Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-03-02

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

All times shown according to UTC.

Time Nick Message
00:00 MTecknology that's what I get. That's easy...
00:00 Cidan joined #salt
00:00 MTecknology saltstack is covering most of it, that's why I'm bigging out the first night
00:03 MTecknology got my taxi for tomorrow taken care of, they'll be here 06:30
00:03 murrdoc where is the salt conf agenda
00:03 MTecknology http://eventmobi.com/saltconf15/
00:07 ocdmw joined #salt
00:08 murrdoc thanks man
00:08 murrdoc whats 'genreal session' on wednesday monring
00:09 p0rkbelly joined #salt
00:09 MTecknology usually it's a kick-off to get people excited about being there
00:10 MTecknology iggy: saltstack is covering breakfast, lunch, and supper Tuesday and Wednesday, and breakfast and lunch on Thursday. I'll be home before my typical supper time on Thursday.
00:10 MTecknology I get there Monday. I'll be there four days. That's $30/day (cumulative)... so $120 to spend when I get there on Monday.
00:10 MTecknology too bad I have to get everything itemized because alcohol gets pulled off
00:16 kaictl joined #salt
00:18 iggy Yeah, I just spent $60 on lunch
00:18 MTecknology holy crap
00:18 MTecknology where?
00:19 iggy sharpies or something?
00:23 MTecknology I think that was one of the recommendations the hotel concierge suggested
00:24 MTecknology I decided that work was paying for it, so screw the taxi and public transportation... they screwed me over last year and I'm gonna enjoy the crap out of this one.
00:25 ajw0100 joined #salt
00:26 iggy are you staying at the grand?
00:27 murrdoc nope
00:27 murrdoc was too late
00:27 MTecknology did it sell out?
00:35 MTecknology iggy: your company website pukes without javascript
00:35 MTecknology actually... just the header and footer do
00:35 lost1nfound joined #salt
00:35 iggy I have nothing to do with it
00:36 MTecknology I was poking around because I looked up your talk, but a junky website turns me off immediately
00:36 iggy but our web guy is an idiot, so not surprising
00:36 iggy I don't think they are targeting people with javascript off
00:36 MTecknology usually not, but....
00:37 iggy there's probably a contact address on the site
00:37 MTecknology I've had higher ups ask me about opinions of companies (which is better than the other) and that's one of the things I use to judge how carefully they code
00:37 MTecknology and how carefully they do other things
00:37 iggy we don't make our own website
00:38 murrdoc what website
00:38 MTecknology iggy: Bamidele Benjamin Adetayo?
00:39 cotton_ joined #salt
00:39 aphorise joined #salt
00:39 MTecknology jquery, google api, "//   =========== how many directories from root is this HTML file located? ==============" ...
00:39 iggy I honestly have no idea who does it
00:42 cotton_ joined #salt
00:46 JDiPierro joined #salt
00:55 bombmask joined #salt
01:13 murrdoc iggy:  hows the kubernetes stuff going
01:14 iggy I'm not currently working on it
01:14 elfixit1 joined #salt
01:15 hackeron joined #salt
01:15 murrdoc ah, u mentioned you guys were considering it
01:16 iggy my boss is all about it, I'm still a little skeptical that's we'll be able to move all of our stuff into it
01:17 query joined #salt
01:17 rypeck joined #salt
01:18 Guest92640 left #salt
01:21 mohae_ For a "file.managed", I'm getting an error: "Unable to manage file: 'OrderedDict' object has no attribute 'find'
01:21 mohae_ " and am unable to find anythinig about it. Any suggestions?
01:26 SheetiS mohae: have a gist or a bpaste of the file.managed state in question?
01:34 g3cko joined #salt
01:35 cotton joined #salt
01:39 toastedpenguin joined #salt
01:45 mohae https://gist.github.com/mohae/20500b7e3197319da53d
01:47 mohae and a gist of the first error encountered. The other parts of this state have the same error occurring: https://gist.github.com/mohae/5701c07bbe8c9f8402dd
01:49 iggy why file.managed? (vs archive.extracted)
01:49 stylica joined #salt
01:50 mohae iggy: good question, I was using the redis-formula as a guide. I'll look up archive.extracted
01:51 mohae as an aside, I've come to dislike the stuff in saltstack formula
01:52 iggy some are good, some are old and crusty, and some are just bad
01:53 mohae iggy: that is what I've found
01:54 iggy they are all meant to be "Starting points"
01:55 mohae well, better starting points would be more useful
01:55 november joined #salt
01:56 harkx joined #salt
01:56 iggy feel free to file bugs
01:58 stylica joined #salt
01:59 SheetiS mohae: I am not sure why you are running into trouble at first glance with file.managed, but archive.extracted would condense and simplify what you have there as iggy said.  Perhaps you can avoid the issue all together by switching which state type you are using.
02:00 mohae I switched to archive.extracted and get a yaml render error. It appears that I'm doing source_hash wrong: https://gist.github.com/mohae/c50b019d09f52afc08b4
02:01 mohae iggy: once I get this working the way I want, I might file some for the redis-formula
02:02 mohae is if_misssing the extracted dir?
02:04 mohae nvm, I should read the docs closer
02:05 mohae I'm still not sure why I'm getting a problem with my source_hash though
02:08 november is there a good example of a Java install with salt?
02:09 mohae SheetIS: I think my first error, before switching to archive.extracted was due to my source_hash, it didn't specify an algo
02:14 michelangelo joined #salt
02:15 nullptr` joined #salt
02:18 Ryan_Lane joined #salt
02:31 fsheraton joined #salt
02:32 fsheraton left #salt
02:37 bombmask left #salt
02:37 scoates joined #salt
02:42 cotton joined #salt
02:45 cooldude_ joined #salt
02:47 ilbot3 joined #salt
02:47 Topic for #salt is now Welcome to #salt | SaltConf 2015 is Mar 3-5! http://saltconf.com | 2014.7.2 is the latest | Please be patient when asking questions as we are volunteers and may not have immediate answers | Channel logs are available at http://irclog.perlgeek.de/salt/
02:49 favadi joined #salt
02:52 dude051 joined #salt
02:58 clintberry joined #salt
02:58 jhauser_ joined #salt
03:01 iggy cooldude_: you need the libssh-dev package if you want to use git+ssh uri's
03:01 badon_ joined #salt
03:03 chamunks Is it possible to reload a salt-minion via salt rather then restart the salt-minion daemon.
03:03 ShibMcNe joined #salt
03:03 tmh1999 joined #salt
03:04 iggy chamunks: salt minion service.restart salt-minion
03:04 chamunks iggy you'd imagine that might unplug it.
03:05 iggy the minion connects to the master on startup
03:05 gnord joined #salt
03:05 iggy so yeah, they will lose contact briefly
03:06 chamunks iggy seems that in a test it just kind of hangs.
03:07 chamunks there we go nvm a test.ping seems to acknowledge its still alive.
03:07 tmh1999 joined #salt
03:09 cooldude_ I installed libssh-dev and still seen the same error : when I run python -c 'import pygit2'   : ERROR Import Error: no Module named _pygit2
03:09 murrdoc pip freeze | grep git
03:10 iggy I meant you need to have the headers installed before you build and install libgit2
03:13 aparsons joined #salt
03:14 cooldude_ ok, anyone have the instruction to follow since the one from github seems to be not working correctly
03:15 murrdoc pip install pygit2
03:16 murrdoc totes srs
03:17 murrdoc http://www.pygit2.org/install.html says u need to "sudo ldconfig"
03:17 murrdoc to make python -c 'import pygit2' work
03:18 bigpup3 joined #salt
03:22 murrdoc time change affect u yet iggy ?
03:25 iggy not really
03:25 iggy I woke up earlier than normal
03:25 iggy but it's only 1 hour difference
03:26 bigpup3 joined #salt
03:27 murrdoc i was in la all week last week
03:27 murrdoc thursday body switched to west coast time
03:28 murrdoc today i think it switched back to chicago time
03:28 murrdoc wed/thu i dont know wtf will happen
03:28 murrdoc still excited about the talk
03:28 murrdoc talks*
03:29 iggy You speaking?
03:30 murrdoc nah
03:30 murrdoc have nothing interesting to present
03:31 murrdoc maybe next year after we salt our servers, got 15k+, probably should have something worthwhile
03:31 murrdoc mebbe
03:31 murrdoc u gonna come to monitarama ?
03:32 iggy nope
03:32 iggy too close
03:33 evle joined #salt
03:33 signull monitarama...? battle of the nagios replacement?
03:34 murrdoc zenoss all day
03:34 murrdoc too close to what ?
03:38 iggy realistically, everything... I'm going to Italy in August
03:38 iggy for like 2 weeks
03:38 murrdoc nice!
03:38 murrdoc vacation ?
03:38 iggy that's basically all my PTO for this year
03:39 iggy music festival
03:39 murrdoc whats in italy
03:39 v0rtex does anyone know approximately how long the SSCE exams take? I am planning to do it at saltconf on tuesday but am trying to figure out my schedule that evening
03:40 signull website says usually its 3 full business days. for the conf they said its early evey morning each morning. dont know how long persay
03:41 iggy v0rtex: less than 2 hours?
03:41 iggy I think the schedule only says 2 hours
03:42 v0rtex iggy: thanks. I saw on the schedule that exams open at 5:30 but just didn't see anything that said how long it takes.
03:45 iggy oh, I guess they don't... I was just going off the next thing after starting at 7
03:45 iggy but I guess that doesn't necessarily assume everyone testing will be done by then
03:46 v0rtex true
03:46 iggy I'd hate to try to ping anyone right now
03:46 v0rtex agreed - I will probably ask tomorrow
03:46 iggy it's almost 10 here and they've got a busy week ahead of them
03:48 dude051 joined #salt
03:50 iggy some googling says an hour? Hard to say for sure
03:52 v0rtex okay thanks for looking iggy
03:58 Jahkeup joined #salt
04:03 subsignal joined #salt
04:04 favadi left #salt
04:04 favadi joined #salt
04:19 murrdoc joined #salt
04:25 sfxandy joined #salt
04:32 subsignal joined #salt
04:35 nullptr` joined #salt
04:40 davidbanham Hello friends! I am running salt-minion on a brand new host, and it is saying "[CRITICAL] The Salt Master has rejected this minion's public key!"
04:40 davidbanham On the salt master, however, salt-key -L tells me there are no keys in the rejected state.
04:41 davidbanham I would expect to see one key in the Unaccepted state, but there is also nothing there.
04:41 davidbanham Halp?
04:43 davidbanham Ah, it had an old hostname cached in /etc/salt/minion_id.
04:43 davidbanham Thanks for listening!
04:44 tligda joined #salt
04:53 aqua^mac joined #salt
05:00 Ryan_Lane joined #salt
05:00 kanaka_ joined #salt
05:01 bhosmer_ joined #salt
05:07 TyrfingMjolnir joined #salt
05:10 mosen joined #salt
05:13 kermit joined #salt
05:24 kanaka_ joined #salt
05:44 kanaka_ joined #salt
05:44 dude051 joined #salt
05:46 MTecknology Good night SoDak!
