Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-06-07

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

All times shown according to UTC.

Time Nick Message
00:01 ahammond I'm developing a reactor using the #!py renderer. I need some debug logging. Do I import logging or is there a salt specific way to do this?
00:02 jfindlay that's a good question.  I think I would start with `import logging` and see if it goes where you expect
00:03 MK_FG joined #salt
00:04 flowstate joined #salt
00:05 paydro joined #salt
00:06 jfindlay Vaish: https://github.com/saltstack/salt/issues/33811
00:06 saltstackbot [#33811]title: Document and test configs in saltrc | ### Description of Issue/Question...
00:07 amcorreia joined #salt
00:11 ajw0100 joined #salt
00:20 lilvim joined #salt
00:20 mikecmpbll joined #salt
00:22 edrocks joined #salt
00:23 DanSlimmon joined #salt
00:25 lilvim joined #salt
00:26 lilvim joined #salt
00:30 lilvim joined #salt
00:33 shoemonkey joined #salt
00:33 subsignal joined #salt
00:35 lilvim joined #salt
00:37 mosen joined #salt
00:38 barmaley joined #salt
00:39 jhujhiti at some point recently, the salt client started giving me 'no response' errors unless the minion finishes executing quickly. it times out very fast. i have not changed the timeout value. it used to wait for connected minions to finish. did behavior change recently?
00:39 jhujhiti i know the minion is working, i can see it doing things in its logs. the client just doesn't wait for it to finish before returning an error
00:39 lilvim joined #salt
00:40 lilvim joined #salt
00:43 lilvim joined #salt
00:48 fxhp joined #salt
00:48 mikecmpbll joined #salt
00:49 edrocks_ joined #salt
00:50 Nahual joined #salt
01:04 edrocks joined #salt
01:05 flowstate joined #salt
01:06 catpigger joined #salt
01:14 pyropoptrt joined #salt
01:22 toastedpenguin joined #salt
01:23 toastedpenguin joined #salt
01:25 tracphil joined #salt
01:31 toastedpenguin joined #salt
01:36 RS joined #salt
01:37 Guest6241 left #salt
01:41 toastedpenguin joined #salt
01:47 shoemonkey joined #salt
01:47 fgimian joined #salt
01:49 ilbot3 joined #salt
01:49 Topic for #salt is now Welcome to #salt! | Latest Versions: 2015.5.10, 2015.8.10, 2016.3.0 | Support: https://www.saltstack.com/support/ | Logs: http://irclog.perlgeek.de/salt/ | Paste: https://gist.github.com/ (please don't multiline paste into channel) | See also: #salt-devel, #salt-offtopic | Ask with patience as we are volunteers and may not have immediate answers
01:53 noraatepernos joined #salt
02:05 flowstate joined #salt
02:09 shoemonkey joined #salt
02:11 barmaley joined #salt
02:12 racooper joined #salt
02:14 brotatochip joined #salt
02:26 noraatepernos joined #salt
02:31 pyropoptrt joined #salt
02:39 paydro joined #salt
02:39 evle joined #salt
02:40 paydro_ joined #salt
02:42 toastedpenguin joined #salt
02:51 tuxx joined #salt
03:02 mikecmpbll joined #salt
03:03 flowstate joined #salt
03:04 Ayo joined #salt
03:07 felskrone joined #salt
03:14 BadgerOps joined #salt
03:23 ajw0100 joined #salt
03:23 noraatepernos Is there anything bad about this grain?  It’s my first one.  It’s a simple one for digital ocean droplets.  http://pastie.org/10867175
03:31 ageorgop joined #salt
03:39 manji joined #salt
03:43 barmaley joined #salt
03:46 favadi joined #salt
04:03 flowstate joined #salt
04:04 riceandbeans joined #salt
04:04 riceandbeans as someone who knows chef and puppet well, how hard would it be to learn salt?
04:07 mosen joined #salt
04:08 nmadhok joined #salt
04:14 sauvin joined #salt
04:20 jfindlay riceandbeans: my guess is probably not that hard, https://docs.saltstack.com/en/latest/
04:35 lompik joined #salt
04:40 madpenguin joined #salt
04:45 rhodgin joined #salt
04:52 smcquay joined #salt
05:05 flowstate joined #salt
05:06 nmadhok joined #salt
05:07 nidr0x joined #salt
05:22 nmadhok1 joined #salt
05:24 nmadhok joined #salt
05:25 nmadhok joined #salt
05:30 rdas joined #salt
05:31 manji joined #salt
05:37 nmadhok joined #salt
05:43 johje349 joined #salt
05:54 _JZ_ joined #salt
05:57 krymzon joined #salt
05:57 nmadhok joined #salt
05:58 anmol joined #salt
06:00 favadi joined #salt
06:03 flowstate joined #salt
06:04 anmol joined #salt
06:07 mikecmpbll joined #salt
06:10 kawa2014 joined #salt
06:13 nmadhok1 joined #salt
06:26 om joined #salt
06:28 Miouge joined #salt
06:30 impi joined #salt
06:32 nmadhok joined #salt
06:35 west575 joined #salt
06:37 sjorge joined #salt
06:42 linjan joined #salt
06:46 smcquay joined #salt
06:47 nmadhok1 joined #salt
06:49 nmadhok joined #salt
06:50 fracklen joined #salt
06:50 dmaiocchi joined #salt
06:51 fracklen joined #salt
06:52 KermitTheFragger joined #salt
07:00 slav0nic joined #salt
07:02 nmadhok1 joined #salt
07:04 _JZ_ joined #salt
07:04 flowstate joined #salt
07:04 toanju joined #salt
07:11 dmaiocchi joined #salt
07:14 mrueg joined #salt
07:15 kshlm joined #salt
07:16 duncanmv joined #salt
07:18 nmadhok joined #salt
07:21 colegatron joined #salt
07:22 colegatron good morning
07:24 smcquay It's morning somewhere ;)
07:25 colegatron I need to regex_replace inside a jinja template. puttin myhelper.py inside /srv/salt/_modules, following the example in https://groups.google.com/forum/#!msg/salt-users/EUbnCGlroAA/i3Q3d0FOSx0J I get an error "'salt.utils.templates.AliasedLoader object' has no attribute 'ivan-helpers.re_replace'"
07:25 manji joined #salt
07:26 colegatron this is my first attempt to write custom code, so I am doing something wrong, but not sure what
07:26 manji joined #salt
07:26 colegatron any advice from a more experienced eyes?
07:29 colegatron the line with the error is: location {{ salt['my-helper.re_replace'](route, ':(.*?)(?=/|$)', '(.*)' ) }}
07:30 mikecmpbll joined #salt
07:33 om joined #salt
07:34 colegatron the example is from '12 maybe it is not the right way to do it nowadays
07:36 punkoivan joined #salt
07:38 deniszh joined #salt
07:42 mikecmpbll joined #salt
07:52 duncanmv joined #salt
07:56 krymzon joined #salt
07:57 punkoivan joined #salt
07:59 mikecmpbll joined #salt
08:03 rim-k joined #salt
08:05 flowstate joined #salt
08:07 Miouge joined #salt
08:07 duncanmv joined #salt
08:08 manji joined #salt
08:09 mpanetta joined #salt
08:12 lero joined #salt
08:14 iceyao joined #salt
08:16 s_kunk joined #salt
08:28 GreatSnoopy joined #salt
08:34 om joined #salt
08:34 om2 joined #salt
08:38 punkoivan joined #salt
08:38 ajw0100 joined #salt
08:45 EvaSDK hello there, anyone of the postgres-formula around here ?
