Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-01-05

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

All times shown according to UTC.

Time Nick Message
00:01 whytewolf retake: typically they are done top to bottom. unless things like includes and order are used. which can throw them out of alignment. orchestrate tends to be more heavy handed about ordering. also you can use state.show_highstate to see how the order is being thought out
00:02 huddy joined #salt
00:03 denys_ joined #salt
00:04 cyborg-one joined #salt
00:07 Vaelatern joined #salt
00:10 lompik joined #salt
00:13 otter768 joined #salt
00:21 Vaelatern joined #salt
00:35 zmalone joined #salt
00:48 Vaelatern joined #salt
00:49 dyasny joined #salt
00:49 subsignal joined #salt
00:59 auzty joined #salt
01:02 cberndt joined #salt
01:19 yomilk joined #salt
01:21 otter768 joined #salt
01:27 zmalone joined #salt
01:29 hasues joined #salt
01:29 hasues left #salt
01:30 PeterO joined #salt
01:33 linjan_ joined #salt
01:40 hasues joined #salt
01:41 hasues left #salt
01:46 yomilk joined #salt
01:47 vit1251 joined #salt
01:51 ronrib joined #salt
01:54 baweaver joined #salt
01:56 yetAnotherZero joined #salt
01:58 wt joined #salt
02:18 comjeito joined #salt
02:22 PeterO joined #salt
02:23 catpigger joined #salt
02:31 racooper joined #salt
02:40 favadi joined #salt
02:49 vit1251 joined #salt
02:53 subsignal joined #salt
02:59 favadi joined #salt
03:01 ryau joined #salt
03:02 ryau Hi I ran into this problem while trying to upgrade my salt using the bootstrap script:
03:02 ryau curl -L https://bootstrap.saltstack.com | sudo sh -s -- -Z -M git 2014.7
03:02 ryau --> Finished Dependency Resolution
03:02 ryau Error: Package: yum-utils-1.1.30-30.el6.noarch (repo-saltstack-el6)
03:02 ryau Requires: yum >= 3.2.29-56
03:02 ryau Installed: yum-3.2.29-43.el6.centos.noarch (@updates)
03:02 ryau yum = 3.2.29-43.el6.centos
03:02 ryau Available: yum-3.2.29-40.el6.centos.noarch (base)
03:02 ryau yum = 3.2.29-40.el6.centos
03:02 ryau You could try using --skip-broken to work around the problem
03:02 ryau You could try running: rpm -Va --nofiles --nodigest
03:02 ryau * ERROR: Failed to run install_centos_git_deps()!!!
03:02 ryau any idea what the problem or fix might be? Thank you in advance.
03:03 malinoff joined #salt
03:04 amcorreia joined #salt
03:07 iggy ryau: did you check the issue tracker?
03:08 ryau what and where is the issue tracker?
03:09 iggy ryau: on github
03:09 iggy https://github.com/saltstack/salt-bootstrap/issues
03:11 yomilk joined #salt
03:12 ryau iggy: thanks I'll take a look.
03:15 tzero joined #salt
03:16 ryau_ joined #salt
03:21 fgimian joined #salt
03:24 yomilk joined #salt
03:27 Vaelatern joined #salt
03:31 jeddi joined #salt
03:32 otter768 joined #salt
03:35 jaybocc2 joined #salt
03:35 quasiben joined #salt
03:42 tzero joined #salt
03:43 jaybocc2 joined #salt
03:46 zifnab so i'm missing something here
03:46 zifnab https://vomitb.in/0oXXTPy217
03:46 zifnab git repo has a 'states' and 'pillar' folder, each with a 'top.sls' inside of it
03:46 zifnab for some reason, i'm seeing the pillar stuff in /var/cache/salt/git_pillar, but i'm not seeing anything in there for states
03:47 zifnab any thoughts on what to look at?
03:47 zifnab nevermind
03:47 zifnab backend, not backends, i'm stupid
04:03 scoates joined #salt
04:08 scoates joined #salt
04:15 scoates joined #salt
04:20 nethershaw joined #salt
04:20 yomilk joined #salt
04:51 XenophonF joined #salt
04:53 XenophonF left #salt
04:54 subsignal joined #salt
04:55 jaybocc2 joined #salt
05:03 yomilk joined #salt
05:04 Bryson joined #salt
05:06 kshlm joined #salt
05:09 Bryson joined #salt
05:12 qman__ joined #salt
05:16 PeterO joined #salt
05:17 anmol joined #salt
05:29 scoates joined #salt
05:33 yomilk joined #salt
05:40 scoates joined #salt
05:46 rdas joined #salt
06:04 calvinh joined #salt
06:08 freelock joined #salt
06:20 XenophonF joined #salt
06:41 [dee] joined #salt
06:56 subsignal joined #salt
06:58 jxm_ joined #salt
07:03 colegatron joined #salt
07:06 jaybocc2 joined #salt
07:17 otter768 joined #salt
07:22 jaybocc2 joined #salt
07:29 calvinh joined #salt
07:31 calvinh_ joined #salt
07:33 KermitTheFragger joined #salt
07:35 jbax joined #salt
07:39 AlberTUX joined #salt
07:41 cyborg-one joined #salt
07:46 elsmo joined #salt
07:47 slav0nic joined #salt
07:47 evle1 joined #salt
07:59 Hydrosine joined #salt
08:00 impi joined #salt
08:05 hydrosine joined #salt
08:08 dgutu joined #salt
08:13 sybix joined #salt
08:33 harkx joined #salt
08:42 felskrone joined #salt
08:44 yomilk joined #salt
08:45 TyrfingMjolnir joined #salt
08:48 TyrfingMjolnir_ joined #salt
08:49 huds joined #salt
08:51 huds joined #salt
08:52 ingslovak joined #salt
08:55 AnotherMind|Away joined #salt
08:57 AnotherMind|Away hi
08:57 AnotherMind|Away i search info to share 1 chrome tab in other chrome tab via internet (no wifi)
08:57 AnotherMind|Away who can help me?
08:59 subsignal joined #salt
09:01 Rumbles joined #salt
09:03 mikeywaites joined #salt
09:07 cberndt joined #salt
09:08 Rumbles joined #salt
09:12 dijit wat
09:12 bhosmer joined #salt
09:18 otter768 joined #salt
09:19 s_kunk joined #salt
09:19 s_kunk joined #salt
09:23 Xevian joined #salt
09:25 GreatSnoopy joined #salt
09:27 felskrone joined #salt
09:32 darxmurf joined #salt
09:32 darxmurf hi all
09:33 eseyman joined #salt
09:34 jaybocc2 joined #salt
09:34 darxmurf my SALT memory is a bit rusty, I want to apply a package.sls file to my clients from command line, how should I procede ? as I have a /base/ folder with my main conf and the sls I want to apply is in /folder1/packagr.sls
09:39 dijit salt 'hostname' state.apply packag
09:39 dijit probably.
