Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-08-29

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

All times shown according to UTC.

Time Nick Message
00:01 patrickcp Should be a search() command that will allow you to look for content in a file.
00:02 Setsuna666 If I want to get all its content ?
00:02 Outlander joined #salt
00:03 bnikolaus rlarking: if '*' targets all servers what is 'web*prod*': tarting?
00:04 bnikolaus sorry im retarted... targeting :)
00:05 Ryan_Lane bnikolaus: anything that matches that glob
00:05 Ryan_Lane so: web-production-001.example.com would match
00:09 Setsuna666 Well, just made my one function :P
00:10 bezeee joined #salt
00:12 bnikolaus Thats what I figured, thanks for the confirmation
00:15 aquinas_ joined #salt
00:21 ckao joined #salt
00:21 oz_akan joined #salt
00:23 oz_akan_ joined #salt
00:26 murrdoc joined #salt
00:28 catpiggest joined #salt
00:28 murrdoc anyone here test out using an external enc, like redis, for their salt installs
00:30 murrdoc aight cool
00:30 saurabhs left #salt
00:34 debian112_ salt-ssh crashes: http://paste.debian.net/118111
00:35 debian112_ salt-ssh 2014.1.1
00:35 murrdoc joined #salt
00:36 kingel joined #salt
00:38 packeteer debian112_: second line has your error and possible answer
00:39 debian112_ works fine in: salt-ssh 2014.1.10 (Hydrogen)
00:39 debian112_ same command: packeteer
00:40 debian112_ not sure why it is trying to bind ssh port
00:40 debian112_ locally
00:43 martoss1 joined #salt
00:48 darrend joined #salt
00:49 schimmy joined #salt
00:51 schimmy joined #salt
00:51 sectionme joined #salt
00:52 schimmy1 joined #salt
00:54 TyrfingMjolnir joined #salt
00:57 packeteer so a possible bug then?
00:58 otter768 joined #salt
01:00 kuffs joined #salt
01:02 mechanicalduck_ joined #salt
01:05 Ryan_Lane http://ryandlane.com/blog/2014/08/29/a-saltstack-highstate-killswitch/ :D
01:05 manfred nice
01:07 pled76 joined #salt
01:07 to_json joined #salt
01:10 to_json1 joined #salt
01:13 [diecast] joined #salt
01:13 [diecast] left #salt
01:23 pled76 joined #salt
01:23 oz_akan joined #salt
01:25 rallytime joined #salt
01:30 pled76 joined #salt
01:40 CeBe1 joined #salt
01:41 mgw2 joined #salt
01:48 TTimo joined #salt
01:52 icebourg joined #salt
01:52 sectionme joined #salt
01:52 pled76 joined #salt
01:53 icebourg joined #salt
02:05 pled76 joined #salt
02:10 _ikke_ joined #salt
02:12 dvestal joined #salt
02:12 skullone is there a good way to have pillar setting inheritance similar to hiera, per node?
02:13 TyrfingMjolnir joined #salt
02:13 skullone ie. if undefined, hosts get X, but if a node is define in a pillar, merge/inherit settings?
02:13 TTimo joined #salt
02:13 dimeshake sure
02:14 dimeshake look at defaults and context for yaml
02:14 dimeshake set a default if pillar isn't present
02:16 jalaziz joined #salt
02:17 dimeshake skullone: you might want to look at some of the salt formulas to see how others do it
02:22 pled76_ joined #salt
02:23 schimmy joined #salt
02:24 mgw2 joined #salt
02:24 oz_akan joined #salt
02:26 schimmy1 joined #salt
02:27 FeatherKing joined #salt
02:31 pled76 joined #salt
02:34 dccc joined #salt
02:34 floWenoL joined #salt
02:50 pled76_ joined #salt
02:52 skullone is there a way to have a per-node pillar file that is automatically looked up?
02:52 skullone ie. to store interface config, outside of the main pillars
02:53 sectionme joined #salt
02:54 tmh1999 joined #salt
02:54 favadi joined #salt
03:02 bhosmer joined #salt
03:05 skullone hrm reclass may be what i want
03:06 madduck \o/
03:07 skullone is that your approval? :p
03:07 madduck yessir ;)
03:07 skullone i use puppet with puppet dashboard as ENC, but its pretty shitty as an ENC
03:08 skullone hiera is ok, but it still is clunky to me
03:08 tmh1999 joined #salt
03:09 nhubbard_ joined #salt
03:09 madduck reclass was designed to be lightweight
03:10 madduck and puppetlabs are interested, i learned today at #DebConf 14
03:10 malinoff joined #salt
03:10 skullone yah?
03:10 skullone hiera is fairly lightweight too
03:11 icebourg joined #salt
03:12 whytewolf joined #salt
03:17 mosen yeah hiera and reclass seem similar ish :)
03:17 mosen except for the fact pillar stores the parameters
03:18 mosen lemme revise and say hiera is the same if you use it purely for ENC purposes :)
03:20 jab416171 joined #salt
03:20 anotherZero joined #salt
03:20 schimmy1 joined #salt
03:20 anotherZero joined #salt
03:23 pled76 joined #salt
03:25 oz_akan joined #salt
03:32 oz_akan joined #salt
03:33 oz_akan joined #salt
03:41 CryptoMe1 Is this the correct method to execute grain targetting in the top file? I can't seem to get it to work
03:41 CryptoMe1 'service:web': - match: grain - sa
03:42 CryptoMe1 imagine a \n after each ':'
03:44 malinoff CryptoMe1, http://pastie.org
03:44 CryptoMe1 good call:
03:44 CryptoMe1 http://pastie.org/9511981
03:45 malinoff CryptoMe1, this is a correct one: http://pastie.org/9511983
03:46 CryptoMe1 ah, it's part of the base ! gotcha
03:47 malinoff CryptoMe1, top-level dict key (e.g. 'base') must specify salt environment
03:47 malinoff and all states should be specified within environments
03:49 CryptoMe1 Ya, that makes sense. thx for teh assist.
03:50 tmh1999 joined #salt
03:54 sectionme joined #salt
03:56 bnikol joined #salt
04:00 tmh1999 joined #salt
04:02 tmh1999 joined #salt
04:03 pled76 joined #salt
04:03 oz_akan joined #salt
04:04 schimmy joined #salt
04:05 Ryan_Lane joined #salt
04:05 oz_akan_ joined #salt
04:07 mosen joined #salt
04:07 schimmy1 joined #salt
04:12 bnikol joined #salt
04:18 tmh1999 joined #salt
04:19 tmh1999 joined #salt
04:21 tmh1999 joined #salt
04:27 Sway joined #salt
04:30 xcbt joined #salt
04:30 murrdoc joined #salt
04:31 ianmcshane joined #salt
04:41 tmh1999 joined #salt
04:50 tmh1999 joined #salt
04:51 bhosmer joined #salt
04:51 kermit joined #salt
04:53 murrdoc joined #salt
04:53 pled76 joined #salt
04:54 ramteid joined #salt
04:55 sectionme joined #salt
05:06 TTimo joined #salt
05:06 oz_akan joined #salt
05:08 melinath joined #salt
05:20 Sway joined #salt
05:25 dvestal_ joined #salt
05:27 occup4nt joined #salt
05:27 pled76 joined #salt
05:30 Sauvin joined #salt
05:31 occupwnt joined #salt
05:49 occupwnt joined #salt
05:50 tmh1999 joined #salt
05:53 calvinh joined #salt
05:54 occup4nt joined #salt
05:55 mpoole joined #salt
05:55 sectionme joined #salt
05:55 dccc_ joined #salt
05:57 17SAA1J7Y left #salt
06:04 harkx joined #salt
06:07 oz_akan joined #salt
06:07 ianmcshane joined #salt
06:11 simonmcc___ joined #salt
06:12 Kelsar_ joined #salt
06:12 thayne joined #salt
06:14 jgelens joined #salt
06:15 pled76 joined #salt
06:17 Jarus joined #salt
06:17 Emantor joined #salt
06:21 mariusv joined #salt
06:24 jhauser joined #salt
06:25 catpigger joined #salt
06:25 kingel joined #salt
06:30 felskrone joined #salt
06:31 pixman joined #salt
06:33 mariusv joined #salt
06:33 roolo joined #salt
06:39 bhosmer joined #salt
06:40 mechanicalduck joined #salt
06:40 melinath joined #salt
06:40 Ryan_Lane joined #salt
06:47 the_lalelu joined #salt
06:51 Sway joined #salt
06:56 sectionme joined #salt
06:59 cym3try joined #salt
06:59 kaictl joined #salt
07:00 melinath joined #salt
07:00 ghartz joined #salt
07:05 scalability-junk joined #salt
07:06 lcavassa joined #salt
07:06 mechanicalduck_ joined #salt
07:08 oz_akan joined #salt
07:08 mr_chris joined #salt
07:09 TTimo joined #salt
07:18 rawzone joined #salt
07:20 ml_1 joined #salt
07:21 skullone sweet, moved all my pillar data into reclass
07:21 skullone this is nice
07:21 pled76 joined #salt
07:22 thayne joined #salt
07:23 pled76_ joined #salt
07:24 slav0nic joined #salt
07:25 skullone one odd thing, the 'environment' flag is set under reclass, and not under salt
07:26 skullone environment: none, when using reclass, but instead its defined under __reclass__ environment:
07:26 fredvd joined #salt
07:31 skullone https://groups.google.com/forum/#!topic/salt-users/pFTU3bJy3sU
07:32 beebeeep ‚
07:32 skullone looks like i should try the dev git branch of reclass to get this
07:33 mekstrem joined #salt
07:35 intellix joined #salt
07:41 mekstrem ~~~~~~~~
07:41 mekstrem oops
07:42 skullone madduck: so i have a use case for $environments, if you were looking for one
07:42 skullone at least for setting environment for the state tree
07:43 skullone for now, i might go back to using standard tops for state, but keep reclass for pillar data
07:48 tmh1999 joined #salt
07:49 Ryan_Lane joined #salt
07:50 Ryan_Lane joined #salt
07:53 chiui joined #salt
