Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-08-23

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

All times shown according to UTC.

Time Nick Message
00:03 xbglowx joined #salt
00:05 subsignal joined #salt
00:08 ninjada joined #salt
00:09 pipps joined #salt
00:09 ffredrikk joined #salt
00:10 ralish left #salt
00:12 infrmnt joined #salt
00:17 akhter joined #salt
00:18 edrocks joined #salt
00:28 iamtew joined #salt
00:32 UtahDave joined #salt
00:42 armonge joined #salt
00:46 ffredrikk joined #salt
00:46 vodik joined #salt
00:58 lorengordon joined #salt
01:05 edrocks_ joined #salt
01:15 shadoxx is there a way to autoprovision hosts based on a reverse lookup against dns?
01:15 shadoxx the whole auto adding to salt thing is throwing me through a loop
01:16 John_Kang joined #salt
01:26 tsia joined #salt
01:27 ajw0100 joined #salt
01:30 amcorreia joined #salt
01:37 UtahDave shadoxx: what exactly are you trying to accomplish?
01:45 mikecmpbll joined #salt
01:48 sp0097 joined #salt
01:54 catpigger joined #salt
02:00 quasiben joined #salt
02:02 rome_390 joined #salt
02:11 skeezix-hf joined #salt
02:28 stephanlooney joined #salt
02:29 bastiandg joined #salt
02:41 Vaelatern joined #salt
02:43 ajw0100 joined #salt
02:55 badon joined #salt
03:02 edrocks_ joined #salt
03:08 Vaelatern joined #salt
03:08 edrocks joined #salt
03:09 akhter joined #salt
03:10 amy_ joined #salt
03:17 Vaelatern joined #salt
03:26 subsignal joined #salt
03:33 evle joined #salt
03:38 majikman joined #salt
03:45 onlyanegg joined #salt
03:47 stephanl_ joined #salt
03:48 ffredrikk joined #salt
03:54 kshlm joined #salt
03:59 stephanlooney joined #salt
04:03 dyasny joined #salt
04:15 k_sze[work] Is it possible to make salt output sorted by minion name?
04:16 k_sze[work] e.g. if I do `salt 'node*' cmd.run 'ffmpeg -version'`, I get the results in random orders (depending on the order in which the minions finish)
04:17 ffredrikk joined #salt
04:19 k_sze[work] looks like it's an open issue on github.
04:20 M-liberdiko joined #salt
04:26 stephanlooney joined #salt
04:30 iggy k_sze[work]: maybe output to json and see if `jq` can do what you want?
04:32 jerredbell joined #salt
04:36 pipps joined #salt
04:42 Vaelatern joined #salt
04:47 DEger joined #salt
04:47 stephanlooney joined #salt
04:49 morissette joined #salt
04:58 juanito joined #salt
05:04 lompik joined #salt
05:09 ffredrikk joined #salt
05:11 edrocks joined #salt
05:16 garthk joined #salt
05:20 John joined #salt
05:26 hlub_ joined #salt
05:29 necronian joined #salt
05:29 M-MadsRC joined #salt
05:29 ageorgop joined #salt
05:33 stephanlooney joined #salt
05:35 UtahDave joined #salt
05:42 ffredrikk joined #salt
05:44 bocaneri joined #salt
05:46 sybix joined #salt
05:54 stephanlooney joined #salt
06:02 ivanjaros joined #salt
06:02 iceyao joined #salt
06:03 ageorgop joined #salt
06:07 colttt joined #salt
06:08 jxm_ joined #salt
06:11 meirw joined #salt
06:12 meirw I'm having an issue with include
06:13 meirw I have 3 sls files (that setup redis on 3 different ports).  I need to set a variable at the top.  I'd like to make 3 files that simply include a base, but it seems the include fires first, and I need it to go last.
06:14 meirw https://gist.github.com/anonymous/0f9463e55f5e99be1a1589df3974689d
06:15 stephanlooney joined #salt
06:19 pipps joined #salt
06:20 meirw this is the base file: https://gist.github.com/anonymous/9949cd2d219091aeec47ea8ffaebc20e
06:22 jaybocc2 joined #salt
06:33 stephanlooney joined #salt
06:34 SpX joined #salt
06:36 seena joined #salt
06:37 seena Hi Guys
06:37 seena I am facing an issue with the reactor
06:37 seena I have two  reactor sls file to run on minion startup
06:37 seena 1) Execute high state  2) Deploy some app
06:38 seena Both of them are working individually , I have checked and confirmed
06:38 seena However When I am adding both
06:39 seena return:                 - The function "state.sls" is running as PID 24692 and was started at 2016
06:40 seena my highstate is not getting executed and getting the above error
06:40 alexhayes joined #salt
06:40 seena Looks like its not allowing to execute multiple state file at the same time
06:40 seena Any solution ?
06:43 DEger joined #salt
06:45 seena @AndreasLutro Any help
06:48 t0m0 joined #salt
06:48 iggy meirw: use jinja includes/macros? or extend?
06:49 iggy seena: run an orchestrate from the reactor that in turn runs one and then the other
06:50 meirw @seena - I will try extend
06:50 saintpablo joined #salt
06:52 stephanlooney joined #salt
06:53 meirw not sure I understand how that would work - I want to combine my variable declaration with the base.  WOuldn't extend be used to overwrite a declaration?
06:53 seena @iggy thanks I will try tha
06:54 UtahDave seena: have the highstate kick off the app deploy
06:54 cheus joined #salt
06:55 seena app deploy getting executed without any issue  , but highstate getting stuck
06:55 babilen Make it part of the highstate?
06:56 babilen (if you don't want to use orchestrate)
06:59 seena Ok thanks , don't want to run as part of  highstate , let me explore orchestrate
07:02 babilen Any particular reason why? It being part of the highstate sounds perfectly adequate
07:03 alexhayes joined #salt
07:05 ninjada_ joined #salt
07:05 Brijesh2 joined #salt
07:06 raspy joined #salt
07:06 Herugrim joined #salt
07:06 watersoul joined #salt
07:06 CeBe joined #salt
07:07 hillna joined #salt
07:07 fracklen joined #salt
07:08 Herugrim goodmorning people, i have a question, salt-call key.finger --local on a minion gives me the md5 sum of the public key, salt-key -f $minion on the master gives me a very long value, 128 bytes
07:08 Herugrim did i do a misconfig or is there a bug?
