Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-01-30

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

All times shown according to UTC.

Time Nick Message
23:01 vejdmn joined #salt
23:01 jtang1 joined #salt
23:05 BonoBGood joined #salt
23:05 dlindquist1 joined #salt
23:05 vejdmn joined #salt
23:06 rojem joined #salt
23:06 Guest45437 joined #salt
23:06 elfixit joined #salt
23:06 tvle joined #salt
23:07 BonoBGoo_ joined #salt
23:08 dlindquist joined #salt
23:09 markizano joined #salt
23:11 DanGarthwaite joined #salt
23:12 DanGarthwaite ....and I've met basePi.  Good to meet you.
23:13 DanGarthwaite forrest is still unverified as human.
23:15 Mua_ joined #salt
23:18 kalloc joined #salt
23:19 druonysus joined #salt
23:20 toofer joined #salt
23:22 toofer_ joined #salt
23:27 reign joined #salt
23:31 rgbkrk joined #salt
23:32 basepi DanGarthwaite: facing the stage, forrest is on the right side, second row from the back
23:32 basepi Alllll the way to the right.
23:32 basepi I'll introduce you after Corey stops talking. ;)
23:33 Nazca__ so, next week? :D
23:41 bemehow joined #salt
23:42 BonoBGood joined #salt
23:42 andrej heh
23:46 justBob joined #salt
23:58 kermit joined #salt
23:59 BonoBGood joined #salt
00:01 justBob joined #salt
00:02 dlindquist joined #salt
00:06 Katafalkas joined #salt
00:11 fllr joined #salt
00:14 dlindquist joined #salt
00:21 higgs001_ joined #salt
00:24 micah_chatt joined #salt
00:27 jtang1 joined #salt
00:27 Ryan_Lane joined #salt
00:32 mgw joined #salt
00:32 DanGarthwaite joined #salt
00:34 Ryan_Lane joined #salt
00:36 gnu-man joined #salt
00:37 wkf_ joined #salt
00:40 melinath joined #salt
00:41 melinath joined #salt
00:44 melinath joined #salt
00:52 jeremyBass joined #salt
00:53 fllr joined #salt
00:56 backjlack joined #salt
00:57 n8n joined #salt
00:58 zfouts joined #salt
00:59 Gordonz joined #salt
01:00 fllr joined #salt
01:01 diegows joined #salt
01:05 gadams9991 joined #salt
01:10 dlindquist joined #salt
01:11 StDiluted joined #salt
01:15 bemehow joined #salt
01:19 n8n joined #salt
01:20 gnu-man_ joined #salt
01:20 zzzirk joined #salt
01:21 jtang1 joined #salt
01:23 mgw joined #salt
01:24 vejdmn joined #salt
01:30 juicer2 joined #salt
01:31 DanGarthwaite joined #salt
01:33 jtang1 joined #salt
01:38 warmwaffles joined #salt
01:38 allanparsons joined #salt
01:38 allanparsons hello - i had a quick question.  If i pass a variable to jinja, how do I check it via if?
01:38 allanparsons actually .. hang tight... might be a syntax err
01:41 elithrar joined #salt
01:43 cachedout joined #salt
01:44 CheKoLyN joined #salt
01:49 fragamus joined #salt
01:52 wkf_ joined #salt
01:52 xmltok joined #salt
01:53 codysoyland joined #salt
01:57 brendangilmor joined #salt
01:59 mgw joined #salt
02:03 bemehow joined #salt
02:05 Katafalkas joined #salt
02:06 jeremyBass quick question for someone.. i need to use an onlyif: .. i have [ -n "$\{ADDED{{ ext_key|replace("-","") }}\}" ] .. but the { } in the command it freaking out.. it needs to render as onlyif: [ -n "${ADDEDvarValue}" ] .. I'm not clear how to escape it right
02:07 wkf_ joined #salt
02:08 jsm joined #salt
02:09 rgbkrk joined #salt
02:11 TronPaul_ joined #salt
02:12 TronPaul_ ids "minion" for salt-ssh come from the minions' fqdn right?
02:13 TronPaul_ trying to use state.highstate with salt-ssh, but the box isn't being matched properly
02:13 TronPaul_ match.glob also does not match the fqdn w/ salt-ssh
02:14 TronPaul_ but target selection works properly with globs/fqdns
02:14 jeremyBass ok nevermind it's akward i think but i think '[ -n "${{ '{' }}ADDED{{ ext_key|replace("-","") }}{{ '}' }}" ]' works
02:14 mwmnj joined #salt
02:17 mgw joined #salt
02:18 yomilk joined #salt
02:26 Ashtonian joined #salt
02:28 jtang1 joined #salt
02:44 dlindquist joined #salt
02:52 cachedout joined #salt
03:01 mpanetta joined #salt
03:04 mpanetta joined #salt
03:08 LowWalker joined #salt
03:09 StDiluted joined #salt
03:12 anuvrat joined #salt
03:13 Gordonz joined #salt
03:13 DanGarthwaite joined #salt
03:13 bemehow joined #salt
03:22 jtang1 joined #salt
03:33 cachedout_home joined #salt
03:42 AdamSewell joined #salt
03:43 dlindquist joined #salt
03:44 rgbkrk joined #salt
03:45 tempspace joined #salt
03:45 tempspace man, it's nasty outside at saltconf
03:46 Katafalkas joined #salt
03:46 tempspace winds were so nasty it ripped a street sign off the pole and sent it flying into the street
03:46 Katafalkas joined #salt
03:46 kedo39 wow, sounds rough
03:48 Gareth w 7
03:48 Gareth erm
03:54 g3cko joined #salt
03:57 Hefeweizen joined #salt
04:05 Furao joined #salt
04:17 jtang1 joined #salt
04:24 mhahn joined #salt
04:25 mhahn hey, i'm really new to salt and am getting the follow error when trying to set up masterless minion with vagrant: http://pastebin.com/JMSZaQf2
04:30 tvle joined #salt
04:31 codekoala anyone here know what the pass/fail line is for the ssce?
04:31 codekoala just got an 85%
04:34 helderco joined #salt
04:35 dave_den 80% i hear
04:35 helderco Hey guys... can someone help me setting up saltstack formulas with gitfs?
04:35 dave_den congrats
04:36 codekoala woo, thanks!
04:36 codekoala if I can do it, anyone can. this is day #2 with salt for me
04:37 helderco They need to be served from a subdirectory, right? (gitfs_root)
04:39 dave_den sorry helderco, i dont use any salt-formula states myself
04:39 helderco Perhaps some other state to test gitfs out
04:41 helderco Ok, tried to add vim from github.com/saltstack/salt-states.git, but still doesn't work. I added git to fileserver_backends, then added git://github.com/saltstack/salt-states.git to gitfs_remotes
04:41 helderco Am I missing anything?
04:42 helderco I have GitPython
04:44 davet joined #salt
04:45 munderwo joined #salt
04:49 mgw joined #salt
04:57 jacksontj joined #salt
04:58 marcinkuzminski joined #salt
05:03 jacksontj joined #salt
05:04 joshe joined #salt
05:05 bob2 joined #salt
05:05 middleman_ joined #salt
05:06 anuvrat joined #salt
05:09 __number5__ you know the saltstack/salt-states repo is for testing, right?
05:10 fllr joined #salt
05:10 jtang1 joined #salt
05:12 Ryan_Lane joined #salt
05:19 mgw joined #salt
05:23 Thiggy joined #salt
05:29 DanGarthwaite joined #salt
05:32 pdayton joined #salt
05:39 txpv joined #salt
05:45 jalbretsen joined #salt
05:45 cachedout_home joined #salt
05:46 anuvrat joined #salt
05:48 tvle joined #salt
05:49 alunduil joined #salt
05:53 mhahn joined #salt
05:54 apergos joined #salt
05:59 matanya joined #salt
06:04 jtang1 joined #salt
06:08 mhahn joined #salt
06:10 ravibhure joined #salt
06:13 snuffeluffegus joined #salt
06:17 twiedenbein joined #salt
06:18 anuvrat joined #salt
06:20 Psyke joined #salt
06:25 druonysus joined #salt
06:27 cloudpie joined #salt
06:32 BenCoinanke joined #salt
06:47 mhahn joined #salt
06:48 Ryan_Lane joined #salt
06:52 LowWalker joined #salt
06:56 cachedout_home joined #salt
06:56 giantlock joined #salt
06:59 alunduil joined #salt
06:59 fllr joined #salt
07:14 cachedout_home joined #salt
07:20 ndrei joined #salt
07:23 ndrei joined #salt
07:25 mhahn i'm getting the following error when trying to install something with pip.installed: http://pastebin.com/FA7616t6
07:26 mhahn any ideas?
07:26 malinoff mhahn, do you have python-pip installed?
07:28 mhahn going pretty much exactly off the example here: http://docs.saltstack.com/ref/states/all/salt.states.pip.html
07:30 fllr joined #salt
07:30 malinoff mhahn, run salt your-minion-id cmd.run 'dpkg -l | grep pip' (if you have a debian minion), or salt your-minion-id cmd.run 'yum list installed | grep pip' (if centos)
07:31 mhahn malinoff, i thought python3-pip would take care of that, but after reading: "As of Salt 0.17.0 the pip state needs an importable pip module. This usually means having the system's pip package installed or running Salt from an active http://www.virtualenv.org/." I'm guessing it doesn't
07:32 fllr joined #salt
07:32 cachedout_home joined #salt
07:33 fllr joined #salt
07:36 ckao joined #salt
07:37 mhahn malinoff, including python-pip got it working
07:38 malinoff mhahn, It looks like a bug
07:39 malinoff Create an issue
07:39 mhahn malinoff, thanks for the heads up
07:40 malinoff mhahn, hm, wait a bit
07:40 malinoff can you log in your minion instance and see which python is configured in the venv you installed?
