Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2017-03-14

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

All times shown according to UTC.

Time Nick Message
00:12 woodtablet left #salt
00:21 mikecmpbll joined #salt
00:59 feld joined #salt
01:01 shoemonkey joined #salt
01:04 dendazen joined #salt
01:09 jimklo joined #salt
01:09 felskrone joined #salt
01:10 mavhq joined #salt
01:19 jimklo joined #salt
01:21 jimklo joined #salt
01:24 jimklo joined #salt
01:24 cachedout joined #salt
01:28 gnomethrower joined #salt
01:28 netcho joined #salt
01:29 jimklo joined #salt
01:34 Poppabear anyone have a good working supervisor formula with a proper map.jinja ?
01:34 Poppabear the one on saltstack is well ...
01:34 Nahual joined #salt
01:35 Poppabear incomplete ? i guess is the right word
01:57 robinsmidsrod joined #salt
02:07 Tanta joined #salt
02:10 Poppabear https://github.com/saltstack-formulas/supervisor-formula/issues/1
02:10 saltstackbot [#1][OPEN] How to extend the map.jinja | I'm sorry but i don't understand how to set the nested values for the defaults.yaml...
02:10 Poppabear any help would be greatly appreciated
02:11 mcspiff joined #salt
02:21 cachedout joined #salt
02:22 mikecmpbll joined #salt
02:27 nethershaw joined #salt
02:32 jerrykan[m] Poppabear: I haven't use the supervisor formula, but with formulas in general you shouldn't need to do anything with the map.jinja file
02:33 jerrykan[m] if there is anything in there that needs overriding you would use pillar to do it
02:33 jerrykan[m] https://github.com/saltstack-formulas/supervisor-formula/blob/master/pillar.example
02:34 jerrykan[m] looking closer at the map.jinja file, it it is indeed lacking support for non debian systems
02:34 Poppabear i use the pillar, but to generate my config file, but not sure how to overwrite the defaults based on OS ?
02:38 hemebond joined #salt
02:38 jerrykan[m] the `lookup` section at the bottom of the `pillar.example` file shows an example of how to override the name of the service defined in `map.jinja`/`default.yaml`
02:38 hemebond Is anyone using the state.test.* modules to create dependencies between other states?
02:39 Poppabear jerrykan[m]: right, but how do i do that based on the different OS's ?
02:42 jerrykan[m] you could do it using a lot of if statements in your pillar (not recommended)
02:46 jerrykan[m] or fork the formula and use something like the apache formula as an example for how they handle multiple different distros
02:46 Poppabear yea, the other formula's like apache uses the map.jinja file ;)
02:46 Poppabear thats why i was trying to work that out
02:46 jerrykan[m] I think it just looks like the supervisor formula isn't very mature, so you'll probably have to do most of the work yourself... sorry
02:48 ilbot3 joined #salt
02:48 Topic for #salt is now Welcome to #salt! <+> Latest Versions: 2016.3.5, 2016.11.3 <+> 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
02:50 evle joined #salt
02:51 masber joined #salt
02:52 antpa joined #salt
02:53 DEger joined #salt
03:02 shoemonkey joined #salt
03:05 sh123124213 joined #salt
03:05 felskrone joined #salt
03:37 stooj joined #salt
03:39 eprice joined #salt
04:01 Klaus_Dieter joined #salt
04:13 scsinutz joined #salt
04:21 ahrs_ joined #salt
04:30 sh123124213 joined #salt
04:31 netcho joined #salt
04:33 tom29739 joined #salt
04:35 Klaus_D1eter_ joined #salt
04:40 mage_ joined #salt
04:43 preludedrew joined #salt
04:43 ahrs joined #salt
04:45 jimklo joined #salt
04:46 bocaneri joined #salt
04:46 leonkatz joined #salt
04:52 swa_work joined #salt
05:03 shoemonkey joined #salt
05:12 j4son joined #salt
05:13 stooj joined #salt
05:18 j4son joined #salt
05:18 msn joined #salt
05:22 Praematura joined #salt
05:28 stooj joined #salt
05:32 netcho joined #salt
05:34 mage_ joined #salt
05:36 stooj joined #salt
05:39 sh123124213 joined #salt
05:44 onlyanegg joined #salt
05:46 sh123124213 joined #salt
05:47 golodhrim|work joined #salt
05:52 rdas joined #salt
05:54 armyriad joined #salt
05:59 onlyanegg joined #salt
05:59 ninjada_ joined #salt
06:08 ninjada joined #salt
06:10 impi joined #salt
06:13 jas02 joined #salt
06:19 ninjada joined #salt
06:23 jas02 Hello, are there some examples howto manage entOS 7 firewalld with Salt? Documentation is very bad in case of state.firewalld.
06:35 sh123124213 joined #salt
06:38 ninjada joined #salt
06:46 cyteen joined #salt
07:04 shoemonkey joined #salt
07:10 candyman88 joined #salt
07:26 mavhq joined #salt
07:27 jas02 states.firewalld is broken. It's unable to handle salt.states.firewalld.service (I have newest version of Salt). Error message while applying:  firewall-cmd: error: unrecognized arguments: --service
07:36 duncanmv joined #salt
07:39 jas02 I got it. I have CentOS 7, where is older version of firewalld, I have to upgrade to the lates CentOS 7.3
07:46 antpa joined #salt
07:47 antpa joined #salt
07:47 impi joined #salt
07:48 antpa joined #salt
07:49 antpa joined #salt
07:50 antpa joined #salt
07:50 antpa joined #salt
07:51 antpa joined #salt
07:52 antpa joined #salt
07:53 antpa joined #salt
07:54 antpa joined #salt
07:54 ninjada joined #salt
08:13 netcho joined #salt
08:18 dariusjs joined #salt
08:25 darioleidi joined #salt
08:26 JohnnyRun joined #salt
08:27 ravenx joined #salt
08:27 ravenx hey guys, anyone have eperience here wiht the supervisord state?
08:27 ravenx when i use file.managed to add a supervisor app.conf there, it doe snot automatically start
08:27 ravenx there is no 'reread'/'update' option it seems and as a result the app doesn't get started
08:45 fracklen joined #salt
08:48 ninjada joined #salt
08:50 Rumbles joined #salt
08:54 o1e9 joined #salt
08:55 Hybrid1 joined #salt
09:00 MattP_ joined #salt
09:00 teclator joined #salt
09:02 John_Kang joined #salt
09:05 shoemonkey joined #salt
09:05 John_Kang our saltstack is super slow but i don't know which point shoud I look into
09:05 John_Kang can someone give me some advice ?
