Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-04-14

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

All times shown according to UTC.

Time Nick Message
00:01 lytchi joined #salt
00:02 KyleG joined #salt
00:02 KyleG joined #salt
00:03 otter768 joined #salt
00:04 Guest89_ joined #salt
00:07 lytchi joined #salt
00:12 lytchi joined #salt
00:21 stooj joined #salt
00:22 bhosmer joined #salt
00:24 hasues joined #salt
00:24 hasues left #salt
00:26 dendazen joined #salt
00:29 MatthewsFace joined #salt
00:31 yomilk joined #salt
00:31 raygunsix joined #salt
00:34 baweaver joined #salt
00:36 heksi joined #salt
00:40 p66kumar joined #salt
00:41 evilrob joined #salt
00:44 notnotpeter joined #salt
00:46 baweaver joined #salt
00:46 Ssquidly joined #salt
00:46 Tyrm joined #salt
00:55 hal58th joined #salt
00:56 BretFisher joined #salt
00:59 boubou joined #salt
01:00 Singularo joined #salt
01:04 dalexander joined #salt
01:05 catpig joined #salt
01:08 beneggett joined #salt
01:14 p66kumar joined #salt
01:14 yomilk joined #salt
01:17 yomilk joined #salt
01:19 baweaver joined #salt
01:19 bhosmer joined #salt
01:22 TyrfingMjolnir joined #salt
01:25 MatthewsFace joined #salt
01:31 baweaver joined #salt
01:32 ckao joined #salt
01:46 crazysim_ joined #salt
01:51 scbunn joined #salt
01:57 bhosmer_ joined #salt
01:59 JlRd joined #salt
02:01 fllr joined #salt
02:03 otter768 joined #salt
02:14 dl_nz joined #salt
02:16 stooj joined #salt
02:20 baweaver joined #salt
02:22 vauxia joined #salt
02:24 moos3 joined #salt
02:24 SeeDickCode joined #salt
02:25 nene joined #salt
02:27 evle joined #salt
02:30 favadi joined #salt
02:34 BretFisher anyone know of an easy way to search saltstack-formulas (besides googling 'topic site:https://github.com/saltstack-formulas')?
02:37 litwol BretFisher: 'salt formula [what you want]'
02:37 litwol google is smart.
02:37 BretFisher :) :(
02:38 BretFisher guess was hoping for "npm of salt" with likes/downloads, etc
02:38 donmichelangelo joined #salt
02:40 iggy nope... salt-starters was shaping up to be that way, but it's poorly maintained
02:43 iggy some others have talked about it... but that's not really what salt-formulas is all about
02:44 iggy it's more "almost ready pieces" rather than plug-and-play stuff
02:45 BretFisher iggy I see your point, thanks.
02:46 iggy more states/modules/etc can be added to do what's necessary
02:47 iggy personally a lot of the formulas are hugely over-engineered for yaml+jinja
02:47 iggy we'll see where things are in a year I guess (or 8 if we want to catch up to puppet/chef)
02:48 luke joined #salt
02:48 sirtaj_ joined #salt
02:50 writtenoff joined #salt
02:54 BretFisher Yea I looked at the apache one but just pulled out a piece or two and made own.
02:55 iggy yeah, some of the web ones are particularly tough to do
02:56 iggy nginx is pretty messy too
02:56 iggy not a fan of the zookeeper one either
02:56 iggy (or any of them that have 200+ line long settings.sl)
02:58 BretFisher thanks for headsup on nginx, that's on my shortlist
02:58 BretFisher haproxy is another
02:58 iggy haven't looked at that one
02:59 BretFisher mongodb, redis, and elasticsearch should be easy i hope
02:59 iggy I'm working on the mongo one
02:59 iggy so if you see anything missing (like actual working replication support that I have locally)
02:59 BretFisher most of my issues right now are issues on SmartOS. meh: https://github.com/saltstack/salt-bootstrap/issues/473
02:59 iggy oh...
03:00 iggy nah, jk
03:00 BretFisher but finding more that i may have to put on that list
03:00 iggy I'm sure the devs would love to have more active users on !linux
03:00 BretFisher oh, so your improving the mongo one?
03:00 iggy yeah, like I said, I have working replication support locally
03:01 iggy I'm adding support for other distros and updating things right now
03:01 BretFisher yea we'll see. Just using Joyent/SmartOS for this project, and it's my first dive into Solaris in a decade, so got a ways to go before I'm throwin' down PR's :)
03:02 BretFisher yea for sure, I'm a "MongoDB Guy" I suppose for last 3 years, and Salt will be going on 4+ replica sets, 3 nodes each, so I'll check it out
03:02 davroman1ak joined #salt
03:02 fusionx86 joined #salt
03:02 iggy oh, nice, that'll be a good test, we're only looking at one replica set to start off
03:04 BretFisher yea no sharding, just running 3-node replica's in various networks
03:04 iggy ahh... so nothing too crazy
03:05 iggy we'll probably be replicating+sharding at some point in the next year, but I'm not going to worry about that just yet
03:07 jalaziz joined #salt
03:09 aw110f joined #salt
03:12 numkem joined #salt
03:13 yomilk joined #salt
03:13 sirtaj joined #salt
03:13 buMPnet joined #salt
03:13 kutenai joined #salt
03:13 joehh1 joined #salt
03:13 modafinil joined #salt
03:13 LinuxHorn joined #salt
03:13 illern joined #salt
03:13 ek6 joined #salt
03:13 synical joined #salt
03:13 abele joined #salt
03:13 babilen joined #salt
03:13 arif-ali joined #salt
03:13 leszq joined #salt
03:13 pcdummy joined #salt
03:13 paolo joined #salt
03:13 ThomasJ joined #salt
03:13 forze joined #salt
03:13 mackstick joined #salt
03:13 Damianz joined #salt
03:13 mortis_ joined #salt
03:13 beauby joined #salt
03:13 clintberry joined #salt
03:14 buMPnet joined #salt
03:32 otter768 joined #salt
03:33 evilrob joined #salt
03:38 p66kumar joined #salt
03:40 srufle joined #salt
03:41 Furao joined #salt
03:50 raygunsix joined #salt
03:52 enarciso_ joined #salt
03:52 beneggett joined #salt
03:52 aw110f joined #salt
03:54 monkey66 left #salt
03:59 armguy joined #salt
04:12 armguy joined #salt
04:15 bhosmer_ joined #salt
04:20 evilrob joined #salt
04:24 TyrfingMjolnir joined #salt
04:35 enarciso joined #salt
04:37 nene hi all.. i have setup a test master and minion..and wrote servers.sls file to install mysql and a test file  and  then ran salt '*' state.highstate... it is showing nothing on my screen
04:38 iggy did you accept the key?
04:38 iggy salt '*' test.ping
04:38 nene iggy: yes.. test.ping is working fine
04:39 iggy then your top file isn't matching
04:40 nene iggy: in top file i've writen base: '*' -servers
04:40 nene with appropriate spaces
04:40 BretFisher nene: I think you could salt 'servername' state.apply servers
04:40 BretFisher to test the servers.sls separate from top file matching
04:40 iggy state.apply is 2015.2 only
04:40 iggy state.sls
04:41 BretFisher but iggy is right
04:41 BretFisher ah, really, cool, aren't I lucky :)
04:43 nene i ran salt 'servername' state.apply servers.. it showing "state.apply" is not available.'}
04:44 nene here servername and servers are same?... i have given my minion name at both places
04:45 eliasp nene: "salt servername state.sls servers"
04:45 eliasp nene: "state.apply" is only available in 2015.2 and develop
04:46 eliasp ah… what iggy said ^^ :)
04:46 nene eliasp: ok.. servername and servers are my minion name?
04:47 eliasp nene: servername = minion name
04:47 eliasp nene: servers = a statefile's (SLS) name
04:47 eliasp nene: but usually one just applies a highstate which then evaluates based on the top.sls assignments the states to apply to the minion
04:48 eliasp nene: so usually it's "salt your-minion state.highstate"
04:50 nene i ran "salt my-minion state.highstate"... it is returning nothing other than my minion name in green :(
04:54 nene eliasp: iggy: BretFisher^^
04:54 eliasp this usually means you didn't apply any SLS to your minion in your top.sls
04:55 eliasp s/apply/assign/g
04:55 BretFisher nene: got a gist of  your top.sls
04:55 BretFisher ?
04:55 iggy yeah, we've been though this already
04:56 iggy paste everything you've got to gist (protip: gist allows multiple file in a single paste)
04:56 nene base:
04:56 nene 'my-minionname'
04:56 nene - servers
04:56 nene this is my top.sls ^^
04:57 eliasp syntax error
04:57 iggy : after 'my-minionname'
04:57 stanchan joined #salt
04:57 iggy so...how about you paste what we asked for
04:57 iggy where we asked for it
04:57 iggy (i.e. not in channel)
04:58 nene iggy: ok
04:58 nene what syntax error?
04:58 iggy what you pasted was not valid yaml
04:59 iggy so it would have errored out
04:59 iggy (as a syntax error)
05:01 nene iggy: http://pastebin.com/9FmcyJFc
05:01 nene eliasp: ^^
05:02 iggy you need a : after 'my-minion'
05:02 iggy it should have given you a syntax error
05:03 nene it not giving any syntax error... any way to debug this?
05:03 iggy you need a : after 'my-minion'
05:03 iggy debugged
05:04 eliasp iggy: you might want to define a shortcut for "you need a : after 'my-minion'" :)
05:04 eliasp iggy: or just have a bot repeat it every 60s
05:04 iggy I definitely need one for "paste everything you have to GIST.GITHUB.COM"
05:05 eliasp hehe
05:05 iggy not pastebin.com
05:05 nene no luck after putting : after minion name
05:05 nene should  i restart any services?
