Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-12-13

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

All times shown according to UTC.

Time Nick Message
00:02 ninkotech__ joined #salt
00:03 KyleG joined #salt
00:03 jcsp joined #salt
00:03 zandy joined #salt
00:04 JordanRinke ohai Gato
00:05 JordanRinke KyleG: el GATO!
00:07 logix812 joined #salt
00:07 pdayton joined #salt
00:08 KyleG oh hai
00:09 JordanRinke today is one of those days where, I am reading documentation... and it just seems like random words put together on a web page
00:10 Gareth We've secretly replaced Jordan's regular documentation with some lorem ipsum generated text.  Lets see what happens.
00:11 psyl0n joined #salt
00:12 cdcalef- anybody have problems with salt-cloud building instances in the wrong region in rackspace?
00:12 cdcalef- or, should i say, having done it without problem
00:15 jslatts joined #salt
00:16 ninkotech__ joined #salt
00:18 cdcalef- looks like this might actually be a libcloud problem :/
00:24 mgw does anybody know how data with newlines should be passed to an execution module?
00:24 mgw I'm working on a module that will need that
00:24 ninkotech__ joined #salt
00:25 mgw In the case of grains.setval, it interprets \n literally, not as a new line
00:27 eculver joined #salt
00:34 zandy joined #salt
00:35 ninkotech__ joined #salt
00:35 jcsp joined #salt
00:41 ninkotech_ joined #salt
00:44 dstanek joined #salt
00:49 ajw0100 joined #salt
00:49 ninkotech_ joined #salt
00:51 zandy joined #salt
00:58 ajw0100 joined #salt
00:58 fllr joined #salt
01:00 ninkotech_ joined #salt
01:01 srage joined #salt
01:02 hellome joined #salt
01:04 KyleG joined #salt
01:04 KyleG joined #salt
01:05 hellome joined #salt
01:06 cachedout joined #salt
01:09 ninkotech_ joined #salt
01:10 hellome joined #salt
01:10 freelock_ joined #salt
01:14 halfss joined #salt
01:15 ninkotech_ joined #salt
01:16 mgw joined #salt
01:17 fllr joined #salt
01:19 freelock_ joined #salt
01:21 cocoy joined #salt
01:23 redondos joined #salt
01:23 redondos joined #salt
01:24 Psi-Jack Gareth: heh
01:26 bt joined #salt
01:27 ninkotech_ joined #salt
01:30 zandy joined #salt
01:33 fandikurnia01 joined #salt
01:45 ninkotech_ joined #salt
01:45 dstanek joined #salt
01:48 quickdry21 joined #salt
01:48 favadi joined #salt
01:51 ninkotech_ joined #salt
01:58 dstanek joined #salt
02:01 ninkotech__ joined #salt
02:04 psyl0n joined #salt
02:09 Furao joined #salt
02:12 ninkotech__ joined #salt
02:12 ninkotech__ joined #salt
02:14 davet joined #salt
02:15 fllr joined #salt
02:28 ninkotech__ joined #salt
02:29 pdayton joined #salt
02:32 favadi joined #salt
02:45 jfzhu_us mgw, are you sure its not how the yaml is being interpreted? It looks like the yaml language doesn't really support new lines and you might be looking to use literal blocks. Maybe this might help you http://stackoverflow.com/questions/11582934/new-line-n-in-yaml-file
02:46 jnials joined #salt
02:47 forresta joined #salt
02:50 jfzhu_us mgw:  Apparently they are called literal styles and not literal blocks (http://www.yaml.org/spec/1.2/spec.html#id2795688). Apologies for the incorrect terminology as this is still new to me
02:50 maj888 joined #salt
02:55 lianhaifu joined #salt
02:56 ajw0100 joined #salt
03:02 Tekni joined #salt
03:05 xinkeT joined #salt
03:08 balltongu joined #salt
03:11 mgw jfzhu_us: I'm sure it is yaml, but I can't use the literal style from bash
03:11 mgw The problem is how to do this:
03:12 mgw salt '*' my.module "foo\nbar"
03:12 mgw For now I do just that, and then in the module do .replace('\\n', '\n')
03:13 xinkeT joined #salt
03:16 jnials So apparently this isn't legal, is t here another way to accomplish the same thing?  http://pastebin.com/NTHBA5WR
03:18 forresta jnials, how about this? https://gist.github.com/gravyboat/7939347
03:20 jnials I will give it a try.
03:24 jnials Sigh.  I see now I was looking at the wrong chunk of documentation.  :(
03:24 jnials Thank you forresta!
03:24 forresta yea I was going to ask where you got managed_file from
03:25 forresta err manage_file
03:25 forresta was going 'hmm did I miss this in the docs?' :P
03:25 jnials saltg.modules.file
03:25 jnials errr salt.modules.file
03:25 forresta ahh yea that's why
03:26 forresta gotta be careful between the module and state docs, haha. I've made that mistake myself!
03:26 jnials So in general, if in doubt I should look at the state docs first?
03:26 forresta well, the state docs are for actual .sls files
03:26 forresta the modules are for use on the command line
03:27 jnials Ah.  now THAT makes sense.
03:27 forresta Yea it can be a bit confusing. I've been considering proposing to add a note to every single module page stating this
03:28 jnials At least a disclaimer at the beginning somewhere.  The docs are sorta kinda chaotic right now.
03:28 fllr joined #salt
03:29 forresta jnials, when you find issues on the docs for confusing parts, can you make an issue? I've been trying to fix up a lot of that confusing stuff, and I know the Salt team has as well, but sometimes it's hard to see all the different issues certain people might be confused about
03:30 jnials Sure.  I'll try and do some this week.  But I have to get this problem solved soon so I don't get ripped in the staff meeting.  Again. :)
03:30 forresta haha
03:30 forresta yea no worries, creating an issue is quick, you don't even have to make the pull request if you don't have time, just having the issues is great
03:30 jnials Yup.
03:31 forresta great, I've gotta head to the gym, I'll be back in a bit if you need help.
03:31 redondos joined #salt
03:32 jnials I think I'm good.
03:32 forresta cool!
03:34 KyleG joined #salt
03:34 KyleG joined #salt
03:44 higgs001 joined #salt
03:53 dancat joined #salt
03:54 dancat I've been looking for a beginner resource on salt and I'm coming up empty any thoughts?
03:54 dancat Like a book
03:56 malinoff There are not good books :) http://docs.saltstack.com/index.html and http://docs.saltstack.com/topics/tutorials/walkthrough.html are the best!
03:57 dancat I'll start there. Thanks
03:58 cachedout joined #salt
03:58 jdenning joined #salt
04:00 srage joined #salt
04:02 Gareth hm. it seem incrond doesn't allow comments.
04:02 Gareth in its incrontab
04:04 dvogt joined #salt
04:04 CheKoLyN joined #salt
04:08 jfzhu_us joined #salt
04:27 druonysus joined #salt
04:31 druonysus joined #salt
04:31 druonysus joined #salt
04:36 prooty hello. has anyone here used salt cloud with digital ocean?
04:37 prooty everytime i try to create a droplet through a mapping it errors with "authentication failed".
04:38 jcockhren prooty: o/ but I haven't used mappings for it
04:40 cachedout joined #salt
04:41 prooty jcockhren, you only use the profile? did you define an ssh-key in your provider configuration file?
04:41 forresta dancat, I'm very slowly working on a book, but for the time being the walkthroughs are the best you're gonna get.
04:41 forresta dancat, I've got a blog post as well that walks you through setting up a static nginx app if you're interested in seeing an actual project that uses Salt.
04:42 forresta it's a bit outdated, but should be ok
04:42 sgviking joined #salt
04:43 jcockhren prooty: the 'ssh_key_name' setting must match the name of a key in your DO account
04:44 prooty i'm not using one in mine yet. i'll try it to see if it's better.
04:45 prooty does the example here work for you (https://www.digitalocean.com/community/articles/automated-provisioning-of-digitalocean-cloud-servers-with-salt-cloud-on-ubuntu-12-04)?
04:46 forresta did you try these docs prooty? http://docs.saltstack.com/topics/cloud/digitalocean.html
04:46 jcockhren prooty: I had mine setup before that was written.