07:45 matanya joined #salt
07:45 mhahn malinoff, this is within vagrant
07:45 malinoff mhahn, so you can't? I'm not very familiar with vagrant
07:47 mhahn malinoff, its running python 2.7 and python 3.3
07:49 malinoff mhahn, No, I mean, what will /path/to/your/venv/bin/python --version return?
07:49 mhahn i think the issue is that "which python" -> returns python2.7
07:49 malinoff Yeah
07:50 malinoff That's what i'm talking about
07:50 malinoff So your venv is basically did not configure to use python3
07:50 mhahn so i would guess that salt tries to import pip running under python 2.7
07:50 mhahn and it fails
07:50 mhahn until python-pip is installed
07:51 anuvrat joined #salt
07:54 mhahn yea this is the step to create the virtualenv
07:58 balboah joined #salt
08:00 lacrymology joined #salt
08:00 lacrymology how do I run a template with test=true?
08:01 lacrymology a state template
08:01 lacrymology ok, let me explain better, since I don't know the jargon
08:02 malinoff lacrymology, salt minion-id state.sls yourstate test=True ?
08:02 lacrymology I'm running with salt-call --local state.template /path/to/file.sls.
08:02 lacrymology malinoff: probably?
08:02 lacrymology =)
08:03 lacrymology malinoff: nope
08:03 malinoff Uhm, don't think 'template' function is a 'test' function already :)
08:03 malinoff I think*
08:03 malinoff 'test' case is for states
08:03 jeremyBass joined #salt
08:04 malinoff If you don't want to actually install a package, but you want to ensure there is nothing that prevents from it
08:04 lacrymology malinoff: something that will allow me to run a state from a file by name that will accept the test parameter? Or some conf file I can add the test: True option easily?
08:04 gammalget joined #salt
08:04 lacrymology malinoff: I'm debugging an error that only ocurrs in a custom state with test: True
08:05 malinoff lacrymology, I don't think you can do that. Only via 'salt-call --local state.sls test=True'
08:09 lacrymology malinoff: well, I changed conf/minion for a second
08:09 bdf 4
08:15 ndrei joined #salt
08:20 gurubert1 joined #salt
08:22 harobed_ joined #salt
08:25 anuvrat joined #salt
08:34 giantlock_ joined #salt
08:34 zooz joined #salt
08:36 yomilk joined #salt
08:38 lacrymology I see quite a lot of this warning: "passing a salt environment should be done via '__env__', not 'env', .... bla bla bla"
08:40 lacrymology but salt.modules.file.check_managed accepts env, not __env__
08:40 lacrymology anyone has a clue of what's going on there?
08:41 malinoff Dunno lol :D
08:41 rodvdka joined #salt
08:41 rodvdka left #salt
08:43 ndrei joined #salt
08:46 miguitas joined #salt
08:50 anuvrat joined #salt
08:56 ndrei joined #salt
08:59 mhahn joined #salt
09:02 fllr joined #salt
09:04 LowWalker joined #salt
09:05 tl-ml joined #salt
09:06 dangra joined #salt
09:06 _fllr_ joined #salt
09:08 fllr joined #salt
09:14 cellofellow left #salt
09:18 dbanck joined #salt
09:21 cast joined #salt
09:22 martoss joined #salt
09:28 fllr joined #salt
09:32 _fllr_ joined #salt
09:36 Simon____ joined #salt
09:36 kalloc joined #salt
09:39 mike251 joined #salt
09:47 Simon____ joined #salt
09:50 Simon____ joined #salt
09:52 Simon____ joined #salt
09:56 che-arne joined #salt
09:59 shiin joined #salt
09:59 shiin service salt-minion restart seems broken
09:59 shiin but /etc/init.d/salt-minion restart still works
09:59 shiin did I miss something there?
09:59 mike251 shiin: what OS?
10:00 shiin debian squeeze
10:00 mike251 shiin:  sometimes.. also on debian i have seen that the service command fails
10:00 shiin well, I do have more squeeze machines that dont have the issue, it might just be affecting new installs
10:00 mike251 not sure why..
10:00 shiin or another random reason, like you said.
10:01 mike251 i have no idea why .... and NOT on all machines
10:01 mike251 i use... turnkeylinux version of debian... and on the new ones it works..
10:15 ndrei joined #salt
10:16 tonthon Hi
10:16 tonthon Is there any difference between configuring firewalls for salt-syndic or for salt-master/mininon communciation ?
10:17 linjan joined #salt
10:17 mike251 tonthon:  sorry ... i am a newbie myself.. i can not answer that... i don't use syndic
10:18 mike251 i am also curious .. what ports is the master-minion used to communicate via? ...
10:18 tonthon mike251: the salt minion connect to the master through 4505 and 4506 through tcp
10:18 mike251 tonthon:  thanks! very useful
10:19 n8n joined #salt
10:20 malinoff http://docs.saltstack.com/ref/topology.html
10:20 malinoff http://docs.saltstack.com/topics/tutorials/firewall.html
10:20 mike251 malinoff: thank you kindly
10:32 fllr joined #salt
10:34 Cidan joined #salt
10:50 ndrei joined #salt
10:54 Nexpro joined #salt
10:58 anuvrat joined #salt
11:06 yomilk joined #salt
11:09 david_a joined #salt
11:13 andyshinn` joined #salt
11:14 david_a joined #salt
11:31 gwmngilfen joined #salt
11:32 fllr joined #salt
11:38 elfixit joined #salt
11:40 jfxura joined #salt
11:42 jfxura folks - I understand, file.managed allows me a rollback, but in general can i rollback actions performed by other state functions?
11:43 taion809 joined #salt
11:45 diegows joined #salt
11:46 helderco joined #salt
11:47 ekristen joined #salt
11:49 stephanbuys joined #salt
11:51 stephanbuys hi all, is there a difference between Salt 2014.1.0 and 0.17.5? I've been trying to access the (beta) Dockerio state, but it doesn't seem to be included in EPEL (0.17.5), so I'm wondering if I'm installing the wrong version somehow?
11:56 beardo joined #salt
11:57 tinuva joined #salt
12:03 helderco Hey guys, can someone help setting up gitfs? I'm trying to use a saltstack formula but can't get it to work. I have git and GitPython, added git to fileserver_backend and git://github.com/saltstack-formulas/docker-formula.git to gitfs_remotes in the master config. Am I missing anything?
12:07 Nazzy stephanbuys, considerable difference, yes
12:08 stephanbuys Nazzy: which would you recommend? Should I abandon the EPEL based install and rather grab 2014.1.x through PIP?
12:09 Nazzy stephanbuys, 0.17.5 is a bug fix point release on the 0.17 branch so it's missing a lot of the new features
12:10 Nazzy stephanbuys, I'd say grab the latest RC or git if you want to try out something that isn't in 0.17.5
12:10 stephanbuys Nazzy: I see, thanks!
12:10 Nazzy more testers are always a good thing :)
12:11 stephanbuys Nazzy: so can I assume that 2014 is not quite "production" yet?
12:12 Nazzy stephanbuys, I'm running it in production, but it's likely still got bugs lurking... YMMV unfortunately
12:14 Nazzy there also seems to be some compatibility issues with minions running 0.17, so that's also something to consider
12:14 stephanbuys Nazzy: ok, thats good to know. thanks for the info. your preferred installation method? are you running from git (that is if you don't mind saying :) )
12:15 Nazzy stephanbuys, I am indeed running from git... I get an easy out because gentoo has a live ebuild I can use.  There's a really nice doc for setting up virtualenv for it though
12:15 Nazzy http://docs.saltstack.com/topics/hacking.html#installing-salt-for-development
12:17 stephanbuys Nazzy: thanks, I've tried that before, but being a rookie I was still getting stuck with the basics and I couldn't pin down why things didn't seem to add up. I'll try it again
12:18 pdayton joined #salt
12:18 Nazzy stephanbuys, esky is also an option but I've not used it so I can't help there
12:22 justBob joined #salt
12:27 crane can i use an http url in file.managed as source?
12:27 crane or how should i download a file?
12:29 ksk joined #salt
12:29 ksk hey guys
12:30 ksk lets say ive got like 15k machines and like to do saltstack - is there a cool way to divide them onto serveral salt-masters?
12:31 shiin ksk: when setting up salt, make their minion config have the right master name?
12:32 fllr joined #salt
12:32 Nazzy ksk, with a pool that size I'd setup syndic or multimaster and use a DNS round robin to spread them out
12:33 ksk so you never know which minion is on which master? bad idea
12:34 ksk i thought about using split horizon dns but that might be not too easy
12:34 Nazzy with 15k I wouldn't be trying to figure it out very often :p
12:35 worstadmin joined #salt
12:35 Nazzy the would be to have it look like one big pool from the master side, though that does have the drawback that targeting could get painful
12:37 Nazzy *thinks* iirc, you could seed them with a fixed address and have salt update that address for you
12:38 ksk yap, targeting is a problem if we use multiple masters. that does sound good, too. thanks for the idea
12:39 Nazzy if you split the minions over separate servers you need to know which master to run on, if you merge the pool via syndic you have a ton of nodes to search ... yeah, sucks either way
12:42 helderco Can someone help setting up gitfs? I'm trying to use a saltstack formula but can't get it to work. I have git and GitPython, added git to fileserver_backend and git://github.com/saltstack-formulas/docker-formula.git to gitfs_remotes in the master config. Am I missing anything? I'm running a minimal Ubuntu precise (12.04.4 LTS).