09:06 dunz0r John_Kang: I noticed a significant speedup when I gave the salt-master more threads
09:07 bdrung_work joined #salt
09:08 jas02 joined #salt
09:09 John_Kang salt-master has 128 threads at the moment, should I give it more ?
09:12 dariusjs joined #salt
09:21 s_kunk joined #salt
09:25 Hybrid2 joined #salt
09:29 impi joined #salt
09:29 babilen John_Kang: How many cores/ram and how many minions does it have?
09:31 John_Kang 16 cores / 32GB
09:33 babilen And minions?
09:35 John_Kang about 200
09:35 John_Kang 256
09:36 babilen That should be perfectly fine .. I'd read up on https://docs.saltstack.com/en/latest/topics/tutorials/intro_scale.html
09:37 babilen Here we have good experiences with, in particular: Switching to TCP transport, enabling random reauth and recon and running jobs in batch mode
09:37 babilen Also what is it about your saltstack that's slow?
09:37 babilen Applying the states? Running simple commands/ping?
09:37 Mattch joined #salt
09:38 John_Kang running salt state on minion
09:38 John_Kang it's too slow to use, it tasks about 5min or so
09:38 babilen Well, that has little to do with the master, but more with the states themselves.
09:39 babilen Running ten thousand states that install a bazillion packages will take time
09:39 babilen What are you doing in your states that takes longer than it should?
09:40 John_Kang If I have bash script instead of salt's state, it will be faster than state, I think --;
09:40 mike25de joined #salt
09:41 John_Kang our state just do build java application, so installing tomcat, deploy war files and run it
09:41 John_Kang that's all
09:42 babilen Did you run it on your minion with "salt-call -ldebug state.apply" to see where the delay comes from? Is this reproducible if you run the states masterless?
09:43 John_Kang it was the same :(
09:43 John_Kang we usually use salt-call from minion
09:43 babilen So, what is it spending time on?
09:44 John_Kang sorry?
09:44 babilen Yeah, but that's not masterless (the data is still being requested from the master)
09:44 babilen So, you ran "salt-call -ldebug state.apply" ?
09:44 babilen Shouldn't be too tricky to spot what causes the delay
09:45 Hybrid1 joined #salt
09:45 John_Kang failed to authenticate message
09:46 John_Kang I can see some of that message from minion
09:46 John_Kang but It had connected successfully, do you have an idea about that message ?
09:47 babilen I would recommend the following: 1. Run your states masterless (as in: https://docs.saltstack.com/en/latest/topics/tutorials/standalone_minion.html) 2. Set ping_on_rotate: True in your master configuration and ping from your minion before executing a highstate 3. Running the state and analysing each step in regard of the time it takes to complete it and if that is an unacceptable amount of time
09:47 toanju joined #salt
09:48 John_Kang okay, let me try
09:48 babilen Once you've done that you could collect (pseudo-anonymised?) logs and paste them to http://paste.debian.net, https://gist.github.com, http://sprunge.us, … for us to investigate further
09:48 John_Kang got it
09:48 babilen (feel free to change the order of those things)
09:49 John_Kang babilen: thanks :)
09:49 babilen Good luck!
09:52 Hybrid joined #salt
09:54 antpa joined #salt
09:56 jas02 joined #salt
09:58 John_Kang babilen: hey sir, I've got to go now, I will try your advice tmw and give you an update
09:59 babilen Sure :)
10:01 Hybrid joined #salt
10:02 jas02 joined #salt
10:06 Tanta joined #salt
10:28 netcho joined #salt
10:29 MattP_ Hello everybody, does anyone knows if I can detect if a state was called via state.sls versus state.highstate ?
10:32 ronnix joined #salt
10:35 babilen MattP_: Whatfor? And when?
10:36 Tanta yes, you can inspect the process tree and find the invocation
10:36 ronnix joined #salt
10:36 Tanta that should have either salt-call state.sls or salt-call state.highstate in it
10:36 babilen Even after it ran? Or if called from the master?
10:36 Tanta only for an active run
10:38 N-Mi joined #salt
10:38 N-Mi joined #salt
10:38 MattP_ basically I want to put a require in state, but only if it was called via highstate
10:40 MattP_ @tanta how do you do this ?
10:41 Tanta onlyif: ps uax | grep "[s]alt-call state.highstate"
10:41 bo-12 joined #salt
10:42 Tanta that's the ghetto way
10:42 MattP_ yes, but I can't do it from the master
10:43 bo-x joined #salt
10:43 Tanta well that is the method I use
10:44 babilen MattP_: Why do you want to do this?
10:44 Tanta http://pastebin.com/raw/xc9WgkJz
10:44 Tanta here's my example
10:45 Tanta you can tell this runs from a masterless environment though, there's no remote master
10:46 babilen Tanta: Why would you do that?
10:46 Tanta because I provision grains from an S3 bucket
10:46 mocky joined #salt
10:46 Tanta they are managed separately
10:46 babilen And why don't you use reload_grains?
10:46 Tanta it doesn't work
10:47 babilen Or orchestrate?
10:47 Tanta salt-call appears to read /etc/salt/grains once-per-run
10:47 Tanta it's a nasty hack, just like whatever mattp is diong
10:47 Tanta because the use case for that solution is really narrow and tough to solve
10:48 babilen I am not sure we know MattP_'s use case yet
10:48 Tanta yeah I run masterless and from an autoscaled environment, there is a chicken + egg issue with provisioning the grains securely
10:48 Tanta it's where I store secrets
10:48 Tanta it has to be self-processing
10:49 msn joined #salt
10:49 MattP_ babilen: I have some nodes with two roles: graylog and mongodb; Graylog requires mongodb, so I can either put a require in graylog or a require_in in mongodb. The problem is that I want to be able to call the states independently for orchestration purpose, and that if I want to be able to call state.sls on either graylog or mongodb, the require will fail on one of them
10:49 Guest90567 how to get local primary ip in salt
10:50 XenophonF just to close the loop on my problems yesterday w/r/t gitfs, the correct fileserver_backend name is "git" not "gitfs" *smh*
10:51 MattP_ guest90567: not ideal, but salt-call grains.get ipv4:0 works
10:51 babilen Guest90567: network.ip_addrs type='public'
10:51 MattP_ babilen: much better :)
10:51 babilen (assuming that is what you think of as primary ip address)
10:52 XenophonF oh so that's how you get one list value out of the cli!
