Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-06-14

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

All times shown according to UTC.

Time Nick Message
00:07 ahammond woodtablet hey, if you figure out how to join, please let me know. :)
00:07 woodtablet ok
00:08 woodtablet i just made the repo, but ill come back here tomorrow to paste the link to ask for feedback
00:15 woodtablet left #salt
00:17 mpanetta_ joined #salt
00:20 iceyao joined #salt
00:20 mpanetta joined #salt
00:27 west575 joined #salt
00:27 phx joined #salt
00:29 squishypebble1 joined #salt
00:30 Sketch joined #salt
00:34 dendazen joined #salt
00:42 brotatochip joined #salt
00:44 squishypebble1 joined #salt
00:47 jmickle joined #salt
00:47 jmickle hello
00:48 jmickle does anyone know how i can use status.nrpoc in a state file so i can pass it as a var to a file template?
00:48 flowstate joined #salt
00:54 Nahual joined #salt
00:59 brotatochip joined #salt
01:01 bltmiller joined #salt
01:02 sagerdearia joined #salt
01:07 thejrose1984 joined #salt
01:07 Edgan joined #salt
01:11 mosen joined #salt
01:16 mpanetta joined #salt
01:17 mpanetta_ joined #salt
01:18 mpanetta_ joined #salt
01:29 edrocks joined #salt
01:38 justanotheruser joined #salt
01:47 ilbot3 joined #salt
01:47 Topic for #salt is now Welcome to #salt! | Latest Versions: 2015.5.10, 2015.8.10, 2016.3.0 | Support: https://www.saltstack.com/support/ | Logs: http://irclog.perlgeek.de/salt/ | Paste: https://gist.github.com/ (please don't multiline paste into channel) | See also: #salt-devel, #salt-offtopic | Ask with patience as we are volunteers and may not have immediate answers
01:49 flowstate joined #salt
01:54 mapu joined #salt
01:55 jwang joined #salt
01:56 catpigger joined #salt
02:09 gngsk joined #salt
02:27 davidhoude joined #salt
02:27 davidhoude Can anyone tell me if the following is correct: {{ salt['pillar.get']('nodes:' ~ grains['host'] ~ ':ipaddress') }}
02:28 davidhoude It seems to work for one node but not the other. When I do a grains.fetch host, I am returned the correct hostname, but when I refer to it like that in a template, it only gets parsed for one host and not the other.
02:28 davidhoude probably fatfingered something but just trying to make sure I am not missing something obvous
02:29 evle joined #salt
02:34 davidhoude damn
02:34 davidhoude it was a typo
02:35 davidhoude that I checked multiple times... should probably sleep
02:36 manji joined #salt
02:42 fgimian joined #salt
02:44 berserk joined #salt
02:48 flowstate joined #salt
02:57 onlyanegg joined #salt
02:57 fannet_ joined #salt
03:12 iceyao joined #salt
03:20 favadi joined #salt
03:21 jwang joined #salt
03:21 Vishvendra joined #salt
03:29 writtenoff joined #salt
03:32 edrocks joined #salt
03:32 Edgan joined #salt
03:40 fracklen joined #salt
03:40 darvon joined #salt
03:44 kshlm joined #salt
03:50 flowstate joined #salt
03:57 dfinn joined #salt
03:59 racooper joined #salt
04:07 brotatochip joined #salt
04:11 subsignal joined #salt
04:12 onlyanegg joined #salt
04:15 jeddi joined #salt
04:18 rem5 joined #salt
04:19 Edgan joined #salt
04:23 rdas joined #salt
04:26 Edgan joined #salt
04:32 aharvey joined #salt
04:33 ageorgop joined #salt
04:48 yuhlw joined #salt
04:49 flowstate joined #salt
04:55 Ayo_ joined #salt
04:57 calvinh joined #salt
04:58 DEger joined #salt
04:59 DEger joined #salt
05:05 DarkKnightCZ joined #salt
05:07 Edgan_ joined #salt
05:13 iggy ahammond: you interested in saltstack-formulas? It's a thankless job
05:13 subsignal joined #salt
05:18 paydro joined #salt
05:19 paydro joined #salt
05:20 punkoivan joined #salt
05:20 cyborg-one joined #salt
05:30 jwang joined #salt
05:30 iceyao joined #salt
05:32 brotatochip joined #salt
05:33 jwang Is it recommended to run salt-minion on the salt masters?
05:33 iggy yes
05:34 edrocks joined #salt
05:40 jwang So the bootstrapping process is kind of like this, right? (1) spin up the master node, (2) get the state files on the master node (git or tarball or whatever), then (3) run salt-call --local state.apply
05:43 whaity joined #salt
05:44 ernestonavarro joined #salt
05:48 flowstate joined #salt
05:48 sauvin joined #salt
05:55 armyriad joined #salt
06:00 manji joined #salt
06:02 ivanjaros joined #salt
06:15 pfallenop joined #salt
06:18 babilen jwang: I use salt-ssh to install new minions or masters, but you could use a masterless approach if you prefer
06:18 k_sze[work] joined #salt
06:19 jwang babilen: interesting. I'm looking into those docs now
06:20 Ayo_ joined #salt
06:21 babilen jwang: https://www.refheap.com/120295 is a suitable Saltfile config
06:22 babilen It expects copies of your states in /srv/salt/file_roots and pillars in /srv/salt/pillars -- feel free to change that to something that fits your setup.
06:25 jwang babilen: neato. Are there official docs for Saltfile? I searched and only see proposals about it
06:29 rdas joined #salt
06:32 kawa2014 joined #salt
06:34 favadi joined #salt
06:37 favadi joined #salt
06:39 duncanmv joined #salt
06:41 favadi joined #salt
06:44 whaity joined #salt
06:46 favadi joined #salt
06:47 slav0nic joined #salt
06:47 ernestonavarro joined #salt
06:48 favadi joined #salt
06:49 flowstate joined #salt
06:55 krymzon joined #salt
06:57 cetex joined #salt
06:58 dmaiocchi joined #salt
06:58 cetex so, i hit tihs: when doing pip.installed: 'Requirement' object has no attribute 'project_name'
06:58 cetex tried upgrading to 2016.3.0 but didn't work.
06:59 fannet_ joined #salt
07:00 AndreasLutro interesting... what does your requirements file look like? and what pip version do you have installed?
07:00 guedressel joined #salt
07:04 cetex six>=1.5.2:\n  pip.installed:\n    - require:\n      - cmd: python-pip
07:05 cetex I wonder if it has to do with pip 8.1.2
07:05 babilen jwang: It is documented on https://docs.saltstack.com/en/latest/topics/ssh/#define-cli-options-with-saltfile
07:06 AndreasLutro doubt it
07:06 toanju joined #salt
07:07 guedressel hello #salt
07:07 cetex seems like it..
07:07 cetex https://github.com/saltstack/salt/issues/33163
07:07 saltstackbot [#33163]title: Salt 2015.8.5 incompatible with Pip v8.1.2 | ### Description of Issue/Question...
