Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-08-01

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

All times shown according to UTC.

Time Nick Message
00:01 dthom91 joined #salt
00:06 oz_akan_ joined #salt
00:06 jslatts joined #salt
00:06 z0rkito UtahDave: Looks like removing the run() fixed the multiple runs, but not the 0 info returning from the minion to the master.
00:08 UtahDave z0rkito: I don't think you're returning anything from your #!py sls file
00:08 z0rkito no i'm not, but the yaml/jinja sls files are also not returning data.
00:10 z0rkito putting together a paste.
00:12 aat joined #salt
00:12 Linz joined #salt
00:15 kstaken joined #salt
00:15 z0rkito http://pastebin.com/qBXEyDtK
00:17 z0rkito does look like it's running the salt-call from the #!py twice still, ~30 seconds appart.
00:20 andrew joined #salt
00:20 UtahDave z0rkito: Your sls files here are kind of confusing.  Can you rename things so you don't clobber names?    like where is test1.sls included?
00:21 z0rkito UtahDave: sure.
00:22 ahammond any news on salt 0.16.1?
00:23 UtahDave ahammond: we're releasing 0.16.2 tonight or tomorrow.
00:24 cewood joined #salt
00:26 juanlittledevil joined #salt
00:29 z0rkito UtahDave: http://pastebin.com/Sbvgbgcv
00:30 druonysuse joined #salt
00:32 UtahDave z0rkito: So is your init.sls being called from your top.sls?
00:33 UtahDave z0rkito: test.echo needs a string to echo back to you
00:33 ahammond @UtahDave exciting! :)
00:34 UtahDave z0rkito: Hey, I'm going to have to return to this tomorrow.  I'm late getting home already.
00:34 z0rkito UtahDave: yeah, top.sls only has test in it.  I'm not so much conserned with echo not coming back, i'm conserned about file not returning anything.
00:35 z0rkito UtahDave: for all of the #!py files i can just have python log it to the local system, or write proper returns.
00:35 z0rkito UtahDave: Have a good night.
00:35 z0rkito UtahDave: thanks for the help :)
00:35 UtahDave z0rkito: It should return telling you that it succeeded or not.
00:35 UtahDave I'll keep your paste open and test it tomorrow, or maybe late tonight if I have a chance tonight.
00:53 jalbretsen joined #salt
00:53 mmilano joined #salt
00:54 aat joined #salt
00:54 kstaken joined #salt
01:07 Nexpro joined #salt
01:07 napperjabber joined #salt
01:14 Gifflen joined #salt
01:17 kstaken joined #salt
01:20 jacksontj question, through the localclient API, how do i pass in kwargs?
01:22 oz_akan_ joined #salt
01:34 mgw joined #salt
01:44 minaguib joined #salt
01:45 dthom91 joined #salt
01:49 Furao joined #salt
01:57 xl1 joined #salt
02:07 Lue_4911 joined #salt
02:11 xl1 left #salt
02:13 xl2 joined #salt
02:29 KyleG joined #salt
02:29 KyleG joined #salt
02:39 syngin z0rkito: as an aside, you shouldn't parse the output of ls like you are in that pastebin dump
02:40 syngin z0rkito: for i in *sls; do echo $i; cat $i; echo; done
02:42 hellome joined #salt
02:45 StDiluted joined #salt
02:47 xl1 joined #salt
02:49 FreeSpencer joined #salt
02:58 alunduil joined #salt
02:59 alunduil Is this the place to ask questions about vagrant-salt or is there another channel for issues that may be related to that project?
02:59 jessep joined #salt
03:02 oz_akan_ joined #salt
03:05 jessep joined #salt
03:08 StDiluted alundull: what questions do you have
03:08 afx_hero joined #salt
03:08 afx_hero joined #salt
03:09 StDiluted sorry alanduil
03:09 StDiluted err alunduil
03:09 StDiluted hehe
03:09 afx_hero left #salt
03:09 alunduil StDiluted, no worries.  I'm trying to use vagrant-salt to deploy a Gentoo box with salt pre-installed but the bootstrap script fails when trying to reconfigure the system.
03:10 alunduil I'm digging through the plugin now to see how I can track it down (the deployment doesn't need a bootstrap at all).
03:10 alunduil The error I'm getting is this:
03:10 alunduil * ERROR: No dependencies installation function found. Exiting...
03:11 StDiluted ah, I haven't seen that one. Have you tried a different distro to see if it works?
03:11 StDiluted I've been using it with Ubuntu 12.04 Precise64 with no issues
03:11 StDiluted not recommending you switch but it could be good to test?
03:12 alunduil No, I'm assuming that it will work.  I can give an ubuntu box a go to verify but I noticed an open bug with the same thing for Ubuntu already.
03:12 alunduil Of course, it doesn't help me if I get Ubuntu working and Gentoo is still failing (playing with Gentoo deployments).
03:13 StDiluted which bootstrap are you using?
03:13 alunduil The one that comes with vagrant-salt when doing vagrant plugin install vagrant-salt.
03:14 alunduil Let me see if the version is in the header.
03:14 xl1 joined #salt
03:14 alunduil Looks like 1.5.2
03:17 StDiluted so it looks like the function name comes from the distro name: install_${DISTRO_NAME_L}${PREFIXED_DISTRO_VERSION_NO_DOTS}_${ITYPE}_deps
03:17 alunduil I just got to that.
03:17 StDiluted but if it doesn't get that info it defaults to null
03:17 alunduil I do see gentoo listed.  Checking to see what might be happening.
03:17 StDiluted which results in your error
03:18 alunduil Hmm, the grains reported the correct information at the beginning of the bootstrap.
03:18 malinoff joined #salt
03:20 alunduil Looks like it uses lsb_release to get that information.
03:20 UtahDave joined #salt
03:20 alunduil Guess that will need to be a dependency for vagrant boxes using salt.  Trying it after installing that.
03:21 StDiluted interestingly, in the latest version of the bootstrap-salt-minion.sh (1.5.5) it just calls 'bootstrap-salt.sh'
03:21 kstaken joined #salt
03:21 StDiluted looks like it works completely differently
03:21 StDiluted hey UtahDave
03:22 alunduil StDiluted, really?  That's interesting.  So no more lsb_release?
03:22 StDiluted looks like it tests to see if that binary exists and if it doesn't it has workarounds
03:23 alunduil Hmm, I'll double check mine but not seeing that.
03:23 StDiluted https://github.com/saltstack/salt-bootstrap/blob/develop/bootstrap-salt.sh
03:23 alunduil Is there anything special to switch to 1.5.5 in vagrant-salt?
03:23 StDiluted not sure, looks like it was done in the salt sprint
03:24 wc- joined #salt
03:24 logix812 joined #salt
03:25 alunduil Hmm.  I'll try doing an update of the vagrant plugin and see what happens.
03:25 alunduil I'll let you know.
03:26 alunduil StDiluted, you have a link to that bootstrap script handy?
03:26 StDiluted just posted it, from github
03:26 StDiluted ^^
03:27 alunduil StDiluted, I'll learn how to read, I swear.  Thanks!
03:27 StDiluted hahaha, no worries. Not sure if that's included in the latest release
03:28 alunduil It's not.  I'm going to override it in vagrant-salt and see if that fixes my issue.
03:31 alunduil That works!  Thanksa gain StDiluted.  I'll file a bug with vagrant-salt.
03:32 StDiluted sure, no problem, glad we tracked it down
03:33 crashmag joined #salt
03:33 alunduil Me too.  I'll be back if I run into more issues.
03:33 whit joined #salt
03:47 HashNuke joined #salt
03:48 HashNuke hey all
03:50 Lue_4911 joined #salt
03:51 jaequery joined #salt
03:51 HashNuke I'm going thru the Salt Minion tutorial - http://docs.saltstack.com/topics/tutorials/quickstart.html. Should I create the /srv/salt/top.sls in the VM in which I installed Salt Minion?
03:51 syngin top.sls is on the master
03:52 syngin in fact, all of your SLS files live on the master.
03:52 syngin Salt then renders them into a configuration (state) and pushes that config to the minion
03:52 HashNuke syngin: Oh so the wget bootstrap command, should I also run that on the master?
03:52 HashNuke I was going thru the masterless minion walkthrough
03:53 syngin you need to have salt running on the master and all minions, so you should bootstrap each machine appropriately.
03:53 syngin i.e., get salt (as a master) running on your salt master, then bootstrap each minion.
03:54 HashNuke Oh so even if I want to go thru the masterless minion walkthrough, I need to install salt on my host machine? (OS X)
03:54 syngin oh, masterless...
03:54 syngin just noticed that at the top of the document.
03:54 syngin sorry, guy
03:54 HashNuke syngin: no problem.
03:54 HashNuke syngin: ya masterless :) Thats why I thought why I need salt on my OSX.
03:54 syngin yes, put the top.sls on the minion.
03:54 HashNuke at /srv/salt/top.sls?
03:55 syngin yep
03:55 syngin doesn't have to be that location though
03:55 syngin you can alter the location via the salt config file in /etc/salt/{master,minion}
03:55 HashNuke Ya was about to ask that. Can I just install it in the ~/salt dir?
03:55 syngin (in this case, master)
03:55 syngin yep
03:55 HashNuke nice.
03:56 HashNuke so I have a ~/salt/top.sls file now.
03:56 syngin after bootstrapping, what do you have in /etc/
03:56 syngin erp
03:56 syngin /etc/salt/
03:56 jessep joined #salt
03:57 HashNuke ls /etc/salt says "minion  minion.d  pki"
03:57 HashNuke is there any place I have to specify the ~/salt dir as the place for sls files?
03:57 syngin edit the minion file and look for "file_roots"
03:58 syngin you might want to change "pillar_roots" too if you're going to have a look at the salt pillar
03:59 berto- joined #salt
04:00 HashNuke ah I think I'll just use /srv/salt it looks to be standard, and if I use anything from within the home dir, I'll have to specify full paths in the minion config file. Looks like it might get dirty that way.
04:06 napperjabber joined #salt
04:07 syngin HashNuke: you could always symlink /srv/salt to your homedir for convenience.
04:09 HashNuke got it ~!
04:13 oz_akan_ joined #salt
04:28 jacksontj joined #salt
04:30 gadams Can I have my states execute sequentially ?
04:39 mianos joined #salt
04:40 malinoff http://docs.saltstack.com/ref/states/ordering.html?highlight=order
04:41 Nitron joined #salt
04:41 gadams malinoff: thank you
04:42 mianos I have this bug: https://github.com/saltstack/salt/issues/6003
04:42 mianos but my new servers are not on the internet
04:42 xl1 joined #salt
04:42 mianos I have an apt proxy so normal bootstrap works
04:42 mianos but I cant use git to get a developer version
04:42 axisys joined #salt
04:43 mianos is there a way to switch to the developer build to get around the bugs in the recent release from salt?
04:44 mianos or maybe use salt to patch the source code?
04:45 gadams Why can't you use git to get the developer version ?
04:45 mianos because the machines are in a VPC
04:45 mianos the only way to connect to the packages is via an APT cache
04:46 gadams Gotcha.
04:47 mianos I've been away for 3 weeks, now I come back and try to create a new app server using everything that worked before I left and of course it's fubar
04:48 mianos at least the the bug in pip has been found and fixed
04:55 jessep joined #salt
04:55 mianos any idea how long it is going to take to get the fix to pip deployed?
04:57 jessep joined #salt
04:57 racooper joined #salt
05:04 mianos solution: wget the new pip.py directly from github into the tree
05:06 cxz joined #salt
05:12 Newt[cz] joined #salt
05:22 druonysuse joined #salt
05:22 druonysuse joined #salt
05:22 Psi-Jack_ joined #salt
05:33 kstaken joined #salt
05:33 koolhead11|away joined #salt
05:44 xl1 joined #salt
05:45 middleman_ joined #salt
05:45 HashNuke When I run "salt '*' state.highstate" on my OSX, I get "Could not access /var/cache. Please give root read permissions". The /var/cache dir doesnt even exist. Should I create it?
05:46 rmt Last time I looked, SaltStack would simply block when a server was inaccessible.. has this been addressed?
05:47 mianos it times out
05:47 mianos never saw it block forever
05:47 mianos I get them when machines are dropped but I still have the keys
05:48 intchanter HashNuke: It looks like I created it just over a year ago on my system.
05:49 intchanter So probably yes.
05:49 rmt mianos, is the timeout configurable now?
05:49 HashNuke intchanter: I keep getting permission errors for each dir... like /var/cache, then /var/cache/salt and now /var/cache/salt/master. Any way to create this quickly, without me having to create them manually after looking at the errors?
05:50 intchanter I'd love to say that I know how to do that more quickly.  The best I can suggest is:
05:50 intchanter sudo mkdir -p dir_1 dir_2 dir_3
05:50 intchanter The -p creates parents along the way.
05:51 intchanter So you only need specify the deepest directory along each path.
05:51 HashNuke intchanter: ya But I didn't know what dir it needs. So I had to run once, see the error, then create each nested dir.
05:51 intchanter They may be in the configs.
05:51 HashNuke looks like I'll just learn how to use a masterless minion.
05:52 intchanter grep dir /etc/salt/{master,minion}
05:52 intchanter Looks like most of the ones you mentioned are set in those two files.
