Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-10-15

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

All times shown according to UTC.

Time Nick Message
00:06 andrej joined #salt
00:07 andrej Trying to set up a minion on a Mac OS/X 10.8.5 machine. Following the homebrew install from http://docs.saltstack.com/en/latest/topics/installation/osx.html
00:07 andrej Doco is somewhat rudiemntary :) ... how do I make it soe the minion starts on reboot?
00:08 pravka joined #salt
00:09 mosen andrej: need a launchdaemon
00:09 mosen andrej: the salt source distribution has a minion and master plist, im not sure if homebrew installs that to /usr/local/Cellar.. checkin
00:10 dude051 joined #salt
00:10 rzzldzzl joined #salt
00:10 pipps joined #salt
00:11 conan_the_destro joined #salt
00:12 fragamus joined #salt
00:13 rzzldzzl joined #salt
00:13 andrej Thanks mosen
00:13 mosen andrej: heh, looks like they dont take care of that in the HomeBrew. But you can use the ones from the saltstack source tree on github
00:13 andrej I'll check it out
00:13 andrej mosen - awesome
00:14 andrej Thanks
00:14 mosen andrej: https://github.com/saltstack/salt/tree/develop/pkg/darwin
00:14 mosen andrej: place into /Library/LaunchDaemons
00:14 mosen just the minion obviously if you only want that to start
00:15 rostam joined #salt
00:18 nitti joined #salt
00:24 mapu joined #salt
00:26 masterkorp Guys any varnish formula ?
00:30 cads joined #salt
00:30 pipps joined #salt
00:33 murrdoc use edgecast (shameless plug)
00:34 masterkorp murrdoc: you work there ?
00:34 murrdoc :D
00:34 masterkorp we were looking at CDNetworks
00:34 murrdoc thats why the plug
00:35 murrdoc edgecast ftw
00:35 masterkorp well i need a quick solution but i am looking at better solutions
00:36 murrdoc not sure about a formula for varnish
00:36 murrdoc but the apache one would be close
00:36 kusams joined #salt
00:37 pravka joined #salt
00:39 obimod joined #salt
00:43 sarlalian joined #salt
00:43 rzzldzzl left #salt
00:44 rzzldzzl joined #salt
00:49 joehh MTecknology: hello
00:55 dude051 joined #salt
01:00 malinoff joined #salt
01:00 ramishra joined #salt
01:00 hasues joined #salt
01:03 higgs001 joined #salt
01:07 jalaziz joined #salt
01:09 aparsons joined #salt
01:11 anotherZero joined #salt
01:11 anotherZero joined #salt
01:11 aparsons_ joined #salt
01:12 Mso150 joined #salt
01:16 ramishra joined #salt
01:18 andrej Hmmm ... the minion seems to ignore the config  in /etc/salt/minion
01:18 andrej the plist file starts with an error code
01:18 andrej I wanted to enable debug, but the log level seems to stay at warning
01:19 swa_work joined #salt
01:28 nitti_ joined #salt
01:28 pravka_ joined #salt
01:29 basepi_ joined #salt
01:29 codekobe_____ joined #salt
01:29 lkannan_ joined #salt
01:30 abele_ joined #salt
01:30 Hazelesque_ joined #salt
01:30 moderation_ joined #salt
01:31 wavis_ joined #salt
01:31 _ikke__ joined #salt
01:32 nebuchad` joined #salt
01:32 mfournier1 joined #salt
01:32 mosen andrej: yeah you would have to change the plist to get a different log level
01:32 mosen andrej: let me take a look at their plist
01:32 adsisco joined #salt
01:32 stewba joined #salt
01:33 akitada___ joined #salt
01:33 rcsheets_ joined #salt
01:33 kwmiebach joined #salt
01:33 ujjain joined #salt
01:33 jgelens joined #salt
01:33 asmaps joined #salt
01:33 kwmiebach joined #salt
01:33 rcsheets_ joined #salt
01:33 mschiff joined #salt
01:33 mschiff joined #salt
01:33 lyddonb joined #salt
01:33 fannet ?
01:33 Jarus joined #salt
01:33 ikanobori_ joined #salt
01:33 whitenite joined #salt
01:33 cb joined #salt
01:33 GrueMaster joined #salt
01:33 juice joined #salt
01:34 jbub joined #salt
01:34 Nazzy joined #salt
01:34 xsteadfastx joined #salt
01:34 crane joined #salt
01:34 Emantor joined #salt
01:35 APLU joined #salt
01:38 Mso150 joined #salt
01:38 codysoyland joined #salt
01:39 scoates joined #salt
01:40 kwmiebach_ joined #salt
01:40 modafinil_ joined #salt
01:40 funzo_ joined #salt
01:40 lyddonb_ joined #salt
01:41 bofhit joined #salt
01:41 asmaps_ joined #salt
01:41 nhubbard_ joined #salt
01:41 rcsheets__ joined #salt
01:42 mrtrosen_ joined #salt
01:45 Bateau- joined #salt
01:45 cwright_ joined #salt
01:45 eliasp_ joined #salt
01:45 wedgie_ joined #salt
01:45 mortis_ joined #salt
01:45 etw_ joined #salt
01:45 rblackwe joined #salt
01:45 colon_D joined #salt
01:45 gladiatr_ joined #salt
01:45 Nazzy joined #salt
01:45 ksalman joined #salt
01:45 Nazzy joined #salt
01:45 platforms joined #salt
01:45 mschiff joined #salt
01:45 mschiff joined #salt
01:45 jxqz joined #salt
01:45 Emantor joined #salt
01:45 kusams joined #salt
01:45 whitenite joined #salt
01:45 irated joined #salt
01:45 ujjain joined #salt
01:45 stevednd joined #salt
01:45 anteaya joined #salt
01:45 ecdhe joined #salt
01:45 anotherZero joined #salt
01:45 Twiglet joined #salt
01:45 Jarus joined #salt
01:45 blast_hardcheese joined #salt
01:45 n8n_ joined #salt
01:46 mpoole joined #salt
01:48 pipps_ joined #salt
01:52 cb joined #salt
01:53 dude051 joined #salt
01:53 UForgotten joined #salt
01:54 JoeyJoeJo joined #salt
01:55 Nexpro1 joined #salt
02:02 ajprog_laptop joined #salt
02:10 nahamu joined #salt
02:20 n8n joined #salt
02:27 MTecknology joehh: sup?
02:29 n8n joined #salt
02:42 davidjb joined #salt
02:48 dude051 joined #salt
02:49 ndrei joined #salt
02:57 davidjb is there any reason I shouldn't be able to call salt['pillar.get']('value') from within an Jinja-templated state file? trying to call the state with salt-ssh results in a rendering error, saying the SaltDotLookup  object has no attribute 'pillar.get'.  I'm using 2014.7.0rc3.
03:05 nitti joined #salt
03:08 otter768 joined #salt
03:12 fannet try  {{pillar[‘key’]|default("6379")}}
03:19 higgs001 joined #salt
03:20 dude051 joined #salt
03:22 ingwaem joined #salt
03:22 mohae joined #salt
03:23 davidjb @fannet: unsure if that's to me, but it works.  However, I rely on numerous formulae / states that use salt['pillar.get'] ...and it should seemingly work as per http://salt.readthedocs.org/en/latest/topics/pillar/#pillar-get-function
03:24 TyrfingMjolnir joined #salt
03:32 n8n joined #salt
03:32 Flusher joined #salt
03:32 ramishra joined #salt
03:33 ingwaem` joined #salt
03:35 anotherZero joined #salt
03:35 ramishra joined #salt
03:36 pravka joined #salt
03:43 kickerdog joined #salt
03:43 higgs001 joined #salt
03:43 kickerdog just salted a complete Oracle 11g install :)
03:48 joehh MTecknology: you were asking who looks after debian packaging - joehh in irc
03:50 MTecknology joehh: ooooh! HI!
03:50 MTecknology kickerdog: holy shit man! impressive! you gonna post a formula for it?
03:50 MTecknology joehh: I filed a bug report and it was apparently a duplicate
03:51 pravka_ joined #salt
03:51 kickerdog MTecknology: yeah, it'll be the first formula i've ever posted.
03:52 kickerdog from salt-cloud to operational database takes about 50 minutes though.
03:52 MTecknology I've yet to ever touch salt-cloud
03:52 kickerdog Been using it for over a year now.
03:53 eliasp_ kickerdog: Salt + Oracle aka "The Beauty and the Beast" :)
03:53 MTecknology never had the need and can't even find one
03:53 MTecknology eliasp_: +
03:53 MTecknology eliasp_: +'
03:53 kickerdog I do everything in AWS
03:53 MTecknology :(
03:53 MTecknology +1
03:53 eliasp_ :)
03:53 MTecknology I've yet to ever touch AWS either
03:53 kickerdog ah, your lucky
03:54 kickerdog I wish I hadn't blown off my salt interview, could have logged a lot of bugs.
03:54 anotherZero joined #salt
03:57 otter768 joined #salt
04:00 n8n joined #salt
04:13 Outlander joined #salt
04:15 packeteer joined #salt
04:16 tpaul joined #salt
04:18 MTecknology Any of you use salt to manage/monitor networking equipment?
04:19 XenophonF MTecknology: managing my Cisco configs is next on my list
04:19 XenophonF i'm hoping to do it via salt-ssh
04:19 gmcwhistler joined #salt
04:19 XenophonF are you?
04:20 MTecknology I'd like to know how, but I have no clue where to start and don't really have any equipment for testing
04:20 pravka joined #salt
04:20 ramishra joined #salt
04:27 davidjb left #salt
04:32 ramishra joined #salt
04:47 ramteid joined #salt
04:52 tkharju joined #salt
04:59 kermit joined #salt
05:05 fannet anyone use mount.mounted to attach a tmpfs ?
05:07 dotz joined #salt
05:13 aparsons joined #salt
05:14 aparsons_ joined #salt
05:15 Ryan_Lane joined #salt
05:21 Ryan_Lane joined #salt
05:26 jeddi joined #salt
05:27 kickerdog joined #salt
05:28 eliasp fannet: yes, works just fine here… any problems you run into?
05:29 fannet the “device’ requirement threw me off
05:29 fannet kicked it to /dev/null
05:29 eliasp fannet: http://pastebin.kde.org/pbbnkpsad
05:29 iggy hey, who's ready to see how good salt is at fixing another massive SSL vulnerability?