07:08 AndreasLutro aha, but it seems to be fixed/handles in all new versions of salt
07:09 cetex huh
07:09 cetex so why am i getting that error? ;D
07:09 cetex tried latest stable and tried 2016.3, same thing.
07:09 AndreasLutro dunno, double check your version I guess?
07:09 AndreasLutro or share the full stack trace
07:09 AndreasLutro maybe it's not where I think it is
07:11 guedressel following the documentation of "publisher acl system" doesn't let my user execute salt commands...
07:12 guedressel ...what could I be missing?
07:12 guedressel $ salt '*' test.ping\n Failed to authenticate!
07:13 guedressel the publisher_acl directirve in /etc/salt/master allows my user to execute ".*" (regex for everything)
07:15 subsignal joined #salt
07:15 cetex http://pastebin.com/m4KhqFzt
07:16 cetex tried with both: "Get:16 http://ppa.launchpad.net/saltstack/salt/ubuntu/ trusty/main salt-minion all 2015.5.3+ds-1trusty1 [29.1 kB]"
07:16 cetex and Get:11 http://repo.saltstack.com/salt_rc/apt/ubuntu/14.04/amd64/ trusty/main salt-minion all 2016.3.0+ds-1rc2 [24.4 kB]
07:17 cetex sorry for some bad formatting, the logs are stored under layers of buildsystem which b0rks *
07:18 dmaiocchi joined #salt
07:22 cetex and pip is 8.1.2, seems to have worked well on 8.1.1
07:22 cetex so i'll try to downgrade unless you have any ideas for how to make salt run :)
07:28 ivanjaros3916 joined #salt
07:29 AndreasLutro cetex: well both of those versions seem old
07:30 AndreasLutro salt versions I mean
07:30 AndreasLutro 2016.3.0+ds-1rc2 is an RC, not a stable release
07:30 AndreasLutro as indicated by the url... http://repo.saltstack.com/salt_rc
07:30 cetex hm.
07:31 cetex right.
07:31 cetex i'll switch to stable 2016.3
07:31 cetex and see if it owrks
07:31 cetex *works
07:32 linjan joined #salt
07:33 fracklen joined #salt
07:35 cetex i had missed that, was using an old ppa previously and had no idea about salt's own repo :)
07:36 edrocks joined #salt
07:40 lero joined #salt
07:41 AndreasLutro I don't blame you, it's been a lot of back and forth - but checking version numbers will get you far :)
07:47 rrei joined #salt
07:48 flowstate joined #salt
07:48 manji joined #salt
07:49 manji joined #salt
07:49 keimlink joined #salt
07:54 cetex that seems to have worked better
07:55 babilen But it was broken with the RC?
07:55 cetex yeah.
07:58 rrei good morning
07:58 rrei any pillar experts here? I'm trying to understand how pillar files are merged
07:59 AndreasLutro rrei: just ask
07:59 rrei I have a file that includes a couple of other files, and then defines some keys
07:59 AndreasLutro includes how?
07:59 rrei I was expecting the included files to be added before the keys defined in the including file
07:59 rrei but it appears that's not the case
07:59 AndreasLutro include: [ foo, bar ] ?
07:59 rrei include:
08:00 rrei yes
08:00 rrei salt pillar includes
08:00 rrei not jinja includes
08:00 AndreasLutro right, I completely went away from sls includes in pillars because the order seems to be completely random
08:00 rrei oh
08:00 rrei damn
08:00 rrei :D
08:00 babilen cetex: You might want to file a bug with saltstack so that they don't break it again in .1
08:01 DEger joined #salt
08:01 AndreasLutro you just have to live with bloat/duplication in your top.sls
08:01 rrei it doesn't look in my case to be completely random, it just seems like the includes are added on top of the pillar keys defined in the including file
08:02 rrei which is opposite of what happens in formulas
08:02 AndreasLutro by formulas you mean state sls files?
08:03 rrei yes
08:03 AndreasLutro in my experience the include order is equally random in states, but in states you have requisites to force the order to be correct
08:04 AndreasLutro (by random I mean I can't find a pattern in the order of things)
08:04 rrei wow, now you have me suspecting everything. perhaps I never noticed that because I have a lot of explicit requisites which force the correct order of states
08:05 rrei I always thought that since it's a list, the order of includes should follow appearance in the list
08:06 AndreasLutro there's probably some complicated stuff with the sls compiler
08:06 SpX joined #salt
08:06 rrei well, I'll try to concoct something magical to work around this strange behavior :D
08:07 AndreasLutro just don't use includes in pillars would be my advice, specify everything in the pillar top.sls
08:08 rrei I had that, but it was getting out of control because of its size. so I spent a looooot of time breaking it up into multiple files to have a better organization
08:08 rrei and it works
08:08 rrei mostly
08:08 rrei until now :D
08:09 rrei oh, i misunderstood what you wrote
08:09 rrei you mentioned top.sls
08:10 rrei I was thinking you suggested to have a single pillar file with everything in it
08:10 rrei okay, I'll try something out. AndreasLutro: thanks for the help
08:11 SpX joined #salt
08:12 s_kunk joined #salt
08:12 Edgan_ joined #salt
08:17 ninkotech__ joined #salt
08:25 lempa joined #salt
08:26 deniszh joined #salt
08:30 Guest58804 joined #salt
08:46 fracklen joined #salt
08:48 flowstate joined #salt
08:50 fracklen_ joined #salt
09:00 felskrone joined #salt
09:08 atmosx Hello great people how are you?
09:10 CeBe1 joined #salt
09:11 atmosx I have some configuration files that do not get updated when I run state. The /var/cache/salt/minion/files has the old version. Is there anything I can do about this?
09:12 przemek_ joined #salt
09:16 subsignal joined #salt
09:35 rrei joined #salt
09:39 edrocks joined #salt
09:41 atmosx is this safe? https://gist.github.com/aventurella/7048284
09:42 Rumbles joined #salt
09:43 Martijn-- joined #salt
09:44 manji it should be yes
09:44 manji blunt
09:44 manji but yes
09:45 atmosx still doesn't work, so the problem is no tchace for me..
09:46 manji tchace ?
09:46 manji atmosx, when this has happend to me, I had a typo in some pillars
09:46 Martijn-- Hi folks, i am having some issues with Salt and a svn backend, i have a default master-file, i did add the file_roots: and svnfs_remotes: to a filesystem.conf in masters.d/; I also added a fileserver_backend: -svn to the same file.  My svn cache gets filled but the content from my salt state directory which was moved to the svn doesn't seem to be reachable.  Who has a good how to i could check? The salt docs don't get me where i want to be. I seem to be mis
09:46 atmosx manji "not cache"
09:47 manji atmosx, yes, I had cleared the caches like you did, as I thought it was a cache prob
09:47 manji before you go  any further, I'd enable debug on master and minion
09:48 manji clear caches
09:48 manji and read carefuly
09:48 atmosx manji thnx for the tips effie, long time no see btw.