04:47 jcockhren there's a typo in there in the settings anyway I think
04:47 forresta oOOooo look at jcockhren, setting his up before the docs were done :P
04:47 jcockhren yeah. magic
04:47 jcockhren ha
04:48 jcockhren ssh_key_name: digital-ocean-salt-cloud.pub
04:48 jcockhren remove the ".pub"
04:48 jcockhren forresta: I mean... you're taking too long with your book. ;)
04:49 forresta I know
04:49 forresta I've been slacking on it, I've got a presentation next week for a local meetup, and then the saltconf one that I'm working on, so it's lower on my priority list.
04:49 prooty forresta, i took a look at that documentation but there was no mention of how to use an ssh key.
04:49 forresta prooty, yea I'm looking at it now and these docs are pretty short
04:50 jcockhren how does one become an author for the book? work for saltstack?
04:50 forresta You fork my repo, then add some content and add yourself ot the author's list: https://github.com/gravyboat/salt-book
04:50 prooty jcockhren, how about the ssh_key_file value?
04:51 forresta jcockhren, I don't work for saltstack, I just want to try and make an 'all in one' book talking about getting started and working on real projects as you go through the chapters.
04:51 forresta it's still pretty bare bones :(
04:51 forresta like I said, slacking
04:52 seanz joined #salt
04:52 prooty sorry, nevermind. i guess it's for the private ssh key then.
04:53 jcockhren forresta: but forreal. chat-ops with salt
04:54 forresta jcockhren, are you still working on that?
04:54 jcockhren writing the bot... now
04:54 forresta awesome
04:55 jcockhren yep... the fix for my autodeploy blocker is in the develop branch. someone was awesome/annoyed enough to solve it
04:55 forresta lol
04:56 jcockhren the beauty of opensource
04:56 forresta yup
04:56 kermit joined #salt
05:00 prooty nice, the ssh method worked.
05:01 forresta prooty, cool.
05:01 forresta I feel like the digital ocean docs should either have more details, or link to the digital ocean docs
05:05 forresta actually nevermind I'll just add it
05:05 jcockhren haha
05:05 forresta I'm already fixing a typo in those docs
05:05 forresta why not
05:05 forresta is there no way to wrap a hyperlink between lines with RST?
05:06 forresta the docs don't say...
05:06 forresta https://www.digitalocean.com/community/articles/automated-provisioning-of-digitalocean-cloud-servers-with-salt-cloud-on-ubuntu-12-04
05:06 forresta I can't believe how long that is
05:06 forresta lol
05:08 Nazca__ jcockhren: bot ... https://github.com/saltstack/salt/issues/8682 bot?
05:09 jcockhren Nazca__: yes-ish
05:09 jcockhren actually didn't know that thread existed
05:09 jcockhren thanks!
05:10 Furao forresta: DigitalOcean API server in NYC2 had been broken for around 48 hours
05:10 forresta Furao, really??
05:10 Furao you can't use API to create new instances, they get created (and billed for)
05:10 Furao but they're network interface is down
05:10 Furao it works fine in all other datacenter
05:10 forresta hmm, there's nothing on the network status page
05:10 forresta did you tweet at them
05:10 Furao I open a ticket, they confirmed the issue and engineer are working on it
05:11 forresta Furao, laaaaaaaame
05:11 forresta digital ocean is usually quick on that stuff
05:11 forresta is that in the public network section?
05:11 Furao I did most of the troubleshooting for them myself already
05:11 Furao I don't think
05:11 Nazca__ jcockhren: I've been meaning to publish my multiprotocol bot and put salt support in to it, but time has not been my friend lately
05:11 forresta bummer
05:12 prooty Furao, i just created 2 droplets using salt cloud. i guess it works for some?
05:12 Furao prooty: in NYC2?
05:12 prooty yes.
05:12 Furao weird
05:12 jcockhren Nazca__: oh!
05:12 Furao "I have seen a couple similar issues reported by other customers, and I am trying to track this down."
05:13 forresta at least they recognize it and are working on it
05:13 Furao I got jenkins create 30 VMs per day using salt-cloud to test all our formulas and nothing worked for more than a day
05:13 forresta lol
05:20 KyleG joined #salt
05:20 KyleG joined #salt
05:21 ajw0100 joined #salt
05:24 joehoyle joined #salt
05:28 y0j joined #salt
05:29 dstanek joined #salt
05:32 y0j joined #salt
05:39 prooty when defining a custom script in salt cloud, do i put in the path of the custom script?
05:39 anuvrat joined #salt
05:46 ckao joined #salt
05:46 cachedout joined #salt
05:47 dstanek joined #salt
05:52 Nazca__ jcockhren: here's the bot framework: https://github.com/kaithar/muhubot ... it's intended to be cleanly extensible, but I've not put much time in to adding shiny toys to it yet
06:01 KyleG joined #salt
06:01 KyleG joined #salt
06:16 nmistry joined #salt
06:16 mgw joined #salt
06:19 cachedout joined #salt
06:29 Furao joined #salt
06:32 KyleG joined #salt
06:32 KyleG joined #salt
06:35 nmistry joined #salt
06:39 nmistry joined #salt
06:42 ConceitedCode joined #salt
06:46 MZAWeb joined #salt
06:51 jkleckner joined #salt
06:58 higgs001 joined #salt
07:09 Teknix joined #salt
07:11 Destro joined #salt
07:11 mgw joined #salt
07:15 rojem joined #salt
07:16 seanz joined #salt
07:17 jdenning joined #salt
07:27 gammalget joined #salt
07:27 q4brk joined #salt
07:28 q4brk hi! can anyone tell me if there's a way of distinguishing between the first deployment of a file using file.managed or an update of an existing one?
07:29 higgs001_ joined #salt
07:36 prooty q4brk: when you run state.highstate and it's a first deployment, the comment will be "new file" or something similar. when you update it the comment will be the diff of the file.
07:40 q4brk prooty: yeah, but I'm trying to figure that inside the salt file... my use case is something like:  for the first deployment, i need to initialize something and start a service,  for file updates, i only need to reload a service
07:42 juasiepo joined #salt
07:43 prooty oh. the service.running state has a reload parameter. so you can enable it and have the service.running state watch the file.managed state. (http://docs.saltstack.com/ref/states/all/salt.states.service.html)
07:44 ntt_ joined #salt
07:45 ntt_ Hi. how can i remove a grain? i'm using grains.remove but i have an error: "The key xxx is not a valid list". Someone can help me?
07:45 q4brk prooty: won't that attempt to reload the service when the file is first created?
07:45 zandy_ joined #salt
07:48 prooty q4brk, yeah i guess. unless it's not started before.
07:48 q4brk that's my problem..
07:48 q4brk some services complain on reload if they're not started
07:51 kh4z joined #salt
07:56 zandy joined #salt
07:57 zandy joined #salt
08:00 balboah joined #salt
08:03 y0j joined #salt
08:05 ajw0100 joined #salt
08:06 harobed_ joined #salt
08:12 slav0nic joined #salt
08:17 MrTango joined #salt
08:19 giantlock_ joined #salt
08:29 ninkotech__ joined #salt
08:30 kh4z joined #salt
08:39 ninkotech__ joined #salt
08:50 favadi_ joined #salt
08:52 ninkotech__ joined #salt
08:55 mike251 joined #salt
09:00 ninkotech joined #salt
09:01 elsmorian joined #salt
09:03 achileuss joined #salt
09:09 kh4z_ joined #salt
09:09 anuvrat joined #salt
09:10 zandy joined #salt
09:12 FL1SK joined #salt
09:12 junedm joined #salt
09:15 zandy joined #salt
09:16 ninkotech joined #salt
09:19 q4brk joined #salt
09:19 junedm left #salt
09:20 aleszoulek joined #salt
09:21 q4brk does anyone know if I want to use salt's gitfs backend with a repository with submodules, does that work properly ?
09:25 fllr joined #salt
09:26 Iwirada joined #salt
09:29 Iwirada hi
09:30 Iwirada is there a possibility to pass a dictionary in the file state to a jinja template?
09:32 jinnko joined #salt
09:39 nocturn00 Hi, how can I force the latest version of a python pip package in salt?
09:39 saltsa joined #salt
09:47 nocturn00 nm found it.  Didn't realize how to put the option in the salt format
09:49 pengunix joined #salt
09:51 sylvinus joined #salt
09:52 ninkotech joined #salt
09:55 fllr joined #salt
09:57 fllr joined #salt
10:05 krak3n` joined #salt
10:08 carlos joined #salt
10:20 balboah_ joined #salt
10:32 kh4z joined #salt
10:33 dstanek joined #salt
10:51 psyl0n joined #salt
10:51 Veticus joined #salt
10:53 Veticus joined #salt
10:54 Veticus joined #salt
10:55 Veticus joined #salt
10:57 fllr joined #salt
11:02 carlos joined #salt
11:04 whiskybar joined #salt
11:06 rojem joined #salt
11:06 pengunix joined #salt
11:08 sunand joined #salt
11:11 sunand_ joined #salt
11:14 ntt_ Hi. i'm using cmd.run for mount an nfs partition. How can i check if it is already mounted?