05:47 MTecknology Good afternoon SLC!
05:47 MTecknology iggy: ^ :D
05:49 MTecknology v0rtex: it's timed, i think it's something like 1.5 hr ish... the schedule is start time, not time to complete
05:49 MTecknology It might be one hour...
05:49 v0rtex MTecknology: I found the answer in an email just now - 60 minutes
05:49 MTecknology perfect :)
05:49 MTecknology I'll be taking it as well. See you there!
05:51 ramteid joined #salt
06:03 clintberry joined #salt
06:06 pravka joined #salt
06:15 iggy I'll be there too
06:22 gibmachine joined #salt
06:27 Furao joined #salt
06:28 calvinh joined #salt
06:30 cberndt joined #salt
06:34 krelo joined #salt
06:43 calvinh joined #salt
06:45 wintamute joined #salt
06:51 I3olle joined #salt
06:52 stoogenmeyer joined #salt
06:55 tmh1999 joined #salt
06:55 calvinh_ joined #salt
06:56 mikeywaites joined #salt
07:16 nullptr` joined #salt
07:18 iggy so yeah... computer just ate my entire presentation
07:18 iggy I quit
07:21 pravka joined #salt
07:26 MTecknology iggy: OUCH!
07:26 pravka joined #salt
07:26 MTecknology no backup copy or anything?
07:27 iggy no
07:27 MTecknology :(
07:27 iggy and it's not like it just crashed and I lost a couple hours work, it overwrote the entire file (the one I started with was 5M, after crash, 44k)
07:28 MTecknology ouch!
07:28 MTecknology at least you remember most of it, right?
07:29 MTecknology why the crap am I still awake? I have a taxi showing up in 5hr
07:30 MTecknology iggy: sorry man.. wish I could say something helpful.
07:35 KermitTheFragger joined #salt
07:37 flyboy joined #salt
07:41 xist joined #salt
07:43 JlRd joined #salt
07:43 trikke joined #salt
07:49 dRiN joined #salt
07:49 Furao joined #salt
07:52 stylica_ joined #salt
07:57 otter768 joined #salt
07:58 Ryan_Lane joined #salt
07:58 iggy I'll survive, I just have people at work asking me to see my slides... like last week
08:04 jri joined #salt
08:06 malinoff joined #salt
08:06 eseyman joined #salt
08:13 krelo joined #salt
08:17 pf_moore joined #salt
08:21 skim_ joined #salt
08:28 markm joined #salt
08:30 Auroch joined #salt
08:32 jtang joined #salt
08:34 Andre-B joined #salt
08:35 kawa2014 joined #salt
08:41 mikeywaites joined #salt
08:41 pravka joined #salt
08:43 lb1a joined #salt
08:44 mikeywaites1 joined #salt
08:46 JlRd joined #salt
08:47 ajw0100 joined #salt
08:51 _ale1_ joined #salt
09:01 chiui joined #salt
09:02 malinoff joined #salt
09:05 linjan joined #salt
09:07 I3olle joined #salt
09:15 forrest joined #salt
09:19 cberndt joined #salt
09:19 intellix joined #salt
09:27 Xevian joined #salt
09:31 j-saturne joined #salt
09:33 jhauser joined #salt
09:34 kiseok7 joined #salt
09:39 wincyj joined #salt
09:39 wincyj hello
09:40 wincyj i recive this kind of error
09:40 wincyj Comment: Source file salt://files/config_db.php not found
09:40 wincyj as i suppose the path i provided is wrong but tbh i dont know what's the structe that salt takes is there somewhere explanation on inet? i havent found one.
09:41 phx wincyj, http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html
09:42 wincyj im in
09:42 wincyj but it says it takes paths from master
09:43 wincyj ../salt/glpi/files/file.log
09:43 intellix joined #salt
09:44 wincyj and i've attach this path in master file
09:44 phx wincyj, relative to the file_roots directive
09:45 dkrae joined #salt
09:45 wincyj should it be defined overt?
09:45 sfxandy joined #salt
09:48 wincyj ok i got master as following: http://pastebin.com/zbdJ8X8m
09:48 wincyj what's wrong with it?
09:49 N-Mi_ joined #salt
09:52 Furao joined #salt
09:52 mikeywaites joined #salt
09:55 iggy wincyj: try "salt-call cp.list_master" on the minion
09:56 wincyj i use salt-ssh
09:57 iggy there's probably some sort of similar functionality... I dn't use salt-ssh, so I don't know tbh
09:58 wincyj its the same conf i suppose
09:58 otter768 joined #salt
09:58 wincyj i just dont understand why it doesnt work since i provided full path
10:04 cberndt joined #salt
10:15 j-saturne joined #salt
10:19 tomspur joined #salt
10:22 _mel_ joined #salt
10:26 slav0nic joined #salt
10:30 j-saturne joined #salt
10:38 giantlock joined #salt
10:45 ocdmw joined #salt
10:53 __gotcha joined #salt
10:54 __gotcha What is the way to make a function from a custom module fail ?
10:54 __number5__ throw an exception
10:55 __gotcha __number5__: how will the exception be reported ?
10:56 __gotcha for instance when using orchestrate
10:57 evle1 joined #salt
10:58 bluenemo joined #salt
10:58 bluenemo joined #salt
10:59 lietu joined #salt
10:59 __gotcha __number5__: thanks
11:01 __number5__ __gotcha: this blog post might help you http://ryandlane.com/blog/2014/09/08/saltstack-development-behavior-of-exceptions-in-modules/
11:05 ocdmw joined #salt
11:06 __gotcha __number5__: thanks again
11:06 __number5__ no worries
11:12 giantlock joined #salt
11:15 amcorreia joined #salt
11:16 mikeywaites left #salt
11:16 __gotcha1 joined #salt
11:21 fredvd joined #salt
11:25 linjan joined #salt
11:37 refnode joined #salt
11:37 refnode hi all
11:55 refnode hi, i'm new to salt, coming with experience from puppet
11:56 refnode are there any plans for something like puppet forge for salt?
11:56 Auroch joined #salt
11:59 otter768 joined #salt
12:01 babilen refnode: What is that?
12:03 refnode babilen, https://forge.puppetlabs.com/ the central place for registering puppet modules
12:03 babilen https://github.com/saltstack-formulas ?
12:04 babilen Or https://github.com/saltstack/salt-contrib ?
12:04 babilen What is "a module" exactly?
12:04 davidbanham joined #salt
12:04 babilen refnode: What are you looking for exactly? What are you trying to do?
12:04 refnode babilen, first link I know, I'm checking your second link, thanks
12:05 huddy joined #salt
12:05 babilen In what way is the first link not what you are looking for?
12:06 refnode I read the stuff first, thanks for the hint
12:06 bhosmer joined #salt
12:06 babilen sure
12:06 refnode puppet modules <=> salt formulas
12:07 babilen There is also http://docs.saltstack.com/en/latest/ref/modules/all/index.html#all-salt-modules and http://docs.saltstack.com/en/latest/ref/states/all/
12:07 stoogenmeyer_ joined #salt
12:07 babilen refnode: So you were looking for https://github.com/saltstack-formulas weren't you?
12:07 refnode I'm checking if salt becomes a way for us, have a foreman/puppet  environ up, but we do some realtime tasks with salt already
12:08 refnode babilen, yes I already checked https://github.com/saltstack-formulas therefore I was wondering if another central "formula place" would exists
12:08 ocdmw joined #salt
12:09 babilen No, various people naturally have their own collection, but saltstack-formulas is the official one
12:09 bhosmer_ joined #salt
12:10 CeBe joined #salt
12:10 refnode ok, I have to "map" some concepts starting with salt, but it looks great yet
12:10 babilen Just ask if there is something you need help with :)
12:10 babilen (e.g. understanding the difference between execution modules and states that many new users find confusing)
12:11 refnode thanks, starting a project now on it, in a few days I see if I got trouble ;-)
12:21 JDiPierro joined #salt
12:24 tkharju joined #salt
12:28 stoogenmeyer joined #salt
12:42 diegows joined #salt
12:43 johtso joined #salt
12:43 sfxandy joined #salt
12:44 JlRd joined #salt
12:46 shorty_mu joined #salt
12:46 frikkie joined #salt
12:47 JDiPierro joined #salt
12:49 shorty_mu Hi all, I have a Python newbie question. I want to copy a file an the "name" should be the "source" with the unix timestamp(epoch) appended to it. Can anyone give me a hint on how todo that? Cheers
12:51 aqua^mac joined #salt
12:54 ojintoad joined #salt
12:58 nullptr` joined #salt
13:02 lietu shorty_mu, https://docs.python.org/2/library/shutil.html and https://docs.python.org/2/library/time.html
13:02 shorty_mu Damn! I forgot to say that I need this in an SLA state.
13:04 TyrfingMjolnir joined #salt
13:06 shorty_mu I found things like int(time.time()) or {{ "now"|strftime  %s}} but I don't know how to assign this to a Jinha variable.
13:09 lothiraldan joined #salt
13:10 nighter joined #salt
13:14 netzmonster joined #salt
13:15 babilen shorty_mu: You cannot do that in jinja
13:15 babilen (afaik)
13:15 netzmonster left #salt
13:16 shorty_mu babilen: *Brilliant*  I use cmd.run now...
13:16 babilen jinja simply isn't particularly powerful when it comes to calling Python code
13:17 babilen (in particular using modules in the standard library)
13:17 bastiandg joined #salt
13:17 nighter joined #salt
13:17 shorty_mu babilen: ack and thx
13:18 cotton joined #salt
13:19 saltymoli joined #salt
13:20 babilen The "easiest" way if you insist on jinja is probably to write an execution module that provides a frontend to the time module and which you can then call from jinja. That sounds like more work than it will probably be, but is not particularly beautiful
13:21 babilen The next approach would be to write your state in mako which allows you to include Python blocks directly: http://docs.makotemplates.org/en/latest/syntax.html#python-blocks -- The (only) drawback of this is that, while gaining easy access to Python, you also add states that are written in yet another templating language.
13:21 babilen I personally would have much preferred mako over jinja fwiw
13:22 babilen Which leaves two other options: 1. Write the state in #!py or #!pyobject or 2. Just use cmd.run with a suitable command that simply does what you want it to
13:22 babilen At this point I'd probably recommend the latter (and the approach you seem to have chosen)
13:24 nighter joined #salt
13:27 _sfxandy hi everyone.  i might have missed something very very obvious, but within an SLS template is there any built in variable or execution module that will let me list the hostnames that are configured as salt minions...?
