Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-05-04

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

All times shown according to UTC.

Time Nick Message
00:02 AdamSewell joined #salt
00:06 hasues joined #salt
00:06 hasues left #salt
00:08 jscott joined #salt
00:09 rem5 joined #salt
00:20 flowstate joined #salt
00:23 ajw0100 joined #salt
00:24 murrdoc joined #salt
00:27 dyasny joined #salt
00:30 spuder joined #salt
00:30 kus joined #salt
00:36 ajw0100 joined #salt
00:42 murrdoc joined #salt
00:42 flowstate joined #salt
00:42 spuder_ joined #salt
00:53 akio left #salt
00:55 ssplatt joined #salt
00:55 edrocks joined #salt
01:03 pipps joined #salt
01:06 iceyao joined #salt
01:06 spuder joined #salt
01:12 iceyao_ joined #salt
01:15 akitada joined #salt
01:18 bowhunter joined #salt
01:18 stone_BOI joined #salt
01:20 brotatochip joined #salt
01:23 stone_BOI left #salt
01:25 racooper joined #salt
01:28 sagerdearia joined #salt
01:28 lungaro joined #salt
01:29 spuder_ joined #salt
01:37 iceyao_ joined #salt
01:38 iceyao__ joined #salt
01:38 iceyao___ joined #salt
01:46 pipps joined #salt
01:46 pipps99 joined #salt
01:46 catpigger joined #salt
01:47 ilbot3 joined #salt
01:47 Topic for #salt is now Welcome to #salt! | Latest Versions: 2015.8.8, 2016.3.0rc2 | 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"
01:51 michelangelo joined #salt
01:52 mohae joined #salt
02:01 spuder joined #salt
02:05 quasiben joined #salt
02:11 hrumph yo i'm working with salt
02:11 hrumph with salt 2015.8 it's finally possible to install ms office with no msi file
02:11 hrumph that's really really neat
02:11 hrumph i can confirm its working
02:12 Castor__ joined #salt
02:13 hrumph but i have a question about best practices i can include certain configuration files and have the installer include them in the ms customization file and drop them into the place where i want them.......*OR* i can have salt drop the files in when the install is complete....
02:13 hrumph what would you guys reccommend?
02:13 hrumph you can do the same thing with registry settings
02:13 hrumph so should it be microsoft customization of should it be salt?
02:13 hrumph i mean
02:14 hrumph should it be microsoft customization or should it be salt that does the post install adjustments?
02:15 hrumph that's a tough call
02:15 hrumph you can make and argument for either one
02:16 hrumph ms installer customization or salt post install formulas....
02:20 meekrab joined #salt
02:24 mapu joined #salt
02:27 mapu joined #salt
02:28 meekrab joined #salt
02:36 estahn morning
02:36 hrumph hi eseyman
02:36 hrumph did you read my question?
02:36 hrumph what do you think?
02:37 estahn do you need to manage the configuration files afterwards? e.g. something changes in a configuration file and it needs to be rolled out to everyone?
02:38 estahn if not i'd pass it to the installer and be done with it.
02:38 hrumph ok i guess that's what i'll do
02:38 hrumph there won't be any changes needed
02:38 hrumph estahn, this is really awesome that this is working i'll mention this on github
02:39 jscott joined #salt
02:39 estahn :)
02:39 hrumph i'll spread the word because other people who don't know the power of salt might be interested
02:40 estahn question: salt.states.archive.extracted can use an s3 path (s3://test.tgz) as source. unfortunately i need to provide a hash. is there a reason salt can take the hash from the etag? or is it just not implmented?
02:40 hrumph estahn, i don't know hopefully someone else can help you
02:40 hrumph estahn, the salt devels have been good at answering github issues within about one business day
02:41 hrumph but bugfixes could take a long time
02:44 cyteen joined #salt
02:49 evle joined #salt
02:50 sjmh joined #salt
02:51 Gitaxis joined #salt
03:10 catpig joined #salt
03:16 kevinquinnyo joined #salt
03:16 raygunsix joined #salt
03:18 flowstate joined #salt
03:20 bevilaqua joined #salt
03:21 bevilaqua left #salt
03:23 vpm_ joined #salt
03:24 __alex joined #salt
03:24 xMopxShell joined #salt
03:26 kevinquinnyo in an orchestration, is it not possible to call one of my modules in jinja?  like:  {% set staging_site = salt['tss.get_staging_domain']({{ possibly_another_var_here }}) %}
03:26 kevinquinnyo because i can't seem to make that work
03:42 kevinquinnyo ohhh right.  orchestrate runner, as a runner, runs on the saltmaster, so it doesn't have _modules/ lazy loaded into it's namespace
03:42 kevinquinnyo dangit
03:42 subsignal joined #salt
03:43 kevinquinnyo i guess i'll have to make a gen-purpose utility module that the saltmaster can use for these kind of things
03:49 meekrab joined #salt
03:51 jfelchner joined #salt
03:59 mowntan joined #salt
04:01 keldwud joined #salt
04:08 ageorgop joined #salt
04:11 ssplatt joined #salt
04:16 keldwud joined #salt
04:19 flowstate joined #salt
04:29 h2osso joined #salt
04:33 hoonetorg joined #salt
04:51 godlike joined #salt
04:51 godlike joined #salt
04:51 pdayton joined #salt
05:01 meekrab joined #salt
05:06 sauvin joined #salt
05:12 ajw0100 joined #salt
05:13 rdas joined #salt
05:14 kevinquinnyo "It's always a typo" -- remember that kids
05:14 kevinquinnyo if you're banging your head against the wall and breaking out strace, and the like, consider that it might be a typo, yet again.
05:16 honestly And before that check with df if any disks have run full, because that causes the strangest errors
05:16 kevinquinnyo yeah i've seen that before
05:16 kevinquinnyo bad memory will cause weird errors too
05:18 flowstate joined #salt
05:18 honestly Bad memory tends to crash the system eventually, but yeah
05:19 honestly It's also not easy to rule out
05:20 kevinquinnyo yeah, you have to use the old "take out half", try to duplicate
05:20 kevinquinnyo take out the other half, try to duplicate
05:20 kevinquinnyo recurse, until you find the bad stick
05:21 kevinquinnyo and curse
05:21 kevinquinnyo a lot
05:21 honestly Or run memtest
05:22 kevinquinnyo or just run the code on another machine and take the bad one out back for some office-space printer treatment
05:23 hoonetorg joined #salt
05:27 honestly Yup
05:28 kevinquinnyo all i know is that i seem to never learn the lesson about the typo.  I will find myself questioning the framework I'm working on, or the underlying programming language, or computer science itself, or the laws of nature, before i stop and consider that there might be a typo that's causing me to chase red herrings.