09:48 manji hehe :p
09:48 flowstate joined #salt
09:51 rlarkin joined #salt
09:52 Martijn-- I do have a full gist available of my salt master configuration.
09:55 Martijn-- How can i tie a svnfs_remotes  rrepository to an environment like base ?
09:55 Martijn-- As thats seems to be what i am missing...
09:57 Ayo_ joined #salt
10:00 fredvd joined #salt
10:10 fredvd_ joined #salt
10:17 subsignal joined #salt
10:24 punkoivan joined #salt
10:25 punkoivan joined #salt
10:25 mikecmpbll joined #salt
10:27 punkoivan joined #salt
10:29 punkoivan joined #salt
10:31 felskrone joined #salt
10:48 dendazen joined #salt
10:57 toastedpenguin joined #salt
11:00 fannet_ joined #salt
11:10 teryx510 joined #salt
11:13 amcorreia joined #salt
11:22 Rumbles joined #salt
11:28 toastedpenguin joined #salt
11:29 evle1 joined #salt
11:30 rdas joined #salt
11:40 edrocks joined #salt
11:47 Muchoz joined #salt
11:49 keimlink joined #salt
11:57 amcorreia_ joined #salt
12:03 DEger joined #salt
12:10 noraatepernos joined #salt
12:16 numkem joined #salt
12:18 danielcb joined #salt
12:19 subsignal joined #salt
12:20 manji joined #salt
12:21 Ixan joined #salt
12:21 manji joined #salt
12:22 antpa joined #salt
12:27 Ixan hi, i'm having some issues figuring out how to use publish.publish inside a state
12:27 Ixan I'm using {%- set ticket = salt['publish.publish'](master, 'icinga2.create_ticket', hostname) %}
12:27 TooLmaN joined #salt
12:28 Ixan where icinga2.create_ticket is a module I've made.
12:28 Ixan I can use salt from the master to trigger the module to create the ticket and salt-call locally on the minion works fine with publish.publish, giving me a ticket
12:28 Ixan but when in the state file, the command never triggers and ticket is set to none
12:29 subsignal joined #salt
12:29 jeddi joined #salt
12:29 Ixan is there any magic i need to apply to be able to run publish from within a state?
12:30 abednarik joined #salt
12:34 Muchoz joined #salt
12:35 XenophonF joined #salt
12:39 gh34 joined #salt
12:44 dendazen joined #salt
12:51 nofxrok joined #salt
12:53 rem5 joined #salt
12:55 TyrfingMjolnir joined #salt
12:56 DanyC joined #salt
12:56 DanyC anyone used http module ?
12:56 squishypebble joined #salt
12:56 babilen DanyC: Be the first!
12:56 babilen But yes, many people have used it. Do you have a problem with it?
12:57 babilen bbl
12:58 DanyC babilen: i tried and i'm a bit confused as i run the execution module as per the doc where i passed the header_file param and the response is that it fails to find the file
12:59 DanyC babilen: should the header file be already present on the minion?
13:01 fannet_ joined #salt
13:01 subsignal joined #salt
13:03 DanyC_ joined #salt
13:04 subsigna_ joined #salt
13:04 DanyC_ joined #salt
13:09 dorei joined #salt
13:09 dorei is there some document explaining the syntax of .sls files? like those [] and ~
13:10 babilen yaml and jinja specifications
13:10 dorei is ~ and [] part of the yaml spec?
13:10 babilen ~ isn't, [] is
13:10 babilen http://jinja.pocoo.org/docs/dev/templates/
13:11 babilen http://yaml.org/ + https://en.wikipedia.org/wiki/YAML
13:11 saltstackbot [WIKIPEDIA] YAML | "YAML (/ˈjæməl/, rhymes with camel) is a human-readable data serialization language that takes concepts from programming languages such as C, Perl, and Python, and ideas from XML and the data format of electronic mail (RFC 2822). YAML was first proposed by Clark Evans in 2001, who designed it together..."
13:11 komputes_ joined #salt
13:12 dorei i guess since i'm using ~ outside a {{ }} block it isnt part of jinja, right/
13:12 dorei ?
13:13 epcim_ joined #salt
13:13 babilen I have no idea where you use it. Could you paste an example?
13:13 dorei yeah, let me find one
13:13 babilen http://jinja.pocoo.org/docs/dev/templates/#other-operators has ~ for string concatenation. I was thinking of that.
13:14 Muchoz joined #salt
13:16 dorei hmm, cant find one, maybe i mistook it with something else
13:16 dorei checking all the links u gave me, thanx a lot :)
13:22 dfinn joined #salt
13:22 edrocks joined #salt
13:25 mapu joined #salt
13:35 tapoxi joined #salt
13:38 gngsk joined #salt
13:40 perfectsine joined #salt
13:42 hasues joined #salt
13:42 hasues left #salt
13:43 rbjorklin Is it possible to do something like `consul.get` in a state, save the return value and then use it later in the state?
13:44 babilen rbjorklin: You could use {{ set foo = ... }}
13:46 nofxrok joined #salt
13:46 Tanta joined #salt
13:48 izibi how can I get a list of all known minions (i.e. accepted keys) from an ext_pillar?
13:49 ekristen joined #salt
13:49 iceyao joined #salt
13:50 ferbla joined #salt
13:50 rbjorklin babilen: Are there any examples of doing that?
13:52 flowstate joined #salt
13:55 gngsk rbjorklin: try something like .. {{ set foo = salt['consul.get']() }} ?
13:56 babilen ... and there are many many examples of using "set" in jinja templates in formulas and the jinja template documentation.
13:56 rbjorklin gngsk: Ah that makes sense!
13:57 babilen izibi: Why does it have to be from an ext_pillar ?
13:57 rbjorklin I wasn't sure how to call the function but seeing gngsk example makes me feel silly...
13:57 rbjorklin Thanks!
13:58 slav0nic joined #salt
13:58 babilen Sorry, I assumed that you knew how to call consul.get, but not how to save its return value for reuse later
13:58 Muchoz_ joined #salt
14:05 perfectsine joined #salt
14:05 mpanetta joined #salt
14:06 mpanetta joined #salt
14:07 kaptk2 joined #salt
14:08 kawa2014 joined #salt
14:09 CeBe1 joined #salt
14:09 andrew_v joined #salt
14:11 izibi babilen: because I'm writing an ext_pillar and need a list of alle minions
14:14 jcherndon joined #salt
14:15 jcherndon hello, have a question about salt-cloud
14:15 jcherndon concerning vm folder paths on vcenter
14:19 DarkKnightCZ joined #salt
14:19 duncanmv joined #salt
14:23 jcherndon in the cloud profile how do I tell it to create a vm in an arbitrarily long folder depth?
