Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-02-18

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

All times shown according to UTC.

Time Nick Message
00:00 fannet_ joined #salt
00:00 iggy after you re-generate the resolv.conf file, do you see that line? if so, where is it?
00:00 fannet_ yes - if I put it in head its above the comment if I put it in base its below
00:01 iggy is the uppermost nameserver line in resolv.conf?
00:01 iggy *is it
00:02 fannet_ yes - so it worked now. apparently L3 is blocking my dns requests :-(
00:02 fannet_ chagned 4.2.2.1 to 8.8.8.8
00:03 fannet__ joined #salt
00:04 fannet__ well that was annoying :-/
00:11 tombibb joined #salt
00:11 jameswarren joined #salt
00:12 bluenemo joined #salt
00:12 bluenemo joined #salt
00:13 otter768 joined #salt
00:15 younqcass joined #salt
00:19 iwishiwerearobot joined #salt
00:20 Xopher joined #salt
00:23 TheoSLC joined #salt
00:24 meylor joined #salt
00:25 baweaver joined #salt
00:25 raygunsix joined #salt
00:27 otter768 joined #salt
00:34 waddles left #salt
00:35 mfournier joined #salt
00:36 ghanima anyone on
00:37 ghanima question about syndir and output
00:38 murrdoc syndir ?
00:39 murrdoc not sure what that be
00:39 ghanima syndic_wait
00:39 ghanima sorry I mean syndic
00:39 ghanima I think I have it configured directly
00:39 ghanima but on my master of masters
00:39 ghanima I don't see the output
00:39 ghanima but looking at the logs the masters of masters gets the response in the logs
00:39 ghanima but the salt command doesn't return the output
00:41 n8n joined #salt
00:41 ghanima murrdoc: any thoughts
00:42 timoguin joined #salt
00:42 murrdoc uh timeouts ?
00:42 murrdoc like the salt command has a -t flag
00:42 murrdoc try salt -t 60 <command>
00:42 murrdoc see if u get output
00:44 ghanima so to be clear the master of masters would have no minions associated with it right?
00:44 murrdoc nope
00:44 murrdoc its a remote executioner
00:44 murrdoc so to speak
00:45 ghanima murrdoc: so specified the timeout gave it 60 seconds
00:45 ghanima and still nothing
00:46 ghanima So to be clear the master does execute the command in question and I see the result in the logs on the masterofmasters
00:46 ghanima but I don't see any output from the salt command on the master of masters
00:46 murrdoc one second
00:46 ghanima just doing a simple salt '*' test.ping
00:46 carlos_felix joined #salt
00:46 ramishra_ joined #salt
00:47 murrdoc and u arent seeing the salt output on it
00:47 MugginsM joined #salt
00:47 murrdoc do u have a returner setup in the salt mom servers's master config
00:47 ghanima no I don't
00:48 ghanima and in regards to the output
00:48 ghanima the only thing I have
00:48 ghanima so I have minion going to the master of masters
00:48 ghanima that is the only output I have
00:49 murrdoc reading docs
00:49 murrdoc ./me is intrigued
00:50 murrdoc and u are on the latest version of salt
00:50 MugginsM joined #salt
00:50 ghanima salt-2014.7.1-1.el6.noarch
00:50 bfoxwell joined #salt
00:50 ghanima I think that is the latest version
00:51 murrdoc yup
00:52 MugginsM joined #salt
00:54 ghanima according to the docs the masterofmaster didn't require any custom setup
00:54 ghanima I just configured the master to point to the masterofmaster
00:54 ghanima and that's it
00:55 murrdoc salt-key
00:55 murrdoc wgat does that return
00:56 ghanima so I have the salt-master configured as a minion
00:56 ghanima and the masterofmaster connected to it
00:56 ghanima and that is the only entry I see
00:56 MugginsM joined #salt
00:57 murrdoc restarted the salt master after making its config change '?
00:57 ghanima yes
00:57 ghanima I have
00:57 ghanima but for good measure let me try again
00:57 cberndt joined #salt
01:00 desposo joined #salt
01:02 aparsons joined #salt
01:07 ajw0100 joined #salt
01:10 MatthewsFace joined #salt
01:10 baweaver joined #salt
01:10 dude051 joined #salt
01:17 ghanima order_masters
01:17 ghanima is what I was missing
01:28 mohae joined #salt
01:29 clintberry joined #salt
01:44 CeBe2 joined #salt
01:47 ramishra_ joined #salt
01:50 Kelsar joined #salt
02:01 TTimo joined #salt
02:02 yomilk joined #salt
02:07 jtang joined #salt
02:08 markm joined #salt
02:08 iwishiwerearobot joined #salt
02:14 clintberry joined #salt
02:14 baweaver joined #salt
02:18 evle joined #salt
02:20 Alec joined #salt
02:22 Alec Hi there. I have a tiny test environment of 1 master & 2 minions. All looks well except that when I use 'cmd.run' to execut a (coincidentally) .run file, the job hangs. I can't find what I'm looking for in the docs, although I'm still looking. Any suggestions?
02:24 Furao joined #salt
02:27 favadi joined #salt
02:29 aparsons_ joined #salt
02:33 nitti joined #salt
02:34 aparsons joined #salt
02:41 flupke joined #salt
02:46 timoguin joined #salt
02:48 ilbot3 joined #salt
02:48 Topic for #salt is now Welcome to #salt | SaltConf 2015 is Mar 3-5! http://saltconf.com | 2014.7.1 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/
02:49 sijis_ joined #salt
02:50 mosen joined #salt
02:52 dude051 joined #salt
02:57 otter768 joined #salt
02:59 stylica joined #salt
03:05 alexhayes Is there a way of ensuring a state is never executed on a particular host or group of hosts?
03:05 alexhayes or rather, can't be executed?
03:05 rawzone joined #salt
03:15 TyrfingMjolnir joined #salt
03:18 bhosmer joined #salt
03:25 raygunsix joined #salt
03:34 mikaelhm joined #salt
03:36 desposo joined #salt
03:39 TyrfingMjolnir joined #salt
03:44 timoguin joined #salt
03:49 ramishra_ joined #salt
03:50 mapu joined #salt
03:51 TyrfingMjolnir joined #salt
03:57 iwishiwerearobot joined #salt
03:58 ramishra_ joined #salt
03:59 SheetiS joined #salt
04:06 TyrfingMjolnir joined #salt
04:07 JlRd joined #salt
04:09 tligda joined #salt
04:12 TyrfingMjolnir joined #salt
04:14 tligda joined #salt
04:25 Ryan_Lane joined #salt
04:27 forrest joined #salt
04:29 lynxman joined #salt
04:37 TyrfingMjolnir joined #salt
04:40 dude051 joined #salt
04:40 TyrfingMjolnir joined #salt
04:42 ramishra_ joined #salt
04:44 timoguin joined #salt
04:51 TyrfingMjolnir joined #salt
04:59 ramishra_ joined #salt
05:00 tligda joined #salt
05:00 TyrfingMjolnir joined #salt
05:08 tligda joined #salt
05:13 ramishra_ joined #salt
05:22 jtang1 joined #salt
05:22 TyrfingMjolnir joined #salt
05:22 tligda joined #salt
05:25 jalaziz joined #salt
05:31 felskrone joined #salt
05:34 ajw0100 joined #salt
05:38 holms joined #salt
05:40 bhosmer joined #salt
05:44 timoguin joined #salt
05:46 Ryan_Lane joined #salt
05:46 iwishiwerearobot joined #salt
05:50 javaserver joined #salt
05:53 aparsons joined #salt
05:55 TTimo joined #salt
05:59 aparsons joined #salt
06:01 ramteid joined #salt
06:04 Furao joined #salt
06:08 AirOnSkin joined #salt
06:08 CeBe joined #salt
06:08 AirOnSkin Good morning everyone. I need help figuring something out...
06:09 AirOnSkin I have a file for Rsyslog which reads local log files. For example: MySQL logs to /var/log/mysql.log ... rsyslog reads that, so I may do something with it.
06:10 AirOnSkin The same goes for HTTPD logs. I have 5 different websites on a host, Rsyslog shall read both access & error log for all 5 services
06:10 Furao for httpd logs, you can make them goes straight into rsyslog
06:11 AirOnSkin Furao: I know. The problem lies elsewhere
06:11 AirOnSkin I would like to be able to configure everything via Salt and it's Pillars
06:11 Furao but yeah mysql suck, i try to avoid it as much as possible, it’s a pain and a shameful db
06:11 AirOnSkin -.-
06:12 Furao so for rsyslog + mysql we implemented it that way:
06:12 AirOnSkin I haven't finished explaining my problem Furao
06:12 Furao ok go ahead
06:14 AirOnSkin I'd like to be able to configure this (for now imaginary) web server with all five websites via Salt & Pillars.
06:15 AirOnSkin So, for the rsyslog config that reads local logs, I'd like some logic that goes: for as many configured websites there are, do "input file xxx"
06:15 Furao ah ok you have a more theorical problem than rsyslog + httpd :P
06:15 AirOnSkin Yes
06:15 Furao because in both your example I have simple workaround :P
06:15 Furao https://gist.github.com/bclermont/fee301810dc0bb681f43 for mysql
06:15 Furao and apache can log to syslog and recent version or nginx too
06:15 Furao but yes I understand your problem
06:16 Furao we do the same to have rsyslog read input log file from upstart /var/log/$service.log
06:16 AirOnSkin So, if I define a Pillar: webservice1: intra
06:16 AirOnSkin For our intranet (just an example)
06:16 Furao and service can be defined in pillar, in some cases
06:17 AirOnSkin "webservice1" will serve as a variable for configuring the vhost, the name of the logfiles, etc...
06:17 AirOnSkin But - an now comes the actual question - how does Pillar know how many webservices there are?
06:18 Furao you list them in your pillars first
06:18 AirOnSkin How?