11:15 ninkotech joined #salt
11:15 uomobonga joined #salt
11:20 elsmorian ntt_: could you use http://docs.saltstack.com/ref/states/all/salt.states.mount.html ?
11:21 elsmorian ntt_: also, http://www.kaffeinatedx.com/2013/04/saltstatesmount.html seems to be a thing
11:21 elfixit joined #salt
11:22 ntt_ elsmorian: mount in substitution of cmd.run?  What happens if nfs filesystem is already mounted?
11:24 pengunix_ joined #salt
11:30 aleszoulek joined #salt
11:42 ntt_ selinux.setenforce  is not available..... but i'm reading here -> http://docs.saltstack.com/ref/modules/all/salt.modules.selinux.html and it seems to work. Same result if i use states
11:42 ntt_ someone can help me?
11:46 rafael joined #salt
11:57 fllr joined #salt
12:08 sgviking joined #salt
12:18 oz_akan_ joined #salt
12:18 sylvinus joined #salt
12:24 dstanek joined #salt
12:26 NotreDev joined #salt
12:28 NotreDev joined #salt
12:34 sylvinus joined #salt
12:34 johtso joined #salt
12:35 johtso Hey guys, I'm having trouble getting git backed ext_pillar working
12:36 johtso no problems when I have the pillar states in /srv/pillar, but when using a git repository, it always says "Specified SLS 'base' in environment 'master' is not available on the salt master"
12:37 johtso the pillar states are all in /var/cache/salt/master/pillar_gitfs/0 .. so that seems fine
12:41 zooz joined #salt
12:47 jfzhu_us joined #salt
12:47 johtso brb, heading to lunch
12:53 vortec joined #salt
12:57 fllr joined #salt
12:59 sylvinus joined #salt
12:59 _fllr_ joined #salt
12:59 ninkotech joined #salt
13:00 blee joined #salt
13:00 sylvinus_ joined #salt
13:01 Sheco joined #salt
13:03 sylvinus joined #salt
13:09 [diecast] joined #salt
13:10 dstanek joined #salt
13:12 NotreDev joined #salt
13:13 sylvinus_ joined #salt
13:17 sylvinus_ joined #salt
13:17 sylvinus__ joined #salt
13:20 sylvinus joined #salt
13:27 elfixit joined #salt
13:30 elfixit1 joined #salt
13:37 jslatts joined #salt
13:50 vortec hi there. is there any built-in functionality to download a tarball and unpack it if a file/directory does not exist?
13:51 oz_akan_ joined #salt
13:53 oz_akan_ joined #salt
13:53 pengunix joined #salt
13:54 ertac joined #salt
13:54 diegows joined #salt
13:57 fllr joined #salt
13:59 ipmb joined #salt
14:00 higgs001 joined #salt
14:03 quickdry21 joined #salt
14:03 Gifflen joined #salt
14:04 krak3n` @vortec you could use unless: test -d /path/to/dir in your cmd state which downloads the tarball
14:05 vortec "unless" is a requisite?
14:05 brianhicks joined #salt
14:05 wkf joined #salt
14:06 Psi-Jack Is there a better way to write this jinja?  if 'phx' in grains['localhost'] or 'ewr' in grains['localhost']
14:06 juicer2 joined #salt
14:10 zandy joined #salt
14:10 vortec krak3n`: ah, you suggest using cmd. I was using file.managed
14:11 vortec Psi-Jack: for the python part, that looks good
14:11 Psi-Jack I was just hoping for a shorter way to provide a list like ['phx', 'ewr'] in grains['localhost'] or something, if the list would be or, not and.
14:12 krak3n` yeah, tho if your tarball is not remote and a file in your states, then instead you can use file.managed with makedirs: True, this will create the required directories for the file being managed
14:16 vortec Psi-Jack: try this: set(('phx', 'ewr')) >= set(grains['localhost'])
14:16 pass_by_value joined #salt
14:17 Psi-Jack Hmmm
14:17 Psi-Jack That's.. Not as human readable. :)
14:18 vortec hehe
14:21 racooper joined #salt
14:23 mpanetta joined #salt
14:25 johtso how would I achieve the equivalent of "pillar_roots: base: - /srv/pillar" using ext_pillar?
14:26 Psi-Jack johtso: What external implementation?
14:26 johtso Psi-Jack, git
14:26 amahon joined #salt
14:27 Psi-Jack Then, as documented, it says your master branch == base
14:27 johtso currently I'm getting a "Specified SLS 'base' in environment 'master' is not available on the salt master" error
14:27 Psi-Jack johtso: Keep in mind, I found a bug that causes gitfs backends to not refresh its cache until AFTER compiling the current highstate, making you have to push highstate twice before it actually applies.
14:28 johtso Psi-Jack, the pillar states have been pulled into /var/cache/salt/master/pillar_gitfs/0
14:28 Psi-Jack What's your full ext_pillar look like?
14:29 AdamSewell joined #salt
14:30 zandy joined #salt
14:30 johtso Psi-Jack, http://hastebin.com/cekupebuta
14:30 Psi-Jack Yeah, change "base" to master
14:30 johtso I've tried with a branch named master, and the pillar state environment named master
14:31 johtso still doesn't work
14:31 Psi-Jack What "doesn't work" about it?
14:32 jankowiak joined #salt
14:32 Psi-Jack johtso: Also, have you checked your salt master logs to see if there's any errors? Is salt-master running? GitPython is required.
14:33 johtso Psi-Jack, I'm running salt-master in the foreground with debug logging
14:33 Psi-Jack Any git errors, or GitPython errors? Show logs..
14:33 johtso I'll paste bin the relevant bit
14:33 Psi-Jack Being intentionally vague does not help.
14:33 Psi-Jack Just FYI. :)
14:34 johtso Psi-Jack, http://hastebin.com/ceyecaripa
14:36 mgw joined #salt
14:36 balboah meh. ImportError: libzmq.so.1: cannot open shared object file: No such file or directory. What could the cause be? I use the same version as on other instances and I have libzmq3 installed and pyzmq from salt 0.17.2 ppa
14:36 Psi-Jack Uhh, why does your top.sls have master: in there?
14:38 johtso Psi-Jack, I was trying with the environment (I think that's the correct term?) and the git branch named master
14:38 Psi-Jack That should be base, base == master.
14:38 johtso oh right
14:38 johtso didn't realise base was special
14:38 Psi-Jack heh
14:38 Psi-Jack It is special. base is always loaded, no matter what environment you use.
14:38 NotreDev joined #salt
14:39 Psi-Jack I'm not, personally, recommending gitfs to anyone at this time with salt, though, just an FYI.
14:39 Psi-Jack Filesystem level git is, IMHO, better and saner.
14:40 Psi-Jack Especially if you do anything with enviroments. :)
14:40 johtso Psi-Jack, hmm, so just manually pulling the pillar data onto your salt master when you change it?
14:41 Psi-Jack Yep. Or using a Makefile to push it to origin and/or login to the saltmaster, cd to the directory, pull from git. :)
14:41 Psi-Jack I have a very simple but effective Makefile for just that if you want an example.
14:42 johtso does sound much more straightforward
14:42 johtso could have the whole /srv directory as a git repo with a sub repository for the pillar data
14:43 Psi-Jack http://paste.linux-help.org/view/0f197e8c
14:43 Psi-Jack That's my Makefile. :)
14:43 dstanek joined #salt
14:43 johtso ah, nice one, thanks
14:44 Psi-Jack I just have that in my master repository, I have symlinks to top_states.sls and top_pillars.sls to their respective base/{states|pillar}/top.sls for centralizing top.sls to the 1 file.
14:44 johtso great, all working after changing the env name to base, thanks!
14:44 pdayton joined #salt
14:45 johtso going to go with gitfs for now I think, unless I find any issues
14:45 Psi-Jack You will. If you use any environments. :)
14:45 johtso just the one at the moment
14:46 Psi-Jack Well, good.
14:46 Psi-Jack What I find painful about gitfs is the obfuscation between environments in branches.
