Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-03-10

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

All times shown according to UTC.

Time Nick Message
00:02 desposo joined #salt
00:02 baweaver joined #salt
00:12 jespada joined #salt
00:13 bhosmer joined #salt
00:22 Georgyo joined #salt
00:23 raygunsix joined #salt
00:30 techdragon joined #salt
00:33 dooshtuRabbit joined #salt
00:37 triplecorn joined #salt
00:37 schristensen joined #salt
00:38 zzach joined #salt
00:41 enarciso joined #salt
00:44 nich0s joined #salt
00:53 cberndt joined #salt
00:57 tmh1999 joined #salt
00:57 forrest joined #salt
00:57 packeteer joined #salt
00:57 catpig joined #salt
01:00 teall joined #salt
01:04 Ahrotahntee is anyone presently and successfully using ext_pillar with gitfs?
01:08 bhosmer_ joined #salt
01:08 aqua^mac joined #salt
01:09 plameniv joined #salt
01:13 Ahrotahntee is it not possible to do with GitPython? I keep seeing a message that it could not be loaded, but only when I'm attempting to use it for ext_pillar
01:13 Ahrotahntee gitfs works fine for states
01:15 badon joined #salt
01:20 otter768 joined #salt
01:23 subsignal joined #salt
01:26 bhosmer_ joined #salt
01:32 MatthewsFace joined #salt
01:34 Ahrotahntee ah, I'm missing libssh2
01:34 Ahrotahntee how silly of me
01:37 * Ahrotahntee rebuilds pygit2
01:39 Whissi joined #salt
01:40 jespada joined #salt
01:44 enarciso joined #salt
01:46 fllr joined #salt
01:47 shaggy_surfer joined #salt
01:48 viq joined #salt
01:48 viq joined #salt
01:48 zwi joined #salt
01:55 druonysus joined #salt
01:57 _JZ_ joined #salt
01:58 teall joined #salt
02:00 mgw joined #salt
02:10 \ask joined #salt
02:14 _JZ_ joined #salt
02:18 __JZ__ joined #salt
02:18 michelangelo joined #salt
02:22 dooshtuRabbit joined #salt
02:22 p0rkbelly joined #salt
02:22 _JZ_ joined #salt
02:25 yomilk joined #salt
02:30 catpigger joined #salt
02:30 tmh1999 joined #salt
02:31 Pixionus What's the best way to edit a single character in a file on a minion?  It's production so I only really get one chance at this.
02:31 Pixionus was looking at file.replace and blockreplace but not sure if they are the best choice here.
02:31 Vynce test it on not-production? ( :   (or on a copy of the file that is not *used* by production)
02:33 * Pixionus sigh..
02:33 Pixionus but I want to be lazy about it (and it's not supper important production, but production none the less
02:33 Pixionus either way still wondering if there's a better way to do this.
02:34 Vynce no idea.  so far i have gotten salt to distribute two files each to two servers, and tha ttook me a week of reading a stressing and beating my head plus half an hour of actually writing an sls file.
02:35 Pixionus lol
02:35 _JZ_ joined #salt
02:36 Pixionus Is that all you've gotten salt to do for you?
02:37 Pixionus It's pretty nice when you can work with it.  Docs are nice when you know how to navigate them correctly (they are a little nonstandard but I like the way they are set up now that I've seen a pro using them and have fallen into more correct habits)
02:38 Pixionus Not a programmer at all so it's all kinda denser to me than it should be but I've been getting the hang of it
02:38 Pixionus course it's not programming.  Just that a programmer is more used to this sort of thing
02:40 Vynce yes, that is all i've gotten salt to do for me.
02:40 Pixionus What else would you like it to do for you?
02:40 Vynce uh… i cannot fathom what would make someone call the docs nice. the problems with them are not jsut in layout.
02:40 Vynce nothing yet, really. ( :
02:41 Pixionus The key to the docs (or the part that's nice) is the modules docs: http://docs.saltstack.com/en/latest/salt-modindex.html
02:42 Pixionus They have hired someone to do just documentation so the rest should be improving.
02:45 Ryan_Lane joined #salt
02:45 Vynce this is what i hear. I hope so.
02:45 Ligthert joined #salt
02:45 Pixionus But basically when I need Salt to do something for me I generally google what I need and 'saltstack' which will take me to an appropriate module (hasn't failed me yet)
02:45 Vynce but i hope by the time it improves, i will have handed off everything salt is responsible for to someone else.
02:45 Vynce you're much farther up the learning curve than i am.  I have not gotten my head around the basics enough to understand how to use a module.
02:45 Pixionus If I know what module I need I can go to it in the index pretty easily using the letter and a quick find (ctrl+f) then in the module's page the TOC is on the right (most important thing I never saw before seeing the pro use it)
02:47 ilbot3 joined #salt
02:47 Topic for #salt is now Welcome to #salt | SaltConf 2015 is Mar 3-5! http://saltconf.com | 2014.7.2 is the latest | Please be patient when asking questions as we are volunteers and may not have immediate answers | Channel logs are available at http://irclog.perlgeek.de/salt/
02:47 Pixionus Most often I find use for the cmd.run module and general salt-key management
02:47 Pixionus That and highstate though I didn't write ours like I afore mentioned
02:49 pdayton joined #salt
02:49 TyrfingMjolnir joined #salt
02:49 Vynce i still don't really even understand what "highstate" means.
02:50 nich0s joined #salt
02:51 Pixionus It executes the state as needed for that particular server
02:51 Pixionus http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.highstate
02:52 Pixionus one of the most important modules in my experience
02:52 sym-jp joined #salt
02:52 Vynce yeah, i read the docs.  still don't understand what they're on about.
02:52 Pixionus So we use it to install our proprietary software and a few other tools if they are not installed already
02:53 Pixionus I kinda wish things were interlinked in these docs so some of the terms you should have learned earlier are "wiki-able"
02:54 Pixionus like test-only(dry-run) mode
02:54 Pixionus Not sure what that is, but I'm pretty sure I should be using it.
02:55 Pixionus I know a lot of people don't like Wiki's and I admit the edit psudocode is kinda annoying to read and write sometimes, but been kinda wanting to set up a salt wiki.  Think it would really help noobs like us get started.
02:56 Pixionus Another thing is they may eventually provide us with some good practice examples so that would be nice.
02:56 Pixionus You've read the tutorials?
02:56 yomilk joined #salt
02:57 Pixionus if not, they are probably the best way to get started.
02:57 Vynce yes.
02:57 desposo joined #salt
02:57 Pixionus would you be interested in making a list of all the things that you find difficult to figure out and get working as you are working through this?
02:58 Pixionus If you send it to me, I think I would be interested in making tutorial blog posts to better explain and help people through as I figure them out.
02:58 Pixionus Or eventually put them on a wiki or something.
02:59 cberndt joined #salt
02:59 Vynce well, i did everything i did with salt a few days ago and didn't keep notes of that sort.
03:00 Pixionus Please keep me in mind if you think of any major gripes or have to do more in the future.
03:01 Pixionus I'm in a similar boat where I'm forgetting my major gripes while I move past them but it could help new users and the Salt team a lot if we could figure out the worst problems
03:03 Vynce one big problem: entry point. even the instructions to bootstrap salt give a ton of examples with no explanation about why you would choose one over another.
03:03 Vynce easiest tutorial to find installs minion and master on the same server, with no explanation why this would ever be useful int he real world, and thus making it difficult to tell which steps are minion steps and which are master steps.
03:04 Pixionus oh yeah.
03:04 Pixionus Forgot about that.
03:04 Vynce in general, the docs are full of "so is you look at this example, for it to work, you'd have to A and B" … without explaining what it would mean for it to work.
03:04 Vynce there's just very little "this is what this actually means"
03:04 Pixionus I figured out it can be helpful to manage your master :P  Not sure if that's good practice or not but found other folks doing it in production
03:05 Pixionus yes.
03:06 Pixionus They are very brief and skip past a lot that I don't know without explanation.  I don't think I would want it all in one place but linked between them in a wiki sort of format would be nice
03:06 Pixionus what do you think?
03:07 thehaven_ joined #salt
03:07 Vynce well, the problem is they're leaving a ton of stuff out *without* then becoming brief.  if it were brief, you could at least say "ah, i can read all this and get a good overview even if i'm missing some detail" — instead, reading it all takes forever, is very redundant, and leaves you missing details and not even knowing where to go to get them.
03:08 Vynce wikis are OK but tend to have organiszation problems similar to what this already has.
03:08 Pixionus Any ideas on what might be better?
03:09 Vynce not that fit in a chatroom line, necessarily.
03:10 Pixionus I like the idea of sourcing our beginner community to make a open set of companion docs but not sure what the best way to go about doing it is
03:10 Pixionus *an open
03:12 ksalman joined #salt
03:12 jespada joined #salt
03:13 Vynce a wiki is a fine idea if someone is dedicated to organizing it as it grows and keeping it structured. A useful set of basic useful usecase walkthroughs wih complete explanations would work too.  I really don't think it's mostly or firstly a matter of generating *more* though — i think it's a matter of curating and refining and fixing what's there.  if they've hired someone to do that full time, great. after they have some time to imp
03:14 Pixionus Yeah.  guess I'm just impatient
03:16 clintberry joined #salt
03:16 dooshtuRabbit joined #salt
03:18 Ryan_Lane joined #salt
03:18 dooshtuRabbit joined #salt
03:20 Pixionus On a related note, they are coming out with a book
03:20 Pixionus There is already one out there too
03:21 monkey66 joined #salt
03:21 dooshtuRabbit joined #salt
03:23 mosen joined #salt
03:23 paolo joined #salt
03:24 dooshtuRabbit joined #salt
03:25 jerematic joined #salt
03:30 \ask joined #salt
03:30 tmh1999 joined #salt
03:31 monkey66 joined #salt
03:32 CeBe1 joined #salt
03:32 CeBe2 joined #salt
03:32 hasues joined #salt
03:33 hasues left #salt
03:33 Ahrotahntee I am having trouble with ext_pillar accessing a git repo. The configuration is accepted, and I'm now using pygit2, however none of the pillar data loads. https://gist.github.com/Ahrotahntee/780cdb6ebc1bc2f272c2
03:34 Ahrotahntee I have the root set to pillars/ and the top.sls file in .git/pillars/top.sls, I'm probably missing something extremely simple
03:34 otter768 joined #salt
03:35 Ahrotahntee is there anything odd/incorrect that jumps out at anyone? this is pretty much by-the-docs example
03:36 tmh1999 joined #salt
03:36 pdayton joined #salt
03:40 favadi joined #salt
03:40 ajw0100 joined #salt
03:41 __JZ__ joined #salt
03:43 monkey66 joined #salt
03:43 smcquay joined #salt
03:43 SheetiS joined #salt
03:43 raygunsix joined #salt
03:44 zzach joined #salt
03:49 _JZ_ joined #salt
03:52 tmh1999 joined #salt
03:52 desposo joined #salt
03:53 iggy Ahrotahntee: ext_pillar doesn't use the same code as the gitfs backend, so I think you still need GitPython installed for ext_pilar
03:57 pcdummy joined #salt
03:57 pcdummy joined #salt
03:57 \ask joined #salt
03:59 _JZ_ joined #salt
04:00 __JZ__ joined #salt
04:02 __number5__ quick question: does ext_pillar works for masterless setup?
04:04 hal58th1 joined #salt
04:05 _JZ_ joined #salt
04:05 raygunsix joined #salt
04:10 iggy check the 2015.2 release notes, more stuff was added to masterless, don't know if that was
04:12 thayne joined #salt
04:14 enarciso joined #salt
04:24 Ryan_Lane joined #salt
04:25 mafro joined #salt
04:29 cberndt joined #salt
04:40 pdayton joined #salt
04:45 \ask joined #salt
04:55 hal58th joined #salt
05:01 jespada joined #salt
05:03 jax0n joined #salt
05:03 jax0n left #salt
05:05 jax0n joined #salt
05:06 jax0n good day folks!
05:07 jax0n just wondering if the saltstack github team is still accepting code volunteers ;)
05:07 jax0n pls dont laugh at my question
05:07 iggy fork the repo, make changes, submit a pull request
05:07 iggy that's how git works
05:08 jax0n OK cool! thanks for answering!
05:09 iggy what were you thinking about doing?
05:10 jax0n well, i was just wondering cos i browsed salt/salt/auth... was actually trying to look for pgsql auth (before actually installing salt)
05:10 jax0n im not claiming to be good .. just  wondering if i had to code my own
05:10 jax0n cos our app uses pgsql
05:11 jax0n or i might have looked at wrong path
05:11 iggy auth
05:11 jax0n yup
05:13 iggy what is your app?