05:52 knightsamar joined #salt
06:03 matanya joined #salt
06:04 druonysus joined #salt
06:06 HashNuke I followed the walkthrough and the masterless minion article - http://docs.saltstack.com/topics/tutorials/quickstart.html When I completed running "sudo salt-call --local state.highstate -l debug" this is what I got - https://gist.github.com/HashNuke/c6ea1bfb02fc835be6aa
06:06 HashNuke Am I doing anything wrong?
06:10 HashNuke Maybe because I was using ubuntu v10.04. I'm now recreating a Ubuntu 12.04 VM
06:14 malinoff Are you sure your repos have 'apache' package? In Ubuntu 13.10 'apache' package is 'apache2'
06:15 xl1 joined #salt
06:16 HashNuke malinoff: ah ok. so I need to figure out how to install nginx from source. I dont use apache anyway. Thought this was a copy-paste example to tryout.
06:17 malinoff HashNuke: Why do you need to install from source?
06:18 malinoff I mean, it's much better to use package manager
06:20 HashNuke malinoff: The versions in the package manager are usually older.
06:24 malinoff malinoff: isn't mainline (http://nginx.org/en/linux_packages.html#mainline) version enough for you?
06:24 dthom91 joined #salt
06:25 HashNuke malinoff: Thanks. Didn't know about that.
06:25 balboah joined #salt
06:26 malinoff HashNuke: np
06:26 malinoff Take a look on pkgrepo.managed
06:26 HashNuke malinoff: is there any cheatsheet for saltstack? Like quick examples of how to run commands, etc?
06:27 HashNuke I just started with salkstack. So most of the vocabulary still hasn't settled in my head.
06:27 malinoff HashNuke: documentation, i think
06:27 HashNuke Very detailed docs but so much to chew.
06:28 HashNuke Would love something in the lines of http://docs.saltstack.com/topics/tutorials/quickstart.html but a little more elaborate
06:28 malinoff Well, http://docs.saltstack.com/index.html - this page helped me a lot :)
06:29 malinoff http://docs.saltstack.com/search.html?q=&check_keywords=yes&area=default and that
06:30 qba73 joined #salt
06:32 HashNuke I'm read the states tutorial. What does "highstate" mean?
06:35 robinsmidsrod joined #salt
06:42 liuyq joined #salt
06:52 malinoff You have top.sls file which is mapping between minions and states (sls files) that must be applied to.
06:53 malinoff salt \* state.highstate simply means "walk over ALL minions and IF current minions is in top.sls apply states on it
06:53 malinoff current minion*
06:54 HashNuke Just realized saltstack also does monitoring
06:56 HashNuke thanks goodness I found this - https://github.com/saltstack-formulas
06:57 dthom91 joined #salt
06:57 mike25_ joined #salt
06:58 mike25_ hi guys - i am REALLY new to salt ... i have read through the documentation and i think salt is awesome
06:58 mike25_ i just want to congratulate the minds behind this brilliant product
06:58 mike25_ is this  a good place to ask STUPID/newbie questions ?
06:59 syngin mike25_: sure is
06:59 mike25_ i have set up a master + 5 minions
07:00 HashNuke mike25_: I've been doing that for a few hours now.
07:00 mike25_ just for the sake of understanding ... pillars grains ... etc
07:00 HashNuke I'm hoping there's some lay man's explanation of these stuff.
07:00 auser hey all
07:00 mike25_ i am interested in your opinion.. and best practice.. i want to START it the right way
07:00 auser HashNuke: that's what I'm working on… slowly
07:01 auser http://saltstat.es
07:01 HashNuke auser: HEy ~!
07:01 auser holla HashNuke
07:01 HashNuke auser: I just found your book and was about to tweet you about the status of the book.
07:01 auser ah, nice
07:01 syngin mike25_: anything in particular?
07:01 mike25_ auser: nice site... i can help out if you need  - since i am a beginner - i am writing everything down
07:01 HashNuke auser: You are Ari right?
07:01 auser I am Ari, and you are?
07:02 auser thanks, that'd be awesome
07:02 HashNuke auser: Akash Manohar
07:02 auser I'm working on two at the same time, so it's going a bit slower than I'd like
07:02 auser ahh hey Akash
07:02 auser we met at the SF meetup, right?
07:02 mike25_ syngin: yeah: thanks for asking... i am thinking what is the best purpose of pillars ? just to store unique data for each minion? if i have 200 minions.. for each one is recommended to have a pillar file?
07:02 HashNuke auser: Is there a way I can get a preview of the book (paid is fine)
07:02 syngin mike25_: It
07:02 syngin erp
07:02 HashNuke auser: No auser. I'm in Bangalore.
07:03 auser Yeah, you may have to wait a few weeks before I'll have a preview copy ready to go
07:03 auser ahh, okay
07:03 auser nm
07:03 HashNuke auser: And going by the listing on your leanpub profile, I would guess the other book is on angularjs :)
07:03 auser mike25_: no
07:03 syngin mike25_: it's up to you. i use pillars for networking config (interfaces, ips, gateways, smtp, nameservers) and users.
07:03 auser yep
07:03 auser ng-newsletter.com/articles
07:03 syngin mike25_: you can do the whole lot in /srv/salt if you prefer.
07:03 auser you should think of pillars as applying data to a targeted minion
07:04 auser meaning you should build your pillars up for different 'types' of minions
07:04 auser not one for each
07:04 syngin mike25_: yeah, what auser is saying.
07:04 mike25_ and if i want for example to ... group minions ... with pillar data ... is it possible?
07:04 syngin mike25_: e.g., i have network config pillar files for each data center.
07:04 syngin mike25_: yep
07:04 syngin mike25_: I use grains for that.
07:04 mike25_ for example.... in a pillar to store data like... let me think :)
07:05 mike25_ pfff now i am confused :) why do we need both grains and pillar?
07:05 syngin mike25_: i keep role, deploymode, and datacenter in grains for each minion and then i have pillar and salt set up to leverage those groupings (or combinations thereof)
07:05 auser yeah, totally
07:05 syngin mike25_: grains are per minion.
07:05 auser grains are something totally different
07:05 auser think of grains as like minion-specific information
07:06 mike25_ i see that the flexibility of the setup is really awesome
07:06 auser like how many cores does it have, how much memory
07:06 auser how many networking devices, etc.
07:06 mike25_ yeah i saw that...
07:06 HashNuke auser: Any usage guide for the salt formulae? They look straight forward. But I get some errors when I use the openssh formulae - https://gist.github.com/HashNuke/036ab2029bdd77440b65
07:06 syngin mike25_: they, by default, contain specific information about a minion (like number of cpus, vm type, ram, ip addresses etc)
07:06 mike25_ yep .. i see that ... i run a grains.items
07:06 auser are you running it as root HashNuke?
07:06 syngin mike25_: you can add information to that data set so that you can target minions or slice your pillar/salt files according to grain information.
07:06 auser you *can* use them to store info
07:06 HashNuke auser: Ah sorry. just noticed that and tried it. And still get another error.
07:07 auser mike25_: http://www.saltstat.es/posts/role-infrastructure.html
07:07 HashNuke auser: Now my error is smaller but different - https://gist.github.com/HashNuke/ec868fd9352d6940fa2b
07:07 berto- is there an easy way to dump content from pillar data into a file?
07:07 auser what os HashNuke?
07:07 auser all pillar data berto- ?
07:07 HashNuke auser: I'm on ubuntu 12.04 VM with vagrant
07:08 berto- auser: no, just one key.
07:08 auser hm, check your /var/logs/syslog
07:08 auser ah
07:08 auser {{ pillar['key'] }} ?
07:08 mike25_ auser: awesome
07:08 berto- auser: it has indents in it and file.managed's -contents barfs.
07:08 HashNuke auser: dont bother. Now I understand what it is. vagrant hooks up to the ssh port. So that's the problem.
07:09 auser ah
07:09 auser with what error berto-
07:09 HashNuke auser: May I suggest a change to the file structure on the salt formulae organisation?
07:09 mike25_ auser: i am just reading through your site.... so as far as i can see ... is valid to use grains to add roles to a server... like for example i want to create salts to install mysql , php, apache etc... so in a grain i can add roles for a server to be just mysql or to have also mysql, apache, php - so a full lamp install
07:10 berto- auser: ScannerError: mapping values are not allowed here
07:10 auser yep
07:10 berto- in "<unicode string>", line 16, column 28:
07:10 syngin mike25_: i also recommend having a look through other people's state directories: http://docs.saltstack.com/topics/community.html#example-salt-states
07:10 auser berto-: can you paste your whole state?
07:10 unicoletti_ joined #salt
07:10 HashNuke auser: If you can make the formulae dir the root of the repo then, people can just clone it and add it as submodules to their formulae book. Else it has to be cloned, then the dir has to be copied, etc
07:10 mike25_ syngin: i did that - i had a look around really nice things.
07:11 HashNuke oh i think it's still pretty straight forward if I use openssh.openssh
07:11 auser yeah, actually, the formulae structure was dictated before I got there :/ But I think that's a much better approach
07:11 auser I'll see what I can do
07:11 berto- auser: https://gist.github.com/rca/6129090
07:11 auser no, your state
07:11 auser not the error
07:12 berto- auser: oop: https://gist.github.com/rca/6129090
07:13 mike25_ can the grains be all kept on the master? for all minions?
07:14 auser you can collect them on the master
07:14 auser but each minion knows about it's own grains
07:14 auser ah berto-
07:14 berto- mike25_: maybe look at pillar data?
07:14 mike25_ auser:  so in your example : http://www.saltstat.es/posts/role-infrastructure.html grains are set up on the minions?
07:14 auser I didn't realize you were collecting it in the state
07:14 berto- mike25_: you can manage and configure pillar data on the server and serve it to minions.
07:15 syngin mike25_: i manage all the grains configs on my master and use salt to manage them.
07:15 auser you need to do {{ salt['pillar.get']('key') }}
07:15 mike25_ berto-: yeah that i already know ... but i am just wondering about all possibilities... i think salt is TOO flexible :)
07:15 syngin erp. i keep them on my masker
07:15 syngin master. dammit.
07:15 auser and you don't need indent(0)
07:15 auser I use salt to manage salt mike25_
07:16 mike25_ syngin:  so you have a /etc/salt/grains folder and inside grains files for each minion? HOW do you send them to the minions?
07:16 berto- auser: ok, trying ...
07:16 berto- auser: is pillar.get() special and gets rid of indents?
07:16 syngin mike25_: i have my grains somewhere like this: /srv/salt/files/grains/HOSTNAME
07:16 mike25_ auser: using salt to manage salt .. pretty interesting .... i have to think about it a bit :)
07:17 auser I have an example of that
07:17 auser salt-formulae/salt
07:17 auser on github
07:17 mike25_ syngin:  ok....
07:17 syngin mike25_: then in my state file i just use the hostname of the minion to help that minion find its' own grains config
07:17 mike25_ syngin: and you just send the files to each minion via salt '*' state.highstate ?
07:18 syngin mike25_: yep. i don't change my grains config very often at all
07:18 mike25_ yeah i can imagine that once the grain is set up... is done.
07:18 syngin mike25_: once you have a setup that works, you'll add a new host, write the grain file once, and that'll pretty much be it.
07:18 mike25_ syngin:  true- taht is what i am looking for ... easy scalability
07:18 syngin mike25_: i think of grains as being "all the bits of metadata about a server that i will need to identify it correctly.
07:19 mike25_ but first i have to understand ... all possible ways of doing this
07:19 mike25_ syngin: yeap really well said!!!
07:19 HashNuke auser: How do I run commands for installing a package? For example, I would like ruby to be installed via rbenv. I would love to update the ruby formulae too if I knew that.
07:19 syngin mike25_: just phrase your management questions when deisgning what data to put in there. for me, it's basically "i need to manage all the web servers in dc1 which are prod mode"
07:20 syngin or "run this on all of the syslog collectors in dc2"
07:20 syngin and so on.
07:20 auser HashNuke: run through this real quick: http://www.saltstat.es/posts/quickstart.html
07:20 mike25_ yeah ... the Question has to be well put.. then the answer can follow
07:20 syngin once you know *how* you want to target minions for day-to-day remote execution, that'll define your grains.
07:20 auser should give you a good idea for how to start
07:21 mike25_ syngin: /etc/salt/grains is a file on the minion? is it a standard thing?
07:21 syngin mike25_: yep
07:21 HashNuke auser: Thanks ~! I was looking for something of those lines since I started with start.
07:21 mike25_ i am not sure i have seen this in doc... but i might be wrong
07:21 syngin mike25_: not sure what that looks like on windows though - still haven't gone that far with salt yet
07:22 mike25_ syngin:  so the grains you have in your /srv/salt/files/grains/HOSTNAME will be send to minion to /etc/salt/grains ?
07:22 syngin mike25_: bingo
07:22 syngin where HOSTNAME is that name of the minion.
07:22 mike25_ ok.. didn't know about the /etc/salt/grains file
07:22 mike25_ yeah sure :) i GOT THAT :)
07:23 syngin mike25_: you can also set grains in state files using the grain module.
07:23 syngin mike25_:  ;)
07:23 mike25_ syngin:  yeah ... and from cli directly
07:23 syngin mike25_: but i prefer a flatfile for grains.
