoonSweep I'll send a mail to the mailing list
00:03 UtahDave I wonder if having a space in your ssh_key_name is causing the issue, TaiSHi
00:04 UtahDave thanks, MoonSweep!
00:04 TaiSHi UtahDave: it's an underscore
00:04 forrest that joehh guy, what a slacker, it's like he's not paid to be here or something, jeez.
00:04 whitepaws forrest, sure
00:04 forrest whitepaws, thanks
00:04 UtahDave TaiSHi: Salt_Deploy    ?
00:04 forrest whitepaws, tag me on it and I'll try to do it if you don't want to, but at least create the issue with the example
00:04 TaiSHi Oh, sorry
00:04 TaiSHi Well, it did work perfectly before
00:05 TaiSHi I'm re-running it with -l debug
00:05 manfred forrest: and this is why i remember it http://openbuildservice.org/2012/09/10/arch-linux-support/
00:05 UtahDave TaiSHi: cool
00:05 forrest oh yea the open build service is the one I was thinking ovf
00:05 TaiSHi Sanitizing THAT output will be a pain
00:05 forrest *of
00:05 forrest manfred, sorry it's opensuse, not gentoo
00:05 UtahDave TaiSHi: does your key file have the proper permissions?
00:05 manfred forrest: ahh
00:05 garthk joined #salt
00:05 TaiSHi UtahDave: yep, as I said, it was working without the -- git develop
00:06 TaiSHi Well, same error but I see nothing weird
00:06 TaiSHi I'm killing the VM and starting a new one without script_args
00:07 TaiSHi Ahh, there we go
00:07 manfred TaiSHi: what if you just do script_args: git develop
00:07 manfred and no --
00:08 TaiSHi I think I found the issue
00:08 TaiSHi Might be a layer 8 one
00:09 manfred human?
00:09 manfred pebkac? what is layer 8?
00:09 TaiSHi All 3 you mentioned
00:09 ml_1 joined #salt
00:09 pfallenop joined #salt
00:09 forrest lol
00:09 TaiSHi http://en.wikipedia.org/wiki/Layer_8
00:10 MoonSweep UtahDave: should I CC joe healy ?
00:10 manfred TaiSHi: what did you do?
00:10 TaiSHi I pulled the ssh key files from git
00:10 MoonSweep left #salt
00:10 TaiSHi And it set them as 644
00:10 manfred nice!
00:10 MoonSweep joined #salt
00:10 UtahDave MoonSweep: yeah, good idea
00:10 TaiSHi Yeah, told you I was not amused by using keys on the master
00:11 MoonSweep Ok
00:11 manfred TaiSHi: ♥
00:11 UtahDave ok, time to head home.  Catch you all tomorrow.
00:11 manfred peace o/
00:11 TaiSHi See ya Dave
00:11 forrest I feel like you are leaving 'early'
00:11 forrest in a good way
00:11 forrest lol
00:11 TaiSHi Any way to pull from private repos w/o keys?
00:12 manfred TaiSHi: not in 2014.1 iirc
00:12 TaiSHi No no, I mean, manually
00:12 manfred oh, yes
00:12 TaiSHi Leaving MY key in the server is like, no gusta.
00:12 manfred TaiSHi: https:// links
00:12 manfred assuming your git repo supports them
00:12 TaiSHi (it's a key I generated for this project, but has access to my gh user)
00:13 forrest TaiSHi, how about generating a new deployment key that you specifically use for this purpose.
00:13 TaiSHi And how do I auth ?
00:13 manfred TaiSHi: http://stackoverflow.com/questions/13298198/git-clone-http-with-password
00:13 TaiSHi forrest: I did, but gh keys are 'global'
00:13 TaiSHi Hmm, same issue applies
00:14 TaiSHi I guess I could create an alternative gh user
00:14 TaiSHi Just to pull
00:14 patarr you guys think salt will overtake chef and puppet? :)
00:14 manfred TaiSHi: doesn't look like you can have private repos with like git:// that could only pull
00:15 TaiSHi manfred: I could create a "fake" github user, give it permissions to view the repo
00:15 manfred sure, that would work too
00:16 TaiSHi Great, I trust their main dev
00:16 TaiSHi But still, I'm not fond of providing access to my accounts
00:17 manfred yar
00:18 MoonSweep left #salt
00:18 stevednd Does anyone know if there are any plans to allow passing data(template variables, some kind of modifiable context) to statefiles called with include?
00:19 Rebus joined #salt
00:20 manfred what would be the usecase?
00:21 stevednd manfred:  statefiles that are templates for repetitive activities. First and foremost for me is deploying an application
00:21 manfred stevednd: so
00:21 manfred stevednd: i got this for you
00:21 manfred https://github.com/saltstack/salt/pull/13489
00:21 mmarch joined #salt
00:22 manfred stevednd: in that, i got the names thing that will be in helium, so you can keep the same state, but override certain things in a dictionary inside names:
00:22 MZAWeb joined #salt
00:22 manfred stevednd: and that pr also has a comment where I show how to template things directly with yaml
00:22 pdayton joined #salt
00:22 Ryan_Lane joined #salt
00:22 manfred oh we went over this
00:23 stevednd yeah, the repepition of the yaml
00:23 manfred you figured out how to do the yaml one that oneday iirc
00:23 stevednd this is sort of different
00:23 manfred stevednd: names will solve part of that, but there isn't a plan to do it so you can repeat stuff in different state files, yet
00:23 manfred stevednd: could you not just pull it down from pillars?
00:24 stevednd Well let me describe it quickly
00:25 manfred go ahead and send it, I am actually going to pack up my stuff and head home, but I will check what you say when I get home
00:26 mastrolinux joined #salt
00:27 TaiSHi yay dinner, bbiab
00:27 whitepaws forrest: here's the https://github.com/saltstack/salt/issues/13710
00:27 whitepaws it's my first github issue.
00:27 whitepaws i've heard you never forget your first time
00:33 bensons_ quoting is always a pita
00:33 stevednd manfred: I have app/myapp.sls which defines it's app key('myapp' in this case) variable, and loads up a config variable using import_yaml. Then I jinja include app/base.sls which does the majority of the deployment. After base.sls has finished running, app/myapp.sls continues with some custom state stuff specific to myapp itself. If I use a salt include statement, I have no way to tell the included sls that the app being deploye
00:33 stevednd d is 'myapp'. I had a discussion about a similar, yet different topic with whiteinge the other day. Basically some way to maintain a sort of writeable global context that is accessible by all state files might be the best way to make this work. Either that or the ability to set specific context or pillar information when calling include:
00:36 yomilk joined #salt
00:37 Ryan_Lane joined #salt
00:42 jslatts joined #salt
00:45 MatthewsFace joined #salt
00:45 ifmw joined #salt
00:46 jonbrefe joined #salt
00:46 manfred stevednd: makes sense
00:46 Shenril joined #salt
00:46 manfred stevednd: if some how salt state files used __context__ dunder, that would work i think
00:47 manfred http://docs.saltstack.com/en/latest/topics/development/dunder_dictionaries.html
00:51 MZAWeb joined #salt
00:55 joehoyle joined #salt
00:59 TaiSHi Anyone using sublime text to edit sls ?
00:59 ghartz joined #salt
01:01 mmarch I’m trying to setup one saltmaster to control multiple environments and I’m looking for examples of exactly how to setup (and then invoke) the whole thing..
01:01 mmarch this link is sorta helpful:  http://comments.gmane.org/gmane.comp.sysutils.salt.user/15586
01:01 mmarch but are there any other good examples out there?
01:01 Guest1501 joined #salt
01:03 Guest1501 joined #salt
01:12 Guest1501 joined #salt
01:12 Ryan_Lane is there any way to give a pillar the value of another pillar in the pillar sls?
01:13 Guest1501 joined #salt
01:13 ipalreadytaken joined #salt
01:14 Ryan_Lane jinja?
01:15 Guest1501 joined #salt
01:17 pdayton joined #salt
01:20 seventy3 joined #salt
01:20 Guest1501 joined #salt
01:26 Ryan_Lane1 joined #salt
01:27 mastrolinux joined #salt
01:27 renothing joined #salt
01:29 alikins joined #salt
01:31 Outlander joined #salt
01:33 kossy joined #salt
01:40 bmatt pillar sls files are templated automatically
01:40 bmatt so in pillars/foo_pillar.sls
01:40 bmatt you just put "bar: {{ salt['pillar.get']('quux') }}"
01:45 shaggy_surfer joined #salt
01:49 mgw joined #salt
01:54 brain5ide__ joined #salt
01:55 ajolo joined #salt
01:59 Ryan_Lane1 bmatt: excellent, thanks
02:00 manfred does that work?
02:00 bmatt works a treat
02:00 manfred i guess it might, but i was under the impression that pillars rendered on the minion
02:01 manfred they might just renders into a bundle to send to the minion for the jinja templating to take place...
02:01 manfred ¯\(°_o)/¯
02:02 manfred https://github.com/saltstack/salt/issues/4326
02:02 manfred oh i see what is going on
02:02 manfred never mind
02:07 druonysus joined #salt
02:07 druonysus joined #salt
02:10 mosen joined #salt
02:13 sashka_ua joined #salt
02:14 CeBe1 joined #salt
02:14 CeBe3 joined #salt
02:18 tkharju joined #salt
02:25 yomilk joined #salt
02:27 mastrolinux joined #salt
02:46 MZAWeb joined #salt
02:52 happytux_ joined #salt
02:55 acabrera joined #salt
03:03 ipalreadytaken joined #salt
03:09 mmarch joined #salt
03:10 schimmy joined #salt
03:12 schimmy1 joined #salt
03:19 jonbrefe joined #salt
03:19 TyrfingMjolnir joined #salt
03:20 thayne joined #salt
03:23 mastrolinux joined #salt
03:26 MZAWeb joined #salt
03:40 quanta_ joined #salt
03:40 mosen joined #salt
03:44 Guest7481 joined #salt
03:45 Guest7481 joined #salt
03:48 stevednd manfred: yeah, that is along the lines of what whiteinge mentioned. He thought there was an issue out there for it already, but wasn't able to find it.
03:48 manfred yeah
03:50 quanta_ a minion used to be run with a master, then I switch to run in standalone mode, and change the minion id
03:51 quanta_ looks like this cannot get the refreshed pillar
03:51 quanta_ the question is: how do I know which minion id is using in masterless mode?
03:52 Guest7481 joined #salt
03:53 bhosmer joined #salt
03:56 ksalman quanta_: what do you mean which minion id is using masterless mode?