07:08 LordOfLA joined #salt
07:08 Herugrim salt 2016.3.2 on both minion and master
07:09 norii joined #salt
07:10 fracklen joined #salt
07:11 ivanjaros3916 joined #salt
07:12 stephanlooney joined #salt
07:14 edrocks joined #salt
07:19 ninjada joined #salt
07:27 amy_ joined #salt
07:30 Electron^- joined #salt
07:31 stephanlooney joined #salt
07:36 fracklen joined #salt
07:40 manji joined #salt
07:46 keimlink joined #salt
07:47 fracklen joined #salt
07:49 tsia joined #salt
07:52 seena Guys I need to run the highstate on the machine where the minion is  getting started/restarted. But in orchestrate sls file I need to mention -tgt . Ideally this should be passed by the reactor to orchestrate sls
07:53 alexhayes joined #salt
07:55 Brijesh1 joined #salt
07:55 krymzon joined #salt
08:00 lero joined #salt
08:04 mikecmpbll joined #salt
08:06 pipps joined #salt
08:08 Rumbles joined #salt
08:08 fracklen joined #salt
08:14 Qlawy Can I use value returned from custom salt module in jinja template?
08:15 iggy {% sure = salt['module.tryit']('foo') %}
08:16 fracklen joined #salt
08:17 haam3r joined #salt
08:17 velikan joined #salt
08:19 Qlawy iggy: yeah, just forget about ()
08:19 Qlawy :D
08:19 Qlawy thank you
08:19 haam3r Hi! Can anybody tell me how to specify both familys in an iptables state declaration? Or am I going to have to duplicate all of my iptables rules for both familys?
08:19 GreatSnoopy joined #salt
08:26 ninjada joined #salt
08:32 iggy jinja loop(s)?
08:33 s_kunk joined #salt
08:33 richardl sounds like a brand of muffin
08:33 fracklen joined #salt
08:37 stephanlooney joined #salt
08:40 ravenx joined #salt
08:43 haam3r okey…was thinking about that as well but hoped for something simpler in the states.iptables…I'll try it out.
08:43 haam3r thanks
08:56 dunz0r Can I do something like "if user exists: do stuff" in a state?
08:56 dunz0r I've got machines where the name of a user differs, and I'd like to try to make it try different usernames.
08:57 fracklen joined #salt
08:57 stephanlooney joined #salt
08:59 jab416171 joined #salt
09:02 tsia joined #salt
09:03 fracklen joined #salt
09:05 majikman joined #salt
09:08 fredvd joined #salt
09:09 meirw @iggy - I got it working with a jinja import
09:09 meirw i meant include
09:09 meirw thank you for the help
09:10 lero joined #salt
09:11 lero joined #salt
09:12 pipps joined #salt
09:12 lero joined #salt
09:17 edrocks joined #salt
09:17 stephanlooney joined #salt
09:18 fracklen joined #salt
09:23 felskrone joined #salt
09:24 jaybocc2 joined #salt
09:29 fracklen joined #salt
09:33 aalmenar joined #salt
09:35 stephanlooney joined #salt
09:44 fracklen joined #salt
09:48 keimlink joined #salt
09:54 badon joined #salt
09:56 alexhayes joined #salt
10:00 fracklen joined #salt
10:05 JohnnyRun hi all. I'm using this formula https://github.com/saltstack-formulas/graphite-formula that need to run a command on the host before highstate.
10:05 JohnnyRun the command is:
10:05 JohnnyRun salt 'graphitemaster' grains.append roles monitor_master
10:05 JohnnyRun I really don't like it. Can I set a grain in the sls ?
10:06 babilen You can .. but then the formula really shouldn't force you to use grains for that
10:06 babilen grains.append is a state
10:07 JohnnyRun ah ok.. so i user grains.append in the sls, right?
10:08 babilen https://docs.saltstack.com/en/latest/ref/states/all/salt.states.grains.html
10:09 JohnnyRun tnx
10:09 babilen But then ... I guess the formula uses that grain for targeting, doesn't it?
10:10 JohnnyRun Yes. But at the moment I need to targeting via hostname match, so setting the grain in the sls should be fine
10:13 fracklen joined #salt
10:13 lempa joined #salt
10:19 fracklen joined #salt
10:26 ninjada joined #salt
10:34 fracklen joined #salt
10:34 stephanlooney joined #salt
10:50 Brijesh2 joined #salt
10:52 teryx510 joined #salt
10:54 stephanlooney joined #salt
10:55 scc I want to do a match for something like if "*dev*web*" in grains['id'] in pillar, but this does not seem to work
10:56 scc Examples I've found would point in the direction of using a custom module for loading fnmatch and using that, but is it really not possible to do a match like this in the basic jinja tools available?
11:13 pipps joined #salt
11:15 stephanlooney joined #salt
11:20 edrocks joined #salt
11:21 keimlink joined #salt
11:25 jaybocc2 joined #salt
11:26 alexhayes joined #salt
11:35 stephanlooney joined #salt
11:35 colegatron sure someone knows how to archive this: my master and minions are on AWS. when a state is run in a minion I need to attach a EBS volume BEFORE the action is run. The minion has no rights to run the ec2AttachVolume api call, so should be done by the master (it will have enough rights). How can this be done with salt?
11:36 kidneb joined #salt
11:37 colegatron with a reactor?
11:43 colegatron oooh, just discovered orchestration.
11:44 colegatron new toy to play with :)
11:45 teryx510 Morning all. Is it possible for processes running as a non-root user to use the minion event bus to fire events to the master?
11:49 colegatron as far as I know you can run the minion only with the user you configured for it. once you set a non-root user to run it, you'll be able to fire whatever with it
11:51 teryx510 Ah, of course.
11:52 teryx510 Thanks colegatron, apparently I need more coffee.
11:56 quasiben joined #salt
11:56 colegatron wecome
11:56 dijit we, come!
11:56 colegatron here we go!
11:57 colegatron you're welcome! :P
11:58 stephanlooney joined #salt
11:59 west575 joined #salt
12:01 welker joined #salt
12:04 welker hi guys, a quick question: what does the "_" mean in state and module names. The documentation writes the names with underscore the run seems to ignore them sometimes and the states ignore them again. Is there a concept behind that or does one have to learn all the different variants of the names?
12:08 kidneb joined #salt
12:09 hlub_ welker: give an example?
12:09 welker hlub_: salt.modules.mac_defaults module
12:10 welker Docu: salt.modules.mac_defaults
12:10 welker cli: salt '*' macdefaults.read com.apple.CrashReporter DialogType
12:10 welker I did not figure out how to address in in a state statement
12:12 welker and mac_shadow, too
12:13 cyborg-one joined #salt
12:13 fracklen joined #salt
12:15 edrocks joined #salt
12:18 numkem joined #salt
12:21 nicksloan joined #salt
12:21 hlub_ wrong version of the docs maybe?