05:28 kevinquinnyo every single time.
05:38 bbradley joined #salt
05:39 jeddi joined #salt
05:39 honestly That's human and the only way to fix it is designing your system in a way that makes typos cause syntax / type errors that make it obvious that the problem is a typo
05:40 honestly Most systems are terrible at that
05:43 kevinquinnyo it's usually when something looks very close to something else, so you look at it and think "There's no difference!" so you think you've ruled out typos
05:43 ninjada joined #salt
05:43 kevinquinnyo but it's do-something vs do_something
05:44 kevinquinnyo or do-somtheing vs do-something
05:44 honestly Yeah that bites me regularly
05:45 honestly Yaml isn't very good about telling you why parsing failed either
05:45 kevinquinnyo i really like the "Did you mean?" feature that things like git have
05:45 kevinquinnyo yeah it fails pretty hard
05:46 subsignal joined #salt
05:47 Niqui joined #salt
05:47 kevinquinnyo i actually avoid yaml completely -- i only use it for orchestration because *i think* (i haven't dug into the orchestrate runner code) that it reads your file line by line, so you can't use something like the py renderer (which i prefer)
05:47 Niqui left #salt
05:48 kevinquinnyo i've tried to use the py renderer with an orchestration and it failed as i expected
05:50 honestly salt-ssh behaves differently when using py renderer instead of yaml...
05:53 kevinquinnyo sometimes i wish salt didn't have all of the renderer abstractions, and you just had to extend core classes to do what you need, and define the output you want
05:54 kevinquinnyo the output format*
05:54 honestly I wouldn't mind the abstractions if they worked :)
05:55 kevinquinnyo yeah i fight with them
05:56 onlyaneg1 joined #salt
05:57 kevinquinnyo i feel like a lot of the time, i'm not fighting with the core of salt (which is solid), i'm fighting with the wizard behind the curtain who's providing me with all of these super high level abstractions and "easy-buttons"
05:57 kevinquinnyo if that makes sense
05:59 honestly Yup
05:59 honestly The core of salt is just a nicer way of running scripts on a remote host
06:00 dunz0r joined #salt
06:02 kevinquinnyo on *many* remote hosts
06:03 kevinquinnyo and it's async / jobs magic i certainly don't want to attempt to write myself
06:03 kshlm joined #salt
06:03 kevinquinnyo i'm also abusing the framework as i'm learning how it works.  A refactor of my code is on the horizon -- I can feel it.
06:04 kevinquinnyo but for now i'm just going to make it work
06:08 nidr0x joined #salt
06:10 ivanjaros joined #salt
06:14 zer0def joined #salt
06:18 flowstate joined #salt
06:19 slav0nic joined #salt
06:30 duncanmv joined #salt
06:34 brianfeister joined #salt
06:39 dober joined #salt
06:51 subsignal joined #salt
06:53 fracklen joined #salt
06:55 aw110f joined #salt
06:55 garphy joined #salt
07:00 toanju joined #salt
07:04 CeBe joined #salt
07:05 ronnix joined #salt
07:06 garphy joined #salt
07:08 justg joined #salt
07:11 fracklen joined #salt
07:11 ronnix joined #salt
07:16 dgutu joined #salt
07:18 flowstate joined #salt
07:18 ivanjaros3916 joined #salt
07:22 remyd1 joined #salt
07:22 aw110f_ joined #salt
07:24 Sylvain31 joined #salt
07:24 josuebrunel joined #salt
07:29 ninjada_ joined #salt
07:30 zer0def joined #salt
07:34 lero joined #salt
07:40 manji joined #salt
07:41 ninjada joined #salt
07:43 dmaiocchi joined #salt
07:47 Sylvain31 hi, is there a common wait to customize satl-bootstrap.sh notably to recreate the saltmaster, or rename minion hostname?
07:49 ashrov joined #salt
07:52 dmaiocchi joined #salt
07:53 kawa2014 joined #salt
07:56 Erik-P joined #salt
07:56 fracklen joined #salt
08:08 meekrab joined #salt
08:08 punkoivan joined #salt
08:10 Rumbles joined #salt
08:18 flowstate joined #salt
08:19 Chris____ joined #salt
08:19 rrei joined #salt
08:19 Chris____ What exactly is salt?
08:19 Erik-P when i try to target a minion i get "Minion did not return." from minions i didnt targeted
08:20 teatime Chris____: well it's pretty flexible and has grown to incorporate a lot of use cases
08:21 teatime but most people would probably say, a configuration management system similar in concept to puppet, chef, ansible
08:23 teatime Sylvain31: you can just properly configure your system hostname before bootstrap and that will be used for the minionid.  or if you want a custom one you can create the file /etc/salt/minion_id with your desired id in it before bootstrapping.
08:23 brianfeister joined #salt
08:25 Erik-P salt -G "os:Windows" service.status XXXX i get respond from all windows but i get Minion did not return. [Not connected] from all linux
08:27 GreatSnoopy joined #salt
08:28 Chris____ so saltstack would let me run my code from "another computer"? could I have it set on a schedule to run without my computer being on?
08:29 fracklen joined #salt
08:30 fracklen joined #salt
08:33 teatime I don't know enough about what you're doing to say.
08:33 teatime you don't need salt to execute stuff on a remote host, or to schedule stuff to run
08:38 stooj joined #salt
08:39 keimlink joined #salt
08:42 onlyanegg joined #salt
08:43 Sylvain31 teatime: thanks, I saw that I need to customize the minion_id, I just wanted to know if there is a "common" way to customize the bootstrap process.
08:45 evle joined #salt
08:50 ninjada joined #salt
08:52 Erik-p joined #salt
08:55 jvblasco joined #salt
09:01 garphy joined #salt
09:04 iceyao joined #salt
09:08 iceyao_ joined #salt
09:11 Erik-p i get jusnk respond from salt -G anyone know how to fix it?
09:16 favadi joined #salt
09:18 flowstate joined #salt
09:31 subsignal joined #salt
09:33 kbaikov joined #salt
09:39 dmaiocchi joined #salt
09:42 kbaikov joined #salt
09:47 linjan joined #salt
09:51 Rumbles hi, can someone give me an idea what this error means? I'm finding it a bit cryptic: "Requisite declaration nginx-config-test in SLS nginx.sites is not formed as a single key dictionary"
09:51 Rumbles this is the state: https://paste.fedoraproject.org/362452/
09:52 s_kunk joined #salt
09:57 Sylvain31 Rumbles: is the YAML valide?