03:56 ksalman the minion that you set up to be masterless
03:56 possibilities joined #salt
03:57 quanta_ ksalman: after swiching to run in masterless mode, I tried to run something like `salt-call -c /path/to/config/dir pillar.item blah` but it returned nothing
03:57 quanta_ I'm sure this key exist in the pillar
03:58 ksalman quanta_: pillar.items should give you a list of all the pillars
03:58 ksalman i think you want "pillar.get blah"
03:59 quanta_ the thing is in the pillar/top.sls I have something like this:
03:59 quanta_ base:
03:59 quanta_ 'integration*':
03:59 quanta_ - integration
03:59 quanta_ but the old hostname is not started with `integration`
03:59 quanta_ by changing this to:
03:59 quanta_ base:
04:01 quanta_ 'old_host_name*':
04:01 quanta_ - integration
04:01 quanta_ I can get the pillar
04:02 quanta_ ah, looks like `/etc/salt/minion_id` is the culprit
04:03 ksalman how did you change the minion id?
04:03 quanta_ ksalman: by changing the hostname
04:04 terminalmage quanta_: we cache the minion_id
04:04 terminalmage in /etc/salt/minion_id
04:04 terminalmage it's not necessarily desired to have the minion_id change when the hostname changes
04:04 anuvrat joined #salt
04:05 terminalmage also, if we change how we obtain the minion_id, and some edge case causes someone's minion_id to change, they tend to get angry
04:05 renothing joined #salt
04:05 ksalman can't make everyone happy =)
04:07 quanta_ terminalmage: so, my question is: how do I know which minion_id is used in the masterless mode?
04:08 ksalman quanta_: you can do a 'salt-call grains.get id'
04:09 quanta_ ksalman: ah, ok, thanks
04:10 ksalman it's nothing to do with it being masterless, you'll see the same behaviour if you had a master and a minion
04:14 fllr joined #salt
04:14 smcquay joined #salt
04:19 londo__ joined #salt
04:20 z3uS joined #salt
04:23 mastrolinux joined #salt
04:26 ramishra joined #salt
04:27 catpig joined #salt
04:32 TyrfingMjolnir joined #salt
04:32 Prateek_ joined #salt
04:42 aquinas joined #salt
04:54 fllr joined #salt
04:56 oz_akan_ joined #salt
05:03 TyrfingMjolnir_ joined #salt
05:05 ramishra joined #salt
05:07 malinoff joined #salt
05:09 ekristen joined #salt
05:10 MZAWeb joined #salt
05:15 renothing exit
05:21 ndrei joined #salt
05:23 mastrolinux joined #salt
05:25 ajolo joined #salt
05:32 cruatta joined #salt
05:37 ramishra joined #salt
05:38 MZAWeb joined #salt
05:55 fllr joined #salt
05:55 mmarch_ joined #salt
05:57 anuvrat joined #salt
05:57 fllr joined #salt
06:05 sygibson joined #salt
06:10 vu joined #salt
06:11 felskrone joined #salt
06:20 picker joined #salt
06:23 mastrolinux joined #salt
06:23 sygibson hey all - I was about to submit 2 feature requests in relation to targeting minions - but wanted to touch base with everyone here first ... to run the eficacy of the request by y'all
06:24 sygibson feature request 1:  use a file as input with a list of minion names for targeting
06:24 sygibson for example - the output of ... say .... 'salt-run manage.up > ./my_minions'
06:25 sygibson would be usable as 'salt -F ./my_minions <cmd>'
06:25 sygibson (presuming "-F" was "use File")
06:26 sygibson feature request 2:  the ability to use Nodegroups within compound matchers
06:27 sygibson I frequently want to do something like 'salt -C 'G@os:Ubuntu and N@compute_nodes' <cmd>'
06:27 fllr joined #salt
06:27 sygibson (again; assuming "N@" was the "nodegroup matcher
06:28 malinoff sygibson, the first one looks very similar to ansible's inventory file
06:28 malinoff I guess, in some situations that would be useful
06:28 sygibson I realize a nodegroup *is* a compound matcher, so logic would have to be imiplemented to do something lie "expand the nodegroup compound match; then run the other compound matchers specified"
06:28 yomilk joined #salt
06:29 fllr joined #salt
06:30 sygibson malinoff: yes - I often need to dynamically assemble a list of nodes I want to initiate "something" on/against, and dumping it to a file, is fast and easy, it'd be nice to be able to "ingest" that file of targets and execute against them
06:31 sygibson I can clearly do something with a shell/for loop, and read the node names in,and run a single salt command against them ... but that's so ... un-salty
06:31 _fllr_ joined #salt
06:31 malinoff sygibson, can't you use pillars and/or grains for that? Because salt assembles such list internally already
06:31 quanta_ joined #salt
06:32 sygibson I do often use grains in conjunctions with targeting - more frequently, I want to mix grains with other compound matchers - see feature request 2 - generally speaking, against a nodegroup definition
06:33 sygibson I could set custom grains or pillar data, but that's (in my mind) a "heavy" approach, when I want todo stuff in a more adhoc basis
06:33 jeddi joined #salt
06:34 sygibson I'm still very new to salt - so that's why I wanted to run it by the list here - and see if there was a better approach
06:34 sygibson I haven't played around with pillars much yet
06:49 ampex joined #salt
06:49 felskrone1 joined #salt
06:51 picker joined #salt
06:51 jpaetzel joined #salt
06:54 mastrolinux joined #salt
06:54 pfallenop joined #salt
06:55 krow joined #salt
07:01 schimmy joined #salt
07:02 ml_1 joined #salt
07:05 schimmy joined #salt
07:05 wigit any tips on how to create a bunch of files on the minion from a single template file? My first thought was to define all the variables in a loop in the state file and then use these variables in the template. That doesn't seem to be possible. Is there an easy way to see the name of the file that is currently being processed from within the template? I could move much of the logic there in that case.
07:13 oz_akan_ joined #salt
07:15 eofs is it just me or does "contents_pillar" really only use "base" pillar env?
07:17 yomilk joined #salt
07:18 alanpearce joined #salt
07:21 dkfjhs joined #salt
07:22 xmj joined #salt
07:22 dangra joined #salt
07:24 chiui joined #salt
07:25 schimmy joined #salt
07:26 johanek joined #salt
07:27 TyrfingMjolnir joined #salt
07:27 slav0nic joined #salt
07:27 johanek hello, i have a custom module that doesn't seem to be available on RH6.5 (works in Centos 6.5 and RH6.4)
07:27 johanek Jinja variable 'dict' object has no attribute 'helpers.re_replace'; line 11
07:27 johanek when i set logging to trace, it says it's loaded:
07:27 johanek [TRACE   ] Added helpers.re_replace to module
07:27 johanek any advice on how to debug this more?
07:27 felskrone joined #salt
07:28 felskrone joined #salt
07:29 fllr joined #salt
07:32 jdmf joined #salt
07:32 Kenzor_ joined #salt
07:38 elfixit joined #salt
07:43 n8n joined #salt
07:43 chiui joined #salt
07:47 rjc joined #salt
07:47 rjc joined #salt
07:49 Lomithrani joined #salt
07:51 it_dude joined #salt
07:52 krow joined #salt
07:53 ndrei joined #salt
07:55 TheThing joined #salt
07:59 fragamus joined #salt
08:02 Lomithrani Hi , I think it's not related to salt but it could be so I still hope your expertise will help me. I've got an issue on an apt-get update with a lock http://pastebin.com/vVzBg71K. I thought it might be because of modifying the sources so I set a require thinking that it would wait for the fil finished to be edited but it doesnt seem to work either    http://pastebin.com/gMMphwjU Sometimes a state.highstate will work fine , sometimes it does
08:03 Lomithrani could be to the pkg installed to
08:04 Lomithrani but again it has a watch , so it's supposed to be after isn't it
08:04 Lomithrani ?
08:05 darkelda joined #salt
08:05 darkelda joined #salt
08:05 malinoff Lomithrani, you don't have to write a custom cmd.run command to run apt-get update, just use refresh argument http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkg.html?highlight=pkg.installed#salt.states.pkg.installed
08:05 quanta_ left #salt
08:07 Lomithrani malinoff: thanks :) I'll do that then and perhaps you can help me with the issue I had yesterday
08:07 Lomithrani I'm trying to so something that would do something like that ::   {% test= (salt['test.ping']('App*', expr_form='pcre')|length) %}  any idea how I could ?
08:07 malinoff Lomithrani, what is the issue?
08:07 Lomithrani I want in a minion to count the other minions corresponding to "App*"
08:08 Lomithrani I could use mine , but I wondered if there were an faster way
08:08 Lomithrani (because I'm not sure the mine functions was enable on all the minions)
08:09 malinoff {% set test = salt['test.ping']('App*', expr_form='pcre') | length %} can you try this? Also, 'App*' does not look like a valid prce pattern, maybe you want to use glob?
08:10 Lomithrani can't believe I forgot the set though :D
08:10 malinoff :)
08:10 Lomithrani and yes might be glob
08:14 Lomithrani ok but pings take no argument so I need ot filter the result of ping
08:14 oz_akan_ joined #salt
08:15 Lomithrani oh ok and it executes only for the minion , so I guess I do have to use mine :(
08:16 malinoff Lomithrani, right, I forgot that it executed on the minion. You can use http://docs.saltstack.com/en/latest/ref/peer.html
08:20 davidone joined #salt
08:29 Lomithrani well thats is weird http://pastebin.com/Z6pJgb6B
08:29 fllr joined #salt
08:29 Lomithrani the minion conf has the function but the mine.get doesnt return anything
08:30 malinoff mine is something that I couldn't handle :(
08:30 malinoff I tried to use it, but I found that I really can't use it to gather dynamic info
08:30 malinoff peer communication is more suitable for it
08:30 malinoff though it is VERY slow
08:31 malinoff (actually, it was in v0.17, don't know is it slow right now)
08:31 picker joined #salt
08:33 Lomithrani mine works fine if you send the function through master cli
08:33 Lomithrani salt '*' mine.send network.interfaces
08:33 Lomithrani if I do this
08:33 Lomithrani then everything works fine , problem is I don't want to tell my boss to do this everytimes he had a minion
08:34 malinoff Can't help you there, sorry
08:34 Lomithrani I could use reactor though but I don't know if my salt version supports it
08:42 dgk joined #salt
08:43 xinkeT joined #salt
08:46 Rebus I am trying to find a salt module that handles openldap runtine configuration. Anyone can point me in the right direction?
08:51 jpaetzel_ joined #salt
08:54 kermit joined #salt
09:07 martoss joined #salt
09:09 zain_ joined #salt
09:15 oz_akan_ joined #salt
09:18 bhosmer joined #salt
09:19 giantlock joined #salt
09:20 pfallenop joined #salt
09:22 zain_ joined #salt
09:25 dkfjhs joined #salt
09:28 ggoZ joined #salt
09:29 fllr joined #salt
09:31 _fllr_ joined #salt
09:35 Hell_Fire hrm, guess module.keystone does some caching somewhere, doesn't like creating the tenant and a user role against that tenant in the same pass
09:36 bhosmer joined #salt
09:41 chamunks joined #salt
09:42 sdfsdfsd joined #salt
09:44 krow joined #salt
09:44 slav0nic joined #salt
09:45 seventy3 joined #salt
09:45 zain_ joined #salt
09:53 pdayton joined #salt
09:54 CeBe joined #salt
09:58 picker Hi Guys
09:58 picker https://github.com/saltstack/salt/issues/13670
09:59 picker Is temp fix could be applied to resolve above
09:59 picker Is there any temp fix that could be applied to resolve above issue?