12:22 welker 2016.3.2
12:23 welker the difference is in the same document
12:24 edrocks joined #salt
12:24 welker e.g. https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.shadow.html
12:24 stephanlooney joined #salt
12:25 patrek joined #salt
12:36 antpa joined #salt
12:36 amcorreia joined #salt
12:36 keimlink joined #salt
12:37 ssplatt joined #salt
12:39 edrocks joined #salt
12:41 gh34 joined #salt
12:44 keimlink joined #salt
12:49 DEger joined #salt
12:49 krymzon scc: you should be able to use any python function, e.g. in my case I do {% if grains['id'].startswith('en') %} I'd guess regexp is available, haven't tried though
12:54 sxar joined #salt
12:55 DEger joined #salt
12:56 welker joined #salt
12:58 scc krymzon, regex is not available by default in jinja
12:58 scc too bad, but can be done with custom module and linking that to be usable on master
12:59 krymzon as far as I understand here you are calling python functions, not just what is in jinja
12:59 krymzon even without the regex module you should be able to do it with standard substring functions
13:00 phx joined #salt
13:00 N-Mi joined #salt
13:00 N-Mi joined #salt
13:00 scc Sure, but regex is not available as a method on string
13:01 scc I guess it may be possible to use .find(), but it didn't look it it would take a match pattern
13:01 pppingme joined #salt
13:02 StolenToast So I would like to run highstate and if state X is run then after the whole highstate I want to reboot the system.  Can this be done?
13:02 scc So just ended up doing salt['regex-re_search('.dev.web.', grains['id']) with "regex.re_search being the custom module and method
13:02 StolenToast Nothing needs to happen after reboot, one step installs a new kernel and I want it to be loaded
13:03 scc *salt['regex.re_search']
13:04 StolenToast hmm, I guess I could just reboot at the end of that state.  the system will run highstate again when it boots up and finish everything else
13:05 StolenToast oh but that will leave me with some ugly unfinished highstates at first...
13:07 tiwula joined #salt
13:08 stephanlooney joined #salt
13:08 scc StolenToast, I think the network state is able to reboot on certain settings. Maybe you can find some inspiration in there.
13:08 krymzon scc: I see, as far as I know there is no better built-in method available yet, I would sure like to be wrong
13:09 subsignal joined #salt
13:09 scc krymzon, yup, I would like some regexy thing to be available by default as well :)
13:09 evidence hmm 'Source hash http://thing.com/cgi-bin/webconfig?hostname=xxx&amp;os=FreeBSD&amp;hash=True format is invalid. It must be in the format <hash type>=<hash>, or it must be a supported protocol: salt, http, https, ftp, swift, s3, file'
13:09 evidence this recently broke in 2016.3.0 .. not sure what it doesn't like about the url
13:12 DEger joined #salt
13:12 krymzon scc: there is 'match' module, have you looked into it? I don't know if you can call it from jinja though
13:12 lero joined #salt
13:12 Tanta joined #salt
13:13 impi joined #salt
13:13 scc krymzon, nope. But I should have, it seems!
13:14 scc that should be available from jinja via i.e. salt['module.compound']
13:14 krymzon that's what I'd hope, it would be very useful this way
13:15 pipps joined #salt
13:15 scc I'll keep that in mind for the next time I need do targeting and filtering on minion id :)
13:16 krymzon if you try it, please let us know if it works :) could be very helpful
13:16 ivanjaros joined #salt
13:17 zer0def uh, quick question: how do i enforce ordering without depending on whether the dependency state failed or not? was thinknig aobut using `require` and `onfail`, but it might get in each other's way
13:18 west575_ joined #salt
13:18 krymzon scc: I guess it would work that way, correct me if I'm wrong: salt['module.compound']('L@cheese,foo and *', grains['id'])
13:20 antpa joined #salt
13:22 anotherZero joined #salt
13:22 akhter joined #salt
13:24 akhter joined #salt
13:25 jaybocc2 joined #salt
13:28 kshlm joined #salt
13:35 akhter joined #salt
13:41 akhter joined #salt
13:47 subsignal joined #salt
13:47 jhauser joined #salt
13:48 keimlink joined #salt
13:49 jhauser_ joined #salt
13:52 welker joined #salt
13:57 TooLmaN joined #salt
14:00 jerredbell joined #salt
14:01 pppingme joined #salt
14:02 stephanlooney joined #salt
14:03 impi joined #salt
14:04 jhauser joined #salt
14:05 armonge joined #salt
14:05 racooper joined #salt
14:06 phx anyone has at hand how can I query the renderred pillar tops?
14:07 raqua joined #salt
14:07 ssplatt pillar.items?
14:08 phx doesn't seem like it
14:09 ssplatt well, that will give you all of your pillar data.
14:09 raqua Hi all
14:09 welker joined #salt
14:09 raqua I have this simple pillar: http://pastebin.com/6zxUKwyS
14:09 phx i know, but what i want is not any pillar data, but the pillar top.sls rendition
14:09 ssplatt although there is something strange with the way it doesn’t render out things from our defaults.yml. something with our map.jinja that we use in our formulas
14:10 raqua pillar.get replacement works fine when run from vagrant/masterless salt, but does not work when run from regular master
14:10 raqua Do you know why please?
14:11 ssplatt becuase referencing pillar in pillar is a dangerous game
14:11 ssplatt basically a race condition
14:11 jhauser joined #salt
14:12 raqua I would expect race condition in multiple pillars, but why in the same ?
14:12 raqua is there any workaround then ?
14:12 ssplatt you either need to use jinja and {% set myvar = “versionstring” %}  or use grains, or external pillar.
14:13 ssplatt or render out the url in states
14:13 ssplatt using apache-tomcat-{{ version }}.zip
14:13 ssplatt raqua: because it can’t pull pillar before it’s rendered
14:13 ssplatt it doesn’t go back and render it twice each run
14:14 raqua this example can be solved easily but I have other, more complex pillars where that is not an option
14:14 raqua when it can not pull a pillar before it is rendered, then why it works reliably in vagrant for a year?
14:15 raqua and reliably fails in real world :)
14:15 ssplatt because of the way that master vs masterless works i’d suspect.
14:16 ssplatt but basically you are coding in bad habbits
14:16 pipps joined #salt
14:16 lovecraftian joined #salt
14:16 lovecraftian joined #salt
14:16 ssplatt and should move away from that
14:16 Rumbles joined #salt
14:17 tapoxi joined #salt
14:18 jerredbell1 joined #salt
14:18 ssplatt in this exact example, i’d just render the url in the state file. so -src: https://apache.org/downloads/apache-{{ version }}.zip   then you’d just have   version: .2.2.2  in pillar.