10:53 babilen MattP_: What I don't understand is "for orchestration purposes" -- What would be the problem if the states are both run as, as you rightfully point out, they actually depend on each other
10:54 Guest90567 I have different interfaces I need the primary ip of that machine
10:54 babilen Guest90567: What is the primary ip?
10:54 babilen Would that be 31.11.22.34, 185.1.22.185 or 10.0.1.123 ?
10:55 Guest90567 185.1.22.185
10:55 babilen And network.ip_addrs type='public' doesn't get you that?
10:55 MattP_ babilen: highstate can be slow, so being able to run just the needed state saves time
10:56 babilen MattP_: Sure, but what is the problem if that state actually triggers running of other states that are, well, required?
10:57 MattP_ babilen: because if makes the state fail, ie:
10:57 MattP_ ID: graylog-service     Function: service.running         Name: graylog-server       Result: False      Comment: The following requisites were not found:                                  require:                                      service: mongodb-service
10:57 babilen MattP_: Did you include the SLS that contains the mongodb-service state?
10:57 MattP_ sorry for the crappy output
10:58 MattP_ no, the graylog service does not include mongodb: the node has the two roles
10:58 Guest90567 network.ip_addrs type='public' it doesn't work
10:58 babilen MattP_: If you want to declare that dependency you have to include it
10:58 babilen Guest90567: Why not?
10:59 MattP_ babilen: I wanted to keep them separate, but you are right, I could do that
10:59 Guest90567 It doesn't give any thing
10:59 babilen MattP_: Well, I understand your sentiment, but my impression is that you want to express a dependency without actually expressing that dependency
11:00 babilen As graylog actually requires mongodb I see no harm in including the mongodb SLS/role in your graylog SLS/role
11:00 babilen (but not the other way round)
11:01 MattP_ babilen: I'm definitly testing this :) thanks
11:01 oaken_chris joined #salt
11:02 babilen Guest90567: How did you test it?
11:04 Guest90567 if I give  "network.ip_addrs " it will be giving 10.0.2.15, 172.28.128.4
11:04 Guest90567 but I need only 172.28.128.4
11:05 Guest90567 I test it by salt-call network.ip_addrs
11:05 MattP_ Guest90567: Do you have the subnet information of the target network available somewhere  ? If yes, you could use it to filter the results
11:06 shoemonkey joined #salt
11:10 babilen Guest90567: What's the exact command you ran?
11:11 babilen Ah, so what happens if you run "salt-call network.ip_addrs type='public'" ?
11:11 Guest90567 sudo salt-call network.ip_addrs
11:12 Guest90567 if I run sudo salt-call network.ip_addrs  type='public' It doesn't give any output
11:12 Guest90567 It seems to be blank
11:13 babilen Isn't 172.28.128.4 a private ip address?
11:14 Guest90567 Its a private address
11:14 babilen Which brings us back to the question: How do you decide which is your "primary" IP address?
11:14 babilen Is it an address within a specific CIDR?
11:14 babilen (and obviously type='public' won't give you a private address)
11:15 Guest90567 But it has two private addresses
11:16 babilen Yes, so how do you know which of the two is your "primary" ?
11:17 impi joined #salt
11:17 Guest90567 In these two address 10.0.2.15 is common for all the machines
11:17 evle joined #salt
11:18 babilen So, is it an address in a specific CIDR? (even if it's just 172.16.0.0/12 )
11:19 Guest90567 Yes
11:19 babilen Okay
11:20 babilen Run "salt-call network.ip_addrs cidr='172.16.0.0/12'" then
11:22 Guest90567 Okay
11:30 Guest90567 babilen: thank u
11:31 haam3r Is the top.sls file shorthand really an all or nothing thing, or might this be a bug?
11:31 Guest90567 its working
11:31 babilen Guest90567: Wonderful
11:31 babilen Guest90567: In case you need this information on other minions you can easily setup a mine function alias for earch CIDR you are interested in
11:32 babilen haam3r: Which file shorthand?
11:32 Guest90567 babilen: like?
11:32 haam3r babilen: I mean this: https://docs.saltstack.com/en/latest/ref/states/top.html#shorthand
11:33 theblazehen joined #salt
11:33 babilen Guest90567: http://paste.debian.net/919823/ would be an example -- https://docs.saltstack.com/en/latest/topics/mine/#mine-functions
11:33 haam3r babilen: If I only use it for some minions, interspersed with minions that have multiple states, it seems to break top file compiling
11:34 babilen haam3r: I guess it is "all or nothing"
11:34 haam3r babilen: Meaning I start getting  'Malformed topfile (state declarations not formed as a list)'
11:34 babilen If that constitutes a bug, I can't say
11:34 haam3r ack, thanks for the clarification
11:35 babilen Judging from the documentation I would, like you, expect it to be possible to intersperse the shorthand with non-shorthand definitions *iff* you use shorthand for systems to which you assign a single SLS
11:35 babilen Personally I wouldn't use the shorthand
11:37 haam3r babilen: Because?
11:38 babilen Primarily because I have very few boxes that only get a single SLS and because I'm switching to automatically generating the top.sls based on pillar data
11:38 cyborg-one joined #salt
11:38 babilen So I don't really care about vertical real estate
11:40 haam3r okey, makes sense. That auto generating of the top file sounds pretty interesting.
11:40 babilen Ping me in a week or two
11:45 haam3r Thanks. I'll try and remember :D
11:47 ronnix joined #salt
11:51 stewgoin joined #salt
11:55 dariusjs joined #salt
11:57 ronnix joined #salt
12:01 _Cyclone_ joined #salt
12:06 shoemonkey joined #salt
12:18 oaken_chris joined #salt
12:38 sh123124213 joined #salt
12:40 rylnd all, i am trying to run a sls based on the salt-cloud deploying event. i am deploying instances with FQDN, so i extract the subdomain from the name and try to address some pillar data with it. unfortunately it doesnt work http://pastebin.com/c9Twym10. can anyone help me out? thanks.
