Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-05-15

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

All times shown according to UTC.

Time Nick Message
00:00 forrest pedromaltez: np
00:00 rager I've specified user, group, and ruby version
00:00 forrest ugh the docs being down is a real pain in the as.
00:00 forrest rager: I can't remember the syntax for gem.installed, what if you try to use the gem command as root, does it fail?
00:01 pedromaltez forrest: use google chrome and prepend with `cache:`?
00:01 rager probably - ruby is installed as vagrant
00:01 forrest pedromaltez: I don't have the docs cached unfortunately, thanks for the suggestion though!
00:01 rager oh
00:01 pedromaltez no no
00:01 rager system has 1.9.3 installed
00:02 pedromaltez google caches the site
00:02 rager so root can install gems and whatnot
00:02 forrest rager: Hmm, well I am confused then how it's not installing
00:02 forrest pedromaltez: Oh I see what you mean
00:02 rager me, too
00:02 rager I'm confused as to what is actually happening in the error: somehow, it fails when calling "gem list bundler" with some cryptic missing file error
00:03 pedromaltez chrome will resolve the special prepended url to whatever google has, if it has it cached
00:04 forrest rager: is that the only one that fails?
00:04 forrest pedromaltez: Yeah that worked, had a URL issue that was causing it to fail before, good suggestion
00:04 rager it's the first and only gem that's installed with a state
00:04 rager since the rest come from a gemfile
00:04 rager well, with the gem.installed state, that is
00:04 forrest rager: I just mean are there any other IDs that fail
00:04 rager no
00:04 rager just that
00:05 forrest rager: also I don't see a group directive
00:05 rager might've removed it
00:05 rager at some point, I got a complaint about group not being an option and removed it
00:05 forrest can you try to drop it and see what happens?
00:05 rager at some point, I tried adding it back
00:06 rager yeah
00:06 rager no dice
00:06 rager it stands that it *should* succeed since the "gem" command is resolvable in the PATH var
00:06 VR-Jack hmmm. how bad is state concurrency? ie, I'd like a highstate sls to signal reactor then wait for reactor to signal it back, then in reactor issue a state.sls to the target
00:08 forrest rager: Yeah, what version of salt are you running? I assume masterless since this is vagrant?
00:08 rager masterless, using stable from ubuntu repos
00:08 rager I think it's 2014 something
00:08 forrest okayu
00:08 rager decided to see if the problem persisted on a completely different OS
00:09 forrest *okay
00:09 rager going with jessie
00:09 forrest sounds good. I'm running centos6.4 with no issues. I DO have to install rubygems though for the gem command to be available
00:09 rager because it worked fine until I got the latest ubuntu box
00:09 forrest I'm not using rubyenv for this though
00:09 rager ^
00:09 forrest rager: Yeah very odd.
00:09 rager rbenv installs rubygems
00:10 forrest okay
00:13 pedromaltez I'm running 14.04 (ubuntu/trusty64) with lithium v2015.5.0 installed from git through the salt provisioner. ruby works out of the box
00:14 rager I was having issues with ubuntu/precise64
00:14 rager I guess it's a bit old by now, anyway
00:14 pedromaltez or maybe it's cause I got rbenv installed... in any case I haven't had any issues with it.
00:15 rager I used rbenv
00:15 rager it worked fine for a few weeks
00:15 rager then I updated to the latest precise64 box and... nothing
00:15 pedromaltez I had issues with vivid64 because of upstart/systemd
00:16 pedromaltez so I had to go back to trusty
00:16 __number5__ what do you mean by latest precice64? 12.04 is EOF
00:16 pedromaltez __number5__: presumably the latest box from vagrant boxes. they get updated, but presumably precise hasn't been in a while
00:17 pedromaltez https://atlas.hashicorp.com/ubuntu/boxes/precise64
00:17 pedromaltez I lied. it was updated 2 days ago.
00:18 rager exactly
00:18 rager I'm not... completely insane
00:18 __number5__ pedromaltez: got it. oh 12.04 is not EOL yet
00:18 jpic i wonder if it would be possible to change the pillar.example standard to something more useful ie. top.sls. Then you could test from within the formula repo with just salt-call --file-root=$PWD --pillar-root=$PWD state.sls $STATE_TO_TEST
00:19 rager perhaps I should xpost to #vagrant
00:19 jpic that would be insanely conevient me thinks
00:20 jpic or at least something include-able directly (not with a symlink) ie. example_pillar.sls
00:20 rager forrest: since that rails state was offensive to you
00:20 rager what's the proper way to do it?
00:22 kunersdorf joined #salt
00:22 JDiPierro joined #salt
00:23 pedromaltez joined #salt
00:23 forrest rager: I'm just joking with you. It was just everything being lumped in together, it wasn't really a big deal :)
00:23 kunersdorf is httpd: service.running: - reload: True a good way to reload at end of a state?
00:23 forrest kunersdorf: If you want it to reload every time the state is run, sure.
00:23 forrest kunersdorf: Might be better to use 'watch' or 'watch_in'
00:24 rager or listen?
00:24 forrest also an option
00:24 forrest many ways to do it
00:25 jpic i mean that this is not the most straightforward way of testing the example pillar: https://github.com/martinhoefling/rsyncd-formula/tree/dc09a0a65f9153e79b95b9394950539c2c673b0d/test/pillar we probably wouldn't want that in every repo
00:26 kunersdorf ok, can I predicate the reload if the state did something and not have it reload if state didn't do anything?
00:28 jespada_ joined #salt
00:35 rager trying my states on trusty, now
00:35 rager debian's just different enough to require more headaches than I want right now
00:35 jpic kunersdorf: with watch maybe
00:36 kunersdorf or listen as prev stated.  ty
00:37 kunersdorf oh, watch is for other states
00:37 rager watch/watch_in vs listen/listen_in
00:38 rager iirc, listen is only different from watch in that it makes the listening state run after everything else
00:40 bfoxwell joined #salt
00:42 otter768 joined #salt
00:43 kusams joined #salt
00:44 rager aha!
00:44 rager it failed the exact same way with trusty
00:44 rager so on the one hand, it's possibly just a coincidence that it failed right when I upgraded the box
00:44 rager on the other, it's highly inconvenient
00:45 pedromaltez rager: whats the problem exactly?
00:45 rager http://hastebin.com/ajuxibulub.vhdl
00:45 rager with
00:45 subsignal joined #salt
00:45 rager http://hastebin.com/enekavibif.sm
00:45 rager it tries to install bundler but fails at "gem list bundle"
00:46 superseb joined #salt
00:46 rager but `vagrant ssh -c 'gem list bundle'` works fine
00:47 pedromaltez I wish hastebin would let me wrap text...
00:47 pedromaltez so it fails when it's trying to find bundler?
00:48 pedromaltez you need to remove the explicit ruby version
00:48 pedromaltez from gem.installed
00:48 pedromaltez I think
00:48 pedromaltez yes.. that's what I had to do.
00:49 pedromaltez just set the version through rbenv...
00:49 pedromaltez which you're already doing with default: true
00:49 * rager vagrant provision
00:49 rager let's spin the wheel and see what we get!
00:49 pedromaltez try it and see what happens
00:49 aparsons joined #salt
00:50 rager colors spinning round and round, lets' see what the gems decide to do
00:50 rager ooh, it works!
00:50 pedromaltez haha
00:50 pedromaltez yay
00:50 __number5__ rager: any reason you need rbenv in this case? a custom build ruby deb might save you lots of trouble
00:50 pedromaltez I just had the same problem earlier
00:50 rager let's see if this means I kicked the errors further down the road or that it's fixed
00:50 rager __number5__: didn't want to dig into yet another tooling system yet
00:51 __number5__ we use both precise/trusty and ruby 1.9.3 and 2.1.5 at the same time
00:51 rager this is just for a vagrant box
00:51 pedromaltez for development using rbenv is a blessing
00:51 rager also, it slims down the repo to keep things like deb packages out of the vagrant repo
00:51 __number5__ for dev on my laptop I'm using chruby
00:52 rager what's weird is that these states worked fine until yesterday
00:52 rager ghost in the machine
00:52 pedromaltez chruby is cool, but everyone has different tastes I suppose. rvm/rbenv/chruby/whatever-else
00:52 __number5__ maybe bundler have been upgraded?
00:53 rager rbenv is less painful than rvm, and it's baked into salt
00:53 pedromaltez also chruby doesn't autoswitch or install them for you, which is what I really want from MY ruby switcher.
00:53 rager that's why I went with it, in the end
00:53 __number5__ pedromaltez: yes, I do like rbenv but I don't that much featutes, RVM on the other hand is troublesome
00:53 pedromaltez it's got nothing to do with bundler
00:54 pedromaltez at least I don't think so. it's got something to do with gem.installed and using non default ruby versions
00:54 pedromaltez https://github.com/saltstack/salt/issues/23215
00:54 pedromaltez I think that's what it is
00:55 pedromaltez as in, even if it's the default version and you're explicitly specifying it, you're gonna break it.
00:55 dalexander joined #salt
00:56 pedromaltez that person is having issues with `backup`, which obviously isn't bundler
00:58 evle joined #salt
00:58 __number5__ it's rvm in that ticket, lol
00:59 pedromaltez rvm? rbenv. no? maybe I'm confused...
01:00 __number5__ I think the rubygems state support both rvm and rbenv and  vanilla
01:00 solidsnack joined #salt
01:01 pedromaltez but yes, installing a gem with a non default ruby version set by rbenv makes bad stuff happen, apparently.
01:03 pedromaltez rager: do you have a way to run db:create;db:migrate and so on, sequentially and only once? or are you doing those manually?
01:03 rager well
01:03 pedromaltez they each require the previous to be successful
01:04 rager I'm actually loading a db dump before rails gets to the db
01:04 pedromaltez I tried to heed forrest's advice but... :/
01:05 pedromaltez ah, ok
01:06 pedromaltez btw rager, were you using a specific version of salt?
01:07 pedromaltez or just the latest? or how were you doing it?
01:07 pedromaltez it might not happen with older versions of salt.
01:07 rager I'm using "stable"
01:07 pedromaltez ah. well that would've changed a few days ago
01:07 pedromaltez to lithium
01:08 pedromaltez you can lock it at a specific version in your provision block
01:08 pedromaltez salt.install_type = "git"
01:08 pedromaltez salt.install_args = "v2015.5.0"
01:09 pedromaltez that's lithium's version number. but if you need 'correct'(?) gem.installed behaviour you might need to go to a previous version
01:09 rager I'm... just sticking to the workaround
01:09 baweaver joined #salt
01:10 rager it installs faster from apt than it does compiling
01:10 MindDrive Ugh... trying to properly size a Salt master for a large environment is not so fun.  Current system is apparently too low on some resource and a race condition is causing both authentication errors, along with https://github.com/saltstack/salt/issues/23047 when trying to use batch more on a relatively large (>50) number of hosts. :(
01:10 pedromaltez haha, okay. :)
01:20 amcorreia joined #salt
01:21 julez joined #salt
01:24 snaggleb joined #salt
01:28 yexingok joined #salt
01:29 jespada_ joined #salt
01:33 beauby joined #salt
01:35 slavz joined #salt
01:35 bhosmer joined #salt
01:37 thayne joined #salt
01:46 pipeep joined #salt
01:48 timoguin joined #salt
01:57 badon_ joined #salt
02:00 diegows joined #salt
02:02 pedromaltez joined #salt
02:05 rideh joined #salt
02:07 beauby joined #salt
02:10 subsignal joined #salt
02:10 mdupont joined #salt
02:16 clintberry joined #salt
02:22 dendazen joined #salt
02:25 favadi joined #salt
02:30 jespada_ joined #salt
02:32 beauby joined #salt
02:37 _JZ_ joined #salt
02:40 clintber_ joined #salt
02:43 desposo joined #salt
02:43 otter768 joined #salt
02:56 donmichelangelo joined #salt
02:59 salty_to_the_cor joined #salt
03:00 thayne joined #salt
03:01 salty_to_the_cor Question: if i have a salt master, having multiple formulae inside /srv, and suppose all new formales are mapped in file_roots, and if i want to add a new formulae, do i really need to restrt the salt master?
03:02 writtenoff joined #salt
03:03 salty_to_the_cor i am guessing anything being added to fileroots, needs a master restart?
03:03 robawt salty_to_the_cor: i believe you can send it a read message salty_to_the_cor
03:03 robawt i gotta look it up
03:04 Furao joined #salt
03:04 salty_to_the_cor robawt: what do you mean by a read message? could you link me to some docs
03:04 Parabola why are you guys salt'ing at 11pm ?