14:19 _JZ_ joined #salt
14:21 raqua Unfortunatelly, my use case is a bit more complex than that. But thanks for your help anyway. I have to thing about how to solve this a little bit more then.
14:22 canci joined #salt
14:24 teryx5101 joined #salt
14:24 mapu joined #salt
14:26 ssplatt raqua: you may want to look into setting variables in map.jinja to get the desired effect
14:28 stephanl_ joined #salt
14:29 welker joined #salt
14:29 armonge joined #salt
14:31 raqua looks interesting
14:32 jaybocc2 joined #salt
14:34 mikecmpb_ joined #salt
14:37 notCalle joined #salt
14:38 tapoxi joined #salt
14:39 DEger joined #salt
14:40 akhter_1 joined #salt
14:43 Brew joined #salt
14:48 akhter joined #salt
14:49 hasues joined #salt
14:50 spuder joined #salt
14:50 impi joined #salt
14:53 west575 joined #salt
14:53 JPT joined #salt
14:53 martin_hynar joined #salt
14:55 JPT Hello :-) Is there a common way to fire salt events as a non-root user on a minion? We have a local python script that is supposed to feed events into the salt event system using the salt Caller.
14:57 JPT I found hints about changing the salt minions file system permissions using an additional group, but i hope there is a better way.
14:57 ssplatt maybe a beacon? but in general, salt things must be run as root, or a user with proper privs
14:57 tapoxi JPT salt-api maybe?
14:57 jaybocc2 joined #salt
14:57 ssplatt or yeah, salt api.  make a webhook
14:58 ssplatt some info here http://bencane.com/2014/07/17/integrating-saltstack-with-other-services-via-salt-api/ , used that as the basis for my webhooks
14:59 JPT Thank you :) I will look into it. I still would prefer something that hooks into the locally available salt-minion. A beacon would be hard to integrate into the software that shall fire the events later on.
15:00 tapoxi JPT maybe you could also write an engine
15:01 tapoxi basically just a long-running script started by the minion that listens for events from your script(s)
15:02 tapoxi then fires them onto the bus
15:02 JPT Sounds interesting - but isn't that almost like a beacon? :)
15:03 tapoxi haven't written a beacon, but I think those are yaml. engines are python
15:03 JPT Aah, okay that is a huge difference then.
15:03 tapoxi so you could create your own simple API for your scripts to use
15:04 JPT That way, my engine is providing whatever is neccesary and whenever my third party software pokes it with some data, it then can fire it through the salt event bus
15:05 JPT And i don't need to expose the salt-api and set up credentials for a script
15:05 tapoxi exactly
15:06 JPT That sounds really good. I will have a go at it. Thanks a lot for your help! :-)
15:06 tapoxi you're welcome
15:09 hasues joined #salt
15:09 hasues left #salt
15:10 Electron^- joined #salt
15:11 west575_ joined #salt
15:12 fyb3r joined #salt
15:14 pipps joined #salt
15:14 kaptk2 joined #salt
15:17 ePrVRSBBhG joined #salt
15:18 pipps joined #salt
15:22 DmZDsfZoQv joined #salt
15:28 akhter joined #salt
15:28 fyb3r1 joined #salt
15:31 akhter joined #salt
15:34 DEger joined #salt
15:36 armonge joined #salt
15:39 autofsckk joined #salt
15:40 antpa joined #salt
15:41 akhter joined #salt
15:42 anotherZero joined #salt
15:42 dendazen joined #salt
15:43 UtahDave joined #salt
15:48 keimlink joined #salt
15:51 UtahDave morning, everyone!
15:53 raspado hi UtahDave
15:53 raspado hopefully someone can answer a mystery for me, I have a salt state with a cron job, its fairly simple example: http://pastebin.com/viTVriGm , I cannot find this cron anywhere on the minion, how exactly is this cron working?
15:54 raspado looks like it calls a function in a python script from _modules_
15:56 twork_ ok, nobody here has ever done this i'm sure, but you can imageine... that sinking feeling when you've just hit 'enter' and realize you forgot to leave off 'test=True' at the end; or otherwise, "i just told salt to do a bad, bad thing." <ctl-c> doesn't save one from a big long bad salting.  anything...?
15:56 heaje joined #salt
15:58 ronnix joined #salt
15:58 welker left #salt
15:59 twork_ (i've managed to reduce my incidence of that from once very 2-3 days or so, down to once every, oh, 2-3 weeks or so, but still, it's something i'd really like to have the antidone in my kit.)
15:59 stanchan joined #salt
15:59 egil raspado: well, it's a module, written in python
16:00 Electron^- joined #salt
16:00 raspado egil: do the minions then call the python script from /var/cache/salt/*
16:02 amy_ joined #salt
16:02 raspado egil: one thing i dont understand is where is this cron stored? I cannot find it in any crontab for the users on the host
16:02 fyb3r1 twork_, I doubt there is anything that can be done, as once you hit that enter button the message is blasted out to all the minions instantly.
16:03 stephanlooney joined #salt
16:03 fyb3r1 short of telling them all to cancel the job by jobID
16:04 fyb3r1 but even then I do not know if that instantly kill execution
16:04 twork_ fyb3r1: something like that was what i figured was my last ditch chance
16:04 twork_ "ok, user, you might be screwed but we'll make a swing..."
16:05 egil raspado: I might be wrong, but I imagine you are telling salt to run that specific command or job at minions defined in your grains
16:05 egil https://docs.saltstack.com/en/latest/ref/states/all/salt.states.cron.html
16:06 raspado egil: thx, just dont know where this cron is being added
16:06 raspado i see it in var cache but its not on any user or in cron.d
16:06 sp0097 joined #salt
16:06 egil its not being stored on the minion
16:07 fyb3r1 twork_, lol yeah unfortunately I believe that would be your only hope. Id have to dig through the source code to see how it handles incoming jobs to see if its possible to kill it mid execution or not though.
16:07 raspado oh? does this mean that the minion is calling the salt-master to run the modules every 6 hours?
16:10 twork_ i suppose the Salt Way<tm> would be more like, "save your states often, and if you send one out that you don't like, back up to the last one and send it out right quick."
16:10 raspado egil: i think this read will help me understand this better https://docs.saltstack.com/en/latest/ref/states/all/salt.states.schedule.html
16:10 egil again, I might be wrong, I don't know the module particularly - but the cron module executes states from the master
16:11 egil that might be correct
16:11 twork_ ...which is what i end up doing anyway even if when i have had a minion window open and killed salt there, or whatever
16:11 bltmiller joined #salt
16:15 onlyanegg joined #salt
16:16 bltmiller joined #salt
16:16 UtahDave twork_: salt-run jobs.active  (get the job id)   salt \* saltutil.kill_jobs <job id >
16:16 twork_ UtahDave: bless you.