12:49 ndrei joined #salt
12:57 gadams999 joined #salt
12:58 lacrymology left #salt
13:02 zzzirk joined #salt
13:13 grim76 joined #salt
13:14 elfixit joined #salt
13:17 imaginarysteve joined #salt
13:21 juicer2 joined #salt
13:22 kadel joined #salt
13:27 LowWalker joined #salt
13:30 jeremyBass left #salt
13:32 fllr joined #salt
13:33 harobed joined #salt
13:34 harobed joined #salt
13:37 babilen helderco: Did you restart the salt-master ? IIRC you need a newer version of GitPython on precise (>= 0.3 IIRC)
13:38 helderco Details here https://github.com/saltstack/salt/issues/4210#issuecomment-33687472
13:38 helderco babilen: I've restarted, but I do have GitPython 0.1.7
13:39 helderco Installed with pip
13:39 babilen You need GitPython 0.3.0 or newer as mentioned on http://docs.saltstack.com/topics/tutorials/gitfs.html
13:40 faust joined #salt
13:40 Nazzy it = just; jmp(&left); return(right);
13:41 helderco babilen: oh, didn't notice that.
13:41 babilen I am trying to include a bar.sls from my foo/init.sls file located in $PILLAR/foo/bar.sls. I am using an external (git) pillar for this. If I do that, however, I get the error "Specified SLS 'bar' in environment 'base' is not available on the salt master"
13:41 yota_ is it a good practice to add own grains?
13:41 babilen helderco: I am referring to the "Note GitFS requires the Python module GitPython, version 0.3.0 or newer. If your Master runs  Ubuntu 12.04 LTS, you will likely need to install GitPython using pip." in there
13:41 yota_ or using Pillar is better way?
13:42 Nazzy yota_, both are acceptable but depend on what you're trying to do
13:43 babilen I should be able to include a foo.sls file located in bar/foo.sls from bar/init.sls by using "include: \n  - foo" shouldn't I?
13:43 Nazzy yota_, grains are best for things the minion wants to define about itself, pillars are best for things the master wants to define for the minion
13:43 Thiggy joined #salt
13:43 Nazzy babilen, probably not, I think all includes are absolute
13:44 babilen So I would have to use "include: \n bar.foo" ?
13:44 yota_ Nazzy: ok thanks. I want add a minion "characteristic" so grains seems better
13:47 Nazzy yota_, as long as the information doesn't need to be kept private to that minion, that works.  I consider things like roles to be master side pillar stuff though, even though it's a "characteristic" of the minion, since it's something the master should tell the minion
13:48 Nazzy babilen, think so... try it?
13:48 babilen Nazzy: I just did (the good old TIAS), but now I am getting "The specified SLS "bar.foo" ..." :)
13:51 babilen Nazzy: Let me prepare some actual data - I might have made a mistake somewhere that is lost in the foo.bar abstraction. I want to define different "profiles" or "roles" for minions. Would the name "role" be more frequently used in the salt community?
13:52 Ryan_Lane joined #salt
13:53 Ryan_Lane1 joined #salt
13:53 babilen Nazzy: http://paste.debian.net/79213/
13:54 ndrei joined #salt
13:55 che-arne joined #salt
13:55 jslatts joined #salt
13:55 helderco babilen: I have 0.3.2-RC1, but it still doesn't work
13:56 helderco Is there a version known to work?
13:56 babilen helderco: Define "doesn't work" and provide additional information please (preferably in a pastebin). Did you restart the salt-master this time around?
13:57 helderco babilen: details are here https://github.com/saltstack/salt/issues/4210#issuecomment-33687472
13:57 Iwirada joined #salt
13:57 helderco I did even destroy de container and try again
13:58 helderco Restarted the salt-master just in ase
14:00 wkf_ joined #salt
14:02 helderco but now gitfs shows in cache
14:02 LowWalker joined #salt
14:06 bhosmer_ joined #salt
14:06 mwmnj joined #salt
14:06 babilen Nazzy: Hmm, I can include it if I have "profiles.sls" in the root and include that directly, but apparently global.profiles does not work
14:08 zooz_ joined #salt
14:09 babilen Hmm, I can include pillars just fine if I am not including sls files in the same subdirectory -- What am I missing?
14:10 Nazzy babilen, curious ... it sounds like a bug but I'm not familiar enough with the mechanic
14:12 babilen Nazzy: I started using salt a week ago ... I was simply under the impression that it *should* work and I am surprised that it doesn't
14:13 babilen It might very well be a bug .. In the sense the a "sub-pillar" can only be included once it finished compiling. If I include foo.profiles in foo/init.sls as that would require foo to be "compiled" / available already.
14:14 babilen It is a shame though as it forces me to use additional hierarchies simply to work around this. The fact that I am running into this might hint at the fact that what I am doing is not very common though.
14:14 helderco babilen: now it just hangs on LocalClientEvent PULL. Too long for git pulling 2 repos. How can I troubleshoot further?
14:15 babilen helderco: nfc
14:15 babilen helderco: It hangs when you do what?
14:15 dangra joined #salt
14:15 helderco salt -l all orion state.highstate
14:15 helderco On [DEBUG   ] LocalClientEvent PULL socket URI: ipc:///var/run/salt/master/master_event_pull.ipc
14:18 timc3 joined #salt
14:20 AdamSewell joined #salt
14:20 AdamSewell joined #salt
14:21 Nazzy babilen, I think you're right, but I'm not sufficiently certain ... might be worth to raise it as an issue outright actually
14:22 shiin mike251: I solved the problem
14:22 lahwran joined #salt
14:23 shiin the file /etc/init/salt-minion.conf does strange stuff that causes the service command to fail, moving it away makes it work again. I have lots of minions where that file is missing for some reason, but some have it (including squeeze ones) and it still works.
14:23 babilen Nazzy: Will do ...
14:23 fejjerai joined #salt
14:23 crane can someone explain me why my build fails? http://dpaste.com/1577086/
14:24 sirtaj joined #salt
14:24 teebes joined #salt
14:24 vejdmn joined #salt
14:25 sulky joined #salt
14:25 jeddi joined #salt
14:26 sgviking joined #salt
14:27 maci joined #salt
14:27 harobed joined #salt
14:27 mike251 shiin: awesome!
14:27 helderco joined #salt
14:28 mackstick joined #salt
14:28 mike251 shiin: my setup is based on centos mostly .. only a few debian (just for test purposes)
14:28 tessellare joined #salt
14:31 Nazzy crane, line 19
14:32 quickdry21 joined #salt
14:32 fllr joined #salt
14:33 Nazzy crane, mysql grant requires a scope specified as database.table or database.* ... not sure if it's dropping the value as invalid, but try sticking a .* on the end of the value and see if that works
14:36 crane Nazzy, thx. i will try this. and what about my file? i told him that the file is required and it exists as seen in the previous step
14:37 diegows joined #salt
14:37 jsm joined #salt
14:38 crane Nazzy, yep, .* fixed my problem with grant :)
14:39 GradysGhost joined #salt
14:42 rlarkin joined #salt
14:43 kalloc joined #salt
14:45 Nazzy crane, as I read it, you haven't got a file.exists state for the prereq?
14:45 kalloc_ joined #salt
14:45 jeremyBass1 joined #salt
14:48 crane Nazzy, nope... i thought that would do it?
14:48 Nazzy crane, you probably want "archive.extracted: /tmp/sql/", though it won't work because line 11 isn't below the line 10 path
14:49 Nazzy crane, http://docs.saltstack.com/ref/states/requisites.html :) the requisites refer to the condition of other states
14:49 zfouts joined #salt
14:49 crane Nazzy, ah okay thx :)
14:49 pdayton joined #salt
14:50 timc3 I have started to look into Salt and the first thing I want to do is to set my minions, /etc/security/limits.conf file to change the number of file descriptors. But I am unsure how to start this configuration? It doesn't seem to sit in states..
14:50 crane Nazzy, i'm defently new to salt. but it's pretty easy if you get how it works :)
14:51 Nazzy crane, yeah, the docs aren't the clearest thing in the world sadly, but they're being improved :)
14:51 crane Nazzy, seen things which are more worse than them :D
14:52 * GradysGhost has seen worser things, too
14:52 crane Nazzy, just a small hint for me... how can i negate file.exists?
14:52 GradysGhost ElasticSearch docs, f/ex
14:52 crane GradysGhost, haha! true! damn true! :D
14:53 GradysGhost crane: The answer to your query: file.absent
14:53 kalloc joined #salt
14:53 crane GradysGhost, ty :)
14:53 MTecknology As saltconf concludes, I want to remind each and every one of you to drink a lot of water. It's going to take a long time to get all of that salt out of your system... if ever it's possible.
14:53 GradysGhost Yes, one would think that a website documenting a fulltext search engine would produce reasonable search results.
14:53 MTecknology :D
14:54 Ahlee timc3: You're right, a state to manage that would make the most sense.
14:54 Nazzy crane, http://docs.saltstack.com/ref/states/all/salt.states.file.html :)
14:54 Ahlee timc3: a file.managed state for /etc/security/limits.conf is what you're likely looking for, one second
14:54 GradysGhost MTecknology: Next year, have a giant community salt lick in the middle of the lobby.
14:54 timc3 Ahlee: thanks. And then some kind of template which adds lines I want
14:54 Ahlee timc3: http://docs.saltstack.com/ref/states/all/salt.states.file.html#salt.states.file.managed
14:55 Nazzy Ahlee, might be something to knock up, a state module for limits.conf, shouldn't be too hard eh?
14:55 babilen I need a bit of help as I can't quite see the big picture yet and have no idea what is being idiomatic. We want to automate administration of hosts for multiple custumers which follow a fairly distinct naming scheme (i.e. $CUST-$ROLE$NUM, e.g. foo-db1) ... What we want to be able to do is to define configuration on four abstraction levels: 1. Global configuration (every host will have these users, these packages, ...) 2. Per-customer configuration ...