10:04 malinoff picker, I guess, the most easiest way is to take the whole git.py state module (https://github.com/saltstack/salt/blob/v2014.1.5/salt/states/git.py), make a fix here (https://github.com/saltstack/salt/blob/v2014.1.5/salt/states/git.py#L234), put module with fix in /srv/salt/_states and run salt \* saltutil.sync_all
10:06 picker malinoff: thanks
10:12 beando joined #salt
10:14 robinsmidsrod joined #salt
10:16 oz_akan_ joined #salt
10:16 bhosmer joined #salt
10:19 krow joined #salt
10:20 TyrfingMjolnir joined #salt
10:29 fllr joined #salt
10:32 millz0r joined #salt
10:38 _mel_ joined #salt
10:38 ramishra joined #salt
10:45 Tekni joined #salt
10:47 bhosmer_ joined #salt
11:02 Outlander joined #salt
11:04 Sypher_ joined #salt
11:07 Sypher__ joined #salt
11:09 krow joined #salt
11:09 6JTAAI72K joined #salt
11:12 wb024 joined #salt
11:14 jalaziz joined #salt
11:16 dangra joined #salt
11:16 oz_akan_ joined #salt
11:18 oz_akan__ joined #salt
11:21 wigit btw - I see that some people use a "-%}" when ending blocks. Anyone care to tell why the dash is there? I tried googling but you get surprisingly few hits on that sequence of bytes
11:21 malinoff wigit, http://jinja.pocoo.org/docs/templates/#whitespace-control
11:22 krow1 joined #salt
11:22 wigit malinoff: thanks
11:22 malinoff wigit, people use that because sometimes extra empty lines break yaml parser
11:24 wigit yeah, that figures.
11:26 bhosmer joined #salt
11:26 jrdx joined #salt
11:29 fllr joined #salt
11:38 Lomithrani joined #salt
11:39 ramishra joined #salt
11:40 dfgfd joined #salt
11:41 dfgfd left #salt
11:46 diegows joined #salt
11:54 ramishra joined #salt
12:05 zz_keyvan joined #salt
12:05 keyvan joined #salt
12:06 ndrei joined #salt
12:09 derito joined #salt
12:09 chiui joined #salt
12:20 acabrera joined #salt
12:21 rawzone joined #salt
12:21 alanpearce joined #salt
12:29 to_json joined #salt
12:29 fllr joined #salt
12:31 agend joined #salt
12:34 blarghmatey joined #salt
12:37 jhauser joined #salt
12:39 brandon joined #salt
12:39 Outlander joined #salt
12:39 gildegoma joined #salt
12:40 ramishra joined #salt
12:41 laubosslink joined #salt
12:42 bhosmer joined #salt
12:45 blarghmatey joined #salt
12:49 jslatts joined #salt
12:50 Rebus I am trying to find a salt module that handles openldap runtime configuration. Anyone can point me in the right direction?
12:54 jslatts joined #salt
13:00 eykd joined #salt
13:01 Lomithrani Can I do that in a reactor  : {% set test= salt['mine.send']('network.ipaddrs')%}   ?
13:01 Lomithrani or just  salt['mine.send']('network.ipaddrs')
13:01 Lomithrani I just want to enable network.ipaddrs mine function on my minion when they start
13:05 ramishra joined #salt
13:05 TheThing joined #salt
13:07 quickdry21 joined #salt
13:07 oz_akan_ joined #salt
13:09 eliasp Rebus: the problem with OpenLDAP runtime config management: the old file-based configuration is considered deprecated and the DB-based onlineConfig is something, which is barely manageable through something like SaltStack… sucks big-time, but that's the way it is ;(
13:09 Rebus eliasp: I know that there are some difficulties with the runtime config
13:10 Rebus using the file based one was easy, e.g. with file.managed and service with a watcher
13:10 eliasp exactly
13:10 Rebus the puppet people have a module that lets you define entry states, e.g. entry.present etc.
13:10 Rebus and I was hoping salt has something similar
13:10 TyrfingMjolnir joined #salt
13:11 viq eliasp: deprecated, but still working AFAIK
13:11 eliasp viq: yes, but don't even dare to enter #openldap with a file-based config ;)
13:11 viq heh
13:11 Rebus that way one could define config entries in a state and the module would take care of diffs, adding, deleting etc.
13:12 eliasp Rebus: I don't know of any existing modules to do that, but you're right… based on this approach a basic management of onlineConfig should be possible
13:12 viq got link for the module?
13:12 Rebus the python-ldap library looks rather complete to me so it should be feasible ^^
13:13 eliasp Rebus: yes, python-ldap is rather nice to work with…
13:13 eliasp Rebus: also continuously and properly maintained
13:14 Rebus viq: https://github.com/datacentred/datacentred-ldap
13:14 the_lalelu i have written my own module(s) for openldap entries ... i manage dhcp and dns entries within ldap.
13:15 Rebus hmm maybe I can find some time next month to try my hands at a basic module to support CRUD for ldap entries
13:15 viq Rebus: interesting
13:18 eliasp if somebody would "port" this over to SaltStack, I'd be more than happy… looks really nice
13:18 Rebus that way one could start publishing formulas for common oenldap use cases, e.g. caching-proxy, slave-replicator, delta-syncrepl multimaster, etc.
13:18 eliasp +1
13:18 oz_akan_ joined #salt
13:24 racooper joined #salt
13:25 jalbretsen joined #salt
13:29 miqui joined #salt
13:29 mgw joined #salt
13:29 fllr joined #salt
13:32 UForgotten joined #salt
13:38 Lomithrani guys how can I declare an array (or something similar) to send through context on a file.managed
13:38 toastedpenguin joined #salt
13:38 Lomithrani {% set seeders= "Cassandra01" %} that works for 1 value
13:38 Eugene you want the "for" keyword.
13:39 Eugene See the jinja docs for the full usage
13:39 Lomithrani all I can find
13:39 Lomithrani is how to use the array
13:39 Lomithrani like for item in seq
13:40 Lomithrani but I want to know how the build the said seq
13:40 Eugene Typically you define the array in Pillar and then pass it into the statefile
13:40 Lomithrani I can't defin the array in the sls ? I don't use pillar so far :(
13:41 Eugene Pretty sure you can create a Python dict, yes
13:41 ml_1 {% for repo in [‘xxx','bbb','ccc'] %}
13:41 Eugene (array is not in the vocab here)
13:41 jaimed joined #salt
13:41 ml_1 {% endfor %}
13:42 Lomithrani so can I just  :  {% set seeders= ["Cassandra01","Cassandra03"] %}    ?
13:42 Eugene looks right
13:42 thayne joined #salt
13:43 Lomithrani ok I'll try that then
13:43 Eugene I'm not at a laptop, so this is all from memory
13:43 ml_1 @Lomithrani try that too: {% for repo in [‘xxx','bbb','ccc'] %}  … {% endfor %}
13:44 Lomithrani and then what name would have my variable ?
13:45 ml_1 Just say {% for seeders in [ ..
13:45 Eugene Or "for seeder in seeders...."
13:45 ml_1 {{ repo }} or {{ seeders }}
13:46 Lomithrani yeah but here you write the array  so that should work
13:46 Lomithrani sounds strange but I'll try too
13:46 Lomithrani thanks guys
13:47 timoguin he's trying to pass the whole array as context to a template
13:47 ml_1 {% for repo in ['a','b'] %}
13:47 ml_1 /etc/zypp/repos.d/{{ repo }}.repo:
13:47 ml_1 file.managed:
13:47 ml_1 - source: salt://files/common/zypper_repos.d/{{ repo }}.repo
13:47 ml_1 - mode: 644
13:47 ml_1 {% endfor %}
13:47 timoguin so the template will have the for loop in it
13:47 timoguin but the state just needs to set the variable
13:47 Lomithrani exactly timoguin
13:48 fllr joined #salt
13:48 Lomithrani timoguin: so wich syntax is the right one according to you ?
13:48 timoguin use set in the state
13:48 Lomithrani {% set seeders= ["Cassandra01","Cassandra03"] %}     is fine ?
13:48 timoguin yea
13:48 dude051 joined #salt
13:48 chiui joined #salt
13:48 Lomithrani good :
13:48 tristianc|Alt joined #salt
13:49 aquinas joined #salt
13:50 ml_1 just try it, my solution is working, maybe come back to that.
13:53 happytux joined #salt
13:55 byronschaller joined #salt
13:58 byronschaller Having issues with docker.create_container... error saying image does could not be resolved to an existing container. I am using the image id: ad892dd21d60
13:59 Bandikoto left #salt
14:00 CeBe1 joined #salt
14:01 tyler-baker joined #salt
14:04 eykd Anyone know the correct URL to use for the public registry in my docker-registries authentication pillar? The docs seem to suggest https://index.docker.io/v1 but I’m still getting authentication errors. :\
14:06 mastrolinux Hi guys, I have this serious error in salt-master http://pastebin.com/sinG0mu8
14:06 mastrolinux and the app is not listening on port 4505 anymore
14:09 mastrolinux even after the update to salt-master_2014.1.5-1 this is still happening
14:11 happytux joined #salt
14:15 ramishra joined #salt
14:15 baj joined #salt
14:15 kaptk2 joined #salt
14:15 baj Hello world !
14:16 baj i'm new to salt, and may be a noob :)
14:16 baj i'm trying to use proxy minion
14:16 baj for exercice, i try it with VMWare API
14:17 baj and now, I want to launch some commands
14:17 baj but nothing in documentation
14:17 baj :/
14:17 baj any idea ?
14:18 racooper baj,  did you go through the salt tutorial?
14:18 baj this http://salt.readthedocs.org/en/v2014.1.1/topics/proxyminion/index.html ? yes
14:19 baj I can see my proxy on the minion with salt myminion pillar.items
14:19 baj i got a /srv/salt/_modules/proxy/vmware.py
14:20 baj implementing Proxyconn
14:20 racooper see, you didn't say anything about proxy when you started asking.
14:20 baj in /srv/prillar/top.sls & /srv/prillar/vmwareconfig.sls
14:20 racooper oh wait, you did
14:20 baj ;)
14:21 racooper nvm I'm not reading well apparently this morning. need caffeine....
14:21 baj i'm french, it's 4 PM, need too :p
14:21 rallytime joined #salt
14:21 babilen It's okay to be French ;)
14:22 laubosslink joined #salt
14:22 lionel joined #salt
14:22 yano joined #salt
14:22 baj so, any idea with my noob-problem ?