03:04 Parabola ya'll mofos need jesus
03:04 salty_to_the_cor cz i am salty_to_the_core?
03:04 robawt that's what i'm looking for one moment
03:04 pedromaltez it's 4am here
03:05 Parabola your non freedom, pro terrorism timezones
03:05 pedromaltez lol
03:05 Parabola i guess my question is even more valid, why are you salt'ing at 4am?
03:05 pedromaltez cus I feel like it
03:05 pedromaltez :D
03:06 Parabola fair enough
03:06 pedromaltez I'm making a VM provisioner. it's been a learning experience
03:06 pedromaltez I'm having fun
03:07 salty_to_the_cor pedromaltez: what exactly, i was working on it somedays back, what are u using? vbox?
03:08 pedromaltez virtualbox via vagrant, with salt provisioner
03:08 pedromaltez setting up a ruby on rails app as much as possible for development purposes
03:08 beauby joined #salt
03:08 salty_to_the_cor hah cool
03:08 salty_to_the_cor let me know if i can help anyway
03:09 pedromaltez I think I got it for now, thanks though :D
03:09 salty_to_the_cor robawt : ping
03:09 pedromaltez btw NFS all the way. omg it was so slow before I turned on NFS... rake db:create took 5 minutes.
03:10 salty_to_the_cor or rather robawt: test,ping
03:10 pedromaltez now it takes 5 seconds.
03:11 Parabola pedromaltez thats really cool, my devs could use something like that
03:11 pedromaltez I was like... oh ok I'll just go get a coffee now. or I'd go and read some documentation and come back later way after it had finished cause I got distracted D:
03:12 Parabola <-- new @ company, not much of anything in place admin wise. no monitoring, ticketing, documentation or centralized auth :) getting that solved then onto something like vagrant (and get salt setup, its only in dev atm)
03:12 Parabola pedromaltez what made you pick vagrant over say docker
03:13 patto joined #salt
03:14 pedromaltez Parabola: that may have sounded a bit more sophisticated than it really is... haha. I mean, I'm just making it to learn.
03:14 Parabola i played with docker a bit at a previous job, just curious why you picked vagrant :)
03:15 Parabola for prod you'd want docker, just curious for dev why :)
03:15 salty_to_the_cor why for prod docker?
03:15 Parabola overhead
03:15 salty_to_the_cor its the same docker vs vagrant debate i guess :/
03:15 Parabola vagrant is a virtual machine manager, still relies on the hypervisor
03:16 Parabola docker uses some namespacing trickery
03:16 Parabola and kernel cgroup
03:16 pedromaltez Parabola: let's see... so basically this is for an existing open source app/initiative. It's a coding workshop for people who don't know even html. So you teach them html and progress onto css and javascript. Anyway I'm involved in that. They have a website, obviously, and it's a rails app.
03:17 Parabola oh i dont know much about the debate, i dont use either product, im just stating some technical bits that i think make it clear which i'd pick :)
03:17 Parabola pedromaltez oh cool
03:17 pedromaltez So I wanted to help on that front too, and figured having a VM available for anyone to grab would make it easier than installing postgres and everything that's needed
03:17 pedromaltez and running all the commands
03:18 pedromaltez this way all they need to do is make sure they have their own environment variables set.
03:18 pedromaltez and that's just dropping them into a .env file in the ENV=value format, pretty simple...
03:18 pedromaltez but yeah, so in this case docker doesn't make sense.
03:19 pedromaltez I can't be sure who is going to be loading up the VM. And I want to make it as easy as possible.
03:19 pedromaltez docker needs linux
03:19 pedromaltez and many people are on OS X
03:19 pedromaltez or... god bless them... windows
03:21 pedromaltez I this there's something you can install on OS X for docker, now, too. But it's just easier to use vagrant/virtualbox still.
03:22 Parabola gotcha
03:22 Parabola yeah boot2docker, i've got it on my mac
03:22 Parabola it works, again i just played with it a bit, but it ran
03:23 pedromaltez and yes, docker for prod, if it gets to that. they're just running it in heroku.
03:23 pedromaltez which is a git push away
03:24 salty_to_the_cor any updates for my question ?: robawt: ?
03:25 dendazen joined #salt
03:28 pedromaltez and the vm I'm making is probably not what you'd want for production. I'm just making arbitrary decisions of memory size and number of cpus available to the vm. and the database setup is being hacked to work for the initial provision, but after that it'll have to be manually updated.
03:30 jespada_ joined #salt
03:33 hasues joined #salt
03:33 VR-Jack manual? eh? who does manual? salt it!
03:36 bhosmer joined #salt
03:36 tmmt joined #salt
03:38 hasues left #salt
03:38 pedromaltez VR-Jack haha, well, this would be a development VM... so kinda you'd be expected to manually run the database changes yourself if you've changes what the application expects or pulled in changes that made the application's expectations change
03:43 agentnoel joined #salt
03:48 rojem joined #salt
03:50 ageorgop joined #salt
03:53 TyrfingMjolnir joined #salt
04:00 pedromaltez joined #salt
04:01 pickledpig Hay.  So, I am trying to configure interfaces in some servers, and due to recent changes in linux naming policies, I am finding it difficult.
04:01 pickledpig I exepcted p1p2, etc.  However, i have one serfver with p3p1 and p3p2
04:02 pickledpig I need a way to test for these pXpX devices, and respond accordingly.  I have a state which works for all but one.
04:02 pickledpig There may be more later with confusing naming, but alas, there is one for now. I would rather adapt my cofniguration practice than change the post.
04:02 pickledpig port, even
04:03 VR-Jack jinja can call salt['file.file_exists'] to set for a file
04:04 VR-Jack s/set/check/
04:04 pickledpig hrm
04:04 pickledpig that is a novel approach....
04:05 VR-Jack can also use onlyif test -f <file> in a state, but that might worse
04:05 pickledpig well, I was hoping for a way to cleanly search a dictionary for a group of keys.
04:06 pickledpig maybe map.jinja can help me, but I am new to thw whole affair.
04:13 thayne joined #salt
04:27 salty_to_the_cor joined #salt
04:28 Singularo Has anybody got an working example config or article for using mysql to store pillar data. My googling hasn't found one and its not accepting the user/pass in the master when using salt-ssh
04:31 jespada_ joined #salt
04:42 cptsupermrkt joined #salt
04:43 aawerner joined #salt
04:43 _Cyclone_ joined #salt
04:44 otter768 joined #salt
04:48 I3olle joined #salt
04:57 joeto joined #salt
04:57 smcquay joined #salt
04:58 julez joined #salt
05:03 snuffychi joined #salt
05:04 patto joined #salt
05:07 stoogenmeyer joined #salt
05:10 ITChap joined #salt
05:13 rdas joined #salt
05:16 slavz joined #salt
05:28 hasues joined #salt
05:30 stoogenmeyer joined #salt
05:32 jespada_ joined #salt
05:32 Chadk joined #salt
05:37 bhosmer joined #salt
05:37 rojem joined #salt
05:46 ITChap joined #salt
05:54 kusams joined #salt
06:09 AndreasLutro joined #salt
06:12 solidsnack joined #salt
06:23 lb1a joined #salt
06:33 KermitTheFragger joined #salt
06:33 jespada_ joined #salt
06:33 AndreasLutro good morning
06:36 flyboy joined #salt
06:45 otter768 joined #salt
06:46 Furao joined #salt
06:47 malinoff joined #salt
06:47 ndrei joined #salt
06:47 JayFK joined #salt
06:54 joeto joined #salt
06:54 viq joined #salt
06:55 kusams joined #salt
07:00 Auroch joined #salt
07:05 yuhl_work___ joined #salt
07:10 leszq joined #salt
07:12 patto joined #salt
07:14 kawa2014 joined #salt
07:24 yggdrasi1 joined #salt
07:25 lietu- joined #salt
07:25 snuffychi joined #salt
07:29 supersheep joined #salt
07:34 jespada_ joined #salt
07:37 bhosmer joined #salt
07:45 thehaven joined #salt
07:52 TyrfingMjolnir joined #salt
08:00 stoogenmeyer joined #salt
08:03 thalleralexander joined #salt
08:08 Chadk joined #salt
08:08 Chadk joined #salt
08:14 Xevian joined #salt
08:19 supershe_ joined #salt
08:22 al joined #salt
08:34 jespada_ joined #salt
08:36 julez joined #salt
08:37 cberndt joined #salt
08:39 yexingok joined #salt
08:46 otter768 joined #salt
08:58 badon joined #salt
09:01 chiui joined #salt
09:01 monkey66 joined #salt
09:01 CeBe joined #salt
09:05 Furao joined #salt
09:07 thayne joined #salt
09:08 ITChap joined #salt
09:10 RobbieAB|laptop joined #salt
09:10 djinni` joined #salt
09:11 Furao joined #salt
09:15 microch1p joined #salt
09:17 microch1p Hey guys, I'm having problems trying to create an ACL with the acl.present state. This is the output -> Comment: State 'acl.present' found in SLS 'acl' is unavailable
09:18 microch1p anyone has an idea or an example state?
09:18 microch1p I'm on a CentOS 6.5 with this version of salt-minion: salt-minion 2014.7.5 (Helium)
09:19 Andre-B joined #salt
09:23 ninkotech joined #salt
09:24 Tremad87 joined #salt
09:24 Tremad87 Hello
09:24 Tremad87 I know this question comes up a lot, but i’m trying to work out the best way to manage environments for dev/stage/prod and local
09:25 Tremad87 I am using the split pillar environment method, with only one source for the salt state files.
09:26 Tremad87 The issue I see is that we definitely have conflicting keys that will be merged between different environments and the documentation says that key merges from pillars will be merged in a non deterministic way.
09:27 Tremad87 This scares me as there’s no way to ensure a hierachy.
09:27 Tremad87 Is the normal suggestion to use a pillar ENC ?
09:33 ljanne_ joined #salt
09:35 jespada_ joined #salt
09:36 txomon|home joined #salt
09:38 bhosmer joined #salt
09:42 slav0nic joined #salt
09:42 slav0nic joined #salt
09:43 peters-tx joined #salt
09:43 slav0nic joined #salt
09:46 AndreasLutro Tremad87: can you link to the documentation that mentions non-deterministic merges?
09:46 Tremad87 Yeah sure
09:46 Tremad87 http://docs.saltstack.com/en/latest/topics/pillar/
09:46 Tremad87 Under pillar merges
09:47 Berty_ joined #salt
09:47 Tremad87 I have found the merge setting order too, but this seems to be when you’re merging the same key with multiple values.
09:47 Tremad87 ie. the smart merge option.
09:47 Tremad87 http://docs.saltstack.com/en/latest/ref/configuration/master.html#std:conf_master-pillar_source_merging_strategy
09:48 Tremad87 not conflicting values.
09:48 ndrei joined #salt
09:51 AndreasLutro https://github.com/saltstack/salt/issues/19332
09:53 AndreasLutro for what it's worth, I've been using pillar overrides without problems
09:55 evle joined #salt
10:05 Grokzen joined #salt
10:05 Grokzen joined #salt
10:09 Tremad87 Ah ok. This makes things better then. I guess the documentation was written this way to ensure that people are aware that the order of the merge is a complicated beast!
10:09 Furao joined #salt
10:10 peters-tx joined #salt
10:14 AndreasLutro I agree that it's kinda misleading though, merging/overriding of pillar data is super useful and scaring people away from using it doesn't really do any good
10:17 giantlock joined #salt
10:19 stephanbuys joined #salt
10:27 monkey66 left #salt
10:32 kusams joined #salt
10:36 stephanbuys1 joined #salt
10:36 jespada_ joined #salt
10:38 Tremad87 Yeah. It’s something that is definitely useful and required if you want to have proper separation between environments.
10:40 dendazen joined #salt
10:47 otter768 joined #salt
10:48 stephanbuys joined #salt
10:48 Berty_ joined #salt
10:56 stephanbuys joined #salt
10:57 Tremad87 i still think the layout of the pillar top file with multi environments is not very elegant
10:57 Tremad87 I think I will investigate using reclass as an enc
10:59 supersheep joined #salt
11:00 inad922 joined #salt
11:03 JDiPierro joined #salt
11:05 pdayton joined #salt
11:11 jhauser joined #salt
11:14 evle joined #salt
11:19 johndeo joined #salt
11:29 bhosmer joined #salt
11:30 mbah joined #salt
11:34 Berty_ joined #salt
11:37 jespada_ joined #salt
11:38 kusams joined #salt
11:38 stephanbuys joined #salt
11:39 bhosmer_ joined #salt
11:44 markm_ joined #salt
11:45 bmcorser joined #salt
11:47 denys joined #salt
11:51 bhosmer_ joined #salt
11:59 amcorreia joined #salt
12:00 Furao joined #salt
12:02 johndeo In the windows repository, cache_file doesn't seem to work, howev's from the command line it does work. Is it a bug?  This is the command that works: salt -t 60 '<MACHINENAME>*' cp.cache_file salt://win/repo/ntp/install.ini -l all
12:04 johndeo or maybe it's something I'm overlooking.
