Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-11-14

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

All times shown according to UTC.

Time Nick Message
00:00 rattmuff joined #salt
00:00 whiteinge haha
00:00 TheoSLC When I update a grain value with grains.setval I have to restart the salt-minion service before I see the change in valide.  Does anybody know if this was fixed in 2014.7?
00:00 iggy you shouldn't
00:00 murrdoc fixeded
00:01 TheoSLC murrdoc: was that an affirmative for my question?
00:02 murrdoc from my tests with 2014…rc7
00:02 murrdoc it looked* fixed
00:02 TheoSLC murrdoc: thanks
00:07 bhosmer joined #salt
00:08 bigpup joined #salt
00:10 MugginsM joined #salt
00:11 TTimo joined #salt
00:11 rattmuff joined #salt
00:17 seanz joined #salt
00:17 seanz whiteinge: Greetings. I was reminded of you today because we're hiring a developer who has some of your mannerisms. :)
00:17 ifmw joined #salt
00:17 genediazjr joined #salt
00:17 seanz whiteinge: I hope you're doing well. If you're going to SCaLE this coming year, I'm looking forward to saying hi.
00:17 whiteinge what? hey, those are _mine_.
00:18 whiteinge seanz: i'll be there almost definitely
00:18 al joined #salt
00:21 seanz whiteinge: Most excellent. I will be looking forward to it.
00:23 rattmuff joined #salt
00:23 juanlittledevil joined #salt
00:24 bigpup joined #salt
00:24 robawt seanz: you'll be at scale?
00:24 iggy I knew it had to be the same person that wrote the formulas for zookeeper and graphite (and now I've seen the jmxtrans one)
00:25 johtso joined #salt
00:26 seanz robawt: Yah, we're going to SCaLE and PyCon this year.
00:31 murrdoc yes on scale
00:31 seanz robawt: Shall you also at SCaLE be?
00:35 rattmuff joined #salt
00:36 mbrgm joined #salt
00:38 Pixionus Is there a simple way to return just the first line of a single text file from all minions?
00:38 younqcass_ joined #salt
00:38 murrdoc cmd.run
00:40 bigpup joined #salt
00:40 pacopablo joined #salt
00:42 TTimo joined #salt
00:43 perfectsine joined #salt
00:46 techdragon joined #salt
00:46 rattmuff joined #salt
00:48 elfixit joined #salt
00:53 jotacepea joined #salt
00:54 yetAnotherZero joined #salt
00:55 jotacepea left #salt
00:56 bigpup joined #salt
00:58 rattmuff joined #salt
00:59 glyf joined #salt
01:04 genediazjr joined #salt
01:05 Eugene Sure, something like `salt '*' cmd.run "head -n1 /path/to/file"`
01:06 Pixionus thanks murrdoc.  Yeah stupid question.  Used cmd.run grep.
01:08 thayne joined #salt
01:09 Rucknar joined #salt
01:10 rattmuff joined #salt
01:12 bigpup joined #salt
01:12 jprostko joined #salt
01:16 techdragon joined #salt
01:16 genediazjr joined #salt
01:19 Ned_ joined #salt
01:22 rattmuff joined #salt
01:23 davidone joined #salt
01:26 skyler Hopefully we can get the salt-formulas mailing list going. I want to start getting people's opinions on good style.
01:26 skyler I am starting with style for accessing pillars, for those interested: https://groups.google.com/forum/#!topic/salt-formulas/A0tfeuE0lP8
01:26 jcockhren is there already a mailing-list?
01:28 jcockhren while I love email and mailing lists as much as the next guy, is there any chance we can create a saltstack/rfc repo? whiteinge, etc
01:28 jcockhren the formulas/code is on GH so maybe we can have the discussions where the code sits?
01:28 bigpup joined #salt
01:29 jcockhren just thought I'll throw that out there
01:34 rattmuff joined #salt
01:34 bhosmer joined #salt
01:43 Pixionus Not really the biggest fan of mailing lists since I get too much mail already, but an indexed central repo (probably based around GH since there's a bunch there) would be nice.  I've been tossing around the idea of starting up a wiki for Salt to help index things like this, provide access to tools, and help n00bs get started faster.
01:43 genediazjr joined #salt
01:44 Pixionus Not sure I'm the best person to do this though, since I'm little more than a noob myself
01:45 bigpup joined #salt
01:46 rattmuff joined #salt
01:46 Pixionus Mailing list wouldn't be a bad idea if it doesn't already exist.  Just saying that an indexed resource for it all to filter into would be nice
01:48 Pixionus (jcockhren: looks like it doesn't since David made it an official mailing list on the 12th)
01:48 jcockhren nice
01:49 _JZ_ joined #salt
01:49 jcockhren basepi: ^^
01:49 jcockhren basepi: if you can see the scrollback
01:52 micah_chatt joined #salt
01:54 whiteinge jcockhren: sure, creating a repo is easy enough. What kinds of things do you think will go in it?
01:56 jcockhren whiteinge: a place to write out community discussed best practices for forumals, custom modules/states, reactors, usage, etc etc
01:56 jcockhren topics discussed as issues, improvements made via PRs
01:56 jcockhren as an alternative option to a mailing-list
01:57 rattmuff joined #salt
01:57 jcockhren not even sure if it's the 'best' way, but wanted to at least mention it
01:59 genediazjr joined #salt
02:00 jcockhren at times recommendations on practices from those who write the code appears here. other times, folks appear in here when their info about their uses. it's always informing and would be great to have to have the discussion in a searchable way
02:00 logandg joined #salt
02:00 jcockhren with labels as tags or something
02:01 bigpup joined #salt
02:01 jcockhren reactors: green, custom modules: yellow
02:01 logandg has anyone had any luck using salt-ssh with vyos/vyatta?
02:01 whiteinge Interesting idea.
02:02 jcockhren every other day someone asks: "What are the best practices for..."
02:02 jcockhren :D
02:03 whiteinge Kind of a "snippets" style collection.
02:04 glyf joined #salt
02:05 Nexpro1 joined #salt
02:08 rattmuff joined #salt
02:12 malinoff joined #salt
02:17 bigpup joined #salt
02:19 rattmuff joined #salt
02:24 genediazjr joined #salt
02:25 mbrgm joined #salt
02:26 Pixionus whiteinge: among other things.  That's why I was thinking a wiki.  Then we could have community generated primers and content to help get started from zero to hero.
02:27 Pixionus but if we could paste solutions somewhere and tag them that would be good in general.
02:30 rattmuff joined #salt
02:32 whiteinge Pixionus: gotcha. Do you think the need for such a place would exist if the docs were 1) better organized, and 2) more opinionated?
02:33 whiteinge (At least in some places.)
02:33 bigpup joined #salt
02:33 mosen joined #salt
02:34 Pixionus not sure what you mean by opinionated, but I think something like a moderated wiki might allow users to provide the documentation they think is needed (since this can be different person to person) without increasing workload at Salt too much.
02:34 Pixionus It would be important that the Main docs be better but they wouldn't have to be as extensive.
02:34 Pixionus when folks request documentation that you just don't have time to write, maybe someone out here does.
02:36 _JZ_ hi guys, sorry to jump in here, but i'd like to second the idea of having a best practices guideline for using saltstack 'correctly'.
02:36 druonysuse joined #salt
02:36 _JZ_ im in the process of redoing our current puppet setup to use saltstack
02:36 Ned_ on the subject of documentation
02:36 Ned_ http://docs.saltstack.com/en/latest/ref/states/all/index.html
02:36 Ned_ "latest" seems to include stuff in 2014.7
02:36 Ned_ is there a version of the docs that are just for the current stable ?
02:37 _JZ_ i know one of the biggest hurdles i'll face is getting people to understand the saltstack 'way' of doing things
02:37 _JZ_ rather than falling back into the spaghetti modules that has become puppet
02:38 Ned_ _JZ_: is there anything that describes the saltstack "way" at the moment ?
02:38 Ned_ I mean, I'd read that if it existed :D
02:38 _JZ_ not that i've found
02:39 Pixionus Whether it's done in a wiki or not, I think it would be good to get some sort of best practices doc from some of the central members of the community.  We are setting up our system atm and basically just using states for everything.  I'm pretty sure that we aren't doing it right.
02:39 _JZ_ i've been reading docs.saltstack.com and the saltstack/salt-formulas git repo to kind of get a feel for what others are doing
02:39 _JZ_ but nothing concrete
02:39 _JZ_ i know that my mgmt are gonna be asking for a sdlc for this as well
02:39 Leonw joined #salt
02:39 Ned_ so err, is there a version of the reference documentation just for current stable ?
02:39 _JZ_ which will be 'fun'
02:41 rattmuff joined #salt
02:45 whiteinge Ned_: the docs are currently built from develop. We're switching them to default to the current release instead.
02:45 whiteinge (On my phone, don't have the Github issue handy.)
02:45 Ned_ ok
02:45 Ned_ whiteinge: slightly unrelated, but is there an eta on 2014.7 ?
02:45 Ned_ I've noticed a few things that I'd quite like from it :p
02:45 whiteinge It's tagged, we're just waiting on packages to get built.
02:46 Ned_ whiteinge: oh, that sounds promising ...
02:46 whiteinge :-)
02:47 nitti joined #salt
02:49 bigpup joined #salt
02:51 whiteinge _JZ_: we've got a series of blog posts planned that will discuss directory layout, customization,  and some best practices. I expect work on those to begin after thanksgiving. I think that will address some of what you're asking for
02:51 genediazjr joined #salt
02:52 rattmuff joined #salt
02:54 druonysus joined #salt
02:54 druonysus joined #salt
02:55 yetAnotherZero joined #salt
02:55 catpigger joined #salt
02:57 whiteinge Pixionus: makes sense. The docs are definitely fair game for community contributions. But I understand there's a "weight" barrier to entry there that a wiki would not have
02:58 Rucknar joined #salt
03:00 Pixionus Ive been doing documentation for 2 years now and at some point someone at a university through up a wiki that quickly took away a lot from my work load while improving on what I had already written.
03:00 Pixionus threw up a...
03:00 Pixionus It was awesome!
03:00 Pixionus :P
03:01 tmh1999 joined #salt
03:03 whiteinge what's the current state-of-the-art for wikis these days? is it still mediawiki?
03:03 rattmuff joined #salt
03:04 alexr joined #salt
03:05 Nilium joined #salt
03:06 bigpup joined #salt
03:11 Mso150 joined #salt
03:12 jcockhren not confluence
03:12 jcockhren </fin>
03:12 whiteinge haha
03:12 Pixionus Not sure.
03:13 Pixionus Some guys at my company are pushing confluence but haven't had a chance to look at it.
03:13 marv__ joined #salt
03:14 malinoff Confluence is a monstrous thing, it is slow, big and enterprise
03:14 Pixionus I'd be interested inn hearing about what you hate jcockhren?  (off topic so I would be happy to receive it as a msg if people here mind stuff like that)
03:15 rattmuff joined #salt
03:15 jcockhren Pixionus: it hard to find documents/posts that are nested similar topics
03:15 jcockhren namely, it seems like the search isn't full text
03:16 jcockhren naturally, it allows for a nested navigation of documents. But that doesn't mean you'll be able to find your doc if you use the search capability
03:17 jcockhren search is usually the primary experience people expect nowadays. when that's not so awesome...
03:17 Pixionus I've been using one note for all my internal documentation and it seems to work great for my needs.  Others were looking for a more centralized and I guess formal thing, so they brought in confluence
03:17 Pixionus nested docs, search works fine, and formatting/images/whatever to a limited degree.  Even encryptable pages.