08:45 mjimeneznet joined #salt
08:45 mjimeneznet Hi! In a .sls file can I do this? {% if grains['host'] in ['*frontend*','*integration*','*staging*'] %} bla bla bla {% endif %}
08:46 EvaSDK mjimeneznet: you probably can't do globbing like this no
08:47 mjimeneznet If i can't globbing... in case that I have for instance 10 frontends, 10 integrations... I have to write all them down in the "if" ?
08:48 EvaSDK multiple {% if 'frontend' in grains['host'] %} maybe
08:48 EvaSDK can't leave that to top.sls btw ?
08:50 mjimeneznet EvaSDK is my first time with Salt, is a project that I have to get .... I'm fighting with that :)
08:53 mjimeneznet EvaSDK: {% if 'frontend' in grains['host'] %} . Writting this matchs all frontends* ... :D
09:04 flowstate joined #salt
09:05 AndreasLutro if grains['host'].startswith('frontend')
09:19 colegatron I want to add a function to all it as salt["myhelper.function"] inside a jinja template. I have been reading the doc, but not sure if I should think on it as state module, execution module, or a completely different thing.
09:19 colegatron what is the good practice to accomplish this?
09:20 manji use _modules and _states
09:20 manji I suppose you mean to write your own salt module, yes?
09:21 colegatron I need to run a regexp replace inside a jinja template, but jinja has no regexp replace, so I've read that could be done writting a 'module' (https://groups.google.com/forum/#!msg/salt-users/EUbnCGlroAA/i3Q3d0FOSx0J) but I can't get this working
09:22 colegatron maybe because it is a post of 2012 and nowadays that is not the right way  to do it, idk
09:22 manji oh I did that once, about a year ago
09:22 colegatron not sure if that is a 'state module', 'execution module' or 'whatever module'. I am confused with that
09:23 manji iirc
09:23 manji I found the jinja filters that salt is using
09:23 manji and I added the re filter
09:23 manji there was a snippet somewhere
09:23 manji hmm
09:24 manji let me see if I can find it again
09:24 colegatron if it can be done within salt, instead adding functionality to jinja, would be nice (to learn)
09:25 impi joined #salt
09:25 AndreasLutro colegatron: salt['mymod.myfunc'] would be an execution module
09:25 AndreasLutro for that example you would just create /srv/salt/_modules/mymod.py and define myfunc in there
09:26 manji that would be surely more proper yes
09:26 AndreasLutro a jinja filter/function would be nice but I don't think salt lets you add those so easily
09:27 om2 joined #salt
09:27 Guest71106 joined #salt
09:28 colegatron AndreasLutro, and it can be used as a regular python function, right?
09:28 manji has anyone found a way to grant mysql privileges to database *.* ?
09:30 colegatron Jinja variable 'salt.utils.templates.AliasedLoader object' has no attribute 'myhelpers.re_replace'
09:30 colegatron :-(
09:30 kshlm joined #salt
09:31 colegatron but it is there http://pastebin.com/AYkuhS3R
09:32 josuebrunel joined #salt
09:32 colegatron I tried to restart the minion service (it is masterless), but no luck
09:33 manji colegatron, my guess is that you need to make that into a string
09:33 manji before you can apply you function to it
09:33 MadHatter42 joined #salt
09:34 manji or something in that directrion
09:34 manji and you need make sure that all modules are synced
09:35 colegatron it says nothing about an execution problem, it is about it does not 'see' the function
09:35 EvaSDK mjimeneznet: well, you did write '*frontend*' so that is what I expected you wanted
09:35 EvaSDK :)
09:42 Ayo joined #salt
09:45 duncanmv joined #salt
09:46 colegatron AndreasLutro, should the modules be added in the minion config in some way (its masterless)?
09:46 colegatron a 'salt-call my.module' says it is not available, but if I do 'salt-call -m /srv/salt/_modules' it
09:46 colegatron works
09:47 babilen colegatron: Did you put it on the minion or master?
09:47 colegatron minion, it's masterless
09:48 babilen Did you sync modules or restart the minion?
09:48 colegatron restarted
09:48 babilen I guess debug output is needed and I have no idea what you've done so far :)
09:48 colegatron there is any specific to sync modules in masterless minio?
09:49 colegatron babilen, I'll exlain: put a myhelpers.py in /srv/salt/_modules with a function re_replace(a,b,c)
09:50 babilen Sounds about right, run the minion in debug mode and check for errors ..
09:51 colegatron a 'salt-call myhelpers.re_replace a b c' says it is not available, but a 'salt-call myhelpers.re_replace a b c -m /srv/salt/_modules' it works
09:51 babilen I don't run masterless, but no obvious error is apparent
09:54 om joined #salt
09:54 om2 joined #salt
09:57 permalac joined #salt
09:58 colegatron babilen, tnx. I'm going to try debug.
10:04 flowstate joined #salt
10:07 kaushal_ joined #salt
10:12 kaushal_ joined #salt
10:14 liskl joined #salt
10:21 kshlm joined #salt
10:21 manji colegatron, did you find anything ?
10:21 colegatron nothin on the debug log about my module. there are errors of modules not loading because dependencies, but nothing about my custom module
10:21 colegatron the curious thing is that works if you specify the path to the _module folder
10:22 colegatron like if the default configuration was not reading the file_root: /srv/salt/_modules
10:22 manji aaa
10:22 manji then
10:22 manji have you added that to file_roots ?
10:23 manji same key where you declare where your states are
10:23 colegatron it is declared by default; file_root: - /srv/salt and salt will (theoretically) read the /srv/salt/_modules folder
10:24 colegatron I tried to declare it explicitly, but no luck
10:25 manji probably useless in masterless, but in master we run a saltutil.sync_modules
10:26 colegatron there is no saltutil afaik, but let me see...
10:27 josuebrunel joined #salt
10:29 colegatron oh, wait; sudo salt-call saltutil.sync_modules
10:30 colegatron :facepalm:
10:30 manji it worked ?
10:31 colegatron yes, it worked
10:31 manji hehe
10:32 colegatron well... learned by the hard way. I'll not forget :)
10:32 manji haha
10:33 manji you're not alone mate trust me
10:39 tuxx joined #salt
10:40 colegatron :) I know, but it not makes it sweeter :)
10:41 Ayo joined #salt
10:41 om joined #salt
10:41 om2 joined #salt
10:54 babilen colegatron: So you didn't sync?
10:55 * babilen should have been more insistent earlier
10:59 felskrone joined #salt
11:02 flowstate joined #salt
11:11 amcorreia joined #salt
11:11 duncanmv joined #salt
11:19 RandyT joined #salt
11:21 subsignal joined #salt
11:24 iceyao joined #salt
11:29 arnaud left #salt
11:38 evle joined #salt
11:39 DanSlimmon joined #salt
11:47 punkoivan left #salt
11:52 om joined #salt
11:52 om2 joined #salt
11:52 josuebrunel joined #salt
11:53 anmol joined #salt
12:10 abednarik joined #salt
12:13 rgrinberg joined #salt
12:20 west575 joined #salt
12:33 source47 joined #salt
12:35 DanSlimmon joined #salt
12:35 source47 hi guys, i am trying to run a docker container in salt. but i am getting errors. https://gist.github.com/developerinlondon/02042dc02db59d67ba3f83a9bf6db353 any ideas?