12:07 jpic hi all, could anybody explain this to me ? https://github.com/saltstack-formulas/jenkins-formula/commit/09f1e448d68b08bfd835c338ac2495e6fcdbd573 is it a mistake from the author as i believe or did i really miss something ?
12:08 jpic here's the pr https://github.com/saltstack-formulas/jenkins-formula/pull/13 and revert pr https://github.com/saltstack-formulas/jenkins-formula/pull/14
12:08 jpic i really wonder what happened there, commit messages are not sufficiently informative so if anyone had a clue it'd be appreciated ;)
12:09 emaninpa joined #salt
12:12 johndeo It seems everybody is off. :)
12:12 Ph-x joined #salt
12:13 julez joined #salt
12:16 jpic i'm sorry but windows hates me i have no idea about anything in the windows world since 2008 or something
12:17 jpic everytimes i touch a windows server or workstation it crashes in less than 15 minutes
12:18 inad922 joined #salt
12:19 pedromaltez hahahaha :D
12:22 johndeo It's delicate, you have to whine & dine it first. Then the outcome might be good.
12:23 Parabola incorrect, you have to treat it like a hooker, just bend it over and shove it in, thats how windows likes it
12:23 Parabola she acts like shes a good girl, shes a dirty dirty one though, bring a collar for the best results
12:24 pedromaltez jpic I just think for whatever reason they didn't understand something and wanted you to explain. gravyboat didn't want to merge until you explained what was happening - that's all I see to it.
12:24 johndeo Damn... You're mentioning my second option.
12:25 johndeo At least everything will be solved with windows 10.
12:25 jpic pedromaltez: oh well, their nginx config breaks jenkins-cli with the latest jenkins version anyway, and since this formula isn't under ci they don't yet know about it. But I'm really not enclined to submiting a PR in this kind of condition.
12:25 Parabola lol..
12:26 inad922 joined #salt
12:26 jpic i know because my salt repo which uses this formula is under CI
12:26 Parabola i'm an avid mac and linux user, i love OSX, i think it provides a stellar stable environment for laptops and getting work done. having said that windows 8.1 minus a few qwerks isnt that bad
12:26 johndeo But I'll think I'll file a bug, for my problem.
12:26 Parabola its largely stable, the superfetch in windows is currently broken, and should just be disabled, power management isnt great, and metro is terrible, otherwise good desktop platform
12:27 johndeo I thought about buying the new macbook 12 inch, but looking at it closely, that one sucks for my purposes.
12:27 ndrei joined #salt
12:27 jpic i don't even use a desktop platform so who am i to criticize ?
12:27 Parabola yeah, its not a good "working" machine
12:27 pedromaltez jpic: well... if it's the latest version you can see how they might not have fixed it yet. I don't know. If you know how to fix the formula and it's current state is causing you problems it seems like a good idea to fix it or you might be waiting a while. Just make sure to keep an eye on the issue.
12:27 johndeo And is too expensive.
12:27 Parabola its not too expensive for its size, if you havent held one its crazy.
12:28 Parabola you're paying a premium in cost and loss of functionality (hardware wise) to get that size and weight, yet still keeping a retina display and crazy battery
12:29 johndeo Well, 1800 euro's for a core-m with, a bit too much for what you get to me. And it's lacking connections.
12:29 pedromaltez no no no it's super overpriced.
12:29 jpic pedromaltez: you know it really depends on the background of the repo owner. If (s)he doesn't gaf about backward compatibility and it's not unders CI at all then I think it's a bad idea to rely on it
12:30 Parabola 1800 euros?!
12:30 jpic so any PR to it is going to be time inefficiently spent imho
12:30 jpic there's no good reason to by new computers, i buy everything second hand, it's half cheaper.
12:30 pedromaltez Parabola: you're not paying a premium for that. not really. well yes. but mostly it's the novelty factor. they've gotta make it seem luxurious. when in actuality you're getting an ipad in the form of a laptop.
12:31 Parabola the new macbook is 1100 euro in the states.
12:31 Parabola well, 1147 to be exact
12:31 pedromaltez euro?
12:31 Parabola so yes, you're paying crazy VAT or markup or something
12:31 pedromaltez well we always pay crazy VAT
12:31 Parabola yes, its 1300 USD for the new macbook 12"
12:31 jpic i bought a thinkpad x240 with core i7 7G of RAM and 180 of SSD with fingerprint reader for 600 or 700 bucks and it's awesome and linux compatible
12:31 Parabola 1300 USD is 1100 euro
12:32 pedromaltez and extra taxes that aren't in the VAT
12:32 Parabola so, yes, in your case its crazy expensive, and shouldnt be purchased
12:32 jespada_ joined #salt
12:32 Parabola for us however, 1100 euro for that super tiny retina macbook, if portability is key, its not bad
12:32 Sketch jpic: it's awesome unless you've used a previous thinkpad
12:33 pedromaltez because they need to pay for some eu certification or other, I think (or something... don't quote me, blah blah blah), so electronics are always much more expensive over here
12:33 Parabola thats too bad, i suspect american greed comes into play as well
12:33 jpic Sketch: i did have the x230 and x220 even before that (now my son has it xD)
12:33 jpic so, not sure what you mean ...
12:33 Parabola "hey they are used to paying more for things, mark it up!"
12:34 jpic also i don't think a coder can be efficient on a laptop without trackpoint
12:34 jpic it's also bad for health
12:34 pedromaltez I remember when I was younger thinking I'd fly to the states and get a macbook, and it'd be about the same price as a macbook here plus I'd get a holiday out of it too
12:34 nthcode joined #salt
12:34 jpic hehehehe
12:35 jpic dvorak + trackpoint and i'm good to go, on a mousepad + qwerty i'm going to cry my mother xD
12:35 c10b10 joined #salt
12:35 Sketch jpic: i'm still using my x201s, because everything that's come after it has been worse
12:35 Sketch though the x250 appears to fix most of the issues, finally
12:35 pedromaltez generally the staggered keyboard isn't good for your health
12:35 Sketch https://ttboj.wordpress.com/2014/02/02/scathing-review-of-the-lenovo-x240/
12:36 pedromaltez hahaha, honest review name!
12:37 Parabola pedromaltez where do you live?
12:37 pedromaltez UK
12:38 scristian joined #salt
12:38 Parabola at least you get good food, some proper cars we don't have access to, and uh.. john lewis?
12:38 Parabola yes, i think that makes up for 40% markup
12:38 pedromaltez England, London region. But the prices online are obviously the same.
12:38 pedromaltez everywhere in the country
12:38 pedromaltez and hahaha, what???
12:38 Parabola :P
12:39 pedromaltez yes John Lewis.
12:39 pedromaltez at least they make good adverts for christmas
12:39 jpic well yeah breaking bc is always going to piss someone off just like the writer of this article apparently
12:39 jpic and just like me when i see this https://github.com/saltstack-formulas/jenkins-formula/pull/14
12:39 Parabola the best adverts you'll find, they are good enough that even some of us fat yanks know about them
12:40 jpic but yeah probably the x250 is going to be awesome, it'll probably be possible to get a good one in the next 6 months for 600 or 700 bucks
12:40 jpic (second hand)
12:41 pedromaltez jpic, they wanted more activity from you on the PR thread. just wait for them to get back to you now. they probably have a lot of stuff to look over. they figured it was best to wait until you clarified their concerns. not a big issue in my opinion.
12:41 jpic Sketch: that said, i don't see much difference visually between the x201 and x210/220/230
12:41 rideh joined #salt
12:42 jpic pedromaltez: that's the way you see it, i don't see anything to clarify on that pr, it's self explainatory: {% if 'server_name' in jenkins %}server_name = {{ server_name }}; {% endif %}
12:43 jpic maybe i'm just unable to communicate properly then because i see nothing to clarify
12:43 jpic if the pillar was set, use it as expected, if not, behave like before (maintain BC)
12:43 pedromaltez I see it that way because that's what they said they wanted.
12:43 jpic yeah, cause they didn't read the PR
12:43 jpic and that's why their comments are so false
12:43 jpic "will it break if the pillar was not set ?" come on ;)
12:44 Sketch jpic: visually, no.  but the screen resolution sucks in the x220+ (they finally introduced a 1080p as an option on the x250)
12:44 Sketch https://ttboj.wordpress.com/2014/02/02/scathing-review-of-the-lenovo-x240/
12:44 pedromaltez it's just a question. they might not necessarily know the innards of jenkins
12:44 Sketch and there are issues with the keyboard and mouse buttons
12:44 jpic Sketch: as long as i have redshift i'm happy with the screen xD
12:45 pedromaltez they want to be sure it won't break. I obviously don't know what went on in their heads - but just be calm and answer. no need to get mad at them.
12:45 pedromaltez it'll be merged back in eventually
12:45 jpic pedromaltez: but if they want to be sure it won't break then they should setup travis !
12:45 pedromaltez then suggest it
12:46 jpic you're right i'm probably thinking far away from them
12:46 pedromaltez or better yet set it up yourself
12:46 pedromaltez and submit it to them
12:47 Sketch jpic: i've been waiting for a high dpi display.  at least in the x250, 1080p is about halfway there...
12:47 catpig joined #salt
12:47 pedromaltez no it's not...
12:47 pedromaltez well.. ok. halfway there. but i'll still see all the pixels
12:47 jpic Sketch: yeah so it depends on your needs i guess
12:47 Sketch ok, not quite halfway, but it's a big improvement over 1366x768
12:48 otter768 joined #salt
12:48 pedromaltez using apple's hDPI screens has made everything else look fuzzy D:
12:49 pedromaltez but yes, much better. how do you even work? does a window even fit?
12:49 jpic pedromaltez: it's not my fault people code without TDD these days and do not read PRs and do reverts based on supersticion, this kind of setting is pretty special and i don't think i have the level for that challenge
12:49 jpic i'm thinking i should instead wait and let it fail cause it's the only really good way to learn
12:50 pedromaltez that's just spiteful
12:50 TOoSmOotH I like tacos
12:50 pedromaltez I like burritos!
12:50 jpic i'm sorry ?
12:51 pedromaltez idk - random dude
12:51 JDiPierro joined #salt
12:53 jpic why does everything have to be complicated with you guys, jeez
12:54 jpic left #salt
12:54 TOoSmOotH someone is seriously butt hurt today
12:56 cpowell joined #salt
13:01 mapu joined #salt
13:02 ferbla joined #salt
13:06 omegamike joined #salt
13:07 elfixit joined #salt
13:09 rick_ joined #salt
13:12 pkimber joined #salt
13:12 timoguin joined #salt
13:17 dendazen joined #salt
13:19 primechuck joined #salt
13:19 jdesilet joined #salt
13:20 primechu_ joined #salt
13:20 tristianc joined #salt
13:20 Tremad87 Ok, I have reclass working now, but how do you use the environment parameter?
13:20 guanophobic joined #salt
13:21 Tremad87 It seems like the marking of a node for it’s particular environment works, but you can use that in reclass to assign certain parameters??
13:22 dyasny joined #salt
13:26 Furao joined #salt
13:28 tkharju joined #salt
13:28 subsignal joined #salt
13:33 pedromaltez does anybody know what P1-P4 labels mean? eg. https://github.com/saltstack/salt/labels/P2
13:33 AndreasLutro prioritt
13:33 AndreasLutro y
13:34 pedromaltez Ah, great. thanks. lower being higher priority?
13:34 pedromaltez lower number
13:34 pedromaltez or the other way round?
13:34 pedromaltez haha
13:35 racooper joined #salt
13:35 troyready joined #salt
13:38 AndreasLutro lower = higher priority, I think
13:40 drico joined #salt
13:41 drico Hi, does anyone managed to merge variable from two pillar file, one including the other one ?
13:42 inad922 joined #salt
13:42 drico I always get some conflicting ID
13:43 JDiPierro joined #salt
13:44 ntropy drico: this is a frequent question, pillar merges are described here http://docs.saltstack.com/en/latest/topics/pillar/#pillar-namespace-merges
13:44 drico yes I've read it a hundred, times and went thru yaml anchoring etc
13:45 drico may I paste you the sls to have your opinion ?