03:18 jcockhren Also, I like markup type languages rather than built-in editors
03:18 Pixionus Same here.
03:18 Pixionus but I hate the limitations of some of them that try to "simplify" markup
03:18 jcockhren a good example happened to me 2 days ago
03:19 Pixionus they just fail at things like ordered lists with nested shit
03:19 jcockhren here I go, using their key bindings. in fact, "!" allows for a quick dialog to appear to embed images/files in your doc
03:19 jcockhren should be a shortcut from using the mouse to select the menu item
03:20 Pixionus ha
03:20 jcockhren instead, it sends you to different dialog where it isn't even possible to select images previously uploaded (even though you can normally)
03:20 Pixionus oh wow.
03:20 jcockhren so, it's painful when trying to upload all the images you need before you write, then trying to reference them later
03:21 jcockhren 2. meh on the nested comment threads
03:21 Pixionus with onenote, I just drag and drop to get that done.
03:22 jcockhren I like the idea of discussion around a doc I guess. but, then the nested comments allow for too much chance for diverence AND it's hard to follow
03:22 Pixionus I work on a linux machine and a windows one and use synergy to keep my desk uncluttered.   Since OneNote is Windows and since Synergy allows for copy and paste between computers, it generally does everything I need and have learned to work in its bounds
03:23 jcockhren 3. they allow optional 'commit/change' messages.
03:23 jcockhren optional is silly to me
03:23 jcockhren either require it or not. easpecially with a comment section
03:23 Pixionus Oh, and it looks like tikiwiki is the up and coming open source wiki.  It's written in PHP but it has a lot built in that you would need extensions for in MediaWiki
03:23 jcockhren 4. using confluence on their servers is SLOOOOW
03:23 jcockhren they = atlassian
03:23 srage_ joined #salt
03:24 Pixionus DokuWiki is made for documentation specifically and seems to have a nice following.
03:24 jcockhren I took some time trying out various ones
03:24 Pixionus Oh shit, sorry jcockhren, somehow missed that you were in the middle of a list...
03:24 Pixionus teach me to read better...
03:25 jcockhren nah you good. I have a long list
03:25 Pixionus didn't mean to cut you off
03:25 jcockhren nah. it's fine. I've listed enough.
03:26 jcockhren but the main issue is that if you have enough documents, you totally miss things when you search
03:26 rattmuff joined #salt
03:26 jcockhren enough being a few hundred
03:26 Pixionus Jesus...
03:27 Pixionus Does it allow for tagging or anything to help mitigate that weakness?
03:28 jcockhren one of the reasons I suggested git based is b/c it would require yet another system for the saltstack team to manage 100% of the time
03:29 Pixionus But Git isn't the most accessible since not everyone is on board.
03:29 jcockhren on GH the community already has access and has a workflow to make comments and make changes
03:29 Pixionus I only just started in on GH and am still learning my way around.  A wiki is much more usable and searchable for the noobs
03:30 marv__ joined #salt
03:30 jcockhren while that might be true, putting it on github is the grestest chance to empower the community b/c it's not fair to have the saltstack team have to maintain yet another thing
03:31 jcockhren eveything else is on github, includes the site that contains the docs/tutorials
03:31 Pixionus Would there be a lot of maintenance in a wiki?  I mean they set up and maintain the server but that shouldn't be a problem for them.  The rest of the maintenance would be crowd sourced for the most part.
03:32 jcockhren how do you manage access for content modification?
03:32 Pixionus Can you link me to a good place to start digging into the GH resources?
03:32 Pixionus leave it open like wikipedia
03:32 Pixionus not many folks out there are going to trash the Salt wiki since the majority that would even know about it are sysadmins with a vested interest
03:33 Pixionus Lots of us are likely Linux nuts and we tend to get behind community projects as if they were of huge importance, so finding some community moderators wouldn't be too much of a problem.
03:34 jcockhren my point is that isn't already done in other forms on GH for other parts of the stack.
03:34 jcockhren s/isn't/IS
03:34 jcockhren haha
03:34 Pixionus I give it a few months before the community starts making bots to patrol the wiki and make remove vandalism if it occurs... if someone doesn't just use some of the ones that are already made
03:34 jcockhren Pixionus: https://help.github.com/categories/collaborating/
03:34 Eugene We have a wiki now?
03:34 jcockhren heh
03:35 Pixionus Not officially and havent found any unofficial ones yet
03:35 jcockhren Eugene: we were discussion options for have a central location for crowd sourced 'best practices'/rfcs and such
03:35 Pixionus Nah, meant a link to the salt stuff you were talking about.  Is there a good central place to start in on it?
03:35 jcockhren whew... typos
03:35 Eugene The issue I have with wikis is that they're a whole different editing system from basically everything else
03:35 Pixionus (salt stuff on GH)
03:36 Eugene I'd much prefer to see a salt-docs repo with an easy-to-join Github docs group, say.
03:36 jcockhren https://github.com/saltstack/salt/tree/develop/doc
03:36 Eugene But that's just me hating documentation
03:36 jcockhren Pixionus: ^
03:36 Pixionus Eugene: while that's true, they are supper simple and some use more common markups
03:36 Pixionus Thanks jcockhren
03:37 Eugene Yeah, but then it's "in the wiki" :-|
03:37 jcockhren Pixionus: https://github.com/saltstack/salt/tree/develop/doc
03:37 Eugene Similar to being "in the database"
03:37 rattmuff joined #salt
03:39 jcockhren Pixionus: on GH discussions around the changes/submission can happen
03:39 Pixionus a wiki is much more user friendly and linkable.   To be fair, I don't think it would be the primary location of all the documentation.  Salt would maintain their docs as they already do but focus more on what the want to focus on/see need for and jack what's good from the wiki.  Users would still be able to use the GH and host discussions and changes, etc.
03:39 jcockhren on a wiki, it's gated by 'moderators' and discussion _may_ happen offline
03:40 Pixionus A wiki isn't a good place for discussions.  I agree.  GH would be better for that.
03:40 jcockhren I think that's the key.
03:40 Pixionus The wiki would be more a place where we can put up good introductory and basic documentation like best practices.
03:41 Pixionus It can and should freely link to GH
03:41 jcockhren Gh has that built in
03:41 Pixionus but again, I see little to no reason for everything to be actively moderated on a Salt wiki
03:42 jcockhren yeah. the moderationed itself would be crowd sourced not just the contributions
03:42 Pixionus I think GH and a wiki serve two different uses.
03:42 jcockhren s/moderationed/moderation/
03:42 jcockhren and and
03:42 jcockhren saltstack team already has jenkins ;)
03:43 Pixionus jenkins?
03:43 Pixionus still too new to know the community very well
03:44 robinsmidsrod joined #salt
03:44 jcockhren jenkins can be hooked up the check basic stylistic things on PR submission. though that'll too much honestly
03:44 TTimo joined #salt
03:45 jcockhren fwiw, GH repos have wikis
03:45 jcockhren more like "wikis"
03:45 jcockhren lol
03:45 Pixionus .. that I did not know.
03:46 marv__ joined #salt
03:46 jcockhren yep. A lot of projects use them to have guides/docs on GH with everything else but not in the main repo
03:46 Pixionus but if that works, then it would take care of both wouldn't it
03:46 jcockhren yep
03:46 mortis_ joined #salt
03:47 nhubbard joined #salt
03:47 Tahm joined #salt
03:47 dwfreed joined #salt
03:47 mirko joined #salt
03:47 dh__ joined #salt
03:48 Pixionus I have a hell of a time navigating docs on the GH right now.  I'm sure Ill get better once I figure out hwo you guys go about it.
03:48 rattmuff joined #salt
03:48 schristensen joined #salt
03:50 anteaya joined #salt
03:50 otter768 joined #salt
03:50 Pixionus A wiki has no learning curve and I think that's probably my strongest reason for liking it as an option.  I'm in the middle of learning so many new systems as well as everything about electricity (I now work for a smart grid company) and our products that an extra learning curve is just getting to be too many
03:51 jcockhren I believe in you Pixionus.
03:51 Pixionus pfft.
03:51 jcockhren You'll absorb it all
03:51 jcockhren :D
03:51 jcockhren ride the wave, which has curves
03:51 Pixionus I just don't think I have enough time in the day.  (Been handling it alright though, if I don't say so my self)
03:52 Pixionus I wouldn't pick jobs like this if I didn't need that sort of challenge.
03:52 hunmonk_ joined #salt
03:54 jcockhren Pixionus: get up the speed with git and github
03:54 Pixionus So whiteinge: the current wiki contenders are: GH wiki, mediawiki, tikiwiki, or doku wiki
03:54 TTimo joined #salt
03:54 hunmonk_ running into a weird jinja2 error with formatting during loops. any white space/line endings after the end of the loop tag are being clobbered, leading to formatting errors in my SLS. example here: http://pastebin.com/xAGcz0xc — anybody have a thought on what i can do differently?
03:54 Pixionus jcockhren: Yeah, I really do need to.
03:55 jcockhren I bet once you do, you too will also think a repo would be a best way to do
03:55 jcockhren s/to do/to go/
03:55 jcockhren RFC is the key term here
03:56 whiteinge Pixionus: ty for the recap
03:56 marv__ joined #salt
03:56 jcockhren non-wiki things: markdown in a repo
03:56 jcockhren that's my vote
03:57 whiteinge jcockhren: what kinds of things are you thinking would go in issues (that could then accept labels), vs things that go in the repo itself, vs things that go in the repo's wiki?
03:58 whiteinge hunmonk_: the dash at the start/end of each jinja tag will kill any leading/trailing whitespace *including* newlines
03:58 whiteinge hunmonk_: your best bet is to just leave the dash off entirely
03:59 whiteinge jinja will generate a few empty lines but salt will ignore them.
03:59 jcockhren whiteinge: issues would be discussion topics that may or may not span multiple markdown documents
03:59 rattmuff joined #salt
03:59 Pixionus jcockhren: I assume the GH wiki would be findable with a simple google search if it was adopted.. (I still feel that a wiki and the repo serve different users/uses).  If they behave like most other wiki's, I think we should look into improving/starting/publicizing a GH wiki
04:00 whiteinge jcockhren:  markdown documents being actual files in the repo?
04:00 jcockhren inside the repo would be the full write-ups of the solutions/ideas discussed.
04:00 jcockhren the readme.md containing the navigation
04:01 Ryan_Lane joined #salt
04:01 jcockhren I'm not really fond of the GH wiki idea since permissions aren't clear PRs aren't possible
04:02 Pixionus whiteinge: strengths of the different wiki's that I would look at: GH wiki because GH seems to have a following here already, but only if it is as strong as it would need to be; Tiki Wiki because it is pretty preconfigured and has the most in one package while still being pretty light; and Doku Wiki because it is specifically designed for this purpose and should cater to the needs.
04:03 Pixionus jcockhren: Pull Requests?
04:03 jcockhren Pixionus: have a github account? if not. get one, it's free
04:04 * Pixionus has one
04:04 jcockhren aah. yeah. PR = Pull Requests
04:04 jcockhren that will give the grammar pirates to step in an ensure ideas are expressed clearly
04:05 jcockhren technical writing is nothing to scoff at
04:05 whiteinge "grammar pirates" hahaha
04:06 jcockhren the key for me is that it's all happening in the open
04:07 jcockhren they way it benefits to largest pool of people
04:07 jcockhren and Gh already hosts it so no additional cost to the saltstack co.