06:18 Furao don’t mix unmanaged and managed stuff, let salt do everything, it need to know as much as possible. so file.managed your http config files
06:19 Furao pillars can be any data you want
06:19 zer0def joined #salt
06:20 Furao here is an (old) example of virtualhost in rabbitmq: https://github.com/bclermont/states/blob/master/states/rabbitmq/init.sls#L116
06:20 Furao i use pillars to define all my vhosts and their config
06:20 Furao today that formula is a more inteligent https://doc.robotinfra.com/rabbitmq/doc/pillar.html
06:21 AirOnSkin I'm really talking about this file: http://hastebin.com/fiqajiyubo.vhdl
06:22 AirOnSkin How do I manage to only have to define the include once, but have Salt loop through all webservices for this host?
06:22 Furao http://hastebin.com/jedexejuyo.vhdl
06:23 calvinh joined #salt
06:23 AirOnSkin And how would a Pillar matching that look?
06:24 Furao webserver:
06:25 Furao - bleh1
06:25 Furao - bleh2
06:26 calvinh joined #salt
06:26 MTecknology Furao: I made a thing!
06:26 Furao MTecknology: ?
06:27 Furao what??
06:27 MTecknology I made a mysql state such that I can use a pillar like this - http://dpaste.com/0B9RXWV
06:28 Furao ah, that can be useful
06:28 Furao I tend to define permission into formula specifics and not in a global pillar key
06:28 MTecknology database: [user1, user2]    but I can optionally assign a specific grants   database: [user1, user2, user3: {grants: select}]
06:28 Furao myapp:grants: *here goes permissions*
06:29 MTecknology I considered that, but it would take more work to add new databases/users
06:29 calvinh_ joined #salt
06:29 MTecknology there's also a mariadb_users dictionary with  user: 'pass'
06:30 MTecknology It was a pain in the ass to make each user optionally be a dictionary, though.
06:30 CryptoMer joined #salt
06:30 AirOnSkin Furao: That really doesn't solve the problem (I think). Say my services really are named bleh1 & bleh2, would {{ webserver }} match them both? So I would have 4 lines in my config (2 for each service)?
06:31 jhauser joined #salt
06:31 Furao MTecknology: what if you include 10 different .sls to build the global pillar dict for a single host? you can’t “modularize” pillar values across multiple based on different condition
06:32 Furao but well if it works for you it’s great, in my case, our formula are used by many different clients that have different requiements. i need to keep them flexible
06:33 MTecknology Mine are all built specifically for me. They only need to be as flexible as I want them to be. If I had to build something that would be used by many unknowns, I'd consider it
06:33 MTecknology The way things are written now, I could VERY easily add a host option to the grants list
06:33 CeBe joined #salt
06:34 MTecknology database: [user1, user2, user3: {grants: select, host: 1.2.3.4}]
06:34 MTecknology I won't, though... It's impossible for my mariadb cluster to get a connection from anywhere other than the load balancer.
06:34 Furao AirOnSkin: you can achieve pretty much everything with pillars and salt mine (auto-configuration). but I don’t understand your problem about webserver match them both, it’s a loop so once it’s bleh1 and the other loop run it’s bleh2
06:36 AirOnSkin Furao: Nevermind. Thanks anyway for the help
06:36 Furao AirOnSkin: if you look in github.com/bclermont/states you can see some real world example that might fit your use case
06:37 Furao for rsyslog and upstart I have this: https://gist.github.com/bclermont/d40874fccfef9560143a
06:37 Furao it’s macro called in different formula
06:37 MTecknology Furao: you gonna be at saltconf?
06:38 Furao MTecknology: nope im in vietnam, it’s too far. and saltlake city is too winter’y. and anyway I had been hit by a motorbike yesterday and can’t even walk :)
06:38 Furao maybe in 2016 i’ll be in south or central america and closer and go there
06:39 MTecknology dang
06:39 Furao they should do it in socal in july
06:40 Furao i didn’t left canada for ever to go to utah in march :)
06:40 MTecknology 1) I'll be sad to not see you 2) It's past the cold time of year. Heck, it's 35 F there. That's shorts weather! 3) Dang, dude. Feel better quickly!
06:40 * MTecknology will be giving a talk there this year.
06:41 Furao oh really? which subject?
06:41 MTecknology I flat out told work that if they don't send me or they pull it from me last minute this year, I'll quit on the spot.
06:41 MTecknology "From Chaos to Order" .. discussing the pile of garbage I started with and the environment I've been able to build
06:42 Furao I recently publicly launched a devop/salt consulting shop that we had been working on for > 3 years. and I wished to talk about it there and maybe sponsor beers. but not this year :(
06:42 MTecknology I went from pretty much not being able to manage 200 servers to managing 500 without much troubles and the more salt, the better things get.
06:42 Furao hehe that is great
06:43 grosjean hello all
06:43 grosjean anyone got idea why sometimes cmd_iter return me an empty generator ?
06:43 MTecknology It's been one heck of an insane ride and I'm more than excited to discuss it.
06:43 Furao ask for spoilers, what was your biggest challenge? mine is unstable new release :(
06:43 MTecknology I need to practice it a *LOT*, though. I *HATE* speaking
06:43 Furao grosjean: empty like return None? or don’t iter?
06:43 grosjean i check Furao
06:44 MTecknology my biggest challenge is the number of new things I need to keep building that keep me from being able to go back and further saltify things.
06:44 stoogenmeyer_ joined #salt
06:44 MTecknology I have ~200 servers that are 100% managed/deployed using salt
06:44 timoguin joined #salt
06:44 Furao MTecknology: ahaha i understand that feeling, I did a speak to Malaysia Python Conference last year about using python (mostly salt) to manage infra, and it turned out bad, I should had just not prepared anything and speak from my mind/heart without any slides. and the challenge increased as english isn’t my native language
06:44 MTecknology as in - if I find out anyone ever logged into one of these boxes and did ANYTHING other than read-only, I'd murder them, kill them, and then strangle them.
06:45 grosjean Furao: empty like empty :)
06:45 grosjean look:
06:45 grosjean {'ret': ''}} ret <generator object cmd_iter at 0x3265640>
06:45 jtang joined #salt
06:46 MTecknology Furao: English isn't my native language either. IRC is.
06:46 Furao grosjean: please show the sls that return that way it’s… weird
06:46 san4 joined #salt
06:46 san4 joined #salt
06:46 grosjean Furao: i hit this problem when using salt python api
06:46 grosjean saltLocal.cmd_iter(.....)
06:46 Furao MTecknology: than sit in the room and talk to an other IRC client plugged to projector :P
06:47 Furao you probably need to do:
06:47 Furao for item in saltLocal.cmd_iter(…):
06:47 Furao process_item(item)
06:48 grosjean this is what i'm doing
06:48 Furao ok so cmd_iter return other generator
06:48 Furao funny
06:49 otter768 joined #salt
06:54 Furao i wonder if it’s hard to hack orchestrate runner to make minion wait for other state on other minions to finish
06:54 linjan joined #salt
06:55 Furao that will break salt-ssh
06:56 colttt joined #salt
06:59 Ilja joined #salt
07:00 Ilja joined #salt
07:01 dkrae joined #salt
07:02 Ilja joined #salt
07:02 toanju joined #salt
07:02 froztbyte MTecknology: I'm quite curious about that talk; will you be putting slides or something online for those who can't attend?
07:05 ghanima joined #salt
07:05 MTecknology froztbyte: I think all talks end up online
07:06 MTecknology bed time!
07:08 san4 left #salt
07:10 hojgaard joined #salt
07:11 yomilk joined #salt
07:12 san4 joined #salt
07:28 flyboy joined #salt
07:32 lb1a joined #salt
07:34 iwishiwerearobot joined #salt
07:37 Auroch joined #salt
07:39 cberndt joined #salt
07:39 jalaziz joined #salt
07:46 iwishiwerearobot joined #salt
07:46 kawa2014 joined #salt
07:46 che-arne joined #salt
07:46 wnkz joined #salt
07:47 slafs joined #salt
07:48 slafs left #salt
07:49 Terminus_ hello. i'm trying to get a minion to work but it crashes with http://paste.ofcode.org/tm7ucAphfsUiEXfkaHcj9G. any way i can fix that?
07:50 Terminus_ durrr... i should have googled that first.
07:51 trikke joined #salt
07:53 KermitTheFragger joined #salt
07:54 krelo joined #salt
07:55 lb1a :)
08:02 hemebond joined #salt
08:19 AndreasLutro joined #salt
08:20 martoss joined #salt
08:21 teogop joined #salt
08:21 cberndt joined #salt
08:23 iwishiwerearobot joined #salt
08:24 jhauser joined #salt
08:26 nullptr` joined #salt
08:30 Furao joined #salt
08:31 intellix joined #salt
08:35 eseyman joined #salt
08:37 overyander joined #salt
08:37 roolo joined #salt
08:40 crazysim joined #salt
08:41 Linuturk joined #salt
08:41 jY joined #salt
08:42 wolog joined #salt
08:42 wm-bot433 joined #salt
08:43 kaictl joined #salt
08:46 JlRd joined #salt
08:48 wm-bot433 joined #salt
08:49 monkey66 joined #salt
08:50 otter768 joined #salt
08:51 Ilja joined #salt
08:52 markm joined #salt
08:57 TTimo joined #salt
08:58 clintberry joined #salt
09:03 bluenemo joined #salt
09:03 bluenemo joined #salt
09:07 nullptr`` joined #salt
09:08 yomilk joined #salt
09:08 tombibb joined #salt
09:09 monkey66 joined #salt
09:18 tombibb_ joined #salt
09:21 jrluis joined #salt
09:22 alanpearce joined #salt
09:27 tombibb joined #salt
09:31 Grokzen joined #salt
09:32 sieve joined #salt
09:35 paulm- joined #salt
09:38 paulm-- joined #salt
09:40 jiperdepip joined #salt
09:52 sieve joined #salt
09:58 badon joined #salt
09:58 TTimo joined #salt
10:03 sieve joined #salt
10:05 yawniek_ somehow on my local vagrant salt cluster often i loose minions or commands do not go trough, e.g. test.ping succeeds only every so often. is this something known?
