Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-04-15

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

All times shown according to UTC.

Time Nick Message
00:00 enarciso joined #salt
00:01 rhodgin joined #salt
00:03 sergiotocalini joined #salt
00:08 esogas_ joined #salt
00:08 sergiotocalini joined #salt
00:08 sergiotocalini W 1
00:12 otter768 joined #salt
00:17 SeeDickCode joined #salt
00:19 aw110f_ joined #salt
00:22 dendazen joined #salt
00:25 stanchan joined #salt
00:25 hemebond joined #salt
00:26 bhosmer_ joined #salt
00:26 hemebond Does the salt minion cache or remember the IP address of the master?
00:26 jalaziz_ joined #salt
00:28 iggy yes
00:28 hemebond Well that solves my "mysterious" connectivity issues :-)
00:29 TheLoeki joined #salt
00:30 bhosmer__ joined #salt
00:32 iggy I think there's a setting for it
00:33 hemebond Oh? I'll have a look. Thanks.
00:33 hasues joined #salt
00:34 hasues left #salt
00:35 iggy nvm, I was thinking of retry_dns, but that's for if the minion can't resolve (not if it's already got something resolved)
00:35 iggy probably have to bounce the minions
00:36 hemebond Ah okay. Yeah I've been restarting the service. It only just occurred to me that it probably doesn't recheck DNS even after it loses connection to the master.
00:37 iggy yeah, I think it's something they want to improve upon when they start revamping the multi-master stuff
00:38 iggy there was at least some mention of it at saltconf
00:48 TheLoeki joined #salt
00:55 MatthewsFace joined #salt
00:56 MatthewsFace joined #salt
01:03 echo joined #salt
01:08 murrdoc joined #salt
01:08 ingwaem joined #salt
01:09 Tyrm joined #salt
01:12 yomilk joined #salt
01:13 djcrabhat joined #salt
01:25 catpig joined #salt
01:30 dleedyna joined #salt
01:33 loz-- joined #salt
01:33 ckao joined #salt
01:35 ITChap joined #salt
01:45 mgw joined #salt
01:53 hasues joined #salt
01:53 hasues left #salt
01:54 scbunn joined #salt
01:56 bhosmer joined #salt
01:58 TyrfingMjolnir joined #salt
02:01 SeeDickCode joined #salt
02:04 TyrfingMjolnir joined #salt
02:05 Tyrm joined #salt
02:06 subsignal joined #salt
02:08 evidence what's the highest number of workers recommonded for a master?
02:11 evidence can't really find much mention of it, other than some people tuning it up to 10/15
02:11 evidence i believe linkedin said they were running something like 30 at saltconf?
02:13 quintinadam joined #salt
02:13 otter768 joined #salt
02:22 evle joined #salt
02:25 ajmath joined #salt
02:30 writtenoff joined #salt
02:35 iggy kind of depends on your master
02:35 iggy I tend to keep it around #cpu's * 1.5-2
02:36 evidence i figure there is a limitation or inefficiency within salt though after a certain point
02:36 evidence cool that works
02:37 evidence somewhere between 20 and 30 is likely our sweet spot, i'll mess around with it.  just wanted to see what the upper limits were
02:38 michelangelo joined #salt
02:43 p66kumar joined #salt
02:45 subsignal joined #salt
02:48 scbunn joined #salt
03:00 viq joined #salt
03:01 _JZ_ joined #salt
03:04 hemebond left #salt
03:05 otter768 joined #salt
03:08 murrdoc joined #salt
03:12 jalaziz joined #salt
03:14 Furao joined #salt
03:15 pipeep Was salt-ssh removed from pypi, and if so, why?
03:16 cro pipeep: It's part of the salt package in pypi
03:16 pipeep cro, okay. Can I install the salt package in a virtualenv?
03:17 cro yes
03:17 cro for production use?
03:17 pipeep cro, I'm planning on deploying to a single host.
03:18 cro ok
03:18 pipeep So full salt would be *way* overkill
03:20 iggy masterless?
03:28 murrdoc joined #salt
03:30 yomilk joined #salt
03:43 _JZ_ joined #salt
03:44 pipeep iggy, Good idea, but not ideal for my workflow
03:46 Furao joined #salt
03:51 dalexander joined #salt
03:56 pipeep okay, my virtualenv/pip installation was just a little bit borked.
03:56 pipeep This is what I get for mixing pip and debian packages
03:57 mgw joined #salt
04:00 murrdoc debian packages dont affect virtuan env setups
04:00 murrdoc also u shouldnt be installing more that python-apt, python-dev and python-setuptools and python-pip
04:00 murrdoc rest should all be pip
04:00 murrdoc for sanity sakes
04:00 murrdoc not cos it akes sense
04:00 Vynce joined #salt
04:01 timoguin joined #salt
04:03 pipeep murrdoc, I had setuptools fucked up, there was an old pip package installed on wheezy, breaking thing when I moved to jessie
04:03 pipeep the whole pip+apt thing is a clusterfuck
04:03 Vynce is there a command you can run on a minion that will highstate that minion?
04:04 Vynce s/run on/run from/
04:06 favadi joined #salt
04:08 stanchan joined #salt
04:33 bhosmer_ joined #salt
04:38 iggy Vynce: sudo salt-call state.highstate
04:38 Vynce ah, good, thanks
04:55 randomuser joined #salt
04:56 desposo joined #salt
04:57 rdas joined #salt
05:06 Furao joined #salt
05:09 enarciso joined #salt
05:20 buMPnet joined #salt
05:21 _JZ_ joined #salt
05:28 elfixit joined #salt
05:32 yomilk joined #salt
05:35 glyf joined #salt
05:37 catpigger joined #salt
05:43 tomasfejfar joined #salt
05:49 yomilk joined #salt
05:52 piru_ joined #salt
05:54 AndreasLutro joined #salt
05:59 krelo joined #salt
06:00 Karlthane_ joined #salt
06:01 colttt joined #salt
06:02 otter768 joined #salt
06:02 I3olle joined #salt
06:03 antonn joined #salt
06:04 antonn hello, anyone with strong orchestrate runner understanding can describe me why salt-run state.orch is restricted with explicit environment specification?
06:12 Trades joined #salt
06:15 rdas joined #salt
06:16 stoogenmeyer__ joined #salt
06:19 __number5__ antonn: it is http://docs.saltstack.com/en/latest/ref/runners/all/salt.runners.state.html#salt.runners.state.orchestrate have a saltenv parameter
06:20 flyboy joined #salt
06:22 lempa joined #salt
06:23 Furao joined #salt
06:28 dopesong joined #salt
06:31 KermitTheFragger joined #salt
06:32 harkx joined #salt
06:34 dopesong_ joined #salt
06:34 antonn __number5__ Yes, maybe i wasn't strict enough on my question: when u want to use orchestrate with special envinronment u MUST specify saltenv. If not - in orchestrate code is strictly written: "saltenv='base'" and during highstate ur declarative enviroment desciption in top.sls will be ignored. it's a big problem for me.
06:39 jhauser joined #salt
06:42 soren joined #salt
06:44 __number5__ yep, salt have the assumption each run will under one environment
06:44 jY joined #salt
06:44 __number5__ antonn: what's your case required to run two environment at the same time?
06:49 kawa2014 joined #salt
06:53 bdf joined #salt
06:57 CeBe joined #salt
07:02 tomasfejfar joined #salt
07:02 _1_Abcd joined #salt
07:06 Romlok joined #salt
07:07 jeblair joined #salt
07:09 zz_cro joined #salt
07:16 eseyman joined #salt
07:21 Furao joined #salt
07:24 lb1a joined #salt
07:25 jeddi joined #salt
07:26 peters-tx2 joined #salt
07:27 jalaziz joined #salt
07:32 peters-tx3 joined #salt
07:34 fredvd joined #salt
07:43 ggoZ joined #salt
07:47 yggdrasi1 joined #salt
07:50 CeBe joined #salt
07:51 favadi left #salt
07:51 CeBe1 joined #salt
07:52 a7p joined #salt
07:53 faust joined #salt
07:58 bhosmer joined #salt
08:01 tajh joined #salt
08:02 otter768 joined #salt
08:12 Xevian joined #salt
08:20 ujjain joined #salt
08:21 antonn __number5__ yes, it's destructing me. In highstate u have clear declarative description, where is told in which environment what targeting and what states should we try to achieve. During orchestrate we can do highstate, yes? During highstate we can not supply saltenv explicit, but ruther rely on declarative top.sls spec of our envs, yes? Double yes. So why can't we transitivly say that we want orchestrate minions on all environments.
08:22 antonn We don't want to explicitly tell in each command what environment to run, but ruther we already described it declarative in top.sls for highstate and we want reuse in during orchestration too.
08:24 antonn Usecase: u have different environments, let me say dev, dev2, prod. Each environment has a map of different minions with crossing roles. Ur states describes those roles. When u orchestarate and what to make highstate for ALL of ur environments, u want to be shure that one role will be availible before another. On ALL environments(as roles are similar),
08:25 antonn But u can;t do this. U should do 3 repetative comands, with saltenv argumets which in this case will just REPEAT ur declaretive environment description in top.sls
08:27 huddy joined #salt
08:27 antonn More over specifying state explicit means that u r pevented from base environment reusage for example in pillars, without duplicating states in exact environment.