05:05 iggy now paste what you have
05:05 iggy and the output of "salt-call -l debug state.highstate"
05:06 * eliasp better switches the channel now… otherwise my passive aggressiveness turns into something else ;)
05:06 iggy run on the minion
05:06 iggy remove the -require line and the one after it
05:07 iggy because it can't work
05:07 nene you mean in my servers.sls file?
05:07 keimlink joined #salt
05:08 iggy sure... I closed the tab already
05:09 nene iggy: i should replace  -require?
05:09 nene in my servers.sls
05:10 iggy just remove it
05:10 iggy and the line after
05:10 iggy that says to require a state that you don't have
05:13 nene iggy: working after removing the -reqire and pkg line...but i want to create that file after installing the mysql pkg
05:14 iggy you need a state that installs mysql then (and then require that)
05:18 iggy and it needs to be in the same file or include'd
05:18 * iggy sleep
05:18 nene iggy: thankyou....working fine.. gud ni8
05:21 Guest70 joined #salt
05:25 rdas joined #salt
05:28 yggdrasi1 joined #salt
05:30 I3olle joined #salt
05:33 p66kumar joined #salt
05:38 catpigger joined #salt
05:48 hillna joined #salt
05:54 malinoff joined #salt
05:58 colttt joined #salt
06:04 otter768 joined #salt
06:07 speedlight joined #salt
06:11 robothands joined #salt
06:11 seev joined #salt
06:12 eseyman joined #salt
06:13 freelock joined #salt
06:13 Tahm joined #salt
06:14 oxae joined #salt
06:19 monkey66 joined #salt
06:20 harkx joined #salt
06:20 AndreasLutro joined #salt
06:23 dopesong joined #salt
06:27 dopesong_ joined #salt
06:31 mosen joined #salt
06:32 zz_cro joined #salt
06:34 bhosmer joined #salt
06:36 clintberry joined #salt
06:36 lietu joined #salt
06:38 KermitTheFragger joined #salt
06:46 flyboy joined #salt
06:48 soren joined #salt
06:51 dRiN joined #salt
06:55 hebz0rl joined #salt
06:57 froztbyte "A more involved example involves pulling from a custom repository. Note that the pkgrepo has a require_in clause. This is necessary and can not be replaced by a require clause in the pkg."
06:58 froztbyte what's that about?
06:58 froztbyte does it do package pinning or something
06:58 froztbyte ?*
07:04 crashmag joined #salt
07:04 faust joined #salt
07:05 Romlok joined #salt
07:11 cheus joined #salt
07:12 davromaniak joined #salt
07:12 zz_ashmckenzie joined #salt
07:13 robinsmidsrod joined #salt
07:13 krelo joined #salt
07:15 kawa2014 joined #salt
07:21 bhosmer joined #salt
07:22 froztbyte (doing a require_in for every package one might need from, say, dotdeb seems like an antipattern)
07:22 ITChap joined #salt
07:24 apchinku joined #salt
07:33 babilen froztbyte: Where is that from? I was always under the impression that require and require_in are symmetric
07:34 a7p joined #salt
07:36 babilen Ah, found it in http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkg.html -- that is weird to say the least
07:39 babilen That was added in fc8d90f3c, I wonder if that really is true
07:40 Romlok hmm, I think I read somewhere else that you needed to do it that way to make sure the repo gets updated before trying to install anything
07:40 gchao joined #salt
07:40 Romlok yeah, http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkgrepo.html
07:40 babilen require/require_in should simply change the state oder
07:41 Romlok right at the end
07:41 fusionx86 joined #salt
07:41 babilen That's just bonkers
07:41 froztbyte babilen: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkg.html
07:41 froztbyte ah, you found it
07:41 froztbyte well, it's a semi viable point
07:42 babilen Why not set refresh_db if one wants that?
07:42 froztbyte yeah, something like that
07:42 babilen (I also don't see the point of running "apt-get update" n times when I install n packages, let alone that is changes the semantic of require_in)
07:43 yomilk joined #salt
07:43 froztbyte getting that kind of state ordering nice in puppet was also a bit hairy
07:43 ITChap joined #salt
07:44 froztbyte so, on one hand, I'd rather just get my things done as quick as possible, but on the other I'd really rather not have that kind of boilerplating in my states
07:44 babilen Setting 'refresh_db' and using either require or require_in for the state ordering should do the same or am I missing something?
07:44 froztbyte not sure, haven't dealt with that yet :D
07:45 froztbyte will be doing so in the next couple of minutes
07:45 babilen Please report on your findings. I am tempted to open a bug for this.
07:45 chiui joined #salt
07:46 jhauser joined #salt
07:48 yggdrasi1 joined #salt
07:54 CeBe joined #salt
07:55 MagiCrazy joined #salt
07:56 a7p left #salt
08:00 k_sze[work] joined #salt
08:02 k_sze[work] What does SLS stand for?
08:02 hebz0rl joined #salt
08:03 AndreasLutro SaLt State
08:05 otter768 joined #salt
08:05 ITChap joined #salt
08:06 jhauser joined #salt
08:06 yomilk joined #salt
08:06 gchao joined #salt
08:07 k_sze[work] And it's not really explained in the glossary.
08:07 lb1a joined #salt
08:07 the_lalelu uh, how unexpected
08:07 k_sze[work] http://docs.saltstack.com/en/latest/glossary.html has "SLS Module", but it points to "state declarations". It would be helpful to just spell out that SLS stands for SaLt State so newbies can stop wondering.
08:08 malinoff http://docs.saltstack.com/en/latest/topics/tutorials/starting_states.html#how-do-i-use-salt-states the second paragraph
08:08 AndreasLutro shrug
08:09 k_sze[work] malinoff: well, the first time it's encountered among the tutorials is actually in the walkthrough.
08:09 k_sze[work] http://docs.saltstack.com/en/latest/topics/tutorials/walkthrough.html
08:10 k_sze[work] (that is, if someone is coming from the salt documentation front page.
08:11 k_sze[work] It's the first link under the "Getting Started" section.
08:12 AndreasLutro the salt documentation has general structural issues, I personally wouldn't single out the meaning of a file extension as the missing piece :p
08:14 MagiCrazy hi, I'm looking for some kind of usage examples of saltrc or .config/salt files.
08:14 monkey66 joined #salt
08:15 stoogenmeyer joined #salt
08:15 stoogenmeyer joined #salt
08:17 keimlink joined #salt
08:18 MagiCraz1 joined #salt
08:20 Romlok there are saltrc or .config/salt files?
08:20 keimlink joined #salt
08:20 ujjain joined #salt
08:23 MagiCrazy joined #salt
08:25 MagiCrazy https://github.com/saltstack/salt/blob/6ae9fe6dff67e26b6553317ed9484b2f8cbfb3b7/salt/config.py#L2571
08:25 MagiCrazy I think so :)
08:26 ingwaem joined #salt
08:26 pf_moore joined #salt
08:27 AndreasLutro :o
08:28 AndreasLutro it looks like just an extra master config location
08:28 Romlok ah, okay
08:28 Romlok yeah, what AndreasLutro said
08:33 ggoZ joined #salt
08:34 jamesog joined #salt
08:35 MagiCraz1 joined #salt
08:38 zer0def quick question: in salt-ssh's rosters, can i attach a grain to a particular target?
08:38 zer0def i'm either dumb, or the docu is rather vague about it
08:38 wnkz joined #salt
08:39 MagiCraz1 Romlok & AndreasLutro: I'm looking for a way to have returners always set without retyping them in my shell...
08:39 ITChap joined #salt
08:45 dopesong joined #salt
08:45 Grokzen joined #salt
08:48 wvds-nl joined #salt
08:49 kawa2014 joined #salt
08:49 ktosiek joined #salt
08:49 stoogenmeyer Hi, can you call on other grains from inside a py file defining another grain ? it doesn't seem to be working for me
08:50 ujjain joined #salt
08:51 bhosmer_ joined #salt
08:52 JayFK joined #salt
08:59 jrluis joined #salt
09:01 denys joined #salt
09:01 froztbyte babilen: so, aside from my vagrant machine being ubuntu (and thus dependencies are exploding), salt seems happy with https://gist.github.com/froztbyte/867b61657a9490b91440
09:01 froztbyte babilen: I feel like I shouldn't be writing the file key, though. that should probably be abstracted away
09:02 froztbyte I plan to test that against an EC2 instance in a couple of minutes
09:04 bluenemo joined #salt
09:04 bluenemo joined #salt
09:04 Anton joined #salt
09:04 Anton Hallow
09:05 Guest96116 Hey, is some1 there&
09:05 babilen froztbyte: I'd still include a require in the packages state, but that is what I expected to work and suffice
09:06 froztbyte err, yeah, I do have that as a mental checklist
09:06 froztbyte just so that ordering is definitely fixed
09:06 froztbyte just wanted to see if it exploded at me
09:06 Guest96116 Can some one help me with orchestrate runner question?
09:06 froztbyte babilen: should I do a PR to clean up the examples on pkg/pkgrepo?
09:06 froztbyte Guest96116: don't ask to ask, just ask
09:06 Guest96116 =)
09:06 Guest96116 https://groups.google.com/forum/#!topic/salt-users/hXjiIF8TvTw
09:07 antani joined #salt
09:07 Guest96116 i asked there and them understood that it ll be faster to go in irc
09:08 antani joined #salt
09:09 ITChap1 joined #salt
09:13 stoogenmeyer Hi, can you call on other grains from inside a py file defining another grain ? it doesn't seem to be working for me
09:15 viq_ joined #salt
09:15 k_sze[work] This is causing me some confusions: http://docs.saltstack.com/en/latest/topics/tutorials/starting_states.html#getting-to-know-the-default-yaml-jinja
09:15 is_null joined #salt
09:15 k_sze[work] Why doesn't the pkg: apache under "watch" need to be surrounded by {% if grains['os'] == 'RedHat' %}?
09:16 k_sze[work] (or != 'RedHat', actually)
09:16 k_sze[work] Since the package name is different under Red hat.