14:27 skeebump joined #salt
14:28 dps joined #salt
14:28 keltim joined #salt
14:29 dps duuuumb question: is it a bad practice to manage VARIABLE=somesetting in config files using salt.states.file.replace? :-)
14:30 aharvey joined #salt
14:30 babilen dps: No, it is not, but you get to the point where it typically would have been easier to just manage the entire file in the first place
14:31 dps babilen: by using jinja templating sort of management, or?
14:31 manji if you need templating yes
14:31 babilen Using file.managed
14:31 babilen Templating is optional, but you could use it
14:31 dps so file.managed
14:31 dps does that manage the contents of the file?
14:32 manji yes
14:32 dps i thought that only managed permisisons and stuff
14:32 dps i can look at the api docs again
14:32 manji you specify the source
14:32 babilen No, no .. it managed the contents.
14:32 manji https://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html
14:32 dps i basically have a single file that i want to transform key/value attribute pairs from a pillar
14:33 dps babilen: i didn't know that
14:33 manji {% for key, value in blabla.items() %}
14:33 manji {{key}}: {{value}}
14:33 dps isn't that templating at that point though?
14:33 manji yes, if you have your data in pillar
14:35 dps i will look into these docs guys thank you for your help
14:35 manji you must use jinja to extract that data
14:35 manji or whatever works for you anyway if not jinja
14:35 dps it seems easier to just use file.replace
14:35 dps maybe its just because this is the first time i have done this
14:36 babilen dps: https://docs.saltstack.com/en/getstarted/config/index.html (and the rest of https://docs.saltstack.com/en/getstarted/) might be interesting
14:36 dps im really tyring to determine a best practice for this.  i basically have a file that has 10 variables set via KEY=VALUE, i want to store basically all of them in pillar and then transform some stubbed out config to assign real values
14:37 dps babilen: fwiw you helped me last week regarding gitfs and the reccomendations you made work great.
14:37 dps thank you for your help.
14:37 babilen Perfect
14:38 babilen dps: When you use file.replace you rely on that file being there and the fact that it has specific content. If the content isn't what you think it is your "transformations" won't work. It is therefore a lot easier to just say "I want the file to look like this" and let sort make whatever changes are necessary.
14:38 dps babilen: i see, that makes sense.
14:39 babilen dps: The file.replace approach also gets tiresome if you make changes in tens of different states and you can never figure out again why the file looks like it does
14:40 babilen It might be perfectly fine though. It really depends on the actual usecase here
14:40 dps yeah that sounds about right
14:40 dps i mean i just want to do this the right way
14:40 dps file replace does sound a little hacky
14:40 dps if i can just do the entire thing using file.manage i think that will work fine
14:41 dps i think part of the problem is that i need to make a leap to the other side of the chasm in terms of how this file is going to be managed.  i have a stubbed out config in git and i'm still clinging to that.  i think file.manage will allow me to free myself, it seems reasonable to do it this way.
14:42 ajv joined #salt
14:42 dps anyway i appreicate your help both manju abnd babilien, thank you for pointing me in the right direction
14:42 * dps back to work
14:48 Muchoz joined #salt
14:50 ageorgop joined #salt
14:53 scoates have any of you ever created an AWS autoscale group (specifically launch config) in not-`us-east-1` with Salt? Can't find our image_id (AMI) because pretty sure it's looking in the wrong region
14:53 scoates ( ref http://boto.cloudhackers.com/en/latest/autoscale_tut.html#a-note-about-regions-and-endpoints )
14:56 corichar joined #salt
14:57 kawa2014 joined #salt
14:57 flowstate joined #salt
15:02 fannet_ joined #salt
15:03 perfectsine joined #salt
15:04 komputes_ joined #salt
15:06 renaissancedev joined #salt
15:06 favadi joined #salt
15:08 impi joined #salt
15:08 ivanjaros joined #salt
15:14 flowstate has anyone used salt to register/deregister instances with ELBs? My ELB has connection draining of up to 300 seconds, and I'm worried about salt timing out the request
15:16 flowstate and I can set the --timeout option, but then I may have to wait for 15 minutes to deregister all 3 instances
15:16 flowstate so I guess I need to concurrently execute the jobs, but wait for up to 300 seconds for each
15:16 flowstate and the return value is important to act as a gate for the next action in my deployment
15:16 flowstate hmm
15:17 mpanetta joined #salt
15:26 vexati0n joined #salt
15:27 vexati0n question re: SaltStack host state. I can add/remove hosts in /etc/hosts, but how do i force multiple aliases for an IP to listen in the correct order (fqdn should always be first)?
15:29 mpanetta joined #salt
15:29 ageorgop joined #salt
15:29 tapoxi joined #salt
15:30 brotatochip joined #salt
15:32 tapoxi hi everyone, thoughts on how to organize pillars? I have my top file, a 'common' file, and then 'roles'. everyone inherits from common, then roles overwrite. trying to figure out how to have append instead of overwrite though, a role specific namespace?
15:33 ElasticApps joined #salt
15:36 user743204321 joined #salt
15:36 woto joined #salt
15:38 permalac joined #salt
15:38 woto left #salt
15:39 _JZ_ joined #salt
15:41 whatevsz joined #salt
15:41 berserk joined #salt
15:44 slimmons joined #salt
15:45 flowstate joined #salt
15:48 Eugene tapoxi - here's how I do it. https://github.com/EugeneKay/srv-salt/blob/pepper/pillar/top.sls
15:49 slimmons If I want to watch a directory on my master, and any time a file in there changes, I want to move all of those files that have changed onto a minion, would the best way to do that be file.recurse, with a watch on that directory?
15:49 Eugene Worth a read if you haven't yet https://docs.saltstack.com/en/latest/topics/pillar/#pillar-namespace-merges
15:49 Eugene slimmons - sounds right
15:50 Rumbles joined #salt
15:51 vexati0n slimmons: a watch watches to see if a state changes somewhere else in the same state run. it won't monitor a folder for changes to files that occur outside of a salt state, and it wouldn't watch a folder on the master anyway unless the state itself is being run on the master.
15:53 favadi joined #salt
15:53 slimmons So, I guess I need the master to be a minion also,
15:54 slimmons hmm, I'll think about it a bit
15:54 hasues joined #salt
15:54 slimmons thx for the help
15:54 hasues left #salt
15:54 vexati0n yeah, and even then, the copy would only execute if it's the salt state that's making the changes to the files in question.
15:55 perfectsine joined #salt
15:55 vexati0n but why not just run the state with a simple file.recurse often enough to keep up with whatever you're doing? if there are no changes, nothing will change on the minion, but if the files on the master are different, they'll be copied.
15:56 slimmons I guess that would work just as well
15:56 vexati0n of course if you're trying to do something like a realtime filesystem sync or the like, salt is really the wrong tool for the job anyway.
15:56 fxhp joined #salt
15:56 slimmons yeah, I don't need real time, I think every few minutes would be fine
15:56 slimmons so just cron the state or something, I guess
15:57 vexati0n yeah so just schedule the state to run every X minutes. salt has a scheduler built in that you can use.