08:27 glyf joined #salt
08:28 antonn It just goes bad with the highstate's top.sls environment concept. They are crossing each other. Like two different approaches.
08:28 lietu joined #salt
08:31 lietu- joined #salt
08:34 Xevian joined #salt
08:37 favadi joined #salt
08:38 bhosmer joined #salt
08:41 faust left #salt
08:45 Andre-B joined #salt
08:45 dalexander joined #salt
08:47 dharper_ii joined #salt
08:48 keimlink joined #salt
08:48 malinoff joined #salt
08:52 FineTralfazz joined #salt
08:57 Grokzen joined #salt
08:59 keimlink joined #salt
09:02 losh joined #salt
09:03 plindgren joined #salt
09:05 plindgren hi guys
09:06 plindgren if i configure my /etc/salt/master according to the first yaml example here
09:07 plindgren should i not be able to switch linux users to thatch and run something like salt -a pam 'webserver' test.ping ?
09:07 antonn __number5__ hmm, have u gone?
09:07 zer0def joined #salt
09:08 bytemask joined #salt
09:11 adelcast joined #salt
09:12 plindgren sorry i was looking for the access control system
09:12 plindgren not external auth
09:19 brayn joined #salt
09:19 plindgren anybody got client_acl 's to work?
09:21 denys joined #salt
09:21 babilen joehh1: Could I bribe you with a pint and cake?
09:21 denys joined #salt
09:26 lietu- joined #salt
09:37 plindgren http://docs.saltstack.com/en/latest/ref/clientacl.html had to actually read 1.3.1
09:38 plindgren who would guess that its a good idea to read the docs
09:40 ]V[ joined #salt
09:41 bhosmer joined #salt
09:42 grrrrr ^^
09:42 Puckel_ joined #salt
10:03 otter768 joined #salt
10:06 bhosmer joined #salt
10:07 is_null hi all ! where are the docs for the whitelist parameter of state.highstate please ? also, how do you recommend running a highstate excluding anything that's not starting with 'foo.*' ?
10:08 sheltonash joined #salt
10:08 Romlok is_null: you mean targeting minions?
10:09 is_null Romlok: rather state names
10:09 babilen is_null: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.highstate is what you are looking for
10:09 babilen Not sure if you can exclude globs
10:09 is_null babilen: that's what i was reading actually, the example mentions the whitelist argument
10:09 is_null but the doc don't ...
10:10 babilen That just seems to take a list of SLS files that are being whitelisted
10:10 giantlock joined #salt
10:11 is_null absolutely, thanks, it doesn't take globs nor regexps, maybe a PR for that would be reasonnable ?
10:17 sheltonash Hey guys, After some advice or examples please? Windows minion - I am making a visual studio state and one of the install flags is the product key. Just wondering what the solution for this sort of thing is? I guess I need to do some templating in the package of the windows package manager? The goal though is to fully install a developers machine by calling state.highstate - how would i  pass a key to highstate or whats the best s
10:18 sheltonash any examples of this sort of thing would be a real help, only been poking salt for a few days, seems really powerful! :)
10:19 AndreasLutro sheltonash: put the key in the pillars somehow
10:20 chiui joined #salt
10:22 lahwran joined #salt
10:23 sheltonash different for every developer though and its sort of private to them, well their msdn account
10:23 sheltonash can you have local pillars on a minion?
10:24 sheltonash don't really want a treasure trove of keys somewhere i guess
10:24 AndreasLutro I'd have thought the keys belonged to the company, not the devlopers
10:24 AndreasLutro if you want to store the keys on the minion you can use grains
10:25 Romlok though you'd still have the "treasure trove of keys" issue with using grains
10:26 echo joined #salt
10:27 Romlok that said, any content of any file on any minion is fair game to be read by the master afaik
10:28 lietu joined #salt
10:28 sheltonash Iwell they do i guess, but to get them a developer usually logs into their msdn grabs a key and enters it in the product key field while they're installing visual studio. i don't know if im thinking of things incorrectly. maybe i need a masterless minion, they checkout a repo containing the states, edit some placeholder line with their visual studio key then run them
10:29 AndreasLutro shrug
10:29 Romlok is there no corporate MSDN admin account, which has access to all the employees' keys?
10:29 AndreasLutro if you want to store it on the master, use pillars. if you want to store it on the minion, use custom grains
10:29 AndreasLutro there is no right and wrong
10:30 Romlok except maybe pushing those pillars to a public github repo
10:30 sheltonash :)
10:30 AndreasLutro even then you could gpg encrypt them and probably be fine
10:31 Romlok true
10:34 sheltonash i guess i want to avoid a directory/list of them some place as maintaining it would be someones job. a lame solution would be to have a c:\dir\saltproductkeys.xml or something that they fill in before running anything and then stuff just gets pulled in on each minion from that known place. I want the dev to do the key work basically rather than an admin guy
10:34 N-Mi_ joined #salt
10:35 sheltonash dunno, i'll try the things you've suggested and see what i like/prefer. Cheers for your time guys
10:37 Romlok custom grains can be defined on the minion machine, so can be essentially what you just described
10:38 sheltonash ah okay, will look at them then, thank you
10:43 markm joined #salt
10:44 CeBe joined #salt
10:44 mike25de left #salt
10:49 aquassaut joined #salt
10:52 Andre-B_ joined #salt
10:58 evle joined #salt
10:58 antonn joined #salt
10:59 antonn __number5__ are u here?
11:06 CeBe joined #salt
11:08 ndrei joined #salt
11:18 bhosmer joined #salt
11:19 stoogenmeyer__ hey guys, is there a way to make the salt mine become ephemeral of sorts? if a minion did not report for a while ?
11:21 stoogenmeyer__ based on the official example (http://docs.saltstack.com/en/latest/topics/mine/), let's say one of the minions is killed, I would still have his network_ip kept in the mine
11:31 otter768 joined #salt
11:31 dRiN joined #salt
11:33 irctc181 joined #salt
11:34 N-Mi_ joined #salt
11:39 loz-- joined #salt
11:44 tajh joined #salt
11:46 CeBe1 joined #salt
11:47 elfixit joined #salt
11:54 jeddi joined #salt
11:55 hojgaard joined #salt
12:01 is_null hi all, how do you recommend to run all salt.* states (from salt-formula) -> restart salt master -> '*' state.highstate ?
12:12 arif-ali joined #salt
12:22 yggdrasi1 joined #salt
12:22 lothiraldan joined #salt
12:35 lothiraldan joined #salt
12:37 faust joined #salt
12:37 yuhl_work_ joined #salt
12:39 murrdoc joined #salt
12:43 otter768 joined #salt
12:45 faust left #salt
12:45 I3olle joined #salt
12:46 bhosmer joined #salt
12:47 zipkid sheltonash: how high is your dev turnover if the VS bootstrap time becomes an issue...?
12:48 * zipkid would think that in that case investing in company culture over salt recipes would be a good idea....
12:49 cmcmacken joined #salt
12:53 Zachary_DuBois joined #salt
12:59 babilen is_null: I simply provision the master before other nodes (salt-formula will take care of the restarting)
13:00 MAbeeTT_ joined #salt
13:01 sheltonash zipkid: well not even that high actually. literally ramping up on this stuff. I mean to question the crazy involved in provisioning a devs machine (surely they'll want it customised, use their own ides/builders whatever). I wanted to know the solution though.
13:02 Arrca joined #salt
13:02 lothiraldan joined #salt
13:02 zipkid well, in that case, conceptually, i'd store all the dev/keys in gpg encrypted files or in an ext_pillar source....
13:03 zipkid some db you could feed via a (web) gui ...
13:03 Arrca Hello guys :) does any know hat's wrong with {% if match.compound('N@solr') %} ? Any way to match a node this way? Thanks !
13:04 sheltonash yeah thats a decent idea too. thanks
13:06 subsignal joined #salt
13:06 niseak joined #salt
13:09 Arrca problem resolved : Need to put salt['match.compound']('N@solr')
13:11 dyasny joined #salt
13:14 fusionx86 joined #salt
13:15 vstoniest joined #salt
13:16 cpowell joined #salt
13:18 toastedpenguin joined #salt
13:19 toastedpenguin left #salt
13:20 primechuck joined #salt
13:23 Tecnico1931 joined #salt
13:23 racooper joined #salt
13:26 sergiotocalini joined #salt
13:28 stoogenmeyer__ is there a way to have {% %} blocks in template files not leave an empty line behind them ?
13:28 phx {% -%}
13:28 phx iirc
13:30 bhosmer joined #salt
13:33 mpanetta joined #salt
13:34 Tyrm joined #salt
13:35 bluenemo joined #salt
13:36 rhodgin joined #salt
13:38 babilen stoogenmeyer__: http://jinja.pocoo.org/docs/dev/templates/#whitespace-control
13:38 stoogenmeyer__ babilen: thanks!