05:14 jax0n its private app. not in github.  its a rest api and were using pgsql.  so i heard about salt (good stuff).  and browsed it on github.  i was particularly looking for pgsql auth
05:15 jax0n i may be wrong and theres a pgsql auth somewhere?
05:15 iggy okay, well, I was going to say, there's django auth, so if it used django's ORM, you could go that route
05:15 iggy but you might have to write it yourself
05:15 mgw joined #salt
05:16 iggy should be pretty easy to alter the mysql one... just change up some modules, clean up the SQL, BAM!
05:16 jax0n Ahh... yeah. that should work. I will try to write one though. and if it works will submit a pull req
05:17 jax0n Yeah i saw the mysql one! nicely written.
05:17 jax0n Thanks a lot mate you were very open / helpful.  one of these days my pull req should come.
05:18 iggy the devs are pretty good about getting stuff cleaned up and merged back upstream
05:18 iggy they realize it's helpful not to have a bunch of outside stuff floating around with different versions, etc.
05:18 jax0n OK... again i wont claim my code will be good. but i just wanted to write for our app, and who knows if my pull will be accepted.  will be installing Salt very soon
05:19 jax0n Cheers!
05:19 iggy good luck!
05:19 jax0n ty
05:20 jax0n left #salt
05:21 desposo joined #salt
05:25 xsteadfa1 joined #salt
05:31 yomilk joined #salt
05:34 tmh1999 joined #salt
05:44 otter768 joined #salt
05:48 ramteid joined #salt
05:50 \ask joined #salt
05:54 SheetiS joined #salt
06:02 zer0def joined #salt
06:14 SheetiS joined #salt
06:20 malinoff joined #salt
06:21 krelo joined #salt
06:31 andreyst joined #salt
06:31 enarciso joined #salt
06:33 markm joined #salt
06:40 andygrunwald joined #salt
06:47 yomilk joined #salt
06:47 evle joined #salt
06:50 jespada joined #salt
06:53 felskrone joined #salt
06:58 colttt joined #salt
07:01 aw110f joined #salt
07:02 stoogenmeyer joined #salt
07:03 jtang joined #salt
07:08 markm joined #salt
07:10 I3olle joined #salt
07:10 aparsons joined #salt
07:10 perfinion joined #salt
07:14 mattiasr joined #salt
07:21 perfinion joined #salt
07:21 AndreasLutro joined #salt
07:23 lietu joined #salt
07:28 triplecorn_ joined #salt
07:29 flyboy joined #salt
07:29 mfournier joined #salt
07:33 bash1245_ joined #salt
07:33 KermitTheFragger joined #salt
07:34 perfinion joined #salt
07:43 toanju joined #salt
07:43 Auroch joined #salt
07:45 otter768 joined #salt
07:47 xsteadfastx joined #salt
07:48 dRiN joined #salt
07:57 pf_moore joined #salt
08:02 jri joined #salt
08:04 linjan joined #salt
08:05 andygrunwald joined #salt
08:06 trikke joined #salt
08:07 lb1a joined #salt
08:07 lytchi joined #salt
08:09 eseyman joined #salt
08:12 trikke joined #salt
08:14 toanju joined #salt
08:14 kawa2014 joined #salt
08:21 Sypher hmm why can't events be sent (event.fire_master, event.send) without data ending up underneath 'data'
08:23 devweasel joined #salt
08:25 wolog joined #salt
08:27 Guest___ joined #salt
08:28 sfxandy hi everyone
08:31 sfxandy two part question here folks so appreciate any help....
08:32 sfxandy i) how do I add specific grains into the Salt mine and ii) how do I reference those in, for example, Jinja?  I have a custom grain labelled roles (lets ignore the fact this may be better implemented in Pillar).  my roles grain can contain multiple values so I'd like to know how to add that grain to the Salt mine and how to reference it in one of my Jinja templated files
08:33 wincyj joined #salt
08:38 illern_ joined #salt
08:38 saltuser joined #salt
08:38 slav0nic joined #salt
08:39 jespada joined #salt
08:40 andreyst joined #salt
08:41 saltuser Hi! Is it possible to set a variable and then include an SLS and it will use that variable?
08:43 saltuser Something like {% set command = "date" %}, include - test.sls. test.sls would contain something like test: \n cmd.run: \n name: {{ command }}
08:44 saltuser I have tried but it fails
08:48 sfxandy hi saltuser.  no idea from me.  am waiting for a more knowledgable person to help me with my query!
08:49 sfxandy interestingly, what error do you get saltuser?
08:51 saltuser "Rendering SLS 'base:test.include_test.include' failed: Jinja variable 'command' is undefined"
08:52 sfxandy ok so when you include the state file, at that point it gets executed.  not sure how it would see the "command" variable as its not in scope
08:52 jri joined #salt
08:52 _mel_ joined #salt
08:54 illern joined #salt
08:54 sfxandy hi everyone
08:55 sfxandy two part question ....
08:55 sfxandy 1) how do I add specific grains into the Salt mine and 2) how do I reference those in, for example, Jinja?  I have a custom grain labelled ‘roles’.  my roles grain can contain multiple values so I'd like to know how to add that grain to the Salt mine and how to reference it in one of my Jinja templated files
08:55 melvin joined #salt
08:56 zer0def joined #salt
08:56 _mel_ joined #salt
08:56 lietu joined #salt
08:57 Sypher saltuser, yes it is possible
08:57 Sypher but not with include:
08:58 Sypher instead with {% include 'some/file' %}
08:59 _mel joined #salt
08:59 Sypher saltuser, some formulas are using the same approach such as: https://github.com/saltstack-formulas/php-formula/blob/master/php/ng/json.sls
09:00 saltuser My current non-working testcase http://pastebin.com/H25QSiGt
09:01 saltuser Sypher: i'll have a look
09:01 illern joined #salt
09:03 wolog joined #salt
09:04 jri joined #salt
09:05 saltuser Sypher: hm, it seems to work indeed :)
09:06 Sypher nice :)
09:07 sfxandy can anyone kindly help me with a Salt mine/grains query please?
09:09 andreyst joined #salt
09:12 sfxandy 1) how do I add specific grains into the Salt mine and 2) how do I reference those in, for example, Jinja?  I have a custom grain labelled ‘roles’.  my roles grain can contain multiple values so I'd like to know how to add that grain to the Salt mine and how to reference it in one of my Jinja templated files
09:13 zipkid Hi, could anyone give me a hint as to what i am missing here ? https://gist.github.com/zipkid/14550b7893d73d465919
09:17 kawa2014 joined #salt
09:18 monkey66 left #salt
09:25 VSpike joined #salt
09:26 che-arne joined #salt
09:28 babilen sfxandy: You define a mine function as detailed in http://docs.saltstack.com/en/latest/topics/mine/#mine-functions (you can use any execution module). Why don't you manage your roles centrally? It feels a bit weird that you are saving this information on the minions only to have them send it back to the master. Why don't you keep it on the master to begin with (pillar, ext pillar, nodegroups, naming scheme + minion id parsing, ...) ?
09:29 babilen sfxandy: Mine lookups use http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.mine.html#salt.modules.mine.get
09:29 sfxandy hmm, just seemed more natural to have it defined as a grain i guess...
09:30 babilen Feel free to use the grains.get execution function and make that a roles function alias, but that feels overly contrived
09:31 babilen sfxandy: Many people do use grains for roles, but make sure that you never ever make sensitive data available based on grains. Other people prefer to centrally manage their roles and state assignment (rather the minion "requesting something" you tell it to "do something")
09:31 Xevian joined #salt
09:31 sfxandy i hear what you're saying babilen .... just getting to grips with Salt at the moment.
09:34 sfxandy as you say, the majority of cases use grains for roles.  i already referred to the mine functions but.  i guess the bit I was missing was how to specify the grains i need in the mine_functions: config bit
09:35 babilen sfxandy: You can define roles in pillars and target based on them also. This is *much* more secure and allows you to easily bring in external pillars such as databases, but comes with the drawback that you cannot target pillars themselves based on that. Given that pillars frequently contain sensitive information (certificates, password hashes, ...) and that you don't want to target this based on grains (as that is easily compromised) it wouldn't be ..
09:35 babilen ... much of a problem though
09:35 I3olle joined #salt
09:36 babilen sfxandy: salt['grains.get']('yourfoograin', ....)
09:38 sfxandy hmmm ok.  might  have a think about this.  it does seem a little contrived
09:40 losh joined #salt
09:40 sfxandy babilen, when you refer to setting roles in Pillar .... do you mean using Pillar lookups to set the grains as opposed to putting the custom grains in /etc/salt/grains ... or do you mean using some othjer technique to set roles via Pillar?
09:42 jri left #salt
09:46 otter768 joined #salt
09:46 perfinion joined #salt
09:49 Vynce joined #salt
09:49 johanek joined #salt
09:50 intellix joined #salt
09:52 zipkid ok, me stupid...
09:54 jespada joined #salt
09:57 stoogenmeyer_ joined #salt
09:57 lietu joined #salt
09:58 markm joined #salt
10:04 aquinas joined #salt
10:07 andygrunwald joined #salt
10:07 johanek joined #salt
10:12 aquassaut joined #salt
10:24 lothiraldan joined #salt
10:25 illern joined #salt
10:39 seshan joined #salt
10:40 seshan I am trying to setup multiple masters, referring here ., http://docs.saltstack.com/en/latest/topics/tutorials/multimaster_pki.html
10:40 seshan But everytime I get an error saying, "[CRITICAL] The Salt Master server's public key did not authenticate! The master may need to be updated if it is a version of Salt lower than 2014.7.1, or If you are confident that you are connecting to a valid Salt Master, then remove the master public key and restart the Salt Minion. The master public key can be found at: /etc/salt/pki/minion/minion_master.pub"
10:41 seshan I found this statement : "If multiple masters are in use and should sign their auth-replies, the signing key-pair master_sign.* has to be copied to each master. Otherwise a minion will fail to verify the masters public when connecting to a different master than it did initially. That is because the public keys signature was created with a different signing key-pair."
10:41 seshan It says copy master_sign.* to the other, but where to copy it to ??
10:42 aparsons joined #salt
10:43 matthew-parlette joined #salt
10:48 teambrad1 joined #salt
10:51 giantlock joined #salt
10:52 zipkid seshan: i would think to the same place you found it on one to all the other masters....
10:53 evle1 joined #salt
10:57 linjan joined #salt
10:59 djn joined #salt
11:05 kawa2014 joined #salt
11:07 _ale1_ joined #salt
11:10 andygrunwald joined #salt
11:12 yomilk joined #salt
11:17 jespada joined #salt
11:20 diegows joined #salt
11:21 kawa2014 joined #salt
11:21 johtso joined #salt
11:23 djn Hello. Not sure whether this is the right place to mention this, but I notice that the version of Salt in Homebrew on OS X is out of date. ‘brew info saltstack’ reports 2014.7.1, though with that installed, ‘salt --version’ reports ‘salt 2014.7.0’.
11:26 lothiraldan joined #salt
11:32 clintberry joined #salt
11:34 jri joined #salt
11:38 huddy joined #salt
11:40 amcorreia joined #salt
11:44 Morbus joined #salt
11:47 otter768 joined #salt
11:54 ktosiek joined #salt
11:56 bluenemo joined #salt
11:56 bluenemo joined #salt
12:00 emirozer joined #salt
12:01 karimb joined #salt
12:02 denys joined #salt
12:04 bfoxwell joined #salt
12:06 andygrunwald joined #salt
12:09 thayne joined #salt
12:10 jerematic joined #salt
12:15 zwi joined #salt
12:18 lothiraldan joined #salt
12:25 DorfOnGolf joined #salt
12:26 mikkn joined #salt
12:27 yomilk joined #salt
12:35 cmcmacken joined #salt
12:35 numkem_ joined #salt
12:41 Exploit joined #salt
12:45 JlRd joined #salt
12:45 Ahrotahntee iggy: in case you were wondering; file.recurse + clean + exclude_pat doesn't behave as expected. Assuming my expectation that an excluded pattern should not be cleaned is correct.
12:46 aparsons joined #salt
12:51 subsignal joined #salt
12:54 subsignal joined #salt
12:57 subsignal joined #salt
12:59 markm joined #salt
12:59 yuhl_work_ How can I easily add/remove a kernel parameter in grub.conf ?
13:00 yuhl_work_ for all the kernels
13:01 damon_ joined #salt
13:05 damon_ I am a salt begin.  Use Puppet for a few years. In puppet you can create a sudo module and in this have a function define sudo:config_line.  Another module to install software X you can call  sudo:config_line(<sudo line>)  How do you get the define behavior in salt.
13:05 andreyst joined #salt
13:05 yuhl_work_ damon_: I see that I can use grubby command to do the job.
13:05 yuhl_work_ damon_: oups sorry.