16:17 UtahDave that should be    saltutil.kill_job
16:17 UtahDave you're welcome!  I've had to do that myself
16:17 twork_ close enough, i'd have figured that out... soon enough, one way or another...
16:18 UtahDave raspado: Yeah, that's not a cron job you're looking at.  That's a "Salt Scheduler" state.  https://docs.saltstack.com/en/latest/ref/states/all/salt.states.schedule.html#salt.states.schedule.present
16:19 UtahDave So there's no actual cron job.  That state is telling Salt itself to run that job on the schedule you set.
16:23 * Sketch wonders why there's no pillar module for csv
16:23 Sketch i guess i could import into sqlite and the use the sqlite module
16:28 amcorreia joined #salt
16:30 rm_jorge joined #salt
16:32 brotatochip joined #salt
16:34 SpX joined #salt
16:37 ronnix joined #salt
16:37 Brijesh1 joined #salt
16:40 ferbla joined #salt
16:43 antpa joined #salt
16:44 raspado UtahDave: can I call a specific minion to list its jobs?
16:44 raspado I see salt-run jbos.list_jobs but looking to target a specific minion
16:45 ferbla Hello everyone
16:45 anotherZero joined #salt
16:45 UtahDave Sketch: it wouldn't be too hard to create a csv pillar module.
16:45 UtahDave Sketch: actually, you might be able to have jinja import it into a regular pillar
16:45 pipps joined #salt
16:46 bltmiller joined #salt
16:46 ferbla I wrote a python script to take snapshots in our VMware environment, but it errors out. If I run only the part of the code that snapshots a single VM it works.
16:46 __number5__ joined #salt
16:46 ferbla Here is the link to my gist. https://gist.github.com/Ferbla/0b9a78a0759824e7051c2adb64919a2b
16:46 ferbla Thoughts?
16:47 impi joined #salt
16:47 UtahDave raspado: salt <minion id >  saltutil.running
16:47 raspado thx
16:48 woodtablet joined #salt
16:49 UtahDave ferbla: can you add a gist of you calling the command that fails as well as the output?
16:49 stephanlooney joined #salt
16:50 tiwula joined #salt
16:52 onlyanegg joined #salt
16:52 tapoxi joined #salt
16:54 DammitJim joined #salt
16:55 ferbla UtahDave: Yes, here is the error gist https://gist.github.com/Ferbla/2d0f38b4871fdbd54ca7f07717c16381 and here is the command I run https://gist.github.com/Ferbla/92ddd057c415aa5405ba945f2edab9c5
16:59 lilvim joined #salt
17:00 SpX joined #salt
17:03 akhter joined #salt
17:05 SpX joined #salt
17:05 sp0097 joined #salt
17:07 akhter joined #salt
17:07 edrocks joined #salt
17:07 nikolassilva joined #salt
17:08 UtahDave ferbla: Hey, so I don't have time today to help troubleshoot your python script, but I would look at the following:
17:08 s_kunk joined #salt
17:09 stephanlooney joined #salt
17:09 UtahDave ferbla: first, when you're calling your script on the command line I would put single quotes around any argument that has an asterisk in it.  To use your example:
17:09 UtahDave python /root/salt_scripts/snapshot_datacenter.py '*dev' glob
17:10 UtahDave Your shell will expand the '*dev' to be a list of filenames that end in 'dev'.   That might be messing you up right there.
17:10 jaybocc2 joined #salt
17:11 ferbla UtahDave: Okay, no worries. When I do that I still get the same error. I will keep digging though
17:11 UtahDave Second, i would simplify your script to make sure you have a function that does what you want correctly.  Then another function that will collect the list of VMs and call the first function with the action you want to take.
17:11 UtahDave make sure the second function provides a proper list of vm names to call
17:12 UtahDave OK.  sorry, I couldn't help more today on that.
17:13 ferbla Yeah, I will start with the base function and work my way out to see what I am missing. No worries, thanks for your help
17:13 pipps joined #salt
17:17 beardedeagle joined #salt
17:18 hardwire joined #salt
17:19 hardwire Ahoy!  I'm using salt for some system automation and provisioning but I'm very interested in using another instance of it on the same hardware for use in web application workflow processing.. trying not to muddy up a single instance.
17:20 hardwire Anybody found a way to do this, or possibly found a way then found a better way?
17:20 hardwire I'm guessing there's a wonderful philosophy of how to properly handle task delegation along with system automation
17:21 tapoxi hardwire what sort of workflow processing?
17:22 wendall911 joined #salt
17:22 hardwire in this situation we have a web application server acting as a process dispatcher that reaches out to internal and external nodes that attempt to reach in to a resource and audit it.
17:23 hardwire We're prepping an audit tool for a large enterprise that individual sysadmins can use to check their framework/systems from inside/outside and check credentials for SNMP access from multiple points without having to do a lot of work.  Audited data collection.
17:23 UtahDave hardwire: have you tried running the salt master in an LXC container on the same box?
17:24 hardwire that seemed like a proper, yet possibly unnecissary, method.
17:24 hardwire and some of these boxes are winders based.
17:25 hardwire only for some corner case access requests.
17:29 krymzon joined #salt
17:30 Tyrm joined #salt
17:30 stephanlooney joined #salt
17:37 hardwire UtahDave: Thanks tho.. we are just seeing where things fit at this point and how to utilize existing infrastructure and configuration to suppliment our needs.  I'm going to peek at some minion-to-minion communication pathways.  I had done a bit of that way back in the day when performing LUKS operations on secured gateway equipment and haven't touched it since.
17:39 UtahDave hardwire: I'm still not clear on exactly what you're trying to accomplish.  Are you wanting to run 2 instances of the salt-master on one server?
17:39 hardwire I was mostly just using syslog and database tailing (mongodb) back then.. having the master push return JSON around.
17:39 west575 joined #salt
17:39 hardwire UtahDave: perhaps the more broken down requirement we are trying to achieve is to have a clear path of JSON RPC between minions.
17:40 hardwire Ideally without making a request on a master.
17:40 UtahDave You can configure the salt master to allow minions to send commands to other minions, through the Salt Master.
17:40 hardwire hot dog!
17:41 UtahDave otherwise you can have the salt master configure each of the minions to communicate with eachother through whatever tool you wish to use.
17:41 hardwire yeh.  the latter is what I'd done in the past
17:41 hardwire Just had applications use a different database which formed a redundant (as in redundant) communication requirement.
17:41 sp0097 joined #salt
17:42 UtahDave https://docs.saltstack.com/en/latest/ref/peer.html
17:42 hardwire oh nice
17:42 UtahDave The peer interface lets a minion send a command to other minions, but it's not a JSON RPC interface.