13:27 murrdoc joined #salt
13:28 murrdoc iggy or MTecknology u guys should use http://gfycat.com/TepidFavorableHammerkop at the end of your presentation
13:28 murrdoc for the lulz
13:30 nighter joined #salt
13:30 linjan joined #salt
13:32 bastiandg joined #salt
13:35 pdayton joined #salt
13:37 TinuvaMac joined #salt
13:40 ndrei joined #salt
13:42 Damien___ joined #salt
13:43 holms joined #salt
13:44 dyasny joined #salt
13:44 holms maybe someone have 2min for checking this? https://github.com/saltstack/salt/issues/21140 it's just 2 lines..
13:47 numkem joined #salt
13:48 Furao joined #salt
13:51 zer0def uh, is there any reason for salt-ssh states misbehaving on grains on debian?
13:51 CeBe1 joined #salt
13:52 babilen zer0def: Are they giving you the fingers?
13:53 zer0def babilen: somehow; referring to grains['oscodename'] in-state returns None, however when called from grains.items, it returns the proper codename
13:54 bhosmer joined #salt
13:54 babilen zer0def: How do you refer to it?
13:55 zer0def {{ grains['oscodename'] }} ?
13:55 babilen How do you know that it is None?
13:56 zer0def i looked up the file which was supposed to be populated with that data
13:56 zer0def sometimes it's None, sometimes it's empty string
13:57 zer0def i suppose you'd like specific states i refer to, right?
13:58 babilen It would probably allow us to figure out what happens. It might be a bug and I don't use salt-ssh a lot so I am not too familiar with its limitations, but a precide specification of the problem one is facing always helps in debugging.
13:59 zer0def hold up, just putting it up
13:59 babilen And I'm not sure if you can use grains on the minion over salt-ssh
14:00 otter768 joined #salt
14:00 JDiPierro joined #salt
14:00 babilen I would have expected them to be available, but I never really verified that
14:00 Furao joined #salt
14:01 murrdoc joined #salt
14:02 zer0def http://codepad.org/nfck5XHm
14:02 zer0def other grain usage works fine
14:02 zer0def if they wouldn't work altogether, i'd probably just get over it and find an alternative way to target those rigs
14:02 nighter joined #salt
14:03 murrdoc is oscodename a grain ?
14:03 zer0def yes
14:03 murrdoc salt 'minoin' grain.get oscodename
14:04 dude051 joined #salt
14:06 babilen zer0def: Where do you get that result? (the "dist == -pgdg")
14:07 zer0def on the same host
14:07 JDiPierro joined #salt
14:07 zer0def murrdoc: http://codepad.org/7pbrL6lb
14:07 babilen zer0def: "on the host" ? Where?
14:07 babilen be specific please
14:07 zer0def in the apt list file
14:07 zer0def that is mentioned in the pkgrepo state.
14:07 murrdoc line 38 babilen
14:07 babilen Which apt list file? What is the name of that file and what is its exact content?
14:08 murrdoc babilen:  http://codepad.org/nfck5XHm
14:08 babilen murrdoc: That doesn't help
14:08 babilen (as in: I can see that)
14:08 subsignal joined #salt
14:08 murrdoc one thing you could try is chaging it from grains{} to salt[grains.get]('oscodename', 'didntwork')
14:09 zer0def i'm not even sure what you're confused about, but (i think that) pkgrepo for debian distros requires a "dist" option
14:09 babilen So you have /etc/apt/sources.list.d/pgdg.list with "deb http://apt.postgresql.org/pub/repos/apt/ -pgdg main" ?
14:09 zer0def babilen: exactly.
14:09 zer0def murrdoc: didn't think of that, let me check
14:09 babilen What happens if you *don't* set "file: /etc/apt/sources.list.d/pgdg.list" and tried murrdoc's suggestion?
14:10 murrdoc file is neeeded
14:10 murrdoc otherwise it puts it in /etc/apt/sources.list
14:10 zer0def you need "file"
14:11 robothands personally, I use a file append to sources.list
14:11 babilen You don't need file, it only puts it all in sources.list if you consolidate
14:12 robothands http://hastebin.com/azetaruvug.django, that works for me
14:12 zer0def something i don't really want, babilen
14:12 robothands meh, nvm then
14:12 mpanetta joined #salt
14:13 babilen https://github.com/saltstack/salt/blob/develop/salt/modules/aptpkg.py#L1465
14:13 babilen I am simply interested in the name it is choosing
14:13 zer0def besides, finding workarounds doesn't really solve problems
14:13 babilen ok, forget that I asked
14:14 zer0def let me quickly reset those rigs
14:14 debian112 joined #salt
14:15 stylica joined #salt
14:15 ughloru joined #salt
14:16 ughloru Hi. I'm trying to set up salt so that when I run state.highstate, services will only restart if there have been any changes to their configuration.
14:17 Damon Is it possible witha salt module to copy a file from the salt master to multiple minions ? It's only a 1 time copy so dont wanna make a state for it. Something like salt "*" file.copy salt://my.file /root/my.file ?
14:17 babilen ughloru: Yes, that is the normal behaviour if you use watch_in or listen_in on the service states in the state that manage the configuration file(s)
14:18 linjan joined #salt
14:18 babilen Damon: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.file.html#salt.modules.file.manage_file
14:18 zer0def Damon: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.file.html#module-salt.modules.file
14:18 ughloru babilen: Maybe I'm making some mistake. I have tried 'watch' and 'require' requisites in my service.running formula, but this formula will be executed even when there are no changes to the configuration.
14:19 ughloru I've not tried it using watch_in or require_in. Maybe that will yield different results.
14:20 babilen ughloru: First and foremost: It is typically preferred to use watch_in and listen_in the the states that manage the files rather than watch or listen in the service.running state. Could you provide your states and output that exemplifies the problem on http://refheap.com please?
14:20 babilen (also: what do you mean by "this formula will be executed" ?)
14:20 mpanetta joined #salt
14:20 Damon babilen: zer0def : i know those docs, i just cant see to find the right command. file.copy doesnt seem to allow a source on salt master.
14:21 ughloru babilen: it'll take me a little while to get together a sanitized version of my config and the output. I'll come back when I have that.
14:22 babilen Damon: You would, naturally, use a salt:// source that is in your file_roots
14:22 babilen ughloru: Perfect
14:23 ughloru babilen: '(also: what do you mean by "this formula will be executed" ?)' Hmmm.... maybe I'm using the terminology wrong. I have a formula with a service.dead state, and another with a service.running state. I want these to only be enforced if some underlying configuration is changed. Is 'executed' the wrong word to use here?
14:24 babilen Oh, why do you have a service.dead state?
14:24 zer0def Damon: file.copy
14:24 mpanetta joined #salt
14:24 ughloru For that matter, I don't even know if "formula" is the right word.
14:24 nitti joined #salt
14:24 ughloru I have a dead state to ensure that the service is killed, so that the service.running call will launch a new process.
14:24 zer0def babilen: murrdoc's suggestion produces the same result
14:25 babilen ughloru: When do you want that to happen?
14:25 babilen (the killing that is)
14:25 babilen (don't quote out of context please)
14:26 zer0def as in salt['grains.get]('oscodename', 'didntwork') results in None for the nginx sls and -pgdg for pgdg sls
14:26 zer0def also, retyped that reference, there's obviously a quote mark after "get"
14:27 zer0def but then again, it should be 'didntwork-pgdg' now.
14:27 ughloru If any of my configuration has changed. i.e. all of my services are written in Python, and live in python virtualenvs. On state.highstate, a `pip.installed` will try to update the package. If the package isn't changed, then the pip state will return {'changes': {}}. In that event, I don't want the service to restart. On the other hand, if any changes are made to the virtualenv, pip.installed will return {'changes': {some stuff}}, and I
14:28 ughloru So, if I have several services living on one machine, and there are code changes to one, I want that one service to restart after a state.highstate, but none of the others should be affected.
14:29 debian112 is there a way to use node groups in states?
14:31 babilen ughloru: Your message was truncated at "{some stuff}}, and I"
14:31 ughloru babilen: Oh, sorry: "return {'changes': {some stuff}}, and I do want the service to restart [14:28] <ughloru> So, if I have several services living on one machine, and there are code changes to one, I want that one service to restart after a state.highstate, but none of the others should be affected."
14:33 babilen ughloru: You would use a "listen_in: - service: THESERVICE" or "watch_in: - service: THESERVICE" in the pip.installed state where THESERVICE is your service.running state. You would *not* use any service.dead state explicitly as that state will *always* ensure that the service isn't running.
14:33 jdesilet joined #salt
14:34 furball365 joined #salt
14:34 babilen ughloru: Salt knows how to restart your service. You don't have to teach it.
14:34 ughloru Aha, I might have been overthinking that point.
14:34 murrdoc joined #salt
14:35 lietu joined #salt
14:35 racooper joined #salt
14:36 shorty_mu left #salt
14:36 ughloru babilen: does that mean that a service.dead state will *always* kill the service, regardless of any watch_in/require_in statements attached to it?
14:37 bluenemo joined #salt
14:37 bluenemo joined #salt
14:38 babilen ughloru: If you define a service.dead state unconditionally then it will always execute. It will therefore *always* kill the process. If you want a service to be running you use "service.running".
14:38 murrdoc yeah
14:38 zer0def murrdoc: your suggestion ended in the same result, which it shouldn't
14:38 murrdoc right
14:38 zer0def since it obviously should've been "didntwork" and "didntwork-pgdg"
14:39 murrdoc next question
14:39 murrdoc does jinja work in your states ?
14:39 murrdoc maybe add an explicit !#jinja|yaml to the top of the state file
14:39 babilen ughloru: The logic of restarting service is handled by the mod_watch functionality and triggered by watch_in and listen_in (or watch/listen). If you use watch_in on a service in a file.managed and that file has changes it would restart the service automagically (as salt knows that that is the correct way to resolve a watch_in on a service)
14:39 ughloru Okay, sorry if this question seems really stupid, but I've clearly not understood. When you say "unconditionally", is that the same as "has no  watch/require requisites"?
14:39 murrdoc yes
14:40 murrdoc service.dead is like calling the one hitman
14:40 murrdoc who once u call
14:40 murrdoc u cant un call
14:40 babilen ughloru: No, I mean that you don't only include that state if you want to stop the service. As soon as you include it salt will always ensure that it is being reached.
14:40 murrdoc not that i know what hitmen are
14:40 murrdoc or how one would call them
14:40 zer0def murrdoc: yes, they do, they differentiate between distros based on grains['os'] and it properly renders grains['os']|lower
14:40 zer0def i'll try adding an explicit reference, but i'm not hopeful about this
14:40 babilen ughloru: If you have both service.running and service.dead in your SLS file it would always essentially restart the service (or start and then kill it again depending on the order)
14:41 murrdoc zer0def:
14:41 murrdoc next option would be to do explicit jinja sets at the top of the page
14:41 babilen ughloru: You don't have to script the restart in salt as that is being handled for you automatically
14:41 ughloru babilen: okay. Thanks. I'll try it again without the service.dead, and with the requisites pointing only to service.running.
14:42 zer0def murrdoc: like adding "!#jinja|yaml" on top?
14:42 murrdoc or #!