13:07 tomspur left #salt
13:11 hasues joined #salt
13:12 hasues left #salt
13:12 cpowell joined #salt
13:13 _ale1_ joined #salt
13:13 scoates joined #salt
13:16 I3olle_ joined #salt
13:18 perfectsine joined #salt
13:18 jdesilet joined #salt
13:19 elfixit joined #salt
13:21 Exploit left #salt
13:21 fredvd joined #salt
13:22 racooper joined #salt
13:23 intellix joined #salt
13:24 murrdoc joined #salt
13:27 MWheelz joined #salt
13:29 saltuser left #salt
13:30 teall joined #salt
13:31 mpanetta joined #salt
13:32 murrdoc joined #salt
13:33 VSpike I have a pkgrepo.managed state that uses an insanely slow keyserver (subkeys.pgp.net) which slows down the entire state and often causes it to fail completely because it times out....
13:33 VSpike Is there an easy way to include the keydata directly? Would be nice if it could be in the SLS itself.
13:34 murrdoc do you have a local server u can put the key on, or can u put the key on the salt master itself, if so then use key_url to fetch the key
13:35 AndreasLutro VSpike: gpgkey as shown in the official example maybe?
13:36 AndreasLutro or maybe that's just for yum
13:36 ThomasJ AndreasLutro: Just for yum
13:36 ThomasJ VSpike: key_url is possibly what you want.
13:36 pdayton joined #salt
13:36 ThomasJ Not sure if it supports salt://, but you could throw it up on a fast webserver
13:37 murrdoc it usees cp.cache_file, salt:// will work
13:38 ThomasJ Saw your reply too late, thanks :)
13:39 VSpike Oh, I'll test it now. Thanks!
13:44 VSpike doesn't seem to like a salt url there
13:44 murrdoc thats weird
13:45 murrdoc http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.cp.html#salt.modules.cp.cache_file
13:45 dyasny joined #salt
13:46 VSpike I dunno, am I doing anything obviously stupid? https://bpaste.net/show/3b8ffbf51e93  .. I get "Failed to configure repo 'deb http://downloads.opsview.com/opsview-core/latest/apt trusty main': Error: key retrieval failed: apt-key add ''"
13:48 VSpike ignore me
13:48 otter768 joined #salt
13:48 murrdoc :) what happened
13:48 murrdoc i just opened the bpaste
13:49 VSpike I just wrote my salt:// path wrong. "base" in the env so should start salt://linux/...
13:49 VSpike works perfectly now
13:49 murrdoc :)
13:49 VSpike Awesome. Thanks! That's so much neater
13:54 Jouke Does salt store the output of jobs somewhere?
13:55 yomilk joined #salt
13:56 murrdoc joined #salt
13:56 murrdoc Jouke:  it does cache the results on disk
13:57 murrdoc but if u want something to persist longer, i recommed looking at saltstack returners
13:57 Jouke nah, I just had a job with 86 states fail on 1 state
13:57 Jouke and my console backlog is not big enough
13:58 murrdoc switch your 'output' config
13:58 murrdoc and maybe setup failhard: True
13:59 Jouke thanks I am going to look into that.
14:00 Jouke murrdoc: would it be possible to look up the previeous job still?
14:00 pdayton joined #salt
14:00 murrdoc i think so, look up the jobs module
14:00 Jouke oh sorry, I allready rtfm ;P
14:01 murrdoc do u have the jid?
14:01 murrdoc salt-run jobs.lookup_jid JID outputter=highstate or salt-run jobs.print_job JID might help
14:02 murrdoc other option is go into your cache dir on the master and minion
14:02 murrdoc and try to find the most recent job
14:06 debian112 joined #salt
14:06 Jouke murrdoc: found it! Thanks!
14:06 murrdoc phew
14:07 murrdoc where did u find it
14:08 Jouke did a jobs.list_jobs, found the jid  and then did a print_job
14:09 murrdoc nice
14:09 murrdoc phew
14:09 jtang joined #salt
14:11 denys joined #salt
14:11 andrew_v joined #salt
14:13 MorbusIff joined #salt
14:15 brick_ joined #salt
14:19 aphorise joined #salt
14:20 jalbretsen joined #salt
14:20 pdayton1 joined #salt
14:25 faliarin joined #salt
14:26 FeatherKing joined #salt
14:27 FeatherKing is anyone familiar with the states.archive module?
14:27 FeatherKing not sure if im using it correclty
14:29 wormwood Greetings all, using salt-ssh and getting the string '_edbc7885e4f9aac9b83b35999b68d015148caf467b78fa39c05f669c0ff89878' in the stdout section. The command I'm running is: HOME=/home/ec2-user sudo salt-ssh -i --roster-file=./roster '*' cmd.run 'ls /'
14:29 aphorise joined #salt
14:29 wormwood its not listing the output regardless off the cmd.run args
14:30 wormwood just that RSTR string ('_edbc7885e4f9aac9b83b35999b68d015148caf467b78fa39c05f669c0ff89878')
14:33 murrdoc weird
14:33 markm joined #salt
14:33 murrdoc FeatherKing:  what u finding with  state.archives
14:33 teall joined #salt
14:35 Xevian joined #salt
14:36 FeatherKing murrdoc: i may have found the issue, i was using tar_options: z but in the main docs it says dont use any of the normal operators. so i guess i dont need to specify gzip compression?
14:36 krelo_ joined #salt
14:36 FeatherKing when i removed it, it finally extracted for me
14:37 Brew joined #salt
14:39 murrdoc nice
14:39 murrdoc my recommendation when the doc is confusing just read the code on github
14:39 murrdoc and submit a pull to the documentation with your clarification
14:40 murrdoc all of this is unnecessary but tis how a community code grows
14:41 zwi joined #salt
14:41 mgw joined #salt
14:45 damon_ I am a salt beginer.  Use Puppet for a few years. In puppet you can create a sudo module and in this have a function define sudo:config_line.  Another module to install software X you can call  sudo:config_line(<sudo line>)  to add a line to sudoers.  How do you get the define behavior in salt.
14:45 murrdoc look up saltstack requisities
14:45 Corey I hate that anti-pattern, damon_.
14:46 Corey damon_: Particularly in Puppet, which is declarative by definition. Tell me what the finished file looks like, not "how to build it piece by piece."
14:47 damon_ Can you in salt "build a file piece by piece", a collection of input
14:49 murrdoc sorta, using a template file and some version of file.replace
14:49 LtLefse damon_: it can be done with e.g. file.replace, file.blockreplace
14:49 murrdoc but going off what corey said, use the conf.d directories
14:49 Corey (It's still terrible, mind you.)
14:49 murrdoc puppet still does this whole, munge file from sections thing, its so much more readable to use conf.d directors
14:49 LtLefse I second what murrdoc said, use sudoers.d
14:49 damon_ Sudo the an example everyone should understand....
14:50 Corey murrdoc wins, /etc/sudoers.d/ it is.
14:50 damon_ their are other config files which do not support .d
14:50 krelo joined #salt
14:50 murrdoc i am running into issues getting people to use 40_custom in grub instead of just shoe horning shit in to 10_linux
14:51 murrdoc pain
14:52 damon_ limits.conf is another example
14:52 murrdoc for limits ?
14:52 murrdoc file limits ?
14:52 Corey Render it as a template, stuff values into pillar.
14:52 murrdoc security.d directory can be used for that
14:52 damon_ You can define limits for individual users/applications
14:52 __number5__ file.accumulated can do .d style config for you
14:52 LtLefse damon_: "man limits" suggests otherwise
14:52 damon_ OK
14:53 LtLefse man limits.conf
14:53 murrdoc thats sexist
14:53 murrdoc should be person
14:53 murrdoc amirite!
14:53 * murrdoc shuts up
14:53 LtLefse anyhow...
14:53 rnts joined #salt
14:54 murrdoc hah
14:54 LtLefse there are times you really do want/need to have Salt manage only part of a file
14:54 kaptk2 joined #salt
14:54 LtLefse if it's just one line, try file.replace
14:54 murrdoc .replace is good for that
14:54 raygunsix joined #salt
14:54 murrdoc or what LtLefse said
14:54 devweasel Die Zeichen heissen aber überall gleich :-)
14:55 devweasel sorry; lost window focus
14:55 TyrfingMjolnir joined #salt
14:55 yomilk joined #salt
14:56 bpb joined #salt
14:56 damon_ ok
14:56 damon_ Thanks
14:57 murrdoc damon_: srsly tho, file.accumulated and file.replace is what u should read up on
14:58 murrdoc also look at saltstack-formula/user-formula to see how u can require sudo confs from different states
14:58 damon_ Is their a page documenting pillar naming standards  or best practice e.g. like puppet   <module>::<variable> so you do not get the same variable being reused for a different process.
14:58 Krzysztof_ joined #salt
14:59 damon_ i.e. Does salt stack have the concept of naming spaces.
15:00 pravka joined #salt
15:00 Krzysztof_ Hello,  is there possiblity to test.ping minion with interval(lets say 10s), and after ping succeeded stop this activity and switch to next state? Thank you
15:02 jerematic joined #salt
15:03 damon_ http://makina-corpus.com/blog/metier/2014/saltstack-use-accumulated-with-file-manage-and-file-blockreplace
15:04 nesv joined #salt
15:04 _ale1_ joined #salt
15:07 devweasel_ joined #salt
15:12 favadi joined #salt
15:15 lietu joined #salt
15:16 rojem joined #salt
15:17 MTecknology Corey: murrdoc: Fun story, in my poorly presented presentation, I specifically mentioned doing as little change as possible. If you can change things using just debconf, do that. Otherwise, use .d directories. Always do everything you possibly can to avoid messing with package maintained files.
15:18 Corey MTecknology: Shenanigans.
15:18 MTecknology You're discussion above prompted me to replay that here. :)
15:18 murrdoc MTecknology:  slides ?
15:18 SheetiS joined #salt
15:18 Corey MTecknology: In Puppet-land, this is called the Package-FIle-Service design pattern.
15:18 smcquay joined #salt
15:18 MTecknology murrdoc: I can probably find them... gimme a sec
15:18 toanju joined #salt
15:18 Corey MTecknology: The idea being you let the package manager do its thing. When it quiesces, you then manage the file. Then, you restart the service.
15:20 trikke joined #salt
15:23 MTecknology murrdoc: http://p.lustfield.net/SaltConf15_Chaos_to_Order.pptx
15:23 murrdoc eugh is that eugh
15:23 murrdoc D:
15:23 murrdoc thanks MTecknology, totes kidding on the eugh
15:23 RedundancyD joined #salt
15:24 MTecknology You mean because it's .pptx?
15:24 murrdoc yeah, like i said kidding, you should probably put these on slide share
15:24 murrdoc get that internet fame
15:24 MTecknology never heard of it before
15:25 MTecknology Corey: lol.. yup, and then never perform a large upgrade ever again.
15:25 TyrfingMjolnir joined #salt
15:26 murrdoc omg dat preseed, is that real
15:26 cmcmacken is it possible to test if a file exists in salt:// from within an execution module? If so how?
15:26 MTecknology murrdoc: it's sooooooo much worse than what I shared, I just stuck with that part
15:27 MTecknology cmcmacken: I don't think so (could very well be wrong, though)
15:27 clintberry joined #salt
15:29 mafro joined #salt
15:29 Corey MTecknology: Depending upon how you manage your environment, "don't upgrade, just reprovision" isn't the worst thing in the world.
15:29 malinoff joined #salt
15:30 MTecknology for some that might be acceptable, for me .. It'll be 10+ years before I reprovision most of my boxes
15:30 rojem joined #salt
15:31 che-arne joined #salt
15:31 triplecorn_ joined #salt
15:32 MTecknology cmcmacken: I suppose you could utilize cp.cache_file...
15:33 MTecknology or cp.get_file
15:34 StDiluted joined #salt
15:34 bluenemo joined #salt
15:34 felskrone joined #salt
15:35 cmcmacken Thanks for the suggestion, that might work. Wondering if I'm better off writing a state module instead
15:36 murrdoc yes
15:36 murrdoc u are
15:36 schristensen joined #salt
15:36 murrdoc and then put it in salt/salt-contrib for the next person who needs it
15:36 murrdoc https://github.com/saltstack/salt-contrib/tree/master/states
15:37 conan_the_destro joined #salt
15:39 timoguin joined #salt
15:40 mangas joined #salt
15:42 Ryan_Lane joined #salt
15:44 iggy cmcmacken: {% if 'foo/bar.sls' in salt['cp.list_master'] %}
15:45 MTecknology the master has spoken!
15:45 thayne joined #salt
15:46 thedodd joined #salt
15:46 MTecknology I'm writing a bunch of states that basically rip out ldapd and stick in sssd. I'm trying to do it with as little disruption as possible. Turns out... that's difficult.