10:05 hemebond yawniek_: I see that too.
10:05 hemebond Try using --verbose (I think) and it might show that it actually failed.
10:05 hemebond Or perhaps it took too long to respond.
10:06 yawniek_ the minions seem not even to be connected
10:10 mdupont joined #salt
10:11 monkey66 joined #salt
10:11 amcorreia_ joined #salt
10:12 jtang joined #salt
10:15 tombibb_ joined #salt
10:18 Furao joined #salt
10:25 Xevian joined #salt
10:31 hemebond ah
10:31 hemebond That'd do it.
10:31 hemebond I have seen my Windows minions struggle to connect if the master is restarted.
10:32 bhosmer joined #salt
10:34 giantlock joined #salt
10:35 tombibb joined #salt
10:36 SpX joined #salt
10:37 jtang1 joined #salt
10:37 michelangelo joined #salt
10:40 Ilja1 joined #salt
10:48 eliasp hemebond: what version? I've seen this with 2014.1.x releases, but I think never again with 2014.7.x
10:50 eliasp what does "return ret['retcode'] == 0" do Python-wise? set ret['retcode'] to 0 and then return ? or is there something special about this?
10:51 otter768 joined #salt
10:51 seshan joined #salt
10:52 seshan I have two machines, with mode 4 teaming/bonding enabled, and both are able to ping each other
10:52 seshan but when I accept the key from the 2nd machine and try a test.ping, I get htis errir "Attempt to authenticate with the salt master failed"
10:53 seshan has anybody faced this
10:56 calvinh joined #salt
10:56 bhosmer_ joined #salt
11:06 OptimusPrime joined #salt
11:09 seshan babilen:  you there ?
11:09 babilen Who disturbs my slumber?
11:11 seshan Its me :D
11:11 fredvd joined #salt
11:12 seshan I was stuck with an issue
11:12 seshan for quite sometime now
11:13 kbyrne joined #salt
11:15 seshan Issue told above ., any idea ?
11:16 babilen That sounds tricky .. we had something similar and it turned out to be due to different MTR settings on links in each direction ..
11:16 seshan here the mtu settings are both 9000
11:16 seshan the point is if we remove the bonding and connect with normal eth0 alone, it all works fine
11:17 seshan only when we bond, i get this issue
11:17 yomilk joined #salt
11:18 babilen Sorry, not sure and no time to diagnose that right now
11:18 seshan babilen: ok :)
11:18 seshan will research more :)
11:24 N-Mi joined #salt
11:24 N-Mi joined #salt
11:25 bhosmer joined #salt
11:32 danemacmillan joined #salt
11:33 nitti joined #salt
11:33 Furao joined #salt
11:36 johtso joined #salt
11:38 agend joined #salt
11:42 linjan joined #salt
11:47 calvinh joined #salt
11:48 Ilja joined #salt
11:48 seshan left #salt
11:48 Ilja joined #salt
11:50 tombibb joined #salt
11:52 Nazca joined #salt
11:52 tombibb_ joined #salt
11:53 yomilk joined #salt
11:53 sieve joined #salt
11:54 holms State 'flle.replace' found in SLS 'common.motd' is unavailable
11:54 holms why
11:55 Furao file.replace
11:55 Furao not flle
11:56 markm joined #salt
11:59 holms :D
12:00 TTimo joined #salt
12:06 rudi_s Hi. Is it possible to run python code before/after a salt run? I'd like to run code like systemd-inhibit to prevent shutdowns while e.g. package installs take place. Thanks.
12:08 rudi_s *python code on the minion
12:08 tombibb joined #salt
12:14 rattng rudi_s: I think you can use Reactors for that purpose
12:16 roolo joined #salt
12:19 jacksontj joined #salt
12:21 jav_ec2 joined #salt
12:22 evle1 joined #salt
12:22 calvinh_ joined #salt
12:28 AirOnSkin I need some help figuring something out...
12:28 AirOnSkin I work with multiple environments in Salt: base, dev, test, prod
12:29 AirOnSkin base is for states that apply to all minions (standards)
12:29 AirOnSkin the other three have redundant information (which is by design... for our case)
12:30 hobakill joined #salt
12:30 AirOnSkin Even though a host might get a role "webserver" the open ports needed by the web application might not be the same for all hosts
12:31 tombibb_ joined #salt
12:31 AirOnSkin I'd like to keep all these "differences" or configuration parameters (environment, roles, open ports, crontab entries) in Pillar
12:32 AirOnSkin Now to the real question: Is it possible to have a separate environment where theses host configuration Pillars lie, that is accessible from the other environments? (especially dev, test & prod)
12:34 kermit joined #salt
12:36 rudi_s rattng: I haven't used reactors yet, will the reactors (or the code I run from them) be run on the master or minion?
12:38 Ilja joined #salt
12:39 paulm- joined #salt
12:39 linjan joined #salt
12:43 sypher_ joined #salt
12:43 holms just a stupid question
12:43 holms how do you split different environment minion in here?
12:44 holms in ansible you can have different playbooks, and you same roles (formulas)
12:44 holms use*
12:44 Ilja joined #salt
12:44 holms also as playbooks would include different config (pillar)
12:44 holms how can I achieve this in saltstack?
12:46 alanpearce joined #salt
12:51 sypher__ joined #salt
12:51 otter768 joined #salt
12:53 yomilk joined #salt
12:53 TTimo joined #salt
13:00 rattng rudi_s: you can decide in your reactor. Both are ok. It will be triggered on the master but what actions it takes is up to you.
13:01 rudi_s rattng: Ah, thank you.
13:04 aphorise joined #salt
13:07 tkharju joined #salt
13:09 yomilk joined #salt
13:12 aphorise joined #salt
13:12 bhosmer joined #salt
13:14 tombibb joined #salt
13:19 linjan joined #salt
13:21 scottpgallagher joined #salt
13:21 scottpgallagher i am running my salt master in a docker container, but how do i accept the minion keys then? a $ docker exec <container_id> salt-key -A won't let me input a Y
13:23 rattng scottpgallagher: I think you can pass -y or --yes  as well
13:23 Puckel_ add -y cf man (Answer 'Yes' to all questions presented, defaults to False)
13:23 rattng " -y, --yes             Answer Yes to all questions presented, defaults to False"
13:23 scottpgallagher Thanks!!! much appreciate #saltrocks
13:25 sieve joined #salt
13:34 babilen holms: I don't quite understand your question. Could you rephrase it?
13:34 mpanetta joined #salt
13:35 babilen holms: What is the meaning of "splitting an environment minion" ?
13:36 alanpearce joined #salt
13:37 tomh- joined #salt
13:38 ksk joined #salt
13:39 ksk hey guys. i have written an sls to store a config from an template, using highstates. that works great so far. now Id like to issue some bash-command, after the config was written - is there a way? thanks!
13:42 Puckel_ ksk: maybe w/ cmd.run ? http://docs.saltstack.com/en/latest/ref/states/all/salt.states.cmd.html#module-salt.states.cmd
13:42 SneakyPhil ksk: you can use something similar to salt '*' cmd.run "your command goes here"
13:42 ksk how would i combine that with file.managed in a state?
13:44 SneakyPhil ksk: this may be of assistance http://bencane.com/2013/09/23/remote-command-execution-with-saltstack/
13:44 ksk thanks
13:44 kawa2014 joined #salt
13:49 ramishra_ joined #salt
13:51 murrdoc joined #salt
13:51 _prime_ joined #salt
13:56 jdowning joined #salt
13:57 dyasny joined #salt
13:57 paulm- joined #salt
13:57 CeBe joined #salt
13:57 TTimo joined #salt
13:59 holms babilen:
13:59 babilen holms
13:59 racooper joined #salt
13:59 holms maybe you familiar with any automation system except saltstack?
14:00 paulm-- joined #salt
14:00 jeremyr joined #salt
14:00 babilen Assume that I am and go on
14:00 holms i'll compare to ansible, as it most simplier way of doing this split
14:00 FRANK_T joined #salt
14:00 holms i have vagrant local dev vm which we'll call dev
14:00 babilen Although I would prefer if you could tell me what you want to achieve without referring to other implementations
14:00 holms and we have prod env on aws/do
14:01 babilen sure
14:01 holms in ansible we use playbooks, a file which gathers roles (formulas in saltstack) in a way you want and with params you want
14:01 holms same goes for configuration inclusion specifically for those environments
14:01 holms prod playbook will include default-config.yml which will be overrriden by prod-config.yml
14:01 cpowell joined #salt
14:02 holms (where needed, var names will be overriden)
14:02 holms so now, when i want to provision AWS instance for example
14:02 holms i specify prod.yml playbook, and for local dev machine dev.yml playbook
14:02 holms how can you specify for saltstack, which scenario i want to use?
14:02 babilen By targeting different boxes in different environments
14:03 bhosmer_ joined #salt
14:03 tligda joined #salt
14:03 babilen http://docs.saltstack.com/en/latest/ref/states/top.html#environments
14:03 subsignal joined #salt
14:03 holms what can I pass to provisioenr, i'd use different set of states
14:03 holms ok
14:03 tligda1 joined #salt
14:03 bhosmer_ joined #salt
14:04 holms now probably this will be vagrant specific question
14:04 holms how can I specify which env to use from top file?
14:04 holms or atleast what param salt provisioenr has for that
14:04 holms maybe i'll find that in vagrant config
14:04 babilen base will be the, well, common base for all and then you have other environments. You would have your data in pillars and rely on http://docs.saltstack.com/en/latest/ref/configuration/master.html#pillar-source-merging-strategy
14:04 holms manual*
14:05 holms is this specified in minion?
14:06 babilen You target minions based on some attribute that allows you to differentiate between "dev" and "prod" minions. It is entirely up to you what you use in that case. One, very basic, way would be to include a minion config in your vagrant setup that sets a grain to, say, "dev: True" and you can then target based on that. There might, however, be better ways to differentiate, but that really depends on your setup.
14:06 babilen "this" ?