07:58 sectionme joined #salt
07:58 ianmcsha_ joined #salt
08:08 oz_akan joined #salt
08:09 che-arne joined #salt
08:10 oz_akan_ joined #salt
08:11 kingel joined #salt
08:14 kingel joined #salt
08:20 kaictl joined #salt
08:22 ml_1 joined #salt
08:25 sectionme joined #salt
08:28 bhosmer joined #salt
08:32 N-Mi joined #salt
08:32 N-Mi joined #salt
08:34 packeteer how do clients handle not being able to talk to the server for an extended time? ie. the server is on the other side of a VPN link
08:34 pled76 joined #salt
08:45 TheThing joined #salt
08:47 sctsang joined #salt
08:47 jhauser joined #salt
08:53 MrTango joined #salt
08:58 sctsang1 joined #salt
09:02 pled76 joined #salt
09:11 TTimo joined #salt
09:11 oz_akan joined #salt
09:15 darkelda joined #salt
09:16 Kenzor joined #salt
09:17 rawzone joined #salt
09:17 harkx joined #salt
09:17 dccc_ joined #salt
09:17 jab416171 joined #salt
09:17 CeBe1 joined #salt
09:17 viq joined #salt
09:17 toddnni joined #salt
09:17 aberdine joined #salt
09:17 zartoosh joined #salt
09:17 claytron joined #salt
09:17 drogoh joined #salt
09:17 smkelly joined #salt
09:17 ifmw joined #salt
09:17 clone1018__ joined #salt
09:17 ede joined #salt
09:17 user___ joined #salt
09:17 dhwty_ joined #salt
09:17 ngo joined #salt
09:17 kalessin joined #salt
09:17 homelinen joined #salt
09:17 rblackwe joined #salt
09:19 mechanicalduck joined #salt
09:20 anotherZero joined #salt
09:21 fxdgear joined #salt
09:22 rmnuvg joined #salt
09:22 jerrcs joined #salt
09:22 wangofett joined #salt
09:22 mirko joined #salt
09:22 Heggan joined #salt
09:22 sdebot joined #salt
09:22 mugsie joined #salt
09:22 zsoftich joined #salt
09:22 mikkn joined #salt
09:22 teepark joined #salt
09:22 Doqnach joined #salt
09:22 GnuLxUsr joined #salt
09:22 ksalman joined #salt
09:23 jut joined #salt
09:23 cym3try joined #salt
09:26 darkelda joined #salt
09:26 esogas joined #salt
09:27 jmccree joined #salt
09:27 redondos joined #salt
09:27 garphy joined #salt
09:30 N-Mi joined #salt
09:30 17SAA16YQ joined #salt
09:30 ramteid joined #salt
09:30 WarP|onwork joined #salt
09:30 17SAA0TK4 joined #salt
09:30 johngrasty joined #salt
09:30 blast_hardcheese joined #salt
09:30 rhand joined #salt
09:30 codysoyland joined #salt
09:30 roo9 joined #salt
09:30 stan_k joined #salt
09:30 notpeter_ joined #salt
09:30 Deevolution joined #salt
09:30 Doqnach`` joined #salt
09:30 z3uS joined #salt
09:30 EWDurbin joined #salt
09:30 renoirb joined #salt
09:30 MaZ- joined #salt
09:30 jayne joined #salt
09:30 Schmidt joined #salt
09:30 crazysim joined #salt
09:30 JoeHazzers joined #salt
09:30 nliadm joined #salt
09:30 al joined #salt
09:30 jY joined #salt
09:30 jforest joined #salt
09:33 TheThing joined #salt
09:34 ianmcshane joined #salt
09:35 elfixit1 joined #salt
09:38 nebuchad` joined #salt
09:40 badon joined #salt
09:41 teunb joined #salt
09:45 ggoZ joined #salt
09:45 EntropyWorks joined #salt
09:46 martoss joined #salt
09:46 Xiao joined #salt
09:46 grepory joined #salt
09:47 Dinde joined #salt
09:47 g3cko joined #salt
09:47 ClausA joined #salt
09:47 svx_ joined #salt
09:47 eculver_ joined #salt
09:47 kriberg joined #salt
09:47 anteaya joined #salt
09:47 colon_D joined #salt
09:48 Eliz_ joined #salt
09:48 izibi_ joined #salt
09:48 imanc_ joined #salt
09:48 jeblair_ joined #salt
09:49 blast_hardcheese joined #salt
09:50 madduck_ joined #salt
09:50 freelock_ joined #salt
09:50 mikepea_ joined #salt
09:50 ZombieTwiglet joined #salt
09:50 mihait_ joined #salt
09:50 creich_ joined #salt
09:50 creich_ hi there
09:50 creich_ i currently try to create a symlink in several homedirectories using salt
09:50 creich_ but i am not sure how to cycle throught the currently installed users
09:50 creich_ i am not able to run throught a list using FOR at all
09:50 creich_ how can i create a list of items?
09:50 goki_ joined #salt
09:50 pviktori_ joined #salt
09:50 ikanobori_ joined #salt
09:51 analogbyte joined #salt
09:51 spader joined #salt
09:51 skullone joined #salt
09:51 modafinil_ joined #salt
09:51 pled76 joined #salt
09:52 joehh joined #salt
09:52 dcolish joined #salt
09:52 sashka_u1 joined #salt
09:52 rcsheets joined #salt
09:52 agj joined #salt
09:52 akitada_ joined #salt
09:52 alainv joined #salt
09:53 eliasp creich_: you could deploy a custom grain to your minions which returns you a list of the users using "getent passwd" or a similar Python builtin
09:54 eliasp creich_: depending on your environment, you should be aware that not all users might have a local home directory, but could also have a home directory on NFS which is mounted by PAM on login
09:54 EugeneKay joined #salt
09:54 CaptTofu__ joined #salt
09:54 creich_ so i have to use grains..
09:55 vbabiy_ joined #salt
09:55 akoumjian_ joined #salt
09:55 creich_ ok i thought it might be possible to e.g. use a FOR statement and check if /home/ hase subdirs and place the symlink in every subdir
09:55 nlb_ joined #salt
09:55 eliasp creich_: yes, you can't do this in Jinja, as the Jinja template is rendered on the Master before it is shipped to the Minion
09:55 jcristau_ joined #salt
09:55 mikjaer_ joined #salt
09:55 creich_ ah ok, that makes sence ;)
09:55 platforms joined #salt
09:55 _pipecork joined #salt
09:56 eliasp creich_: using /home/* to iterate over is fragile… it will fail for users with another basedir for their homedir… use "getent passwd" to retrieve the list of users + their attributes
09:56 oznt joined #salt
09:56 creich_ might it be possible to place a statement for the symlink creating as (sub)-part of the user definition?
09:56 malinoff eliasp, jinja was not rendered on the master in 0.17
09:56 malinoff i guess its not in 2014
09:57 malinoff so you can use stuff like {% __salt__['cmd.run']('getent passwd') %}
09:57 bofhit joined #salt
09:57 malinoff directly in states
09:57 agend joined #salt
09:57 Ahlee_ joined #salt
09:57 tru_tru_ joined #salt
09:57 maber joined #salt
09:57 creich_ getent passwd lists a lot if users
09:57 balltongu_ joined #salt
09:58 eliasp creich_: that's different… cmd.run will be issued by the Master, executed by the Minion, returned to the Master to be rendered into Jinja
09:58 eliasp eh, sorry… malinoff was meant
09:58 cyberjam3s joined #salt
09:58 malinoff eliasp, you're wrong. cmd.run will be executed locally
09:58 whiteinge_ joined #salt
09:58 arapaho_ joined #salt
09:58 zach_ joined #salt
09:58 eliasp sure, the command executed by cmd.run is executed locally by the Minion, but only after the Minion was told by the Master to do so
09:59 scarcry_ joined #salt
09:59 oznt hi guys, I am seeing a very weird behaviour with salt. I have a virtual machine machine called data2 (the machine was cloned from core1), when I start salt minion, and try to add it's key on the master it registers as core1, even though I updated the hostname! can someone help me find why is it registering with the wrong name?
09:59 kwmiebach____ joined #salt
09:59 arnoldB_ joined #salt
09:59 syst3mw0rm_ joined #salt
09:59 robawt1 joined #salt
09:59 Sp00n oznt, is the id in the minion file still set to core1?
09:59 eliasp oznt: check /etc/salt/minion for a minion_id file
09:59 bigmston1 joined #salt
10:00 mapet_ joined #salt
10:00 andrej_ joined #salt
10:00 bezaban_ joined #salt
10:00 creich_ eliasp, malinoff thx for your help! i'll try some stuff according to your information :)
10:00 malinoff nope, cmd.run does not issue the master to run that command
10:00 gamingrobot_ joined #salt
10:00 malinoff eliasp, __salt__ is just a dictionary contains python functions
10:00 malinoff all these functions does not know anything about 'salt master'
10:01 scooby2_ joined #salt
10:01 mariusv_ joined #salt
10:01 malinoff you have 3 ways to issue the master from a minion: peer interface, event system and salt mine
10:01 malinoff that's all
10:01 oznt eliasp: thanks! I didn't know about this file before
10:02 malinoff eliasp, please, don't mislead
10:02 eliasp malinoff: not entirely true… how could the dictionary predict the outcome of salt[cmd.run](foo bar)
10:02 codekoala-simann joined #salt
10:02 codekoala-simann joined #salt
10:03 rreboto joined #salt
10:03 jmccree_ joined #salt
10:03 Ryan_Lane joined #salt
10:03 georgemarshall_ joined #salt
10:04 sadbox_ joined #salt
10:04 xnxs joined #salt
10:04 creich_ eliasp, malinoff, now i get this error: Rendering SLS "base:bash.bashcolors" failed: Jinja syntax error: Encountered unknown tag '__salt__'.; line 20
10:05 creich_ so the __salt__ seems not to work in my sls file
10:05 eliasp creich_: could you nopaste the SLS?