14:42 murrdoc sorry just waking up
14:42 babilen ughloru: You are welcome. Just use a watch_in or listen_in (depending on when you want to restart the service (watch_in executed immediately (and potentially multiple times) while listen_in simply restarts the service once at the end)
14:43 zer0def murrdoc: that's fine, glad for your help
14:44 ughloru I believe that listen_in is only from 2014.7.0 onwards. Due to a bug in pip_state for Windows, I'm stuck at 2014.1.13, so I'll have to use watch_in.
14:44 babilen ughloru: That is correct
14:45 ndrei joined #salt
14:46 babilen 2014.1.13 is ... ugh
14:46 babilen Which bug if I may ask?
14:47 ughloru babilen: https://github.com/saltstack/salt/issues/8405
14:47 ughloru Actually, that bug goes back to before 2014.1.10, at least
14:47 ughloru But I had a work around in place (I'm tensorproduct, in that thread)
14:48 babilen Ah, why is Windows so broken?
14:48 zer0def now i'm stumped
14:48 babilen Ah, hello tensorproduct!
14:48 ocdmw joined #salt
14:48 ughloru Right, sorry. Bad form to be using two different handles. :-P
14:48 zer0def murrdoc: running a state.sls on those files now works fine, including examples i've given before
14:49 ughloru Sometimes I forget who I am.
14:49 babilen ughloru: obviously ;)
14:49 babilen zer0def: That is really weird
14:49 murrdoc zer0def:  http://codepad.org/nfck5XHm/fork
14:49 zer0def babilen: i know, right?
14:49 murrdoc anytime i am using a grain more than once in a state i set it at the top of the file
14:49 murrdoc cos u know readability
14:49 murrdoc :D
14:50 zer0def yeah :D
14:50 babilen Sounds as if the default renderer is somehow not working how it should .. but only partially. /me is confused
14:50 zer0def in conditional blocks, murrdoc?
14:50 zer0def since rpm repos don't really like codenames
14:50 zer0def or at least centos.
14:50 ughloru But, yeah. I have a custom version of pipwin_state that works up til 2014.1.13, but something in 2014.7.0 makes it even more broken. I've not managed to fix it (nor to clean up my code enough to get the initial workaround pulled in)
14:50 babilen I mean I am all in favour of explicitly stating the renderer everywhere, but that is just wird
14:51 zer0def babilen: we all are, especially me
14:51 zer0def since it's like "wat?"
14:51 babilen indeed
14:52 zer0def and i don't like the feeling of being a dog chasing it's own tail
14:52 yoq joined #salt
14:53 zer0def oh, another run results in the codename not being put in
14:53 yoq S.a
14:53 yoq :D
14:54 babilen zer0def: Hmm, ... that's at least ... consistent ... Are you sure that you are not running of some weird cache? Have you updated grains?
14:54 hobakill joined #salt
14:54 getling joined #salt
14:55 zer0def i'm not aware if there's any cache in salt in the first place
14:55 zer0def and those grains are default ones
14:56 murrdoc there is a cache and the grains are default
14:56 murrdoc cache wont affect you here tho
14:56 murrdoc but for sanity
14:56 murrdoc salt '*' saltutil.clear_cache
14:56 murrdoc and what babilen said
14:56 murrdoc check out the page for saltutil
14:56 zer0def sure
14:57 babilen http://docs.saltstack.com/en/latest/ref/cli/salt-ssh.html#cmdoption--refresh
14:57 murrdoc u are using salt-ssh
14:58 zer0def i should probably add the option in the Saltfile to remove it's cache when it's done with a task
15:01 ughloru babilen: yep, taking out the service.dead state did the trick. Thanks! You've been a great help as usual.
15:01 babilen ughloru: Rock on! :D
15:02 ughloru I'm afraid that contrary to best practise, I've had to use watch rather than watch_in. Due to our current way of deploying virtualenvs and services, using watch_in would require more of a refactor than I can devote my time to right now.
15:02 ughloru But such is life.
15:03 babilen Use whatever is best. There is a reason why both exist .. I just find it much better to keep the "triggers" with the states that will trigger them rather than a long list of other states with the one that is being triggered
15:04 ughloru Oh, I don't disagree.
15:04 ughloru We've got a fairly crazy assemblage of jinja macros that manage a lot of our states.
15:04 Damon salt "*" file.copy salt://my.cnf /root/.my.cnf , gives a "File path must be absolute" , is it now allowed to do this ? Yes the file is in salt file_root dir
15:05 ughloru I've been trying to move this code into proper state and execution modules where appropriate.
15:05 ughloru But it's one of those never ending battles.
15:06 berto- joined #salt
15:07 germs_ joined #salt
15:08 murrdoc tt]
15:09 tvinson joined #salt
15:09 Brew joined #salt
15:10 dude051 joined #salt
15:11 andrew_v joined #salt
15:11 aparsons joined #salt
15:13 stylica joined #salt
15:14 berto- joined #salt
15:16 getling <- Brand new to salt (v. impressed so far), looking to roll out as orchestration to our teams...
15:17 getling Wondering about architectures/security, whether folks lock things down more via sudo, by binding to localhost, or by ACL
15:18 zer0def ok, so i've found out the pattern
15:19 tvinson joined #salt
15:19 zer0def murrdoc, babilen: this happens only the first time that particular state is run
15:19 zer0def regardless of cache refreshing or config wiping
15:19 _JZ_ joined #salt
15:21 murrdoc are u using salt or salt-ssh
15:21 murrdoc and what version
15:21 murrdoc u should totally make a github issue for this
15:21 zer0def salt-ssh, 2014.07.1
15:22 babilen And yeah, issue! bug! heisenbug!
15:24 hungry_bear5000 joined #salt
15:24 polar_robot_9000 joined #salt
15:25 stylica joined #salt
15:25 polar_robot_9000 Anyone ran into an issue where running a bash script, which installs some software (in this case, installs a ruby version through rbenv), does not persist after the salt process finishes high state?
15:25 babilen you?
15:26 polar_robot_9000 Everytime I run a high state, it runs my bash script, which installs a ruby version through Rbenv (confirmed in log) but the version is gone after the salt process dies, and the changes don't persist
15:26 babilen Could you show us applicable details on http://refheap.com ? (state, output, ...) -- Also: Why do you think that it shouldn't behave that way?
15:27 polar_robot_9000 Sure, will do, and thanks
15:27 murrdoc babilen:  laying down the law
15:27 * murrdoc likes it
15:28 polar_robot_9000 Also, for instance, I'm exporting that path salt needs (export PATH='blah blah blah') and the install works, and it's downloading a file, so while I expect that the shell session will and its exports will end, I'm surprised that a file it downloaded doesn't persist.  One sec for those details though.
15:31 elfixit joined #salt
15:31 polar_robot_9000 hmmm... you know what, I think the build is failing ;) This might be my fault. My bad :)
15:34 Eureka_ joined #salt
15:36 XenophonF joined #salt
15:36 XenophonF howdy
15:36 esogas_ joined #salt
15:36 XenophonF i'm getting SSL errors when the Windows minion tries to download stuff from a https URL
15:37 XenophonF this is version 2014.7.1
15:37 XenophonF give me a sec and i'll have a traceback posted for your pleasure
15:38 XenophonF https://bpaste.net/show/ab3a56b5e694
15:38 seanz joined #salt
15:38 babilen polar_robot_9000: #salt will provide complimentary rubber ducks after your twentieth question ;)
15:38 XenophonF i stow installers and stuff in S3,
15:38 XenophonF i could use http URLs, but i wouldn't like it
15:41 blackhelmet joined #salt
15:43 XenophonF perhaps i'm better off using the s3 URLs and trying to configure read-only access for minions via API key
15:52 bhosmer_ joined #salt
15:53 gladiatr joined #salt
15:53 gladiatr joined #salt
15:55 esogas_ joined #salt
15:56 clintberry joined #salt
15:56 iggy murrdoc: If I didn't know people from my work are going to go back and watch this, I totally would
15:57 murrdoc :)
15:58 ecdhe Has anyone installed MS Office on windows with salt?
15:59 Morbus joined #salt
15:59 smcquay joined #salt
16:00 otter768 joined #salt
16:01 ALLmightySPIFF joined #salt
16:01 ecdhe I know thatch was excited years ago when he got Firefox to install on windows, unattended...
16:01 ecdhe But I haven't seen anyone installing MSO that way.
16:01 murrdoc thats impressive
16:02 hobakill ecdhe, basic packages is as far as i've gotten on windows. not office.
16:03 ecdhe hobakill, any notes I could look off of?
16:03 hobakill i could post my windows repo somewhere if you'd like.
16:03 ndrei joined #salt
16:04 ecdhe if it's not too hard to sanitize, I'd appreciate it!
16:05 hobakill ecdhe, lemme check. otherwise you should look into chocolatey <sp?>
16:05 hobakill it's far less of a pain in the ass then a winrepo
16:05 hobakill (IMHO)
16:06 ecdhe hobakill, I hadn't heard of that.  Thanks!
16:06 NET||abuse joined #salt
16:07 hobakill oh yeah man. check out the docs .... http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.chocolatey.html
16:07 hobakill it doesn't have everything by a long shot but it has a lot of the essentials. for the most part it has let me ignore the winrepo i made.
16:08 JDog joined #salt
16:08 victor- joined #salt
16:09 anotherZero joined #salt
16:10 NET||abuse joined #salt
16:11 ocdmw joined #salt
16:13 ocdmw joined #salt
16:14 jmccree So apparently the salt iptables policy state does not respect test=True
16:14 ecdhe jmccree, what os are you on?
16:14 jmccree Ubuntu
16:15 jmccree It's kind of sucky, because it set policy to DROP without running the ones to let in ssh etc.
16:16 brianfeister joined #salt
16:16 SheetiS joined #salt
16:16 iggy the iptables state modules are basically crap (for most distros)
16:18 jmccree A giant "WARNING DONT USE THIS" might be a recommended thing to add the docs. ;)
16:18 schristensen joined #salt
16:19 ecdhe the issue I had with iptables was non-idempotency on CentOS...  the same rules would just get added repeatedly every time I enforced highstate.
16:19 iggy I could tell you where the issue tracker was, but I feel like you're smart enough to find it
16:20 jmccree Yeah, searching now to see if this is already known.
16:22 SheetiS I'd be curious to see how the module is verifying if an iptables rule is already applied.  I might look @ the source.
16:22 aparsons joined #salt
16:23 jmccree Well for rules it apparently is using -C check which works right. But for the policy it runs the command for some reason, even under test=True
16:23 jmccree Which means it bricks the vm.
16:24 StDiluted joined #salt
16:25 kanaka_ joined #salt
16:27 jeremyr joined #salt
16:28 edrocks joined #salt
16:28 aparsons joined #salt
16:33 iggy Sounds bad
16:33 iggy Who gets into SLC today?
16:36 perfectsine joined #salt
16:37 nicksloan my master's IP address changed. Any chance my minions will pick up on that automatically?