14:23 babilen I've never used proxy minions, so unfortunately: No
14:25 perfectsine joined #salt
14:25 ganitsu joined #salt
14:26 ganitsu joined #salt
14:29 vejdmn joined #salt
14:29 bhosmer_ joined #salt
14:31 nahamu proxy minion?
14:31 nahamu is that the syndic ?
14:36 baj no
14:36 baj it's 2 tools
14:39 fllr joined #salt
14:40 jas-_ joined #salt
14:44 kermit joined #salt
14:45 CheKoLyN joined #salt
14:46 Lomithrani how can I do that ? http://pastebin.com/pbiWCLa6
14:46 Lomithrani I can print in a for loop but I can't seem to do any function
14:47 timoguin baj: yea docs are quite sparse for the proxy minion. good luck?
14:48 baj need the force
14:50 timoguin fyi, salt-cloud supports libvirt and vsphere if you didn't know
14:51 baj just for testing because i now how to use vmware api
14:51 dude^2 joined #salt
14:51 baj afeter, lot of apps without salt integration will be addressed with the proxy
14:51 * timoguin nods
14:52 baj so, i think i will use a minion as a "proxy" and use state with rest implmentation
14:52 baj "ça fait chier" :)
14:53 timoguin Lomithrani: possibly the do statement: http://stackoverflow.com/questions/4870346/can-a-jinja-variables-scope-extend-beyond-in-an-inner-block/4880398#4880398
14:53 timoguin but that may not be available from salt
14:53 alanpearce joined #salt
14:54 timoguin looks like that extension has to be explicitly enabled, but i'm not sure how that would be done inside of salt, or if it's currently possible
14:55 Lomithrani looks like a good way around thanks
14:55 timoguin if it works!
14:58 dvogt joined #salt
14:59 blarghmatey joined #salt
14:59 tristianc|Alt joined #salt
15:01 jonbrefe joined #salt
15:03 dude051 joined #salt
15:03 gothix joined #salt
15:03 gothix joined #salt
15:04 andabata joined #salt
15:06 mastrolinux I am going to change a salt-master host, where are stored minion keys on the master?
15:06 mastrolinux i copied over the whole /etc/salt dir
15:06 mastrolinux but salt-key -L does not find keys
15:08 baj in /etc/salt/pki/master/
15:08 seventy3 joined #salt
15:09 baj and minion /etc/salt/pki/minion/
15:09 bastion1704 joined #salt
15:10 baj check if on the master the content of cat /etc/salt/pki/master/master.pub
15:10 logix812 joined #salt
15:10 baj and on minion /etc/salt/pki/minion/minion_master.pub are same
15:10 wendall911 joined #salt
15:12 mgw joined #salt
15:14 ecdhe Does anyone salt their firefox config?
15:14 ecdhe For instance, installing noscript, bookmark backups, and setting policy?
15:14 racooper I know there is a command module for quota, but is there a way to set user quotas as a state?
15:17 mastrolinux baj: thanks a lot, the problem was that my files disappeared in some way, I needed to recover from backups
15:20 baj nw ;)
15:20 Lomithrani joined #salt
15:21 dude051 joined #salt
15:27 Lomithrani can I target all the minions in a reactor ?
15:28 Lomithrani like for example if I want to trigger a state.highstate for all the minions when I add a new minion
15:28 timoguin sure
15:28 timoguin tgt: *
15:28 Lomithrani sweet :)
15:28 Lomithrani the more I know about salt the more I love it
15:31 ndrei joined #salt
15:33 mgw joined #salt
15:36 ndrei joined #salt
15:38 n8n joined #salt
15:40 thayne joined #salt
15:45 davet joined #salt
15:45 rojem joined #salt
15:46 davet1 joined #salt
15:47 tinuva joined #salt
15:48 ipmb joined #salt
15:48 fneves joined #salt
15:49 fneves Hi all
15:49 fneves I'm currently trying to understand how environment work for pillars
15:50 fneves lets say I want to have global, dev, qa and production
15:50 fneves is it possible to have pillars with the same ids on different environments
15:51 fneves that will get picked up or fallback into base if they don't exist?
15:51 ipmb joined #salt
15:53 perfectsine joined #salt
15:54 BrendanGilmore joined #salt
16:01 chiui joined #salt
16:03 tligda joined #salt
16:08 ndrei joined #salt
16:09 GnuLxUsr joined #salt
16:10 racooper joined #salt
16:11 eykd joined #salt
16:12 KyleG joined #salt
16:12 KyleG joined #salt
16:17 UtahDave joined #salt
16:17 bemehow_ joined #salt
16:19 tinuva joined #salt
16:21 ckao joined #salt
16:21 troyready joined #salt
16:30 analogbyte joined #salt
16:32 joehillen joined #salt
16:34 perfectsine joined #salt
16:41 alanpearce joined #salt
16:47 analogbyte joined #salt
16:50 forrest joined #salt
16:50 ajolo joined #salt
16:53 alanpearce joined #salt
16:56 Guest7481 joined #salt
16:59 shaggy_surfer joined #salt
17:01 ml_1 joined #salt
17:02 byronschaller_ joined #salt
17:03 alanpearce joined #salt
17:03 Rojematic joined #salt
17:05 alanpearce joined #salt
17:06 forrest manfred, any luck on the sprint?
17:06 chrisjones joined #salt
17:07 manfred forrest: i got the first part of permission to do it, now i just gotta figure out who can help
17:07 forrest manfred, what do you mean?
17:07 forrest it should be easy to run, just need a conference room
17:07 manfred right
17:07 manfred i gotta get that part reserved
17:08 manfred and i need someone to do that
17:08 Ryan_Lane joined #salt
17:08 forrest you can't just reserve a room on a calendar?
17:08 manfred i do not have any clue how
17:08 forrest did you ask Jason?
17:08 forrest he probably knows
17:09 manfred i will hit him up, i have been slammed with this cluster that broke on saturday all morning cause we just figured out why it failed
17:09 dude051 joined #salt
17:10 lude is there a way for salt to set a flag in the state tree based on a file existing on the minion?
17:10 lude eg., i have a file /root/something
17:10 lude and want to be able to {% if that_file_exists %} in my state tree
17:16 timoguin {% if salt['file.file_exists']('/root/something') %}
17:16 forrest manfred, you shouldn't pee on the servers
17:17 forrest that would probably resolve the issue.
17:17 manfred heh
17:19 notpeter_ joined #salt
17:23 ksalman how would use this in an state? http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.win_useradd.html#module-salt.modules.win_useradd
17:23 marco_en_voyage joined #salt
17:23 ksalman i want to add a user
17:24 Ryan_Lane UtahDave: I'm getting errors when trying to use listen/listen_in
17:24 timoguin ksalman: that module will load on windows as just "user"
17:24 timoguin so it looks just like adding a user on other platforms
17:24 UtahDave Ryan_Lane: What kind of errors?
17:24 googolhash joined #salt
17:24 Ryan_Lane stacktrace
17:24 Ryan_Lane UtahDave: https://gist.github.com/ryan-lane/88fec4264d134e25e5ea
17:25 ksalman timoguin: I don't understand, can i not add a user in a state then?
17:25 UtahDave Ryan_Lane: I'll pass that on. thanks!
17:25 Ryan_Lane yw
17:25 Ryan_Lane should I open an issue?
17:25 aw110f joined #salt
17:26 UtahDave yeah, that would be great, Ryan_Lane. thanks!
17:26 davet joined #salt
17:27 shaggy_surfer joined #salt
17:27 n8n joined #salt
17:27 ksalman i think i figured it out
17:27 timoguin ksalman: yes, you can. you'd use the user state. http://docs.saltstack.com/en/latest/ref/states/all/salt.states.user.html
17:27 eliasp any ideas, why my docker.installed state only returns 'image "pctue/mysqld" does not exist' while it is listed just as expected in 'docker images'? the SLS: http://pastebin.kde.org/pvjatk8py
17:28 mastrolinux1 joined #salt
17:28 ksalman timoguin: thanks
17:28 viq hm
17:28 viq salt minion cmd.run "at +1 minute <<< 'logger this is a test'"
17:29 viq Syntax error: redirection unexpected
17:29 viq ideas?
17:29 xmj viq: what shell is that?
17:29 xmj looks a bit like csh.
17:29 Guest7481 joined #salt
17:29 viq freebsd's sh
17:30 viq erm
17:30 viq I think sh...
17:30 viq I'm not even sure how to check
17:30 xmj cmd.run "echo $SHELL"
17:30 eliasp viq: try cmd.run "echo $SHELL"
17:30 eliasp :)
17:30 oeuftete Is there a module for Julia in the works anywhere?  At my work, we're vaguely starting to think it might be worth writing one.
17:31 viq no output
17:31 krow joined #salt
17:31 Guest7481 joined #salt
17:31 xmj blargh
17:31 dvogt joined #salt
17:32 viq adding shell=/bin/sh didn't help
17:33 logix812 joined #salt
17:33 xmj incomplete time
17:33 * viq attempts to do http://www.rackeroz.com/2013/10/how-to-restart-salt-minion.html
17:33 eliasp viq: are you sure about the "triple" redirection? I thought that's only supported by zsh + bash > 3.0
17:33 lude timoguin: much delayed thank you
17:34 xmj viq: what is it you're trying to achieve?
17:34 viq eliasp: good point, indeed
17:34 xmj you could just sleep 60 && service salt-minion restart?
17:34 viq xmj: I could just as well just service.restart salt_minion, but that makes highstate etc not return anything
17:35 xmj echo foo | at + 1 minute works, too
17:35 xmj ah.
17:35 viq ah, let me try that
17:36 viq yeah, that seems to work, thank you
17:36 xmj de nada
17:37 viq :)
17:43 UtahDave Ryan_Lane: Tom's posted a fix. He was wondering if he can see the relevant portion of your sls
17:44 viq UtahDave: are there /will there be somewhere recordings of your talks about OpenStack deployments?
17:46 vejdmn joined #salt
17:47 UtahDave viq: Those talks weren't recorded
17:47 UtahDave but maybe I could do a Salt Air for them...
17:47 viq UtahDave: thanks. shame
17:47 bemehow joined #salt
17:48 viq That would be cool :)
17:50 ndrei joined #salt
17:53 LordMike91 joined #salt
17:55 schimmy joined #salt
17:58 druonysus joined #salt
17:58 druonysus joined #salt
17:58 schimmy1 joined #salt
17:58 googolhash joined #salt
18:00 kballou joined #salt
18:02 unixman9000 joined #salt
18:02 bhosmer joined #salt
18:03 ramteid joined #salt
18:03 ipalreadytaken joined #salt
18:04 _jslatts joined #salt
18:05 rojem joined #salt
18:06 moos3 I have a question when you have a state/init.sls then in there sub sls files, how can I use the init.sls to include those
18:06 meteorfox joined #salt
18:07 bemehow_ joined #salt
18:07 viq moos3: - state.substate
18:07 moos3 inside the init.sls ?