04:08 jcockhren I'm sure most people here already know of people that'll actually propose something for inclusion and contribute
04:08 jcockhren so... as the codebase changes and it effects thr workflows of folks. tweaks to how to do X can be made to match
04:09 jcockhren for example, salt-cloud using AWS spot instances
04:09 jcockhren in-depth
04:09 Pixionus Not the biggest fan of the GH file viewer layout.. Like many sites, it doesn't make use of different size windows and the like.  Also, these aren't so easy to find when searched for on Google.. At least my searches yield few results for documentation hosted on GH
04:09 jcockhren search the repo itself
04:09 Pixionus One of the first things I searched when I found the documents to be lacking was "Salt wiki"
04:10 jcockhren GH search is pretty powerful
04:10 jcockhren it's backed by elasticsearch. ;)
04:10 Pixionus that's only good if you already know it exists and know your way around (which is simple now that I use it).  I think docs should be readably findable by the whole of the internets
04:10 rattmuff joined #salt
04:11 jcockhren people repos are public to the internet
04:12 Pixionus ok, starting to train google.  I think that's part of it (google doesn't know I would be interested in GH... FU personalized bs searches)
04:12 jcockhren Pixionus: true
04:13 Pixionus Really don't like markup's limitations, but I guess we can't all just use full bracketed syntax's
04:14 mbrgm joined #salt
04:14 Pixionus I do feel that wiki markup is MUCH stronger than what I'm seeing so far on the github rst stuff.
04:14 jcockhren all we need is internal links, basic formatting, images and of course emojis
04:14 yomilk joined #salt
04:14 jcockhren lol
04:15 Pixionus (I also hate how people feel the need to format the new lines in text in docs they write on line but that's a personal pet peeve)
04:15 Pixionus lol emojis
04:16 Pixionus We need ordered lists, text size and strength, and even font or code boxes
04:16 jcockhren the great part about this. IF we try just plain ol markdown and it turns out not to work, there are wiki systems out there that understand markdown
04:16 jcockhren however, it'll be much more painful to go in the opposite direction with the wiki markups
04:17 Pixionus It won't turn out not to work, it will become a system some hate and some couldn't care less about
04:17 Pixionus I thought everyone knew wiki markup  ;)
04:17 jcockhren unless the _some_ hate github and all the tools it provides, I find that hardely a thing
04:18 jcockhren we are just talking about text in a collection of files
04:18 jcockhren K.I.S
04:18 jcockhren in the full pragmatic glory
04:19 whiteinge i like how forgiving markdown is compared to rST. i really wish they'd steal extensibility (roles & directives). rST is so freakin' powerful with those.
04:19 whiteinge ("they" == markdown)
04:19 jcockhren whiteinge: that's certainly true
04:20 tligda joined #salt
04:20 jcockhren I suggest markdown since it renders most correctly on GH so there's no need to "build it" and host it somewhere
04:20 jcockhren but rST could be just a static site in s3
04:21 jcockhren or cloudfiles if that's you guy's speed. ;)
04:21 whiteinge no i'm with you on that, rST is _awesome_ but it's too high a barrier to entry for this kind of a project
04:21 Pixionus Am I missing something?  It looks like some of these files are rst..
04:22 jcockhren they are
04:22 whiteinge all the salt docs are built with rST
04:22 rattmuff joined #salt
04:22 Pixionus and it doesn't render correctly in GH?
04:22 jcockhren the rendering is off quite a bit thought
04:22 Pixionus ah
04:22 whiteinge i'm just lamenting the pros of one with the cons of the other. :-)
04:22 whiteinge github's rST rendering doesn't take any of the sphinx extensions into account
04:23 TyrfingMjolnir joined #salt
04:23 Pixionus See, GH obviously sux!  (I kid, I kid)
04:23 jcockhren heh
04:24 Pixionus yeah, I see how it could be a good usable resource.
04:24 whiteinge haha
04:24 Pixionus I still feel a wiki is a stronger tool for documentation, but I could deal with GH limitations if I needed though I doubt I would use many of its strengths for the most part
04:25 Pixionus In the end, it's what wiki is designed to do, and GH can just be made to do it.  But I think the most important factor here is the larger community
04:25 Pixionus How much of a following would each solution have?
04:25 whiteinge i'm gonna conk out. very interesting discussion though. thanks for starting it. i'll keep thinking on it and chat with y'alls more later.
04:26 jcockhren whiteinge: cool cool
04:26 jcockhren thanks!
04:26 Pixionus Maybe have a vote posted on the main Salt site for a while to find out what everyone would use the most?
04:26 Pixionus aight.  Night whiteinge
04:26 jcockhren Pixionus: a survey is a good idea
04:26 Pixionus Thanks for discussing!
04:27 Pixionus jcockhren: I think in the end it's vital.  Otherwise we could end up with a GH no one ever uses or a wiki that everyone hates.
04:27 Pixionus or something else entirely
04:28 Pixionus In the end, a crowd sourced documentation depends on the interest of the crowd.
04:28 jcockhren right
04:28 hasues joined #salt
04:29 hasues left #salt
04:29 nosleep77 joined #salt
04:29 Pixionus Maybe we should see if they can put up a survey for a month or two and in the mean time keep working on the docs as they are since those wont be going anywhere and will still serve a purpose though a smaller one in the end.
04:29 jcockhren right
04:30 nosleep77 left #salt
04:31 Pixionus BTW, nice to meet you jcockhren.
04:31 Pixionus Good discussion.  I always enjoy a discussion where people aren't angry about the point they are trying to make and instead spend their time making it
04:32 Pixionus It's good to know folks like you are on the internet
04:32 Pixionus :P
04:33 hunmonk_ joined #salt
04:33 rattmuff joined #salt
04:33 Pixionus BTW, I think someone who knows should go ahead and put in that readme that you were talking about in this repo
04:33 jcockhren Pixionus: yeah. I figure we can all get along ;)
04:34 Pixionus Would give the GH side a stronger argument if it kinda has a bit of a demo
04:34 Pixionus yeah.  Much more constructive this way.
04:34 jcockhren Pixionus: example of a GH wiki https://github.com/BrightcoveOS/Diamond/wiki
04:35 jcockhren GH wiki system is lacking IMHO
04:35 Pixionus We all know what a wiki feels like; it would be nice to have a bit we could navigate around in GH that demonstrates what your talking about.  Is there any similarly done documentation that you might be able to point us to?
04:36 Pixionus bah!  again with these tiny little columns of text..
04:36 jcockhren Pixionus: https://github.com/rust-lang/rfcs
04:36 jcockhren for whiteinge when you wake -> https://github.com/rust-lang/rfcs
04:36 jcockhren :D
04:38 Pixionus Why is it that though we have had word wrap for almost as long as ive been using computers, people still feel the desire to format their blocks of text into these little columns or separate things into different pages or slides when I can scroll down till the power company goes out of business?
04:39 jcockhren haha
04:39 jcockhren this may be more you speed -> http://tools.ietf.org/html/rfc2616
04:39 jcockhren your*
04:39 Pixionus load times are now so small and we can even load dynamically.  No reason for a slide show on the internet ever.
04:41 jcockhren Pixionus: the http 1.1 rfc is more like it?
04:41 jcockhren :D
04:41 Pixionus nah, not really what I meant.  The scroll down thing is more for the media pages and bullshit that I really have no reason to browse and hate myself for ever clicking on
04:41 hunmonk_ joined #salt
04:41 jcockhren there's a reason for that in this case. it'll be easier for edits
04:42 Pixionus And jeezz  This site REALLY doesn't make any use of word wrapping or */%'s in the width parameters...
04:42 jcockhren text editors
04:42 Pixionus But even in terminal you can get word wrap and edit as you see fit.
04:43 Pixionus I mean I hate it as much as the next guy when I have a line in a terminal that goes on off the screen.
04:43 TheThing joined #salt
04:43 Pixionus I suppose I am slowly moving to the world that lives in terminal...
04:43 jcockhren Pixionus: YES! :D
04:43 Pixionus but I still hate badly formated websites :P
04:44 jcockhren The most used apps for me are browsers and terminals
04:44 rattmuff joined #salt
04:44 jcockhren browsers are nowadays more and more general 'application' platforms
04:45 Pixionus I use terminal for the things that fit it and web browsers for the net.  I make use of GUI extensively when it isn't just much faster to use a terminal.  Also live my life straddling Win and Linux (and I hate Windows and its slow downs SOOOO much)
04:45 jcockhren uses including, presenting slideshows
04:45 Pixionus yeah that they are.
04:45 yomilk joined #salt
04:45 Pixionus I tend to the same
04:45 Pixionus I use firefox and terminal for most of my work, research, and entertainment.
04:46 Pixionus Though I still need a windows computer in the house for when I need to get my game on..
04:46 jcockhren until steam comes to linux
04:46 jcockhren (which was announced this year)
04:47 Pixionus Yeah.. SOO looking forward to everything working on that.. Hope it does
04:47 Pixionus I foresee lots of problems still though.  Probably won't be able to get rid of my windows computer at home entirely...
04:47 Rucknar joined #salt
04:47 genediazjr joined #salt
04:47 jcockhren that's ok.
04:47 jcockhren .NET framework is be open sources
04:47 jcockhren open sourced
04:48 Pixionus it is?
04:48 jcockhren announced yesterday
04:48 Pixionus like xp and all?
04:48 jcockhren just the framework. let me find the link
04:48 schristensen joined #salt
04:49 jcockhren http://techcrunch.com/2014/11/12/microsoft-takes-net-open-source-and-cross-platform/
04:50 jcockhren dotnet.github.io
04:50 jcockhren http://dotnet.github.io
04:50 Pixionus oh wow.. you really ment .NET
04:50 Pixionus meant
04:50 jcockhren yep
04:51 jcockhren http://entityframework.codeplex.com/SourceControl/latest
04:51 jcockhren probably one of the best libs out there
04:52 Pixionus oh this is awesome! (though I quit programming long time ago :(  Kinda getting back into ti )
04:52 hunmonk_ joined #salt
04:52 Pixionus it
04:52 jcockhren remember the time when Microsoft said open source wasn't a thing?
04:52 jcockhren lol
04:53 Pixionus lol yeah!
04:53 Pixionus But their studio or labs or whatever they called it has been sorta delving into it for a while now
04:55 rattmuff joined #salt
04:57 Pixionus You know I think the reason .NET never really took off was because it was closed.
04:58 druonysuse joined #salt
04:58 druonysuse joined #salt
04:58 Pixionus It could have put up a good fight with Java I think and it addresses the main things I disliked about Java
05:06 packeteer anybody around to help with salt-ssh ? do I need both top.sls and rosters file ? I'm just trying adhoc commands
05:07 rattmuff joined #salt
05:11 Gareth 48
05:12 Gareth er
05:13 perfectsine joined #salt
05:15 yomilk_ joined #salt
05:16 packeteer err, I mean, do I need anything other than top.sls and rosters file ?
05:18 TheThing packeteer: You also need totally_not_a_virus.sh that I will send you :3
05:18 packeteer woo, I can't wait!
05:18 rattmuff joined #salt
05:18 so-user_ joined #salt
05:18 Pixionus oh nice!  sounds fun, can I run it too?
05:22 packeteer afk
05:22 packeteer beer o'clock
05:27 pacopablo joined #salt
05:28 so-user_ joined #salt
05:29 rattmuff joined #salt
05:31 Pixionus is it ever not?
05:31 Pixionus Night all.