07:23 mike25_ i tested from cli
07:23 syngin lets me bulk edit changes with sed et al.
07:23 mike25_ syngin: me too.. i would prefer a flat file + separation of concepts
07:24 mike25_ syngin:  yeah.. i just read about the sed module - really good one.
07:24 mike25_ i have to say .. that with such a nice community like you guys... SALT will be huge in 2-3 years
07:24 mike25_ puppet will suck big time :)
07:25 syngin heh
07:25 syngin i like salt for three reasons: Python, ZMQ, flexibility.
07:25 mike25_ i worked a bit with puppet... and seems like dark ages compared to salt
07:25 bud joined #salt
07:25 syngin i also like the lack of a DSL for defining things. jinja + yaml is so handy.
07:26 mike25_ i just started learning python as well :) that is how i come across salt
07:26 syngin ahh, python's a great language. certainly not without its warts, but very classy.
07:26 mike25_ i used to be a web devel for 7 years .. lamp environment
07:27 syngin gotcha
07:27 mike25_ then i moved a bit towards sysadmin :)
07:27 mike25_ now i am doing both :)
07:27 syngin i did my time in the lamp trenches.
07:27 syngin now i push python/django
07:27 mike25_ still good though...
07:27 mike25_ i am still  giving php a chance... and symfony2 rocks
07:28 syngin the right tool for the right job. and skillset too.
07:28 mike25_ symfony2 has a lot from django ... twig ... logic behind .. etc
07:28 syngin much easier to hire a PHP dev than a Python one.
07:28 mike25_ TRUE... i am in Romania ... and we can not find here python devs...
07:28 mike25_ php devs.. are a lot
07:28 mike25_ where are you guys from?
07:29 malinoff I'm from Russia, if you want to know :)
07:30 bud mike25_: .ro, sweet, where from .ro?
07:30 cowyn hello, does salt support yum plugin?
07:30 mike25_ nice to meet you malinoff
07:31 mike25_ bud:  timisoara
07:31 HashNuke I'm from India. And looks like I'm the odd one out in this channel - I use ruby
07:31 bud No way mike25_ :)). me too.
07:31 mike25_ haha HashNuke Ruby is the future
07:31 syngin mike25_: I'm from .nz, but I live in .au
07:31 mike25_ bud: yeah you ass i know you Vali - i am mihai
07:31 mike25_ nice to meet you syngin
07:32 syngin mike25_: you too, mike25_
07:32 HashNuke mike25_: I'm not sure of that. I always go with the flow. I spend my spare time learning erlang.
07:32 bud mike25_: which mihai? I have a hunch but I am not sure...
07:32 mike25_ not bad :)
07:33 Furao this place used to be dead during the night in US
07:33 malinoff >ruby end end end end end end end end end end end end end end end end end end
07:33 mike25_ thank God i came here :)))
07:33 bud Hey Furao.
07:34 syngin Furao: indeed. I'm only online here when I'm in the office.
07:34 Furao hi bud :)
07:34 Furao I'm actually in Malaysia and for months there were only me, auser and few others who were not sleeping at that time
07:34 berto- mike25_: salt is pretty flexible; is the flexibility causing you grief?
07:35 auser heh, it's true Furao
07:35 mike25_ berto-: i think so ... there are more ways of doing one thing... and i am just looking for the best way - which is always relative depending on the setup
07:36 berto- mike25_: i know what you mean.  i decided on going with pillars for minion data and so far it's been good.  i don't use grains for configuration data.
07:40 mike25_ berto-:  i was thinking to do the same thing
07:40 mike25_ berto-:  since pillar data is automatically pushed (as far as i can see) is a lot easier
07:41 berto- mike25_: yep, every time highstate is run is my experience.
07:44 malinoff Can some one help me? I'm trying to use pkgrepo.managed: require_in with yum - and my repo file in /etc/yum.repos.d contains "require_in=[{'pkg': 'my-cool-pkg'}]"
07:45 malinoff Should i create an issue?
07:46 bud malinoff: can you share your sls files?
07:46 cowyn joined #salt
07:49 druonysus joined #salt
07:50 malinoff https://gist.github.com/malinoff/6129285
07:51 jhauser joined #salt
07:53 malinoff And i'm using 0.16.0 salt
07:54 bud malinoff: as far as I know that looks good. I have never extended a pkgrepo though and I had problems the other day including a pkgrepo in a state.
07:54 bud Not the same error though, the state was crashing.
07:57 malinoff Well, i can put require_in clause right into my-repo.sls, it's still broken
07:57 bud That how I have done it and the state crashed.
07:58 nil-yang joined #salt
07:58 malinoff What's the output?
07:58 bud I ended up in putting the pkgrepo.managed in the same state as the package.
07:58 malinoff Not the solution for me
07:58 bud Just a second to modify the state so I can show you.
07:58 cowyn joined #salt
07:58 bud I agree that's not elegant.
08:00 carlos joined #salt
08:00 malinoff
08:01 bud Cannot extend ID unbound in "base:apt.repo.squeeze-backports". It is not part of the high state.
08:02 bud Data failed to compile ^^
08:02 nil-yang left #salt
08:02 malinoff Uh, just type "squeeze-backports", not all the path
08:03 bud malinoff: where?
08:03 malinoff This path is just like python's "import a.b.c.d.e.f as f"
08:03 felixhummel joined #salt
08:03 malinoff Can you share your sls?
08:07 bud malinoff: yes I can - http://pastebin.com/DFNmz8ha
08:09 malinoff I'm not sure, but my exprerience says you cant 'require' a pkgrepo, only pkg
08:09 malinoff It's the first thing
08:10 bud So I should use just fromrepo and that would do it?
08:11 bud Is anybody using gitfs along side file roots?
08:11 Furao bud: yes
08:11 malinoff No, you 're specifying 'require_in' clause for this purpose
08:11 Furao bud: i use non-gitfs only for my top.sls
08:11 bud malinoff: thanks, I'll give it a try in a sec.
08:14 bud Furao: my setup doesn't want to work. I have a file_roots:\n  base:\n  /root/code/salt-states. fileserver_backend:\n  - git\n  - roots and a gitfs_remotes:\n  - file:///var/lib/git/repositories/salt-states.git.
08:14 Xeago joined #salt
08:14 bud The /root/code/salt-states is a clone of the repo from /var/lib/git/repositories/salt-states.git.
08:15 bud The setup should pick up file from git and fallback to file roots. Am I wrong on this assumption?
08:15 Furao yes, because you put git first
08:15 bud Yes, I am wrong, or yes the files come from git first?
08:16 Furao yes the file should be picked first from git
08:16 Furao and then file roots
08:17 bud Ok. I have a file nox.sls under /root/code/salt-states/mine/nox.sls to learn about the mine.
08:17 bud I don't want to add, commit, push to the repo every time.
08:18 krak3n` joined #salt
08:18 bud salt '*' state.sls mine.nox gives empty output on minions.
08:19 bud salt-call throws out: [ERROR   ] No matching sls found for 'mine.nox' in env 'base'
08:20 bud switching only to file_roots works :).
08:20 kchr joined #salt
08:23 ml_1 joined #salt
08:23 bud I have also remove /var/cache/salt, restarted the daemons and tried again, same result.
08:28 it_dude joined #salt
08:29 HashNuke auser: Nice tutorial. Completed it. (I did it with digitalocean)
08:29 auser thanks HashNuke
08:29 auser did you take notes?
08:30 auser I wanna get a quickstart for digitalocean up
08:30 bud auser: what tutorial are talking about?
08:30 auser saltstat.es
08:30 HashNuke auser: Yes I have my whole command history. I can write notes for you.
08:30 auser awesome! If you're too busy, feel free to send me the history (with sensitive info blocked out)
08:30 HashNuke auser: maybe also change the AMI to use a micro API instead of a small. That way anyone who is just getting started with a free account can do it too.
08:30 auser I have little experience with digitalocean
08:31 auser yeah, good idea
08:31 HashNuke AFAIK, AWS gives a free micro instance to everyone for the first year.
08:31 malinoff So, any ideas about my problem? :)
08:31 HashNuke auser: same here. digitalocean is another nicer throwaway host. I havent hosted anything serious, so have nothing to complain.
08:32 HashNuke auser: I'll send you the notes in a few minutes. <goes off to look thru history>
08:32 auser thanks HashNuke
08:32 HashNuke auser: No problem. happy to help.
08:33 Jason-AVST joined #salt
08:39 lulala joined #salt
08:47 HashNuke auser: Here are my notes - https://gist.github.com/HashNuke/6129613
08:48 auser awesome, thanks HashNuke
08:48 auser I'll get a post up this weekend
08:48 auser with your work
08:48 auser (have a twitter handle/github so I can credit you?)
08:48 auser or I'll just use HashNuke
08:48 HashNuke auser: sure. @HashNuke on twitter.
08:48 auser rgr
08:49 HashNuke auser: Can you add a small note at the bottom of the post about how to run all states at once if you have multiple states.
08:50 auser to which post?
08:50 HashNuke auser: the 60 seconds post.
08:50 auser ah
08:50 auser sure, I will
08:50 auser I have about 10 more minutes of life in me for today
08:50 HashNuke That would be a good hint to continue to refer to the rest of the saltstack on what top or base means.
08:50 auser so I'll do it tomorrow
08:50 auser good idea
08:50 HashNuke auser: No problem :)
09:05 berto- joined #salt
09:06 aleszoulek joined #salt
09:12 bud Furao: can I pm you, e-mail, or other kind of means?
09:13 Furao bud: I wasn't there sorry
09:15 bud No problem Furao, I want to ask you some things and I was wondering if I can pm you.
09:15 Furao I PM'ed you
09:15 Furao you don't see it?
09:28 jacksontj joined #salt
09:38 fredvd joined #salt
09:44 helderco joined #salt
09:46 whiskybar joined #salt
09:53 bemehow joined #salt
10:00 ggoZ joined #salt
10:14 fredvd joined #salt
10:21 lyddonb_ joined #salt
10:27 giantlock joined #salt
10:31 Katafalkas joined #salt
10:32 jslatts joined #salt
10:36 it_dude joined #salt
10:37 backjlack joined #salt
10:39 * liuyq is away: 我很忙
10:42 Furao liuyq: ?
10:42 Furao public away are so 1998 :)
10:43 Furao you're the first from HK I see here
10:45 david_a joined #salt
10:54 bhosmer joined #salt
10:58 robbyt joined #salt
10:59 lazyguru joined #salt
11:01 lemao joined #salt
11:01 Furao https://github.com/saltstack/salt/pull/6469 <- 0.16.2 is coming
11:07 Katafalkas joined #salt
11:10 liuyq Furao, no, I am from China
11:10 Furao liuyq: oh, sorry you're IP whois is in HK
11:10 mike25ro Q: grouping minions into nodes can be done only in the master conf? or can be set up in a different file?
11:11 Furao probable the owner of the IP block
11:11 liuyq Furao, maybe, the IP should be the public ip of my company
11:12 liuyq or the database the ip is not right:)
11:20 cowyn joined #salt
11:21 Ivo joined #salt
11:23 felixhummel hi all! how many salt-minion processes are spawned on your boxes? i have 300 minion processes running - seems a little too many...
11:24 Furao felixhummel: I have a monitor check that make sure never more than 2 run
11:26 felixhummel could be this one: https://github.com/saltstack/salt/issues/5729
11:28 felixhummel after killing them all and restarting the service, i have one process. strange. i'll keep an eye on it and report back here, if i find something.
11:29 Katafalkas joined #salt
11:30 bemehow joined #salt
11:32 mike25ro is this valid? : salt '*' --pillar 'salt23:saltpillar' grains.items   i have the salt23 pillar set up i can see it
11:32 mike25ro i just want to filter based on that pillar
11:33 bhosmer_ joined #salt
11:40 mike25ro2 joined #salt
11:41 mike25ro2 hey again
11:41 drawsmcgraw joined #salt
11:41 koolhead11 left #salt
11:43 mike25ro2 bud:  still around?
11:45 tomeff joined #salt
11:49 mikedawson joined #salt
11:50 com48 joined #salt
11:58 bud mike25ro2: yes
12:00 mike25ro2 bud:  i was just testing .... irc / pidgin :)
12:00 krak3n` joined #salt
12:03 bhosmer_ joined #salt
12:04 blee joined #salt
12:11 mike25ro2 guys...
12:11 mike25ro2 i have set up a pillar for 2 hosts :
12:12 mike25ro2 salt2:
12:12 mike25ro2 ----------
12:12 mike25ro2 salt23:
12:12 mike25ro2 saltpillar
12:12 mike25ro2 from the master
12:12 mike25ro2 salt '*' match.pillar 'salt23:saltpillar'
12:13 mike25ro2 i tried to do that ... but it returns FALSE for salt2
12:14 mike25ro2 it should return true.. since the pillar exists on minion with id salt2
12:14 mike25ro2 am i missing something?
12:17 Nexpro joined #salt
12:26 fredvd joined #salt
12:26 ninkotech joined #salt
12:28 felixhummel i' begin with inspecting the pillars, i.e. salt '*' pillar.items. maybe salt '*' saltutil.refresh_pillar before that...
12:30 Nitron left #salt
12:34 jefferai joined #salt
12:34 drdran joined #salt
12:35 bhosmer joined #salt
12:39 drdran hi! did anyon have used ldap for acl authorization of salt?