09:18 glyf joined #salt
09:20 babilen froztbyte: I guess that would be a great way to start dicussions about this
09:23 babilen k_sze[work]: I take it that you answered your own question. Is that correct?
09:23 ujjain joined #salt
09:23 k_sze[work] babilen:
09:23 k_sze[work] no
09:25 k_sze[work] In fact, why is pkg.installed initially an "empty list"?
09:26 babilen k_sze[work]: Okay. References in requisites adhere to the form "state_module:[state_id|state_name]" which means that if you define both name and id for a state and they differ you can use both for references. I would, however, recommend to always use the state id for this
09:27 babilen k_sze[work]: "apache: pkg.installed: []" means that pkg.installed uses the state id (i.e. "apache") as name argument. It is essentially a shorthand for "apache: pkg.installed: - name: apache"
09:28 k_sze[work] I see.
09:29 ITChap joined #salt
09:31 k_sze[work] But the ID could actually be whatever I want to name it, not corresponding to any package name, right? It could even be foobar_blabla, in which case I would always need to specify the -name in pkg.installed, etc., right?
09:31 mdupont joined #salt
09:31 wnkz_ joined #salt
09:32 stoogenmeyer would this work in grain file ? http://pastebin.com/6BhzhBsr
09:33 test12 joined #salt
09:33 test12 left #salt
09:34 babilen k_sze[work]: That is correct. In fact that is my preferred style while I do ensure that I have no spaces in the id. I typically use something along the lines of foo_pkg: foo_service: foo_conf: ...
09:34 k_sze[work] babilen: because explicit is better than implicit? (when you use a package name as the ID, there's a lot of implicit magic around names)
09:35 babilen stoogenmeyer: That should work, but __grains__ are monkey patched into the module IIRC
09:36 MagiCrazy joined #salt
09:37 stoogenmeyer so I don't have to create it from scratch ? I can just call __grains__ magically ?
09:39 stoogenmeyer So if you wanted to get a grain from inside a custom grain you would just use __grains__ ?
09:39 BtbN left #salt
09:39 stoogenmeyer Right now I'm getting an error RuntimeError: maximum recursion depth exceeded when trying that grains
09:40 glyf joined #salt
09:42 matthew-parlette joined #salt
09:44 ITChap joined #salt
09:49 kawa2014 joined #salt
09:52 rubenb_ joined #salt
09:52 gdm85 joined #salt
09:53 ja-s_ joined #salt
09:54 Heartsbane_ joined #salt
09:54 wolog_ joined #salt
09:54 litwol_ joined #salt
09:54 MrMarv_ joined #salt
09:54 Ryan_Lan1 joined #salt
09:54 canci_ joined #salt
09:54 jamesog_ joined #salt
09:54 jerrcs_ joined #salt
09:55 Guest96116 guys, seems no one want's to answer on ggroups.. can u tell me why orchestrate runner usage is restricted with either base either explicit env specification?
09:57 patrek_ joined #salt
09:58 MagiCraz1 joined #salt
09:59 froztbyte babilen: confirmed, salt is happy with that
09:59 froztbyte Guest96116: I haven't yet gotten to play with those parts, maybe stick around for the US timezone to wake up
09:59 froztbyte or late euro afternoon folks
10:00 imanc_ what does a tilde do in pillars.get?  Cant' seem to find any documentation for it
10:02 mr-op5 joined #salt
10:02 AndreasLutro imanc_: example?
10:02 imanc_ AndreasLutro: {% if salt['pillar.get']('apps:' ~ my_role ~ ':cron_highstate:'
10:02 AndreasLutro ah that's not salt
10:02 AndreasLutro that's jinja
10:03 AndreasLutro it does string concatenation
10:03 imanc_ presumably it's just used to concat vars and strings?
10:03 AndreasLutro you can use + as well
10:03 AndreasLutro but if my_role is an int
10:03 AndreasLutro it will error with +, but not with ~
10:04 nadley joined #salt
10:05 Guest96116 \@BaseApi
10:05 imanc_ AndreasLutro: awesome, thanks
10:06 otter768 joined #salt
10:06 froztbyte babilen: E_BRAIN_TIME_ORDERING; I've added a note in my booklet, will do some poking at the PR a bit later
10:07 stoogenmeyer how can I have a custom grain that's based on other grains?
10:07 stoogenmeyer I can't seem to get to the grains from inside a custom grain
10:08 dean joined #salt
10:08 stoogenmeyer salt.modules.grains.get(GRAIN_NAME) always returns ''
10:08 MagiCrazy joined #salt
10:08 yomilk joined #salt
10:09 Guest96116 froztbyte @BaseApi whois tells he is from RO.. so he should be somewhere around i think
10:10 Guest96116 xD
10:10 MagiCraz2 joined #salt
10:19 ingwaem joined #salt
10:20 faust joined #salt
10:24 bhosmer joined #salt
10:27 MagiCraz2 joined #salt
10:28 stoogenmeyer Can someone verify that it's not possible to get a grain from inside a custom grain file?
10:33 giantlock joined #salt
10:33 bhosmer_ joined #salt
10:34 kwork joined #salt
10:36 keimlink_ joined #salt
10:39 quist` left #salt
10:41 ShadowHntr joined #salt
10:48 aquassaut joined #salt
10:50 Hell_Fire joined #salt
10:52 ITChap joined #salt
10:55 OnTheRock joined #salt
10:58 bmcorser_ joined #salt
11:01 bmcorser ah got my name back
11:03 _mel_ joined #salt
11:03 ITChap1 joined #salt
11:06 mindscratch joined #salt
11:08 mindscratch Is it possible to target minions on the command-line using grain pcre?
11:08 glyf joined #salt
11:08 Guest96116 yep
11:09 mindscratch I have used: salt -G "roles:web" .....   but I need to use pcre to do something like: salt <flag> "roles:(web|db)"
11:09 Guest96116 see compound targeting
11:09 mindscratch is it still just "-G"
11:09 mindscratch ah, got it: http://docs.saltstack.com/en/latest/topics/targeting/compound.html
11:09 mindscratch thanks Guest96116
11:10 Guest96116 np, just use -C
11:11 mindscratch looks like -P might work too
11:12 mindscratch nope, -C it was:  salt -C 'G@roles:web or G@roles:db'
11:13 Guest96116 yes, -C is just like the general solution for all kind of targeting
11:13 patrek joined #salt
11:13 Guest96116 u can use -C " P@..." i think
11:16 echo joined #salt
11:17 mindscratch another question, if I execute a state from the CLI: salt -C "..." state.sls myservice.start  ...if it fails, how can I tell? For example, I had myservice/start.sls just do a "cmd.run" that did "exit 5"....when I executed taht state it failed, but from the CLI i can't tell other than looking at the output
11:17 mindscratch i'd like to see "echo $?" return something other than 0
11:17 mrchris joined #salt
11:18 mrchris Hi. I want to use salt on certain machines, but each machine has a different primary user (username).  Some of the states need to know about the primary user.  Where's the best place to store this information in salt?
11:22 bhosmer joined #salt
11:25 gdm85 joined #salt
11:27 mrchris * solved
11:27 mrchris Not solved!
11:28 ndrei joined #salt
11:29 mage_ mrchris: in the pillar ?
11:31 mrchris mage_: Aren't pillars generally used for using data with batches of machines rather than targeting specific machines?
11:31 mrchris Each machine will have a different primary username
11:32 AirOnSkin I have the following Pillar data: http://hastebin.com/ohevicipik
11:33 Guest96116 mrchris nope, pillars a especially used for targeting specific minions
11:33 AirOnSkin I know how to loop through this data with iteritems to write a single vhost,conf file for each vhost
11:34 mrchris But I won't know in advance which machine has which user on.
11:34 AirOnSkin What I don't know is, how to access a single value in the file. The domain for example. In the first file, I'd like to do something like: ServerName {{ salt['pillar.get'](virtual_host:vhost_1:domain) }}
11:34 mrchris It somehow needs to intelligently work out what the primary user is
11:34 Guest96116 how this can be?
11:35 bhosmer_ joined #salt
11:35 AirOnSkin But the vhost_1 part is hardcoded here... how can I access such a value dynamically?
11:35 mrchris The users can use their laptop at different physical locations,  so their IP address will change
11:36 Guest96116 u can use hostnames
11:40 giannello joined #salt
11:40 mrchris maybe a grain with their username in is the answer.
11:43 Guest96116 u should understand one thing about grains: it was designed to be a static data, strictly it is bound to the start of the minion
11:43 Morbus joined #salt
11:43 yomilk joined #salt
11:44 Guest96116 u can update it on running minion using states or modules from master, but it can be not updated when needed yet or cached
11:44 babilen Well, not quite, but the data are expected to be relatively static
11:45 Guest96116 yes, i mean, u can't update it and be sure that it is updated without restart of minion
11:56 refnode joined #salt
11:56 MaZ- joined #salt
11:57 programmerq joined #salt
11:58 froztbyte mmmmmmmm
11:59 froztbyte babilen: what if I only wanted to reload package indexes every 2h or so?
12:00 favadi left #salt
12:02 echo joined #salt
12:02 babilen froztbyte: I wouldn't include the reload in the pkgrepo definition at all. If new packages are isntalled they are refreshed anyway and updates is something I prefer to do manually
12:02 babilen s/reload/refresh
12:07 otter768 joined #salt
12:09 stoogenmeyer joined #salt
12:13 bhosmer joined #salt
12:13 agend joined #salt
12:19 denys joined #salt
12:21 Furao joined #salt
12:26 is_null hi all, what's the standard indentation caracters ? it's not consistent accross docs ie. sometimes 4 spaces sometimes 2
12:26 is_null http://docs.saltstack.com/en/latest/ref/states/top.html#environments 4 spaces
12:26 is_null http://docs.saltstack.com/en/latest/topics/tutorials/standalone_minion.html 2 spaces
12:27 cmcmacken joined #salt
12:27 is_null makes my eyes bleed and probably kills a kitten xD
12:28 denys joined #salt
12:30 zipkid Hi, does anyone know where i could find a graphical overview of a salt(-call) process, what happens when/where ? grains, pillar, reactor, compile, apply....?