09:40 darxmurf well nop as the .sls file is not in the /base/ folder
09:41 darxmurf in my setup I have /base/top.sls and the file I want to apply is in /folder1/package.sls
09:41 darxmurf and when I try to apply it I get : - No matching sls found for 'packages' in env 'base'
09:42 jhauser joined #salt
09:42 dijit oh, you need to add the path to your master config then.
09:42 darxmurf it's in
09:42 XenophonF joined #salt
09:43 dijit then it should work. I guess I'm wrong.
09:43 darxmurf but when I run my cli I should give him the root isn't it ?
09:44 keimlink joined #salt
09:44 XenophonF darxmurf: to run a single state on a minion, use state.sls
09:45 av_ joined #salt
09:45 XenophonF e.g., given base: /srv/salt/foo.sls and minion crimson.example.com, salt crimson.example.com state.sls foo
09:45 darxmurf yep but when the .sls is in a different file_roots ?
09:46 XenophonF the same environment or a different environment?
09:46 XenophonF all of the folders in the same environment get merged
09:46 darxmurf this is my master conf : http://pastebin.com/3jXMYc9a
09:47 XenophonF don't use pastebin - too much adware/tracking
09:47 XenophonF use gist or something instead
09:47 XenophonF ok so you have multiple environments set up
09:47 XenophonF so add the 'saltenv' keyword argument to state.sls
09:48 XenophonF e.g., salt crimson.example.com state.sls foo saltenv=lcbc
09:48 darxmurf sweet thanks !
09:49 darxmurf in the end I don't know if this multiple environment is the best choice
09:49 XenophonF if you haven't already, check out the documentation for the state module
09:49 XenophonF lots of helpful debugging tools there
09:49 XenophonF !state.sls
09:49 XenophonF !saltbot
09:49 jaybocc2 joined #salt
09:49 XenophonF i can never remember how to summon the thing
09:50 darxmurf :)
09:51 XenophonF on the topic of multiple environments, i have them set up so that each is a different DTAP phase
09:51 XenophonF so development, testing, staging, production
09:51 XenophonF with corresponding branches in a github repository
09:51 darxmurf yep
09:51 darxmurf in my case it's just because I have machines in 2 different departments
09:52 darxmurf but it's maybe not the best setup :)
09:52 XenophonF that's perfectly logical
09:52 darxmurf ok, so I'll keep this like that
09:52 XenophonF you could have one environment per customer, or per ITIL service, or per OS
09:52 XenophonF whatever makes the most sense to you and matches your workflow
09:52 darxmurf the main problem is that I'm using salt once every 3 months and always forget commands :x
09:53 XenophonF heh
09:53 XenophonF we're implementing something similar to github flow, where the production env will always mirror what's actually deployed
09:53 XenophonF so if we ever have to do DR, we just redeploy the production environment from salt
09:53 XenophonF or if we have to back out changes
09:59 RobertChen117 joined #salt
10:03 darxmurf thanks for you help
10:12 Jimlad joined #salt
10:13 ryau joined #salt
10:15 LondonAppDev joined #salt
10:18 rotbeard joined #salt
10:18 ryau joined #salt
10:23 cyborglone joined #salt
10:26 amcorreia joined #salt
10:33 sjorge_be joined #salt
10:35 sjorge joined #salt
10:35 sjorge joined #salt
10:41 sjorge joined #salt
10:41 sjorge joined #salt
11:02 Vaelatern joined #salt
11:02 subsignal joined #salt
11:08 yomilk joined #salt
11:19 The_Loeki joined #salt
11:19 otter768 joined #salt
11:22 linjan_ joined #salt
11:32 Rumbles is it possible to apply an sls against a machine without updating the top_file ?
11:34 bluenemo joined #salt
11:39 jbax joined #salt
11:46 LostSoul joined #salt
11:46 LostSoul Hi
12:01 jaybocc2 joined #salt
12:01 anmol joined #salt
12:06 bigfoot_ joined #salt
12:07 bigfoot_ Hi guys. How do I use modules (like these examples https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.pw_user.html) in state pillars?
12:13 bigfoot_ for example I want to describe that users full name in passwd file would be correct and I find a command to do it but (user.chfullname) but I don't see how to wirte it into sls file.
12:16 jaybocc2 joined #salt
12:19 ingslovak joined #salt
12:26 dijit Rumbles: yes, salt 'hostname' state.apply <statefileName>
12:30 huds joined #salt
12:44 colegatron joined #salt
12:48 mapu joined #salt
12:49 Ssquidly joined #salt
12:52 Ssquidly joined #salt
12:53 solene joined #salt
12:53 solene Hi, I just discovered that using when using file.line on a file, its perm goes from 644 to 600 is it something normal ?
12:53 dijit check this out for inspiration https://github.com/darkscience/ds-salt/blob/master/states/users/init.sls
12:54 dijit solene: would depend on the umask of the user running salt 'root' in this case most likely.
12:54 dijit but you can specify the modes to use.
12:54 dijit https://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html
12:56 solene dijit: I can't find the syntax to combine file.line with something else in the same sls on the same file :(
12:56 dijit just add it underneath.
12:56 dijit https://github.com/darkscience/ds-salt/blob/master/states/users/sshkeys.sls
12:56 dijit check that out for an example.
12:58 bigfoot_ dijit can you please answer my question too?
12:58 bigfoot_ :)
12:59 Rumbles thanks dijit !
12:59 dijit 1) don't PM me. I'm not affiliated with salt, I just use it.
12:59 bigfoot_ sry
12:59 dijit 2) bigfoot_: dijit | check this out for inspiration https://github.com/darkscience/ds-salt/blob/master/states/users/init.sls
12:59 dijit that's how I manage users.
12:59 solene dijit: I tried to add -mode: 644 to the file.line config but I get an error because this does not exists for file.line  I don't use file.managed but I think I will end with it
12:59 ericof joined #salt
13:00 dijit solene: file.line and file.managed should take the same args.
13:00 dijit since they're part of the same module.
