Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-03-11

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

All times shown according to UTC.

Time Nick Message
00:03 Yaann joined #salt
00:04 goodwill run kiorky, ruuuunnn
00:04 goodwill okie, who here is looking for a job
00:04 forrest goodwill, he's probably not around, it's after 1 in France
00:04 * goodwill looks at forrest
00:06 zz_Cidan ya, if anyone wants a fun devops position in LA...
00:09 coolj joined #salt
00:10 drogoh those words mean something to me. I am in the LA area
00:10 zz_Cidan orly?
00:10 ajw0100 joined #salt
00:11 drogoh currently working solely on chef right now, but I have recent exposure to salt
00:12 zz_Cidan don't really care about what deploy tools you've used
00:12 zz_Cidan just care if you know what you're doing, are quick on your feet, and are generally a superstar
00:13 Voziv left #salt
00:13 drogoh ah, that disqualifies me. I openly admit I have no idea what I'm doing
00:13 zz_Cidan :(
00:13 forrest superstar is a bad term
00:13 zz_Cidan yeah it's a shitty term
00:13 goodwill zz_Cidan: I asked first
00:13 forrest I hate when people want superstars :\
00:14 zz_Cidan i agree fully
00:14 timoguin_ joined #salt
00:14 zz_Cidan but working with sysadmins that are less than stellar is painful
00:15 zz_Cidan so that will be my new word
00:15 zz_Cidan stellar sysadmin
00:15 timoguin_ zz_Cidan, I would only take a job like that if I could dress like Marlon Brando and call myself a Stella Sysadmin
00:16 zz_Cidan sold
00:16 forrest hah
00:16 zz_Cidan come on by
00:17 goodwill zz_Cidan: do not piggy back on my questions ... get your own
00:17 zz_Cidan fine
00:17 goodwill :-P
00:17 goodwill jk
00:17 timoguin_ snap!
00:17 zz_Cidan does anyone want to stop working for goodwill and work for me?
00:17 goodwill zz_Cidan: where in LA are you at
00:17 zz_Cidan santa monica
00:20 ipalreadytaken joined #salt
00:21 failed_array joined #salt
00:26 l0x3py joined #salt
00:31 faldridge joined #salt
00:32 [diecast] joined #salt
00:33 rojem joined #salt
00:34 kickerdog joined #salt
00:35 oz_akan_ joined #salt
00:38 justlooks joined #salt
00:40 toguin__ joined #salt
00:40 justlooks hi,i try Hydrogen new ferture, minionFS ,i copy file from minion to master using "# salt 'saltslave2' cp.push /tmp/myfile" but i can not see any file in /var/cache/salt/master/minions/saltslave2/ dir ,why?
00:44 elfixit joined #salt
00:47 oz_akan_ joined #salt
00:50 justlooks my mistake ,it should reconfig master ,and restart it
00:53 bryanl joined #salt
00:56 jnials joined #salt
00:58 robawt lol @ zz_Cidan
00:58 robawt "does anyone want to stop working for goodwill and work for me? "
01:07 rojem joined #salt
01:11 pydanny joined #salt
01:19 elfixit joined #salt
01:23 cedwards joined #salt
01:23 jslatts joined #salt
01:30 ajprog_laptop joined #salt
01:35 bhosmer joined #salt
01:37 schimmy joined #salt
01:38 schimmy1 joined #salt
01:44 baniir i have "- order: last" in one of my states but others are preceding it
01:46 baniir correction. i mean it precedes other states
01:49 greg joined #salt
01:50 helderco joined #salt
01:55 greg Hi Everyone, I'm having a difficult time setting up the gitfs backend with my private github repository. I've setup ssh to authenticate to github, ssh -T git@github works but I'm getting an exception by GitPython in the master log file.
01:55 greg Someone had a similar issue here: https://github.com/saltstack/salt/issues/8684
01:55 greg But it isn't helping for me.
01:57 Networkn3rd joined #salt
01:57 greg I was wondering if I'm missing something. gitfs_remotes looks like - git+ssh://git@github.com/foobar/states.git
01:58 budrose joined #salt
01:59 greg I'm running salt-master as root...
02:02 generj joined #salt
02:03 [diecast] joined #salt
02:03 greg ok, thx anyways
02:04 Yaann joined #salt
02:07 budrose joined #salt
02:10 Ryan_Lane joined #salt
02:11 sroegner joined #salt
02:12 tmandry Is it possible to see the post-jinja version of an SLS file during execution? Trying to debug why this formula won't work
02:13 defunctzombie having issues with git repo updates on salt 2014.1.0
02:13 defunctzombie seeing a lot of
02:13 defunctzombie Command 'git rev-parse staging ^{commit}' failed with return code: 128
02:13 defunctzombie [ERROR ] output: fatal: ambiguous argument '^{commit}': unknown revision or path not in the working tree.
02:16 yomilk joined #salt
02:17 tmandry got it, salt-call -l debug
02:23 MTecknology http://dpaste.com/1713136/   that's... interesting
02:23 rostam joined #salt
02:27 MTecknology any other box appears to be fine
02:27 MTecknology it's just this one that seems royally ticked off
02:30 cewood joined #salt
02:38 Furao joined #salt
02:44 d3vz3r0 joined #salt
02:47 d3vz3r0 joined #salt
02:52 d3vz3r0 joined #salt
02:52 higgs001 joined #salt
02:54 d3vz3r0 left #salt
02:55 d3vz3r0 joined #salt
03:05 justBob_ joined #salt
03:10 Furao joined #salt
03:15 Networkn3rd joined #salt
03:21 [diecast] salt + amazon vpc + dhcp dns = me confused
03:22 kermit joined #salt
03:26 elfixit joined #salt
03:29 smcquay joined #salt
03:32 ckao joined #salt
03:38 CeBe1 joined #salt
03:38 pydanny joined #salt
03:44 xl1 joined #salt
03:50 elfixit joined #salt
03:52 faldridge joined #salt
03:54 ajw0100 joined #salt
03:55 ravibhure joined #salt
04:01 defunctzombie anyone got a tip on how to install my own salt-minions on my servers via salt itself?
04:01 defunctzombie should I use the python packaging process or something else?
04:02 manfred salt-cloud can deploy for you
04:06 xmltok joined #salt
04:08 schimmy joined #salt
04:12 defunctzombie right
04:12 defunctzombie but I am looking for a way to change the version of salt installed after a deploy
04:12 manfred ahh
04:12 defunctzombie lets say I make a patch to salt-minion
04:12 manfred i have a state for that
04:12 defunctzombie I would like to use my version
04:12 defunctzombie ideally right from github
04:12 defunctzombie tips on how to do that?
04:12 manfred defunctzombie: that is what i do
04:12 manfred defunctzombie: http://ix.io/b1W
04:13 defunctzombie manfred: and that will just do it?
04:13 manfred yeah
04:13 manfred it will make sure that the git repo is up to date
04:13 defunctzombie manfred: amazing.. gonna try it out now on a test box
04:13 manfred then python setup.py install in it, then sync everything
04:13 defunctzombie what is the sync for?
04:13 manfred need to change the name for git repo, and then the branch in the pillar data
04:14 defunctzombie right
04:14 manfred just to sync all pillar data, grains, modules, states, etc
04:14 defunctzombie I see
04:18 defunctzombie manfred: any idea if this will play nice with any salt-minion packages installed by my system package manager?
04:18 defunctzombie manfred: (debian in this case)
04:18 defunctzombie I suspect I will need to uninstall those?
04:18 manfred meh, no idea about that, python setup.py install should take precidence
04:19 defunctzombie k
04:20 manfred i just bootstrap them all with -LP -g git://github.com/gtmanfred/salt.git <branch> for salt-bootstrap
04:21 schimmy joined #salt
04:21 defunctzombie meaning?
04:22 defunctzombie I use salt-cloud to bring up servers
04:22 Furao left #salt
04:28 apergos joined #salt
04:29 zain_ joined #salt
04:33 sroegner joined #salt
04:38 gnugnu9 joined #salt
04:38 Gifflen joined #salt
04:46 favadi joined #salt
04:46 kolaman joined #salt
04:52 ipalreadytaken joined #salt
04:59 cewood joined #salt
04:59 jalbretsen joined #salt
05:01 echos joined #salt
05:09 pydanny joined #salt
05:12 Gifflen joined #salt
05:13 manicouman joined #salt
05:16 yomilk joined #salt
05:33 kickerdog joined #salt
05:54 ajw0100 joined #salt
06:00 andr386 joined #salt
06:01 kiorky goodwill: pong
06:02 manicouman joined #salt
06:04 linjan joined #salt
06:20 kickerdog joined #salt
06:21 sroegner joined #salt
06:22 Yaann joined #salt
06:23 ndrei joined #salt
06:29 kickerdog joined #salt
06:33 CeBe joined #salt
06:37 Furao joined #salt
06:45 ajw0100 joined #salt
06:48 ndrei joined #salt
06:52 schimmy1 joined #salt
06:54 schimmy joined #salt
06:57 dave_den joined #salt
07:00 bhosmer joined #salt
07:04 justlooks joined #salt
07:06 justlooks i am testing salt-api ,i use obsolete token to get minion info ,it still work ,why? https://gist.github.com/justlooks/9480880
07:17 Furao token expires after a while
07:17 carlos joined #salt
07:18 jalaziz joined #salt
07:19 Furao https://gist.github.com/bclermont/5212641
07:19 Furao I wrote that last year, I think it don't work with 0.8.11 but this renew the token for you
07:19 Iwirada joined #salt
07:23 roolo joined #salt
07:38 vortec_ joined #salt
07:39 Yaann joined #salt
07:40 lessismore joined #salt
07:42 alfborge joined #salt
07:44 alfborge Does anyone know of documentation on how to use salt and salt cloud for continous integration with jenkins?
07:45 ChaosPsyke joined #salt
07:50 ndrei joined #salt
07:51 mikeywaites joined #salt
07:58 balboah joined #salt
08:02 favadi joined #salt
08:05 Kenzor joined #salt
08:10 sroegner joined #salt
08:13 ml_1 joined #salt
08:14 elsmorian joined #salt
08:15 Furao alfborge: good luck! I had to wrote my own testing framework for that (to continuously test all formulas)
08:16 kadel joined #salt
08:16 svx joined #salt
08:20 Shenril joined #salt
08:21 alfborge Furao: Thanks, I was hoping there was some community knowledge we could leverage since we've just started looking into saltstack.  It seems a great fit, but it's always nice to learn from others experiences. :)
08:27 faldridge joined #salt
08:27 harobed joined #salt
08:30 slav0nic joined #salt
08:30 davidk joined #salt
08:32 Guest38292 Hi - quick question from a salt newbie - when would you write modules and when would you write salt state files?