14:55 babilen ... (for foo-*) 3. Per-custumer + Per-Role (all database server of customer foo, foo-db*) and 4. Per host ... (customers are pesky and might need very specific adaptations)
14:55 babilen What are sensible ways to tackle this?
14:56 Katafalkas joined #salt
14:57 harobed joined #salt
14:57 Nazzy babilen, wildcards in top.sls perhaps?
14:57 babilen My idea so far was to write something along "roles" or "profiles" that are backed by state files. The roles should be set via pillars (* → global.sls, foo-* → customer/foo.sls, ...) that each reference actual states/formulars.
14:58 babilen The basic idea was to reference fairly generic formulas in the state files and to customize them with different data in the pillars as that also ensures that data is specific to minion classes.
14:58 Nazzy that's also a good option if states are being shared in a sufficiently complex fashion
14:59 Nazzy ah rats, I should have gone to grab food 10 minutes ago *disappears*
15:00 babilen Well, right now I am trying to work on a Install packages and try to implement that with *one* state files that reads "packages" from the pillar ... the problem is now that I cannot extend that in the pillar. I mean I start with "global" but can't add additional packages in more specific pillars.
15:00 Ahlee babilen: Yeah, that sounds feasible RE: pushing hte logic complications into pillar
15:01 cachedout joined #salt
15:02 ekristen wow no PGP commands for salt?
15:02 ccase joined #salt
15:02 mgw joined #salt
15:02 babilen So if I want the hierarchy of packages that will be eventually installed are comprised of GLOBAL_PACKAGES (pillar: global.packages) + CUSTOMER_PACKAGES (foo-* in the pillar foo.packages) + ROLE_PACKAGES (foo-db* in the pillar foo.db.packages) ...
15:02 MTecknology GradysGhost: I was going to buy a salt lick when I got there and bring it along.
15:03 babilen I can't really do that as I cannot easily extend the same pillar variable (i.e. packages) -- All I can see now is that I actually need states on each of these abstraction layers.
15:04 babilen What I would like is a general way to do this however in the sense that CUSTOMER_PACKAGES will be installed if I define the foo.packages pillar data or not if that is not there.
15:04 foxx joined #salt
15:05 babilen So to allow me to do things: Hey, customer wants $PKG on all boxes so I can add it to foo.packages and $OTHER_PACKAGE on all db servers so lets add it to foo.db.packages (id based match on foo-db*)
15:06 babilen foo-db1 would have different roles: global, foo, foo-db and foo-db1 and corresponding states.
15:06 jinnko_ joined #salt
15:06 pythopticon joined #salt
15:07 jinnko joined #salt
15:07 babilen But I can't seem to do this as pillar data is not extendable (seems to not be) ... so my plan is somewhat thwarted and I have the feeling as if I want to do something unidiomatic. What are other/better approaches?
15:07 babilen Sorry for EWALLOFTEXT :)
15:08 miguitas joined #salt
15:10 bhosmer_ joined #salt
15:11 mlen joined #salt
15:11 pythopticon joined #salt
15:12 mlen joined #salt
15:13 miguitas Hi, I using salt-api and return a weird thing using wheel module http://paste.openstack.org/show/62172/
15:13 simpleirc4 joined #salt
15:14 miguitas thanks in advance
15:14 timc3 salt.states.file.managed I can pass the file through a template, render it and then pass it back. But how do I get the current file contents into the template?
15:15 timc3 Is there a variable that gets included?
15:16 ndrei joined #salt
15:19 cachedout_home joined #salt
15:22 martoss left #salt
15:23 alunduil joined #salt
15:24 abe_music joined #salt
15:24 tempspace joined #salt
15:26 xmltok joined #salt
15:28 dlindquist joined #salt
15:31 higgs001 joined #salt
15:33 dccc joined #salt
15:33 rojem joined #salt
15:37 Katafalkas joined #salt
15:37 mlen joined #salt
15:37 bhosmer_ joined #salt
15:37 cnelsonsic joined #salt
15:40 miguitas nobody?
15:41 tvle joined #salt
15:41 cedwards miguitas: most of the devs and core contributors are currently at Salt conference so things are quiet in here.
15:41 cedwards miguitas: I might suggest trying the mailing list.
15:42 miguitas oh, thank you
15:43 cedwards miguitas: https://groups.google.com/forum/#!forum/salt-users
15:44 ekristen timc3: you can’t get the current file contents, its a managed file
15:45 ekristen timc3: salt is managing it with the contents you provide it
15:45 timc3 ekristen: thanks, just figured that out so I provided new file contents for it
15:46 ekristen yeah I usually put a files/template.jinja and use salt://state/files/template.jinja for the source
15:46 justBob_ joined #salt
15:46 fllr joined #salt
15:46 timoguin joined #salt
15:49 NotreDev joined #salt
15:56 ccase joined #salt
16:03 Iwirada left #salt
16:05 ndrei joined #salt
16:09 [diecast] joined #salt
16:10 Thiggy joined #salt
16:14 n0c so if I have, say, multiple wordpress sites that I'm managing with saltstack, would it be best to create a directory with an init.sls (plus nginx config, etc) for each one, or is there a better way (possibly with pillar) to organize everything?
16:16 che-arne joined #salt
16:17 timoguin n0c, so they'll have the same configuration, just config values changing.
16:17 timoguin write one state that will loop through a list of the sites defined in pillar
16:17 timoguin and do the same operations for each one
16:17 n0c they should be /mostly/ the same
16:18 timoguin anything that needs to be configurabble per-site you should be able to set in pillar
16:18 harobed joined #salt
16:18 hunter_ joined #salt
16:18 chadhs joined #salt
16:19 timoguin so each site would be a key in pillar somewhere, and then per-stuff like 'wordpress_user', 'wordpress_password", etc. can be set under it
16:19 timoguin *per-site stuff
16:19 n0c cool, thanks
16:20 jimallman joined #salt
16:22 MindDrive joined #salt
16:25 shiin left #salt
16:26 rajeshver joined #salt
16:26 rajeshver how is it working?
16:28 hunter joined #salt
16:28 carmony go IRC channel! :D
16:29 tempspace yay IRC :)
16:29 hunter bah.
16:29 hunter All human communication should be through twitter.
16:29 whiteinge gah!
16:29 zfouts joined #salt
16:30 timoguin this is all the microbloggin I need
16:30 nicksloan joined #salt
16:32 n0c so timoguin building off that idea of storing each site's configuration in pillar
16:32 n0c do I need to have all the configuration in one file in order to loop over it?
16:33 Mua joined #salt
16:33 whiteinge Ahlee: http://salt-api.readthedocs.org/en/latest/topics/netapis/index.html#saltapi.APIClient.local_batch
16:33 n0c that is... do site1 and site2 need to be in, say, /srv/pillar/websites.sls
16:33 n0c or can I somehow do /srv/pillar/websites/site{1,2} and iterate over the contents of the websites dir
16:33 timoguin n0c, i would do one file. and you have one key defined in there called websites
16:34 timoguin so for pillar, the SLS files are really only used so the top-file can assign specific files to minions
16:34 timoguin but the key-value pairs within those files are really all that matter
16:34 nicksloan I've seen a bit of griping about the use of socket.getfqdn() vs socket.gethostname() for determining the minion ID. In my case, the fqdn is pretty much useless, and the hostname could solve all of my problems. Has there been any discussion of workarounds for this circumstance?
16:35 nicksloan the best I can come up with is forgo the minion ID altogether and match on a custom grain.
16:35 timoguin nicksloan, if you set an id in the minion config, that overrides everything
16:35 bhosmer_ joined #salt
16:35 cachedout joined #salt
16:36 nicksloan timoguin: sure. But how do I set a minion ID for every instance? I'd prefer not to modify base image. Is minion id supposed to be unique among all minions?
16:37 timoguin yes, it has to be unique
16:37 Nazzy where the heck did the base ext_pillar disappear to? ¬_¬
16:37 dangra joined #salt
16:38 dlindquist joined #salt
16:38 Nazzy ah, found ya
16:38 tvle joined #salt
16:38 anuvrat joined #salt
16:38 timoguin nicksloan, it depends on how you're deploying your minions. i use salt-cloud and set the minion ids when i spin up the new instances.
16:38 timoguin and for vagrant i have a unique id set in the minion config for each box
16:39 mlen joined #salt
16:39 CheKoLyN joined #salt
16:39 Nazzy aaaah, /thats/ what the pillar variable is for
16:40 CheKoLyN joined #salt
16:40 cachedout joined #salt
16:41 Katafalkas joined #salt
16:41 Ahlee whiteinge: *hugs*
16:41 Ryan_Lane joined #salt
16:42 Ahlee course, starting tomorrow i'm gonna start ripping out all the stuff hitting salt-api and move it to reactor calls
16:42 Ahlee but I continue to appreciate your hard work.
16:42 whiteinge ha :)
16:42 dlindquist joined #salt
16:43 Nazzy https://github.com/saltstack/salt/blob/develop/salt/pillar/mysql.py#L102 is very confusing
16:43 Nazzy pillar is a dict
16:44 tedski joined #salt
16:44 mlen joined #salt
16:45 Nazzy time to upgrade the pillar ^_^
16:46 jeremyfelt joined #salt
16:47 forrest joined #salt
16:47 gnu-man joined #salt
16:47 tedski this is frustrating... i'm having issues with state.highstate where all i get is Minion did not return, but nothing in the debug logs on either minion or master
16:47 tedski state.sls works fine if i explicitly identify the state to apply
16:47 forrest tedski, did you already restart the master/minion services in debug mode?
