Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-10-14

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

All times shown according to UTC.

Time Nick Message
00:00 aqua^c joined #salt
00:04 woodtablet left #salt
00:07 Edgan joined #salt
00:08 ageorgop joined #salt
00:12 otter768 joined #salt
00:12 thalleralexander joined #salt
00:19 pm90_ joined #salt
00:21 aron_kexp joined #salt
00:24 pravka joined #salt
00:25 iggy moogyver: yeah, that's what happens when you have thousands of people contributing to a project
00:36 wryfi left #salt
00:42 otter768 joined #salt
00:47 scoates joined #salt
00:50 ashirogl joined #salt
00:51 TyrfingMjolnir joined #salt
00:54 ageorgop joined #salt
00:58 TyrfingMjolnir joined #salt
01:11 chequers joined #salt
01:23 ageorgop joined #salt
01:24 mike2512 left #salt
01:40 teryx510 joined #salt
01:42 aparsons joined #salt
01:45 catpigger joined #salt
01:45 ageorgop joined #salt
02:13 bfoxwell joined #salt
02:18 writteno1 joined #salt
02:41 evle joined #salt
03:04 clintberry joined #salt
03:18 ageorgop joined #salt
03:20 jaybocc2 joined #salt
03:23 moo joined #salt
03:25 otter768 joined #salt
03:26 stanchan joined #salt
03:29 rdas joined #salt
03:33 papertigers joined #salt
03:34 papertigers So im looking at targeting specific nodes.  It looks like salt does that via minion id.  Anyway to change an existing minion id or is it remove the key and recreate it?
03:35 _JZ_ joined #salt
03:36 moogyver you can rename the key on the master and then change the /etc/salt/minion_id file on the minion
03:36 beardedeagle joined #salt
03:36 moogyver master stores the keys in /etc/salt/pki/master/minions/<minionname>
03:38 papertigers That worked great moogyver
03:38 papertigers perfect
03:39 papertigers coming from chef, so mindset is a bit different
03:39 moogyver yeah it's a bit different
03:39 moogyver we still use chef for our cfg mgmt but salt for all our execution/orchestration
03:39 moogyver use salt to tell our chef-solo boxes to do a chef run :)
03:40 chequers left #salt
03:42 papertigers im starting fresh in this env, so I figured salt for everything
03:42 papertigers see how I like it
03:43 ageorgop joined #salt
03:43 nzero joined #salt
03:44 papertigers moogyver: is there anything tricking to passing in the target?  salt foo-*  vs 'foo-*'
03:44 papertigers one way is failing for me ;)
03:45 moogyver you have to use single quotes or you have to escape the * - ie - salt foo-\*
03:45 breakingmatter joined #salt
03:45 moogyver otherwise the shell with interpret the *
03:46 moogyver with=will
03:47 ramteid joined #salt
03:47 favadi joined #salt
03:48 papertigers right, shell expansion is happening first
03:48 papertigers its late :)
03:52 moogyver papertigers: you can also target minions based off grain data, pillar data, and some other things ( or combinations )
03:52 moogyver https://docs.saltstack.com/en/develop/topics/targeting/index.html
03:52 moogyver https://docs.saltstack.com/en/latest/topics/targeting/compound.html
03:52 beardedeagle I use hostnames for cicd pipeline I am building out
03:52 beardedeagle as an example
03:54 jaybocc2 joined #salt
03:57 papertigers how do you add specific nodes to a pillar
03:57 papertigers or is that done through things like base
03:58 moogyver it's done in base ( /srv/pillar/top.sls ).  you specify a matcher to target certain hosts, then assign them pillar files
03:58 otter768 joined #salt
03:58 papertigers ahh perfect, just found that in the doc as you replied
03:59 moogyver https://gist.github.com/sjmh/ff5768e258b4eafaedd1
03:59 moogyver simple example from our lab
03:59 moogyver so the 'steve' host has the pillar data 'provzone: qcy-zone-b'
04:00 papertigers top.sls can define many pillars I take it then
04:00 papertigers alpha and steve in this case
04:00 moogyver yup - top.sls is for matching the minions to their pillar files
04:00 moogyver but you can also do stuff with external pillars
04:02 moogyver where you get pillar data from external systems like oconsul, foreman, etcd, s3, etc
04:02 beardedeagle left #salt
04:04 moogyver you don't have to target by minion-id in the top.sls file though, you can use grains and stuff there as well
04:09 zmalone joined #salt
04:13 pogotech_ joined #salt
04:29 ageorgop joined #salt
04:33 Norbell joined #salt
04:35 Norbell joined #salt
04:36 pogotech joined #salt
04:37 pogotech joined #salt
04:42 pogotech left #salt
04:50 viq joined #salt
04:50 aparsons joined #salt
04:53 aparsons_ joined #salt
04:55 hasues joined #salt
04:58 aparsons joined #salt
04:58 jalbretsen joined #salt
05:01 evle joined #salt
05:06 rdas joined #salt
05:20 mehakkahlon joined #salt
05:20 opensource_ninja joined #salt
05:25 kinetic joined #salt
05:30 rdas joined #salt
05:37 anmol joined #salt
05:45 kinetic_ joined #salt
05:45 ronrib joined #salt
05:45 teebes joined #salt
05:46 evle joined #salt
05:54 favadi joined #salt
05:55 kinetic_ joined #salt
05:57 kinetic joined #salt
05:58 mpanetta joined #salt
05:58 felskrone joined #salt
05:58 impi joined #salt
06:07 malinoff joined #salt
06:09 jaybocc2 joined #salt
06:11 edulix joined #salt
06:13 sirex joined #salt
06:14 lb joined #salt
06:17 kinetic_ joined #salt
06:19 kinetic joined #salt
06:21 cornfeedhobo does anyone know if it is possible to deploy using a pre-made, self signed intermediate CA in place of the `master_sign` pair? i see info on salt generating keys, but nothing about deploying pre-existing ones
06:26 kinetic_ joined #salt
06:28 slav0nic joined #salt
06:31 shiriru joined #salt
06:34 KermitTheFragger joined #salt
06:36 kinetic joined #salt
06:46 stephanbuys joined #salt
06:46 elsmo joined #salt
06:48 favadi joined #salt
06:49 felskrone joined #salt
06:55 ramteid joined #salt
07:07 linjan joined #salt
07:10 Roee joined #salt
07:10 Roee Hi, I'm tying to run mongodb.find from salt and getting syntax error
07:10 Roee my query is: salt P_EDITIS_DIRECT_DB mongodb.find (accounts find({},{_id:1,url:1,name:1}) jenkins jenkins localhost admin)
07:11 Roee can someone tell me what is the problem ?
07:13 hasues left #salt
07:15 ^C how are people syncronising their syndic nodes' file_roots with the master master? rsync on cron? salt-schedule?  reactor?
07:17 eseyman joined #salt
07:18 minfi joined #salt
07:18 Ztyx joined #salt
07:19 kinetic joined #salt
07:20 permalac_ joined #salt
07:20 malinoff joined #salt
07:20 jaybocc2 joined #salt
07:21 scc joined #salt
07:21 Ztyx left #salt
07:21 stephanbuys1 joined #salt
07:23 keimlink joined #salt
07:27 toanju joined #salt
07:36 mpanetta joined #salt
07:37 Grokzen joined #salt
07:38 jaybocc2_ joined #salt
07:41 kinetic joined #salt
07:51 katyucha joined #salt
07:51 kinetic joined #salt
08:05 kinetic joined #salt
08:05 TyrfingMjolnir joined #salt
08:06 rdas joined #salt
08:09 SheetiS1 joined #salt
08:09 PI-Lloyd ^C: we use gitfs to sync the file_roots
08:10 MadHatter42 joined #salt
08:10 kinetic joined #salt
08:12 ^C means your constantly git pushing tho, yes?
08:12 PI-Lloyd only when we make changes to the states
08:12 trapha joined #salt
08:12 ^C which unfortunately is quite often while i build :(
08:12 ^C was thinking about nfs
08:14 PI-Lloyd while you "build" ?
08:14 ^C adding states, configuring pillars
08:14 PI-Lloyd use vagrant for developing your states
08:14 PI-Lloyd all done locally that way, then push up when it all works
08:16 toanju joined #salt
08:16 elsmo joined #salt
08:20 jaybocc2 joined #salt
08:20 AndreiStaicu joined #salt
08:20 kinetic joined #salt
08:23 rdas joined #salt
08:24 chiui joined #salt
08:25 kinetic joined #salt
08:28 AndreiStaicu hi, i'm trying to dsitribute generated ssh keys via mine. after generating them, i can publish them to the mine with this state: https://gist.github.com/smarmie/d41d382a3c41753633d4 i see them in the mine.p file on the master, but i can't seem to be able to get them back. salt minion mine.get 'ssh_key_root' glob returns nothing.
08:29 AndreiStaicu any idea what i'm missing here? thanks.
08:29 ziro` joined #salt
08:29 VSpike If I want a file with only Jinja macros in it to share between sls files, should i have my macro file with a .sls filename, and when doing {% import 'foo/bar/baz.sls' as baz %} how do the paths work?
08:29 VSpike What are they relative to?
08:29 thefish joined #salt
08:30 s_kunk joined #salt
08:32 armyriad joined #salt
08:32 thefish left #salt
08:36 AndreiStaicu ah, wait. my bad. i completely missed how mine.get works. it should be salt minion mine.get '*' another_minion ssh_key_root . sorry for the confusion.
08:38 babilen AndreiStaicu: Yeah, you always have to include a target expression in which you specify the minions in which data you are interested
08:38 malinoff joined #salt
08:39 kinetic joined #salt
08:45 CeBe joined #salt
08:48 pi3r joined #salt
08:50 pi3r Is there a way to use __salt__['pkg.install'] (or the equivalent) in a module and getting back a return dict with the 'retcode' or 'result'. I am so surprised that the ret dict is empty
08:51 pi3r I can use __salt__['cmd.run'] instead but it is going to be quite 'adhoc' ...