12:42 rylnd the timeout, port and algorithm values dont work. the subdomain works fine. i also tried ('envinfo:' + subdomain + ':dns:timeout') directly in the pillar.get, but that did not work either
12:46 Nahual joined #salt
12:47 sh123124213 joined #salt
12:59 golodhrim|work joined #salt
13:08 jas02 joined #salt
13:11 sh123124213 joined #salt
13:11 ssplatt joined #salt
13:18 jas02 joined #salt
13:20 CrummyGummy joined #salt
13:24 ronnix joined #salt
13:25 remijouannet joined #salt
13:30 scsinutz joined #salt
13:31 cachedout joined #salt
13:36 Electron^- joined #salt
13:36 sh123124213 joined #salt
13:45 scoates joined #salt
13:46 stewgoin joined #salt
13:50 racooper joined #salt
13:51 mpanetta_ joined #salt
13:54 ssplatt joined #salt
13:56 rylnd anyone an idea how to use a variable to get pillar data? similar to http://pastebin.com/c9Twym10. all my tries were not successful so far
13:59 antpa joined #salt
14:03 max42 I recently used pillar[key]
14:04 aldevar joined #salt
14:04 aldevar left #salt
14:04 mpanetta joined #salt
14:10 dariusjs joined #salt
14:13 morsik joined #salt
14:13 morsik Hi. Is there a ways to get Jinja rendering error >line<?
14:19 mike25de joined #salt
14:19 babilen rylnd: You seem to be doing that inside a string
14:20 netcho joined #salt
14:21 rylnd babilen: yes. i do. i should test that outside of the string. thanks!
14:21 cachedout joined #salt
14:26 mike25de hi guys - i have a weird issue. I use salt-api  to call salt states from my own app. So far so good.  I get from the salt-api a return json with the status of the job and output.  NOW i have enabled the mysql returner on the salt-master and used master_job_cache: mysql   While using this, the salt-api does not return the data anymore - probably because now is in my mysql database. IS it possible to somehow keep the local cache and also
14:30 Brew joined #salt
14:31 promorphus joined #salt
14:32 mikecmpbll joined #salt
14:34 jauz I wonder if you disable the returner entry but keep the master_job_cache entry if it will output it that way... ?
14:35 Ricardo1000 joined #salt
14:35 Ricardo1000 Hello
14:35 Ricardo1000 Does any body home ?
14:36 sh123124213 joined #salt
14:37 babilen Sorry, I'm at work
14:37 sarcasticadmin joined #salt
14:40 jauz Also at work! Hello!
14:40 Roelt My cat is at home
14:41 Praematura joined #salt
14:43 sh123124213 joined #salt
14:45 rylnd babilen: even when i do it outside of the string, it does not work unfortunately
14:45 jas02 joined #salt
14:47 mike25de jauz: good idea
14:48 babilen rylnd: What's the actual problem?
14:48 babilen But let me fix ENOCOFFEE first
14:48 rylnd babilen: babilen
14:48 babilen I am?
14:48 rylnd lol
14:49 rylnd babilen: i would like to use a variable inside the pillar.get and i thought it was possible
14:49 rylnd babilen: but either it is not or i just did not find the right way yet
14:50 rylnd babilen: i need to look up some pillar data that is different for each environment. i am getting the environment from the subdomain of the FQDN of the host.
14:50 Ricardo1000 How often new releases are roll out ?
14:50 mike25de rylnd: how do u use it?
14:52 rylnd mike25de: see http://pastebin.com/c9Twym10. i tried to define it outside the string as well
14:54 muxdaemon joined #salt
14:56 babilen http://pastebin.com/raw/c9Twym10
14:56 gableroux joined #salt
14:57 DEger joined #salt
14:58 babilen rylnd: So, what's the issue exactly and where is 'data' coming from?
14:59 ronnix joined #salt
15:00 Tanta joined #salt
15:01 rem5_ joined #salt
15:06 promorphus joined #salt
15:07 rylnd babilen: i think i found the issue. i am running that in a sls that is triggered by a reactor. i just re-read the reactor documentation and saw that pillar data is not available in the reactor context
15:08 babilen aye
15:11 rem5 joined #salt
15:13 promorphus joined #salt
15:15 rem5_ joined #salt
15:18 ronnix joined #salt
15:23 scsinutz joined #salt
15:29 Shirkdog joined #salt
15:29 Shirkdog joined #salt
15:31 seanz joined #salt
15:34 ny joined #salt
15:38 rem5 joined #salt
15:38 fracklen joined #salt
15:38 edrocks joined #salt
15:38 ny Good morning. I was wondering if anyone can advise me on something. I have set up a cloud provider config on salt master and am able to connect to any of the vcenters defined. When using the salt-cloud command I can reset a vm in by running it like this from salt master "salt-cloud -a reset <vm hostname>"
15:39 ny how can I define the above salt-cloud command in an sls file?
15:39 netcho hi all, i have set  'Ebs.VolumeType: standard' for root volume in my cloud.profile but it creates gp2(SSD) instead
15:39 gtmanfred you would need to get the provider available on the minion
15:39 netcho docs say standard is default value
15:40 netcho but looks like not
15:40 dariusjs joined #salt
15:40 ny Thbe provider is vmware
15:40 gtmanfred yeah, the provider would need to be available on the minion, either in pillars or in /etc/salt/cloud.profiles.d on the minion and then use a module.run and https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.cloud.html#salt.modules.cloud.action
15:41 raspado joined #salt
15:41 oaken_chris joined #salt
15:41 ny Thanks gtmanfred. I have already setup a vmware.conf file under /etc/salt/cloud.profiles.d and when running any salt-cloud command on cli these all work fine
15:41 jesusaur joined #salt
15:42 gtmanfred did you set it up on the minion or on the master?
15:42 ny i have set all of that up on the master
15:42 whytewolf ny, what he is saying. in order for that to work. the system that has salt-cloud working also needs a salt-minion that is accepted by the master
15:42 ny so via cli all salt-cloud commands work ok and I can connect to any vcenter
15:42 Inveracity joined #salt
15:42 gtmanfred if you are using the minion on the master to reset it, then you are fine, but if you are using another minion to reset it, it needs to be setup on the salt-minion
15:43 ny gtamfred yes I am wanting to use the minion on the master to reset it
15:43 gtmanfred ok, then you are good
15:43 ny thanks
15:43 gtmanfred just use a module.run and the cloud.action module
15:44 ny do you have an example of how this hsould be defined in an sls file using module.run?
15:44 gtmanfred https://docs.saltstack.com/en/latest/ref/states/all/salt.states.module.html
15:44 gtmanfred reset minion:
15:44 gtmanfred module.run:
15:44 gtmanfred - name: cloud.action
15:44 gtmanfred - fun: reset
15:45 gtmanfred - instance: minionname
15:45 whytewolf might wanna make sure there is something there that makes sure it doesn't happen every state run
15:45 ny brilliant, thanks alot for your help :-)
15:46 shoemonkey joined #salt
15:47 dariusjs joined #salt
15:50 sh123124213 joined #salt
15:51 tkojames joined #salt
15:51 netcho gtmanfred: can youer please take a glance? https://hastebin.com/owewawosez.php
15:51 netcho this profile for some reason creates gp2 volume
15:51 netcho instead fo standard
15:53 whytewolf is it supposed to be sda? not sda1?