12:37 AndreasLutro source47: I think it's dockerng, not dockereng
12:39 source47 nope. dockerng doesnt work either ... https://gist.github.com/developerinlondon/02042dc02db59d67ba3f83a9bf6db353
12:41 shoemonkey joined #salt
12:45 gh34 joined #salt
12:46 AndreasLutro what version of salt are you running?
12:47 teryx510 joined #salt
12:49 fgimian joined #salt
12:51 Ayo joined #salt
12:52 Guest25178 joined #salt
12:52 om2 joined #salt
12:54 source47 AndreasLutro, that seems to be it. I have an older version. Let me try updating it.
12:54 fredvd joined #salt
12:57 impi joined #salt
13:00 _JZ_ joined #salt
13:04 dendazen joined #salt
13:05 source47 upgrading helped a bit but still getting some problems:
13:05 ngs joined #salt
13:06 source47 https://gist.github.com/developerinlondon/02042dc02db59d67ba3f83a9bf6db353
13:07 flowstate joined #salt
13:07 flowstate joined #salt
13:07 MeltedLux joined #salt
13:11 tomasfejfar joined #salt
13:12 josuebrunel joined #salt
13:13 tomasfejfar Hi, I'm using salt in master-less setup and it seems to be updating the gitfs every time it touches it. That makes it very slow. Any ideas how to fix it? (relevant part of the log: https://gist.github.com/tomasfejfar/04e21748e6d2558ec9e66fe0dbb4e16c)
13:14 subsignal joined #salt
13:16 tomasfejfar It does that after every YAML render. ("
13:16 tomasfejfar [DEBUG   ] Results of YAML rendering:
13:17 tomasfejfar I'm running salt-call --local state.highstate test=True
13:18 racooper joined #salt
13:19 Tyrm joined #salt
13:21 west575 joined #salt
13:22 amcorreia d
13:22 * amcorreia ops
13:22 flowstate joined #salt
13:24 slav0nic joined #salt
13:24 hlub joined #salt
13:28 hlub cp.cache_file cannot be used together with salt-ssh. is there other ways to get contents of a file from under the file root?
13:29 tomasfejfar hlub: cp.get_file_str does not work?
13:31 ageorgop joined #salt
13:31 dyasny joined #salt
13:32 hlub oh, that one works
13:33 tomasfejfar hlub: ;)
13:33 protoz joined #salt
13:35 mcxian joined #salt
13:35 flowstate joined #salt
13:35 mcxian Howdy all. I'm having a problem writing a reactor with the py renderer. I need it to call a runner, but am getting stuck on the opts = salt.config.master_config('/etc/salt/master')  line needed in salt.runner.RunnerClient(opts)
13:36 mcxian When attempting the salt.config.master_config() bit, I get back an error: AttributeError: 'OrderedDict' object has no attribute 'master_config'
13:36 mcxian Does anyone have any advice?
13:37 subsigna_ joined #salt
13:37 edrocks joined #salt
13:38 AndreasLutro why are you trying to use a runner in an sls?
13:40 mcxian I would like the reactor to wheel accept a minion key. But I don't want it to accept just all keys, so I've written a small runner that checks to see if a host is in our cmdb
13:40 mcxian so I would like the reactor to call that runner to see if it should accept the minion, and if so calls the appropriate wheel accept
13:41 mcxian so my reactor is written as a #!py renderer
13:41 iceyao joined #salt
13:42 AndreasLutro okay... but you don't do that by instantiating a runner in the sls, you just return a dict from the run() function that tells the salt master what to do (which can include invoking a runner)
13:42 traph joined #salt
13:42 traph joined #salt
13:42 mcxian Here's what I'm trying to do: https://gist.github.com/mchugh19/7be27ea18b5585da96e331d527d94d6b
13:43 mcxian can you show me an example of the run() function?
13:43 jdubski joined #salt
13:44 AndreasLutro you should move the check_autosign logic into the custom runner you've written
13:45 rhodgin joined #salt
13:46 AndreasLutro I don't have one lying around, but it'd basically just be a python function that returns the same dict that a yaml sls would render into
13:46 mcxian The authminion.servicenow runner I've created does have all necessary logic. You just call salt-run authminion.servicenow hostname and it returns True or False
13:47 mcxian I just don't know how to properly call that runner from the reactor
13:47 mcxian so I don't think there is any logic to move over. I'm just having trouble calling that runner
13:48 flowstate joined #salt
13:48 AndreasLutro because you're not supposed to do that
13:48 AndreasLutro you can't use complex logic in reactors
13:49 mcxian the logic is in the runner. I just need the reactor to call the runner
13:49 AndreasLutro no, you have logic in your check_autosign function
13:49 AndreasLutro that won't work in a reactor
13:49 AndreasLutro move it to a runner
13:50 mcxian The only thing in that function is attempting to call the runner
13:50 AndreasLutro I won't repeat myself :)
13:50 mcxian the logic of the reactor is in the run function and is: if data.get('act') == 'pend' and check_autosign(minion_id)
13:51 mcxian where check_autosign is just trying to call the runner
13:51 jad_jay joined #salt
13:52 mcxian So if I have a runner that returns True or False and contains all the necessary logic, how might I call that runner from within my reactor?
13:52 mcxian I'm using the py renderer
13:55 heewa joined #salt
13:55 mcxian In that gist I pasted, there already is a run function. I've truncated the return at the bottom, but it does return the same dict that a yaml sls would render into
13:56 mcxian just inside of that run function, I would like to call a salt runner which contains all the complex logic
13:56 mpanetta joined #salt
13:56 perfectsine joined #salt
13:56 mikeywaites joined #salt
13:57 mikeywaites Hey! What is the accepted process for upgrading salt master / minions to specific versions ?
13:57 mpanetta joined #salt
13:57 manji mikeywaites, read the changelog first
13:58 manji carefully
13:58 manji then update master, then update minions
13:58 manji if you have a staging master-minion, try there first
13:58 Ayo_ joined #salt
13:59 mward joined #salt
14:01 source47 i am getting " Failed to create new container: Unable to perform create_container: create_container() got an unexpected keyword argument 'port_bindings'" when using dockerng
14:01 source47 but port_bindings is a valid parameter
14:02 andrew_v joined #salt
14:02 perfectsine_ joined #salt
14:03 DanSlimmon joined #salt
14:03 mikeywaites manji:  thanks for the info.  I meant to a specific version though.  Obviously updating via apt or whatever is just gonna bump me top the LTS
14:03 mikeywaites i'd like to go to a specific version.
14:04 manji mikeywaites, do an apt-cache policy salt-master
14:04 manji to check all available versions on your repo
14:05 manji before you go ahead and do manual work
14:05 mikeywaites manji:  Great thanks!