16:48 kermit joined #salt
16:48 tedski forrest: yes
16:48 tedski the same state is identified in the top.sls
16:48 forrest tedski, hmm ok. Did you go through all the states that are getting applied? Are any long running?
16:48 helderco joined #salt
16:48 mgw joined #salt
16:48 forrest and are job ids being generated during the highstate?
16:48 tedski i'm setting -t 0
16:48 seanz joined #salt
16:48 cedwards tedski: have you tried running salt-call state.highstate on the minion for more output?
16:48 tedski job ids are being created
16:49 tedski forrest: if i run state.sls, it finishes up fine with -t 0
16:49 n0c timoguin: sorry to keep bothering you with this, but if I keep all my sites configuration in a single websites.sls file in pillar, how can I say 'these sites should be on minion1, these on minion2'?
16:49 pdayton joined #salt
16:49 tedski cedwards: not yet, let me do that
16:49 n0c the obvious answer being one sls file for each minion, but that could end up in a lot of copy/pasting
16:49 n0c say I want the same site on two hosts, do I define it's configuration twice?
16:49 jsm joined #salt
16:50 gnu-man joined #salt
16:50 tedski hrmm... state.highstate works fine via salt-call
16:50 tedski *facepalm* and now retrying it via salt from the master works fine
16:51 forrest tedski, did you need to sync over pillar or anything?
16:51 gadams999 left #salt
16:51 fllr How can I get a list of host and ip addresses of machines whose keys have been accepted by my master on a state file?
16:52 scoates n0c: I do a lot of shenanigans with pillar and a custom module to match groups, to manage that. it's not pretty.
16:52 timoguin n0c, hmm, maybe have a pillar dict that defines what host patterns or grain patterns get what sites
16:52 timoguin loop over that, and then have another pillar dict that defines the configs for each site
16:52 tedski forrest: nope
16:52 matanya joined #salt
16:53 tedski forrest: and the custom grains were already latest versions, so they were skipped of course
16:53 n0c that could work, I was actually just messing with something along those lines
16:53 forrest tedski, you're running the same version across systems right?
16:53 tedski forrest: all the state files were latest in cache, too
16:53 StDiluted joined #salt
16:54 tedski forrest: yup, 0.17.5 on ubuntu 12.04 across the board
16:54 forrest weird
16:54 tedski tell me about it
16:55 StDiluted_ joined #salt
16:55 scoates n0c: I have this http://paste.roguecoders.com/p/95960a98e60b7313a58f7710807a9ad8.txt … like I said… not pretty.
16:56 fllr How can I get a list of host and ip addresses of machines whose keys have been accepted by my master on a state file? (I basically wanna create a /etc/hosts file using salt)
16:56 tedski here's another question:  if i have a nodegroup with host1 in it, should salt 'host1' state.show_highstate work?
16:56 tedski and show the highstate for the nodegroup, too?
16:56 tedski in other words, does the master find what nodegroups the host belongs to and apply those states?
16:56 forrest fllr, no need to repaste, if someone knows they usually answer, a lot of the regulars are at saltconf this week so we're very busy/distracted.
16:56 scoates fllr: maybe this can help you? I haven't actually used it yet. http://docs.saltstack.com/topics/mine/
16:57 n0c scoates: that's not too bad, where would something like this go?
16:57 n0c I've not written any modules or anything yet
16:57 fllr scoates: Thanks!
16:57 fllr forrest: Sorry, I won't do it again...
16:57 scoates n0c: I have it in roots/_modules/fksite.py … and I call it in a weird way with jinja
16:57 n0c :)
16:58 scoates I'd really love it if 1) pillar was a two-pass render, and 2) there was a helper for host globbing available to pillar, without module hackery.
16:58 forrest fllr, no worries I wasn't trying to slap your wrists or anything, just giving you a heads up! :D
16:58 forrest fllr, sometimes it's totally justified.
16:58 fllr Lol. No worries...
16:59 fllr I used to maintain an irc channel before, so I understand it.
16:59 forrest psssh, I don't maintain, you think they could sucker me into that??
16:59 jrdx Speaking of, I hope they post that presentation about standards and best practices... I'm really struggling with that over here. I've found a lot of bad, and some sketchy, ways to do things... I'm not sure I've found a lot of great ways to do things.
16:59 mgw joined #salt
17:00 scoates speaking of questionable practices: http://twitter.com/coates/status/428932071164092416 (-;
17:02 nicksloan scoates: whose mom is that repository named after?
17:02 scoates nicksloan: Salt -> Pillar of Salt -> http://en.wikipedia.org/wiki/Lot%27s_wife
17:03 nicksloan haha
17:03 nicksloan I should have expected something like that
17:05 forrest jrdx, there should be slides on that talk even if the video isn't available.
17:07 bemehow joined #salt
17:07 jrdx Awesome, looking forward to that! I'm pretty new to the project, and I've been struggling with NOT applying a "this is how you would do it in code" flow. Trying to wrap my head around the "proper" way to do things.
17:08 toofer joined #salt
17:08 forrest jrdx, yea totally, what portions are you struggling with the most?
17:08 xmltok joined #salt
17:08 forrest jrdx, I have slides for writing formulas (and therefore states) that focuses on a few things as well, so I'd be happy to share those with you
17:09 forrest depending on where you are having problems of course, no reason to spam you with crap you don't need :P
17:09 bhosmer_ joined #salt
17:10 markizano joined #salt
17:13 simio joined #salt
17:15 dangra1 joined #salt
17:16 Mua_ joined #salt
17:17 jrdx I've only been using Salt for a few weeks. We've got a lot working, especially stuff that's default linux stuff... we're struggling a bit when it comes to deploying our custom code artifacts. I've got it working, but it seems sketchy, I'm not sure what the "best" way to do something like:  download a new tomcat binary zip if the version has changed, extract it, stop the apps, clear out the old version, move it to the target directory,
17:17 jrdx apply permissions, start the app. It's somethign I can see as a very clear set of steps in a script, but I'm not sure how to translate that to Salt logic. It seems to be a mess of requires, prereqs, and watches at the moment to link all those actions together in a chain.
17:17 Nazzy scoates, nicksloan: salts do really weird stuff if you get a saturated solution and evaporate it... you can under certain cases end up with fairly solid pillar shapes
17:18 forrest jrdx, so right now does that stuff all happen within a single state file?
17:19 markizano I'm in a bit of an architectural wall here... I'd like to think this would work, but it's just not panning out...
17:19 markizano I've turned the top.sls in the pillar and the states to python, and it's the same among all branches/environments.
17:19 jrdx forrest, yeah, but more for me to easily proof-of-concept and figure out how things should be working and tie together. I'm sure a lot of that can/will be chuncked up and reused once I figure out the flow.
17:20 forrest ok cool, just so we can get an idea of what you are doing now, can you pastebin/gist it? And does it currently work?
17:20 tedski doh
17:20 Nazzy scoates, nicksloan: given the location and composition of the Dead Sea, it's gotta be a valid option for the origin of that particular part of the bible
17:20 tedski i just read the source code
17:20 tedski i assumed '-t 0' disabled the timeout
17:20 tedski silly noobsauceski
17:20 markizano The issue I run into is I have <base>, <production>, <integration>, and <FeatureA> - When I'm telling a minion you apply to base and FeatureA - it won't get FeatureA's pillar even though it's being returned.
17:21 forrest Nazzy, you realize you are talking in the main channel right?
17:21 forrest jrdx ^
17:21 forrest sorry I didn't tag you previously
17:22 SEJeff_work Nazzy, pretty sure you're in the wrong channel. This is about some software named salt
17:22 forrest tedski, oh no no no
17:22 forrest tedski, that just makes it instantly return to the console,
17:22 forrest *.
17:22 tedski forrest: hence all the "Minion did not return"s :)
17:22 forrest yea
17:22 tedski doh
17:22 forrest tedski, sorry, I thought you were expecting it to drop and it was giving you an extra error, hah.
17:22 whiskybar joined #salt
17:26 Nazzy lol, I was responding to a remark from like 20 minutes ago as to where "Pillar of Salt" comes from
17:26 forrest Tom actually explained where the name 'Salt' came from at the keynote today, and that should go up as a vdeo
17:26 forrest *video
17:26 jeremyBass1 joined #salt
17:26 Nazzy and yes, I know where I am ... this is the channel where I random remarks and occasionally say something helpful
17:27 forrest Yea I was gonna say, Nazzy is a normal around here unless his account was hacked.
17:28 cedwards forrest: hey, I noticed the nginx-formula doesn't have a map.jinja
17:28 forrest really?
17:28 Nazzy only one hacking me is my code ... wait, is that the right way round?
17:29 quickdry21 joined #salt
17:29 jrdx forrest, here's my current attempt, which works... but most reliably from a clean start. It's my attempt at A) building in some variables so I can reuse it later, B) naming the IDs in a way that wont class at other projects, and C) tying all this together in a chain. http://pastebin.com/FxjkkEu0
17:29 forrest jrdx, ok cool
17:29 tedski since network.managed doesn't work on ubuntu yet, what's the canonical way to manage bonding with saltstack
17:29 forrest cedwards, I see it now, I think I overlooked that one last week as 'too much stuff to do before the conference' :\
17:30 tedski the problem i run into is that networking must be stopped before /etc/network/interfaces is edited
17:30 tedski and that breaks master-minion connections
17:30 cedwards forrest: are you in C.R.'s lab?
17:30 forrest I am
17:30 markizano http://markizano.net/files/salt-run.txt <- Here's some relevant snippets to the code I'm writing in the top.sls files.
17:30 ch0psti95 joined #salt
17:30 markizano In the pillar: I return data so that it can apply the base and FeatureA branches.
17:30 markizano but I don't get data in FeatureA branch.
17:30 markizano I get base pillar...