10:05 creich_ is it new? maybe my salt-version is to old?
10:05 creich_ sure
10:06 creich_ eliasp, http://nopaste.info/5a9f029849.html
10:06 creich_ this is just a test
10:06 creich_ maybe my whole idea is wrong
10:07 creich_ i just want to create a symlink to the color-file for every user in it's home
10:07 eliasp creich_: use {% salt['cmd.run']('getent passwd') %}
10:07 eliasp creich_: but this will still fail, as it doesn't produce valid YAML
10:08 creich_ Rendering SLS "base:bash.bashcolors" failed: Jinja syntax error: Encountered unknown tag 'salt'.; line 20
10:08 eliasp creich_: so you'd have to place the output of the cmd.run in a variable {% set userlist = salt['cmd.run']('getent passwd') %}
10:08 jasonrm_ joined #salt
10:08 eliasp and then process the result in whatever way you want to generate YAML SLS from it
10:08 creich_ yeah i tried that before but wanted to have a simple line that at least work
10:08 eliasp creich_: which version are you on? 2014.1.x?
10:09 creich_ ah ok you're right
10:09 creich_ now the command works
10:09 eliasp ok, great
10:09 creich_ if i put it into a variable
10:09 creich_ great
10:09 creich_ so about the way i try to achive this... is there a better way to place the symlink in every home-dir?
10:10 creich_ or is it more or less the way to go?
10:10 creich_ i know the code looks messy right now, but i'm testing for hours and "collected" some garbage allready ;)
10:10 eliasp doesn't sound that wrong…
10:10 InAnimaTe joined #salt
10:10 [vaelen] joined #salt
10:10 vukcrni joined #salt
10:11 Jahkeup joined #salt
10:11 creich_ ok thanks for the help :)
10:11 eliasp it sure isn't completely generic (NFS homes etc), but might work perfectly for your local setup
10:11 Gareth joined #salt
10:11 eliasp yw
10:11 mfournier joined #salt
10:11 buhman joined #salt
10:11 kaictl joined #salt
10:12 oz_akan joined #salt
10:12 TTimo joined #salt
10:13 creich_ is there any way to easily test the statement without calling state.highstate?
10:16 snuffeluffegus joined #salt
10:17 bhosmer joined #salt
10:21 ianmcshane joined #salt
10:21 eliasp creich_: salt your-minion state.sls bash.bashcolors test=True
10:22 creich_ ahh great
10:22 creich_ thx again :D
10:23 mechanicalduck_ joined #salt
10:23 bhosmer joined #salt
10:25 tmmt_ joined #salt
10:26 packeteer joined #salt
10:26 rogst_ joined #salt
10:26 alaskabear joined #salt
10:26 d3vz3r0 joined #salt
10:26 nomad_fr_ joined #salt
10:26 jacksontj_ joined #salt
10:28 rlarkin|2 joined #salt
10:30 tru_tru joined #salt
10:30 cym3try_ joined #salt
10:31 eofs_ joined #salt
10:31 martinp joined #salt
10:32 zach joined #salt
10:32 creich_ can i just output the content of the variable to the log somehow?
10:33 djaykay_ joined #salt
10:33 mariusv_ joined #salt
10:33 crashmag_ joined #salt
10:33 ThomasJ|o joined #salt
10:34 dcolish_ joined #salt
10:35 Guest74423 joined #salt
10:37 ajw0100_ joined #salt
10:38 tmh1999 joined #salt
10:39 eliasp creich_: try "salt your-minion state.show_low_sls bash.bashcolors" to get the rendered SLS
10:39 calvinh joined #salt
10:39 eliasp creich_: see also similar functions listed here: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.show_low_sls
10:40 eliasp creich_: I don't know of a better way to "inspect" Jinja/SLS while being rendered
10:40 creich_ how?
10:40 creich_ ah don't ;)
10:40 TheThing joined #salt
10:40 creich_ ok thx
10:41 creich_ uuh it looks hard to format the list in a correct manner
10:41 creich_ i allready have some unix-cmd line that brings a pretty good list
10:41 davidone joined #salt
10:41 creich_ but the salt puts some "local:" before the output :(
10:43 ml_11 joined #salt
10:43 eliasp creich_: you could some Python code to return you a nice dictionary of all the users and their attributes into a custom grain (http://docs.saltstack.com/en/latest/topics/targeting/grains.html#writing-grains) and then simply access those data in your Jinja using something like {% for user in grain['users'] %} .... {{ user.homedir }} ... {% endfor %}
10:44 creich_ hm sounds good :)
10:46 mariusv joined #salt
10:46 mariusv joined #salt
10:47 mechanicalduck joined #salt
10:48 msil joined #salt
10:48 sctsang_ joined #salt
10:49 londo__ joined #salt
10:49 Kelsar_ joined #salt
10:49 Hazelesque_ joined #salt
10:49 fxdgear_ joined #salt
10:50 gamingrobot_ joined #salt
10:50 rlarkin joined #salt
10:50 cyberjames joined #salt
10:51 kuffs_ joined #salt
10:51 ShibMcNe_ joined #salt
10:51 CryptoMer joined #salt
10:51 the_lale1u joined #salt
10:51 imanc__ joined #salt
10:51 mschiff_ joined #salt
10:51 mschiff_ joined #salt
10:51 mariusv_ joined #salt
10:52 jalaziz joined #salt
10:52 schristensen_ joined #salt
10:52 ianmcshane joined #salt
10:53 akl_ joined #salt
10:53 akl_ joined #salt
10:54 _ikke__ joined #salt
10:54 _blackjid joined #salt
10:54 ujjain2 joined #salt
10:54 flebel_ joined #salt
10:55 JPau1 joined #salt
10:55 hopthrisC joined #salt
10:56 freeminds joined #salt
10:56 MatthewsFace joined #salt
10:56 urtokk joined #salt
10:56 kedo39 joined #salt
10:56 martinp joined #salt
10:56 cb joined #salt
10:56 nahamu joined #salt
10:57 synical joined #salt
10:57 pass_by_value joined #salt
10:57 xsteadfastx joined #salt
10:57 APLU joined #salt
10:57 txmoose joined #salt
10:57 Ssquidly joined #salt
10:57 munhitsu_ joined #salt
10:57 bezaban joined #salt
10:57 nlb joined #salt
10:57 repl1cant joined #salt
10:58 Sauvin joined #salt
10:58 pled76 joined #salt
10:58 TheThing joined #salt
10:58 ggoZ joined #salt
11:00 ianmcshane joined #salt
11:00 ajprog_laptop joined #salt
11:01 marcinkuzminski joined #salt
11:01 CaptTofu__ joined #salt
11:02 mariusv__ joined #salt
11:02 Nazca__ joined #salt
11:03 mfournier joined #salt
11:03 basepi joined #salt
11:03 codekobe__ joined #salt
11:03 octarine joined #salt
11:03 kedo39 joined #salt
11:03 davet joined #salt
11:04 bofhit joined #salt
11:04 urtokk joined #salt
11:04 freeminds joined #salt
11:04 xsteadfastx joined #salt
11:06 Sauvin joined #salt
11:06 aberdine left #salt
11:07 mariusv__ joined #salt
11:08 Nazzy joined #salt
11:08 Nazzy joined #salt
11:10 darkelda joined #salt
11:12 oz_akan joined #salt
11:19 acabrera_ joined #salt
11:20 xmj joined #salt
11:20 msil joined #salt
11:20 martoss joined #salt
11:21 _sifusam_ joined #salt
11:21 emostar_ joined #salt
11:21 rglen__ joined #salt
11:22 mortis joined #salt
11:22 nyov joined #salt
11:22 pfallenop joined #salt
11:23 Sauvin joined #salt
11:23 seblu joined #salt
11:23 denstark joined #salt
11:23 Bosch[] joined #salt
11:23 eightyeight joined #salt
11:23 goodwill joined #salt
11:24 Yoda-BZH joined #salt
11:24 Yoda-BZH joined #salt
11:25 bhosmer_ joined #salt
11:26 xmj_ joined #salt
11:27 calvinh_ joined #salt
11:27 mike25de1 joined #salt
11:27 _mrtango_ joined #salt
11:27 yetAnotherZero joined #salt
11:27 TaiSHi joined #salt
11:28 ajolo_ joined #salt
11:29 msciciel joined #salt
11:29 davroman1ak joined #salt
11:29 t0rrant_ joined #salt
11:29 bensons joined #salt
11:29 steveoliver joined #salt
11:30 gfa_ joined #salt
11:30 tedski- joined #salt
11:30 phx__ joined #salt
11:30 tru_tru_ joined #salt
11:30 bdf_ joined #salt
11:30 alaskabe1r joined #salt
11:30 Damon__ joined #salt
11:30 avn_ joined #salt
11:32 intellix joined #salt
11:32 pled76 joined #salt
11:33 manytrees_ joined #salt
11:33 drags joined #salt
11:33 Bryanstein joined #salt
11:33 _|d0m|_ joined #salt
11:33 drawks joined #salt
11:33 lazybear joined #salt
11:33 lahwran joined #salt
11:34 nlb joined #salt
11:34 hellerbarde joined #salt
11:34 dcmorton joined #salt
11:35 ecdhe joined #salt
11:36 eofs joined #salt
11:37 ajolo_ joined #salt
11:39 dh joined #salt
11:39 lcavassa joined #salt
11:45 Flusher joined #salt
11:45 timbermaniac joined #salt
11:45 xmj joined #salt
11:48 diegows joined #salt
11:52 ndrei joined #salt
11:53 TTimo joined #salt
11:54 kbyrne joined #salt
11:55 phx joined #salt
11:55 pled76 joined #salt
11:56 calvinh joined #salt
11:56 redondos joined #salt
11:58 calvinh_ joined #salt
11:58 darkelda joined #salt
12:01 calvinh_ joined #salt
12:01 TTimo joined #salt
12:02 kbyrne joined #salt
12:06 ianmcshane joined #salt
12:08 arapaho joined #salt
12:09 TheThing joined #salt
12:09 arapaho joined #salt
12:11 Doqnach`` joined #salt
12:11 msciciel1 joined #salt
12:12 martoss hey folks,
12:12 clone1018 joined #salt
12:13 martoss how can I extend / overwrite a  file.recurse by a git.latest? I have a formula that contains a file.recurse for a directory which I wanna checkout from git instead.