13:00 stolenmoment joined #salt
13:01 solene dijit: seems no => http://pastebin.com/g3vp5fAV
13:01 stolenmoment left #salt
13:01 dijit solene: https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.file.html#salt.modules.file.line
13:01 dijit seems to exist.
13:02 solene the mode isn't the permission here
13:02 dijit oh damn.
13:02 bigfoot_ so salt '*' user.chfullname foo "Foo Bar" becomes something like user: \  - chfullname: Full Name
13:02 dijit you are quite right solene.
13:02 dijit bigfoot_: yeah
13:02 dijit in theory.
13:03 solene this is disturbing to edit files with file.line and get a chmod of 600 on the file... I have been asking for help on #opensmtpd because I didn't think about a perm issue because I never modified it...
13:04 solene I may create an issue on github maybe ?
13:04 fas3r joined #salt
13:04 fas3r hello
13:05 subsignal joined #salt
13:06 fas3r Then defining the servers and rights for other users than root using the client_acl functionnality, is it possible that this specific user run : salt '*' test.ping where * will only match the server define in the acl.conf file for this specific user ?
13:12 DanyC joined #salt
13:13 quasiben joined #salt
13:13 dijit solene: yeah if you edit a file in posix you should be returning it to the proper permissions it was taken with.
13:13 dijit definitely open a ticket.
13:13 dijit what version are you running? 2015.8.3?
13:15 Seanie joined #salt
13:15 fas3r dijit: you talk to me ?
13:15 fas3r sorry :D
13:15 dijit fas3r: I don't have a solution to your problem, sorry.
13:15 Seanie Hey guys just wondering what is the best way to configure worker threads right now we have roughly a thread per minion(20). We seem to be getting time outs alot.
13:16 CeBe joined #salt
13:16 solene dijit: no 2015.8.1
13:17 akhter joined #salt
13:18 cliluw joined #salt
13:18 yomilk joined #salt
13:18 solene oh I have some 2015.8.3 I will check with them
13:19 solene I have the same issue with 2015.8.3
13:19 morissette joined #salt
13:19 dijit oki, make a ticket.
13:19 dijit sounds like they're doing something explicitly non-posix.
13:21 akhter joined #salt
13:21 TyrfingMjolnir joined #salt
13:22 DanyC hi all, any Jinja experts arounds who can help me streamline the vpc_id grain output http://hastebin.com/lagusadoce.vhdl
13:23 DanyC i'd like a simple way to not require hardcoding the mac value nor having to specify all recursive keys, any idea?
13:24 BlackBishop joined #salt
13:25 solene dijit: is it the salt_master version or salt_minion version which matter ?
13:25 BlackBishop Once I've updated a template file in a module .. what can I do to make it reload that module so it applies the new template file ?
13:26 BlackBishop do I have to restart the whole master ? is there a cache I can clear or something ?
13:27 dijit solene: minion
13:27 solene dijit: fine then, it's up-to-date
13:27 dijit ^_^
13:28 BlackBishop :-s
13:28 dijit if it's jinja that fails, it's the master, if it's the state that fails, it's the minion.
13:28 dijit BlackBishop: it should be checking the full hierarchy everytime.
13:28 BlackBishop doesn't seem so .. I've updated the template file and added some comments
13:29 BlackBishop did a salt 'host' state.sls module
13:29 BlackBishop and it still applies the old template .. no comments that I've added
13:30 BlackBishop and if I do state.sls module twice .. it does the same changes all over again .. it doesn't seem to see that it was applied a sec ago
13:31 dijit there's no caching in salt.
13:31 dijit so idk why it would do that.
13:33 BlackBishop I see some caches though ...
13:33 BlackBishop https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.saltutil.html
13:33 BlackBishop and refresh stuf
13:36 BlackBishop [INFO    ] Fetching file from saltenv 'base', ** skipped ** latest already in cache 'salt://module/templates/file.conf'
13:36 BlackBishop so there is some kind of cache it seems !
13:36 BlackBishop ( this is the result when I do a salt-call state.sls module from the minion itself )
13:39 dijit oh.
13:39 dijit salt-call behaves differently.
13:39 BlackBishop restarted the minion, same thing.
13:39 dijit I always assume central master.. sorry.
13:39 BlackBishop well, I assume it's the same thing that happens .. for some reason
13:40 dijit https://github.com/saltstack/salt/issues/24050
13:40 saltstackbot [#24050]title: state.highstate does not refresh minion cache pillar data | ***NOTE:*** The original test steps do not work anymore because "salt-call" always gives you the latest data. But this still is broken when you run the pillar calls (pillar.item and pillar.get) from the master....
13:40 dijit could be related.
13:41 dijit it's just failing to invalidate caches appropriately.
13:41 BlackBishop heh :) so I should restart the master, right ?
13:43 BlackBishop but as the latest comment says, this is an awful bug !
13:44 BlackBishop restarting the master didn't help either ! :|
13:48 alvinstarr joined #salt
13:50 subsignal joined #salt
13:51 BlackBishop and fixing by hand doesn't help since when I apply again it replaces my changes.
13:51 dyasny joined #salt
13:51 BlackBishop where's the cache dir ? I'm goin' ninja on it.
13:52 subsignal joined #salt
13:53 BlackBishop ... /var/cache/salt/master ..
13:53 BlackBishop is this safe to delete ? :D
13:54 racooper joined #salt
13:56 jaybocc2 joined #salt
14:02 job Who manages the "INSTALLATION FROM THE COMMUNITY REPOSITORY
14:02 job https://docs.saltstack.com/en/latest/topics/installation/debian.html
14:02 DammitJim joined #salt
14:03 job because the debian packages on debian.saltstack.com are out of date
14:11 bhosmer joined #salt
14:17 LotR job: why not use the packages that are in debian itself? unstable has 2015.8.3, which the /topic says is latest
14:17 mpanetta joined #salt
14:17 job because i don't want to run debian unstable
14:19 LotR so configure the sid repo with a lower priority, so you have to explicitly get stuff from there before it gets used
14:19 cpowell joined #salt
14:19 job i feel it is simpler if modern packages are available on either repo.saltstack.com or debian.saltstack.com
14:21 LotR me, I don't trust third parties to get packaging right :)
14:21 job that much is clear
14:22 DanyC joined #salt
14:24 darvon joined #salt
14:24 eichiro joined #salt
14:25 darvon joined #salt
14:25 eichiro joined #salt
14:25 XenophonF job: http://repo.saltstack.com/#debian
14:25 XenophonF i'm guessing those instructions are outdated
14:26 XenophonF everything's been moving to repo.saltstack.com
14:26 XenophonF their PPA and COPRs haven't been kept up to date
14:27 LondonAppDev joined #salt
14:28 job 'everything moved' is not true
14:28 viq joined #salt
14:29 job repo.saltstack.com does not contain packages for armhf CPU architecture
14:39 kbyrne joined #salt
14:46 perfectsine joined #salt
14:47 winsalt joined #salt
14:49 scoates joined #salt
14:51 eliasp yeah, coverage/support in repo.saltstack.com is rather thin… no packages for recent Fedora, Ubuntu only 12.04 and 14.04, nothing else… missing pyopenssl deps on Ubuntu etc.