09:57 deus_ex joined #salt
09:57 Rumbles I think so Sylvain31
09:59 yomilk joined #salt
10:01 Rumbles does it look valid Sylvain31 ?? https://paste.fedoraproject.org/362456/
10:01 Rumbles just wondering if reload-nginxwould owkr, I want it to watch the nginx-config-test section
10:01 Rumbles not sure if that's correct
10:01 Rumbles the rest looks right though :/
10:04 Sylvain31 Rumbles: cant help on that yet, I too new to salt, right now. I just given that hint… ;)
10:06 Rumbles ok, np, thanks anyway
10:08 liqw left #salt
10:11 garphy joined #salt
10:11 Rumbles think I figured it out
10:12 brianfeister joined #salt
10:12 dmaiocchi joined #salt
10:18 N-Mi joined #salt
10:18 N-Mi joined #salt
10:19 flowstate joined #salt
10:23 fracklen joined #salt
10:28 babilen I'd like to upgrade a bunch of jessie minions from 2015.5 to 2015.8.8, but am unsure how to best deal with the missing "KillMode=process" in the systemd service file (cf. http://repo.saltstack.com/#debian → Debian 8)
10:28 TomJepp joined #salt
10:28 babilen I guess I could just write a state that adds that line, but would appreciate input if anybody else did this beforehand
10:28 Knuta babilen: that
10:29 Knuta babilen: that's what I was thinking to do as well
10:29 fracklen joined #salt
10:32 akhter joined #salt
10:33 wmp joined #salt
10:33 babilen Knuta: Yeah, I'd simply be happy if somebody has some proven states for this :)
10:35 Knuta babilen: would have been nice to provide on the website, actually. It has to be a common usecase, and I'd expect the affected systems to be able to run salt states :-P
10:35 brianfeister joined #salt
10:35 babilen You sort of do, yeah
10:35 babilen I was a bit surprised by the lack of a salt state too
10:40 estahn joined #salt
10:44 fracklen joined #salt
10:45 cyborg-one joined #salt
10:45 Rumbles can anyone tell me how I can watch part of a state fle within a statefile? https://paste.fedoraproject.org/362469/
10:49 babilen Rumbles: That would be "watch: module: nginx-config-test"
10:50 babilen The reference is MODULENAME.{STATE_ID,NAME_ARGUMENT}
10:50 fracklen joined #salt
10:54 Rumbles thanks babilen
10:56 Rumbles any idea why I am getting the following error still? https://paste.fedoraproject.org/362471/
10:57 amcorreia joined #salt
10:58 babilen Rumbles: What is the error when you start that service on minion?
10:59 babilen You could also run the state locally on the minion with "salt-call -ldebug state.sls THE_STATE" (or state.highstate in lieu of state.sls)
11:04 ninjada joined #salt
11:17 flowstate joined #salt
11:19 Rumbles the service is running on the minion
11:19 Rumbles I did a reload (which is what that's trying to do) and it worked fine
11:20 catpig joined #salt
11:20 Rumbles I don't really understand what you're saying in your second comment babilen
11:21 babilen You can login to your minion and run the saltstate locally (which gives you more output)
11:21 babilen So, nginx is running?
11:21 babilen "the service is running on the minion" == "nginx is running" ?
11:21 gladiator joined #salt
11:21 Rumbles I just want it to reload the service
11:22 babilen I understand, but *is* nginx currently running on the minion?
11:22 Rumbles yeah, 1 master 4 workers
11:22 shiriru joined #salt
11:22 Rumbles checked with ps
11:22 babilen So "service nginx status" returns what?
11:23 Rumbles * nginx is running
11:23 babilen And what happens if you run the state locally on the minion? Could you paste the output of that?
11:25 Rumbles hmmmm I'm not doing this right
11:25 Rumbles my state file is called sites.sls it's in the nginx folder
11:25 Rumbles and the module is reload-nginx
11:25 Rumbles I tried: "sudo salt-call -ldebug nginx.sites.sls reload-nginx"
11:25 babilen So you are running "salt-call -ldebug state.sls nginx.sites" ?
11:25 Rumbles 'nginx.sites.sls' is not available.
11:25 Rumbles ah
11:25 Rumbles wrong way round
11:25 Rumbles 1 sec
11:28 fracklen joined #salt
11:30 Rumbles babilen, https://paste.fedoraproject.org/362478/
11:31 babilen Check the errors on line 210 and 211
11:33 subsignal joined #salt
11:34 mavhq joined #salt
11:34 babilen You could also run state.sls_id (which allows you to specify a state id such as "reload-nginx" -- https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.sls_id )
11:34 babilen (or the highstate)
11:35 averell joined #salt
11:35 jvblasco joined #salt
11:37 fracklen joined #salt
11:39 Rumbles I tried removing the 2 lines it was complaining about, and it tells me it failed to start the service again
11:40 Rumbles ah, I was being stupid, sorry, got it now :)
11:40 Rumbles thanks, the ldebug stuff is useful!
11:43 garphy joined #salt
11:44 babilen Rumbles: So .. what was it?
11:44 Rumbles so, I was hoping that this way of doing this would mean that the configtest would only be run if the files being watched changed. But t seems to run every time, regardless of whether the file changed: https://paste.fedoraproject.org/362480/
11:45 Rumbles I didn't name the service to be reloaded babilen
11:45 babilen heh
11:45 babilen So it tried to reload "reload-nginx" ?
11:45 punkoivan joined #salt
11:45 Rumbles yeah it does the reload every time according to the highstate results
11:46 punkoivan left #salt
11:46 rem5 joined #salt
11:48 Rumbles am I right in thinking that the configtest should only run if the files change, and the reload should only happen is the config test has run and been passed?
11:50 aqua^c joined #salt
11:57 Rumbles sorry, yeah I see what you mean babilen, it was trying to reload "reload-nginx" :)
11:58 Rumbles but now it appears to reload whether there has been a config change or not...
11:58 TyrfingMjolnir joined #salt
11:58 josuebrunel joined #salt
11:58 babilen You could use "onchanges" -- https://docs.saltstack.com/en/latest/ref/states/requisites.html
11:58 flowstate joined #salt
11:58 fgimian joined #salt
11:59 babilen But I'm not sure what you are doing exactly and a "service.running" shouldn't reload all the time. It sounds as if you want to make the syntax checker a require and define listen_in/watch_in on the service.running state for all files you want to watch
12:00 babilen If you watch the syntax checker it will reload every time as that state presumably has changes
12:00 deus_ex1 joined #salt
12:00 jab416171 joined #salt
12:03 Rumbles I want to run the configtest every time a config file changes, and if the config test passes I want the service to be reloaded...