08:54 kinetic joined #salt
08:55 Roee getting the following error in log when runing mongodb.find : "TypeError: can't serialize ValueError('No JSON object could be decoded',)"
08:56 Roee did someone ever seen this ?
08:59 babilen pi3r: It should return a dictionary with keys for all packages whose state changed. The value would be a dict in which you have 'old' and 'new' keys for the respective version before and after the run.
08:59 babilen (at least that's how aptpkg behaves)
09:03 pi3r @babilen I am logging the content of "ret =__salt__['pkg.install'](cmd, refresh=True, skip_verify=True)" and it is empty. The python file is a custom module (called with func in an orchestrate sls)
09:04 pi3r @babilen I meant it is an empty dict
09:04 babilen cmd?
09:05 babilen You don't seem to pass 'name' (i.e. the package to install)
09:05 pi3r @babilen cmd is just the name of the package
09:05 babilen Well, it would be cmd, but could you elaborate on your code a little?
09:06 babilen Okay, which package management system is this?
09:06 pi3r @babilen I am on centos, so I expect it to call yum at the end
09:07 babilen Okay, so looking at aptpkg is pointless
09:08 kinetic joined #salt
09:08 babilen There are various places in which an empty dict is returned. What does https://github.com/saltstack/salt/blob/develop/salt/modules/yumpkg.py#L967 return for you?
09:09 babilen https://github.com/saltstack/salt/blob/2015.8/salt/modules/yumpkg.py#L980 (assuming you are on 2015.8, adjust branch as needed)
09:10 babilen And: Does it return an empty dict or just None?
09:11 babilen Is something actually done on the minion? Could you check its debug log? (e.g. run salt-call -ldebug ..... on the minion)
09:11 pi3r @babilen I will try it but what is the best way to install a pkg in a py module if I do want to know why it would fail ?
09:12 pi3r @babilen yes it is done on the minion at the end (but it comes from an orchestration sls)
09:14 pi3r @babilen So to be more precise, I am calling an orchestration job (so it is on the master) and the master is going to target minion through the orchestration script
09:15 babilen Well, it would raise a CommandExecutionExeption (you can check that it is an instance of six.stringtypes and then print the exception)
09:15 babilen Oh, and you don't need the @ on IRC. It is customary to simply use "$NICK:" or "$NICK," at the beginning of the line.
09:16 babilen https://github.com/saltstack/salt/blob/2015.8/salt/states/pkg.py#L1072 that's how it is used in the pkg state
09:18 kinetic joined #salt
09:23 jcristau joined #salt
09:26 larsfronius joined #salt
09:26 KermitTheFragger joined #salt
09:27 chrismckinnel joined #salt
09:27 kinetic joined #salt
09:27 chrismckinnel hey guys, i have a nodegroup defined nodegroups: group1: 'L@some*.domain.co.uk' and it's not matching any of my servers when i do 'salt -N group1 test.ping'
09:28 chrismckinnel but when i define the nodegroup as 'L@some-full.domain.co.uk' it works
09:28 chrismckinnel thoughts?
09:28 chrismckinnel hmm, removing the L@ seems to have fixed it, i guess with L@ you need to make it a list?
09:30 chrismckinnel now only about half return with a successful ping, but hitting them individually works
09:31 pi3r $babilen: So pkg_params is {'puppet-middleware-testing': None} and pkg_type is repository.
09:32 pi3r $babilen: I am actually on 2015.5.5 (2015.8 is not yet available from epel)
09:33 ziro` joined #salt
09:34 kinetic joined #salt
09:39 babilen pi3r: Well, that would explain why you get back an empty dictionary.
09:40 pi3r $babilen: Yes but anyhow I can see that I won't get a retcode ... I am in an orchestration script, the install of the pkg work for some minions, not for the other, so I really need to know what's up.
09:40 babilen Well, that return value is alright at that point
09:40 N-Mi joined #salt
09:41 babilen I guess that you'll have to catch the CommandExecutionError and print it (comment in the return)
09:41 babilen Sorry, this is getting quite deep into code that I haven't memorised yet ;)
09:42 pi3r babilen: ok I will try that. Thx for your help. I initially thought that __salt__['whatever'] would always return the same dict shape.
09:43 babilen All the best
09:44 babilen I'd take a closer look at the pkg.installed *state* and the way it uses pkg.install. I'd say that if you mimic that you are on a good way
09:44 pi3r babilen: I am used to Haskell so this kind of lose way of returning surprises me a bit ;-)
09:46 keimlink joined #salt
09:47 kinetic joined #salt
09:49 breakingmatter joined #salt
09:51 babilen pi3r: Well, I wouldn't necessarily consider it good style in Python either.
09:52 babilen But then: __salt__['....'] is essentially a dictionary of all execution functions salt uses. They naturally differ. You are *not* calling __salt__[...] as a single function with a defined and stable return value, but you are using that collection to gain access to the entirety of execution function in a unified way.
09:53 babilen The functions themselves naturally differ in what they do and return. This is just like functions in haskell.
09:53 brent3 joined #salt
09:54 babilen This is due to excessive, hard to track and easy to get wrong monkey patching of functions into the __salt__ dictionary. The normal way would be to ensure that functions can be defined completely when a module is imported and then to import those modules and call the functions directly.
09:55 babilen I would have even preferred an explicit import salt.foo  ... foo.run_execution_function(...) way, rather than rely on __salt__ (and other dunder dicts) being monkey patched into the modules later on
09:59 Ztyx joined #salt
10:05 pressureman joined #salt
10:06 pi3r babilen: Yep I agree wholeheartedly ;-)
10:07 kinetic joined #salt
10:10 pi3r babilen: as a joke the interface is actually {trymeandfigureout}=__salt__['youtellme']({hellodict}) ;-)
10:20 kinetic joined #salt
10:21 linjan joined #salt
10:22 elsmo joined #salt
10:22 honestly is there a simple way to produce a file that remains on the master?
10:28 sgargan joined #salt
10:35 jaybocc2 joined #salt
10:36 kinetic joined #salt
10:38 larsfronius joined #salt
10:40 Roee does someone has any experience with mongodb.find commands ?
10:42 kinetic joined #salt
10:43 losh joined #salt
10:43 Roee @babilen ?
10:43 giantlock joined #salt
10:47 av___ joined #salt
10:47 ziro` joined #salt
10:53 pfallenop joined #salt
10:54 Ztyx joined #salt
10:57 Norrland joined #salt
10:59 Norrland Is it possible to structure states without any init.sls in it? mytools/tool1.sls, tool2.sls and so on? Or would that break stuff?
11:00 linjan joined #salt
11:03 VSpike Norrland: certainly is. In that case the state's paths would be mytools.tool1 and mytools.tool2
11:04 VSpike Norrland: if you had mytools/init.sls then the state path would just be mytools .. in other words, if you refer to a directory it will look for the init.sls in it. Otherwise, you can just refer to files.
11:07 Norrland VSpike: ah great. I want to group them due to their similarity/use case, but I don't necessarily need all of the tools on every minion.
11:10 _aboe joined #salt
11:13 bluenemo joined #salt
11:14 edulix joined #salt
11:16 kinetic joined #salt
11:19 lb joined #salt
11:23 jaybocc2 joined #salt
11:32 VSpike Norrland: yeah, that's a good use case. Maybe have an init.sls that just includes all the other files in the directory too for convenience
11:34 VSpike Norrland: note you can do this in the init.sls (and elsewhere) https://bpaste.net/show/00b29287e61d
11:34 kinetic joined #salt
11:35 ziro` joined #salt
11:37 VSpike That relative include path doesn't work in pillars though, for some reason
11:37 VSpike Also there's a gotcha if you use require with SLS files. https://bpaste.net/show/6c2f782baf7e
11:38 Norrland VSpike: okey.
11:38 VSpike Both of those are probably bugs, TBH :)
11:38 Norrland :)
11:41 Mohan_ joined #salt
11:42 breakingmatter joined #salt
11:43 Mohan_ where to get the scripts to deploy a application in IIs
11:44 anmolb joined #salt
11:45 amcorreia joined #salt
11:48 slicer joined #salt
11:50 kinetic joined #salt
11:50 tkharju joined #salt
12:00 izibi joined #salt
12:00 babilen Mohan_: IIs as in the webserver?
12:00 babilen (IIS that is)
12:02 Mohan_ I am trying to deploy an web application in IIS using the salt script ,where can I get those script?
12:03 kinetic joined #salt
12:03 Ztyx joined #salt
12:03 izibi joined #salt
12:04 babilen Are you sure that one exists?
12:06 Ztyx joined #salt
12:06 babilen There are well supported formulas for nginx, apache or lighttpd, but IIS is a bit of an exotic choice in this environment. I found https://github.com/saltstack/salt-contrib/blob/master/modules/iis.py but have no idea how much that'll help. That being said, you might be able to write your own states/formula simply by using cmd.run, file.* and other standard modules.
12:06 Mohan_ I cant get you
12:08 Mohan_ I am new to this, did .sls files support python ?
12:08 tmclaugh[work] joined #salt
12:09 babilen Forget about that module, it is simply the only thing I found. You could place it in _modules and use it in states though.
12:10 babilen Normally you'd copy files somewhere, run some commands and similar things to configure an "application" on a webserver. You can do that with the file.* https://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html states, then run a few commands with https://docs.saltstack.com/en/latest/ref/states/all/salt.states.cmd.html and similar things.
12:10 babilen I would never dream of running a webserver on Windows though nor am I familiar with IIS so I can't really help you
12:11 babilen What would you do to configure IIS without saltstack though?
12:13 kinetic joined #salt
12:13 falenn joined #salt
12:14 Mohan_ thanks for your help babilen, I am trying to automate deployment application using the saltstack
12:20 babilen Mohan_: Sure, I can't really help you with IIS. If you could tell us what you would do *manually*, we could see how you'd do the same in saltstack. (e.g. "copy files from A to B, run command "foo", ...)
12:24 stephanbuys joined #salt
12:26 zwi joined #salt
12:27 kinetic joined #salt
12:29 Mohan_ I need to create the following things 1.enable IIS 2.creating an app pool 3.creating an web site, 4.creating an application
12:34 Norrland Mohan_: are you able to trigger these events by manually running scripts on the target server?