15:54 whytewolf sda1 would be a partition on sda
15:54 netcho yes
15:54 netcho but it creates volume
15:54 gtmanfred sorry, i cannot, i am working on a customer issue
15:54 netcho if i comment out line 14 it should use default values
15:55 netcho docs say default is standard, but it still creates gp2 one
15:55 netcho i can try removing '1' from sda1 but it was working fine so far
15:56 * whytewolf shrugs. take anything i say about aws with a large bag of salt.
15:56 sh123124213 joined #salt
15:56 whytewolf i don't use it for systems. only push things to s3 and use it for route53
15:56 rylnd thanks for putting me onto the right track babilen. it helps to "talk" to people when sitting alone in a home office environment
15:57 rylnd can anyone explain to me how i must format my kwargs when i get 'DeprecationWarning: kwargs must be passed inside the low under "kwargs"' warning? i am at a bit of a loss here. thanks.
15:58 nbuchanan joined #salt
15:58 babilen rylnd: np :)
15:58 jauz How do people usually hide their sensitive data in that needs to exist in config files without pillar data?
15:59 babilen jauz: Why 'without pillar' ?
15:59 netcho whytewolf: when /dev/sda is set instead of /dev/sda1 salt returns profile error
15:59 gtmanfred jauz: config.get?
15:59 gtmanfred or sdb.get
15:59 gtmanfred https://docs.saltstack.com/en/latest/topics/sdb/
16:00 jauz Well, I have my returner working for example in my own case but the api_key has to be mentioned in the config file of the minion or master, so I can't reference it in a safer way as far as I know since I can't put a pillar reference in a config file, can I?
16:00 gtmanfred use sdb://
16:00 jauz Ahhh.
16:00 impi joined #salt
16:00 jauz I had seen sdb but didn't know I could use it in config.
16:01 whytewolf jauz: ... a lot of settings can be pillars also
16:01 gtmanfred i use it here, https://github.com/gtmanfred/blog-sls/blob/master/salt/reactor/blog.gtmanfred.com.sls#L1
16:01 whytewolf such as mysql config settings
16:01 promorphus joined #salt
16:01 gtmanfred but you should also be able to put the returner configs in pillar data
16:05 PatrolDoom joined #salt
16:05 whytewolf there should be a list of settings that use config.get so we know what can use pillars as a config option
16:07 fracklen joined #salt
16:08 gtmanfred :+1:
16:09 dariusjs joined #salt
16:10 netcho ok so i commented out whole block_device_mapping and it created gp2 by default
16:10 netcho is it a bug?
16:10 whytewolf could it be a change in aws?
16:11 jauz I had originally setup my returner in pillar entirely, but I was using jinja and then my config wouldn't run. I see now I just need to reference it another way. Neato.
16:13 netcho whytewolf: when creating machine from aws console defualt is gp2. does salt use that default option or it forces type:standard? if it uses aws as a default fine, but can i then force it to use magnetic(standard)?
16:16 Praematura joined #salt
16:16 whytewolf netcho: I do not know enough about aws to know the answer to that
16:18 netcho roger that, but salt docs say 'type: choose between standard (magnetic disk), gp2 (SSD), or io1 (provisioned IOPS). (default=standard)'
16:19 netcho but looks like default is gp2, so error in docs or? i cannot manage to create a machine with magnetic root volume
16:19 javoskamp left #salt
16:20 whytewolf netcho: it could be an issue with the aws api also. which is why i say i don't know enough about aws to say
16:20 netcho sure no problem :)
16:21 rylnd babilen: you were the lucky charm the last time ;-). any idea how to prevent this from happening? http://pastebin.com/tLFcFxsR. it runs fine, but it seems i have to escape the colon maybe.
16:23 netcho creating volume itself with: 'salt-cloud -f create_volume my-ec2-us-east-1a size=10 type=standard zone=us-east-1a' creates a volume with correct type
16:23 fracklen joined #salt
16:23 DammitJim joined #salt
16:24 woodtablet joined #salt
16:24 netcho but doing the same thing from profile it does not
16:24 babilen rylnd: I will if you stop using pastebin.com -- There are many other nice ones such as http://paste.debian.net, https://gist.github.com, http://sprunge.us, … that don't drown you in ads, require captchas or are ugly
16:25 gtmanfred and that don't use flash to display their ads
16:25 rylnd babilen: deal
16:25 rylnd i dont see the ads, so i thought it was ok
16:26 gtmanfred also, it adds windows line endings to the raw paste
16:26 whytewolf actually i know many places block pastebin.com so i typically use that as an excuse not to go there. even though i work from home
16:26 whytewolf ;)
16:26 gtmanfred yeah, the started blocking it when one of their ads was spreading malware from flash
16:27 babilen rylnd: The indentation seems to be off and you might want to use double quotes
16:27 babilen Can't really tell as pastebin.com loves to fuck up indentation also
16:27 babilen But to me it looks as if line 20 has an extra leading space
16:28 rylnd babilen: https://gist.github.com/jbfriedrich/df8ed93ce33d7ddfdfa3e5b67ff433ea better?
16:28 babilen muchos
16:28 rylnd babilen: tried double and single quotes, same error message
16:28 ronnix joined #salt
16:28 babilen rylnd: What about message|json ?
16:29 rylnd babilen: let me try that
16:29 babilen Unfortuantely I have to leave soon :-/ -- Escaping the colon might also be worth a try
16:30 babilen err .. yaml naturally
16:31 rylnd babilen: i'll run that now. thanks so far for your help!
16:36 rylnd babilen: seems like escaping doesnt work :-/
16:36 dariusjs joined #salt
16:36 gtmanfred rylnd: kwarg should be :...
16:37 gtmanfred channel : '#channel'
16:37 muxdaemon joined #salt
16:37 gtmanfred commented
16:38 rylnd gtmanfred: thanks! but when i do it that way than i get the deprecated warning that kwargs is not low. what would be the right to do it, so that its not deprecated?
16:39 ronnix joined #salt
16:39 gtmanfred not low?