05:29 fannet haha already done on our end iggy
05:29 eliasp ;)
05:29 fannet couple hundred nginx instances - poof
05:30 iggy I can't find anyone else online tonight to do the code review
05:30 iggy aside from that it's done
05:31 iggy well, merge, wait a minute, salt '*something*' state.sls nginx
05:31 calvinh joined #salt
05:32 kusams_ joined #salt
05:34 fannet elias - mkmnt does that mkdir if not exist?
05:36 eliasp fannet: exactly
05:36 fannet sweet
05:37 eliasp fannet: so you don't need another file.directory requisite
05:37 fannet just blew that away =)
05:41 calvinh joined #salt
05:42 calvinh_ joined #salt
05:48 AirOnSkin joined #salt
05:50 oyvjel joined #salt
05:50 joehh MTecknology: ok
05:58 CeBe joined #salt
06:00 ttrumm joined #salt
06:01 baconbeckons joined #salt
06:08 gyre007_ joined #salt
06:12 jalaziz joined #salt
06:18 kickerdog joined #salt
06:24 schimmy joined #salt
06:28 schimmy joined #salt
06:31 linjan joined #salt
06:35 felskrone joined #salt
06:36 hotbox joined #salt
06:36 mechanicalduck joined #salt
06:36 hotbox hi
06:38 TheThing joined #salt
06:42 trikke joined #salt
06:43 hasues joined #salt
06:51 kickerdog joined #salt
06:53 CycloHex joined #salt
06:54 CycloHex Anyone else having trouble setting the locale to utf-8? When installing postgres, it creates a default template with the ascii encoding.. This makes my application crash whenever I'd like to create a new db, because it wants to make a utf-8 table, but uses the ascii template
06:55 pravka joined #salt
06:55 akafred joined #salt
06:56 fannet anyone seen this error w/ salt-cloud -u : http://pastebin.com/wPugLFhk
06:58 felskrone joined #salt
07:00 kickerdog joined #salt
07:08 martoss joined #salt
07:09 gyre007_ joined #salt
07:09 Hydrosine joined #salt
07:10 martoss1 joined #salt
07:11 slav0nic joined #salt
07:11 slav0nic joined #salt
07:20 flyboy joined #salt
07:26 chiui joined #salt
07:29 mikeywaites joined #salt
07:30 bhosmer joined #salt
07:36 CycloHex Is it possible to require: a cmd.run state??
07:37 superted joined #salt
07:39 adsisco joined #salt
07:40 adsisco how do you write 'ps.pkill logstash' in an sls file?
07:42 n8n joined #salt
07:42 tomspur joined #salt
07:42 tomspur joined #salt
07:43 adsisco anyone have any idea?
07:45 aquinas joined #salt
07:50 snuffeluffegus joined #salt
07:55 CeBe joined #salt
07:57 intellix joined #salt
07:59 wnkz joined #salt
07:59 colttt joined #salt
07:59 baconbeckons joined #salt
08:00 nocturn joined #salt
08:01 rattmuff joined #salt
08:02 fragamus joined #salt
08:03 lcavassa joined #salt
08:03 student__ joined #salt
08:03 Damon_ joined #salt
08:04 PI-Lloyd joined #salt
08:04 wnkz_ joined #salt
08:05 calvinh joined #salt
08:05 student__ Hi all, Could you help please with my may be stupid question - when I tried to run some salt modules into the ipython, I had global name '__salt__' is not defined. How can I define it?
08:06 calvinh__ joined #salt
08:07 student__ or may be I need to import modules in right way...
08:07 student__ with __salt__
08:09 Nexpro1 joined #salt
08:10 Damon_ Is there any timetable when 2014.7 will be released(not as RC) and available as (ubuntu) package?
08:11 wnkz joined #salt
08:14 wpot joined #salt
08:17 Mso150 joined #salt
08:18 PI-Lloyd Damon_: there has been talk of it being released in the next week or two
08:19 PI-Lloyd but nothing has been officially announced afaik
08:19 krak3n` joined #salt
08:20 n8n_ joined #salt
08:20 fredvd joined #salt
08:21 nocturn joined #salt
08:22 CeBe joined #salt
08:22 VSpike I keep changing the way I organise stuff as I learn more about salt, but at the moment part of my /srv/salt/ looks like this https://bpaste.net/show/1c8d8f3d63a7
08:22 VSpike common, iis and iis/premium all contain SLS files...
08:23 gyre007_ joined #salt
08:23 VSpike however, being windows a lot of them just cmd.script a powershell script that's in win/files/scripts/... but confusingly, some of the files in there are powershell scripts to install to boxes, e.g for scheduled tasks (cron jobs)
08:24 VSpike I was thinking of moving all the powershell scripts that are spawned by states into the same place as the states themselves. I think it will be easier to maintain, and more logical.
08:24 cberndt joined #salt
08:24 VSpike Any thoughts on why this is or isn't a good idea?
08:25 VSpike I can try it and see how it goes ... fairly easy to change ... but thought someone might have experience to offer
08:25 wnkz_ joined #salt
08:25 agend joined #salt
08:25 eliasp adsisco: if you're on 2014.7/develop, you can use a cmd.run state to execute "ps.kill logstash" and then use onlyif/unless (new in 2014.7, http://docs.saltstack.com/en/latest/topics/releases/2014.7.0.html#global-onlyif-and-unless) to only execute it when logstash is running
08:26 eliasp student__: https://github.com/saltstack/salt/issues/16600 - the issue regarding the pkg.installed + slot … unless it was you who already commented there, then simply ignore this ;)
08:29 student__ <eliasp> Yyes, it was I)
08:32 linjan joined #salt
08:34 student__ <eliasp> Thank You for addition issue
08:35 eliasp student__: I tried to fix it, but the problem I ran into was, that it would need some Gentoo specific additions directly in state.pkg to deal with this and I wasn't sure whether that's the proper way to go…
08:35 eliasp student__: … so waiting for either a fix or a statement on how to proceed by the devs
08:39 CeBe1 joined #salt
08:40 AirOnSkin with salt.modules.file.replace I don't suppose I can define two paths, so it'll do the same change on two files?
08:40 calvinh joined #salt
08:42 eliasp AirOnSkin: use a simple Jinja loop to do this: {% for file in ['/foo/bar', '/boo/far'] %} {{ file }}: \n file.replace: \n -.... {% endfor %}
08:42 student__ as far as I understood, If write smth like "if slot in kwargs:version=slot+"*" " in check versions and check current version
08:42 martoss1 joined #salt
08:43 eliasp student__: yeah, sure… it's no problem at all to grab the slot from kwargs… the problem is mostly, that the generic pkg handling is simply not made to deal with a package which can be installed more than once
08:43 student__ then all will ok)
08:46 AirOnSkin eliasp: I see. Thanks :)
08:48 baconbeckons joined #salt
08:49 ggoZ joined #salt
08:50 student__ but if We are using slots, then in particular slot can be only one package... in default uses 0 slot as now, but if slot is defined we check if pkg not installed in this slot then pass slot and package into ebuild
08:51 CeBe joined #salt
08:51 student__ or I cant understand you correctly
08:52 Mso150_h joined #salt
08:54 eliasp yes, sure… one slot can only contain one package… and as long as no slot is defined, it will be left to Portage which one to pick (just like 'emerge media-libs/tiff' would do)
08:54 eliasp only when a slot is defined, this one will be used as in 'emerge media-libs/tiff:5'
08:55 eliasp but to do this, the usage of slots needs to be exposed to pkg as if they were package versions
08:55 eliasp will have a look at it this evening again… maybe I'll have an idea how to properly solve this
08:55 eliasp need to take care of some other stuff right now
08:56 oznt joined #salt
08:56 student__ If slot will not defined - salt will runs like it runs now
08:57 calvinh_ joined #salt
08:57 gerardjp joined #salt
08:57 oznt hi guys, I am installing python packages via salt.states.pip_state, and it reports success, by when I issue python -m <pkg> I get import error. So where does salt install the packages?
08:58 iondrip joined #salt
08:59 TheThing joined #salt
08:59 student__ eliasp , ok, sorry and thanks)
09:00 student__ I'll come back on this evening
09:00 bhosmer joined #salt
09:06 N-Mi joined #salt
09:10 CeBe joined #salt
09:14 wnkz joined #salt
09:18 _mel_ joined #salt
09:30 Twiglet_ joined #salt
09:31 ntt joined #salt
09:31 wnkz joined #salt
09:32 ntt Hi, someone can help me with mysql-server installation on centos with saltstack? More precisely, how should i manage the "mysql_secure_installation" script with saltstack?
09:38 Sacro ntt: i used an expect script to do it
09:44 Outlander joined #salt
09:46 giantlock joined #salt
09:47 ntt Sacro: can you explain how? Is there any tutorial?
09:48 CeBe joined #salt
09:51 intellix joined #salt
09:51 colttt joined #salt
09:55 iamtew joined #salt
09:59 glyf joined #salt
10:00 colttt joined #salt
10:01 tld_wrk joined #salt
10:02 AirOnSkin ntt: http://www.admin-magazine.com/Articles/Automating-with-Expect-Scripts
10:03 arthurlutz joined #salt
10:04 arthurlutz #poodlebleed + #saltstack http://www.logilab.org/blogentry/267847
10:04 viq joined #salt
10:04 ntt AirOnSkin: thanks
10:04 TheThing joined #salt
10:09 fragamus joined #salt
10:17 abele joined #salt
10:22 k3ys3r arthurlutz: nice one! just ran it with a test=True and it looks to try and change it both in mods-available AND mods-enabled, despite mods-enabled just being a symlink back to available. Might be worth a tweak.
10:22 bra1nout joined #salt
10:22 bra1nout_ joined #salt
10:24 jeddi joined #salt
10:24 Sacro ntt: https://gist.github.com/zanematthew/6230817
10:24 Sacro line 62 to 77 ish
10:25 arthurlutz k3ys3r: you're right, let me take a look
10:26 wnkz_ joined #salt
10:26 ntt So, in my .sls file i should before call "mysql_secure_installation" script and then run the expect script?