12:35 Norrland Mohan_: 1, sounds like "just enabling" the IIS service. 2 & 3. No idea, perhaps calling some command with proper arguments. 4, same as 2&3
12:36 Ztyx1 joined #salt
12:36 Mohan_ yes, I done that using the chef commands
12:37 babilen So you have a set of chef commands and want to translate them into salt states?
12:37 kinetic joined #salt
12:37 Mohan_ absolutely
12:38 Mohan_ you have any idea?
12:40 aqua^c joined #salt
12:40 babilen Sure, start with the first: "How do I install IIS and enable its service?" -- For that read https://docs.saltstack.com/en/latest/topics/windows/windows-package-manager.html and https://docs.saltstack.com/en/latest/ref/states/all/salt.states.service.html#salt.states.service.running
12:45 Mohan_ I already enabled the IIS and its services now I am wondering how to create website and application in it.
12:47 kinetic joined #salt
12:50 Ztyx1 left #salt
12:50 babilen You tell us what you do and we can tell you how you do that with Saltstack
12:53 stephanbuys1 joined #salt
12:53 TC joined #salt
12:55 subsignal joined #salt
12:55 jdesilet joined #salt
12:57 kinetic joined #salt
12:57 giantlock joined #salt
12:58 Mohan_ 1.first creating an app pool 2.creating an web site, 3.creating an application on IIS this are things I need to do
12:58 tru_tru joined #salt
13:00 babilen How do you do that?
13:01 favadi joined #salt
13:01 Mohan_ https://technet.microsoft.com/en-us/library/cc731784(v=ws.10).aspx please refer this
13:01 pfallenop joined #salt
13:02 brandhauser joined #salt
13:03 ViciousLove joined #salt
13:03 numkem joined #salt
13:04 anmolb joined #salt
13:05 brandhauser Hi everybody, noob question: is it possible to overwrite a list assigned by pillar instead of the 2 lists being merged?
13:05 JDiPierro joined #salt
13:08 malinoff_ joined #salt
13:08 eliasp brandhauser: v
13:08 eliasp brandhauser: https://docs.saltstack.com/en/latest/ref/configuration/master.html#pillar-source-merging-strategy
13:09 brandhauser eliasp: Thank you very much!
13:11 brandhauser eliasp: if you have a btc address, i'll send you a beer ;-)
13:13 bhosmer joined #salt
13:16 jaybocc2 joined #salt
13:18 eliasp brandhauser: no BTC :) yw
13:21 brandhauser eliasp: ack
13:21 armguy joined #salt
13:25 brandhauser Ok, i got it working now. But having some strange output. I have now the proper list showing up. But i am seeing it twice. Am i doing something wrong? http://pastebin.com/75v5MiZw
13:28 brandhauser (So this is with the setting "pillar_source_merging_strategy: overwrite" in /etc/salt/master)
13:29 trph joined #salt
13:29 Norbell joined #salt
13:30 pfallenop joined #salt
13:30 pfallenop joined #salt
13:30 kant joined #salt
13:30 _JZ_ joined #salt
13:30 MadHatter42 joined #salt
13:31 pm90_ joined #salt
13:31 kinetic joined #salt
13:32 jaybocc2 joined #salt
13:35 bhosmer joined #salt
13:36 aparsons joined #salt
13:37 pfallenop joined #salt
13:37 giantlock joined #salt
13:38 brandhauser joined #salt
13:38 brandhauser Whoopsy, power failure :). But i got it working
13:38 peters-tx joined #salt
13:39 elfixit joined #salt
13:39 RandyT Greetings, could some kind soul give me some hints as to how to test S3fs access through salt tools?
13:39 mapu joined #salt
13:39 RandyT I am able to test this use awscli, but would appreciate some guidance on how to test it at the command line with salt.
13:41 teebes joined #salt
13:42 TooLmaN joined #salt
13:45 kinetic joined #salt
13:49 ashirogl joined #salt
13:52 kaptk2 joined #salt
13:53 DanyC joined #salt
13:53 DanyC hi all, from a salt master what is the easiest way to find out the minions version ?
13:55 zmalone joined #salt
13:55 teryx510 joined #salt
13:55 Norrland DanyC: I'd grab the grain 'saltversion'
13:56 Norrland DanyC: salt 'minion' grains.get
13:56 ferbla joined #salt
13:57 VSpike DanyC: also salt-run manage.versions
13:58 DanyC Norrland: VSpike thanks!
13:58 VSpike or salt 'minion' test.version (or even test.versions for more output)
13:59 ashirogl1 joined #salt
14:00 kinetic joined #salt
14:02 ambientsound does salt have any mechanism for automatically testing salt code on a fresh VM?
14:02 ambientsound e.g. vagrant or openstack
14:03 ambientsound preferably the latter.
14:03 timoguin_ joined #salt
14:03 JDiPierro Is there any known issues around the salt mine for 2014.7.6? I've deleted hosts, cleared them from salt-key, used the cache.clear_all runner and the mine.flush module but the mine is still returning info about nonexistant minions.
14:03 DanyC ambientsound: i doubt, everything i've seen people mentioning here is using vagrant
14:03 sunkist joined #salt
14:04 pm90__ joined #salt
14:06 ageorgop joined #salt
14:06 brandhauser joined #salt
14:07 kinetic joined #salt
14:07 brandhauser Hi all, is there a way to ignore the error that a package is already installed with pkg.installed: <CR> - pkgs: <CR> [list of packages]?
14:07 JDiPierro ambientsound: DanyC: Yeah, vagrant is your best friend for developing with salt.
14:08 DanyC JDiPierro: on this topic i'm curious how oyu guys are testing salt-cloud using vagrant/ virtualbox ?
14:08 DanyC *you
14:09 JDiPierro DanyC: Can't speak to that :( I haven't found a great way. I've been wanting to make a virtualbox provider for salt-cloud but haven't gotten around to it. I recall seeing in an issue that someone else was toying with that though.
14:11 I joined #salt
14:11 zwi joined #salt
14:11 DanyC JDiPierro: ah i see ;) i'll dig on the issue list then... cause i started to play with salt-cloud and i do have a urgent need to test my stuff, even installing minion while the VMs have no internet access. Having a handy dev env will make life easier ;)
14:12 N-Mi joined #salt
14:12 elsmo joined #salt
14:13 winsalt joined #salt
14:13 stephanbuys joined #salt
14:15 kinetic joined #salt
14:15 mpanetta joined #salt
14:15 DanyC VSpike: fyi salt 'minion' test.versions is not working as i see the output saying "not available". I'm on 2015.5.5
14:22 JDiPierro How can I reliably clear the salt mine?
14:23 ambientsound hm, thanks.
14:23 MadHatter42 joined #salt
14:24 anotherZero joined #salt
14:24 impi joined #salt
14:25 kinetic joined #salt
14:25 sunkist joined #salt
14:25 babilen JDiPierro: https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.mine.html#salt.modules.mine.flush
14:26 quasiben joined #salt
14:28 larsfronius joined #salt
14:29 JDiPierro babilen: I've called that so many times D: and the cache.clear_mine runner. I've removed the minions from salt-key.. they're still being returned in mine calls though.
14:29 stephanbuys left #salt
14:29 kinetic joined #salt
14:31 babilen JDiPierro: So you do know how it should work, but you are having a problem with it!
14:31 VSpike DanyC: that might be a 2015.8 thing. I only only found it just now by a typo :)
14:32 JDiPierro babilen: Seems so :( Thanks for the help. I'll see if I can reliably recreate it and open a ticket.
14:32 babilen JDiPierro: My understanding is that this should reliably clear the mine
14:32 babilen (if it doesn't then we are either mistaken or salt is buggy)
14:33 DanyC VSpike: heh, thanks :)
14:35 edulix joined #salt
14:35 Akhter joined #salt
14:36 quasiben joined #salt
14:37 Akhter joined #salt
14:37 Ahlee anybody know how to auto-roll the minion log file on windows minions?
14:41 Norbell joined #salt
14:41 openfly oddball salt question
14:42 polishdub joined #salt
14:42 openfly if I am doing a cmd.run, is there a way to return a value that will show up as a 'clean' state in salt?
14:43 kinetic joined #salt
14:43 Phil-Work openfly, you'll need to selectively run the command
14:44 Gareth o/
14:44 Phil-Work with an "unless", that is ^^
14:45 Nazca__ joined #salt
14:47 pfallenop joined #salt
14:47 pfallenop joined #salt
14:50 beardedeagle joined #salt
14:50 dkrae joined #salt
14:53 openfly Phil-Work unless eh?
14:53 openfly nah
14:53 openfly that can't work
14:53 openfly literal use of can't.
14:53 openfly =/
14:55 kinetic joined #salt
14:55 JDiPierro openfly: If a file is created by the cmd.run you can use - creates: /path/to/file. 'unless' is for what you're asking about though.
14:56 openfly i get what you are saying
14:56 openfly but that's not the game
14:56 Ahlee Move the work you're doing with cmd.run into a _module, and you'll need to call the module and have it populate hte return without a changes section to make it 'clean'
14:56 openfly i'm basically executing a script that does a bunch of dependency related plugin installs
14:56 openfly i mean i can live with it not being clean when it makes no changes.
14:57 openfly but it's kinda ugly
14:57 openfly salt should probably have like a return code for a clean state.
14:57 Ahlee God ain't that the truth
14:57 JDiPierro So if that script isn't going to do anything if it's job is already done can't you do an unless that runs the same script and greps for some output like "already installed" or something?
14:58 openfly i could but frankly i don't trust unless to run interatively inside of a jinja loop
14:58 openfly s/inter/iter
14:58 DammitJim joined #salt
14:58 Ahlee do you delete python objects before returning, too?