13:45 ntropy fine by me :)
13:45 babilen http://refheap.com
13:47 ntropy babilen: muscle memory? :)
13:48 babilen ntropy: abbreviation in irssi
13:49 babilen "ref" expands to it. I have a couple of those :)
13:49 Furao joined #salt
13:50 faliarin joined #salt
13:50 drico ntropy, https://gist.github.com/anonymous/626b5932073712136989
13:51 joeto joined #salt
13:54 andrew_v joined #salt
13:55 numkem Deevolution: Did you managed to get any feedback beside the issue on github concerning the pillar.get vs pillar.items problem?
13:55 hasues joined #salt
13:56 hasues left #salt
13:56 drico it looks like it's working if the two sls are included by a top SLS like in the example
13:56 drico but not from one another
13:58 Deevolution numkem:  Nothing yet.
13:58 Deevolution numkem: I was planning to ping them again today.  I believe I've now provided a Vagrant setup that will duplicate the issue on demand.
13:58 numkem Deevolution: alright, thank you. It's a pretty big issue.
13:59 Deevolution numkem: For us also.
14:00 kaptk2 joined #salt
14:02 Tyrm joined #salt
14:02 julez joined #salt
14:02 Tremad87 who here actually uses reclass?
14:03 mpanetta joined #salt
14:04 mpanetta joined #salt
14:05 fyb3r joined #salt
14:06 babilen drico: I wouldn't include one in the other, but simply list *both* (with overlapping keys) in the top.sls
14:08 dendazen joined #salt
14:09 WildPikachu joined #salt
14:10 ntropy would include without context achieve the same?
14:14 babilen ntropy: I am not sure in which cases the smart merging will be used.
14:14 babilen I know that it is if you simply include pillars with overlapping keys in top.sls and if you pass "merge=True" to pillar.get
14:15 babilen I think the "include" is just 'normal' jinja stuff that does whatever it does
14:15 TyrfingMjolnir joined #salt
14:16 icksa joined #salt
14:17 kusams joined #salt
14:17 drico babilen, yes I think I have no other alternative, tks
14:24 bhosmer_ joined #salt
14:25 salty_to_the_cor joined #salt
14:25 salty_to_the_cor Morning: anyone knows how to update file_roots dynamically?
14:26 cpowell you have to restart the master when you change them
14:26 salty_to_the_cor i am adding a new formula, so dont want to restart the master
14:26 salty_to_the_cor so this implies we need to have a downtime whenever we promote a formula?
14:27 supersheep joined #salt
14:27 cpowell no, but if you change the file_roots section in your /etc/salt/master then you need to restart
14:28 cpowell if you are just changing the contents of the directories, then you don't
14:28 salty_to_the_cor no i am adding a new formula, so this implies an addition to /etc/salt/master right?
14:29 cpowell no
14:29 salty_to_the_cor is there any other way for formula promotion?
14:29 cpowell just copy the formula
14:29 cpowell thats fine
14:29 salty_to_the_cor and how would salt know where to look for the new formula?
14:29 cpowell are you putting it into a directory already listed as a file_roots?
14:29 cpowell or creating a new file_roots
14:29 salty_to_the_cor its like it maps into file roots: /srv, /srv/formula/nginx
14:30 salty_to_the_cor new formula , so i assume a new file roots
14:30 Brew joined #salt
14:30 cpowell no, you only need one file_roots
14:30 cpowell unelss you are doing envs
14:30 cpowell or something more complicated
14:31 salty_to_the_cor just to be clear. Say i have two formulas, nginx and postgres, and they reside in /srv/formulas/nginx, and /srv/formula/postgres, how should the file roots look like?
14:31 cpowell # /srv/formulas
14:32 salty_to_the_cor and when i add another formula, say "testformula", how will it change?
14:32 cpowell if you have /srv/formulas as your fileroots, then nothing changes
14:32 cpowell salt will find it
14:33 salty_to_the_cor ok, one last question on it, how will it affect the _modules and _states, does one manually need to copy them to /srv/salt/_modules or they should reside inside the formula? like /srv/salt/formula/nginx/_module
14:33 cpowell they should reside in your file_roots
14:33 cpowell so /srv/salt
14:33 cpowell by default
14:34 salty_to_the_cor so one needs to manually copy them?
14:34 salty_to_the_cor a formula by def cant have _modules?
14:34 cpowell you will need to create those directories
14:34 cpowell no
14:34 cpowell _modules and _states are shared acrorss the salt env
14:35 salty_to_the_cor so its not a good practice to have _modules and _states per formula?
14:35 cpowell you can't
14:35 salty_to_the_cor like /srv/salt/formula/nginx/_module
14:35 salty_to_the_cor well if we map our file roots to /srv/salt/formula/nginx
14:35 salty_to_the_cor _module will be it in
14:35 cpowell no, it would be {file_roots}/_states
14:36 salty_to_the_cor yup, exactly /srv/salt/formula/nginx/_states
14:36 cpowell # /srv/salt/formula/nginx/ is not your file_roots
14:36 cpowell #/srv/salt/formula is
14:37 cpowell open your master file and see what you put there
14:37 cpowell if you havent changed it, then the default is /srv/salt
14:37 salty_to_the_cor ok but here http://docs.saltstack.com/en/latest/topics/development/conventions/formulas.html
14:37 salty_to_the_cor file_roots:   base:     - /srv/salt     - /srv/formulas/apache-formula
14:37 salty_to_the_cor they have this example, so they map to apache-formula
14:38 salty_to_the_cor apache-formula is the new file roots
14:38 salty_to_the_cor so it can have its _modules?
14:39 cpowell so thats a poor example
14:39 cpowell by default you have file_roots of /srv/salt
14:39 cpowell you can place ALL of your formuls/states there
14:39 cpowell so you would have /srv/salt/nginx
14:39 cpowell or /srv/salt/postgres.sls
14:39 cpowell and /srv/salt/_modules
14:40 cpowell in that example they added a new file_roots for a single formula...not sure why
14:40 salty_to_the_cor and modules etc have to be copied in one directory /srv/salt/_modules
14:40 salty_to_the_cor ok
14:40 r900 joined #salt
14:40 cpowell if you write your own modules then yes
14:40 cpowell they might need to be synced to your minions with saltutil.sync_all
14:41 bhosmer joined #salt
14:41 salty_to_the_cor but the other way, its a separation of concern, i just liked that, for anything corresponding to one formula was in one unit ( like all modules & states). If you are gettign my point
14:41 markm joined #salt
14:42 cpowell trust me, that won't scale well
14:42 cpowell what happens when you have 50 formulas
14:42 cpowell you then have to restart the master each time you add a new one
14:42 cpowell the other way, it will find them
14:42 cpowell no restart
14:42 salty_to_the_cor haha! thanks so much!
14:42 farcaller joined #salt
14:43 bhosmer joined #salt
14:43 c10b10 is there any way of making salt-ssh show output as it is created, instead of spitting it all at the end?
14:43 salty_to_the_cor yea, i was assuming i was missing something? to load formulas
14:43 farcaller hi. can someone give me an advice how to debug a 'pillar not working' issue? I have absolutely no idea where to start
14:43 gladiatr joined #salt
14:43 salty_to_the_cor like a util or something or salt has an sls which it check everytime to see the file_roots
14:43 cpowell c10b10: it shows the output when the server has finished, sorry
14:44 c10b10 ok
14:44 salty_to_the_cor that would be cool though :/
14:44 cpowell farcaller: any more information besides 'pillar not working'?
14:44 jhammons joined #salt
14:44 farcaller my pillar/top.sls is `base:  /  '*':  /  - test`, and pillar/test.sls is `test: key`. salt saltutil.refresh_pillar && salt pillar.items returns nothing
14:44 eriko joined #salt
14:45 cpowell can you paste them into a gist?
14:45 farcaller cpowell: https://gist.github.com/farcaller/1913e03ac33b28e891e0
14:45 masterkorp Hello
14:46 cpowell farcaller: that looks right...
14:46 farcaller cpowell: also added the part from master conf
14:46 masterkorp I want to match a minion on the top file that contains the x keyword, but not the the y keyword on the name
14:47 farcaller cpowell: is there any debug option I can set to see what's going on?
14:47 cpowell masterkorp: I know you can do OR's with the matchers, don't think you can do NOT's
14:48 cpowell farcaller: turn off the salt-master and run it from the cli with salt-master -d debug
14:48 cpowell sorry -l not d
14:48 stoogenmeyer joined #salt
14:48 cpowell -l debug
14:48 cpowell it will spit everything out
14:48 farcaller yeah, did that
14:48 otter768 joined #salt
14:49 farcaller [DEBUG   ] Results of YAML rendering:
14:49 farcaller OrderedDict([('test', 'key')])
14:49 farcaller it parses it as expected
14:49 cpowell are you running it from the master?
14:49 cpowell the salt command?
14:49 farcaller yup
14:49 cpowell are you targeting a minion?
14:49 farcaller yup
14:49 PI-Lloyd check the master log for any errors
14:49 pedromaltez masterkorp, farcaller: can't you do AND NOT?
14:50 farcaller and not?
14:50 masterkorp pedromaltez: i think so
14:50 pedromaltez yes, NOT by itself means nothing, but AND NOT does
14:50 pedromaltez if I remember correctly
14:51 cpowell farcaller: that sounds correct
14:51 PI-Lloyd top file would look something like 'hostname1 and not hostname2':
14:51 farcaller well, I have * in there
14:51 pedromaltez indeed
14:52 cpowell masterkorp: Matchers can be joined using boolean and, or, and not operators.
14:52 masterkorp cpowell: can you give am example
14:52 pedromaltez (sorry farcaller - that was meant for cpowell, not you)
14:52 cpowell salt -C 'not web-dc1-srv' test.ping
14:52 cpowell http://docs.saltstack.com/en/latest/topics/targeting/compound.html#compound-matchers
14:52 PI-Lloyd ^^
14:52 cpowell sorry salt -C '* and not G@kernel:Darwin' test.ping
14:53 cpowell becareful, there are some docs that reference beta versions
14:53 pedromaltez yay for obsolete documentation
14:53 pedromaltez :P
14:53 cpowell in this case, its actually too new
14:54 pedromaltez mmm, yes, that :D
14:54 pedromaltez what would the name be...
14:54 cpowell which is worse becuase then I get excited I can do something, but then I cant :(
14:54 pedromaltez hahaha
14:54 PI-Lloyd I know that pain
14:55 pedromaltez ultracurrent documentation??
14:55 pedromaltez :S
14:56 PI-Lloyd the docs are like a mystic ball.. allows you see things that aren't really there
14:56 icksa joined #salt
14:56 VR-Jack heh, or are there but only accessible by python sometimes. :(
14:56 CedNantes joined #salt
14:56 CedNantes Hello
14:57 txomon|home joined #salt
14:57 lothiraldan joined #salt
14:57 txomon|home_ joined #salt
14:58 VR-Jack most wheel.key and event functions are only accessible externally. :(
14:58 CedNantes quick questions : when using salt-cloud with profile, i you want a custom ip address you actually have to set it in the profile configuration file. Is there any mean to pass it as an argument ?
14:59 bhosmer joined #salt
14:59 clintberry joined #salt
14:59 pedromaltez clairvoyant!
14:59 pedromaltez clairvoyant docs.
14:59 PI-Lloyd ikr
14:59 cpowell pedromaltez: haha, 5 minutes late
15:00 PI-Lloyd better late than never ;)
15:00 VR-Jack unless kicking a dead horse
15:00 farcaller okay, apparently I had a typo in states/top.sls that prevented pillars to sync
15:00 farcaller :)
15:00 PI-Lloyd lol Jack
15:00 VR-Jack especially if it is burried. I hate to redig a hole
15:00 pedromaltez cpowell: well I tried another word first... :P
15:01 PI-Lloyd VR-Jack: You are one disturbed individual :p
15:01 pedromaltez haha what is this...
15:01 cpowell so I am trying to add custom grains during an orchestration and the salt.function yaml syntax is kicking my utt
15:02 VR-Jack PI-Lloyd: hard to argue with that. lol
15:02 PI-Lloyd heh
15:02 jlhammons joined #salt
15:03 farcaller now I have a problem that pillar key is returned by pillar.items, but not by pillar.get. https://gist.github.com/farcaller/1913e03ac33b28e891e0 any ideas?
15:03 al5d joined #salt
15:04 al5d hi there
15:05 al5d i'd like to know if it's possible to dump the Publisher messages between master and minions?