12:05 quasiben joined #salt
12:05 CeBe1 joined #salt
12:08 babilen aye
12:09 Rumbles I tried setting both to onchanges, but even when the file changes "the req did not change"
12:09 babilen I think you want the setup I sketched above
12:17 onlyanegg joined #salt
12:17 flowstate joined #salt
12:18 teryx510 joined #salt
12:18 Rumbles okay, I'm trying to understand how to use watch_in, I started with this: https://paste.fedoraproject.org/362491/ I have changed it so the config files are watch_in the nginx-config-test https://paste.fedoraproject.org/362492/
12:18 Rumbles does that look right?
12:19 babilen You don't want onchanges on line 35
12:19 imanc joined #salt
12:19 Rumbles have that as a watch_in in nginx-onfig-test ?
12:21 babilen Ah, no .. that might work. Have you tried it?
12:21 Rumbles ran test=True using both
12:21 babilen Are you testing this *live* ?
12:22 Castor__ joined #salt
12:22 Rumbles yes, but it's a beta server, so I can break it wihtout someone shouting at me
12:22 honestly welldoitlive.gif
12:23 losh joined #salt
12:23 Rumbles when I ran it before with onchanges nginx-config-test wanted to run, even though the config files hadn't changes
12:23 babilen I still think that you might want to require the config-test in the service.running state and have watch_in: service: ... in the file ones
12:24 babilen You want to make sure that the nginx-config-test passed and don't care that that state had changes
12:24 kojiro joined #salt
12:24 pdayton joined #salt
12:25 Rumbles I don't think I understand what you mean
12:27 babilen Sorry, you will have to try different requisites and see which combination results in the behaviour you want. I can tell you that you do not want an onchanges on the syntax checker
12:29 Rumbles so using watch_in is wrong then?
12:34 subsignal joined #salt
12:34 ronnix joined #salt
12:36 pdayton joined #salt
12:38 pdayton1 joined #salt
12:38 CeBe1 joined #salt
12:39 ssplatt joined #salt
12:40 iceyao joined #salt
12:41 viq joined #salt
12:41 yomilk joined #salt
12:43 AdamSewell joined #salt
12:43 SpX joined #salt
12:46 akhter joined #salt
12:47 gh34 joined #salt
12:48 pdayton joined #salt
12:48 cyteen_ joined #salt
12:51 TooLmaN joined #salt
12:53 fracklen joined #salt
12:54 fracklen joined #salt
13:00 JohnnyRun joined #salt
13:01 toastedpenguin joined #salt
13:07 kojiro Is there a module for configuring an application through its web api?
13:10 sjorge joined #salt
13:12 flowstate joined #salt
13:13 dendazen joined #salt
13:14 pdayton joined #salt
13:19 fracklen joined #salt
13:20 Rumbles what application kojiro ?
13:20 cyteen_ joined #salt
13:20 kojiro Rumbles: gitlab.
13:20 babilen https://docs.saltstack.com/en/develop/ref/states/all/salt.states.http.html allows you to perform http queries -- https://docs.saltstack.com/en/latest/topics/tutorials/http.html
13:21 kojiro I want to preconfigure gitlab using salt, but I can't figure out how they want me to do it. Direct database manipulation seems to be the best option. (And that _can't_ be right.)
13:21 kojiro So I was looking at curl for a hail-mary
13:22 babilen https://github.com/saltstack-formulas/gitlab-formula exists, but I haven't used it yet
13:22 kojiro It doesn't really work. First, there's the problem of the postgres requirement that isn't met
13:23 babilen https://github.com/gepoch/salt-gitlab + https://github.com/bechtoldt/gitlab-formula + ...
13:23 kojiro the bechtoldt one seemed better, but I got a lot further using gitlab's own installer
13:23 kojiro The problem isn't installing it - it's configuring it after install
13:23 kojiro none of those formulas really tackle application settings.
13:23 babilen Their all-in-one-installer is rather cumbersome in that it packages everything and the kitchen sink
13:24 kojiro Yeah, and runs Puppet behind the scenes heh
13:24 kojiro My plan is to get a working application then iteratively improve my states until I am not using the omnibus installer anymore
13:24 kojiro but either way I want to preserve certain settings in states
13:29 perfectsine joined #salt
13:31 fracklen joined #salt
13:32 subsignal joined #salt
13:32 bowhunter joined #salt
13:35 estahn joined #salt
13:38 cpowell joined #salt
13:39 ivanjaros joined #salt
13:39 flowstate joined #salt
13:40 DammitJim joined #salt
13:40 cpowell joined #salt
13:41 numkem joined #salt
13:41 mapu joined #salt
13:46 racooper joined #salt
13:48 subsigna_ joined #salt
13:49 ifsp joined #salt
13:49 ZiLi0n joined #salt
13:52 ZiLi0n Hello everyone. I have started to use "onchanges" and I am seeing that when running the state.sls the output changes its format, there is a __run_num__ parameter in the output and the output order is not the same as in the sls file... it makes very difficult to follow the order of the tasks... any option to show the output properly? I would say it does not happen in all cases where onchanges is used....
13:53 tkharju joined #salt
13:57 ferbla joined #salt
13:58 dyasny joined #salt
13:58 Rumbles ZiLi0n, I think using onchanges you would expect things to run out of order, as a state can be run reactively when a state it was watching changes
14:00 _JZ_ joined #salt
14:00 mage_ any idea how could I run a state on machine B when state on machine A executes sucessfully?
14:00 fracklen_ joined #salt
14:00 zmalone joined #salt
14:01 mage_ should I use the Reactor system?
14:02 mage_ or the Orchestrate runner
14:05 simon_c joined #salt
14:05 meekrab joined #salt
14:05 jerredbell joined #salt
14:05 simon_c hi, what's the best way or reporting documentation bugs ? (this is with the getting started guide)
14:08 mage_ create a pull request on github?
14:10 simon_c even if I knew git and pull requests, it's not a simple fix. https://docs.saltstack.com/en/getstarted/fundamentals/states.html the screen shots and side bar refer to a state with two things init, but the state file only has one.
14:12 patarr joined #salt
14:15 favadi joined #salt
14:16 tharkun joined #salt
14:17 ZiLi0n Rumbles, thanks, yeah it makes sense, I though the output should continue being in the same order it is written though. Let's say the a state does not run because of onchanges, so the output order would be the same, just that state was not run
14:19 Rumbles simon_c, the page you pointed to has a text box that described 3 packages to install. the screenshots seem to match up with the text.... I don't see any mention of how many "thing" the state file contains....
14:19 Rumbles np ZiLi0n
14:19 ZiLi0n Rumbles, it changes the order a lot, not just a bit hehe it changes the order of all the whole state.sls, even the includes, and the output is like running task: 0 then 10 then 5 then 18, then 2... it is difficult to follow
14:20 meekrab joined #salt
14:20 mage_ shoud Ihttps://dpaste.de/7vT2/raw any idea ?