14:58 JDiPierro The cmd docs mention a "stateful" flag that might help here? I haven't used it and don't totally understand how to use it from just reading the dogs, but it's here: https://docs.saltstack.com/en/latest/ref/states/all/salt.states.cmd.html
14:59 JDiPierro s/dogs/docs/ lol
14:59 JDiPierro I need coffee
14:59 openfly give it a read
15:01 kinetic joined #salt
15:06 linjan joined #salt
15:07 cpowell joined #salt
15:07 _mel_ joined #salt
15:08 clintberry joined #salt
15:08 zmalone joined #salt
15:10 ziro` joined #salt
15:10 kinetic joined #salt
15:10 hasues joined #salt
15:11 pm90_ joined #salt
15:12 hasues left #salt
15:13 pm90__ joined #salt
15:17 overyander joined #salt
15:18 toastedpenguin joined #salt
15:18 RandyT I'm seeing the following in my logs as I try to get s3fs working: https://gist.github.com/rterbush/f1de923c27ba909a82e1
15:19 chiui joined #salt
15:19 RandyT I am able to list and copy from this bucket using awscli executed on the master.
15:19 RandyT What salt commands allow me to debug this a bit more from the command line?
15:23 aqua^c joined #salt
15:24 kaptk2 joined #salt
15:24 cpattonj joined #salt
15:25 cpattonj where is the documentation for debugging the salt-minion on a windows host?
15:25 zmalone RandyT: That sounds like a AWS permissions error, I'd go download s3cmd, set it up with the same API keys, and begin troubleshooting your per-user S3 ACLs
15:26 zmalone Once you are able to put and get files from S3 like that, I'd move back to Salt
15:27 RandyT zmalone: thanks, as I mention above, I've verified that I can list and copy files from this bucket using awscli from the same machine that the master is running on.
15:27 RandyT now looking for some hints as to how to execute salt commands to attempt to access the same resource.
15:28 Gareth o//w 62
15:28 Gareth erm
15:29 cpattonj my windows minion isn't returning a response when I try to apply states but is returning a response to ping
15:31 ziro` joined #salt
15:31 kinetic joined #salt
15:31 DanyC all, on salt-cloud openstack provider - any idea what delete_sshkeys: True does as is not in doc? (i can assume but if anyone can point me to the doc will be great)
15:32 RandyT zmalone: two things unique to my bucket. 1. I am using iam profiles to control access to the resource and 2. the contents of the bucket are using kms to encrypt
15:32 RandyT as I mentioned though, these things all work using the awscli
15:32 zmalone That's a straight http error, so I'd really look closely at whether your awscli keys match the keys that salt is using, or if salt is using keys at all right now
15:32 zmalone (or if it's just attempting to pull as though the buckets are public)
15:33 RandyT not using keys at all. I am using iam profile
15:33 RandyT key info is pulled from instance metadata
15:34 RandyT and I know this works with other salt actions and comments in git issues suggest that support for iam profiles works.
15:34 kinetic joined #salt
15:36 RandyT zmalone: just as an example https://github.com/saltstack/salt/pull/25767
15:36 saltstackbot [#25767]title: Add AWS KMS support to S3 |
15:37 zmalone Sure enough, or https://docs.saltstack.com/en/latest/ref/file_server/all/salt.fileserver.s3fs.html
15:37 RandyT zmalone: any hint as to what salt command I can execute to try to list an s3fs from the command line?
15:38 clintberry joined #salt
15:38 deus_ex joined #salt
15:38 RandyT I see the api calls for dir_list(), etc. How do I execute these from the command line?
15:38 zmalone I'm sorry, I have not used KMS before.
15:38 RandyT salt-run?
15:39 RandyT salt-call?
15:40 RandyT not convinced this is a KMS issue. I would expect to get either garbage back or some decryption error if it is KMS.
15:41 RandyT and have tried the same config against a bucket that has no contents and is not encrypted.
15:41 tomspur joined #salt
15:44 scoates joined #salt
15:44 wendall911 joined #salt
15:45 kaptk2 joined #salt
15:46 aparsons joined #salt
15:47 beardedeagle cpattonj: https://docs.saltstack.com/en/develop/topics/troubleshooting/master.html#commands-time-out-or-do-not-return-output
15:48 pfallenop joined #salt
15:48 pfallenop joined #salt
15:49 pravka joined #salt
15:51 tercenya joined #salt
15:52 RandyT zmalone: some progress, this at least replicates the error at the command line, salt-run -l debug fileserver.dir_list
15:53 RandyT sometimes it returns the error, other times it does not
15:54 Cavlim joined #salt
15:54 Cavlim left #salt
15:55 morsik joined #salt
15:55 morsik hi, i'm trying to generate file using Jinja + YAMLex as context
15:55 morsik the problem is: doing {{ some_var }} causes to print that variable inside double quotes.
15:55 morsik always.
15:56 morsik why? and how can i prevent this behavior?
15:57 morsik doing {{ show_full_context() }}  prints everything without double quotes… :<
16:01 morsik doing {{ some_var|yaml }} causes to get this:  !!python/object/new:salt.utils.serializers.yamlex.SLSString [my value]
16:01 moogyver RandyT: dir_list is part of the runner fileserver, so you'd call it via 'salt-run fileserver.dir_list'
16:04 writtenoff joined #salt
16:08 morsik maybe this is problem…? https://github.com/saltstack/salt/blob/develop/salt/serializers/yamlex.py#L357,L363
16:08 morsik but… why? O.o
16:10 dyasny joined #salt
16:11 jaybocc2 joined #salt
16:13 stanchan joined #salt
16:14 v0rtex joined #salt
16:15 kinetic joined #salt
16:16 dyasny_ joined #salt
16:17 dyasny__ joined #salt
16:18 edrocks joined #salt
16:18 Pouldenton joined #salt
16:18 breakingmatter joined #salt
16:20 moogyver morsik: maybe post a gist of the template?
16:20 kukacz joined #salt
16:22 morsik moogyver: https://dpaste.de/4i84
16:22 CeBe joined #salt
16:22 jalbretsen joined #salt
16:22 morsik and i'm getting:    host_name "somehost"      alias "somehost"    address "127.0.0.1"
16:22 morsik everywhere things are quoted…
16:22 morsik looks like because of lines above ↑
16:23 scoates joined #salt
16:23 JDiPierro joined #salt
16:25 Pouldenton Hello! Tell me please, is there way to render salt templates from CLI before it will be assigned to the minion? For example, I am writing Salt state for ntp and want to see how exactly generated config file (ntp.conf) will be looking at the custom minion after deploy.
16:26 morsik Pouldenton: i use docker for that with localfiles, and set hostname of that docker to specified host :P
16:27 pravka joined #salt
16:27 moogyver morsik: perhaps your data is quoted then?  using jinja2, seems to work fine for me? https://dpaste.de/t0Uj#L26
16:27 whytewolf Pouldenton: I just use this https://github.com/whytewolf/salt-debug
16:28 morsik moogyver: i'm using yamlex…
16:28 moogyver ah didn't see that
16:28 moogyver it's early, i just read 'YAML' :)
16:28 morsik and yamlex adds this. see link to GH i posted above ↑
16:28 morsik moogyver: for me it's 18:28 :P
16:28 moogyver 9:30 AM here
16:29 morsik looks like there's no way to use yamlex for template rendering… :<
16:31 malinoff_ joined #salt
16:33 Pouldenton Wow, I thought this will be something like "salt "*" render <path to file> ..." very close to whytewolf method. Thanks a lot!
16:34 morsik Pouldenton: quite impossible, cause templates are rendered on client-side.
16:34 morsik Pouldenton: also, templates depends on grains (which are inaccessible on master)
16:34 Azulez Hey guys -- just kind of getting started with Salt.  Is there any advantage to using Vagrant on top of Salt, or is that kind of unnecessary for most use cases?
16:34 Azulez Specifically I'm looking to deploy environments, configure them, and execute stuff on them.
16:35 Pouldenton @morsik Ok, It can be salt-call instead salt...
16:35 sgargan joined #salt
16:35 whytewolf Pouldenton: that is why i wrote that. because of all the client side needed info for rendering jinja. it is just an added module.
16:36 morsik Pouldenton: i'm using docker for testing, so I can just modify that file anytime and it doesn't hurt me :P
16:36 whytewolf it uses the built in salt render engine
16:36 morsik Pouldenton: (btw remember you need always minimum 4 ntp servers :>)
16:36 nkuttler_ joined #salt
16:37 morsik eh… so last hope to get this working is use python instead of yaml. damn.
16:38 fridder joined #salt
16:38 whytewolf Pouldenton: also with that module it doesn't need to be setup to run. it just needs to be in the salt:// directory tree.
16:38 pm90_ joined #salt
16:39 jaybocc2 joined #salt
16:39 hacfi_ joined #salt
16:39 nobrak_ joined #salt
16:39 garphyx joined #salt
16:39 canci_ joined #salt
16:40 sirtaj_ joined #salt
16:40 murkey_ joined #salt
16:40 beardedeagle Azulez: I use salt to build out my environments via salt cloud/openstack. salt-sandbox could be used as an example for what you are doing: https://github.com/elasticdog/salt-sandbox
16:40 ]V[_ joined #salt
16:40 the_lale1u joined #salt
16:41 beardedeagle Azulez: or more specifically, a kind of template of what you are wanting to accomplish.
16:41 Azulez beardedeagle: Yeah, this looks like it could be a great reference
16:41 pm90__ joined #salt
16:41 Azulez thanks!
16:41 Akhter joined #salt
16:42 beardedeagle Azulez: np, glad to help.
16:43 teryx510 Hey guys, for windows minions, is there a minion_id file?
16:43 Bryson joined #salt
16:43 etw_ joined #salt
16:43 brucewang joined #salt
16:44 ageorgop joined #salt
16:44 Pouldenton whytewolf: Yes, it's great, It souns like a simplest variant. I will try it tomorrow, thanks
16:44 mrtrosen_ joined #salt
16:44 Karunamon|2 joined #salt
16:44 ze-_ joined #salt
16:44 emostar_ joined #salt
16:45 JoeHazzers_ joined #salt
16:45 jbub_ joined #salt
16:47 teryx510 Nevermind, looks like the salt-minion process didn't actually start.