13:44 yggdrasi1 joined #salt
13:45 joehh1 packages for ubuntu going up to launchpad now, debian to debian.saltstack.com very shortly
13:46 joehh1 still waiting in new queue for "official" debian so not uploading 2014.7.4 to experimental yet
13:46 murrdoc thanks joehh1
13:46 mgw joined #salt
13:46 murrdoc whats the version u are pushing
13:47 murrdoc i actually have a test setup now where i can test it for the salt states we use
13:47 murrdoc if u want
13:47 murrdoc i didnt the last time we spoke cos work derailed my ass
13:48 joehh1 2014.7.4
13:48 murrdoc cool , i can test it out on trusty
13:48 murrdoc once i mirror down the ppa
13:48 zwi joined #salt
13:48 murrdoc and precise
13:49 murrdoc thanks again for doing all this btw, packaging and stuff
13:49 joehh1 still building for trusty - I'm watching it before I upload the others to see if there are any build issues
13:49 joehh1 not expecting any, but best to do one first then the others
13:50 murrdoc precise it is
13:50 murrdoc is it in the ppa ?
13:50 murrdoc i need to use a local mirror becuase of secops
13:50 joehh1 doing it next - will be in the salt-testing ppa initially
13:51 murrdoc got it
13:51 joehh1 plan/hope/wish is that I upload it to salt-testing ppa, go to bed, people/saltstack test it, I wake up and copy to main ppas
13:51 murrdoc u have my vote to goto bed
13:51 andrew_v joined #salt
13:51 murrdoc not that that affects anything
13:51 murrdoc but u have it
13:52 StDiluted joined #salt
13:52 joehh1 not yet! haven't uploaded precise yet
13:52 joehh1 looks like trusty build succeeded (https://launchpad.net/~saltstack/+archive/ubuntu/salt-testing/+packages) so precise et al on the way
13:53 babilen Splendid. We do need a better method for preventing such buggy releases as 2014.1.5 or 2014.7.2 from making it into production repositories though
13:53 babilen The last two weeks were rough because of that
13:53 babilen And I am not sure I won't to deal with that ever again
13:53 murrdoc 2014.7.2  works for me
13:53 murrdoc where was it buggy for u
13:54 AndreasLutro .4 still has a critical iptables bug
13:54 Furao joined #salt
13:54 Ligthert_ murrdoc: git-backend bombed out for me
13:54 murrdoc ah, so bugs not related to packaging
13:54 murrdoc cool cool
13:54 babilen murrdoc: https://github.com/saltstack/salt/issues/21480
13:55 AndreasLutro encourage more people to run from source to test new releases
13:55 AndreasLutro I found it easy enough to git clone and run locally
13:55 AndreasLutro but certainly wasn't aided by the salt docs
13:55 timoguin joined #salt
13:56 murrdoc :)
13:56 babilen I am seriously considering switching to pip/bootstrap, but I'm not sure how to deploy that
13:56 murrdoc i hacked up bootstrap to give me the binaries and other setups in a tmp location
13:56 babilen The changes in the "stable" packages are simply way too large and intrusive for me to just deploy them and I don't really want to maintain packages locally
13:57 AndreasLutro does salt not set feature freezes for its releases?
13:57 murrdoc it does
13:57 babilen AndreasLutro: I am talking about changes in minor versions
13:58 murrdoc but it also works with changes from all
13:58 murrdoc i think thats where the issue comes in
13:58 murrdoc so i dont mind it, since its the nature of crowd sourcing
13:58 echo joined #salt
13:58 dleedyna joined #salt
13:58 AndreasLutro babilen: what defines a minor version? I'd imagine that any 2014.7.x release should be purely bugfixes for example
13:59 babilen Yes, you would expect that, wouldn't you?
13:59 numkem joined #salt
13:59 babilen The problem is that the changes in those versions are simply quite intrusive
13:59 joehh1 babilen: i would like to be corrected, but I suspect that the changes in the stable packages are really just changes in the branch
13:59 bhosmer joined #salt
14:00 murrdoc yeah
14:00 joehh1 I suspect you will have the same or similar issues, just maybe with more direct control
14:00 babilen joehh1: Yes, that has nothing to do with the packaging per-se. I'd love a "testing" → "prod" repository, but the diffs between minor version are *massive*
14:01 hasues joined #salt
14:01 mgarcia joined #salt
14:01 hasues left #salt
14:02 joehh1 babilen: It has only taken me several years, but I think I am understanding the parts of the debian infrastructure you are appreciating
14:02 hal58th joined #salt
14:02 joehh1 automatic migration, blocked by RC bugs
14:02 babilen exactly
14:02 murrdoc u just made babilen 's day
14:02 murrdoc :)
14:03 mpanetta Anyone know if it is possible to match against a node group in mine?
14:03 joehh1 It would certainly change the way salt was developed if that was the case
14:03 babilen Right now I am tracking bleeding edge, don't like that as minor releases are anything but minor
14:03 joehh1 agreeed
14:03 babilen But right now I would be happy if I could use salt again
14:04 joehh1 day job tends to work a major release back (ie 2014.1) so there are fewer changes
14:04 joehh1 precise building in ppa:saltstack/salt-testing
14:05 babilen I guess I won't get around either hosting those packages myself or simply deploying with pip
14:06 debian112 joined #salt
14:06 joehh1 My gut feel is that building packages yourself with your own patches is likely to be the most sustainable
14:06 joehh1 that way you can fix the bugs you care about
14:07 babilen Well, I thought I'd get away with tracking a specific branch and testing major releases *sob*
14:08 babilen What branches/suites are there for Debian at the moment btw?
14:08 babilen (in the repo)
14:08 ujjain joined #salt
14:09 joehh1 repo being the git buildpackage repo - unstable and jessie (master), wheezy and squeeze
14:09 joehh1 I didn't manage to build squeeze for 2014.7.2, so it is a little out of date
14:10 joehh1 *more cursing about big changes in point releases*
14:10 Twiglet Clearly following the Apache method of point releases
14:10 Twiglet ;)
14:11 Brew joined #salt
14:12 joehh1 biggest issue for building packages is typically the docs though rather than the code itself
14:12 pdayton joined #salt
14:12 ndrei joined #salt
14:12 joehh1 changes in the code typically don't disrupt the packaging process
14:13 Brew joined #salt
14:13 berserk joined #salt
14:13 perfectsine joined #salt
14:14 Heartsbane joined #salt
14:14 babilen joehh1: No, I meant in the saltstack repositories
14:17 stoogenmeyer__ why is jinja failing on this line {% set main_private_ips = ','.join(salt['mine.get'](query, 'private_ip',  expr_form='grain').values()) %}
14:18 ggoZ joined #salt
14:18 iggy I wonder if we shouldn't rethink the current method of "whatever can go in older branches should go in there and then forward port to newer branches"
14:18 stoogenmeyer__ sorry, correction - it's failing on this line {% set main_private_ips = ','.join(map(lambda ip: ip + ':4151', salt['mine.get'](query, 'private_ip',  expr_form='grain').values())) %}
14:18 murrdoc god damn
14:18 glyf joined #salt
14:18 murrdoc stoogenmeyer__:  write a module function ? to do that in python
14:19 iggy jinja doesn't support lambdas (pretty sure)
14:19 murrdoc and in jinja just call your module
14:19 stoogenmeyer__ iggy: the lambda worked, it's the second line i posted that fails
14:19 stoogenmeyer__ this is the error: Unable to manage file: Jinja syntax error: expected token ',', got 'ip'
14:20 joehh1 not exactly sure how to answer, but for squeeze 0.16.4, 0.17.5, 2014.1.13, 2014.7.1, wheezy 0.16.4, 0.17.5, 2014.1.13, 2014.7.2 (4 in wheezy-testing)
14:20 mapu joined #salt
14:20 daemonkeeper murrdoc: How would I do that? I try to find out right that.
14:20 joehh1 2014.1.13 and 2014.7.4 for jessie
14:20 SpX joined #salt
14:20 iggy lambda is in the second line...
14:21 babilen joehh1: What I mean is: I can use "deb http://debian.saltstack.com/debian wheezy-saltstack main" and jessie-saltstack and 201 ... What is available?
14:21 joehh1 packages for ubuntu utopic, trusty and precise in salt-testing ppa
14:21 babilen +1
14:21 babilen ta
14:21 favadi joined #salt
14:21 stoogenmeyer__ iggy: sorry, missed that
14:22 debian112 is this still experimental on debian systems? http://docs.saltstack.com/en/latest/ref/states/all/salt.states.network.html
14:22 debian112 in 2014.7.0
14:22 debian112 ?
14:22 iggy debian112: yes
14:22 murrdoc daemonkeeper:  mkdir (whateveryourfileroot)/_modules, edit (whateveryourfileroot)/_modules/mymodule.pu, add a  __virtual__ function that returns your module name, add a function that does what u need in python and then in jinja {% set varname =  __salt__['mymodule.sexyfunction'](params) %}
14:22 debian112 ok thanks iggy
14:22 murrdoc and then go from there
14:22 daemonkeeper murrdoc: Thanks!