12:30 Romlok is_null: AFAICT 2 characters seems to be what the docs *almost* always use
12:34 is_null Romlok: unlike python, good to know, thanks
12:34 is_null or did you mean i should *almost* always use 2 spaces ? jk xD
12:35 AndreasLutro yeah yaml is 2 spaces
12:36 stoogenmeyer_ joined #salt
12:36 Romlok is_null: actually yes, you should *almost* always use 2 spaces
12:36 Romlok as I found out yesterday, multi-line strings need 4 spaces...
12:37 jdesilet joined #salt
12:37 Romlok but IIRC YAML doesn't actually care how many spaces, as long as you're consistent within a block
12:38 Romlok YAML parser implementations notwithstanding
12:40 is_null yep, so that's more like ansible then, makes sense
12:41 timoguin joined #salt
12:50 dyasny joined #salt
12:54 yggdrasi1 joined #salt
12:54 fredvd joined #salt
12:56 TyrfingMjolnir joined #salt
12:57 jdesilet joined #salt
12:59 drico joined #salt
13:00 JDiPierro joined #salt
13:00 drico Hi, i'm using the apache formula, with configuration in the pillar.
13:00 drico If I suppress the definition of a website inside the pillar, it doesn't get removed on the minion
13:01 drico any idea on this removal, is there a standard way to remove those configuration file done by salt between states ?
13:02 AndreasLutro drico: I always put an "enable: true" on all of my pillar stuff, that way I can switch it to false to make salt remove it, then remove it from the pillar later
13:02 subsignal joined #salt
13:02 circ-user-pATR3 joined #salt
13:04 subsigna_ joined #salt
13:05 racooper joined #salt
13:05 tajh joined #salt
13:05 circ-user-pATR3 hello salt masters, I'm very new at this so please be kind :-) . I have a minion on my raspberry pi 2 with deb jessie installed and want my salt master to automagically install kodi on it. Did anyone create a SLS which does this?
13:05 circ-user-pATR3 Very, very cool framework BTW
13:07 kerberos joined #salt
13:08 fusionx86 joined #salt
13:08 bhosmer_ joined #salt
13:08 tajh Hi Folks! I am wondering if anyone can suggest a best practice for dealing with salt-formulas that have hard coded paths in the pillar.example, I have already changed the map.jinja to reflect paths for my different platforms.  The docs hint that it is acceptable to do this for pillars too, but I have not seen that in practice yet.
13:09 ksdncksjnkjn joined #salt
13:10 nadley joined #salt
13:10 racooper circ-user-pATR3,  there's a PPA for Kodi. Just set up the repo then pkg.install kodi should work?
13:10 drico AndreasLutro, thanks !
13:13 slav0nic joined #salt
13:13 slav0nic joined #salt
13:13 is_null hi all, any idea how to get 2014.7 on jessie with the bootstrap script without git ? ie. using the saltstack repos for jessie
13:14 * is_null trying to get rid of our own bootstrapping spagetti boilerplate xD
13:16 wnkz joined #salt
13:17 cpowell joined #salt
13:17 dendazen joined #salt
13:18 randomuser joined #salt
13:25 AndreasLutro is_null: just follow the install instructions, create an /etc/apt/sources.list.d/saltstack and apt-get update + upgrade
13:26 circ-user-pATR3 thank racooper, will try it
13:27 stoogenmeyer_ can you access grains from inside a custom grain file?
13:27 stoogenmeyer_ I'd like to base a few grains on the minion's id
13:27 is_null that's not the install instructions with the bootstrap script apparently but it sounds fair, i mistakenly though jessie was in end of life until a few minutes ago anyway, sorry, real noob here
13:28 Tecnico1931 joined #salt
13:29 iggy stoogenmeyer_: it's tricky, but can be done... you have to be very careful of ordering
13:29 is_null but not using the bootstrap script means re-inventing/pasting the rest of the logic inside the official bootstrap script into my own which is ugly
13:30 hasues joined #salt
13:30 hasues left #salt
13:31 AndreasLutro is_null: don't worry about it, follow the debian instructions. bootstrap scripts are usually just hacky shortcuts
13:32 yomilk joined #salt
13:32 drico AndreasLutro, have you an example of a file removal by putting the enable switch to false please ?
13:32 stoogenmeyer_ iggy: I figured as much, I kept running into trouble with infinite recursion. Can you explain why this happens?
13:33 elfixit joined #salt
13:33 is_null AndreasLutro: i've got reasons to be super worried about it when this is part of a test script that's supposed to run around 40 times a day
13:33 is_null i think we don't have quite the same needs, it doesn't matter, thanks for your advice :)
13:34 drico stoogenmeyer_, I ended up doing http://pastebin.com/1ZKEf8Dy
13:34 glyf joined #salt
13:35 AndreasLutro drico: no but I can make a simple one for you
13:36 drico AndreasLutro, that would be awesome, I just don't get how you make salt rm the /etc/apache2/sites-available/example.com.conf file
13:36 drico I mean on the minion
13:36 AndreasLutro drico: https://bpaste.net/show/4951cd79ea66
13:36 primechuck joined #salt
13:37 AndreasLutro you can apply the same principle to services
13:37 stoogenmeyer_ drico: didn't know about /etc/salt/minion_id, but that doesn't exactly work for me because /etc/salt/minion_id contains the host name and the minion id for some reason. Maybe I need to delete that file and restart the minion ?
13:37 AndreasLutro service.{{ 'running' if enabled else 'dead' }}
13:37 stoogenmeyer_ drico: also, thats not reading from an actual grain
13:38 drico stoogenmeyer_,  that's a custom grain file in fact
13:38 drico thanks AndreasLutro !
13:38 stoogenmeyer_ drico: a custom grain file -> /etc/salt/minion_id ?
13:39 Deevolution Anyone tried using the Salt python API (local client) with 2015.2.0rc1?  I'm trying to work around the change to cmd.run in; i.e. https://github.com/saltstack/salt/issues/22052  I can't seem to get cmd.run in the local client to take the python_shell = True parameter.
13:40 drico stoogenmeyer_,  no that's a file that is in /srv/salt/_grains and that is then deployed on the minion
13:40 drico stoogenmeyer_,  and will be used when you ask the minion for his grains with salt 'vm01' grains.items per example
13:40 Deevolution Oddly enough, if you supply the python_shell kwarg, it runs the command but doesn't take any | output.
13:41 drico stoogenmeyer_,  there is probably some other solution, probably better, but that's the one I found
13:42 stoogenmeyer_ drico: no I understand the file you posted on pastebin is a custom grain file, just saying you didn't actually get the minion id using grain.get
13:42 stoogenmeyer_ drico: thanks though!
13:43 bhosmer joined #salt
13:43 tuor joined #salt
13:44 drico stoogenmeyer_, right, I finally went to creating a custom grain to get the manipulation I wanted on the minion_id string, that was more simple to use in a pillar then
13:44 Furao joined #salt
13:44 Deevolution nevermind, dharper_ii figured out my problem
13:45 davidbanham joined #salt
13:45 stoogenmeyer_ I understand. Sorry for repeating myself but do you see any way to write a custom grain which will itself access the minion id through a call to grains.get ?
13:45 Tyrm joined #salt
13:45 stoogenmeyer_ When I try to do that I get errors about infinite recursion
13:46 stoogenmeyer_ and I don't really understand why
13:46 moskitone joined #salt
13:46 toastedpenguin joined #salt
13:49 jakubm joined #salt
13:49 jakubm hola
13:51 bhosmer joined #salt
13:54 martoss joined #salt
13:54 niseak joined #salt
13:55 dharper_ii joined #salt
13:57 dharper_ii left #salt
13:57 dharper_ii joined #salt
13:58 iggy stoogenmeyer_: use socket.getfqdn() ?
13:59 iggy stoogenmeyer_: oh... a custom grain file (vs custom grain module... i.e. /srv/salt/_grains)? No, you can't do that
13:59 bhosmer_ joined #salt
14:03 stoogenmeyer_ iggy: http://pastebin.com/6p83ec4b, this is pretty much what I was trying to do. but ended up doing this instead http://pastebin.com/AnkWgrue
14:03 andrew_v joined #salt
14:05 perfectsine joined #salt
14:05 Guest96116 is there anyone, who understands orchestrate runner's mechanic?
14:05 scoates joined #salt
14:05 subsignal joined #salt
14:08 otter768 joined #salt
14:09 babilen Guest96116: What would you ask that hypothetical person?
14:09 babilen IRC 101
14:11 Brew joined #salt
14:11 overyander joined #salt
14:11 dabb joined #salt
14:12 Guest96116 i wish to know why in orchestrate runner's code environment is pre-set to base. So that u have to explicit pass saltenv argument or use base. It doesnt go well with the declarative concept of environments in top.sls of highstate. Also i wish to know what orchestrate_high (which is added in fresh release candidate) for..
14:12 Guest96116 j
14:15 n joined #salt
14:16 evle joined #salt
14:18 timoguin Guest96116: it's quite annoying to me too.
14:18 timoguin i want my servers who are tied to specific environments to use those.
14:18 roedie joined #salt
14:18 timoguin It's been more than once I've forgot to specify saltenv and the wrong shit runs
14:19 Guest96116 timoguin is it a bug or smth? i wish it can be fixed, as it ruines the highstate declaretive concept for me..
14:19 roedie Hi all, I'm trying to create a bond interface using salt 2014.7.1 and it keeps giving me 'Invalid option -- Interface: bond0, Option: 'mode' 'active-backup', Expected: [str within ['gre', 'GRE', 'ipip', 'IPIP']]' is this some kind of bug?