10:27 arthurlutz k3ys3r: didn't see it since i'm using a recent version of grep http://lists.gnu.org/archive/html/grep-commit/2012-03/msg00048.html
10:27 bra1nout Hello guys, do you know any of modules providing Web UI to SaltStack? Heard of Halite, but not sure if it is what I am looking for
10:27 bra1nout Maybe something based on Django
10:27 superted halite is a add on product to give a web gui to execute salt commands. There isn't a module as such but it's likely what your looking for...
10:28 superted Not sure how active the dev on it is though
10:28 bra1nout 7 month of inactivity :)
10:29 ramishra joined #salt
10:30 arthurlutz bra1nout: according to the mailing list there should be a new version (or product) at the end of the year
10:30 superted Custom web front end built upon the API shouldn't be too hard but i've not seen anything done around this myself. It's on our 'to-do' list where i work.
10:31 superted APi ships as part of 2014.7.0 too
10:31 bra1nout wow, I am new to python world, so my questions could be a little annoying. Thanks, Arthur
10:33 bra1nout superted: same for me, trying to learn what I can do, what can be contributed. Our company decided to start use open products instead of proprietary :)
10:35 trikke joined #salt
10:36 bra1nout And one more question to go. Is there any tools to generate reports from grains?
10:36 bra1nout *tool
10:37 colttt joined #salt
10:39 giantlock joined #salt
10:47 calvinh joined #salt
10:55 shookees joined #salt
10:57 CeBe joined #salt
10:59 arthurlutz bra1nout: we have a proof-of-concept using salt with http://sphinx-doc.org/
10:59 arthurlutz bra1nout: unfortunately not published yet
11:04 calvinh_ joined #salt
11:06 bhosmer joined #salt
11:07 glyf joined #salt
11:20 ingwaem joined #salt
11:21 ingwaem joined #salt
11:25 ramishra joined #salt
11:26 ramishra joined #salt
11:30 bigl0af joined #salt
11:37 johtso joined #salt
11:39 johtso If I have a directory of code that I have symlinked for the purposes of local development, is there a way to have salt watch for changes in that directory recursively?
11:39 glyf joined #salt
11:40 ingwaem yes, file.recurse
11:40 ingwaem and file.managed
11:40 johtso Ah okay, so I'll want to not use a symlink and do copying instead
11:40 ingwaem the following has a nice example: http://intothesaltmine.org/blog/html/2012/12/18/using_watch_with_file_recurse.html
11:41 ingwaem johtso: I don't think you need to symlink, but rather just watch the source
11:41 johtso I suppose that makes sense.. either I want to be able to change code and have it be reflected instantly.. or I want salt involved.. I can't have it both ways
11:41 ingwaem well salt can watch a source and instantly do something because of it
11:42 ingwaem so if it's copying, salt can do that
11:42 johtso I'm just used to using things like django's runserver which watches for file changes and automatically restarts, which is handy
11:42 ingwaem ahh ok :)
11:43 johtso trying to reduce the time to seeing your changes in your web browser as much as possible during development
11:43 ingwaem the age old problem :)
11:43 intellix joined #salt
11:47 ramishra joined #salt
12:00 tld_wrk joined #salt
12:01 CeBe1 joined #salt
12:06 dccc joined #salt
12:07 wnkz joined #salt
12:09 jmdcal joined #salt
12:12 jrluis joined #salt
12:14 berserk joined #salt
12:21 mrh1717 joined #salt
12:24 duncanmv joined #salt
12:24 mrh1717 Hi there, I've got an error "Invalid requisite type 'npm.bootstrap' in state 'nodejs', in SLS 'nodejs'. Requisite types must not contain dots, did you mean 'npm'?" on sls: https://www.refheap.com/91775 , can someone tell me the right syntax?
12:25 gerardjp People, I install 2014.7RC3 and am wokring on it .. Where can I address my findings? ... got a funky error
12:25 viq mrh1717: npm: {{.....
12:25 gerardjp sorry for the typos :)
12:26 mrh1717 viq, without any "bootstrap"?
12:26 viq mrh1717: just like above you have require_in: - pkg: ...
12:26 viq or - watch: - file: ...
12:26 viq mrh1717: yes
12:27 viq mrh1717: so exactly what the error message suggested to you ;)
12:28 mrh1717 viq, thank you! It passed. But I've got another one to my:  - git.latest: {{ pillar["app_url"] }} - Invalid requisite type 'git.latest' in state '/tmp/nodeapp', in SLS 'nodeapp'. Requisite types must not contain dots, did you mean 'git'?
12:29 mrh1717 viq, salt don't like "latest"?
12:30 viq mrh1717: git.latest, pkg.installed, npm.bootstrap - all of them are in form module.function. You can't have two instances of same module under a single state ID, and you require module: name
12:32 viq Does that answer your question/
12:32 linjan joined #salt
12:32 mrh1717 vlq, if I get it right, in my state: https://www.refheap.com/91776
12:32 mrh1717 I should change git.latest to git ?
12:33 viq mrh1717: yes, you should. Also, you need to have somewhere defined a state named git, and nodejs
12:34 viq because '- require: - pkg: git' does not mean "make sure a package called git is installed", it means "make sure a state called 'git' using module 'pkg' is executed first"
12:35 mrh1717 vlq, Right. Let me check it in real environment...
12:38 mrh1717 vlq, Package is bootstrapped and running. Thank you and have a great day!!
12:38 ramishra joined #salt
12:42 cpowell joined #salt
12:46 vejdmn joined #salt
12:47 ericof joined #salt
12:48 gmcwhistler joined #salt
12:49 miqui joined #salt
12:50 spo0nman joined #salt
12:51 spo0nman hello, i'm not able to run and sync custom modules, any ideas on how to debug?
12:51 spo0nman i created a module in /srv/salt/_modules/something.py
12:52 tld_wrk joined #salt
12:52 spo0nman sudo salt-call something.test  gives me Function something.test is not available
12:54 cpowell spo0nman: is there a function in your something.py called test?
12:54 spo0nman cpowell: yes
12:54 cpowell and have you done a saltutil.sync_all to sync your modules to your systems
12:55 spo0nman http://pastebin.com/raw.php?i=a3N5iZG0
12:55 superted Few thoughts Spo0nman. 1. Check you've done a module sync. 2. Check the minion log to see if it's loaded. 3.use salt-call -m /srv/salt/_modules/ which should just load the files in that directory directly if 1/2 fail
12:55 cpowell systems = minions
12:56 spo0nman superted: $ sudo salt-call -m /srv/salt/_modules/ something.test local:     True
12:56 superted I've an open bug that's fixed in 2014.7.0 where new modules are synced but are not accessible until the minion is restarted, could be related
12:56 spo0nman superted: -m /srv/salt/_modules/ works
12:56 cpowell you need the saltutil.sync_all
12:57 superted Yea, so the minion isn't aware of the plugin. Did it load it if you look in the minion log?
12:57 cpowell to push custom modules/states to your minons
12:58 spo0nman cpowell: when i do a sync_all i don't see something being copied
12:58 spo0nman something.py that is
12:58 Jellyfrog joined #salt
12:58 superted Is that module name listed anywhere in the minion log?
12:58 cpowell it should come back with _modules: \n something
12:59 Jellyfrog is it possible to include additional ssh-config with salt-ssh?
12:59 spo0nman i'm running latest stable 2014.1.11
12:59 Jellyfrog i want to do agent-forwarding with salt-ssh
12:59 otter768 joined #salt
12:59 cpowell nope, downgrade to 2014.1.10
13:00 spo0nman cpowell: superted : there is no mention of something.py on the minions
13:00 vejdmn joined #salt
13:00 spo0nman or minion log on the master
13:00 superted It may be the bug i'm seeing then. Restart the minion, wait a minute then try again. Any luck?
13:01 spo0nman oh! i forgot to mention, this is a syndic master
13:01 cpowell i would downgrade to 2014.2.10 also, .11 has a few issues
13:01 spo0nman and the other masters are running syndic
13:01 cpowell there you go, you have to manually sync formuals to syncids
13:01 cpowell or manage them with a minion
13:02 nitti joined #salt
13:02 cpowell Each Syndic must provide its own file_roots directory. Files will not be automatically transferred from the master-master.
13:02 spo0nman cpowell: you mean make the other masters a minion of the main master?
13:02 cpowell http://docs.saltstack.com/en/latest/topics/topology/syndic.html
13:03 cpowell yes, in addition to being a syndic. Or you have to manually sync the file_roots
13:03 spo0nman cpowell: i got it. won't it make the masters run the commands twice?
13:04 cpowell no, you make the minion running on the syndic report to your main master
13:07 spo0nman cpowell: wokie
13:07 spo0nman thanks
13:10 vejdmn joined #salt
13:13 racooper joined #salt
13:16 Ahlee do runners execute returners?
13:20 ajolo joined #salt
13:24 che-arne joined #salt
13:26 mapu joined #salt
13:26 vejdmn joined #salt
13:30 dude051 joined #salt
13:31 sgate1 joined #salt
13:34 aranhoide joined #salt
13:37 elfixit joined #salt
13:39 MTecknology This is what I love about open source. I wrote the entire first version of this file. https://github.com/saltstack/salt/blame/develop/salt/modules/debconfmod.py
13:39 MTecknology six lines are still mine
13:42 primechuck joined #salt
13:43 kusams joined #salt
13:46 tafa2 joined #salt
13:50 kickerdog joined #salt
13:51 wendall911 joined #salt
13:52 StDiluted joined #salt
13:52 wendall911 joined #salt
13:53 ponpanderer joined #salt
13:53 bdrung_work joined #salt
13:54 higgs001 joined #salt
13:55 VSpike I'm trying to use this in a pillar file, but I don't think I have it quite right {{ salt['grain.get']('nodename').split("-")[-1:] }}
13:56 thebishop joined #salt
13:56 VSpike What am I doing wrong there?
13:57 bdrung_work hi, i have an issue with salt 2014.1.10 when booting up a minion. /var/log/salt/minion log shows these errors: https://paste.debian.net/126954/ Consecutive salt calls work and the minion_master.pub is correct. can someone help me to troubleshoot this problem?
14:00 oyvjel1 joined #salt
14:01 oyvjel joined #salt
14:01 wnkz joined #salt
14:01 oyvjel1 joined #salt
14:02 adsisco joined #salt
14:02 adsisco left #salt
14:02 ramishra joined #salt
14:02 JoeyJoeJo left #salt
14:03 superted I presume the master is 2014.1.10 as well, if so has it had any changes to it's public key recently?