14:05 menglish joined #salt
14:05 manji :)
14:07 DarthAT joined #salt
14:08 DarthAT hellp
14:08 BadgerOps joined #salt
14:08 babilen mikeywaites: I would recommend to upgrade minions using the procedure discussed on https://docs.saltstack.com/en/latest/faq.html#what-is-the-best-way-to-restart-a-salt-daemon-using-salt
14:09 heewa @ecdhe Try: `salt '*' cmd.run 'hostname {{ grains["id"] }}' template=jinja`
14:09 shoemonkey joined #salt
14:09 heewa Woops, meant eckod. They're gone :(
14:10 DarthAT when i run salt-master it sits on this for a very long time (hours) and never reaches completion... [INFO    ] Worker binding to socket ipc:///var/run/salt/master/workers.ipc [INFO    ] Worker binding to socket ipc:///var/run/salt/master/workers.ipc [INFO    ] Worker binding to socket ipc:///var/run/salt/master/workers.ipc [INFO    ] Worker binding to socket ipc:///var/run/salt/master/workers.ipc [INFO    ] Found minion id from gener
14:18 hlub cp.get_file_str raises an odd erro when usedvia salt-ssh: "Passed invalid arguments: coercing to Unicode: need string or buffer, bool found." I definitely gave it a string. Tiis happens only with salt-ssh. :O
14:20 mpanetta_ joined #salt
14:21 tomasfejfar hlub: could be that the file was not found (returned false) which is not "displayable" so thus the warning
14:21 hasues joined #salt
14:21 hasues left #salt
14:22 dmaiocchi joined #salt
14:23 rhodgin_ joined #salt
14:24 hlub tomasfejfar: it is the very same file in both cases.
14:29 tkharju joined #salt
14:30 tomasfejfar hlub: but the way you access it could not work with ssh
14:30 tomasfejfar *may not
14:30 tomasfejfar hlub: can you share the exact command?
14:32 tapoxi joined #salt
14:32 hlub okay, that same error message appears when: salt some_machine cp.get_file_str salt://some/non/existing/url
14:32 hlub and the file I tested is not cached by salt-ssh.
14:36 ageorgop joined #salt
14:36 hlub so, does that mean there is no access to those files when using salt-ssh?
14:37 tkharju joined #salt
14:39 Tyrm_ joined #salt
14:40 Tyrm__ joined #salt
14:41 MadHatter42 joined #salt
14:44 dmaiocchi joined #salt
14:46 wiqd joined #salt
14:53 DaV joined #salt
14:53 DaV Hi everyone
14:54 _JZ_ joined #salt
14:54 sten_ joined #salt
14:55 brew joined #salt
14:56 shiriru joined #salt
14:57 _JZ_ joined #salt
14:58 ageorgop joined #salt
14:58 west575 joined #salt
14:58 MadHatter42 joined #salt
14:59 johtso joined #salt
14:59 komputes joined #salt
15:00 MadHatter42 joined #salt
15:02 rhodgin joined #salt
15:04 Tyrm joined #salt
15:04 coleman joined #salt
15:05 noraatepernos joined #salt
15:06 M-liberdiko joined #salt
15:08 shoemonkey joined #salt
15:09 noraatepernos Is it a bad idea to write files with cloud-init like /.platform with contents “digitalocean” and /.roles with comma separated values?  These could then be read by a grain.  I use only digital ocean droplets and aws ec2 right now.
15:17 ssplatt joined #salt
15:22 ssplatt hey guys, tryign to figure out why my external pillar module isn’t loading on my prod salt master and syndics (and minions). in my vagrant environment it’s loading fine. i’ve set module_dirs: - /srv/saltstack/modules/mymod   and the actual module is in /srv/saltstack/modules/mymod/pillar.  i see i get the .pyc file which makes me think it is loading, but i don’t see the LazyLoading: mymod.ext_pillar  in the debug output
15:22 ssplatt don’t see anything when i call pillar.get mymod
15:23 TooLmaN joined #salt
15:25 TooLmaN joined #salt
15:26 ssplatt maybe i’m looking in the wrong spot.
15:30 ssplatt on the minion, should i see it get pulled into /var/cache/salt somewhere?
15:39 Ayo joined #salt
15:43 ssplatt should i be settings the external pillar dir in the master config or the minion config?
15:43 ssplatt module_dirs:
15:44 jorr-el joined #salt
15:46 flowstate joined #salt
15:50 hasues joined #salt
15:51 hasues left #salt
15:52 scottk_ joined #salt
15:52 abednarik joined #salt
15:52 M-MadsRC1 joined #salt
15:54 Tyrm joined #salt
15:54 flowstate joined #salt
15:55 iggy noraatepernos: there are grains for aws and do already in salt-contrib... what does writing out files and then reading them buy you?
15:56 abhi_ joined #salt
15:57 ssplatt should external pillar modules sync to all minions when saltutil.sync_modules is run?
16:00 UtahDave joined #salt
16:01 Tyrm joined #salt
16:02 Tyrm joined #salt
16:02 Tyrm joined #salt
16:03 nmadhok joined #salt
16:03 heewa joined #salt
16:03 bltmiller joined #salt
16:04 ageorgop joined #salt
16:04 edrocks joined #salt
16:09 mward Heya.  If I wanted to restrict '@wheel' permissions in my external_auth configuration to say, only wheel.key functions, what would the syntax for that be?
16:10 iggy ssplatt: ext_pillar's run on the master
16:11 iggy mward: I've never seen any syntax for that (i.e. it may not be possible)... but wait around for someone a little more knowledgeable
16:12 flowstate joined #salt
16:16 sten_ I have an orchestrate runner that fails when a minion is already running a scheduled highstate, is there anyway to make the orchestrate wait instead of failing?
16:16 josuebrunel joined #salt
16:17 UtahDave mward: I believe that at the moment @wheel is all or nothing
16:18 mward Got it, thanks!
16:19 UtahDave sten_: I think there's a "queue" option. From what I understand it's not a real queue, but will cause the orchestrate runner to just sit and wait for the other highstate is finished.
16:19 shoemonkey joined #salt
16:19 onlyanegg joined #salt
16:20 flowstate joined #salt
16:22 aharvey joined #salt
16:23 tracphil joined #salt
16:23 tracphil Hi Salters!
16:23 aharvey mornin’ mornin'
16:23 tracphil Is anyone using Docker to test states?
16:24 tracphil I am wanting to use Vagrant with either Docker or LXC.
16:25 aharvey We’re using openstack for VMs in general but I test-driven-dev locally using docker for states
16:25 tracphil aharvey, cool and it works fine for you? That is what I am looking for as well, test locally before push to a git repo.
16:26 cyborg-one joined #salt
16:26 iggy doesn't vagrant only do 1 VM at a time?
16:26 aharvey yeah it smokes.  It was a bit of work getting it to behave like an openstack VM but states run quick
16:26 tracphil No, it can do multiple.
16:26 aharvey you’ll vagrant ssh into the VM and kick off docker containers from within
16:26 aharvey sec lemme find the repo I edited for my usage case
16:26 tracphil man that would make my day!
16:27 aharvey https://github.com/amolenaar/salt-formula-testing
16:27 abednarik joined #salt
16:28 tracphil I really appreciate the validation from you that it works well and the pointer to get started with, that helps alot.