12:40 mike25ro felixhummel: all seems fine.. i have the pillar displayed when i run salt '*' pillar.items  -> i have it set up on 2 hosts
12:40 krissaxton joined #salt
12:41 mike25ro oh shit.. it works now... WTF? :)
12:41 mike25ro anyway ... hmm ... thanks for wasting your time reading my crap :)
12:45 drawsmcgraw mike25ro: I pulled one of those yesterday with some yum/rpm oddness. Don't feel too bad :)
12:45 mikedawson joined #salt
12:45 mike25ro another Q ... how can i ... add a command after this one:
12:46 mike25ro salt '*' match.pillar 'salt23:saltpillar'   -> it returns 2 hosts .. which is correct ... i just want to get their ip addresses... and i thought this would work:
12:46 mike25ro salt '*' match.pillar 'salt23:saltpillar' grains.item ip_interfaces
12:46 mike25ro but it doesn't do anything...
12:46 mike25ro i suppose i didn't put the correct sequence or smth
12:48 drawsmcgraw mike25ro: Not sure what your use case is, but you could swap out 'grains.item ip_interafaces' for something like 'cmd.run ifconfig | grep <something useful>'
12:48 drawsmcgraw kludgy though....
12:48 bemehow joined #salt
12:48 mike25ro drawsmcgraw:  i was just wondering if is possible to .... concat more actiopns... commands..
12:49 mike25ro salt '*' grains.item ip_interfaces   works just fine
12:49 mike25ro I am just testing what salt can do ... and how to do things with it.
12:49 mike25ro and some of these things... can be tested directly into cli...
12:49 mike25ro not sure how to .. pipe them together...
12:50 __Jahkeu_ joined #salt
12:52 mike25ro it seems that : this is how it works
12:52 mike25ro salt -C 'I@salt23:saltpillar' grains.item ip_interfaces
12:52 mike25ro with a compound match
13:02 _ilbot joined #salt
13:02 Topic for #salt is now Welcome to #salt - http://saltstack.org | 0.16.0 is the latest | Please be patient when asking questions as we are volunteers and may not have immediate answers - Channel logs are available at http://irclog.perlgeek.de/salt/
13:03 bhosmer joined #salt
13:12 bud How can I require a state that has an ID like the following tinc_host_{{ id }}_config in an extend service declaration?
13:13 dinoop joined #salt
13:14 krissaxton joined #salt
13:14 jslatts joined #salt
13:14 juicer2 joined #salt
13:15 cron0 joined #salt
13:16 dinoop We are automating the complete SDLC can tat be possible with saltstack ?
13:17 brianhicks joined #salt
13:19 mlister2006 joined #salt
13:19 bemehow joined #salt
13:22 Furao dinoop: I did, but salt can't do everything by itself. I had to use things such as Jenkins to power developers to manage their own builds. but yes once they set some flag on, it ends in a VM and finally someone can go on my custom inventory web ui and promote it to prod. salt do everything in background
13:22 QauntumRiff joined #salt
13:22 mgw joined #salt
13:23 bhosmer_ joined #salt
13:23 QauntumRiff is there a 'best practices' for layout of the /srv/salt directory?  sub-directories for each application (ie, ldap, http, etc) or configs in one directory, files in another? Or something else entirely?
13:23 QauntumRiff as I grow my usage of Salt, I keep finding I don't like the way I set it up for my 1 test server (I have about 40 minions now) :)
13:24 tempspace Does anybody have any iptables rules for salt minions?
13:24 drawsmcgraw QauntumRiff: I'm *really* eager to see something like this
13:24 drawsmcgraw I'm seeing issues down the road where I'll likely have conflicting state decs and such as well
13:25 QauntumRiff I'm kind of leaning towards a "ldap/config and ldap/files" kind of structure (ie, httpd/conf and httpd/files)
13:25 QauntumRiff and then a 'test' directory, with a whole other set of app directories inside of it
13:25 tim_s joined #salt
13:26 tim_s left #salt
13:26 mike25ro this is what i am looking here as well... being a newbie... i am looking to start the layout using the best layout practice :)
13:27 mike25ro salt is flexible... but messing it up with lots of configs... seem easy :)
13:27 QauntumRiff mike25ro: yeah, some 'best practices' docs would be very nice in general.. all the tutorials cover one or two applications :)
13:27 QauntumRiff its a big leap between that, and really using it
13:27 dinoop Thanks Furao got hope:)
13:28 mike25ro QauntumRiff: indeed... i was reading a lot in the last days... and just small tests...
13:28 Khollowa_ joined #salt
13:28 mike25ro there are some git examples,,,, but everyone seem to set up states and  pillars differently
13:29 mike25ro QauntumRiff: just this morning i was asking how is the best way to set up minion-roles etc.. some said using grains, some using pillars - i would go for pillars
13:30 mike25ro i am just learning salt... and i really want to waste as much time as possible... just to get things right... and easy to scale.
13:31 mike25ro I am still wondering how to ... set up different environments... if to use ldap or just salt-user management etc :) still a lot to test.
13:31 iMil morning/evening there
13:31 mike25ro iMil: hey there
13:31 bud Morning iMil.
13:32 iMil is it possible to pass **kwargs from a state to a module?
13:32 iMil I found that example:
13:32 iMil https://github.com/bclermont/states/blob/master/states/git/server/init.sls
13:32 JasonSwindle joined #salt
13:33 QauntumRiff mike25ro: 2 months ago, I was in a similar situation: http://www.reddit.com/r/saltstack/comments/1e12tm/newbie_need_help_with_rolestargeting/
13:33 QauntumRiff My question, some answers, and then my "what I did" are all in there
13:33 HashNuke hey guys, how do I install git-python via saltstack?
13:34 mike25ro QauntumRiff:  thanks for the tip
13:34 iMil HashNuke: via pip ?
13:34 HashNuke iMil: ah everytime I run salt-call, I get "Git fileserver backend is enabled in configuration but could not be loaded, is git-python installed?"
13:34 QauntumRiff :) its about time I answered a question, Since I have asked so many here
13:34 iMil HashNuke: probably an older git-python version
13:34 HashNuke iMil: should I install it that manually via a command or via saltstack?
13:35 iMil HashNuke: is it for the git repo ?
13:35 iMil HashNuke: if yes, install it by hand
13:35 iMil and restart salt-master when it's done
13:35 HashNuke iMil: I have no idea. Whenever I run salt-call I get that error. So decided to look into it.
13:35 JasonSwindle HashNuke:  THis is how I do it
13:35 JasonSwindle https://gist.github.com/JasonSwindle/1a9ac61ff7e5dcb64583
13:35 bud And from pip because, at least on Debian, it doesn't work with the one from packages.
13:35 iMil indeed
13:35 iMil been bitten by that one too :)
13:36 JasonSwindle sorry, I was thinking you said pip-python
13:36 HashNuke JasonSwindle: And that's fine. I found something useful :)
13:36 racooper joined #salt
13:36 JasonSwindle Yes, install everything you can from PIP.  Ubuntu / Deb packages can be WAY out of date
13:36 JesseC If I have this -> /srv/salt/webserver/init.sls. Will the calls in init.sls have paths starting from /srv/salt/webserver or /srv/salt ?
13:36 HashNuke I never want to use the OS packages. They are generally build for the last century
13:36 JasonSwindle I had a HORRIBLE issue because of it
13:37 HashNuke do you guys have your salt states up online on github for reference?
13:37 JesseC those packages normally have bugs and issues backported and repaired so that the distro can stay stable
13:37 mike25ro is there a way to force the pillars to get a refresh? a command that can be run from master to refresh all minions pillar data?
13:37 JesseC if you had everything as the latest version you wouldn't have a stable distro
13:38 iMil HashNuke: there are a *lot* of states examples on github
13:38 HashNuke iMil: I found none installing nginx from the nginx ppa.
13:38 JasonSwindle Using pip state?
13:38 JasonSwindle HashNuke:  I have one
13:39 iMil HashNuke: you didn't search hard
13:39 iMil https://www.google.fr/#sclient=psy-ab&amp;q=inurl:sls+nginx+ppa&amp;oq=inurl:sls+nginx+ppa&amp;gs_l=serp.3...115803.118614.0.118829.9.9.0.0.0.0.108.703.8j1.9.0....0...1c.1.23.psy-ab..9.0.0.TDMrIal5-Do&pbx=1&bav=on.2,or.r_cp.r_qf.&fp=b578430e69a6f924&biw=1358&bih=591&bvm=pv.xjs.s.en_US.seW1cfrvSKg.O
13:39 iMil 1st try
13:39 JasonSwindle HashNuke:  https://gist.github.com/JasonSwindle/d3587b419d790ede6949
13:39 HashNuke JasonSwindle: iMil: Thanks.
13:39 iMil oh np, my google-fu is strong
13:39 iMil :)
13:39 HashNuke JasonSwindle: That's erroring out.
13:40 HashNuke ah no. it's just github.
13:40 JasonSwindle Yeah
13:40 JasonSwindle it is being odd lately
13:40 HashNuke I've been getting too many of those errors lately.
13:41 HashNuke JasonSwindle: you install pip from the OS package?
13:42 JasonSwindle Yes
13:42 JasonSwindle See above
13:43 JasonSwindle https://gist.github.com/JasonSwindle/1a9ac61ff7e5dcb64583
13:43 JasonSwindle I also supply a pip.conf because PIP can time out like crazy
13:44 __Jahkeup__ joined #salt
13:44 HashNuke Wow, I just found out that saltstack handles installation of most required tools/packages easily.
13:44 JasonSwindle Salt does A LOT.
13:44 JasonSwindle It is crazy awesome for being version 0.16.x
13:44 bud JesseC: /srv/salt
13:44 JasonSwindle I don't do the version number game.
13:44 bud mike25ro: you want to refresh pillar from a minion?
13:44 HashNuke JasonSwindle: I was happy that rbenv install is so easy with salt :)
13:45 mike25ro bud: from the master to force the minion to refresh the pillar
13:45 bud salt '*' saltutil.refresh_pillar ;)
13:45 mike25ro bud:  from the master?
13:45 bud Yeap.
13:45 JasonSwindle HashNuke:  I am doing a complex DJANGO + NGINX + UWSGI + POSTGRES clustered setup
13:45 JasonSwindle it is awesome
13:45 mike25ro i did that :)
13:45 mike25ro ok then :)
13:46 HashNuke JasonSwindle: awesome ~!
13:46 bud JasonSwindle: that sounds sweet. I have been having fun setting up a tinc VPN state that uses the mine to autoconnect to other minions that are part of the VPN.
13:47 mike25ro bud:  seems complicated :)
13:47 bud Can't do it in one run though :|. It takes two runs to complete.
13:47 JasonSwindle The peer and publish.publish is awesome
13:47 JasonSwindle bud:  Why so?
13:47 HashNuke so a salt master must always be on a server online, or can it be on my local machine (which i disconnect from the net often)?
13:47 bud JasonSwindle: are you familiar with tinc VPN?
13:48 HashNuke I've been using a masterless minion to learn salt
13:48 JasonSwindle I am not
13:48 JasonSwindle HashNuke:  Vagrant will be your friend
13:49 Odd_Bloke joined #salt
13:49 HashNuke JasonSwindle: Yes I am using vagrant to play with a masterless minion.
13:49 ipmb joined #salt
13:49 bud I am asking because if you know the tinc specifics we can talk the same language, if not I will try my best to explain it.
13:49 HashNuke JasonSwindle: But what I meant was... is salt master like chef, where the chef server has to be hosted somewhere.
13:50 bud Each tinc node has a key pair. I gen them using openssl and do mine.update to push via a custom made module the public key.
13:50 JasonSwindle HashNuke:  My VagrantFile
13:50 JasonSwindle https://gist.github.com/JasonSwindle/4552855c7e7c35069a5e
13:50 JasonSwindle bud:  Tried an overstate?
13:50 JasonSwindle I am going to start work in mine soon
13:50 bud No, I haven't, not just yet.
13:51 HashNuke JasonSwindle: now... need to move all my chef-managed VMs to salt ones.
13:51 JasonSwindle yeah
13:51 bud I don't mind it takes two runs. It's kind of logic to take two runs, because the async/parallel nature of salt.
13:51 bud *of
13:52 __Jahkeu_ joined #salt
13:53 anteaya joined #salt
13:54 [diecast] joined #salt
13:54 bemehow joined #salt
13:56 Katafalk_ joined #salt
13:59 opapo joined #salt
14:01 __Jahkeup__ joined #salt
14:04 HashNuke Is there a way to remove states from a minion? I've installed apache while playing around. Now I want to remove it and install nginx.
14:05 HashNuke I can get a new VM, but I'm just curious if it's possible. because I saw some "absent" stuff mentioned in the docs
14:06 JasonSwindle HashNuke:  http://docs.saltstack.com/ref/states/all/salt.states.pkg.html?highlight=pkg#salt.states.pkg.removed
14:06 JasonSwindle The Docs will be your friends
14:06 LyndsySimon joined #salt
14:07 JasonSwindle Welcome LyndsySimon
14:07 HashNuke JasonSwindle: thanks
14:07 QauntumRiff hashNuke: http://docs.saltstack.com/ref/states/all/salt.states.pkg.html#module-salt.states.pkg  Looks like in addition to pkg.installed there is also pkg.removed and pkg.purged
14:07 LyndsySimon o/
14:07 bemehow joined #salt
14:07 quantumsummers|c joined #salt
14:07 denstark Can you put comments in a state file?