14:04 AviMarcus joined #salt
14:04 rojem joined #salt
14:05 hobakill trying to install 2014.1.11 on windows. getting a 'zmq.error.ZMQError: Address in use' error. thoughts? i've uninstalled and reinstalled and the problem persists.
14:05 rypeck joined #salt
14:07 bdrung_work superted, yes, the master was 2014.1.10 (is now 2014.1.11). the master key has never changed.
14:07 CeBe joined #salt
14:08 calvinh joined #salt
14:08 CeBe joined #salt
14:08 bdrung_work the minion boot from network and therefore are fresh every time (they only have  /etc/salt/minion, /etc/salt/pki/minion/minion.pem and /etc/salt/pki/minion/minion.pub in /etc/salt at boot)
14:09 calvinh_ joined #salt
14:11 dude051 joined #salt
14:12 dude051 joined #salt
14:12 TheThing joined #salt
14:14 Deevolution Has anyone seen a situation where minions can no longer run test.ping?  They return: "Function test.ping is not available"
14:15 djstorm joined #salt
14:15 MTecknology I think I've seen that once before, but can't remember why. It was something really screwy.
14:15 MTecknology Any other time, it's because of a typo.
14:16 Deevolution Definitely not a typo...  LOL.  I'm digging into what else it could be.
14:16 sroegner joined #salt
14:16 KennethWilke joined #salt
14:18 MTecknology I want to have ext_modules and reactor modules in git!!!
14:18 rojem joined #salt
14:18 VSpike I also tried {{ grains['nodename'].split("-")[-1:] }}foobar which gives me an error https://bpaste.net/show/aceb99c6d7ec
14:18 MTecknology then, the only thing I'll have on disk for the salt master are large files that I don't want in git.
14:19 iggy we just went ahead and put the large files in git
14:19 KennethWilke put 'em in a state
14:20 iggy just not in github since they like to randomly delete binary files (even in private repos)
14:20 MTecknology didn't know they did that
14:20 KennethWilke first i've heard of that, that sounds awful
14:20 iggy I think it's based on size and other factors
14:20 MTecknology iggy: git does pretty bad over the long haul of handling large binary files so I decided I just don't want to ever fight it
14:20 ramishra joined #salt
14:21 iggy MTecknology: yeah, our other option (and still is an option) is to put them in object storage
14:21 MTecknology currently, this includes VMwareTools-9.0.5-1065307.tar.gz  .. and nothing else
14:21 iggy git just seemed easier in the short term
14:21 workingcats hobakill, seems something else is already running on that port. check taskmanager if anything salty is in there, if so, kill it (with fire, preferably)
14:22 TimCrider joined #salt
14:22 workingcats if not, not sure how to check what program is on a particular port in windows, in linux i'd use netstat -anp
14:22 MTecknology iggy: so, how long until we can store reactors and ext_pillars in git? hm? now? hm? I want now. Now is good for me.
14:22 hobakill workingcats: agreed. thanks.
14:22 workingcats netstat exists on windows, too, so it'll probably be that, possibly with other options
14:23 TimCrider Quick question... does anyone know how to escape {{ and }} in jinja templates
14:23 iggy MTecknology: ours is in git now, but I want to say I did something rather dirty to do it
14:23 TimCrider Tried using {% verbatim %} and it yelled at me :(
14:23 iggy {% raw %} I think
14:23 hobakill workingcats: that was it. another salt-minion task was running.
14:24 VSpike duh. Just worked out the problem
14:24 workingcats cool
14:24 hobakill i just couldn't remember 'task manager' for the life of me! i'm a linux guy and rarely touch windows....obviously.... thanks again!
14:24 workingcats hobakill, hehe yw
14:24 workingcats iggy, github deletes binary files? form paid-for private repos? o.O
14:25 TimCrider @iggy Thanks, that did it
14:25 MTecknology iggy: I don't really want to hack at salt, so I guess it's still out for me. However, nice to know it's been done.
14:26 iggy MTecknology: yeah, I just have a reactor dir in our main state repo that I salt.recurse on
14:26 MTecknology ah, ya, no thanks. :P
14:27 micah_chatt joined #salt
14:27 iggy I would like to see pillars, reactors, etc. all get the same backend as states (especially with the new features coming in 2014.7)
14:28 jmdcal joined #salt
14:28 MTecknology and top.sls logic modified
14:28 iggy yeah, well... I don't see that happening unless someone figures out a very clever way to maintain backward compatibility
14:30 MTecknology heh... does the salt file server only handle things from the master branch?
14:31 MTecknology I guess that makes sense... I expected salt:// to grab from the current environment, but I guess it doesn't really have any way to pick from that environment.
14:32 MTecknology I need to deploy a gitlab-ci server
14:33 jaimed joined #salt
14:33 iondrip joined #salt
14:37 MTecknology weird, it seems to use environments on disk, but not from git
14:38 MTecknology I guess this is motivation to stick large files in git
14:40 ipmb joined #salt
14:43 jergerber joined #salt
14:47 fannet anyone know what this error is all about in salt-cloud -u :  NameError: global name 'is_writeable' is not defined    MORE: http://pastebin.com/pf6PUEcr
14:47 rojem joined #salt
14:49 adsisco joined #salt
14:49 iggy anybody know who runs intothesaltmine.org?
14:49 adsisco hi, for https://github.com/saltstack-formulas/sun-java-formula, where do i place the pillar.example after installing it?
14:50 jalbretsen joined #salt
14:50 iggy fannet: do you not see "from salt.utils.validate.path import is_writeable" on line 26 of salt/cloud/cli.py?
14:50 nitti_ joined #salt
14:51 fannet iggy I just realized that my salt-cloud is v0.8.11
14:51 fannet is that the latest on centos
14:51 iggy and your salt is?
14:51 fannet 2014.1.10
14:51 iggy uninstall the salt-cloud package
14:51 iggy salt-cloud is part of salt now
14:52 iggy wherever that v0.8.11 came from, uninstall it
14:52 fannet from pip
14:52 fannet should i re-install the salt-master package after?
14:52 adsisco can anybody please teach me how to make use of the formulas after setting them up?
14:53 iggy fannet: no, probably just uninstall that package
14:54 fannet i did then did a yum reinstall salt
14:54 fannet [root@saltmaster .ssh]# salt-cloud
14:54 fannet bash: /bin/salt-cloud: No such file or directory
14:56 arthurlutz can a scheduler take --out-file=/tmp/file as an argument ?
14:56 jnials joined #salt
14:56 pdayton joined #salt
14:56 arthurlutz and an outputter argument ?
14:57 honestly hrm, is there a counterpart to salt-cp that copies from minion to master?
14:57 iggy fannet: does that file exist?
14:57 kickerdog joined #salt
14:57 iggy I wouldn't expect it to be in /bin
14:59 slav0nic joined #salt
14:59 slav0nic joined #salt
14:59 repl1cant joined #salt
14:59 fannet fixed it - with the pip installer can I force 2014.1.10?
14:59 calvinh joined #salt
15:02 yaru22 joined #salt
15:02 mohae joined #salt
15:03 anotherZero joined #salt
15:04 pipps joined #salt
15:04 dude051 joined #salt
15:06 pipps joined #salt
15:06 thedodd joined #salt
15:07 ramishra joined #salt
15:09 UtahDave joined #salt
15:11 flyboy82 hey guys! no matter what I've tried with identation, salt keeps saving a provate key with \n instead of actual newline and prints everything in a single line
15:11 flyboy82 anyone seen this before?
15:11 fannet what version
15:11 flyboy82 I followed the YAML identation in my pillar key with pipe and 2 spaces
15:11 flyboy82 2014.1.11
15:12 fannet can you post your yaml to pastebin?
15:13 cpowell_ joined #salt
15:13 flyboy82 http://pastebin.com/hXqkhn1s
15:13 StDiluted joined #salt
15:13 Ozack1 joined #salt
15:14 flyboy82 and what I get in my file: http://pastebin.com/qNq63ta3
15:18 iggy flyboy82: nothing weird like \r\n in your files?
15:18 eliasp flyboy82: are you probably editing your YAML on windows?
15:18 flyboy82 nope... debian Konsole
15:18 flyboy82 should I use vi?
15:19 fannet try it
15:19 fannet what happens when you cat this yaml file
15:19 flyboy82 shows it as it should
15:20 flyboy82 multiline
15:20 fannet I dont see anything wrong with you yaml at first glance
15:20 zooz joined #salt
15:20 jxqz salt ftw!
15:21 flyboy82 thanks for the effort guys
15:21 flyboy82 gotta go so I'll see what I can do with it tomorrow
15:21 flyboy82 cheers
15:22 SheetiS joined #salt
15:23 higgs001 joined #salt
15:24 fannet installing salt 2014.7 from pip on centos didn’t put the init scripts in place
15:24 fannet am I missing a step
15:25 tmh1999 joined #salt
15:25 favadi_ joined #salt
15:26 conan_the_destro joined #salt
15:30 tmh1999_ joined #salt
15:30 RobSpectre joined #salt
15:31 jgelens joined #salt
15:32 linjan joined #salt
15:33 tmh1999_ joined #salt
15:35 tmh1999 joined #salt
15:39 fannet nevermind I re-install w/ bootstrap script and all is well….. I just can’t get salt-cloud to use /etc/salt/cloud.providers.d/*
15:41 rattmuff Is there a known good workaround for the following issue: https://github.com/saltstack/salt/issues/15240 ?
15:42 zerthimon joined #salt
15:42 ckao joined #salt
15:42 kaptk2 joined #salt
15:46 jxFive joined #salt
15:46 ollins joined #salt
15:47 hardwire joined #salt
15:48 tligda joined #salt
15:48 mechanicalduck_ joined #salt
15:50 jxFive hi, I'm having trouble using 'grains.get', in a template I am trying to bind to the ip address, but when testing in vagrant I have two ipv4 addresses, how can I just get the index of the first?