14:46 johtso currently using a grain for the environment name (develop, staging, production), and using a ninja if-else in the pillar states
14:46 Psi-Jack That makes things much more daunting and less useful.
14:46 johtso not sure if that's a completely crazy way of doing things though..
14:47 johtso right, I really dislike the idea of it requiring you to do specific things with your branches
14:47 Psi-Jack Heh, you should define an environment grain, like environment: develop
14:47 oz_akan_ joined #salt
14:47 Psi-Jack In /etc/salt/grains
14:47 cyberbob joined #salt
14:47 johtso I'm currently doing that in the minion conf
14:47 Psi-Jack Yeah, you should do it in /etc/salt/grains ;)
14:47 Psi-Jack Much cleaner, and requires no changes, except to your minion conf to pull grains back out.
14:47 * johtso checks out the documentation
14:48 Psi-Jack My salt manages my grains by managing /etc/salt/grains, and this file is auto-reloaded every time a highstate is pushed. ;)
14:49 johtso hmm
14:50 johtso I really wish I could find a non trivial example salt project that uses multiple environments and pillar data
14:50 kaptk2 joined #salt
14:51 Psi-Jack heh
14:51 Psi-Jack The way environments work now, that's not really possible, IMHO. :)
14:51 johtso still really not sure what the "correct" way is to manage configuration of multiple environments
14:51 johtso oh right :(
14:52 johtso I saw something in the documentation that seemed to suggest that you would have a different top file in the different git branches.. but that sounds horrible..
14:52 Psi-Jack Right now, if you have a salt state in two environments that have the same salt state name, it will conflict if you include that into one host. I recently recommended stacking and overriding the state so you could do real development, and stacking of environments properly. ;)
14:52 Psi-Jack johtso: No, they recommend not having different top files, because it can quickly get lost.
14:52 johtso right
14:53 fllr joined #salt
14:53 StDiluted joined #salt
14:54 Psi-Jack It's the conflict bit that gets me. I want to write salt states that go from dev to qa, to prod, in that order. Including updating an existing state to include new functionality. I Want to copy the state to dev, edit it, change it, target it to systems in dev, and test it. Currently if you do that, it will conflict because the state exists and is defined in prod.
14:54 th3reverend joined #salt
14:54 th3reverend left #salt
14:55 fllr joined #salt
14:55 flebel joined #salt
14:57 fllr joined #salt
14:57 johtso yeah, it's pretty confusing, you almost have two dimensions of environment. you have your different git branches, with your salt configuration in different states of development, and then you have different pillar data for different environments
14:59 Psi-Jack Well, pillar data is stacked.
14:59 Psi-Jack Which ... Is awesome. :)
14:59 Psi-Jack You can replace pillar data based on ordering matches, environments, etc.. Last-matching pillar state wins the actual conflict.
15:00 Psi-Jack That's how states should be, IMHO, too, so you could do the same thing, and really truely use environments. :)
15:00 Psi-Jack I almost thing environments should be renamed to.. Stacks. :)
15:01 Psi-Jack You have your base stack, then you have your prod stack, your qa stack, and dev stack, and eventually it all gets back to at least prod, in most cases, base if you just want it applied to every system, like common system authentication, common ntp time syncing, etc..
15:01 whiskybar joined #salt
15:06 tyler-baker joined #salt
15:06 Veticus joined #salt
15:07 Veticus joined #salt
15:08 Veticus joined #salt
15:09 prooty is there a way when calling salt.highstate to only see the changed states?
15:09 Veticus joined #salt
15:09 prooty sorry, nevermind.
15:10 Veticus joined #salt
15:10 Veticus joined #salt
15:11 Veticus joined #salt
15:12 Veticus joined #salt
15:13 Veticus joined #salt
15:13 quickdry21 joined #salt
15:15 vejdmn joined #salt
15:15 vejdmn joined #salt
15:17 mr_chris joined #salt
15:23 forresta joined #salt
15:24 andresr joined #salt
15:26 ninkotech_ joined #salt
15:31 FL1SK joined #salt
15:32 NotreDev joined #salt
15:35 lineman60 joined #salt
15:48 redondos joined #salt
15:48 redondos joined #salt
15:51 smccarthy joined #salt
15:52 teskew joined #salt
15:53 fllr joined #salt
15:54 kermit joined #salt
15:56 cyberbob joined #salt
15:57 slack3er joined #salt
15:58 shinylasers joined #salt
16:02 cachedout joined #salt
16:04 mgw joined #salt
16:05 jalbretsen joined #salt
16:07 Iwirada left #salt
16:07 rojem joined #salt
16:08 rojem joined #salt
16:10 NotreDev joined #salt
16:12 Ryan_Lane joined #salt
16:12 meteorfox joined #salt
16:12 redondos joined #salt
16:12 elfixit joined #salt
16:13 jkleckner joined #salt
16:23 Shenril joined #salt
16:24 pdayton joined #salt
16:26 shinylasers joined #salt
16:28 shinylasers joined #salt
16:29 shinylasers joined #salt
16:30 pentabular joined #salt
16:31 pentabular left #salt
16:33 meteorfox joined #salt
16:36 pentabular1 joined #salt
16:37 sh3rp4 joined #salt
16:38 tyler-baker joined #salt
16:38 coolj joined #salt
16:38 cyberbob joined #salt
16:38 jankowiak joined #salt
16:38 hotbox joined #salt
16:38 opapo joined #salt
16:38 juice joined #salt
16:38 scristian joined #salt
16:38 zz_cro joined #salt
16:38 pmcg joined #salt
16:38 aptiko joined #salt
16:38 eskp joined #salt
16:38 bdf_ joined #salt
16:38 basepi joined #salt
16:38 cedwards joined #salt
16:38 melinath joined #salt
16:38 scooby2 joined #salt
16:38 __number5__ joined #salt
16:38 quickdry21 joined #salt
16:38 wunki joined #salt
16:39 jdenning joined #salt
16:39 flebel joined #salt
16:39 drags joined #salt
16:41 StDiluted joined #salt
16:41 kaptk2 joined #salt
16:42 baffle_ joined #salt
16:42 kula_ joined #salt
16:42 Anb joined #salt
16:42 Anb joined #salt
16:42 canci_ joined #salt
16:43 pmcg_ joined #salt
16:43 amahon joined #salt
16:43 opapo_ joined #salt
16:44 defunctzombie joined #salt
16:44 defunctzombie joined #salt
16:44 bdf joined #salt
16:44 dpac|away joined #salt
16:45 scooby2_ joined #salt
16:45 eskp_ joined #salt
16:45 quickdry21_ joined #salt
16:45 aptiko_ joined #salt
16:45 hotbox_ joined #salt
16:48 tsantero joined #salt
16:48 luminous joined #salt
16:48 juice- joined #salt
16:48 brucewang joined #salt
16:49 lynxman joined #salt
16:49 melinath_ joined #salt
16:50 coolj joined #salt
16:51 anuvrat joined #salt
16:53 blast_hardcheese joined #salt
16:53 alekibango joined #salt
16:53 jankowiak joined #salt
16:53 scristian joined #salt
16:53 zz_cro joined #salt
16:53 basepi joined #salt
16:54 cedwards joined #salt
16:55 justBob joined #salt
16:55 vejdmn joined #salt
16:58 hotbox joined #salt
17:04 mwmnj joined #salt
17:05 luminous joined #salt
17:05 scristian joined #salt
17:07 UtahDave joined #salt
17:09 redondos joined #salt
17:10 KyleG joined #salt
17:10 KyleG joined #salt
17:11 Psi-Jack OKat, this is frustrating. Now that I moved back pillar_roots, pillars from different environments aren't overriding anymore.
17:12 AdamSewell joined #salt
17:12 AdamSewell joined #salt
17:12 Psi-Jack ext_pillars, it seemed to work, but was annoying to infrastructure, now, it just doesn't work.
17:12 dstanek_afk joined #salt
17:13 jslatts joined #salt
17:13 kamal_ joined #salt
17:14 tsantero joined #salt
17:15 slav0nic joined #salt
17:15 jcockhren Psi-Jack: you gotta fix it brother
17:15 Psi-Jack Heh.
17:15 Psi-Jack It's salt itself, that's... likely.. the broken... aspect...