17:31 markizano Is the pillar only ever applied once in a state run so if base ever catches - it's done searching, ever ?
17:31 forrest ok so jrdx, the first thing I notice is there are a lot of values getting set in this state, and it would be cool if you moved those to pillar.
17:32 jrdx forrest: yeah, that's the plan.. it's just easier to deal with in here for proof-of-concepting. I figure that's easy to break out and set properly later.
17:32 forrest so you could create a pillar associated with the app, OR better yet, you could break that out so you had something like a users pillar, an app pillar (or multiple app pillars0
17:32 forrest jrdx, yea for sure
17:32 KyleG joined #salt
17:32 KyleG joined #salt
17:32 markizano How can one specify which environment he wants to get data ?
17:32 jacksontj joined #salt
17:33 markizano pillar.get(key, default=None, saltenv='base') ?
17:33 Mua joined #salt
17:34 KyleG morning #salt
17:35 cedwards KyleG: morning
17:36 KyleG cedwards; You play around with BSD 10 yet
17:36 forrest so jrdx, some of this stuff is good, like using {{ dir }} to create the dir since it's very modular, there is a tomcat state you can use: http://docs.saltstack.com/ref/states/all/salt.states.tomcat.html
17:36 cedwards KyleG: I have it in a VM is all.
17:36 KyleG Ah cool.
17:36 KyleG I love how finally I can setup ZFS as the root FS without going into the shell and having to type a ton of commandslol
17:36 KyleG so much time saved
17:37 jrdx forrest: yeah, I saw that.. we're not quite up to Tomcat7, and pull from the binary zip as the base... so I wasn't sure I could use that.
17:37 mpanetta joined #salt
17:38 cedwards KyleG: have you used the 2014.1/develop on FreeBSD yet? I'm tempted to post an RC patch for testing.
17:38 KyleG Not develop I don't think
17:38 dlindquist joined #salt
17:38 KyleG We're running 2014.0.1 I believe on 9.2 still
17:39 KyleG I only have 2 servers on BSD 10 though, and they're DB servers so I don't even manage them w/ salt…although I probably should. I just don't like DB's breaking due to bugs;)
17:39 forrest jrdx, yea gotcha, none of this really seems that bad. I think what you are really missing here is a more modular idea of how you can accomplish stuff, so breaking it into multiple states, and structuring it clearly. I think when you move these variables to pillar, and split stuff into multiple states it will become a lot more clear. Take a look at how UtahDave is doing some stuff with users/groups: https://gist.github.com/UtahDave/3785738
17:39 forrest like all those dirs should live in pillar
17:40 forrest and then you move over them and so on
17:40 cedwards wow. super crowded in the docker presentation
17:41 simio joined #salt
17:43 hunter joined #salt
17:43 jrdx forrest: so is it preferable to build dependencies top down, or bottom up? For instance, bottom up: version changed, so: clear the local zip cache since it's not valid... if the cache is cleared, redownload the proper one, if one was downloaded extract it, if one was extracted clear out the "live" one (which needs apps stopped first), if the "live one was cleared" move the cached one in there, if there's a new one move in start the ap
17:43 jrdx ps.... etc. That seems a bit complicated and hard to explain to someone.
17:43 troyready joined #salt
17:44 jrdx Top down seems to make more sense I think... but it seems equally as complex.
17:44 forrest cedwards, I went to halite, let me know ho that one is, and if docker takes a crap on itself for no reason.
17:46 dangra joined #salt
17:48 jsm joined #salt
17:49 aleszoulek joined #salt
17:49 forrest so jrdx, I'd consider doing something like /srv/salt/app/init.sls which just installs the tomcat stuff, then a second app/update.sls which handles updating and clearing if you can't just overwrite. You want to think of it in terms of chunks that you could easily include in other items to achieve a task, so an install task, a cleanup task, etc.
17:50 forrest I don't have a ton of great examples of doing this with java which makes it a bit tough, Even with this data it might be a good idea to put that pastebin up on the mailing list, to see what others say, since then some java gurus might be around
17:51 mgw joined #salt
17:51 jrdx cool, I may do that... cause it's practically the same flow for any of that stuff: java, tomcat, jboss.
17:51 joehillen joined #salt
17:52 harobed joined #salt
17:52 forrest jrdx, yea exactly.
17:53 mpanetta joined #salt
17:53 zfouts joined #salt
17:53 forrest jrdx, I'd be interested to see how people handle that, because it's such a mess deploying java apps generally.
17:53 miguitas joined #salt
17:54 zfouts I provide much love to whoever wrote http://docs.saltstack.com/ref/states/all/salt.states.xmpp.html#module-salt.states.xmpp
17:54 zfouts forrest: whoa whoa, java apps? Thats what I just joined for :)
17:54 forrest lol
17:55 forrest zfouts, if you scroll up jrdx is just starting with Salt, we've discussed moving stuff into pillar, but there's still the issue of cleaning dirs and all that crap: http://pastebin.com/FxjkkEu0
17:55 markizano lol
17:55 zfouts I have to deploy a java app
17:55 zfouts forrest: I just joined :(
17:55 forrest have you written states for it yet?
17:55 forrest what a slacker
17:55 zfouts no, not yet!
17:55 zfouts Getting ready to
17:55 jrdx zfouts: that xmpp state seems pretty handy
17:56 zfouts jrdx: indeed! I was writing a bot a few months ago and never finished, so happy someone else did it for me now ;)
17:56 fllr joined #salt
17:56 forrest zfouts, looks like that one was techhat, s0undt3ch, and terminalmage
17:56 dangra1 joined #salt
17:56 zfouts forrest: awesome, cheers to them
17:57 zfouts The nastiness I am contending with is a jar installer file. Said jar installer deploys all the other goodies…kind of makes for a difficult automation process
17:57 forrest Yea Pedro is actually here
17:57 forrest s0undt3ch
17:57 forrest zfouts, yea that's the same thing jrdx and I were talking about, because he needs to check for versions to see if it's an update, then clear all the dirs, and re-deploy
17:57 zfouts I am trying to convince them to split the jar file up into rpm
17:57 forrest I was thinking about that same thing for jrdx, but that's kind of an annoying amount of work
17:58 forrest for a small team :P
17:58 zfouts if we can split it into an rpm for the initial installer, we can either A) update the RPM, or B) deploy just the war/jar files that need to be moved
17:58 zfouts yeah, tons of work…my whole company is like 30 people total
17:58 forrest you know...
17:58 forrest it's dirty
17:58 forrest but FPM might work
17:58 forrest I hate it
17:58 zfouts FPM is great for PHP & NGINX
17:58 forrest does it fail for java stuff?
17:58 zfouts i've never used it for java
17:59 zfouts I have thankfully avoided java up until this point in my career
17:59 forrest https://github.com/jordansissel/fpm if anyone is curious what we are discussing it
17:59 forrest zfouts, lucky you
17:59 forrest I'd like to see how that handles java..
17:59 mpanetta_ joined #salt
17:59 forrest because if you could package it into an RPM, you could easily control versions, AND remove all excess files when you moved from release to release, which solves almost all the issues.
18:00 zfouts our app used to be an rpm
18:00 tvle joined #salt
18:00 zfouts they just changed it 2 release cycles to a .jar installer
18:00 forrest that's dumb
18:00 forrest why would they do that?
18:00 forrest packaging the jar is easy
18:01 forrest if they already had it happening
18:01 zfouts no idea, I wasn't part of that discussion since I've been here <1 week :)
18:01 zfouts well, prior it was python
18:01 jacksontj joined #salt
18:01 zfouts they rewrote it in java
18:01 zfouts so maybe they just don't know how to package it into an rpm
18:01 zfouts we'll see
18:01 zfouts I've been tasked with automating the install so we can pick a version and auto-deploy a cloud server and a version
18:02 forrest well, an RPM would do that for sure
18:02 zfouts yep
18:02 forrest are they using jenkins?
18:02 zfouts nope
18:02 zfouts no CI here yet
18:02 forrest you should get them using jenkins
18:02 zfouts I hate jenkins
18:02 forrest to build their stuff out
18:02 zfouts :\
18:02 forrest well, you're in a java shop
18:03 forrest it would be easy to run their jobs after they finish getting the jar together to package it up
18:03 forrest and push it to the repo
18:03 zfouts yeah
18:03 forrest or you could write me a python alternative, and I will use that
18:03 zfouts I need to get some hardware to build this stuff, right now I only have AWS and a macbook
18:03 forrest aren't you a manager now
18:03 forrest *?
18:03 forrest so your team should be doing this?
18:03 zfouts yes
18:03 zfouts yes, but they don't know how
18:03 zfouts they're still focused on their normal every day stuff
18:03 forrest pass out the books and tell them to learn then :P
18:04 forrest otherwise you're gonna get bogged down doing all the support for that stuff, and being a manager
18:04 ch0pstick joined #salt
18:04 zfouts I'm just doing a proof of concept right now
18:04 zfouts not a big deal
18:04 forrest gotcha
18:04 tvle joined #salt
18:04 zfouts setup automated install process, show them, let the devs improve on it and write salt modules
18:05 markizano I'm at a point now where we've got a decent setup working, just need things to be consistent now...
18:05 zfouts markizano: that's the most difficult part!
18:05 zfouts good luck
18:05 jrdx zfouts, ha... I'm in that boat as well
18:05 markizano Indeed :/
18:06 markizano It's like... if I could just get my feature branch to get its data from the pillar...
18:06 zfouts jrdx: it's a great boat to be in
18:06 markizano I'm about ready to start reading stuff off the filesystem and stop asking the pillar nicely for the stuff i want ....