14:20 kaptk2 joined #salt
14:20 roedie If I try something bogus it gives me 'Invalid option -- Interface: bond0, Option: mode, Expected: [0|1|2|3|4|5|6|balance-rr|active-backup|balance-xor|broadcast|802.3ad|balance-tlb|balance-alb]'
14:20 roedie Which is what I tried in the first place...
14:20 yml hello
14:20 timoguin Guest96116: I don't think it's technically a bug... but yea
14:21 yml anyone at pycon sprint today ?
14:21 yml I have a sticky issue that I can't quite figure out how to deal with.
14:22 Guest96116 timoguin in highstate code there is like saltenv='none' and in orchestrate saltenv='base'. But i see no point to restrict it: highstate are allowed in orch, declarative env are allowed in highstate, but trasitive logic doesnt work for us there..
14:23 yml I need to be able to store the `git.revision` in a variable once my `git.latest` is done
14:23 Guest96116 babilen So, u cant give me a hint?
14:23 yml this git revision will then be used to in few other states like creating a virtualenv
14:23 mapu joined #salt
14:24 stoogenmeyer__ joined #salt
14:25 yml or updating the few files
14:25 rhodgin joined #salt
14:26 babilen yml: Expose it in a custom grain or via the salt mine?
14:26 yml babilen: I was thinking about the mine
14:27 MTecknology How can I have a null state that's always true?.. hrm...
14:27 yml but it seems silly to set a variable in the mine in an sls file and use it 3 times in the same .sls
14:28 Guest96116 timoguin the problem is not only in explicit saltenv specifiaction, but moreover in regression of base env reusability
14:28 yml babilen: what I want is just a local variable inside the .sls
14:29 yml https://dpaste.de/XnSG
14:30 yml this is an example of places where I would like to get access to the git revision
14:30 nich0s joined #salt
14:30 nich0s joined #salt
14:31 bhosmer joined #salt
14:31 diegows joined #salt
14:31 conan_the_destro joined #salt
14:32 debian112 joined #salt
14:32 ndrei joined #salt
14:32 kusams joined #salt
14:33 Guest96116 hmm..
14:33 stoogenmeyer__ how do I use a formula to install docker on a minion? Can I call the formula like a normal state file, like this: sudo salt -G 'env:staging' state.sls docker, and have this refer to the docker formula? I don't have it cloned locally
14:35 StDiluted joined #salt
14:36 TyrfingMjolnir joined #salt
14:37 tajh left #salt
14:37 antonn joined #salt
14:38 niseak joined #salt
14:39 yml babilen: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.grains.html
14:39 jalbretsen joined #salt
14:39 yml I think grains is what make the more sense then
14:40 yml I can have a grain like git_revision set as a states
14:40 yml and other states grabbing the value from this grains
14:41 * yml is going to give this a try
14:41 GreyGnome Hello. I created a simple module, then I removed the module from /salt/srv/_modules (the entire _modules directory is gone now). Then I did saltutil.sync_modules, but the module is still on the minion.
14:41 GreyGnome I am using last-2014.7.1 on CentOS 7.
14:41 GreyGnome ...oops, I mean salt-2014.7.1
14:42 kawa2014 joined #salt
14:42 jalbretsen1 joined #salt
14:42 GreyGnome I also tried saltutil.refresh_modules but that did not help either.
14:45 free_beard joined #salt
14:48 babilen yml: yeah, that's why I mentioned both (depends on the question if you need this elsewhere too)
14:48 babilen elsewhere == on other minions
14:49 yml babilen: it is a pair minion thing
14:49 David________ joined #salt
14:49 babilen per minion?
14:49 yml so now I am looking into writing a custom string for this
14:49 babilen Or "a pair of minions" ?
14:50 yml babilen: per minion because the git revision is local to a minion at any given time
14:50 yml so as I was looking into writing a custom grain python function
14:50 babilen grains are appropriate then
14:50 David________ joined #salt
14:50 yml I realized that I need to data from the pillar
14:51 yml like what is the path to the git repo
14:51 babilen You don't have access to that
14:51 yml babilen: :-(
14:52 thedodd joined #salt
14:54 yml where the repo is located is in the pillar something like /srv/myproject
14:55 yml I would like to store in the grains or something else the value of the git rev-parse head for this repo
14:55 yml so I can use it in later states
14:55 babilen yes, I understand what you are trying to do.
14:55 yml babilen: I am in a chicken and egg issue :-(
14:56 gladiatr joined #salt
14:57 stoogenmeyer__ I'm trying to use dockerio and to pull a public image as well as private images, and keep getting a 500 error. What's wrong with this docker.sls pillar file http://pastebin.com/ewbV6cs6
14:57 niseak joined #salt
14:58 babilen yml: Your basic problem is that your deployment method is not the most powerful and hard to integrate. If you were to use, say, pip, virtualenvs, distribution packages, hell *something* that adheres to a standard and allows you to access information about "installed" things you'd be in a better position.
14:58 babilen yml: In a way you run into the same problem when you are trying to set passwords in salt that you want to generate.
14:59 babilen Salt simply doesn't have a nice "generate some information and then make it stick" interface at the moment.
15:00 babilen grains (not custom ones, but writing it explicitly with salt.states.grains.present) would probably be your best bet, but note that those would not updates automagically if you decide to update your git checkout
15:01 ajw0100 joined #salt
15:01 lahwran joined #salt
15:02 bmac2 joined #salt
15:03 clintberry joined #salt
15:03 fusionx86 joined #salt
15:05 Fiber^ joined #salt
15:06 I3olle joined #salt
15:07 timoguin joined #salt
15:08 David________ I'm getting an error when using git.config error and code here: https://gist.github.com/davidblackzipier/36a9c442367b55ad0dba
15:09 scbunn joined #salt
15:10 David________ Has anyone experienced the same issue and/or resolved it?
15:13 tvinson what are the downsides to using UUIDs as minion id's? minions are still addressable in state and execution modules using grains. are there pitfalls to configuring things that way? i'm trying to address hostname changes/vm cloning/other reasons a key with the same name would already be accepted on the master.
15:18 iggy stoogenmeyer__: are your minion_id's different than fqdn?
15:19 eliasp_web joined #salt
15:19 eliasp_web hi
15:19 MTecknology wow.. salt-run fileserver.update takes a looong time
15:19 stoogenmeyer__ iggy: what's fqdn ?
15:19 babilen fully qualified domain name
15:20 iggy MTecknology: state.test.*?
15:20 yml babilen: we are using pip and virtualenv
15:21 iggy yml: they aren't doing a sprint at pycon (despite what it says)
15:21 yml the issue we are trying to fix is deploy to a new virtualenv every time
15:21 Grokzen joined #salt
15:21 babilen yml: Why don't you make, say, "pip show FOO" output available?
15:21 yomilk joined #salt
15:21 yml and not reuse the same virtualenv on every highstate
15:21 stoogenmeyer__ iggy: looking at /etc/hosts, the first line (which i understand to be the fqdn?) is 127.0.0.1 localhost. Also, I named my minions manually for now, using a <env>.<role>.<index> scheme
15:21 yml because it can be let in an inconsistent state.
15:22 babilen yml: So delete it beforehand?
15:22 yml babilen: we want to keep the history of the 5 last viable deploy
15:22 yml for easy rollback if something bad happen
15:23 babilen So, keep explicit backups?
15:23 iggy stoogenmeyer__: python -c 'import socket; print socket.getfqdn()'
15:23 yml babilen: you can change the name of the virtualenv after the fact
15:23 eliasp_web any ideas, why the merge of this map.jinja fails? http://pastie.org/10091972 all I get is the 'os=Ubuntu' section + lookup pillar data, but the 'default' section isn't merged…
15:23 iggy it's actually the first name with dot's that matches an IP assigned to the current host
15:23 babilen yml: Sorry, there is obviously much more to your setup than what has been discussed so far.
15:23 yml and thus why I am looking for a solution to have a dynamic name
15:24 smcquay joined #salt
15:24 faust joined #salt
15:24 iggy eliasp_web: are you on a version that supports a default in grains.filter_by?
15:25 eliasp_web iggy: 2014.7.2 - let me check whether it already had 'default' support
15:25 stoogenmeyer__ iggy: i'm on ec2, the fqdn is "ip-xxx-xxx-xxx-xxx" (replaced x)
15:25 drawsmcgraw left #salt
15:25 yml I have much simplified the problem to I want to save a git revision of a particular repo and reuse it in mulitple places
15:25 stoogenmeyer__ that was the initial minion id, but I changed it in /etc/salt/minion
15:25 MTecknology DANGIT!
15:25 iggy stoogenmeyer__: ahh, okay then... your way probably isn't wrong
15:26 eliasp_web iggy: default (defaulting to default) has been there since 2014.1, so it should work just fine on 2014.7.2
15:26 stoogenmeyer__ iggy: I was just confused as to why it's not easy to call on previously defined grains from inside a custom grain file. It looks like each grain is trying to load all the grains
15:29 MTecknology heh... How do you tell mount.mounted to NOT create an fstab entry?...
15:29 iggy stoogenmeyer__: no, but you are when you try to import salt.modules.grains
15:29 eliasp_web MTecknology: persistent: false
15:30 * MTecknology hugs eliasp_web
15:30 MTecknology persist:
15:30 eliasp_web oh, sorry… this was just off the top of my head
15:30 bhosmer joined #salt
15:30 stoogenmeyer__ iggy: I understand. But how would I get a grain if I don't import salt.modules.grains ? (-:
15:30 iggy eliasp_web: do it the right way ;)
15:31 iggy that shit makes my eyes bleed
15:31 eliasp_web iggy: my map.jinja? whats "the right way"?
15:31 tomasfejfar joined #salt
15:32 MTecknology eliasp_web: zero apologies requested! I'm thrilled that you said that
15:32 eliasp_web :)
15:32 amcorreia joined #salt
15:34 SeeDickCode joined #salt
15:34 tvinson now that i'm thinking about it, targeting minions by hostname sort of undermines the idea behind pillar. hmmm
15:35 babilen How so?