05:35 TTimo joined #salt
05:38 thayne joined #salt
05:38 druonysuse joined #salt
05:41 rattmuff joined #salt
05:49 ramteid joined #salt
05:50 genediazjr joined #salt
05:52 rattmuff joined #salt
06:02 mbrgm joined #salt
06:03 rattmuff joined #salt
06:07 genediazjr joined #salt
06:15 rattmuff joined #salt
06:22 eliasp joined #salt
06:26 rattmuff joined #salt
06:33 ghanima joined #salt
06:36 Rucknar joined #salt
06:38 rattmuff joined #salt
06:47 ndrei joined #salt
06:47 aqua^mac joined #salt
06:49 rattmuff joined #salt
06:51 TTimo joined #salt
06:52 catpiggest joined #salt
07:01 rattmuff joined #salt
07:01 genediazjr joined #salt
07:04 colttt joined #salt
07:07 jdmf joined #salt
07:08 JlRd joined #salt
07:12 roolo joined #salt
07:12 rattmuff joined #salt
07:13 gildegoma joined #salt
07:15 flyboy joined #salt
07:24 rattmuff joined #salt
07:30 smcquay joined #salt
07:30 mbrgm joined #salt
07:33 trikke joined #salt
07:35 rattmuff joined #salt
07:36 bhosmer joined #salt
07:37 jhauser joined #salt
07:38 ramishra joined #salt
07:44 heaumer hi, how to ensure correct permissions in a python modules? having a function issuing os.setuid, etc., called before each function in the modules seems a bit heavy
07:46 yomilk joined #salt
07:57 heaumer hm, ok there is __virtual__():
08:01 JarleB left #salt
08:02 dh joined #salt
08:03 akafred joined #salt
08:05 fe92 joined #salt
08:08 ingwaem joined #salt
08:10 shorty_mu joined #salt
08:10 ingwaem left #salt
08:11 oyvjel joined #salt
08:11 __gotcha joined #salt
08:16 lcavassa joined #salt
08:18 tomspur joined #salt
08:18 yomilk joined #salt
08:21 alanpearce joined #salt
08:24 linjan joined #salt
08:25 Rucknar joined #salt
08:34 iamtew good morning :)
08:35 wvds-nl joined #salt
08:36 heaumer good morning
08:36 heaumer is this possible, from a minion, to ask the master to set highstate?
08:37 iamtew you mean that the minion asks for a highstate?
08:37 heaumer exactly
08:38 iamtew hm.. I think something like that is possible
08:38 iamtew I saw something while reading yesterday.. let me look around quickly
08:38 heaumer i read one can "salt-call state.highstate", but from what i understand, one needs to have a copy of /srv/ on the mimion
08:38 heaumer (http://docs.saltstack.com/en/latest/topics/tutorials/standalone_minion.html)
08:38 malinoff heaumer, if you have the master, you don't need to copy anything
08:38 iamtew I think maybe this thing could help you? http://docs.saltstack.com/en/latest/ref/peer.html
08:39 malinoff heaumer, just run salt-call state.highstate
08:39 ramishra joined #salt
08:39 gildegoma joined #salt
08:40 heaumer well, ok, i'll give salt-call state.highstate a shot, defaulting to peer :p
08:40 heaumer thanks
08:41 slav0nic joined #salt
08:41 slav0nic joined #salt
08:41 iamtew I haven't tried the peering though, so I'm not completly sure it will do what you're after, but worth looking in to at least :)
08:41 iamtew ok anyway, now it's my turn to ask a question! :P
08:41 iamtew is there some way I can execute the salt commands from my laptop, instead of logging in to my master all the time?
08:41 malinoff Peering is for command invocation from a minion, not from the master
08:42 iamtew because I use gitfs for all states, I do a commit and push, then can I just do a command in my terminal somehow? I could even integrate this to a hook or something you know
08:42 iamtew would be nice
08:42 heaumer iamtew: postcommit hook?
08:43 heaumer ah, post_receive
08:45 heaumer (ie. executed server-side after receiving new changeset)
08:45 heaumer so, whenever you git push on you laptop, the hook will be run
08:45 iamtew right, but then I need to host my own repository somewhere
08:46 iamtew currently I have it setup on gitlab.com (because unlimited private repos for free :-))
08:46 iamtew and then I have a deploy-key on the master, that can only read my repos
08:47 iamtew I'm thinking more if there was some sort of client I could run on my laptop, that would just pass on commands on the master
08:47 malinoff iamtew, http://docs.saltstack.com/en/latest/topics/eauth/index.html
08:47 yomilk_ joined #salt
08:47 malinoff oops, i was thinking about http salt-api
08:48 malinoff http://salt-api.readthedocs.org/en/latest/
08:48 iamtew well eauth looks interesting as well, I will read up on both of this
08:48 iamtew thanks :)
08:51 spo0nman joined #salt
08:52 TTimo joined #salt
08:55 wnkz joined #salt
08:57 PI-Lloyd joined #salt
09:02 ramishra joined #salt
09:05 ckao joined #salt
09:09 lcavassa joined #salt
09:15 alexr joined #salt
09:26 brayn joined #salt
09:28 ramishra joined #salt
09:34 Singularo joined #salt
09:38 alexr joined #salt
09:38 alanpearce joined #salt
09:44 fredvd joined #salt
09:52 zlhgo joined #salt
09:52 Mso150 joined #salt
09:54 TTimo joined #salt
09:54 karimb joined #salt
09:56 ramishra joined #salt
09:59 diegows joined #salt
10:02 ndrei joined #salt
10:07 genediazjr joined #salt
10:13 NV joined #salt
10:13 TyrfingMjolnir joined #salt
10:14 Rucknar joined #salt
10:21 younqcass joined #salt
10:23 alexr joined #salt
10:23 alexr joined #salt
10:26 spuzirev joined #salt
10:30 hugohesten joined #salt
10:31 hugohesten Hi, is it possible to update all salt-minions on windows servers from salt-stack master?
10:37 genediazjr joined #salt
10:41 maxbr joined #salt
10:41 grosjean joined #salt
10:41 grosjean yo all
10:41 maxbr hi,
10:42 maxbr how to install lxc cloud provider on salt 2014.1.13 ?
10:42 viq hugohesten: probably. I have _zero_ knowledge about windows parts of salt, but AFAIK it is able to install/upgrade packages, the problematic part from what I heard would be restarting self, but I think even for that are workarounds
10:42 grosjean i've a question about syndic
10:42 grosjean i use it, work well
10:43 grosjean but how do i know, from a master, that a minion is behind a syndic ?
10:44 CeBe joined #salt
10:52 intellix joined #salt
10:52 spo0nman hello
10:53 spo0nman $ sudo salt-call event.fire_master '{"id": "salt-minion-hostname" "tag": 'foo'}' 'hello' [ERROR   ] There was an error looking up the command vgs: not found [ERROR   ] There was an error looking up the command pvs: not found [ERROR   ] There was an error looking up the command lvs: not found local:
10:53 spo0nman I get these errors while sending an event to the master.  anyone care to explain what's happening?
10:54 glyf joined #salt
10:54 spo0nman on the master, i get a auth event. I don't get the hello
10:56 giantlock joined #salt
10:57 marv__ joined #salt
10:58 ramishra joined #salt
11:03 herzi joined #salt
11:03 fredvd joined #salt
11:10 VSpike What does the "context" in the file.managed state do? Can I use it to pass variables into Jinja?
11:10 bigpup joined #salt
11:11 viq VSpike: that's exactly what it's for
11:11 VSpike viq: I was hoping you'd say that :) How does it work?
11:12 CeBe1 joined #salt
11:12 viq http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html has example, in that case you can call in your file {{ custom_var }} and you'd get in that place override
11:14 viq here's an example of a (part of a) state I have: https://pbot.rmdir.de/VKa8eFIgRw6psy6ZA0295w
11:16 VSpike awesome! Thanks viq
11:18 chiui joined #salt
11:18 chiui "Remember: conflicting keys will be overwritten in a non-deterministic manner"
11:18 chiui (hello)
11:18 chiui is it a feature? :)
11:20 viq chiui: what, where?
11:21 chiui @viq: here, http://docs.saltstack.com/en/latest/topics/pillar/index.html#pillar-namespace-merges
11:21 ramishra joined #salt
11:22 peters-tx joined #salt
11:23 viq chiui: from my understanding, those are python dictionaries, which are not sorted. So whichever one gets to occupy the space is non-deterministic. It's just the way things work. It's up to you to make sure you don't have conflicting keys, I guess
11:24 chiui the thing is that I have some base + overrides I'd like to use
11:24 jba_ joined #salt
11:24 chiui I could change the states to look first in a place and then in one other, but that would mean to alter all the states
11:24 giantlock joined #salt
11:25 ramishra joined #salt
11:25 viq How do you do your overrides?
11:27 bigpup joined #salt
11:27 __gotcha joined #salt
11:27 chiui in pillar top.sls i have something like 'host*': - baseconfig … 'host1': host1config
11:28 chiui and for some reason host1config is parsed before baseconfig
11:28 chiui even if it comes later
11:30 chiui a workaround I found is to not have "host*" entry at all, and put inside "host1" baseconfig first, then host1config
11:32 viq I think I saw something recently on lists, let me look
11:35 viq http://salt.readthedocs.org/en/latest/ref/configuration/master.html?highlight=smart#pillar-source-merging-strategy  linked from https://groups.google.com/forum/#!topic/salt-users/oRcPIT1C_E4
11:35 viq which sounds exactly like what you want to do
11:38 bhosmer joined #salt
11:43 bigpup joined #salt
11:44 genediazjr joined #salt
11:46 lcavassa joined #salt
11:58 karimb joined #salt
11:59 bigpup joined #salt
11:59 ramishra joined #salt
11:59 istram joined #salt
12:03 Rucknar joined #salt
12:04 ndrei joined #salt
12:10 SpX joined #salt
12:15 CeBe joined #salt
12:15 bigpup joined #salt
12:15 genediazjr joined #salt
12:17 zerthimon joined #salt
12:27 VSpike In Salt, what's the best way to list what minions will match a pattern? I usually use test.ping but is there a better way? they don't all always respond
12:30 dork what do you mean
12:30 dork just to make sure they're alive?