14:23 mpanetta murrdoc: s/pu/py/ :P
14:23 joehh1 with apologies to all:
14:23 joehh1 jessie-saltstack
14:23 joehh1 jessie-saltstack-2014-01
14:23 joehh1 jessie-saltstack-2014-07
14:23 joehh1 jessie-testing
14:23 joehh1 squeeze-saltstack
14:23 joehh1 squeeze-saltstack-016
14:23 joehh1 squeeze-saltstack-017
14:23 joehh1 squeeze-saltstack-2014-01
14:23 joehh1 squeeze-saltstack-2014-07
14:23 joehh1 squeeze-testing
14:23 joehh1 unstable
14:23 joehh1 unstable-testing
14:23 joehh1 wheezy-saltstack
14:23 joehh1 wheezy-saltstack-016
14:23 joehh1 wheezy-saltstack-017
14:23 joehh1 wheezy-saltstack-2014-01
14:23 joehh1 wheezy-saltstack-2014-07
14:23 joehh1 wheezy-testing
14:23 joehh1 babilen: is that what you are after?
14:24 babilen daemonkeeper: http://docs.saltstack.com/en/latest/ref/modules/
14:24 babilen joehh1: Exactly (couldn't find that anywhere)
14:24 iggy well... at least we have that list of debian repos in the channel logs now ;)
14:24 joehh1 I'll put them on the debian.saltstack.com page
14:24 babilen +1
14:24 Romlok it would've been faster to just say http://debian.saltstack.com/debian/dists/
14:25 murrdoc also daemonkeeper https://www.youtube.com/watch?feature=player_detailpage&v=YP73LM8mzL0
14:25 murrdoc mpanetta:  i had poo on my brain
14:25 murrdoc brb
14:25 murrdoc :D
14:25 Xevian joined #salt
14:25 babilen o_O
14:25 mpanetta heh
14:36 rhodgin joined #salt
14:37 MartinMakarov joined #salt
14:38 daemonkeeper Could it be murrdoc meant salt for the jinja context, not __salt__?
14:40 babilen yeah
14:41 babilen http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.jinja.html#calling-salt-functions
14:41 iromli joined #salt
14:42 wendall911 joined #salt
14:42 niseak_ joined #salt
14:43 MartinMakarov left #salt
14:43 MartinM_ joined #salt
14:45 evilrob joined #salt
14:45 sk_0_ is there a way to look at diff of files across nodes with salt?
14:46 jalbretsen joined #salt
14:46 otter768 joined #salt
14:47 _JZ_ joined #salt
14:48 murrdoc joined #salt
14:49 conan_the_destro joined #salt
14:50 c10b10 joined #salt
14:50 murrdoc .py ftw
14:50 berserk joined #salt
14:51 c10b10 hello
14:51 daemonkeeper Thanks, that worked.
14:51 _JZ_ joined #salt
14:52 primechuck joined #salt
14:53 c10b10 any idea why i would be getting "sys.state_doc" is not available. when I run sudo salt '*' sys.state_doc service.running
14:54 evilrob joined #salt
14:54 enarciso joined #salt
14:55 iggy what version of salt?
14:55 c10b10 salt 2014.7.2 (Helium)
14:55 Twiglet is there a way to see if a master is loading gitfs_remote are loading?
14:55 Twiglet nm debugmode
14:55 Twiglet derp
14:57 iggy Twiglet: salt-run -l debug fileserver.update
14:57 iggy c10b10: that version everywhere? (master and minions?)
14:57 gladiatr joined #salt
14:57 Twiglet iggy: nice, cheers
14:58 Twiglet and no they're not
14:58 Twiglet odd
14:59 iggy Twiglet: do you have the deps installed (gitpython, pygit2, dulwich)
15:00 Twiglet Installing with the bootstrap script, not too sure
15:00 iggy then no, it's not
15:00 Twiglet ah ok, cheers will get them installed
15:00 iggy curl -s -L https://bootstrap.saltstack.com | sudo sh -s -- -M -N -p python-git git 2015.2
15:00 iggy that's what we use on our masters
15:01 iggy (the -p python-git being the important part here)
15:01 Twiglet awesome cheers
15:01 Twiglet in the middle of trying to work out how to test our salt states with jenkins
15:02 Twiglet so far the idea is vagrant with a vagrant master, will see how it goes
15:02 c10b10 @iggy, seems the minions were rocking 0.17.5, ty
15:03 bhosmer joined #salt
15:06 bhosmer_ joined #salt
15:07 thedodd joined #salt
15:07 glyf joined #salt
15:08 scbunn joined #salt
15:09 evilrob joined #salt
15:10 user0901 joined #salt
15:12 user0901 hello, I am having an issue with salt-bootstrap script on RHEL 6.5 - basically the script doesn't honor -U option, so it doesn't update the system before trying to install salt. Any ideas?
15:13 MTecknology iggy: It looks like even if you got raet working, you'd still be missing some of the coolness of it that's planned for 2015.2.0. I'm likely going to hold off on raet until that release and hope that dependencies get better in that time.
15:13 jdowning joined #salt
15:14 MTecknology I'll also be on a pure debian-stable environment by then! :D  (because debian should definitely probably release 8 by then, maybe)
15:14 MTecknology nacl and sodium look pretty sexy too ... patience is hard
15:15 luke joined #salt
15:15 MTecknology I might start doing some testing in the next few months, though.
15:15 ek6 patience is hard...having to have a totally different set of keys made it a bit easier for me to wait
15:19 mwpher joined #salt
15:20 user0901 just updated the salt bootstrap script wit salt-cloud -u and latest version doesn't work either...
15:21 out180 joined #salt
15:23 mwpher Hey folks, I'm wanting to load a set of default options and merge them with specified pillar settings. according to the docs I can pass kwargs 'default' and 'merge' to pillar.get, but it says it doesn't take kwargs and only expects two arguments. Are the docs out of date? And how could I merge defaults with pillar data otherwise?
15:24 Tecnico1931 joined #salt
15:25 murrdoc {% set magic  = salt['pillar.get'](
15:25 murrdoc 'somepillar', <— pillar key
15:25 murrdoc default=default_settings, <<— needs to be a dict
15:25 murrdoc merge=True)
15:25 murrdoc %}
15:28 mgw joined #salt
15:30 mwpher Thanks murrdoc, but that's exactly what I'm doing. It fails saying that pillar.get doesn't accept keyword arguments. If I remove the keywords and just try "'somepillar', default_settings, True", it says it only takes two arguments.
15:30 rdas joined #salt
15:30 murrdoc i have no keywords
15:31 murrdoc check the version of salt
15:33 mwpher It's 2015.2.0rc2. Moreover, when I look at the source code for modules.pillar.get(), I see it clearly accepts the keyword arguments, so I'm not sure what's going on
15:35 mwpher Oh!
15:35 mwpher Ok, that's strange
15:35 raygunsix joined #salt
15:35 stanchan joined #salt
15:35 mwpher I was getting these errors when using the format salt.pillar.get instead of salt['pillar.get']
15:36 mwpher From what I've been told, these should be the same
15:36 mwpher Maybe the former uses an outdated version of pillar.get?
15:37 peters-tx joined #salt
15:40 clintberry joined #salt
15:40 timoguin joined #salt
15:40 murrdoc i havent used salt.pillar.get ever
15:41 rm_jorge joined #salt
15:41 murrdoc that looks like a dict.get
15:43 mwpher I thought that salt.x.y was supposed to be equivalent to salt['x.y']
15:43 murrdoc doubt it
15:43 mwpher Maybe I'm wrong, or maybe it's not implemented correctly
15:44 niseak joined #salt
15:45 dopesong joined #salt
15:45 mwpher It looks like that's what it's supposed to be: http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.jinja.html#calling-salt-functions
15:46 TheoSLC joined #salt
15:46 murrdoc ouch
15:46 murrdoc thank u for rtfm mwpher
15:46 seev joined #salt
15:47 TheoSLC Greetings.  I need some expert eyes on this bug https://github.com/saltstack/salt/issues/22685  Any takers?
15:48 cpowell Greetings, I am looking into the salt reactor. Specifically the 'mapping events to reactor sls files' and I see in the documentation example that you can have multiple sls's run
15:48 cpowell http://docs.saltstack.com/en/latest/topics/reactor/#mapping-events-to-reactor-sls-files
15:48 cpowell are they guarenteed to run in that order?
15:49 iggy MTecknology: oh, I'm running 2015.2 already
15:49 user0901 if [ "$DISTRO_NAME_L" = "oracle_linux" ]; then that should certainly be             if [ "$DISTRO_NAME_L" == "oracle_linux" ]; then ...
15:50 user0901 not that it helps RHEL but still...
15:51 ndrei joined #salt
15:51 iggy mwpher: probably a lazyloader issue
15:51 iggy mwpher: look at the 2 open bugs for 2015.2 and see if the one about lazyloader might apply to your situation
15:52 iggy cpowell: probably not
15:52 iggy cpowell: a lot of areas still use normal dict's (vs OrderedDict's)
15:52 mwpher How do I filter the bugs to 2015.2?