15:50 pipps joined #salt
15:52 abe_music joined #salt
15:55 babilen jxFive: Do you always want the first? I would probably use network.ip_addr like http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.jinja.html#calling-salt-functions -- http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.network.html#salt.modules.network.ip_addrs
15:55 babilen jxFive: Alternatively use http://jinja.pocoo.org/docs/dev/templates/#first
15:56 dude051 joined #salt
16:00 nocturn Hi, I'm running salt on RHEL6 (version 2014.1.10).  Salt seems to randomly have contact with minions, and loose them again afterward
16:01 mschiff joined #salt
16:01 mschiff joined #salt
16:02 troyready joined #salt
16:03 pipps joined #salt
16:03 beneggett joined #salt
16:04 jgelens_ joined #salt
16:05 higgs001 joined #salt
16:06 rlarkin|2 I see there's a tar file for 2014.1.12 , is that a "release" ?  I'm using salt-cloud and want to get passed 11 , but I don't want to jump the gun
16:07 viq there appears to be even a .13
16:07 viq Depends where you're looking, but https://github.com/saltstack/salt/tags and https://pypi.python.org/packages/source/s/salt/ are good places
16:11 KyleG joined #salt
16:11 KyleG joined #salt
16:11 iggy nocturn: what kind of hardware setup?
16:11 __TheDodd__ joined #salt
16:11 hasues joined #salt
16:11 iggy rlarkin|2: I don't think there were any salt-cloud fixes in .12 (fwiw)
16:12 nocturn iggy, virtual
16:13 iggy rlarkin|2: oh, yeah, there's one for scp_file
16:13 iggy rlarkin|2: you can probably just copy that file over the installed version
16:14 aparsons joined #salt
16:15 pdayton1 joined #salt
16:15 yaru22 left #salt
16:16 babilen rlarkin|2: AFAICT there are no official packages yet
16:16 ekristen joined #salt
16:16 iggy nocturn: be more specific...
16:17 nocturn iggy, this in on a virtual data center
16:17 iggy of what type?
16:17 nocturn ES X managed with Apache cloudstack and some customized middleware
16:18 Nazzy anyone know of a nice list of companies using saltstack?
16:18 iggy nocturn: hmm, do you have a way to rule out network connectivity issues?
16:19 nocturn iggy, not really.  I don't see network issues on other types of connections though (MySQL, Postgress, ActiveMQ, SSH, NFS, ...)
16:19 viq Nazzy: how about at the bottom of http://www.saltstack.com/ ?
16:20 iggy add linkedin and reddit to that list (along with many others)
16:20 Nazzy I'd missed that bit, but yeah, was hoping for a longer list heh
16:21 viq iggy: linkedin is on the list ;)
16:21 kickerdog joined #salt
16:22 apergos out of curiosity: is there an allowed character set for characters in a grain value?
16:22 apergos or can it be any arbitrary bytestring (presumably printable characters)?
16:23 linjan joined #salt
16:23 ze- I'd guess anything that can be be serialized via yaml.dump() and yaml.load()
16:24 aparsons joined #salt
16:24 bashcoder left #salt
16:24 anotherZero joined #salt
16:25 apergos wfm
16:25 iggy viq: not for me
16:25 Nazzy I apparently have to explain why puppet is not a suitable replacement for salt ¬_¬
16:25 iggy oh, it scrolls
16:25 apergos heh good luck with that
16:26 iggy "it's not written in python" at least that's enough for me
16:26 Nazzy somehow I don't think "are you having a laugh?" is a valid reason
16:28 Nazzy shockingly "I have an understanding of the internals and can fix bugs/write new code for it" also wasn't a valid reason
16:28 SheetiS Nazzy: https://developer.rackspace.com/blog/why-i-use-saltstack/ <== use saltstack because Rackspace uses it :D
16:28 apergos are they already invested in a puppet infrastructure?
16:28 SheetiS If there is already a heavy puppet infra, I can understand not wanting to switch (can be a LOT of work).
16:28 baconbeckons joined #salt
16:29 SheetiS I'd still try and switch myself, but I can at least understand
16:29 Nazzy my ruby favoring colleague uses it in a different system
16:29 higgs001 joined #salt
16:29 Nazzy I'm being asked to convert systems /from/ salt ¬_¬
16:30 iggy linkedin uses cfengine and salt together
16:30 iggy there's nothing that says you can't have both
16:30 schimmy joined #salt
16:30 Nazzy cfengine ... I have nightmares about that thing lol
16:31 apergos what is their rationale for converting?
16:32 schimmy1 joined #salt
16:32 Nazzy standardizing on a single platform
16:32 beneggett joined #salt
16:33 spookah joined #salt
16:33 apergos so it's a choice between alrey existing deployment / cofinguration systems?
16:33 Nazzy pretty much, which seems like it misses the point of having the choice in the first place
16:33 iggy go with puppet and make the other guy do all the work
16:34 iggy any time anybody asks you to do anything with it "I don't know ruby"
16:34 viq Nazzy: "salt also provides remote execution"
16:34 Nazzy puppet is not, so far as I can tell, in any way suitable for use as an orchestration system
16:35 davet joined #salt
16:35 SheetiS viq: The puppet user will say 'just use mcollective' to that unfortunately.
16:35 Nazzy viq, I spent half an hour looking at puppet's docs thinking "ok, where's the rest of it?" :p
16:35 viq Nazzy: though TBH there are things that are easier to do in puppet than in salt
16:36 iggy and probably vice versa
16:36 iggy what are the other guys arguments for going with puppet?
16:36 iggy and is he easily bullied?
16:36 viq bullied/bribed ;)
16:37 Nazzy damned if I know actually
16:37 SheetiS viq: have an example of what is easier in puppet than salt?  I've used both quite a bit and I am genuinely curious.  I don't count pre-existing modules/formulas since I mostly roll my own.
16:38 viq SheetiS: virtual resources, specifically "here's a definition for 20 users, but I only want those 5 on those systems"
16:39 Nazzy I have found exclusion based selection is a bit annoying in salt, but really that's because it's applying things backwards
16:40 viq "exclusion based selection"?
16:41 Gareth morning morning
16:41 rlarkin|2 iggy: thanks, I might do that  ( copy over the scp_file ) , I never looked into the code and didn't realize I could manage it with a single clobber
16:41 SheetiS viq: I suppose out of the box that's true.  I have some macros to cherry-pick stuff from my pillars that kinda emulate that functionality.
16:41 cb joined #salt
16:41 whitenite joined #salt
16:41 Nazzy "apply policy X to all except group Y" vs "skip policy X for system in group Y"
16:42 iggy rlarkin|2: well, it depends on the change... that particular one seemed minimally invasive
16:42 rypeck joined #salt
16:42 Nazzy the former is an exclusion rule, you're removing servers from a population
16:42 utahcon Using a reactor to call a runner... how do I pass data from the reactor to the runner?
16:42 Nazzy the latter is an inclusion rule, you're making a group and applying a rule based on it (which in this case is opting out of a policy)
16:43 Nazzy the latter can be done easily with grains and an if statement
16:43 viq Nazzy: uhm, I can't really think of the way to do "create users on all but X servers" either
16:43 SheetiS utahcon: in 2014.7 you can pass a pillar kwarg.  in 2014.1 I have to do a cmd.run on salt-call and pass the kwargs on the command line.
16:43 iggy compound targeting is quite powerful (even in the negating case)
16:44 viq Ah, that way... Well, it becomes more complicated (and much messier) if you have users in pillars
16:44 iggy '* and not dev-*': matches everything except dev machines for example
16:44 SheetiS "* and not G@categories:prod_server"
16:44 SheetiS iggy++
16:44 * viq waits for 2014.7 and pillar merging
16:44 giantlock joined #salt
16:44 Nazzy ah, yes, that works nicely
16:45 viq Though the reverse users formula seems to be interesting
16:45 Nexpro1 joined #salt
16:45 Mso150 joined #salt
16:46 Nazzy generally I'm a fan of the notion of having to explicitly define as much as possible... too much implied config can bite you later
16:47 tpaul joined #salt
16:47 Nazzy but that's the message that paranoid security processes like to drum in to you
16:47 * viq evacuates
16:51 ze- is there a known way to get a list of what would be done with a pkg.install ?
16:51 ze- aka which packets would be installed/removed/upgraded/...
16:52 utahcon ze-: depends on your distro
16:52 ze- Debian.
16:53 utahcon apt-cache depends <package name>
16:53 ze- yeah, but i meant with salt :)
16:53 ze- sorry if it wasn't clear :)
16:53 utahcon pkg.install will run: "apt-get install <package name>"
16:53 ze- apt-get install xxx would show it (and probably ask confirmation) if run manualy
16:54 utahcon ze-: pkg.install adds the yes flag, so it just accepts for you
16:54 ze- but if i want to run it on alot of hosts, pkg.install would not ask, and install/remove others stuffs without letting me see what would be done.
16:54 ze- is there something like pkg.install test=True ?
16:55 jgelens_ joined #salt
16:55 utahcon ze-: AFAIK no
16:56 Nazzy hmmm, test=True works with the portage pkg provider, but it's not super useful as I remember
16:56 utahcon ze-: state.sls has a test=True so you could run: salt \* state.sls your.state test=True
16:56 Ryan_Lane joined #salt
16:56 utahcon I don't think you will get package output though
16:57 ze- utahcon: yeah, but the state would not display exactly what would be done though.
16:57 * ze- nods.
16:57 ze- That's what I thought. Just asking before opening the issue :)
16:59 aparsons_ joined #salt
17:02 gyre007_ joined #salt
17:04 anotherZero joined #salt
17:04 ozzzo joined #salt
17:04 BrendanGilmore joined #salt
17:05 n8n joined #salt
17:08 ozzzo I am managing a list of files, and I need to require another block before they are copied. Is there a way to group them so that I can use one require statement instead of one per file? Here's what I have now: http://pastebin.com/ArHDPJtM
17:10 kermit joined #salt
17:12 eriko joined #salt
17:14 renoirb Hey, i’ve been told I can talk to manfred. That’s what Ryan_Lane told me :)
17:14 renoirb manfred, i have issues to make my openstack configuration to work with salt-cloud. Can you help me?
17:15 manfred I cannot today
17:15 manfred Too much work doing all the sslv3 changes.
17:15 renoirb oh, right :/
17:15 renoirb No hurry. I’ll ping you later.