14:51 job yeah lets not forget it was broken for 2 or 3 weeks
14:51 eliasp and the repos (at least those for APT-based distributions) get quite often out of sync
14:51 eliasp exactly
14:53 LotR all the more reason to stick with packages from debian proper :)
14:53 kbyrne joined #salt
14:57 teryx510 joined #salt
15:00 ALLmightySPIFF joined #salt
15:00 kbyrne joined #salt
15:00 edrocks joined #salt
15:02 lumtnman joined #salt
15:02 zmalone joined #salt
15:02 kbyrne joined #salt
15:04 DammitJim joined #salt
15:11 _JZ_ joined #salt
15:14 bhosmer joined #salt
15:16 malinoff joined #salt
15:17 malinoff joined #salt
15:18 ALLmightySPIFF joined #salt
15:19 BlackBishop left #salt
15:20 Brew joined #salt
15:20 otter768 joined #salt
15:21 thalleralexander joined #salt
15:21 fredvd joined #salt
15:23 TomJepp_ joined #salt
15:24 eightyeight joined #salt
15:25 hasues joined #salt
15:26 hasues left #salt
15:26 tehsu running a for loop, is it possible to read a list of numbers?
15:26 tehsu from a file that is
15:28 tehsu {% for usr in ['moe','larry','curly'] %} pretty much this but the users are in a file listed on a new line
15:29 andrew_v joined #salt
15:31 thalleralexander joined #salt
15:32 yetAnotherZero joined #salt
15:45 chadhs joined #salt
15:45 winsalt is the file a pillar file?
15:45 tehsu yeah
15:46 tehsu it can be
15:47 winsalt it should be, then its easy with salt['pillar.get']('users') where users is the key
15:50 spuder joined #salt
15:51 bhosmer joined #salt
15:52 dfinn joined #salt
15:54 numkem I keep having memory issues with the minion taking a lot of ram. How can make it dump it so I could figure out whats happening?
15:54 numkem Or at least send it to someone who would know more
15:58 tracphil joined #salt
16:00 CheKoLyN joined #salt
16:01 gcorey joined #salt
16:01 ALLmightySPIFF joined #salt
16:02 tracphil when using map.jinja as so: {% from "apache/map.jinja" import apache with context %} does word after import need to be name spaced across SaltStack or can I make it the same word such as map?
16:02 ALLmightySPIFF joined #salt
16:03 dfinn joined #salt
16:04 sdm24 joined #salt
16:07 chadhs joined #salt
16:11 tehsu winsalt: the pillar file would look like this? https://gist.github.com/tehsu/8e6911b91aff61b95f1a
16:12 spuder joined #salt
16:15 beardedeagle joined #salt
16:15 winsalt I made a comment on  your gist, you want a list, so that you can iterate over it
16:16 akhter_1 joined #salt
16:17 ingslovak teshu: that is a bit redundant, the "users" key can contain a list of users, like this: http://pastebin.com/HuyxGbGe . Since list is an iterable object you can also do {% for user in pillar['users'] %}
16:21 tehsu ah thank you, thats what I was trying to understand
16:25 PsionTheory joined #salt
16:26 grumm_servire joined #salt
16:31 calvinh joined #salt
16:33 calvinh joined #salt
16:33 amcorreia joined #salt
16:36 vlad___ joined #salt
16:37 vlad___ hello
16:38 vlad___ I'm receiving the following error when starting a proxy minion for a junos device. I don't know if someone can shed some light?
16:38 vlad___ [ERROR   ] No proxy key found in pillar for id srx.
16:38 breshead joined #salt
16:40 Tanta joined #salt
16:41 dfinn joined #salt
16:44 teryx510 joined #salt
16:49 jrgochan joined #salt
16:54 Bryson joined #salt
16:56 ksj joined #salt
16:56 TyrfingMjolnir joined #salt
16:57 jrgochan Hello all! Anyone know when Boron is planned for release?
16:59 ksj hi, I'm trying to test salt-virt on arch linux, following the "salt as cloud controller" official doc. I get the following failure, which causes other failures because it's a requisite: http://dpaste.com/10GDWDK  any ideas?
16:59 ksj interface exists and is up
17:00 ksj I have set up bridges and many virtual machines with qemu (though they're not running now), so I doubt it's missing packages or anything
17:02 ekristen joined #salt
17:02 impi joined #salt
17:03 spiette joined #salt
17:04 dfinn joined #salt
17:04 onlyanegg joined #salt
17:07 murrdoc joined #salt
17:08 gcorey joined #salt
17:08 keimlink joined #salt
17:16 iggy jrgochan: nobody can say for sure, hopefully by Feb
17:19 DammitJim joined #salt
17:20 upb hi, what can i use in windows in "PS C:\salt> ./salt-call cp.get_template salt://win/repo-ng/lprkiosk/init.sls /dev/stdout" instead of /dev/stdout ?
17:20 DammitJim is there a way to use Salt with Bamboo?
17:21 buhman upb: literally anything
17:21 otter768 joined #salt
17:22 upb buhman: yeah but i mean without using a temp file
17:22 jfindlay jrgochan: should be in the next few months.  I would like to be more specific, but we've not committed to anything specific yet.  It should be about 6 months since 2015.8 (Beryllium)
17:23 buhman windows doesn't represent stdout as a file, so you can't do that without writing your own module
17:23 upb ah, thx
17:24 upb it does actually but maybe python doesn't have a filename for it
17:29 jfindlay upb: I did some work with one of our windows engineers on this.  I still need to bug him about possibly sending some code upstream into python :-)
17:29 patchedmonkey joined #salt
17:29 jfindlay my impression was that the platform support was incomplete
17:30 upb ic
17:30 patchedmonkey joined #salt
17:31 evle1 joined #salt
17:31 patchedmonkey joined #salt
17:32 jrgochan iggy and jfindlay, thanks for the update. I was itching to use the LDAP management state. Might just set up a development version and start playing with it
17:35 writtenoff joined #salt
17:41 iggy jrgochan: you might be able to get away with copying the module from devel into _states/_modules
17:43 stupidnic Alright, does anybody know if the network state supports pre-up/post-up directives? I looked at the code and they are referenced, but I can't seem to get them to actually trigger.