15:47 iggy yeah... imagine that
15:47 kawa2014 joined #salt
15:49 otter768 joined #salt
15:51 MTecknology install packages, update config files, restart samba, rejoin domain, wipe sssd cache, update file system permissions, update nsswitch, remove old stuff, update other fs permissions, restart samba ... done...
15:52 MTecknology This currently looks like 'salt-call state.highstate; <long bash one-liner>; service samba restart; salt-call state.highstate'
15:53 iggy orchestrate that shizzle
15:55 murrdoc is orchestrate real
15:55 tligda joined #salt
15:55 iggy not really
15:55 iggy it's kind of a poorly documented pita right now
15:56 murrdoc no honest question, is it real, like are the saltstack crew committed to it ?
15:56 MTecknology I could. This stuff is all one off, though
15:56 murrdoc like i like the idea, but will it be supported or halited
15:56 iggy I have one off states/orchs
15:56 iggy lol, halited
15:56 murrdoc MTecknology:  u can also do state.sls state.that.does.long.bash.and.samba
15:56 murrdoc that way the one off is in git, but not in highstate
15:57 iggy well, they dumped overstate for orchestrate (and I think they claim complex orchestration as a selling point in SSE)
15:57 bluenemo joined #salt
15:57 cwright where am i supposed to drop a custom ext_pillar module for it to be loaded automatically without setting extension_modules ?
15:57 iggy yeah, what murrdoc said... you don't necessarily put it in a highstate, but write it anyway
15:58 iggy you know you're going to find some system that someone had powered off in like 6 months
15:58 murrdoc /srv/salt/pillar/ext_pillar.py
15:58 iggy you just made that up
15:59 cwright murrdoc: thanks
16:03 iggy I would think it'd be /srv/salt/_pillars but the link script in salt-contrib doesn't do anything with pillars
16:03 murrdoc no its pillar
16:03 murrdoc ireadthecode
16:04 cwright yea it is strange that there isn't a _directory style directory somewhere for them
16:04 cwright like the rest of the /srv/salt/_{modules,grains}/ dirs
16:04 MTecknology iggy: I didn't realize orchestrate replaced overstate. I hated overstate when I tried to use it.
16:04 Fiber^ joined #salt
16:05 iggy MTecknology: I never used it... but orchestrate has some issues... (lack of) docs being one of them
16:06 murrdoc docs ?
16:06 murrdoc read the code
16:06 MTecknology I can just rest happy knowing that docs are something they're putting heavy effort into fixing. What's funny is that I actually like all of the documentation except when it's lacking.
16:06 murrdoc hit up based pi
16:06 aparsons joined #salt
16:07 MTecknology murrdoc: We can't all be russian like you!
16:07 iggy there used to be a site called saltstarters.org ... that appears to be working again
16:07 iggy should put up some of my examples (from my talk) there
16:07 murrdoc russian ?
16:07 MTecknology I'm going with russian
16:07 MTecknology iggy is chinese
16:08 murrdoc cos of his low tolerance for alcohol ?
16:08 iggy lol
16:08 murrdoc why russian tho
16:08 MTecknology I was so effing ticked off when I realized there was an open bar. I wanted to abuse the crap outta that.
16:08 murrdoc did bechtoldt setup this site
16:09 iggy you'd think so
16:09 MTecknology because I've seen some really crazy coders come from russia that could read code like it was documentation
16:09 hybridpollo joined #salt
16:09 murrdoc code is documentation man
16:09 murrdoc http://blog.codinghorror.com/learn-to-read-the-source-luke/
16:09 MTecknology not the code I usually deal with.
16:09 MTecknology you saw that preseed, no?
16:10 murrdoc preseed made my eyes bleed
16:10 MTecknology Lemme go dig up some of our quality code that exists on a public facing website...
16:11 gladiatr joined #salt
16:12 jalbretsen joined #salt
16:12 MTecknology actually... no, can't do it
16:14 murrdoc :)
16:14 MTecknology murrdoc: I'll summarize it. It's full of gaping security holes. Our developers wrote it. It's public facing. The mysql root password is embedded multiple times into the source. If I showed you the scary, you would then see how to manipulate payroll information for the last 7 years without even needing to log in.
16:14 murrdoc nah u guys are a nice company
16:14 murrdoc dont want to mess with your payroll
16:15 MTecknology the version control system is cp <file> <obfuscated_file_name>
16:15 iggy give him a raise?
16:15 iggy so yeah, I'm going ahead with the 2015.2 upgrade (doing it remotely via salt)
16:15 iggy I'll let you know how it goes
16:16 * MTecknology is rather curious
16:16 murrdoc u crazy
16:16 murrdoc upgrading before a point release
16:16 murrdoc :nope:
16:16 iggy well, we haven't been running a point release
16:16 markm_ joined #salt
16:17 iggy we've been running straight from git
16:17 iggy because they never did another release in 2014.1.x
16:17 MTecknology Hrm.....
16:18 iggy and the last release in 2014.1 had DISABLED COMPOUND MINE MATCHES
16:18 iggy bahahahah... never gonna live that one down
16:18 MTecknology I have 5 DB servers that should NEVER restart a service at the same time. In general, that would only happen from executing a highstate. It seems to make sense that I could use mine to let one DB server see if other DB servers are restarting the service.
16:18 * iggy throws darts at the base of pi
16:19 MTecknology potential race issues... all could check all at exactly the same time
16:19 iggy MTecknology: publish is probably better for that
16:19 iggy unless your mine update interval is super low
16:20 amcorreia joined #salt
16:20 MTecknology I've never touched publish
16:21 dthorman joined #salt
16:21 wolog joined #salt
16:23 iggy went fairly well
16:23 MTecknology iggy: My concern is that if I ever issue  salt '*' state.highstate  then I'll potentially have all 5 boxes wanting to restart at the exact same time. I'm trying to figure out how to prevent that issue.
16:23 iggy sudo salt -C '* and not G@tags:master' cmd.run "curl -s -L https://bootstrap.saltstack.com | sudo sh -s -- git 2015.2"
16:23 iggy curl -s -L https://bootstrap.saltstack.com | sudo sh -s -- -M git 2015.2
16:23 iggy don't judge
16:24 Ahrotahntee this is a judgement free zone. if we start to judge eachother, we will have to judge ourselves; nobody wins that one
16:24 iggy MTecknology: we don't put watches on services in our highstate and we manage those things through orchestrate
16:24 desposo joined #salt
16:24 MTecknology this is the only service that doesn't have a watch on it.
16:24 MTecknology I'll take a look at doing it that way
16:25 MTecknology meeting time, then food time
16:26 wnkz joined #salt
16:27 iggy I always forget about salt.modules.saltutil
16:31 murrdoc http://www.slideshare.net/ThomasJackson4/salt-stack-at-web-scale-better-stronger-faster great talk
16:32 KyleG joined #salt
16:32 KyleG joined #salt
16:32 murrdoc still not clear how mcuh of it was pushed up to salt upstream
16:32 iggy he did more than one talk I guess... I don't think I caught that one
16:36 ksalman murrdoc: is there a video to go with those slides?
16:37 scoates Frustrated. I have these states (and more) http://paste.roguecoders.com/p/7e5c7b96b87e817efe2fa7cb209e774e.txt and I get a failure on pkg: libc-ares2 as "is unavailable" but I can immediately `apt-get install libc-ares2`; any idea how to really-really force that refresh?
16:37 murrdoc refresh: True
16:37 iggy ksalman: not yet... watch the youtube channel for videos
16:37 ksalman k
16:38 lietu joined #salt
16:39 edrocks joined #salt
16:40 spookah joined #salt
16:41 MatthewsFace joined #salt
16:41 intellix joined #salt
16:41 amcorreia joined #salt
16:41 MatthewsFace joined #salt
16:42 edrocks has anyone else had dns issues after upgrading a salt master on gce? I can only get test.ping to work with nodes I restarted after upgrading my master
16:43 scoates murrdoc: I did have refresh=True, but I stupidly used package.installed instead of pkg.installed. all fixed. thanks.
16:43 iggy edrocks: I just upgraded our entire QA from 2014.1 to 2015.2 and had no issues
16:43 iggy scoates: omg! I do that all the time
16:43 scoates very non-obvious. /-:
16:43 edrocks iggy: I upgraded instance type not salt
16:44 iggy scoates: it's more obvious in 2015.2
16:44 scoates oh? nice.
16:44 Twiglet how can I get mine info (eth0) for a single host? every single reference I can find online is looping over stuff
16:44 iggy edrocks: ahhh, yeah, never tried that... didn't even know you could do that
16:44 iggy scoates: yeah, the output is much improved in 2015.2
16:45 yomilk joined #salt
16:45 scoates … just updated to 2014.7 /-:
16:45 iggy Twiglet: try pasting an example
16:45 scoates actually, this project is on 0.17.2, even.
16:45 edrocks iggy: you have to shutdown then remove your instance without deleting the drives then make a new one which is a pain
16:45 edrocks but I'm running my ci on my salt master and it was maxing out everything because I used a small instance
16:46 Twiglet iggy: every single reference online is basically: {% for host, ip in salt['mine.get']('role:production-server', 'network.ip_addrs', expr_form='grain').items() %}
16:46 hal58th joined #salt
16:46 Twiglet whereas I'm just trying to get trhe infor for one: {{ salt['mine.get']('id:sensu-haproxy', 'network.ip_addrs', 'eth0' }}
16:47 Twiglet which gets nothing
16:47 iggy Twiglet: you can loop over a dict with 1 entry
16:47 denys joined #salt
16:47 iggy salt-call mine.get id:sensu-haproxy network.ip_addrs grain
16:48 Twiglet iggy: I could, however that seems completely wrong, if I were matching on grain and something came on with the same grain that could very easily break a config
16:48 shaggy_surfer joined #salt
16:49 Twiglet iggy: that looks to work, thanks
16:49 iggy well, unless you name your hosts "id:sensu-haproxy" (which is totally f'ed btw), that's already a grain lookup
16:49 shaggy_surfer joined #salt
16:49 Twiglet why is it f'ed?
16:50 iggy you can't put : in a hostname
16:51 Twiglet ah, yes. that was meant to match on hostname not the id grain
16:52 iggy then just take out the id: and it automatically does a "glob" lookup
16:52 Twiglet ah ha, excellent. many thanks
16:53 iggy salt-call mine.get sensu-haproxy network.ip_addrs
16:53 Twiglet the documentation gets a little sketchy as soon as you move away from baisc states, that or I am really bad at reading docs
16:54 aron_kexp joined #salt
16:54 iggy ehh, it just expects you to read a lot and be able to put a lot together
16:55 iggy but yeah, extended examples are an area the docs could use some help in
16:55 murrdoc saltstarters.org man
16:55 _prime_ joined #salt
16:55 iggy that's why I like the idea of saltstarters, but bechtoldt's states are always fugly
16:55 neogenix joined #salt
16:56 zer0def joined #salt
16:56 iggy and 90% of that site appears to be stuff from him
16:57 neogenix Ola Folks. Anyone have a good example of a testing suite to run through yaml parsing in sls files?
16:57 iggy nothing
16:57 edrocks is there any way to refresh the master ip without restarting?
16:58 hal58th probably not edrocks
16:58 iggy neogenix: you can get the rendered yaml from salt-call -l debug and run it through a yaml parser
16:58 lifelearner_ joined #salt
16:58 edrocks hal58th: thats crappy I'll file an issue
16:59 iggy edrocks: there's a 30 min timeout... but that might only be if you have multiple masters specified
17:00 lifelearner_ Are there any CM system that supports Windows and OS-X? Cfengine and Puppet only, maybe?
17:01 hal58th edrocks, a lot of changes to the master config require a restart. You are binding to an IP when the service starts. I don't think they would care if you have to restart a service.
17:01 iggy salt should support anything where python runs (theoretically... at some point...)
17:02 edrocks hal58th: I didnt change the config I upgraded my masters instance type. But now I have to restart each minion to get them to connect again
17:02 lifelearner_ iggy - But, Salt currently doesn't run on Windows and OS-X, right?
17:02 Ryan_Lane joined #salt
17:03 edrocks this is a crappy situation especially if your running things like influxdb which dont have good cluster support yet so your sol while it restarts
17:03 hal58th So you have the same IP or a different one edrocks?
17:03 iggy lifelearner_: yes it does
17:03 edrocks same hostname but it resolve to a different ip
17:04 debian112 I am targeting by grains in a state file, not sure if I am heading down the right road.
17:04 debian112 http://paste.debian.net/160596/
17:04 amcorreia joined #salt
17:04 edrocks at least i went a little overkill with my upgrade so I wont have to do it again soon
17:04 lifelearner_ iggy - even for the free community edition Salt?