16:48 UForgotten_ joined #salt
16:48 THE_BOULDER joined #salt
16:49 mapu joined #salt
16:50 indispeq_ joined #salt
16:50 Vye joined #salt
16:50 giantlock joined #salt
16:52 larsfronius joined #salt
16:53 moogyver morsik: how are you trying to generate that file?
16:54 mlanner joined #salt
16:55 kiorky joined #salt
16:55 nate_c joined #salt
16:58 Brew joined #salt
16:59 rawzone joined #salt
16:59 Ryan_Lan1 joined #salt
16:59 aparsons joined #salt
16:59 overyander joined #salt
16:59 cpowell joined #salt
16:59 keimlink joined #salt
16:59 flebel joined #salt
16:59 shnguyen_ joined #salt
16:59 NV joined #salt
16:59 garthk joined #salt
16:59 Zachary_DuBois joined #salt
16:59 jacksontj joined #salt
16:59 stupidnic joined #salt
16:59 voileux joined #salt
16:59 ablemann joined #salt
16:59 baffle joined #salt
16:59 `chris joined #salt
16:59 Reinhart_ joined #salt
16:59 eightyeight joined #salt
16:59 blueyed joined #salt
16:59 muep joined #salt
16:59 sybix joined #salt
16:59 paha joined #salt
17:01 amcorreia joined #salt
17:03 zer0def joined #salt
17:03 jmreicha joined #salt
17:04 ashirogl joined #salt
17:04 Striki joined #salt
17:05 riftman joined #salt
17:06 XqueZme joined #salt
17:06 ShibMcNe joined #salt
17:06 robawt joined #salt
17:06 nahamu joined #salt
17:06 notnotpeter joined #salt
17:06 tzero joined #salt
17:07 itsamenathan joined #salt
17:07 darix joined #salt
17:07 Riz joined #salt
17:07 cpattonj joined #salt
17:07 dandelo joined #salt
17:07 rnts joined #salt
17:07 Hipikat joined #salt
17:07 zemm joined #salt
17:07 utahcon joined #salt
17:07 daemonkeeper joined #salt
17:07 chamunks joined #salt
17:07 Norrland joined #salt
17:07 mullein joined #salt
17:07 hypnosb joined #salt
17:07 seblu joined #salt
17:07 titilambert joined #salt
17:07 toddnni joined #salt
17:07 papertigers joined #salt
17:07 aRkadeFR joined #salt
17:07 danemacmillan joined #salt
17:07 XqueZme Hi everybody, just started with salt and going through tutorial and pillars. If I issue pillar.items it seems to return none, but if I run it again it returns pillars. Anyone can help me out?
17:07 cwyse joined #salt
17:07 CaptainMagnus joined #salt
17:07 davidbanham joined #salt
17:07 xaviern joined #salt
17:07 dec joined #salt
17:08 breakingmatter joined #salt
17:08 monokrome joined #salt
17:08 a7p joined #salt
17:08 LeProvokateur joined #salt
17:08 iamtew joined #salt
17:08 hypnosb joined #salt
17:08 karlthane joined #salt
17:09 kalessin joined #salt
17:09 andi- joined #salt
17:10 terinjokes joined #salt
17:10 bregalad joined #salt
17:10 phx joined #salt
17:10 froztbyte joined #salt
17:10 notnotpeter joined #salt
17:10 rawzone joined #salt
17:10 Ryan_Lan1 joined #salt
17:10 aparsons joined #salt
17:10 overyander joined #salt
17:10 cpowell joined #salt
17:10 keimlink joined #salt
17:10 flebel joined #salt
17:10 shnguyen_ joined #salt
17:10 NV joined #salt
17:10 garthk joined #salt
17:10 Zachary_DuBois joined #salt
17:10 jacksontj joined #salt
17:10 stupidnic joined #salt
17:10 voileux joined #salt
17:10 ablemann joined #salt
17:10 baffle joined #salt
17:10 `chris joined #salt
17:10 Reinhart_ joined #salt
17:10 eightyeight joined #salt
17:10 blueyed joined #salt
17:10 muep joined #salt
17:10 sybix joined #salt
17:10 paha joined #salt
17:10 terinjokes joined #salt
17:11 UtahDave joined #salt
17:11 aqua^c joined #salt
17:12 daschatten joined #salt
17:12 Edur joined #salt
17:12 fbettag joined #salt
17:12 jhujhiti joined #salt
17:12 rofl____ joined #salt
17:12 SneakyPhil joined #salt
17:12 nihe joined #salt
17:14 Akhter joined #salt
17:14 ronny joined #salt
17:14 ronny hi
17:15 ronny is there a process for writing tests for salt declarations?
17:16 ronny also is there a process to run salt rosters from a certain folder, i dont want it in etc, i want it in a git repo
17:17 cpattonj_ joined #salt
17:20 armguy joined #salt
17:22 Akhter joined #salt
17:22 kinetic joined #salt
17:23 iggy symlink?
17:24 robawt joined #salt
17:25 whidbeywalker joined #salt
17:28 PredatorVI joined #salt
17:32 joshin joined #salt
17:32 joshin joined #salt
17:36 kinetic joined #salt
17:37 PredatorVI I'm having problems using salt-api for state.orchestrate runner ... The following gist has the salt-run command-line that works along with the CURL call that doesnt:  https://gist.github.com/PredatorVI/5397a9c184c8f4bfa433
17:38 PredatorVI The gist has the error and it seems to be targeting a minion randomly
17:39 PredatorVI I'm having a hard time finding examples for this scenario
17:41 PredatorVI Do I need to pass in a target even though the orchestration state uses the pillar data for targeting?
17:42 baweaver joined #salt
17:43 baweaver joined #salt
17:44 moogyver PredatorVI: sec, looking
17:45 Aleks3Y joined #salt
17:45 stupidnic I am wondering if somebody can help me with salt.state.network. I am trying to do a bond0 (balance-rr) type and it says I need arp_ip_target, but if I put it as a text string it doesn't like it, but if I put it as a list the formatting (I think) is wrong in the interfaces file on the minion
17:46 Akhter joined #salt
17:46 Brew joined #salt
17:47 ferbla joined #salt
17:48 kinetic joined #salt
17:49 PredatorVI @moogyver  ty
17:51 winsalt joined #salt
17:52 darvon joined #salt
17:52 dabb joined #salt
17:52 eichiro joined #salt
17:53 mapu joined #salt
17:53 kinetic joined #salt
17:56 aboe joined #salt
17:58 Akhter joined #salt
18:04 Gareth Is there something that one has to do before a custom runner is available?
18:04 iggy sync_all?
18:04 pfallenop joined #salt
18:04 Gareth did that.
18:05 Gareth doesn't mention runners.
18:05 iggy set runner_dirs?
18:06 iggy yeah, thinking about it, sync_all sync's stuff to the minions, runners run in the master only
18:06 Gareth Another custom runner is available in the same path.
18:07 iggy reboot
18:07 * iggy runs
18:07 linjan joined #salt
18:10 Ryan_Lane joined #salt
18:10 geekatcmu your master will need to highstate
18:10 geekatcmu And that highstate has to copy the runners out of your state tree into "the correct location".
18:11 geekatcmu I have no idea if that's part of the standard master highstate or not
18:11 katyucha joined #salt
18:13 baweaver joined #salt
18:13 Gareth nm.  looks like someone put the custom runner that is working in with the built in runners.
18:13 Fiber^ joined #salt
18:15 iggy is that someone with the initials GG?
18:15 baweaver joined #salt
18:16 tkharju joined #salt
18:17 Gareth Definitely not.  This was done BG (Before Gareth) :)
18:22 ajw0100 joined #salt
18:23 bcrockett joined #salt
18:26 bcrockett Greetings. I'm having a bit of trouble with salt minions pulling the correct 'environment'. I'm trying to combine two salt configurations into one environment (not by choice), but need to keep the logic of both for now. One is driven by a top.sls using the example '*', the other does some very interesting stuff with grains. The issue i'm having is it seems once it 'hits' the first environment; it no longer cares to proceed to the 2nd
18:26 bcrockett When I force either config A, or config B however; they both work flawlessly
18:26 moogyver PredatorVI: sorry, had to present for a meeting
18:29 iggy bcrockett: sounds like expected behavior (and yes it sucks)
18:30 bcrockett hmm
18:31 bcrockett I've even tried making two different masters so it can ask each one individually "hey,what should I do?"
18:31 bcrockett but again, it only stops at one
18:31 iggy minions can't talk to 2 masters
18:31 tanta they can't serve 2 masters either
18:32 bcrockett I expected as much
18:33 iggy this is why I avoid environments like the plague
18:33 bcrockett What do you do instead? I thought that was at the heart of salt logic
18:34 bcrockett I *can* change my side very easily if there is a better (more compatible) way
18:34 iggy I... don't use environments
18:35 iggy not sure what you're asking
18:35 bcrockett Well, as simply as "I want this server to be a webserver"
18:36 bcrockett I'm using this method currently: https://docs.saltstack.com/en/latest/ref/states/top.html#a-basic-example
18:37 babilen That has nothing to do with environments. You simply target specific states to it that *make* that box a webserver. You can use the entirety of https://docs.saltstack.com/en/develop/topics/targeting/index.html for that, but "minion id" and "role in pillar (or grains if you have to)" approaches are common
18:37 iggy in the 2 major Salt deployments I've worked in, there's been an external store that was a source of truth for "host1 is web, host2 is db, etc"
18:37 babilen bcrockett: The "web" in that example is not an environment, but a target expression in the "base" environment.
18:37 iggy one used cloud metadata (as an example)
18:38 pm90_ joined #salt
18:38 bcrockett Okay, so it is entirely possible that I'm only really using the targeting whereas I thought that was considered an environment
18:39 bcrockett I'm still getting my feet wet with the language here
18:39 iggy and a host can match multiple targets
18:39 babilen bcrockett: Environments are used to essentially divide minions into different groups that correspond to stages in your environment. It's the "prod", "dev" and "qa" bits later on.