17:45 wangofett Does anyone know if it's possible to install an RPM without knowing the package name? I'm not really finding anything, and when I tried of course it failed :P
17:46 jrgochan iggy: would that just go in /srv/salt/_states/_modules ?
17:46 zmalone wangofett: You need to know the package name, but it's possible to look up a package name (using yum whatprovides etc.) if you don't know it
17:47 wangofett well, I've got an RPM (https://www.softwarecollections.org/en/scls/rhscl/rh-python34/epel-7-x86_64/download/rhscl-rh-python34-epel-7-x86_64.noarch.rpm)
17:47 wangofett but when I ran 'rpm -qpil <rpm>' on it it gave me 'rh-python34'
17:48 wangofett but that didn't work
17:49 jrgochan perhaps an rpm -Uvh ?
17:51 mfournier joined #salt
17:52 wangofett I don't want to install it, I want to know what package it provides
17:53 felskrone joined #salt
17:56 jrgochan hrm. it adds this repo. so all of the packages in there https://www.softwarecollections.org/repos/rhscl/rh-python34/epel-7-x86_64/
17:58 whytewolf wangofett: rhscl-rh-python34-epel-7-x86_64 is the package name.
17:58 whytewolf yum is a pain when it comes to package name. but typically it is the name of the rpm to the first dot
17:59 HardWall joined #salt
18:00 wangofett Hm. "The following package(s) were not found, and no possible matches were found in the package db: rhscl-rh-python34-epel-7-x86_64"
18:00 wangofett though if all it's doing is adding a repo, I don't really need the package
18:01 wangofett I just want to add the repo :P
18:01 whytewolf lol. then just add the repo :P
18:01 tehsu winsalt: is this for loop correct, I assume it's correct but it doesn't run through, {% for num in salt['pillar.get']('stop') %}
18:02 chadhs joined #salt
18:02 winsalt looks alright
18:03 chadhs hi fellow salters, where do you like to store your pillar data?
18:03 tehsu and then I have stop: - 123456
18:03 tehsu two lins
18:03 tehsu lines*
18:03 chadhs with your salt configs in a git repo, or store them in s3, or some other option for secure pillar data
18:04 winsalt like the users, you need to make it a list.  '123456' will be read as a string
18:04 zmalone chadhs: I'm cleaning up problems that resulted from sticking salt configs in the same repo as my formulas right now
18:05 zmalone in particular, file_roots is accessible to all minions, so because my pillar_root was in file_roots (which was a git repo), my pillar was accessible to all clients
18:05 tehsu like you showed, https://gist.github.com/tehsu/8e6911b91aff61b95f1a
18:05 chadhs we're deciding between pillar data in our private git repo OR is it really secret enough to place elsewhere
18:05 zmalone regardless of targetting
18:06 tehsu and then I have the for loop which doesn't run and my output from salt is empty
18:06 whytewolf chadhs: typically i have several repos. one private one for pillar. with some extra secret stuff in ext_pillar sources. one for configs. one for states. then seperate repos for my custom mods.
18:07 wangofett whytewolf: yeah, I didn't know that was all rpm did >.<
18:07 winsalt tehsu, do you have {% endfor %} at the bottom
18:07 tehsu yes
18:08 whytewolf wangofett: probley just created a file in yum.repo.d and imported an rpm key :P
18:08 chadhs whytewolf: im trying to wrap my head around how can i ensure servers can access pillar data they need and Ops can but Dev users on servers cannot
18:08 tehsu I can post my script if you need
18:09 whytewolf chadhs: if dev uses have root access to the minions they have access to the pillar data of that minion.
18:09 chadhs no root access
18:09 chadhs special dev accounts
18:09 chadhs with sudo for specific commands
18:09 whytewolf do they have access to salt?
18:10 xmj 750 root:wheel with ops being wheel but dev not?
18:10 chadhs whytewolf: if they don't on minions and they don't have access to the repo with secure data…
18:10 chadhs then im good
18:11 chadhs so… whytewolf: could i just have a special repo for secure pillar data?  would that be a good strategy
18:11 chadhs im also reading the best practices doc
18:11 whytewolf chadhs: well i meant the salt master. if they have any kind of access to run the salt command vs salt-call then they can't dig info about pillars from anywhere
18:11 chadhs im running masterless
18:11 whytewolf ahhh good.
18:12 chadhs so there's a dedicated salt users with no privs
18:12 chadhs that pulls the repo
18:12 chadhs and then a root priv cron call to call highstate at a time interval
18:12 tehsu i updated the gist with the script itself, winsalt
18:13 wangofett whytewolf: I'm kind of suspcious about importing an rpmkey - it has gpgcheck=0
18:13 whytewolf wangofett: really? sigh ....
18:13 bhosmer_ joined #salt
18:14 Brew Hello!  quick question, when using a beacon to fire a salt event, can i customize the tag?  or is it possible from within a reactor to read event details?  eg "tag": "salt/beacon/<minion_id>/service/" where the name of the service is contained in the json info
18:15 cyborg-one joined #salt
18:15 Brew i need the name of that service in order to fire off the correct sls
18:16 egor joined #salt
18:17 winsalt tehsu. those are dashes right, i know git is changing them to dots
18:17 Brew n/m rtfm'ing solved this one
18:17 Brew https://docs.saltstack.com/en/develop/topics/reactor/index.html
18:17 tehsu yes
18:17 Brew The data variable is the event's data dict.
18:17 Brew thanks for playing, have a great day :)
18:17 whytewolf lol Brew I was just about to say that
18:17 chadhs whytewolf: would this be sane? have non secure pillar stuff accessible and having a secure pillar data locked down (referring to repositories).  can you pull from two pillar sources
18:18 winsalt have you set up your top file yet, tehsu.  You will need to match your minion to the name of the pillar file
18:18 whytewolf chadhs: you can pull from as many as you need.
18:18 tehsu ah god damnit, forgot, sorry
18:18 chadhs sweet!
18:18 whytewolf chadhs:  just watch for how they get merged. sometimes it can be tricky
18:19 tehsu testing now
18:19 chadhs whytewolf: essentially i'd only put secure stuff there like tokens, passwords, etc
18:19 whytewolf chadhs: yeap.