17:04 iggy debian112: lol, nope
17:05 hal58th edrocks, your minions should self heal when looking up the DNS address. Not sure how long it takes. I misunderstood your issue
17:06 iggy debian112: {% if salt['grains.get']('nodename').endswith('net') %} maybe
17:06 edrocks hal58th: do you know how long it takes? iggy was saying maybe 30min
17:06 TheoSLC joined #salt
17:06 debian112 iggy. thanks. I will check that out
17:08 iggy lifelearner_: the only thing SSE has is a couple custom integrations (salesforce or something) and a dashboard/UI
17:09 hal58th not sure edrocks. look through minion config to see defaults.
17:10 younqcass joined #salt
17:10 TheRealBill joined #salt
17:11 edrocks hal58th: ok thanks for the info
17:11 iggy I also said that might only be when you have multiple master's specified
17:14 TheoSLC I just noticed that http.query is not available on windows?  It's all python, right?  why not?
17:15 wincyj joined #salt
17:16 shaggy_surfer joined #salt
17:18 Vynce joined #salt
17:19 iggy TheoSLC: the only thing I can see that disables http.query is if you have requests_lib = True and don't have it installed
17:21 theologian joined #salt
17:24 ajw0100 joined #salt
17:25 forrest joined #salt
17:27 aw110f joined #salt
17:27 sarlalian joined #salt
17:28 mpanetta Hey guys, quick Q... Does the creates attribute work in 2014.1.x?  It seems to be ignoring it for me...
17:28 iggy nope
17:28 mpanetta Ahh that would be why...
17:29 mpanetta Thanks iggy
17:29 iggy most of those requisites (beyond require and watch) were added in 2014.7
17:30 mpanetta Ah crap, including unless?
17:30 mpanetta I was just going to replavce it with an unless: test -e <file>
17:31 iggy yes
17:31 iggy if you are stuck with 2014.1, your best option is probably jinja if's
17:32 mpanetta well, right now the state just fails if the file is already there...
17:32 mpanetta Which is ok for now.
17:33 TheoSLC iggy: Here is what I think is going on with http.requests.  2014.7.2 comes with python 2.7 which does not include urllib2.  And requests is not installed.  I guess I need to install requests and use requests_lib = True.
17:33 iggy heh, I have a bunch like that "oh it'll stop failing when you let me upgrade salt"
17:33 mpanetta haha
17:34 Gareth morning morning
17:34 mpanetta apparently 'unless' works in this particular 2014.1.x version...
17:34 iggy TheoSLC: maybe, I'd have to look further, but that sounds like something to test I guess
17:35 iggy mpanetta: cmd.run?
17:35 mpanetta iggy: Yeah
17:35 cheus joined #salt
17:36 mpanetta So I replaced 'creates:' with 'unless: test -e'
17:36 iggy mpanetta: so yeah... that (and others) went global in 2014.7... before that, some of them did actually work with cmd.*
17:36 mpanetta Ah ok cool
17:36 iggy so it's kind of a crap shoot
17:37 mpanetta Well hopefully we will switch to 2014.7.x or whatever it is called now soon...
17:37 iggy look at the docs for the 2014.1 version if you want to see what should/shouldn't work with 2014.1
17:37 iggy we just jumped from 2014.1 to 2015.2 today
17:37 iggy it was painless
17:37 mpanetta Ok cool
17:38 TheoSLC iggy: I have a memory that at saltconf last week somebody said something about needing to install urllib2 or requests for http.query to work on windows.
17:38 TheoSLC iggy: I just can't find any documentation on it.
17:42 Saltuser124321 joined #salt
17:43 Salted54 joined #salt
17:43 baweaver joined #salt
17:43 jalbretsen joined #salt
17:45 ajw0100 joined #salt
17:45 hal58th The person to ask would be Joseph something. The writer, don't know his IRC name or last name though… iirc
17:49 leszq joined #salt
17:50 dcmorton left #salt
17:50 otter768 joined #salt
17:51 Salted54 I'd like to use the salt.client.LocalClient API from a computer that is not salt master. It seems to me that I accomplish this same functionality by ssh tunneling to the salt master and mounting a few key directories over sshfs. Would this bring up other issues?
17:51 linjan joined #salt
17:53 pmcg joined #salt
17:55 notbuu joined #salt
17:56 jespada joined #salt
17:58 rojem joined #salt
17:58 BrendanGilmore joined #salt
17:58 notbuu hi! in /srv/pillar/top.sls it is not possible to use wildcards for sls files?
18:00 Krzysztof joined #salt
18:00 Deevolution notbuu:  Not sure if you can wild card the file names or not, but you could certainly accomplish something along those lines via Jinja includes if need be.
18:02 aphoriser joined #salt
18:07 TheoSLC joined #salt
18:08 wolog joined #salt
18:08 jri joined #salt
18:11 jespada joined #salt
18:13 notbuu thanks! I will have a look!
18:15 MWheelz joined #salt
18:16 hal58th no wild cards for sls files notbuu
18:18 jbub joined #salt
18:18 ericof joined #salt
18:20 thekabal joined #salt
18:21 thekabal Hello. I am trying to get salt working, but upon launching it errors out with “global name ‘RSA’ is not defined”
18:22 hal58th upon launching what? Way more info needed
18:22 notbuu ok i will try to wrap around because i've to build the pillars out of an external system
18:22 Brew joined #salt
18:23 ltsampros joined #salt
18:23 ltsampros asd
18:23 ltsampros hallo
18:23 thekabal When I launch /usr/local/bin/salt-minion , it gives a warning, and then four info’s and then that error, and then restarts. This is on Ubuntu, and I’ve built from source, including installing zeromq. (I had no choice but to install from source)
18:24 clintberry joined #salt
18:25 ltsampros is there anyone here that specialized in salt-api ?
18:26 forrest thekabal: https://github.com/saltstack/salt/issues/16808
18:27 thekabal That.. was incredibly helpful. Thank you very much forrest
18:27 forrest thekabal: yeah np
18:31 TheoSLC Could somebody from Salt let me know what is the simplest way to get http.query to work on Windows?
18:31 cberndt joined #salt
18:31 hal58th ping basepi
18:32 basepi yo
18:32 TheoSLC basepi: what is the simplest way to get http.query to work on windows?
18:32 basepi hrm, not sure, I haven't done anything with salt and windows. let me do some checking
18:33 yomilk joined #salt
18:35 Vynce joined #salt
18:39 wolog joined #salt
18:40 TheoSLC basepi: Thanks.
18:41 mndo joined #salt
18:42 intellix joined #salt
18:45 baweaver joined #salt
18:46 thekabal It seems it is down to just ZeroMQ & PyZMQ. I’ve gotten the zeromq source, and run .configure, make, make install, and it installed without issue - but salt-minion versions report does not show it as installed. What did I miss?
18:49 Vynce did you restart salt-minion?
18:53 murrdoc joined #salt
18:54 badon_ joined #salt
18:54 basepi TheoSLC: I've been told you should install requests and set `request_lib: True` in your minion config.
18:55 TheoSLC basepi: okay.  now I just need to find an easy to way to install requests.
18:55 thekabal I got it! Many thanks folks!
18:57 rickflare joined #salt
18:58 TheoSLC basepi: why isn't requests packaged with the windows installer?
18:58 basepi TheoSLC: ya, from what I understand it can be a little bit difficult because we don't use the system python, we install python ourselves. I thought requests was actually packaged with the windows installer......
19:00 TheoSLC basepi: it's not.  I'll create an issue for it.
19:01 schlueter joined #salt
19:02 murrdoc basepi:  whats the nicest way to request info on issues
19:02 murrdoc is *bump* enough
19:02 murrdoc or
19:03 murrdoc totally a srs question basepi
19:04 basepi Bump can help. The problem is the number of issues we have open right now is incredibly overwhelming, so to date we've had to heavily prioritize core salt issues, and so some bugs in modules and so forth have been open for months or even years.
19:04 murrdoc i am ok with being told that it works as designed
19:04 murrdoc or that .. here go fix this here
19:04 murrdoc i just want to know whats the right way to do
19:05 murrdoc i ll *bump*
19:05 basepi Ya, if you're looking for guidance on how to fix an issue, first just ask on the issue, and if you never get an answer please feel free to ping me on it.
19:05 murrdoc i can work with that
19:05 murrdoc thanks basepi u alright despite what iggy goes on about
19:05 basepi Yes, definitely ignore iggy. ;)
19:06 basepi iggy: <3
19:08 SheetiS Anyone have a problem wherein running a salt \* saltutil.refresh_pillar causes a percentage of their minions to become incaccessible from the master until the minion service restarts when using an ext_pillar?
19:09 SheetiS This is with 2014.7.1 salt on all minions/masters for reference
19:09 Mena joined #salt
19:09 TheoSLC Is Salt moving to python 3 any time soon?
19:09 murrdoc SheetiS:  does it not do the same with the ext_pillar removed ?
19:09 basepi TheoSLC: Not any time soon. We have long term plans to support it alongside py2, but probably not for some time.
19:10 SheetiS It never did it until I added the ext pillar.
19:10 murrdoc what does the ext_pillar look like, python includes and things it does
19:10 TheoSLC https://github.com/saltstack/salt/issues/21506  Everybody give this a +1 :)
19:11 cheus Anyone know if pillar.get is cached locally on a minion during the run of a state/highstate or if each pillar.get is calling back to master?
19:12 murrdoc TheoSLC:  it doesnt use requests, it uses urllib2 from what i heard at presentation
19:12 murrdoc no ?
19:12 SheetiS It's super simple actually.  Just builds 'roles' from a yaml file (with the plan to move it to something like mongo going forward with a web_ui to assign roles to the minions). https://github.com/rfairburn/salt-roles-pillar/blob/master/pillar/roles.py
19:13 TheoSLC murrdoc: Requests or urllib2.  both are missing from the windows package.
19:13 murrdoc SheetiS:  file locking ?
19:14 murrdoc theo yeah
19:15 bhosmer joined #salt
19:15 bhosmer joined #salt
19:16 SheetiS murrdoc: good idea, but it's not showing open on lsof, and restarting the master doesn't seem to matter.
19:19 basepi murrdoc: joseph was telling me that he doesn't have it working on windows without requests yet. something about getting the certs on the windows box being difficult.
19:19 basepi urllib2 is stdlib, so it should be there.
19:20 nesv Has anyone experienced salt-cloud using a lot of memory to do it's thing?
19:20 nesv Guh, and by "its thing" I mean, launching a new node
19:21 nesv On my salt-master, I was noticing salt-cloud consuming 1.2GB of RAM when trying to launch just one node. On a server with 4GB of RAM, and idling at ~800MB used, trying to start three nodes in parallel failed because the salt-master ran out of RAM
19:22 murrdoc yeah i agree basepi
19:22 murrdoc he mentioned something along those lines in teh conf too
19:22 Pixionus Anyone have anything to say on Learning Saltstack Paperback by Colton Myers?  Like anyone have the book and can attest to it being a good one?
19:22 forrest TheoSLC: Regarding that issue, why can't you just install requests?
19:23 SheetiS murrdoc: also wouldn't a file lock raise an Exception on my open('file.yaml', 'r') or whateve ranyhow and then log as critical to the master?
19:23 notbuu @Deevolution Jinja magic works
19:23 notbuu Deevolution Jinja magic works
19:24 basepi Pixionus: Most people I've talked to say it's a good starting resource. Of course, I don't know if I can trust them -- they may just be being nice to me since I'm the author.
19:24 basepi ;)
19:24 notbuu left #salt
19:24 murrdoc SheetiS:  i started reading the pyaml code, then went back to work, but basically when u do salt '*' refresh_pillar i am assuming bunch of threads all open up to read the same file
19:25 murrdoc and that *might* be causing issues
19:25 murrdoc i cant say for a fact they are , but i think might
19:28 iggy I heard that...
19:29 Pixionus Well basepi I think I'm going to go ahead and buy it and Ill give you candid feedback when I actually use it.
19:30 Pixionus Wish there were more used copies though ::grumbles about book prices and low pay::
19:30 ckao joined #salt
19:33 Deevolution notbuu:  Excellent.
19:34 basepi Pixionus: I appreciate it! And yes, it was only released in January, so I don't expect there to be a thriving used market yet.
19:34 basepi Though it is a pretty short book.
19:35 basepi So that market may start up soon.
19:36 Pixionus How in depth do you go into writing states and managing things like pillars etc?
19:37 Pixionus That's about where I am these days.  Remote execution and the like I use daily and am comfortable with.
19:37 N-Mi joined #salt
19:38 andygrunwald joined #salt
19:38 SheetiS murrdoc: I'll try and import salt.utils and use salt.utils.fopen, but from what I can see in the read-only case I'm not gaining myself anything.  In any case, it would look more 'salty' to handle that way I suppose.