15:52 cpowell thanks iggy
15:52 murrdoc iggy:  how does one 'look at 2 open bugs for 2015.2 a'
15:53 iggy milestone in github
15:53 iggy https://github.com/saltstack/salt/issues?q=is%3Aopen+is%3Aissue+milestone%3ALithium
15:54 mwpher Ok, thank you. Not sure why they use those names though when the version numbers are so short
15:54 mwpher :P
15:54 iggy because the version numbers are based off the release date and they don't know when they are going to release 3 versions from now
15:55 iggy supposedly anyway
15:55 iggy the last couple of releases have not been like that at all
15:55 iggy 2014.7 came out in 2014.12 and 2015.2 isn't out yet
15:56 iggy (and it's a little past February)
15:56 Twiglet heh watching Tom's docker talk with him stating "Dat versioning makes more sense" did make me laugh out loud
15:57 Twiglet Date*
15:57 ek6 iggy: nah 2014.7 was more like 2014.9
15:57 iggy https://groups.google.com/forum/#!topic/salt-announce/Il7YbDPZJEc
15:57 iggy 11/19
15:57 iggy so I was off by 2 weeks
15:58 ek6 hm..guess i pulled an iggy because i was running 7 in production in early sept
15:58 iggy yeah.... don't be like me
15:59 rick-copeland joined #salt
16:00 rick-copeland Has anyone tried using the DigitalOcean v2 salt-cloud driver with more than 100 droplets? It appears that it does not handle pagination...
16:00 mwpher Hey iggy, you're totally right. That bug was the issue. I guess that the lazyloader doesn't translate those notations properly. My confusion came from the fact that it wasn't loading a module at all, but an internal salt method (I guess?!?). I would've known what was up if I had tried some module that didn't exist as a method, like salt.test.ping
16:00 cpowell does anyone have an example of using the reactor to call an orchestration? The syntax is killing me.
16:01 agend joined #salt
16:01 iggy cpowell: yes... hold on let me find it
16:03 denys joined #salt
16:04 clintber_ joined #salt
16:04 Heartsbane joined #salt
16:06 iggy cpowell: https://gist.github.com/iggy/b7599a97c18ddd6e6ca4
16:06 iggy kind of messy, so let me know if you have any questions
16:06 mccricardo joined #salt
16:07 cpowell ahh, you are using the runner
16:07 cpowell thank you very much!
16:07 cpowell I was using local.state.orchestrate
16:09 pahko joined #salt
16:10 mccricardo left #salt
16:10 robinsmidsrod joined #salt
16:12 aparsons joined #salt
16:13 mccricardo joined #salt
16:13 mccricardo left #salt
16:14 kawa2014 joined #salt
16:14 KyleG joined #salt
16:14 KyleG joined #salt
16:15 ndrei joined #salt
16:16 MTecknology iggy: running the dev branch?
16:17 iggy MTecknology: 2015.2 branch
16:17 MTecknology err, I meant that.. that's a development branch, isn't it?
16:17 murrdoc nope
16:18 iggy It's supposedly feature frozen
16:18 murrdoc its blocked-on-iggy status right now
16:18 murrdoc thankfully
16:18 bhosmer joined #salt
16:18 MTecknology I assumed it was feature frozen and such, but I thought it was basically HEAD for 2015.2.X
16:19 iggy well, yeah
16:19 iggy guess it depends how you look at it
16:19 dopesong joined #salt
16:19 iggy it's supposed to just be fixes at this point
16:19 iggy it hasn't really been that way
16:21 Workflow left #salt
16:22 bezeee joined #salt
16:23 scott_ joined #salt
16:23 iggy oooh, fancy {{ show_fill_context() }}
16:23 enarciso joined #salt
16:23 iggy show_full_context
16:24 pahko joined #salt
16:24 iggy whoever put i and u next to each other obviously didn't live in the 21st century
16:25 fyb3r joined #salt
16:25 scott_ I have a state file that checks if a host is VMware, using grains.item virtual. It uses cmd.run with - check_cmd: /bin/true  or /bin/false to determine if the state should succeed or fail.
16:26 scott_ I want to use that state in another state
16:26 scott_ Is it possible to have a state use the output of another state, and do something based on that output?
16:27 p66kumar joined #salt
16:28 bhosmer joined #salt
16:30 mgw joined #salt
16:30 iggy you can have a state run only when another one "changes"
16:32 stoogenmeyer__ joined #salt
16:32 DaveQB joined #salt
16:32 ndrei joined #salt
16:35 bluenemo_ joined #salt
16:35 keimlink_ joined #salt
16:36 tkharju joined #salt
16:39 bezeee joined #salt
16:39 bhosmer joined #salt
16:41 nickdew joined #salt
16:42 desposo joined #salt
16:43 writtenoff joined #salt
16:45 bhosmer_ joined #salt
16:45 MatthewsFace joined #salt
16:47 otter768 joined #salt
16:47 gthank Is Salt well-suited to the task of spinning up new instances in AWS? This includes things like launching a new instance of given type using a specified AMI, creating new EBS volumes, attaching them, and then doing config management?
16:48 gthank Or would it handle just the CM part, and I'd use something different for the rest?
16:56 TaiSHi joined #salt
16:57 iggy salt-cloud (part of salt) can do the cloud managment side
16:58 iggy the last time I tried it, it didn't have great support for volume handling (I assume because that functionality varies much more wildly across cloud providers)
16:59 bhosmer joined #salt
16:59 gthank iggy Thanks
16:59 iggy that was over a year ago though, so...
17:00 raygunsix joined #salt
17:00 iggy at the time it looked like the functionality existed, but just hadn't been tied into the cloud map functionality
17:00 iggy for aws, there's also all the boto modules/states
17:01 evle1 joined #salt
17:02 catpig joined #salt
17:03 murrdoc joined #salt
17:04 murrdoc joined #salt
17:05 spookah joined #salt
17:05 Nazca joined #salt
17:06 rick-copeland joined #salt
17:09 murrdoc joined #salt
17:10 tajh joined #salt
17:10 Andre-B joined #salt
17:11 nickdew Are there any tools for pulling in salt formulas into my local repo?
17:12 evilrob anyone here play with salt-cloud?
17:13 iggy nickdew: most people use them as gitfs remotes
17:14 nickdew iggy: won't that automatically pull in remote changes?
17:14 iggy there is no npm/berkshelf/etc
17:14 iggy nickdew: not if you fork your own copy (like the formula docs say)
17:15 nickdew Okay. Thanks.
17:15 nickdew * Hating stash right now
17:18 KyleG man
17:18 KyleG fucking stash
17:18 KyleG ugh
17:18 murrdoc git stash ?
17:18 murrdoc why hate on something that nice
17:18 KyleG atlassian stash
17:18 nickdew Atlassian Stash
17:18 murrdoc oh wtf is that
17:18 KyleG ^
17:18 KyleG GIT hosting
17:18 nickdew Atlassian Git server
17:19 KyleG https://www.atlassian.com/software/stash
17:19 desposo1 joined #salt
17:21 nickdew GitHub is just so much easier.
17:21 smcquay joined #salt
17:21 KyleG we can't all use github
17:21 KyleG I work for a SaaS company
17:21 bhosmer joined #salt
17:21 KyleG Our code is protected by patents and whatnot.
17:22 lz-dylan evilrob: salt-cloud is nifty and functional, but I find it syntactically difficult
17:22 nickdew Someone here just really likes Atlassian, we use everything.
17:23 evilrob lz-dylan: problem I'm having is I can create an instance just fine until I swap out auth to the role_id with a '.' in it.
17:23 evilrob command-line it all works fine.
17:23 hal58th_ joined #salt
17:23 evilrob I can do a salt-cloud --list-images with the role ID
17:23 murrdoc could be worse
17:23 evilrob but not create images
17:24 murrdoc TFS or whatever the fuck Microsoft shiped with visual studio
17:24 evilrob so I question the "functional" part of that statement.  syntactically difficult? I can deal with that.
17:24 _1_Ovawais joined #salt
17:27 lz-dylan evilrob: hm. salt-cloud isn't really a fresh topic for me, as it's beem...probably four, five months? I do recall having managed to get IAM roles for provisioning to work.
17:27 lz-dylan is that roughly what you're going for, or am I drastically misreading?
17:27 smcquay joined #salt
17:29 kawa2014 joined #salt
17:30 lz-dylan tbh I may just not know what you're talking about :)
17:31 evilrob lz-dylan: corporate environment.  the group that runs the openstack servers uses corporate ID for auth.  with my ID, I can deploy just fine using salt-cloud or nova commandline.  with the role ID (that doesn't have my intranet password in config files) I can create instances with the nova command-line, but not salt-cloud
17:31 kaictl joined #salt
17:32 iggy the company that owns mine also owns atlassian (or a majority of it)... so they keep trying to get us to switch
17:32 evilrob when I try, I get "The following exception was thrown by libcloud when trying to run the initial deployment: 400 Bad Request Invalid key_name provided."