17:16 th3reverend joined #salt
17:17 jcockhren Nazca__: I looked at the bot. it nice. I started on one since I need something not exactly salt-centric
17:18 defunctzombie joined #salt
17:18 defunctzombie joined #salt
17:20 crane joined #salt
17:20 th3reverend1 joined #salt
17:21 cyberbob joined #salt
17:22 troyready joined #salt
17:23 jdenning joined #salt
17:26 cyberbob joined #salt
17:26 robawt joined #salt
17:27 baffle joined #salt
17:27 hotbox joined #salt
17:27 UtahDave jcockhren: which bot?
17:27 drags joined #salt
17:30 robawt joined #salt
17:31 th3reverend1 left #salt
17:32 bemehow joined #salt
17:33 jcsp joined #salt
17:33 hotbox joined #salt
17:36 KyleG1 joined #salt
17:37 jacksontj joined #salt
17:39 sroegner joined #salt
17:39 hotbox joined #salt
17:39 nmistry joined #salt
17:43 Nazca__ jcockhren: yea, that framework was started on the basis that hubot does a lot of really useful stuff but the connection adapters and general structure are horrific
17:49 shinylasers joined #salt
17:53 bhosmer joined #salt
17:54 cro joined #salt
17:57 MK_FG joined #salt
17:57 NotreDev joined #salt
17:58 jcockhren UtahDave: Nazca__ has one
17:58 jcockhren Nazca__: yeah. AND I don't want to have to deal with coffeescript dependecies. see nodejs
17:59 forresta I was actually just gonna ask Nazca__ how he was planning to link that up with all the coffeescript fun that hubot does
17:59 forresta but now that I actually look at the repo I see that isn't the case :P
17:59 NotreDev joined #salt
18:00 higgs001 joined #salt
18:00 jcockhren repost the repo plz. apparently, I didn't star it before crashing last night
18:00 forresta https://github.com/kaithar/muhubot
18:00 forresta jcockhren, I had to go back to the IRC logs to find it
18:01 jcockhren yeah. my ideal bot is first written in python
18:02 jcockhren and leverages decorators for adding plugins/commands easily
18:03 jcockhren Nazca__ bot is clearly more mature than mine will be for a while
18:03 jcockhren I'll be good as soon as I can get twisted.words to connect to gtalk
18:04 quanta_ joined #salt
18:05 mgw joined #salt
18:06 jcockhren also, for my purposes, I'd like it to be chat-ops ish during commandline usage
18:06 jcockhren interactive salt in a way
18:06 Nazca__ jcockhren: gtalk being jabber, you'll want to read how I did it ... particularly the ping bit, cause apparently t.words doesn't have it built in lol
18:07 rgarcia_ joined #salt
18:08 quanta_ What can I do to debug in this case: https://gist.github.com/quantonganh/7948532
18:09 micall joined #salt
18:09 Nazca__ I can't quite remember which server we're using ... something fire ... but in any case the protocol module is really low level, not like the one it has for irc
18:11 Nazca__ part of the reason I went with tornado is that it makes it really trivial to build web hook framework, so that github, redmine and things like CI packages can be added on as separate modules but neatly co-existing
18:11 quanta_ I meant how can I find the line number in the template file that caused above error
18:12 meteorfox joined #salt
18:12 Nazca__ I already have a bunch of python code that hooks directly in to salt, I use it for a panel that queries grains info and uses salt for it's orchestration layer
18:12 Nazca__ just need to bul the two in heh
18:12 th3reverend joined #salt
18:13 Nazca__ quanta_: cute trace.... gist carbon.sls?
18:13 jcockhren Nazca__: hmm.
18:14 jcockhren my salt panel is returner-based. viewing stuff returned to syslog long that is then thrown into elasticsearch
18:15 Nazca__ "bul the two in" ... *scratches head* that was probably meant to be tie
18:17 Nazca__ yeah, I'm going to do some returner magic once I've got my weird pydsl issue sorted, but for initial work I found it to be handy to issue calls directly... for things like checking if an arbitrary file is in a cache directory on any of the nodes
18:18 Nazca__ cmd is a really handy thing when you need to do remote management and don't want the hassle of paramiko
18:19 jcockhren yep. cmd is cool
18:19 oz_akan_ hi guys, what is the best practice to upgrade salt version on minions?
18:19 jcockhren working out great during my initial POC
18:19 oz_akan_ has anyone here done this at scale ?
18:19 Nazca__ oz_akan_: depends how you installed them, but probably pkg
18:20 jcockhren https://github.com/sophicware/benderbot
18:20 jcockhren b/c everyone needs a better bender
18:20 quanta_ Nazca__: https://gist.github.com/quantonganh/7948734
18:20 oz_akan_ Nazca__: yes pkg
18:21 oz_akan_ Nazca__: from which version to which version have you upgraded?
18:22 Nazca__ oz_akan_: I've found that they're reasonably compatible, so I'd stick with the strategy of having it target the latest version and doing a single host test deploy first to make sure it upgrades
18:22 zandy joined #salt
18:22 cro joined #salt
18:23 utahcon Too many functions declared in state "pkg" in sls
18:23 utahcon I didn't know there was a limit
18:25 justBob_ joined #salt
18:25 dvogt joined #salt
18:26 prooty hello. when i launch an instance, i'd like my other instances to automatically (or with a single command) add the new instance's ip address to the list of allowed hosts in their own firewall.
18:26 prooty how can i go about doing this?
18:26 cdk is it possible to have query a salt minion for all packages installed/managed via Salt? I don't want all the system packages, just those that Salt has installed (so pkg.list_pkgs won't do)
18:26 redondos_ joined #salt
18:27 Nazca__ quanta_: I don't know if the 'File "<template>", line 61, in top-level template code' is going to be strictly accurate, but with that big of a state I'd test those jinja files one at a time to work out which one is faulty then look for a parse error or bisect the file to narrow it down
18:27 Nazca__ cdk: probably not
18:30 amckinley joined #salt
18:32 mattmtl joined #salt
18:34 Nazca__ jcockhren: this is what I meant by use of cmd: https://gist.github.com/kaithar/7948949
18:35 justBob_ joined #salt
18:35 jergerber joined #salt
18:39 forresta prooty, how about using the salt mine?
18:39 forresta whoever owns intothesaltmine.org makes me sad :(
18:39 jcockhren Nazca__: ahhhh
18:39 forresta always shows up when I'm trying to get actual docs on the salt mine itself.
18:39 oz_akan_ Nazca__: I was asking how you installed new version of salt on all the minions. do you just upgrade the package and then restart minion via salt?
18:40 UtahDave prooty: here's an example using the peer interface  (publish.publish)
18:40 forresta prooty, check out: http://stackoverflow.com/questions/17158665/how-to-get-a-list-of-all-salt-minions-in-a-template
18:40 Nazca__ oh, I knew I was going to do something else ... prooty instance of what exactly, and how is it being launched?
18:40 nmistry joined #salt
18:41 Nazca__ I was going to suggest using peer and reactor, but Dave beat me to it :p
18:41 forresta yea that's a better suggestion than the mine, overcomplicating it :P
18:41 JordanRinke Morning
18:41 forresta hi
18:42 rgarcia_ joined #salt
18:42 Nazca__ oz_akan_: yup, that's what I'll be testing for rolling out... just using the pkg module to update and restart
18:42 brianhicks joined #salt
18:43 JordanRinke forresta: Hello.
18:43 * Gareth waves
18:43 oz_akan_ Nazca__:  I had issues with that, salt-minion wouldn't start after the update. good if it works now
18:43 forresta Gareth, do you have your client set to automatically wave?
18:44 Gareth forresta: hah. no.
18:44 oz_akan_ Nazca__: https://github.com/saltstack/salt/issues/5721
18:44 forresta Gareth, ok
18:44 Gareth forresta: I could see how one would be led to that impression though :)
18:45 aleszoulek joined #salt
18:45 Nazca__ oh the dark and depressing days of the mirc autogreeter *shudder*
18:45 forresta heh
18:46 * Gareth slaps Nazca__ around a bit with a large trout
18:46 oz_akan_ Nazca__: this is how I restate the minion, http://www.rackeroz.com/2013/10/how-to-restart-salt-minion.html
18:46 oz_akan_ restate = restart
18:46 Nazca__ oz_akan_: aaaah, interesting ... would be good to know if it's still an issue then
18:47 Nazca__ Gareth: I lack a slap button on xchat .. I do have a Punt button though :p
18:47 oz_akan_ 17.2 might be fine
18:48 oz_akan_ though to go there, we have to upgrade at least one :)
18:48 oz_akan_ omce
18:48 forresta lol damn oz_akan_, this workaround is ghetto!