16:40 gtmanfred rylnd: from this https://docs.saltstack.com/en/latest/ref/states/all/salt.states.saltmod.html#salt.states.saltmod.function it should be dictionary
16:40 gtmanfred kwarg
16:40 gtmanfred The dict (not a list) of keyword arguments to pass into the functio
16:40 gtmanfred if there is a deprecation message, please open an issue
16:41 seanz1 joined #salt
16:42 dariusjs joined #salt
16:43 muxdaemon joined #salt
16:43 rylnd gtmanfred: thanks. so you say your example, that you put in your reply under the gist should work and should not produce a deprecatedWarning. but if it does, i should open an issue, correct?
16:43 muxdaemon joined #salt
16:44 Trauma joined #salt
16:44 gtmanfred yes
16:45 Satyajit joined #salt
16:46 sh123124213 joined #salt
16:48 fracklen joined #salt
16:50 lilvim joined #salt
16:54 cdunklau hmm, is there a way to get the minion's id using salt-call --local?
16:55 gtmanfred salt-call --local grains.get id
16:55 cdunklau sweet thanks
16:56 _JZ_ joined #salt
16:56 gtmanfred np
16:56 gtmanfred that can be overwritten fyi, if someone hard codes the id to /etc/salt/grains
16:57 gtmanfred if you actually use config.get it will check the minion config first
16:57 gtmanfred hrm, but that doesn't work, nevermind
16:57 gtmanfred cdunklau: you can just cat /etc/salt/minion_id
16:57 cdunklau gtmanfred: config.get id   works for me...
16:57 gtmanfred yeah, but that is still pulling it from the grain, not the minion config
16:57 cdunklau right, but if that's set somewhere...
16:58 cdunklau like, in the minion conf
16:58 gtmanfred the definitive way find it, is just check /etc/salt/minion_id
16:58 whytewolf cdunklau: order of config.get has grains after the config file
16:58 gtmanfred that is the minion_id it is connected to the master with
16:58 cdunklau gtmanfred: ah, okay, so if i explicitly set the minion id in /etc/salt/minion, when the minion starts it will set /etc/salt/minion_id ?
16:59 gtmanfred yes
16:59 cdunklau gotcha. so that opens another question: what's the best way to explicitly set the minion id?
16:59 cdunklau in a config file, or with /etc/salt/minion_id ?
17:00 gtmanfred i like /etc/salt/minion, but i think salt-cloud sets it in /etc/salt/minion with the id: key
17:00 gtmanfred i like /etc/salt/minion_id...
17:00 cdunklau that's what i'm using now
17:01 cdunklau i also avoid using the main conf file and instead stick stuff in minion.d
17:01 netcho joined #salt
17:01 cdunklau because boo parsing and hacking config files automatically. i feel like it's much easier to just segregate partial config into separate files
17:01 whytewolf i stick stuff in minion.d but also use pillar conf for anything i know i can get away with
17:03 cdunklau realistically i'm doing minimal config on the minions though. master dns name, master fingerprint, and an explicit minion id prefix-<UUID1> so it has the MAC address
17:04 cdunklau i don't think i'll need anything else
17:09 leonkatz joined #salt
17:14 WKNiGHT joined #salt
17:14 brakkisath joined #salt
17:17 seanz joined #salt
17:18 promorphus joined #salt
17:22 Edgan joined #salt
17:23 cscf joined #salt
17:24 sh123124213 joined #salt
17:30 sp0097 joined #salt
17:33 yuhl______ joined #salt
17:33 sp0097 joined #salt
17:34 cliluw joined #salt
17:35 jrgochan hello. is there a way in the /etc/salt/master file to specify an event_return_whitelist with only certain functions? For example, I'm currently returning all "salt/job/*/ret/*", but only want "state.apply" and "state.highstate" functions to be returned
17:35 jrgochan right now it's returning all functions, which is getting... verbosey
17:36 sh123124213 joined #salt
17:39 scsinutz joined #salt
17:42 uk00123 joined #salt
17:45 scsinutz joined #salt
17:48 shoemonkey joined #salt
17:49 s_kunk joined #salt
17:52 GnuLxUsr joined #salt
17:53 brakkis__ joined #salt
17:58 Rumbles joined #salt
18:04 edrocks joined #salt
18:05 jhauser joined #salt
18:08 KyleG joined #salt
18:08 KyleG joined #salt
18:26 Rumbles joined #salt
18:28 Electron^- joined #salt
18:31 netcho_ joined #salt
18:32 DammitJim joined #salt
18:33 pipps joined #salt
18:43 Rumbles joined #salt
18:47 theblazehen joined #salt
18:51 s_kunk joined #salt
18:53 scsinutz joined #salt
18:57 nbuchanan joined #salt
19:00 fracklen joined #salt
19:02 sh123124213 joined #salt
19:13 jrgochan joined #salt
19:18 sh123124213 joined #salt
19:18 shakalaka joined #salt
19:21 edrocks joined #salt
19:21 leonkatz joined #salt
19:21 cyborg-one joined #salt
19:25 jas02 joined #salt
19:27 ALLmightySPIFF joined #salt
19:32 debian112 joined #salt
19:34 pipps joined #salt
19:36 nixjdm joined #salt
19:40 gtmanfred whytewolf: if you get a minute, can you validate that I am not crazy with this https://github.com/saltstack/salt/pull/39527#issuecomment-286536308
19:40 saltstackbot [#39527][MERGED] Add splay support for all scheduling criteria | What does this PR do?...
19:42 debian112 joined #salt
19:43 netcho_ joined #salt
19:46 prg3 joined #salt
19:48 shoemonkey joined #salt
19:49 gtmanfred nevermind, i made a bug
19:49 gtmanfred https://github.com/saltstack/salt/issues/40022
19:49 saltstackbot [#40022][OPEN] [Develop] Scheduler only runs jobs one time | Description of Issue/Question...
19:50 rylnd gtmanfred: thanks for your help so far. it almost works now. the only other issue i have is how to call the ddns.add_host runner from the orchestrator. i tried several things, this is where i am right now https://gist.github.com/jbfriedrich/ceeddb2b19bc8b28f9dfa065bbb33fcc. you prob see instantly what i am doing wrong
19:50 rick_ joined #salt
19:51 Rumbles joined #salt
19:54 whytewolf kawrgs with the runner in saltmod are funny. you don't actually put kargs in everything passed to the runner that isn't normal to the runner is passed in as part of the kawrgs
19:54 gtmanfred don't put it in kwargs
19:54 gtmanfred just put it directly on the same line
19:54 rick_ I'm interested in writing a Salt event streaming service to Kafka.  Anyone know if there is a module out there that does already?  Otherwise I was going to write an Engine to do it.  Is that something I should do on a Salt GH fork and give back w/ a pull request or just do in my own repo?