18:19 chadhs sweet
18:20 chadhs i think i trust that more than s3
18:20 chadhs and then there's history etc
18:20 whytewolf history is good
18:24 anotherZero joined #salt
18:26 tehsu winsalt same thing, no output when I run it with the for loop
18:29 winsalt i dunno, it looks fine.  you can try running salt with "-l debug" to see whats going on
18:31 tehsu yeah tried that as well, nothing really outputted from there, I'll see, maybe I can find something wrong on my side
18:31 AllSySt3msG0 joined #salt
18:32 ryau joined #salt
18:38 baweaver joined #salt
18:38 lemur joined #salt
18:40 lemur joined #salt
18:40 brianfeister joined #salt
18:41 oida Hi, I'm having an issue using salt-cloud and lxc with download templates. My lxc_profile looks like "lxc_profile: \n backing: btrfs \n template: /usr/share/lxc/templates/lxc-download -- -d debian -r jessie -a amd64"
18:43 oida Now salt seems to making it "lxc create ... -t /usr/share/lxc/templates/lxc-download -- -d debian -r jessie -a amd64 -B btrfs" so lxc complains "Container could not be created with cmd 'lxc-create -n m-fileshare -f /tmp/tmp_1grO0 -t /usr/share/lxc/templates/lxc-downlo    ad -- -d debian -r jessie -a amd64 -B btrfs': getopt: invalid option -- 'B'". Any ideas?
18:43 colegatron joined #salt
18:44 ryau_ joined #salt
18:47 aea joined #salt
18:51 subsigna_ joined #salt
18:54 baweaver joined #salt
19:01 cberndt joined #salt
19:04 aron_kexp joined #salt
19:05 ajw0100 joined #salt
19:05 stomith joined #salt
19:09 iggy jrgochan: /srv/salt/_modules (for the execution module which the state most likely needs) and /srv/salt/_states for the state module
19:10 ryau_ Hi... anyone seen this error before while installing Salt:
19:10 ryau_ Error: Package: zeromq3-3.2.4-1.el6.x86_64 (@epel)
19:10 ryau_ Requires: libpgm-5.1.so.0()(64bit)
19:10 ryau_ Removing: openpgm-5.1.118-3.el6.x86_64 (@epel)
19:10 ryau_ libpgm-5.1.so.0()(64bit)
19:10 ryau_ Updated By: openpgm-5.2.122-2.el6.x86_64 (repo-saltstack-el6)
19:10 ryau_ Not found
19:10 ryau_ Thank you in advance...
19:11 subsignal joined #salt
19:12 wangofett has anyone ever used salt to install python3 packages a la pip.installed?
19:12 iggy wangofett: yes
19:13 wangofett iggy: do you have an example? Mine is saying that it needs a pip module but it doesn't exist :P
19:13 iggy wangofett: that means you need to install python-pip first
19:13 zmalone ryau_: yes, typically while installing salt
19:14 subsigna_ joined #salt
19:14 zmalone sorry, upgrading salt on *EL platforms
19:14 wangofett iggy: even if I'm passing bin_env?
19:14 zmalone it's a packaging mistake
19:14 ryau_ what should I do?
19:14 iggy wangofett: yes, you still need the pip module (for certain things) in the python path for the same python interpreter that salt uses
19:15 jrgochan Thanks iggy!
19:15 wangofett aahhh. Sneaky.
19:15 ryau_ zmalone: what should I do, any suggestions?
19:15 zmalone ryau_: I think uninstalling any prior packages of salt, zeromq, etc. will allow you to install from the repo.saltstack.com repo
19:16 zmalone (I'm not on EL, but it keeps coming up in this channel)
19:16 wangofett iggy: looks like that did it, thanks!
19:22 otter768 joined #salt
19:32 perfectsine joined #salt
19:37 jaybocc2 joined #salt
19:37 perfectsine joined #salt
19:37 grumm_servire joined #salt
19:38 wt joined #salt
19:40 ageorgop joined #salt
19:42 forrest joined #salt
19:44 huds joined #salt
20:00 perfectsine joined #salt
20:01 ryau joined #salt
20:03 * wangofett is having problems pip.installing lxml via salt: http://stackoverflow.com/q/34620267/344286
20:04 forrest wangofett, http://stackoverflow.com/questions/15759150/src-lxml-etree-defs-h931-fatal-error-libxml-xmlversion-h-no-such-file-or-di
20:04 forrest you need the xml libs installed on the system
20:05 ryau__ joined #salt
20:05 wangofett forrest: yeah... they are :(
20:06 ageorgop joined #salt
20:06 justanotheruser joined #salt
20:06 forrest wangofett, :(
20:06 wangofett the build can't find them. I'm 90% sure that's due to it being the "software collection" version of Python
20:06 ryau joined #salt
20:07 forrest wangofett, What about that answer the guy has in response of the comments about using a symlink?
20:07 wangofett but in theory there's *some* environment variable that I should be able to use to say, "Hey, use this thing too!"
20:07 forrest for libxml2 versus libxml
20:07 wangofett hmmm... lemme try that
20:08 forrest wangofett, Cool, also might be worth trying to use: https://docs.saltstack.com/en/latest/ref/states/all/salt.states.environ.html
20:09 keimlink joined #salt
20:11 blckbit10 joined #salt
20:12 ryau_ joined #salt
20:15 ryau__ joined #salt
20:17 wangofett forrest: awwww, yeah - the pip log shows an -I/opt/rh/rh-python34/root/usr/include
20:18 wangofett adding `file.symlink` from that `include/libxml` to `/usr/include/libxml2/libxml/` took care of it \o/
20:18 forrest Ahh
20:18 forrest That's good
20:18 otter768 joined #salt
20:18 forrest I should add that as the answer to your post, lol
20:19 forrest Actually I wll, hang on
20:21 bhosmer_ joined #salt
20:22 wangofett +1
20:25 tehsu winsalt: thanks for your help, I did figure out the issue which was my pillar was not in my specific environment it was in the base only
20:26 tehsu Is there a way to have a schedule job run against certain servers at certain times instead of creating 15 different schedules per server
20:27 tehsu just doing a simple restart for each server and its the same command that needs to be ran at different times
20:29 s_kunk joined #salt
20:31 cberndt joined #salt
20:32 keimlink joined #salt
20:36 GreatSnoopy joined #salt
20:41 jhauser joined #salt
20:50 ryau joined #salt
20:52 enigma99a joined #salt
20:53 enigma99a hey all, I wasn't having any issues with pygit and 2015.8.1 but 8.3 is.. Which version of pygit should I be using now?