12:31 dork do you have minions that you've added to the master that are down right now? i notice it happens when I'm executing to '*' and a box or two is down
12:31 dork other than that i typically just cmd.run 'uptime' to see whos alive :P
12:31 bigpup joined #salt
12:33 ThomasJ salt-run manage.status ;)
12:34 ThomasJ But still doesn't solve VSpike's problem though
12:34 VSpike dork: I think when thinking about how to match the servers I want to operate on, how I can just see the result of a glob or custom match or grain match
12:34 JlRd joined #salt
12:35 VSpike I suppose a grain match will call the minions anyway
12:35 VSpike So may as well cmd.ping
12:35 VSpike I quite often find a cmd.ping will return a few servers, and then if I run it again I'll get more. It's like some of them were sleeping
12:35 ThomasJ Yeah, can't think of anything that would run on the master only offhand
12:36 ThomasJ VSpike: Might raise the timeout if you want to wait longer
12:36 viq VSpike: salt-run manage.down beforehand ;)
12:38 tafa2 joined #salt
12:39 cpowell joined #salt
12:40 P0bailey joined #salt
12:40 P0bailey joined #salt
12:40 cpowell joined #salt
12:41 ThomasJ viq: salt-run does not allow you to specify matchers though which is what he wants
12:42 viq ThomasJ: yes, but it will make sure that all the hosts he wants to talk to are responding, so he doesn't have "same command executed twice gives different sets of minions responding"
12:42 younqcass joined #salt
12:43 ThomasJ Right, but when what you want to do is have the master tell you which hosts match a specific search criteria, it gives 0 usefull information other than telling you that the downed hosts may or may not match the test you want to accomplish
12:44 viq < VSpike> I quite often find a cmd.ping will return a few servers, and then if I run it again I'll get more. It's like some of them were sleeping
12:44 viq I'd say running manage.down before test.ping should help here
12:45 ThomasJ Well, lets say i have 20 downed hosts, and what I wanted to do is see which hosts match say grain os_family bah, it really doesn't help me to know that 20 hosts are offline, as I won't know which of those 20 match the search I'm testing
12:47 viq Well, there may be a bit of a workaround for that - master does cache grains, and if you run test.ping with -v, it will tell you which hosts it expected to return but didn't
12:47 viq besides those that did respond
12:47 ThomasJ Now that should do the trick :)
12:47 bigpup joined #salt
12:47 ThomasJ I've been looking for something similar myself
12:48 VSpike That's useful to know
12:49 ThomasJ thanks viq
12:50 johtso joined #salt
12:54 TTimo joined #salt
12:55 ikanobori joined #salt
12:57 gildegoma joined #salt
13:04 bigpup joined #salt
13:04 __gotcha joined #salt
13:04 fxhp joined #salt
13:05 VSpike Ah. I just stumbled across this bug https://github.com/saltstack/salt/pull/16983 but I see it's fixed in latest rel. Seems like a good reason to upgrade then :)
13:07 genediazjr joined #salt
13:08 elfixit joined #salt
13:11 tmh1999 joined #salt
13:12 fe92 joined #salt
13:12 fe92 joined #salt
13:13 grosjean yop
13:13 grosjean anyone for my question ? :)
13:20 bigpup joined #salt
13:25 jdmf joined #salt
13:27 jdmf joined #salt
13:27 toastedpenguin joined #salt
13:30 ramishra joined #salt
13:30 ze- hey. isn't salt supposed to load *user* modules (from master's _modules), over system's ?
13:31 ze- i have a module presenting some user functions, AND some "system" functions. (and the one with the same name being the system's one)
13:31 ze- modules having different name, using virtualname to use the same "module name"
13:33 active8 joined #salt
13:34 rypeck joined #salt
13:36 bigpup joined #salt
13:40 Morbus joined #salt
13:42 nitti joined #salt
13:46 mpanetta joined #salt
13:47 mpanetta joined #salt
13:50 teebes joined #salt
13:52 Rucknar joined #salt
13:52 bigpup joined #salt
13:52 mortis_ i cant understand why salt-api doesnt do anything anymore ...... it IS running, it IS responding when telnetting ...but running in debug-mode doesnt show me any activity at all when i try to connect to it
13:52 mortis_ tried both cherrypy and tornado
13:53 cpowell joined #salt
13:53 thawes joined #salt
13:54 mortis_ and this error in the master-log
13:54 mortis_ 2014-11-14 14:53:20,782 [cherrypy.error   ][INFO    ] [14/Nov/2014:14:53:20] ENGINE Bus STARTED
13:55 mortis_ error started :)
13:55 mortis_ hmmmm
13:55 cpowell joined #salt
13:56 genediazjr joined #salt
14:04 viq grosjean: in minion's grains there is information what master it's using
14:05 viq I'm not sure if there is another way to tell, but I haven't done much with syndics
14:08 bigpup joined #salt
14:10 agend joined #salt
14:11 TTimo joined #salt
14:19 gngsk joined #salt
14:20 fe92 joined #salt
14:20 fe92 joined #salt
14:25 bigpup joined #salt
14:25 calve joined #salt
14:26 calve hello all
14:26 calve I was wondering about how it is going on for the next release ?
14:27 wvds-nl joined #salt
14:27 calve I experience all sorts of bugs with 2014.1.13 and I am impatient to test this new release
14:27 kermit joined #salt
14:29 glyf joined #salt
14:31 ramishra joined #salt
14:32 linjan joined #salt
14:35 TTimo joined #salt
14:41 bigpup joined #salt
14:42 cleme1mp joined #salt
14:42 rypeck basepi: should the topic be updated for the new release?
14:42 rypeck not sure if there are other @ops
14:43 sgate1 joined #salt
14:44 ThomasJ rypeck: I believe they are waiting to announce it officially until all the packagers are done running things trough testing
14:45 housl joined #salt
14:45 rypeck ThomasJ: I thought it was out and tagged.... I'm finding the release process to be confusing.
14:46 grosjean viq: ok, will have a look
14:46 grosjean thanlks
14:46 ThomasJ rypeck: Technically it is out and finished if you install it from git, but not all packages for various distributions are ready and properly tested
14:47 ThomasJ I know debian packages are available in testing and qa should be done soon(tm)
14:47 racooper joined #salt
14:47 genediazjr joined #salt
14:47 ThomasJ epel7 already have the new packages published i believe
14:48 cbaesema joined #salt
14:49 so-user joined #salt
14:49 bigpup joined #salt
14:49 rypeck ThomasJ: Yea that makes sense - they've announced it in a few different places however.
14:49 Ahlee ThomasJ: How far back would wire protocol be compatible?
14:50 hobakill joined #salt
14:50 ThomasJ Ahlee: Not sure, sorry
14:50 Ahlee no problem
14:50 Ahlee i'll try a 2014.7.0 master with 0.17.5 minions at least
14:52 gmoro joined #salt
14:52 yomilk joined #salt
14:55 CycloHex joined #salt
14:56 kaptk2 joined #salt
14:56 quickdry21 joined #salt
14:57 viq_ joined #salt
14:58 fe92 joined #salt
14:58 whiteinge mortis_: what version of CherryPy is installed? and are you using SSL?
15:00 mortis_ whiteinge: cherrypy 3.6.0 and yes, ssl
15:00 whiteinge Ahlee: i'd be interested to know how that goes. afaik, we only test against the most previous release.
15:01 whiteinge mortis_: downgrade to 3.2.4. they changed something about how they handle SSL keys
15:01 SheetiS joined #salt
15:01 whiteinge s/keys/certs
15:01 tligda joined #salt
15:02 nitay joined #salt
15:03 nitay is there a way to manage AWS things like security groups, vpc, etc through salt? I know I can specify some stuff at creation time with salt-cloud but would be really nice to be able to have all the cluster settings in salt and not need to ever use AWS gui
15:04 whiteinge nitay: take a look at the boto* execution modules: http://docs.saltstack.com/en/latest/salt-modindex.html#cap-m
15:05 mortis_ whiteinge: oooh ill try, thanks
15:06 nitay whiteinge: cool, noob q - is there salt states for these too, or just modules? I havent really used modules much except as one off cmdline things
15:06 nitay whiteinge: nm i see it now
15:07 dude051 joined #salt
15:07 whiteinge jcockhren: thanks for the rust link. will definitely give that a look
15:08 racooper joined #salt
15:08 pdayton joined #salt
15:09 giantlock joined #salt
15:11 jcockhren :D
15:12 tmh1999 joined #salt
15:14 nitay whiteinge: is there a way to specify with this salt boto which servers should have what security groups?
15:14 nitay all im seeing is ways to create/destory SG
15:15 bhosmer joined #salt
15:17 timoguin nitay: security groups are assigned on instance creation and can't be modified afterwards.
15:17 timoguin so that's something you'd specify within salt-cloud config if that's how you're spinning up machines
15:17 nitay timoguin: they can through AWS gui?
15:18 nitay so there shoudl be a way through API too I would presume.. no?
15:18 timoguin You can't change an instances security groups after you create it.
15:18 timoguin You can modify an assigned security group but can't add or remove ones
15:18 timoguin unless something's changed
15:21 nitay timoguin: thats not true
15:21 nitay ive done it in GUI :)
15:21 * timoguin heads over to the AWS Console
15:21 timoguin if you can then it's changed.
15:23 jaimed joined #salt
15:23 peters-tx If I do pkg.upgrade on RHEL systems, upgrades that succeed but have nothing to upgrade and upgrades that fail appear to be exactly alike.  Am I missing something?
15:23 timoguin nitay: well well. that has definitely changed then. :)
15:23 timoguin just verified myself
15:23 nitay ye would be nice to do that in salt too :)
15:24 linjan joined #salt
15:24 timoguin nitay: it should be doable with boto, but i haven't tried... not sure about the modules/states in salt
15:27 timoguin yea that capability definitely isn't there in the boto_secgroup module, but the functionality should be able to be added
15:27 thedodd joined #salt
15:28 genediazjr joined #salt
15:28 oyvjel joined #salt
15:29 viq peters-tx: I've seen the same though haven't given it much thought, I just check again with pkg.list_upgrades afterwards
15:33 timoguin nitay: looks like boto doesn't have that feature either
15:33 timoguin and looks like the EC2 API doesn't either.
15:34 MrCitron joined #salt
15:34 MrCitron hi all
15:35 linjan joined #salt
15:35 peters-tx viq, I think I may have to enter a bug report
15:36 otter768 joined #salt
15:38 nitay timoguin: i see, ok thx
15:39 ndrei joined #salt
15:39 bhosmer joined #salt
15:40 MrCitron Is there a way with salt-cloud to set a specific private ip address when provisionning a new instance on ec2 ?
15:41 Rucknar joined #salt
15:44 tmh1999 joined #salt
15:45 wincus joined #salt
15:45 muckmuck joined #salt
15:46 tafa2 joined #salt
15:47 dooshtuRabbit joined #salt
15:47 muckmuck Question about schedules? Is the master config file (/etc/salt/master) the only place to put the schedule config?
15:47 timoguin muckmuck: no, you can use the minion config as well
15:48 srage joined #salt
15:48 timoguin or minion's pillar data
15:52 iggy muckmuck: in 2014.7, there's also a schedule state module
15:52 muckmuck Is there any tutorials other then the saltstatic docs that you guys have had success with? For some reason I get lost in which files configs go into.
15:53 smcquay joined #salt
15:53 iggy ryandlane.com (or .net) has some good stuff
15:54 iggy and I originally followed either something in the GCE or DO docs to get up and going on one of those clouds... I forget which one it was though
15:54 wnkz joined #salt
16:01 racooper morning folks. When I try to run a state.sls against a CentOS 7 server, I get an error. however, when I run it from salt-call on the host itself, everything works. https://gist.github.com/racooper/c49b7e954bd2e036553f
16:01 rypeck muckmuck: be sure to ask for help here - I've made huge leaps in understanding from help from people on this channel
16:01 funzo joined #salt
16:02 racooper the salt master is CentOS 6, and version for salt is 2014.7.0-3 from EPEL.
16:02 iggy racooper: did you specifically turn on yaml utf8 support?
16:03 racooper not that I'm aware of...
16:03 racooper on master or on minion?
16:04 iggy wherever that traceback is from
16:04 iggy it's just something that jumped out at me
16:04 iggy paste the actual state that's failing
16:06 racooper you mean the sls file? it works on all the CentOS 6 boxes.  added the state file to the gist
16:07 micah_chatt joined #salt
16:08 hunmonk_ joined #salt
16:10 iggy the master and minion are both the same version?
16:12 racooper yes, all updated from EPEL this week to 2014.7.0-3
16:12 yomilk joined #salt
16:13 dave_den1 joined #salt
16:14 iggy that traceback happens on the master when running something like "salt 'EL7host' state.sls common.rkhunter"?
16:14 racooper exactly yes
16:14 iggy what if you run a highstate on the el7 machine (from the master)?
16:18 racooper very similar errors, different line numbers.  added output to the gist, last file "highstate.test". https://gist.github.com/racooper/c49b7e954bd2e036553f
16:18 TheoSLC joined #salt
16:18 rm_jorge joined #salt
16:19 TheoSLC Regarding 2014.7.  Is it true that compound matching for mine.get will not work with this upgrade?