19:54 whytewolf also sorry gtmanfred was in a meeting, glad you figured it out
19:55 gtmanfred rylnd: commented
19:55 gtmanfred whytewolf: no worries :D
19:55 rylnd gtmanfred: life saver! thanks!
19:55 gtmanfred rick_: pull request would be great!
19:55 gtmanfred and engine would be correct!
19:56 rick_ great, i'll give it a shot.  Just didn't want to waste my time if it wasn't the right place to go...
19:56 gtmanfred note, there is a kafka returner, so you might look at that for configuration options
19:56 rylnd gtmanfred: i am just a newbie or does every runner feel a little bit different? :-/
19:56 rick_ i want to stream everything to kafka, so i thought an engine would be best
19:56 gtmanfred every way to do something is just slighly different that it is annoying
19:56 gtmanfred I have to read the docs to figure it out still
19:57 rylnd gtmanfred: then i am a little bit relieved
19:57 gtmanfred rick_: engine would be perfect, I wrote an engine that takes all the events on the hpe one view event stream and puts it on the salt event stream
19:57 rick_ great, thanks for the info.  Hopefully you'll see a PR in the near future
19:58 gtmanfred rick_: also, check out the hipchat engine, it shows streaming the salt event stream to hipchat, but allows to customize which tags to send
19:58 gtmanfred err, logstash*
19:58 gtmanfred not hipchat
19:59 gtmanfred https://github.com/saltstack/salt/blob/develop/salt/engines/http_logstash.py
20:00 rick_ i'll model it after that, should be pretty straight forward to put a kafka producer in there in place of the ES http call
20:00 rylnd gtmanfred: i am still confused, after https://docs.saltstack.com/en/latest/ref/states/all/salt.states.saltmod.html#salt.states.saltmod.runner and https://docs.saltstack.com/en/latest/ref/runners/all/salt.runners.ddns.html#salt.runners.ddns.add_host i would have never guessed you just write it on the same line
20:01 gtmanfred rylnd: **kwargs means to pack all other key word arguments that don't have a spot in the () list into the `kwargs` key
20:01 gtmanfred so, you don't have to specify kwargs:, just everythign else in the definition will be put into kwargs
20:01 gtmanfred and then unpacked onto ddns.add_host
20:02 rylnd gtmanfred: thanks!
20:02 gtmanfred no problem
20:02 rylnd gtmanfred: i need to convince my company to hire you for consulting
20:02 gtmanfred if you see *args,with one asterisk, it packs up positional arguments that do not have a key into a list
20:03 gtmanfred pay #salt :P
20:03 rylnd gtmanfred: i am working on it :-)
20:03 gtmanfred https://saltstack.com/support/
20:03 rylnd gtmanfred: i got the message that we are "doing too good" so that no one followed through with that support option
20:04 gtmanfred yeah that seems to be the problem...
20:04 gtmanfred it is just too easy to use
20:05 whytewolf ugh. I remeber a time when managers would rather cut off their own arm then go with out a support. regardless of ease of use. because it meant they had an insurence that could be relied on
20:05 pipps joined #salt
20:05 gtmanfred now we are all devops all in one monkesy
20:05 rylnd gtmanfred: we could have saved two days of frustrations. i am most def working on it as i would love to have a monthly budget to sit down with you guys to ask stupid questions like this and learn and get best practises and stuff. i hope there is some movement on this matter this quarter
20:05 gtmanfred :D
20:06 rylnd all in one monkeys that can be outsourced
20:06 rylnd because the overseas people are as good, right?
20:07 brakkisath joined #salt
20:07 gtmanfred look at the keystone tests in our test suite and you will see they are not
20:08 rylnd lol
20:08 whytewolf humm. I have never had my job threatened to be outsourced.
20:08 rylnd the higher you go up, the more think that anyone with a pulse can do everything and you need at least one devops teams, more than one if you can...
20:09 rylnd whytewolf: depends - i guess - how much the VP understands of IT. i have found many specimens who think that you just need a "devops team" somewhere overseas and they will be awesome as anyone can do that IT work :(
20:09 rylnd every time i hear "devops team" my toenails roll up a little bit more
20:10 gtmanfred i just cringed at teh description of that...
20:10 gtmanfred the toenails thing
20:10 rylnd haha
20:10 rem5 joined #salt
20:10 gtmanfred s/description/thought
20:12 jmedinar joined #salt
20:13 toanju joined #salt
20:14 jmedinar Q: How can I set a delay after before I move to the next state into an SLS?
20:14 jmedinar the first state generates a file from a DB and is copied into an SSHFS that takes a few seconds to complete
20:14 whytewolf I think it is the fact that I generally am on the forfront of the company i work at. generally i am pushing the ideas that drive a company. [when i can get away with it] so i am in the C*O sites a lot.
20:15 pipps joined #salt
20:15 whytewolf unforchantly i do tend to take a brunt if things go south
20:16 jmedinar I can just execute a delay in the minion and wait for it to come back... but is there something that doesnt depend on the minion
20:17 pipps joined #salt
20:17 * whytewolf shrugs module.run with a test.sleep and a require that it returns
20:18 rylnd whytewolf: yes, i am similar. but here we dont get into the C** circles at all. they dont mingle with peasants i guess. its hard to get heard here. maybe because i am remote, different country as main office etc.
20:18 jauz I wonder if he could do some sort of "requires..." that specifies that the output file in SSHFS must exist before executing the next step of the state file?
20:19 whytewolf jauz: no, that kind of require would just fail. not wait for it to be done
20:20 jmedinar yeah problem is it does exist right from the moment the first state starts...
20:20 jmedinar the file is completed and then transferred into SSHFS
20:20 jmedinar I will try the test.sleep
20:25 quique joined #salt
20:27 netcho_ joined #salt
20:31 GothAck joined #salt
20:36 pidydx joined #salt
20:37 pidydx Does anyone here have experience with using Hashicorp Vault with Salt?  I can't seem to find a way to handle pillars that are nested
20:39 pipps joined #salt
20:42 Louis_ joined #salt
20:43 promorphus joined #salt
20:44 ninjada joined #salt
20:44 rylnd gtmanfred: i declared victory too early lol. just updated the gist with the new error message and how the sls looks now https://gist.github.com/jbfriedrich/ceeddb2b19bc8b28f9dfa065bbb33fcc
20:45 Louis_ Hello, I'm new to SaltStack and trying to implement a beacon that returns the state of specific docker containers via the dockerng execution module. When I call dockerng.state(...), an exception is thrown within dockerng saying global name __context__ is not defined. Is there something that I need to do to initialize the __context__ in my beacon?