19:40 baweaver joined #salt
19:40 Pixionus Also need a sort of introduction to the cooler and most useful of the modules since I'm sure there are a lot I don't know much about and the docs are a bit dense.
19:41 murrdoc SheetiS:  like i said, i dont know but i wont be surprised that deep in cyaml
19:41 anotherZero joined #salt
19:43 scoates meta question: what are sane people using for self-hosted monitoring/alerting?
19:43 iggy I always find it weird when I hear about people using the remote execution side of salt well before states
19:44 iggy scoates: sane people don't...
19:44 denys joined #salt
19:44 ericof joined #salt
19:46 Pixionus iggy: Someone else in the company set up our highstate and I never had to dive into it.  However the remote exec comes in pretty handy when I'm troubleshooting servers that aren't behaving (have a fleet of embedded servers that can act up for any number of known reasons that I haven't gotten around to automating for yet since I'm not a programmer yet and the programmers are all busy doing actual dev)
19:46 Pixionus Kinda underskilled for this stuff but i'm new and we are a start up.  Ill catch up.
19:47 tmh_ joined #salt
19:47 basepi Pixionus: It starts by assuming no salt knowledge, and works through remote execution and the state system, and also has a quick chapter on salt cloud, and one on the reactor/event system. It doesn't explore very many modules, because it's going for core concepts, not an exhaustive guide.
19:47 Mena left #salt
19:47 basepi It does have quite a bit about the state system.
19:48 murrdoc what book
19:49 lothiraldan joined #salt
19:49 Pixionus fair enough.  Hopefully it will be what I need for now and if there's anything I'm still lacking at the end of it, I'm sure Salt Essentials will come in handy then.
19:49 Pixionus murrdoc: Saltstack by Colton Myers
19:50 basepi "Learning SaltStack", specifically
19:50 Pixionus Oops yeah.
19:50 murrdoc this is new
19:50 basepi Joseph Hall just started writing Mastering SaltStack, which should be out later this year.
19:50 basepi So that will be cool
19:51 otter768 joined #salt
19:51 basepi And obviously Salt Essentials from O'Reilly will be out in May, supposedly. Though I've heard it may be behind schedule.
19:51 amcorreia_ joined #salt
19:52 Pixionus It looks to be a somewhat small book as well.
19:52 Pixionus 100 more pages then yours, but not one of the huge tomes I'm used to seeing
19:53 basepi Nope, salt's still pretty new as far as tech goes, hopefully those books will come in the coming years.
19:53 basepi It's hard because salt moves SO FAST
19:53 iggy salt is also a moving target still... the more specific you get, the quicker it gets out of date
19:53 Pixionus Speaking of which anyone have any advice on a good python book?  Let's pretend I didn't have a couple of years of C++ in college and may as well be new to programming these days
19:54 Pixionus Yeah.  It's one of the things I like about it but it makes documentation and diving in all the more difficult.
19:54 SheetiS Pixionus: I've not done it but I have heard good things about http://learnpythonthehardway.org/book/.
19:54 Pixionus (the speed at which it is changing and adapting)
19:54 MatthewsFace joined #salt
19:55 SheetiS #python here on freenode seems to recommend it 24x7.
19:55 Pixionus Cool.  thanks
19:55 basepi I've also heard good things about "the hard way"
19:56 basepi Back in the day I went through head first python. It was decent
19:56 tmh_ joined #salt
19:56 evidence 2015.2.0rc1 looking solid, or waiting on any fixes?
19:56 basepi And the official tutorial on the python website is pretty good if you're already a programmer. It's boring, but fairly thorough
19:57 Pixionus I like physiucal books because the tactileness helps my memory for some reason and I can easily note and mark things of note
19:57 basepi evidence: it's pretty solid. We should cut 2014.2.0rc2 in the next day or two
19:57 basepi Has a number of fixes, but nothing too critical I think.
19:57 gar joined #salt
19:57 evidence basepi: good deal, thx
19:57 Pixionus basepi is the 0mq replacement "just usable" or pretty stable?
19:58 Pixionus We've been looking to move over but can't put production on the bleeding edge
19:58 basepi As far as I know it's actually getting to be pretty awesome. Still beta, for sure, but pretty stable from what I understand.
19:58 iggy in 2015.2, probably "works for some"
19:58 basepi I haven't used it heavily yet
19:58 Pixionus Cool.  I'll start playing with it in preparation for when it moves out of beta.
20:00 iggy we probably won't move to it until Boron(2015.8?)
20:00 SheetiS I probably do so many things terribly in python (and other languages) since I tend to mainly come with a task or a problem and 'just do it' as oppose to do course work or school.  I just find I lam motivated if I need to accomplish something specific.
20:01 iggy I'm hoping just the keepalives on both sockets helps
20:02 iggy (thanks for routing _everything_ GCE)
20:02 murrdoc bi directional pips
20:02 TaiSHi https://speakerdeck.com/ipmb/monitoring-infrastructure-with-saltstack <-- that was awesome, anyone has news on this? If it had further development or so ?
20:02 Pixionus What can you tell me about the keepalives issue?  Is there a ticket you can point me too?
20:02 basepi iggy: Let's see, since 2015.2 didn't hit February, Boron would now be targeting 2015.9
20:02 basepi With Beryllium in 2015.6
20:02 Pixionus I assume it's part of the communication thing in 0mq where sometimes things weren't responding first time
20:03 iggy no, it was something Mike mentioned in his troubleshooting talk
20:03 basepi Pixionus: cachedout is currently working on some keepalive work which apparently is improving stuff
20:03 basepi cachedout == mike
20:03 basepi jacksontj is also working on a raw tcp replacement in salt for ZMQ, so that should be cool.
20:03 iggy said it was in devel (and indeed we've seen improvements in devel)
20:04 iggy so we'll have 0mq, raet, and something_else?
20:04 basepi my_awesome_transport
20:04 Pixionus I mean what was the issue being solved there?
20:04 basepi ZMQ is fairly.....opaque. Hard to debug
20:04 iggy we "lose" minions a lot
20:05 Pixionus TaiSHi: I was looking at that a while back but I think the beacon should actually work better.  Haven't had the chance to dive into it yet though
20:05 iggy they usually come back alive after a few test.ping's
20:05 Pixionus iggy: ok, so it is related to what I was thinking.
20:05 Pixionus Cool.   Ill be glad when it comes out.
20:06 Pixionus will the keepalive fix be in the next release?
20:06 iggy I'm not sure
20:06 TaiSHi Pixionus: Beacons still need a lot of dev ahead, although they look really good
20:06 basepi If we continue to see these improvements, we may backport it to 2015.2. We shall see.
20:07 basepi Ya, what you guys might not realize is that beacons and engines both went in only a few weeks ago. Engines were only committed last Monday. =P
20:07 basepi They're *very* new.
20:07 Pixionus Looking forward to it.  Ill keep the hope alive.
20:07 iggy I'm running git 2015.2 in qa and it's been pretty good today (although I may not have given them enough time to die off)
20:07 TaiSHi basepi: I agree, so for now I might have to stick with icinga2
20:08 Pixionus yeah.  they were JUST out when I went to SCALE so I've left them to wallow at the bottom of my list of tolearn's
20:08 murrdoc TaiSHi:  did u make a formula for adding tests to icinga2
20:08 murrdoc i d like to check it out if possible
20:08 Pixionus But that said, since they will likely be my solution of choice I'm content with our scripted shitty monitoring for now till I can use them
20:09 TaiSHi murrdoc: I'm stuck lately with icinga2 formula, haven't moved forward with it. Although I cleaned up nginx-formula :P
20:09 TaiSHi murrdoc: I plan on advancing at least with basic agent install this week
20:09 Pixionus by the time I install and learn a new set of tools for this, I think beacons might be usable and a bit more fleshed out.
20:10 basepi Ooh, if someone new forks salt right now, they'll be the 2000'th fork.
20:10 TaiSHi Pixionus: nrpe is quite the standard, problem with it is scaling (in previous experiences at least)
20:11 iggy there you go basepi
20:12 iggy (I was suprised I didn't already have it forked)
20:12 murrdoc iggy:  https://groups.google.com/forum/#!topic/salt-formulas/2B_7EwgCT7s
20:12 iggy I'm subscribed
20:12 murrdoc my bad
20:12 * murrdoc retreats slowl
20:12 thedodd joined #salt
20:13 Pixionus I was looking at nagios, but decided like I said above.
20:13 iggy I haven't read the last post (although I did see it)... it looked like something I probably need to read on an empty stomach
20:13 Pixionus Right now I use Monit and I was thinking of throwing some ELKstack in the mix
20:13 Pixionus Figure it's a little less redundant with the roadmap of Salt
20:14 Pixionus eventually I want to replace Monit with Salt as well
20:15 Pixionus We run on embedded machines so I want to keep my software stack super slim and as long as Saltstack doesn't become top heavy/take on too much, it looks to become my swiss army knife.
20:19 basepi iggy: yesssssss
20:22 yomilk joined #salt
20:26 jdesilet joined #salt
20:29 giantlock joined #salt
20:29 cody__ joined #salt
20:31 SheetiS murrdoc: Looks like using salt.utils.fopen() instead of open() took care of it.  Not yaml's fault then after all.
20:31 murrdoc SheetiS:  nice find
20:32 bhosmer joined #salt
20:33 opennode joined #salt
20:33 SheetiS it defaults to not locking "lock = kwargs.pop('lock', False)", and I didn't turn it on, but it does other 'magic' with the file handle that I haven't looked into too closely.
20:34 SheetiS So it wasn't file locks per say, but something that someone had already experienced in the past and fixed I guess :D
20:35 SheetiS https://github.com/saltstack/salt/blob/v2014.7.1/salt/utils/__init__.py#L1051-L1078 in case you are curious.
20:35 aphorise joined #salt
20:36 murrdoc nice
20:36 murrdoc btw there is a project to handle pillars like classes
20:36 andreyst joined #salt
20:36 murrdoc i forget what its called, its improved a lot, might be close to what u trying to do with your roles yaml
20:37 _JZ_ joined #salt
20:37 iggy reclass?
20:37 murrdoc ^
20:38 murrdoc SheetiS:  thanks for link
20:39 preachermanx joined #salt
20:39 preachermanx Greetings all.
20:40 preachermanx I am having some issues with a ~2000 node saltstack was wondering if anyone is around who could answer some brief questions.
20:40 numkem_ has anyone ran into issues with services (like asterisk) that won't return anything? It just hangs there
20:41 SheetiS murrdoc, iggy:  I might look into reclass.
20:42 druonysus joined #salt
20:42 druonysus joined #salt
20:42 iggy preachermanx: shoot
20:42 SheetiS numkem_: I've seen some sysvinit scripts that do "Bad Things"™ cause a python subprocess to never return.  I'm not saying that this is the case, but most of my service.running items that have hung in the past have been due to that.
20:43 numkem_ SheetiS: have you managed to do anything about it?
20:43 SheetiS numkem_: for me anyhow.
20:43 matthew-parlette joined #salt
20:43 relidy joined #salt
20:43 SheetiS numkem_: when I had a problem with the nginx init script on Amazon Linux doing it, I just switched it to an upstart job :(
20:44 numkem_ SheetiS: thats what I was thinking but I wasn't sure if I wanted to do that or not
20:45 micko joined #salt
20:46 rojem joined #salt
20:46 preachermanx iggy: thanks; Well i am constantly getting my saltmaster flooded even after adding the reccomended tweaks to my minions, also things done against all nodes only return values for the first ~20-30 hosts then then it silently stops.
20:46 cody__ left #salt
20:46 faliarin /znc disconnect
20:49 neogenix Thatch talked about multi-master at SaltConf, anyone here recall what sort of timeframe he was talking about ?
20:49 iggy neogenix: long term
20:49 neogenix iggy: I thought he was saying salt.next();
20:49 iggy with what all they have left to do, I wouldn't expect that this year
20:49 neogenix (not .2, but .2++
20:50 iggy I can't see that happening
20:50 nickg for some reason salt does not like my service..when i have service.enabled start the service, the salt-minion hangs.  it's stuck with a zombie under it that is the init.d script for the service.. any ideas?
20:50 preachermanx nickg: pastebin of the service script?
20:50 neogenix iggy: ah. Buggery. Let me to work out if I can get some sort of multi-master, or HA dohicky.
20:51 hal58th nickg, I ran into something similar for a version of mariadb. Just had to roll back to a previous version because it seemed like they changed something with the init script that python was having trouble with
20:51 iggy preachermanx: what's your master look like ("hardware" wise)?