08:34 topochan joined #salt
08:35 Guest38292 no takers?
08:37 yomilk joined #salt
08:39 malinoff joined #salt
08:41 alfborge I could try.
08:41 alfborge But I'm a newbie as well, so take what I say with a grain of salt...
08:41 alfborge state files are a declarative description of what your environment should look like
08:42 alfborge modules are python "scripts" that do actions in your environment
08:42 giantlock_ joined #salt
08:42 alfborge So if you want to perform an action in your environment that isn't handled by existing modules, then you should write a module.
08:43 alfborge However, if you want to specify how your environment should look like then you're better off with a state file.
08:44 alfborge In other words: If the module you're creating contains code that checks whether what you want to do has already been done, then this sounds like something you want to from a state file.
08:46 alfborge Hope that helps.
08:59 ipalreadytaken joined #salt
09:08 viod joined #salt
09:15 Nexpro joined #salt
09:22 mikeywaites Hey guys having a bit of a weird issue using git file_server.  When the repo is being cached on the master it seems to be creating some of the sls files as directories opposed to files.
09:22 alfborge joined #salt
09:26 mikeywaites take for example, i have an elasticsearch module with an init.sls file and an install.sls file.  when looking in /var/cache/salt/master/salt/states/elasticsearch/ dir the install.sls file is actually a directory
09:33 helderco joined #salt
09:34 Yaann joined #salt
09:35 sgflt joined #salt
09:50 aleszoulek joined #salt
09:51 ndrei joined #salt
09:56 bhosmer joined #salt
09:56 helderco joined #salt
09:59 sroegner joined #salt
10:02 tmandry joined #salt
10:06 mike25ro hi guys.. i am deploying some file with a state. I have a service that need to restart ONLY if BOTH files have been deployed.  Is there a way to watch for 2 state definitions .. but not using OR ... i need using an AND.. if  file1 and file2 then reload the service
10:07 helderco joined #salt
10:11 sroegner joined #salt
10:13 johtso joined #salt
10:13 helderco joined #salt
10:16 alfborge joined #salt
10:24 nn0101 joined #salt
10:28 faldridge joined #salt
10:37 yota joined #salt
10:38 yomilk joined #salt
10:40 helderco joined #salt
10:44 bbinet mike25ro: would it work if you make you second file require your first file, then your service would only watch only the second file to deploy?
10:59 Furao joined #salt
11:01 manicouman joined #salt
11:09 manicouman joined #salt
11:13 mike25ro bbinet: i have actually used the require .. i have added require for the service... so it requires both files.
11:14 mike25ro bbinet:  your idea is good should work
11:18 viod joined #salt
11:22 viod joined #salt
11:28 vbabiy joined #salt
11:29 ravibhure1 joined #salt
11:31 vbabiy joined #salt
11:36 jcsp joined #salt
11:41 gnugnu9_ joined #salt
11:42 gnugnu9__ joined #salt
11:44 diegows joined #salt
11:52 hhenkel Hi all, I'm not completly sure how to use environments in salt.
11:53 hhenkel As far as I understand there is the base environment and I'm able to target servers by different methods.
11:53 budrose joined #salt
11:54 hhenkel But I'm wondering how to pin a minion to an environment.
11:55 hhenkel Or are all defined environments checked for each host? Therefore gets base first applied and then something from "dev" if there is a match?
11:55 ksk hhenkel: take a look at grains
11:56 ksk no idea about your question though
11:56 hhenkel ksk: I'm aware of grains so far....
11:57 mike25ro hhenkel: in my setup .. i use my own environments: each minion has a pillar file (based on the grains['fqdn'] and in that pillar file i have environment:office or prod or test ... and based on that i deploy things using salt -I "environment:office" salt.highstate
11:57 hhenkel ksk: I'm just thinking on how to design environments with gitfs and recognized that I'm not sure if I did understand it fully.
11:58 hhenkel mike25ro: Okay, so you would say to prefer central groups over setting the env on the minion?
11:58 mike25ro hhenkel: yeah, that is how i did it..
11:59 hhenkel mike25ro: Why not with this: http://docs.saltstack.com/topics/targeting/nodegroups.html ?
12:01 mike25ro hhenkel:  i know that there are nodes, but i prefered to have for each minion a pillar file... and i have created a web frontend for myself where i can see all minions + their environment  or i can easily add a minion to another environment
12:01 hhenkel mike25ro: Also when using the pillar, how many environments are defined in your top.sls ? Or do you only have one single base env and just seperate them with your pillar?
12:02 mike25ro hhenkel:  yes .. one base... and everything else is in the pillar
12:03 tmandry joined #salt
12:03 hhenkel mike25ro: hmm, okay...I'll try to figure out how that may work with me having gitfs and runtime mapping of branches...
12:04 Furao joined #salt
12:04 hhenkel Anyone aware if there is the possibility to see the "current view" of saltstack on the complete config tree (merged from several git repos)?
12:04 mike25ro hhenkel: i am not saying my solution is better, but after setting up 3 times and testing.. this is what i have preferred using.
12:07 Furao config tree?
12:10 hhenkel mike25ro: Wasn't judging...
12:10 hhenkel Furao: Yes, as far as I undestand salt checks out the git repo every n seconds.
12:12 hhenkel Furao: Is there a way to see the complete "picture" like salt sees it (something similar to what you have when using roots)
12:12 oz_akan_ joined #salt
12:14 hhenkel mike25ro: is your frontend available on github?
12:16 Furao hhenkel: you mean gitfs for states or pillars?
12:17 Furao the master only update references, it don't actually checkout all files, it do that only if the file is requested by a minion
12:18 hhenkel Furao: Currently states...so any idea what a good way to go is with gitfs and salt?
12:19 hhenkel As far as I understand it is a good idea to have a seperate repo for the top.sls with only a master branch.
12:19 manicouman joined #salt
12:20 hhenkel Also it seems like a good idea to seperate things in different repos.
12:21 Furao my top.sls is the only file not in gitfs :)
12:21 hhenkel But what I'm not completly sure is, how one could use a base env (where the master is mapped to) and different branches.
12:21 hhenkel It seems like one would then not use a wildcard match for base anymore, right?
12:22 Yaann joined #salt
12:22 Furao hhenkel: https://gist.github.com/bclermont/18d278439041726ee1b2
12:22 Furao this is my top.sls maybe it can help you
12:22 hhenkel Furao: So, do you pin your minions via the minion conf?
12:22 Furao I have a pillar key for the branch name in git
12:23 hhenkel Furao: thanks for the link, I guess it will take a moment to look that through... :)
12:24 Furao it's quite complicated as it cover a lot of requirements that I had
12:24 Outlander joined #salt
12:25 Outlander what's the way that salt keeps a regular/consistent state, as per puppet's agent runs every 30mins?
12:25 [diecast] joined #salt
12:31 mike25ro you can cron the  salt '*' state.highstate
12:32 malinoff Or use scheduler: http://docs.saltstack.com/topics/jobs/index.html#scheduling-jobs
12:32 mike25ro yeah malinoff i forgot about that one :)
12:37 jslatts joined #salt
12:38 baniir joined #salt
12:41 faldridge joined #salt
12:44 ndrei joined #salt
12:55 R_Shackleford joined #salt
12:56 agenteo joined #salt
12:56 agenteo hi guys, how do you do a sanity check on a salt state syntax before applying it?
12:56 ambientsound agenteo: yaml lint?
12:57 agenteo fair enough, I'll try that
12:58 agenteo a separate question about cron (http://docs.saltstack.com/ref/states/all/salt.states.cron.html) is the cron.present just a label that I can personalize?
12:59 agenteo I am getting an error on its line that I called cron.publish_article:
12:59 dln joined #salt
13:00 HeadAIX joined #salt
13:00 agenteo my bad I got confused between that and identifier
13:03 oz_akan_ joined #salt
13:05 racooper joined #salt
13:07 Doqnach joined #salt
13:08 cpenner461 joined #salt
13:09 gammalget joined #salt
13:10 mikkn joined #salt
13:13 harobed_ joined #salt
13:14 rojem joined #salt
13:18 harobed_ joined #salt
13:21 mugsie joined #salt
13:21 jrdx joined #salt
13:26 harobed_ joined #salt
13:27 harobed joined #salt
13:27 juicer2 joined #salt
13:28 budrose joined #salt
13:32 harobed joined #salt
13:33 Gifflen joined #salt
13:33 harobed joined #salt
13:34 harobed joined #salt
13:35 harobed joined #salt
13:37 harobed joined #salt
13:38 harobed joined #salt
13:39 harobed joined #salt
13:44 ajprog_laptop joined #salt
13:46 GradysGhost joined #salt
13:46 harobed joined #salt
13:47 elfixit joined #salt
13:47 mgw joined #salt
13:48 harobed joined #salt
13:52 jaimed joined #salt
13:53 harobed joined #salt
13:53 higgs001 joined #salt
13:54 lessismore joined #salt
13:54 harobed joined #salt
13:55 gcbirzan joined #salt
14:01 vejdmn joined #salt
14:02 harobed joined #salt
14:03 harobed joined #salt
14:08 mikeywaites joined #salt
14:08 harobed joined #salt
14:09 jeremyBass joined #salt
14:09 ndrei joined #salt
14:13 quickdry21 joined #salt
14:13 harobed joined #salt
14:14 Gordonz joined #salt
14:14 harobed joined #salt
14:15 Gordonz joined #salt
14:16 harobed joined #salt
14:17 harobed joined #salt
14:19 harobed joined #salt
14:21 choopooly joined #salt
14:24 rostam joined #salt
14:25 danielbachhuber joined #salt
14:28 JesseCW joined #salt
14:28 harobed joined #salt
14:29 xmltok joined #salt
14:30 harobed joined #salt
14:30 fllr joined #salt
14:32 homelinen joined #salt
14:34 pcarrier joined #salt
14:35 harobed joined #salt
14:36 harobed joined #salt
14:39 yomilk joined #salt
14:42 rwazar joined #salt
14:43 rwazar New to ss... in declaring a state for an installed package is it possible to specify the rpm name & location?