20:46 Louis_ I am using the latest docker-py python module as my docker api
20:46 edrocks joined #salt
20:49 aldevar joined #salt
20:50 scsinutz joined #salt
20:56 whytewolf also, w00t my new laptop just got here
20:56 rylnd gtmanfred: is it possible that two times "name:" just breaks that runner call?
20:57 rylnd whytewolf: woohoo! what did you get?
20:57 whytewolf a system76 oryx pro 15"
20:58 rylnd ok, i had to google that. never heard of system76 before
20:58 rylnd looks neat!
20:59 whytewolf it is replacing my old macbook pro retina, so it better perform great
20:59 lamer14895246260 joined #salt
21:00 rylnd whytewolf: what os are you running? ubuntu? or can you do windows too?
21:00 rylnd i have to admit, i am not a fan of linux as a desktop/laptop os
21:00 whytewolf it can do windows.... but i have exactly 1 windows system in my home and that is my gaming desktop. that will always remain the only windows box in my place
21:01 pipps99 joined #salt
21:01 pipps_ joined #salt
21:01 whytewolf it is more of a workstation for me then a true desktop. I am reaplcing a system i speand 90% of my time looking at a console anyway
21:02 rylnd whytewolf: i am waiting what apples does for hardware this year. after 6 years i need a hw upgrade. but if they dont come out with something decent i have to build my own with windows 10 i am afraid
21:02 rylnd whytewolf: gotcha
21:05 antpa joined #salt
21:05 whytewolf honestly i find i can never work on windows. i get to distracted by flashing lights.
21:06 rylnd for me its a real alternative, especiallyt when they include ubuntu 16.04. subsystem with their next update
21:06 ninjada joined #salt
21:06 nemi_ joined #salt
21:07 whytewolf eh, I've used the subsystem and it is meh compared to a real linux
21:07 rylnd you mileage may vary, but it worked for all i did on my local machine. i like it though that they implemented it
21:07 whytewolf it isn't even half as powerful as cygwin is.
21:08 rylnd i always had trouble with cygwin, it never fully behaved the same way. very minor things but very annoying. but that was ~13 years ago. maybe i need to take another look at cygwin
21:09 pipps joined #salt
21:14 funabashi joined #salt
21:18 promorphus_home joined #salt
21:19 xet7 joined #salt
21:23 mschiff "onchanges" seems buggy in 2015.5.10: Docs say "If a state has multiple onchanges requisites then the state will trigger if any of the watched states changes." but in fact, *every* watched state needs to have changes on order to trigger the state... anyone with the same experience?
21:28 scsinutz joined #salt
21:29 * whytewolf doesn't remeber enough about 2015.5.10
21:34 AndreasLutro mschiff: that behaviour was changed fairly recently, check docs for your version
21:36 mschiff AndreasLutro: Thanks. Indeed I was looking at a more recent version of the docs and I was relying on my assumption that such a change would have been mentioned with a note since which version...
21:41 scsinutz joined #salt
21:44 pidydx How do people store secrets in their pillars here?
21:44 ninjada joined #salt
21:49 pipps joined #salt
21:49 shoemonkey joined #salt
21:50 sh123124213 joined #salt
21:57 rpb joined #salt
21:58 dqx joined #salt
21:58 dqx left #salt
21:58 Praematura joined #salt
21:59 ninjada joined #salt
22:00 pipps joined #salt
22:00 ninjada joined #salt
22:00 impi joined #salt
22:05 pipps99 joined #salt
22:10 pipps joined #salt
22:13 relidy pidydx: gpg
22:15 Ryan_Lane pidydx: I use a secret management system
22:16 Ryan_Lane and have nodes directly access the secret management system, rather than relying on the salt master
22:16 relidy Would love to have something like that available here.
22:16 Ryan_Lane I'm not sure if salt ever made it possible to have minions use external pillars
22:17 Ryan_Lane we use masterless, so each minion has an external pillar configured to fetch secrets from the secret management system
22:17 pidydx @Ryan_Lane:  So you have to rewrite states/formulas to fetch keys?
22:17 Ryan_Lane we specifically use https://github.com/lyft/confidant
22:18 Ryan_Lane which has support in salt directly, as an SDB: https://github.com/saltstack/salt/blob/develop/salt/sdb/confidant.py
22:18 Ryan_Lane and as an external pillar: https://github.com/saltstack/salt/blob/develop/salt/pillar/confidant.py
22:18 antpa joined #salt
22:19 pidydx Does confidant handle nested pillars?
22:19 Ryan_Lane (confidant is only usable in AWS, since it depends on AWS services, but you could do the same thing using vault)
22:19 pidydx mysql: database: user: password=foo
22:20 Ryan_Lane confidant is just a dumb crud server, for the most part. you store values, which can be anything, and the pillar values are based on the string representation of the values
22:20 Ryan_Lane so if you want the data to have meaning, you'd need to json serialize it
22:21 Ryan_Lane that said, confidant supports extensible metadata, so in theory you could specify that a particular key has json data, and have the client auto-serialize it for you
22:21 Ryan_Lane but that's not built into the client by default.
22:21 Ryan_Lane you'd probably want to write your own client for any service you use
22:22 pidydx So the root of what I am trying to figure out is if a Formula expects a secret in a nested set of keys how do I store that secret somewhere and get it to populate that piece of the pillar
22:23 Ryan_Lane json serialize/deserialize is probably your best option
22:23 Ryan_Lane you could also use the gpg support in salt
22:23 Ryan_Lane I think when it decrypts it also deserializes into a proper pillar structure
22:25 scsinutz joined #salt
22:25 pidydx I guess I will keep looking.  I really wish Salt had a better variable management system :(
22:31 hatifnatt joined #salt
22:46 Trauma joined #salt
22:48 seanz joined #salt
22:49 debian112 joined #salt
22:52 netcho_ joined #salt
23:01 Renich joined #salt
23:07 leonkatz joined #salt
23:11 bocaneri joined #salt
23:12 rem5 joined #salt
23:12 rem5 joined #salt
23:17 scsinutz1 joined #salt
23:22 rschneberger joined #salt
23:32 pipps joined #salt
23:32 whiteinge sounds like he was looking for sdb.
23:34 mavhq joined #salt
23:35 oaken_chris joined #salt
23:36 pipps joined #salt
23:45 shoemonkey joined #salt

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