20:51 preachermanx nickg: i would recon there may be a problem with the service script returning properly
20:51 hal58th nickg https://github.com/saltstack/salt/issues/16153
20:51 garthk joined #salt
20:51 opennode joined #salt
20:52 aparsons joined #salt
20:52 iggy neogenix: do you _require_ multi-master right now? For us it's a nice to have, but we can wait till it's ready.
20:52 nickg preachermanx: i know it's properly starting the service and returning 0, or 1 based on the service's startup.
20:53 _prime_ joined #salt
20:53 SheetiS nickg: I've been replacing sysvinit scripts with upstart jobs when I run into this to avoid the python subprocess from never returning.
20:53 SheetiS from what i've seen is it has to do with something in the init script and subprocess not getting along.
20:53 preachermanx iggy: a lot.   I was doing this on a vmware vSphere host but since moved it to a physical host and are still seeing the issues.  Intel(R) Core(TM) i7-5930K CPU @ 3.50GHz  64GB of DDR4 128GB SSD for root disk SL 6.5
20:54 aparsons joined #salt
20:54 nickg SheetiS: I tried to use upstart, but it was a disaster
20:54 neogenix iggy: it's more I'd prefer to pull stuff out of salt from a local master, instead of a remote one. I'll give it a shot with multi-master, and see what happens.
20:54 iggy preachermanx: and you've done the normal stuff in the docs (increase open files, etc.)?
20:55 iggy neogenix: isn't that more of a syndic thing (than multi-master)?
20:55 SheetiS nickg: if the init script is busy, switching to upstart can be tricky for sure.
20:55 preachermanx iggy: yep
20:55 nickg hal58th: it looks like that may have helped it.. I added 2>&1  to the tail end of the startup line that had > /dev/null and it worked fine
20:56 hal58th nickg, yeap but now you have to fix the package, if you are using a package that is
20:56 murrdoc1 joined #salt
20:56 nickg SheetiS: I couldn't locate a restart feature with upstart... which was a hinderance.  also upstart would think a service was running that wasnt running and the only way to fix it is to run some ghetto script to try and fill up the PID list to get back to that pid
20:56 nickg hal58th: yeah it's a custom service so it's all internal.
20:56 murrdoc1 joined #salt
20:57 hal58th ah, then I am glad I could help nickg :)
20:57 _JZ_ joined #salt
20:57 nickg really weird situation.. since nothing is sent over stderr
20:59 iggy preachermanx: you might just be screwed till raet... if you've got 2000 minions, that's going to be like 8-12k context switches per second when the minions are doing stuff
20:59 preachermanx iggy: https://gist.github.com/preachermanx/3bbc3fc9045173d93dd6
21:01 nickg did anyone add a delay interval to the -b BATCH setting for the cli salt yet?
21:01 preachermanx iggy: well that would suck.... what happens to the "tested daily with 10,000 minions" :/
21:02 iggy that's using syndics I'd guess
21:02 iggy or multi-master
21:02 iggy or probably both
21:03 SheetiS preachermanx: is it when you try and 'do things' that you have issues, or is just the traffic of keys, pings, etc that are barfing?
21:03 numkem_ SheetiS: that worked, thanks man
21:03 preachermanx SheetiS: do things
21:03 preachermanx SheetiS: at least thats when its noticible
21:04 iggy use batch?
21:04 SheetiS preachermanx: I typically try and batch my highstates to help with that.  say "salt \* state.highstate --batch 10" if I were running 10 processes.  You are beefy enough you might be able to batch much bigger.
21:04 opennode If useful to anybody - we have developed a light wrapper for safeguarding against common human mistakes when working with salt environments in production deployments - details are here: http://opennodecloud.com/howto/2015/02/15/howto-salt-safeguard.html
21:04 SheetiS iggy: you think as I do :D
21:04 chalcy0n yo iggy :)
21:05 iggy Hey!
21:06 iggy chalcy0n: You recovered from the conf yet?
21:07 chalcy0n iggy: yep finaly lol
21:08 TheoSLC joined #salt
21:08 hal58th Just in the states long enough to get used to the time difference.
21:08 iggy ^
21:09 iggy I was going to say, almost over the jetlag from flying here before you flew back
21:09 chalcy0n haha yep that was so yrue
21:09 timoguin joined #salt
21:10 chalcy0n but the conf was awesome
21:12 iggy yeah, I had a great time (halfway) looking forward to next year
21:12 chalcy0n yeah me too
21:12 aw110f joined #salt
21:13 chalcy0n already played with the beacons?
21:14 iggy not yet, got back, did my yearly review and started on the next ops release later this month :/
21:14 iggy but part of that release is getting everything upgraded to 2015.2
21:14 iggy so it'll make more sense after that anyway
21:15 chalcy0n yeah true
21:17 * iggy wishes he'd gotten more interest in the open positions
21:17 murrdoc1 that one dude was interested
21:17 aboe joined #salt
21:17 murrdoc1 the argentinian dude
21:18 iggy I got 1 resume so far (looks good, but I don't think my company is setup to employee people outside .us)
21:18 iggy yeah
21:18 dendazen joined #salt
21:19 dendazen when i run ‘salt * state.highstate’
21:20 dendazen does it log the output anywhere to the disk?
21:20 thedodd joined #salt
21:20 dendazen besides stdout?
21:20 andreyst joined #salt
21:20 evilrob so according to the docs, mysql is a default included module.  when I try mysql.db_list I get '"mysql.db_list" is not available.'  I've made sure the mysql config is in the minion config, restarted the minion, still get the "not available" message.
21:20 dendazen or it is stdout only
21:20 evilrob ideas?
21:20 iggy dendazen: job cache
21:20 dendazen where would i get it from
21:20 iggy evilrob: install the requisites
21:21 dendazen “/var/cache/salt/proc”?
21:21 iggy dendazen: look at the saltutil module and jobs runner
21:21 SheetiS http://docs.saltstack.com/en/latest/topics/jobs/ (salt-run jobs.list_jobs and salt-run jobs.lookup_jid <jid>)
21:21 hal58th evilrob check the note http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.mysql.html
21:22 iggy python-MySQLdb or whatever
21:22 dendazen Thank you.
21:23 peters-tx joined #salt
21:25 schlueter joined #salt
21:25 preachermanx Thanks btw guys for the input
21:26 whiteinge ltsampros: did you get your salt-api question answered?
21:27 evilrob so I installed python-mysqldb and it started working (thank you) but this exposes something I misunderstood.  I thought sys.doc showed you modules on the master that were available, but nothing mysql-ish showed up until I installed that package on a minion.
21:27 evilrob so is sys.doc showing what's available on the master or on a minion?
21:30 mafro joined #salt
21:31 chalcy0n evilrob: on the minions which you selected with the salt command
21:32 chalcy0n i.e salt minion1 sys.doc shows the modules on minion1
21:33 chalcy0n some modules will not show if you dont install the required python libs
21:34 baweaver joined #salt
21:38 yomilk joined #salt
21:38 Vynce joined #salt
21:43 Singularo joined #salt
21:49 shaggy_surfer joined #salt
21:52 otter768 joined #salt
21:52 aw110f joined #salt
21:57 schlueter joined #salt
21:58 bluenemo joined #salt
22:00 conan_the_destro joined #salt
22:00 mimianddaniel joined #salt
22:00 baweaver joined #salt
22:02 jcsp joined #salt
22:05 aquinas joined #salt
22:06 mafro joined #salt
22:06 schlueter joined #salt
22:07 linjan joined #salt
22:08 murrdoc joined #salt
22:11 debian112 question for the group. I have a state that works on new centos 7 and 6 servers, but fail on debian 7 servers? How can troubleshoot this:
22:11 debian112 http://paste.debian.net/160665/
22:11 rojem joined #salt
22:11 schlueter1 joined #salt
22:12 debian112 Output: http://paste.debian.net/160666/
22:12 aquassaut joined #salt
22:13 IOMonster joined #salt
22:13 IOMonster joined #salt
22:13 iggy protip: gist allows multiple files per paste
22:14 iggy you need an else
22:15 forrest Yeah seems like it just isn't defining dns1 because it doesn't match either of those two checks you have
22:16 iggy not sure if that's a direct C&P, but you're missing salt['grains.get'] in a few places
22:22 mimianddaniel hey iggy did you say you are running rc2?
22:23 iggy installed from git
22:23 iggy so I guess, yeah
22:23 mimianddaniel seeing this error in the master log ?
22:23 mimianddaniel The module 'salt.loaded.int.module.ipmi' is showing some bad usage. It's __virtualname__ attribute is set to 'ipmi' yet the __virt ual__() function is returning (False, 'No module named pyghmi.ipmi'). These values should match!
22:24 beauzeaux joined #salt
22:25 iggy the log file is huge, but nothing yet
22:26 murrdoc iggy:  if i include a variable in an sls, how do i pass it on to included sls files
22:26 murrdoc like is there a way
22:26 iggy with context?
22:26 iggy but I think that's default
22:27 murrdoc this is my current state
22:27 zer0def joined #salt
22:27 shaggy_surfer joined #salt
22:27 murrdoc {% from blah include var_blah with context %}
22:27 murrdoc include:
22:27 murrdoc - formula.blah.packages
22:27 murrdoc - formula.blah.configs
22:28 murrdoc from what i can see, i need to put the {% from blah %} in each sls file
22:28 iggy oh, that direction
22:29 iggy not sure
22:30 esogas_ joined #salt
22:32 TaiSHi murrdoc: Do you have a current icinga2 setup? I might use you later for beta testing :P
22:32 peters-tx Ugh.  http://fpaste.org/196468/   I'm trying to get a state such that systems affected get etckeeper installed.  And if they get etckeeper, if they don't have a '/etc/.git/' directory, it will run init
22:32 peters-tx Very much not working :(
22:33 peters-tx Tried a bazillion different methods
22:34 iggy weird requires
22:34 peters-tx iggy, Well, now yes, but that's through trying everything else 8/
22:34 yomilk joined #salt
22:35 iggy why not reorder and use salt's natural ordering to your advantage
22:35 peters-tx Well, other things at least
22:35 iggy what version of salt?
22:35 peters-tx Ok, so just Install the pkg and then Run the command?
22:35 peters-tx Latest via EPEL, 2014.7.1 I think
22:36 druonysuse joined #salt
22:36 aw110f joined #salt
22:39 timoguin joined #salt
22:41 pravka joined #salt
22:41 esogas_ joined #salt
22:41 iggy http://fpaste.org/196470/02729914/
22:42 iggy install etckeeper and only run etckeeper init if /etc/.git doesn't exist
22:42 peters-tx Woah
22:42 peters-tx Creates
22:42 peters-tx Hmmmm
22:43 peters-tx BAM
22:43 peters-tx yay :)
22:43 peters-tx iggy, Yessir, that was the magic I was missing
22:43 peters-tx thanks, "create" /me makes notes
22:44 iggy SSCE!
22:45 peters-tx Well me too, somehow... apparently  :)  accident?
22:45 baweaver joined #salt
22:45 * iggy doesn't mention that he barely passed
22:45 peters-tx I think it was light on the specifics, like this, of states
22:45 iggy it was light on anything post 0.17 I think
22:46 peters-tx Good thing too
22:46 iggy (not sure if on purpose or just old questions)
22:46 peters-tx I also meant to ask someone "how long does the SSCE last"
22:47 iggy *
22:48 peters-tx NaN  ?  well, that's what it is right now I guess
22:48 iggy I'm not sure they have any plans of making them end right now, but you should probably verify that with someone from SSE
22:48 murrdoc as long as needed peters
22:48 murrdoc as long as needed
22:48 peters-tx 8D
22:48 StDiluted joined #salt
22:48 murrdoc are u in texas
22:49 murrdoc so is iggy
22:49 murrdoc he needs friends
22:50 iggy mimianddaniel: grep finally finished, nope
22:50 peters-tx Yah, but he's like 5 hours away
22:50 iggy mimianddaniel: but I have noticed some chatty log messages in general
22:50 * peters-tx waves
22:51 iggy lol, I get that from recruiters all the time "I have an open position in DFW"... This isn't f'ing delaware, you can't get from one side to the other in <30 mins
22:52 sarlalian joined #salt
22:52 murrdoc 5 hours in texas is nothing
22:52 murrdoc jump in your truck
22:53 murrdoc hopefully a lifted raptor
22:53 murrdoc and march on
22:53 iggy on a bad day it takes that long to get from one side of Houston to the other
22:53 hal58th iggy, you should start every response with "SSCE POWERS ACTIVATE"
22:53 preachermanx iggy: so. so. true.    Austin here
22:53 murrdoc or
22:53 murrdoc solution
22:53 murrdoc — SSCE SAID SO
22:58 KyleG joined #salt
22:58 KyleG joined #salt
22:58 murrdoc thoughts on including a variable once and passing it around between states ?