14:43 harobed joined #salt
14:43 rwazar or rather a pkg to be installed
14:44 jankowiak joined #salt
14:47 pass_by_value joined #salt
14:47 harobed joined #salt
14:48 seapasulli joined #salt
14:49 harobed joined #salt
14:49 bbinet rwazar: you mean the "sources" option of pkg.installed?
14:49 baniir joined #salt
14:49 bbinet http://docs.saltstack.com/ref/states/all/salt.states.pkg.html#salt.states.pkg.installed
14:50 MTecknology Do any of you guys use salt for managing static network addresses?
14:50 yes456 joined #salt
14:50 rwazar Thanks bbinet, I'll give that a look
14:51 bezaban I do, but using static leases, on the ones I don't want configured using dhcp I've just been avoiding salt
14:51 bbinet MTecknology: yes I used network.managed rencently
14:53 ccase_ joined #salt
14:53 CeBe joined #salt
14:54 rwazar Thanks bbinet... The "sources" option on that doc page is exactly what I was looking for.
14:55 skynet joined #salt
14:55 MTecknology bbinet: How well does that scale to managing >500 servers across 200 subnets?
14:56 MTecknology and... some interfaces that have multiple vlan interfaces hanging off of them?
14:56 bbinet MTecknology: sorry, I can't answer to this, I've only a little experience with network.managed on few machines
15:00 mrchrisadams left #salt
15:00 yes456 salt 'abc.com' state.sls syslog-ng.install  env='stage' saltenv='stage'  when i run this file client fecthing file from base env.  for log & sls file http://pastebin.com/kC4x5jWL
15:01 ipmb joined #salt
15:02 wendall911 joined #salt
15:02 dangra joined #salt
15:03 timoguin joined #salt
15:03 jslatts joined #salt
15:05 jalbretsen joined #salt
15:06 zirpu joined #salt
15:06 pydanny joined #salt
15:08 opapo joined #salt
15:15 ccase_ joined #salt
15:15 arapaho joined #salt
15:16 teskew joined #salt
15:17 ndrei joined #salt
15:17 hunter joined #salt
15:21 dave_den joined #salt
15:23 bitmand I like Salt :)
15:23 helderco joined #salt
15:24 rwazar Just now learning Salt... installed it on a few nodes last week
15:29 seapasulli joined #salt
15:30 derelm joined #salt
15:32 baniir joined #salt
15:34 ChaosPsyke joined #salt
15:36 ndrei joined #salt
15:36 zirpu salt and chocolate, yum yum.
15:36 favadi joined #salt
15:36 timoguin that's a good combo
15:37 vimalloc I'm having problems getting mysql_user to utilize the defaults file. I have a file on the minion (/etc/salt/minion.d) which has "mysql.default_file: '/etc/mysql/debian.cnf'", verified that the creds in debian.cnf are correct, but when I try to do a mysql_user.present I get the error "   MySQL Error 1142: SELECT command denied to user 'salt'@'<ip>' for table 'user'. Any ideas?
15:38 vimalloc (17.5 for salt)
15:38 bhosmer_ joined #salt
15:38 [diecast] joined #salt
15:41 mgw joined #salt
15:42 forrest joined #salt
15:42 ndrei joined #salt
15:44 nix_ joined #salt
15:45 lessismore I am just starting with salt formulas, the template-formula repo seems wrong to me: https://github.com/saltstack-formulas/template-formula/blob/master/template/map.jinja should end with merge=salt['pillar.get']('template')) %} instead of merge=salt['pillar.get']('template:lookup')) %} according to the pillar example provided: https://github.com/saltstack-formulas/template-formula/blob/master/pillar.example or am I missing something?
15:45 conan_the_destro joined #salt
15:46 bhosmer joined #salt
15:48 forrest lessismore, no you're correct.
15:48 davet joined #salt
15:49 forrest lessismore, the pillar doesn't seem to have been updated when the map file was.
15:49 forrest lessismore, in this case the pillar is incorrect, not the map file.
15:49 frasergraham joined #salt
15:50 yes456 salt 'abc.com' state.sls syslog-ng.install  env='stage' saltenv='stage'  when i run this command file client trying fecth file from base env.  for log & sls file http://pastebin.com/kC4x5jWL
15:51 smcquay joined #salt
15:51 forrest I fixed it real quick lessismore
15:52 lessismore forrest: thanks, I see that also in the doc there is a mention of "lookup": http://docs.saltstack.com/topics/conventions/formulas.html#map-jinja, so basically the "lookup" is only a sort of optional namespace
15:52 forrest correct, we use it specifically for stuff related to the map
15:52 forrest you can put additional items under it or in other areas, but it's just a standard since we know that stuff is getting pulled from somewhere.
15:52 mgw joined #salt
15:53 lessismore forrest: thanks again for the clarification
15:53 forrest lessismore, np
15:55 xmltok joined #salt
15:57 [diecast] joined #salt
15:57 higgs001 joined #salt
15:58 rallytime joined #salt
15:59 Gareth morning.
16:00 forrest hey Gareth
16:00 utz joined #salt
16:01 mpanetta joined #salt
16:04 pvial joined #salt
16:05 pvial hello, anyone ever run into this error when trying to create a VM with Salt-Cloud v0.8.11?  TypeError: string indices must be integers, not str
16:05 tmandry joined #salt
16:05 Gareth forrest: hey. hows it going?
16:06 forrest Gareth, starting to heat up over here! How about yourself?
16:06 rysch joined #salt
16:07 Gareth forrest: not bad :)
16:08 KyleG joined #salt
16:08 KyleG joined #salt
16:08 forrest Gareth, yea, I'm sure having SCALE finished is nice
16:08 forrest you've been able to recover :P
16:08 KyleG I wish I could've gone :\
16:09 KyleG I live near LAX and everything lol
16:09 KyleG But I was busy being a sys admin that weekend :(
16:09 KyleG Duty calls
16:10 mpanetta joined #salt
16:10 pvial are there plans for salt-cloud to move away from libcloud?  every error i seem to ever get is from libcloud and salt-cloud guys can't do much about that.
16:11 mgw joined #salt
16:12 bastion1704 joined #salt
16:13 forrest pvial, I have no idea, you'd probably want to make a post on the mailing list about that since the main guy who works on it isn't in this channel.
16:13 forrest KyleG, That's why you have tethering!
16:13 Saint-N joined #salt
16:14 KyleG tis' true…I have this.
16:14 KyleG lol
16:15 pvial forrest, thanks but i haven't had much luck with the mailing list on previous attempts.  never any luck on IRC too.
16:15 forrest pvial, yea it happens, sometimes questions slip through the cracks, or we just don't have an answer!
16:16 Saint-N thats cuz yer jerks forrest  ;)
16:16 Saint-N heh
16:16 pvial gotta switch to ansible
16:16 forrest Saint-N, yea I know, it's like I work a regular job :P
16:16 Saint-N madness! that is unpossible
16:17 forrest pvial, that's ok, if one tool works better for you than other no reason not to use it
16:17 bemehow joined #salt
16:18 pvial forrest, it's necessarily the tool...it's the support around the tool
16:18 Saint-N free tools have a way of having that flaw
16:18 [diecast] joined #salt
16:18 Saint-N usually the irc rule works for it though ;)
16:18 Gareth forrest: yeah.  good to be able to relax a bit :)
16:18 Saint-N if someone asks a question answer it with something blatantly wrong
16:19 Saint-N nerds cant help themselves but to one up and provide the real answer
16:20 timoguin very true
16:21 dave_den pvial: yes, i believe the goal is to move away from libcloud eventually, but i don't have an official word on that
16:21 rwazar lol true words
16:21 forrest pvial, in what sense? We've put quite a bit of work in the documentation, but it's still a free and open source piece of software, so sometimes you have to dig in.
16:21 forrest Gareth, yea I'm sure.
16:22 forrest timoguin, I saw your post on reddit last night about the Windows support. As a heads up I have this issue: https://github.com/saltstack/salt/issues/9066 open regarding that.
16:22 gnugnu9 joined #salt
16:22 hunter joined #salt
16:23 forrest timoguin, I'll bug Dave about it again when I see him :P
16:23 timoguin forrest, nice. just subscribed.
16:23 forrest be like THE INTERNET WANTS THIS DAVE, MAKE IT HAPPEN!
16:23 pvial forrest, i think it's just my bad luck.  every IRC post i've ever made goes unanswered.  my guess is this is due to no one knowing the answer and always figure it out anyway.
16:23 forrest pvial, yea sometimes we just don't know, or there's no one around that does know.
16:23 musinsky_ joined #salt
16:24 harobed joined #salt
16:25 Philip joined #salt
16:26 dave_den pvial: what questions have you not been able to find answers?
16:27 pvial dave_den, i've been able to answer my own questions, eventually.  the one i'm stuck on now is: TypeError: string indices must be integers, not str (when trying to create a VM with salt cloud)
16:27 lessismore still starting with formulas, I add https://github.com/saltstack-formulas/template-formula.git in the master conf, restarted the master, but  salt -v 'beta' state.sls template ends with No matching sls found for 'template' in env 'base', in /var/cache/salt/master/gitfs/<hash> I can see the git checkout
16:28 lessismore any hints?
16:28 pvial googling shows that it was fixed in some devel branch release but the posts weren't specific.  i'm running 0.8.11
16:28 dave_den pvial: what version of salt?
16:29 pvial 0.14.0
16:29 pvial dave_den: i'm guessing that could be my problem
16:29 dave_den haha, yeah
16:29 timoguin yea
16:29 musinsky_ hi all, im getting an HAVP error from the python-zmq and salt-common packages from the ppa, which is preventing salt from installing correctly
16:29 musinsky_ http://ppa.launchpad.net/saltstack/salt/ubuntu/pool/main/s/salt/salt-common_2014.1.0-1precise1_all.deb
16:29 musinsky_ anyone able to resolve that?
16:30 pvial when does salt become 1.0?  i'm tired of upgrading all the time...not so good for production environments.
16:30 dave_den pvial: i would recommend 0.17.5.
16:30 pvial dave_den:  i'll give it a shot and pray it doesn't break my prod environement.