16:38 iggy if you used domain names
16:39 iggy or some sort of name resolution
16:39 iggy I think the DNS lookup timeout is 30 mins?
16:40 iggy so it'll be a long wait
16:40 rypeck joined #salt
16:44 Ahrotahntee huh, there is a case sensitivity bug with salt-cloud & ssh keys
16:44 Ahrotahntee oh no wait, still not working. something else must be wrong
16:45 tligda joined #salt
16:50 brianfeister joined #salt
16:52 tligda1 joined #salt
16:53 racooper joined #salt
16:53 aparsons joined #salt
16:54 brianfei_ joined #salt
16:55 aparsons joined #salt
16:56 aparsons joined #salt
17:00 tligda joined #salt
17:00 ndrei joined #salt
17:05 diegows joined #salt
17:10 murrdoc still not a fan of not being able to include pillars for formulas
17:10 murrdoc only when using a formula
17:11 jmccree iggy, Just an FYI, I found the bug I experienced has already been patched in a newer version. https://github.com/saltstack/salt/commit/97ef8e45aa74f4d9d7d262c3029a11d833cfecb3
17:12 iggy jmccree: oh, nice
17:12 iggy speaking of versions... did 2015.2 get renamed to 2015.3?
17:14 murrdoc :D
17:14 murrdoc are u waving your 'called it' flag yet
17:16 iggy no... I'd never...
17:17 iggy Just curious
17:17 conan_the_destro joined #salt
17:18 murrdoc salt devs right now http://i.imgur.com/i6l470Z.jpg
17:19 MatthewsFace joined #salt
17:22 iggy I'd post a gif of a chicken running around with it's head cut off, but that might be a tad gruesome
17:22 SheetiS salt is always growing and evolving so much/fast that consolidating to get a release out the door would be challenging at best imo.
17:22 Brick__ joined #salt
17:25 iggy We know, we just like giving the devs a hard time
17:25 aparsons joined #salt
17:25 iggy (and I hope they know that)
17:26 murrdoc i think they think that u think u are a wiseass
17:26 iggy I am
17:27 murrdoc SheetiS:  just didnt know
17:27 murrdoc also that its all love
17:28 SheetiS lol I know.  I'm just amazed they can get releases out at all with all the chaos :)
17:29 KyleG joined #salt
17:29 KyleG joined #salt
17:29 iggy We've had this running dialog with them about the weird release process
17:30 iggy now that you mention it, I should probably drunkenly apologize to the pi of base this weekend
17:30 lifelearner joined #salt
17:31 lifelearner If I want to compare between different CM solutions, what should be the criteria? Easiness of use, popularity, etc.?
17:32 shaggy_surfer joined #salt
17:32 ecdhe lifelearner, corporate support, e.g, if you run into trouble, is it an option to pay for help?
17:33 iggy scope, language
17:33 fsero joined #salt
17:33 fsero hi everybody :)
17:33 iggy method of function (push/event driven, vs agent's pulling periodically)
17:34 fsero i have a problem with salt-ssh, my current version is salt --version
17:34 fsero salt 2014.7.0 (Helium)
17:34 shaggy_surfer joined #salt
17:34 fsero and when i do a file.recurse on a folder i got this
17:34 fsero Function: file.managed
17:34 fsero Result: False
17:34 fsero Comment: An exception occurred in this state: Traceback (most recent call last):
17:34 fsero File "/tmp/.root_982293__salt/salt/state.py", line 1533, in call
17:34 fsero **cdata['kwargs'])
17:34 fsero File "/tmp/.root_982293__salt/salt/states/file.py", line 1316, in managed
17:34 fsero __env__
17:34 fsero File "/tmp/.root_982293__salt/salt/modules/file.py", line 2375, in source_list
17:34 fsero mfiles = __salt__['cp.list_master'](saltenv)
17:34 fsero File "/tmp/.root_982293__salt/salt/modules/cp.py", line 505, in list_master
17:34 fsero return __context__['cp.fileclient'].file_list(saltenv, prefix)
17:35 fsero File "/tmp/.root_982293__salt/salt/fileclient.py", line 1058, in file_list
17:35 fsero return channel.send(load)
17:35 fsero File "/tmp/.root_982293__salt/salt/fileserver/__init__.py", line 571, in send
17:35 fsero return getattr(self.fs, cmd)(load)
17:35 fsero File "/tmp/.root_982293__salt/salt/fileserver/__init__.py", line 455, in file_list
17:35 fsero ret.update(self.servers[fstr](load))
17:35 fsero File "/tmp/.root_982293__salt/salt/fileserver/roots.py", line 300, in file_list
17:35 iggy whoah... refheap.com
17:35 fsero return _file_lists(load, 'files')
17:35 fsero File "/tmp/.root_982293__salt/salt/fileserver/roots.py", line 288, in _file_lists
17:35 fsero __opts__, ret, list_cache, w_lock
17:35 fsero File "/tmp/.root_982293__salt/salt/fileserver/__init__.py", line 147, in write_file_list_cache
17:35 fsero fp_.write(serial.dumps(data))
17:35 fsero File "/tmp/.root_982293__salt/salt/payload.py", line 115, in dumps
17:35 fsero return msgpack.dumps(msg)
17:35 fsero NameError: global name 'msgpack' is not defined
17:35 fsero sorry is a file.managed
17:35 fsero any ideas?
17:36 fsero sorry :(
17:36 iggy fsero: this is what refheap/gist were made for
17:36 clintber_ joined #salt
17:36 anotherZero ^
17:36 jalbretsen joined #salt
17:36 fsero yeah yeah iggy you are absolutely right
17:37 fsero sorry about that
17:37 fsero :)
17:37 fsero https://www.refheap.com/98022 here you go
17:37 murrdoc ./kick fsero
17:37 iggy you said file.recurse, that says file.managed
17:37 fsero thanks in advance!
17:37 fsero yeah i got confused, it is file.managed
17:37 lifelearner OK - great ideas, thanks! ecdhe & iggy!
17:37 iggy fsero: paste your state too
17:37 iggy (this is where gist is stronger... it allows multiple files per paste)
17:38 iggy but this looks vaguely like a version mismatch?
17:39 iggy or something msgpack is something you see when zmq gets wonky
17:39 lifelearner iggy: Most of them are pushing from the master to clients, right? Any of them using the pull method?
17:39 fsero iggy here you go
17:39 fsero https://gist.github.com/Fsero/af3331fed2a40dfb16d6
17:39 fsero mmm
17:40 iggy lifelearner: chef and puppet both poll afaik
17:40 nicksloan iggy: you're certain about minions reconnecting to a master on a different IP address as long as the domain name resolves?
17:41 lifelearner OK.
17:41 fsero iggy, i got the exact same version on both sides for zmq ii  python-zmq                           13.1.0-1~bpo70~dst+1          i386         Python bindings for 0MQ library
17:41 iggy nicksloan: nope... I thought I saw something in the config file about name lookup timeout
17:41 Ahrotahntee salt-cloud is awesome
17:41 iggy fsero: zmq doesn't get used in salt-ssh
17:42 iggy fsero: whcihc is the weird part
17:42 lifelearner I think Cfengine also pulls, right?
17:42 iggy fsero: - template: jinja
17:43 iggy fsero: I don't think that's your problem, but it's _a_ problem
17:44 fsero iggy, iep  you are right i just modified the file and it gives me the same output
17:44 fsero with the inclusion of template:jinja
17:45 iggy yeah, like said, not your problem, but a problem
17:46 Konijntjes joined #salt
17:47 lifelearner As long as we use Community Editions with no support, each of those 5 CM solutions is free for unlimited number of nodes, right?
17:47 desposo joined #salt
17:47 iggy fsero: you'll probably want to hit the mailing list or wait around for someone that's more knowledgeable in that area... from what I understand, with salt-ssh, you shouldn't be hitting msgpack code paths at all
17:49 fsero ok, iggy thanks for your time :) i'll be do that
17:49 Ryan_Lane joined #salt
17:49 fsero im sorry for the paste again, murrdoc :)
17:49 murrdoc u better BE
17:49 murrdoc (totes kidding)
17:53 SheetiS :P
17:54 iggy we're apparently a rough crowd sometimes
17:55 fsero iggy, i've just updated to tha last rc and the error is fixed apparently
17:55 fsero :)
17:56 SheetiS fsero: I saw this just before you said that https://github.com/saltstack/salt/issues/20276.  Fairly recent commit that addresses some salt-ssh msgpack issues.
17:57 SheetiS I would guess https://github.com/saltstack/salt/pull/20304/files is your fix.
17:57 netzmonster joined #salt
17:57 sfxandy joined #salt
17:59 Gareth morning morning
18:00 murrdoc :)
18:00 murrdoc o/
18:00 SheetiS :D
18:01 otter768 joined #salt
18:02 Guest45602 hi guys
18:03 hackel_ joined #salt
18:03 sfxandy_ joined #salt
18:04 SheetiS wb sfxandy_
18:04 sfxandy_ flaming IRC client!!
18:04 JDiPierro joined #salt
18:05 sfxandy_ have been puzzling over a "how do i do that" for a few hours with no luck
18:05 SheetiS What's your thought?  We'll see if I can steer you off on a wrong tangent and everyone else here can help out :)
18:06 sfxandy_ i'd like to be able to iterate over all of the hosts running salt minions and based on some grain information, produce a templated file
18:06 SheetiS so you want a file.managed on one (or more) minion(s) that contains data about multiple minions?
18:07 sfxandy_ essentially yes ....
18:07 SheetiS the salt mine is what would be used to handle that kind of magic.  http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.mine.html
18:08 SheetiS if you had more specifically what you wanted to do, I might have a more specific example.
18:08 sfxandy_ hmm, never even heard of the salt mine before :/
18:09 sfxandy_ ok, so.  i have a file in /etc/salt called grains in which - for each minion - I specify a few items of information.  server role, etc.
18:10 murrdoc thats not the right place
18:10 murrdoc is it
18:10 sfxandy_ well it finds the grains ok so it seems the right place...
18:11 murrdoc nice
18:11 sfxandy_ lets say i have a grain called "roles"
18:11 SheetiS /etc/salt/grains definitely works.  The bad part about relying on /etc/salt/grains though is that you have to trust that the data the minion is giving you is correct which is inherently insecure.
18:11 sfxandy_ well i was thinking of moving it to Pillar...
18:12 spookah joined #salt
18:12 SheetiS You definitely do that, but in any case continue with your roles example and what you want to achieve.
18:12 sfxandy_ ok ....
18:12 sfxandy_ so i have a grain called "roles" which could is basically a list of roles the server can fulfil
18:13 Andre-B joined #salt
18:13 SheetiS (I've used grains for roles myself for months and only migrated to a pillar model this weekend (and I knew it was insecure the whole time))
18:14 sfxandy_ and for any server that has matches  the role "foo-*" (so it matches foo-head and foo-node roles), I want to have a file that contains the FQDN of each host that mathces
18:14 sfxandy_ mathces
18:14 sfxandy_ matches
18:15 wendall911 joined #salt
18:16 SheetiS ok so looking at the link I gave on the mine, you could add grains.item (roles) and network.get_hostname to your mine functions
18:17 sfxandy_ ok....