17:42 hardwire yeh this is new since I'd used salt last ;)
17:43 hardwire UtahDave: what sort of return are you looking at with this?
17:43 UtahDave ;) it came out in Salt 0.9.0.
17:43 UtahDave hardwire: what do you mean?
17:43 hardwire This allows publishing commands, but the return result just heads back to the master?
17:44 DEger joined #salt
17:44 UtahDave Oh no.  The calling minion gets the return.
17:44 UtahDave The return does go through the master for security and authorization purposes.
17:44 hardwire sure no big deal
17:45 hardwire but it's not specifically JSON formatted like you'd expect from a master salt call?
17:46 Lionel_Debroux joined #salt
17:47 tuxx joined #salt
17:49 stephanlooney joined #salt
17:54 Pulp joined #salt
17:55 GreatSnoopy joined #salt
17:58 west575_ joined #salt
18:01 armonge joined #salt
18:03 mikecmpbll joined #salt
18:03 west575 joined #salt
18:04 tapoxi what's the goal of spm? to move away from the current fork & git clone from the formulas repo?
18:07 iter joined #salt
18:08 west575_ joined #salt
18:11 iter I have a cifs mount that gets remounted every highstate because I have "opts: rw,credentials=/root/creds" in the statefile
18:11 iter latest master/minion from ppa
18:12 iter I do see extra_ignore_fs_keys as a workaround but wonder if other people haven't solved this
18:12 west575 joined #salt
18:19 amy_ joined #salt
18:20 armonge joined #salt
18:21 howdythere joined #salt
18:21 dendazen joined #salt
18:22 tapoxi joined #salt
18:25 iggy tapoxi: you'd have to ask someone with SSE... that's something they've been pushing almost exclusively
18:26 tapoxi enterprise?
18:26 iter ok for posterity I fixed it by adding hidden_opts: credentials=/root/creds
18:28 ronnix joined #salt
18:28 iggy tapoxi: yeah, that's my abbreviation for the company behind saltstack
18:28 pipps joined #salt
18:28 tapoxi ahh ok
18:30 pfallenop joined #salt
18:30 pfallenop joined #salt
18:32 pipps joined #salt
18:34 orionx joined #salt
18:34 fracklen joined #salt
18:36 stephanlooney joined #salt
18:37 UtahDave hardwire: you can use whatever outputter you want.  If you want json just add   `--out json` to your call
18:37 UtahDave tapoxi: yeah, spm will allow you to package and deploy salt-formulas
18:38 UtahDave iggy: that's changing soon, actually
18:39 tapoxi @UtahDave is the plan for salt to run a repo?
18:39 UtahDave tapoxi: https://repo.saltstack.com
18:39 tapoxi @UtahDave no for SPM
18:39 lovecraftian joined #salt
18:40 orionx how do you trigger a mine.update from a state?  i tried this with module.run, but i don't get the expected results.  it seems to run ok, but when i reference the mine later in the sls file, i don't see the updated data.
18:40 tapoxi right now I'm forking formulas and using gitfs from my fork, but spm sounds way cleaner
18:41 UtahDave tapoxi: I believe so. I haven't been involved much in that project so I don't know if there's a timeline for that or not
18:42 tapoxi ok thanks
18:42 UtahDave orionx: The problem you're running into is that Salt runs all the sls files through the jinja renderer first, then runs everything through the state compiler, so likely everything is rendered before you've updated the mine data
18:43 amy_ joined #salt
18:44 orionx UtahDave: yep, that would explain it.  any thoughts on how i could add an interface IP to a host and then get that IP added to the network.ip_addrs mine all in the same run?  i was thinking a workaround might be to just use a reactor to add the IP as soon as the minion registers, then do a mine.update, then run highstate... thoughts?
18:44 antpa joined #salt
18:45 orionx UtahDave: i was trying to avoid reactors for this application, but will use them if needed
18:46 pipps joined #salt
18:46 hardwire UtahDave: thanks.  It looks like I'd most likely want to use saltnado on a master anyways and use http clients from my applications due to some missing async requirements in the salt client classes (according to the docs) where minions can't do async or queued command calls.  Doesn't operate well with the tornado workers.
18:46 hardwire -> far far away from keyboard
18:46 west575_ joined #salt
18:47 UtahDave hardwire: gotcha.   salt-api is awesome
18:47 UtahDave orionx: Yeah, that you could have reactors do that for you. Or you could have a startup state that made sure the IP was added and then ran a mine update
18:48 amy_ joined #salt
18:50 orionx UtahDave: by "startup state", do you mean a state that is triggered by a reactor when the minion starts or is there another way to execute a startup state?  thanks
18:52 armonge joined #salt
18:55 LeProvokateur joined #salt
19:00 raspado what is the best way to restart the salt-master service? I added another environment in /etc/salt/master
19:07 catpig joined #salt
19:09 noob joined #salt
19:09 UtahDave orionx: no, there's an option in the master config called startup_state.  You can have it run a highstate or a particular state that you want.
19:09 UtahDave So you could have it run the function to add the IP and the mine.update
19:15 joshin joined #salt
19:16 tapoxi joined #salt
19:16 alexhayes joined #salt
19:19 orionx UtahDave: oh cool.  i can't find a reference for this for master config, only minion config.  i can try it out, but do you know if it should work on master?  (I can't find a reference to this option in either the master or minion config file docs - https://docs.saltstack.com/en/latest/ref/configuration/minion.html, https://docs.saltstack.com/en/latest/ref/configuration/master.html)  i do see it in the minion config file example though (https://doc
19:19 orionx s.saltstack.com/en/latest/ref/configuration/examples.html#example-minion-configuration-file)
19:24 ronnix joined #salt
19:27 pipps joined #salt
19:29 pipps99 joined #salt
19:32 dendazen joined #salt
19:36 stephanlooney joined #salt
19:38 teco joined #salt
19:39 TooLmaN joined #salt
19:40 ruxu joined #salt
19:41 t0m0 joined #salt
19:43 raspado is there a way for file_roots to be called from gitfs based on each environment?
19:44 keimlink joined #salt
19:45 raspado in my file_roots, I have defined dev: - /srv/salt/dev but I want the files from my git repo in /dev to be the file system for /srv/salt/dev
19:49 bltmiller joined #salt
19:52 ruxu joined #salt
19:57 stephanlooney joined #salt
19:57 Tyrm_ joined #salt
19:59 giany hi, I'm using this top.sls file : http://pastebin.com/agzQf1YB , thing is that I also use the gitfs fileserver_backend and the docker.mon recipe is called first..is there a way to reorder or prioritize recipes in gitfs?