15:35 MTecknology There we go! cmd.run with two options depending on server group to a single mount.mounted and now it no longer always evaluates as changes having happened. :D
15:35 huddy joined #salt
15:36 wendall911 joined #salt
15:37 peters-tx joined #salt
15:38 MTecknology I'm getting really incredibly excited for RAET
15:38 MTecknology crypto changes sound awesome, but raet has me really excited
15:39 iggy weird... I got over that about a week after I got back from saltconf ;)
15:39 denys joined #salt
15:40 MTecknology I haven't built the updated deb yet and that's how I distribute changes
15:40 MTecknology Maybe I should just do it
15:40 iggy the raet deps are a pita
15:41 iggy (just like the pygit2 and dulwich deps)
15:42 * MTecknology shudders
15:42 MTecknology Getting git working was a nightmare
15:42 iggy well, now I've added the deps to the salt-formula
15:42 iggy so it's as easy as adding salt.gitfs.pygit2 to your top file
15:43 dabb joined #salt
15:43 MTecknology nifty
15:43 MTecknology What did the raet deps look like?
15:43 rm_jorge joined #salt
15:46 iggy I don't remember... none of them were packaged for Debian (shocker!), so I abandoned the idea
15:47 MTecknology ah.. crap
15:47 MTecknology in that case, raet won't be happening.
15:48 iggy hopefully when SSE rolls their own repos, they'll supply those deps
15:48 MTecknology My environment is on a debian stable/testing hybrid at the moment. Once the new stable is released I'll switch to that fully.
15:49 iggy supposedly sometime this month
15:49 iggy but they are about as good at hitting release dates as salt
15:49 braqoon joined #salt
15:49 iggy heyo
15:49 dalexander joined #salt
15:49 braqoon hi
15:49 MTecknology LOL!
15:49 MTecknology Three people looked up when I laughed at that.
15:51 p66kumar joined #salt
15:53 babilen .4 will never happen
15:53 babilen :(
15:55 braqoon I'm trying to do a simple template file generation in salt and can't find how to do it with passing of some static data. I'm getting an error "Context must be formed as a dict"
15:55 braqoon Can some can tell me how to mark up static data in the file ? my current steup is : https://gist.github.com/anonymous/1c324f5e0c2fdf4a31a2
15:56 braqoon so in the example hashed message is just added to a file
15:56 iggy braqoon: replSet :
15:56 iggy should just be replSet:
15:56 babilen remove that space
15:56 braqoon oh, ok
15:56 iggy yaml-online-parser.appspot.com
15:57 eliasp_web I have the slight feeling Jinja doesn't like me today: http://pastie.org/10092046
15:57 iggy braqoon: did you look at the mongo-formula?
15:57 eliasp_web used saltstack-formulas/sysctl-formula as a blueprint for the defaults handlign
15:58 iggy that's murdock's... that was your first mistake
15:58 * iggy runs
15:58 * babilen chuckles
15:58 eliasp_web :)
15:59 braqoon iggy, yes i did. Looks way to complicated for me to use it. I don't like to use something that i don't understand
15:59 iggy braqoon: fair enough...
16:00 ingwaem joined #salt
16:01 iggy eliasp_web: maybe try the way salt-formula does it (similar to sysctl-formula, but slightly different)
16:02 eliasp_web iggy: thanks, I'll have a look
16:04 elfixit joined #salt
16:05 stanchan joined #salt
16:06 eliasp_web iggy: result: absolutely nothing… giving up for today, will try again in 13h ;)
16:07 iggy that sucks :/
16:07 iggy hate those kind of days
16:09 eliasp_web once this shit's working, I'll have to take care of tunneling my OpenVPN connection through my workplace firewall, so I can finally use my QuasselClient for IRC again ;)
16:09 otter768 joined #salt
16:09 iggy irccloud?
16:10 eliasp_web nah… I just love my Quassel way too much… for now my QuasselCore @home is online as 'eliasp' and @work I'm using webchat.freenode.net for now as 'eliasp_web'
16:10 eliasp_web as I need my OpenVPN tunnel anyways for git and some other stuff, I'll stick with that ;)
16:10 iggy they should really have a $50/yr package... think that would be more appealing to a lot of people
16:11 eliasp_web just the usual "making your regular tools work at your new workplace again"
16:11 raygunsix joined #salt
16:11 iggy love overbearing coroporate firewalls
16:12 iggy I have ssh listening on my_vps:443 just for situations like that
16:13 coval3nce joined #salt
16:17 GreyGnome Has anyone tried to perform user provisioning with salt? I looked at http://serverfault.com/questions/538218/deploy-ssh-key-from-master-to-minion-via-salt-pillars but I keep getting "Jinja variable 'str object' has no attribute 'uid' "'
16:17 GreyGnome I am trying to do a subset of the accepted answer... no joy at all...
16:18 GreyGnome It's a couple of years old, I wonder if the for loop syntax is acceptable anymore? I have salt 2014.7.1
16:19 iggy salt-call pillar.item users
16:19 iggy make sure it matches reality
16:20 spookah joined #salt
16:20 GreyGnome looks like it to me.
16:20 dalexander joined #salt
16:20 GreyGnome I see "             root:"
16:20 GreyGnome Then indented below there, I see "           uid:"
16:21 GreyGnome and a 0 below and indented from that.
16:22 ajw0100 joined #salt
16:22 yomilk joined #salt
16:22 GreyGnome My latest run of   salt myminion state.highstate test=True      says, "Rendering SLS 'base:system/base.users' failed: Jinja variable 'uid' is undefined"
16:23 JDiPierro joined #salt
16:23 iggy paste the state file
16:23 iggy it sounds like you put {{ uid }} somewhere instead of {{ user.uid }}
16:23 KyleG joined #salt
16:23 KyleG joined #salt
16:24 TheRealBill joined #salt
16:24 GreyGnome Sorry, yes I did. But user.uid doesn't help... it says    "Jinja variable 'str object' has no attribute 'uid' " . I will get the state file.
16:25 GreyGnome https://gist.github.com/GreyGnome/d012fcf499fc1ac066e2
16:25 GreyGnome thar she blows.
16:25 iggy output of salt-call pillar.item users (sanitized) wouldn't hurt either
16:27 GreyGnome Here she goes: https://gist.github.com/GreyGnome/55697d74a5418f71141d
16:27 iggy protip: gist supports multiple files per paste (and updating existing pastes)
16:28 writtenoff joined #salt
16:29 GreyGnome <insert appropriate newbie thanks here>
16:29 speedlight joined #salt
16:29 enarciso joined #salt
16:30 GreyGnome Looks like it sports a handy "Edit" button. I missed that... :-P
16:31 iggy take out the .uid part and run "salt-call -l debug state.highstate test=True"
16:31 iggy it'll show you the rendered jinja
16:31 iggy that might give you a clue what's going on
16:35 bluenemo_ joined #salt
16:35 tkharju joined #salt
16:37 GreyGnome Take it out of the salt state file? from both user.uid: and {{user.uid}} ?
16:37 evilgeenius joined #salt
16:38 evilgeenius Can anyone explain this error, it seems strange :
16:38 iggy just {{user.uid}}
16:38 evilgeenius [TRACE   ] '/usr/local/rvm/bin/rvm' could not be found in the following search path: ['/usr/local/sbin', '/usr/local/bin', '/usr/sbin', '/usr/bin', '/sbin', '/bin', '/home/chrisr/.rvm/bin']
16:38 evilgeenius Rvm does actually exist in "/home/chrisr/.rvm/bin"
16:38 MatthewsFace joined #salt
16:39 iggy evilgeenius: are you using rvm?
16:39 iggy (or trying to)
16:39 evilgeenius iggy:  yes, rvm.installed
16:39 iggy I think I actually have an issue open about the rvm module
16:39 GreyGnome It complains that "ID user in SLS system/base.users is not a dictionary"
16:39 GreyGnome But no other startling revelations exposed...
16:39 iggy https://github.com/saltstack/salt/issues/19105  "rvm module is whack"
16:40 MatthewsFace joined #salt
16:40 conan_the_destro joined #salt
16:40 iggy evilgeenius: that sounds like unrelated problems
16:41 iggy what does base.users state look like?
16:41 evilgeenius iggy: to me?
16:41 iggy yes
16:42 evilgeenius I don't have a base.users
16:42 evilgeenius I'm not sure what you mean
16:42 iggy system/base?
16:42 quintinadam joined #salt
16:43 GreyGnome I think that's to me.
16:43 iggy salt doesn't like one of your state files... and it would seem to
16:43 iggy oh shit
16:43 evilgeenius ha
16:43 iggy you guys are both the same color
16:44 iggy okay... so let's try again
16:44 Number6 joined #salt
16:44 iggy GreyGnome: base.users, go
16:44 GreyGnome my base has an init.sls.  In there i have include:   - .users
16:44 Number6 I'm running salt -G 'os:Ubuntu' test.ping , but I'm only seeing one Ubuntu server out of 8... a salt '*' test.ping returns all the machines... Where did I mess up the syntax?
16:45 iggy evilgeenius: you need to set runas if you want to use a user's rvm... ignore the trace message
16:45 babilen Number6: And all of them return "Ubuntu" on "salt '*' grains.get os" ?
16:45 babilen (mind pasting it on http://refheap.com ?)
16:45 GreyGnome @iggy I just updated https://gist.github.com/GreyGnome/55697d74a5418f71141d
16:46 evilgeenius iggy: runas isn't an option for the state rvm.installed
16:46 evilgeenius iggy: I think runas only applies to the module
16:46 Number6 babilen: Yep - all report Ubuntu for get os
16:46 iggy evilgeenius: my comment stands though
16:47 evilgeenius iggy: ok. The state says use "user:"
16:47 iggy evilgeenius: looks like user maps to runas
16:47 evilgeenius I am using user:
16:48 iggy GreyGnome: the error points at something else that you haven't pasted
16:48 evilgeenius It's looking for  '/usr/local/rvm/bin/rvm'  within '/home/chrisr/.rvm/bin' which makes no sense.  I do have a rvm within '/home/chrisr/.rvm/bin'
16:48 babilen Number6: Could you show me your exact commands and their output?