17:33 ty2u joined #salt
17:33 lz-dylan evilrob: shoot, I think that's beyond me. I've really only used salt-cloud with AWS, so other auth systems aren't familiar.
17:36 yggdrasi1 joined #salt
17:38 bmac2 joined #salt
17:38 TOoSmOotH joined #salt
17:39 TOoSmOotH So I have been having a weird issue of late.. I run salt-call state.hihgstate every 15 minutes via cron
17:39 TOoSmOotH and every once ans a while the salt-call will run for days
17:39 TOoSmOotH if I kill the process and rerun everything is fine
17:39 TOoSmOotH Anyone else seen anything like that?
17:41 TOoSmOotH I found this but there is no resolution just how to kill it: https://groups.google.com/forum/#!topic/salt-users/7QEGnSmK4-4
17:42 denys joined #salt
17:44 cpowell joined #salt
17:45 timoguin joined #salt
17:49 notbmatt joined #salt
17:50 notbmatt hi everyone! we're seeing apparently-random failures to source files out of gitfs in file.managed states
17:51 notbmatt is this a known issue, or does anyone have any recommendations on how to diagnose?
17:52 ajw0100 joined #salt
17:58 enarciso joined #salt
18:00 sergiotocalini joined #salt
18:00 druonysus joined #salt
18:00 druonysus joined #salt
18:00 echo joined #salt
18:02 conan_the_destro joined #salt
18:03 ek6 notbmatt: which gitfs provider are you using?
18:04 krelo joined #salt
18:04 notbmatt good question, let me check - fwiw, I also just noticed that I'm getting "Error in function _serve_file" messages in our logs but due to implementation bugs I can't get the actual trace :(
18:05 ek6 and are these git repos that see lots of churn?
18:05 notbmatt ek6: it looks like we've got 'gitfs_provider: GitPython' set
18:05 ek6 you see anything odd when you run the master with -l debug
18:05 notbmatt not a huge amount, no - on the order of 1 commit over tens of minutes
18:05 notbmatt yeah, since logging is broken I'm going to try and repro with debug on
18:06 notbmatt sorry for not getting useful log output beforehand :(
18:07 ek6 ok so my 1c is to look for the gitfs provider...ill get the error wrong...invalidating the cache...i saw lots of issues myself with...after a change on the first refresh the gitfs provider would invalidate and throw everything out...on the next pass it would do a clean pull and everything would be back to normal
18:07 smcquay joined #salt
18:07 ek6 but if you hit it in that middle 1min well..nothing would be there
18:08 speedlight hi ppl, dont find in the docs yet, is posible to "migrate" a master from one machine to another? i mean, i.e. my box is the master now and i already accepted some minions keys, now i need to have the master in other machine (no both)
18:08 speedlight the files are in git no prob about that, the main thing is the keys
18:08 drawsmcgraw speedlight: If I understand correctly, you can copy the contents of /srv/salt over to your new salt master
18:08 drawsmcgraw That's what I do when I back up my masters, anyway.
18:08 murrdoc also the key dir
18:08 drawsmcgraw murrdoc: that's under /srv/salt, right?
18:09 murrdoc nope
18:09 speedlight yes the srv/slat is in git so no issue
18:09 drawsmcgraw oye
18:09 ek6 well i just tar up /etc/salt myself
18:09 drawsmcgraw sorry
18:09 drawsmcgraw not /srv/salt. /etc/salt
18:09 speedlight i just copy the /etc/salt and is done?
18:09 drawsmcgraw my mistake. I had meant /etc/salt/.   Force of habit :/
18:09 murrdoc specifically /etc/salt/pki/
18:09 ek6 well all of etc salt will cover you for the DEFAULT location of config files and keys
18:09 drawsmcgraw speedlight: That's my understanding, yes.
18:10 speedlight yes is all default
18:10 speedlight i now the key s are in pki dir, but wasnt sure if cp that is enough
18:10 speedlight ok
18:10 speedlight tias
18:11 speedlight try it and see XD
18:11 speedlight first lunch jojo
18:11 speedlight thnkz ppl
18:12 Vynce joined #salt
18:13 timoguin joined #salt
18:15 baweaver joined #salt
18:21 mkropinack joined #salt
18:22 elfixit joined #salt
18:23 druonysus joined #salt
18:23 druonysus joined #salt
18:25 tomasfejfar joined #salt
18:28 seev joined #salt
18:30 _ikke_ I suddenly don't get a response anymore from my minion. I run salt -t30 pi2 test.ping, and nothing returns. If I look up the job id, with jobs.lookup_jid, I get no response. On both master and minion, I see the job being executed.
18:31 _ikke_ Any idea why?
18:33 borgstrom joined #salt
18:34 _ikke_ tried restarting master and minion, but to no avail
18:35 Aidan_ joined #salt
18:35 fyb3r experiencing something similar on my minion as well. working on it
18:36 _ikke_ trying to increase logging
18:37 Aidan_ left #salt
18:38 tomasfej1 joined #salt
18:39 aparsons joined #salt
18:40 enarciso joined #salt
18:40 giantlock joined #salt
18:42 _ikke_ oh, suddenly I get a response again, weird
18:43 borgstrom joined #salt
18:47 jeremyalons communication is opened on both ports 4505 and 4506?
18:47 murrdoc jeremyalons:  o/
18:47 murrdoc how u been man
18:47 jeremyalons Busy.
18:48 jeremyalons also, \o
18:48 murrdoc :)
18:48 jeremyalons sorry to leave you hanging
18:48 murrdoc no clue what u talking about
18:48 jeremyalons How are things in your neck of the woods?
18:48 murrdoc but how dare u !
18:48 murrdoc is all good
18:48 otter768 joined #salt
18:48 CeBe1 joined #salt
18:49 timoguin joined #salt
18:49 jeremyalons Currently laughing at just how pathetic iptables NATing is for throughput and wondering how anybody uses this garbage
18:50 murrdoc man
18:50 murrdoc maaan painful stuff
18:51 * jeremyalons disappears again
18:51 murrdoc o/
18:54 terratoma joined #salt
18:58 druonysus joined #salt
18:58 druonysus joined #salt
18:59 stanchan joined #salt
19:00 tajh joined #salt
19:01 seev joined #salt
19:02 hasues joined #salt
19:04 borgstrom joined #salt
19:04 amcorreia joined #salt
19:05 hasues left #salt
19:07 soren joined #salt
19:08 jalbretsen joined #salt
19:08 zircote joined #salt
19:08 piru_ joined #salt
19:08 tajh joined #salt
19:15 berserk joined #salt
19:16 Vynce left #salt
19:16 ndrei joined #salt
19:17 aw110f joined #salt
19:17 robinsmidsrod joined #salt
19:18 spookah joined #salt
19:19 hal58th_ How many rules do you have jeremyalons?
19:20 MaliutaLap joined #salt
19:20 MaliutaLap left #salt
19:20 mgw joined #salt
19:21 bhosmer joined #salt
19:24 timoguin gthank: not sure if you ever got a response, but salt cloud does a fine job of spinning up instances
19:24 gthank timoguin One person recommended it, but only half-heartedly, because their experience was not recent.
19:25 gthank I'm glad to hear a second person vouch for it
19:25 mpanetta salt-cloud rocks my world
19:25 murrdoc u ho
19:25 mpanetta There, that is my recommendation :P
19:25 murrdoc u hold one second
19:25 murrdoc :D
19:25 mpanetta haha
19:28 tomh- joined #salt
19:33 spookah joined #salt
19:37 ndrei joined #salt
19:37 zircote joined #salt
19:38 RabidCicada joined #salt
19:39 bhosmer joined #salt
19:42 ndrei joined #salt
19:43 iggy babilen: basepi: other -formulas people: are there any rules we are supposed to follow? I see people committing to master, approving their own PRs, etc... Seems like a bad idea
19:44 murrdoc thats all bad
19:44 timoguin no one should commit to master. but permissions for that org are weak as fuck.
19:45 iggy especially those particular commits that were questionable to begin with
19:47 nickdew joined #salt
19:48 basepi iggy: whiteinge might have guidance there
19:48 basepi We should probably write up a doc
19:49 iggy I always forget he's in here
19:51 basepi The problem is that github doesn't have a mechanism to stop direct pushing.
19:51 basepi Because you can merge pull requests on the command line, and that requires pushing.
19:51 iggy right, but at least if there's something written down, we can say... You're not supposed to do that
19:51 basepi So it will have to be just a best-practices doc, and maybe some policing and kicking people out of the org if they don't follow them.
19:51 basepi Yep.