12:13 steveoliver_ joined #salt
12:14 rawzone joined #salt
12:15 smkelly joined #salt
12:15 calvinh joined #salt
12:15 t0rrant joined #salt
12:15 brandon___ joined #salt
12:17 akl joined #salt
12:17 akl joined #salt
12:18 jslatts joined #salt
12:20 Flusher joined #salt
12:23 Outlander joined #salt
12:23 TaiSHi joined #salt
12:24 jaimed joined #salt
12:25 ede joined #salt
12:27 pled76 joined #salt
12:28 joehh joined #salt
12:28 pled76 joined #salt
12:28 mike25de1 hi guys - i have a state where i want to copy 7 files from the master salt to the minion.   is there a way to copy all 7 at once.. without creating file.manage for each of them?
12:28 mike25de1 all files are in the same directory
12:28 catpiggest joined #salt
12:30 kbyrne joined #salt
12:30 dccc_ joined #salt
12:30 Setsuna666 joined #salt
12:36 babilen mike25de1: file.recurse could be used
12:37 bhosmer_ joined #salt
12:38 toastedpenguin joined #salt
12:40 ianmcshane joined #salt
12:41 pled76 joined #salt
12:41 aw joined #salt
12:43 pled76_ joined #salt
12:44 ianmcshane joined #salt
12:48 jkaye joined #salt
12:49 ml_1 joined #salt
12:51 mike25de1 babilen:  thank you ,... i just discovered that :)
12:52 Flusher joined #salt
12:53 pled76 joined #salt
12:53 ericof joined #salt
12:56 ndrei joined #salt
12:58 tempspace mornin'
12:58 phx left #salt
13:06 mechanicalduck_ joined #salt
13:06 ianmcshane joined #salt
13:07 wangofett 'morning to you too, tempspace
13:09 jkaye joined #salt
13:09 FeatherKing joined #salt
13:10 ianmcshane joined #salt
13:12 dude051 joined #salt
13:14 nitti joined #salt
13:17 workingcats joined #salt
13:17 TheThing joined #salt
13:19 TTimo joined #salt
13:21 nitti joined #salt
13:23 mike25de1 hi again:  is there a way to check that an user exists.. and run some states depending on that ?
13:23 cpowell joined #salt
13:24 oz_akan joined #salt
13:27 racooper joined #salt
13:30 mapu joined #salt
13:31 pled76 joined #salt
13:33 smcquay joined #salt
13:35 wangofett mike25de1: do you want to just check if a user exists, or make sure they're there (or gone)?
13:36 TTimo joined #salt
13:40 mike25de1 just to check wangofet .... if they don't exist... i can then run the state to create them
13:42 nitti joined #salt
13:43 pled76 joined #salt
13:45 kermit joined #salt
13:45 tmh1999 joined #salt
13:47 ianmcshane joined #salt
13:47 quickdry21 joined #salt
13:49 ndrei joined #salt
13:51 KevinMGranger mike25de1: That's not quite a good way of thinking about it. Salt is meant to be declarative. You should describe what you want your end result to be in your states.
13:52 ianmcshane joined #salt
13:52 Neocid joined #salt
13:53 favadi mike25de1: just user.present, salt will does the magic for you
13:53 pled76 joined #salt
13:53 Nss how I can create a state that transfers a pillar file into all the minions?
13:54 favadi Nss: for what?
13:55 Nss I'm trying to create new users on the minions and I want to trasfer  ssh keys
13:55 mike25de1 KevinMGranger: yeah ... i agree with you
13:55 mike25de1 favadi: thank you!!
13:56 mike25de1 ah ... that's not what i wanted favadi ...
13:57 mike25de1 my create user state fails.. because the user... after is created .. gets some groups from LDAP... but the user is a local user.  so the next time my create user state runs... it fails withi this error: These values could not be changed: {'groups': []}
13:58 mike25de1 i was thinking of adding some jinja to check if the user already exist and to bypass the declaration to create that user... does this make any sense?
13:59 tk75 joined #salt
14:03 tmh1999 joined #salt
14:03 KevinMGranger are there some LDAP states? hmm
14:03 kt76 joined #salt
14:05 KevinMGranger it appears there are not
14:05 mike25de1 http://docs.saltstack.com/en/latest/ref/auth/all/salt.auth.ldap.html
14:05 mike25de1 but that is a limited thing
14:05 mike25de1 so my error is ldap related?
14:05 KevinMGranger Having salt ignore the state of groups for a user might be useful. I'd file a ticket, but keep in mind it might be interpreted as a feature request
14:06 _mel_ joined #salt
14:07 KevinMGranger you might consider writing your own state in the meantime, perhaps wrapping http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.useradd.html#module-salt.modules.useradd
14:08 metaphore joined #salt
14:09 icebourg joined #salt
14:09 pled76 joined #salt
14:11 rallytime joined #salt
14:12 ianmcshane joined #salt
14:12 mike25de1 KevinMGranger: thanks... in the meantime... i will find a solution to bypass those definitions if the user exists.
14:14 jalbretsen joined #salt
14:23 peters-tx joined #salt
14:25 dvestal joined #salt
14:26 mike25de1 i am stupid: how can i use : salt.states.file.exists  to check if a file exists... and if NOT run... other definitions that are in the same state file..
14:28 pled76 joined #salt
14:28 bezeee joined #salt
14:31 TaiSHi Hi all
14:31 TaiSHi This is weird, it has been happening a log: http://dpaste.com/04ZXRNX
14:31 TaiSHi I receive no answer from minions
14:33 cb apt is already running as it seems
14:34 SheetiS joined #salt
14:35 conan_the_destro joined #salt
14:35 murrdoc joined #salt
14:36 TaiSHi Yeah but why the first time it just didn't output anything
14:37 TaiSHi I launched it a second time, and it seems the previous cmd.run WAS running
14:37 jergerber joined #salt
14:37 laxity joined #salt
14:37 martoss joined #salt
14:38 cb TaiSHi: How long did you wait?
14:38 pled76_ joined #salt
14:38 TaiSHi I didn't, the apt isn't the point. It has been happening randomly with everything
14:39 rojem joined #salt
14:40 TaiSHi cmd.run, states, etc
14:41 jeremmm joined #salt
14:42 TheThing joined #salt
14:42 icebourg_ joined #salt
14:43 icebourg_ joined #salt
14:43 rallytime joined #salt
14:43 Heggan joined #salt
14:44 Neocid joined #salt
14:44 davromaniak joined #salt
14:44 davromaniak joined #salt
14:51 Nss favadi, I want to trasfer the ssh keys for the new users
14:52 TOoSmOotH Nss: https://github.com/TOoSmOotH/onionsalt/blob/master/opt/onionsalt/salt/users/init.sls
14:52 wendall911 joined #salt
14:53 TOoSmOotH That is the hack I use
14:54 Ozack1 joined #salt
14:55 Nss TOoSmOotH, thanks but for what I can see you still picking up the files from the file_roots: directory, am I wrong?
14:55 pled76 joined #salt
14:56 dvestal joined #salt
14:56 TOoSmOotH I put the keys in /srv/salt/users/keys/
14:56 Nss that's counterintuitive from my point of view, we are defining the list of users in a pillar and it make much more sense to me to have the keys close to the pillar
14:56 TOoSmOotH as <username>.id_rsa.pub
14:57 TOoSmOotH the suers are in the pillar
14:57 TOoSmOotH you want the key in the pillar?
14:58 Nss yes I would like to have the definition of the users and the keys in the same directory
14:58 TOoSmOotH in your pillar create a - key: <key>
14:59 Nss yes, right now I'm using the users formulas that uses exactly that approach
14:59 TOoSmOotH then you will have to somehow create the file based on that value
15:00 TOoSmOotH so a file.managed with a jinja template
15:00 favadi why not use a simple file.append?
15:01 floWenoL is there a way to use file.blockreplace to consume from files instead of sls?
15:01 babilen Do I have access to map.jinja in a configuration file that I render with jinja? (map.jinja as used in various formula)
15:01 kaptk2 joined #salt
15:01 TOoSmOotH that is assuming the .ssh directory is already there
15:02 TOoSmOotH and the file has been created
15:02 to_json joined #salt
15:03 TOoSmOotH personal preference really.. you can do all that kungfu in the pillar or simple create a file with a key already in it
15:03 TOoSmOotH the keys change
15:03 TOoSmOotH so for me its always been easier to manage keys on the actual key files not pasting it into a pillar
15:05 TOoSmOotH they send me an email with their pub key and I just copy it to that directory.. done
15:07 ndrei joined #salt
15:08 Setsuna666 joined #salt
15:09 to_json joined #salt
15:10 PsionTheory joined #salt
15:12 berserk joined #salt
15:13 vejdmn joined #salt
15:13 berserk joined #salt
15:14 vejdmn joined #salt
15:15 Nss ok different question, how can I get the value of pillar_roots from a state?