16:48 iggy evilgeenius: can you paste the state that calls rvm.installed?
16:49 Number6 babilen: Oh, I figured it out... they're all using the same public key
16:49 iggy lol
16:49 GreyGnome @iggy ok I'll post the (edited and trimmed down to only this issue) whole boat.
16:49 evilgeenius iggy:  http://hastebin.com/yebuwipoha.sm
16:49 aw110f joined #salt
16:50 babilen Number6: Well .... that wouldn't work ;)
16:50 babilen Number6: I mean: They all have the same name? Which public key are you referring to?
16:52 dopesong joined #salt
16:52 rap424 joined #salt
16:52 iggy evilgeenius: from looking at the code, it looks like user: should work (and be passed to the module as runas)
16:52 evilgeenius iggy: yeah I see that
16:52 evilgeenius so not sure what's happenin
16:54 iggy evilgeenius: like I said... that module is whack
16:54 evilgeenius iggy: :-( feck
16:54 evilgeenius s'pose I should fix it
16:54 iggy evilgeenius: open issues and use cmd.run
16:55 iggy or yeah, fix it if you can
16:55 Number6 babilen: Yeah, I messed up my AWS deploy.
17:00 yomilk joined #salt
17:01 babilen Number6: You have to ensure unique minion ids, that's all that's needed
17:02 timoguin joined #salt
17:05 GreyGnome @iggy I have stripped my entire /srv/salt and /srv/pillar to focus on just the users issue. All the files are here:  https://gist.github.com/GreyGnome/55697d74a5418f71141d
17:05 GreyGnome I still get "ID user.uid in SLS system/base.users is not a dictionary"
17:06 GreyGnome ...I can't see anything useful in the DEBUG output...
17:06 iggy that's not valid
17:06 iggy you aren't copying enough of that example
17:07 iggy that's not even valid yaml
17:07 GreyGnome To whom are you typing?
17:07 iggy you
17:08 bhosmer joined #salt
17:08 GreyGnome All the (root)# lines are shell commands.
17:08 iggy yeah, I'm following
17:08 bhosmer joined #salt
17:08 GreyGnome Ok, which file do you mean?
17:08 iggy /srv/salt/system/base/users/init.sls
17:10 GreyGnome I'll go back to the serverfault and see if I can construct poper YAML.
17:10 GreyGnome So far, I've been (non)-YAMLing for a week.
17:11 GreyGnome so to me it comes as little surprise... :-\
17:12 iggy https://gist.github.com/iggy/392111baf56832cd4fe1
17:12 iggy try that
17:12 iggy your top file is also wrong
17:12 iggy should be system.base
17:13 iggy but for whatever reason it seems to be working
17:13 mgw joined #salt
17:13 baweaver joined #salt
17:13 aberdine joined #salt
17:14 ndrei joined #salt
17:14 GreyGnome Agh. Mental transpose. Well your example worked. Now I'm going to try a real value...
17:15 buMPnet joined #salt
17:18 echo joined #salt
17:18 iggy I wanted to see the output from that
17:19 GreyGnome Ok I updated https://gist.github.com/GreyGnome/55697d74a5418f71141d
17:20 iggy I don't see any changes
17:20 GreyGnome Still complains "Rendering SLS 'base:system.base.users' failed: Jinja variable 'str object' has no attribute 'name' "
17:20 GreyGnome I inserted /srv/salt/system/base/users/init.sls: and the file's contents at the beginning.
17:21 GreyGnome The file has my updates to it.
17:21 GreyGnome user.present: - name: {{user.name}}
17:21 iggy what was the output from what I told you to put in there?
17:23 GreyGnome Updated my gist. It was successful.
17:23 theologian joined #salt
17:23 theologian joined #salt
17:23 GreyGnome Somehow, I think, it thinks that users is a string and not a dictionary...?
17:23 GreyGnome Excuse me, "user" is a string.
17:29 iggy https://gist.github.com/iggy/392111baf56832cd4fe1  show me the output of that
17:30 giantlock joined #salt
17:30 iggy GreyGnome: ^
17:33 GreyGnome <insert Startrek-like "working" voice here...>
17:33 GreyGnome Got it:  output was successful: "Comment: User add_root set to be added"
17:34 GreyGnome So that old syntax is not good, it seems? I actually tried his example verbatim and it didn't work.
17:34 iggy what's the actual output?
17:34 GreyGnome 1 second.
17:34 ty2u joined #salt
17:34 iggy the rendered yaml+jinja output is what I'm interested in
17:36 GreyGnome https://gist.github.com/GreyGnome/55697d74a5418f71141d
17:37 toastedpenguin joined #salt
17:37 iggy you should be able to cobble together what you have now and what that post has right?
17:38 GreyGnome Yes, thanks. I added the uid and sshpub and it looks good.
17:39 GreyGnome Thank you very much for your help.
17:41 awerner1 left #salt
17:43 raygunsix joined #salt
17:44 tomasfejfar joined #salt
17:45 JohanR joined #salt
17:45 JohanR Hi, got a very strange issue with a couple of my salt-minions
17:46 JohanR If I run salt-call grains.item id
17:46 JohanR I got the following error from iptables
17:46 JohanR [INFO    ] Executing command 'iptables --help' in directory '/root'
17:46 JohanR [INFO    ] output: iptables v1.4.7
17:46 JohanR ….
17:46 JohanR is it normal to invoke iptables when salt-call grains.item id are issued on a minion?
17:49 diegows joined #salt
17:50 iggy is anybody successfully using the mongodb-formula?
17:50 iggy JohanR: it probably shows a lot more than that, and yes, it's normal
17:51 JohanR iggy: Eeeh, why would some minions run iptables --help when I want to get the minion id??
17:52 iggy because it's run as part of grains/modules being initialized
17:52 iggy there's not a shortcut to bypass all that to get the id grain
17:53 iggy it loads all the modules/grains/etc it needs then outputs what you asked for
17:54 JohanR iggy: but it echo this to the console
17:54 iggy and?
17:55 Guest50339 joined #salt
17:56 JohanR I'm expecting for example: salt-call grains.item osrelease to only give the following output:
17:56 JohanR local:
17:56 JohanR osrelease: 6.5
17:56 iggy salt-call is a debugging tool... it's going to output a fair amount of data
17:57 MorbusIff joined #salt
17:57 timoguin joined #salt
17:59 JohanR ok, but this must be some difference between versions. 2014.1.5 does this, but 2014.1.13 work as I "want it to"
17:59 JohanR tnx anyway
17:59 manfred JohanR:  it was a bug in the code i wrote to implement things
18:00 manfred JohanR:  https://github.com/saltstack/salt/issues/13527#issuecomment-46692055
18:00 manfred was fixed in 2014.1.6
18:00 JohanR manfred: thanks! I will update
18:00 desposo joined #salt
18:01 desposo1 joined #salt
18:01 bhosmer__ joined #salt
18:02 spookah joined #salt
18:10 Vye joined #salt
18:10 otter768 joined #salt
18:11 huleboer joined #salt
18:12 phx joined #salt
18:17 nyx_ joined #salt
18:17 elfixit joined #salt
18:20 denys joined #salt
18:20 yggdrasi1 joined #salt
18:21 heksi joined #salt
18:22 lahwran joined #salt
18:22 soren joined #salt
18:24 martoss joined #salt
18:25 tg_ joined #salt
18:27 drawsmcgraw joined #salt
18:29 clintberry joined #salt
18:30 agend joined #salt
18:30 tg_ Hey all, I'm looking for some advice on best practices for managing file content. I have php files that I need to configure and I'm storing variables in map.jinja and pillars. Currently I'm making use of the file.patch state, but this will obviously break once pillar data changes as the hash will not match. Is the recommendation to use sed or another method instead?
18:31 tg_ The other option I've considered is entire file.managed operations, but it seems excessive.
18:33 fyb3r joined #salt
18:34 stoogenmeyer__ joined #salt
18:34 iggy if you look at the formulas, they almost all use file.managed with templates
18:35 iggy that doesn't necessarily mean that's the best thing for you to do
18:35 lb1a joined #salt
18:35 heksi joined #salt
18:36 tg_ True. It seems to be the most universal choice for this type of configuration, but I'd rather not replace entire php files as I'd like to retain as much forward compatibility as possible with this formula.
18:39 tg_ Hmm, well I suppose I can grab the specific version from the project instead of seeking out the version in the repos. If I need to update in the future it will require an update to the code but I suppose that's the safest bet.
18:40 babilen People I know like augeas, but I'm not really a fan
18:42 babilen Ah, you aren't necessarily talking about configuration files here, are you?
18:43 ndrei joined #salt
18:44 tg_ I'm configuring phpldapadmin. It makes use of a config.php file. I also need to modify some lines in the source to backport fixes that aren't in the ubuntu repo versions.
18:46 krelo joined #salt
18:47 keimlink joined #salt
18:47 writtenoff I'm wondering if anyone can direct me where or what to look to fix Minion did not return error for my initial test.ping test?
18:49 writtenoff I went through steps in Salt documentation's sections: Troubleshooting the Salt Minion, Troubleshooting the Salt Master, Configuring the Salt Minion.
18:49 tg_ The backports are fine as they don't use variables. I'd like to find a way to fully configure the application with pillar data though. The more I think about it the more it looks like a version lock and file.managed is my best choice. Importing an entire 576 line file to change two values seems inefficient though.
18:50 Vye joined #salt
18:50 yggdrasi1 joined #salt
18:50 awerner joined #salt
18:50 aqua^lsn joined #salt
18:51 patrek joined #salt
18:51 babilen tg_: And augeas is no fit?
18:51 tg_ babilen: Augeas looks very interesting. Thanks for the recommendation. I may end up making use of it even if it doesn't work for this scenario.