16:30 dave_den pvial: i would test it in dev first
16:30 cachedout joined #salt
16:31 timoguin yea, 0.14 to 0.17 is a nice jump.
16:31 dave_den there will not be a 1.0 version. the new version scheme is date based - to which there has been a bit of controversy
16:31 pvial interesting...
16:32 dave_den the first release in that scheme being 2014.1.0 which was released last week
16:33 forrest musinsky_, let's see if joehh is around.
16:33 lessismore I have to leave, will start fresh tomorrow debugging formulas!
16:33 forrest musinsky_, he does the Ubuntu packaging.
16:33 dave_den and salt-cloud has been integrated into the salt code in 2014.1.0
16:33 musinsky_ forrest: thanks
16:33 pvial thats good to know.  thanks for the headsup.
16:34 forrest musinsky_, yea he lives in Australia, so it's 3:30 AM right now, doubtful he is around, but he might be later today.
16:34 dave_den pvial: but like i said, i would recommend going to 0.17.5 with libcloud 0.8.11
16:34 dave_den errl slt-cloud
16:35 Philip got a tricky one today you guys. When I call salt-call from jenkins, everything is coming back with a return code of 1, minion running as root, jenkins user has sudo/NOPASSWD set
16:35 forrest Philip, from within a Jenkins job?
16:35 mgw joined #salt
16:35 Philip forrest: yeah, but also if I just su jenkins and call anything
16:36 Philip jenkins ALL=(ALL) NOPASSWD: ALL in sudoers for now (gonna lock it down to just salt-call in a bit once i debug this)
16:36 Philip running salt-call from just root works, as does from any other user account
16:36 forrest Philip, so are you trying to run this from a job, or from the jenkins user directly?
16:37 Philip both dont work right now
16:37 Philip I'm testing as the jenkins user to try and debug where this is happening
16:37 Philip where it also fails
16:37 forrest Philip, what happens when you pass -l debug to the command you're trying to use?
16:38 Philip I've been trying to track it down that way, basically everytime the minion shells out, it gets a return code of 1
16:38 forrest odd, can you restart the minion with debug logging enabled to see if anything else gets pushed into the logs?
16:39 Philip yeah, I've got them all on debug by default
16:39 forrest but nothing in there is hinting at the issue?
16:39 arodizzy joined #salt
16:39 Philip no, just the "return code was 1"
16:40 Philip but if I copy those same commands and run them right from the shell, they return 0 and I get the normal command output
16:40 forrest as the jenkins user?
16:40 Philip as the jenkins user, but using sudo since the minion is running as root
16:40 yomilk joined #salt
16:41 Philip that's the strange part, why would the minion behave any differently depending on the user making calls to salt-call?
16:41 joehillen joined #salt
16:41 forrest Gotcha, ok so what I'd do in that case, as much as it sucks is update your jenkins job that is failing to include an strace in there and output to a file, then strace from the command line on the call, then compare those two files.
16:41 forrest Philip, well, maybe the way it interacts with the shell is different, interactive sessions and such?
16:42 forrest I know people have had issues with that before.
16:42 Philip forrest: I'll give strace a try for this one, but I thought I'd ask in here in case there was a simple "ah well it's because ..." that I was overlooking
16:43 forrest Philip, yea I'm not familiar enough with how the jenkins jobs work to provide an answer :\ Is there seriously nothing in the jenkins job saying why it fails?
16:43 joehillen joined #salt
16:44 Philip well I can get it to fail sans-jenkins, just running say... sudo salt-call state.single pkg.installed name=git
16:44 Philip running that as the jenkins user fails, but works for any other user on that machine
16:44 forrest Philip, oh I thought you were saying it worked from the command line as the jenkins user
16:44 Ryan_Lane joined #salt
16:44 forrest ok yea, then strace the jenkins user, versus the working user :\
16:45 Philip yup, will do
16:45 CaptTofu joined #salt
16:46 [diecast] forrest i have jenkins user with salt-call
16:46 [diecast] you have to allow access to a few directories in visudo
16:46 forrest [diecast], sweet!
16:46 Philip [diecast], I have entries in sudoers for the jenkins user, and had this working on my old master
16:46 Philip which is why this is so strange to me right now
16:47 [diecast] it depends on how restrictive you are
16:47 [diecast] did you give full NOPASSWD: ALL
16:47 [diecast] my guess is forrest did not and you did
16:47 Philip [diecast] yeah, im gonna lock it down to just salt-call as soon as this gets debugged
16:47 Philip jenkins ALL=(ALL) NOPASSWD: ALL
16:48 forrest [diecast], I'm not using jenkins for salt right now at all :P
16:48 [diecast] let me see if i have it
16:48 [diecast] forrest oh, nm.. then ;)
16:48 forrest yea just trying to help out Philip
16:48 [diecast] gotcha
16:48 [diecast] jenkins ALL=NOPASSWD:/opt/saltwrapper, /var/log/salt, /var/cache/salt, /etc/salt
16:49 [diecast] prod/salt/saltcall.sls
16:49 utz joined #salt
16:49 [diecast] one sec
16:49 Philip [diecast] yeah even at the most permissive, when salt-minion shells out everything is returning 1
16:49 [diecast] http://pastebin.com/abXPXfSZ
16:50 Philip [diecast] that's the problem im trying to debug now, not getting sudo to work, that's working fine
16:50 Philip thanks for the help so far [diecast] and forrest, gonna grab some food real fast
16:50 [diecast] oh
16:53 bram__ joined #salt
16:54 bastion1704 joined #salt
16:55 bram__ hi all! salt noob with noob question: I've started using salt -successfully- and I get how SLS'es etc work... I'm however still confused about the different states (low, high, over, ...) and what they exacly mean. I've studied the manual about the states, but I'm STILL confused... can someone either point me to a (simpler??) explanation or a simple example which compares/uses all of these?
16:55 baniir joined #salt
16:56 Iwirada left #salt
16:56 timoguin bram__, did you look at the doc that describes the layers of the state system?
16:56 timoguin this one: http://docs.saltstack.com/ref/states/layers.html
16:56 bram__ timoguin: I did and it confused me even more... it's all exaplined quite metaforically :/
16:56 timoguin it explains things kinda backwards, and it needs a solid diagram
16:57 bram__ imho it needs an example of when one would use each of the states and what you would use them for and how
16:58 bram__ s/states/layers
16:58 timoguin agreed. everything below the SLS layer is generally only needed for debugging. people are probably most concerned with SLS and Highstate
16:58 bram__ and Over as well...
16:58 forrest timoguin, where's the link to your presentation about the layers?
16:59 timoguin overstate is actually being deprecated in favor of Remote Control States
16:59 ndrei joined #salt
16:59 bram__ timoguin: oh no! don;t confuse me even more :-)))
16:59 robbie joined #salt
16:59 harobed joined #salt
17:00 timoguin bram__, i'm only saying that because i don't want you to get too invested in an Overstate setup
17:01 bram__ I'm only interested in these bits because I'm trying to set up vagrant with salt provisioning - just as a test to see if I understand the salt-bits correctly
17:02 timoguin forrest, bram__, http://slid.es/timoguin/pytn2014-salt
17:02 vejdmn joined #salt
17:02 timoguin slide #10 is where i break down the state layers.
17:02 timoguin but i don't think my slides are much better than the salt docs
17:03 bram__ lemme go through (all) the slides, I'll get back to you if I still don;t understand it :)
17:03 timoguin i think explaining highstate first and then working down would be more effective.
17:03 timoguin right now it's organized like, "here's the lowest level of abstraction you won't ever use. let's start here."
17:05 bram__ timoguin: your super simple examples of python modules with source just made a LOT of sense to me - I fully believed that making your own salt task would be a daunting task
17:05 timoguin good! i wanted to show how simple it is.
17:06 harobed joined #salt
17:07 tmandry joined #salt
17:08 utz left #salt
17:08 rupsky joined #salt
17:08 vejdmn joined #salt
17:08 utz joined #salt
17:09 penguin_dan joined #salt
17:13 zach_ Out of the blue salt stopped seeing my state files in /srv/salt
17:14 elfixit joined #salt
17:14 nix_ joined #salt
17:14 forrest zach_, so you're saying you broke it?
17:16 zach_ i'm saying i dont understand why that's happening
17:16 zach_ base is normal /srv/salt
17:16 schimmy joined #salt
17:16 zach_ I have a test.sls file in there
17:16 zach_ Oh...seriously....now it works....
17:17 zach_ guess my minions were all on this weird 2014 version naming scheme and master was 0.17.x
17:17 AviMarcus joined #salt
17:17 forrest did you update the master?
17:17 bhosmer joined #salt
17:18 zach_ yeah
17:18 zach_ just now
17:18 schimmy1 joined #salt
17:19 n8n joined #salt
17:19 amckinley joined #salt
17:20 bram__ timoguin: "insert diagram" ?
17:20 timoguin bram__, lol forrest made sure to point that out too. i didn't have time to create one, so i made a joke instead.
17:20 timoguin but we really need a good visual for that
17:21 sporkd2 am I missing something on my service.running?  "Comment: Service nginx is already enabled, and is dead"
17:21 sporkd2 ^ shouldnt service.running start that?
17:21 sporkd2 when I run test=True I get "Comment: Service nginx is set to start"
17:22 dln left #salt
17:25 CaptTofu HI all!
17:25 CaptTofu I just submitted a pull request
17:26 bhosmer_ joined #salt
17:29 gkze joined #salt
17:29 bram__ timoguin: I just got passed your slides about the definition of the states and I'm really wondering about the naming of them and the explanation in the docs... it's.... stragely confusing
17:30 forrest CaptTofu, great! The devs are usually pretty quick to review and accept or comment.
17:30 CaptTofu forrest: more to come!
17:30 timoguin bram__, yea you're not alone. do you have a specific part you're confused about?
17:30 bram__ timoguin: your slides explain it slightly better, but imho super simple examples for each would be THE way to go!
17:31 timoguin yea I demo'd them during my presentation, so there's not much on the slide portion
17:31 billyogat joined #salt
17:31 bram__ one thing I'm confused about: is the highstate always defined in a file called top.sls or is this just convention?
17:32 bram__ I suppose the latter
17:32 timoguin bram__, you can define a different file, or you can fetch top data from an external system.