17:17 rap424 joined #salt
17:18 kickerdog left #salt
17:20 spookah joined #salt
17:26 tld_wrk joined #salt
17:26 pipps joined #salt
17:28 forrest joined #salt
17:28 jgelens joined #salt
17:29 cpowell joined #salt
17:30 pipps_ joined #salt
17:30 aw110f joined #salt
17:31 mschiff joined #salt
17:31 mschiff joined #salt
17:32 rojem joined #salt
17:32 zooz joined #salt
17:40 Emantor joined #salt
17:41 mschiff joined #salt
17:41 mschiff joined #salt
17:42 druonysus joined #salt
17:44 CeBe joined #salt
17:46 quantumriff joined #salt
17:46 Emantor joined #salt
17:48 micah_chatt joined #salt
17:48 mschiff joined #salt
17:48 mschiff joined #salt
17:48 jgelens_ joined #salt
17:48 quantumriff Long story, but I am hoping to use ssh_auth to add authorized keys ONLY if the home directory exists on the server. (the users do not exist in /etc/passwd, they are external Active Directory users)
17:49 quantumriff is there a way to check if /home/username123 exists, before trying to add their authorized keys?
17:49 KyleG joined #salt
17:49 KyleG joined #salt
17:50 kossy joined #salt
17:51 micah_chatt_ joined #salt
17:51 quantumriff there doesn't appear to be an "unless" in the ssh_auth state
17:52 ericof joined #salt
17:53 tld_wrk joined #salt
17:53 MTecknology Is it possible to test jinja2 rendering in any really easy way? I'm trying to understand how {%- -%} will remove white space.
17:54 forrest MTecknology, not really, I believe there is an example on the jinja2 doc site though
17:55 SheetiS quantumriff: in 2014.7 unless should be available to all states, but not in 2014.1
17:55 quantumriff interesting
17:56 utahcon Aren't these, in theory, doing the same thing: http://pastie.org/9650537
17:57 utahcon running 2014.1.11
17:57 MTecknology forrest: bummer, because it's a little confusing after reading their stuff too. Thanks. :)
17:57 SheetiS you could do something ugly in the mean time though something like this {% if salt['cmd.run']('[ -d "' ~ <home_directory_variable> '" ] && echo True || echo False') == 'True' %} \n <do your state> {% endif %}
17:58 forrest MTecknology, yeah, just use -%} if you don't want the logic to create a whitespace
17:58 forrest MTecknology, or just make a quick managed file and mess with that
18:00 SheetiS basically with the {%- and -%}, you only do both if you want 2 lines with jina in between to be merged to one line after.  Otherwise only use one and use it consistently.  It can be right sometimes depending on what you are doing to start with {%- and then use %} to close.  The main thing to remember is that using both is usually _not_ what you want :D
18:01 timbermaniac joined #salt
18:01 SheetiS {%- says remove the whitespace before the jinja including newlines, where -%} says to remove the whitespace after the jinja including newlines.
18:02 kermit joined #salt
18:02 CeBe1 joined #salt
18:02 SheetiS quantumriff: hopefully that cmd.run is not too abstracted and not _too_ ugly.  If you need a full working example, just let me know and I can spend a few minutes to setup a paste of what I had in mind.
18:03 Steve7314 joined #salt
18:05 MTecknology SheetiS: I maybe did it wrong then... I have this, http://dpaste.com/2YM6NKB, and I want no empty lines at all. If none of those match, I end up with none. If one matches, I end up with whitespace at the top.
18:05 aparsons joined #salt
18:07 SheetiS since the first line has nothing before it, use {%- -%} on just the first line to start with
18:07 SheetiS does only 1 line of whitespace appear?
18:07 MTecknology yup
18:07 SheetiS I think that will do it
18:08 SheetiS you have nothing consuming the whitespace of the first enter ever
18:08 SheetiS err newline
18:09 thedodd joined #salt
18:10 SheetiS You could also switch all of your jinja to use -%} for every line, and it might also do what you want.
18:10 CeBe joined #salt
18:11 MTecknology I still get a blank line at the beginning. It's probably fine, though...
18:11 MTecknology it's just me being pedantic, it doesn't actually affect anything
18:11 debian112 joined #salt
18:11 murrdoc joined #salt
18:12 SheetiS try leaving the first line {%- -%} and then make all the other lines {% -%} and see if that works.  after that, feel free to give up ;-)
18:12 MTecknology trying now
18:12 tpaul joined #salt
18:13 slawek-_ joined #salt
18:14 MTecknology SheetiS: That did it. You can punch me now, but I realized that extra includes should be at the bottom, not top.
18:15 tmh1999 joined #salt
18:15 XenophonF joined #salt
18:15 JoeJulian joined #salt
18:16 MTecknology this is just the reverse, though
18:16 SheetiS just put the jinja at the bottom then and use {% -%} even for the first if
18:16 SheetiS if the includes need to be at bottom
18:16 MTecknology {% -%} and not {%- %} ?
18:17 SheetiS correct
18:17 mindKMST joined #salt
18:17 XenophonF in salt.states.reg.present, what does the "reflection" parameter do?
18:17 tmh1999 joined #salt
18:18 XenophonF oh never mind, i get it
18:18 XenophonF it puts the key/value into the 32-bit registry
18:18 quantumriff SheetiS: Thanks! this works perfectly: {% if salt['cmd.run']('[ -d "/home/newuser123" ] && echo True || echo False') == 'True' %}
18:19 tpaul I am having some trouble using a grain in a jinja file. Everything works as expected using grains.get on the master but doesn't seem to work when the file is rendered: http://hastebin.com/oxagolohel.yaml
18:19 MTecknology SheetiS: one is getting the space removed, and another has one extra space at the bottom
18:19 SheetiS paste me how it looks now if you don't mind?
18:20 tmh1999 joined #salt
18:21 SheetiS tpaul: use salt['grains.get'] instead of just grains.get to be able to traverse dict with :
18:21 SheetiS salt['grains.get']('set_php:output_buffering', '4096')
18:22 MTecknology SheetiS: http://dpaste.com/173FKZD
18:22 tpaul SheetiS: Thanks! I'm actually trying that right now
18:22 tpaul glad to know I was headed in the right direction!
18:23 MTecknology moses-ci get's the space, moses-cr does not
18:23 tmh1999 joined #salt
18:23 MTecknology ooooooh.....
18:23 tpaul SheetiS: What is the difference between the 2 ways of getting grains in the templates?
18:23 tpaul That worked BTW, thanks
18:24 SheetiS grains.get() uses the jinja builtin on the grains dictionary.  salt['grains.get']() uses the salt grains module's get method.
18:24 micko joined #salt
18:24 SheetiS *jinja get builtin
18:25 mindKMST I'm trying to use publish.publish from a jenkins job to apply state to specific host: "sudo salt-call publish.publish 'my-host*' state.sls mystate". It seems to work, the state gets applied but the only result I get on the minion is "local:
18:25 mindKMST ----------"
18:25 mindKMST Is there a way I can check up on the status of the job or have it return the output from applying the state?
18:25 tmh1999 joined #salt
18:26 MTecknology SheetiS: moving from -%} to {%- made the spaces go away
18:27 SheetiS MTecknology: I was thinking it might, but I was trying to play through in my head why it would still have one and hadn't hit the spot where it stays yet.
18:28 SheetiS I'm apparently poor at rendering jinja in my head ;-)
18:28 MTecknology I'm jinja at rendering poor in my head.
18:31 MTecknology exciting stuff - I now have all of my states, files, and pillar data in git and they all have prod/dev/test. Our devs have a group of servers that they have a git repo with prod/dev/test and thats how they manage the servers.
18:31 MTecknology It makes a distinct line between what I manage and what they do.
18:33 utahcon I fire up salt-master -l debug, once it calms down on my minion I run "sudo salt-call event.fire_master '{"server": "myserver.com"}' 'mycompany/glusterfs/peer'", the master does a lot, but nothing in my reactor happens. I call the exact string a second time, reactor fires. I call it a third time, nothing happens. Using 2014.1.11
18:33 snuffeluffegus joined #salt
18:33 jalaziz joined #salt
18:33 utahcon Is that a bug, known, new?
18:33 jeffspeff joined #salt
18:34 Mso150 joined #salt
18:35 snuffeluffegus joined #salt
18:39 cberndt joined #salt
18:42 snuffeluffegus joined #salt
18:46 mindKMST ah...I just have to set a timeout
18:46 mindKMST sudo salt-call publish.publish 'my-host*' state.sls mystate timeout=120
18:46 nickg utahcon: there have been quite a few bug fixes since January
18:47 thedodd joined #salt
18:52 iggy utahcon: I think I remember seeing a bug about that (1st, 3rd, etc)
18:52 thehaven_ joined #salt
18:52 hobakilllll joined #salt
19:11 ilbot3 joined #salt
19:11 Topic for #salt is now Welcome to #salt | 2014.1.10 is the latest | Help us test the 2014.7 RC! http://bit.ly/salt-rc | 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/
19:11 tkharju joined #salt
19:11 ckao joined #salt
19:12 sarlalian joined #salt
19:12 ClausA joined #salt
19:12 SheetiS Ahrotahntee:  I don't see why that wouldn't work assuming that you actually have a servers.<hostname|first> for each minion so that the pillar always compiles.
19:12 teepark joined #salt
19:12 iggy cool, DO now supports a metadata service similar to AWS, GCE, etc.
19:12 SheetiS I guess you could use grains.get to pass a default as well to cover that case though.
19:13 Ahrotahntee iggy: for limited distributions, in limited datacentres
19:13 iggy thanks for raining on my parade
19:13 Ahrotahntee SheetiS: that shortens the crap out of my top.sls; thanks :D
19:13 Ahrotahntee iggy: I'm a big fan of DO
19:14 Ahrotahntee iggy: don't get me wrong; I'm just not overly thrilled about how they've rolled out new features in the last little while
19:14 Ahrotahntee to be fair it's most DCs, and two images
19:19 pdayton joined #salt
19:26 jalaziz joined #salt
19:35 Mso150 joined #salt
19:36 tpaul Is there a php pear module for salt? I see pecl...
19:37 Emantor joined #salt
19:37 utahcon tpaul: not that I have seen
19:38 utahcon tpaul: please contribute one :D
19:39 iggy writing a python module to install php modules...
19:39 tpaul utahcon: thanks, I'm not sure if I want to write and contribute such a thing, I kind of wish pear would just die
19:40 utahcon iggy, are the pear packages you need not available through your package manager?