14:20 mage_ should I configure something on the master ?
14:21 simon_c Rumbles: the last image on the page has a section in the text  "update_lftpd_conf" which is nowhere in the state file., both say "successded: 2",where as when I run it in the demo system, it says successded:1, and the side bar talks about execution order, saying "packages are installed before teh configuration file is updated" when nothing in the state referes to a config file. Looks like the author tried to simplify the page, but didn't do the im
14:26 teryx510 joined #salt
14:26 Rumbles fair point :)
14:27 kaptk2 joined #salt
14:28 simon_c Rumbles: and is there a docs page for "how to contribute to docs" which explains all the git & pull request thing, that I see all over the place, (but which I've always ran away from as I'm not a developer ;) )
14:28 Rumbles hmmmm I think explaining git is outside the scope of salt documentation
14:29 jerredbell1 joined #salt
14:30 Rumbles I don't actually know how to contribute to the documentation myself
14:30 Rumbles looking in the repo there is a doc folder
14:30 simon_c Rumbles: at saltconf the other week, the chap doing the talk about the docs made it look about as easy as a wiki edit, but I couldn't even find what git repo to look at for reporting an issue with that page, never mind how to edit it.
14:30 Rumbles but it all looks like application documentation
14:31 Rumbles do you know who they are?
14:31 Rumbles maybe worth asking them :)
14:31 Sylvain31 salt.loaded.int.module.boto_route53][ERROR   ][6080] boto_route53 requires at least boto 2.35.0.
14:31 Rumbles afraid I'm not a dev, just another user
14:31 simon_c trying to remember :)
14:31 Sylvain31 as far as I can see its related to amazon s3
14:32 Sylvain31 can I configure salt to not output this error in minion log?
14:32 Rumbles route 53 is AWS DNS Sylvain31
14:33 Rumbles not sure about logging, but "pip install boto3" might fix
14:33 Sylvain31 I don't use AWS, can I disable it?
14:33 Rumbles not sure, I guess that's stuff for sal cloud
14:33 Rumbles salt cloud*
14:36 simon_c Rumbles: I think it was Jacob Hammos, does he hang out on here ?
14:36 Rumbles sorry, wouldn't know :)
14:36 simon_c np
14:36 tvinson lastlog ammos
14:37 Rumbles I guess he's ammos simon_c :)
14:38 tvinson no, i was just trying to be helpful and missed the /
14:40 hasues joined #salt
14:40 Brew joined #salt
14:40 Rumbles hah :)
14:40 Rumbles sorry
14:41 remyd1 Hi. Is there a way to test a module like test=True for states (dry run). For example to test system.reboot or something else ?
14:41 Sylvain31 Rumbles: no help yet with "pip install boto3" still the error message logged. on the minion but I'm not sure salt use the pip installed version
14:42 Rumbles maybe pip install boto
14:42 Rumbles boto is the old version which is still maintained
14:42 flowstate joined #salt
14:42 Rumbles I only use boto3 when writing scripts
14:43 Sylvain31 Rumbles: My salt wasn't pip installed but from the bootstrap-salt.sh from debian jessie package, with a boto package
14:44 Rumbles yeah but pip is pythons package manager, salt relies heavily on python, so installin via pip is probably the correct (only?) way to isntall boto
14:45 Rumbles it would make it available for use on the system you are running on anyway
14:47 flowstate joined #salt
14:47 jscott joined #salt
14:47 ageorgop joined #salt
14:48 Sylvain31 apt-get remove python-boto + restart both master and minion seems do the trick… boto3 still installed via pip
14:49 moy joined #salt
14:49 remyd1 otherwise you have whell or conda for python to manage packages installs
14:49 remyd1 wheel*
14:50 ZiLi0n Rumbles, there is requisite called "listen/listen_in" that executes at the end, so it does not change the order. However it is like "watch" so it always executed, and the behavior I need it is like "onchanges
14:50 spuder joined #salt
14:50 fracklen joined #salt
14:50 ZiLi0n "onchanges" so only execute if there are changes, but if it has to execute it can do it at the end, so it does not affect the order, that is what I am looking for... I haven't found a good solution :(
14:56 romanxfifty joined #salt
15:00 knine joined #salt
15:03 josuebrunel joined #salt
15:05 jvblasco joined #salt
15:05 romanxfifty Hi #salt! Whether there is a way to get current version of a package ? I want to delete some package,  but I need to specify exact version , so I executed this command :  salt 'node' pkg.purge pkg1 version= "salt['pkg.version']('pkg1')"  , but it doesn't work at all  :(
15:06 Miouge joined #salt
15:07 ohe_ joined #salt
15:12 impi joined #salt
15:18 merlehaggard I've added this to sensu.client service.running for 7.2 to attempt to use systemd and it still is trying to use update-rc.d
15:18 merlehaggard {% if grains['osrelease'] == '7.2' %}     - provider: systemd {% endif %}
15:18 merlehaggard getting CommandNotFoundError: update-rc.d
15:21 AdamSewell joined #salt
15:21 toastedpenguin joined #salt
15:27 estahn joined #salt
15:31 brotatochip joined #salt
15:32 honestly service has never worked properly for me
15:33 honestly Just use cmd.run
15:33 tristianc joined #salt
15:33 simon_c left #salt
15:39 Fiber^ joined #salt
15:45 RandyT joined #salt
15:47 Miouge joined #salt
15:48 rm_jorge joined #salt
15:49 dayid joined #salt
15:49 linjan joined #salt
15:50 hasues left #salt
15:54 dendazen hey guys can i do if in file.replace state? something like {% if salt['grains.get']('domain') == 'chi1' or  salt['grains.get']('domain') == 'aur1' %} then replace with one value else replace string in file with other value, where would if statement go to?
15:57 dayid joined #salt
15:59 DammitJim dendazen, I think you do it the other way around
16:00 DammitJim you create the state file
16:00 DammitJim and only assign it to the minion you want to target
16:00 DammitJim does that make sense?
16:00 DammitJim guys, correct me if I'm wrong
16:00 DammitJim the value thing, you can pass from pillar, though
16:00 DammitJim so, you define 'chi1' or 'aur1' in pillar where you have your if statement logic
16:01 DammitJim so, I think you were looking more for the pillar way of doing this
16:01 dendazen well, i mean the value gets assigned to all the minion, but just different based on which subodmain they are located in
16:01 bb74352 joined #salt
16:01 DammitJim dendazen, yes, you define this "subdomain" in pillar
16:02 dendazen i i have state like this for example
16:02 seblu joined #salt
16:02 dendazen https://gist.github.com/anonymous/308553eab6c0210a9d33146c6bbf838a
16:02 dendazen I wanted to do something similar for the config file i want to edit the value of parameter.