17:32 timoguin but top.sls is the standard
17:32 logandg joined #salt
17:32 gkze hey guys, I'm trying to use GitFS for Salt states and files, and ext_pillar (git_pillar) for pillar. I have one repo, and I have indicated in the config to use the states subfolder for states, and pillar subfolder for pillar like this, in the config: https://gist.github.com/gkze/9490774
17:33 bram__ so then why isn't it called high_state.sls ? :-)
17:33 blee joined #salt
17:33 gkze However, when I try to run a highstate, I get an error saying 'AttributeError: 'list' object has no attribute 'endswith''
17:34 gkze can anyone help? Does anyone have GitFS working?
17:34 timoguin bram__, maybe it should be. :)
17:35 bram__ timoguin: actually that would kind of simplify things :-) so a highstate is an SLS where the states are hardwired to certain minions (using matching)... correct?
17:36 timoguin right. and it can define the environments as well.
17:37 druonysus joined #salt
17:37 druonysus joined #salt
17:38 bram__ environments?
17:40 timoguin bram__, yes, the salt fileserver can define a different root directory for specific environments (dev, qa, etc.). if minions are assigned to that environment they'll fetch their files from there.
17:41 timoguin by default /srv/salt is the 'base' environment, so you'll probably wanna just stick with that for the time being
17:41 bram__ alrighty!
17:41 timoguin http://docs.saltstack.com/ref/states/top.html
17:41 rupsky joined #salt
17:42 erchn joined #salt
17:43 rgbkrk joined #salt
17:43 bram__ timoguin: thx!
17:43 erchn hi all, I'm attempting to use Salt's mine system with a reactor to auto configure one of our applications, anyone have experience with such a system around?
17:44 bram__ timoguin: now... I know you said it was going to be deprecated, but what does the overstate than mean?
17:44 mikkn Is it possible to reload grains during a highstate? Especially during the first highstate, we push a few grains that are needed for some later states. Just adding a watch to the salt minion doesn't seem to cut it...
17:44 erchn I'm not having trouble necessarily fetching data or triggering actions based on reactor events, but more on the issue of a node going away and the mine cache not know of the change
17:45 mikkn erchn: Maybe you could trigger a mine refresh when a node goes away?
17:45 mikkn Or skip the mine and use state calls directly from other minions
17:46 timoguin bram__, overstate allows you to orchestrate system dependencies like configuring the database server before the web server: docs.saltstack.com/ref/states/overstate.html
17:46 erchn mikkn: I tried forcing mine.update to run but since the minion is down it's not clearing it's own data, so all the other minions still see it's data in the cache
17:46 timoguin think of top.sls / highstate as system level. and then overstate is the next step up at the multi-machine level
17:47 dave_den mikkn: yes, use saltutil.refresh_modules
17:47 erchn My next thought is to enable the presence reactor events and on a node going away just clear all mine data from all nodes, then ask for a mine.update
17:47 mikkn erchn: look into http://docs.saltstack.com/ref/peer.html
17:47 dave_den mikkn: use the state.module state to run saltutil.
17:47 vejdmn joined #salt
17:48 erchn maybe an easier way though...I could use the peer system, yea...hmm
17:48 mikkn dave_den: Will the data be available for the states afterwards in the same highstate run?
17:48 mikkn erchn: I'm facing the same problem and I'm going back and forth too
17:48 erchn I think what's missing from the mine system is a way to tell all minions to clear mine data about a given node
17:49 dave_den mikkn: if you're using jinja then the jinja is evaluated at compile time.
17:50 bram__ timoguin: OK! I think I got it. Thanks to you for being awesome. Last question: is the talk you gave with those slides available somewhere? as in a video recording or audio...?
17:50 timoguin bram__, unfortunately not. the con didn't have the budget for video this first time.
17:50 timoguin recordings from SaltConf should be released before too long though
17:51 bram__ timoguin: you did the talk twice?
17:51 timoguin no, just once at PyTN, but it looked like there were a lot of good talks at SaltConf
17:53 bram__ aaah okay
17:53 bram__ well thx anyway...
17:53 rupsky joined #salt
17:54 hhtpcd joined #salt
17:55 mikkn dave_den: Ah, that's what I suspected. Is it common not to use jinja?
17:55 erchn mikkn: peer system it is, at least it's reliable
17:55 erchn even if slow
17:55 mikkn erchn: Do you have any offline minions?
17:55 erchn I can, yes
17:56 erchn ideas?
17:56 mikkn erchn: That slows it down significantly. There is a ticket up for it, I think they want to merge the roster code for salt-ssh with the normal zmq-powered salt, and keep track of which minions are offline and not.
17:57 mikkn erchn: Right now the problem is that it isn't being kept track of, so every time you trigger a command it needs to wait for the timeout...
17:57 naiyte joined #salt
17:58 erchn OIC, actually on my system it takes 6 seconds regardless
17:58 dimeshake joined #salt
17:59 mikkn erchn: There might've been more to it, I skimmed the issue and pegged it as something I would like to have fixed at least. :)
17:59 erchn alright, thanks for the help, have a good one
17:59 amallais joined #salt
17:59 amallais hey!, What kind of cpu usage should I expect on a salt-master ?
18:00 mikkn erchn: No problem!
18:01 mikkn erchn: https://github.com/saltstack/salt/issues/2361 is the related issue
18:02 ipalreadytaken joined #salt
18:03 amallais my state.highstate did not return any output
18:03 amallais cpu usage on master is hight
18:03 amallais but logs on minion shows that jobs are running
18:04 amallais got lots of "saltreqtimeouterror" of minions
18:04 erchn mikkn: thanks.  looks like a long-standing issue that may not get resolved anytime soon
18:06 mikkn erchn: Not very soon, but I will probably look into it myself in a couple of months if it isn't resolved by then - ZMQ seems to be giving the information required rather recently, so it should be "easier" now. I think it's a rather high-priority issue. :)
18:08 ml_1 joined #salt
18:08 harobed_ joined #salt
18:09 musinsky_ forrest: turns out the ppa is fine, its just my network here at the office which is all kinds of screwed up. thanks for the help though
18:09 gadams999 joined #salt
18:10 mikkn amallais: http://docs.saltstack.com/topics/troubleshooting/master.html#commands-time-out-or-do-not-return-output and https://github.com/saltstack/salt/issues/8647
18:10 hunter joined #salt
18:11 Gordonz joined #salt
18:11 amallais yeah
18:11 amallais already put the timeout at 60
18:11 forrest musinsky_, ahh great!
18:11 forrest joehh, you're free if you see this :P
18:13 amallais if i understand https://github.com/saltstack/salt/issues/8647 correctly
18:13 amallais we have to put very high timeout ?
18:14 erchn mikkn: I wonder if the info already exists in the master, since the introduction of "presence_events"...just isn't used in some places
18:15 agenteo joined #salt
18:17 mikkn amallais: Yeah, but running into the timeout isn't a problem in itself. The problem is to get the results of the run, for me that hasn't been a big-enough problem to find out the solution to yet.. Maybe someone else knows how to do that...
18:18 mikkn erchn: Maybe? I'm not sure. :) I haven't looked into presence_events at all, actually.
18:18 sroegner joined #salt
18:18 amallais I see
18:19 mikkn erchn: Have you tried presence_events?
18:19 hunter Is there a way for one minion to "see" pillar data for all the minions? My use case is to generate kickstart files for all the minions and I need [ ip, arch, some, key, words] etc to generate teh correct kickstart.
18:19 erchn mikkn: I'm still on 0.17.x so it's not available yet
18:19 mikkn erchn: Ah, I see
18:19 hunter Is that what the mine does? Allows _some_ pillar data to be visible to other minions?
18:20 timoguin hunter, you could use the mine or peer systems for that.
18:20 timoguin they both allow you to call specific functions and grab the output
18:20 hunter timoguin: is the mine or peer systems in 0.17.5?
18:20 timoguin yes, both
18:20 hunter And whats the decision tree I should use to choose between them?
18:21 hunter Oh wait - both of those would require the minion to be "up" on the client machine - and I need the kickstarts generated to INSTALL them.
18:21 hunter timoguin: or do one of them allow _some_ pillar data through, BEFORE a minon exists?
18:22 timoguin ah, that sounds like it'd be better as a runner. it only needs to run on the master to generate the kickstart files right?
18:22 dave_den hunter: create a runner and use salt.utils.MasterPillarUtil. allow your kickstart runner to run it by configuring peer_run in the master config.
18:22 timoguin yea that
18:22 Teknix joined #salt
18:23 hunter ok, that introduces about 3 new concepts to me. :)
18:23 hunter I can find the runner docs and I'll read that - I assume peer_run is something related to that I'll find in the docs?
18:23 dave_den yes
18:24 dave_den you'll hve to read the code for masterpillarutil though
18:24 dave_den it's easy to use and i documented some uses in the code.
18:24 rainmanjam joined #salt
18:25 dave_den although now that i think about it, i don't think it will let you get the pillar data without the grains of the minion
18:26 dave_den you might as well just compile the pillar directly in the runner and provide an empty grains hash
18:26 dave_den but that is also pretty easy to do
18:27 n8n joined #salt
18:33 hunter holy moley
18:34 hunter Ok - I'll circle around and look at this in a couple of days (I'm exhausted and about to go home as a sick day)
18:34 hunter dave_den: I'll ask questions once I've had a chance to read the runner docs on the site.
18:34 dave_den cool
18:35 gkze has anyone gotten GitFS to work?
18:36 nahamu gkze: what error are you seeing?
18:36 bhosmer joined #salt
18:36 Ahlee returners are loaded locally, but config is pushed from the master, right?  So if I change a returner, I have to push it to all minions, I can't just update it on the master and expect the next returner call to sync it down
18:36 Ahlee locally to minions, that is
18:37 gkze nahamu: TypeError: string indices must be integers, not str
18:37 timoguin Ahlee, I think that's correct.