18:48 oz_akan_ once
18:48 oz_akan_ :)
18:48 forresta very good thinking, but still hilarious
18:48 oz_akan_ you have to solve the problem one way or another
18:49 forresta of course
18:49 JordanRinke Any of you using python or the pydsl for your state files?
18:49 jcockhren o/
18:49 Nazca__ o/ also :)
18:49 Nazca__ pydsl specifically
18:49 Psi-Jack UtahDave: Ahh, good, you're handy? I'm finding some wierd discrepencies in pillar environments... What is the expected occurance when pillar:base has no ref, pillar:prod defines test in top for '*', and pillar:dev defines test in top for '*' as well, but only the one in prod is actually getting applied, not overridden by the one in dev, which it technically should, according to the docs.
18:49 CheKoLyN joined #salt
18:50 Nazca__ oz_akan_: that is indeed a hack worthy solution hehe
18:51 oz_akan_ I agree, we had to restart the minion as we change grains. and an issue was if you set a grain with grain.setval it wouldn't overwrite the grain that is in the minion configuration file
18:51 younqcass joined #salt
18:52 hotbox joined #salt
18:52 NotreDev joined #salt
18:53 forresta Psi-Jack, once you figure this out I'm expecting that you update the documentation to clarify
18:53 forresta there you can't get away with your 'I do not know Python' excuses! :P
18:53 Psi-Jack Well, I get two completely different outcomes.
18:54 Psi-Jack Depending on use of ext_pillars, and pillar_roots. :)
18:54 JordanRinke I don't know why but I am having a ton of trouble grokking the pydsl documentation
18:57 hotbox joined #salt
18:57 Psi-Jack For example, this: http://paste.linux-help.org/view/ec6adf60    Shows the setup, the top, the pillar states involbved.
18:58 Psi-Jack Yet, what occurs is the value of grains:roles ends up being the value from prod, not dev, for a host in grain environment:test
19:01 NotreDev joined #salt
19:02 Nazca__ JordanRinke: it is a little painful in parts, yes
19:03 Nazca__ JordanRinke: I make an instance via state() and then write states based on the function layout given in the module docs ... also remember that they're like other states so you can't collide them
19:03 lineman60 joined #salt
19:04 nmistry joined #salt
19:04 JordanRinke I must be missing it in the documentation, but how do I do matching when creating a state? Like.. does someone have an example of a top.sls in pydsl?
19:04 Nazca__ I've not done a top.sls via pydsl, only states
19:05 druonysus joined #salt
19:05 jcockhren JordanRinke: what Nazca__ said
19:05 JordanRinke well, looks like I get to go exploring :-D
19:05 nmistry joined #salt
19:07 ajw0100 joined #salt
19:07 forresta JordanRinke, if you figure it out let us know what the repo location is, and then we can get an example into the docs
19:08 JordanRinke Literally just starting with the pydsl, if I switch rendered in salt/master to pydsl do I still have to shebang the pydsl files?
19:10 Nazca__ JordanRinke: pydsl basics: https://gist.github.com/kaithar/7949487 lol ... param to state is the state's id ... also this boilerplate is awesome: https://gist.github.com/kaithar/7357224
19:10 cyp joined #salt
19:12 zandy joined #salt
19:14 xmltok joined #salt
19:14 Nazca__ but now I must fly *gone*
19:14 jcockhren https://github.com/saltstack-formulas/backuptocloud-formula/blob/master/backuptocloud/init.sls
19:14 prooty UtahDave, forresta, thanks. i'll take a look at the publish docs.
19:14 jcockhren JordanRinke: ^^
19:15 * Psi-Jack sighs
19:15 william20111 joined #salt
19:16 JordanRinke Nazca__ & jcockhren : thx
19:22 hotbox joined #salt
19:27 pengunix joined #salt
19:28 cachedout joined #salt
19:31 druonysus joined #salt
19:34 whiskybar joined #salt
19:35 diegows joined #salt
19:37 bhosmer joined #salt
19:41 shinylasers joined #salt
19:44 abhi_ joined #salt
19:44 Sheco joined #salt
19:48 blee joined #salt
19:50 th3reverend joined #salt
19:50 Ryan_Lane joined #salt
19:50 Sypher hmm, is it me or is "salt-run manage.present" not working in the (latest) development branch? I'm getting an "Function 'manage.present' is unavailable"..
19:50 th3reverend left #salt
19:54 Sypher hmm could also be a conflict with the stable packages which are installed via apt-get... running --verisons-report on the checked out files results in it returning 0.17.2 as salt version :/
19:55 rcg_re joined #salt
19:57 vbabiy How can I kill a job that is running from the master
19:58 abhi_ joined #salt
19:59 dstanek joined #salt
20:03 rcg_re left #salt
20:04 derelm joined #salt
20:09 druonysus joined #salt
20:09 druonysus joined #salt
20:10 nebuchadnezzar hi
20:12 zandy joined #salt
20:16 nebuchadnezzar I wonder if adding “size test” when comparing files before reading all lines of both could not speed things, I'm referencing things like https://github.com/saltstack/salt/blob/develop/salt/modules/file.py#L2083
20:16 UtahDave1 joined #salt
20:16 nebuchadnezzar plus comparaison code is duplicated at several places
20:17 tim_ joined #salt
20:19 shadowfax|chc joined #salt
20:22 JordanRinke hour later, no idea how to do it yet
20:24 NotreDev joined #salt
20:24 Psi-Jack I DO NOT GET THIS! This is freaking insane, how such a thing can happen! I have salt.grains in prod and dev environments, I got dev:salt.grains working, then I disabled dev in the base top.sls, and now it wont apply the prod:grains.
20:25 Psi-Jack Yet.. Hmm, odd, it does.. IN the actual grains.items.
20:27 forresta JordanRinke, you might want to post a topic to the mailing list, might get more eyes on it that way
20:27 wkf joined #salt
20:31 KyleG joined #salt
20:31 KyleG joined #salt
20:34 Marion joined #salt
20:34 UtahDave JordanRinke: sorry, I've been afk.  What are you trying to do?
20:34 TheRealBill joined #salt
20:35 JordanRinke pydsl top.sls
20:36 srage joined #salt
20:38 UtahDave JordanRinke: are you asking how to use a top file with pydsl?
20:38 JordanRinke How to make a top file that is pydsl
20:38 andresr joined #salt
20:39 UtahDave I don't think you can do that.
20:39 JordanRinke a yaml top file, with pydsl states, no problem, but if I am using pydsl states, it makes sense to just have a pydsl top as well
20:40 UtahDave So you're wanting to programmatically build your top.sls?
20:40 UtahDave There's a plugin system for the top.sls called Master Tops.
20:40 NotreDev joined #salt
20:40 Psi-Jack http://paste.linux-help.org/view/9c2437a3  -- This.. When I apply this, and dev gets applied successfully, I comment out the whole dev environment from the top, and it still keeps applying the dev pillar. I can make changes to the dev/pillars/salt/grains.sls and even though dev is not matching anything, it's still applying the pillar.
20:40 JordanRinke I don't specifically want to build one programmatically, I just want consistent file language across my installation
20:40 druonysus joined #salt
20:41 UtahDave yaml, bro
20:41 JordanRinke pssshhh
20:42 Psi-Jack The spacing is correct, just that pastebin seems to do funky stuff when using the YAML language parser.
20:42 UtahDave pydsl is for your sls files.  It doesn't have anything to do with the top.sls
20:42 UtahDave Psi-Jack: are you executing   salt \* saltutil.refresh_pillar after commenting out the dev environment?
20:43 jacksontj joined #salt
20:43 younqcass joined #salt
20:43 Psi-Jack Hmm,. no I was not.
20:43 Psi-Jack But, that made no difference.
20:45 JordanRinke UtahDave: Yeah, I assumed as much, but that doesn't prevent me from wanting to try to make a top with it.
20:45 JordanRinke My issue is less of an I need this, help me... and more of a hmmm... let me screw around with this for a while and talk about it thing
20:47 UtahDave cool
20:47 Psi-Jack Yeah, no change at all, it's still applying dev environment's grains/salt.sls instead of prod's
20:49 Psi-Jack However, if I make my top.sls have a blank dev: environment definition, NOTHING runs anywhere anymore.
20:49 UtahDave Psi-Jack: I'm a bit confused.  so are you creating a "grains" dict inside your pillar?
20:50 dstanek how do i get supervisor to restart when a config in /etc/supervisord/conf.d changes?