18:07 nosleep77 joined #salt
18:07 viq let me find you an example
18:07 moos3 thanks
18:07 dvogt joined #salt
18:07 djaime joined #salt
18:08 viq moos3: https://github.com/saltstack-formulas/nginx-formula/blob/master/nginx/init.sls
18:08 moos3 oh nice
18:09 timoguin also, .substate works as a relative path import
18:09 moos3 i'm still trying to wrap my head around pillars
18:10 viq moos3: "secure place to store variables to pass to states"
18:10 moos3 yeah i'm trying to figure out how to handle say different environments, I have a dev, test and production environments with all different constants
18:11 fragamus joined #salt
18:11 moos3 I get with base: '*':  -- but if I have 'dev': '*':  - .... how do I get to the pillars for dev versus base
18:11 fragamus joined #salt
18:17 dvogt joined #salt
18:19 marco_en_voyage joined #salt
18:23 Guest27990 joined #salt
18:24 mgw joined #salt
18:28 mapu joined #salt
18:28 dvogt joined #salt
18:29 viq well, in that case you can't have '*'
18:29 viq you have to assign machines more explicitly to environments
18:31 zartoosh joined #salt
18:31 linjan joined #salt
18:32 Guest7481 joined #salt
18:38 dstokes is `salt-api` the proper command for starting netapi services on a salt master?
18:39 dstokes also, hai :)
18:41 ramishra_ joined #salt
18:44 bhosmer joined #salt
18:45 maboum joined #salt
18:49 bhosmer joined #salt
18:55 Guest7481 joined #salt
18:56 dude051 joined #salt
18:56 tkharju joined #salt
18:57 claytron joined #salt
18:58 bhosmer_ joined #salt
18:58 claytron I'm confused about version numbers based on this post: http://www.saltstack.com/salt-blog/2013/10/27/salt-version-numbers and the actual versions: http://docs.saltstack.com/en/latest/topics/releases/index.html
18:59 claytron is it YYYY.Codename.bugfix ?
19:00 kivihtin joined #salt
19:00 manfred no
19:00 manfred YYYY.M.bugfix
19:00 manfred Hydrogen was the first one
19:01 manfred and it was only hydrogen because we didn't know the month it was going to be released
19:02 claytron I don't quite get the M there
19:02 claytron what would helium be given a specific month
19:03 manfred M is month
19:03 manfred hydrogen was released in january of 2014, so 2014.1
19:03 Eugene Ubuntu style numbering
19:03 rojem joined #salt
19:03 manfred helium will be given a month when the rc is tagged
19:03 manfred and the release is branched
19:04 manfred then we will rc, until it is stable and then we will have the 2014.x.0
19:04 Eugene rah rah rah shoulda just reved to 1.0 and stuck with Semantic
19:04 claytron ok, so the M is just when the next release is cut
19:04 manfred Eugene: better to know whne it was released
19:04 manfred claytron: probably will be 2014.6, but it might be 2014.7
19:05 claytron and the codename is just that, has nothing to do with a number
19:05 Eugene Better to know what was released
19:05 Eugene but hey, not my project to run, so what would I know
19:06 claytron I think that is what was tripping me up since hydrogen is element 1
19:07 ingwaem joined #salt
19:08 jrdx joined #salt
19:09 ingwaem Greetings Ya'll
19:09 bhosmer joined #salt
19:09 Eugene howdy
19:10 linjan joined #salt
19:10 krow joined #salt
19:10 ingwaem Question out to the SaltStack guys (Dave..you there? :) ) I see that Salt-Api has moved into the main salt repo...is there any documentation on this change?
19:11 ingwaem ultimately it's been great to see all the updates, I just have lost a little track of all the changes :)
19:12 beando joined #salt
19:12 anotherZero joined #salt
19:12 Ahlee ingwaem: https://github.com/saltstack/salt/pull/13554
19:13 ingwaem oh cool thanks Ahlee
19:13 whiteinge joined #salt
19:14 ingwaem is salt at Velocity this year?
19:15 forrest ingwaem, http://www.saltstack.com/salt-blog/2014/6/24/saltstack-customers-and-engineers-demonstrate-automation-for-all-the-data-center-things
19:15 forrest it was an ignite talk last night
19:16 forrest I don't know if Mike is still around, but he's not in the IRC so maybe
19:16 ksalman i am managing minion.conf via salt, and i restart salt-minion service if minion.conf changes, via watch. On Windows, somtimes the salt-minion stops instead of restarting, is there no way around it?
19:16 ingwaem whoa cool thanks :) I didn't go to velocity this year but have a bunch of colleagues there that may be able to stop by. I'm sure I'll be there next year though
19:16 forrest ingwaem, damn I didn't realize how expensive velocity was, holy crap
19:17 forrest 1800 bucks is insane
19:17 ingwaem forrest, yea it's nuts
19:17 Ahlee [x] Expense it!
19:17 ingwaem one of the reasons I didn't go cause we have like 10 guys there
19:17 ingwaem bunch from sweden and a bunch from the states, almost like a small country going to visit them :)
19:17 forrest Ahlee, I don't think they'd appreciate me expensing that much for a single conference
19:17 forrest could go to like 4 other conferences
19:18 forrest ingwaem, hah
19:18 ingwaem out of all the conferences I've been to though forrest, Velocity I believe is the most rewarding...both in terms of making contacts (and sales leads) and learning about new tech
19:18 forrest ingwaem, oh really?
19:18 ingwaem it's probably worth going to velocity as opposed to those other 4 instead
19:19 forrest ingwaem, that's good to know, usually I'm not very impressed with what o'reilly offers.
19:20 ingwaem the other conferences are bleh...GDC, while good and huge last year was mainly turned into a job fair...we were inundated by job seekers, and not enough leads...wasn't worth it...but velocity, yea that's the best tech people all in one place...and it's big
19:20 forrest yea santa clara is a pretty large conference area
19:20 ingwaem and it's way bigger than the cloud expos too
19:20 forrest cool, I'll let our guys know
19:21 ingwaem had we not had just a large contingent going I would have definitely been there...but our company is cool...tries to get everyone to a conference at least once, and repeats occur quite often too :)
19:22 forrest ksalman, you could try using the at command
19:22 forrest though I don't know if that works on windows...
19:23 ksalman hmm..
19:23 forrest basically you do at + 1 to restart the minion
19:23 forrest so it does't crap itself during the run
19:23 ksalman okay, i'll see whats available on windows haha
19:23 kivihtin joined #salt
19:23 ksalman wish i didn't have to support windows
19:24 smcquay joined #salt
19:24 forrest ksalman, you could always do a cmd.run with whatever window's equivalent is
19:24 forrest ksalman, you should tell your team no more windows
19:25 ingwaem Windows should burn in the fires of corrupted bytes
19:25 timoguin but then how will they see outside!
19:25 timoguin ksalman: windows has at. :)
19:25 ksalman huh
19:25 ksalman thats news to me lol
19:25 bmatt ReactOS 4 life!
19:26 ksalman wow
19:26 eliasp regarding the Windows Minion restart issue: https://github.com/saltstack/salt/issues/11726
19:26 herlo joined #salt
19:26 timoguin ksalman: http://technet.microsoft.com/en-us/library/cc755618%28v=ws.10%29.aspx
19:26 eliasp there's also a workaround: https://github.com/saltstack/salt/issues/11726#issuecomment-46245881
19:26 ksalman timoguin: thanks
19:28 tkharju1 joined #salt
19:29 tkharju2 joined #salt
19:30 chamunks joined #salt
19:31 racooper trying to get working a modified (cut down) version of the users-formula.  It's working, mostly but I'm getting an error on a group require. https://gist.github.com/racooper/b44e9d1168c1ed52d3a1
19:31 racooper the groups are present, and I'm also not sure why there's a require for them in the first place.?
19:33 tkharju joined #salt
19:33 n8n joined #salt
19:34 aw110f I'm trying to build salt rpm for centos5, I grabbed the salt.spec file from the EPEL source repo https://dl.fedoraproject.org/pub/epel/testing/5/SRPMS/salt-2014.1.5-1.el5.src.rpm, but when I try to build it, it complains:
19:34 aw110f error: File /usr/src/redhat/SOURCES/salt-2014.1.5.tar.gz: No such file or directory
19:35 aw110f I was able to rebuild salt RPM for centos6 using the provided spec file
19:36 herlo aw110f: are you trying to build it as root?
19:36 aw110f no building with a user
19:36 herlo what distro are you using?
19:37 herlo if fedora, install fedora-packager
19:37 aw110f centos 5.8
19:37 herlo use fedora and mock
19:37 herlo mock will give you tooling for centos 5
19:37 racooper you're about two versions of CentOS 5 behind....
19:37 herlo racooper: lol
19:37 herlo aw110f: you don't have a few things defined.
19:38 smcquay UtahDave, whiteinge good morning. I am back, and have verified that the /etc/salt/minion.d/auth_flood_fix.conf has been pushed out to all the hosts, and I am still seeing odd behavior. This behavior is specifically that I can do test.ping on a subset of my machines, but if I do it twice consecutively or on "too large" a subset, then it hangs indefinitely. When I kill the salt command and run the salt-run it
19:38 smcquay suggests it comes back quickly, but missing vms. There is no consistency for what is going on with these missing vms. Sometimes it is msgpack errors (I clean out cache, seems to fix those), other times it seems the minions are happy when queried again.
19:38 * smcquay is ashamed (slightly) for his novel)
19:38 racooper that's like 2 years and who knows how many security fixes and bugfixes.
19:38 n8n joined #salt
19:39 smcquay also, hi herlo !
19:39 herlo smcquay: hi!
19:39 herlo racooper: unfortunately, it may be what he has to work with...
19:39 herlo racooper: I do agree that he should be on the latest 5.x though at a minimum.
19:39 rojem joined #salt
19:39 herlo aw110f: at one time, we had salt packages for epel5
19:39 racooper precisely. I wasn't recommending going to 7 :)
19:40 herlo aw110f: in fact, http://koji.fedoraproject.org/koji/buildinfo?buildID=537902 says they still exist
19:41 n8n_ joined #salt
19:41 herlo aw110f: I recommend installing the epel5 rpm and then installing salt. It 'should just work (tm)'
19:41 whiteinge smcquay: where are you seeing messagepack errors?
19:41 herlo aw110f: unless you had something you thought needed to be fixed?
19:42 smcquay whiteinge: on the minions.
19:42 smcquay I looked up the error yesterday on google and it was a version mismatch issue.
19:42 smcquay so I updated salt on master (it was old and installed via pip, now we're running with debian's 2014.1.5) and same on the hosts.
19:42 smcquay sorry, minion hosts.
19:43 smcquay still I find it odd that eventlisten.py shows nothing even after connecting to the
19:43 smcquay ipc:///var/run/salt/master/master_event_pub.ipc
19:45 whiteinge smcquay: are you seeing msgpack errors after the upgrade?