15:05 drico farcaller, did you try to saltutil.refresh_pillar ?
15:05 * farcaller facepalms
15:06 farcaller okay, final question, how do I match on etcd.server == true in state top.sls?
15:06 farcaller 'etcd:server:true' with `match: pillar` doesn't work
15:08 al5d i'm having trouble with the salt.client.LocalClient API, I guess there's a problem with the messages sent from the master : when trying the grains.setval function, i get 'TypeError encountered executing grains.setval: setval() takes at most 3 arguments (13 given). See debug log for more info.  Possibly a missing arguments issue:  ArgSpec(args=['key', 'val', 'destructive'], varargs=None, keywords=None, defaults=
15:08 al5d (False,))' as an answer
15:08 al5d i get the same answer with grains.get
15:08 cpowell are you trying to set multiple grains with a single call?
15:09 al5d no, only 1 grain on 1 minion on each call
15:09 cpowell I am guessing around 10 grains?
15:09 cpowell hmmm
15:09 ageorgop joined #salt
15:10 mpanetta al5d: http://docs.saltstack.com/en/latest/topics/event/index.html It has been updated since I last looked, but tells you how to watch events.
15:10 cpowell al5d: it should look like cls.client.cmd(server.name, 'grains.setval', ['environment_id', 'test'], 120)
15:10 cpowell mpanetta: Get back to work MIke
15:10 mpanetta cpowell: :P
15:11 al5d cpowell: but i think you have something right, the original list of targets is 10 elements long
15:11 cpowell you need to change the default matcher from a glob to list
15:11 drico farcaller, maybe {% if pillar['etcd']['server'] == 'true' %} -foo {% endif %} ( not sure maybe do salt['pillar.get']('etcd.server') )
15:12 farcaller drico: yeah, used pillar['etcd:server'], should have used pilalr.get
15:12 farcaller thanks all
15:12 cpowell al5d: look at the expr_form and change it to list
15:12 cpowell http://docs.saltstack.com/en/latest/ref/clients/#localclient
15:12 masterkorp cpowell: i think it works
15:12 cpowell nice
15:13 Heartsbane A while ago I heard that there is a fix for sluggish minions on older versions of zmq
15:13 cpowell al5d: salt_client.cmd(server_names,'state.sls', [formula, env], SALT_TIMEOUT, expr_form='list')
15:14 cpowell with server_name being a []
15:14 cpowell if you are passing a literal list of servers rather than matching with a glob or grain
15:17 rm_jorge joined #salt
15:20 kusams_ joined #salt
15:21 al5d cpowell: the problem is that I use a single string as tgt, which is the minion id. the parameter is not a list in this case
15:21 cpowell then you shouldnt have to change anything
15:22 al5d http://pastebin.com/DXJb5Q0g
15:22 al5d here's my script
15:24 elfixit1 joined #salt
15:24 dyasny joined #salt
15:25 cpowell targets will only equal the keys(), not the entire salt output
15:25 cpowell for one
15:26 cpowell other than that, it should work for test.ping
15:27 cpowell wait, nm
15:28 whiteinge FYI: Interruptions to the docs site are expected over the next short while.
15:28 cpowell line 12: the args are not in the proper order
15:28 cpowell client.cmd(server.name, 'grains.setval', ['environment_id', 'test'])
15:28 cpowell the grain you are settings needs to be in a list syntax
15:29 cpowell client.cmd(minion_id, 'grains.setval', ['mysql_pool_id', pool_id])
15:29 cpowell try that
15:29 al5d ook
15:30 rm_jorge joined #salt
15:30 al5d cpowell: thanks a lot! that was the issue
15:31 jlhammons_ joined #salt
15:31 cpowell np :)
15:34 c10b10 so i have a salt-ssh command running for a while now
15:34 c10b10 how do you propose debugging what's going on with it?
15:37 kusams joined #salt
15:37 al5d c10b10: not very salt-ssh specific, but strace can be useful
15:39 c10b10 well, the absence of a python processes running is a clue that the process has finished executing
15:41 jimklo joined #salt
15:41 bhosmer joined #salt
15:41 bhosmer joined #salt
15:41 cpowell sorry c10b10, I haven't used salt-ssh
15:47 Gareth whiteinge: ping
15:47 whiteinge Gareth: pong
15:48 Gareth whiteinge: question for you, if you've got a few minutes.
15:48 whiteinge yah
15:49 martineg_ joined #salt
15:49 theologian joined #salt
15:49 ajw0100 joined #salt
15:50 Gareth whiteinge: Can you think of any reason why the contents of __opts__ would be different if you're using salt-call on the minion to call a function vs salt & targeting against a minion calling the same function?
15:51 whiteinge Gareth: it could just be a mistake. assembling __opts__ has never (unfortunately) been a central, well-known operation. it happens in a bunch of places and happens in different ways.
15:52 thayne joined #salt
15:52 microch1p left #salt
15:52 whiteinge Gareth: are you seeing missing data in one vs. the other?
15:53 Gareth whiteinge: An example of where I'm seeing this.  If you add something into the schedule, then delete it using salt-call.  It's gone, but if you run schedue.list via salt-call it still shows up, and running schedule.delete (again using salt-call) it attempts to delete it.  Dumping out __opts__ shows it's there when running things via salt-call, but now there when run via salt.
15:53 JayFK joined #salt
15:54 whiteinge ah...
15:55 whiteinge my guess as to what's happening is that salt-call is instantiating it's very own copy of __opts__ rather than using the in-memory copy that the already-running salt-minion daemon is using.
15:55 MatthewsFace joined #salt
15:55 whiteinge salt-call is it's own thing/entity and starts up brand new processes for itself
15:56 linjan joined #salt
15:56 arnoldB yeppieh, looks like the major works on the docs do arrive
15:56 MatthewsFace joined #salt
15:56 arnoldB (at least the redesign)
15:56 whiteinge new docs are indeed up!
15:56 Gareth whiteinge: Hmm yeah, I was thinking that might be the case.
15:57 whiteinge let us know if you run into hiccups with the new docs. for some reason the CDN is being a touch slow in caching some of the CSS/JS files....
15:57 Gareth Ooo. new docs look awesome.
15:58 whiteinge Gareth: i'm not sure, offhand, what would be a good way to get those two invocations to share state for the scheduler config. cachedout would probably be a good candidate to bounce ideas off
15:59 iggy arnoldB: and most important, it now defaults to $current_release docs, not devel
15:59 arnoldB whiteinge: the web page search of my browser (opera) doesn't work
15:59 arnoldB iggy: +100
15:59 arnoldB whiteinge: chrome is fine
15:59 iggy of course, now we won't notice all the modules committed without versionadded tags
15:59 stephanbuys joined #salt
15:59 whiteinge heh
16:00 whiteinge that's true :)
16:00 Gareth whiteinge: I'm thinking I just can't rely on __opts__, maybe have to fire an event off to get the latest list of scheduled jobs.
16:00 iggy whiteinge: http://docs.saltstack.com/en/latest/salt-modindex.html has "Table of Contents
16:00 iggy Glossary"
16:00 whiteinge arnoldB: any helpful JS errors in the console?
16:00 iggy x2
16:00 iggy (it was like that before too, fyi)
16:02 oyvindmo joined #salt
16:04 arnoldB whiteinge: there are no errors, looks like it works as extended
16:04 whiteinge If you run into the odd 404 on the latest docs URL, try jumping to the 'develop' docs. Looks like there's a thing or two that need ironing out.
16:04 arnoldB *expected
16:04 arnoldB the side reloads after opening the search bar
16:04 arnoldB *site
16:07 Tyrm joined #salt
16:07 whiteinge that may also be a JS error.
16:08 7F1AA0UK2 joined #salt
16:09 rojem joined #salt
16:09 martintamare joined #salt
16:09 martintamare Hello there ! I've having difficulties understanding pillar matching to targeting minion
16:09 martintamare here is my issue
16:09 martintamare http://pastebin.com/gpbNje8D
16:10 martintamare using list : target is ok
16:10 martintamare using dict : target is not ok
16:11 iggy salt -C 'I@haproxy:listen2:front01.test.alkivi.fr:port' test.ping
16:11 iggy try that
16:12 iggy maybe I@haproxy:listen2:front01.test.alkivi.fr:port:9000
16:13 martintamare it work with the final value, but my goal is to stop without the port. If value is defined target should work, shoud'nt it ?
16:14 oyvindmo joined #salt
16:15 iggy I'd say yes, I was just seeing if where it broke
16:16 iggy try I@haproxy:listen2:front01.test.alkivi.fr*
16:16 martintamare I@haproxy:listen2:front01.test.alkivi.fr:port : ko
16:16 martintamare I@haproxy:listen2:front01.test.alkivi.fr:port:9000 : ok
16:16 martintamare I@haproxy:listen2:front01.test.alkivi.fr* : ko
16:17 JayFK joined #salt
16:18 jvblasco_ joined #salt
16:18 Grokzen joined #salt
16:22 kunersdorf joined #salt
16:22 iggy I'd open an issue
16:23 iggy that's not the kind of output I'd expect
16:23 martintamare Ok, thx for the confirmation
16:24 asoc joined #salt
16:26 KyleG joined #salt
16:26 KyleG joined #salt
16:26 jvblasco_ is it posible to use the python renderer in a pillar?
16:27 smcquay joined #salt
16:29 iggy jvblasco_: yes
16:31 rojem joined #salt
16:31 renoirb jvblasco_, look at the docs at the "salt.renderers.py" part. You’d want to have a #!py on top of the file.
16:31 renoirb http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.py.html#module-salt.renderers.py
16:31 jvblasco_ i'm having an strange issue. I'm implementing a python renderer for a pillar, when i do i state.highstate the pillar data most times renders correctly
16:31 jvblasco_ but when i call pillar.items
16:33 jvblasco_ i get an Attribute error being the full stacktrace: http://pastebin.com/QtWxWSLY
16:33 jvblasco_ in the highstate it renders the pillar correctly
16:33 primechuck joined #salt
16:34 aparsons joined #salt
16:34 bhosmer_ joined #salt
16:35 dalexander joined #salt
16:36 elfixit joined #salt
16:37 iggy I just said it was possible, not highly tested ;)
16:38 jvblasco_ hehehehehe
16:38 writtenoff joined #salt
16:38 jvblasco_ asking because i don't have many experience with salt
16:38 hacfi joined #salt
16:38 spookah joined #salt
16:39 jvblasco_ so, not sure if i arrived to a bug
16:39 iggy it wouldn't suprise me if there were bugs in that area, as I said, I doubt that's a feature that gets a lot of testing
16:39 hemphill joined #salt
16:39 jvblasco_ or i'm missing something in the master or minion configuration
16:41 desposo joined #salt
16:41 jvblasco_ maybe some cache is playing mind games, i'll try to figure something out, thnx for the answers ;)
16:41 iggy can you share the code for your pillar file?
16:42 iggy it kind of looks like someone is doing something like "if not __salt__:\n    __salt__ = {}\n   __salt__['pillar'] = {}"
16:43 iggy I've seen people do stuff like that frequently to get around the dunder dicts not existing at certain stages
16:44 oyvindmo joined #salt
16:45 hemphill What is the best way to get a list of hosts that match a grain to feed into scripts?
16:48 ALLmightySPIFF joined #salt
16:49 jespada_ joined #salt
16:49 otter768 joined #salt
16:51 cheus joined #salt
16:52 dangra joined #salt
16:52 jimmybee joined #salt
16:53 ALLmightySPIFF joined #salt
16:54 ALLmight_ joined #salt
16:54 iggy whiteinge: is there someone we can send docs feedback to?
16:58 wendall911 joined #salt
17:00 thedodd joined #salt
17:00 rdutch joined #salt
17:01 rdutch left #salt
17:02 lothiraldan joined #salt
17:04 thayne joined #salt
17:05 forrest joined #salt
17:05 linjan joined #salt
17:08 Nicop06 joined #salt
17:09 jvblasco_ iggy: sure, np --> https://gist.github.com/anonymous/1bb2f4a7be3175853172
17:14 nthcode joined #salt
17:14 tr_h joined #salt
17:16 dangra joined #salt
17:17 martintamare joined #salt
17:20 robinsmidsrod joined #salt
17:21 iggy jvblasco_: yeah, in that case, I'd guess it's an actual salt issue
17:24 jespada_ joined #salt
17:24 _JZ_ joined #salt
17:24 rdutch joined #salt
17:26 bradthurber joined #salt
17:26 monkey66 joined #salt
17:27 Chadk joined #salt
17:28 Snick joined #salt
17:28 _Cyclone_ joined #salt
17:32 rojem joined #salt
17:33 rm_jorge joined #salt
17:34 druonysus joined #salt
17:34 tedski the new docs look great, but breakage continues to be a problem
17:35 tedski can we help by writing tests for the docs so they don't get pushed in such a broken state?