14:06 fivmo joined #salt
14:07 holms babilen: http://imgur.com/HntXzBz can "this" be specified in minion?
14:08 holms so that "dev" branch would be executed (together with base ofcourse)
14:08 holms vagrant has ability to specify minion path
14:08 sxar joined #salt
14:09 babilen holms: That is a configuration you set on your master. You configure which minions belong to that environment in the top file
14:10 hasues joined #salt
14:10 hasues left #salt
14:12 ALLmightySPIFF joined #salt
14:12 babilen Note that you should use dedicated repositories for your top files if you use GitFS and environments
14:13 holms this is overcomplicated
14:13 holms in compare with others.
14:13 holms i have masterless setup in here
14:13 babilen Then use different top files
14:13 holms and it will be so, for at least while we have <3 nodes
14:14 holms so top-dev.sls? and top-prod.sls?
14:14 holms how can i specify in minion which one to use?
14:14 babilen I also don't do masterless so I am not the ideal person for that discussion. You can, however, use different top files ...
14:14 holms i'm ok with one top file :D
14:14 holms i just don't undestand how to *specify* which env to provision
14:14 ALLmightySPIFF joined #salt
14:14 jamesog joined #salt
14:14 murrdoc master config file
14:14 murrdoc sorry, minion config file
14:16 holms murrdoc: thanks
14:16 holms that what i thought it should be
14:16 holms http://i.imgur.com/BPjKFmY.png
14:16 holms just a question which pameter gonna say what to use in that top file
14:16 babilen http://docs.saltstack.com/en/latest/ref/configuration/minion.html#environment + can't find it
14:16 babilen one secondf
14:16 murrdoc top.sls is used with state.highstate
14:17 JDiPierro joined #salt
14:17 murrdoc so the logic of role to states is done using compound matching
14:17 murrdoc http://docs.saltstack.com/en/latest/topics/targeting/compound.html + http://docs.saltstack.com/en/latest/ref/states/top.html
14:17 murrdoc OR
14:18 murrdoc you can do nodegroups
14:18 murrdoc http://docs.saltstack.com/en/latest/topics/targeting/nodegroups.html
14:18 holms murrdoc: so to run dev (http://imgur.com/HntXzBz) from top.sls, i need to specify environment: dev in minion config file?
14:18 murrdoc yup
14:18 dooshtuRabbit joined #salt
14:18 paulm-- archive.extracted documentation claims that "Main operators like -x, --extract, --get, -c and -f/--file should not be used" in tar_options, but by not specifying those I get this error: tar: You must specify one of the '-Acdtrux', '--delete' or '--test-label' options
14:18 holms oh thanks, that's what i want :)
14:20 primechuck joined #salt
14:23 murrdoc joined #salt
14:24 holms nodegroups similar to ansible hosts file
14:24 holms cool
14:25 Nazca joined #salt
14:26 N-Mi is it possible in a state definition for a minion to generate a file directly on the master ?
14:27 N-Mi generating on the minion then pushing it on minionfs is not a good option as if I understand, the file will be readable by all minions
14:27 rawkode joined #salt
14:31 murrdoc u need data from the minion ?
14:31 ninkotech joined #salt
14:33 brayn joined #salt
14:33 perfectsine joined #salt
14:35 mindscratch joined #salt
14:35 N-Mi murrdoc: what I want to do is quite similar to this link, but I'd like to do it differently : https://hveem.no/icinga-configuration-generation-using-salt
14:35 schlueter joined #salt
14:36 yomilk joined #salt
14:36 berserk joined #salt
14:36 N-Mi first idea waq to generate monitoring config file for this minion, and then push it on minionfs, then the supervision minion would be able to get it
14:36 N-Mi but everything pushed on minionfs is readable by all minions
14:37 mindscratch I have an SLS file that runs a command. One of the arguments to the command should be the hostname of a server, is it possible to use jinja to somehow get the name of the server (I can figure out the hostname by using a grain) and instead of hard coding it?
14:38 dude051 joined #salt
14:38 murrdoc {{ salt['grains.get']('id') }}
14:38 murrdoc N-Mi:  u are trying to generate the ichinga config file and push it to the master ?
14:39 N-Mi murrdoc: yes (shinken in fact, but it doesn't matter)
14:39 murrdoc why do you need it on the master ?
14:39 N-Mi and these files can contain password or secret
14:40 N-Mi to send them then on the shinen minion
14:40 clintberry joined #salt
14:42 iggy did you look at how the formula does it?
14:42 N-Mi something that a few people have asked on forums and mailing-lists according to my research, but no easy/clean solution was provided as I can see
14:43 ramishra_ joined #salt
14:43 N-Mi in puppet it's done with "exported ressources"
14:43 aphorise joined #salt
14:44 murrdoc you ll have to use a custom module for this
14:44 murrdoc store the info in some sort of shared data source
14:44 iggy well, it sounds like puppet has a leg up on us in this one use case... we might as well all pack it up
14:44 tombibb joined #salt
14:44 murrdoc puppet and chef both have a 'db'
14:44 murrdoc salt doesnt
14:44 murrdoc it is known
14:44 iggy pillar wheel and generate off that
14:44 iggy boom!
14:45 murrdoc ?
14:45 murrdoc http://docs.saltstack.com/en/latest/ref/wheel/all/salt.wheel.pillar_roots.html#module-salt.wheel.pillar_roots
14:45 murrdoc doubt it
14:46 murrdoc you could probably do it with grains or mine
14:47 murrdoc but really saltstack needs a db
14:47 JDiPierro joined #salt
14:47 murrdoc i think they are working ion it for the 'paid version'
14:48 mindscratch on a minion, when I run: salt-call state.highstate I get back "No matching sls found for 'mesos-slave' in env 'base'"
14:48 mindscratch however, on the master I have /srv/salt/mesos/slave.sls
14:49 jalbretsen joined #salt
14:49 mindscratch on master I have /srv/salt/top.sls with:  http://hastebin.com/vepuhivufa.sm
14:50 mindscratch doH! I was referencing "mesos-slave" instead of "mesos.slave" (since slave.sls is the file in the mesos directory)
14:52 mindscratch murrdoc: can I get a grain for another minion?
14:52 murrdoc using pillars yes
14:52 otter768 joined #salt
14:53 mindscratch murrdoc: so i have to statically store data in pillar, in my case to say "host3" is the server
14:54 murrdoc uh
14:54 murrdoc http://docs.saltstack.com/en/latest/topics/mine/
14:54 mindscratch so i want to run a command on a minion, that command needs to know the hostname of another server (based on a grain)
14:55 mindscratch so the server in question has a grain "roles: webserver". the command to run on the minion needs to get the name of the host whose "roles" grain matches "webserver"
14:55 timoguin joined #salt
14:56 rawkode joined #salt
14:57 Ilja joined #salt
14:59 Deevolution mindscratch: You can use an execution module from the master that first queries grains data on your minions and then runs the command using the resulting data.
14:59 N-Mi iggy: I didn't know wheel.pillar_root. Il I understand correctly, it can write a pillar file, that I could then use in the shinken minion state to generate monitoring config for the first minion
14:59 JDiPierro joined #salt
15:03 kyriakos joined #salt
15:05 kyriakos hello. is there a way to have the tag/tag values of an ec2 instance represented as grains ?
15:06 raygunsix joined #salt
15:06 dude051 joined #salt
15:07 _JZ_ joined #salt
15:07 mindscratch murrdoc: I modified an SLS file for the webserver to add: mine_functions: \n network.get_hostname: []
15:07 andrew_v joined #salt
15:08 mindscratch murrdoc: on the webserver minion, I tried running salt-call state.highstate  and I get an error 'State 'network.get_hostname' found in SLS 'webserver' is unavailable'
15:08 murrdoc sorry man,work time, you ll have to figure it out from the docs
15:08 murrdoc :|
15:09 kyriakos left #salt
15:09 kaptk2 joined #salt
15:09 mindscratch np
15:11 TheoSLC joined #salt
15:12 nitti joined #salt
15:12 scristian joined #salt
15:16 scristian joined #salt
15:16 stylica joined #salt
15:18 stylica joined #salt
15:19 mapu joined #salt
15:20 claytonk joined #salt
15:26 scristian joined #salt
15:27 che-arne joined #salt
15:28 linjan joined #salt
15:30 TheoSLC joined #salt
15:31 Brew joined #salt
15:31 jrluis joined #salt
15:34 lb1a joined #salt
15:37 mindscratch murrdoc: got it, had to end mine_functions to /etc/salt/minion
15:38 schlueter joined #salt
15:39 murrdoc nice
15:39 chris-m joined #salt
15:41 chris-m good day all!
15:41 chris-m has anyone specific experience with syndics and having mom sync 'custom modules-python scripts' to it
15:42 edrocks joined #salt
15:43 mindscratch murrdoc: is it possible to have 'cmd.run' has one of the salt mine functions?
15:45 murrdoc possibly, i think anything that runs under module.run should work in mine_functions
15:46 trevorj joined #salt
15:47 CeBe1 joined #salt
15:49 ajw0100 joined #salt
15:55 dude051 joined #salt
15:56 mindscratch I tried adding to mine_functions, cmd.run: \n name: "cat /some/file"
15:56 nullptr joined #salt
15:56 ghanima joined #salt
15:56 mindscratch so far not getting what i hoepd
15:56 mindscratch hoped
15:56 pdayton joined #salt
16:00 TheoSLC joined #salt
16:04 * akafred tries a quick poll: Windows clients as salt-minions a) a joy! b) a pain! c) mostly ok, but beware that ... d) haven't tried
16:04 timoguin joined #salt
16:05 Twiglet can you use salt mine in pillars?
16:05 Twiglet e) bollocks to windows ;)
16:08 akafred :-)
16:08 akafred I considered e) jeez ... get a life!
16:08 ghanima anyone familiar with syndic
16:09 ghanima I just implemented it and I am seeing some performance issues that I am not sure if its expected
16:10 edrocks if I'm using cmd.script args do I need to add "--" to my args or does salt do it for me?