19:45 Hell_Fire joined #salt
19:45 aw110f herlo: https://github.com/saltstack/salt/pull/13417/files#diff-0
19:46 aw110f This is what i want fixed ^
19:49 n8n joined #salt
19:49 eliasp UtahDave: any chance of having {states,modules}/dockerio.py backported from develop to 2014.1? latest Docker release doesn't work anymore with 2014.1.5
19:49 aw110f racooper: I have a few straggler machines still not on centos 6 yet unfortunatetly
19:50 racooper well, should at least get your CentOS 5 boxes up to 5.10 :)
19:50 UtahDave eliasp: Hm. good point.   lemme see
19:51 ramishra joined #salt
19:51 eliasp UtahDave: to be more specific: docker.inspect_image doesn't find anything…
19:51 aw110f herlo: which fedora and mock version should I use to build a el5 RPM?
19:51 herlo aw110f: so you want to rebuild the centos5 rpm, okay
19:51 herlo aw110f: the latest should work.
19:52 n8n joined #salt
19:52 Karunamon joined #salt
19:52 herlo aw110f: you just need the el5 srpm and you can install it there, make your modifications, create a new srpm and rebuild it in mock.
19:52 pdayton joined #salt
19:52 herlo if that works, you can submit a patch to salt
19:52 smcquay_ joined #salt
19:53 herlo aw110f: http://fedoraproject.org/wiki/Projects/Mock#Setup
19:53 Karunamon Hi folks - newbie here trying to get their head around how the namespace works. I'm trying to create one sls file per package I install on my minions as detailed here: https://gist.github.com/Karunamon/aa4d2c7c96cbb18b11c5
19:53 aw110f Thanks herlo, I'll give it a shot
19:54 herlo aw110f: let me know if you have further questions
19:54 herlo aw110f: I was the original packager of salt, so I should be able to answer most anything.
19:55 UtahDave eliasp: Have you tested the latest develop version of the docker states and modules?  If they work, we can totally just backport the whole thing
19:55 UtahDave herlo++
19:55 eliasp UtahDave: yes, I deploying now the latest develop module/state here locally to get everything working… wait 5 minutes, need to confirm some stuff, but I think they're fine
19:55 timoguin Karunamon: I'd recommend doing the walkthrough
19:55 UtahDave Karunamon: Yeah, that's not going to work how you've got it set up.
19:56 nhubbard UtahDave: we have been running the latest modules for docker and have not seen any issues
19:56 UtahDave eliasp: cool, let me know
19:56 Karunamon timoguin: I'm actually basing this on the walkthrough - i'm trying to rewrite some puppet modules as salt states that do the same function
19:56 UtahDave nhubbard: awesome!  thanks!  So it sounds like it's working pretty well.
19:57 UtahDave I'll let eliasp do his last confirmation and move forward on that.
19:57 bemehow joined #salt
19:58 n8n joined #salt
19:59 * herlo is verifying that all configs for salt-{master,minion} still reside in /etc/salt by default on RHEL/Fedora systems, correct?
19:59 herlo UtahDave: ^^
19:59 UtahDave herlo: Yep!
19:59 Karunamon the top I understand - all minions will grab the states under '*', you name a file at the top level the same thing as the state in the file, it's getting that far. What I don't understand is how subfolders work in that process. The walkthrough mentions that dots are used for subfolders, so commonpackages.ntp should refer to /srv/salt/commonpackages/ntp.sls, right?
20:00 timoguin Karunamon: "top" in the top.sls should be "base", commonpackages.sls should include commonpackages.ntp, and for the ntp state i recommend reading the pkg state examples in the docs
20:01 rojem joined #salt
20:02 rojem joined #salt
20:02 ksalman so this software i am installing on Windows does not show in the "add/remove programs" in Windows, and I am guess that is why Salt complains that it failed to install it. But it does get installed. Any way to tell Salt to not complain about it?
20:02 ksalman or any better way to do this?
20:03 forrest ksalman, ughhhhhhhhhhh
20:03 ksalman =(
20:03 eliasp ksalman: does it show up in "wmic product get name,version"
20:03 forrest you should tell whoever wrote the software to stop being a lazy PoS
20:04 forrest and add the required entries
20:04 ksalman haha
20:04 ksalman it's cygwin
20:04 timoguin god i hate the cygwin installer
20:04 ksalman eliasp: that command just says "No Instance(s) Available"
20:05 babilen Hi all. I recently filed https://github.com/saltstack/salt/issues/13671 and wouldn't mind implementing this myself, but I am not sure what the most idiomatic way is to get access to data in file_roots. The sls files for reactors are rendered in salt.utils.event.Reactor.render_reaction() and this seems to be the sensible place where to implement this, but how do I get hold of the entire data in file_roots?
20:05 forrest ksalman, timoguin I love cygwin, but yea, their installer is garbage
20:06 forrest I'd still rather use it than powershell
20:06 herlo UtahDave: thanks
20:06 UtahDave you're very welcome, herlo!  anytime.
20:06 UtahDave herlo: We should hit that thai place for lunch sometime soon.  It's been a while.  :)
20:08 timoguin forrest: i uninstalled cygwin to force myself to get better at powershell
20:08 timoguin i actually loves the powershell
20:09 herlo UtahDave: yeah. Maybe when I'm able to eat out again.
20:09 herlo UtahDave: mid-July looks good.
20:09 druonysus joined #salt
20:09 druonysus joined #salt
20:09 racooper what is an appropriate time to wait before repeating a question? sadly, in the last three days, I have had three different questions that nobody's had an answer for (or at least, nobody has responded to...)
20:10 forrest timoguin, can you install virtualenv with powershell?
20:10 taterbase joined #salt
20:10 geekmush joined #salt
20:11 timoguin forrest: yea via pip, https://pypi.python.org/pypi/virtualenvwrapper-powershell/2.7.1
20:11 timoguin i haven't done that though
20:11 ingwaem racooper: could be dependent on who is qualified to answer the question...a lot of times I find if a question is ignored other people may not understand what you're asking, or just may not know entirely...so rather than embarrass themselves by saying the wrong thing they stay mum
20:12 UtahDave herlo: sounds good.  I'm sure I'll be hungry in mid-July.   :)
20:12 * herlo will be STARVING by then. :)
20:12 UtahDave lol
20:12 timoguin racooper: some questions are better for the mailing list too
20:12 ingwaem racooper: the main thing to do is to look out for channel operators...if they are around then chances are someone from SaltStack corporate is in the room...front your questions to them directly if needed, but do so in the main channel with a request to go pm if needed too
20:13 UtahDave racooper: Sorry about that.    Yeah, sometimes the mailing list is the best option if nobody is around that knows how to help you.  Timezones and travel affect who is in here.
20:13 babilen How would I access data in file_roots most idiomatically? (in salt code)
20:13 jhauser joined #salt
20:13 UtahDave racooper: What questions do you have?
20:14 eliasp UtahDave: can't confirm wrt dockerio right now (http://pastebin.kde.org/pp2xuih2i), but given that it runs fine for nhubbard, I'd say: go for it
20:14 racooper of course, I understnad that :) not upset about anything, was just wondering how soon would be too soon to repeat questions :)
20:14 UtahDave babilen: I'd look in other spots in the code.  You'll probably want to grab the opts['file_roots'] option and use that as the base of your file path
20:14 UtahDave eliasp: OK, will do.  Thanks!
20:14 ingwaem racooper: if there's a conversation going on and you see the ops are busy with that and you ask a question...I think if you haven't had a response in 20 minutes would be good to throw it out there again...and maybe include someones name...eg.
20:14 racooper UtahDave,  the most recent question is https://gist.github.com/racooper/b44e9d1168c1ed52d3a1 trying to figure out why I'm getting requisite not found when the group already exists
20:15 babilen UtahDave: Yeah, I am exactly looking for "other spots", but not quite sure where to start. Let me grep for "opts['file_roots']" and hope that I find something sensible in the vicinity
20:15 ingwaem hey UtahDave: I asked a question a little while back but didn't hear anything back...any chance you can help out ?  (and then he'll read up and find it, unless you're kind enough to paste it again :) )
20:15 UtahDave babilen: it might be __opts__['file_roots']
20:15 racooper I do try to ask during the lull times :)
20:16 forrest herlo, I had to give up on buildbot as a heads up, we went with Jenkins
20:16 alanpearce joined #salt
20:16 herlo forrest: jenkins is annoying
20:16 ingwaem hehe :) fair do's
20:16 herlo forrest: but I understand
20:16 UtahDave ingwaem: Yeah, just a minute and I'll go find your question
20:16 forrest herlo, *shrug* so is buildbot considering how poorly it supports multiple projects, wasn't a realistic solution for people to maintain :\
20:16 ingwaem UtahDave: was an example...no worries :) was for racooper :)
20:17 herlo forrest: multiple projects? You mean the same thing jenkins calls jobs?
20:17 babilen UtahDave: Yeah, I am simply looking for the function/method that takes a glob (or something) and returns me all (matching if possible) sls files in file_roots.
20:17 forrest herlo, ye
20:17 forrest *p
20:17 racooper and earlier question this morning: "I know there is a command module for quota, but is there a way to set user quotas as a state?"
20:17 herlo forrest: yeah, I hadn't gotten that far yet.
20:17 babilen UtahDave: I mean I can write the matching myself and would be happy with give me all of file_roots *after* merging all of them and GitFS
20:17 forrest herlo, yea, I looked at what it would take to support it from a front end, and it would require rewriting a good chunk of buildbot internals
20:18 herlo forrest: indeed.
20:18 herlo for me, we're just supporting a customer.
20:18 babilen There is no #salt-dev is there?
20:18 forrest since the conf is where all the jobs live which requires a reload/restart of the service to be picked up. I think it has great potential still, but I like it the best for personal projects
20:18 forrest babilen, yea there is
20:18 babilen ah!
20:19 UtahDave racooper: can you pastebin the sanitized output of  pillar.get 'sas_users'
20:19 UtahDave babilen: there's a salt function that will give you all the files in the file_roots
20:19 UtahDave babilen: yeah, there is a #salt-devel
20:20 babilen UtahDave: I take it there
20:22 racooper UtahDave,  pillar.get isn't returning anything.  pillar.item does, but not sure if that's helpful.
20:22 UtahDave racooper: ah yeah.  sorry, you have the correct syntax
20:23 JoeHazzers joined #salt
20:23 n8n_ joined #salt
20:23 ndrei joined #salt
20:23 debian112 joined #salt
20:23 rmnuvg joined #salt
20:24 masterkorp joined #salt
20:25 racooper UtahDave,  added it to the gist, pillar.item_output
20:25 UtahDave k
20:25 abele joined #salt
20:25 stevednd is there a way to use a file on a minion as a template for file.managed?
20:25 smcquay_ left #salt
20:26 debian112 any tips on reading a file for a key word?