20:02 sp0097 joined #salt
20:03 Tyrm joined #salt
20:04 fracklen joined #salt
20:06 CeBe joined #salt
20:07 DammitJim AndreasLutro, just read your "Managing systemd units with Salt"
20:07 wendall911 left #salt
20:09 UtahDave orionx: Ah, that may be a minion side config option
20:09 wendall911 joined #salt
20:09 wendall911 left #salt
20:10 UtahDave raspado: You can configure your gitfs to mount your git repo to whatever path you want.  Then the minion will see a combined /srv/salt/dev
20:11 Tyrm joined #salt
20:12 raspado UtahDave: thx, do i have to whitelist the git branches?
20:13 UtahDave raspado: gitfs will expose all git branches and tags as salt environments, by default. There are a bunch of options to blacklist and whitelist what branches or tags are available.
20:20 west575 joined #salt
20:21 MTecknology HOLY CRAP
20:21 MTecknology UtahDave!!!!
20:21 * MTecknology jumps on UtahDave
20:21 UtahDave lol
20:21 UtahDave Hey, MTecknology!!
20:22 MTecknology UtahDave: Wanna see my thing? I'm pretty proud of my thing
20:22 UtahDave Let's see it!  lol
20:22 MTecknology http://p.lustfield.net/network_overview.svg
20:23 MTecknology salt-cloud -p do.<dc> <fqdn>; wait... and I have a server that's fully up and finished running highstate. :)
20:23 UtahDave Nice!  looks good.  Whenever I try to put together diagrams like that they always look like crap
20:23 MTecknology lucidcharts!
20:23 UtahDave sweet! automate all the things!
20:24 MTecknology I haven't paid for it yet, but I'm starting to think it's actually worth the cost if you make a lot of diagrams. I was looking for a visio replacement in linux-land, somebody mentioned that site, tried it... better than visio. :)
20:25 UtahDave Nice. I'll have to give that a try.  Looks nice.
20:26 stanchan joined #salt
20:26 joshin joined #salt
20:27 armonge joined #salt
20:28 pipps joined #salt
20:28 Tyrm joined #salt
20:28 lero joined #salt
20:31 anotherZero joined #salt
20:33 viq graphviz? And I've played a bit with neo4j, though that's slightly different animal
20:35 viq but neo4j (a graph database) seems to be really useful for analyzing connections and dependencies
20:35 viq MTecknology: and yeah, certainly nice setup
20:35 * viq waves to UtahDave
20:36 UtahDave hey, viq!  How's it going?
20:36 viq trying to play some with HubbleStack right now
20:37 wendall911 joined #salt
20:38 viq Otherwise - ok, busy, had nice vacation. You?
20:38 Tyrm joined #salt
20:39 MTecknology viq at the same time!
20:39 viq It's too many timezones away for it to be same time!
20:40 MTecknology what's a timezone?
20:42 MTecknology dangit... I need the virtual grain inside of a pillar I'm trying to write.
20:44 UtahDave i'm doing well.  I need a vacation, too!
20:45 armin_ joined #salt
20:45 MTecknology I kinda figured the timezone state existed; didn't expect it to be quite so shiny.
20:46 devopsprosiva joined #salt
20:46 MTecknology Here's a question... all systems the same TZ relative to HQ? all systems use their local TZ? all systems use UTC+0?
20:46 devopsprosiva Hi, Is there any documentation to templatize the salt cloud maps and profiles?
20:47 fracklen joined #salt
20:47 Tyrm joined #salt
20:47 viq MTecknology: unless someone besides you care, I'd say HQ or UTC
20:50 MTecknology my vote was probably HQ time, just so mentally adjusting offsets isn't a pain... and it'd force my IRC server onto my actual TZ
20:51 pppingme joined #salt
20:54 filippos joined #salt
20:58 win_salt would anyone know why a file Always shows diffs during a state run, even when there is no difference
21:00 DammitJim anyone use salt with raspberry pi?
21:02 bltmiller joined #salt
21:02 jhauser joined #salt
21:04 pipps99 joined #salt
21:04 raspado alright.... so i've got my git repos (branches, dev,stage) and i configured the master file for dev: /srv/salt/dev
21:05 raspado but the commits im making in my git repo in the dev branch are not popularting /srv/salt/dev on the master
21:05 raspado populating*
21:06 raspado any other config changes to make the salt master pull the dev branch into /srv/salt/dev ?
21:08 viq raspado: master doesn't put git into /srv/salt but uses it directly (some of it ending up in /var/cache/salt/master)
21:08 devopsprosiva In the salt cloud map file, how can I set the name tag to have the role tage plus the instance id?
21:08 devopsprosiva like this: 'Name': 'Role' + 'instance-id'
21:12 krymzon joined #salt
21:12 raspado viq: oooo so it doesnt actually create a file structure?
21:12 viq raspado: nope
21:14 Tyrm joined #salt
21:14 MTecknology raspado: poke around /var/cache/salt/ .. lots of goodies in there for troubleshooting. :)
21:15 giany hi, I'm using this top.sls file : http://pastebin.com/agzQf1YB , thing is that I also use the gitfs fileserver_backend and the docker.mon recipe is called first..is there a way to reorder or prioritize recipes in gitfs?
21:17 raspado viq: MTecknology thx!
21:19 hemphill joined #salt
21:20 UtahDave raspado: run  salt <minion id >  cp.list_master       to see what the minion sees on the fileserver
21:20 UtahDave You won't be able to see it on the master itself
21:20 UtahDave raspado: it's kind of a virtual overlay
21:22 devopsprosiva Is there a way to dynamically generate the map file in salt cloud instead of creating a static file?
21:23 stomith left #salt
21:24 stanchan joined #salt
21:27 sjmh joined #salt
21:31 viq g'night
21:31 alrayyes joined #salt
21:33 stephanlooney joined #salt
21:37 jrgochan Edgan: Yes, I'm using an internal CA
21:40 DEger joined #salt
21:41 stephanlooney joined #salt
21:56 raspado UtahDave: if I add additional environments but the folders in those environments are empty, would there be a problem with that?
21:57 raspado they would just be placeholders until we start adding top.sls files etc etc
22:05 alexhayes joined #salt
22:05 pipps joined #salt
22:06 subsignal joined #salt
22:06 velikan joined #salt
22:06 bltmiller joined #salt
22:07 orionx joined #salt
22:07 velikan hello everyone! I’m reading docs and don’t understand how to write hostname for /etc/hosts in my topfile…
22:08 pipps joined #salt
22:09 chrichip velikan what are you trying to do?