15:16 mike25de1 can i do a salt['cmd.run']  within a state?
15:18 Nss mike25de1, http://docs.saltstack.com/en/latest/ref/states/vars.html
15:18 pled76 joined #salt
15:19 mike25de1 Nss: thanks buddy
15:19 floWenoL is there a way to use file.blockreplace to provide content from files instead of in the sls
15:24 SheetiS floWenoL: you could do like this: {% import_text "completeworksofshakespeare.txt" as poems %} and then use the imported variable as your "content"
15:25 floWenoL hmmm i will try
15:25 floWenoL thanks
15:26 home_rlarkin joined #salt
15:26 SheetiS http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.jinja.html is where I took the import_text from.
15:26 SheetiS in case you need to see some context.
15:26 floWenoL SheetiS: ty
15:28 yomilk joined #salt
15:30 home_rlarkin http://paste.debian.net/plain/118198 <--my custom grains in _grains/role.  saltutil.sync_grains returns 'grains.role' but nothing changes on the minion. I've been two days on this and could use a clue.
15:31 home_rlarkin I expected that /etc/salt/minion would be created ?
15:31 berserk_ joined #salt
15:32 SheetiS home_rlarkin: is there an /etc/salt/grains after?
15:32 berserk__ joined #salt
15:33 SheetiS I don't handle my custom grains in that fashion, but rather store them in /etc/salt/grains, which is what I _thought_ that method also did.
15:33 berserk__ joined #salt
15:34 intellix joined #salt
15:34 home_rlarkin oh, sorry, I meant /etc/salt/grains.  No, it doesn't appear.  When I do it manually with grains.setval that file does get created
15:35 che-arne joined #salt
15:36 home_rlarkin I followed the example at the bottom of http://docs.saltstack.com/en/latest/topics/targeting/grains.html , and the return is a dic ... my google searches aren't returning any new info
15:36 conan_the_destro joined #salt
15:36 ghartz joined #salt
15:37 martoss joined #salt
15:39 tligda joined #salt
15:39 yomilk joined #salt
15:43 toastedpenguin is there a list of what can be defined in a salt-cloud profile file when deploying in a VPC https://docs.saltstack.com/en/latest/topics/cloud/aws.html#launching-instances-into-a-vpc ?
15:43 toastedpenguin want to specify ip address
15:45 home_rlarkin I just did that yesterday toastedpenquin.  looking for the example....
15:46 toastedpenguin home_rlarkin: sweet, thx!
15:46 analogbyte joined #salt
15:47 sectionme joined #salt
15:47 home_rlarkin http://docs.saltstack.com/en/latest/topics/cloud/aws.html  at 15.5.4.7. Cloud Profiles
15:47 home_rlarkin network_interfaces:  <- anything in that stanza must not be also defined anywhere else.
15:48 home_rlarkin more detail here: https://github.com/saltstack/salt/issues/10816
15:52 toastedpenguin home_rlarkin:  perfect! thx again!
15:53 home_rlarkin sure
15:53 rojem is there a good practice for using salt to put a group of different servers into a maintenance mode state that would be different from its normal running state?
15:53 Sauvin joined #salt
15:54 toastedpenguin ah you can specify AZ.....
15:54 home_rlarkin rojem: ask a different question?  that can mean so many things...
15:54 bezeee joined #salt
15:55 KennethWilke joined #salt
15:55 rojem home_rlarkin I want multiple states defined for a group of servers, a production state where salt ensures everything is up and running, and then a maintenance state that would shut down services, perhaps execute some scripts to put them into downtime in our moniotirng software, etc
15:57 rojem home_rlarkin: one such action would be removing web servers from our load balancer for example
15:57 rojem home_rlarkin where the production state would ensure that they are actually in the load balancer
15:58 berserk joined #salt
15:58 bhosmer_ joined #salt
15:58 home_rlarkin ah.  we're just using boto directly for that so far.
15:58 home_rlarkin elbadmin
15:59 FeatherKing how stable is the windows minion in 2014.1.7 ? seems that it just bluescreened a few production minions of mine
15:59 rojem that for AWS only?
15:59 robawt joined #salt
15:59 home_rlarkin rojem: I think so , yes.
16:00 rojem we're all self hosted, no public cloud
16:00 SheetiS rojem: If you wanted to manage it in salt, you could make a state or orchestration that you only called for maintenance and applied directly to just the minions you wanted.
16:00 rojem healthcare data...
16:00 SheetiS so your maintentance.sls or whatever wouldn't occur in base or any other environment so it is never triggered on a highstate
16:01 rojem SheetiS and the minions wouldn't be put back into highstate unless I manually triggered it?
16:01 SheetiS Well you could work logic into that
16:02 SheetiS What I'd think about doing is a custom grain called maintenance
16:02 SheetiS and have maintenance.sls apply it to the machines
16:02 SheetiS then in my top.sls do something like this:
16:03 SheetiS (assuming that you were to make the grain "maintenance: True")
16:03 home_rlarkin I think I'm going to give up on custom grains in a _grains folder :/  maybe a reactor to minion_start that runs ( a dynamically generated ) salt '*' grains.setvals "{'key1': 'val1', 'key2': 'val2'}"
16:03 rojem SheetiS I see where you're going, that's not a bad idea
16:03 SheetiS home_rlarkin: I can show you what I do for custom grains in a moment.  might not meet your needs, but it worked well for me.
16:04 home_rlarkin Nice. I have struggled for a while now.  I have a pastebin above
16:06 SheetiS rojem: I'd basically wrap the stuff under base: in some jinja that looked like this {% if not grains.get('maintenance', False)  %} .... {% endif %}
16:07 SheetiS In the example there, False is the default in case maintenance grain is not defined.
16:07 pled76 joined #salt
16:07 rojem ya, that looks like it would work well
16:08 madduck skullone: let's hear the environments use case!
16:09 SheetiS home_rlarkin: I have a pillar that defines the custom grains I want applied to each machine.  It groups them into categories for logical assignment groupings.  Let me take a quick look at your pastebin and see if it would apply well
16:10 home_rlarkin thanks.  btw, the double quotes are me flailing.  I tried single and no quotes also.
16:11 dvestal joined #salt
16:11 berserk joined #salt
16:12 SheetiS home_rlarkin: would it be acceptable to have the grains end up in /etc/salt/grains on the minion, or do you need a custom module for your use case?
16:12 home_rlarkin I actually would like them in /etc/salt/grains
16:13 home_rlarkin I considered putting them in pillar, but we're migrating from scripts to salt-cloud and we have a lot of sls that looks for the grain...it would be a big refator.
16:13 home_rlarkin *refactor
16:15 SheetiS https://github.com/rfairburn/salt-grains-formula  <== this is what I did.  It actually does put everything in a pillar and then when you apply the 'grains' sls, it applies the proper roles/categories/etc to the proper minions (and has a 'default' list that get applied if a minion id is not defined in the list).
16:15 skullone madduck: so basically, we assign all nodes the same states, based on regexes, and the environments are git branches of configs
16:15 skullone we do this today with puppet, and it works alright
16:16 skullone that way, we dont have to modify node definitions ever
16:16 SheetiS home_rlarkin: http://devop.ninja/configuration%20management/2014/08/25/the-grains-conundrum/ describes a little bit better what it does...
16:16 home_rlarkin Thx SheetiS
16:16 SheetiS I just make sure that the grains.sls is applied before the first highstate
16:16 skullone madduck: so, with reclass, we'd have to do like dev-exim state, qa-exim, etc, instead of the same state
16:16 troyready joined #salt
16:16 SheetiS so that it knows whta to do with it.
16:17 madduck skullone: environments have been implemented in Git…
16:17 home_rlarkin I suppose even in highstate I could make sure that state is first.
16:17 SheetiS as long as it completed before the others, probably.
16:18 SheetiS I applied it as part of an orchestration, but I've been thinking of trying it the way you mention (give it order 1 or whateveR)
16:18 skullone madduck: which branch?
16:18 madduck skullone: HEAD
16:19 SheetiS the problem is that the top.sls is what uses the grains in my use case
16:19 SheetiS so even a reorder wouldn't always work for me
16:20 home_rlarkin ah. I was doing that , the 'role' example.
16:22 home_rlarkin but now I'm using private environments extensively and everything has to be defined in the config anyway.
16:23 pled76 joined #salt
16:24 ajolo joined #salt
16:25 mechanicalduck joined #salt
16:26 dccc_ joined #salt
16:27 SheetiS I just found it hard to visualize everything and how it applied to a minion in a complex top.sls, so I went with the roles with categories and how I applied those grains.  I can look at the /etc/salt/grains on any minion (or get it via grains.get/grains.items/etc) and know all the states and groupings at a glance.
16:28 pled76_ joined #salt
16:29 bezeee joined #salt
16:31 home_rlarkin I have several custom grains, but only one in top, the primary app name.  all the other grains are 'consumed' inside the environment.
16:31 home_rlarkin the 'other' grains are fairly dynamc...versions etc.
16:32 SheetiS if that one doesn't change, as long as the grains were initially applied before the first highstate, I think using the order would work for everything else.
16:32 thayne joined #salt
16:33 home_rlarkin yeah. I think so too.
16:34 SheetiS and that could be applied on minion start or after the key is first accepted via the reactor.  There are probably other ways as well.