20:29 UtahDave racooper: it's kind of hard to troubleshoot all that jinja.  The error you're getting means what it says; it can find a group.present for wendyliu
20:30 UtahDave racooper: So you're going to need to figure out why you have a require for that group wendyliu, but the group.present  for wendyliu doesn't exist
20:30 * robawt highfives UtahDave
20:31 n8n joined #salt
20:31 ksalman i guess I will set a grain and not install cygwin on subsequent runs if a grains is set.. sigh
20:31 UtahDave yo, robawt!
20:31 racooper the code is stripped down from the user-formula, I didn't write it. the group does exist though, I can see it in /etc/groups
20:32 racooper (this is trying to bring a bunch of existing servers into state compliance...so pretty much everything I'm working on is already set up, I just want to make it salty.)
20:32 timoguin racooper: for the quotas stuff, if the functionality is not available in the state, you can use module.run or module.wait to call any execution module
20:32 timoguin if it's not in an exec module, cmd.run or add it to the module and send a PR. :)
20:33 UtahDave racooper: Salt doesn't care if the group exists on the system. it cares if there is a group.present for wendyliu in your sls file somewhere
20:34 racooper ah. I'll go back and see what else I stripped out of users-formula that might have borked that then.
20:34 anuvrat joined #salt
20:34 racooper or just remove the require, since user.add on RHEL creates those automatically anyway.
20:34 UtahDave cool.  let me know how that goes, racooper.  Sorry I'm not more help, but it's tough to troubleshoot sls files with so much jinja
20:34 n8n_ joined #salt
20:35 racooper yeah. I've been staring at it for three days :P
20:35 racooper and thanks timoguin for the hint on quotas.
20:36 rojem joined #salt
20:36 Ahlee heh, ain't that the truth
20:37 racooper my last question (well, first, from monday is about gitfs and pillars. I can't get it working, though I do have it working for the state tree.
20:37 racooper https://gist.github.com/racooper/373a3a06d2ecaca8c545
20:37 shaggy_surfer joined #salt
20:37 KyleG joined #salt
20:37 KyleG joined #salt
20:38 racooper tried several different variants on the URI for but no luck.
20:39 herlo okay, dumb question, but why would I put the 'base' reference in my top.sls for my 'dev' file_root?
20:39 debian112 How can I get salt to read a value from a file?
20:39 * herlo notes this is documented on http://docs.saltstack.com/en/latest/ref/states/top.html, section '22.25.17.3. How Top Files Are Compiled'
20:40 herlo UtahDave: ^^
20:40 jchen racooper: git+ssh://git@github.tamu.edu:racooper/ocss-salt-states.git
20:40 testt joined #salt
20:40 jchen and the other way around for https
20:41 jchen https://github.tamu.edu/racooper/ocss-salt-states.git
20:41 racooper the states URI works; I had to replace the : with / for it to do so. i'ts the pillars URI that doesn't
20:41 LordMike91 Anyone here have experience with Pkgrepo?
20:42 UtahDave herlo: a base environment is required. I get around that by adding an "empty.sls" that goes under your base environment.
20:42 jchen racooper: i'm /pretty/ sure all of your git+ssh references iether in gitfs_remotes or in pillar settings as it stands are invalid
20:42 jchen you want to make it look like what i just said
20:42 kballou joined #salt
20:43 herlo UtahDave: hmm, that seems like an oversight.
20:43 racooper and I can't use https...our github server requires authentication to get to ANYTHING so ssh using keys is the only option I have.
20:43 jchen racooper: ok, then use the git+ssh uri i gave
20:43 herlo UtahDave: also, if you look at #3 under that same section, there is no 'base'.
20:43 bensons_ jo ich auch bald, n8 digger
20:43 bensons_ er
20:43 timoguin jchen: racooper: I have that exact URL syntax in my setup and it's working
20:44 jchen looks like i lied
20:44 tkharju1 joined #salt
20:45 jchen what happens if you try to clone git@github.tamu.edu:racooper/ocss-salt-pillars.git directly?
20:45 UtahDave herlo: Ah, ok, so you're using a separate top.sls for your dev environment?
20:47 borgstrom joined #salt
20:47 racooper git clone works to the pillar and the state repos.  git fetch does not, for either.
20:48 racooper both give the error "fatal: Not a git repository (or any of the parent directories): .git"
20:48 nineteeneightd joined #salt
20:50 tkharju1 joined #salt
20:50 bhosmer joined #salt
20:50 herlo UtahDave: yes
20:51 herlo dev will have its own stuff, so will prod/master.
20:51 UtahDave OK, I'd avoid doing that.   Just use one top.sls in your file_roots
20:51 herlo why?
20:51 herlo also, I have an interesting branching schema in git that helps with this...
20:51 UtahDave It gets confusing quickly
20:51 UtahDave the main top file can manage all your environments
20:52 herlo UtahDave: so I have a dev, config and master branch which map to the /etc/salt/ /srv/salt/dev and /srv/salt/prod directories respectively.
20:52 herlo UtahDave: right, but then I would have to have a separate branch *just* for the top.sls.
20:53 UtahDave herlo: Hm.  Ok, so you're doing all this with gitfs?
20:54 herlo UtahDave: no
20:54 herlo UtahDave: this is just for managing it with git directly. No gitfs atm. Maybe down the road.
20:54 Gareth 'LO
20:54 * herlo isn't sure he likes the idea of gitfs.
20:54 Gareth erm
20:54 Gareth 'lo :)
20:54 herlo hi Gareth!!
20:54 herlo fancy meeting you here...
20:54 Gareth herlo: Hey :) hows it going?
20:54 Gareth herlo: I am *everywhere*
20:55 herlo Gareth: going well, other than UtahDave is being difficult. :)
20:55 UtahDave lol
20:55 Gareth herlo: UtahDave is *always* difficult.
20:55 * herlo looks around, sees no Gareth in his bedroom. :)
20:55 herlo Gareth: :)
20:55 herlo UtahDave: just joshing you
20:55 Gareth herlo: I'm hiding under the bed.
20:55 UtahDave I know.
20:55 herlo Gareth: looked there...
20:56 Gareth herlo: trap door.
20:56 herlo Gareth: are you microscopic?
20:56 herlo lol
20:56 herlo Gareth: wow, didn't even know that was there!
20:56 gothix joined #salt
20:56 * Gareth nods and smiles in a creepy fashion
20:56 UtahDave herlo: so are you getting a stacktrace with what you're doing right now?
20:57 herlo UtahDave: no, not even getting that far yet.
20:59 rojem joined #salt
21:01 oz_akan_ joined #salt
21:01 rojem joined #salt
21:03 unixman9000 joined #salt
21:05 tinuva joined #salt
21:05 ksalman whats the best way for salt to figure out the windows architecture? amd64 vs i386?
21:07 eliasp ksalman: that's available as a grain
21:07 ksalman oh?
21:07 ksalman huh
21:07 dlam joined #salt
21:07 ksalman i see the grain cpuarch, but I thought that was for the cpu
21:07 ksalman not windows =)
21:08 eliasp ksalman: ah, you're right… might need a custom grain then
21:08 dlam hmmm i got a  file.managed salt state, and i made some updates to it which should change the file but it's not!  anyone know how to debug it??
21:08 racooper I did just find something interesting. if I clone my ocss-salt-pillars.git repo,then go into the cloned directory, the 'git fetch' works.
21:09 dlam on the minion im calling it like:  salt-call state.single file.managed name="/home/dlam/davids-cool-files"
21:09 eliasp ksalman: your custom grain could determine the architecture using "wmic OS get OSArchitecture"
21:10 ksalman eliasp: thats great thanks!
21:12 ksalman does custom grains need to be written in python or can i whip up a batch file?
21:12 timoguin python and just shell out
21:13 ksalman true..
21:14 krow joined #salt
21:15 druonysus joined #salt
21:16 whiteinge whoa, who unbanned herlo from this channel? UtahDave was it you?
21:16 herlo whiteinge: you've got it all wrong. It was I who unbanned the channel from me. :)
21:16 herlo whiteinge: 13:28 -!- Irssi: Unignored #salt
21:16 forrest lol
21:17 herlo whiteinge: you'll see that indeed, I started talking shortly after this time MDT. :D
21:17 whiteinge hmph. likely story. i've got a ban-macro at the ready if you don't behave yourself
21:17 * herlo continues to wait upon UtahDave....
21:18 forrest herlo, Dave hasn't been around for a while
21:18 herlo whiteinge: I dare you. :) I will happily just put you on ignore
21:18 herlo forrest: I was talking to him in a PM until about 5 minutes ago.
21:18 forrest herlo, oh ok
21:18 herlo well, 10
21:18 herlo but yeah.
21:20 Ryan_Lane joined #salt
21:22 blarghmatey joined #salt
21:22 Guest7481 joined #salt
21:22 anuvrat joined #salt
21:25 whiteinge left #salt
21:26 Hell_Lap joined #salt
21:27 mastrolinux joined #salt
21:29 whiteinge joined #salt
21:38 Ryan_Lane joined #salt
21:38 smcquay joined #salt
21:40 cwyse Hey guys.  Anyone know if rest_tornado is supported in salt-api 0.8.2-0?
21:41 * herlo kicks whiteinge
21:42 jnials_laptop joined #salt
21:42 srage_ joined #salt
21:43 Karunamon Can you have multiple top-level definitions in a single sls file?
21:43 whiteinge cwyse: no. it went in 0.8.4 iirc
21:43 whiteinge hm. the docs move obscures that info
21:43 herlo Karunamon: yes
21:44 herlo er...
21:44 Karunamon say, a pkg: definition, and then a related /file/path: definition in the same thing
21:44 Karunamon same file, but both at the top level of the yaml
21:45 herlo oh, I think that's fine.
21:45 cwyse So I see that salt-api is now included with the helium release of salt.  Forgive my ignorance.  But how do I know which version/tag is the helium release?
21:47 herlo can I run salt-call with something like --noop?
21:47 eliasp cwyse: Helium is the current 'develop' branch… it will be (very likely) the 2014.7 release
21:47 whiteinge cwyse: it will be in the *upcoming* helium release. i'll restore the old salt-api docs
21:47 cwyse Thank you.
21:47 eliasp herlo: if you use salt-call to apply a highstate, use test=True
21:47 herlo eliasp: ahh, that's right. Thanks
21:48 * herlo is trying to get his brain to remember stuff from a year+ ago
21:48 mprymek joined #salt
21:49 Hell_Fire_ joined #salt
21:50 juanlittledevil joined #salt
21:50 bhosmer joined #salt
21:51 ajprog_laptop joined #salt
21:53 eliasp yehaw, circular dependency… making sure my Saltmaster Docker container is up'n'running via Salt's Docker state… :)
21:57 btorch do I remove /var/cache/salt/ on the master to make it forget using the cache ? I have changed a state and restarted the master and the minion but for some reason salt is going nuts on me :)
21:58 UtahDave btorch: yeah, you can do that.
22:00 whiteinge cwyse: restored: http://salt-api.readthedocs.org/en/latest/
22:02 brendangilmor joined #salt
22:02 n8n joined #salt
22:05 kballou I must not be asking the googles the right words, but how do I manage several unix groups and user membership which may not be mutually exclusive?