17:35 tedski right now, both docker state module docs are missing
17:35 farcaller I can't seem to make gpg to work in pillar. I added `renderer: jinja | yaml | gpg` to master config, but the distributed pillar value is unencrypted gpg block. no warnings/errors
17:35 Snick Hey all, learning salt and playing with salt virt at the moment. I'm trying to figure out how to get salt virt.init to create a 64-bit VM. It seems to be defaulting to i686 and preventing my 64-bit images from booting.
17:37 bradthurber hi everyone. I am trying reactor for the first time. Seems ok up to the point where I see message like "Compiling reactions for tag salt/cloud/c6-web-dev/created" in master log. I'm trying to troubleshoot with "salt-run state.event pretty=True" but it never returns. Any ideas what to try next?
17:38 forrest Hmm, has anyone noticed that some doc pages just continually load/reload/
17:38 forrest or was that already mentioned this morning by the east coast guys
17:39 forrest oddly enough it's when you go between tabs
17:39 bradthurber regarding docs. I know I'm not the first to mention this but it sure would be nice if the default documentation was for latest stable - not develop
17:39 julez joined #salt
17:39 forrest ahh it's because of trackalyzer.
17:40 forrest bradthurber: It is now
17:40 solidsnack joined #salt
17:40 forrest bradthurber: They're working on a docs update, and that is one of the new pieces of functionality
17:40 bradthurber forrest: great to hear!
17:40 forrest bradthurber: Yeah it's live now, looks like there are still some kinks
17:42 monkey66 left #salt
17:45 fbergroth joined #salt
17:46 forrest basepi: Should we be filing issues for problems with the new docs or wait?
17:46 jvblasco_ iggy: and what "if not __salt__:\n    __salt__ = {}\n   __salt__['pillar'] = {}" that mean?
17:46 spookah joined #salt
17:50 basepi forrest: good question, let me ask
17:50 forrest basepi: Thanks
17:50 ajw0100 joined #salt
17:51 spookah joined #salt
17:51 iggy I wondered the same thing
17:52 bradthurber I really like the docs selector that allows current stable/prior stable/develop. It seems search function only searches current stable even if develop is checked
17:52 iggy jvblasco_: like I said, it's to work around the dunder dicts not being available at all times
17:54 bhosmer joined #salt
17:55 forrest bradthurber: Yeah once we hear back regarding issues we can start filing stuff. I just don't want to start filing a bunch of issues.
17:56 jhauser_ joined #salt
17:57 iggy I swear some people don't look at formula PRs at all before they merge them
17:59 jvblasco_ iggy: so it may be a formula issue?
17:59 iggy jvblasco_: no, that was a completely unrelated comment
17:59 baweaver joined #salt
17:59 jvblasco_ iggy: ok
18:01 supersheep joined #salt
18:02 rojem joined #salt
18:02 denys joined #salt
18:03 edrocks joined #salt
18:05 basepi forrest: file issues, tag @jacobhammons on the issues
18:05 forrest basepi: Sounds good, iggy ^
18:05 cheus Sometime before the most recent release I was using salt + lxc pretty heavily and things seemed stable but in this most recent run of it I can't get it to bring up a new ubuntu container with the simplest possible configuration. Has anyone had recent success with that?
18:06 iggy rgr
18:07 alemeno214 joined #salt
18:07 aparsons joined #salt
18:08 druonysuse joined #salt
18:08 alemeno214 Hey! Quick question. How many worker threads do you guys have configured on your Salt Master? I'd like to increase it on my setup, but I don't want to go crazy.
18:09 iggy I think general concensus goes something like "number of cores + a few"
18:10 neogenix joined #salt
18:10 iggy I've got some masters at default (5), some at 2 * #cores, some at #cores
18:10 baweaver joined #salt
18:11 rdutch I would add a few at the time and see how it behaves, like go from 5 to 7
18:13 alemeno214 Okay awesome, thanks guys!
18:20 amcorreia joined #salt
18:20 hacfi joined #salt
18:21 cpowell can the minion still not restart itself?
18:22 cpowell would this work? https://github.com/python/psf-salt/blob/master/salt/base/salt.sls#L48
18:23 rap424 joined #salt
18:23 iggy I restart the minions all the time (well, salt-formula does)
18:23 cpowell what manner are you using? The docs refer to using 'at' or forking/disowning
18:23 iggy salt-formula... I don't think it does anything special
18:24 iggy yeah, just standard watch statements
18:24 cpowell ok, i was reading http://docs.saltstack.com/en/latest/faq.html#linux-unix
18:25 iggy I can live with possibly losing a return from a minion
18:27 solidsnack joined #salt
18:31 eliasp congrats on the docs.saltstack.com rollout… looks great! and finally access to different versions while defaulting to the current stable release!
18:31 iggy ^
18:32 nthcode joined #salt
18:34 iggy forrest: help! (when you've got some free minutes, not urgent) https://gist.github.com/iggy/cb800f3401a6364cd5bd
18:35 forrest Points 3 and 4 seem to be too much to me. Point 3 primarily because that increases our barrier to entry, it's something I've never done with my contributions to be honest. Point 4 could lead to a lot of spam for some people, and again increases the barrier to entry
18:35 bhosmer_ joined #salt
18:37 iggy well, we need someway to be able to communicate with people that have access to commit to code that a lot of people use
18:37 iggy case in point... how the hell am I going to get people to read this document once I'm done with it
18:37 forrest iggy: We should probably just put it in every readme realistically.
18:37 iggy I can see it now... "Why did you commit to master?" "I didn't know I wasn't supposed to" -> boned
18:38 forrest I'd say link to the discussion forum if people are interested, but don't require it. We've gotten a lot of contributions from one off users.
18:38 forrest Well, we can always revert the commit if we need to.
18:38 iggy well, I'm hoping it ends up with much more info in it than what it currently has
18:38 cromark joined #salt
18:38 iggy oh wait, this is just for us
18:38 forrest I'm fine with an extensive readme.
18:38 forrest ?
18:39 iggy just for https://github.com/orgs/saltstack-formulas/teams/contributors
18:39 iggy (and owners, but I'd like to thing they know better)
18:40 Ryan_Lane in firefox the scrollbar on the right is incredibly slow and interacts poorly. both the left and right panes scroll :(
18:40 Ryan_Lane for new docs
18:40 iggy I was trying to make that clear in the first line, I'll try harder
18:40 iggy Ryan_Lane: 14:05 <@basepi> forrest: file issues, tag @jacobhammons on the issues
18:40 baweaver joined #salt
18:41 forrest Ryan_Lane: ISSUES GOOOOOooooOOoo!
18:41 forrest iggy: Understood!
18:41 Ryan_Lane heh
18:41 Ryan_Lane there's no unified color scheme either :(
18:43 saltycharles joined #salt
18:43 Ryan_Lane ooohhh. you can target releases for the docs!
18:43 forrest Ryan_Lane: Yes there are several areas that need improvement. I'm going to talk to my old coworker who implemented the color coding for different code blocks once we've had a few days with the new version.
18:43 forrest Ryan_Lane: Yeah that is super nice
18:43 forrest Ryan_Lane: The site continually reloads if you block trackalyzer as a heads up in case you see anyone encountering that.
18:44 Ryan_Lane hm, the search goes off to google custom search
18:44 Ryan_Lane would be nice if typing into the box auto-completed to sub-doc items
18:44 iggy forrest: updated, hopefully more clear
18:44 forrest Ryan_Lane: The google custom search was an existing thing. I'm not sure if sphinx supports a search like that, would be awesome if it did.
18:44 sevalind joined #salt
18:45 Ryan_Lane well, you can index sphinx in elasticsearch or some other full text index
18:45 Ryan_Lane have that modify the dropdown, but still have the search go off to google custom search
18:46 forrest iggy: https://gist.github.com/iggy/cb800f3401a6364cd5bd#intro Does that sound too haughty?
18:47 iggy I don't know, mostly just filler text for now... I'm kind of more interested in the actual guidelines
18:47 iggy I didn't want to say "Follow these or else"
18:49 forrest sure
18:50 forrest Ryan_Lane: Could even compile the data into a static list which the search reads every time sphinx rebuilds if you had to :)
18:50 iggy actually, I do want to say that, but I figured people would complain less about wording it the other way
18:50 Ryan_Lane forrest: indeed
18:50 otter768 joined #salt
18:51 forrest iggy: Yeah I'll admit, I'm not going to follow the 2 release rule. I just don't have the time. What we SHOULD do is put the formulas through tests in test-kitchen via travis.
18:51 druonysus joined #salt
18:51 wwwBUKOLAYcom joined #salt
18:51 forrest iggy: That would actually be pretty sick to test on every PR against X number of releases.
18:51 iggy have fun setting it up
18:51 forrest iggy: I'll do some research to see if it's feasible.
18:51 iggy for all 100 something formulas
18:52 forrest Ryan_Lane: Have you set up test kitchen internally by any chance?
18:52 Ryan_Lane nope
18:52 iggy and I'm not installing vagrant anywhere, so it's going to be a lot of "test X to fix stupid vagrant test"
18:52 iggy commit messages
18:53 Ryan_Lane forrest: our container generation is basically our test kitchen :D
18:53 _Cyclone_ joined #salt
18:54 forrest iggy: heh
18:54 forrest Ryan_Lane: Yeah makes sense.
18:54 Ryan_Lane we run highstates via docker run, then do docker commits
18:54 Ryan_Lane which generate our images
18:54 mapu joined #salt
18:55 Ryan_Lane great. looks like I already found a bug in 2015.5
18:55 Ryan_Lane first run, second state
18:55 iggy welcome to my world
18:56 iggy forrest: I'm going to start reviewing all your self merges and we'll see how many issues I find with them
18:57 MatthewsFace joined #salt
18:57 forrest iggy: Self merges?
18:57 jespada_ joined #salt
18:57 iggy "No merging your own pull requests"
18:57 Ryan_Lane either that or my vagrant environment is hosed. I'm hoping it's that
18:57 forrest iggy: I don't think I've merged my own PR in a looooong time.
18:58 supersheep joined #salt
18:59 iggy you said you wouldn't follow that guideline
18:59 forrest what??
18:59 forrest no, points 3 and 4
18:59 forrest iggy: Regarding testing on the prior 2 releases
18:59 iggy "Yeah I'll admit, I'm not going to follow the 2 release rule."
19:00 iggy oh, I see what you meant
19:00 forrest iggy: Yeah, testing 2 releases locally'
19:00 ajw0100 joined #salt
19:01 forrest iggy: lol, not having time to let other people merge my own PR would be hilariously stupid on my par.
19:01 forrest *part
19:01 iggy "When possible, we should attempt a best effort to maintain backward compatibility for the previous 2 releases"
19:01 iggy best effort... better?
19:01 forrest Yeah, that's fair. I'm just saying I wouldn't test the previous 2 releases
19:01 forrest Yep *thumbs up*
19:02 forrest iggy: It's not a big deal, just trying to keep the barrier to entry as low as we possibly can
19:03 iggy again, this is meant for people that have already volunteered to have commit access, they should expect it's going to take some time, and due to the nature of what they have access to, it shouldn't come as a surprise that we don't want an all out free-for-all
19:03 mapu joined #salt
19:03 hybridpollo joined #salt
19:03 forrest iggy: Yep I understand the reasoning.
19:03 ahale_ joined #salt
19:04 iggy updated gist with some other wording changes
19:06 peno joined #salt
19:07 forrest looks pretty good
19:07 Nicop06 joined #salt
19:07 davisj how would one test for the existence of a particular sls file from jinja? i.e. {% if my.state is defined %} - foo?
19:08 WildPikachu http://paste.linuxassist.net/view/e6423f3e  <= I must be doing something wrong, my "os:Debian" section is not being used at all on Debian hosts
19:10 ek6 iggy: about backward compat..maybe just make a comment like 'pull requests should include which major release you have tested against'?
19:11 hemphill joined #salt
19:11 forrest WildPikachu: Your matching is incorrect, look at this example: http://docs.saltstack.com/en/latest/ref/states/top.html#other-ways-of-targeting-minions
19:11 solidsnack joined #salt
19:13 iggy ek6: this isn't for people submitting pull requests, this is for people who have commit access to the formulas org
19:13 iggy I would like to flesh out the contributors documenation too at some point
19:14 ek6 my bad...