20:51 UtahDave dstanek: create a service.running state that watches your file.managed or file.recurse state
20:51 dstanek i tried all sorts of stuff including using watch_in from the config file and just haven't found something that works
20:52 forresta dstanek, can you share your state file?
20:52 forresta post it on pastebin or something?
20:52 dstanek forresta: yes, i'll do that now
20:52 ninkotech joined #salt
20:53 forresta ok cool, I haven't seen someone trying to watch a whole directory, do you manage the files within /etc/supervisord/conf.d?
20:53 Psi-Jack UtahDave: state.grains.sls makes a dict, yes, a text entry to build a /etc/salt/grains file.
20:54 Psi-Jack http://paste.linux-help.org/view/ec6adf60   This shows the grains.sls files from a former attempt, but basically the same.
20:54 cachedout joined #salt
20:54 akoumjian I'm getting an odd error preventing one of my minion processes from running: https://gist.github.com/akoumjian/884a5c294533dcce5a32
20:54 dstanek forresta: my supervisor state http://pastebin.com/P6k61HPX
20:55 abe_music joined #salt
20:55 dstanek forresta: my celeryd state http://pastebin.com/x0Wbm0dF
20:55 UtahDave Psi-Jack: why are you doing this with pillar?
20:56 forresta lol
20:56 dstanek i want to have several states like the celeryd state have all pulled in based on the role so the supervisor state won't know what is running on it
20:56 forresta Psi-Jack, just link the issue, don't go through it all again :P
20:56 forresta dstanek yea I gotcha
20:56 dstanek forresta: i figured the watch in would make the supervisor server reload configs
20:56 UtahDave dstanek: in your service.running, I think you also need   - name: supervisord
20:57 UtahDave also,     supervisord.running:    doesn't make sense  on line 5
20:58 abe_music has anyone attempted to put a rsa private key in pillar and then use that in file.managed? for some reason i can't figure out what i'm doing wrong...it's either formatted weird or it fails with an "Unknown yaml render error"
20:59 akoumjian UtahDave: Any notion of what might be going on here? https://gist.github.com/akoumjian/884a5c294533dcce5a32  Minion won't run with this error
20:59 dstanek UtahDave: what do you mean by doesn't make sense?
20:59 forresta ave_music, did you make sure to use the | (pipe) so that yaml would interpret it exactly as it is? Or is the managed file actually containing the key?
21:00 abe_music forresta: yeah, in my pillar i'm using the pipe for multiline, but then the file.managed: -contents: {{ pillar.private_key }} is failing for some reason
21:00 Psi-Jack forresta: Eh? This is a completely new and different issue, actualluy.
21:00 forresta dstanek, well, what was your goal with the supervisord.running?
21:00 forresta Psi-Jack, oh ok, I didn't read the scrollback
21:00 Psi-Jack lol
21:00 Psi-Jack This is a pillar issue.
21:00 forresta yea I saw that much, but I thought the pillar thing was the workaround you were trying for the other issue
21:01 forresta I may have just forgotten
21:01 Psi-Jack Nope
21:01 UtahDave akoumjian: what version of Salt?
21:01 akoumjian UtahDave: 17.2 on ubuntu ppa
21:01 dstanek forresta: to make sure the celeryd script runs through supervisor
21:01 forresta abe_music, hmm, that is odd. Can you run it through with debug and see if it actually logs what the line is looking like.
21:01 forresta dstanek, but you're trying to treat that like a service right?
21:01 UtahDave dstanek: ah, ok. I didn't realize you were using the supervisord statee
21:01 UtahDave state
21:02 chromake_ joined #salt
21:02 forresta jeez UtahDave! It's like he didn't direct the pastes at you or something...
21:02 akoumjian UtahDave: I think I found a relevant issue https://github.com/saltstack/salt/issues/9138
21:02 UtahDave akoumjian: I'm not sure.  Are you using the scheduler?
21:03 UtahDave akoumjian: does balboah's suggesting work for you?
21:03 forresta so dstanek, when the supervisor service is running, do you reference it via supervisor or supervisord?
21:03 forresta sorry I'm not too familiar with supervisor
21:03 akoumjian UtahDave: Yes
21:03 akoumjian UtahDave: back on track.
21:04 Ryan_Lane joined #salt
21:04 UtahDave cool
21:04 dstanek forresta: UtahDave: i just checked and on debian the supervisor service is just supervisor
21:04 forresta hmm weird
21:04 dstanek when i added 'name: supervisord' it failed to start it
21:04 forresta can you modify the celeryd.supervisor.conf file, add a line, then run it back through with -l debug added on?
21:05 dstanek using 'supervisor' works just file though
21:08 dstanek forresta: how to i turn debugging on? the salt command doesn't seem to have a -l option
21:09 dave_den Psi-Jack:  I commented on your issue https://github.com/saltstack/salt/issues/9186  -  hopefully it will start to explain why pillar_roots/file_roots/ and matching minions to multiple environments works.
21:09 Psi-Jack I see that.
21:09 dave_den i'll post a longer comment explaining why i think the proposed behavior is A Bad Thing (™) when I have time, aybe this weekend
21:10 abe_music forresta: looks like it's this exact issue: https://github.com/saltstack/salt/issues/5480
21:10 abe_music hmmm
21:11 abe_music i'm trying to enable passwordless SSH between a master and bunch of slave nodes (not salt specific)
21:11 Psi-Jack dave_den: I'll look through this more over the weekend, and see if I can wrap my head around this better, but the principal idea on the actual issue itself is still liked, so far. The concept of stacking. ;)
21:12 frosty996 joined #salt
21:12 zandy joined #salt
21:12 frosty996 So, I had a strange thing happen yesterday while attempting to demo mine.get.
21:12 frosty996 it seemed no mine data was coming back to my master
21:13 frosty996 even though normal minion communication and states were working fine
21:13 dave_den Psi-Jack: when we all get on the same page of understanding how salt currently works, I don't think it will be much liked  :)    anyway, I hope i don't come across insulting - it's a complicated issue and will be good to clarify the behavior
21:13 frosty996 then I disabled the firewall on the master and suddenly it was able to get the mine data
21:13 frosty996 does the mine data get sent back on ports other than 4505 or 4506?
21:14 dave_den frosty:  the ret_port, default 4506
21:14 dave_den bbl
21:14 frosty996 dave_den: that's what I thought, but 4505 and 4506 were both open
21:14 frosty996 does it come back as udp?
21:15 forresta dstanek, just add -l debug onto the end of the command
21:16 bhosmer joined #salt
21:18 forresta abe_music, ahh yea, I've looked at that issue quite a few times before
21:18 dstanek forresta: i ran 'salt-minion -l debug'
21:18 dstanek forresta: http://pastebin.com/3uu1H9A0
21:19 forresta line 4 is interesting
21:19 forresta it shouldn't already have the latest file in the cache if you updated it.
21:19 dstanek forresta: actually line 23 may show the issue
21:19 forresta yea that's also good
21:20 dstanek what is '/usr/bin/supervisorctl add celeryd' doing?
21:20 forresta is celery called celery, or celeryd?
21:20 forresta dstanek, I'm not actually sure what that is doing.
21:21 dstanek forresta: it's celeryd
21:21 forresta ok
21:21 dstanek going to run these commands by hand
21:21 forresta ok cool
21:22 Psi-Jack dave_den: Not insulting at all. After I started playing more into pillars today, and racking my head on it in pillar_roots, because pillar_roots and ext_pillar seems to act COMPLETELY different.
21:23 jdenning_ joined #salt
21:23 abhi_ joined #salt
21:25 vejdmn joined #salt
21:29 jcsp1 joined #salt
21:29 abhi_ joined #salt
21:30 frosty996 and now I can't reproduce it
21:32 jcsp joined #salt
21:32 forresta frosty996, my favorite kind of problem!
21:32 frosty996 I just don't get it
21:32 forresta ok I'll give you a hint
21:33 forresta live demo
21:33 forresta =
21:33 forresta issues
21:33 sylvinus joined #salt
21:33 forresta no matter how tested something is :P
21:33 frosty996 I don't understand how test.ping could work fine and then the mine data doesn't come back
21:33 forresta if it's a live demo, it's bound to have issues
21:33 frosty996 forresta: yah, of course
21:33 dstanek forresta: so my problem turns out to be running multiple processes
21:33 forresta dstanek, ahhh ok that makes sense
21:33 frosty996 forresta: but you know, the demo was in this vagrant, so I don't understand how it can work here and not at the office!