22:09 velikan I’m trying to write all the minions hostnames to /etc/hosts of every minion
22:10 chrichip https://docs.saltstack.com/en/develop/ref/states/all/salt.states.host.html#module-salt.states.host
22:10 chrichip and can be templated out for each node
22:11 velikan chrichip: yeah, but I’ve added all this in my topfile and… nothing hapenned
22:11 chrichip your top file isn't where the states go.
22:12 chrichip top file will decide which states get run for which node
22:12 chrichip i.e.
22:12 velikan oh
22:12 onlyanegg My state says that an NFS filesystem is mounted, but it's not. Can I force it to be remounted when running the state?
22:12 chrichip base:
22:12 onlyanegg I've tried removing the mount directory
22:12 chrichip '*':
22:12 chrichip - state to run
22:15 velikan chrichip: thank you it works! you rock man!
22:15 chrichip np
22:16 amcorreia joined #salt
22:19 raspado does anyone know what will happen if I create empty environment directories
22:19 raspado but define them in the salt master file?
22:19 raspado i want to prep my folder structure
22:20 raspado they wont have any top.sls files in them (yet)
22:20 stanchan joined #salt
22:24 chrichip raspado don't believe it would be an issue, the daemon will start up and you'll still be able to work on environments that have been setup
22:25 lero joined #salt
22:27 lero joined #salt
22:27 raspado chrichip: thx!
22:28 chrichip np
22:47 rome_390 joined #salt
22:48 rome_390 joined #salt
22:48 baffle joined #salt
22:49 sjmh is there anyway to target hosts that don't have a specific pillar key?
22:51 Kaji_ joined #salt
22:53 Vaelatern joined #salt
22:55 majikman joined #salt
22:56 rome_390 joined #salt
22:57 __number5__ sjmh: yes, use compound target with `not` keyword e.g. `salt -C 'not I@pdata:foobar' test.ping`
22:57 __number5__ https://docs.saltstack.com/en/latest/topics/targeting/compound.html
22:57 sjmh hm, thought I'd tried it - and it didn't work if the key didn't even exist.
22:58 sjmh ie - it's testing the value, not key existence.
22:58 sjmh I don't care what the value of the key is set to.  I care whether or not the key exists.
22:59 whyteaway salt -C 'not J@<pillar>:.*'
23:00 pipps joined #salt
23:00 rome_390 joined #salt
23:03 whytewolf sjmh: sorry that was meant for you.
23:04 sjmh @whytewolf i saw, thanks :)
23:04 Sammichmaker joined #salt
23:04 Sammichmaker joined #salt
23:04 whytewolf np
23:04 onlyanegg Is there a way to specify the order of an entire sls files in relation to other sls files?
23:04 onlyanegg rather than state by state
23:04 whytewolf onlyanegg: use orchestration
23:05 onlyanegg I have a bunch of common state files in states/common
23:08 onlyanegg ok, I'll look into that. Thanks!
23:08 sjmh and another pillar question - are external pillars still refreshed via the normal maintenance cycle?
23:09 whytewolf sjmh: I do believe so. but as i don't currently use external pillars i can not say for sure
23:11 lero joined #salt
23:11 sjmh @whytewolf hm, I'm trying to think of how that'd work - since the ext pillars are called w/ a minion id, does the maintenance cycle refresh the ext pillar for every minion every 60 sec?  seems like it'd be a huge amount of load if you have alot of minions
23:12 ninjada joined #salt
23:12 whytewolf sjmh: pillars are not refreshed every 60 seconds. they get updated during highstate [i think at the end] and by calling saltutil.pillar_refresh
23:13 sjmh @whytewolf - trying to understand https://github.com/saltstack/salt/issues/19994#issuecomment-95323569
23:13 saltstackbot [#19994][MERGED] Ext pillar refresh interval | According to log file it looks like external pillar with gitfs backend is not refreshed with the same interval as states/formulas (loop_interval) but with every minion highstate. In small deploymonents it's not a problem but with 1200+ minions it's quite big overhead. Proccess list many times contains multiple git-remote-http entries. In my opinion it should be refreshed with the same interval as
23:14 sjmh isn't the loop interval every 60 sec?
23:14 * whytewolf shrugs
23:14 sp0097 joined #salt
23:15 sjmh so, yes, loop interval is 60 sec, but not sure I'm understanding what's meant by the ext_pillar update then.
23:15 whytewolf reread that it is saying it is NOT updated with loop interval
23:15 whytewolf [which it isn't]
23:15 ninjada joined #salt
23:16 sjmh that's what the issue was about, but @cachedout's statement says it is
23:16 sjmh "After talking this over with @thatch45, the proposal is to go with a change to external pillars to put their updates in the master maintanence process and let them update on the loop interval instead of during the state run."
23:17 orionx joined #salt
23:17 whytewolf again. I don't run external pillars. and this stuff all looks like it is recent changes. so I have no idea
23:18 sjmh understood
23:18 devopsprosiva joined #salt
23:19 stanchan joined #salt
23:19 Deliants joined #salt
23:21 alexhayes joined #salt
23:22 giany is it possible to use orchestrate with salt-cloud?
23:22 JPT joined #salt
23:22 whytewolf giany: https://docs.saltstack.com/en/latest/ref/states/all/salt.states.cloud.html
23:23 whytewolf giany: https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.cloud.html#module-salt.modules.cloud
23:24 giany hmm
23:24 orionx joined #salt
23:25 alexhaye1 joined #salt
23:26 giany not sure if i understand, there is nothing related to orchestrate there, that unless I can use the cloud.* in the orchestration file
23:26 iggy you can
23:26 iggy the same as any other states and modules
23:26 whytewolf giany: orchestration can use modules and states. so. those are the modules and states for controlling salt-cloud
23:28 whytewolf oh i forgot you can also use runners with orchestration so this could be used as well. https://docs.saltstack.com/en/latest/ref/runners/all/salt.runners.cloud.html#module-salt.runners.cloud
23:28 orionx_ joined #salt
23:28 devopsprosiva Do the orchestration runners work with aws autoscaling groups?
23:29 bltmiller joined #salt
23:29 devopsprosiva I need a way to automatically scale up and down with minions already configured in the AMI
23:30 giany thx whytewolf, will dig that out
23:31 amy joined #salt
23:32 bltmiller joined #salt
23:32 stanchan joined #salt
23:36 whytewolf devopsprosiva: https://www.packtpub.com/books/content/how-to-auto-scale-your-cloud-with-saltstack read that. might help you plan out better.
23:39 devopsprosiva whtewolf: Thanks. Will check it out.
23:40 edrocks joined #salt
23:41 nidr0x joined #salt
23:52 bltmiller joined #salt
23:58 |aaron joined #salt

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