18:37 bastion1704 joined #salt
18:38 gkze master log - "AttributeError: 'list' object has no attribute 'endswith'"
18:38 timoguin gkze, that may be an error in your highstate somewhere. try seeing if you can apply specific states with state.sls
18:38 timoguin I'm using gitfs just fine with a single base environment. states are in one repo and pillar is in another
18:40 gadams999 left #salt
18:40 gkze I have GitFS roots mapped to different repo branches
18:40 gkze base only has tops, branches have states
18:40 gkze both pillar and states
18:41 gkze so in tops I have no base: block
18:41 gkze only dev: and prod:
18:41 gkze I think that doesn't seem to work
18:44 whidbeywalker joined #salt
18:46 jalaziz joined #salt
18:47 zz_Cidan guys, I have a poem
18:47 zz_Cidan oh salt, how I love thee
18:47 zz_Cidan except when your regressions stop my work in it's fucking tracks
18:47 zz_Cidan and I can't deploy new servers
18:47 zz_Cidan then I want to roll up into a ball
18:47 zz_Cidan and cry
18:48 zz_Cidan the end.
18:48 timoguin pfftt, doesn't even rhyme
18:48 zz_Cidan it's new age
18:48 gkze free verse
18:48 gkze we got a walt whitman of Salt over here
18:48 amallais zz_Cidan: \o/
18:48 Linz joined #salt
18:48 timoguin yo enjambment's week, brah.
18:51 goodwill zz_Cidan: nice haiku
18:52 zz_Cidan thanks
18:52 kzx joined #salt
18:52 raizyr joined #salt
18:53 lahwran joined #salt
18:53 roolo joined #salt
18:55 amallais you guys run a lot of minions ?
18:55 ksalman how come this needs to happen if i install an older version (0.17.2) https://gist.github.com/ksalman/9492498
18:55 ksalman when i install the latest version, 2014, it doesn't install stuff from unstable
18:56 ksalman i am installing the older version like so
18:56 ksalman wget --no-check-certificate -O - http://bootstrap.saltstack.org | sh -s -- -P git v0.17.2
18:56 timoguin ksalman, what os/distro are you running on?
18:57 ksalman timoguin: Debian Wheezy
18:57 ksalman i need to install 17.2 but it breaks stuff by installing lots of things from unstable, as stated in the warning message :P
18:57 racooper joined #salt
18:57 timoguin my guess is that the pyzmq version that's packaged on stable is older, but i'm not sure.
18:58 ksalman but i am wondering why that happens with the old versioin only
18:58 timoguin it doesn't happen if you install 2014.1?
18:58 timoguin on the same system?
18:58 ksalman nope
18:58 ksalman ya
18:58 ksalman it was working fine until salt was upgrded upstream it seems
18:58 pass_by_value joined #salt
18:58 timoguin yea i'm not sure. gotta meeting...
18:59 ksalman i was installing 17.2 on debian wheezy on multiple hosts using
18:59 ksalman wget --no-check-certificate -O - http://bootstrap.saltstack.org | sh -s
18:59 ksalman previously
18:59 ksalman and it did not install anything from unstable
19:00 zirpu left #salt
19:01 gadams999 joined #salt
19:02 vejdmn joined #salt
19:04 hunter I do not understand the changes to https://github.com/saltstack-formulas/template-formula/blob/master/template/map.jinja
19:05 hunter Why is template / default there?
19:07 utz joined #salt
19:07 lahwran joined #salt
19:08 meteorfox joined #salt
19:08 mrigoryan joined #salt
19:09 derelm joined #salt
19:15 bhosmer joined #salt
19:15 schimmy joined #salt
19:17 dogear joined #salt
19:18 schimmy2 joined #salt
19:20 kzx in a windows salt minion I need to put a zipped folder, that I want unzipped on the target (minion) what is the best way to do it with salt ?
19:20 harobed joined #salt
19:21 harobed joined #salt
19:25 helderco joined #salt
19:25 budrose joined #salt
19:26 pydanny joined #salt
19:27 maxleonca joined #salt
19:28 maxleonca hello
19:28 Corey Hello.
19:30 maxleonca can someone please point out if I'm wrong, I have a formula that applies to all servers but 2, so I have setup the file exceptions in the init file as  "soruce: salt://files/config.{{ grains['fqdn'] }}" and "source: salt://files/config" for both the exceptions and in general, but doesn't seem to work.
19:30 maxleonca so I have to assume I'm doing something wrong.
19:30 ndrei joined #salt
19:31 maxleonca and yes there are 3 files the general one, and the 2 with .$server.my.domain, where $server.my.domain are the exceptions.
19:32 rysch and how are you choosing between the different sources in the init file?
19:33 maxleonca well, I list the files setting the .fqdn on top, it it doesn't match it takes the one below
19:33 oz_akan_ joined #salt
19:34 maxleonca the formula works for the bulk but not for the exception.
19:35 mikeywaites joined #salt
19:36 rysch if this was a fairly static thing, I'd do
19:36 rysch {% if grains['host'] in ['hostname1', 'hostname2'] %}
19:36 rysch - source: salt://files/config.{{ grains['fqdn'] }}
19:36 rysch {% else %}
19:36 rysch - source: salt://files/config
19:36 rysch {% endif %}
19:37 rysch or you could used grains['fqdn'] and fully qualified names for consistency, but same idea
19:37 kermit joined #salt
19:38 maxleonca true
19:39 maxleonca I'll try to work it out that way, thanks
19:39 slacker joined #salt
19:39 ze- hi. is there a full documentation of what is required to get tests running ?
19:40 thedodd joined #salt
19:40 ze- I have a pull-request i'd like to do, but i'd rather have the code tested before
19:40 ipalreadytaken joined #salt
19:40 slacker joined #salt
19:41 gkze joined #salt
19:41 slacker hi everyone.
19:42 slacker does anybody know what 'Requisite delacarion in SLS in not formed as a single key dictionary' mean?
19:42 slacker i have validated my sls file and it is valid YAML
19:42 sashka_ua hello guys, I have a question. What if I want pass some data to state through API, how would I do that? besides pillar, I'd like to call the state.sls with some additional variables, using them inside of jinja
19:43 schimmy joined #salt
19:43 gkze joined #salt
19:44 vasyajap joined #salt
19:44 ze- slacker: I would say at some point, you have multiple key dict, where you should have a single-key ? without context, I don't have enough SLS syntax in mind to help.
19:44 n8n joined #salt
19:45 slacker here is the section its complaining about:
19:45 slacker deploy_source:   cmd.run:     - require:       - file: /usr/local/src/git.zip     - name: 'cd /usr/local/src && unzip git.zip && git checkout -f prototype'     - user: root     - group: root
19:45 slacker crap
19:45 slacker deploy_source:
19:45 sashka_ua use pastebin
19:46 erchn ze-: did you look here? http://docs.saltstack.com/topics/hacking.html#running-unit-and-integration-tests
19:46 slacker http://pastebin.com/VAiRYmen
19:46 Yaann joined #salt
19:48 slacker could you explain what you mean by a multiple key dict please?
19:48 [diecast] This minion received a job where kwargs were passed as string'd args
19:48 [diecast] https://github.com/saltstack/salt/issues/10503
19:49 ze- erchn: thanks. the online doc includes a "See here for more informations", which my HACKING.rst did not :)
19:49 Philip joined #salt
19:49 erchn ze-: np
19:49 erchn slacker: I think it looks like a bug, you're state looks fine, to me
19:49 Philip forrest: finally figured this out: on jenkins install, it's shell was set to /bin/false
19:50 erchn your*
19:50 Philip that ended up being a pain to find
19:50 slacker thanks erchn. It could not have anything to do with the required managed file?
19:51 erchn slacker: I'm assuming you have a id declaration matching that name
19:51 mgw joined #salt
19:51 slacker http://pastebin.com/HqPvqU8y
19:52 Philip [diecast] meant to ping you on that last reply too, it was a default shell problem
19:52 Philip in case that one ends up biting you one day
19:52 [diecast] ah =)
19:52 erchn slacker: looks right to me, this is on what version of salt?
19:52 [diecast] ya, i sometimes forget about that
19:52 [diecast] it's important to run the task as the user
19:52 [diecast] usually uncovers issues unseen via sudo and such
19:52 slacker erchn: 2014.1.0
19:53 erchn slacker: your minion and master are the same version?
19:53 ipalreadytaken joined #salt
19:53 slacker erchn: yep
19:53 slacker both running on Ubtuntu 12.04 machines.
19:53 slacker 12.04 minimal
19:54 Philip [diecast] yup, this one was just so weird, I never even think of the default shell not being bash
19:54 slacker erchn: wild idea, but could line endings have anything to do with it?
19:54 slacker the sls file has been modified on a windows machine too
19:54 Philip I still dont understand why the calling user's shell would affect what the minion is doing
19:54 erchn slacker: possibly, does "cat -v <file>" show ^M at the end
19:55 Philip since it's just a daemon running as root
19:55 Philip basically this feels like a bug
19:55 slacker erchn: nope.
19:56 kickerdog joined #salt
19:57 erchn slacker: sounds like the line endings are ok then
19:57 slacker erchn: the master and slave can communicate. test.ping and cmd.run 'echo hi' work.
19:58 slacker erchn: I had updated the salt master before trying this out.
19:58 forrest Philip, hah, nice and easy
19:58 slacker erchn: is it necessary that master and slave be the same version?
19:59 erchn slacker: typically the master must be newer in version than the minion, but there are some cases where too much skew can cause issues
19:59 Philip forrest is there a reason grains shell defaults to /bin/fa;se?
19:59 Philip false*
20:00 erchn slacker: what happens if you quote the /usr/local/src/git.zip string?
20:00 forrest Philip, did you set the shell?
20:00 helderco joined #salt
20:00 forrest when the user was created.
20:00 erchn just wondering, though it shouldn't be necessary
20:00 Philip forrest: it's a user created by installing the jenkins package
20:01 Philip I'm going to create the user and set it's shell using the user.present state, but this was a hard one to track down, and I'm wondering if it should have been easier
20:01 linjan joined #salt
20:01 slacker erchn: same error
20:01 Philip but it seems strange to me that calling salt-call from different users should make the minion behave differently
20:02 Philip why should the minion care about the calling user's shell
20:02 slacker erchn: I also tried double quotes. same error.
20:04 erchn slacker: do you get the same error running "salt-call state.show_sls <state_name>"?
20:04 erchn slacker: (from the minion)
20:04 adizam joined #salt
20:05 amckinley joined #salt
20:05 erchn slacker: maybe run that with "-l debug" also, to see if you can spot something odd in the compiled yaml
20:07 yomilk joined #salt
20:08 slacker erchn: On the minion: "local: - Requisite declaration deploy_source in SLS <name> is not formed as a single key dictionary'
20:08 nicksloan joined #salt
20:10 linjan_ joined #salt
20:11 erchn slacker: not sure, the code looks really simple, it looks at each object under the require section and if the list item isn't seen as a dictionary throws that error...nothing obvious from -l debug in the rendered yaml output?