16:10 rojem joined #salt
16:11 Ymage joined #salt
16:12 mindscratch i am using the salt mine to report the hostname of a webserver so i can use it in an SLS that runs on another minion
16:12 InAnimaTe joined #salt
16:12 theologian joined #salt
16:13 clintberry joined #salt
16:13 mindscratch I also need to get the port the webserver is listening on, which is as simple as doing "cat /etc/webserver/http_port" ...but how i can report that using salt mine? If I can't is there another way?
16:14 Deevolution akafred: c and e
16:17 N-Mi hum... I don't see how to use this : http://docs.saltstack.com/en/latest/ref/states/all/salt.states.saltmod.html#salt.states.saltmod.wheel
16:17 N-Mi here is what I've done : http://dpaste.com/2VXB8ZF
16:17 N-Mi maybe I misunderstood something ?
16:18 drawsmcgraw mindscratch: I've used Salt Mine to get the hostname of a web server, but looking back at my work, I hard-coded the port
16:18 drawsmcgraw If you go here: https://www.digitalocean.com/community/tutorials/how-to-use-salt-cloud-map-files-to-deploy-app-servers-and-an-nginx-reverse-proxy
16:18 drawsmcgraw and go to "Step 5: Define the Reverse Proxy"
16:19 drawsmcgraw You'll find the templated Nginx config where I grab the web servers's IP
16:19 N-Mi oups, conf is here : http://dpaste.com/04JP69R
16:19 mindscratch drawsmcgraw: yea I've got the salt mine working to get the hostname
16:20 mindscratch drawsmcgraw: was hoping "cmd.run" in mine_functions might work, but doesn't seem too
16:20 drawsmcgraw mindscratch: bummer :/
16:20 felskrone joined #salt
16:20 gattie joined #salt
16:21 aphorise joined #salt
16:21 murrdoc have u looked at the aliasing stuff ?
16:21 murrdoc in salt mine doc
16:22 debian112 joined #salt
16:23 mindscratch no
16:23 vexati0n in 2014.7.x , more than half the minions stop responding until i do test.ping like 4 times
16:23 vexati0n is there any "don't fall asleep" option?
16:24 yomilk joined #salt
16:25 schristensen joined #salt
16:25 mindscratch murrdoc: aliasing doesn't seem to help http://www.hastebin.com/amusinoyar.vhdl
16:26 murrdoc oh well
16:26 signull joined #salt
16:29 conan_the_destro joined #salt
16:29 mohae joined #salt
16:30 ccarney_ROCC joined #salt
16:31 robothands any ideas on using salt-cloud with vcenter? pretty sure the vsphere provider will not work the same....
16:32 ccarney_ROCC left #salt
16:32 JlRd joined #salt
16:34 rbjorklin joined #salt
16:39 thedodd joined #salt
16:41 tligda joined #salt
16:45 ghanima murrdoc: thanks for your help yesterday but if you haven't seen the archive I figured out my problem from last night
16:46 murrdoc oh nice
16:46 murrdoc what was it
16:47 StDiluted joined #salt
16:47 edrocks can you have multiple profiles in your cloud.profiles?
16:49 SheetiS has anyone used the salt pagerduty integration with their api keys in the pillar (I do this because I store it with the gpg renderer)?  It seems the way options are grabbed that this won't work since it does a simple opts.get (where opts = __opts__) instead of calling the config.option function which will also check the pillar.  This is in salt 2014.7.1.  The config is in opts['pillar'], but the way it is coded has no way to determine
16:49 sxar joined #salt
16:49 tligda1 joined #salt
16:53 otter768 joined #salt
16:54 dyasny joined #salt
16:58 SheetiS I'm just not seeing it as even possible to configure the options via the pillar at all with how it is currently written.
16:59 murrdoc http://docs.saltstack.com/en/latest/ref/configuration/master.html#pillar-source-merging-strategy
17:02 san4 joined #salt
17:02 san4 joined #salt
17:02 Ozack-work joined #salt
17:03 san4 joined #salt
17:03 san4 joined #salt
17:03 san4 joined #salt
17:03 san4 joined #salt
17:05 SheetiS murrdoc: Was that for me?  If so, I don't think it applies. The data appears in the pillar how I expect it, so I don't need to change how pillar data is merged.  Pagerduty cannot have its configuration in the pillar with how it is currently written from what I can see even though the documentation says otherwise.
17:05 murrdoc i was hoping it would help u
17:05 murrdoc if it didnt :|
17:10 SheetiS murrdoc: most things I see in salt use the __salt__['config.option'] to get configuration items that can exist eitehr in the pillar or the configs.  An example is the mine_functions:  https://github.com/saltstack/salt/blob/2014.7/salt/modules/mine.py#L59 but in the case of the pagerduty module, it just uses __opts__: https://github.com/saltstack/salt/blob/2014.7/salt/modules/pagerduty.py#L106.  Now I know that it tries to be dynamic an
17:10 rgarcia_ joined #salt
17:11 akafred Deevolution: Anything particular I should beware of about salt on Windo$?
17:12 JDiPierro joined #salt
17:12 SheetiS If I were fixing this myself, I'd probably do something like __salt__['config.option'](profile) at that point and pass the dict into the util, then make the util not try and parse __opts__ itself.  The change would not be insignificant and while I could make it match feature-wise with what exists, it might be different enough to get resistance as a PR.
17:13 SheetiS Alternatively, if anyone has gotten pagerduty integration to work with the pagerduty profile configured in the pillar, I would love to talk to you :)
17:13 murrdoc worth a try
17:13 murrdoc the pr that is
17:18 gattie joined #salt
17:18 Deevolution akafred: I've found a number of issues regarding the Salt scheduler (patched one, it's in 2014.7.1; starting to work on some others).  Also just kind of general Windows issues (file permissions, running as the correct user due to lack of sudo, etc...)
17:19 spookah joined #salt
17:19 kiorky joined #salt
17:19 ghanima murrdoc: order_masters: True
17:20 murrdoc nice
17:21 ghanima murrdoc: now I am seeing a performance issue and I am not sure if this is expected overhead or something can be changed
17:21 ghanima I have 600 hosts
17:21 ghanima and when I execute salt commands on the lower master they respond reasonably well somewhere between 10 to 20 seconds
17:21 ghanima if I do it from the mastermaster its like 45 seconds
17:22 ghanima I purposely specified the syndic_wait variable to 35 seconds to make sure I get a response back from all nodes within a certain amount of time
17:22 ghanima but it seems that that there is a level of overhead with syndic
17:22 ghanima just curious your opinion
17:23 mikaelhm joined #salt
17:25 yomilk joined #salt
17:25 akafred Deevolution: Ok. MMMV, then ... but it's mature enough to give it a shot ... (?)
17:25 murrdoc makes sense there is a delay tho
17:26 murrdoc but i dont know if it should be that bad
17:26 tligda joined #salt
17:26 tru_tru joined #salt
17:28 cwright i'm struggling trying to get pillar to reference a previous defined pillar variable. the docs say this is possible, but I have not been able to make it work
17:28 cwright has anyone else succeeded in accomplishing this?
17:30 gattie joined #salt
17:31 aphorise joined #salt
17:32 Deevolution akafred:  We're using it.  The command and control abilities alone sold me on it, even with the flaws.
17:33 Deevolution akafred: Also, since we have a mixed environment, it's doing a good job of allowing us a single tool to control Linux/Windows servers.
17:37 akafred Deevolution: Awesome, thanks  ... we were discussing it at work today, this will make me more confident in going forward with using it for Windows boxes.
17:37 fivmo left #salt
17:42 iggy cwright: don't... where do the docs say it's possible?
17:44 murrdoc left #salt
17:44 JDiPierro joined #salt
17:46 cwright iggy: unless I'm misunderstanding, right here: http://docs.saltstack.com/en/latest/topics/pillar/#including-other-pillars
17:47 murrdoc joined #salt
17:48 neogenix joined #salt
17:48 desposo joined #salt
17:50 neogenix Hallo! I have an odd behavior in archive. Anyone seen anything like this : https://gist.github.com/neogenix/b79520d9415a68c2e441
17:50 iggy oh, that kind of include... cwright so what are you trying to do? paste as much code as you can (to refheap or gist or wherever)
17:59 cwright iggy: this is about as much as I can post: https://gist.github.com/corywright/8eacc85e027a053db332
18:00 hal58th neogenix maybe you need the z flag in tar_options?
18:00 neogenix @hal58th same result.
18:01 neogenix hal58th: I’ve tried z, xvf, zxvf, zv, zx.
18:01 cwright iggy: i need the gateway to be different based on the cabinet the server is in. the cabinet is stored in a grain on each server via network:switch
18:01 aparsons joined #salt
18:01 iggy nope, that's not what includes are for
18:01 Deevolution aar
18:02 cwright what are they for then?
18:02 iggy you can't pillar.get in a pillar
18:02 iggy doesn't matter what sequence of events lead up to it
18:02 douglassssssssss joined #salt
18:02 murrdoc u can possibly do a __pillar__[name] in them
18:03 murrdoc but its a hit/miss
18:03 iggy ^
18:03 iggy there are actually circumstances where pillar.get will work... but more often than not it doesn't
18:03 iggy so don't
18:03 iggy find another way to achieve what you need to do
18:03 murrdoc and really dangerous
18:06 cwright thanks
18:06 MatthewsFace joined #salt
18:06 hal58th neogenix not sure why it's failing to find it. look up github issues?
18:07 iggy cwright: can you not do the lookup in the state instead of needing it done in the pillar?
18:08 hal58th neoginix your files names for /opt/apache-cassandra-1.2.19 don't look like they match
18:09 cwright iggy: that's a good question
18:09 murrdoc iggy , i use grains in pillars to enable different pillars, but not more than lsb_release
18:09 cwright let me give that a shot
18:09 iggy yeah, there's a bug open about custom grains not working in pillars anyway (I opened it)
18:10 murrdoc custom from /etc/salt/grains or from _grains
18:11 iggy _grains
18:11 murrdoc yeah
18:11 murrdoc the ones from /etc/salt/grains work just fine
18:13 iggy _grains worked at one point... don't know exactly when they stopped
18:13 murrdoc i smell a base pi update
18:13 iggy get the pitchforks and torches!