16:34 SheetiS anyhow, I'm about to step out for lunch, so good luck home_rlarkin!
16:34 home_rlarkin I'm think a reactor for minion_start
16:34 home_rlarkin thx again
16:35 SheetiS glad to help when I can.
16:36 berserk_ joined #salt
16:43 pmcg joined #salt
16:43 KyleG joined #salt
16:43 KyleG joined #salt
16:46 miles32 joined #salt
16:46 miles32 I think one of the most infuriating things about salt is when the call you made returns with no output
16:47 sectionme joined #salt
16:48 murrdoc joined #salt
16:48 nitti_ joined #salt
16:50 dstokes morning guys
16:51 vbabiy joined #salt
16:51 v0rtex miles32: for long running jobs the CLI will timeout and exit before a response returns. You can check on the status easily enough: "salt-run jobs.list_jobs" and "salt-run jobs.lookup_jid <ID>"
16:53 v0rtex miles32: I have had some weird instances where nothing is returned for a short-running job and even where it seems nothing ever gets returned from the minion - usually this seems to be some sort of communication issue between the minion and the master though
16:56 ndrei joined #salt
16:58 nitti joined #salt
16:59 dstokes is there any way to control the execution order of states in a topfile? i'm using the same deployment script for several projects, but some have system dependencies and some don't.
17:01 wangofett dstokes: have you looked at the depends thing?
17:02 dstokes requisites? yeah, afaik you can't define one state's dependency on another
17:02 dstokes state === sls in this case
17:02 wangofett there's also includes...
17:02 wangofett I'm not familiar with that part at all
17:02 wangofett I've been using salt for less than a week ;)
17:03 dstokes yeah. so the problem is if i include my state in my deploy script _all_ the apps will run that state file.
17:03 dstokes only one of them needs it. some might argue that a pillar flag would be the way to handle that case, but seems redundant when i'm already defining host dependencies in the topfile
17:04 miles32 I'm aware of the timeout
17:04 Setsuna666 joined #salt
17:04 wangofett you could include based on the pillar...
17:04 miles32 it's the short running ones that complete successfully without returning that annoy me
17:04 wangofett not sure if there's a better way
17:06 kermit joined #salt
17:09 chrisjones joined #salt
17:15 eliasp basepi: ping… still got a keybase.io invite?
17:15 basepi eliasp: indeed I do.  Nine, in fact.  =P
17:15 basepi pm me your e-mail, eh?
17:15 eliasp basepi: awesome, I'll do ;)
17:16 pled76 joined #salt
17:18 schimmy joined #salt
17:18 murrdoc joined #salt
17:20 spookah joined #salt
17:20 martoss joined #salt
17:21 rap424 joined #salt
17:21 shaggy_surfer joined #salt
17:22 zz_chamunks joined #salt
17:25 melinath joined #salt
17:25 Sweetshark joined #salt
17:29 chrisjones joined #salt
17:29 bhosmer_ joined #salt
17:31 schimmy joined #salt
17:33 scoates joined #salt
17:33 thayne joined #salt
17:33 Sweetshark triggering highstate here yields "Comment:   State pip.installed found in sls tb3.master is unavailable" for my fumbling attempt at getting a pip module installed that way. Any hints on how to ensure a pip module installed on a minion are welcome!
17:33 Sweetshark http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.pip.html only has examples for the command line. :/
17:36 UtahDave joined #salt
17:36 v0rtex miles32: ahh yes, very annoying - in my case I have some that do that and the best I could figure is that there must be a bug of some sort causing some random/weird communication issues from the minion
17:37 zz_chamunks joined #salt
17:38 whytewolf Sweetshark: that would be because that is the module and not the state. http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pip_state.html#module-salt.states.pip_state
17:40 murrdoc this module needs a 'manage pip config file'
17:43 dvestal_ joined #salt
17:43 jxqz joined #salt
17:43 sctsang joined #salt
17:43 jxqz holy salt, so many nicks
17:46 Sweetshark whytewolf, murrdoc: thanks a lot, those hints were exactly what I needed -- most appreciated.
17:46 kermit joined #salt
17:48 sectionme joined #salt
17:50 acabrera joined #salt
17:50 pled76 joined #salt
17:51 dvestal joined #salt
17:52 murrdoc :) np
17:55 pled76 joined #salt
18:01 rypeck Can I use file.recurse and leave the executable bit set in files? or should I have another state change the bit?
18:01 Damon joined #salt
18:02 Ahlee_ I have a minion I specifically stopped that matches an ext_pillar, if i salt -t 30 -I pillar on the cli, i get back that the minion did not return in time, however if I run the same through a salt.client.LocalClient() the down minion doesn't show up as not responding
18:02 UtahDave rypeck: file permissions aren't copied down from the master. You can have file.recurse set the perms
18:02 Ahlee_ anybody know how I can get the down minion that matched but didn't respond?
18:02 berserk joined #salt
18:02 UtahDave or have another state set the perm
18:03 rypeck How can I have salt set the executable bit?
18:03 UtahDave rypeck: with the "mode" option
18:04 UtahDave whiteinge: do you know the answer to Ahlee_'s question?
18:05 Ancient Aside from the ability to run commands, what are a couple big selling points salt has over puppet? Benefits rather than just differences of syntax.
18:05 Ahlee_ looks like it's probably -v related on the cli, vs not.  Will poke the LocalClient to see if there's a verbose option
18:06 rypeck UtahDave: ah duh .thanks.
18:06 Setsuna666 UtahDave: Do you know if there is a way to add custom execution modules on the master and use them in pillars ?
18:07 Setsuna666 UtahDave: Or extend/modify a salt built-in one ?
18:08 dude051 joined #salt
18:08 Ahlee_ UtahDave: whiteinge: Looks like I can use cmd_cli and parse the generator it returns as it's present there.  Now to figure out if I can add that to pepper and if cmd_cli is exposed over the rest interface
18:09 Setsuna666 UtahDave: I made a new function for the file execution module and put it in the _modules folder, and synced it with my minions, but should putting it in this folder makes it accessible by the master ?
18:11 Ryan_Lane is there a helper recursive dictionary diff function in salt already?
18:12 Ryan_Lane we have one in the boto_* codebase to provide diffs of changes, but if salt doesn't have one I think it would make sense to move it into utils
18:13 Ryan_Lane we also have one to recursively compare a dict
18:13 Ahlee_ even better, cmd_full_return returns a dict rather than a generator, so I'll roll with that
18:13 drawks anyone have any insight as to why socket.getfqdn is preferred to /etc/hostname for minion id?
18:14 Ahlee_ i guess i'll have to dive into low states though to make this work over the api
18:14 UtahDave Setsuna666: Yeah, you can easily create custom modules
18:14 Ryan_Lane drawks: socket.getfqdn goes through the system resolver, right?
18:14 Ryan_Lane which would hit /etc/hostname first
18:15 Setsuna666 But how can I make them accessible to the master from pillar sls files ?
18:15 whiteinge Ahlee_: sorry, distracted. glad you found it. +1 on adding those extra stuffs to the REST API and pepper. getting the JID in the API return data has been on my wishlist for a while too
18:15 Ryan_Lane and if nothing was defined in /etc/hostname, it would do dns resolution
18:15 Setsuna666 UtahDave: But how can I make them accessible to the master from pillar sls files ? *
18:15 scoates joined #salt
18:15 UtahDave Setsuna666: I'm not sure. So they didn't work on the master?  You might have to drop the module in Salt's main modules directory
18:16 curlingbiathlon joined #salt
18:16 Setsuna666 UtahDave: You mean the /usr/share/pyshared/salt/modules folde r?
18:16 UtahDave Ryan_Lane: I'm not aware of a recursive dict diff in Salt.  It does sound like a good addition to salt.utils
18:17 UtahDave Setsuna666: yeah
18:17 Ryan_Lane ok. we've found it greatly clarifies what's changing in states, rather than old/new :)
18:17 Setsuna666 UtahDave: Already tried that, and it didn't seems to work, but will try again to make sure the problem wasn't on my part
18:17 Setsuna666 UtahDave: Thanks
18:17 Ryan_Lane especially since some of them have giant dicts for old and new
18:17 UtahDave yeah
18:18 drawks Ryan_Lane: are you confusing /etc/hosts and /etc/hostname?
18:18 Ryan_Lane whoops. yes
18:19 bhosmer_ joined #salt
18:19 Ryan_Lane is /etc/hostname standard across distros?
18:19 Ryan_Lane also, for /etc/hostname, don't you only put the short name and not the FQDN?
18:19 drawks the gethostname function is
18:19 Ryan_Lane hm guess not
18:19 drawks yeah, which is sort of my question... why the preference for fqdn as the id
18:21 Eugene FQDN is (hopefully) unique.
18:21 Ryan_Lane well, assume you have two datacenters and you have hosts named: 'myapp-production.a.example.com' and 'myapp-production.b.example.com'
18:21 martoss joined #salt
18:21 Ryan_Lane if you don't use FQDN, you get two hosts trying to use 'myapp-production' as their key names
18:21 Eugene I use the non-qualified hostname, because I know better than to create hostname collision
18:22 dvestal joined #salt
18:22 Ryan_Lane some people use convention based naming and include their domain name in that
18:22 quickdry21 joined #salt
18:22 drawks maybe this is philosophical question. I mean id is required to be unique, but there is no option to easily have a different function called besides socket.getfqdn
18:22 Eugene Eg, "firewall-sitename.mydomain.com", rather than "firewall.sitename.mydomain.com". As a bonus this works with the *.mydomain.com wildcard TLS cert
18:22 murrdoc can you switch the interpreter on a per state basis ?