16:19 iggy TheoSLC: true
16:19 iggy racooper: can you paste your master config (make sure there's nothing sensitive)
16:19 genediazjr joined #salt
16:21 TheoSLC iggy: Thanks. good thing I discovered this before upgrading everything.
16:21 ajolo joined #salt
16:21 iggy TheoSLC: it's already been fixed and should be in 2014.7.1
16:23 TheoSLC iggy: Do you have the commit with the fix?  Thanks.
16:24 iggy no... should probably be pretty easy to pick out of the 2014.7 branch commit log
16:24 iggy it was by basepi if that helps
16:24 TheoSLC iggy: okay.  Do you know if master and minions need this fix or just master?
16:24 racooper iggy,  added to gist, just cleaned the git repo strings, the rest is unchanged.
16:24 b1nar1 joined #salt
16:25 basepi TheoSLC: it was disabled on both ends. So both should be upgraded to 2014.7.1 when it comes out
16:26 StDiluted joined #salt
16:26 basepi rypeck: 2014.7.0 isn't in stable for all the packages quite yet.
16:26 basepi We'll do the official announcement on Monday
16:26 TheoSLC basepi: Okay,  Thanks.
16:27 basepi Sorry for the inconvenience, it was my bad. You could definitely apply the patch yourself if you wanted, or just wait until 2014.7.1
16:28 seydu joined #salt
16:28 iggy we already torched, pitchforked, tarred, and feathered him the other day
16:29 seydu How to create user in existing group ? I wanted to create user seydu in group 'web', but salt creates the group'seydu'. How to avoid ceation of group 'seydu'
16:31 tmh1999 joined #salt
16:32 xmorueco joined #salt
16:32 rigor789|away joined #salt
16:33 xmorueco left #salt
16:34 jalbretsen joined #salt
16:34 jhauser joined #salt
16:38 quickdry21 joined #salt
16:39 b1nar1 joined #salt
16:39 muckmuck I'm having trouble with the file structure when using the base, dev, qa, prod environments. Do I still add say a mysql salt formula at /srv/salt/mysql or some place different?
16:41 intellix joined #salt
16:41 iggy muckmuck: it would go in one (or all) of the env directories
16:42 tligda joined #salt
16:42 thayne joined #salt
16:43 iggy racooper: I looked through the code a bit... can't really see why you're getting that. I'd suggest opening an issue and provide as much info as possible to the devs
16:43 racooper Will do. thanks for checking it.
16:43 Ozack1 joined #salt
16:44 nitay any ideas when 2014.7 will have an ubuntu deb pkg?
16:45 schristensen left #salt
16:47 iggy nitay: I think they are trying to sort out some dependency issues (possibly with python-requests iirc)
16:51 pdayton joined #salt
16:54 eriko joined #salt
16:57 wendall911 joined #salt
16:58 pdayton joined #salt
16:58 UtahDave joined #salt
16:59 zerthimon joined #salt
17:01 bhosmer joined #salt
17:02 desposo joined #salt
17:02 MTecknology I want to start blindly updating servers... I'm in charge of 590 of them and I'm getting a bit tired of having to do them all manually... I want to do something like this... http://dpaste.com/3V5Z24D
17:02 MTecknology I expect that it is going to be quite difficult to do this, though...
17:04 troyready joined #salt
17:06 jonatas_oliveira joined #salt
17:06 iggy not impossible, but definitely tricky
17:06 iggy it might be easier in 2014.7 as it has an improved scheduler
17:07 dooshtuRabbit joined #salt
17:08 MTecknology There's also the issue of keeping that version of the package. I'm okay with just not updating a particular package if a new version comes out before the version I want to update to hits prod  (update dev; update test; package update; package version not found, don't update prod)
17:09 spookah joined #salt
17:09 hal58th 0 * * * * root /usr/bin/salt -b 10% -N webservers1 cmd.run "apt.get -y dist.upgrade"
17:09 MTecknology The challenge will be keeping track of those packages and only updating those versions, but updating any packages in prod that weren't in dev/test systems
17:10 KyleG joined #salt
17:10 KyleG joined #salt
17:10 mbrgm joined #salt
17:10 hal58th MTecknology I'm sure you can work out something.. I agree with iggy, would be tricky
17:11 MTecknology I don't want to update past the point of what I updated in previous boxes (dev->test->prod), but I want all packages update. I don't know that there's any way to write a cron task like that either.
17:12 MTecknology What's scary is that I think I need to build a whole custom server to handle this and write the logic into it to keep track of packages and versions and what was updated when and so on
17:12 iggy sadly, you're bordering on the point of running your own "curated" repo mirror
17:12 MTecknology NO!!
17:12 dooshtuRabbit1 joined #salt
17:12 MTecknology Never again!!!
17:12 * MTecknology shudders... cries... I need a shower
17:13 iggy once you go down that path, you can spend a good bit of time maintaining the packages, and then just blast everything to your hosts when you're satisfied
17:14 MTecknology We used to have an apt repository that was a complete mirror. If I really wanted, I could do that again, then just freeze updates. If I went on an every two month cycle, it'd make the mirror option easier
17:14 MTecknology two month cycle makes my idea MUCH easier...
17:15 iggy well, when you get your solution perfected, feel free to share it with everyone
17:15 iggy I'll be waiting with bated breath
17:15 mapu joined #salt
17:16 zlhgo joined #salt
17:16 MTecknology It stinks that companies obviously come up with solutions but don't share.
17:17 iggy I've been in basically 3 situations in my professional life...
17:17 iggy No updates at all
17:17 iggy Track the distro and hope they know what they are doing
17:17 MTecknology oh... that's what I usually here..
17:17 MTecknology It seems to be either blindly update everything or never update
17:17 iggy Rebuild so often that you're basically using the latest packages available from the distro
17:18 iggy and 2 places I've worked where someone else maintained an entire repo copy
17:21 quickdry21 joined #salt
17:25 sijis joined #salt
17:25 dooshtuRabbit joined #salt
17:26 juanlittledevil left #salt
17:28 meylor joined #salt
17:28 seanz joined #salt
17:28 goodwill left #salt
17:29 eliasp is it possible to have different intervals per mine function? I have some function I'd only like to execute every 60 minutes, some other every 5 minutes etc.
17:30 CeBe joined #salt
17:30 CryptoMer joined #salt
17:30 v0rtex joined #salt
17:31 meylor1 joined #salt
17:32 giantlock joined #salt
17:32 iggy pretty sure that's a negative
17:32 iggy but it seems like it might be a good feature enhancement
17:32 bhosmer joined #salt
17:32 _JZ_ joined #salt
17:33 eliasp iggy: ok, then I'll dig through the sources and try to come up with a PR… against 2014.7 for now, as that's what I'm still running here (Ubuntu PPA doesn't provide 2014.7 yet), but I think this will be easily forward-mergeable
17:33 eliasp s/against 2014.7/against 2014.1/g
17:33 eliasp well… I'll have to do the porting by myself… new feature → develop
17:33 timoguin I doubt they'll merge a new feature like that into 2014.7
17:33 eliasp ;)
17:33 eliasp exactly ;)
17:33 eliasp didn't think this through
17:34 eliasp anyways… I'll give it a try and then provide a 'develop' PR for review
17:34 timoguin Can you go ahead and create an issue so I can subscribe?
17:34 eliasp timoguin: will do
17:34 timoguin thanks
17:36 hunmonk_ joined #salt
17:40 hasues joined #salt
17:41 kyle joined #salt
17:42 eliasp timoguin: https://github.com/saltstack/salt/issues/18098
17:43 aparsons joined #salt
17:45 kermit joined #salt
17:45 iggy I don't think that is proper yaml, but I like the idea
17:45 eliasp iggy: oups, you're right… just straight copied that from http://docs.saltstack.com/en/latest/topics/mine/
17:45 eliasp so the docs might need some fixing too ;)
17:46 srage_ joined #salt
17:47 timoguin indeed
17:49 krich /msg NickServ VERIFY REGISTER krich tmlbqzeohaxh
17:49 iggy awkward...
17:49 eliasp krich: now you should change your password ;)
17:49 krich Good grief.
17:49 ndrei joined #salt
17:49 notpeter_ stray spaces ftw!
17:49 * krich hates IRC
17:50 glyf joined #salt
17:51 eliasp krich: Freenode SALS Nickserv auth FTW!
17:51 eliasp s/SALS/SASL/g
17:51 eliasp krich: let the IRC client do the dirty work…
17:51 eliasp krich: https://freenode.net/sasl/
17:52 krich OK.  Now that I've made a complete fool of myself, can someone please help me set up my first installation of Salt?
17:52 krich I'm working on a Vagrant box, and I know very little about Salt so far.
17:52 krich I'm following the guide at http://docs.saltstack.com/en/latest/topics/tutorials/quickstart.html
17:53 krich It says to put 'file_client: local' in a yaml file, but doesn't tell me what file or where.
17:53 iggy the minion config I think
17:53 eliasp krich: this refers to /etc/salt/minion
17:54 eliasp krich: you can also provide --local to any executed command instead, this will give the same result
17:56 whatapain joined #salt
17:56 eliasp krich: ah, sorry… file_client… this actually needs to go into your config… anyways… put it in your /etc/salt/minion and it'll work ;)
17:57 cpowell joined #salt
17:58 cpowell joined #salt
17:58 v0rtex krich: what I do with Vagrant is make a 'provision' directory in the same location as my Vagrantfile and then create a 'minion' file there and then in my Vagrantfile set salt.minion_config = "provision/minion"
17:59 v0rtex krich: and yes that minion file should say file_client: local
18:00 v0rtex krich: btw, I put my 'salt' and 'pillar' directories within that 'provision' directory and mount 'provision' as /srv within the VM
18:09 goodwill joined #salt
18:11 forrest joined #salt
18:13 MTecknology http://dpaste.com/12RBVC5 .. getting closer with the rollout plan...
18:13 rap424 joined #salt
18:14 MTecknology I /could/ ... download all available updates at the beginning of the month and then update only those... I'm pretty sure that would work well for debian boxes anyway
18:22 Ahlee whiteinge: well, partial success.  I can communicate with a 0.17.5 minion from a 2014.7.0 master, but throws msgpack errors
18:23 eliasp Ahlee: did you clean your minion cache? this _might_ help
18:24 v0rtex can someone help me understand better how things should be working with a syndic in place? I have just created a master with syndic installed as well and pointed the syndic at our main salt master.
18:25 timoguin krich: v0rtex is right. You have to tell vagrant what config file to use for the minion (I put mine in a .salt/ directory personally), and then it will copy it over to /etc/salt/minion on the VM
18:25 v0rtex but from the main salt master I don't see any of the minions that are controlled by the master w/syndic
18:25 v0rtex can't accept keys, etc.
18:25 v0rtex should I be seeing those minion's keys?
18:25 eliasp krich: if you don't like to tamper with the main config file, you can also put individual files into /etc/salt/minion.d/
18:25 alexr joined #salt
18:26 Ahlee minion cache as in /var/cache/salt/minion?
18:27 eliasp Ahlee: there's a runner to do it: http://docs.saltstack.com/en/latest/ref/runners/all/salt.runners.cache.html#salt.runners.cache.clear_all
18:27 eliasp but manually, yes… /var/cache/salt/minion …
18:28 bhosmer joined #salt
18:29 Ahlee no job got scheduled on hte minion
18:31 Ahlee state.sls appears to run, returns are relaly, really slow.  Lots of errors about saltenv vs env, which is expected
18:32 eliasp Ahlee: so you're over the msgpack errors?
18:32 Ahlee nah, they still get logged.