16:28 aharvey There’s some work you’ll need to do - get your formulas/states/pillars sync’d to the VM
16:28 aharvey here’s my vagrant file http://www.heypasteit.com/clip/2PG6
16:29 aharvey once you’re in the VM you’ll want to build the container images for centos/ubuntu/anything-but-windows
16:29 tracphil My hardest part is that I use reclass for pillar data with gitfs, but I think I have that part figured out in theory.
16:30 tracphil keyword being "theory" :)
16:30 aharvey gotcha - I’ve not looked at that.  We use gitfs for states and pillar - Currently going through a refactor thanks to Onno Ebbinge’s awesome presentation at saltconf
16:33 writtenoff joined #salt
16:34 tracphil I use gitfs for both as well and it works very nice pulling from a remote repo. With vagrant, I will have it pull from a local disk instead of a remote repo
16:34 beardedeagle joined #salt
16:34 shoemonkey joined #salt
16:35 tracphil "it" being salt-master
16:38 impi joined #salt
16:43 abhi_ hi all
16:44 abhi_ i am having a strange issue
16:44 abhi_ i created a pillar
16:44 abhi_ valid_groups:         - prod         - qa         - hdpdev         - business
16:44 abhi_ when i call in one of the branch of git it pulls multiple causing same namespace error in state
16:44 abhi_ any idea what could be the issue
16:44 abhi_ ?
16:46 abhi_ master config has below
16:46 abhi_ ext_pillar:   - git: master git@GitLab02.source.com:BigData/salt_pillars.git   - git: prod git@GitLab02.source.com:BigData/salt_pillars.git   - git: integration git@GitLab02.source.com:BigData/salt_pillars.git   - git: dev git@GitLab02.source.com:BigData/salt_pillars.git
16:46 abhi_ 4 diff branches
16:46 abhi_ and salt pillar.items get me output as
16:46 abhi_ valid_groups:         - prod         - qa         - hdpdev         - business         - prod         - qa         - hdpdev         - business
16:47 abhi_ notice repetition
16:48 tomasfejfar joined #salt
16:54 iggy why do you have all the branches listed? I thought the point of using gitfs was to get the branches automatically loaded up as env's?
16:54 iggy also, don't use environments
16:55 iggy they seem great, but...
16:57 cableninja is there a way in jinja to do if a and b ?
16:57 cableninja I havent seen any use examples anywhere so far
16:58 hlub cableninja: {% if a and b %} ... {% endif %}
16:58 iggy {% if a and b %}
17:00 heewa joined #salt
17:00 woodtablet joined #salt
17:01 bltmiller joined #salt
17:04 punkoivan joined #salt
17:09 rhodgin joined #salt
17:10 edrocks joined #salt
17:10 cableninja smh, that should've been stupidly obvious thanks hlub, iggy
17:11 ssplatt iggy: so if i’m seeing the .pyc file being created, then it should be loading right?  why am i not seeing the “LazyLoaded mymodule.ext_pillar” message in debug mode?
17:11 menglish joined #salt
17:12 leev joined #salt
17:14 abednarik joined #salt
17:17 punkoivan joined #salt
17:18 komputes joined #salt
17:19 bltmiller joined #salt
17:23 abhi_ am using gitfs only for states...there is only one master state branch...for pillars we have multiple branches
17:24 flowstate joined #salt
17:24 edrocks joined #salt
17:27 woodtablet BadgerOps: no luck with trying to vagrant up the same demo. i even updated my vagrant images of Trusty to the latest
17:27 mikecmpbll joined #salt
17:27 woodtablet ohhhh
17:27 woodtablet utahdave is here, and this is his demo
17:27 woodtablet i can ask himm
17:27 UtahDave woodtablet: what's going on ?
17:27 woodtablet UtahDave: hey dave
17:28 woodtablet UtahDave: I am trying out this cool salt vagrant demo you have https://docs.saltstack.com/en/getstarted/fundamentals/install.html
17:29 woodtablet UtahDave: but i get stuck at the very beginning, salt '*' test.ping
17:30 woodtablet UtahDave: the master is constantly timing out or something, with errors llike this: root@saltmaster:~# salt '*' test.ping
17:30 woodtablet Salt request timed out. The master is not responding. If this error persists after verifying the master is up, worker_threads may need to be increased. and this is a fresh vagrant up
17:30 UtahDave woodtablet: what's the output of   salt-master --version     ?
17:30 woodtablet UtahDave: salt-master 2016.3.0 (Boron)
17:31 UtahDave ok, let me try real quick.  Just a sec.
17:31 woodtablet ok cool, when i restart the master, i get lots of logs, with an error, see here: http://pastebin.com/VN5SYnjJ
17:31 woodtablet gl
17:32 flowstate joined #salt
17:32 alex_s joined #salt
17:33 punkoivan joined #salt
17:35 alex_s hi everyone! I have a question about file.managed and it would be awesome if you could help me: when I use a HTTP url as source and I provide as source_hash I would expect salt to download the file only if the hash does not match. But it seems to download it every time. How can I prevent this?
17:39 XenophonF well how else is salt supposed to check the checksum?
17:39 XenophonF hashes must be computed locally
17:39 alex_s when the state completed the first time the file is already available locally
17:40 alex_s there is not need to download it again, if the checksum did not change, right?
17:40 XenophonF so file.managed is going to download the file, compute the appropriate checksum, compare the computed checksum against the one provied in the arguments to salt.states.file.managed(), compare the downloaded file to the one on the minion, and copy the download file into place
17:40 komputes joined #salt
17:41 XenophonF i'm not sure file.manage is that clever about caching files
17:41 flowstate joined #salt
17:41 XenophonF but i'd have to RTFS to be certain
17:42 UtahDave woodtablet: It worked fine for me just now.  I'm updating my vagrant box to try again.
17:43 alex_s OK, I'm using version 2016.3.0
17:43 woodtablet utahdave: what version of trusty do you have working ? (vagrant box list )
17:45 UtahDave I just deleted ubunty/trusty64 so it will download it again
17:45 impi joined #salt
17:46 punkoivan left #salt
17:47 nmadhok joined #salt
17:47 mikecmpbll joined #salt
17:47 woodtablet ah ok. thanks again dave
17:47 alex_s I pasted my config and the log output to https://gist.github.com/astehlik/adb8e284cd8974cba0bd2858c5bb038b
17:48 gimpy2938 left #salt
17:48 manji joined #salt
17:48 UtahDave alex_s: is it possible to try putting the source hash in a file on your ftp server and having salt point to that source hash file instead? I wonder if it will then only download the actual file if the hashes are different.
17:49 UtahDave alex_s: i think that otherwise Salt has no way to know if the file has changed on the server unless it downloads it first
17:49 iggy ssplatt: I would think so
17:49 flowstate joined #salt
17:50 hopthrisC joined #salt
17:51 alex_s it could just calculate the checksum of the local file (if there is any) and then compare it to the configured source_hash
17:52 alex_s I try to find a download where a hash file is available on the server and test if it behaves differently
17:52 ssplatt iggy: i think it’s my mistake. my syndics were picking up the master config changes but not pulling down my git repo
17:52 hopthrisC my master changes ip addresses once a day. i tried the settings suggested for "DDNS" in the config file, but that doesn't seem to work. are they still valid?