15:57 slimmons awesome, didn't know that
15:57 slimmons thanks
15:57 slimmons again
15:57 vexati0n np
15:58 flowstate joined #salt
16:00 DEger joined #salt
16:02 dh__ joined #salt
16:03 dh__ Can someone look at this and tell me if it's wrong.
16:03 dh__ http://pastebin.com/4fR8vKve
16:03 dh__ It passes but doesnt target the 'dev' environment
16:03 edrocks joined #salt
16:03 dh__ still targets Base
16:04 bltmiller joined #salt
16:05 vexati0n dh__: i think i need to see the code where "data" is defined.
16:06 brotatochip joined #salt
16:06 dh__ I think that is passed from the minion to the event queue
16:06 dh__ data['host'] is just the minion name
16:06 dh__ er data[name]
16:06 dh__ I think it's the env: dev line
16:06 dh__ i cannot find anything in the documentation though
16:07 vexati0n idk then... i'm a reactor n00b
16:07 vexati0n :(
16:13 dh__ looks like it could be saltenv: dev and not env: dev -- testing now
16:13 dh__ thanks anyways though vex
16:15 adelcast left #salt
16:17 Muchoz joined #salt
16:17 mpanetta_ joined #salt
16:19 dh__ nope =\
16:19 noraatepernos joined #salt
16:20 flowstate joined #salt
16:25 abednarik joined #salt
16:25 flowstate joined #salt
16:27 RandyT joined #salt
16:28 favadi joined #salt
16:30 aharvey joined #salt
16:35 woodtablet joined #salt
16:41 writtenoff joined #salt
16:46 west575 joined #salt
16:46 tylux joined #salt
16:48 dmaiocchi joined #salt
16:49 tylux Hello Everyone, I am trying to do a jinja conditional but am having a hard time finding out if something like this is possible , {% if package['postfix'] is installed %}. Am I going down the right path?
16:50 ageorgop joined #salt
16:55 brotatochip joined #salt
16:55 slimmons This setup isn't working for what I was talking about earlier.  Maybe the files need to be in a different directory?  https://gist.github.com/Slimmons/2cf1082c2c02c9435f5c3da3c9ffef4c
16:55 iggy tylux: {% if salt['pkg.version']('postfix') %} maybe
16:56 iggy or {% if 'postfix' in salt['pkg.list_pkgs']() %}
16:56 tylux k I will give those a try
16:56 iggy you'll have to get creative probably based on your pkg manager's specific functionality
16:56 iggy those functions are both from aptpkg
16:57 impi joined #salt
16:57 kevinquinnyo joined #salt
16:57 tylux ok let me try those out in my test ubuntu and centos boxes, I will report back in a few
16:58 kevinquinnyo I've never found a good workflow for quick debugging with saltstack. For instance, in say PHP, i can do a quick print_r($something); die(); to stop execution there and print the contents of that variable
16:58 mikecmpbll joined #salt
16:59 kevinquinnyo What's the best way to do something like that with salt?
16:59 kevinquinnyo as opposed to something crazy like file.managed and putting the contents into /tmp/debug.out
16:59 nicksloan joined #salt
16:59 nicksloan are pillar top files merged in the same way that state top files are?
17:00 paydro joined #salt
17:00 keltim kevinquinnyo, you and everyone else
17:01 paydro joined #salt
17:01 iggy kevinquinnyo: at this point, I just know what works and what doesn't... but I've resorted to to things like `salt-call -l debug state.sls foo` on the minion because it shows the rendered output of sls files before
17:01 iggy it's also possible to use some bits of jinja on the command line
17:02 garphy joined #salt
17:02 linjan joined #salt
17:02 keltim it'd be awsome to have some kind of IDE that could check all these things for you
17:03 jcherndon joined #salt
17:03 kevinquinnyo iggy i like that salt-call -l debug from the minion
17:03 keltim like show the rendered yml in a sep. window, autocomplete pillar names, etc.
17:03 iggy there's also state.show_*
17:04 iggy yeah, I think lots of people have tried it... it's complex to render correctly or without mocking so much data that you spend more time doing that than actually working
17:05 tylux iggy: {% if 'postfix' in salt['pkg.list_pkgs']() %} seemed to work in Ubuntu, testing centos now
17:05 kevinquinnyo iggy the salt-call is actually useful enough for now because with debug mode, it will display things that i print out in the log, like '___DEBUG___ {0}'.format(some_variable)
17:06 kevinquinnyo thanks
17:06 btorch joined #salt
17:06 iggy there's also state.* with test=True, but that doesn't really show you anything, just lets you know if everything renders correctly
17:06 kevinquinnyo good to know
17:06 iggy but it is something I use for debugging, so I include it
17:07 jcherndon hello, does anyone know how to list arbitrarily deep folder tree for salt-cloud profiles?
17:07 paydro_ joined #salt
17:07 iggy jcherndon: rephrase the question? or... What are you actually trying to achieve?
17:08 btorch how come pkgrepo.managed always append to an exists /etc/apt/sources.list.d/ list file instead of modifying the the current "deb url ..." line ? am I missing something ?
17:08 iggy btorch: that's what it's supposed to do
17:08 gladiatr joined #salt
17:09 iggy btorch: you can use clean_file if you want everything else in that file gone
17:10 mpanetta joined #salt
17:10 paydro joined #salt
17:10 jcherndon I have a vcenter that has VMs in folders. I have subfolders and subfolders. I can clone my vm and place it in a top level folder, but can't get any deeper
17:11 * iggy backs away slowly
17:11 btorch iggy: cool thanks I'll keep that in mind once I upgrade to 2015.8.X :)
17:12 iggy if you are something lower, it's going to lose support pretty soon, so I'd do it sooner than later
17:13 btorch trying :(
17:14 iggy btorch: otherwise, use pkgrepo.absent on the ones you want gone
17:14 bltmiller joined #salt
17:15 berserk joined #salt
17:15 Ayo__ joined #salt
17:18 slimmons This schedule.present isn't working as I expect (first time making one).  Any ideas?  https://gist.github.com/Slimmons/2cf1082c2c02c9435f5c3da3c9ffef4c
17:23 iggy slimmons: do you see it added to /etc/salt/minion.d/_schedule.conf ?
17:23 Vishvendra joined #salt
17:25 slimmons yes, it's there,
17:25 slimmons does the function need to be written on the minion, or is it in the correct place in the init.sls on the master?
17:26 teryx510 joined #salt
17:29 Rumbles joined #salt
17:30 slav0nic joined #salt
17:30 nmadhok joined #salt
17:34 woodtablet1 joined #salt
17:38 charli joined #salt
17:39 patarr Does salt have functionality to generate unique data? As in store it somewhere and make sure you never get the same value twice?