18:39 babilen "stages in your development"
18:39 babilen *sigh*
18:39 cyborg-one joined #salt
18:40 bcrockett okay, then that's what they are doing on the *other* side of this equation
18:40 iggy bcrockett: in that basic example, 'base' is the env, 'web*' is the match (hostname match in this case), apache is the state to target at 'web*'
18:40 babilen Could you paste the respective top.sls files (or suitable examples) so that we can take a look and you could explain what problematic behaviour you see.
18:41 iggy ^
18:41 bcrockett sure
18:41 Akhter joined #salt
18:41 babilen Please use one of http://refheap.com, http://paste.debian.net, https://gist.github.com, http://sprunge.us, http://dpaste.de, … or any ¬ pastebin.com one if you want me to be happy
18:42 bcrockett https://gist.github.com/anonymous/a4ad87cffc4109c8af5c
18:42 bcrockett So that's the 'easy' side
18:42 iggy so you have servers named X-edge-*?
18:43 bcrockett and on it's own, it works beautifully by applying 'common' to everything, and 'common+edge' to the X-edge servers
18:43 babilen Sure, that's what it does
18:43 bcrockett on the other side, it's not using top.sls files, rather using grains on the client
18:43 babilen You *always* use top files. They might match on grains, but that still has to be configured.
18:44 andrew_v joined #salt
18:44 bcrockett https://gist.github.com/anonymous/8e751c058afbc3881e3b
18:45 * iggy gouges his eyes out
18:45 breakingmatter joined #salt
18:45 bcrockett yea - I know
18:46 babilen Yeah, you would still have a top files that matches on those grains... and who on earth came up with those names?
18:46 bcrockett unfortunatly I inhereted this
18:46 babilen *shudder*
18:46 bcrockett well, i'm sanitizing with X
18:46 beardedeagle ow
18:46 babilen "X_-esb-jms" -- wtf?
18:46 beardedeagle that looks painful
18:46 bcrockett that's not REALLY what it says
18:46 babilen Anyway, that doesn't help
18:46 bcrockett :)
18:46 babilen Ah
18:47 babilen Okay, you changed it to those values so they appear uglier than they really are?
18:47 bcrockett something like that
18:47 bcrockett :)
18:47 babilen I feel better now
18:47 bcrockett So the roles: esb-jms line matches to ...../roles/vhx-esb-jms.sls
18:47 iggy commented on 8e
18:48 babilen Either way, you would have a top file that matches on X_-esb-jms and the other grains.
18:48 babilen Or something that creates that top file automagically
18:48 bcrockett Well, all this works perfectly right now when used alone
18:48 babilen In the end you will have logic that says "if grain matches X_-esb-jms then apply states A, B, C and F"
18:49 pravka joined #salt
18:49 iggy thanks for showing us what works?
18:49 bcrockett Well, I'm illustrating the two methods that work independently
18:50 bcrockett but it's combining them, is where I'm having the issue
18:50 babilen Sure, but *something* has to link grains and states together. This is being done in a top file. It has to be somewhere.
18:50 mschiff Does anybody have a good hook to use in git that does yaml/jinja syntax checks?
18:51 iggy there have been various attempts to do a salt-lint like tool, none of which made it very far
18:52 babilen bcrockett: One thing you could do is to set a specific grain "grains_setup: True" on your "grain setup" minions and exclude them from "*" states in the other setup. Something like "* and not G@grain_setup" - match: compound
18:53 moogyver don't need the '* and' anymore ( for 2015.8 )
18:53 babilen bcrockett: There really is little we can do without a more precise and complete paste about the "grain setup".
18:53 babilen moogyver: indeed
18:54 babilen bcrockett: I assume that the problem is that you don't want "- common" on *all* boxes in the "grain setup" ?
18:54 babilen Or .. let me rephrase that: What is the actual problem you are trying to solve?
18:54 babilen (provide pastes as you see fit)
18:55 iggy but it doesn't hurt to keep using it (and since I haven't heard of people upgrading to 2015.8 in droves...)
18:55 bcrockett it's quite the inverse. I want 'common' everywhere, and then whatever it should be doing with the grain method (the -esb-jms bit)
18:55 babilen You would have that.
18:55 iggy if that's not working, show an example that reproduces it
18:55 babilen - common is being targeted to all minions
18:56 iggy it's likely that you have an error somewhere (rather than Salt having a bug that would have hopefully been caught in such a oft used area of the code... the top parser)
18:56 babilen The output of "salt 'someminion' state.show_top" (with 'someminion' being the minion that shows wrong behaviour) would also be helpful
18:57 bcrockett thanks - that's something that I can actually get
18:57 bcrockett the 2nd method is so difficult to read, i'm trying to get you something to see -- but don't know enough to show
18:58 PredatorVI moogyver:  no worries...just got back from lunch.  I noticed that I think I had the wrong key 'args' for the pillar instead of 'pillar', but now I get an error stating that pillar data is not a dictionary.  I tried changing the 'Accept' type to 'application/json' but that didn't help.
18:59 bcrockett It's going to take me a bit to undo the master/master madness I was trying
19:00 babilen You'd have to do that anyway :)
19:00 bcrockett yep
19:00 bcrockett lol
19:00 babilen iggy: Did you hear of any plans to actually support multiple masters?
19:00 iggy next time ask us before you go to such extremes ;)
19:00 bcrockett well funny enough
19:00 bastiandg joined #salt
19:01 bcrockett I thought I was following https://docs.saltstack.com/en/latest/topics/tutorials/multimaster.html
19:01 iggy babilen: you mean in that fashion? no but afaik, they are working on improvements to failover between masters
19:01 bcrockett but just a slightly different use case
19:01 bcrockett by slightly I mean 'way different'
19:01 iggy bcrockett: multi-master is a thing... just not _that_ thing
19:01 iggy it's failover between masters (theoretically)
19:02 iggy not talking to multiple masters at once
19:02 babilen iggy: I was toying around with ideas on how to organise a way in which guests could be managed by two completely independent parties (both of which want to use salt to do so)
19:02 PredatorVI moogyver: I updated the gist https://gist.github.com/PredatorVI/5397a9c184c8f4bfa433 with the current error.  I'm sure I'm missing something simple.
19:02 bcrockett yea - I even tried the master_shuffle bit to see if I could get it to 'flip flop' between methods
19:02 babilen Which would, essentially, necessitate "that thing" ;)
19:02 iggy babilen: would be nice (I'm sure we'd use it here as well)
19:03 babilen bcrockett: hahahaha
19:03 iggy but I haven't heard anything
19:03 babilen iggy: That's cool. Just wanted to make sure that I'm not missing something that has been around for some time :)
19:03 iggy maybe my talk will get accepted this year and I'll ask the head honcho
19:04 iggy babilen: for now I think that amounts to 2 git repos and lots of testing
19:04 cberndt joined #salt
19:04 iggy or 1 git repo and someone that can crossover teams well enough
19:04 moogyver iggy: I thought multi-master already worked ( 2 masters at once ).  seems to work in my lab setup?
19:05 babilen iggy: That's not an option for us in some cases unfortunately (customers want their own salt, but I'd like to manage some stuff with it too). So far I went down the "salt-ssh" + "salt-master" route.
19:05 iggy moogyver: it does(ish), it just doesn't do what someone thought it would
19:06 iggy babilen: nice idea (salt-ssh+salt-master)
19:06 babilen It's the best I could come up with for that usecase so far
19:06 babilen It works well for what we are doing now, but it wouldn't scale well to *all the things" (salt-ssh that is)
19:06 iggy yeah
19:07 babilen And you miss out on a plethora of master only features
19:07 iggy and it still takes someone on one of the teams that has a good handle on salt's "intricacies" and what the other team is doing
19:07 bVector so I'm doing this on the minion to upload the key before starting the service: https://gist.github.com/bVector/2d5bca4a6150eb77d7d1  is there a way to get the minion to generate its keypair inside the state?
19:07 babilen Sure, so far these two setups have been concerned with mutually exclusive areas
19:08 jeffpatton1971 joined #salt
19:08 jeffpatton1971 hey, I have a question about nodegroups and nestedgroup support in 2015.5.5
19:09 jeffpatton1971 is there any special notation in the master file that I need to do for this?
19:09 jeffpatton1971 group1: 'L@server1.domain.com,server2.domain.com'
19:09 jeffpatton1971 group2: 'N@group1'
19:09 jeffpatton1971 works
19:10 jeffpatton1971 but group2: 'N@group1,group3'
19:10 jeffpatton1971 doesn't
19:10 jeffpatton1971 I tried the N@ infront of each group
19:10 jeffpatton1971 and that made salt very unhappy
19:11 jeffpatton1971 can you have multiple groups a member of one?
19:11 babilen Probably not
19:11 iggy N@group1 and N@group2 ?
19:12 iggy or s/and/or/ not really sure what you're after
19:12 jeffpatton1971 use the word and inbetween?
19:12 jeffpatton1971 haven't tried that
19:13 iggy bVector: wouldn't you rather have the master generate it and have the minion pull it from the bucket?
19:13 babilen Well, "or" is probably what you are after if you want to match both
19:13 timoguin joined #salt
19:15 chiui joined #salt
19:15 PredatorVI Bumping a previous question about using the salt-api to call the salt.orchestrate runner.  This gist shows the command-line that works as well as the CURL call I'm trying to construct to do the equivalent:  https://gist.github.com/PredatorVI/5397a9c184c8f4bfa433
19:15 jeffpatton1971 @iggy N@group1 and N@group2 didn't appear to work :(
19:17 babilen jeffpatton1971: Try "or" if you want to match both (otherwise you would get only those minions that are in group1 and group2)
19:18 jeffpatton1971 @babilen OR si the love! thankS!
19:19 iggy "15:12 < iggy> or s/and/or/ not really sure what you're after"
19:19 babilen indeed
19:19 babilen I want to match "foo" and "bar" so I have to use "or"
19:19 babilen logic is hard
19:20 babilen conjunctive normal form all the things :)
19:24 baweaver joined #salt
19:25 jeffpatton1971 I sometimes just miss where logic like that should actually be used...as this looked more along the lines of a list of things
19:25 jeffpatton1971 but thanks ;)
19:25 cpattonj Anyone know what would cause wildcard salt queries to fail all of a sudden?