20:53 enigma99a local/lib/python2.7/dist-packages/cffi/model.py:526: UserWarning: 'git_merge_tree_flag_t' has no values explicitly defined; next version will refuse to guess which integer type it is meant to be (unsigned/signed, int/long)
20:54 ryau_ joined #salt
20:55 felskrone joined #salt
20:55 ryau joined #salt
20:59 AndreasLutro enigma99a: let me look it up in my setup...
20:59 jfindlay enigma99a: https://github.com/saltstack/salt/issues/29049#issuecomment-164862851
20:59 saltstackbot [#29049]title: UserWarning: 'git_merge_tree_flag_t' has no values explicitly defined; next version will refuse to guess which integer type it is meant to be | Using gitpython my salt master is logging the following errors after upgrading my master to 2015.8.2...
20:59 enigma99a I see there is a devel pygit available
21:00 AndreasLutro I'm using libgit2 0.23.4, pygit2 0.23.1
21:00 jfindlay in terms of redhat packages, devel usually means additional package components, like for development use rather than just use
21:01 AndreasLutro enigma99a: though it is only a warning, you could ignore it
21:01 ryau_ joined #salt
21:02 enigma99a hmm
21:02 ryau__ joined #salt
21:03 perfectsine joined #salt
21:03 enigma99a thanks all
21:05 ryau joined #salt
21:07 perfectsine joined #salt
21:08 slav0nic joined #salt
21:11 snarfy joined #salt
21:13 snarfy hay I have a question
21:13 murrdoc HAI
21:14 terratoma is that a Blacked Eyed Peas song
21:14 snarfy every time I do a state run, salt-call says "executing npm --version in directory...."
21:14 snarfy but the state run will have nothing to do with npm
21:14 snarfy or node. at all.
21:14 snarfy what the dilly, yo?
21:20 bonzibuddy hey folks - i have a file that i want present on a minion, but during system operation it gets appended to
21:20 bonzibuddy so i only want it to move over the initial file when it doesnt exist - file.managed will always change the contents
21:20 bonzibuddy at least how i have it configured
21:24 frew https://docs.saltstack.com/en/latest/ref/states/all/salt.states.alternatives.html doesn't have anything to do with debian/ubuntu's "update-alternatives" does it?
21:24 frew I don't see how it would
21:25 eliasp bonzibuddy: use "replace: false" for "file.managed"
21:25 bonzibuddy eliasp: :D
21:26 bonzibuddy knew there had to be an option for that
21:27 RandyT Is there a way to get jinja to provide current date in state file?
21:28 sdm24 RandyT: there's probably a better way, but I use {{ salt['cmd.run']('date') }}
21:28 eliasp RandyT: https://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.jinja.html#filters → strftime
21:28 sdm24 for date only, {{ salt['cmd.run']('date +%D') }}
21:28 RandyT excellent. thanks to you both
21:29 sdm24 no problem
21:30 snarfy frew: of course it does
21:31 frew snarfy: yeah I just don't know the alternatives system from both perspectives very well
21:32 patchedmonkey joined #salt
21:33 snarfy frew: link: is the symlink you want, path is the actual binary
21:33 patchedmonkey joined #salt
21:33 frew yeah, I just did a stupid thing and dind't scroll down
21:33 frew for some reason I thoguht install ~~ set
21:33 snarfy on debian/ubuntu this managed links in /etc/alternatives
21:33 frew thanks
21:37 jaybocc2 joined #salt
21:42 jfindlay snarfy: what version of salt?
21:43 jfindlay also, are you running salt-call with -l debug?
21:43 jfindlay personally, I think those loader log messages should be trace level at best
21:48 DammitJim joined #salt
21:54 anotherZero joined #salt
21:56 jvv joined #salt
22:00 murrdoc if only
22:00 murrdoc u knew someone who could fix it
22:00 murrdoc jfindlay:
22:00 murrdoc if
22:01 murrdoc only
22:01 jfindlay I might
22:01 ryau_ joined #salt
22:01 jfindlay if I feel like it
22:02 stanchan joined #salt
22:02 snarfy^ joined #salt
22:05 subsignal joined #salt
22:11 ryau joined #salt
22:12 ryau__ joined #salt
22:12 dancat joined #salt
22:15 ryau_ joined #salt
22:18 dancat I am having an issue provisioning an Ubuntu (14.04 x64) instance on DigitalOcean. I can provision other instances (Debian, CentOS, Fedora). The error message is at http://pastie.org/10672208 I am not sure if it is a problem with salt-cloud's usage of salt-bootstrap.sh, the DigitalOcean image, Ubuntu 14.04 packages, or a combo of a few of those. What's the best way to approach reporting the issue?
22:18 dancat My bad 15.10 x64
22:20 otter768 joined #salt
22:21 dancat here's a more comprehensive error report http://pastie.org/10672216
22:22 ryau__ joined #salt
22:23 ryau joined #salt
22:25 ryau_ joined #salt
22:33 wangofett Hm.... Is it possible to install a single state? I thought I saw someone say it was the other day
22:33 jaybocc2 joined #salt
22:33 wangofett I recall seeing something like `salt 'minion' your_state.sls state_id`
22:34 whytewolf wangofett: single state. or single state file? [both are possable]
22:34 wangofett just one state from a single file. I know state.apply (I've been building up my state files and using that to test)
22:35 whytewolf wangofett: https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.sls_id
22:35 zmalone joined #salt
22:35 retake left #salt
22:36 murrdoc left #salt
22:36 kevinqui1nyo joined #salt
22:36 murrdoc joined #salt
22:36 wangofett whytewolf: do you have to have an sls file, or does it work if you have <thing>/init.sls?
22:37 wangofett i.e. I tried state.sls_id <mything> <mystate> and it failed
22:37 wangofett where I have <mything>/init.sls
22:37 kevinqui1nyo what's the best practice for including a salt module from an external git repo in my salt config git repo?  Submodule?
22:38 whytewolf wangofett: odd it should work like you state. personally never used it though so wouldn't know
22:38 wangofett it says "no matching sls for 'mything' in env 'base'"
22:39 wangofett oh wait
22:39 wangofett I might have this backwards -_-
22:39 wangofett oh yeah, looks like it
22:39 wangofett So, it's <state> <statefile> :P
22:40 whytewolf lol. opps year. the order is id_, mods on the function call.