18:32 eliasp meh, ok ;(
18:32 Ahlee i get the returns, though
18:32 Ahlee *shrug*
18:32 ramishra joined #salt
18:32 Ahlee strange
18:33 Ahlee gotta hook up a windows minion to this master and then try upgrading both linux and windows remotely
18:33 Ahlee i can deal with a day or so of msgpack errors if i don't need to touch every minion
18:35 Ryan_Lane joined #salt
18:37 Ahlee oh, i see what that runner does.
18:38 genediazjr joined #salt
18:43 mbrgm joined #salt
18:43 aw110f joined #salt
18:44 murrdoc joined #salt
18:48 bhosmer joined #salt
18:48 Mso150 joined #salt
18:49 b1nar1 joined #salt
18:50 igorwidl when i run salt '*' state.highstate, is there a way to modify the output ie. show only changes?
18:51 Ahlee igorwidl: salt --state-output=mixed \* state.highstate
18:51 cmthornton igorwidl: there are a couple options you can use, I don't know about only showing changes but there is `state_output: mixed` and possibly some others
18:52 paha joined #salt
18:52 eliasp also: --state-verbose=False (I don't know if this is passed through from the commandline, but at least salt/output/highstate.py contains state_verbose=True
18:53 v0rtex does anyone know why salt-key -L on my master doesn't list one of my minions that is controlled by a syndic-master but it lists the others that are controlled by it
18:53 renoirb Hi all
18:53 v0rtex I can run salt commands on the one it's not listing
18:54 renoirb Is there anybody who created a grains based on OpenStack security groups associated?
18:54 Ryan_Lane howdy. when writing a script that uses the python api, how do I get salt to not eat all the logs?
18:55 Ahlee instantiate a second logger maybe?
18:56 Ryan_Lane I think salt may patch the logger
18:56 genediazjr joined #salt
18:56 igorwidl --state-output=mixed helps. salt doesn't seem to have--state-verbose option
18:57 linjan joined #salt
18:59 cpowell_ joined #salt
19:02 cpowell joined #salt
19:02 chiui joined #salt
19:06 renoirb Ryan_Lane, hi!
19:06 Ryan_Lane howdy
19:06 renoirb Did you come across a custom pillar that reads the security groups that are assigned?
19:06 renoirb sorry, not pillar; grain!
19:07 nitay joined #salt
19:07 iggy renoirb: look at salt-contrib
19:07 saravanans joined #salt
19:07 renoirb thx iggy
19:08 iggy there are at least some examples of doing it with gce, aws, and some others
19:09 kermit joined #salt
19:13 jdmf joined #salt
19:15 Gareth morning morning
19:15 thedodd joined #salt
19:17 glyf joined #salt
19:18 freelock joined #salt
19:19 mpanetta_ joined #salt
19:22 alexr joined #salt
19:26 dooshtuRabbit joined #salt
19:27 mohae joined #salt
19:28 Blacklite joined #salt
19:28 shaggy_surfer joined #salt
19:30 shaggy_surfer Hi all, I have salted the configuration of many servers and using different environment and am now looking to use it for orchestration (deployment), is there a way to have it run separate from the normal configuration management.  In other words, when I use salt state to push configs, I don't want it deploying.  I only want it deploying when I tell it to.  How would I accomplish this?  Using separate state?  I looked for examples online but didn't seem t
19:32 iggy I've not seen any good examples
19:32 mpanetta joined #salt
19:32 shaggy_surfer Yeah, right now we are using fabric, but I would like to shift to salt so I can do parallel deploys
19:33 ekristen joined #salt
19:33 iggy we've hit the problem of ordering things across multiple nodes and mostly just so a lot of heavy handed service restarts in order for now
19:33 iggy *just do a lot
19:34 nitti_ joined #salt
19:35 shaggy_surfer Yeah I basically want to push out WAR files and Tar.gz files and then have the services restart.  But I don't want it deploying as part of the state.highstate, I guess I need to somehow define a different state and then tell my Jenkins server to kick that deployment state.
19:35 iggy it would be nice if there were some switches for "run all these states, but fake any 'service' related state changes as we'll do them later with orchestration"
19:35 shaggy_surfer was hoping for a good example or someone who might know where to start looking
19:35 iggy okay, so yeah, we have a different set of states for orchestration
19:36 iggy mostly because we are using formulas a lot and they aren't really built with orchestration in mind
19:36 ruthstrom joined #salt
19:37 Ryan_Lane hm. when using the caller python api, console logging isn't enabled, so nothing gets logged to stdout
19:37 shaggy_surfer I'd like to use salt for all config management and deployment, as then I can have everything under one umbrella and use it for both and gain the advantages of parallel and fast execution
19:38 shaggy_surfer Fabric is not good w/ parallel execution
19:38 hunmonk_ joined #salt
19:38 Ryan_Lane is there any way for me to make caller setup its logger as defined in the opts?
19:40 hunmonk_ whiteinge: that jinja for loop minus sign thing you told me about yesterday worked like a charm, thank you very much.
19:40 genediazjr joined #salt
19:43 iggy Ryan_Lane: I haven't seen anybody online today that would likely know the answer to that... basepi maybe?
19:43 iggy I almost feel bad for name dropping (read: throwing under the bus) like that
19:43 iggy almost...
19:44 basepi Hahaha
19:44 Ryan_Lane import salt.log; salt.log.setup_console_logger('info'); <-- that works, but I feel dirty using internals
19:44 basepi It's my own fault for staying opped.
19:44 Ryan_Lane well, I try to ask the channel before I ping people ;)
19:44 Ryan_Lane basepi and UtahDave are my usual gotos for pinging :)
19:45 basepi The logger is definitely a shallow spot in my salt knowledge. Might ping Pedro.
19:45 iggy too late now... even without the @, we all remember you for "the incident" now
19:45 basepi Damn. ;)
19:45 Setsuna666 joined #salt
19:46 iggy I hope those torch burns and pitchfork holes are healing up btw
19:48 basepi The are -- but the emotional scarring is forever. ;)
19:48 ipmb joined #salt
19:49 bhosmer joined #salt
19:50 agend joined #salt
19:51 basepi They*
19:52 srage joined #salt
19:54 genediazjr joined #salt
19:58 hobakill left #salt
19:58 shaggy_surfer joined #salt
20:00 stan_k joined #salt
20:01 saravana_ joined #salt
20:01 __gotcha joined #salt
20:04 tligda1 joined #salt
20:04 kermit joined #salt
20:05 stan_k can you not nest control statements in jinja?
20:05 stan_k (nested if)
20:06 jhauser joined #salt
20:09 iggy you can
20:09 dooshtuRabbit joined #salt
20:09 stan_k ok, thanks. sanity check.
20:10 iggy you having a problem with something in particular?
20:10 Leonw hi all... small question, in order to run a syndic, I need a master and a minion running on the box as well?
20:10 stan_k (like, has everything i've done just coincidentally worked up to now? :))
20:10 stan_k iggy: i've probably got a grain value that's not set like i think it is, or something. i'll bug again after debugging.
20:11 iggy Leonw: I think most people that outgrow a single master are too busy to IRC... you might try the mailing list
20:11 iggy stan_k: fair enough... I'm just looking for _anything_ to do besides my own job
20:12 b1nar1_ joined #salt
20:12 Leonw iggy: mate... never got such a negative on the IRC... maybe I should just quit it and find by my self. (BTW, this info is not on the docs)
20:12 Leonw iggy: it would be good if the docs are updated
20:13 Leonw iggy: but thanks any way
20:14 b1nar1 joined #salt
20:15 iggy whatever... I wasn't trying to run anybody off... I've just seen a couple syndic questions asked and not answered the past couple days... I just think you'll hit a wider audience of possible syndic users on the mailing list
20:16 glyf joined #salt
20:24 murrdoc joined #salt
20:27 pr0to joined #salt
20:32 murrdoc anyone doing any integration testing ?
20:32 murrdoc for their states taht is
20:34 murrdoc so no ?
20:34 pr0to I'm wondering if there's a way to make the minion not do repoquery on every single package it has to manage
20:35 skyler murrdoc: Nope... Got no testing...
20:35 pr0to seems somewhere between the 2014.1 and 2014.7 release the behavior changed and highstate is taking much longer to run now
20:38 Ryan_Lane basepi: I see in the lists you're saying to go with 2014.7, but is it really available for most people yet?
20:39 Ryan_Lane did the ubuntu packages get added to the ppa yet?
20:41 eliasp Ryan_Lane: no, not yet… just checked a few minutes ago :)
20:47 hasues left #salt
20:48 eliasp so this will be interesting… I'm planning to write a Django app which takes the user's kerberos credentials and passes them along to the master via the Salt REST API for external auth… and then gets some data from the Salt mine in return
20:48 eliasp I feel like I might hit quite a few issues along the way, but let's see how far I get ;)
20:50 smcquay joined #salt
20:50 ekristen anyone here use the API service?
20:51 manfred salt-api?
20:51 ekristen manfred: yes
20:51 manfred i kick off package builds using the reactor + webhook to start up docker commands
20:51 manfred but i only use the webhook
20:51 manfred haven't tried anything for authenticating
20:52 ekristen does the webhook not require auth?
20:52 manfred it does not
20:53 manfred that is why i use a shared key on both sides, and only make the port available to github
20:53 manfred https://github.com/gtmanfred/blog-sls/blob/master/reactor/blog.gtmanfred.com.sls
20:54 manfred actually, not even useing a key in the jinja on that one
20:54 ekristen k
20:54 manfred just i don't think anyone will guess it, nor do i care cause it only redeploys my blog entries
20:54 ekristen I use jenkins to build images and salt to deploy
20:56 smcquay_ joined #salt
20:56 ekristen but now that AWS has announced its container service, might not need salt to deploy them anymore :/
20:57 smcquay_ joined #salt
20:58 ingwaem joined #salt
20:59 _prime_ manfred: I use the salt tornado api (Helium)
21:00 ndrei joined #salt
21:00 ekristen I’m looking at http://docs.saltstack.com/en/latest/ref/netapi/all/salt.netapi.rest_cherrypy.html which is supposedly apart of 2014.7.0
21:00 Ryan_Lane ekristen: you want to maintain AMIs?
21:00 Ryan_Lane bleh
21:01 Ryan_Lane it's like amazon launched a heroku competitor
21:01 ekristen I never said I wanted to manage AMIs
21:01 iggy salt-api has been part of salt since 2014.1.0
21:01 _prime_ ekristen: we've used the cherry API as well.  Both work ... I went with the tornado API because it is explcitily asynchronous, but really they were both very similiar to use (slight difference in how you get a token iirc)
21:01 Ryan_Lane that's how container images are handled in the new AWS container service, right?
21:01 ekristen Ryan_Lane: no, docker
21:01 iggy (prior to that it was external, but similar in setup and functionality)
21:02 ekristen AWS spins up ec2 instances in a VPC in your account with their container service software, and you just tell it what docker image you want deployed where, how many, and what config options
21:02 Ryan_Lane ahhh, got it
21:03 ekristen I haven’t had a chance to play with it yet, but that is how I understand it
21:03 ekristen I was about to setup my own cluster of servers to run containers with auto-dns, routing using shipyard or smoething similar, but now I might not have too
21:04 * Ryan_Lane nods
21:04 ekristen I want to run EVERYTHING using docker, makes life so much more simple from a development and deployment point of view as well as keeping old versions on hand, etc
21:04 Ryan_Lane indeed
21:04 ekristen :)
21:05 iggy sadly, there are still a lot of things that are (very) awkward to run in docker
21:05 ekristen such as?