17:55 flowstate joined #salt
17:58 hopthrisC i found the relevant code in minion.py, but those error messages never appear in the log file
17:58 komputes joined #salt
17:58 flowstate joined #salt
17:58 ajw0100 joined #salt
17:59 NV joined #salt
17:59 UtahDave woodtablet: Hm. Ok. I'm getting that same error now, too
18:02 anotherZero joined #salt
18:02 LtLefse joined #salt
18:02 alex_s @UtahDave: OK, I tried it with a GraphicsMagick mirror and it still downloads both files every time. I think I'll use a onlyif state that calculates the checksum of the existing file.
18:03 UtahDave alex_s: that seems like a bug to me.
18:04 LtLefse So, about this warning: "[WARNING ][6483] IMPORTANT: Do not use md5 hashing algorithm! Please set "hash_type" to SHA256 in Salt Master config!"
18:04 cableninja what am I doing wrong here? the cpu.conf keeps ending up with empty serverinfo/globalinfo, even though it should be set (as at least globalinfo should be getting set); http://pastie.org/10868225
18:04 alex_s I'm thinking the same because I'm quite sure that older versions did not download it very time ;)
18:04 LtLefse if it is indeed important not to use md5, why is md5 still the default?
18:05 UtahDave alex_s: would you mind opening a bug on that?
18:05 nmadhok joined #salt
18:05 UtahDave Gareth: you around?
18:05 alex_s sure. Thanks for your help :)
18:05 whaity joined #salt
18:09 fracklen joined #salt
18:10 aharvey joined #salt
18:11 bltmiller joined #salt
18:12 bltmiller quick question, is pillar data available in orchestrator SLS files?
18:13 LtLefse oh, I think I found the answer: "# Warning: Prior to changing this value, the minion should be stopped and all Salt caches should be cleared"
18:14 UtahDave bltmiller: yep!
18:14 bltmiller woohoo!
18:16 flowstate joined #salt
18:16 mikeywaites joined #salt
18:18 rhodgin joined #salt
18:19 woodtablet utahdave: yaaa, i am not alone or insane =D. btw the document for this demo looks awesome, i am trying to use to recruit my coworkers to be salty too, so much appreciated my friend
18:20 UtahDave woodtablet: I think it's how the salt-master daemon is getting started.
18:20 cableninja 'to be salty too'; you could just insult them if thats the goal :)
18:21 woodtablet double entendre's are always a plus ^_~
18:21 UtahDave woodtablet: try stopping the salt-master daemon.   service salt-master stop      then a salt-master daemon is still running.   then   pkill salt-master
18:21 UtahDave woodtablet: then  run    service salt-master start     and everything will work great
18:23 woodtablet whoa i have 3 running
18:23 woodtablet yaaaaa it works !
18:23 cableninja grrr why is this template not working :\
18:23 woodtablet utahdave: thanks, not sure why so many daemons were started
18:24 numkem joined #salt
18:24 UtahDave woodtablet: Yeah, I'm not sure either. still investigating
18:24 fannet joined #salt
18:24 lempa joined #salt
18:26 _JZ_ joined #salt
18:27 s_kunk joined #salt
18:31 flowstate joined #salt
18:31 barmaley joined #salt
18:35 rothsa joined #salt
18:37 mapu joined #salt
18:41 dendazen how do i downgrade package in salt?
18:42 dendazen trying to force a version
18:42 dendazen salt -G 'velio.environment:prod' pkg.install informatica-ums-6.8-1
18:42 punkoivan joined #salt
18:42 amcorreia joined #salt
18:42 dendazen like this, but salt skips
18:44 punkoivan joined #salt
18:46 bltmiller joined #salt
18:47 punkoivan joined #salt
18:49 punkoivan joined #salt
18:51 brew joined #salt
18:51 UtahDave dendazen: version='X.Y'
18:51 punkoivan joined #salt
18:51 nmadhok joined #salt
18:53 deniszh joined #salt
18:55 ajw0100 joined #salt
19:03 UtahDave woodtablet: what version of Vagrant did you have installed?
19:04 woodtablet UtahDave: Vagrant 1.8.1
19:04 UtahDave Ok, I was on 1.7.2.  I just upgraded to 1.8.1 and am testing again
19:05 woodtablet bbl lunch
19:05 tedski joined #salt
19:05 UtahDave It has something to do with vagrant.  If I bring up that master with  --no-provision  and then install following the instructions at repo.saltstack.com,  then it works just fine
19:05 trent__ joined #salt
19:10 t0nyhays joined #salt
19:16 cableninja I'm about to scream :\ this template has been giving me the worst problems, I'm not sure what I'm missing... http://pastie.org/10868315
19:17 mikeywaites joined #salt
19:20 babilen woodtablet: Could you give me the link to the pastebin with the error again?
19:21 UtahDave cableninja: so it looks like that globalinfo isn't getting set at the top of your file
19:21 babilen UtahDave: Do you have a link to woodtablet's problem?
19:22 UtahDave http://pastebin.com/VN5SYnjJ
19:22 bet0x joined #salt
19:22 UtahDave Hm.  woodtablet, after upgrading to 1.8.1 now it's working perfectly for me.
19:22 viq cableninja: how about {% if salt['pillar.get']('monitoring:global:enable:option_servicename', False) %} at the top?
19:23 swa_work joined #salt
19:23 ajw0100 joined #salt
19:23 UtahDave woodtablet: I'm going to get some lunch.
19:23 babilen ta
19:31 morissette joined #salt
19:32 Muchoz joined #salt
19:33 aharvey joined #salt
19:43 hasues joined #salt
19:43 hasues left #salt
19:47 heewa joined #salt
19:51 heewa cableninja: Yes. And you can use salt-call to try out jinja stuff without having to write state files, sync to master, run on minion, etc: salt-call cmd.run 'echo {% if 1 and 0 %}true{% else %}false{% endif %}' template=jinja
19:52 mikecmpbll joined #salt
19:52 heewa damn, I keep not noticing messages are from last time I had my laptop open, and think someone _just_ asked something
19:52 toanju joined #salt
19:53 heewa Question: the first salt minion run on a box fails for me because boto isn't installed, but it's needed for a template evaluation, so it fails at the template render stage, and won't get to the state that would have installed boto. How do people install the python dependencies of salt using salt?
19:55 nkuttler heewa: build it into your os image?
19:57 heewa I'd really rather try to find a salt-only way of doing it, and hopefully without an initial bootstrap phase. I'm hoping I can do my machine config with salt, who's job it is to do this, and not need a separate 2nd config management system to configure salt.
19:58 heewa Like, I recently started using the boto stuff, so it's a new pre-salt requirements. I'd hope to not have to change non-salt stuff every time I decide to use a new salt feature that needs packages installed.
20:00 nkuttler heewa: if salt is already installed, just have one state file that installs it and doesn't use boto?
20:00 nkuttler not sure if that would work though as i don't use other rendering engines
20:01 andrew_v joined #salt
20:01 bltmiller joined #salt
20:02 heewa nkuttler: I tried that, but salt first tries to render all the state files (which makes sense, so it can know what to run & require-dependencies), and fails there, before it ever tries to run any of the states.
20:02 nkuttler heewa: even with state.sls ?