18:06 markizano but that would be extremely hacky :\
18:07 zfouts I still don't use pillars enough
18:07 zfouts I really need to
18:07 markizano When I do include the pillar data in the state run with globals, i get complains about global name conflicts...
18:08 forrest zfouts, you should watch my talk on formulas
18:08 forrest and then start using pillars
18:08 forrest once it is up
18:08 forrest because shame on you
18:08 jrdx forrest: i would be interested in watching that as well
18:08 forrest oh yea let me get you the slides.
18:08 markizano been using the pillar etensively...
18:08 jrdx One other thing I'm not sure how to handle: recursively setting the perms for files in <myapp>/ to 640, then setting <myapp>/bin to 740 fights with eachother. Any way to exclude /bin in the <myapp>/ definition?
18:08 markizano tried using multiple pillars
18:08 markizano but that didn't work :/
18:09 zfouts zfouts: video plz
18:09 forrest zfouts, yea I don't have the video
18:09 zfouts zfouts: oh sadness
18:09 forrest but they were recording it I believe
18:10 forrest http://goo.gl/wg6YZN there are the slides, imagine that with me repeatedly saying 'make your stuff modular!', and that is the gist
18:10 forrest and making the mic buzz at one point
18:10 Nazzy this is a bad sign ... it's taken me most of an hour to write the comment explaining what this function is going to do
18:11 zfouts hah. I love the graphics in this
18:12 Nazzy once it all works it will be hotness, but thinking of writing the actual documents is a scary thought
18:12 forrest Nazzy, lol
18:13 Nazzy query and config shaped multidimensional pillars from MySQL
18:13 tl-ml joined #salt
18:13 zfouts wow this is cool forrest thank you for sharing it
18:13 * zfouts contracts forrest to work for him full time as a salt engineer
18:13 matanya joined #salt
18:14 forrest zfouts, yea np, the slides are better than the talk, where I'm just going 'please use formulas, because they kick ass'.
18:14 zfouts unfortunately, I don't have a use for that currently
18:14 zfouts We have one application, so the state files should be exceptionally small
18:15 zfouts Just need to make sure the JVM is running, sshd is up, and iptables rules are in place
18:15 zfouts dead simple
18:16 Katafalkas joined #salt
18:20 zfouts Where is the salt repo with the states examples?
18:20 Nazzy which one? :p
18:20 LowWalker joined #salt
18:20 zfouts Hmm, the one with awesome in it?
18:20 dangra joined #salt
18:21 Nazzy again, which one ... you're thinking salt-formulas?
18:21 zfouts yeah, thats the one
18:21 zfouts ty
18:22 forrest Nazzy, https://github.com/saltstack-formulas/
18:23 Nazzy in order of approximate age: https://github.com/saltstack/formulae https://github.com/saltstack/salt-states https://github.com/saltstack-formulas :p
18:23 zfouts oh man, so many formulas
18:24 quickdry21 joined #salt
18:24 Nazzy though, some of the ones in the later repo date to before the first repo ... so more accurately it's last commit date
18:25 Nazzy heh, https://github.com/saltstack-formulas/elasticsearch-logstash-kibana-formula ... I got my hopes up for a moment
18:26 fllr joined #salt
18:27 jrdx Nazzy: me too...
18:27 mgw left #salt
18:30 mgw joined #salt
18:31 matanya_ joined #salt
18:35 jrdx One other thing I'm not sure how to handle: recursively setting the perms for files in <myapp>/ to 640, then setting <myapp>/bin to 740 fights with eachother. Any way to exclude /bin in the <myapp>/ definition?
18:35 abe_music joined #salt
18:36 Zimmie joined #salt
18:36 KyleG1 joined #salt
18:36 Zimmie Has anybody here tried Salt on FreeBSD 10? The pkg provider doesn't seem to work for me.
18:37 Nazzy jrdx, hmmm, I've a vague recollection of an exclude somewhere in there, but I could be miles off
18:37 Nazzy jrdx, you could have two states and set the /bin one to run after the parent one via a requisite
18:37 Nazzy jrdx, but that will flipflop every time, which isn't ideal
18:38 Zimmie It worked on 9.2 with the original package manager, then it still worked after I updated to pkgng. FreeBSD 10 defaults to pkgng. I've tried overriding the pkg provider in the minion config as well as calling the pkgng module directly.
18:38 jrdx yeah.. that's what I'm seeing now... the flip-flopping
18:38 ndrei joined #salt
18:40 Nazzy I think my fiancee is going to attempt bodily harm on me o_O Her: "how goes?" Me: "in a non-stationary fashion" H: "how does finishing look?"  M:"depends on how well you polish it" H:"how are you liking the idea of death? :p" M:"it leaves me cold? *shrug*"
18:41 fllr joined #salt
18:41 n8n joined #salt
18:41 dangra joined #salt
18:41 Zimmie_ joined #salt
18:42 Nazzy jrdx, file.recurse
18:44 Nazzy jrdx, see if it preserves modes ... if not, file an issue to extend it with mode preserve and/or file.directory with a better exclude filter heh
18:46 Nazzy jrdx, file.directory has an exclude filter, but only for clean and I don't think it'll be setup as an exclude for anything else
18:46 Nazzy it shouldn't actually be too hard to implement either feature
18:49 jrdx Nazzy, cool thanks... I saw the exclude but came to the same conclusion. Guess it's not just me.
18:49 iShming_ joined #salt
18:50 cachedout_home joined #salt
18:50 baniir joined #salt
18:52 Nazzy jrdx, yeah, I'm basing that off what I remember of the function layout... I could be wrong but I think I have it right.
18:55 Katafalkas joined #salt
18:57 DanGarthwaite joined #salt
18:57 helderco My containers hang when I use multiple file_roots directories. Any idea why or how to troubleshoot?
18:57 taion809 joined #salt
19:01 mgw joined #salt
19:01 Nazzy sneak peek for those who can interpret coder gibberish: https://github.com/kaithar/salt/commit/e7bb6bbe4cd22f854ebb5ff9882a0b5cee43fa08#diff-2e303814613940ed6c8699a88601e551R104
19:02 scoates I want that thing forrest posted about.
19:06 fllr joined #salt
19:07 amckinley joined #salt
19:07 Ryan_Lane joined #salt
19:07 zfouts damn, that's a problem. I can't have a gifts within a folder in a repo?
19:08 forrest joined #salt
19:10 amckinley1 joined #salt
19:10 amckinley2 joined #salt
19:12 zzzirk joined #salt
19:13 aleszoulek joined #salt
19:13 quickdry21 joined #salt
19:13 baniir joined #salt
19:20 timc3 I am trying to add a new debian repo and install a package from it. It never seems to add the repo. Any examples I can look at somewhere?
19:22 anitak joined #salt
19:26 blee joined #salt
19:29 KyleG joined #salt
19:29 KyleG joined #salt
19:30 Sypher in module ps, function top. is there a reason why its doing info['cpu.{0}'] (and mem) rather than mem['cpu'][key] ?
19:30 Katafalkas joined #salt
19:31 Sypher since mem and cpu contains more results it would make more sense if it would group it like that
19:31 shadowh511 joined #salt
19:32 toofer joined #salt
19:32 nliadm joined #salt
19:34 bhosmer_ joined #salt
19:35 sroegner joined #salt
19:35 simio joined #salt
19:37 tvle joined #salt
19:39 dlindquist joined #salt
19:41 hunter joined #salt
19:46 n8n joined #salt
19:51 ajw0100 joined #salt
19:54 Gareth 'lo
19:56 jalbretsen joined #salt
19:58 Ryan_Lane so, I keep seeing people with issues with aws autoscaling and keys. it would be kind of nice if you could send a one-time-password in to a minion that's known by the master.
19:59 Ryan_Lane it could go into user-data and would be sent along with the initial auth
19:59 martoss joined #salt
19:59 Ryan_Lane then you'd know the minion was legit
20:00 Ryan_Lane but you wouldn't need to store anything permanently useable in the user-data
20:00 [diecast] joined #salt
20:01 zfouts I keep getting this error "blah is not a dictionary"
20:02 zfouts https://p.6core.net/p/wArMbKzkwG6r6HF3Ayqs91dX is my state
20:02 zfouts It is located in Folder/Folder/version.sls
20:02 zfouts So state.sls Folder.Folder.version is what I ran
20:06 zfouts weird, I changed the naming of it and it worked
20:07 zfouts Oh I also had a typo - no wonder ;-)
20:10 rchrd joined #salt
20:11 Katafalkas joined #salt
20:11 helderco joined #salt
20:11 zooz joined #salt
20:14 mr_chris joined #salt
20:14 alunduil joined #salt
20:20 dlindquist joined #salt
20:21 ajw0100 joined #salt
20:22 bemehow joined #salt
20:23 cachedout joined #salt
20:23 [diecast] joined #salt
20:25 bhosmer_ joined #salt
20:26 simio joined #salt
20:27 LucasCozy higgs001, is it normal that when I call cmd_async then get_cli_returns on my job. The cache file is still present in /var/cache/salt/master/jobs/
20:28 tedski i'm trying to run a state.highstate on a minion by glob match... test.ping returns True, but state.highstate returns     "state.highstage" is not available.
20:28 tedski any idea what might cause that?
20:28 LucasCozy I have about 7GB of this file right now in a day.
20:28 tedski running salt-call state.highstate from the minion works properly
20:28 scoates tedski: highstage ?
20:28 LucasCozy tedski, type stage??
20:28 LucasCozy tedski, typo stage??
20:28 tedski hahaha
20:28 tedski thanks, guys
20:29 * tedski kicks himself
20:37 Katafalkas joined #salt
20:37 Mua joined #salt
20:37 Katafalkas Is it possible to run ZeroMQ outside the master ?