18:13 aparsons joined #salt
18:13 linjan joined #salt
18:14 iggy I probably should git bisect it at some point
18:14 iggy nobody from saltstack has looked at it (other than the pi of base actually)
18:15 murrdoc based pi god
18:15 smcquay joined #salt
18:15 JDiPierro joined #salt
18:16 nikogonzo wubwub does salt have a big presence in scale13 this week? i see they're a sponsor but is anyone going?
18:16 robawt i hope so
18:17 murrdoc its got a whole day
18:17 iggy there are at least 2 devs I know of going (probably more, that's just what I've picked up in here)
18:17 forrest joined #salt
18:17 fbettag joined #salt
18:18 nikogonzo most excellent
18:18 murrdoc the guy who our company got training from is going to scale
18:18 murrdoc that guy is the man
18:18 iggy dave?
18:19 murrdoc nah
18:19 murrdoc shall remain nameless
18:19 murrdoc super humble dude
18:19 iggy ahh
18:19 iggy yeah, I just said that because I know he's been doing a lot of traveling lately
18:20 murrdoc it was seth
18:22 beneggett joined #salt
18:23 subsigna_ joined #salt
18:25 spookah joined #salt
18:26 spiette joined #salt
18:27 neogenix hal58th: I’m pretty sure they match. The file is in the files dir, and is on the master, as well as appears on the minion in the cache all by itself.
18:28 hal58th not sure what to tell you but good luck
18:29 neogenix hal58th: fwiw, i did create a file.manage using the same source, and works great out the box. I can always convert these to cmd + file, but was hoping to use archive.
18:32 hebz0rl_ joined #salt
18:32 joehoyle joined #salt
18:33 joehoyle Hey, I'm using the s3fs file backend for my salt master - but I don't work out where I should put my custom salt runners?
18:34 germs_ joined #salt
18:35 rvankleeck joined #salt
18:35 amcorreia_ joined #salt
18:35 rvankleeck in version 2014.7.0, I'm trying to create a report of jobs that have run in the last X days.
18:36 rvankleeck Trying to get the output of salt-run jobs.list_jobs in a parseable format using outputter=<something> gives me an error stating that option is not valid
18:37 rvankleeck is that not implemented in my version? it looks like it would be according to: http://docs.saltstack.com/en/latest/ref/runners/all/salt.runners.jobs.html#salt.runners.jobs.list_jobs
18:39 SheetiS should the __salt__ dict be available inside a salt/utils/<some_util>.py?  Would I want it to be, or should I handle everything inside of the module that calls the util before passing it on?
18:42 N-Mi joined #salt
18:42 jalaziz joined #salt
18:45 toanju joined #salt
18:47 schlueter joined #salt
18:49 CeBe1 joined #salt
18:50 spookah joined #salt
18:51 schlueter1 joined #salt
18:52 aparsons_ joined #salt
18:52 SheetiS It seems it is not available in 2014.7.1 but maybe is in newer versions?  https://github.com/saltstack/salt/blob/develop/salt/utils/fsutils.py#L88 but trying to use __salt__ in something on 2014.7.1 results in this: NameError: global name '__salt__' is not defined.   Now obviously I could import all-the-things to make __salt
18:52 SheetiS __salt__ available, but that seems bad juju.
18:52 viq joined #salt
18:53 mikaelhm joined #salt
18:54 otter768 joined #salt
18:54 Zachary_DuBois joined #salt
18:58 mikkn joined #salt
18:59 scoates joined #salt
19:00 s0lar hi ppl, anyone have any idea why I get this when running pip install?`Requested salt==2014.7.1, but installing version 2014.7.0
19:01 forrest is .1 available?
19:11 ckao joined #salt
19:12 s0lar forrest, yep
19:12 s0lar it is on pypi
19:12 s0lar but I cannot install it :/
19:12 forrest s0lar: Do you have an old version of pip?
19:13 forrest could update to the latest with pip install --upgrade pip
19:13 baweaver joined #salt
19:14 yomilk joined #salt
19:15 igorwidl for salt-call , is there a way to specify port number along with master?
19:16 MatthewsFace joined #salt
19:16 timoguin joined #salt
19:20 forrest igorwidl: You would need to set that in the minion config file.
19:23 chris-m reposting: has anyone specific experience with syndics and having mom sync 'custom modules-python scripts' to it
19:24 stylica joined #salt
19:25 chris-m is there a pre-step I need to do on the lower-level master (which has the minion running) to sync the modules?   I run the command ./salt.sh '*' saltutil.sync_modules, and see in the debug logs of the syndic/low-level master a bunch of key exchanges and a job referencing the sync_modules. but no files are actually synced
19:26 scoates_ joined #salt
19:26 eichiro joined #salt
19:27 s0lar forrest, we upgraded pip but now get Requested salt==2014.7.1, but installing version None :)
19:29 dyasny joined #salt
19:29 NVX joined #salt
19:30 forrest s0lar: I've got a system running pip 1.4.1, python 2.6, was able to install to a fresh venv, so it's definitely working there. Maybe try creating a new virtualenv to see if something is screwed up with that one?
19:30 _prime_ joined #salt
19:31 yidhra_ joined #salt
19:32 toanju joined #salt
19:32 spookah joined #salt
19:35 mailo joined #salt
19:36 baweaver joined #salt
19:36 SheetiS is anyone here successfully using the pagerduty integration stuff at all?  it seems that every time I get somewhere on it, I find another issue.  I got the module accepting pillar config options (and will have a PR for it soon), but it appears that the thing wants a pagerduty.service config option to go into the data['service_key'] even if the module provides the data['service_key'] dict separately.  pagerduty.service is completely und
19:38 alanpearce joined #salt
19:39 waltstack joined #salt
19:40 s0lar forrest, thanks. seems more broken than we thought. downgrading to 2014.7.0 for now
19:40 s0lar had some pyzmq problems also
19:41 TTimo joined #salt
19:41 s0lar it might be our local devpi cache that is causing downgrade as well, no time to investigate unfortunately.
19:41 s0lar or something is broken in the setup.py / dependencies in 2014.7.1
19:43 forrest s0lar: That's good you were able to confirm at least. I feel like there is a way with pip to ignore certain repos, but I can't remember what it is.
19:46 hobakill joined #salt
19:48 claytonk is there a way to test a specific state using salt-call?
19:49 murrdoc salt-call state.sls ?
19:49 claytonk murrdoc: thanks
19:49 roolo joined #salt
19:51 edrocks joined #salt
19:51 hal58th claytonk salt-call sys.doc state.sls
19:52 claytonk hal58th: thank you
19:52 joehoyle joined #salt
19:52 hal58th welcome claytonk
19:55 brick_ joined #salt
19:57 jdesilet joined #salt
20:00 Ilja joined #salt
20:01 mschiff is it possible to target minions with regex within a pillar sls file? (e.g. using some {% if  ... %} ?
20:03 pppingme joined #salt
20:03 Ilja1 joined #salt
20:04 hemebond mschiff: Have a look at Salt Mine
20:07 mschiff hemebond: hmmm mybe I should rephrase: How can I do regex comparison in a jinja if-Statement?
20:08 hemebond oh
20:08 nullptr joined #salt
20:09 hemebond http://stackoverflow.com/questions/28327764/saltstack-use-regexps-in-pillar-file/28386071#28386071
20:10 hemebond Apparently not.
20:10 hemebond Though I suppose you could use pyobjects for your SLS
20:10 SheetiS mschiff: with the jinja renderer, I do not believe there is a way.  The easiest way to implement would be a salt module so that you could do something like __salt__['custom.re_match'](pattern, buffer) or something like htat
20:10 SheetiS then just have that match to re.match
20:10 SheetiS pyobjects or py renderer would be another option.
20:11 Ilja joined #salt
20:11 mschiff Hm ok thanks guys. Propably its better to it another way then: SPlitting the file then do the matching in the top.sls ...
20:12 timoguin joined #salt
20:15 subsignal joined #salt
20:15 chris-m maybe I can start with has anyone used a salt-syndic in their linux infrastructure?
20:16 murrdoc have u posted to teh salt users mailing list
20:16 murrdoc https://groups.google.com/forum/#!forum/salt-users
20:16 murrdoc might be better
20:17 chris-m hi murrdoc, yes,  no feedback since I posted on Monday https://groups.google.com/forum/#!searchin/salt-users/Passing$20modules$20down$20from$20MOM$20to$20the$20syndic|sort:relevance/salt-users/hN5E0Uf8wR4/V5UrSIJ4bNkJ.   thx fro the suggestion though
20:17 stylica joined #salt
20:17 bash124512 anyone used renderers.gpg ?
20:18 mschiff SheetiS: but I will remember your suggestions, looks good ;)
20:19 SheetiS bash124512: I use the gpg renderer right now for a lot of my pillar data.
20:21 bash124512 SheetiS : http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.gpg.html
20:21 bash124512 I don't get the documentation
20:21 alanpearce joined #salt
20:21 bash124512 I don't understand how it works
20:21 hemebond left #salt
20:22 murrdoc put #!yaml|gpg on the top of the pillar that may or may not hvae gpg data in it
20:22 murrdoc
20:22 murrdoc success
20:22 cpowell joined #salt
20:23 subsignal joined #salt
20:23 cpowell joined #salt
20:23 edrocks are you supposed to put your installers in highstate?
20:25 bash124512 murrdoc  : lets say I have many keys. would that work ?
20:25 bash124512 would I be able to specify the key for the pillar file ?
20:26 murrdoc SheetiS:  might know more, but yes it should work
20:26 bash124512 so I just import the keys to gpg and it will try all keys ?
20:28 SheetiS bash124512: you have multiple gpg keys and you want to encrypt different parts of your config with different keys?
20:28 SheetiS Let me check my configuration to see how I would go about doing that.