19:14 iggy yeah, it wasn't super clear at first
19:15 iggy (it's still maybe not, but I'm not copy writer)
19:16 ek6 the thought of writing a formula for me is tough simply because I dont keep current alternate os's handy to test against...and im not going to submit an ubuntu/centos only formula
19:16 iggy people do it all the time
19:16 iggy then someone comes in behind them and adds other distro support
19:17 iggy I'm getting ready to submit one that will likely only work on Debian
19:17 iggy then I'll probably go back and add Ubuntu support and maybe CentOS support
19:17 ek6 oh one thing though iggy...i would suggest NOT making it a sep page..from like something contributors would read....just have it further down the page
19:18 ek6 its good for people who are going to contribute to clearly understand the 'checklist' reviewers will go through...imho
19:18 iggy I'm sure we'll work that out with the Salt guys once we have it in a format that at least most of us agree on
19:18 iggy I mostly just want something written down so people can stop playing dumb when they get told
19:19 iggy "I didn't know I wasn't supposed to commit directly to master"
19:19 ek6 'make better idiot traps all you like, i will just build larger idiots'
19:20 iggy because yeah, any project using git with more than one person has people commiting directly to master on a regular basis
19:22 forrest That's mostly an issue because of how poor github permissions are :\
19:22 forrest If they were better this wouldn't even be a problem
19:22 iggy some of it would
19:23 sevalind joined #salt
19:24 ek6 i love gerrit sometimes......i hate gerrit sometimes..for projects like formulas
19:25 iggy forrest: "Don't merge a pull request that someone else has asked for clarification on (questions on pull requests, line comments, etc)." ?
19:25 forrest iggy: It pains me that needs to be said, lol
19:25 forrest Has anyone ever done that?
19:25 iggy didn't someone do that to you recently
19:25 iggy and yes, I've had it done to me a few times
19:25 forrest Oh yeah it did happen recently, I don't know, so easy to revert it isn't usually a huge problem, but that makes sense.
19:26 iggy I'd just assume avoid the reverting and do it right the first time
19:26 iggy less commit log garbage
19:27 iggy to me that seems like more of an instance of people going through and blindly merging stuff without actually reading it
19:27 iggy like earlier today I saw something that somebody merged that clearly broke backward compatibility with no explanation as to why
19:30 joshfu joined #salt
19:31 rojem joined #salt
19:32 hoo joined #salt
19:35 Guest30965 hi
19:36 Guest30965 i found out that the log of salt orchestrate runner significantly changed between 2014.7.5 and 2015.5.0
19:36 solidsnack joined #salt
19:37 Guest30965 (updates were available yesterday on epel)
19:37 ek6 I am finding it a bit ironic that while enjoying going through the new docs that first broken link i find is how to contribute to the documentation
19:37 soren_ joined #salt
19:38 Ryan_Lane -_-
19:38 Ryan_Lane one thing I dislike about the docs change is that there's no sidebar shortcuts to functions in modules anymore
19:38 Guest30965 i find it now more difficult to debug, because the log is not anymore in the order the orchestration is done
19:39 ek6 that just means its perfect...they have declared victory and left the field
19:39 Guest30965 can this be configured the way it was before (2014.7.5)?
19:39 forrest Ryan_Lane: Just open it as an issue. Can always fix the CSS.
19:40 Ryan_Lane sigh. file.copy seems like it's been changed so that it always does a chown?
19:40 iggy Guest30965: highly doubtful
19:41 ckao joined #salt
19:41 WildPikachu forrest, I'm still trying to figure out what I am doing wrong :)
19:41 WildPikachu I'l find it sooner or later!
19:42 Ryan_Lane Comment: Copied "/srv/environment" to "/srv/service/current/tmp/environment"; Failed to change user to root; Failed to change group to root
19:42 Ryan_Lane so full of hate
19:42 WildPikachu forrest, found it!
19:42 forrest WildPikachu: Cool, you see how you had to make sure it was matching against the grain specifically?
19:42 Guest30965 iggy: so the problem is that 1. you lose your overview, because on large orchestrations i must have a view on the code
19:43 WildPikachu forrest, thankyou man :)
19:43 forrest WildPikachu: np.
19:43 iggy Guest30965: I'm not doubting that it's an issue, I just doubt there's a way to revert back to the old way (without actually reverting to the old version)
19:43 Guest30965 and 2. failed states are not in the end anymore (makes me think everything is ok)
19:44 Guest30965 iggy: Ohh, I understand
19:44 Ryan_Lane I guess I need to assume that salt is just completely broken for using the file module on nfs mounts that are squashed?
19:44 Ryan_Lane why does salt need to ensure ownership?
19:45 ajw0100 joined #salt
19:45 iggy Ryan_Lane: if they changed how something worked without documenting the change, open an issue (I did right before the release and held it up for a month)
19:46 Ryan_Lane it's doc'd
19:46 Ryan_Lane but it's idiotic
19:46 Ryan_Lane you can either preserve ownership, or you can set the user/group
19:46 Ryan_Lane of course, when you say preserve, it actively tries to chown after it copies
19:46 Ryan_Lane if you don't, then it defaults to root
19:46 iggy I thought the docs said it would get whatever owner the minion ran as
19:47 Ryan_Lane in either case, if you're copying to a squashed mount, it fails because it didn't chown to the right owner
19:47 iggy (that implies it doesn't actually try to run any chown/chgrp/etc)
19:47 Ryan_Lane --->     Comment: Copied "/srv/environment" to "/srv/service/current/tmp/environment"; Failed to change user to root; Failed to change group to root
19:47 Ryan_Lane that implies otherwise :)
19:48 ek6 you run your salt-minion with non root user?
19:48 iggy right, so someone is lying (or I'm misremembering)
19:48 Ryan_Lane I run with root
19:48 Ryan_Lane I'm actually using salt-call (masterless)
19:48 ek6 iggy: sign me up for team misremember...because i thought the same
19:48 Fiber^ joined #salt
19:49 mario_rc joined #salt
19:49 ek6 well if you have salt installed so it can run as non root user cant just execute as whomever..
19:49 mario_rc hey guys, is SaltStack talking at AWS reInvent this year?
19:50 ek6 ryan_lane: dont get me wrong...the new preserve bits i find less then appealing
19:50 _Cyclone_ joined #salt
19:51 Ryan_Lane it's just annoying that file.copy doesn't just work like cp
19:52 Ryan_Lane now I'm having to workaround this with cmd.run cp source dest
19:53 bhosmer_ joined #salt
19:56 iggy you guys maintain a lot of custom modules anyway, just copy the function from 2014.7?
19:56 saltycharles mario_rc: SaltStack is planning to sponsor it but don't have any employees speaking there.  I imagine there will be talks that include/mention the use of SaltStack, however.
20:01 thayne \me has dynamic targetting working!
20:02 * thayne fumbles with the wrong slash on the keyboard.
20:05 thayne $> salt -D target.module=parm1,parm2,parm3 test.ping
20:06 Guest30965 just compared the output of a salt-run state.orchestrate from 2014.7.5 and 2015.5.0
20:07 thayne $> salt -D jobs.filter=JID,return:path:*foo* test.ping
20:07 Guest30965 it completely changed
20:08 thayne Dynamic targeting can target a salt command based on the results of a previous command.
20:08 Guest30965 is this change somewhere documented
20:08 mario_rc saltycharles: thank you, just asking becasue I'm doing the registration today
20:08 davisj thayne: what's "-D"?
20:08 Ryan_Lane Guest30965: the yaml/json output did, or the pretty output?
20:08 thayne Dynamic target.
20:09 thayne It is a programmatic expansion for targets.
20:09 davisj Cool. Is that a 2015.5 thing or something of yours?
20:09 thayne It can be an extension module that queries a DB in real time.
20:09 Guest30965 Ryan_Lane: the pretty
20:09 Ryan_Lane Guest30965: do you parse that? :)
20:09 thayne It can also target based on the results of a *previous* salt JID.
20:10 thayne I have it working in my development tree.
20:10 thayne I'm cleaning it up to do a pull request for ``develop``.
20:10 thayne It is not in 2015.5.
20:10 davisj sweet :)
20:10 Ryan_Lane Guest30965: it's probably not documented anywhere
20:11 thayne I have some filters for searching previous JIDs, summarizing them and using the matches for targets for new commands.
20:11 Guest30965 Ryan_Lane: no i use it for finding errors during development of orchestration and for checking if orchestration went well on prod systems
20:11 thayne Potentially some of the existing targets/matchers can be rewritten as dynamic targets.
20:12 Ryan_Lane Guest30965: via code, or visually?
20:12 Ryan_Lane Guest30965: in what way did it change? is something missing?
20:12 thayne We'll have some of the SaltStack guys over at Adobe in a few weeks and I'll do some demos for them of what dynamic targets do.  Hopefully I can convince them to take the P.R..
20:13 DJSnick joined #salt
20:13 Guest30965 Ryan_Lane: how should i paste an example (i hv no github account and never used gist.github.com)
20:15 Guest30965 Ryan_Lane: shortly said: 1 missing some information 2. not ordered in order of execution 3. me feeling - it's not complete (subjective)
20:16 karlthane joined #salt
20:16 Guest30965 Ryan_Lane: even shorter: hard to debug (i use it visually - to answer your question)
20:19 joeto joined #salt
20:19 _Cyclone_ joined #salt
20:19 Guest30965 Ryan_Lane: on 2014.7.5 it looks like this (snippet): https://gist.github.com/anonymous/c81f4445c5c94f9ed0f5
20:21 Guest30965 Ryan_Lane: 2015.5.0: https://gist.github.com/anonymous/41f6b28cecde6cef5129
20:21 dingo quick poll, does anybody else use salt to create host roles through bootable images and netboot them into being?
20:22 pravka joined #salt
20:23 pravka dingo, do you mean actually creating the images via salt, or just using salt as the orchestration for programmatic netbooting?
20:23 dingo could be just netbooting or hybrid, post or pre-salting, so to speak
20:24 dingo just curious if anybody else is doing much "salt orchestrates netbooting" specificly
20:25 Guest30965 Ryan_Lane: it seems like they are ordered by the name of the run instead of the order of execution (but i must recheck)
20:26 kohonwn joined #salt
20:26 pravka I'm doing it w/ ipxe w/ salt generating the boot script for each node
20:28 Ryan_Lane it looks like it's switched to a different outpuppet
20:28 Ryan_Lane outputter
20:28 FRANK_I joined #salt
20:28 Ryan_Lane Guest30965: you should file an issue: https://github.com/saltstack/salt/issues/new
20:30 Guest30965 Ryan_Lane: i'll create an account now and do :)
20:30 joshfu joined #salt
20:32 aparsons joined #salt
20:33 dingo cool pravka, me too
20:34 pedromaltez joined #salt
20:36 bhosmer joined #salt
20:37 farcaller left #salt
20:37 ajw0100 joined #salt
20:40 kohonen joined #salt
20:41 solidsnack joined #salt
20:41 kohonen left #salt
20:43 supersheep joined #salt
20:44 qweqwe joined #salt
20:45 Guest30965 left #salt
20:45 pravka joined #salt
20:47 pravka joined #salt
20:47 qweqwe left #salt
20:51 giantlock joined #salt
20:51 otter768 joined #salt
20:52 baweaver joined #salt
20:52 hoo joined #salt
20:52 Abc2 joined #salt
20:53 Abc2 left #salt
20:54 Guest27861 Ryan_Lane: me(former Guest30965) created https://github.com/saltstack/salt/issues/23789
21:01 neogenix_ joined #salt
21:02 pravka joined #salt
21:05 bhosmer joined #salt
21:06 spookah joined #salt
21:07 cromark joined #salt
21:08 druonysus joined #salt
21:11 _Cyclone_ joined #salt
21:11 ajw0100 joined #salt
21:28 saltycharles joined #salt
21:31 baweaver joined #salt
21:32 spookah joined #salt
21:33 bhosmer joined #salt
21:34 Tyrm joined #salt
21:45 Tyrm_ joined #salt
21:50 solidsnack joined #salt
21:56 _Cyclone_ joined #salt
21:57 jhauser joined #salt
22:03 murrdoc joined #salt
22:03 jlhammons_ I'm the doc writer at SaltStack, thanks for the feedback on the new site so far, it is very helpful. As you see issues or have suggestions please put them on github and @jacobhammons.