17:39 patarr And no, uuid won't work in this instance unfortunately
17:40 impi joined #salt
17:40 mapu_ joined #salt
17:41 s_kunk joined #salt
17:41 Guest58804 joined #salt
17:47 charli joined #salt
17:48 Miouge joined #salt
17:48 fredvd joined #salt
17:48 teryx5101 joined #salt
17:49 jesusaur joined #salt
17:50 paydro joined #salt
17:51 tylux joined #salt
17:52 fannet joined #salt
17:54 tylux iggy: {% if 'postfix' in salt['pkg.list_pkgs']() %} works on Ubuntu but not CentOS, fails with a Pillar render error, salt pkg.list_pkgs seems to work on my centos box. I am running this with salt-call, going to spin up a test master box to see more logs
17:55 btorch left #salt
17:55 paydro joined #salt
17:58 ahammond patarr would https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.grains.html#salt.modules.grains.get_or_set_hash solve your problem?
17:59 paydro joined #salt
18:00 ahammond in the context of an execution module, how do I get a file from salt?
18:00 ahammond for example, I have a foo/defaults.yaml checked into my salt repo and I want the data in it on a minion.
18:01 ahammond file.get_managed?
18:06 fracklen joined #salt
18:06 paydro joined #salt
18:11 aharvey joined #salt
18:11 jwang joined #salt
18:12 toanju joined #salt
18:14 flowstate is there any way similar to a highstate to execute a salt-runner?
18:14 flowstate as in, minion does a salt-call, master somehow introspects on minion, and runs a salt-runner?
18:14 flowstate I've moved my deployments over to orchestration, but now I can't just call highstate when my minions start for the first time
18:14 patarr ahammond, sounds almost exactly like what i need, but I can't use it because it's not documented. I don't know how the hash is generated, and its uniqueness guarantees
18:14 Rumbles joined #salt
18:15 flowstate not to be that guy, but you could clone the repo and check out the code
18:15 flowstate I've had to do that for some of the boto stuff, but it's really not hard to read if you write python
18:16 flowstate I totally get where you're coming from though, I had the same issue with the supervisord states
18:19 armguy joined #salt
18:20 berserk joined #salt
18:20 cellofellow joined #salt
18:20 druonysus joined #salt
18:21 jwang flowstate: are you looking for the salt-run command? https://docs.saltstack.com/en/latest/ref/cli/salt-run.html
18:22 iggy tylux:
18:22 flowstate jwang: I'm using that on the master for deployments to my minions, but now I need to somehow invoke the runner from the minion when it bootstraps
18:22 flowstate with my old deployment style, I could just do a salt-call state.highstate
18:23 flowstate but now I'm not really using that mechanism anymore, since I'm deploying with an orchestrator
18:23 iggy terminal locked up there
18:23 iggy tylux: you're not going to get more output from a master
18:24 tylux iggy: Figured out my CentOS issue, my test machine didn't have dmidecode installed so I had another jinja issue on {% if grains['manufacturer'] == 'Dell Inc.' %}
18:24 iggy use salt-call -l debug
18:24 jwang gotcha. Isn't that what something like `salt '*' state.highstate` does? I'm not familiar with orchestrators, though
18:24 druonysus joined #salt
18:24 slimmons I posted a question about the scheduler, on stack overflow.  Feel free to help me out :)  http://stackoverflow.com/questions/37819472/how-do-i-get-schedule-present-to-work-in-saltstack
18:24 tylux my setup is I use vagrant boxes to test/write my states. I now have included installing dmidecode if it isn't installed.
18:25 flowstate yeah, basically I was able to create a single sls that could install many of my services. This creates the problem of all of the states having the same ID when I try and run them multiple times
18:26 brotatochip joined #salt
18:26 flowstate to get around that, I created an orchestrator which calls that single sls once for every service in the list
18:26 scoates joined #salt
18:27 flowstate https://gist.github.com/flowstate/58d4692f67134785e6f112f633506658
18:27 flowstate but now, I can't use highstate, because the salt-run needs arguments, and it is explicitly only able to be executed on the master, as far as I can tell
18:28 flowstate if I could apply that one single sls for each of my services without running into the conflicting IDs problem, then I wouldn't need the orchestrator
18:28 flowstate but alas, here I am
18:28 jwang yeah, AFAIK, salt-run isn't even installed by salt-minion packages
18:29 snergster joined #salt
18:30 tylux iggy: Thanks for all your help!
18:35 jdonahue joined #salt
18:36 tmehlinger joined #salt
18:38 brotatochip joined #salt
18:40 punkoivan joined #salt
18:42 Joe630 I am stuck!  I am using winrm to create an ec2 win2012 image, but winrm is trying to use my proxy as if it is external to my master.  I can't figure out how it is getting proxy info or how to turn it off.  Any advice?
18:42 punkoivan joined #salt
18:42 mapu_ joined #salt
18:45 keimlink joined #salt
18:48 jwang joined #salt
18:48 Bloodshot joined #salt
18:49 Bloodshot Could somebody kindly provide me with an example of how to run an execution module on a minion from a reactor, and use the result in Jinja?
18:50 Bloodshot I've tried this: http://ix.io/SI3
18:50 Bloodshot But 'ip' seems to end up empty
18:57 mpanetta_ joined #salt
18:58 fer_bla joined #salt
18:58 ageorgop joined #salt
19:00 fer_bla joined #salt
19:01 twork_ if a minion is in the midst of a long run to highstate and gets a fresh copy of the pillar pushed to it, does that cause issues, or will it just continue with the version of the pillar it had to begin with?
19:01 DammitJim joined #salt
19:01 mpanetta joined #salt
19:03 flowstate it will continue with the first defined
19:03 flowstate at the beginning of the job run, it builds the full context, including the pillar
19:03 twork_ thanks
19:03 flowstate any changes after that will only affect the next run
19:03 flowstate np
19:05 amcorreia_ joined #salt
19:05 amcorreia joined #salt
19:06 edrocks joined #salt
19:06 s_kunk joined #salt
19:06 s_kunk joined #salt
19:10 kingscott joined #salt
19:10 kingscott Is there a way to require a plugin when running a salt state? I'm trying to only update the security on a centos 6.8 box and it fails because it doesn't have yum-plugin-security installed.
19:24 jwang kingscott: isn't that just a yum package? Can you install it in your states?
19:27 flowstate joined #salt
19:28 whitenoise joined #salt
19:32 BogdanR joined #salt
19:33 Joe630 re: my previous proxy question - I had a dirty environment
19:34 drawsmcgraw I'm running `salt-run fileserver.update` on my Master and getting this nasty -> [ERROR   ] Import pygit2 failed: CompileError: command 'gcc' failed with exit status 1
19:34 drawsmcgraw I've confirmed all packages & configs are correct. Anyone else seen this thing?
19:35 Bryson joined #salt
19:35 viq how can I diff a file between two git branches?