21:34 dstanek forresta: supervisor things my process name is celeryd:celeryd_0[123] and salt thinks it is celery
21:34 forresta frosty996, yea, only solution is to go back to the office and spend your weekend there troubleshooting it :P
21:34 frosty996 perfect
21:34 forresta dstanek, Gotcha, yea that's kind of a bummer.
21:35 forresta dstanek, did you figure out a way around it? Or is there a good way around it?
21:35 forresta some sort of celery setting or something
21:35 dstanek forresta: i'm going to try messing with my state a bit
21:36 forresta dstanek, ok cool, let me know what you find!
21:36 dstanek forresta: thank you for your help!
21:37 forresta dstanek, yea np, always happy to sit around and say 'did you try running it again'!
21:38 dstanek forresta: it often helps to just walk someone else through the problem :-) i've by looking at this on and off since yesterday - probably sunk a good two hours into staring and the screen and drooling
21:39 forresta Yea I don't blame you! I've got a duck on my desk at home that I always talk to, he's pretty useless for suggesting solutions though.
21:40 foxx joined #salt
21:40 jacksontj joined #salt
21:40 forresta meteorfox, did you get google compute engine working yesterday, or file an issue?
21:42 meteorfox forresta: yes, got it working. Found 2 bugs. #9217 and #9219
21:42 meteorfox forresta: Fixes are already upstream
21:43 forresta meteorfox, nice
21:44 forresta that was quick turn around on those two
21:44 viq joined #salt
21:44 meteorfox yeah, they essentially were one liners
21:48 forresta yea
21:50 dstanek forresta: i just have to put the 3 in pillar and then http://pastebin.com/ZutxSdyF
21:51 dstanek i can use the pillar data to also populate the celeryd.conf
21:51 forresta ahh ok, so it's always formatted the same way?
21:51 dstanek forresta: yes, i control the name formatting in my celeryd.conf
21:51 forresta gotcha, awesome then
21:51 dstanek looks like this deserves a blog post :-)
21:52 dstanek only one problem left...it looks like saltstack always thinks something got updated http://pastebin.com/GskUEXn7
21:55 forresta dstanek, yea that would be good to put in a blog
21:59 rgarcia_ joined #salt
22:08 abe_music UtahDave, forresta: getting closer to a release of stackd.io: http://bit.ly/1dbS0R7
22:08 abe_music taking longer than expected :(
22:12 zandy joined #salt
22:14 bhosmer joined #salt
22:17 zandy_ joined #salt
22:20 forresta so you're saying you're slacking?
22:20 jdenning joined #salt
22:21 forresta abe_music, did you write this article?
22:22 forresta or did Charlie?
22:22 abe_music no, a colleage of mine did
22:22 abe_music yeah Charlie
22:22 forresta gotcha, I'm always interested to talk to people who put two spaces after a period.
22:22 amckinley joined #salt
22:23 abe_music LOL
22:23 abe_music i can make the introduction if you want :)
22:23 jcockhren ha'
22:23 forresta lol
22:23 abe_music who notices stuff like that? forresta...that's who
22:23 forresta just give him a picture of WHY YOU NO guy
22:24 forresta abe_music, well, I wasn't sure at first if it was just the font
22:24 forresta There's nothing wrong with it, I just never understood why
22:25 forresta most of the time it seems like people who learned to type on typewriters do it
22:25 abe_music i have no idea...writing is not my strong suit. i have heard of people doing that before
22:26 forresta *shrug*, it's just an interesting thing to discuss with people.
22:26 abe_music you know what, i want to say my wife did it when writing her dissertation. i'll have to ask her
22:26 forresta hah ok
22:32 amahon joined #salt
22:34 vejdmn joined #salt
22:35 dvl forresta: I always do two spaces.
22:35 JordanRinke sigh, going back to using yaml...  :(
22:35 dvl forresta: Learned it in typing class, back in the 70's
22:35 forresta dvl, on a typewriter?
22:36 dvl forresta: indeed.
22:36 forresta dvl, interesting, that's the most common reason I've seen when I've asked people.
22:36 dvl forresta: Don't give me the 'typewriter's needed it bit'.
22:36 dvl forresta: I still think it looks better.
22:36 dvl Especially for mono-spaced fonts, e.g. on the console.
22:36 dvl But then, I do it in all my blogs too.
22:36 forresta dvl, I don't know enough about typewriters to have input there, nor do I 'judge' people because they do it, I'm just interested in the reason why some people do it, and others don't.
22:37 dvl forresta: That's my reason.  And I haven't used a typewriter in over 40 years.
22:37 forresta dvl, fair enough.
22:37 dvl well, perhaps 35 years.
22:38 ajw0100 joined #salt
22:38 forresta dvl, heh
22:39 amahon joined #salt
22:42 dvl The same teacher also taught me that to any number whose digits add up to 3, 6, or 9, is divisible by 3.
22:45 bhosmer joined #salt
22:49 JordanRinke Can't wait for SteamOS to release today :/
22:49 whiteinge +1!
22:50 JordanRinke They *say* it is going to be a bit of a challenge to install... interested to see what that means
22:50 JordanRinke who will get a config manager running on SteamOS first? :X
22:51 forresta JordanRinke, I'm excited about that as well
22:51 NotreDev joined #salt
22:51 forresta Did they update when they were releasing it?
22:52 JordanRinke No, everything still just says today basically
22:52 forresta yea I was hoping they'd have updated with an actual time today
22:53 JordanRinke I hope they release it as a torrent file, trying to host that will be devestating and lead to all night downloads :/
22:53 forresta heh
22:53 forresta I hope they sent out the steambox to some people who are gonna dive into it and check out what the hardware looks like
22:54 JordanRinke I thought they released various beta box specs
22:54 JordanRinke just a range of common pc parts really
22:54 forresta I'm more interested in seeing if there's anything specific they have on there though for the bios and such
22:54 JordanRinke I have a GTX Titan in the office right now, might have to put steam OS on that rig for a few minutes
22:54 forresta nice
22:55 JordanRinke It is supposed to be doing GPU column database stuff right now but I haven't had time to build it :/
22:55 forresta I'd be happy to take that titan off your hands :P
22:58 forresta this might actually justify rebuilding my extra PC with a failing drive though
22:59 rgarcia_ joined #salt
23:04 NotreDev joined #salt
23:08 cocoy joined #salt
23:12 pass_by_value joined #salt
23:13 amckinley joined #salt
23:17 ajw0100 joined #salt
23:18 JordanRinke Interested to see how the game streaming stuff works too
23:19 oz_akan_ joined #salt
23:20 JordanRinke in theory this will be the download link... http://steampowered.com/steamosbeta
23:21 JordanRinke and I am sure that is getting hammered with scripts right now
23:22 JordanRinke ohhh reallly
23:22 JordanRinke http://repo.steamstatic.com/download/
23:24 JordanRinke aanndd it is basically unavailable now hah
23:25 Gareth Seems to be working here.
23:25 JordanRinke I keep getting 20-50mb and then download terminates
23:26 NotreDev joined #salt
23:26 Gareth hm. 20MB here then dead.
23:27 Gareth wget --continue to the rescue :)
23:27 JordanRinke haha yeah doing the same thing
23:28 JordanRinke after the 3rd try in chrome :x
23:28 FL1SK joined #salt
23:28 JordanRinke All this bandwidth doing me no good :/
23:30 JordanRinke you know, that would be a sly way to make sure only linux nerds get it at first, make it so it can't be downloaded without some extra work heh
23:30 Gareth hah
23:33 jslatts joined #salt
23:33 abhi_ joined #salt
23:33 JordanRinke you getting the sys restore too?
23:33 JordanRinke I can't even seem to get headers off that file
23:34 Gareth starting off with just the installer.
23:36 JordanRinke http://steamdb.info/blog/35/
23:36 JordanRinke (Sorry to offtopic the room)
23:36 NotreDev joined #salt
23:37 Marion_ joined #salt
23:38 JordanRinke oh interesting, notice on your wget, reset happens every 60 seconds?
23:38 Gareth roughly.
23:39 FL1SK joined #salt
23:39 abhi_ joined #salt
23:40 JordanRinke server or lb terminating long running sessions probably lol
23:42 jeddi joined #salt
23:48 amahon joined #salt
23:50 amahon joined #salt
23:55 JordanRinke_afk downloads died :/
23:59 FL1SK joined #salt

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