20:30 neogenix joined #salt
20:31 bash124512 I'm not sure either because I have a master, syndinc and minions type of setup.
20:31 bash124512 there are many syndics
20:31 bash124512 I'm not even sure where they passwords are going to be decrypted
20:32 SheetiS bash124512: I currently only use a master and minions, but I will say this.  I only put gpg data in my pillar so it is always decrypted on my master
20:32 SheetiS your syndic connects to some other master for some states/configs?
20:33 bash124512 syndic and second master are in the same server
20:33 bash124512 I generate pillara data from master of masters
20:33 SheetiS I'd definitely have the gpg stuff on the master-of-masters then since that is where you generate your pillar data
20:34 intellix joined #salt
20:34 bash124512 It would seem that I will have to have the keys on both syndic and master-of-masters even if I wanted to decrypt them on the syndic
20:34 SheetiS then your pillar can be encrypted at rest there and be passed out decrypted to the minions that are supposed to know the data
20:34 SheetiS How are you storing pillar data right now?  is it on the filesystem in /srv/pillar?  or using like a git ext pillar, or something else?
20:35 bash124512 filesystem
20:35 bash124512 bah,
20:35 bash124512 yes git
20:35 bash124512 thats why I need to use the thingy :)
20:35 SheetiS if the pillar data lives on /srv/pillar on the masters via ext_pillar via git or on the filesystems
20:36 SheetiS then all the masters will need the gpg private key so that they can decrypt the encrypted pillar data.
20:36 bash124512 master-of-masters is in git
20:36 linjan joined #salt
20:37 SheetiS anywhere the pillar data lives is where you will have to have the gpg keys installed and configured is the short of it.
20:37 bash124512 I think I get it now, so I'll have to move the gpg keys to my syndic ( second master ) and decrypt the data there.
20:37 stylica joined #salt
20:38 bash124512 keys will have to be on both master-of-masters and second master
20:38 SheetiS Best thing I could say is to give that a try and see if it does what you need.  My environment is small enough that I don't feel a big need for syndic or multiple masters (only ~100 virtual machines or so).
20:38 SheetiS bash124512: I think that sounds right.
20:40 stylica joined #salt
20:40 baweaver joined #salt
20:42 bash124512 ok, lets say I have 100 passwords I need to encrypt and have them in pillar. can you think of an easy way to get them generated @ pillar ?
20:43 bash124512 ofc 100 is an exaggeration
20:43 nitti joined #salt
20:43 claytonk is there a method for using grains to get a specific interface IP address for a jinja template?
20:43 hal58th yes claytonk, one second
20:44 hal58th clayonk, not through grains, but the network module.
20:44 hal58th {{ salt['network.interfaces']()['eth0']['inet'][0]['address'] }}
20:45 hal58th bash124512 a looping bash script? Not sure what your problem is.
20:46 chris-m @murrdoc - if you come across anyone that has successfuly deployed a syndic, please forward them to me (via email: pc911consulting@bell.net)  thank-you
20:46 bash124512 hal58th : I don't want to do anything manually
20:46 hal58th still not sure what your problem
20:46 dyasny joined #salt
20:47 cberndt joined #salt
20:49 iggy {{ salt['grains.get']('ip4_interfaces')['eth0'][0] }} should work too
20:49 bash124512 hal58th : lack of imagination :D
20:50 claytonk thanks, gents
20:51 stylica_ joined #salt
20:52 tvinson joined #salt
20:53 ghanima has anyone seen this error before
20:53 ghanima Failed to authenticate!  This is most likely because this user is not permitted to execute commands, but there is a small possibility that a disk error occurred (check disk/inode usage)
20:53 ghanima I have been getting it as of late since I added some new hosts to my environment
20:55 linjan joined #salt
20:55 rvankleeck ghanima, are you using external_auth or client_acl?
20:55 ghanima no
20:55 ghanima I am not
20:55 ghanima I am actually root
20:55 ghanima when I execute these commands
20:55 otter768 joined #salt
20:55 rvankleeck well that's...odd...
20:55 rvankleeck what user is salt running as on the minion/master?
20:56 ghanima root on both sides
20:56 rvankleeck weird. i've only seen that with acl/external auth
20:57 rvankleeck maybe restart the minions/make sure they are running and accepted?
20:57 iggy ghanima: shut down the master and make sure there aren't any rogue salt-master processes running
20:57 iggy also... check inode usage
20:57 stylica_ joined #salt
20:58 ghanima iggy: doing a restart now
20:58 ghanima iggy: for inode usage just the master right?
20:58 baweaver joined #salt
20:59 dooshtuRabbit joined #salt
21:01 iggy yes
21:03 yomilk joined #salt
21:03 joehoyle joined #salt
21:04 joehoyle Is anyone else here using s3fs / gitfs backend with a masterless setup? I can't seem to get custom modules / states / grains working
21:05 iggy joehoyle: what version?
21:06 joehoyle @iggy 2014.7.1 (Helium)
21:06 rojem joined #salt
21:07 jY on a syndic host.. do i need to run syndic+master?
21:09 joehoyle iggy: running saltutil.sync_grains for example gives me this: https://gist.github.com/joehoyle/36819e86ce77f7c2ebea
21:09 joehoyle though then the grains are not available in grains.items
21:10 beneggett joined #salt
21:11 desposo joined #salt
21:14 joehoyle iggy: just tried 2015.2 rc1 also, no better - so I wonder if I'm configuring something wrong
21:15 iggy I don't know, if it was <2014.7 I was going to say upgrade... other than that, I don't use masterless
21:15 spookah joined #salt
21:16 iggy I wouldn't be surprised if that's expected _not_ to work in a masterless setup
21:16 joehoyle hmm right ok
21:16 joehoyle maybe I'll post to the mailing list
21:17 iggy keep an eye out for ryan_lane
21:17 joehoyle cool
21:17 iggy they use masterless extensively
21:17 joehoyle sweet, thnKA
21:18 meylor joined #salt
21:24 ajw0100 joined #salt
21:24 rgarcia_ joined #salt
21:25 mapu left #salt
21:26 ciss joined #salt
21:28 madisonmay joined #salt
21:29 madisonmay Hi there!  Does anyone have experience setting up docker device passthrough with salt?
21:29 madisonmay I see a port mapping, but no settings for device mappings.
21:29 masterkorp joined #salt
21:30 fxhp joined #salt
21:32 _prime_ joined #salt
21:33 igorwidl left #salt
21:35 beneggett joined #salt
21:35 ghanima Is there any progress on this issue: https://github.com/saltstack/salt/issues/12248
21:35 ghanima since the upgrade to latest I am seeing this issue alot
21:36 ghanima I have attempted multiple restarts
21:37 moos3 when using gitfs on centos7 my states fail to run, am I doing something wrong ?
21:37 moos3 I get a error about can't find the state
21:39 Ahlee if you salt \* cp.list_master do you see your state?
21:39 ghanima also can someone tell me which version this was introduced in: https://github.com/saltstack/salt/pull/19350
21:41 murrdoc lazy q, do pillars do includes
21:41 kevinforbes joined #salt
21:41 murrdoc like outside of top.sls can i do include: pillar.name inside of a yaml pillar file
21:42 SheetiS ghanima: The first tag I see it in is 2014.7.1.
21:42 SheetiS I do not see it in 2014.7.0
21:42 bhosmer_ joined #salt
21:43 ghanima SheetiS: so as long as I am on version higher than 2014.7.1 than It should have the fix
21:43 kevinforbes @murrdoc I don’t know about pillars, but in states you can do that inside an include block
21:43 kevinforbes include:
21:43 SheetiS It should be in anything 2014.7.1 or later, yes.
21:43 kevinforbes - pillar.name
21:43 iggy murrdoc: yes
21:44 JDiPierro joined #salt
21:44 murrdoc saweet
21:44 ghanima basepi: are you online
21:44 murrdoc thank u for supporting my lazy
21:44 kevinforbes http://docs.saltstack.com/en/latest/topics/pillar/#including-other-pillars
21:45 murrdoc rtfm'ed !
21:45 murrdoc ./me puts ice on burn
21:45 kevinforbes It’s cool
21:45 murrdoc thnx
21:47 kevinforbes I’ve got a chicken and egg problem; looking for the right solution.
21:47 basepi ghanima: I am
21:47 kevinforbes 2 states: one to install the nginx package, another to put a conf file at /etc/nginx/sites-enabled/default
21:47 kevinforbes conf file state requires pkg state
21:47 schlueter joined #salt
21:47 murrdoc http://docs.saltstack.com/en/latest/ref/states/requisites.html
21:47 jalaziz joined #salt
21:47 kevinforbes package state watches conf state (so it should restart when the conf changes)
21:48 kevinforbes But that’s not happening; I have to manually restart nginx to get it to read the conf file
21:48 ghanima basepi: this change was put in to 2014.7.1 and later correct?
21:48 basepi ghanima: which?
21:48 ghanima basepi: https://github.com/saltstack/salt/pull/19350
21:49 ghanima basepi: and did this only apply to windows only
21:49 baweaver_ joined #salt
21:50 Brew joined #salt
21:50 evidence anyone using the lxc runner on 2014.7.1?  i'm running 'salt-run lxc.init testies host=salt-master profile=ubuntu', and it's failing
21:51 evidence the runner is actually executing 'lxc-create -n testies -f /tmp/tmpPn_jMO', shouldn't it be passing -t ubuntu also?
21:51 basepi ghanima: Yes, 2014.7.1 and later, and yes, only windows.
21:51 smcquay joined #salt
21:51 basepi evidence: huge rewrite of the lxc runner in 2015.2.0. There's an RC out now, v2015.2.0rc1
21:51 rap424 joined #salt
21:52 evidence basepi: thinking 2014.7.1 is buggy?
21:52 raygunsix joined #salt
21:52 basepi Er, sorry, it's not in that rc, but it's in the 2015.2 branch and will be in v2015.2.0rc2
21:52 basepi evidence: Very possible, I wasn't involved in the rewrite.
21:52 evidence is that one not tagged yet?