19:36 jwang viq: `git diff branch1 branch2`, though this probably isn't the best place to ask that
19:37 viq jwang: well, I'm trying to figure out what changed in one of modules before I submit a bug report
19:37 viq jwang: anyway, thank you :)
19:38 jwang viq: fair. good on you for taking the time :)
19:38 fannet joined #salt
19:38 viq Yeah, I noticed it works on one of my minions with 2015.8.8 and fails on 2016.3.0, so seeing what changed
19:39 scoates joined #salt
19:40 viq ah, looks like I need to start with git fetch --all
19:40 fannet_ joined #salt
19:41 ahammond Bloodshot: reactors do not have access to the salt dict. What you probably want to do is configure salt.mine to run network.ip_addrs and then look up the minion's IP in salt mine.
19:41 jwang drawsmcgraw: never seen that error, but I assume it's related to the git fileserver backend...
19:41 Bloodshot ahammond: https://docs.saltstack.com/en/latest/topics/reactor/index.html#jinja-context states that execution modules are available
19:42 ahammond Bloodshot huh, you're right.
19:43 ahammond I wonder why my reactor isn't working? :)
19:43 Bloodshot In fact, I can `salt-master -l debug` lets me know that my module is running, but is confused about its arguments
19:43 ahammond anyway, you _probably_ want to use salt-mine as a layer of caching rather than running network.ip_addrs directly.
19:44 Bloodshot I do use salt-mine, but I'm running this reactor as soon as the minion is brought up, possibly before the salt-mine is updated (before the highstate is run)
19:45 Bloodshot "[ERROR   ] Interface {'interface': 'eth0', 'type': 'public', 'tgt': 'dnstest2'} not found."
19:45 ahammond Bloodshot ah, that make sense.
19:47 brotatochip joined #salt
19:48 komputes_ joined #salt
19:49 slav0nic joined #salt
19:51 viq meh, not enough in the diff for me to go on, so here goes https://github.com/saltstack/salt/issues/34012
19:51 saltstackbot [#34012]title: States mount.* fail on OpenBSD's tmpfs | ### States mount.* fail to take into consideration `mount` output on OpenBSD...
19:57 mpanetta joined #salt
19:58 aharvey joined #salt
20:00 gngsk Is it possible to set use the archive execution module to extract a tar from a URL as you can with the state module?
20:00 bltmiller joined #salt
20:00 PeterO joined #salt
20:00 gngsk s/set//
20:01 andrew_v joined #salt
20:01 mpanetta joined #salt
20:05 gngsk I suppose not, it seems the state actually calls file.managed.
20:07 skeebump joined #salt
20:07 gngsk heh, I'll just do what that execution module does. Use state.single to run a state remotely?
20:07 jwang joined #salt
20:08 mapu_ joined #salt
20:09 amcorreia joined #salt
20:10 amcorreia_ joined #salt
20:14 flowstate joined #salt
20:15 thenut joined #salt
20:15 fannet joined #salt
20:16 iggy slimmons: if it's getting in that file on the minion, I'd crank up debug output and see why it's failing (the scheduled job is running, but it's failing for some reason)
20:17 jwang joined #salt
20:18 subsignal joined #salt
20:18 flowstate joined #salt
20:18 renaissancedev joined #salt
20:19 ahammond Bloodshot I think your problem there is that you're handing it data['name'] and expecting it to run remotely. However, I think this a straight up local call to an execution module.
20:24 flowstate joined #salt
20:28 SpX joined #salt
20:28 flowstate joined #salt
20:31 mapu joined #salt
20:35 flowstate joined #salt
20:37 djgerm joined #salt
20:38 djgerm what happens when one applies a state to a minion that is already applying a state (e.g. two engineers targeting the same minion with different states)?
20:38 skeebump joined #salt
20:39 flowstat_ joined #salt
20:39 djgerm without explicitly saying queue or concurrent or anything like that.
20:40 Trauma joined #salt
20:41 whatevsz joined #salt
20:41 nidr0x joined #salt
20:41 rem5 joined #salt
20:42 lorengordon joined #salt
20:47 brotatochip joined #salt
20:48 garphy joined #salt
20:52 ahammond djgerm the 2nd state fails to apply. you'll get an error message back to the effect that there's already a state running.
20:53 whaity joined #salt
20:53 slimmons iggy: The error from the minion is that the function doesn't exist.  I think it wants the function written inside of the minion config file instead of the master init file.
20:53 noraatepernos joined #salt
20:53 slimmons iggy:  also, I posted the question here, so there's a little more info.  http://stackoverflow.com/questions/37819472/how-do-i-get-schedule-present-to-work-in-saltstack
20:53 fannet_ joined #salt
20:54 iggy we aren't allowed to view stackoverflow at work
20:54 woodtablet joined #salt
20:54 djgerm thanks ahammond that's what I thought.
20:55 ahammond iggy that's draconian! Come work with me, it's less draconian and more Stalinist
20:56 iggy I owe my soul to Verizon for a while longer
20:56 iggy or AOL? Yahoo? Oh, who even knows...
20:57 woodtablet iggy: yesterday you said you dont use salt enviroments. why? and what do you use instead to test stuff ?
20:57 woodtablet iggy: i have bunch of formulas, and testing it going well, but i want to separate my testing from deploying to production (which i want to try to do soon here)
20:58 bltmiller joined #salt
20:58 iggy separate masters
20:58 iggy at $currentjob and $previousjob
20:58 slimmons oh man, not viewing SO at work, that's pretty nuts
21:00 woodtablet iggy: hmm.. how does that work ? you have a set of test machines with a test master? Or are the test machines multi-mastered to try the test master first ?
21:00 Edgan You work for Verizon. Which means you almost surely have a smartphone, probably from Verizon. So view it on your phone.
21:00 ws2k3 joined #salt
21:01 deniszh joined #salt
21:01 aharvey joined #salt
21:01 woodtablet note: it is hard to determine who i hate more.. mobile wireless companies or internet service providers..
21:01 Rkp joined #salt
21:01 devsterknight|ZN joined #salt
21:01 edulix joined #salt
21:02 Vye joined #salt
21:02 slimmons mobile for me
21:02 ajw0100 joined #salt
21:02 iggy I don't work for _that_ verizon
21:02 woodtablet you havent dealings with Cox Cable i beat
21:02 woodtablet bet i mean
21:03 iggy I mean I do, but not really
21:03 iggy look, it's all a bit complicated, but...
21:03 ahammond I like my ISP. Etheric Networks in the Bay Area. They're expensive, competent and neither a telco nor a cable company.
21:04 Edgan ahammond: but availability and speed?
21:05 ahammond Edgan, I'm paying for 20Mbps and getting about 25 + bursts. Latency is reasonable. I monitor for outages > 60s and there's only been one in the last year: when I moved. Since it's wireless, rainstorms can lead to packet loss, but even that seems pretty minimal.
21:06 Edgan ahammond: ah
21:06 Edgan viq: I found a fun 2016.3.0 bug. I am about to test it in an older version to see where it broke.
21:07 ahammond Edgan Kids can still watch netflix / amazon video in HD during rainstorms, so the necessary functionality is there.