14:08 JasonSwindle denstark:  yes
14:08 JasonSwindle I do it all the time
14:08 JasonSwindle it is YAML, so just do it correctly
14:08 denstark Aha, that's right
14:08 JasonSwindle ONE BIGGy
14:09 denstark Google ws not helpful because I was googling the wrong thing ;)
14:09 JasonSwindle JINJA WILL READ COMMENTS!
14:09 LyndsySimon left #salt
14:09 JasonSwindle So, if you comment out jinja, it will still process it
14:09 denstark Okie doke
14:09 krak3n` joined #salt
14:10 denstark So comments are just marked with a #
14:10 JasonSwindle Yeah
14:10 jeffasinger There are also Jinja comments
14:11 denstark Not jusing any jinja at the moment
14:12 JasonSwindle jeffasinger:  Very true.  I wasted an hour on silly JIJNA
14:13 kallek joined #salt
14:15 jeffasinger Jinja is probably the part of salt that I'm most comfortable with, seeing as I spent 3 months writing Jinja for a website. YAML on the other hand …
14:16 whit joined #salt
14:18 mannyt joined #salt
14:19 alunduil joined #salt
14:19 JesseC Is there a document that goes over specifics of syntax? There's like 3-4 examples in the walkthrough where something is done different every time. Ex. pkg.installed or package name then installed, etc.
14:19 JesseC or name pkg: -installed
14:21 kaptk2 joined #salt
14:22 ze- joined #salt
14:23 ze- anyone knows if it's possible to pass **kwargs via a module.run in a state?
14:23 aat joined #salt
14:23 JesseC and the modules document only goes over command line syntax and not the syntax used in state files..
14:27 jaequery joined #salt
14:32 QauntumRiff this might be a silly question.. I need to match a grain for the RHEL major release.. {% if grains['osrelease'] == '5.x' %}  how do I tell it that I only care about the 5, and not the dot and other number?
14:32 QauntumRiff for example, the epel rpm for RHEL 6.x is different than RHEL5.x
14:33 QauntumRiff can I use a wildcard?
14:34 napperjabber joined #salt
14:34 cnelsonsic joined #salt
14:35 QauntumRiff ie: {% if grains['osrelease'] == '5*' %}
14:38 JesseC QuantumRiff: https://groups.google.com/forum/#!msg/salt-users/DQ3RouJF_9Y/DBqF4IzlabUJ
14:38 Furao qauntumriff: {% if grains['osrelease'].startswith('5') %}
14:38 QauntumRiff sweet.. thanks
14:39 JesseC The guy at the end of that thread also does this: grains['osrelease'][0] == '6'
14:43 Furao the important thing to remember is that in jinja variable are python object/instance, so you can use all python method
14:43 Furao {% set x = {} %}{% x.update({1:2}) %}
14:46 teskew joined #salt
14:48 JesseC Does pkgrepo create the repo if it doesn't exist?
14:48 m_george|away joined #salt
14:54 HashNuke JesseC: AFAIK, it uses the package repository you specify to install the package.
14:54 mgw joined #salt
14:57 StDiluted joined #salt
14:58 QauntumRiff jessec: furao: thanks for the help. I threw that into a pull request on the saltstack/salt-states repo for the epel.sls file :)  hopefully, it will help someone else in the future
14:59 JesseC Thank you, I didn't even know there was a salt-states repo
15:01 mannyt_ joined #salt
15:02 bastion2202 joined #salt
15:02 JesseC What are all of the absent.sls files for/>
15:02 JesseC *?
15:04 m_george left #salt
15:04 bastion2202 Hey guys, I have a strange behaviour  with pgk.installed on gentoo. In the sls I have version: ':9.2' to install the latest version on 9.2. The prob is that is reinstall the package every time
15:08 toastedpenguin joined #salt
15:13 jessep joined #salt
15:14 KennethWilke joined #salt
15:16 conan_the_destro joined #salt
15:17 davidone joined #salt
15:20 diegows joined #salt
15:20 dthom91 joined #salt
15:21 devinus joined #salt
15:23 bud I know it's offtopic but is it worth to deploy LDAP in an infrastructure?
15:24 QauntumRiff bud: YES!
15:24 QauntumRiff well, it depends..
15:24 __Jahkeup__ joined #salt
15:25 QauntumRiff having centralized authentication is nice, to force password changes, disable accounts, etc.  I'm not sure if salt can manage user passwords, and I'm sure its not as simple as users typing "passwd" on any server to change their password globally
15:25 bud QauntumRiff: I am thinkins more of long term benefits.
15:25 QauntumRiff we also manage sudo with ldap
15:26 QauntumRiff before ldap, we kept finding all sorts of servers, where accounts were 'left behind' for people that had not been at the company for years...
15:26 bud Now the infrastructure has only a handfull of users but it's growing steadily and we could do lot of things from LDAP, like allowed hosts to SSH, etc.
15:27 bud That's another plus to have LDAP for QauntumRiff.
15:27 HashNuke I created an nginx state and I get an error when I try to watch files - https://gist.github.com/HashNuke/6132396
15:27 QauntumRiff were a centos shop.  if I were to do it again, I would probably use freeIPA, since it also includes certificates, ssh keys, etc
15:27 bud Guess I will go on and create an OpenLDAP state then :).
15:28 QauntumRiff bud: thats actually my next project for this afternoon.  figure out how to get a vanila system to ldap authenticate with salt. So we don't have to run our custom shell scripts on new machines anymore
15:28 bud HashNuke: pkgrepo needs require_in for packages that use it.
15:28 QauntumRiff trying to automate all of our previously semi-automated processes.
15:29 akoumjian HashNuke: The watch command does not watch files themselves, it watches other file.* states, for example file.managed
15:29 bud HashNuke: And you don't need require package in nginx.
15:29 forrest joined #salt
15:30 bud QauntumRiff: sounds like a challenge. You want the system to authenticate the salt minion to the Master?
15:30 bud Or the users than SSH in and stuff.
15:30 QauntumRiff no, just users and sudo
15:30 bud So, kind of an LDAP client state.
15:31 QauntumRiff well, maybe in the future, but right now, we have a wiki page full of scripts to run.. directories that are different depending on what the machines are for, etc.  I'm replacing most of that with salt and pillars
15:31 HashNuke I was under the impression that whenever a file is changed outside of saltstates, even then the service is restarted, if the files are being watched. Was I wrong?
15:31 QauntumRiff yeah. On RHEL 5, its an upadte to the ldap.conf, nssswitch, and right now, and authconfig command that automates a few other things
15:33 akoumjian HashNuke: You are wrong. Salt does not have a built in file watcher daemon. require/watch statements refer to other states, such as files managed by salt. When you run state.highstate, a service watching a file state will restart if the file.managed state modifies the file
15:33 bud No HashNuke, I think you've got all right, but the files need to be under Salt's control.
15:33 bud Way better explanation akoumjian, thanks!
15:33 chrisgilmerproj joined #salt
15:34 HashNuke akoumjian: bud: Thanks a lot. That cleared a lot of questions I had.
15:34 HashNuke Removed the watch lines from my sls file.
15:35 akoumjian HashNuke: You're welcome.
15:35 UtahDave joined #salt
15:35 HashNuke Bud: you mentioned that pkgrepo requires require_in in packages that use it. Can you point out exactly where in my gist?
15:37 napperjabber joined #salt
15:37 HashNuke Whoa ~! I get a 404 from nginx now. Which is good actually. Something is working.
15:37 alunduil joined #salt
15:39 JesseC Would this be the proper way to stop and disable iptables in a state file? https://gist.github.com/JesseC-/6132555
15:44 bud Just a second HashNuke - https://gist.github.com/valentinbud/6132571
15:46 HashNuke bud: thanks ~! now I get it.
15:46 JasonSwindle HashNuke:  https://gist.github.com/JasonSwindle/d66948b3b301a6edaf69
15:47 JasonSwindle My NGINX SLS
15:47 HashNuke JasonSwindle: Yes I made mine out of it when you posted earlier :)
15:48 bud You're wellcome HashNuke. The first try I wrote a pkgrepo state I did it just like you and it crashed on me.
15:48 HashNuke JasonSwindle: But I made some changes to learn.
15:48 JasonSwindle Sure sure
15:48 lempa joined #salt
15:49 JasonSwindle I group everything I can under the ID
15:51 HashNuke JasonSwindle: Ya it makes sense.
15:51 JasonSwindle HashNuke:  Do you have Visio or OmniGraffle?
15:51 HashNuke JasonSwindle: Ah no. why?
15:52 HashNuke JasonSwindle: I have Sketch and Pixelmator.
15:52 kermit joined #salt
15:52 JasonSwindle It helps to map out what you are doing
15:52 JasonSwindle auser, who is not on at the moment is working on this
15:52 JasonSwindle http://www.saltstat.es/posts/dependencygraph.html
15:52 HashNuke for mapping I just use markdown. and I keep indenting stuff. I know it's mundane way of doing mapping :)
15:52 HashNuke Ari again ~! :)
15:53 __Jahkeup__ ari
15:53 __Jahkeup__ oops sorre
15:55 SEJeff_work PEP8 was just updated. Any serious python dev should read this: http://hg.python.org/peps/rev/fb24c80e9afb
15:57 jalbretsen joined #salt
15:58 oz_akan_ joined #salt
15:59 JasonSwindle oz_akan_:  WELCOME!
16:01 UtahDave thanks, SEJeff_work.  reading it right now
16:02 JasonSwindle UtahDave:  WELCOME!
16:02 JasonSwindle EVERYBODY gets a welcome!
16:02 JasonSwindle lol
16:03 SEJeff_work UtahDave, You know, the contents of that post should probably make their way into the official docs. That is much better than the official docs
16:04 JasonSwindle UtahDave:  Salt-Cloud question.  I want salt bootstrap to use git tags, and the tag *I* pick.  How can I do this?
16:07 UtahDave SEJeff_work: I like what i've read so far.
16:07 UtahDave JasonSwindle: yeah, you can.  just a sec
16:08 mikedawson_ joined #salt
16:08 faldridge joined #salt
16:09 UtahDave JasonSwindle: So salt-cloud uses salt-bootstrap to install the minion on the new vm. https://github.com/saltstack/salt-bootstrap
16:09 KennethWilke is there some tricks to ordering file_roots? in this thing i'm doing i have the 'base' state in /srv/salt and a 'mysql' state in gitfs. when git is first in fileserver_backends salt doesn't seem to find my base state, though it does read the top file
16:09 KennethWilke trick to ordering fileserver_backends*
16:09 UtahDave JasonSwindle: So you need to pass in "script_args" in salt-cloud to tell the bootstrap to use git <yourtag>
16:09 UtahDave http://salt-cloud.readthedocs.org/en/latest/topics/misc.html#deploy-script-arguments
16:10 JasonSwindle Looking
16:10 KennethWilke conversely, when i have roots before git it does seem to work properly now
16:10 UtahDave KennethWilke: yeah, it looks in the order their specified
16:10 jkleckner joined #salt
16:11 KennethWilke well i don't have base in git, but it's in roots it looks like it's not actually getting past the git stuff
16:13 JasonSwindle UtahDave:  Testing it out know
16:13 bud I have bas in git and defined a base in file also. fileserver_backend setup with git before roots.
16:13 JasonSwindle We can have our whole deploy node + salt deploy done in 10 to 15 minutes
16:13 bud If the file is only in file_roots doesn't work.
16:14 JasonSwindle UtahDave:  Testing now; We can have our whole deploy node + salt deploy done in 10 to 15 minutes
16:14 KennethWilke bud: so you're seeing the same thing it looks like?
16:15 KennethWilke bud: so if you have it in git, does it combine both base states?
16:15 UtahDave KennethWilke: when looking for an sls file it should search in the order defined in fileserver_backends, and then in file_roots
16:15 JasonSwindle UtahDave:  Is the GIT_FS getting any love?
16:15 UtahDave back soon. Getting called into a meeting
16:15 JasonSwindle kk
16:16 KennethWilke UtahDave: that's what i want it to do :(
16:19 al5d joined #salt
16:19 JesseC Are state files ran in a linear fashion? I have a list of dependencies that I'm installing, but they need the required repos setup beforehand. If I include repos in my state file and put my dependencies installer after than include, then it won't run until the repos state is finished right?
16:19 al5d hi there
16:19 bud KennethWilke: I don't know, I just have some mine states full of tests that I don't want in the repo.
16:20 KyleG joined #salt
16:20 KyleG joined #salt
16:20 bud JesseC: right you can require git.latest repos in states.
16:20 JasonSwindle JesseC: They do not, you will need to setup your requires correctly
16:20 TheRealBill JesseC: no, if you want order you need to use order or requires
16:20 KennethWilke JesseC: what the other spammed you with
16:20 JesseC Can you do a require on a state file?
16:20 KennethWilke yar!
16:21 JesseC or do I need to setup requires on each individual repo file in /etc/yum.repos.d?
16:21 KennethWilke JesseC: http://docs.saltstack.com/ref/states/requisites.html
16:21 bud But is it wrong to include another sls, the one with repos and require the repos when you need them via requisite statement?