18:17 SheetiS or you could skip network.get_hostname if all of your minion ids are their fqdn
18:18 cpowell joined #salt
18:18 SheetiS Then you could use some jinja like this {% for host, roles in salt[
18:18 sfxandy_ no - minion IDs are a concatentaiton of environment ID, short hostname and something else
18:18 SheetiS err
18:18 murrdoc mine is the way
18:18 SheetiS {% for host, roles in salt['mine.get']('foo-*', 'grains.item') %}
18:18 SheetiS and iterate from tere
18:19 nullptr` joined #salt
18:19 SheetiS I could give a full example, but I have a meeting for a little bit here
18:19 * sfxandy_ frantically tries to comprehend what SheetIS has written
18:19 murrdoc http://docs.saltstack.com/en/latest/topics/mine/ read
18:20 sfxandy_ i'll experiment with mine and your starter for 10 above and see how I get on.  anticipate more questions though!! but thanks SheetIS
18:20 eichiro joined #salt
18:21 jngd joined #salt
18:22 bhosmer__ joined #salt
18:27 druonysuse joined #salt
18:27 druonysuse joined #salt
18:29 darien joined #salt
18:31 jeremyr1 joined #salt
18:32 aparsons joined #salt
18:37 lothiraldan joined #salt
18:38 forrest joined #salt
18:39 nullptr` joined #salt
18:42 I3olle joined #salt
18:43 davet joined #salt
18:43 druonysus joined #salt
18:43 druonysus joined #salt
18:43 ajw0100 joined #salt
18:43 baweaver joined #salt
18:45 aparsons joined #salt
18:45 ajw0100_ joined #salt
18:46 juso joined #salt
18:47 bash124512 joined #salt
18:48 juso Hi guys. Have a question, trying to download a file via http from remote host, but get an error 'Unable to determine upstream hash of source file'. I assume hash code is required to determine if the file is still valid or should be replaced (re-downloaded), but I assume it cannot get the hash code from the remote host. What can be done in this situation?
18:51 juso can I make file.manage to ignore the hash issue?
18:51 aparsons joined #salt
18:51 juso meant 'file.managed'
18:53 nitti joined #salt
18:53 stoogenmeyer joined #salt
18:54 lifelearner joined #salt
18:57 sfxandy joined #salt
18:57 sfxandy hi folks ... again!
18:59 murrdoc gawd this dude
19:00 sfxandy hope you're not talking about me murrdoc!
19:00 pacopablo I am trying to install a file in a state and it's giving me a 'file.installed' found in SLS '...' is unavilable
19:00 pacopablo anyone have an idea what I'm doing wrong?
19:00 murrdoc juso:  no clue, but maybe try cp instead of file.managed
19:00 pacopablo https://gist.github.com/pacopablo/644bc57d8aea8863a8b4
19:01 murrdoc juso: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.cp.html#salt.modules.cp.get_url
19:01 juso murrdoc: thank you, will give it a try
19:02 economoum joined #salt
19:02 pacopablo I am assuming that it can't find the file in my salt:// line.  is there a way to test if it can access that file?
19:02 XenophonF hey all the windows minion is throwing sslerrors when i try to use https urls in my formulas
19:03 XenophonF has anyone else seen this?
19:03 XenophonF https://bpaste.net/show/ab3a56b5e694
19:04 XenophonF one of the package definitions looks like this: https://bpaste.net/show/e1913fb9c871
19:06 XenophonF i can download the files anonymously using wget, so i'm pretty confident it isn't a permissions problem
19:08 brianfeister joined #salt
19:09 XenophonF do i need to install a CA bundle or something?
19:10 SheetiS XenophonF: I would guess that the system has an out-of-date CA bundle, yes
19:13 bash1245_ joined #salt
19:13 chiui joined #salt
19:14 _prime_ joined #salt
19:15 pacopablo well, apparently I'm an idiot.  file.installed doesn't exist.  I needed to use file.managed
19:16 linjan joined #salt
19:17 mordonez joined #salt
19:18 lifelearner Anyone knows what the Saltstack standard support (vs. enterprise/premium) costs per node per year?
19:18 brianfei_ joined #salt
19:22 ckao joined #salt
19:22 tmwsiy joined #salt
19:22 lifelearner So far, from my research, support contract for Puppet seems the most expensive ($100/node/year).
19:22 pacopablo lifelearner: I don't know, but I'm guessin that it depends on the number of nodes you have
19:23 pacopablo you could email them and find out
19:23 lifelearner And, Ansible($50/node/year) seems to be the least expensive... for the same # of nodes for apples-to-apples comparison.
19:24 lifelearner (Not counting SaltStack for which I don't know yet)
19:24 XenophonF SheetiS: does the Salt minion use the Windows CA bundle (i.e., via CryptoAPI)?
19:25 Brick__ joined #salt
19:26 XenophonF lifelearner: i got pricing for Saltstack Enterprise a while ago, but i'm sure it's outdated by now
19:26 murrdoc joined #salt
19:26 SheetiS hmm I am not sure about python on windows.
19:26 XenophonF SheetiS: this is a new problem - i don't recall having this issue a month ago, but i'm not sure when it started happening
19:26 hal58th joined #salt
19:29 anotherZero anybody know how I can set a superglobal var in apache/fcgi ?  Like, if I wanted to set $_SERVER['YOUR_DADDY'] = $_SERVER['PHP_AUTH_USER']
19:29 teepark joined #salt
19:29 SheetiS XenophonF: On linux, I'd say to make sure your os was updating /etc/ssl/certs/ca-bundle.crt, but I am not sure how windows python handles it.
19:30 teepark joined #salt
19:30 SheetiS It'd be in the requests python module, I am almost certain.
19:31 XenophonF SheetiS: thanks - i'm rtfs-ing now
19:32 teepark joined #salt
19:35 lifelearner SheetiS - Do you manage Windows servers as well using Salt?
19:35 SheetiS lifelearner: not at work, but I have used windows minions in my test network
19:36 irctc018 joined #salt
19:36 SheetiS XenophonF: I am getting an error message trying to follow the ssl chain from the command line as well: https://bpaste.net/show/fc76530c5e0f
19:36 lifelearner Is it free to manage Windows minions for Salt Community Edition?
19:37 SheetiS It shows a chain, but has a problem with the parent ca cert
19:37 baweaver joined #salt
19:38 SheetiS lifelearner: you should be able manage windows minions with free version http://docs.saltstack.com/en/latest/topics/installation/windows.html is the link to installers.
19:39 stoogenmeyer joined #salt
19:39 jeremyr joined #salt
19:40 lifelearner Thanks, SheetiS - Not that I'm administering any Windows servers, I'm just curious.
19:41 SheetiS The windows installers basically just bundle python + the salt bits to make it easy for you.
19:41 kanaka_ joined #salt
19:41 lifelearner Some solutions like Cfengine, they charge for native Windows clients (although free for the community edition to run on top of Cygwin)
19:47 lifelearner Would it be correct if the ages of the invention are like this: Cfengine --> Puppet --> Chef --> Ansible --> Salt ?
19:47 lifelearner (Not necessarily the maturity)
19:50 murrdoc Puppet --> Chef  around the same time
19:50 jdesilet joined #salt
19:51 lifelearner OK, somehow I thought Chef was inspired by Puppet.
19:51 SneakyPhil there's also for loops and ssh ;)
19:51 lifelearner I've seen it somewhere.
19:51 LtLefse lifelearner: you are right, Puppet predates chef
19:52 SheetiS Once upon an expect-and-telnet loop...
19:52 SneakyPhil hahahaha
19:54 lifelearner If I want to run various UNIX commands on remote hosts, Salt is the easiest alternative to replace scripting (bash/perl/expect), right?
19:54 SheetiS salt '*' cmd.run "something" is definitely a thing.
19:55 douglassssssssss joined #salt
19:55 SheetiS I try and define states for as much as I can so I don't have to remember what I have done w/ cmd.run and what I have not
19:55 LtLefse I'd say it's easier than trying to do that with Puppet (can't speak for chef or ansible)
19:55 LtLefse if you really just want an easy way to fire off unix commands, try pdsh
19:55 lifelearner SheetiS - whereas we should edit files and wait with all the other CMs, right?
19:56 hal58th1 joined #salt
19:56 davet joined #salt
19:57 lifelearner LtLefse - PDSH seems to be a great tool! Thanks.
19:58 lifelearner I used to use "global" before... seems similar.
19:58 clintberry joined #salt
19:59 LtLefse and pdcp for pushing out files
20:00 LtLefse but on the other hand, +1 to what SheetiS said
20:01 jtang joined #salt
20:02 LtLefse keeping everything in idempotent states is must better in the long run
20:02 lifelearner Both seem great tools from the same origin! I'll start using them as well.
20:02 otter768 joined #salt
20:03 Aikar Anyone know why minions cant see other minions in the mine? The minion can see the masters mine information (IP), but not other minions
20:04 baweaver joined #salt
20:04 Aikar oddly it was working in Vagrant, but not working in next tier up
20:07 murrdoc isnt there an acl for that
20:08 murrdoc for minion mine access
20:08 dimeshake can I set ext_job_cache to syslog, and still retain local job cache?
20:08 Aikar it actually just populated, seems to of been an issue with the mine interval
20:10 iggy Aikar: mine_interval is 1hr by default
20:10 Aikar after a mine cache flush, salt '*' mine.update should sync all and within 60s have access to all?
20:11 Aikar we did set it to 60 now and also just found the pillar refresh command
20:11 Aikar wondering if it had a slower mine interval cached
20:12 karimb joined #salt
20:12 iggy I've seen weird things with the mine updating... ours is set to 5mins and we will often wait a half hour or more to see new hosts show up in mine data
20:12 Aikar is interval in seconds or minutes?
20:12 dimeshake to answer my own question, you can, but you must also set job_cache: True because setting an ext_job_cache disables it by default
20:12 Aikar mine_interval:60 is 1 min ?
20:12 iggy I've always just figured I was doing something wrong
20:14 iggy Aikar: it's in minutes
20:15 JDiPierro Any way to use salt-call and the file module to grab a file from gitfs?
20:15 iggy the scheduler only runs every 60 seconds, so sub-minute intervals doesn't make sense
20:15 murrdoc the pillar based scheduler is just amazing
20:15 murrdoc and like dangerous
20:16 iggy JDiPierro: the master grabs files from git, not the minion
20:16 numkem I'm trying to setup roles into my infrastructure and I'm wondering if I should use grains or pillar to do that
20:16 iggy JDiPierro: if the master already has them fetched, you can fetch them from the master
20:16 iggy numkem: depends if you feel like you can trust your minions
20:16 numkem iggy: by trust you mean people wouldn't just change them at will
20:17 JDiPierro iggy: How do you fetch them from the master? Tried doing a file.copy with a salt:// address and it didn't seem to like that.