19:25 cpattonj Like.. "sudo salt 'something*' test.ping"
19:26 baweaver joined #salt
19:26 DanyC joined #salt
19:26 zmalone How are they failing?
19:27 cpattonj No hosts match the query but when I do a system wide ping like '*' it finds them.
19:27 cpattonj something.somesite.com returns true for the ping but against something* it doesn't work
19:34 babilen Show us you exact commands and their output please
19:35 cpattonj sudo salt '*' test.ping
19:35 cpattonj sudo salt 'something*' test.ping
19:35 cpattonj for the first command, it correctly returns something.somedomain.com and something2.somedomain.com
19:36 cpattonj for the second command: "no minions matched the target, no command was sent.."
19:37 babilen That shouldn't be
19:40 opensource_ninja joined #salt
19:43 bhosmer joined #salt
19:43 PredatorVI cpattonj:  This is a long shot but I've seen unexplainable issues (on ubuntu) if I've upgraded from an older version of salt (e.g., 2014.x.x).  I've had to purge the salt packages, remove all unneeded dependencies (apt-get autoremove) then reinstall salt from scratch.  Something to look at if you have nothing else to try...
19:43 cpattonj that's terrifying
19:44 cpattonj I have a ton of hosts, reinstalling across all sounds daunting
19:44 cpattonj I did upgrade my salt master recently >.<
19:44 cpattonj le sigh
19:46 iggy how different are the versions? you could try clearing the cache
19:49 moogyver PredatorVI: i updated your gist
19:49 moogyver PredatorVI: try that out and let me know if that works
19:50 moogyver gonna go grab food
19:50 PredatorVI kk
19:51 mapu joined #salt
19:51 jhauser joined #salt
19:53 dyasny joined #salt
20:04 GabLeRoux joined #salt
20:04 kinetic joined #salt
20:11 CeBe joined #salt
20:13 aparsons joined #salt
20:13 kinetic joined #salt
20:16 bastion1704 joined #salt
20:21 huddy joined #salt
20:23 larsfronius joined #salt
20:24 sir_tyrion joined #salt
20:25 sir_tyrion Hi, the gpg key for ubuntu dists is malformed: SALTSTACK-GPG-KEY.pub
20:25 sir_tyrion https://repo.saltstack.com/apt/ubuntu/ubuntu14/SALTSTACK-GPG-KEY.pub
20:25 sir_tyrion there is that =dtMN at the end that shouldn't be there
20:26 PredatorVI moogyver:  No dice ... 'no client specified'.  I moved the client back to it's own -d parameter and then I got 'not authorized' so I moved the eauth/username/password out of the json and got 'runner takes at least 1 argument (0 given)'.  I've been messing with other combinations, but always the 'runner takes at least 1 argument (0 given)' error.
20:28 * babilen wonders when he'll feel happy with the state of repo.saltstack.com to actually use their packages
20:28 sir_tyrion babilen why wouldn't you use their packages?
20:28 sir_tyrion are you sticking with 0.17.5?
20:28 kinetic joined #salt
20:28 zmalone sir_tyrion: https://github.com/saltstack/salt/issues is probably where you should report that
20:29 zmalone sir_tyrion: because for Ubuntu, that repo is all of a week old
20:29 zmalone prior to that they recommended an independently run ppa that had 2015.5, but never got 2015.5.5
20:30 sir_tyrion oh didn't know
20:30 zmalone Upgrades from 2015.5 to 2015.8 cause some problems though, so it's worth being a little careful
20:30 cpattonj_ is there a list of the problems somewhere?
20:30 babilen sir_tyrion: No, I'm on 2015.5 ... I couldn't install 2018.8.0 because the packages were in contrib and it looks as if there are still problems with the repo. I might wait a little longer for other problems to surface. Can't really risk spending a lot of time on this right now
20:31 zmalone cpattonj_: https://github.com/saltstack/salt/issues
20:31 sir_tyrion well i just went from 0.17.5 to 2015.8.0 just now
20:31 sir_tyrion we'll see how that goes
20:31 babilen quite a jump
20:32 sir_tyrion yup, I wasn't using a custom ppa before
20:32 babilen https://docs.saltstack.com/en/latest/topics/releases/2015.8.1.html
20:36 clintberry joined #salt
20:37 Akhter joined #salt
20:38 kinetic joined #salt
20:39 iggy install from source
20:39 ahammond I have a couple of states that run pkg.install. They both have aggregate: False. However I'm seeing aggregation. How do I force these to be run sequentially?
20:40 RedundancyD joined #salt
20:40 moogyver PredatorVI: You include the -H 'Content-Type: application/json' ?
20:40 moogyver Because that command works for me
20:40 PredatorVI moogyver: yes
20:40 PredatorVI waiut
20:40 PredatorVI no
20:41 PredatorVI Accept
20:41 moogyver yeah, you need to specify both
20:41 moogyver one is saying 'i accept json', the other is saying 'i'm sending you json'
20:41 ahammond https://gist.github.com/ahammond/b1ce6bc7c36b34f913c8 specifically
20:41 PredatorVI if that is it, I'm gonna scream :)
20:42 moogyver well, you also needed to format it as json too, but yeah..
20:46 moogyver but yeah, it works for me to fire off a orch w/ pillar data
20:47 aqua^c joined #salt
20:47 bhosmer joined #salt
20:47 kinetic joined #salt
20:47 keimlink joined #salt
20:51 DanyC joined #salt
20:54 baweaver joined #salt
20:56 beardedeagle left #salt
20:56 ahammond how do I force pkg.installed states to be run sequentially (not aggregated)?
20:57 PredatorVI moogyer:  thanks...I've updated the gist with what I think are all the recommendations and the error is slightly different.  I've run the JSON through a validator and it didn't complain.  Is the pillar value supposed to be in JSON format to or embedded yaml (somehow)?
20:57 kinetic joined #salt
20:57 moogyver it's supposed to be json
20:58 PredatorVI Error updated at https://gist.github.com/PredatorVI/5397a9c184c8f4bfa433
20:58 moogyver PredatorVI: is the gist pasted right?  shows '-H 'ept: application/json'
20:58 moogyver rather than Accept :)
20:58 PredatorVI bad paste...fixing
20:59 PredatorVI fixed
20:59 JDiPierro ahammond: If you're not using the "pkgs" argument or specifying "aggregate: True" then it shouldn't aggregate packages across stanzas.
20:59 moogyver and PredatorVI : what version of salt is this and I'm assuming the rest-cherrypy module?
21:00 ahammond JDiPierro we have aggregation enabled by default in salt master's config.
21:00 ahammond however, we've also added aggregate: False to these states, and yeah, they're separate states, not pkgs
21:00 PredatorVI moogyver:  v2015.5.3 (Ubuntu 14)
21:00 moogyver ok, let me check something ( i've been testing on 2015.8 )
21:01 PredatorVI yes rest_cherrypy
21:01 JDiPierro ahammond: "aggregate: False" doesn't work? :( That should be reported as a bug IMO. Only other thing I'd suggest is having one pkg.installed require another pkg.installed... no idea how that would work with aggregates, but worth a shot.
21:01 PredatorVI moogyver:  the formatting and JSON key names looked okay so far?
21:02 ahammond JDiPierro yeah, that's exactly what we did. https://gist.github.com/ahammond/b1ce6bc7c36b34f913c8
21:02 moogyver yeah, it looks ok.  it's just saying that you're calling the orchestrate function but not supplying any arguments to it
21:02 PredatorVI ok
21:05 PredatorVI moogyver:  got it working...I stumbled on this thread and the 'mods' and 'pillar' keys were shown not part of a kwargs keyword.  When I made them peer to 'fun' it worked:  https://groups.google.com/forum/#!topic/salt-users/NGK6FoeiWXc
21:07 moogyver PredatorVI: nice! looks like that changed in 2015.8 then
21:07 moogyver because if i do that, I get your old error :)
21:07 PredatorVI good to know....
21:07 PredatorVI hopefully I can remember this after I upgrade :/
21:09 aidalgol` joined #salt
21:09 aidalgol` joined #salt
21:10 bhosmer joined #salt
21:15 kinetic joined #salt
21:21 kinetic joined #salt
21:25 wryfi joined #salt
21:25 wryfi how do i depend on a pkgrepo state in a different file?
21:25 baweaver joined #salt
21:26 aurynn you have to include: the other statefile
21:27 moloney joined #salt
21:28 shanemhansen joined #salt
21:29 moloney I am noticing that using "pillar.get('some-key')" in some state file Jinja doesn't work, but "salt['pillar.get']('some-key')" does work. Any idea why?  Or why the first one doesn't raise an error?
21:30 moloney Err, and I am using colon notation for nested access.  So intead of just 'some-key' it is more like 'roles:some-role', maybe that is related?
21:30 sunkist joined #salt
21:31 kinetic joined #salt
21:34 babilen moloney: pillar.get() calls the .get() method on dictionaries which doesn't support nested access (don't ask why), while salt['pillar.get'] is a salt funtion that does
21:35 babilen (standard Python .get() method that is)
21:35 jcristau joined #salt
21:35 Guest89_ joined #salt
21:36 moloney babilen: Thanks, makes sense. I guess I kindof figured it out as I typed my question ;)
21:40 moloney I am currently using the "module.run" state to run "saltutil.refresh_pillar" at the beggining of a highstate (to work around this: https://github.com/saltstack/salt/issues/24050).  This state shows up as "changed" every time I run the highstate. I guess is expected but I would prefer to avoid this as it adds "noise" to the results output, any way to avoid this?
21:40 saltstackbot [#24050]title: state.highstate does not refresh minion cache pillar data | If you change pillar data, and then run state.highstate (via salt or salt-call), the pillar data will not be refreshed on the minion cache. It was my understanding that a state.highstate run should update pillar minion cache....