20:12 slacker erchn: no nothing obvious..just a bunch of what looks like nested OrderedDicts
20:13 slacker will take a closer look at the file and report back in an hour or so. need to go to a stupid meeting now. thanks for your help.
20:13 hunter joined #salt
20:15 erchn slacker: hmm, any chance you could gist or pastebin the relevant portions of that output?
20:15 kickerdog When I create a new instance in aws using salt-cloud, the aws instance hostname comes out as ip-10… (ip-10-0-0-199). Is there a way to have salt-cloud make the hostname match the name specified in the cloud.map?
20:17 erchn kickerdog: the minion id is correct, but the hostname is wrong?
20:17 kickerdog yeah
20:18 Shenril joined #salt
20:18 erchn kickerdog: well, you could definitely use Salt to change the hostname
20:18 higgs001 joined #salt
20:18 kickerdog True
20:18 kickerdog But it would be nice if it was automatic :)
20:19 erchn kickerdog: I don't think salt-cloud will try and do that for you, once you get a very simple state setup, and set highstate to run on minion startup, you'll never even know it didn't happen automatic :-)
20:19 erchn (well you will, but no one else will)
20:20 rawzone joined #salt
20:20 kickerdog Sounds good, I wasn't sure if I was doing it wrong. I'll just over-ride the host file
20:23 mikeywaites joined #salt
20:23 kickerdog Once we series A, I've been given permission to migrate from AWS to Google. I'm glad salt-cloud was updated for it.
20:23 kickerdog I'm over how AWS does things.
20:24 rawzone joined #salt
20:24 hunter joined #salt
20:25 rawzone joined #salt
20:28 gadams999 I love that salt-cloud supports openstack too
20:29 gadams999 left #salt
20:30 nuclid joined #salt
20:30 rupsky joined #salt
20:31 jslatts i have an issue where some servers are intermittently responding to a grain based compound query
20:32 jslatts i have four servers with a role and only two ever respond, but it can be a different two each run
20:32 jslatts anyone ever had that happen before?
20:33 jalaziz joined #salt
20:36 cewood joined #salt
20:37 vlcn suggestions for installing vmware tools?
20:38 amallais ?
20:39 forrest vlcn, do you not have an RPM for it?
20:40 pydanny joined #salt
20:40 vlcn forrest, I've always used the perl install script
20:40 rawzone joined #salt
20:41 Tekni joined #salt
20:41 forrest vlcn, you could still use that, just put it into a cmd.run: \n - name: '/path/to/perl -options script_name.pl
20:41 forrest *'
20:41 forrest and drop the perl script on as a managed file.
20:42 yomilk_ joined #salt
20:44 kseniy joined #salt
20:46 ccase joined #salt
20:47 pydanny joined #salt
20:50 DaveQB joined #salt
20:51 ze- is the develop branch supposed to pass all tests ?
20:51 kickerdog joined #salt
20:56 [diecast] joined #salt
20:57 forrest ze-, sometimes it does, sometimes it doesn't
20:57 forrest ze-, depends on what is being worked on, in theory it should, but it's develop, so a lot of the time it doesn't
20:58 budrose_ joined #salt
20:59 ze- so, what is best-practice to make a pull-request ?
20:59 gkze joined #salt
21:00 ze- getting it from develop might not pass the tests, if develop already doesn't work.
21:01 xDamox Hello, is there any good way to test/debug Jinja files? Can I load the Jinja and Salt within a Python interpreter?
21:01 ndrei joined #salt
21:02 forrest xDamox, yea you can load it through in python, a few guys have been doing that
21:02 forrest xDamox, also it's odd to see you here during the day, I thought you worked nights for some reason.
21:02 dave_den xDamox: you can also use cp.get_template to see what a rendered template looks like form a minion.
21:02 Saint-N he does hes australian ;)
21:02 Saint-N thats my story
21:02 forrest dave_den, ahh yea another good plan
21:02 forrest Saint-N, actually damox is out in san fran
21:03 Saint-N shhhh
21:03 forrest I was creeping on his linkedin profile
21:03 Saint-N australia
21:03 bhosmer joined #salt
21:03 utz joined #salt
21:03 Saint-N :)
21:03 dave_den ze-: i think they still prefer you make the PR against develop
21:04 forrest ze-, just make the PR against develop, you can look through and see why it failed, and whether it's attributed to your code usually.
21:05 xDamox Haha, I am at work
21:06 ze- forrest: well, I have a PR, and the build states errors in something that is NOT what I did.
21:06 ze- but does not show an error I know should be in what I wrote
21:06 forrest ze-, that's fine then
21:06 forrest ze-, if you didn't create the error, don't worry about it
21:06 xDamox Trying to put a big push for SaltStack here
21:06 forrest ze-, worst case, they will say 'hey you broke this, can you please fix it?'
21:06 ze- forrest: I created an error that doesn't show :P
21:06 ze- and I already know about it :)
21:06 forrest hah
21:07 forrest well, you should fix that ze- !
21:07 ze- fix with an additional commit, or an ammend as it's a typo ?
21:07 forrest ze-, just make an additional commit if the merge isn't complete on that branch, and it will show up in the PR
21:07 forrest the PR will be updated when you fix the issue on your branch, and it will be rechecked.
21:08 forrest then you can see the pass/fail status again
21:08 Gordonz joined #salt
21:10 manicouman joined #salt
21:11 tmandry_ joined #salt
21:11 lionel_ joined #salt
21:12 ggoZ joined #salt
21:13 manicouman joined #salt
21:14 andr386 joined #salt
21:16 bemehow joined #salt
21:16 gammalget_ joined #salt
21:19 kickerdog1 joined #salt
21:21 tmmt_ joined #salt
21:21 mikeywaites1 joined #salt
21:21 manicouman joined #salt
21:23 raizyr joined #salt
21:23 HeadAIX joined #salt
21:24 ksk_ joined #salt
21:24 rockey_ joined #salt
21:24 Gareth_ joined #salt
21:24 arapaho_ joined #salt
21:25 seblu joined #salt
21:25 stotch_ joined #salt
21:26 timoguin i've made a few minor documentation PRs, and got the Build Failed message.
21:26 timoguin I just screamed "wasn't me!" and ran away
21:26 redondos joined #salt
21:26 redondos joined #salt
21:26 svx_ joined #salt
21:26 kickerdog1 haha
21:27 copelco______ joined #salt
21:27 herzi__ joined #salt
21:27 goki_ joined #salt
21:27 forrest timoguin, yea I usually make sure the docs build correctly locally, then don't worry about it in those situations.
21:27 lyddonb_ joined #salt
21:27 viq_ joined #salt
21:27 slacker erchn: the debug output from salt-call on the minion looks ok: http://pastebin.com/zCC44w2X
21:28 ThomasJ|d joined #salt
21:28 mpoole joined #salt
21:28 crashmag joined #salt
21:28 drags joined #salt
21:28 galyur joined #salt
21:28 eightyeight joined #salt
21:28 welby_away joined #salt
21:28 slacker still throws the same error about the requisite declaration in SLS is not formed as a single key dictionary
21:29 aarontc joined #salt
21:29 Jarus joined #salt
21:29 grep_away joined #salt
21:29 madduck joined #salt
21:29 madduck joined #salt
21:30 yota joined #salt
21:30 majoh joined #salt
21:31 APLU joined #salt
21:31 Eugene joined #salt
21:32 gkze_ joined #salt
21:33 n8n joined #salt
21:36 DenkBrettl joined #salt
21:37 bemehow joined #salt
21:37 darrend joined #salt
21:37 kula joined #salt
21:37 kvbik joined #salt
21:38 openxmark joined #salt
21:38 pfallenop joined #salt
21:38 jcristau joined #salt
21:38 ay joined #salt
21:38 iMil joined #salt
21:38 vlcn joined #salt
21:38 iMil joined #salt
21:38 canci joined #salt
21:38 travisfischer joined #salt
21:38 tmwsiy_ joined #salt
21:38 davet joined #salt
21:38 sporkd2 joined #salt
21:38 kedo39 joined #salt
21:38 bwq joined #salt
21:38 agenteo joined #salt
21:38 utz joined #salt
21:38 basepi joined #salt
21:38 justBob joined #salt
21:38 whiteinge joined #salt
21:38 ewong joined #salt
21:38 Linz joined #salt
21:38 funzo joined #salt
21:38 vbabiy joined #salt
21:38 sgflt joined #salt
21:38 jgelens joined #salt
21:38 torrancew joined #salt
21:38 cb joined #salt
21:38 aleszoulek joined #salt
21:38 dmwuw joined #salt
21:38 octarine joined #salt
21:38 kaiyou joined #salt
21:38 Cidan joined #salt
21:38 crane joined #salt
21:39 EntropyWorks joined #salt
21:39 faldridge joined #salt
21:40 schristensen joined #salt
21:40 N-Mi joined #salt
21:40 N-Mi joined #salt
21:40 Cidan joined #salt
21:40 paramah joined #salt
21:41 munhitsu_ joined #salt
21:41 SEJeff_work joined #salt
21:41 frasergraham joined #salt
21:41 jab416171 joined #salt
21:41 troyready joined #salt
21:41 meteorfox joined #salt
21:41 Nazzy joined #salt
21:41 dopp joined #salt
21:41 cornmander joined #salt
21:41 mikkn joined #salt
21:41 godog joined #salt
21:41 stewpot joined #salt
21:41 amckinley joined #salt
21:41 bezaban joined #salt
21:41 crazysim joined #salt
21:41 darless joined #salt
21:41 mfournier joined #salt
21:41 jnials joined #salt
21:41 viq_ joined #salt
21:41 lyddonb_ joined #salt
21:41 goki joined #salt
21:41 herzi_ joined #salt
21:41 copelco_____ joined #salt
21:41 svx_ joined #salt
21:41 redondos joined #salt
21:41 ndrei joined #salt
21:41 rupsky joined #salt
21:41 linjan_ joined #salt
21:41 adizam joined #salt
21:41 rgbkrk joined #salt
21:41 druonysus joined #salt
21:41 mpanetta joined #salt
21:41 timoguin joined #salt
21:41 danielbachhuber joined #salt
21:41 mugsie joined #salt
21:41 coolj joined #salt
21:41 rjc joined #salt
21:41 Sypher joined #salt
21:41 tzero joined #salt
21:41 kaictl joined #salt
21:41 gldnspud__ joined #salt
21:41 gparent joined #salt
21:41 emostar joined #salt
21:41 vxitch joined #salt
21:41 sulky joined #salt
21:41 vegardx joined #salt
21:41 jmccree joined #salt
21:41 repl1cant joined #salt
21:41 n0c joined #salt
21:41 Eliz joined #salt
21:41 svs joined #salt
21:41 terminalmage joined #salt
21:41 dcmorton joined #salt
21:41 steveoliver joined #salt
21:41 Guest10196 joined #salt
21:41 rhand joined #salt
21:41 bejer joined #salt
21:41 kevinbrolly_ joined #salt
21:41 nicksloan joined #salt
21:41 carmony joined #salt
21:41 cyrusdavid joined #salt
21:41 mjulian joined #salt
21:41 chuffpdx joined #salt
21:41 hotbox joined #salt
21:41 jero joined #salt
21:41 xt joined #salt
21:41 JordanRinke joined #salt
21:41 abele joined #salt
21:41 bretep joined #salt
21:41 delkins_ joined #salt
21:41 rogst joined #salt
21:41 rcsheets joined #salt
21:41 modafinil_ joined #salt
21:41 gamingrobot joined #salt
21:41 dpacz joined #salt
21:41 pnl joined #salt
21:41 shadowsu1 joined #salt
21:41 runnyspot joined #salt
21:41 alienresidents joined #salt
21:41 balltongu joined #salt
21:41 sashka_ua joined #salt
21:41 andrej joined #salt
21:41 scarcry joined #salt
21:41 magenbrot joined #salt
21:41 msciciel joined #salt
21:41 SaveTheRbtz joined #salt
21:41 sinh joined #salt
21:41 jbub joined #salt
21:41 babilen joined #salt
21:41 bigmstone joined #salt
21:41 ahammond joined #salt
21:41 cewood joined #salt
21:41 Cidan joined #salt
21:41 Cidan joined #salt
21:41 redondos joined #salt
21:42 sulky joined #salt
21:42 tharkun joined #salt
21:43 zdong joined #salt
21:45 rostam joined #salt
21:46 blast_hardcheese joined #salt
21:46 blast_hardcheese joined #salt
21:46 ipalreadytaken joined #salt
21:46 chutzpah joined #salt
21:46 chutzpah joined #salt
21:46 ipmb joined #salt
21:47 Yaann joined #salt
21:47 ze- github kinda down ?