23:00 iggy I'm glad I didn't make that formulas roundtable... I disagree with 90% of that
23:01 warpaint joined #salt
23:01 murrdoc wonder what forrest thinks of it
23:01 TaiSHi iggy: Didn't really follow the post, but I'm guessing they want to make formulas more 'module-like' ?
23:02 murrdoc as they should be
23:02 ajw0100 joined #salt
23:02 murrdoc #devlife
23:02 murrdoc TaiSHi:  no icinga2 setup, i just want tojack your code
23:02 forrest murrdoc: what's up?
23:03 murrdoc on the salt-formulas mailing forrest there is a convo about the impromptu round table
23:03 TaiSHi murrdoc: I'll refrain from answering that with a joke :P
23:04 forrest Ahh
23:04 forrest taking a look now
23:04 __number5__ TaiSHi: is that joke related to King Arthur?
23:04 iggy I feel like Nitin should probably weigh in
23:05 TaiSHi __number5__: No, more of a genital joke
23:05 murrdoc typical south american men
23:05 murrdoc i tell u
23:05 murrdoc *snerk*
23:06 forrest point 5 makes me want to throw up murrdoc
23:06 forrest physically ill
23:06 forrest It's like someone said 'ahh you know how easy salt is to understand, fuck it! Make everything a variabalized template so our lazy ass developers can avoid writing an extra line of code!'
23:07 forrest Basically any time an ng formula is referenced, I'M OUT
23:07 iggy ^5
23:07 iggy I knew I liked this guy for some reason
23:08 murrdoc i agree on the ng part
23:08 forrest Chad is a good dude, but I hate ng formulas, and I hate using jinja when you don't have to
23:09 forrest Seems like a lot of this stems from ng style formulas, and trying to make those more the 'norm'
23:09 forrest Every time that comes up I just want to say 'GOOD SIR, PLEASE DO NOT SCREW THIS UP BECAUSE YOUR APP IS BAD AND REQUIRES BULLSHIT SETTINGS!'
23:09 TaiSHi forrest: hey! I use nginx.ng :P
23:09 murrdoc https://github.com/saltstack-formulas/php-formula/blob/master/php/ng/map.jinja
23:09 murrdoc oh lord
23:09 lothiraldan joined #salt
23:10 forrest TaiSHi: That's fine
23:10 forrest Doesn't mean I like the idea of our standard formulas going that way
23:10 iggy I hate looking at a formula that should be easy, trying to figure out what it's doing and giving up and rolling my own
23:10 forrest the ng stuff to me has always been 'advanced'
23:10 iggy defeats the point of formulas
23:10 forrest iggy: Exactly
23:10 forrest They are supposed to be simple, then you can work on whatever you want after you fork it
23:10 TaiSHi Ah, I agree on that iggy, a while back we had a conversation of the sorts about nginx.ng and php.ng
23:10 forrest they're just a starting poin
23:10 forrest *t
23:11 forrest I mean TaiSHi, think if I showed someone nginx.ng, they would run, versus normal nginx? Pretty easy
23:11 forrest So I can't support ng becoming any sort of standard/norm. I'm too stupid to handle it.
23:11 iggy then there's the people that want dep resolution, installer scripts, etc. for formulas
23:11 murrdoc ng should have been a brnach
23:11 bfoxwell joined #salt
23:11 murrdoc instead of shoe horned into existing formula
23:11 murrdoc talk about clutter
23:11 murrdoc imho
23:12 iggy if formulas become that, I'll stop using formulas
23:12 murrdoc u will stop using saltstack-formulas
23:12 murrdoc not formulas in general
23:12 forrest murrdoc: Agreed, honestly I believe I suggested making ng it's own repo when we first started seeing those
23:12 bluenemo joined #salt
23:13 forrest I don't get why we can't have more repos
23:13 forrest infinite repos!
23:13 iggy well, if that's what formulas become, there will be a void to fill between nothing and that... I wouldn't expect them to be called formulas
23:13 forrest yep
23:14 murrdoc i agree with u guys
23:14 murrdoc but
23:14 iggy maybe we should tell the people that want npm for salt formulas to start calling them something else... formulas are already an existing thing
23:14 murrdoc i am moving all my default pillars to a default.sls
23:14 murrdoc that i pull into my map.jinja
23:14 murrdoc and then merge in the pillar
23:14 murrdoc yes its slow forrest but its the lesser of two evils for me
23:14 murrdoc for now
23:15 forrest murrdoc: Yeah I think we talked about that before right? Just like the settings.sls
23:15 forrest it's not great, but ffs it's better than shoving everything in the map
23:15 murrdoc no not like settings.sls
23:15 iggy my problem with some of the formulas that use settings.sls is the settings file is longer than the damned state file
23:15 iggy that's ridiculous
23:16 murrdoc using http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.jinja.html import_yaml
23:16 murrdoc so its not a map with logic
23:16 forrest iggy: I agree. but really it's just an 'example' settings file
23:16 murrdoc just a default pillar yaml
23:16 forrest yeah that's not bad murrdoc
23:16 forrest like I said, it isn't all in the map, it's okay with me
23:16 murrdoc basically u know the pillar.example
23:16 forrest just grosses me out, non-os stuff in the map? Ugh
23:16 * iggy really wants to see the EMC talk
23:17 forrest is that the one about shoving everything in the map iggy?
23:17 murrdoc i am moving that into a defaults.yml in the formula and then import_yaml it
23:17 forrest murrdoc: gotcha
23:17 forrest murrdoc: Did you test how fast it is to render versus all in the map? I imagine your method would be faster.
23:17 murrdoc so it becomes
23:17 murrdoc set formual_settings to do grain:filterby for os
23:17 iggy it was about formulas in general... I'm not sure what solution they suggested
23:18 murrdoc load in defaults.yml
23:18 murrdoc apply os.map to it
23:18 murrdoc then merge in pillar
23:18 murrdoc same thing thats done as one function
23:18 forrest murrdoc: Yeah that's a good option
23:19 murrdoc phew
23:19 iggy a config.get that was smart enough to look at defaults.yaml?
23:19 * murrdoc totes happier that forrest approves
23:19 murrdoc also http://whatthecommit.com/2461c2e5456a9e486c1a2207f0efdd43
23:19 forrest murrdoc: You gotta do what you gotta do to make it work, at least your method is sensible instead of being lazy.
23:20 iggy and partially understandable
23:20 murrdoc wait 2 ssces approve
23:20 * murrdoc makes himself ssce approved cert
23:20 murrdoc :)
23:21 iggy I hope to never see a settings.sls in formulas that are as simple as the aptly formula
23:21 TheoSLC joined #salt
23:21 bhosmer__ joined #salt
23:21 forrest iggy: Why would you need a settings file for that
23:21 forrest lol
23:21 __number5__ iggy: where is that aptly formula? will need one soon
23:22 iggy __number5__: saltstack-formulas
23:22 forrest __number5__: https://github.com/gravyboat/aptly-formula/tree/master/aptly
23:22 __number5__ thanks iggy forrest
23:22 forrest yup
23:22 forrest aptly is pretty great
23:22 iggy forrest: that seems like what everybody (that isn't us) wants to do
23:23 forrest They can, just make a new repo called <formula-name>-ng
23:23 forrest formulas need to be simple, you can't just use an ng formula out of the box with pillar filled in
23:23 forrest it shits itself
23:23 forrest and even if that was fixed, they're too damn complicated to understand easily and quickly
23:23 murrdoc yes it do
23:23 forrest I find them frustrating
23:24 iggy then we need to figure out a way to filter *-ng formulas out of the index page
23:25 murrdoc they should be their own repo
23:26 forrest oh __number5__ use the main repo not my forked one for aptly, sorry :\
23:26 forrest was on the wrong page
23:27 iggy heh, that whatthecommit.com site is pretty funny
23:28 forrest iggy: I'm updating the aptly integration docs with the link to the saltstack formula, do you want your name first or second?
23:28 murrdoc https://groups.google.com/forum/#!topic/salt-formulas/2B_7EwgCT7s
23:28 mafro joined #salt
23:28 forrest -   [SaltStack Formula](https://github.com/saltstack-formulas/aptly-formula) by Forrest Alvarez and Brian Jackson
23:29 forrest Or do you want it the other way? I don't really care since you did the clean up to make it work you can decide.
23:29 murrdoc who is this brian jackson person
23:29 iggy what you have is fine
23:29 forrest cool
23:31 forrest iggy: https://github.com/aptly-dev/aptly-dev.github.io/pull/10
23:31 forrest we'll be famous in no time!
23:31 iggy big $$$
23:31 iggy I've mentioned it a few times in the aptly channel
23:32 forrest Living off our big dollar donations
23:32 forrest Yeah it makes it way easier/better to configure
23:32 forrest manual configuration is kind of a pain in the ass.
23:32 forrest better than a normal apt repo though that's for sure.
23:33 iggy the formula doesn't do 1/10th of that all aptly can do (but everything I need it to ;)
23:33 murrdoc have they fixed it so u can put a deb and add it to a repo
23:33 Vynce can a file:recures exclude_pat include grain information?
23:34 Vynce (also, include_pat?)
23:34 forrest iggy: yeah it does all the basics which is important.
23:34 Vynce also, how do include and exclud epats play together?
23:34 forrest murrdoc: I thought that was slated to be fixed a few months ago, not running debian/ubuntu any longer so I don't know
23:36 clintberry joined #salt
23:38 iggy they just had a new release, so I'll probably back out the "install nightly" functionality
23:38 iggy I only added it for the --batch no-tty mode
23:40 wintamute joined #salt
23:41 mndo joined #salt
23:42 dude051 joined #salt
23:43 jaloren joined #salt
23:43 jaloren hi all .....this is randomly getting outputted when i do a state.sls https://gist.github.com/jaloren/b19bb22509470242a93b
23:44 jaloren i have salt-minion running in debug mode in the foreground and there's nothing there that would indicate a problem
23:44 jaloren i don't beleive there's anything invalid about the yaml
23:44 iggy try pasting it
23:44 jaloren iggy: was to me?
23:44 iggy yes
23:45 jaloren https://gist.github.com/jaloren/b94a1b087a77b809a839
23:46 jaloren iggy: i verified that python can load that file and convert the yml into a python object
23:46 iggy - module.function:
23:46 iggy top level is a dict, second level is a list
23:46 iggy no it's not
23:47 iggy I don't know what I was looking at that made me say that
23:47 iggy EMULTITASKING
23:47 jaloren lol okay
23:47 jaloren i was confused
23:47 murrdoc jaloren:  paste in state.show_sls of that state
23:48 iggy the require's shouldn't be necessary there (although I don't think that's what's causing your issue)
23:48 jaloren murrdoc: https://gist.github.com/jaloren/7d91518e0358bdb72e0a
23:49 jaloren I am executing this from a custom python script using the python client api...but i don't think that should matter?
23:50 iggy does it do the same if you just "salt 'minion' state.sls whatever.stuff" (i.e. take your script out of the picture)
23:50 jaloren iggy: excellent question....trying that right now...should now in a sec
23:51 iggy I mean it shouldn't matter, but... debugging 101
23:51 murrdoc - esg: reset_esg
23:51 jaloren right
23:51 murrdoc that shouldnt work
23:51 jaloren murrdoc: and yet apparently it works when i execute it from the command line without the script
23:52 murrdoc take out all your requires , try again (its 7pm, i am giving up)
23:52 iggy assuming there's a custom state called esg it should
23:52 murrdoc :)
23:52 murrdoc late
23:52 jaloren iggy: there is in fact ushc a custom state :)
23:52 otter768 joined #salt
23:53 iggy lol, I'm sandbagging at the office (room mate wants me to go grocery shopping... f that I don't eat shit at the house)
23:53 Vynce ping?  can file.recurse exclude_pat include info from grains?
23:54 forrest Vynce: Most people don't respond if we don't know
23:54 desposo joined #salt
23:54 baweaver joined #salt
23:54 jaloren iggy: you don't think it could be something like this do you https://github.com/saltstack/salt/issues/14992
23:54 forrest I don't know, otherwise I would have said something :P
23:55 Vynce yeah, just re-asking in case someone who knew jsut missed it. in a busy room, you never know.
23:55 opennode_ joined #salt
23:56 iggy jaloren: possibly?
23:56 jaloren iggy: i am firing multiple api calls in extremely quick succession. maybe i am hitting a race condition or lock issue
23:57 iggy should be easy to check
23:57 jaloren iggy: oh?
23:57 iggy shutdown salt-minion and make sure there aren't any other copies running
23:57 jaloren iggy: oh already did that...mutiple service restarts and did ps and kills etc
23:58 iggy put a sleep in-between?
23:58 jaloren iggy: yep sounds reasonable trying it now

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