19:40 iggy not me...
19:40 tpaul Usually they are available via yum, but not the ones I need today
19:40 iggy I thankfully get to stay as far away from php as I do from windows these days
19:40 tpaul natually
19:40 Ryan_Lane tpaul: the pecl one will also work for pear
19:41 Ryan_Lane from the cli you can use pecl for pear modules as well
19:41 Ryan_Lane pear is starting to become unusable, though, you should switch to composer when you can ;)
19:41 utahcon learn something new everyday
19:41 higgs001 joined #salt
19:42 ingwaem joined #salt
19:42 tpaul Ryan_Lane: so if I use  pecl.installed with a pear package it should work?
19:43 Ryan_Lane yep
19:43 tpaul Cool, thanks, I'll give that a try
19:43 murrdoc joined #salt
19:44 TheThing joined #salt
19:45 phblj joined #salt
19:47 Mso150 joined #salt
19:51 P0bailey joined #salt
19:51 P0bailey joined #salt
19:53 ndrei joined #salt
19:56 baconbeckons joined #salt
19:57 thehaven joined #salt
19:58 duncanmv joined #salt
20:03 viq_ joined #salt
20:03 hasues joined #salt
20:03 mapu joined #salt
20:04 dude051 joined #salt
20:04 Outlander joined #salt
20:05 druonysuse joined #salt
20:05 druonysuse joined #salt
20:05 fragamus joined #salt
20:07 wt joined #salt
20:07 wt Hi, is the doc at http://docs.saltstack.com/en/latest/topics/development/external_pillars.html#ext-pillar correct?
20:08 wt I am looking at the code for the git pillar
20:08 wt It's args for ext_pillar are confusing me.
20:08 murrdoc depends on version of salt u are using
20:08 wt def ext_pillar(minion_id, repo_string, pillar_dirs)
20:08 murrdoc but yes it should be correct
20:09 wt I am looking at 2014.7
20:09 wt and 2014.1
20:09 wt They both appear to have the same function args
20:10 wt however, the docs describe the second arg as being the pillars from previously-run modules, if I am understanding the doc correctly
20:11 Setsuna666 Anyone have a bug with 2014.7rc3 with the archive.py state module ? Since upgrading to this version of 2014.7, I'm having a python exception: http://pastebin.com/aRAGRYrB
20:11 PI-Lloyd joined #salt
20:11 wt the git pillar works, so I assume that it must be getting what it expects...so how does it get a repo_string instead of the dict of pillars?
20:13 giannello joined #salt
20:14 PI-Lloyd anyone know if this commit on the develop branch is in 2014.7 or when it will be? - https://github.com/saltstack/salt/commit/407e48ae3516de8893f942fecd6599a38d7fd668
20:16 wt Other pillars seem to take the args I'd expect.
20:19 wt Is git_pillar specially handled?
20:19 Cidan has anyone started to use RAET?
20:19 Cidan any thoughts?
20:23 bhosmer joined #salt
20:26 martoss joined #salt
20:29 gyre007_ joined #salt
20:29 giantlock joined #salt
20:31 gyre007_ joined #salt
20:32 fragamus joined #salt
20:32 cpowell joined #salt
20:33 wt I guess git_pillar is specially handled: https://github.com/saltstack/salt/blob/develop/salt/daemons/masterapi.py#L50
20:33 gyre007_ joined #salt
20:34 wt that's funky and really magical
20:35 martoss1 joined #salt
20:35 wt yuck: https://github.com/saltstack/salt/blob/develop/salt/master.py#L146
20:35 faust joined #salt
20:37 whitenite joined #salt
20:38 fragamus joined #salt
20:38 faust left #salt
20:40 baconbeckons joined #salt
20:40 wt Okay, this makes stacking an ext_pillar before the git_pillar not work like I'd expect and not like the ext_pillar docs say.
20:40 wt That's pretty lame.
20:40 fragamus joined #salt
20:41 Ryan_Lane wt: open an issue in github
20:41 Ryan_Lane would be good for that to be fixed
20:41 wt I will. I was trying to see if I could unwind it it my head...maybe figure out a solution.
20:41 Ryan_Lane it's a bit weird that git isn't just another ext_pillar
20:42 wt It looks like someone was trying to get around the limitation of the salt loader where the modules are reloaded every time.
20:42 wt Or maybe there was some other motivation.
20:42 mohae joined #salt
20:44 PI-Lloyd anyone had any joy with getting the digitalocean cloud provider working using either the v1 or v2 digitalocean api?
20:45 dude051 joined #salt
20:45 babilen Are there good theforeman alternatives that work with salt (I am aware of the foreman salt plugin, just looking for alternatives)
20:46 Ryan_Lane are you using foreman to manage hardware?
20:47 Ryan_Lane there's not really much around that handles this well
20:47 Ryan_Lane maybe openstack ironic/nova?
20:47 Ryan_Lane but that's not a minor change
20:47 kballou joined #salt
20:48 babilen I'm essentially looking for active projects that allow us to install VMs before we provision them with salt.
20:48 martoss joined #salt
20:48 phblj is there a way to install npm and then respect npm.installed states in the same pass?  I'm having to run state.highstate twice, the first time it installs npm and tells me "npm.install isn't available" then the second it can execute npm.installs
20:48 Ryan_Lane oh. what are you using for virtualization?
20:48 philipsd6 joined #salt
20:48 Ryan_Lane babilen: salt-cloud is meant for this
20:49 Ryan_Lane salt-cloud is a built-in part of salt now
20:49 Ryan_Lane it'll create vms, install salt, connect to the master, then provision itself
20:49 bhosmer joined #salt
20:50 babilen I know salt cloud, but that doesn't work (yet) with our cloud. Which reminds me: Are there specifications what you have to implement to enable a system to be usable with salt-cloud?
20:50 philipsd6 exit
20:51 babilen Ryan_Lane: We run our own system on bare metal - Just trying to look into alternatives for the provisioning stack as we'd like to be a bit more independent from Xen.
20:51 babilen A nice integration with salt-cloud would be nice, but first I need an alternative for the initial install. :)
20:51 Ryan_Lane heh
20:52 Ryan_Lane you may want to consider openstack ironic
20:52 philipsd6 joined #salt
20:52 Ryan_Lane it's a bare-metal provisioner
20:53 babilen ack, thank you. I'll look into that too.
20:53 martoss1 joined #salt
20:55 babilen Ah, no .. not quite. We are running our own software to create/manage virtual machines, so I'm not really looking for tools to provision bare metal servers per-se, but simply VM provisioning tools that work on bare-metal, but aren't restricted to "If you use cloud provider A, use ..., B use, .."
20:57 dude051 joined #salt
20:58 tafa2 joined #salt
20:58 CeBe joined #salt
21:02 martoss joined #salt
21:03 aparsons joined #salt
21:03 jdmf joined #salt
21:04 jcl joined #salt
21:05 jalaziz joined #salt
21:06 cb joined #salt
21:06 jcl Hey all, as soon as my salt-minions start, I end up with 2 defunct processes right away (Ubuntu 12.04), here are the messages in /var/log/salt/minion: https://gist.github.com/Calygos/f61f4b44bd3201136b7a
21:07 timoguin joined #salt
21:07 jcl any ideas what might be causing that?
21:07 dude051 joined #salt
21:12 fannet guys is salt-master stable on centos 7?
21:12 fannet I’ve been running into all sorts of wierd bugs not seen in debian
21:14 jcl fwiw, i don't generally get the defunct processes on ubuntu 14.04 but one of my few server running that version is getting them
21:16 Outlander joined #salt
21:17 martoss1 joined #salt
21:19 Emantor joined #salt
21:19 Pixionus So odd occurrence.  We have a system where when I run test.ping I don't get back true or false for a number of minions...  What am I missing?
21:19 oyvjel joined #salt
21:19 jcl Pixionus: what do you get with salt-run manage.status on the master?
21:19 iggy Pixionus: happens to lots (all?) of us from time to time
21:19 Ahlee Pixionus: are the non-responsive systems under distress?  Does it take longer than ~5 seconds for the minions to respond (the default the cli will wiat)?
21:21 wt Are you sure the minions are authed/connected to the master?
21:21 Pixionus So they are down, but I don't knwo why.  One of them is sitting on my desk right now and chugging along happily enough.
21:21 nitti joined #salt
21:21 Pixionus yes wt, they are.
21:22 iggy if you run it a few times in a row, do they start responding?
21:22 Ahlee how many minions to how many worker threads on the master?
21:22 Pixionus Don't think they are under distress Ahlee.  trying to rerun a few times iggy
21:22 ingwaem Pixionus: I sometimes find that the ping command will appear to timeout meanwhile it's actually the cli process that timed out. You can change this by using a -t param...example salt '*' test.ping -t 90 will tell the cli to wait 90 seconds for replies
21:23 Pixionus yeah, getting a few more servers to respond now
21:24 Ahlee fannet: no idea on cent7, but i ended up wrapping salt inside supervisord (minions, master, and api) on cent6
21:24 jcl Pixionus: probably the key rotation issue
21:24 Ahlee jcl: key rotation issue?
21:24 jcl Pixionus: run salt-run manage.status a few times, it just does test.ping under the hood but will summarize which hosts are connected (up) and which aren't (down)
21:25 Ahlee jcl: what version of salt is throwing that proxytype KeyError?
21:25 jcl needing to ping after key rotation since the sessions are lost
21:25 jcl Ahlee: 2014.1.10 and 2014.1.11
21:25 Ahlee ah.  That's likely the same issue I get through the API some times
21:26 Ahlee minions will return they're not responding immediately after being added/deleted
21:26 Pixionus these have been on it for some time.
21:26 jcl it should get fixed in 2014.7 with a new variable, ping_on_rotate
21:26 ntropy_ joined #salt
21:26 Ahlee Pixionus: again, how many minions, and what's your maseter's worker threads set to?
21:27 Ahlee I have ~750 minions on my busist master with 50 worker threads, and I don't get test.ping timeouts even with a dog slow returner I'm writing to
21:27 alex-solidfire joined #salt
21:27 fannet ahlee - any reason you did that
21:28 __TheDodd__ joined #salt
21:28 Ahlee fannet: what exactly?  bump the number of workers so high?