18:22 Ryan_Lane drawks: it's possible to set the id explicitly in the config
18:23 Ryan_Lane salt has to default to something
18:23 drawks sure, but that is sort fo a chicken/egg problem
18:23 Ahlee_ or in minion_id
18:23 Eugene I generate my id: line as part of a kickstart or cloning script before Salt ever runs
18:23 murrdoc like jinja-yaml to yaml
18:23 pled76 joined #salt
18:23 drawks you have to localize the config to an id. and the localization would typically be handled by config managment
18:23 SheetiS joined #salt
18:23 Ryan_Lane if you're able to change the function via config, you'd be able to write the id into the file, too
18:23 Eugene "who manages the config management"
18:23 UtahDave murrdoc: - template: mako
18:23 murrdoc yes!
18:24 melinath joined #salt
18:24 druonysus joined #salt
18:24 drawks i was going to open a feature request issue, I just wanted to talk it out a bit first.
18:24 Ryan_Lane drawks: yeah. I use salt to generate its own ID
18:24 drawks Ryan_Lane: how do you do that?
18:24 Ryan_Lane salt-call --local state.sls bootstrap
18:25 drawks i mean if this is already possible that'd be ideal.
18:25 Ryan_Lane the bootstrap state sets the id
18:25 Ryan_Lane then you can start the minion
18:25 Ryan_Lane this assumes your bootstrapping has access to a state that can set your id
18:25 dude051 joined #salt
18:26 Ryan_Lane (I actually name by convention and have a custom grain that I use to set my hostname during my first full run, but I also avoid master/minion)
18:27 skarn joined #salt
18:28 Ryan_Lane basepi: hm. I think one of my backports is missing
18:30 rypeck I'm on salt 2014.1.4 - is exclude_pat broken for file.recurse? I'm looking to exclude a sub directory...
18:30 Ryan_Lane basepi: ah, because I forgot to mark it for backport since it wasn't my PR
18:31 * drawks shakes fist in air... the manpage for hostname contains lies!
18:32 drawks "hostname will print the name of the system as returned by the gethostname(2) function." yet an strace of hostrace shows that it actually never calls gethostname, it instead uses the nodename value from the uname syscall
18:34 tligda1 joined #salt
18:36 rypeck nvm I figured it out looks like.
18:36 bmcorser joined #salt
18:37 drawks Ryan_Lane: was your bootstrapping setup informed by the salt docs somewhere? Or is that something you cooked up yourself?
18:38 Ryan_Lane cooked up
18:38 drawks k
18:38 drawks thanks
18:38 Ryan_Lane yw
18:38 Ryan_Lane drawks: http://ryandlane.com/blog/2014/08/26/saltstack-masterless-bootstrapping/ <-- that's my bootstrap process
18:39 chrisjones joined #salt
18:39 roolo joined #salt
18:41 drawks rad thanks
18:42 eagen Ryan_Lane: I just ready your Salt vs. Ansible comparison and it was quite helpful. Thanks.
18:42 Ryan_Lane thanks. glad it could help :)
18:44 chiui joined #salt
18:44 eagen You don't happen to have a github repo of your code do you? It would be nice to see how someone with experience with Salt structures things.
18:45 tligda joined #salt
18:46 londo__ joined #salt
18:48 cpowell_ joined #salt
18:49 sectionme joined #salt
18:51 mapu joined #salt
18:53 pled76 joined #salt
18:53 floWenoL joined #salt
18:54 floWenoL hi, how can I run a file.managed state only if a file does not have some string
18:54 Ryan_Lane floWenoL: you can use onlyif global state argument
18:54 Ryan_Lane floWenoL: http://docs.saltstack.com/en/latest/ref/states/requisites.html
18:55 floWenoL ty
18:55 Ryan_Lane yw
18:55 SheetiS onlyif is 2014.7.0 or later, right?
18:55 Ryan_Lane I didn't think it was...
18:55 Ryan_Lane I could be wrong, though
18:56 Ryan_Lane I thought onlyif and unless have been around a long while
18:56 floWenoL according to docs is
18:56 Ryan_Lane oh :(
18:56 Ryan_Lane yeah, they're global now. I guess they don't work globally in 2014.1
18:57 dagvl joined #salt
18:57 floWenoL damn
18:57 floWenoL going to need an alternative
18:59 bnikolaus any advice on troubleshooting 'sls.highstate' is not available.'
18:59 floWenoL bnikolaus: i remember seeing a bug around that
18:59 floWenoL not sure if fixed
19:00 floWenoL i had a similar problem, but mine was some sort of service start problem right after installation
19:00 floWenoL a reboot fixed my case
19:01 floWenoL Ryan_Lane: any clue when 2014.7 will be out
19:01 Ryan_Lane hopefully soon. it's in release candidate now
19:03 floWenoL may as well start preparing my sls for 2014.7
19:03 floWenoL with rc
19:03 floWenoL im only going into prod in a few weeks
19:05 Ryan_Lane yeah. it's worth using the rc, especially to find any bugs you may hit
19:06 Ryan_Lane to report them before it's released
19:06 SheetiS Sooo many more features in 2014.7 than 2014.1
19:07 catpiggest joined #salt
19:09 tligda1 joined #salt
19:09 metaphore joined #salt
19:11 Gareth morning
19:11 Gareth ..ish.
19:11 Gareth :)
19:12 tligda joined #salt
19:14 floWenoL Gareth here is afternoon P
19:14 floWenoL :P
19:15 XenophonF joined #salt
19:16 Setsuna666 UtahDave: Are you awere of a bug in 2014.7 where, when you use the salt command on a group of minion like so: 'MINION_NAME-*' and only one minion is in the list, no output is returned to the shell, in everything succeed, only when it fails ?
19:16 Setsuna666 if everything succeed*
19:17 XenophonF in file states, can i do something like context: variable: {{ dict }}, and have {{ variable }} in the file refer to the original dict from the SLS?
19:17 wendall911 joined #salt
19:18 miles32 can you reload grains without restarting the salt-minion?
19:18 XenophonF http://paste.debian.net/118239/ is what I have in mind
19:19 Setsuna666 XenophonF: Yeah, it should work
19:19 murrdoc joined #salt
19:19 XenophonF ok thanks
19:19 TheThing joined #salt
19:20 Gareth floWenoL: yeah...here too...
19:20 Setsuna666 XenophonF: Wait, you are talking about passing it through the file context, or updating the dict from the file context ?
19:20 XenophonF my example is from the SLS file
19:21 XenophonF in Pillar, I want to have a dict full of web sites and their corresponding settings called apache:vhosts
19:21 XenophonF so for example, you might have apache:vhosts:www.example.com
19:21 Setsuna666 XenophonF: Ah ok, then yeah, no problem in doing so, the way you did it
19:22 XenophonF and apache:vhosts:test.example.com
19:22 XenophonF and so forth
19:22 XenophonF each of those would be key-value pairs
19:22 XenophonF hence the for loop at the top of the sls
19:22 racooper joined #salt
19:22 pled76 joined #salt
19:23 Setsuna666 XenophonF: Shouldn't be a problem
19:23 XenophonF i want to pass the entire value of apache:vhosts:www.example.com to the jinja template via the context parameter
19:23 XenophonF ah good
19:23 XenophonF i guess the other way i could do it is pass in just the web site name, then do the pillar lookup again inside the config file
19:24 XenophonF i just didn't want to spend a lot of time writing the config file template only to find out that my first approach didn't work :)
19:24 Setsuna666 XenophonF: Yeah, its up to you to define what standard you want to use. Some people prefer to do the lookup in the template file, so it doesn't clutter too much the sls file
19:24 MatthewsFace joined #salt
19:25 XenophonF gotcha
19:31 mhubbard joined #salt
19:33 pled76 joined #salt
19:38 murrdoc UtahDave:  whats the recommended way to do package lists
19:38 murrdoc well not just Utah dave but anyone
19:38 murrdoc do you guys do a pillar and then pass to the sls ?
19:39 UtahDave murrdoc: be sure to use - pkgs in your pkg.installed state
19:41 otter768 joined #salt
19:42 casey joined #salt
19:42 Setsuna666 murrdoc: It depends on what you want to achieve, at the office we use package or packages list, like UtahDave mentionned, directly in the states, since most states are specific for a certain need
19:43 casey joined #salt
19:48 Ryan_Lane joined #salt
19:50 sectionme joined #salt
19:50 pled76 joined #salt
19:52 berserk joined #salt
19:55 joehillen joined #salt
19:57 catpigger joined #salt
20:02 XenophonF murrdoc: i tend to use pkg.installed states with the pkgs parameter
20:02 XenophonF the package list comes from a map.jinja
20:02 XenophonF but i have the pkg.installed state wrapped in a Jinja if statement
20:03 XenophonF so for example on FreeBSD, ntp is part of the base system
20:03 XenophonF so my ntp states look like this:
20:04 XenophonF http://paste.debian.net/118250/
20:05 XenophonF in my state repo that's from ntpd/client.sls
20:05 XenophonF the corresponding map.jinja file looks like:
20:05 XenophonF http://paste.debian.net/118253/
20:07 Setsuna666 joined #salt
20:07 bhosmer_ joined #salt
20:18 pled76_ joined #salt
20:18 floWenoL anyone having issues running the bootstrap to install rc?
20:22 murrdoc so i have this http://paste.debian.net/118262/ XenophonF
20:22 murrdoc but i want to more with the packages
20:22 regisu joined #salt
20:23 murrdoc basically in the json pass in the package name, and params
20:23 Ryan_Lane joined #salt
20:23 murrdoc not sure what the salt-y way is to do
20:23 murrdoc this*
20:23 pled76 joined #salt
20:23 Ryan_Lane murrdoc: to do what?