16:21 JasonSwindle JesseC:  https://gist.github.com/JasonSwindle/d66948b3b301a6edaf69
16:21 KennethWilke bud: i think it's fine
16:22 JasonSwindle require / require_in
16:22 blee_ joined #salt
16:22 JasonSwindle I require stuff in other states ALL the time
16:22 KennethWilke JasonSwindle: you do require a lot... so needy
16:22 JasonSwindle Oh yes, giveme
16:23 TheRealBill KennethWilke: truth
16:23 napperjabber_ joined #salt
16:23 bud Happy Salting, see you tomorrow.
16:24 KennethWilke goodbye bud!
16:25 JasonSwindle Later!
16:25 JesseC Ok. So to require an entire state. It would simply be -> - require: - state: repos
16:25 al5d i was wondering, how is it possible to manage testing new state formulae within a production environment?
16:26 JesseC The only examples in requisites are using files and pkg
16:26 QauntumRiff This allows me to update my minions, to force some logging options, but the minion restarts, and does not answer back. it forces me to run salt manually twice on a new server. First to do update the minion, and then it restarts and does not answer, and second, to enforce all the other states.
16:26 QauntumRiff https://gist.github.com/briffle/6132956
16:26 QauntumRiff is there a way to do this, and force the minion to restart at the end, after its answered back or something?
16:27 JesseC or.. would it be require: - module: modulename
16:28 devinus joined #salt
16:28 KennethWilke JesseC: it's generally you'll do  - require:\n  - <state_module>: <resouce_id>
16:28 KennethWilke so like they should requires: pkg: vim
16:28 KennethWilke means something named vim needs to be handled by the pkg state module
16:29 JesseC I understand the pkg example
16:29 JesseC but I don't need to require a pkg, I need to require a module that creates some repos
16:29 JasonSwindle module or state?
16:29 KennethWilke what state module are you using to create the repos?
16:30 JesseC lol, let me gist this so it's easier to understand what I'm attempting to do, it's possible I'm going about it the wrong way entirely
16:30 KennethWilke lol sure thing
16:30 JasonSwindle Yep
16:30 JasonSwindle Salt is awesome, but wrapping your head around it does take a little time
16:31 troyready joined #salt
16:31 __Jahkeu_ joined #salt
16:31 al5d JasonSwindle: yep, especially with a puppet / chef background...
16:32 KennethWilke i found it easy to move to salt from chef
16:32 JasonSwindle al5d:  I never had either of these.  I tried Chef for a bit, but not too deep
16:32 KennethWilke but maybe thats because i wanted to be away from chef so badly
16:32 dthom91 joined #salt
16:32 KennethWilke bah other way around there*
16:32 KennethWilke from chef to salt
16:32 KennethWilke bah, nightmares
16:34 al5d one thing i don't quite manage to do with salt, is testing changes in salt states on production environments, where it's better not to disrupt anything
16:34 forrest JesseC, if you create a good example for the repo stuff you should make sure it gets on saltstarters.org
16:34 HashNuke Yeeehaaa ~! nginx setup successfully. Thanks to bud, akoumjian, auser, JasonSwindle and a lot others on this channel.
16:34 KennethWilke gg HashNuke
16:34 JasonSwindle HashNuke:  Any time
16:34 JasonSwindle :)
16:34 HashNuke Now I pretty confident of writing states. Even with a few screwups :)
16:34 JasonSwindle HashNuke:  Are you on Win or OSX?
16:34 KennethWilke we've all shot ourselves in the foot before i am sure!
16:35 JasonSwindle Or Lin?
16:35 HashNuke JasonSwindle: OSX.
16:35 al5d i must be missing something about salt environments, because i can't find a way to have some sort of staging environment for tests to be deployed
16:35 JasonSwindle I found Sublime 3 helps A LOT on YAML
16:35 forrest VertigoRay, did you ever figure out the OSX install?
16:35 HashNuke JasonSwindle: but my vagrant and anything I host on is always linux. Ubuntu mostly. But I did try debian a few times.
16:35 HashNuke JasonSwindle: ya I just set sls files to always open in yaml syntax mode in sublime :)
16:35 aleszoulek joined #salt
16:35 JesseC forrest: I'm doing my repos manually using files because I couldn't figure out pkgrepo
16:35 JasonSwindle Awesome, that helps a lot
16:36 KennethWilke al5d: http://docs.saltstack.com/topics/tutorials/states_pt4.html#environment-configuration might help you out
16:36 forrest I'm still too bitter after buying sublime2 to support sublime3 :\
16:36 KennethWilke al5d: though i still find it odd myself
16:36 JasonSwindle JesseC:  What does your SLS look like?
16:36 JesseC forrest: also because I needed some customization to the repo files, but I'll upload anything that will help as long as I'm not telling someone how to do it the wrong way
16:36 JesseC JasonSwindle: almost got the gist done
16:36 KennethWilke oh JesseC i think i've done the same on ubuntu
16:36 JasonSwindle kk
16:36 KennethWilke i will try to gist it as well
16:37 forrest JesseC, yea I understand
16:38 KennethWilke JesseC: maybe this is similar to what you're going for: https://gist.github.com/KennethWilke/6133060
16:40 KennethWilke JesseC: i added packages_alt.sls as well there to highlight how the top level represents the ID for that resource
16:40 Lue_4911 joined #salt
16:43 JesseC Ok, here is what I have so far. https://gist.github.com/JesseC-/59d43a574cf799619c51
16:43 forrest can you paste your epel.repo file?
16:43 forrest the example for pkgrepo should really have a more thorough example.
16:44 JesseC my epel.repo file is the one generated when you install the rpm
16:45 KennethWilke JesseC: so it looks like you'd do '- require:\n  - file: /etc/yum.repos.d/epel.repo' perhaps inside of 'mypkgs'
16:45 jeffasinger joined #salt
16:45 KennethWilke that way the you know that the file state module has handled  /etc/yum.repos.d/epel.repo before pkg handles 'mypkgs'
16:45 forrest JesseC, above you said that you have some customization to the repo files though, what is that?
16:46 alunduil joined #salt
16:46 JasonSwindle JesseC:  What is in epel.repo?
16:46 JesseC forrest: yes, just this: includepkgs=php-mcrypt libmcrypt mhash php-mbstring php-php-gettext mycrypt-devel libmcrypt-devel
16:46 JesseC I only allow specific items from epel
16:47 berto- joined #salt
16:47 JesseC oh just realized you can do individual files in gist. ;)
16:48 JesseC here is my epel.repo https://gist.github.com/JesseC-/8866d408ac7195976849
16:48 JesseC I have also add nginx and percona repos in the same manner
16:49 JesseC KennethWilke: Does that mean there's no way to require an entire state file?
16:49 JesseC KennethWilke: so I could just require each of the repo files
16:50 JesseC should*
16:50 krak3n` joined #salt
16:52 UtahDave JesseC: Yeah, you can require an entire state file
16:52 JesseC UtahDave: How is that achieved?
16:53 forrest JesseC, ahh yea I can see how that would be a problem now, doesn't seem like extra options are supported.
16:54 m_george|away joined #salt
16:54 UtahDave JesseC: https://github.com/saltstack/salt/issues/5098#issuecomment-18266509
16:55 forrest JesseC, could you mark the includepackages option as a comment, and then just not throw the # out in front?
16:55 JesseC Thanks UtahDave!
16:55 UtahDave you're welcome!
16:56 JesseC forrest: not sure, but you could also use file.sed after the repo is created to do it as well
16:56 JesseC forrest: I'm not sure what the best method is, but I like versioning the repos I use in dev anyways rather than pulling the latest version, lets me know everyone is using the same thing
16:56 JasonSwindle UtahDave:  I had no luck at all with the link you sent trying to use it in the map file.
16:56 m_george left #salt
16:57 UtahDave can you pastebin your map?
16:58 JesseC UtahDave: This would make dependencies require the repos sls correct? https://gist.github.com/JesseC-/8d9356cfaae14d544b8f
17:00 JasonSwindle UtahDave:  SUre
17:00 UtahDave JesseC: No, I don't think that would work.  You'd need to include the repos sls into your dependencies.sls  and then have one or more of your states require the entire repos.sls
17:00 JasonSwindle UtahDave:  https://gist.github.com/JasonSwindle/e3503cddb2d47e64397a
17:01 juanlittledevil joined #salt
17:02 JasonSwindle I took out the script part
17:02 JasonSwindle let me add on what I had
17:03 faldridge joined #salt
17:03 JesseC UtahDave: so more like this -> https://gist.github.com/JesseC-/8d9356cfaae14d544b8f
17:04 UtahDave JesseC: yes, exactly that!
17:04 JesseC awesome, thanks again
17:05 JesseC to everybody
17:05 forrest UtahDave, so the preferred way would be to simply include the repo file itself, as opposed to having odd or one off options as part of the pkrepo state?
17:07 __Jahkeup__ joined #salt
17:08 UtahDave JasonSwindle: Try something like this: https://gist.github.com/UtahDave/a6d8796bd930095dbd48
17:08 mike25ro2 joined #salt
17:08 JasonSwindle Let me try that
17:08 UtahDave forrest: I'm not exactly sure what you mean. I've missed a lot of the above discussion.
17:09 forrest so JesseC had this gist
17:09 forrest https://gist.github.com/JesseC-/8866d408ac7195976849
17:09 forrest He's basically been forced to do a work around, since he requires that 'includepkgs' line, so he has to include the file, instad of using the pkgrepo state
17:09 forrest Is that the appropriate way to handle this instance?
17:10 QauntumRiff for packages.. if I have require: pkg: <blah> does <blah> have to be in a salt-state? or can it just exist on the system?
17:10 mike25ro left #salt
17:10 JasonSwindle UtahDave:  Testing now
17:10 mike25ro joined #salt
17:11 jschadlick joined #salt
17:11 UtahDave forrest: I'm not exactly sure. pkgrepo.managed  can take **kwargs.  Can the includepkgs be passed in as a keyword argument?
17:11 UtahDave QauntumRiff: Salt doesn't look on the file system for requires. It only looks at the results of other states.
17:11 forrest UtahDave, I don't know.
17:11 UtahDave QauntumRiff: so you have to have salt managing that package first.
17:12 forrest UtahDave, I would imagine that it would work. JesseC, if you have time can you try it in your configuration?
17:13 auser joined #salt
17:13 QauntumRiff thanks.. I'm actually reading through some of your other conversations above.. I basically only want to install some packages, if the EPEL or Oracle repo's are installed on the server
17:14 juanlittledevil this is how I'm handling my repos: http://pastebin.com/XfvVc8XH
17:14 bastion2202 joined #salt
17:15 afx_hero joined #salt
17:15 JesseC forrest: Sure. This is my first setup so give me a little while longer to finish the other pieces and I'll test it all and let you know.
17:17 QauntumRiff juanlittledevil: I didn't realize I could just use source, and tell it where the epel rpm's are.. that is handy..
17:17 forrest JesseC, how about something like this: https://gist.github.com/gravyboat/a989c089ab2f84cbd6c5
17:18 JasonSwindle UtahDave:  WIth the example map file, the nodes built but I am unable to match by grain
17:18 mjulian How do I set specific permissions on a file inside of a state? (the file will already be there and isn't being managed from the Salt file server)
17:18 juanlittledevil QauntumRiff: as long as you have an RPM package for the repo. I use this for my locally hosted repo. so far it's working just fine.
17:18 dthom91 joined #salt
17:18 JesseC forrest: I can try that, but afaik pkgrepo doesn't include those lines
17:18 JasonSwindle Mabe
17:19 mjulian I'm able to do it on directories by setting '- directory', but if I remove that, it complains there's no function.
17:19 JasonSwindle *maybe let me test more
17:19 forrest JesseC, right but it takes kwargs
17:19 forrest so it technically should be able to take any number
17:20 auser joined #salt
17:20 Xeago joined #salt
17:20 auser hey all
17:21 forrest JesseC, if that works you or I should update the example in the documentation to provide a second one where you're doing a bunch of crazy stuff.
17:23 hazzadous joined #salt
17:23 JasonSwindle auser:  Hey
17:23 pdayton joined #salt
17:24 austin987 joined #salt
17:25 QauntumRiff after futzing with this for a long time, It just dawned on me that we install salt-minion via epel, so the whole "setup epel" thing is kind of redundant :)
17:25 forrest heh
17:26 mike25ro what is epel?
17:26 QauntumRiff redhat extra repositories
17:26 forrest extra packages for enterprise linux
17:26 QauntumRiff I still need the Oracle ones though, so not all work is wasted
17:26 mike25ro ah right - i am a debian dude
17:26 * forrest gasps
17:26 forrest that was directed at QauntumRiff
17:27 QauntumRiff kind of like universe, multiverse, etc.
17:27 JasonSwindle UtahDave:  https://gist.github.com/JasonSwindle/e3503cddb2d47e64397a <- I cannot highstate on these nodes
17:27 mike25ro QauntumRiff: thanks
17:27 JasonSwindle but yet, I can do a grains.item GRAIN
17:27 JasonSwindle and that returns
17:31 Ryan_Lane joined #salt
17:31 UtahDave JasonSwindle: do they match anything in your top.sls?