21:28 wt Pixionus, how many minions?
21:28 Pixionus Not many.  17 minions and only a few that are showing sometimes and not others.
21:28 wt Pixionus, which version?
21:28 Ahlee should not have an issue with 17 minions and the default (8 i think?) workers
21:28 Pixionus not sure what you mean what my master worker threads are set to though
21:28 pipps joined #salt
21:29 fannet Ahlee - putting salt inside supervisord
21:29 Ahlee fannet: segfaults causing processes to die
21:29 Ahlee so rather than have to go find what minions (or worse the API or master) where dead I just tossed them in supervisord and went about my business
21:30 fannet I’m having an issue w/ the minions barfing when calling salt-call state.highstate
21:30 fannet File "/usr/lib/python2.6/site-packages/salt/fileclient.py", line 923, in get_file
21:30 fannet if not data['data']:
21:31 fannet here’s the full dump http://pastebin.com/i269Um3m
21:31 Pixionus 2014.1.7
21:31 Pixionus wt: ^
21:34 Pixionus actually will be migrating the master to the cloud so Ill find out if maybe it's a matter of the local machine or some such and deal with it if it's a problem then.
21:34 perfectsine joined #salt
21:34 Emantor joined #salt
21:36 peters-tx joined #salt
21:36 fannet actually all these wierd issues only started popping up when I started using gitfs
21:37 iggy fannet: that means data is a string and not a dict... indentation/:/- error maybe?
21:37 fannet iggy: indentation where though? seems like its barfing on core salt stuff?
21:38 iggy seems like it's barfing on your top file or some state file to me
21:38 Guest21651 joined #salt
21:38 fannet hmm
21:38 iggy I normally binary search by commenting out half the top file at a time
21:39 iggy then 1/4, then 1/8, etc
21:41 Pixionus ok, one of these servers, I think the problem is I installed a new bootstream.  Sorry kinda new to this so if anyone has the time to humor a noob, I'd appreciate it.
21:41 fannet iggy (used a sublime macro to re-indent everything)
21:41 fannet lets see what happens
21:41 Pixionus not sure what I need to do to get it running again.
21:42 Outlander joined #salt
21:44 smcquay joined #salt
21:45 Pixionus never done any kernel dev and then I got thrown into it at work.
21:45 mosen joined #salt
21:45 fannet :-/ still the same issue
21:46 Pixionus Trying to figure my way through it and test everything now to make sure I didn't break anything.. Unfortunately I may have done so.
21:46 UForgotten joined #salt
21:50 fannet iggy - I have 2 repos for the states. One with just the top.sls and the other with the state files.. are there any special considerations in that type of setup?
22:00 pipps joined #salt
22:00 fannet found the source of that bug (had a trailing “:” in master config)… but now it just can’t find the top.sls file :-/
22:02 dccc joined #salt
22:03 jcl joined #salt
22:06 NaPs joined #salt
22:07 UForgotten joined #salt
22:09 murrdoc joined #salt
22:11 aparsons joined #salt
22:11 Outlander joined #salt
22:13 whitenite joined #salt
22:17 UForgotten joined #salt
22:17 jgelens_ joined #salt
22:18 kermit joined #salt
22:25 baniir joined #salt
22:29 reversEngineer joined #salt
22:38 pipps joined #salt
22:41 snuffeluffegus joined #salt
22:41 phblj joined #salt
22:42 pdayton1 joined #salt
22:43 UForgotten joined #salt
22:44 jalaziz joined #salt
22:48 baconbeckons in salt-cloud, can i only define subnetid in my providers or can it be defined in my profiles?
22:49 manfred i would assume you could provide it in either
22:50 forrest Hmm, has anyone run the salt centos bootstrap today? Getting a failure on retrieving the epel metadata.
22:53 SheetiS forrest: sometimes I have to do a 'yum clean all' to get epel to behave recently.
22:54 forrest SheetiS, yeah this is on a freshly provisioned machine that the bootstrap is running on, so it should be clean already, I'll look at modifying the script i have though to see if that helps, good idea.
22:55 rawzone joined #salt
22:55 jalaziz joined #salt
22:55 SheetiS There's no downtime indicated on fedoraproject.org http://fedoraproject.org/wiki/Infrastructure/Downtime
22:55 SheetiS Couple weeks ago I had the problem and it was because of a downtime.
22:55 baconbeckons manfred : i don’t see the devision between providers and profiles. is this spelled out in the docs somewhere?
22:55 jab416171 joined #salt
22:55 kalessin joined #salt
22:55 forrest SheetiS, yeah that's why I was asking, it usually only occurs when there is downtime
22:56 Ryan_Lane joined #salt
22:56 manfred baconbeckons: providers should be the configuration to use the api
22:56 manfred and then any defaults you want
22:57 manfred profiles is the profile for how to build a server, and then you can override some of those things in the profiles.
22:57 tafa2 joined #salt
22:57 manfred there are some things that don't overlapp
22:57 manfred like image or flavor
22:57 baconbeckons manfred: can i override whatever i want in the profiles?
22:57 manfred but lots of the extra capabilities override
22:57 manfred not whatever
22:57 manfred just like server specific things
22:58 manfred like the which key to put on the server, what networks to attach
22:58 baconbeckons for instance, i initially had subnetid defined in my providers and ended up with differet providers based on this subnet, but it seems to make more sense to put it into profiles or at least override it as needed in profiles
22:58 manfred yeah
22:58 manfred you should be able to do that
22:58 manfred which provider?
22:58 baconbeckons ec2
22:59 manfred so
22:59 manfred the way it works
22:59 manfred is salt-cloud parses the configs, /etc/salt/cloud, then /etc/salt/cloud.providers.d/* and lastly /etc/salt/cloud.profiles.d/* and then it generates a vm_ object
22:59 manfred profiles.d is loaded last, and will override variables from providers.d
23:00 manfred baconbeckons: then __opts__ could override those, etc
23:00 manfred baconbeckons: https://github.com/saltstack/salt/blob/develop/salt/cloud/clouds/ec2.py#L2011
23:00 baconbeckons manfred: given that i can extend a provider or a profile, why do both exist?
23:01 manfred it uses that function, config.get_cloud_config_value() to get the variable
23:01 baconbeckons i understand what you said about how to divide api config, but it seems like that could be done by extending as well
23:01 manfred the provider is the information with how you interact with the api
23:01 manfred the profile is the server profile
23:02 manfred in the profile... you specify which provider it applies too... i don't understand the question
23:03 jmdcal joined #salt
23:03 aparsons joined #salt
23:03 manfred so
23:04 manfred you start with /etc/salt/cloud, this is information that applies to all providers
23:04 manfred liek display_ssh_output
23:04 manfred etc
23:04 manfred then you have your provider, have one for each ec2 account, then your rackspace account, then your digital ocean account
23:04 manfred then lastly, you configure all the types of servers you want to create on each provider
23:05 manfred a centos one for an ec2 account, ubuntu for your DO account, and then archlinux/gentoo on your rackspace account
23:05 manfred each step could override something from above it
23:05 iggy you also can't infinitely extend profiles
23:05 manfred say on your rackspace account, you do want the ssh output, you add display_ssh_output: True to that one provider
23:05 manfred and that is the only one that overrides it
23:05 manfred or
23:06 manfred you only want ssh out put for the gentoo image on your rackspace account
23:06 manfred add it to the profile
23:06 rlarkin|2 regarding cloud...I terminated an aws instance that was in my map and reran salt-cloud -m , it did get replaced.  but later when I added new instances to my map the "replacement" instance wasn't recognized as pre-existing
23:06 manfred they server different purposes
23:06 manfred that is why both providers and profiles exist
23:07 manfred baconbeckons: does that answer your question?
23:08 baconbeckons manfred: yes, this makes more sense
23:08 jgelens_ joined #salt
23:09 jalaziz_ joined #salt
23:09 baconbeckons i was confused because some of the sample providers had multiple providers for ec2. the example showed different subnets being defined in the provider. so for ec2, 2 providers would exists, one for each subnet
23:10 baconbeckons thanks manfred
23:10 baconbeckons :)
23:10 manfred so
23:10 manfred don't take that 100%, because it may have to be defined in the provider
23:10 manfred but most of the variables don't require it
23:10 manfred (i would have to look closer, but the information should be overwritten by the profile in the vm_ info)
23:11 manfred but theoretically it should work in the profile
23:15 ryuhei joined #salt
23:21 gyre007__ joined #salt
23:23 bhosmer joined #salt
23:24 che-arne joined #salt
23:25 druonysus joined #salt
23:25 druonysus joined #salt
23:34 druonysus joined #salt
23:36 dude051 joined #salt
23:44 aquinas joined #salt
23:51 murrdoc any pointers for salt-minion stuff ?
23:51 murrdoc i want to whip one for interactng with dell dracs
23:52 kermit joined #salt
23:53 iggy you mean like a module/state/grain?
23:53 manfred what do you mean?
23:53 manfred there is a drac module
23:53 manfred or i thought there was
23:54 manfred nevermind, i was just remembering the F5 one
23:54 manfred murrdoc: do you mean modules?
23:54 ph8 joined #salt
23:54 manfred murrdoc: http://docs.saltstack.com/en/latest/topics/topology/proxyminion/index.html
23:54 murrdoc sorry i meant salt proxy minion
23:55 murrdoc yeah that one
23:55 * iggy backs away slowly
23:55 murrdoc have you guys seen any good examples of those
23:56 dude051 joined #salt
23:56 murrdoc no
23:56 murrdoc ok then
23:56 manfred murrdoc: someone went the other way and made a runner for the bigip f5 https://github.com/saltstack/salt/blob/abde8b038c35a568c5996b472bda775087bcd0ff/salt/runners/f5.py
23:57 murrdoc haha
23:57 manfred https://github.com/saltstack/salt/blob/develop/salt/runners/f5.py
23:57 murrdoc is that u ?
23:57 manfred no
23:57 manfred i was thinking about it, but never did
23:57 jgelens_ joined #salt
23:57 murrdoc this is sweet
23:58 manfred i am not aware of any proxy modules, except the rest_service or rest_package or junos modules
23:59 manfred the junos one isn't bad
23:59 murrdoc rest_service ?
23:59 murrdoc got link ?
23:59 murrdoc imma whip something up for dracs
23:59 murrdoc using paramiko

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