22:41 wangofett that seems backwards to me, but *shrug*
22:41 whytewolf kevinqui1nyo: pull it into a a seperate directory and add an extra directory path to you file_roots
22:41 whytewolf wangofett: does to me too. but thats how it was written.
22:43 wangofett ah. I see why you might want that.... I guess(?)
22:43 wangofett It says: "Call a single id from the named modules(s)"
22:43 wangofett so... you might have "state.sls_id someid module_one,module_two"
22:44 wangofett but wouldn't that barf with a duplicate ID? It seems like it would...
22:44 whytewolf everything in salt barfs on duplicate id's
22:44 zenlot1 joined #salt
22:44 kevinqui1nyo whytewolf: perfect, thank you
22:48 jaybocc2 joined #salt
22:49 kevinqui1nyo whytewolf: i wonder if i should use gitfs for that external module or if i should just use file_roots and git clone it myself to the path?
22:49 kevinqui1nyo just looking for an opinion on that really
22:49 whytewolf kevinqui1nyo: up to you reallt. either way FORK IT
22:49 kevinqui1nyo ahh
22:50 kevinqui1nyo whytewolf: probably a good idea so i can pin it to a working state or modify as needed
22:52 spuder_ joined #salt
22:53 jhauser joined #salt
22:57 HardWall joined #salt
22:57 baweaver joined #salt
23:00 lemur joined #salt
23:02 TyrfingMjolnir joined #salt
23:02 kusuriya left #salt
23:05 otter768 joined #salt
23:06 dancat using salt-cloud, I am having difficulty communicating with the minions from the master. I am able to ssh into each of the minions, but I cannot figure out why `salt '*' test.ping` is unsuccessful. I do have a firewall up on the master but I allow for 4505 and 4506 to pass through as far as I can see.
23:06 ryau joined #salt
23:08 murrdoc jfindlay:  u around
23:08 murrdoc i cant merge in https://github.com/saltstack-formulas/msdtc-formula/pull/1 or https://github.com/saltstack-formulas/rdp-formula/pull/1
23:08 saltstackbot [#1]title: Initial commit |
23:08 ryau__ joined #salt
23:08 murrdoc forrest:  ^^
23:08 forrest Hang on, I forgot the perms don't auto set
23:08 forrest fucking github
23:08 murrdoc noob
23:09 forrest Try now
23:10 enigma99a hmm  Exception 'Failed to authenticate SSH session: Callback returned error' caught while fetching gitfs remote 'git@myremote:salt.git'
23:10 enigma99a Is 0.23.1 still broke?
23:11 enigma99a I think they changed the auth schemes in 0.23
23:12 dancat nevermind
23:12 dfinn joined #salt
23:15 enigma99a I have data in /var/cache/salt/master/gitfs ... anyone know why this error might be generating?
23:16 woodtablet joined #salt
23:16 woodtablet who here is going to scale ?
23:19 AndreasLutro I go on the scale every morning before I shower
23:20 whytewolf I glance at the scale and then cry myself to sleep
23:22 spuder joined #salt
23:23 woodtablet lol
23:23 woodtablet i dont look at the scale that often, its just scares me
23:23 woodtablet but there is also the southern california linux expo (SCALE) coming up
23:26 dancat I case anyone sees the errors described. It was a result of me not having the correct ip address in the cloud file on the master.
23:27 woodtablet whytewolf: if you go, i ll buy you a beer
23:27 woodtablet lol
23:27 whytewolf woodtablet: humm. it isn't that far from me. only about a 5-6 hour drive. but it is unforchantly just before a trip to AZ I have to take
23:28 woodtablet whytewolf: no worries. you have helped me before, thats why i was going to offer the beer ^_~
23:28 whytewolf lol. no problem. there may be other chances for that beer ;)
23:31 murrdoc fuck scale
23:31 murrdoc useless fucking conference
23:31 * murrdoc strokes grey beard
23:32 murrdoc (also totally kiding, scale is awesome gonna miss it the first time ever)
23:32 frew I know this is a jinja question but I figure someone here will know the answer
23:32 frew how can I set the value of a key in a hash?
23:32 frew `do hash['key'] = value` errors because of the =
23:32 whytewolf set
23:32 AndreasLutro do hash.update({key: value})
23:33 whytewolf or yeah .update
23:39 frew oooh
23:39 frew set isn't a fucntion, as far as I can tell; I dud update wrong (hash.update(key, value))
23:39 frew thanks
23:41 frew ok NOW to the quesiton I wanted to ask but couldn't get the code to run
23:41 frew it seems like it would make sense for me to be able to do something like
23:41 frew {{variable|yaml}}
23:41 whytewolf frew: http://jinja.pocoo.org/docs/dev/templates/#assignments
23:41 TyrfingMjolnir joined #salt
23:41 frew so I can avoid the whole "if it's zero, print [], otherwise format as a list" silliness
23:42 frew whytewolf: yeah where does that show mutating a hash?
23:46 frew I mean, I'm generating yaml in a templating language; I feel like I'm in the 90s or something
23:46 rihannon joined #salt
23:46 frew there *has* to be a better way
23:46 lompik joined #salt
23:46 whytewolf see there is the differents. mutating isn't the word you used :P but this does work {% set data = {'key': 'value'} %}
23:46 frew well fwiw I didn't say define a hash, I said set the value of a key
23:46 frew but w/e
23:47 * whytewolf shrugs .update
23:48 whytewolf but yeah, I would find a simpler way
23:48 frew so like, I am doing [{{pkgs|join(',')}}]
23:48 onlyanegg joined #salt
23:48 frew so if pkgs is empty it'll still work
23:48 frew but I'd much rather just do {{pkgs|yaml}}, but that's nto a thing
23:49 frew I just don't understand the template driven nature of salt (and let me be clear: I am new to both salt and CM in general, so I certainly don't understand the decisions that went into it)
23:51 * frew tries |yaml_encode
23:51 frew nope.
23:52 iggy frew: https://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.jinja.html#salt.utils.jinja.SerializerExtension
23:53 frew huh
23:53 frew so maybe that is working and I'm just getting an unrelated error
23:54 frew I did {{pkgs|yaml}} (and pkgs is ['fontcustom'])
23:54 frew and I get `State 'fontcustom' in SLS u'ruby' is not formed as a list`
23:54 iggy try a gist
23:55 frew yeah on it; sorry
23:55 iggy much easier for us to see what's going on
23:55 frew please be gentle: https://gist.github.com/frioux/ddc26c88199e812a8556
23:57 frew this is in a pillar called ruby.sls

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