18:51 babilen tg_: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.augeas.html
18:52 babilen We found that it is useful for those "adjust foo value in file /a/b/c" but I typically prefer the full "template approach"
18:53 tg_ Oh, wow look at that. this looks very useful
18:53 tg_ I think this may be the perfect answer. Thank you!
18:53 babilen Sorry, I should have provided more information right away.
18:53 babilen Enjoy :)
18:53 yomilk joined #salt
18:56 the_lalelu babilen: i do not "trust" augeas. so templating is also my choice.
19:00 iggy lsb_release -sc
19:00 ek6 joined #salt
19:00 iggy can someone tell me what the output of that is on ubuntu trusty/precise
19:01 awerner joined #salt
19:01 ek6 iggy: just on...output of what?
19:01 iggy lsb_release -sc
19:01 awerner joined #salt
19:02 ek6 trusty and...
19:02 ek6 precise :)
19:02 iggy okay, that's what I thought
19:02 iggy danke
19:03 ipmb joined #salt
19:07 fusionx86 joined #salt
19:08 keimlink_ joined #salt
19:09 I3olle joined #salt
19:12 yorjo joined #salt
19:15 eliasp iggy: German? join #salt-de! :)
19:22 GreyGnome Ich kann nicht Deutsch sprechen.
19:22 drawsmcgraw I need to maintain a veth pair on some machines. They're created with an 'ip' command, which works great but doesn't survive reboot
19:22 drawsmcgraw i.e. the pair is gone after a reboot
19:22 drawsmcgraw That causes..... strife. When we have to reboot a machine.
19:23 drawsmcgraw Is there some way to make that veth pair more permanent? I'm not certain network.managed() would do it....
19:23 bhosmer joined #salt
19:23 Eugene I would file.managed your OS' network-configuration file
19:24 Eugene Eg, /etc/network/interfaces or /etc/sysconfig/network-scripts/ifcfg-whatever
19:26 jalaziz joined #salt
19:27 awerner joined #salt
19:28 smcquay joined #salt
19:29 drawsmcgraw Eugene: Agreed, but the veth devices are ports on bridges
19:29 drawsmcgraw so... can I say something like "TYPE=Port" in the config file?
19:29 drawsmcgraw CentOS 6, btw
19:29 Eugene THat was my next question, heh
19:30 Eugene I'm not familiar with "the CentOS way" of doing veth devices
19:30 amcorreia joined #salt
19:30 Eugene This is for a VZ host?
19:31 _ale1_ joined #salt
19:31 Eugene It looks like you don't really need to, but I've not played with VZ in literal years. https://wiki.openvz.org/Virtual_Ethernet_device#Making_a_veth-device_persistent
19:31 drawsmcgraw No, it's an openstack cluster
19:31 Eugene Ya got me, then.
19:31 drawsmcgraw And yeah, I don't even know anything about VZ hosts but that's all that comes up in my Google searches :)
19:31 Eugene Surely they have a persistence option
19:32 drawsmcgraw For what it's worth, the incantation is:  ip link add name veth0 type veth peer name veth1
19:32 drawsmcgraw You would think so. I'm sure they do, I just need to do more research....
19:32 Eugene You could add that to the end of rc.local :v
19:32 drawsmcgraw huhn... That's a thought. Basically ensure it's run on each startup?
19:33 Eugene Yeah
19:33 drawsmcgraw I'll look into that. Thanks!
19:33 Eugene file.append would do the trick
19:35 soren joined #salt
19:35 baweaver joined #salt
19:35 yorjo I run a command using the salt api ( local = salt.client.LocalClient()) and this synchronous call returned a VALUE_TRIMMED instead of the true return output of the command. Am I hitting some sort of limit?
19:40 fxhp joined #salt
19:42 dopesong_ joined #salt
19:47 cberndt joined #salt
19:49 iggy eliasp: took some in high school... that and expletives are about all I remember
19:50 lb1a joined #salt
19:50 JPT joined #salt
19:51 florinandrei joined #salt
19:51 jalaziz joined #salt
19:51 berserk joined #salt
19:53 martineg_ joined #salt
19:54 smcquay joined #salt
20:03 schristensen joined #salt
20:03 mgw joined #salt
20:11 otter768 joined #salt
20:12 rihannon joined #salt
20:17 evilrob joined #salt
20:20 timoguin joined #salt
20:21 yorjo Just saw the trim_dict function in the salt.utils.dicttrim module. Mostly explain what I am seeing
20:24 BretFisher joined #salt
20:24 spookah joined #salt
20:30 ajw0100 joined #salt
20:37 rap424 joined #salt
20:39 yomilk joined #salt
20:40 Tecnico1931 joined #salt
20:50 I3olle joined #salt
21:04 sijis left #salt
21:12 a7p joined #salt
21:12 tg_ Any advice on removing blank lines that are inserted when using {% from 'foo/map.jinja'...
21:14 lietu- joined #salt
21:15 I3olle joined #salt
21:15 iggy http://jinja.pocoo.org/docs/dev/templates/#whitespace-control
21:16 a7p left #salt
21:18 hojgaard joined #salt
21:23 aparsons joined #salt
21:23 paha joined #salt
21:23 murrdoc joined #salt
21:25 jalaziz joined #salt
21:27 bVector joined #salt
21:28 tg_ That works perfectly. Thanks!
21:29 nyx_ joined #salt
21:35 fyb3r left #salt
21:36 xrc joined #salt
21:37 BretFish_ joined #salt
21:37 baweaver joined #salt
21:38 agend joined #salt
21:39 xrc left #salt
21:40 yomilk joined #salt
21:43 xrc joined #salt
21:47 nickdew joined #salt
21:48 fusionx86 joined #salt
21:51 nyx_ joined #salt
21:52 buMPnet joined #salt
21:54 PI-Lloyd joined #salt
21:55 xrc joined #salt
21:56 sarlalia1 joined #salt
21:59 buMPnet joined #salt
22:00 raygunsix joined #salt
22:00 nickdew left #salt
22:00 yggdrasi1 joined #salt
22:10 nickdew joined #salt
22:12 otter768 joined #salt
22:12 jalaziz joined #salt
22:14 baweaver joined #salt
22:14 jalaziz joined #salt
22:17 hal58th joined #salt
22:20 raygunsix joined #salt
22:27 bhosmer_ joined #salt
22:31 giantlock joined #salt
22:36 Bevo joined #salt
22:36 ipmb joined #salt
22:37 Bevo does anyone know if its possible to use puppet facts as grains?
22:39 Singularo joined #salt
22:39 murrdoc factor yes
22:39 murrdoc puppet facts maybe
22:39 babilen external pillar
22:39 iggy https://github.com/saltstack/salt-contrib/tree/master/grains
22:40 murrdoc output = __salt__['cmd.run']('facter -p -j')
22:40 murrdoc that still work :D
22:40 murrdoc iggy:  what came out of your ticket ?
22:40 murrdoc for cmd.run in 2015.2
22:41 iggy I haven't seen anything on it since your comment
22:41 murrdoc good
22:41 iggy I'm hoping they are having some real talk about that "decision"
22:41 timoguin joined #salt
22:42 alemeno214 joined #salt
22:44 bfoxwell joined #salt
22:45 alemeno214 Hey! Question regarding Salt Syndic. We have a private network with a full Salt setup, and we have another network up in Digital Ocean that we use for dev. We want to be able to run the same forumlas set up on our main network in Digital Ocean. One of our former engineers had mentioned Salt Syndic as a possible option, but that it was still pretty buggy. Any insights?
22:46 iggy we just have multiple masters (that all pull from the same gitfs)
22:46 yomilk joined #salt
22:47 echo joined #salt
22:47 viderbit joined #salt
22:47 alemeno214 That's what I was thinking of doing since our config is in Git, but my team lead wanted to investigate Syndic first
22:47 murrdoc tell him it stinks
22:47 murrdoc </research>
22:47 aquassaut joined #salt
22:47 murrdoc :)
22:48 murrdoc btw u could totally use syndic
22:48 alemeno214 Ah okay thanks, any specific pain points that I can bring back to him? He's going to ask
22:48 murrdoc security is one
22:48 murrdoc and rest u ll have to test out to see what breaks in your setup
22:50 alemeno214 okay great, thanks for the insight!
22:50 primechuck joined #salt
22:56 iggy a lot of people have trouble with the concept of how it works (i.e. it doesn't work the same as when you just have a single master... commands route differently, returns route differently, there's lag, etc.)
22:57 murrdoc its easier btw to use a master of masters
22:57 alemeno214 master of masters?
22:57 murrdoc to maintain the different individual masters' code
22:57 alemeno214 ah
22:57 murrdoc yeah so put a minion on your masters and point it at a central master
22:57 iggy it's even easier to just have separate masters
22:57 * iggy runs
22:57 murrdoc lies
22:57 murrdoc like u can run
23:02 micko joined #salt
23:05 hal58th joined #salt
23:06 jdesilet joined #salt
23:10 mosen joined #salt
23:10 ndrei joined #salt
23:15 aw110f joined #salt
23:17 elfixit1 joined #salt
23:18 clintber_ joined #salt
23:20 alemeno214 What would be the potential downsides of running a redundant master setup like this: http://docs.saltstack.com/en/latest/topics/tutorials/multimaster.html
23:20 alemeno214 Have one of the masters residing on Digital Ocean with a VPN connection to our private network
23:23 zwi joined #salt
23:24 iggy the possibility that your minions could connect to the wrong master would probably be the big one
23:30 rhodgin joined #salt
23:30 yomilk joined #salt
23:32 mosen joined #salt
23:32 quintinadam joined #salt
23:36 kusams joined #salt
23:39 baweaver joined #salt
23:40 bluenemo joined #salt
23:42 raygunsix joined #salt
23:48 iggy I just know someone is going to ask me to split up my mongodb-formula changes
23:49 Daemonik joined #salt
23:54 ipmb joined #salt

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