20:40 pydanny joined #salt
20:40 druonysus joined #salt
20:41 martoss joined #salt
20:41 tvle joined #salt
20:42 elfixit joined #salt
20:44 alekibango joined #salt
20:50 dlindquist joined #salt
20:50 Ryan_Lane joined #salt
20:52 Mua_ joined #salt
20:55 kalloc joined #salt
20:56 kalloc joined #salt
21:03 KyleG joined #salt
21:03 KyleG joined #salt
21:05 vejdmn joined #salt
21:12 martoss joined #salt
21:15 kermit joined #salt
21:16 sroegner is there any way to pass a dictionary through the context to a jinja template?
21:17 jrdx Anyone have a better way to download/install java via binary zip, and/or see anything I can do better in my watch/prereq chain? I tried to get clean_java_home to be a prereq for install_java, but I couldn't get it to trigger properly off the resulting state if install_java (tried moving install_java to a cmd.wait as well, and had the same issues): http://pastebin.com/ZQYqjpKN zfouts
21:17 LowWalker joined #salt
21:18 jrdx Is using watch AND require redundant, or is it need to tright both "do this" and "do this before"
21:18 jrdx *tright = trigger
21:21 cachedout joined #salt
21:24 markizano Whelp....
21:24 markizano Reading off the filesystem proves to be more reliable....
21:24 markizano Only kunundrum seems to be now that all minions will be responsible for tracking the pillar upon which the branches they are applied....
21:25 markizano ^^ e.g. I'll have to git-clone the pillar onto each minion in order for it to know what's what ...
21:28 ajw0100 joined #salt
21:29 mgw joined #salt
21:29 sroegner jrdx: you can look at the formulas in the saltstack-formulas org but neither does what your code tries to do
21:30 jrdx stronger, that's the java-formula and sun-java? Or were there others you were thinking of specifically?
21:31 cachedout joined #salt
21:31 jrdx ^ sroegner
21:31 tedski is there a way to kick off a job on a nodegroup in serial?
21:31 sroegner these two
21:32 giantlock_ joined #salt
21:32 sroegner jrdx: i am not sure you want to do the cleanup part
21:34 jrdx sroegner: my thinking is that it always forces a "fresh" install, which seemed safer when dealing with stateless zips. Seemed easier to guarantee consistency and reliability of the files in that directory that way.
21:35 tedski ahh, --batch
21:35 sroegner when i wrote the sun-java formula, my only point of reference is the name of the top level directory in the tarball
21:35 ipmb joined #salt
21:36 sroegner anything beyond that i fell i'd have trouble dealing with
21:36 sroegner so if it already exists it doesn't do anything
21:36 sroegner especially not a download every time
21:38 vejdmn joined #salt
21:38 martoss1 joined #salt
21:40 rubberne_ joined #salt
21:42 helderco joined #salt
21:42 rubberne_ New to Salt, Have a hopefuly quick question. I want to have something in my top file like '*' -standardstuff   I want salt to do its thing on eveyrhint in the standardstuff directory. but it appears I have to do something like -standardstuff.vim etc.... What can I do to include everything in a directory?
21:42 kalloc joined #salt
21:43 QuantumRiff joined #salt
21:44 kalloc_ joined #salt
21:45 QuantumRiff I have salt creating directories for me, and setting up the correct permissions ( http://pastebin.com/XQZCmShg ) however, is it possible to tell it to only do the permission on creation?  Its overriding my user permissions, and I don't want that for auditing reasons
21:45 Nazca__ rubberne_: don't think you can
21:46 Nazca__ QuantumRiff: using file.directory?  would turning off recurse and enforcing just that directory itself be ok?
21:47 QuantumRiff ooh, that is worth a try..
21:47 QuantumRiff I wanted it to recurse the 2775 permissions the first time
21:48 Nazca__ you're putting things in it?
21:48 QuantumRiff Nazca__: thanks.. that was it
21:49 Nazca__ sweet :) np
21:49 QuantumRiff when I was creating it by hand, I was doing a chmod -R 2775, so I carried the recursive into salt
21:49 Nazca__ *nod*
21:54 david_a joined #salt
21:56 mr_chris Finally making good use of the py renderer. Just reduced this http://paste.linux-help.org/view/4cf7251e into this http://paste.linux-help.org/view/68cccc7d
21:56 mr_chris It manages our MySQL users and grants.
21:56 mr_chris Still a bit more refactoring that could be done, though.
21:56 Katafalkas joined #salt
21:57 zfouts Hmm, is there a way to pass input on to a command executed from Salt?
21:59 scoates mr_chris: python renderer? is there a doc page?
21:59 Nazca__ zfouts: in state or module?
22:00 zfouts either or
22:00 mr_chris http://docs.saltstack.com/ref/renderers/all/salt.renderers.py.html
22:00 scoates :stare:
22:00 scoates is that available in 0.17? and all that I have to do is shebang "py"?
22:01 mr_chris Yes.
22:01 Nazca__ yes and yes scoates
22:01 scoates I had no idea.
22:01 Nazca__ also check out pydsl :)
22:01 mr_chris There's also this bit of niceness http://docs.saltstack.com/ref/renderers/all/salt.renderers.pydsl.html
22:01 mr_chris Nazca__: Beat me to it.
22:01 scoates I had modules generating yaml… and that sucked. Just finished a big refactor to work around that mess, but this… this is so much better.
22:01 scoates is it states only? or pillar, too?
22:01 Nazca__ zfouts: modules.cmdmod.run has a stdin param
22:02 Nazca__ http://docs.saltstack.com/ref/modules/all/salt.modules.cmdmod.html?highlight=cmd#salt.modules.cmdmod.run
22:02 scoates :stare: again
22:02 helderco joined #salt
22:02 Nazca__ last option
22:02 mr_chris Nazca__: I didn't know about cmdmod
22:03 mr_chris Does it differ much from cmd.run?
22:03 Nazca__ it is the cmd.run module :)
22:03 mr_chris Ah.
22:03 mr_chris Read it wrong.
22:04 NotreDev joined #salt
22:04 justBob joined #salt
22:04 Nazca__ called cmdmod, likely for reserved name reasons heh
22:04 Nazca__ sadly states.cmd has no equiv to the stdin param
22:05 cachedout joined #salt
22:06 mr_chris Nazca__: Was zfouts asking about stdin or interactive input?
22:07 scoates guess I'll be porting pillar and states to #!py (-;
22:07 mr_chris scoates: I think just states.
22:07 scoates oh? no pillar?
22:07 mr_chris scoates: Wait, I think I am wrong there.
22:07 Nazca__ mr_chris: "passing input" though I tend to interpret that as just pipe to stdin since interactive is very messy
22:07 scoates I have so much mess because pillar's jinja+yaml renderer is one-pass
22:07 mr_chris Pillars can be dynamic and so they can use the py renderer if I am correct.
22:07 mr_chris My mistake.
22:09 scoates this makes me happy. wish I hadn't spend so many hours this week beating jinja into submission.
22:09 scoates thanks, mr_chris
22:09 mr_chris I only went with py over pydsl because the I was able to pass full dictionaries full of state data to a single command whereas with pydsl I would have had to have handled them individually.
22:09 mr_chris scoates: You and me both.
22:09 jsm joined #salt
22:10 scoates sucks: https://dl.dropboxusercontent.com/1/view/0pzl3h23ls2g11z/Captured/UhsLD.png
22:11 Nazca__ oh, if you're really adventurous, there's https://github.com/saltstack/salt/blob/develop/salt/utils/vt.py
22:12 zfouts mr_chris: I was asking about interactive input
22:13 zfouts mr_chris: I have to provide it a "1" at a specific prompt…thought about doing an expect script
22:13 mgw joined #salt
22:14 mr_chris zfouts: I don't think it can do that.
22:14 zfouts mr_chris: yea, I don't think so either
22:14 mr_chris My research lead me here. http://bencane.com/2013/09/23/remote-command-execution-with-saltstack/
22:15 mr_chris See the bottom of the page.
22:15 Napster joined #salt
22:16 hunter joined #salt
22:16 jslatts joined #salt
22:20 mm239 joined #salt
22:21 zfouts thanks mr_chris
22:22 helderco joined #salt
22:23 Nazca__ zfouts: there is that whole vt.py thing I linked
22:24 zfouts thanks I'll look over it
22:24 zzzirk joined #salt
22:25 Ryan_Lane joined #salt
22:29 SEJeff_work Are there going to be videos of saltconf online at any point?
22:31 timoguin SEJeff_work, I've been told yes.
22:31 timoguin Not sure how soon they'll be up though..
22:32 SEJeff_work Certainly hope so
22:32 SEJeff_work I'd like to see the material but couldn't make it this year
22:32 busted joined #salt
22:32 scoates mr_chris: with a `#!py` state, does the python run on the master or the minion?
22:33 Nazca__ scoates:  should be minion, like the others
22:33 timoguin SEJeff_work, same
22:33 scoates cool, thanks, Nazca__
22:33 rchrd joined #salt
22:33 scoates I have great power and great responsibilities (-;
22:34 Nazca__ so, you're in the web spider business :p
22:34 SEJeff_work scoates, pillar data renders on the master and state data renders on the minions
22:34 scoates SEJeff_work: thanks!
22:35 SEJeff_work np
22:37 zz_cro joined #salt
22:38 MindDrive joined #salt
22:38 MindDrive left #salt
22:38 MindDrive joined #salt
22:39 StDiluted joined #salt
22:42 dangra joined #salt
22:43 kalloc joined #salt
22:49 mr_chris scoates: Minion I think.
22:49 diegows joined #salt
22:50 scoates mr_chris: thanks
22:50 martoss joined #salt
22:51 TronPaul_ joined #salt
22:55 bhosmer_ joined #salt
22:57 yomilk joined #salt

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