22:04 jlhammons_ A few responses to the issues so far:
22:04 jlhammons_ arnoldB: re: search issues in opera: are you seeing issues in Opera with the sidebar search? Or the search.html page (< this page is not working right, I'll fix)
22:04 badon_ joined #salt
22:04 jlhammons_ forrest: re: trackalyzer: this should be fixed now
22:04 jlhammons_ bradthurber: re: unique searches for each branch: I've set up indexes for each branch, since we just made the content live I'm waiting to switch to make sure google has indexed.
22:04 forrest jlhammons_: I already commented and closed the issue, thanks for the quick resolution!
22:05 jlhammons_ ek6: re: contributing to docs page is broken: :facepalm: I set up a temp page here: http://docs.saltstack.com/en/latest/topics/development/conventions/contribd.html - I think our web server is redirecting that page, I'll figure it out.
22:05 jlhammons_ Ryan_lane: re: search autocomplete - that is very high on my list of things to implement
22:05 jlhammons_ Ryan_lane: re: functions in TOC - also high on my list, I think I have a good solution to this
22:05 jlhammons_ Ryan_lane: re: firefox nav issues: can you send me more details here or in an issue?
22:06 forrest jlhammons_: Oh that reminds me, I need to open an issue for code boxes not rendering properly: http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.py.html It's noticeable there for the python code block
22:06 jlhammons_ ok that is the end of the flood, if I missed any suggestions please let me know
22:06 forrest jlhammons_: I can ask my buddy who wrote the functionality initially to take a look, just not sure if he has time
22:06 iggy jlhammons_: on salt module index page, there's a double "Table of Contents Glossary"
22:07 micko joined #salt
22:07 iggy jlhammons_: and I'd suggest moving the "Social" links to the bottom on the right
22:08 jlhammons_ thanks forrest, I wonder if it is the line numbers? I'll see what I can find out from a css perspective
22:08 jlhammons_ iggy: below the TOC?
22:08 forrest jlhammons_: Yeah I'm not sure, I created this issue: https://github.com/saltstack/salt/issues/23793 and pinged my buddy on it as well.
22:08 iggy (and personally, I'd like to see the nav on the right a bit slimmer)
22:08 iggy jlhammons_: yeah
22:09 iggy it's a docs site, using for other uses (which is understandable) should not be a primary focus
22:09 bfoxwell joined #salt
22:09 tony3 joined #salt
22:09 tony3 howdy
22:10 forrest I actually like that idea as well if the details were on the bottom iggy. jlhammons_ I'm already seeing people in the IRC complaining about clicking on a social link when trying to search ;)
22:10 tony3 can someone help me with jinja macros on salt
22:10 tony3 http://paste.debian.net/177831/
22:10 jlhammons_ makes sense: re: the right nav, the width changes as you resize the window, do you happen to know what width you are at?
22:10 iggy I also think the main container (i.e. the left half) should have smaller text
22:10 tony3 I am created a macro that adds a given hostname and ip to the /etc/hosts
22:10 peno joined #salt
22:11 iggy probably ~1000x1000
22:11 tony3 while I can properly import it, (so I think, states come back green) I am not able... well do not know the proper syntax to execute it
22:11 forrest jlhammons_: regarding the python render thing, yeah I agree it's probably the line numbers, they're outside the object now which is weird, wouldn't be surprised if that was the issue.
22:11 tony3 the syntax I have on the link above shows how I am calling the macro
22:11 tony3 on the receiving end, I just get a print out of the macro name and its arguments
22:12 tony3 nothings happens to the /etc/hosts
22:12 tony3 any ideas?
22:12 iggy there's nothing on the nav that's even half the width of that div (Except the social links)
22:13 iggy tony3: {{ ... }}
22:13 iggy I think most people avoid macros
22:14 iggy once you start getting into that kind of stuff, you may as well just start using the pyrenderer
22:14 jlhammons_ iggy: the width was set up for the nested topics like this one: http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.py.html - where there are a bunch of children
22:14 tony3 ah
22:14 bhosmer joined #salt
22:15 tony3 yea thanks, the {{ prints out the function on the file
22:16 iggy jlhammons_: okay, then you probably want some sort of cues that indicate that might happen (and when stuff is expanded like that, it's a little clear what level you are at, maybe something that shows where you are)
22:16 iggy jlhammons_: either way, it's an improvement, I'm just giving you some things that jump out at me... I'm certainly not the typical person that's going to be looking at the docs
22:17 jlhammons_ thanks iggy, these are all great ideas
22:17 baweaver joined #salt
22:18 iggy jlhammons_: do you want bugs too? (so you can show your boss you are actually doing stuff
22:18 prwilson joined #salt
22:19 jlhammons_ haha, bugs are great, mostly so I can keep track of them and they don't get lost in the sea of one-line text files in my home directory
22:19 iggy I started to do some earlier, but got sidetracked
22:20 _Cyclone_ joined #salt
22:21 forrest jlhammons_: Are you not pushing the CSS changes to a repo on github? I looked at your profile and didn't see many commits, wasn't sure where you were working.
22:22 tony3 thanks guys
22:22 keimlink joined #salt
22:23 jlhammons_ forrest: for now I'm pushing the sphinx theme I'm using to my fork at https://github.com/jacobhammons/salt - I'll merge it to the main project soon (I'm still getting the hang of sphinx)
22:24 forrest jlhammons_: Ahh that's why I didn't notice it, it's one giant commit, okay makes sense
22:24 forrest I was going 'man how are all these changes in here with so few commits!' :)
22:25 Abi_ joined #salt
22:26 JayFK joined #salt
22:26 jlhammons_ :)
22:26 Abi_ :)
22:30 jlhammons_ ek6: this page works again: http://docs.saltstack.com/en/latest/topics/development/conventions/documentation.html - I had an over-eager redirect rule set up on the webhost
22:35 peters-tx Just upgraded to 2015.5.0 ....  Everything seems.. FASTER
22:35 peters-tx O_o
22:36 gwmngilfen joined #salt
22:48 JDiPierro joined #salt
22:48 _Cyclone_ joined #salt
22:49 solidsnack joined #salt
22:50 ajw0100 joined #salt
22:52 otter768 joined #salt
22:55 diegows joined #salt
22:56 solidsnack joined #salt
22:57 sygibson joined #salt
22:58 sygibson question on docs.saltstack.com - it seems we've "lost" the view in to a specific modules functions as a menu on the right side of the browser, with the new site upgrade ...
22:58 sygibson is there an option I'm not seeing somewhere, to be able to see a menu of the modules functions ... like before the upgrade??
22:59 jlhammons_ sygibson: we are working on adding that back after the toc change
22:59 forrest sygibson: Yep there are some definite things missing! I believe jlhammons_ is working on some of that stuff, so this question might be better directed at him, otherwise please create an issue so we can see about getting it prioritized to get that sort of thing back in :)
22:59 sygibson cool beans!
22:59 forrest thanks jlhammons_!
23:00 jlhammons_ you are welcome!
23:00 sygibson good to see some work being done to overhaul the docs - nice to know it's getting some focus - the info is very useful there already - and easier navigation and exposing more of the information seamlessly in a session is a good thing - I like being able to see all of the modules available, just was despondent to have lost the functions of a specific module !!
23:02 jlhammons_ sygibson: so agree, especially for the modules that have many many functions
23:02 KyleG joined #salt
23:02 KyleG joined #salt
23:03 Nicop06 joined #salt
23:07 [arx] joined #salt
23:07 [arx] joined #salt
23:10 Brew joined #salt
23:10 keimlink joined #salt
23:14 paha joined #salt
23:14 jlhammons_ iggy: here is a screencap of the site with the social icons moved to the bottom right: http://docs.saltstack.com/en/stage/social-icons-moved.png
23:15 iggy looks better to me, but I know it's all kind of personal preference
23:15 iggy I'm certainly not a usability expert... just some nerd who read a HIG book once... like 10 years aog
23:16 forrest jlhammons_: I like that better!
23:16 forrest jlhammons_: I'm just imagining someone clicking on a social button instead of the search
23:16 forrest but with them there it's a non issue
23:16 jlhammons_ I like it too
23:16 iggy Unless they watn "Salt Best Practices" ;)
23:18 Nazca joined #salt
23:18 forrest oh that's a good point iggy, should the stuff focused around beginners be up at the top?
23:18 forrest jlhammons_: How was that list on the right generated?
23:19 iggy I imagine the same way it was before
23:19 iggy i.e. something in docs/
23:20 jlhammons_ there is a toctree inserted there in the sphinx theme
23:20 forrest jlhammons_: But did sphinx generate the toctree?
23:21 forrest It's not alphabetically sorted.
23:21 jlhammons_ yes
23:21 forrest hmm
23:21 jlhammons_ oh, it is generated from the contents.rst file
23:23 iggy doc/contents.rst
23:23 forrest iggy: Yeah
23:23 iggy sorry :*(
23:23 iggy had to wait for the git pull to finish
23:23 forrest jlhammons_, iggy: Opinions on reordering that? Understanding YAML would probably be good to have further up
23:24 forrest iggy: I wasn't giving you a hard time, just acknowledging
23:24 iggy I was under the impression that actual content changes were still to come
23:24 iggy this was just reskinning what was already there
23:25 iggy (and making 2015.5 the default)
23:25 forrest iggy: I'm not sure about that, when was that discussed/noted?
23:26 jlhammons_ yes, we are working on the content changes. Reordering the TOC will be a part of that
23:26 forrest if content changes are coming later, we should hold off and just work on layout until that is looking good, then redo the content when we can exclusively focus on that.
23:26 iggy at least at the conf, someone said something about revamping all the intro stuff
23:26 forrest jlhammons_: Awesome.
23:26 iggy not crazy!
23:27 jlhammons_ forrest: agree
23:27 iggy jlhammons_: can you adjust the scrollbar size? I keep trying to grab and scroll the TOC and end up selecting the whole page on accident
23:28 forrest iggy: you mean more dead space on the right hand side correct?
23:29 iggy like the little slider that appears when you mouse over the TOC
23:29 jlhammons_ iggy: re: intro stuff - we are working on new intro content here: http://docs.saltstack.com/en/getstarted/ - I need to review the tutorials to see where they fit
23:29 iggy yeah, dead space on the right might help that too (it dissappears pretty quick)
23:29 forrest Also this is a ridiculous nitpick (makes me feel like some kind of guy doing the 'move this two pixels' thing), but jlhammons_ does the blue on the headers feel too light to you? I notice when I scroll fast I have a harder time focusing on the headers than the black text, might just be me.
23:29 forrest jlhammons_: That graphic/structure is great.
23:30 jlhammons_ forrest: I think I'm used to the light blue headers now
23:30 forrest okay cool, might just take me a bit of time to get used to then.
23:31 iggy if remote execution doesn't make an appearance in "Orchestration", I'd say add a separate section for it
23:31 cwyse joined #salt
23:31 jlhammons_ our graphics guy is going to sit down with me next week to look over colors and whatnot
23:31 iggy it's a pretty fundamental selling point of salt
23:31 forrest jlhammons_: Sounds great! Almost time to kick out since it's Friday too
23:32 baweaver joined #salt
23:32 jlhammons_ thanks for the feedback forrest
23:32 iggy well, was I guess... guess I haven't looked at the saltstack site in a while
23:32 jlhammons_ iggy: I cover remote execution in the fundamentals section, I do think it should be in the graphic somewhere
23:32 forrest jlhammons_: Yeah of course, I figure it is probably one of those annoying things when you've been designing it for a while though, so no need to be polite if I'm full of crap.
23:33 iggy ^
23:33 iggy we definitely aren't polite to each other, no need to be either
23:33 forrest lol
23:34 jlhammons_ sometimes I feel like I've looked at the design so long that I'd like it no matter what, so other perspectives are always refreshing
23:34 jlhammons_ iggy: haha
23:34 forrest jlhammons_: Yeah, regardless though. I've known too many designer friends who just 'take it' and then push it back, so better to say something so we won't keep commenting on it and will leave you be.
23:42 druonysus joined #salt
23:42 nthcode joined #salt
23:43 mpanetta joined #salt
23:46 baweaver joined #salt
23:47 forrest I'm outta here, have a good weekend
23:48 jlhammons_ you too, thanks again
23:50 _Cyclone_ joined #salt
23:51 baweaver joined #salt
23:55 kusams joined #salt
23:56 shadowsun joined #salt
23:57 shadowsun A bunch (50+) minions just updated to the new salt-minion 2015.5.0, and it looks like jinja templating has been changed. The following is not being replaced
23:57 shadowsun # {% include 'common/template.py' %}
23:58 shadowsun How can I fix this without uncommenting it, because jinja template isn't valid code, and I need this to be runnable one way locally and another via salt

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