20:02 UtahDave heewa: I'd create a state to install the other deps and have that run at startup time.  Then run your highstate
20:02 nkuttler right
20:02 GreatSnoopy joined #salt
20:03 heewa Ok, so on my machine boot/init/setup script I'd boostrap with salt, then run something like `salt-call state.apply my_bootstrap_state && salt-call state.apply` ?
20:03 mikecmpbll joined #salt
20:03 heewa Yea, that works. Then I wouldn't have to change my box setup script, just that my_bootstrap_state.sls file. Thanks!
20:05 edrocks joined #salt
20:06 manji joined #salt
20:07 UtahDave yeah, that should work.
20:08 fannet joined #salt
20:08 noraatepernos joined #salt
20:08 noraatepernos I’m new to salt and wondering where to debug KeyError: 'mysql_query.run' at the end of a “the minion function caused an exception” trackback.  Will this be in the mysql logs on the minion?
20:10 fannet joined #salt
20:10 iggy noraatepernos: key error usually means you're missing some dependency
20:10 iggy the mysql stuff specifically requires the mysqldb python module
20:11 fracklen joined #salt
20:11 spuder joined #salt
20:14 ahammond is there a reasonable way to get salt running on coreos?
20:14 ahammond or is that as ugly as it's looking?
20:18 woodtablet utahdave: back, really it works for you ? this is my 3rd time, and it fails like it did for you with multiple masters running
20:21 UtahDave woodtablet: Ah, interesting.  bumping up the ram to 1024 seems to solve the problem.
20:21 UtahDave for the master
20:21 c4rc4s joined #salt
20:22 woodtablet UtahDave: i tried more threads, and that didnt help. ahhhh i ll update my vagrant file
20:23 woodtablet utahdave: hmm... how do i do that in the vagrant file ? I dont see memory specified
20:23 woodtablet UtahDave: v.customize ["modifyvm", :id, "--memory", 2048]  ?
20:24 UtahDave just a second, I'm going to push my fix to a branch on that repo
20:26 UtahDave ok woodtablet,  https://github.com/UtahDave/salt-vagrant-demo/compare/moar_ram?expand=1
20:26 josuebrunel joined #salt
20:27 woodtablet UtahDave: nice =D, tahnks
20:27 ahammond with salt-cloud, I seem to remember there was a way to pass in a cloud-config.yaml for openstack based providers. But I can't find it again. I remember it being at the bottom of a web-page somewhere... Can anyone give me a pointer please?
20:27 UtahDave gtmanfred: ^^
20:33 bltmiller joined #salt
20:34 mohae joined #salt
20:36 cableninja_ joined #salt
20:36 c4rc4s joined #salt
20:36 sjorge joined #salt
20:37 c4rc4s joined #salt
20:38 c4rc4s joined #salt
20:38 cableninja__ joined #salt
20:42 Eureka703 joined #salt
20:42 ahammond uh oh. I might have confused this: -C --cloud-config with the ability to pass along a cloud-config.yaml file.
20:46 woodtablet utahdave: i destroyed my instances, incorporated your changes to the vagrant file, and it works, i ll add a comment to your pull request. Thanks!
20:46 subsignal joined #salt
20:47 Tyrm_ joined #salt
20:47 UtahDave cool
20:48 BadgerOps hey, missed your comment earlier woodtablet - but I see that @UtahDave hooked you up :)
20:48 UtahDave LtLefse: the hash default will change in a future release after a deprecation period
20:49 flowstate joined #salt
20:49 woodtablet hey badgerops, no worries, i felt better when someone was able to replicate my issue
20:50 LtLefse UtahDave: thanks. figured it was something like that
20:50 BadgerOps I'm wondering why it worked fine for me locally following the same steps you did. (shrug)
20:50 noraatepernos Still get a KeyError: ‘mysql_user.present’ sls: http://pastie.org/10868407
20:51 UtahDave woodtablet: i found another open issue referencing that problem.  changing the box to  hashicorp/precise64   also seemed to work. I'm testing that right now
20:53 noraatepernos Does it mean one of my requirements is not being met?  Completely shooting in the dark here haha.
20:53 UtahDave although, I don't really want to go back to precise...
20:54 cableninja UtahDave: I know that, but -why- everything is where it should be, and option_servicename/option_servername are being set, and properly.
20:54 cableninja sorry for the delay, I went for food
20:54 woodtablet BaderOps: It depeneded on your version of vagrant and your version of the trusty64
20:54 UtahDave noraatepernos: yeah, most likely
20:54 cableninja to recap: "<UtahDave> cableninja: so it looks like that globalinfo isn't getting set at the top of your file"
20:54 noraatepernos UtahDave: Is there a way to get salt to tell me which requirement isn’t able to be met?
20:54 woodtablet BadgerOps: it just ran out of memory so i am guess it was always close to running out, and it just did with the newer version
20:55 UtahDave cableninja: yeah, try running the exection module on the commandline
20:55 cableninja sorry, can you elaborate?
20:56 cableninja you mean do like 'salt-call _minion_ servers.monitoring ?
20:56 UtahDave oh, sorry, cableninja.  That was for noraatepernos
20:57 noraatepernos UtahDave: Heh I think you answered this here: http://stackoverflow.com/questions/30306693/using-saltstack-to-create-a-mysql-database
20:57 cableninja oh lol np
20:57 _JZ_ joined #salt
20:57 UtahDave cableninja: can you provide the url to your file again? I've lost it.
20:58 cableninja UtahDave: http://pastie.org/10868315
20:59 UtahDave cableninja: I don't see option_servicename in your pillar file
21:01 aharvey joined #salt
21:02 flowstate joined #salt
21:03 hopthrisC minion-restart on ping timeout is clearly not working
21:03 hopthrisC is somebody here familiar with tornado.stack_context?
21:04 shoemonkey joined #salt
21:05 UtahDave hopthrisC: sorry, I haven't really dug into that code at all
21:07 hopthrisC UtahDave: i wouldn't even know where to begin a bug report :)
21:07 hopthrisC i mean, apart from "it doesn't work"
21:08 abhi_ joined #salt
21:08 UtahDave hopthrisC: I think that's good enough to start with.  Just make sure to provide all pertinent configuration and version info
21:08 cableninja joined #salt
21:08 cableninja sorry UtahDave I fell off
21:08 cableninja didnt see anything after pasting my link
21:08 UtahDave cableninja: no problem. I do have a question for you
21:08 noraatepernos joined #salt
21:09 cableninja I probably have an answer :)
21:09 UtahDave I'm not seeing the  option_servicename in your pillar file
21:09 UtahDave if pillar['monitoring']['global']['enable'][option_servicename]
21:09 aharvey joined #salt
21:09 cableninja option_service name is passed as a default in the sls
21:09 UtahDave so that will not be true
21:09 hopthrisC UtahDave: the version doesn't seem that important. the bug must have been there for years (:
21:09 cableninja lines 76/77
21:09 edrocks joined #salt
21:10 UtahDave cableninja: sure, but   if pillar['monitoring']['global']['enable'][option_servicename]    is still going to be false
21:10 cableninja how so?