21:40 pravka joined #salt
21:42 iggy doubtful
21:42 jhauser joined #salt
21:44 iggy moloney: maybe a dummy state at the beginning of your code that does a test.succeed_without_changes and has reload_pillar: True ?
21:45 baweaver joined #salt
21:45 subsignal joined #salt
21:46 kinetic joined #salt
21:47 hoonetorg hi, are there alternatives to cp.push
21:47 teebes joined #salt
21:47 hoonetorg (for people who are not allowed to set file_recv: True)
21:47 wryfi aurynn: thanks
21:48 hoonetorg I want to copy some generated config from a controller node to the master and then distribute it from the master to slave nodes
21:49 babilen mine
21:49 babilen events
21:49 moloney iggy: Thanks for the suggestion, unfortunately the "reload_pillar" part doesn't actually refresh the cache.  I guess I will live with my current setup until that bug is fixed
21:49 hoonetorg babilen: thx
21:50 hoonetorg never tried that
21:50 babilen hoonetorg: I'm sending around generated certificates by including it in the event data and then distribution it via suitable reactors on the master
21:50 NightMonkey joined #salt
21:50 babilen hoonetorg: https://docs.saltstack.com/en/latest/ref/states/all/salt.states.event.html#salt.states.event.send
21:50 hoonetorg babilen: hv u some good snippets of what you do there?
21:51 hoonetorg babilen: will read doc link from you 1st
21:51 hoonetorg babilen: thx
21:51 babilen hoonetorg: Ping me tomorrow, it is not easy to boil it down to the essentials and it is late now.
21:52 APLU joined #salt
21:53 hoonetorg i'm gmt+2 i understand
21:53 hoonetorg babilen: will be around tomorrow and ping you
21:54 overyander hey, am i reading lines 193 through 201 correctly that the bootstrap will install .net 4 for me? https://github.com/saltstack/salt/blob/2015.5/salt/modules/chocolatey.py
21:54 hoonetorg babilen: what is a good time, what tz
21:56 babilen cest
21:56 hoonetorg ok my tz
21:57 Brew joined #salt
21:57 pm90_ joined #salt
21:57 hoonetorg babilen: sleep well, enjoying my beer now
21:59 pm90__ joined #salt
22:00 kinetic joined #salt
22:09 scoates joined #salt
22:13 lude joined #salt
22:15 baweaver joined #salt
22:16 UtahDave overyander: that appears to be the case.
22:20 overyander Hey, UtahDave,  just the person i was wanting to speak to. referencing this https://github.com/saltstack/salt/issues/20685 and  https://github.com/jfindlay/salt/commit/330b43e340209b2ab455b31ec8234bd8109baefd the patch doesn't seem to exist in 2015.5.5 or 2015.8.x or am i looking at this wrong?
22:20 saltstackbot [#20685]title: The next version of Chocolatey (v0.9.9) will require -y | Update chocolatey.install module to use -y flag for compatibility with latest version....
22:20 moogyver for acl stuff, is it possible to target based off pillars or grains?
22:23 kinetic joined #salt
22:23 UtahDave overyander: This PR fixes that. https://github.com/saltstack/salt/pull/21058     Let me see if it got merged forward
22:23 saltstackbot [#21058]title: Support Chocolatey 0.9.9+ | The upcoming release will require "--yes" to execute commands that change...
22:23 pm90_ joined #salt
22:24 UtahDave overyander: Yeah, it appears to be in 2015.8.0 as well
22:24 kinetic joined #salt
22:24 overyander i have the code pulled up https://github.com/saltstack/salt/blob/2015.8/salt/modules/chocolatey.py but don't see the line where the install commands had -y
22:24 overyander *has
22:25 UtahDave look at line 54
22:25 UtahDave oh wait.  that's not it.
22:26 overyander i expected it to be on 376
22:26 ahammond @UtahDave any thoughts about https://github.com/saltstack/salt/issues/27961
22:26 saltstackbot [#27961]title: aggregate: False should disable aggregation even when state_aggregate: True enabled | On a salt-master with state_aggregate: True, a formula like:...
22:28 PredatorVI On ubuntu I have salt-minion 2015.5.2 (installed via the PPA) and want to upgrade but a standard 'apt-get update; apt-get upgrade' doesn't pick it up I have to do 'apt-get install salt-minion'.  Is this expected behavior?
22:28 UtahDave overyander: look at line 389
22:29 UtahDave ahammond: I think you have a valid point there.
22:29 otter768 joined #salt
22:30 ahammond @UtahDave we're removing state_aggregate for now. which is gonna slow things down. :(
22:30 ahammond fortunately most of our formulae specify aggregate: True, so hopefully not too much.
22:30 ashirogl joined #salt
22:33 * hacks waves to UtahDave
22:35 UtahDave hey, hacks!
22:35 KyleG joined #salt
22:35 KyleG joined #salt
22:35 * hacks stares at KyleG
22:36 KyleG I came here to ask a salt question, indeed. LOL
22:36 KyleG https://gist.github.com/kylegato/3fda566158dbbcc3c1d8
22:36 KyleG What the heck am I doing wrong :(
22:37 KyleG I know my salt_minion $PATH variable is set correctly, so it knows where the sysrc binary is
22:37 KyleG and it does add amd_enabled="TRUE" to /etc/rc.conf so I know it's able to fin dit
22:37 KyleG find it*
22:37 KyleG really puzzling
22:38 hacks KyleG: Getting anything in minion debug logs?
22:40 KyleG Nope all it says is
22:40 KyleG 2015-10-14 15:34:21,390 [salt.state                                          ][INFO    ] Running state [amd_flags] at time 15:34:21.389810
22:40 KyleG 2015-10-14 15:34:21,390 [salt.state                                          ][ERROR   ] State sysrc.managed found in sls amd.dev is unavailable
22:40 KyleG in debug mode
22:40 KyleG Maybe I should try garbage
22:40 wryfi left #salt
22:41 KyleG no difference in garbage debug mode.
22:41 KyleG "Unavailable" is a wonderfully descriptive error -_- lol
22:41 iggy the log message says you don't have sysrc.managed available
22:41 hacks Yeah, I am not sure if it is saying it can't be found, or if for some reason it can't be used
22:41 iggy either way
22:41 iggy that's going to fail
22:42 KyleG Is there any situation where it won't fail?
22:42 KyleG because that'd be ideal, obviously. heh
22:42 iggy salt 'minion' cmd.has_exec sysrc
22:43 iggy if it's not in a system path, I wouldn't count on Salt finding it
22:43 KyleG returns true.
22:43 KyleG As I said earlier, I already verified the $PATH
22:43 KyleG and it successfully uses sysrc to add amd_enabled="YES" to the /etc/rc.conf
22:43 KyleG so I know it can find and use the binary
22:44 iggy well, that error message seems to say otherwise
22:44 Eureka703 joined #salt
22:44 KyleG https://gist.github.com/kylegato/cb6e13d9b95232f97e76
22:44 KyleG there's a few more lines above
22:44 KyleG you can see where it checks the RC vars from the init script
22:45 KyleG then it says, hey this is already enabled in rc.conf
22:45 KyleG then it moves on to try to add amd_flags but says that unavailable message
22:45 Eureka703 Hi all, ive got an interesting question. Is is possible for me to use file.managed to reference a file in a minion cache to manage the file on the local salt master?
22:45 Eureka703 Basically, Im using cp.push on a minion to get a file to the master. Then I am using the minion running on my master to update the file in the filesystem.
22:46 Eureka703 (pulling it out of the location it ends up in cache)
22:46 iggy Eureka703: there's an issue open about adding file:// uri support to file.managed
22:46 sgargan joined #salt
22:47 Eureka703 @iggy ah, so it may not be currently doable at the moment?
22:47 iggy KyleG: try moving /etc/rc.conf to a backup file and try again
22:50 mohae joined #salt
22:50 KyleG Same behavior
22:50 baweaver joined #salt
22:50 KyleG Running 2014.1.13 (Hydrogen) if that helps
22:51 Eureka703 @iggy Oops! Looks like i solved it myself. I apparently need to specify file:///path rather than just file://path Sorry about that!
22:52 TyrfingMjolnir joined #salt
22:54 hacks KyleG: Probably not the best approach but I would modify the sysrc.py state file, to remove the load exclusion
22:54 hacks and see what happens
22:55 hacks https://github.com/saltstack/salt/blob/develop/salt/states/sysrc.py
22:55 hacks The only thing I can think of there is that it isn't loading, because it isn't seeing the exec, but you already demonstrated that isn't the case
22:56 hacks I would modify that, and log if/when a load decision was made, and back track from there
22:56 bfoxwell joined #salt
23:00 KyleG so yeah that's the problem….the version of salt I'm on does not have sysrc.py
23:00 KyleG explains a lot
23:00 hacks lolol
23:01 KyleG According to the commit history, it was added Dec 2014
23:01 KyleG and i'm on 2014.1 so yeah.
23:01 KyleG thanks anyways hacks
23:02 hacks Well, I feel like I sort of helped
23:02 KyleG indeed, saved me some pain in hunting
23:02 bougie joined #salt
23:03 hacks bah the whole reason I have the nick 'gg' is for right now, and I can't switch to it.
23:05 breakingmatter joined #salt
23:08 otter768 joined #salt
23:09 Guest89 joined #salt
23:10 pravka joined #salt
23:10 iggy I really hate that for a while, nobody was enforcing the versionadded annotations (probably aren't again)
23:12 ashirogl joined #salt
23:17 big_area joined #salt
23:19 _JZ_ joined #salt
23:19 borgstrom joined #salt
23:26 falenn joined #salt
23:29 Singularo joined #salt
23:30 aqua^c joined #salt
23:32 Guest89 joined #salt
23:40 ahammond iggy you said it!!!
23:41 keimlink joined #salt
23:47 falenn joined #salt
23:54 bhosmer joined #salt
23:57 baweaver joined #salt
23:57 thehaven_ joined #salt
23:57 Eureka70_ joined #salt
23:58 bhosmer joined #salt
23:59 brent4 joined #salt

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