20:17 MatthewsFace joined #salt
20:18 MindDrive joined #salt
20:19 MatthewsFace joined #salt
20:19 iggy numkem: correct... think if someone could change grains, they could possibly get access to pillar data they shouldn't have
20:19 iggy numkem: i.e. add web to roles and get ssl secret keys
20:20 numkem iggy: It would indeed be the case... In this case pillar would be the best way than. My problem comes where I would need to match the beginning of a hostname inside jinja and I can't find doc about that
20:20 iggy cp.get_file ?
20:20 dimeshake i use roles via grains, but i don't send anything sensitive via roles to protect from that. i use pillar to assign the sensitive stuff
20:20 ajw0100 joined #salt
20:20 murrdoc cp.get_file for the win
20:20 murrdoc cp.cache_file even better
20:21 iggy dimeshake: how do you assign pillars?
20:21 iggy numkem: {% if "prenamething" in salt['grains.get']('nodename') %}
20:22 numkem iggy: i keep forgetting about in... thanks!
20:22 iggy numkem: yeah, works for simple sub-string searches
20:22 dimeshake iggy: by machine id at the moment. but my infra is kinda different.
20:22 numkem iggy: wouldn't it be easier to do this with an external piller with files?
20:22 dimeshake we have many different clients and so i want to be specific about that sort of thing
20:22 evilrob joined #salt
20:23 iggy dimeshake: cool, just as long as you don't use grains in your pillar/top.sls
20:23 dimeshake but a client may have many machines that would share, ie, ssl keys
20:23 perfectsine joined #salt
20:23 iggy some people don't think that deep into it
20:24 murrdoc "as long as you don't use grains in your pillar/top.sls"
20:24 murrdoc i use a grain in my pillar/top.sls
20:24 murrdoc if grain.get('servertype')|length
20:24 murrdoc - {{ grain.get('servertype') }}
20:24 iggy cool, just as long as you are careful about your use of grains in your pillar/top.sls
20:24 iggy better?
20:24 murrdoc i just wanted to know why not
20:25 sdm24 joined #salt
20:25 iggy you knew what I meant -_-
20:25 murrdoc :| sorry
20:28 sdm24 I have an issue trying to get the 'list_nodes_min' function to work with my vsphere provider. When I try "salt-cloud -f list_nodes_min vsphere", I get "error running the function: list_nodes_min() takes no arguments (1 given)"
20:29 sdm24 When I try "salt-cloud list_nodes_min" I get an error "Nothing was done. Using the proper arguments?"
20:29 sdm24 However, when I try "salt-cloud -Q" or "salt-cloud -F" they are both able to query and return a list of my nodes
20:30 sdm24 Is there a shorthand command for a min query? I tried searching online and could not find one
20:31 SergeantWagner joined #salt
20:31 dimeshake sdm24: that seems to work for me using ec2 provider
20:32 SergeantWagner Hello
20:32 iggy shouldn't it be -p vshpere?
20:33 SergeantWagner Is anyone still working on Halite?
20:33 sdm24 I have tried "salt-cloud -f list_nodes_min -p sphere" and get an error "--function expects two arguements: <function-name> <provider>"
20:34 samnmax joined #salt
20:35 Ahrotahntee isn't -p provision?
20:35 aquinas joined #salt
20:35 getling joined #salt
20:36 getling So I'm setting up a salt master/syndic hierarchy
20:36 sdm24 and if I try "salt-cloud llist_nodes_min -p vsphere", I get an error "Profile vsphere is not defined"
20:36 getling When using syndic, should the minions be registered to only one master/syndic, or all?
20:39 hal58th joined #salt
20:40 hal58th joined #salt
20:41 jeremyr joined #salt
20:44 dimeshake -p is indeed provision
20:46 numkem does anyone have more documentation regarding the redismod external_pillar? The docs I have only state that I have to put a function with values. Any examples?
20:48 sdm24 I tried using "salt hostname cloud.query", where hostname is the same of the salt master and minion (both are installed on the same machine). Whether I run "cloud.query", "cloud.full_query", "cloud.query list_nodes_full", etc, it only returns the hostname of each server followed by "True" underneath
20:50 sdm24 However using "salt-cloud -Q" and "salt-cloud -F" returns the proper results
20:50 viq joined #salt
20:51 shaggy_surfer joined #salt
20:54 babilen getling: Only one AFAIUI
20:54 MTecknology iggy: Guess where I am!
20:56 fsero_ joined #salt
20:56 iggy at the airport?
20:59 babilen iggy: Did you sort out your presentation?
20:59 getling babilen: Keys get passed through to the master of masters, yes?
20:59 babilen getling: master of masters simply talks to other masters, but tias ;)
21:00 iggy babilen: uhh... mostly... I'm still missing a slide that I can't remember what it was
21:00 iggy and I still need to tweak some bits of it, but it's at least pretty close to what I had before (and I was only up till 5am doing it)
21:00 babilen iggy: One could argue that this slide might not have been that important. Good job, but stressful and unnecessary. Which program caused this?
21:01 iggy calligra
21:01 iggy stage
21:01 iggy or whatever it is
21:01 getling babilen: so salt-key -L run on the syndic (master) should or should not show the key of a registered minion?
21:01 babilen getling: I don't know
21:01 MTecknology iggy: I'm in my temporary residence.
21:02 babilen getling: My understanding was that you could talk to all minions, but that you don't necessarily manage all their keys on MoM
21:03 getling babilen: OK, good then I am seeing correct behaviour
21:03 auser joined #salt
21:03 getling I got thrown by the output of salt-key -L on the syndic-master
21:04 auser left #salt
21:04 getling babilen: hm, maybe not ok. unable to issue commands
21:04 getling from the syndic
21:10 getling babilen: OK, found a blocking config. Registered through the syndic, accepted key from the syndic...now the MoM cant see it...yay.
21:12 subsignal joined #salt
21:12 Aikar is there no file.chown state?
21:13 Aikar namely wanting to touch+chown for log file access
21:13 Aikar recommendations?
21:14 pdayton joined #salt
21:15 dimeshake you can enforce permissions on files and directories with file.managed and file.directory
21:15 Aikar right but for a log file i dont want it to wipe out the contents
21:15 dimeshake it won't if it exists
21:15 dimeshake just don't specify a source
21:15 getling babilen: Which is actually a version and key issue...2014.7.2 landed on my minion
21:15 getling babilen: thanks for the help
21:15 dimeshake test it first if you are uneasy :)
21:16 Aikar yeah we have a vagrant environment to test it all
21:19 baweaver joined #salt
21:21 bauser joined #salt
21:23 baweaver Is there a way to hit a salt minion using its IP address instead of its hostname?
21:24 dimeshake baweaver: you could use the ip grain
21:24 dimeshake salt -G 'ipv4:x.x.x.x' test.ping
21:24 perfectsine joined #salt
21:25 LtLefse or Compound matching with "S@"
21:25 LtLefse salt -C 'S@192.168.1.0/24'
21:25 baweaver Yeah, just saw that. Thanks!
21:26 baweaver Trying to get a host name out of AWS is a pain it seems.
21:26 dimeshake you do have to set an aws hostname yourself, yeah
21:27 adrianhannah joined #salt
21:27 baweaver they're set, but getting them isn't fun with the way it was implemented here.
21:27 daschatten joined #salt
21:28 LtLefse can't you just "host <ipaddr>"
21:28 baweaver odd security settings make it a pain
21:29 baweaver the problem with locking everything down is fixing a bug
21:29 baweaver Anyone know how I could target a grain using the Rest API?
21:31 baweaver That one's going to be a bit more fun
21:31 druonysuse joined #salt
21:31 baweaver I hadn't seen anything in the saltnado code that looked like it would to that, though hopefully I missed something.
21:31 iggy didn't we cover this the other day?
21:31 baweaver different issue on that one. I was trying to send commands and it was mad that I put it in an array
21:32 baweaver as far as specifying it's targeting on grains, looking into that one.
21:32 baweaver expr_form
21:32 baweaver that might be it.
21:32 iggy probably
21:33 baweaver default is glob
21:33 baweaver let's see if it works with grain
21:33 AnxiousGarlic joined #salt
21:34 AnxiousGarlic left #salt
21:36 baweaver Whenever I figure this out I'll write it out a bit and push it back as documentation for the APIs
21:37 wincyj joined #salt
21:38 wincyj hello why is vars from pillar put in php file without replacement names with values of that vars?
21:39 ajw0100 joined #salt
21:39 rap424 joined #salt
21:41 iggy wincyj: best to post your states, and the output (and what you expect if it isn't obvious)
21:42 murrdoc we should make that part of the topic
21:42 iggy It's in my talk ;)
21:43 murrdoc all coming to #saltconf, remember to bring iggy a beer or 6
21:43 murrdoc as a thank u
21:44 signull iggy: is your name Brian or Sagi perhaps?
21:44 iggy Brian
21:44 signull iggy: does your last name begin with an H
21:44 iggy but nobody calls me that... I even tried to get work to put Iggy on my business cards
21:45 iggy Jackson
21:45 signull ahh ok
21:45 signull I had an old coworker from like a decade ago who's handle was digi
21:45 signull his name was Brian too
21:45 iggy my nick has been iggy since high school
21:46 iggy (which was sadly more than a decade ago)
21:46 signull same here :-/ not by much though
21:47 daschatten anyone responsible for saltstack-formulas on github here? i would like to ask some questions about when and how to transfer formulas to saltstack-formulas organization.
21:49 babilen Just ask
21:50 iggy wait till next week
21:50 iggy j/k
21:50 babilen heh
21:51 daschatten hrhr ok, when should a formula be transferred to github organization? are there any rules?
21:51 babilen 1. They must be perfect
21:51 iggy when you think it'd be useful for others?
21:51 daschatten ok, 1 is done :-)
21:52 babilen Wonderful, which formula is it and will it be of use to somebody else? Do you plan to maintain it in the future?
21:52 babilen I mean there really aren't any strict rules. If a formula is in a decent state and other people might find it useful I see no reason not to include it
21:53 babilen It's nice if people provide some ongoing support for it and ensure that it covers multiple distributions/releases and that it isn't too opinionated
21:53 auser left #salt
21:54 Deevolution Anyone running multi-master Salt answer a question or two?
21:54 jhauser joined #salt
21:54 babilen ... just ask
21:54 iggy support for multiple distros usually isn't the case right off the bat (unless you are in an env where you actually use more than 1)
21:54 Deevolution I've got a Vagrant setup of 2 masters and 2 minions for testing.  When I issue a test.ping against all nodes it's back in 800 ms on either master.  When I stop a master and do the same thing it now takes 45 seconds from the remaining master.
21:55 babilen Sure, it's by no means a must. Just a(nother) nice to have
21:55 Deevolution Can anyone explain why that would be?