17:31 JasonSwindle They should
17:32 JasonSwindle Before I added the one line, it worked
17:32 Xeago joined #salt
17:32 JasonSwindle going to keep poking around
17:32 JasonSwindle re-deploy from git
17:32 JasonSwindle reboots
17:32 JasonSwindle all of that jazz
17:33 whit joined #salt
17:33 JasonSwindle UtahDave:  it is not even doing * on top.sls
17:34 UtahDave JasonSwindle: can you execute one single sls file?    salt 'theminion' state.sls yourstate
17:35 JasonSwindle Let me try again
17:35 JasonSwindle UtahDave:  Rebuilding the nodes
17:37 __Jahkeu_ joined #salt
17:37 __Jahkeup__ joined #salt
17:39 bemehow joined #salt
17:42 JesseC Is there a way to reboot using salt and then continue when the system comes backup automatically?
17:42 auser reboot the machine JesseC?
17:42 JesseC an example would be for disabling selinux
17:42 JesseC auser:  yes
17:43 forrest JesseC, I've been looking at using http://docs.saltstack.com/ref/python-api.html for that
17:43 JasonSwindle UtahDave:  It will match, but not hs
17:43 JesseC forrest: I'll give that a look
17:43 JesseC forrest: thanks
17:44 forrest I wrote a script yesterday to move a file (pulling a server out of an LB), and then tail the log for specific things. I didn't get far enough to actually reboot once certain conditions were met, and then continue after that though
17:44 forrest yea np
17:45 forrest My big concern as well is how to get the logic in there regarding monitoring the restart, I've considered just having the script wait, then try to hit the server, and if that fails after a specific timeout, it loops back
17:45 JasonSwindle UtahDave:  [INFO    ] Running command "sshpass -p 9zh2WVuWNixR ssh -oStrictHostKeyChecking=no -oUserKnownHostsFile=/dev/null -oControlPath=none root@xx.xx.xx.xx '/tmp/deploy.sh -c /tmp/ git v0.16.1'"
17:46 UtahDave JasonSwindle: what's the output of    salt \* test.versions_report   ?
17:47 tempspace if you all want to run a command within a virtualenv, what's the best way to do that?
17:47 JasonSwindle UtahDave:  the staging nodes got                Salt: 0.16.1
17:47 devinus joined #salt
17:47 JasonSwindle vs the others with 0.16.0
17:48 berto- joined #salt
17:49 wpc joined #salt
17:49 JasonSwindle So, it setup
17:49 JasonSwindle correctly
17:49 timl0101 joined #salt
17:50 ipmb joined #salt
18:03 faldridge joined #salt
18:04 Ryan_Lane joined #salt
18:05 juanlittledevil OMG why was I not using pillars before!! they are awesome!
18:07 faldridge joined #salt
18:08 napperjabber joined #salt
18:09 bemehow joined #salt
18:10 m_george|away joined #salt
18:13 auser :) juanlittledevil
18:15 dthom91 joined #salt
18:18 druonysuse joined #salt
18:18 druonysuse joined #salt
18:24 tsheibar joined #salt
18:27 JesseC the whitespace in salt is supposed to be tabs and not spaces, or does it matter?
18:27 faldridge joined #salt
18:28 JesseC the whitespace in an sls
18:28 tsheibar hoping to get this into 0.16.2: https://github.com/saltstack/salt/issues/6427
18:28 whit joined #salt
18:29 tsheibar my company is evaluating Salt and getting this in the latest version will certainly make their evaluation smoother :)
18:29 m_george left #salt
18:29 KennethWilke JesseC: i think yaml wants spaces
18:29 KennethWilke at least i've had issues with tabs before
18:29 JesseC KennethWilke: Any idea what's wrong with this then? https://gist.github.com/JesseC-/a304b1810fe251805cb0
18:31 JesseC I added a comment with the error I get on it
18:31 KennethWilke i'm really not sure if this matters, but i usually use two spaces for the indentation
18:33 JasonSwindle JesseC:  Try giving it an ID
18:33 jeffasinger YAML does not want tabs, it wants spaces
18:33 druonysuse joined #salt
18:33 druonysuse joined #salt
18:33 JasonSwindle two spaces for the win
18:33 johntron joined #salt
18:33 johntron left #salt
18:33 QauntumRiff I have one last pesky question for the day: I have a subdirectory structure I want to build on a new server.. is it possible to do something similar to "mkdir -p /path/to/dir"  I have tried  listing it with file.directory and setting 'makedirs: True' but they are not being created
18:34 tsheibar QauntumRiff: cmd.run ?
18:34 JesseC JasonSwindle: same error even when adding an id
18:35 qba73 joined #salt
18:35 juanlittledevil you can use the resurse option to file.directory
18:36 tsheibar cmd.run: \ - name: mkdir -p /your/awesome/path
18:36 juanlittledevil sorry nevermind. was thinking of somethign else.
18:36 QauntumRiff i thought the recurse was for permissions
18:36 KennethWilke QauntumRiff: i thought makedirs was supposed to do what mkdir -p does
18:36 JasonSwindle JesseC:  https://gist.github.com/JasonSwindle/81b736eec20564050389
18:37 QauntumRiff KennethWilke: That was how I interpreted it as well, but at least so far, it does not seem to be creating the structure for me. they keep coming back red
18:37 sinh joined #salt
18:38 tsheibar QuantumRiff: what's that error message?
18:38 tsheibar *QauntumRiff
18:38 JesseC JasonSwindle: same error
18:38 juanlittledevil QauntumRiff: this totally works for me: http://pastebin.com/YqwqjEJy
18:38 QauntumRiff I don't actually seem to get an error
18:38 JesseC expected <block end>, but found '<block sequence start>'
18:39 jeddi joined #salt
18:39 mannyt joined #salt
18:40 JasonSwindle Why are you doing a yum update?
18:40 QauntumRiff I see this in debug mode, at the end, in red:
18:40 QauntumRiff local: Name: /path/to/dir - Function: file.directory - Result: False
18:40 JasonSwindle Salt will do apt-get update / yum update after adding a repo.
18:41 tsheibar can you pastebin your sly?
18:41 tsheibar *sls
18:41 QauntumRiff nevermind.. :)
18:41 Nazzy joined #salt
18:41 Nazzy joined #salt
18:41 JesseC JasonSwindle: I noticed when I installed it, but just trying to learn in general on how it all works
18:41 tsheibar ??
18:42 QauntumRiff I'm setting its owner and group to ones that do not yet exist.. /facepalm
18:42 JesseC JasonSwindle: I added extra spaces and it seems like its kind of working now, moving onto the next error
18:42 tsheibar ha...pebcak errors are frustrating, but easily fixed
18:42 JasonSwindle kk
18:42 JasonSwindle I never use -include
18:43 JasonSwindle Never have seen a point to it IMHO
18:43 bhosmer joined #salt
18:43 JesseC pkg.installed doesn't seem to like require
18:44 mannyt joined #salt
18:45 JasonSwindle JesseC:  Eh?
18:45 JesseC JasonSwindle: Nothing. :) It was because I was missing more spaces after I doubled up on everything earlier
18:45 JesseC seems like I just need to get my whitespace down
18:46 JasonSwindle What text editor are you using?
18:46 JesseC nano at the moment
18:46 JasonSwindle shun
18:46 JesseC I'll bring it into an idea later and configure my tabs for spaces
18:46 JesseC lol
18:46 JasonSwindle There is a salt VIM
18:46 JesseC just testing righ tnow
18:46 JasonSwindle or sublime
18:46 JasonSwindle anything that has some power to understand YAML and spaces
18:49 JasonSwindle I was in your same boat once, and moved to sublime
18:51 JasonSwindle JesseC:  The Salt Vim did give me a bit of issues
18:51 JasonSwindle It kept trying to auto format every
18:52 JesseC Eh. I'm down to this now. -> Name cmd.run in sls lnpp.dependencies is not a dictionary
18:54 JasonSwindle can you paste the whole warning?
19:01 david_a joined #salt
19:01 alunduil joined #salt
19:02 alunduil Does anyone know how to specify a testing version of an ebuild to be installed via the ebuild state?  Is there another state for manipulating keywords or similar?
19:05 diegows joined #salt
19:07 helderco joined #salt
19:07 forrest JesseC, were you able to try that repo sls?
19:11 devinus joined #salt
19:16 faldridge joined #salt
19:21 cfortier joined #salt
19:22 [diecast] wanting to use a pillar inside a jinja template, but says dict object has no attribute
19:22 forrest [diecast], can you toss a link out to the related files?
19:23 [diecast] http://pastebin.com/H5pRUNkA
19:23 [diecast] thats my jinja template
19:23 Xeago joined #salt
19:24 cfortier hi everyone. i was just searching through the history. iptables as a state are not implemented yet?
19:24 [diecast] cfortier http://docs.saltstack.com/ref/modules/all/salt.modules.iptables.html
19:25 SEJeff_work cfortier, It isn't in 0.16, it will be in 0.17
19:25 [diecast] forrest did you need to see another file?
19:25 forrest [diecast], unless my jinja is just terrible, that looks fine, can we see your pillar
19:25 cfortier i read that, and I was able to run it through the cli. but i'd like to create a state file with specific ports open. that won't be available until .17?
19:27 [diecast] forrest well its on me this time… the pillar data wasn't in my top file for that host =/
19:27 [diecast] all should be good now
19:27 forrest [diecast], oh cool, nice catch!
19:27 [diecast] more like rookie error ;P
19:28 [diecast] thanks for the assist
19:28 forrest yea np, I am still new myself, so trying to find that stuff helps me out as well.
19:31 berto- joined #salt
19:33 bemehow joined #salt
19:44 JesseC forrest: not yet, I had to go to a meeting
19:44 forrest JesseC, meetings, where work doesn't get done at work!
19:44 JesseC forrest: I'm also getting this error I'm going to try and work through now. -> Name cmd.run in sls lnpp.dependencies is not a dictionary
19:45 forrest did you provide the gist for that earlier?
19:45 JesseC I didn't
19:45 JesseC Had to walk out as soon as I posted it here
19:45 forrest gotcha
19:46 UtahDave joined #salt
19:46 JesseC https://gist.github.com/JesseC-/eaeecc9bcd17241bd01c
19:46 JesseC I tried adding IDs above the cmd.run but it didn't make a difference
19:46 juanlittledevil joined #salt
19:47 jschadlick joined #salt
19:47 JesseC nm, ids worked, next error. :D
19:47 forrest oh cool
19:49 forrest so why are you running yum update -y?
19:49 forrest instead of doing pkg.latest for whichever ones should be updated.
19:53 devinus joined #salt
19:53 whiskybar joined #salt
19:54 david_a joined #salt
19:56 auser Hey UtahDave, does the reactor system accept compound matching?
19:56 Lue_4911 joined #salt
19:59 Katafalkas joined #salt
20:00 auser anyone?
20:01 auser erp, anyone know if the reactor system can take compound matches?
20:01 forrest Not sure
20:01 JesseC forrest: it's just there for testing
20:01 auser I thought they could, but can't seem to get it to actually work
20:01 mike25ro auser:  sorry .... i am a newbie mysel
20:01 mike25ro f
20:01 auser <~ embarrassingly not a newbie
20:01 forrest JesseC, ok that makes more sense, was cringing at the thought of a yum update -y in production
20:01 JesseC forrest: I'd like to get the commands working in general so I can move on, but they can't resolve the sls require
20:01 auser just being lazy and don't want to dig into the source I guess
20:01 JesseC forrest: none of this is for production, I'm using salt with vagrant to spin up development vms
20:02 forrest ahh ok
20:02 forrest what error was it throwing for the sls?
20:02 JesseC The following requisites were not found: require: {'sls': 'lnpp.repos'}
20:02 faldridge joined #salt
20:03 forrest so that seems like your include isn't working properly
20:03 JesseC the include ran fine
20:03 JesseC and all of the files its supposed to create are create
20:03 JesseC created*
20:03 forrest oh
20:03 forrest that's weird then
20:03 forrest lol
20:03 Ryan_Lane joined #salt
20:03 JesseC I'll gist the entire log
20:03 forrest sounds good
20:03 erezool joined #salt
20:04 dthom91 joined #salt
20:04 forrest If we can't figure it out hopefully someone around can
20:04 erezool Hey all!
20:04 forrest hey
20:04 erezool i'm interested to remove any package of openjdk
20:04 JesseC Here it is: https://gist.github.com/JesseC-/886b86c90f0a31b4d405
20:04 krissaxton joined #salt
20:04 erezool i'm trying pkg.purge
20:04 erezool - name: openjdk*
20:05 erezool and that doesn't work.
20:05 JesseC My sed failed because I put enabled over enforcing so that can be ignored.
20:05 forrest ok JesseC
20:05 forrest erezool, I don't know for sure if you can use a wildcard like that
20:05 erezool i've tried some other patterns like .* and that doesn't work as well..
20:05 erezool would appreciate any creative idea!
20:06 drawsmcgraw I'm using the 'watch' requisite in a state dec. Even though the watched state is *not* returning any value for 'changed', the state that's doing the watching runs anyway
20:06 erezool thanks forrest, i validated i cannot..
20:06 drawsmcgraw Do I misunderstand how to use 'wait' ?
20:06 forrest you've confirmed that just using pkg.purge works right erezool? If you pass it a single value?