21:10 UtahDave because passing default data in your sls file isn't going to modify your pillar data
21:10 cableninja I know that
21:10 cableninja wait
21:10 cableninja what do you mean
21:10 UtahDave so that will resolve to false, and then   globalinfo will not get set
21:11 UtahDave and so that's why   you're getting the   Unable to manage file: Jinja variable 'globalinfo' is undefined
21:12 cableninja the option_server/service name gets set (and appears properly when it shows up in the service name /server comments
21:12 gtmanfred ahammond: i do not ever remember that, but we are are going to be able to use the /etc/openstack/clouds.yaml with the next release
21:12 gtmanfred ahammond: cause we are going to be using the openstack shade project, so checkout os-client-config for options
21:13 cableninja UtahDave, not sure that made sense, did you understand that?
21:13 UtahDave cableninja: I know, but in the sls file, not in your pillar data.  You aren't modifying your pillar data when you modify your state sls files.
21:15 cableninja I'm not trying to modify pillar data, I'm trying to read it
21:16 cableninja so that I can get the underlying lists/dicts
21:17 UtahDave cableninja: OK. let's walk through this. Maybe I'm missing something.   I'm looking at line 8.   Line 8 is going to be False.  Am I correct?
21:18 cableninja no it shouldnt be
21:18 cableninja look at the pillar
21:18 punkoivan joined #salt
21:18 cableninja monitoring:global:enable:(cpu)
21:18 cableninja is defined
21:19 UtahDave ah, crap.  I didn't notice that option_serviceaname didn't have quotes around it.
21:19 UtahDave ok
21:23 UtahDave cableninja: can you paste the output of   salt '<minion id>' state.show_sls servers.monitoring      ?
21:29 cableninja UtahDave, http://pastie.org/10868447
21:31 anotherZero joined #salt
21:32 tharkun joined #salt
21:32 cableninja UtahDave, I found one issue (a spacing issue with the defaults), but it didnt resolve the actual problem
21:42 Sylvain31 joined #salt
21:43 scsinutz joined #salt
21:44 rgrinberg joined #salt
21:48 nmadhok joined #salt
21:48 cableninja UtahDave, I circumvented the issue by looping the needed data into defaults entries
21:48 cableninja I think its dumb, but it works, so meh
21:51 noraatepernos Is it accurate to say that a KeyError occurs when a - require: references a package/sls/service that is not pkg.installed etc?
21:51 noraatepernos Chasing my tail now.  Officially
21:52 ageorgop joined #salt
21:52 noraatepernos Also, do I need to restart salt-master when my sls files change?  Guessing no but really have no idea why this keyerror is occurring now.
21:52 UtahDave noraatepernos: can you pastebin what you're seeing?
21:53 noraatepernos UtahDave: Thanks http://pastie.org/10868470
21:53 UtahDave No, you don't need to restart the salt-master when you change sls files. Only when you change the master config file
21:54 UtahDave noraatepernos: can you include the actual command you ran, too?
21:54 josuebrunel joined #salt
21:56 shoemonkey joined #salt
21:56 noraatepernos UtahDave: I’m running salt '*' state.highstate from the master.  I just have one minion.
21:58 UtahDave noraatepernos: Usually that means mysql_user.present couldn't be loaded because either the required python libraries or some dependency isn't satisfied
21:59 west575 joined #salt
21:59 woodtablet i have been introduced to cowsay, the world will never be the same
22:02 flowstate joined #salt
22:02 menglish1 joined #salt
22:05 noraatepernos UtahDave: Ok I think I got past it.  Working on “TypeError encountered executing state.highstate: object of type 'bool' has no len(). See debug log for more info.” now
22:06 rhodgin joined #salt
22:06 UtahDave noraatepernos: a good way to debug those kinds of things is to log into the minion, stop the salt-minion daemon, and then run the salt-minion in the cli in debug mode:      salt-minion -l debug
22:06 UtahDave you'll often see stacktraces and errors that way
22:06 fannet joined #salt
22:07 noraatepernos UtahDave: Awesome.  I didn’t know that.  I use events to drive everything (activemq or rabbit) so I’m used to doing that.
22:20 TyrfingMjolnir joined #salt
22:22 ahammond Hmm, so what's the Right Way to handle a CloudOS box?
22:24 aharvey joined #salt
22:24 amcorreia joined #salt
22:28 rhodgin joined #salt
22:31 UtahDave what is cloudOS?
22:33 ahammond UtahDave docker oriented distro.
22:33 ahammond UtahDave pretty decent design, actually. Especially if you're looking to run a whole bunch of docker instances and use something like Kubernetes or Mesos.
22:36 ahammond although, if you want to check out nifty new designs for docker host OSs, I think RancherOS is really cool. Brief version is that instead of systemd they're using docker, and there is only dockerized containers outside of kernel space.
22:37 UtahDave interesting
22:40 ahammond Anyone heard of a dockerized salt-minion? I was thinking that it would make a super slick solution to this problem.
22:41 ahammond We're very seriously looking at containerizing our salt-master with docker, but getting the minion into a container would be... awesome
22:43 tomasfejfar joined #salt
22:44 iggy docker isn't really well suited to running a minion alongside whatever else
22:44 iggy I mean there are hacky ways to run multiple things in a single docker instance, but that's not really the point of docker
22:44 iggy I'd look at something else like lxd for that
22:57 tracphil joined #salt
22:58 jhujhiti joined #salt
22:59 cwyse joined #salt
23:00 aharvey joined #salt
23:01 flowstate joined #salt
23:05 andrew_v joined #salt
23:08 hopthrisC UtahDave: it wasn't the async code, after all
23:08 hopthrisC UtahDave: i think…
23:08 hopthrisC UtahDave: https://github.com/saltstack/salt/issues/33841
23:08 saltstackbot [#33841]title: minion is not restarted on ping timeout | The code that should restart the minion when it can't successfully ping the master is never called....
23:09 UtahDave ah, interesting. It will be good to get that sorted out.
23:09 whatevsz joined #salt
23:10 bltmiller I have a question about orchestraters. I want to script out a rolling upgrade of each Elasticsearch node in a cluster. I'm targeting the ES cluster with a node group, and I only want to upgrade one at a time. As far as I can tell, `salt.state` takes the batch argument, but if my rolling upgrade has multiple steps (stop replication -> delete instance -> upgrade -> recover) that all need to happen in order, one minion at a time, I don't see any way
23:10 bltmiller within an orchestraters to make that happen. Am I correct in this analysis?
23:10 bltmiller yeesh, long message...
23:10 MikaT joined #salt
23:11 ageorgop joined #salt
23:14 bltmiller since my steps are (function -> function -> state -> function -> function), I could probably setup some requisites I'm thinking...
23:16 heewa joined #salt
23:20 * woodtablet listening to the answer to bltmiller's question =D
23:21 bltmiller heh this channel seems to get a little quieter into the evening hours. either that or I phrased my question poorly. brand new with using orchestration in salt, so I'm still trying to grasp the ins and outs :)
23:26 snc joined #salt
23:27 UtahDave bltmiller: you should be able to accomplish what you want.
23:28 UtahDave I'm kind of under a time crunch right now, so I don't think i'll be able to put together an example until tomorrow
23:28 bltmiller no worries! I typically idle in here, I'll try pinging you tomorrow :)
23:43 UtahDave bltmiller: cool!
23:54 punkoivan joined #salt
23:55 fannet joined #salt
23:58 whaity joined #salt

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