21:55 pacopablo anyone have experience with the docker.io module?
21:55 babilen pacopablo: I'm sure you are the only one
21:55 daschatten i created some formulas (https://github.com/search?q=user%3Adaschatten+form) and one of them, asterisk, has also been forked to saltstack-formulas. yesterday i got an invitation to saltstack-formulas organization and accepted it. so now i'm wondering what to do with my formulas now. in my oppinion transferring them (at least the production ones) should be the goal, or not?
21:56 pacopablo babilen: sweet, so I'll ask my question so that I can answer it later
21:56 babilen pacopablo: Don't forget about time travelling and answering your own question before you asked it
21:56 iggy daschatten: if they'd be useful for others... do it... makes them easier to find
21:56 MTecknology Calling all drunks!!! All drunks: you are being called!   iggy and I are gonna hit registration and then go to Squatters. All drunks (and non-drunks) are welcome!
21:56 iggy I hear they have food as well
21:56 Deevolution pacopablo: I have a minimum of Docker running via Salt, but it's all through custom states, rather then docker.io module.
21:57 pacopablo I am following the instructions here: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.dockerio.html#salt.modules.dockerio.pull regarding authenticating to docker repositories
21:57 daschatten ok iggy, then i will do that, thanks
21:57 pacopablo but when running my state, it's complaining about KeyError: 'username"
21:58 peters-tx joined #salt
21:58 Singularo joined #salt
21:59 MTecknology I imagine other attendees are around and would be interested. I'm going to go wander around, though. It's too dang nice around here.
21:59 murrdoc hows the weatehr
21:59 murrdoc weather*
22:02 nich0s joined #salt
22:02 monkey66 joined #salt
22:02 babilen pacopablo: That's not a state
22:02 pacopablo https://gist.github.com/pacopablo/3938e9dc5527279a31e8
22:03 pacopablo no, it's a pillar
22:03 pacopablo ^^ is the pillar and state
22:03 baweaver joined #salt
22:03 pacopablo from what I can tell, login info has to be in the pillar
22:03 toastedpenguin joined #salt
22:03 otter768 joined #salt
22:04 babilen pacopablo: user != username
22:05 pacopablo thanks
22:05 babilen (in your pillar)
22:05 iggy murrdoc: cold
22:06 murrdoc cold like socal cold or actual code
22:06 murrdoc cold*
22:06 iggy They don't even sell the kind of clothes you need around here where I live
22:06 toastedpenguin left #salt
22:06 * murrdoc checks googles
22:07 iggy think it's like 45 now... was 37 last night
22:08 auser joined #salt
22:09 pacopablo man, been staring at that pillar and state file too long.  was blind to my idiocy.  thanks for the eyes
22:14 daschatten iggy, i can't transfer or fork my repos to saltstack-formulas because of insufficent permissions. who can do that for me?
22:14 babilen pacopablo: You are welcome, hope it works now
22:14 pacopablo well, the login seems to work
22:15 pacopablo I've got other issues that seem to be more related to running a private docker registry to fix
22:16 ndrei joined #salt
22:17 iggy daschatten: think you need someone higher up to do that
22:18 daschatten iggy, how do i find someone higher up?
22:19 clintberry joined #salt
22:19 martaius joined #salt
22:19 iggy can you see the contributors list?
22:20 gladiatr_ joined #salt
22:20 daschatten yeah
22:20 daschatten ah, owner is next higher one?
22:20 daschatten should be
22:21 gladiatr joined #salt
22:21 daschatten can i send messages over github?
22:22 druonysuse joined #salt
22:22 druonysuse joined #salt
22:23 mosen joined #salt
22:24 bhosmer__ joined #salt
22:26 daschatten joined #salt
22:29 mpanetta joined #salt
22:30 mpanetta joined #salt
22:32 baweaver joined #salt
22:35 jdesilet joined #salt
22:36 druonysuse joined #salt
22:37 auser left #salt
22:41 shaggy_surfer joined #salt
22:41 shaggy_surfer joined #salt
22:44 ipmb joined #salt
22:46 kermit joined #salt
22:49 XenophonF eliasp: did utahdave share the fix for https://github.com/saltstack/salt/issues/20502 with you?
22:49 XenophonF i'm running into the same issue with Salt 2014.7.2 on Windows Server 2012 R2
22:49 XenophonF SheetiS: this is the SSLError i was reporting earlier today
22:49 eliasp XenophonF: I don't have a fix for that, sorry…
22:50 XenophonF eliasp: any indication what might be causing the error?
22:51 eliasp XenophonF: if development on Windows wouldn't be such a clusterfuck, I'd given it a try… but I didn't plan on admitting myself to a psych ward any soon
22:51 eliasp XenophonF: I think it's as before the requests package
22:51 eliasp XenophonF: the requests package does a lot of horrible things like bundling all its dependencies, which makes it hard to fix things systemwide
22:51 XenophonF eliasp: that's definitely what the stack trace implies
22:52 XenophonF eliasp: ugh really? that's pretty gross
22:52 btorch I always forget this... do we upgrade minions first or master ?  I'm going from 0.17.5 to latest
22:52 XenophonF eliasp: if it uses cryptoapi, the bug could be in python-win32
22:52 eliasp XenophonF: there have been plenty of complaints about this to the requests maintainer, but he's completely unwilling to change that
22:52 yomilk joined #salt
22:52 eliasp XenophonF: probably… but I didn't have the time/patience yet to dig into this Win-hell
22:53 XenophonF eliasp: i was really hoping to use salt in a domain-controller/federation-server/sharepoint-server deployment
22:53 XenophonF eliasp: i think as a workaround i'll have to put the installers etc. onto the master server directly
22:53 XenophonF and switch to salt:// urls
22:54 eliasp yes, should do as a workaround
22:54 XenophonF eliasp: do you know if this bug affects s3fs?
22:54 XenophonF on the master?
22:54 eliasp XenophonF: sorry, no idea… never used S3FS
22:55 XenophonF eliasp: ah ok - thanks though, appreciate it
22:55 theologian joined #salt
22:56 iggy this hotel is pretty pimp
22:57 StDiluted joined #salt
22:58 perfectsine joined #salt
22:58 iggy I'll just go ahead and throw this out there now to anybody that gets in tonight, we'll be out sampling the town hit me up 832.421.2464
22:58 eliasp iggy: have fun… I hope to make it to saltconf next year
22:59 iggy if i get spam shit from you giys, i'll track you down
22:59 eliasp ;)
22:59 * murrdoc totes signs up for everything
22:59 murrdoc wahts your social iggy
23:00 murrdoc need it for uh reservations
23:00 murrdoc 555-415555
23:00 murrdoc anyone have an example of custom logging
23:00 murrdoc in modules
23:01 thunderbolt left #salt
23:02 mosen murrdoc: writing your own execution modules?
23:03 Brew joined #salt
23:03 murrdoc yeah
23:03 murrdoc a few now
23:03 mosen I've just been using the old import logging logging.getLogger etc.. you using a different package?
23:04 murrdoc nope
23:04 SheetiS I also do the log = logging.getLogger(__name__)
23:05 SheetiS then i can log.critical('foo {0}'.format(whatever)) and the like
23:06 mosen yep
23:08 rgarcia_ joined #salt
23:16 bhosmer joined #salt
23:20 tmwsiy so besides the obvious zeromq HUGE advantage. What are some bullet points as to why salt is better than ansible?
23:20 aparsons joined #salt
23:21 XenophonF left #salt
23:22 XenophonF joined #salt
23:23 XenophonF tmwsiy: from my perspective, salt's open source, not open core
23:23 XenophonF that's kinda meta though
23:23 XenophonF also, support for windows
23:24 XenophonF although it's definitely a second-class citizen compared to RHEL or Ubuntu
23:24 XenophonF or even FreeBSD
23:24 murrdoc u get remote execution, reactor based system, ability to chain tasks, and the warm fuzzy feeling of fixing the tools u use
23:24 ahale joined #salt
23:25 XenophonF anyway, those are the main reasons i chose salt over ansible
23:25 XenophonF "fixing the tools u use" LOL
23:25 tmwsiy right on, thanks
23:25 murrdoc its true tho xeno
23:25 XenophonF absolutely!
23:25 murrdoc the thereifixedit feeling is nice
23:25 vegardx XenophonF: I think the open source / open core is actually way more important than people think. Just look at how chef just went downhill unless you do it exactly the Opscode-way.
23:25 XenophonF that reminds me, i have patches for freebsd i need to contribute back
23:26 vegardx There is only one way of doing chef. If you do it any other way you're fucked. In all meanings of the words.
23:26 XenophonF vegardx: too true, see also zenoss, where at lot of really useful monitoring capability is behind a paywall
23:26 murrdoc one thing chef and puppet still have on salt, is the central db for exported resources
23:26 XenophonF makes me thing twice and three times before contributing anything back
23:26 vegardx Yeah!
23:26 murrdoc puppet db , the http fronted postgres db is nice
23:26 vegardx Speaking of monitoring, take a look at sensu!
23:26 murrdoc chef db is decent
23:27 XenophonF i really, really wish halite was a thing
23:27 murrdoc vegardx:  likes pain
23:27 XenophonF it'd be nice to track things through a pretty UI
23:27 murrdoc salt-pad is doing well
23:27 XenophonF salt-pad?
23:28 murrdoc salt-pad
23:28 murrdoc i think saltstack devs are hard working to write a new halite
23:28 murrdoc for the conf
23:28 * murrdoc puts $1 on it
23:28 murrdoc XenophonF:  https://github.com/tinyclues/saltpad
23:29 XenophonF oh that's interesting, thanks murrdoc
23:29 mpanetta salt pad actually looks very interesting
23:30 murrdoc personal hate of mongodb notwithstanding
23:30 XenophonF hah well
23:32 douglassssssssss joined #salt
23:33 mpanetta I have a love hate relationship with mongodb...
23:33 mpanetta Love it because it is so simple to config, hate it because it breaks in odd ways at times...
23:33 UtahDave joined #salt
23:36 druonysuse joined #salt
23:36 druonysuse joined #salt
23:40 murrdoc i hate it because its not operationally easy to maintain
23:40 murrdoc like come on, if the unnatural growth on disk is something thats expected
23:40 murrdoc at least provie a simple way to fix it
23:41 druonysus joined #salt
23:41 druonysus joined #salt
23:41 murrdoc but i digress
23:48 CheKoLyN joined #salt
23:51 douglassssssssss joined #salt
23:53 baweaver joined #salt
23:53 XenophonF left #salt
23:53 nicktrs joined #salt
23:59 vschum1 joined #salt
23:59 nicktrs Installing salt-minion on ubuntu via PPA results in failures %20 of the time in Europe (Depends: salt-common (= 2014.7.1+ds-3trusty1) but it is not going to be installed).  We launch all servers the same way so I'd expect either a very high number of failures or a very low number (something deterministic).  Does anyone have an idea of what may be causing this?

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