20:23 murrdoc to cascade package lists for servers
20:24 murrdoc there is a list of common packages to install on all servers
20:24 murrdoc then for each server type there are packages added/removed
20:24 murrdoc also 2014.7 has option to hold packages so i want to pass that too alongwith version
20:25 regisu #mikrotik
20:25 jkaye joined #salt
20:25 regisu left #salt
20:32 Tahm joined #salt
20:39 smcquay_ joined #salt
20:40 longdays joined #salt
20:40 longdays I need to restart a service before run a cmd state. I can certainly just chanin the service restart with the command with bash && but it seems I may be able to tell salt to do this with a requisite. Is this possible?
20:41 manfred yes
20:42 floWenoL i believe bootstrap script has a bug
20:42 floWenoL what idk
20:43 floWenoL yet
20:44 jkaye joined #salt
20:44 longdays I have used a require to ensure a state is running but not sure I know how to use a requisite to force a restart. Is there an example config or should I go back to reading the requisite docs?
20:44 floWenoL hmm latest works rc does not
20:45 floWenoL *  WARN: Upstart does not appear to know about salt-minion
20:45 floWenoL guess not
20:45 nyx joined #salt
20:46 bhosmer_ joined #salt
20:47 KyleG1 joined #salt
20:47 flappp I have a pillar in my pillars top.sls  base: '*': foo and I can access it on the master but not on any other minion why could that be?
20:48 murrdoc joined #salt
20:50 sectionme joined #salt
20:53 pled76 joined #salt
20:54 smcquay joined #salt
20:56 robawt hey cool salt kids
20:57 robawt can i pass "--nogpgcheck" as a kwarg in pkg.installed?
20:57 kermit joined #salt
20:57 spookah joined #salt
20:58 floWenoL robawt sorry but idk
20:58 murrdoc skip_verify
20:59 robawt ty murrdoc
20:59 murrdoc Skip the GPG verification check for the package to be installed
20:59 robawt for those following along at home: don't do this unless you own your repos
20:59 murrdoc http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkg.html :) np
20:59 floWenoL need an alternative for the only_if on a file.managed, cant install rc
21:00 nitti joined #salt
21:05 bryguy joined #salt
21:06 ndrei joined #salt
21:09 otter768 joined #salt
21:09 bezeee joined #salt
21:10 otter768 joined #salt
21:13 KyleG joined #salt
21:13 KyleG joined #salt
21:14 elfixit joined #salt
21:16 nitti joined #salt
21:18 otter768 joined #salt
21:18 otter768 joined #salt
21:20 colinjohnson joined #salt
21:22 kballou joined #salt
21:23 pled76 joined #salt
21:24 viq joined #salt
21:25 wedgie Trying to install an RPM on a CentOS machine. SLS file and error: https://gist.github.com/anonymous/955f8fb7d234f48ac672 Trying to follow this http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkg.html ('sources' section) but not sure where i'm going wrong
21:26 Corey wedgie: Can the minion successfully wget http://yum.postgresql.org/9.3/redhat/rhel-6-x86_64/pgdg-centos93-9.3-1.noarch.rpm ?
21:27 wedgie Corey: yes
21:27 Corey wedgie: Can you get debug output?
21:27 wedgie well, strictly speaking i used curl to test...
21:27 wedgie Corey: on master or minion?
21:29 Corey wedgie: The minion.
21:31 Outlander joined #salt
21:32 koyd Ryan_Lane: are you around?
21:32 wedgie [ERROR   ] Command 'yum -y     install install_postgresql_pgdg' failed with return code: 1
21:32 wedgie that doesn't look right at all
21:32 Ryan_Lane koyd: kind of. about to go into a meeting
21:32 Ryan_Lane what's up?
21:32 koyd am I correct in assuming ryandlane.com is yours?
21:33 koyd I wanted to understand better the reasons for not having a master in your infrastructure, been playing with that tought myself
21:34 Ozack1 joined #salt
21:34 wedgie Corey: so, based on that it would seem that it's ignorning sources and just using the "name" of the section as the package locatoin (not sure what the proper term for the unindented line in the SLS file is)
21:37 whytewolf joined #salt
21:38 Ozack1 joined #salt
21:43 jkaye_ joined #salt
21:45 WarP|work joined #salt
21:45 home_rlarkin ah..just the other day someone posted a link with a cmd.yaml reactor example.  does anyone remember?
21:45 EWDurbin_ joined #salt
21:45 jayne_ joined #salt
21:46 clone1018__ joined #salt
21:46 Ryan_Lane koyd: yep
21:46 Ryan_Lane koyd: we don't want one because it's a single point of failure
21:46 shaggy_surfer joined #salt
21:47 Ryan_Lane and making a master not a single point of failure is a decent amount of work
21:47 Ryan_Lane that said, we'll probably still end up with a master for remote execution
21:47 Ryan_Lane and operate on a model that the remote execution is only there to speed things up
21:47 Ryan_Lane push/pull model
21:48 ShibMcNe joined #salt
21:48 Damon__ joined #salt
21:49 jY- joined #salt
21:50 miles32 for some reason I had to explicitly use mine.send to get the network interfaces off of 60% of the machines, that should have had that provided by a pillar
21:50 drawks ha, nice short term solution to using just the hostname as an id. I just drop a symlink into /etc/salt that links minion_id > ../hostname
21:50 miles32 upon further inspection, the pillars for the machines that failed do report they have the correct mine function
21:50 miles32 but the mine isn't updating?
21:51 z3uS| joined #salt
21:51 nlb_ joined #salt
21:51 JoeHazzers_ joined #salt
21:51 that_guy_ joined #salt
21:51 johngrasty_ joined #salt
21:51 sectionme joined #salt
21:51 renoirb_ joined #salt
21:51 crazysim_ joined #salt
21:51 nitti joined #salt
21:51 goodwill_ joined #salt
21:52 roo9- joined #salt
21:52 codysoyland_ joined #salt
21:53 Ryan_Lane koyd: hm. did I lose you in the net split?
21:54 pled76 joined #salt
21:54 shaggy_surfer joined #salt
21:55 t0rrant joined #salt
21:55 ndrei joined #salt
21:58 Bryanstein joined #salt
22:01 stan_k joined #salt
22:02 miles32 interface is being listed at 0.0.0.0
22:02 N-Mi joined #salt
22:02 N-Mi joined #salt
22:03 martinp joined #salt
22:03 _mel_ joined #salt
22:03 ajprog_laptop joined #salt
22:04 CeBe joined #salt
22:06 SheetiS joined #salt
22:07 bhosmer_ joined #salt
22:07 zach joined #salt
22:11 Damon joined #salt
22:11 thayne joined #salt
22:11 jergerber joined #salt
22:11 joehh joined #salt
22:11 17SAA2F7L joined #salt
22:11 repl1cant joined #salt
22:11 packeteer joined #salt
22:11 ClausA joined #salt
22:11 rhand joined #salt
22:11 notpeter_ joined #salt
22:11 17SAAYPRB joined #salt
22:11 renoirb joined #salt
22:11 MaZ- joined #salt
22:11 Schmidt joined #salt
22:11 17SAASO9W joined #salt
22:11 nliadm joined #salt
22:11 al joined #salt
22:14 mapu joined #salt
22:14 Ryan_Lane does the concurrent argument to state.sls allow me to run multiple salt processes with state.sls at the same time?
22:15 jforest joined #salt
22:15 Setsuna666 joined #salt
22:16 mihait_ joined #salt
22:16 pwiebe_ joined #salt
22:16 natewalck joined #salt
22:16 thunderbolt joined #salt
22:24 koyd Ryan_Lane: sorry, I was away. Reading now :)
22:24 SheetiS Ryan_Lane: I don't have a definitive answer, but looking at https://github.com/saltstack/salt/blob/2c0c11a6abb46bba49d0be328325c222114365e2/salt/modules/state.py and the sls function, I would guess that you could.
22:24 drawks hrm.. this is probably a stupid question, but... if salt-minion is run as non-root it obviously can't do a ton of things, right i.e. install software packages or modify network interfaces etc... is there any way to configure salt-minion to run as root but escalate via sudo or other mechanism before attempting to do things which require root privs?
22:24 koyd Ryan_Lane: ah, are you considering salt-ssh for remote execution, dropping the master completely ?
22:25 XenophonF joined #salt
22:25 Setsuna666 joined #salt
22:25 skarn joined #salt
22:26 SheetiS drawks: It looks like it can be done, but requires a slight modification
22:26 SheetiS https://groups.google.com/forum/#!topic/salt-users/0D0fUjmFxAU was the only thing I'd seen related
22:29 scalability-junk joined #salt
22:29 drawks ah, cool
22:39 ndrei joined #salt
22:42 Guest21595 joined #salt
22:44 sectionme joined #salt
22:49 oncallsucks joined #salt
22:50 aquinas joined #salt
22:54 SheetiS joined #salt
22:55 Corey What would cause a salt run to return things it didn't change as "succeeded" rather than leaving it off of the summary output? They show up green...
22:55 Corey (This environment is really, really backwards)
22:56 TTimo joined #salt
22:57 mhubbard joined #salt
22:59 possibilities joined #salt
23:01 colinjohnson1 joined #salt
23:02 metaphore joined #salt
23:03 bezeee joined #salt
23:04 shaggy_surfer joined #salt
23:16 icebourg joined #salt
23:16 sctsang joined #salt
23:19 conan_the_destro joined #salt
23:24 bezeee joined #salt
23:29 badon_ joined #salt
23:31 ipmb joined #salt
23:33 smcquay joined #salt
23:39 miles32 left #salt
23:44 sectionme joined #salt
23:52 badon_ joined #salt
23:55 bhosmer_ joined #salt
23:55 badon__ joined #salt

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