21:07 woodtablet iggy: so just a separate test infrastructure ?
21:08 Edgan ahammond: I have Comcast, and had packet loss last night, intermittently. Everything was glitching, except Amazon Video on my TV.
21:08 iggy woodtablet: yeah, prod/qa/dev masters
21:08 woodtablet iggy: ok, cool, thanks, giving me ideas for architecture of my soon tobe salt stuff
21:09 iggy at the last job, we pulled git branches for each master separately... at this one, our salt tree's are handled by packages using a system that predates salt here
21:09 ahammond woodtablet we're using a vagrantfile that's checked into source control for all devops guys. I'm not particularly satisfied with that solution.
21:09 iggy we are moving more that route for dev stuff ^
21:09 woodtablet ahammond: that is my current solution for myself on multiple computers for the development of the formulas
21:10 iggy but it's complicated because we have some things that can only be tested with like 10 servers up
21:10 ahammond what I'd really like is a docker solution where the salt-master is dockerized. It'd make upgrades (and more importantly rollbacks) about a billion times easier
21:10 iggy and some of them require dedicated hardware, so VMs are a non-starter
21:11 ahammond iggy 10+ servers? yikes. My most complicated test scenario is our build / cd stuff, and that requires 7 boxes. My laptop can manage, but only barely.
21:12 iggy we are hopefully getting more and more away from huge, complicated, complex, interwoven setups... but until then ^
21:13 Edgan iggy: are you using 2016.3.0?
21:13 iggy also, the fact that some things require actual hardware bits makes vagrant a no-go for those teams as well
21:13 iggy personally, yes, at work, no
21:13 mrMute joined #salt
21:13 nledez joined #salt
21:13 iggy we don't go to new releases until like .8
21:13 Edgan iggy: vagrant has it's limitations, especially if you use EC2 and stuff is EC2 only
21:14 Edgan iggy: so personally, https://github.com/saltstack/salt/issues/34013
21:14 saltstackbot [#34013]title: top.sls not processed in order | ### Description...
21:14 ahammond Edgan we are on 2016.3.0 at work. It's not going very well.
21:14 deniszh1 joined #salt
21:15 Edgan ahammond: The problem I have is I run into so many bugs that if I didn't stay on the latest I would be backporting a ton of patches myself.
21:15 ahammond Edgan yeah, well... we really want that Thorium reactor for a project I'm looking to start in a month or two.
21:15 ahammond also, we really really really wanted off 0mq
21:16 Edgan ahammond: oh, do tell
21:17 ahammond tcp transport is The Future. :)
21:18 Edgan Lately I have found workarounds for any bugs that I haven't found existing patches for. With the current exception of the one above.
21:22 nidr0x joined #salt
21:26 ninkotech__ joined #salt
21:32 XenophonF anyone out there using salt to provision sharepoint?
21:32 XenophonF i don't want to jump down that rabbit hole if someone else has already done it
21:36 flowstate joined #salt
21:36 CeBe joined #salt
21:36 nidr0x joined #salt
21:36 fredvd joined #salt
21:37 mikecmpbll joined #salt
21:37 fannet_ joined #salt
21:38 haaj joined #salt
21:43 ninkotech__ joined #salt
21:44 hlub joined #salt
21:46 flowstate joined #salt
21:48 Jaremy joined #salt
21:48 ajw0100 joined #salt
21:48 Jaremy trying to set the python version when creating a virtualenv state but when i apply the state I keep getting the follow:  'Cannot find requested python ({0}).'.format(python)               CommandExecutionError: Cannot find requested python (python3.4).      Started: 21:30:41.659945     Duration: 93.665 ms
21:49 Jaremy I have tried pointing the python argument in the virtualenv.managed: python: /usr/bin/python3, /usr/bin/python34, python3, etc to no avail
21:51 nidr0x joined #salt
21:52 ahammond Jaremy post a gist please.
21:56 ahammond Jaremy https://gist.github.com/ahammond/5555a812089c0057880a6bc9868325ed if you want a working example for a few years back.
22:00 edrocks joined #salt
22:00 Jaremy https://gist.github.com/jcheetwood/9905a7cf28bb145a28766cb729da47a4
22:00 Jaremy That is my current sls for virtualenv
22:02 user3280943 joined #salt
22:02 user3280943 hey all
22:02 freelock joined #salt
22:03 user3280943 i've got a docker running with /etc/nginx/nginx.conf bound to /home/docker/nginx/etc/nginx.conf and a cmd with the name /.../nginx -s reload and i'm trying to have it fire every time the nginx.conf is saved/modified
22:03 ekristen joined #salt
22:04 user3280943 this is what i'm doing now: http://pastebin.com/8EZ3J3RB
22:04 user3280943 doesn't seem to be working
22:05 user3280943 (assume the docker is in the right state, still have to add that requirement in)
22:07 user3280943 i'll try using watch
22:07 deniszh joined #salt
22:09 user3280943 salt probably isn't continually watching the file for changes (most likely the source of my problem)
22:10 user3280943 rather changes from the last state run
22:10 user3280943 continuously**
22:12 brotatochip joined #salt
22:16 nledez joined #salt
22:27 robawt is there a nice way to highstate machines as part of a deployment?
22:27 robawt salt orchestrate runner seems too complicated for my priority:  highstate an environment of machines based on role
22:28 robawt i think i might be better off just highstating them from the master on a grain match salt call
22:37 Edgan robawt: I am doing what you are talking about with salt-ssh
22:37 Edgan robawt: It is broken into two pieces
22:38 Edgan robawt: The base stuff like users, salt minion, timezone, repositories, etc I handle with salt master highstate.
22:39 Edgan robawt: The app dependent stuff like nodejs/python/ruby/java, language modules, deployed code, etc, I handle with salt ssh highstate
22:39 juanito_ joined #salt
22:39 robawt thanks for the input Edgan
22:46 flowstate joined #salt
22:53 manji joined #salt
22:57 jwang joined #salt
22:59 Deliants joined #salt
23:03 repl1cant joined #salt
23:04 scoates joined #salt
23:07 Robert__ joined #salt
23:10 keimlink_ joined #salt
23:12 nZac joined #salt
23:16 ElasticApps joined #salt
23:22 noraatepernos joined #salt
23:23 ahammond robawt we use a combination of a grain and a reactor. When a server is first imaged, they have a firstboot grain set. When a minion starts, it generates an event that triggers a reactor which matches on the minion name and that firstboot grain.
23:23 ahammond robawt once the first configuration run is completed, the firstboot grain is removed.
23:24 ahammond robawt it's not perfect, but it's reasonable.
23:25 fannet_ joined #salt
23:30 Deliant joined #salt
23:46 flowstate joined #salt
23:47 ahammond salt 2016.3.0 is expected to install from repo.saltstack.com via salt-bootstrap, right?
23:47 druonysus joined #salt
23:48 dendazen joined #salt
23:59 ronrib joined #salt

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