21:52 basepi It is not tagged yet.
21:52 evidence i guess i could grab develop also
21:52 MugginsM joined #salt
21:53 evidence basepi: let me try to find that rewrite and give it a quick go, ty
21:53 ghanima basepi: there was a google thread: https://groups.google.com/forum/#!topic/salt-users/br4UXQEz3VE that led to -> https://github.com/saltstack/salt/issues/18515 -> that referneced your fix -> https://github.com/saltstack/salt/pull/19350
21:53 ghanima This issue seems to be happening on my Linux server when I go above 360 minions
21:53 ghanima basepi: curious since your familar with the issue if you think this issue could happen on any os
21:54 basepi ghanima: Well, I'm not super familiar. UtahDave was the one who fixed it.
21:54 basepi I would create an issue, sounds like it's probably a different issue.
21:55 iggy kevinforbes: paste code... so much easier than trying to explain with words
21:55 basepi Link to it, though, it's possible they're related.
21:55 madisonmay Did some digging into salt's support of Docker `--device` flags and it doesn't look like there is any current support for this.    Can anyone recommend a path forward?  Is there any way to pass store raw docker run configuration in a .sls file?
21:55 kevinforbes @iggy, ok here it is
21:55 kevinforbes nginx:
21:55 kevinforbes pkg:
21:55 kevinforbes - latest
21:55 kevinforbes service:
21:55 kevinforbes - running
21:55 kevinforbes - watch:
21:55 iggy paste it somewhere else
21:55 kevinforbes - file: nginx.conf
21:55 kevinforbes nginx.conf:
21:55 kevinforbes file.managed:
21:56 kevinforbes - name: /etc/nginx/sites-enabled/default
21:56 kevinforbes - require:
21:56 kevinforbes - pkg: nginx
21:56 kevinforbes ha ha, right
21:56 peters-tx joined #salt
21:56 iggy next time refheap.com or gist.github.com or something
21:56 kevinforbes http://pastebin.com/5fNirXnr
21:57 iggy A. the id should probably be something other than nginx.conf
21:57 spookah joined #salt
21:57 kevinforbes How come? Seems to work just fine
21:57 iggy because it's confusing at best
21:58 kevinforbes mm, ok I can see that
21:58 iggy nginx.conf looks like a filename
21:58 douglassssssssss left #salt
21:59 iggy try doing the watch_in on the file.managed instead of the watch on the service
21:59 babilen It's typically preferred to use "pkg.latest" rather than "pkg: - latest". In this case you'd use "pkg.latest: [] \n service.running: ..."
21:59 iggy that too
21:59 kevinforbes ok I’ll check it out
22:00 kevinforbes @babilen awesome, I love best practices, thanks
22:00 iggy but do what I said to fix the actual service restart problem
22:00 babilen np
22:00 kevinforbes right
22:04 kevinforbes joined #salt
22:04 denys joined #salt
22:06 TTimo joined #salt
22:08 cpowell_ joined #salt
22:08 karimb joined #salt
22:10 mosen joined #salt
22:12 blackhelmet joined #salt
22:14 desposo joined #salt
22:16 joehoyle joined #salt
22:16 scoates joined #salt
22:16 Brew joined #salt
22:18 perfectsine joined #salt
22:19 ghanima has anyone done in evaluation of raet vs zeromq
22:21 slafs joined #salt
22:21 slafs left #salt
22:22 aquinas joined #salt
22:23 TheoSLC joined #salt
22:25 manfred raet is awesome
22:26 manfred but isn't supported in salt-bootstrap yet
22:26 manfred from the stuff I have tried raet is significantly faster
22:26 manfred (i did end up writing my own bootstrap scripts for just my cloud images, and it was pretty cool)
22:27 manfred i haven't used it in anything in production yet, partly because I don't have anything that would be large enough to need it
22:28 manfred I am interested in trying it with python3, since zmq won't be python3 compatible, because it requires M2Crypto which hasn't been ported.
22:32 Singularo joined #salt
22:35 MatthewsFace joined #salt
22:36 conan_the_destro joined #salt
22:36 ghanima when running salt-run jobs.lookup_jid
22:36 ghanima can you is possible to tell how long it took for jobs to get a response
22:36 ghanima and if anything node was missed
22:39 iggy not really
22:39 ghanima iggy: so does any command do that if you launch a job in the background
22:40 dyasny joined #salt
22:42 __number5__ "to tell how long it took for jobs to get a response" <- you'll need real magic to get that :P
22:42 jeremyr joined #salt
22:43 ghanima __number5__: ok would it be possible to tell if all nodes responded when looking at the job history
22:44 iggy there isn't anything that does that out of the box
22:45 iggy but you could probably write a runner that did something like what manage.down/up does but with jobs
22:45 __number5__ ghanima: that's possible, returner kept all results somewhere, but jobs.lookup_jid is not doing that for you
22:46 __number5__ or maybe some additional returners to store the results in db, then you can check the db for results
22:46 iggy I suspect that's probably what most people end up doing
22:47 iggy once you get to the point of caring about that much detail, you probably have some sort of dashboard for all this cooked up
22:47 iggy too bad nobody has decided to open that stuff up yet
22:47 iggy (at least not that I've heard of)
22:48 ghanima I understand that saltstack product enforces configuration state and is able to to do parallel based execution and orchestrate certain action. I am currently using it for parallel execution of commands using the salt command and the cmd.run module.
22:49 ghanima In my use case I am trying to confirm that everything executed however there atimes when I executing a salt command that some nodes are missed either due to master didn't want to wait for the reply or the minion was down.
22:49 analogbyte iggy, __number5__: I got a kind of dashboard for that, we use it at the company... it isn't much, but it serves for our purposes
22:49 analogbyte https://github.com/analogbyte/saltobserver
22:49 ghanima In those cases its important for me to see those errors
22:50 baweaver joined #salt
22:50 __number5__ analogbyte: nice, thanks
22:51 ryuhei joined #salt
22:51 * iggy takes a peek
22:52 analogbyte __number5__: if theres something you want to know or need with it just highlight me here and I'll try to help, but I gotta go right now, sorry
22:53 ghanima I mean the salt --summary option is extrememly useful no
22:54 ryuhei hi all
22:55 kermit joined #salt
22:55 ryuhei when I spin up minion,, does it take configuration from minion config file or.. cloud config file?
22:55 stylica joined #salt
22:56 otter768 joined #salt
22:56 stylica joined #salt
22:57 Alec I'm having no luck executing a .run file using cmd.run. Is there another way I should try to run the shell script?
23:03 baweaver joined #salt
23:03 babilen Alec: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.cmd.html#salt.states.cmd.script maybe?
23:04 babilen (no experiece with Windows)
23:06 Alec hmm...I'll give that a try, babilen. Thanks. I'm super new to Salt and read that page before but I think I overlooked that bit.
23:08 babilen Alec: cmd.run literally takes a command that you would execute. That would necessitate that you put the script on the minion beforehand (e.g. with file.managed) and then give the command to run it (no idea what that might be on Windows). cmd.script takes a salt:// source and lets you execute the content of that file as "script" (whatever that actually entails on Windows I do not know, does it support shebangs?)
23:11 Alec I'm actually accessing the servers via Putty (SSH), so I can avoid any of the Windows idiosyncrasies.
23:11 iggy also... define "having no luck"
23:12 iggy and most .run files I've seen are interactive commands
23:12 Alec Oh. No luck meaning that after I do something like "salt '*' cmd.run 'sh /root/thisfilehere.run'" the master just sits there and never returns anything.
23:12 Alec This one isn't interactive. I did check that bit.
23:15 rojem joined #salt
23:20 subsignal joined #salt
23:21 yomilk joined #salt
23:21 neogenix joined #salt
23:25 karimb joined #salt
23:26 steveoliver left #salt
23:27 steveoliver joined #salt
23:27 steveoliver joined #salt
23:27 sjol joined #salt
23:28 spookah joined #salt
23:28 steveoliver left #salt
23:29 jstorey_ joined #salt
23:32 klaatu joined #salt
23:34 wnkz_ joined #salt
23:34 klaatu hello, I'm working on a POC for my company/team to setup an automation framework. I read about salt stack and was wanting try this one out. I managed to configure it on my system running mac os x (mavericks) and another machine that I borrowed which os x as well (yosemite). I successfully started salt-master on my machine and salt-slave on the other machine. The slave machine exits salt with this error - [ERROR   ] Attempt to authentic
23:35 klaatu I did google for some help but none solution suits my problem. Can someone in the group help me? Thank you.
23:36 __number5__ klaatu: better to do you POC on linux unless your goal is managing all macs
23:37 klaatu Unfortuantely I'd have to run slaves on mac machines
23:37 murrdoc or vagrant
23:38 klaatu The project is to help test ios and android applications. The devices are mostly connected to mac machines (a few win machines)
23:38 igorwidl joined #salt
23:39 CeBe2 joined #salt
23:40 aphoriser joined #salt
23:43 mosen klaatu: im running salt on all macs
23:43 mosen brb a minute :)
23:47 jstorey_ klaatu: did you accept the minion's certificate on the salt master?
23:48 roolo joined #salt
23:48 ryuhei Does master_sign.pub has to be copied to minions manually ?
23:49 klaatu I tried to do that but do not have permissions to the pki folder.
23:50 klaatu I was able to move a step forward. I'm not sure what changed, I restarted salt master a couple of times and now I see the master has accepted the key
23:51 klaatu But the slave throws the flowing exception -
23:51 klaatu [ERROR   ] An un-handled exception was caught by salt's global exception handler: TypeError: string indices must be integers, not str Traceback (most recent call last):   File "/usr/local/bin/salt-minion", line 9, in <module>     load_entry_point('salt==2014.7.0', 'console_scripts', 'salt-minion')()   File "/Library/Python/2.7/site-packages/salt/scripts.py", line 57, in salt_minion     minion.start()   File "/Library/Python/2.7/site-pa
23:51 iggy you also know that just for testing/PoC, you can run a minion on the same system as the master

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