16:03 DammitJim dendazen, I think you can do that, but a more "elegant" way of doing it would be through pillar
16:03 DammitJim that's my take on it
16:03 DammitJim no if statements in your state, but the if statement goes in your pillar instead
16:04 DammitJim does that make sense to you?
16:05 DammitJim maybe you are just used to doing it in the state, but in the long run, you'll see it works better in pillar
16:06 cro joined #salt
16:07 iggy ZiLi0n: watch/listen aren't always executed, only when changes happen
16:08 iggy ZiLi0n: oh, I see what you're saying about the difference between the two
16:08 iggy ZiLi0n: sounds like a good candidate for a feature addition (that's how we got listen in the first place)
16:08 Vaelatern joined #salt
16:10 merlehaggard honestly: that seems to work for starting but I don't really like that, because now it thinks the service is running even when I manually stop it
16:11 merlehaggard so I'd have to set a service.status section also I assume
16:11 onlyanegg joined #salt
16:11 wych joined #salt
16:12 meekrab joined #salt
16:17 dendazen I can do it in pillar, i have some states which are done through the pillar
16:18 lero joined #salt
16:19 bltmiller joined #salt
16:20 lumtnman joined #salt
16:20 Aleks3Y joined #salt
16:21 chadhs joined #salt
16:22 chadhs does this have to be configured in the minion file? https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.mysql.html#module-salt.modules.mysql
16:23 pdayton joined #salt
16:23 chadhs can't seem to get the values to be used/picked up when i put it in pillar data instead
16:23 spuder joined #salt
16:25 gtmanfred chadhs: it does have to be in the minion config
16:25 chadhs @gtmadfred thnx. so is it best practice to have different minion files for prod vs dev if that's the case
16:26 chadhs @gtmanfred is there a way to securely store that mysql password?
16:26 gtmanfred no
16:26 gtmanfred you might be able to use the sdb
16:26 gtmanfred but in the end you could pass it on the command line
16:26 gtmanfred basically everything is stored somewhere
16:26 gtmanfred store it in /etc/salt/minion is only readable by root
16:26 gtmanfred so that is ... i mean, no worse than having it in pillars
16:26 gtmanfred which if you are the root user, you can call salt-call pillar.items and get the same thing
16:27 chadhs and just have different minion files for different envs?
16:27 chadhs with the appropriate mysql creds
16:27 gtmanfred so, it looks like that module hasn't been touched in a while
16:27 gtmanfred and needs to have some features added
16:28 gtmanfred what we did with the keystone module was to allow for pillar data to be used, and for the multiple profiles to be specified
16:28 gtmanfred that module would need some TLC before that could happen
16:28 gtmanfred s/that module/the mysql module/
16:29 chadhs gotcha
16:29 chadhs thanks for the info @gtmanfred, saved me more than an hour of messing around with it most likely
16:29 gtmanfred :)
16:29 chadhs i have it working via the minion file already
16:29 writtenoff joined #salt
16:29 chadhs was pulling my hair out saying "put i can see the pillar data on the CLI!!!"
16:30 chadhs ;-P
16:30 gtmanfred hehe
16:30 gtmanfred yeah, that module needs to be updated to use __salt__['config.get']
16:30 gtmanfred so that it can look in grains, pillars, minion config, master config
16:31 gtmanfred and also use the sdb so you don't have the password saved in the file, but can have it off in a remote database (still not more secure, but nice to not have it in a config file)
16:31 zmalone joined #salt
16:32 disbound joined #salt
16:35 disbound Can I use salt:// as the file path for runner_dirs ?
16:35 gtmanfred uhh, maybe?
16:35 gtmanfred i use it for my reactor files, so i would guess yes
16:36 disbound I figured so too but when reactor calls the runner it throws and exception saying 'custom.runner' is not available.
16:36 gtmanfred lame
16:37 gtmanfred in 2016.3, you will be able to put them in salt://_runner and run salt-run saltutil.sync_runners
16:38 zmalone I read that .8.9 is supposed to be tagged next week, is .3.0 doing okay?
16:38 disbound thanks gtmanfred
16:39 mortis isnt this how i schedule a highstate in the minion-config? (posting 4 lines! :o)
16:39 mortis schedule:
16:39 mortis highstate:
16:39 mortis function: state.apply
16:39 mortis minutes: 10
16:39 mortis doesnt seem to run a highstate
16:39 disbound I put that under pillar, but that is just me
16:40 mortis yeah, docs says it should work as pillar or in minion-config
16:41 pipps joined #salt
16:49 gtmanfred zmalone: we have like 2 more bugs we think, then rc3
16:49 bb74352 joined #salt
16:49 gtmanfred that is what we want to be the release
16:49 zmalone Thank you!
16:49 pdayton1 joined #salt
16:50 gtmanfred mortis: it should work in both minion config and pillar data
16:50 mortis gtmanfred: thought so too ... trying it as pillar now
16:51 gtmanfred did youi restart your minion after setting it in /etc/salt/minion?
16:51 mortis yeah
16:51 gtmanfred pillar will also require a salt-call saltutil.pillar_refresh
16:51 mortis checked the minion logs and they didnt run anything
16:51 mortis but that should happen on restart?
16:51 gtmanfred i don't know, it could be 10 minutes after it starts
16:51 mortis :)
16:51 gtmanfred https://docs.saltstack.com/en/latest/topics/jobs/#highstates your setting looks correct though
16:51 mortis i will run it anyway, to make sure :)
16:52 mortis yeah, it should work
16:52 flowstate joined #salt
16:53 mortis running refresh on all now
16:53 mortis nagios-alarms should tick in with OK shortly if it actually works :)
16:53 ZiLi0n iggy thanks very much! sorry I have just seen your reply right now. Yes, it would be great to have the same for onchanges
16:54 flowstate joined #salt
16:54 ZiLi0n iggy I think watch/listen are always executed, they add extra behavior (they execute again) is the watched state changes. That is different than onchanges, but onchanges does not have its respective "listen" (end of execution)
16:55 mortis im checking fileage of /var/cache/salt/minion/highstate.cache.p to verify if highstate has been ran
16:57 kojiro joined #salt
16:58 merlehaggard also salt 'minion-id' service.* sensu-client works as expected but for some reason sensu states are looking for the wrong provider
17:03 bltmille_ joined #salt
17:04 ageorgop joined #salt
17:05 Lionel_Debroux joined #salt
17:06 tmclaugh[work] joined #salt
17:07 rem5 joined #salt
17:09 iggy ZiLi0n: yeah, listen was added by the lyft guys a while back, they probably just didn't see a need for an onchanges version of the same... I'd say open a feature request
17:11 bltmiller joined #salt
17:11 ZiLi0n thanks iggy
17:12 ZiLi0n for your great feedback and help!