21:47 harobed joined #salt
21:48 darless left #salt
21:48 yano joined #salt
21:48 kermit joined #salt
21:50 xDamox Hey Forrest, is there any documents of loading Jinja and Salt into the python interpreter for testing?
21:50 agenteo @ze- http://www.downforeveryoneorjustme.com/github.com
21:50 agenteo yep
21:51 elfixit joined #salt
21:51 xDamox 21:39 UTCWe continue to investigate connectivity problems to GitHub.com. We'll update when we know more.
21:51 xDamox https://status.github.com/
21:52 cro joined #salt
21:53 rawzone joined #salt
21:54 zain joined #salt
21:54 forrest xDamox, I don't think so.
21:56 xDamox forrest, is there any interface to load it all up in one swoop?
21:57 forrest xDamox, I don't believe so :\
21:58 xDamox I guess it's not a simple import salt and import Jinja :(
21:59 ipalreadytaken joined #salt
21:59 bhosmer joined #salt
22:00 forrest xDamox, I wish!
22:01 ze- damn, it's down, and i just deleted my salt-testing stuffs... so now i can't test my code anymore :P
22:04 younqcass joined #salt
22:04 whidbeywalker joined #salt
22:05 mpanetta Quick question...
22:06 mpanetta When you have a requires or watch section in a state, does it do the things in order they are listed?
22:06 thayne joined #salt
22:06 forrest mpanetta, what do you mean?
22:06 forrest mpanetta, can you show an example?
22:06 mpanetta I think so. sec
22:06 forrest ok
22:07 mpanetta https://gist.github.com/anonymous/fbdaff3a1548f389dfec
22:07 mpanetta The require bit in that state, does it run the 2 things in order they are listed?
22:08 mpanetta Like will it do the File: state and then the cmd: state?  Or is it random?
22:08 whiteinge xDamox: take a look at the salt/testing/saltsh.py script. it has (very) basic jinja + salt stuffs
22:08 whiteinge (i'd like to the file on github, but...)
22:08 forrest mpanetta, honestly I'm not sure. Usually you'd have a require in cmd: rs-initiate that it requires file: tmp/.ismaster.py so that would be a non-issue
22:08 forrest whiteinge, github seems to have recovered
22:09 mpanetta hmm
22:09 forrest does that make sense mpanetta?
22:09 whiteinge forrest: i'll believe it when i see it!
22:09 forrest mpanetta, if not I can add an example
22:09 mpanetta forrest: Yeah, I think...
22:09 forrest whiteinge, hah, I'm seeing it right now
22:09 mpanetta In the case above the order is irrelevent...
22:09 forrest mpanetta, let me write up a quick example, hang on
22:09 mpanetta Ok
22:10 mpanetta Thanks :)
22:10 forrest yea np
22:10 CeBe joined #salt
22:11 yomilk joined #salt
22:11 coolj left #salt
22:12 forrest mpanetta, ok refresh that gist
22:13 forrest I've added a comment
22:13 whidbeywalker joined #salt
22:13 forrest mpanetta, you see how we have this 'chain' established because of the requires? It ensures that we're executing things in that specific order.
22:13 rupsky_ joined #salt
22:13 [diecast] joined #salt
22:14 SEJeff_work joined #salt
22:18 ksalman how come version 0.17.2 isn't available here? http://ftp.us.debian.org/debian/pool/main/s/salt/
22:19 mikeywaites joined #salt
22:20 vejdmn joined #salt
22:20 utz joined #salt
22:22 forrest ksalman, looks like only the latest release of each major release is available.
22:22 ksalman :(
22:26 n8n joined #salt
22:27 pcarrier joined #salt
22:28 mpanetta forrest: Taking a look now
22:28 mpanetta Sorry I stepped away for a moment
22:28 forrest mpanetta, no worries
22:28 l0x3py joined #salt
22:29 slacker joined #salt
22:30 mpanetta forrest: Ok, so that makes sense
22:30 forrest mpanetta, cool
22:33 fllr joined #salt
22:35 darless joined #salt
22:35 slacker hi, just need a little help with my sls file: http://pastebin.com/jtEEHyUR
22:36 slacker in the deploy_source section, what should I use instead of the question marks?
22:36 darless left #salt
22:39 mikeywaites joined #salt
22:40 akoumjian joined #salt
22:42 forrest slacker, pkg: prerequisites
22:43 mikeywaites joined #salt
22:44 slacker forrest: excellent. thank you. it worked! :)
22:44 forrest slacker, yea np.
22:45 younqcass joined #salt
22:48 manicouman joined #salt
22:49 hunter joined #salt
22:49 vejdmn1 joined #salt
22:51 vejdmn joined #salt
22:58 che-arne joined #salt
22:59 fwiles joined #salt
22:59 ipalreadytaken joined #salt
23:01 fwiles So pretty confused by an error I'm getting.  Getting 'No Top file or external nodes data matches found', despite the being used showing up in salt-keys and have verified (cut-n-pasted even) to ensure there wasn't a typo between top.sls and the commandline. All the errors I find via Google point to gitfs, but we're not using that on this system.  Any ideas what might be going wrong?
23:01 ipalreadytaken joined #salt
23:01 fwiles sorry despite the name being used showing up in salt-keys
23:01 fwiles Can even target it for things like 'cmd.run uptime'
23:09 sporkd2 fwiles: you check the minion ids?
23:10 fwiles that's a good idea
23:10 fwiles sporkd2: wait you mean the ID in top.sls?
23:10 quickdry21 joined #salt
23:10 sporkd2 what are you matching your minions w/ in your top file?
23:11 fwiles not actual hostname but 'es.dev.example.com'
23:11 fwiles salt 'es.dev.example.com' cmd.run uptime is fine
23:11 fwiles salt 'es.dev.example.com' state.highstate fails with "No Top file or external nodes data matches found"
23:12 fwiles despite it having 'es.dev.example.com':\n- elasticsearch in top.sls
23:12 fwiles we're using the elasticsearch on another system 'es1.example.com' so I don't think it's related to the state itself just the targeting
23:13 sporkd2 i would maybe do a salt 'es.dev...' grains.items and make sure their returning that
23:14 yomilk joined #salt
23:14 fwiles yup that returns data
23:20 fllr joined #salt
23:25 hunter joined #salt
23:25 fwiles hmmm the box is newer than the rest of ours and is running 2014.1.0 where the master is 0.17.4. Maybe that's the issue?
23:26 bhosmer joined #salt
23:27 ajprog_laptop joined #salt
23:27 allanparsons joined #salt
23:27 fwiles yup that's it, top of the Release notes
23:27 allanparsons is there a way i can get avail sys memory in pillar data?
23:28 allanparsons i have a redis config, and i want to set max memory = 50% of physical memory
23:29 gnugnu9 joined #salt
23:31 brucewang allanparsons: status module, status.meminfo
23:31 Fin1te joined #salt
23:32 allanparsons how would i do math within the pillar?
23:32 allanparsons (or state file)?
23:32 brucewang use python
23:33 allanparsons is it best to place that in pillar or the state file?
23:33 brucewang better to do in state file, pillar is for more "static" data
23:34 allanparsons ok
23:34 allanparsons thx
23:37 allanparsons that look right: maxmemory {{ redis.get('maxmemory', 0.5*status.meminfo-(0.2*status.meminfo)) }} ?
23:37 manicouman joined #salt
23:39 tr_h joined #salt
23:40 fllr joined #salt
23:40 greg joined #salt
23:41 montyboy joined #salt
23:43 jssjr joined #salt
23:48 Yaann joined #salt
23:51 [diecast] joined #salt
23:57 seblu joined #salt

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