22:08 shaggy_surfer joined #salt
22:09 kballou actually, UtahDave has the gist I needed: https://gist.github.com/UtahDave/3785738 <- thanks
22:09 UtahDave :)  I'm glad you found it!
22:11 unixman9000 left #salt
22:11 unixman9000 joined #salt
22:12 bemehow_ joined #salt
22:14 Guest7481 joined #salt
22:17 tkharju2 joined #salt
22:17 dstokes hey guys. i'm trying to get salt-api working on the dev branch with rest_cherrypy config, but when i install cherrypy and bootstrap salt to develop, the salt src directory is missing the cherrypy files which causes salt-api to exit silently
22:18 dstokes am i doing something wrong here?
22:18 londo_ joined #salt
22:20 jcsp1 joined #salt
22:24 jalbretsen joined #salt
22:26 whiteinge dstokes: are you seeing an error message?
22:27 dstokes nope, exit code 0. writes a pidfile, but never starts a process
22:27 dstokes that's w/ and w/o -d (daemon mode)
22:28 whiteinge dstokes: sounds like it's missing either req'd config or req'd deps. start it with higher logging:  salt-api -l debug
22:28 dstokes i looked at the code and there's a check for a start func in the netapi module that seems to not be fufilled
22:28 dstokes -lall ends with "[TRACE   ] Skipping /var/cache/salt/master/extmods/netapi, it is not a directory"
22:28 whiteinge you can ignore that
22:28 dstokes last debug message is pidfile creation
22:29 whiteinge nothing about the rest_cherrypy module?
22:29 dstokes nope
22:29 whiteinge what does the rest_cherrypy section in your master config look like?
22:29 dstokes line i mentioned above is https://github.com/saltstack/salt/blob/develop/salt/client/netapi.py#L30
22:30 shaggy_surfer joined #salt
22:30 dstokes config is http://paste.ubuntu.com/7702979/
22:30 whiteinge yeah, that's ok. it's looking for this: https://github.com/saltstack/salt/blob/develop/salt/netapi/rest_cherrypy/__init__.py#L74
22:31 dstokes which is not in my salt src dir after bootstrap
22:31 dstokes i've tried reinstalling before & after cherrypy install (w/ pip) and it never copies over netapi/rest_{cherrypy,tornado}
22:31 whiteinge hm. maybe the bootstrap installed the wrong version?
22:31 dstokes only the rest_wsgi.py file
22:32 whiteinge interesting...
22:32 dstokes bootstrapping w/ `curl -L http://bootstrap.saltstack.org | sudo sh -s -- -M git develop`
22:32 whiteinge i'll run that now on a fresh VM. what distro are you running?
22:33 torrancew joined #salt
22:34 dstokes whiteinge: Ubuntu 12.04.3 LTS
22:34 * whiteinge nods
22:34 matrix3000 joined #salt
22:34 dstokes also, looking at src files in /usr/lib/python2.7/dist-packages/salt (pretty sure that's right)
22:35 krow joined #salt
22:35 allanparsons joined #salt
22:35 whiteinge dstokes: are you running bootstrap over an existing install perhance?
22:36 dstokes yes-ish. on every install i've removed everything i can find except the bin & config files
22:37 dstokes at one pt i had several installs (apt, pip, bootstrap) but that should all be cleaned up now
22:39 whiteinge sometimes it can be tricky to clean out old installs. my VM is bootstrapping now. sec...
22:40 dstokes whiteinge: you using vagrant + salt plugin? got one of those running as well but haven't tested there yet, as i believe it does a stable install
22:41 whiteinge no, rackspace VM
22:42 dstokes ah. just found an option for install_type in vagrant \0/
22:42 krow joined #salt
22:42 smcquay joined #salt
22:44 gothix Anyone know long until RHEL 7 packageas get released
22:44 mosen joined #salt
22:45 whiteinge dstokes: confirmed. not sure what the disconnect is yet. looking into it...
22:46 dstokes whiteinge: got it, thx for your help so far
22:47 pdayton joined #salt
22:47 ipmb joined #salt
22:47 Outlander joined #salt
22:48 Hollinski joined #salt
22:48 oz_akan_ joined #salt
22:48 kwmiebach joined #salt
22:48 to_json joined #salt
22:49 octarine joined #salt
22:49 whiteinge np. sorry for the trouble
22:49 seventy3 joined #salt
22:49 akoumjian joined #salt
22:50 whiteinge looks like those subdirs aren't getting included in salt's python sdist build
22:50 gldnspud joined #salt
22:51 gamingrobot joined #salt
22:54 dude051 joined #salt
22:59 dude051 joined #salt
23:00 analogbyte joined #salt
23:01 dstokes whiteinge: i was digging around for that logic in setup.py but got pulled away
23:02 whiteinge i'm not super familiar with setup.py but i'm very much under the impression that this is all that is needed: https://github.com/saltstack/salt/blob/develop/setup.py#L505
23:03 mrchrisadams joined #salt
23:03 smcquay whiteinge: I have worker_threads set to 16; is that not high enough? 400 machines.
23:04 whiteinge smcquay: i think that's plenty. you'll still get all the returns with a low worker count, it'll just take a little bit for them all to get processed
23:05 smcquay alright; thanks.
23:06 whiteinge dstokes: dang, i'm wrong. we need setup.py entries for each subdir
23:06 whiteinge adding...
23:07 dstokes whiteinge: if this works, i owe you beer ;p
23:07 smcquay yeah, even after the updates to minon configs I get salt commands running for a while, then stopping, then timeout for subset of hosts.
23:07 dstokes been on it for 2 days now
23:10 whiteinge dstokes: https://github.com/saltstack/salt/pull/13746
23:11 whiteinge dstokes: no, i owe you a beer for catching that before we got any closer to the helium releease :-P
23:11 dstokes :D
23:11 tkharju joined #salt
23:13 mrchrisadams joined #salt
23:17 stevednd whiteinge: is there a way to use a file on a minion as a template for file.managed?
23:17 analogbyte joined #salt
23:18 ajolo joined #salt
23:19 anuvrat joined #salt
23:20 dstokes stevednd: you try `source: /path/to/file`?
23:20 kivihtin joined #salt
23:21 stevednd dstokes: last time I tried that I recall salt complaining that it couldn't find the file. it seemed like salt only wanted a file from the master
23:23 dstokes "The source file to download to the minion, this source file can be hosted on either the salt master server, or on an HTTP or FTP server"
23:23 yomilk joined #salt
23:24 dstokes may be better off using one of the file mod methods (i.e. patch, blockreplace, prepend, append etc..)
23:24 stevednd yep, I'm looking for alternatives or a workaround. whiteinge is usually pretty good about detailing those
23:27 stevednd I'm also trying to think of a way to do fallback for files to manage
23:27 oz_akan_ joined #salt
23:28 mastrolinux joined #salt
23:28 dstokes whiteinge: just pulled your change, *queues "it's working!" gif*
23:30 acabrera joined #salt
23:31 whiteinge woot!
23:31 logix812 joined #salt
23:31 whiteinge thanks again for finding that
23:32 whiteinge stevednd: i don't think that can be done with file.managed (would be a nice addition though -- not sure why it's not in there)
23:32 whiteinge i *think* it can be done via an execution module. sec...
23:32 dstokes i think the question is how you would get the template on the minion in the first place (if it was coming from http or the master)
23:33 rudolfsteiner joined #salt
23:33 dstokes doesn't seem like a common use case to bootstrap a server that already has jinja templates on it
23:33 nan2nan joined #salt
23:33 stevednd whiteinge: even if it could I think I'm still back to the problem I've had in previous days in that I need to dynamically determine the template files to use based on the project checkout
23:34 stevednd dstokes: the files would come as the result of a git checkout of the application being deployed needing some custom configuration based on the environment they're benig deployed to
23:35 whiteinge stevednd: can you make that determination with jinja checking the {{ env }} var and changing the ``source`` argument to file.managed accordingly?
23:36 dstokes stevednd: crap, that makes sense ;p
23:36 stevednd whiteinge: no, the files reside in the application repo, which gets fetched as one of the states that is  run
23:36 Luke__ joined #salt
23:37 rallytime joined #salt
23:37 dvogt joined #salt
23:37 Blacklite joined #salt
23:38 dstokes there's probably a way to could invoke the salt jinja render w/ custom context on a local file
23:38 jonbrefe joined #salt
23:38 bhosmer joined #salt
23:40 bhosmer_ joined #salt
23:40 Blacklite We have a number of servers with LAMP stacks set up on them. on one of these servers we want to have a suhosin config file, but i don't want it on all servers. And if i want to set up suhosin on a server in future, i'll probably want a different config file. What's the best way to have custom per-host config files?
23:41 Blacklite I've googled around a bit but I can't find anything, which probably means I'm going about it the wrong way. I just need steering in the right direction :)
23:42 stevednd Blacklite: have a different config file for each server named or prefixed with the server id or something. Use grains to get the id, and use the right file
23:42 dstokes stevednd: that may be a pipe dream (using renderer). no module exposed for jinja
23:42 dstokes from what i can see
23:42 Blacklite ok I'll give it a go and get back to yo
23:42 Blacklite u
23:42 Blacklite thanks
23:43 stevednd dstokes: yeah, I would probably have to drop to the command line to do it, or a custom module to load jinja and the context myself
23:46 anuvrat joined #salt
23:47 dstokes stevednd: wonder if you could copy the files to cache dir and render that way..
23:47 aw110f herlo: while waiting for my fedora virtualbox image to download, I went ahead and tried to use mock on Centos6 to rebuild salt RPM for centos5 with the changes i needed... and that worked
23:47 aw110f thanks herlo:
23:48 sdebot joined #salt
23:48 stevednd on a somewhat different topic does anyone by chance know what order 'includes' run in? Is it when they appear, or before the other states?
23:48 dstokes stevednd: actually, this might work.. http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.py.html#module-salt.renderers.py
23:49 rallytime joined #salt
23:49 dstokes "It is also possible to use the py renderer as a templating option. The template would be a python script which would need to contain a function called run(), which returns a string. The returned string will be the contents of the managed file. For example:"
23:49 ajolo_ joined #salt
23:51 dvogt joined #salt
23:51 druonysus joined #salt
23:51 druonysus joined #salt
23:56 stevednd dstokes: I like how it then neglects to show how to actually call the script. Would it just be salt://app/files/script.py
23:57 dstokes it's a module, so salt['renderers.py'](template, env, path) _should_ work..
23:57 bhosmer joined #salt
23:57 dstokes sry __salt__[...
23:58 stevednd yeah, that won't work then
23:58 dstokes wait, it's not a module?... you're right, this is weird.
23:59 stevednd because that will execute during the templating part of the jinja doc. The file won't actually exist until the state is run to fetch from git
23:59 dstokes right

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