17:12 tmclaugh[work] Hi, I'm having an issue with file contents not being populated.
17:12 tmclaugh[work] https://github.com/threatstack/threatstack-formula/blob/fix-extra-args/threatstack/init.sls#L96-L97
17:12 rem5 joined #salt
17:12 tmclaugh[work] File's being created but the contents are empty
17:14 tristianc joined #salt
17:18 Tanta joined #salt
17:24 Ch3LL tmclaugh[work]: try contents instead of content
17:26 akhter joined #salt
17:26 GreatSnoopy joined #salt
17:30 AdamSewell joined #salt
17:33 tercenya joined #salt
17:36 brotatochip joined #salt
17:38 flowstate does anyone know if I can use a source file for the contents of a multiline string in a state file?
17:38 flowstate specifically, for the contents of my userdata script in my launch configuration?
17:39 flowstate wait, I could include it from an external file as a string using jinja... maybe?
17:40 cnk joined #salt
17:45 Idrot joined #salt
17:46 pipps joined #salt
17:46 mirko joined #salt
17:47 aw110f joined #salt
17:52 fracklen joined #salt
17:55 bluenemo joined #salt
17:56 josuebrunel joined #salt
17:58 blu_ joined #salt
17:59 Trauma_ joined #salt
18:02 ajw0100 joined #salt
18:02 spuder_ joined #salt
18:04 zmalone joined #salt
18:05 AdamSewell joined #salt
18:08 flowstate anyone here have experience including multiline strings with jinja?
18:08 flowstate I somehow need to bring it in at a certain indention
18:08 flowstate friggin yaml
18:11 dyasny joined #salt
18:11 AdamSewell joined #salt
18:11 kojiro flowstate: http://stackoverflow.com/a/21699210/418413
18:14 babilen flowstate: Could you give an example of what you want to do?
18:14 flowstate this actually describes it perfectly
18:14 flowstate https://github.com/saltstack/salt/issues/5480
18:14 saltstackbot [#5480]title: Multiline strings tricky when used with ``file.managed`` ``contents`` param | This one is best articulated with an example. In short, if the ``contents`` param of ``file.managed`` is a function call that returns a multi-line string the template indentation is wrong....
18:15 flowstate I'm trying to put the bootstrap script for my instances into another script, and then include it as a jinja variable, but the indentation is super wacky
18:19 kojiro Speaking of file.managed, I'm confused by the warning in the docs to use quotes around octal values for `mode`.
18:19 AdamSewell I had asked yesterday about some issues i've been having with my master timing out. I get the error that the master is not responding and to increase the worker_threads. Resources are not an issue and I've increased both resources and worker_threads. Someone had suggested to cehck the log of the minion and in doing so, the command is never getting to the minion apparently.
18:19 flowstate figured it out
18:19 flowstate I'll paste the solution
18:19 kojiro Speaking of file.managed, I'm confused by the warning in the docs to use quotes around octal values for `mode`. Modes _should_ be octal. And indeed, an unquoted octal value yields what I expect. So is the warning unnecessary?
18:22 flowstate http://pastebin.com/NvwicTEU
18:22 flowstate I've never quoted them
18:22 flowstate that's super strange
18:22 flowstate maybe I should be?
18:22 kojiro flowstate: er, why isn't that file just included with salt://path/to/file?
18:23 jfindlay kojiro: I think the quoting is to protect octal numbers from yamlyness, but I could be wrong
18:23 hoonetorg joined #salt
18:23 kojiro jfindlay: yeah, maybe an example would help. But I can't produce an actual problem, so I can't open a pr
18:23 kojiro jfindlay: like, all I can say is, "there doesn't seem to be a problem to warn about"
18:23 flowstate kojiro: because normally it expects a multiline yaml string
18:23 flowstate not a file
18:24 babilen https://docs.saltstack.com/en/latest/topics/troubleshooting/yaml_idiosyncrasies.html#integers-are-parsed-as-integers
18:24 flowstate so I'm injecting that multiline string via jinja, so that all my instances get the same userdata script
18:24 flowstate because remembering to set them retroactively was driving me nuuuuuts
18:24 babilen kojiro: ^^ that's the link discussing the issue
18:25 kojiro babilen: aha, so it was fixed, that's why I can't reproduce it.
18:25 flowstate that would certainly explain it
18:26 kojiro It was confusing me, because modes _should_ be octal, and I would've expected the instructions to say something like "either use quotes _or_ a leading zero, or know what a mode is"
18:26 kojiro Or maybe just "use symbolic modes"
18:26 nethershaw joined #salt
18:27 rem5 joined #salt
18:32 jfelchner joined #salt
18:33 pipps joined #salt
18:34 s_kunk joined #salt
18:36 flowstate joined #salt
18:36 [Kernel_Panic] joined #salt
18:36 [Kernel_Panic] joined #salt
18:39 teryx510 joined #salt
18:39 bb74352 joined #salt
18:40 ponpanderer joined #salt
18:42 akhter joined #salt
18:44 akhter joined #salt
18:48 M-liberdiko joined #salt
18:50 M-liberdiko joined #salt
18:52 M-liberdiko joined #salt
18:52 linjan joined #salt
18:52 M-liberdiko joined #salt
18:55 AdamSewell joined #salt
18:56 ageorgop joined #salt
18:56 bltmiller joined #salt
19:00 brotatochip joined #salt
19:00 bVector anyone used pepa for pillar? I see an explanation of how to use it but no mention of why its useful
19:02 GreatSnoopy joined #salt
19:04 akio joined #salt
19:05 akio im trying to use a grain as a variable in a config file and i don't know how i should do it properly, its working but it leaves brackets and single quotes around the rendered config
19:06 akio tried syslocation {{ salt['grains.get']('location') }} and syslocation {{ grains['location'] }}
19:06 akio both work but they leaev the brackets and single quotes
19:06 akio syslocation ['location']
19:07 bltmiller joined #salt
19:07 bVector akio: is the grain a list or a string?