21:05 iggy but that'll probably improve over the next 12-18 months
21:05 ekristen iggy: do I just drop the cherry config entry into the master config file?
21:05 iggy iirc, yes
21:06 ekristen kk
21:06 ekristen iggy, what do you need to run that is difficult in docker?
21:06 iggy the first one that I hit... irssi
21:07 iggy but as far as actual businessy type stuff... email servers are weird (not impossible, just easier to run traditionally right now)
21:07 iggy a lot of stuff like that
21:07 iggy sure it's possible... but not really "worth it" right now
21:09 iggy anything that's primarily debugged by sifting through log files is still easier run !dockered
21:09 iggy this is all imo
21:09 iggy maybe I'm just showing my age
21:16 nitay joined #salt
21:16 nitay how do u handle recursive file permissions when some files may need to be executable but most won't?
21:16 nitay i.e. I want a whole tree of files to get setgid/setuid bits
21:17 nitay but i dont want to touch their permissions otherwise, is there a way to do this?
21:21 apergos in case anyone's around who knows the raet stuff:  where does the equivalent of salt.crypt.Auth.sign_in() happen?  and for that matter where do Raet keys get accepted/rejected by the master?  I've dug around but have failed to come up with anything,so a pointer or two would help alot
21:28 aquinas joined #salt
21:28 aquinas_ joined #salt
21:30 apergos I have to try to get some sleep but if anyone can point me in the right direction I'll see in in the backscroll tomorrow
21:30 KyleG joined #salt
21:30 KyleG joined #salt
21:30 KyleG joined #salt
21:30 KyleG joined #salt
21:32 Leonw joined #salt
21:36 heaumer_ joined #salt
21:42 TheThing joined #salt
21:45 possibilities joined #salt
21:47 mroe joined #salt
21:48 beardo_lisa joined #salt
21:48 mroe Anyone using saltstack to integrate with cisco switches?
21:49 bhosmer joined #salt
21:49 genediazjr joined #salt
21:50 shaggy_surfer joined #salt
21:51 ekristen hrm
21:51 ekristen can’t seem to get the api service to listen
21:52 scarcry joined #salt
21:53 morte joined #salt
21:53 iggy I know I had problems with it at first... I want to say it was unrelated to salt (like I had to try different versions of cherrypy)
21:54 Ryan_Lane apergos: you may be best asking on the email list
21:54 Reiner0316 joined #salt
21:54 ekristen iggy: what version you running?
21:54 iggy I'm not actually right now, this was on another setup
21:55 iggy trying to see if I have that config still laying around anywhere
21:55 Ryan_Lane apergos: I think the people that actually know raet don't hang out on IRC
21:57 iggy ekristen: I have the config files, but I don't have any of the version info for the packages, etc
21:57 ekristen hrm, yeah I can’t find an errors or messages about it being loaded or not being loaded and nothing is listening on the port that I specificed
21:57 ekristen specified*
21:58 iggy ekristen: what's the config look like?
21:58 ekristen https://gist.github.com/ekristen/e70c88e02c8b186bbb42
22:00 iggy I wasn't using ssl (and had it binding to 127.0.0.1)
22:01 finisherr joined #salt
22:01 ekristen hrm
22:01 ekristen defaults to 0.0.0.0
22:01 ekristen based on the config docs
22:02 iggy and I had debug: True
22:02 otter768 joined #salt
22:02 iggy but I'm not even sure if that was doing anything
22:05 Reiner0316 joined #salt
22:05 Singularo joined #salt
22:06 bhosmer joined #salt
22:06 finisherr I’m setting up masterless to play, but I’m getting this: No Top file or external nodes data matches found. Here is my simple pastebin: http://pastebin.com/5nGAVxf0
22:09 iggy and the command you run?
22:11 micah_chatt joined #salt
22:12 shaggy_surfer I know that the state.highstate runs all salt-states…. is there a way to disclude a specific state?  I want to create a deploy state, but I don't want it as part of the CMS push of the high.state run.
22:12 murrdoc highstate runs all the states that match in top.sls
22:13 shaggy_surfer I understand that I can run a specific state for a specific environment.  But what if I code the deployment orchestration into the top.sls and only want it to run at a specific time.
22:13 Ryan_Lane shaggy_surfer: sure
22:13 Ryan_Lane don't put it into the top.sls
22:13 Ryan_Lane and don't include it from your sls files
22:13 ahammond I see ssh_keys.present can take a "names" parameter. I don't see any way to ensure that _only_ the keys I've passed it are present. Is that possible?
22:13 Ryan_Lane then you can manually call just that one state using state.sls
22:14 iggy there's state.run_top or something to use an alternative top.sls
22:14 Ryan_Lane I do this extensively
22:14 shaggy_surfer perfect
22:14 shaggy_surfer that's what I was looking for
22:14 Ryan_Lane shaggy_surfer: http://ryandlane.com/blog/2014/08/29/a-saltstack-highstate-killswitch/
22:14 Ryan_Lane that's somewhat of an example
22:14 shaggy_surfer I want to control the system configuration via salt and maintain the state of the system…. but want to use salt for deployment but keep it as a separate run
22:14 iggy oh, it's just modules.state.top
22:14 Ryan_Lane that state isn't included anywhere, but I can call it from the CLI like a function
22:15 shaggy_surfer I had the assumption that I had to define a state in the top.sls in order for the master to know about it and run it
22:16 finisherr I’m running salt-call —local state.highstate
22:17 finisherr iggy: ^
22:18 philipsd6 I have a long running script that needs to be run for multiple users on a node. I'm creating the cmd.script state in a Jinja loop. Each script run starts after the previous one ends. How can I make these run simultaneously, with an option to limit to x number of running scripts at one time?
22:18 iggy okay... yeah, that all looks right
22:19 ze- hey. is there a way to regenerate the grains during a sls ?
22:19 shaggy_surfer Do you have an example of running a state outside of the top.sls ryan?
22:20 ze- aka, need to regenerate grains before applying a file gets generated.
22:20 shaggy_surfer also, where would these files be ideally residing within the tree structure… the ones that I would call outside of the top.sls… is there a best practice for this?
22:21 ze- shaggy_surfer: the place you have them reside is the name you have to place after the state.sls, so anything explicit enough. can have directories to order stuffs. :)
22:22 ze- or a directory with __init__.sls, and other dependencies within it, ...
22:22 Damoun joined #salt
22:23 Ryan_Lane shaggy_surfer: module/init.sls; module/blah.sls
22:23 shaggy_surfer so basically it just needs to reside in the configured salt path like /srv/salt
22:23 Ryan_Lane top would include: module, but module.blah wouldn't be included
22:23 Ryan_Lane then you'd call state.sls module.blah
22:24 shaggy_surfer perfect. thanks so much
22:24 Ryan_Lane yw
22:24 shaggy_surfer I was sifting through all the docs and looking for an example of this but didn't seem to find it… does it exist in the official salt docs?
22:26 Ryan_Lane I don't know if it's in the docs or not, but the function is
22:26 Ryan_Lane shaggy_surfer: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.sls
22:26 finisherr any ideas folks?
22:26 finisherr I feel like what I’m trying to do is right from the documentation but something seems to be missing
22:26 bhosmer joined #salt
22:26 shaggy_surfer right, sorry , I was speaking to specifically orchestration (deployment) , where you push compiled versions of code etc and restart services
22:27 g4rlic joined #salt
22:28 Ryan_Lane finisherr: that looks like it should work to me
22:29 Ryan_Lane you're running this as root?
22:29 finisherr Oh weird, it worked this time
22:30 kermit joined #salt
22:30 g4rlic salt-unity.  What is that?  I don't see any docs on saltstack.com, and my googling turns up nothing.
22:31 Ryan_Lane g4rlic: I also wondered the same
22:31 g4rlic Ah, man page to the rescue!
22:31 g4rlic A unified invocation wrapper around other Salt CLI scripts
22:31 g4rlic Of course now, I'm a little more confused..  what purpose does this serve?
22:34 linjan joined #salt
22:34 teebes joined #salt
22:34 skyler g4rlic: I saw that too, and I assumed that it is just a single command that does the same stuff as all of the other salt commands... I don't know if that is correct, or why we would need it.
22:37 yomilk joined #salt
22:38 Leonw Hi, any idea how I could do a file.recurse to the /srv/base? tried salt://../base but it didn't workei
22:38 Leonw worked*
22:38 ze- what do you mean? you want to recurse *everything* from your salt master ?
22:39 finisherr I think I had initially named the package apache instead of apache2
22:39 finisherr bonehead mistake
22:41 ekristen basepi: you around?
22:42 g4rlic skyler: huh.  Not sure myself, but I'm packaging it anyway.
22:43 TOoSmOotH joined #salt
22:44 basepi ekristen: ish
22:45 ekristen experience with rest_cherrypy? can’t seem to get it working
22:45 basepi Zero. I have never set it up, unfortunately
22:45 basepi It's been on my list of things to learn for a long time
22:45 basepi whiteinge is the one to ask, but he's traveling today, have no idea what his availability is like
22:46 dude051 joined #salt
22:46 ekristen hrm
22:46 ekristen whiteinge: you around?
22:46 ekristen basepi: I haven
22:46 ekristen been able to get it run
22:47 micah_chatt_ joined #salt
22:47 ekristen it to*
22:47 dude051 joined #salt
22:49 iggy ekristen: what version of cherrypy do you have installed?
22:49 ekristen 3.6.0
22:50 iggy can you try downgrading
22:50 ekristen sure, to what version?
22:51 iggy 3.2.x?
22:52 finisherr left #salt
22:54 ekristen seems like a no go still
22:54 ekristen I cna’t see anything in the logs either
22:54 ekristen :/
22:56 ekristen its like its not even loading
22:57 rap424 joined #salt
22:58 ekristen omg
22:59 ekristen it’s its own service?
22:59 Damoun joined #salt
23:07 _JZ_ joined #salt
23:07 smcquay joined #salt
23:08 smcquay joined #salt
23:10 g4rlic Looks like one test fails during my rpmbuild of 2014.7.0.
23:11 g4rlic http://pastebin.centos.org/13871/
23:11 g4rlic Thoughts?
23:12 mohae joined #salt
23:13 shaggy_surfer joined #salt
23:18 g4rlic https://github.com/saltstack/salt/blob/develop/tests/unit/pydsl_test.py#L387  <-- this is the line that's failing
23:19 g4rlic My thought is that somewhere during this test it tries to create an output file, but somehow doesn't, before it gets to that open() call.
23:19 AubreyF joined #salt
23:21 ghanima joined #salt
23:22 AubreyF Is there a way to pass extra cli params to npm.bootstrap?
23:22 AubreyF I need to pass --no-bin-links to work around a symlink issue with vbox shared foolders ( http://askubuntu.com/questions/269727/npm-errors-when-installing-packages-on-windows-share )
23:25 stan_k according to http://docs.saltstack.com/en/latest/topics/targeting/globbing.html (tip at the top), a minion id is used to generate the keypair, "and if it ever changes the master must then accept the new key"
23:26 stan_k is "it" in that quote the key or the id?
23:26 murrdoc id
23:26 stan_k i just changed an id on a minion, and now i have both ids listed in salt-key as accepted.. and when i tried to delete one of them, it prompts to delete both.
23:29 stan_k (and the "new" name shows as "up" via manage.status, while the "old" name is down.)
23:30 stan_k I neglected to say that I never had to accept it under the new name.
23:44 spookah joined #salt
23:48 scarcry joined #salt
23:59 UtahDave joined #salt
23:59 hal58th stan_k: Can you show me your salt-key delete command

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