19:51 iggy I think everybody there is mature enough to follow rules
19:52 iggy if they know what said rules are
19:53 timoguin Bitbucket has branch-level permissions. :)
19:53 murrdoc that is so nice
19:53 murrdoc github still doesnt
19:53 murrdoc well git doesnt
19:54 timoguin Well obviously we need to move all the formulas to bitbucket
19:54 timoguin Probably should switch to Mercurial too
19:54 murrdoc yes and no
19:55 evidence anyone ever tweak file_buffer_size on the master?  does that relate to the cache for talking with the backend (ie gitfs), or is it similar to the send window for the clients
19:55 iggy I've been following a couple formulas for a while (that are apparently fairly popular) so I always assumed other formulas were fairly popular
19:56 iggy I'm starting to see that's not the case
19:56 evidence gitfs is constantly failing as i ramp up.. only ~200 minions so far
19:57 iggy I'm pretty sure in the last couple days I've had more commits to the formulas than everyone else combined
19:58 ggoZ joined #salt
20:00 baweaver joined #salt
20:00 druonysuse joined #salt
20:00 druonysuse joined #salt
20:05 markm joined #salt
20:07 subsignal joined #salt
20:09 bezeee joined #salt
20:10 speedlight why if i set the id of a minion the master show the minion key in unaccepted when was already accepted?
20:10 dynamicudpate joined #salt
20:10 speedlight i mean the minion is in both, accepted and unaccepted
20:11 timoguin speedlight: because you changed the id
20:11 murrdoc <schrodinger jokes>
20:11 timoguin the master ties minion id to the key
20:11 speedlight oh
20:11 speedlight clear enough XD
20:11 timoguin reject it and reaccept and you should be good
20:11 speedlight thnkz again
20:13 bezeee 0anyone know if there are known bugs when using multiple “source” values in file.managed?
20:13 bezeee https://gist.github.com/brianz/0e63ed502974e0762e7a
20:13 bezeee seems like specifying muliple values isn’t working at all
20:13 bezeee when i set it to just use a single source it works fine
20:13 bezeee v2014.7.4
20:13 evidence is this right? https://github.com/saltstack/salt/blob/develop/salt/config.py#L776
20:14 evidence the config states the default is a sane 60
20:14 evidence that's for the minion.. the master has a proper default of 60
20:14 timoguin bezeee: I think you need "sources" for that to work
20:14 timoguin actually doesn't look like that's possible eyt
20:15 timoguin https://github.com/saltstack/salt/issues/7532
20:15 timoguin not with file.managed at least
20:16 hal58th__ joined #salt
20:16 hal58th_1 joined #salt
20:16 bezeee hrmm….docs state it’s possible using “source”
20:17 bezeee “A list of sources can also be passed in to provide a default source and a set of fallbacks. The first source in the list that is found to exist will be used and subsequent entries in the list will be ignored.”
20:17 ipmb joined #salt
20:17 bezeee “  file.managed:
20:17 bezeee - source:
20:17 bezeee - salt://file_that_does_not_exist
20:17 bezeee - salt://file_that_exists
20:17 manfred bezeee:  where does it say that?
20:17 timoguin file.append can do it.
20:17 timoguin Who broke the file state docs? http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html
20:17 timoguin Page is empty for me
20:18 bezeee yeah…in there
20:18 bezeee looking at the google cached version
20:18 manfred is that available in 2014.7?
20:19 manfred be8474086a4ba047e39fddade31df00a6c737ce4
20:19 manfred it was added in that commit
20:19 bezeee docs I’m reading…here’s a screen grab
20:19 manfred which was made in Date:   Tue Sep 2 14:07:27 2014 -0600
20:19 bezeee https://www.dropbox.com/s/y4awfjwtdr3e31n/Screenshot%202015-04-15%2014.19.01.png?dl=0
20:20 manfred is it in the man 7 salt page for your salt install?
20:20 manfred cause it looks like that is only in develop?
20:20 manfred and then 2015.2
20:20 bezeee looking...
20:20 manfred or actually, the documentation was added then
20:21 ndrei joined #salt
20:21 timoguin According to that doc, the first source in teh list will be used and the rest ignored
20:21 timoguin It's a fallback system
20:21 manfred but that wasn't added to the documentation until september of last year
20:21 bezeee right…i’m getting errors though when source is a list rather than a single value
20:21 manfred so it wouldn't have been backported to 2014.7
20:23 manfred fa7e9fc6
20:23 manfred it was added last January
20:23 manfred so it should have been in there
20:24 bezeee in the man page it states the same thing: “A list of sources can also be passed in to provide a default source and a set of fallbacks. The  first  source  in
20:24 bezeee the list that is found to exist will be used and subsequent entries in the list will be ignored.”
20:25 bezeee There error I get: “Comment: Unable to manage file: none of the specified sources were found”
20:27 timoguin But works fine with just one item?
20:29 bezeee not as a 1-item list
20:29 bezeee but lemme verify that again…
20:30 bezeee “source: salt://“ works fine
20:30 baweaver joined #salt
20:31 DanGarthwaite joined #salt
20:33 manfred bezeee:  i would report a bug with that
20:33 tomasfej1 joined #salt
20:34 bezeee manfred: k…i’ll do a bit more testing and make sure I have it
20:34 mapu joined #salt
20:34 bezeee thanks!
20:34 davisj is it possible to reference nodegroups in a sls template? i.e. {% for node in some_node_group %}
20:36 manfred no problem
20:36 timoguin bezeee: looks like this is the part that's failing: https://github.com/saltstack/salt/blob/2014.7/salt/modules/file.py#L2395
20:36 manfred ^^ yeah
20:36 manfred if you can also include the -l debug information from a salt-call with that
20:36 manfred in the bug report, would be super helpful
20:36 timoguin Called here in the state: https://github.com/saltstack/salt/blob/2014.7/salt/states/file.py#L1366
20:37 bezeee ok…cool…helpful.  I’ll read through the source and see if I can spot the issue
20:40 Enwony joined #salt
20:40 druonysuse joined #salt
20:41 DanGarthwaite @davisj: I've never tried it, or would have considered it.  Is something preventing you from setting and using grains?
20:43 DanGarthwaite salt -N nodegroup grains.append roles webserver
20:43 DanGarthwaite Then you can use the grain in templates
20:44 davisj DanGarthwaite: I get a bad feeling about the source of truth for that being the client. Also nodegoups seem more, er... centralized
20:44 murrdoc joined #salt
20:48 DanGarthwaite You can use pillars exactly the same way, and then truth is on the master.
20:49 otter768 joined #salt
20:53 JPaul joined #salt
20:53 murrdoc can i do an - unless  based on a state ?
20:54 murrdoc like if i state to ensure something is a symlink, can i put that as an unless in another state to not run
20:54 murrdoc googling
20:55 losh joined #salt
20:56 murrdoc there isnt
20:56 murrdoc unless only uses commands and not states
20:56 JPaul joined #salt
20:57 dendazen joined #salt
20:58 kunit joined #salt
20:58 kunit "Unable to determine upstream hash of source file http://artifactory.blah"
20:58 kunit I've added the md5 hash to the state
20:59 kunit - source_hash: md5=72b7ef~
20:59 kunit I'm using archive.extracted
20:59 kunit should I switch to "salt.modules.artifactory"?
21:01 hal58th joined #salt
21:01 pdayton joined #salt
21:02 hal58th_ joined #salt
21:03 smcquay joined #salt
21:04 fyb3r left #salt
21:06 Xeryuz joined #salt
21:11 timoguin joined #salt
21:12 murrdoc this a blank page for anyone else http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html
21:13 kunit only title a second ago for me
21:13 timoguin yep
21:13 raygunsix joined #salt
21:13 timoguin I don't see anything wrong with the source file though unless I'm missing something
21:13 Xeryuz left #salt
21:14 eliasp is there a straightforward salt-ish way to change the SELinux context of a directory (besides cmd.run/-unless/…)?
21:19 c10b10 joined #salt
21:22 gladiatr joined #salt
21:22 glyf joined #salt
21:26 p66kumar joined #salt
21:33 murrdoc joined #salt
21:35 murrdoc joined #salt
21:42 Tyrm joined #salt
21:42 glyf joined #salt
21:47 robinsmidsrod joined #salt
21:52 theologian joined #salt
21:56 Crick1 joined #salt
21:56 Crick1 left #salt
22:01 pahko joined #salt
22:02 baweaver joined #salt
22:10 druonysuse joined #salt
22:11 desposo joined #salt
22:17 racooper execuction module or state? either way, check the "file" module for the appropriate useage
22:18 racooper exec module: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.file.html
22:18 Andre-B_ joined #salt
22:19 Vynce joined #salt
22:19 Vynce i can't help but feel the documentation her eis missing something.  http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html
22:20 ingwaem joined #salt
22:21 racooper um....heh. put a bug on github for that probably...
22:23 racooper it's actually missing from the list of states on http://docs.saltstack.com/en/latest/salt-modindex.html too
22:26 Vynce sadly, the page does not seem to have a link for where to file a bug.
22:27 ingwaem` joined #salt
22:27 racooper github.com/saltstack
22:28 racooper it's the same place you file for any other salt bugs.
22:28 Vynce yeah, i found that… i don't *see* a separate repo for docs, so i guess it's the /salt repo?