19:07 akio string
19:07 bVector sounds like its a list, you could try salt['grains.get']('location')[0]
19:07 merlehaggard can I create a pillar to set a grain value, ie set a role based on minion-id using pillar data
19:08 akio oh well i think the next line style might make it a list
19:09 bVector yeah you would be better served making sure its interpreting it as a string, but indexing into the list would work short term
19:09 akio bVector: TIL, ty sir
19:09 bVector merlehaggard: if you're using config.get, that treats everything under opts, grains, pillar as the same thing
19:09 akio so yeah it was a list
19:10 bVector akio: np :D
19:11 flowstate joined #salt
19:11 bVector merlehaggard: if you're using pillar to set a grain, commonthink first reaction would be "perhaps you're approaching it wrong"
19:12 bVector salt is a deep beast, theres probably a way of getting what you want without taking that specific path, if you want to expound on what you're trying to accomplish
19:12 garphy joined #salt
19:12 merlehaggard ok so best to target the minions to mod the /etc/salt/minion file and restart client
19:12 merlehaggard fair enough
19:14 bVector /etc/salt/grains even
19:14 bVector or drop something in /etc/salt/minion.d/mycoolgrains.conf to be even more modular
19:15 zmalone joined #salt
19:15 merlehaggard yes reading that now, I want to set a role but yes same principles
19:15 merlehaggard thx
19:15 bVector but what I was getting at initially, is that maybe you could just use pillar without touching the grains
19:15 AdamSewell joined #salt
19:15 bVector what is the end result you're looking for
19:17 flowstate joined #salt
19:18 merlehaggard well I want roles, ie also to be able to target role based minions from the server or other
19:18 bVector e.g. if you're using grains in jinja templates like {{ grains['configvar'] }}  you could instead use something like {{ salt['config.get']('configvar') }} and that would work for a value set in grains, pillar, or opts
19:19 merlehaggard in this specific case I need to target a certain minion role for my autofs formula
19:20 Fiber^ joined #salt
19:20 merlehaggard so in the top.sls file
19:21 bVector hmm, I havent dealt with a case for targeting, you could definitely do a compound matcher like: salt -C 'I@role:web or G@role:web' test.ping
19:21 bVector that would match anything with pillar or grain set to role: web
19:23 merlehaggard so 2 options, create a pillar named 'role' or mod the minion config and add an standard role grain
19:24 ageorgop joined #salt
19:24 ajw0100 joined #salt
19:24 merlehaggard it seems like salt bootstrap should have a switch to create roles during install
19:26 AdamSewell joined #salt
19:27 DammitJim joined #salt
19:28 jvblasco joined #salt
19:30 merlehaggard I'm also looking at nodegroups for some of what I need but my setup can't find nodegroups
19:31 ajw0100 joined #salt
19:37 pipps joined #salt
19:42 mapu joined #salt
19:47 bltmiller joined #salt
19:48 Jimlad joined #salt
19:52 AdamSewell joined #salt
19:53 punkoivan joined #salt
19:54 losh joined #salt
20:00 brotatochip joined #salt
20:02 tharkun joined #salt
20:03 toanju joined #salt
20:03 mapu joined #salt
20:04 AdamSewell joined #salt
20:05 manji joined #salt
20:14 subsigna_ joined #salt
20:15 AdamSewell joined #salt
20:16 subsign__ joined #salt
20:17 pipps joined #salt
20:19 manji joined #salt
20:21 viq joined #salt
20:25 pipps joined #salt
20:26 pipps99 joined #salt
20:27 AdamSewell joined #salt
20:36 kevinquinnyo joined #salt
20:36 jfelchner joined #salt
20:38 rem5 joined #salt
20:39 ahammond is there a __mine__ for execution modules, or would I use __salt__['mine.get']() ?
20:40 iggy yes
20:40 ahammond merlehaggard for targeting roles, nothing beats nodegroups
20:40 ahammond iggy yes there's a __mine__? :)
20:41 jfelchner_ joined #salt
20:41 kevinquinnyo i think he meeant yes, use __Salt__['mine.get']()
20:41 kevinquinnyo the list of dunder dictionaries avaialble are listed here:  https://docs.saltstack.com/en/latest/topics/development/dunder_dictionaries.html
20:42 ahammond thanks kevinquinnyo  :)
20:45 manji joined #salt
20:45 flowstate joined #salt
20:48 brianfeister joined #salt
20:49 brianfei_ joined #salt
20:51 AdamSewell joined #salt
20:52 ahammond is there any chance salt-mine emits events... and hope beyond hope, events which talk about changes to the data?
21:00 bowhunter joined #salt
21:00 pipps joined #salt
21:02 jvblasco joined #salt
21:04 manji joined #salt
21:04 AdamSewell joined #salt
21:05 pipps joined #salt
21:10 pipps joined #salt
21:10 brotatochip joined #salt
21:14 shiriru joined #salt
21:19 subsignal joined #salt
21:21 AdamSewell joined #salt
21:23 RandyT joined #salt
21:23 bVector ahammond: I see salt mine events a lot, let me check
21:26 bVector hrm, all I see are job returns from __mine_interval without the mine data
21:27 bVector but my mine data isnt changing
21:42 devops joined #salt
21:47 LtLefse_ joined #salt
21:51 cliluw joined #salt
21:53 cliluw joined #salt
22:10 Garyx joined #salt
22:19 tpaul joined #salt
22:21 viq joined #salt
22:22 tpaul In what file should I set config options for returners (e.g. couchdb.url)? Settings are being ignored in the master config.
22:22 bVector are you doing master returns on minion returns?
22:23 bVector *or minion returns
22:23 tpaul bVector: master returns
22:24 bVector so you have master_job_cache: couchdb in master config
22:25 tpaul ^yes
22:26 bVector looking at the source, the returner will log the following message as DEBUG 'Using default url.' if it cant find couchdb.url in salt.returners.get_returner_options
22:33 bVector tpaul: http://pastebin.com/P4LGFycZ
22:34 bVector er, add a print to line4 if you save that as a file, otherwise just paste into python repl
22:36 bVector https://www.irccloud.com/pastebin/6R2nY7cf/
22:37 bVector thats what I get for my redis return, if the module itself isnt picking up those values it would be a bug in the module
22:37 zenlot6 joined #salt
22:38 pdayton joined #salt
22:41 bVector or something else is overwriting those config values
22:42 pdayton1 joined #salt
22:45 tpaul bVector: It appears I have some hardcoded values to /etc/salt/master, I'm investigating those now
22:50 Gitaxis joined #salt
23:05 devops_ joined #salt
23:06 teryx510 joined #salt
23:08 hasues joined #salt
23:09 Garyx joined #salt
23:16 hasues left #salt
23:17 west575 joined #salt
23:17 flowstate joined #salt
23:19 ahammond bVector thanks. I'll probably trigger off the __mine_interval job. Not loving it, but... it'll work.
23:21 brianfeister joined #salt
23:24 ahammond is the targeting functionality accessible in execution modules? If so, any examples on how to use it?
23:25 kevinquinnyo1 joined #salt
23:33 gtaylor left #salt
23:34 blu__ joined #salt
23:37 favadi joined #salt
23:45 nZac joined #salt
23:49 \ask joined #salt
23:49 estahn joined #salt
23:50 ajw0100 joined #salt
23:51 flowstate joined #salt
23:54 ssplatt joined #salt

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