22:28 racooper yes, there's not a separate docs repo
22:29 Vynce OK.  filed as #22720
22:31 SeeDickCode joined #salt
22:31 druonysus joined #salt
22:31 racooper in the meantime... for eliasp: https://salt.readthedocs.org/en/v0.17.4/ref/states/all/salt.states.file.html
22:32 iggy the docs are mostly inline in the source (with some longer and freeform docs separate)
22:34 murrdoc iggy:  but the last release broke docs
22:34 murrdoc like the pages are empty
22:34 murrdoc except for headers
22:34 iggy I know
22:34 murrdoc must be for security reasons
22:34 murrdoc @see: cmd.shell
22:34 racooper though..it doens't appear that the states.file module actually supports selinux contexts, or if so there's no example of it
22:34 iggy I was referring to "where do we file bugs for docs... I don't see a docs repo"
22:35 iggy lol... cmd.shell can go to hell
22:35 aparsons_ joined #salt
22:36 iggy both of the f states are gone
22:36 murrdoc f state?
22:37 iggy salt.states.f*
22:37 iggy file and firewalld
22:37 iggy both are gone from the index
22:38 ek6 i mean file? who uses that anyway
22:39 shoemonkey joined #salt
22:40 Vynce another docs related contribution at https://github.com/saltstack/salt/pull/22722
22:40 murrdoc (clap)
22:44 baweaver joined #salt
22:45 hal58th__ joined #salt
22:45 hal58th_1 joined #salt
22:46 aparsons joined #salt
22:46 Vynce f the state
22:48 aquassaut joined #salt
22:48 murrdoc did we tell basedpi shits broke
22:50 otter768 joined #salt
22:50 nafg_ joined #salt
22:51 iggy I didn't
22:51 iggy I already invoked him once today
22:51 murrdoc hes probably heard
22:51 Vynce OK, so question unanswered by the file.managed docs, as far as i can tell — can the destination path be relative?  if it is, what's it relative to?
22:52 iggy ion, .ninja domains are super cheap now (<$3 ... mine was like 15)
22:52 iggy no
22:52 iggy there's an issue open about that
22:52 robawt Vynce: why would a relative path be better than a pillar value?
22:52 Vynce robawt: becaus ei know how to use a relative path?
22:52 Vynce and i have been unable to make heads ro tails of anything involving pillars?
22:53 murrdoc hahahaha
22:53 murrdoc mans got jokes
22:53 murrdoc needs to write better commit messages
22:53 robawt if you follow the tutorial it'll show a very easy use of pillar, i'd start there
22:53 murrdoc but hes got jokes
22:53 Vynce murrdoc: what's wrong with my commit message?
22:53 hal58th_1 damn, vicious.ninja is already taken
22:53 murrdoc nothing
22:53 * murrdoc slowly backs awya
22:54 Vynce robawt: what tutorial?
22:54 robawt the official salt tutorial on the doc page
22:56 robawt Vynce: http://docs.saltstack.com/en/latest/#getting-started
22:56 Vynce so, as far as i can tell, this would involve looking up the correct information from teh minion for each minion to put it into the pillar
22:57 Vynce so, i can answer that it is inferior to a relative path for that reason.
22:57 robawt sounds like your problem is not specific to config management, but is in fact a bigger problem
22:58 Vynce mmm… well, system administration is a big problem, yes. ( :
22:58 iggy "I know how to do it with some other tool, and I'm unwilling to learn the best way to do it in Salt"
22:59 Vynce no, more "there's more than one way to do things; is there a reason this way isn't supported"
22:59 iggy (note: I know nothing about the problem, that's just what the last 10 lines paraphrased to me)
23:00 bfoxwell joined #salt
23:01 Singularo joined #salt
23:01 robawt Vynce: perhaps you can tell us what kind of files or relation of files you're trying to copy or enforce on the system, we could be using the wrong module
23:01 Vynce also, not unwilling — i read pillar walkthrough twice and learned nothing from it.  also also, the data would have to be put into pillar, which means duplication, which means it has to be managed or potentially get out of sync. relative paths would JFW.
23:01 catpig joined #salt
23:03 iggy cmd.run cp foo bar
23:03 iggy bam
23:03 Vynce also also also, often the questions i ask are to learn what the tool can do, not to try to solve a particular problem; "can i use relative paths?" was one such. I believe i can rely on the absolute path to this file being the same on all the machines.
23:03 mpanetta joined #salt
23:03 robawt cmd.run is a hammer that makes everything else a nail
23:03 Vynce iggy: that had occurred to me. ( :
23:04 murrdoc cmd.run is evil
23:04 murrdoc also broke
23:04 murrdoc but largely evil
23:06 mosen joined #salt
23:08 arapaho joined #salt
23:08 tedski is the docs being borked a known issue?
23:08 tedski http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html
23:08 iggy yes
23:08 tedski any etr?
23:08 notbmatt the link from http://docs.saltstack.com/en/latest/ref/states/all/ appears to go... to the wrong place
23:09 TheoSLC joined #salt
23:09 iggy no, the devs are probably all gone for the day
23:09 amcorreia joined #salt
23:11 Vynce 31 minutes ago the bug report got a "working on it right now"
23:11 cb joined #salt
23:11 krobin joined #salt
23:11 hal58th joined #salt
23:11 Vynce meanwhile, workaround is o read it on readthedocs: https://salt.readthedocs.org/en/v0.17.4/ref/states/all/salt.states.file.html (h/t/ racooper)
23:11 mfournier joined #salt
23:12 mosen ah crap
23:12 Vynce (or another version there, but not latest)
23:13 druonysuse joined #salt
23:13 druonysuse joined #salt
23:15 ipmb joined #salt
23:16 otter768 joined #salt
23:16 kaiyou joined #salt
23:17 hal58th joined #salt
23:18 Daemonik joined #salt
23:20 keimlink joined #salt
23:23 bhosmer_ joined #salt
23:26 baweaver joined #salt
23:31 donmichelangelo joined #salt
23:31 monkey661 joined #salt
23:31 bhosmer__ joined #salt
23:32 phx__ joined #salt
23:32 clintberry joined #salt
23:32 johnkeates joined #salt
23:33 bash124512 joined #salt
23:33 patrek_ joined #salt
23:34 buMPnet_ joined #salt
23:34 yomilk joined #salt
23:34 Ryan_Lan1 joined #salt
23:34 litwol_ joined #salt
23:34 ShadowHn1r joined #salt
23:34 tmmt joined #salt
23:35 lytchi_ joined #salt
23:35 Hipikat_ joined #salt
23:35 wolog_ joined #salt
23:36 gerl1ng_ joined #salt
23:36 eichiro_ joined #salt
23:36 Aidan__ joined #salt
23:37 mr-op5 joined #salt
23:38 egil_ joined #salt
23:38 awerner_ joined #salt
23:38 seblu42 joined #salt
23:38 |borat| joined #salt
23:38 signull_ joined #salt
23:38 N-Mi__ joined #salt
23:38 djinni`_ joined #salt
23:38 Nazca__ joined #salt
23:38 CeBe joined #salt
23:38 emid joined #salt
23:38 peters-tx joined #salt
23:39 ecdhe_ joined #salt
23:39 harkx joined #salt
23:39 Ouzo_12 joined #salt
23:39 sjohnsen- joined #salt
23:39 SeeDickCode joined #salt
23:39 iggy bahaha... found another blocker
23:39 troyready joined #salt
23:39 murrdoc what meow
23:39 hasues joined #salt
23:39 jesusaurus joined #salt
23:40 twiedenbein joined #salt
23:40 runnyspot joined #salt
23:40 runnyspot joined #salt
23:41 kaictl joined #salt
23:43 cwright joined #salt
23:43 pahko joined #salt
23:44 sarlalian joined #salt
23:45 Ryan_Lane joined #salt
23:45 sarlalian joined #salt
23:47 iggy https://github.com/saltstack/salt/issues/22725
23:47 iggy basepi: https://github.com/saltstack/salt/issues/22725 appreciate the appropriate tags, etc.
23:48 iggy now I feel bad having summoned him twice in a day
23:48 murrdoc [INFO    ] Setting up the Salt Minion "None"
23:48 murrdoc damn
23:48 murrdoc not using gethostname
23:48 basepi Ya, Mike broke it.
23:48 murrdoc thats some nkotb shit
23:48 basepi I have a pull request in to fix it.
23:48 murrdoc thank u based pi
23:49 basepi For now just define id, and you'll be good.
23:49 iggy <3 basepi
23:49 iggy sadly, this is on some automated stuff spinning up about 20 nodes at a time... I'm not logging into them all to set an id
23:50 * iggy packs it in for the day
23:50 basepi Yea sorry. It cause me a fair amount of grief today as well. Anyway, I'm going home.
23:50 iggy nn
23:51 hasues left #salt
23:51 hal58th Nice catch iggy
23:52 iggy yeah... wasted half of my day on that... kept thinking I was doing something wrong
23:52 iggy need to start having more faith in myself (and less faith in others ;)
23:53 hal58th I remember when I found the double minion key bug a long time ago. That was so frustrating to nail down why
23:59 zircote joined #salt

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