20:06 erezool i'm very new to saltstack and trying my first thing to remove any openjdk package and install java..
20:06 erezool not easy..
20:07 erezool yes, i confirmed that forrest
20:07 it_dude joined #salt
20:07 forrest We'll have to wait for someone with more knowledge then erezool, I don't know about wildcard usage for that sort of thing
20:07 UtahDave auser: Yeah, the reactor should be able to take compound matches
20:07 drawsmcgraw erezool: You could just install Java and then update-alternatives instead, yeah?
20:08 auser UtahDave: hm, can I not call highstate in the reactor then?
20:08 erezool that's correct drawsmcgraw..
20:08 erezool have't researched how to change alternatives, but that would be my last resort..
20:08 UtahDave JesseC: change that to  - require:\n   - lnpp
20:08 UtahDave auser: Yeah, you can.  cmd.state.highstate
20:09 auser that's what I thought
20:09 forrest - lnpp.repos UtahDave? Or do you not even have to include that
20:09 auser 'cept highstate isn't running
20:09 JesseC UtahDave: even though its here -> /srv/salt/lnpp/repos ?
20:09 erezool i've seen another module/state called pkgng that has that capability..
20:09 drawsmcgraw erezool: update-alternatives --install /usr/bin/java java /path/to/java/install 1  <-- is one way that I know of
20:09 auser Actually, to narrow it down, I'm just doing a file.touch, that's not running on the compound matcher
20:09 auser do I need to specify I'm doing a compound match?
20:09 erezool but it seems my installation (vagrant) is missing it and i'm not yet advanced enough to understand it all..
20:09 kstaken joined #salt
20:10 UtahDave auser: yes, you do
20:10 auser ah
20:10 auser well, that'd do it
20:10 auser http://pastie.org/private/trfebduvzudatiu2pmvng
20:10 erezool i know how to run it from command.. i'm kind of interested to stick to the modules as i want support for multiple distros..
20:10 auser ahh
20:11 auser you can bet your bottom dollar a new post on this will go up next week
20:11 UtahDave auser: basically, anywhere in Salt where you're using a matcher that's not the default globbing matcher, you have to specify which matcher.  Except in the overstate, which just uses the compound matcher for everything, I believe
20:11 auser just saying it 'out loud' helped resolve that for me
20:11 erezool could i use some recursion to handle that maybe ?
20:14 auser UtahDave: doesn't seem to make a difference even specifying the master :/
20:14 auser matcher*
20:15 erezool any further ideas?
20:15 drawsmcgraw UtahDave: When you say that's correct, do you mean that's the correct behavior of 'wait' ?
20:15 drawsmcgraw Or did you mean my comment about update-alternatives?
20:16 UtahDave auser: - match: compound          does that work?
20:16 auser naw, that's what I have in there
20:16 auser it works with the default glob, unfortunately that's not the behavior I want
20:17 Ryan_Lane joined #salt
20:17 tekspot joined #salt
20:17 UtahDave auser: try   - tgt_type: compound
20:18 auser k
20:19 jimallman joined #salt
20:19 erezool the comment about update-alternatives, drawsmcgrae
20:19 erezool the comment about update-alternatives, drawsmcgraw
20:20 auser no, unfortunately
20:20 UtahDave what??.  Hm. let me do some source diving auser
20:21 auser that doesn't work
20:21 auser k, thx man
20:21 forrest JesseC, did that work? I guess I don't exactly understand the difference, is it not needed because the include has already added it in, so it's now considered local?
20:21 UtahDave drawsmcgraw: Hey, sorry, I missed the convo you were having before. I wasn't referring to your situation.  What problem are you running into?
20:21 JesseC forrest: I'm not sure what you mean.
20:21 JesseC forrest: I'm trying to require the sls lnpp.repos
20:22 JesseC forrest: but it's not finding it even though it's include
20:22 whit joined #salt
20:22 forrest JesseC, regarding what UtahDave was saying that you just need to use - require:\n    - lnpp
20:22 UtahDave auser: try  - expr_form: compound
20:22 forrest Yea I know the requires aren't grabbing it, UtahDave made that suggestion above, and I'm trying to undrestand why that would fix it
20:22 JesseC forrest: Ah, I didn't try that because I don't understand how that could work. lnpp isn't what I need to require, repos is what needs to be required
20:22 forrest what I'm saying is does that include: at the top automatically treat it as though it's local
20:22 auser k
20:22 JesseC lnpp is the entire state
20:23 auser where in the source are you looking UtahDave; I'm peeking into it
20:23 erezool thanks all!
20:23 forrest I know, that's my confusion as well
20:23 erezool i'll keep seeking and keep you all posted once i have found something..
20:23 erezool cheers!
20:23 UtahDave auser: I'm peeking into Tom's brain...
20:23 auser haha, okay
20:23 auser well… I'd love to know why there are 3 different ways to specify match type
20:24 UtahDave JesseC: So do you have a   /srv/salt/lnpp.sls    or /srv/salt/lnpp/repos.sls
20:24 auser btw, Tom's brain… I'm sure it's easy to get lost in there
20:24 UtahDave auser: Yes, I agree. we need to normalize those.
20:24 drawsmcgraw UtahDave: It involves using the 'wait' requisite but the end of my day is here. If I haven't tackled it by the time you get back on here tomorrow I'll be sure to stalk you :)
20:24 drawsmcgraw Thanks!
20:25 JesseC UtahDave: It's /src/salt/lnpp/repos/init.sls
20:25 JesseC UtahDave: everything is being ran based on /srv/salt/lnpp/init.sls
20:25 forrest JesseC, can you try - require:\n     - lnpp.repos?
20:25 dave_den was there a bug with file.directory setting the correct mode permissions when using recurse in 0.15.3? http://pastebin.com/BAqShc5A
20:25 ze- hey. Are you supposed to be able to call a module's function passing **kwargs from a state file ?
20:25 JesseC forrest: that's what I have it as already that doesn't work.
20:26 forrest oh you changed it from the - sls: ?
20:26 JesseC forrest: I tried lnpp.repos and repos
20:26 UtahDave ze-: sometimes that works.
20:26 JesseC oh no, not the sls
20:26 JesseC I'll give that a try
20:26 JesseC but that sls is mentioned here and by UtahDave. https://github.com/saltstack/salt/issues/5098#issuecomment-18266509
20:26 forrest yea what I'm saying is that you don't need to qualify it as an sls, because you already marked it as an include at the top
20:26 ze- UtahDave: from my tests, I couldn't get anything within the kwargs. just an empty dict.
20:26 dthom91 joined #salt
20:27 forrest JesseC, Hmmm.
20:27 UtahDave JesseC: this should work: http://pastebin.com/M6qigmmC
20:27 forrest He's got that
20:27 forrest https://gist.github.com/JesseC-/eaeecc9bcd17241bd01c
20:27 auser expr_form works UtahDave
20:27 auser on file touch
20:28 UtahDave auser: Thanks!  I'm adding some docs right now on that.
20:28 auser thanks dude
20:29 JesseC UtahDave: this is all I get -> The following requisites were not found: require: {'sls': 'lnpp.repos'}
20:30 auser UtahDave cmd.state.highstate: however does not
20:30 UtahDave JesseC: This might be a bug.  try requiring against   lnpp.repos.init
20:30 auser can I call a module inside the reactor system? I just need to restart a service
20:31 JesseC UtahDave: no dice
20:31 forrest same error JesseC?
20:31 UtahDave auser: just sent a pull req with a doc update: https://github.com/saltstack/salt/pull/6480
20:31 JesseC forrest: yes
20:32 auser rad
20:32 troyready joined #salt
20:32 UtahDave JesseC: Could you rename that init.sls to something else and require against the file directly?  I'm thinking this may be a bug with init.sls files.
20:32 dthom91 joined #salt
20:32 UtahDave auser:   yeah   cmd.service.restart
20:33 JesseC UtahDave: yep,  I'll create another test scenario tht doesn't use the init.sls
20:33 UtahDave cool, thanks, JesseC.
20:34 forrest Has anyone ever used the api to monitor a system all the way through reboot and complete tasks once the reboot is complete?
20:36 auser UtahDave: it also seems as if you can't have more than 1 state declaration in a reactor.sls, is that intended?
20:37 nliadm been getting this lately: zmq.error.ZMQError: Operation cannot be accomplished in current state
20:37 forrest what versio is that nliadm
20:37 forrest *version
20:37 nliadm 0.16.0
20:39 forrest I meant for zeromq nliadm
20:39 nliadm PyZMQ: 13.1.0
20:39 nliadm ZMQ: 3.2.2
20:39 forrest can you try zmq 3.2.3?
20:39 UtahDave auser: Huh. I would think that you can have more than one state declaration. Are you getting an error?
20:39 JesseC UtahDave: forrest: no dice -> https://gist.github.com/JesseC-/0654686dccf7030ee095
20:39 ze- UtahDave: what would the proper way to talk about kwargs not working properly (for states module.run), and suggesting a patch ?
20:40 JesseC UtahDave: Same error even without the init.sls and both files in the base directory. /srv/salt
20:40 auser UtahDave: I'm not seeing an error, in fact I only see output (watching the event system + logs) if it's successful. I'm also watching the service logs
20:40 oz_akan_ joined #salt
20:40 UtahDave ze-: I would include detailed examples.   github.com/saltstack/salt/issues/new
20:40 auser I get nothing :/
20:41 UtahDave JesseC: Ok, could you open an issue on that?
20:41 JesseC UtahDave: Maybe it's because I'm on version 0.15.3 ?
20:41 Ryan_Lane joined #salt
20:41 JesseC I think there is something mentioned about 0.16 in that post.. checking
20:41 UtahDave JesseC: Yes, that's exactly it. I'm sorry. I should have asked which version you're on
20:41 forrest that's the latest available from the repo JesseC :\
20:41 UtahDave 0.16.0 is the first version that has that.
20:41 JesseC I just thought about checking it because I thought bootstrap would install the latest version
20:41 JesseC dog
20:41 JesseC doh*
20:43 forrest That's odd the repo is behind though
20:43 auser UtahDave: http://pastie.org/private/l6mb76s3rgaclcx7z7ybw
20:43 UtahDave JesseC: what OS are you on?
20:43 auser from the cli, that works, from the reactor file, it does not
20:44 JesseC UtahDave: CentOS 6.4
20:44 forrest I'm running RHEL 6.4
20:44 teskew any way you can resend the keys from minions to salt master if you've deleted them from salt master?
20:44 dthom91 joined #salt
20:44 UtahDave teskew: the minions will attempt to authenticate after a little while.  Or you can manually restart the minion so the attempt to reauthenticate.
20:45 forrest This is odd though JesseC
20:45 UtahDave teskew: then you can use   salt-key -a <minionid>
20:45 forrest http://dl.fedoraproject.org/pub/epel/6/x86_64/repoview/salt-master.html
20:45 forrest 16 is in there, and I installed 0.15.3 from EPEL
20:45 forrest at least according to repoquery
20:45 giantlock joined #salt
20:46 teskew UtahDave: that doesn't seem to work. i've waited for a few hours and restarted salt-minion multiple times
20:46 auser <~ stuck
20:46 auser lame
20:46 JesseC forrest: I just used the wget one liner to pull the boostrap script down
20:46 forrest JesseC, yes I used the bootstrap as well
20:46 auser brb, gotta get in transit -- be back in ~30
20:46 forrest You can see on the centos section though that it's installed EPEL though
20:47 forrest https://raw.github.com/saltstack/salt-bootstrap/stable/bootstrap-salt.sh
20:47 forrest do you see if after cleaning yum metadata and trying a yum list salt?
20:48 faldridge joined #salt
20:49 forrest http://download.fedoraproject.org/pub/epel/6/x86_64
20:49 forrest the version available there is 15.
20:50 forrest so the epel list got updated but the mirrors didn't?
20:53 forrest JesseC, have you ever talked to the EPEL guys about ensuring the mirrors have the same package available from the repoview? I haven't done so before.
20:53 nliadm is there a way to turn off the mine?
20:53 devinus joined #salt
20:54 UtahDave nliadm: yeah, I think if you set mine_interval: 0
20:55 oz_akan_ joined #salt
20:58 mianos joined #salt
20:58 jefimenko joined #salt
20:59 mianos this bug https://github.com/saltstack/salt/issues/6003 with the pip.py is in the ubuntu ppa versions, I have a lot of machines that I need to install but they are not connected to the internet so devel via git is not a solution
20:59 mianos has anyone a better suggestion that maybe install a new pip.py as a managed file into the install
21:00 mianos saltstack minion is installed as part of the cloud init process
21:00 mianos via a local apt cache normally
21:01 carlos joined #salt
21:02 oz_akan_ joined #salt
21:03 dthom91 joined #salt
21:03 tomeff_ joined #salt
21:03 UtahDave mianos: you can drop the latest pip.py   in /srv/salt/_modules/    and run   salt \* saltutil.sync_modules
21:04 bemehow_ joined #salt
21:04 mianos hey! thanks, top idea
21:04 mianos that will sync over the top of the normal ppa install?
21:05 mianos hmm, hang on sync_modules oprganises the sever to obtain files for sync?
21:06 mianos oops, sorry, just read
21:06 JesseC forrest: had to bail for another meeting, lol
21:06 lyddonb_ joined #salt
21:06 mianos I was looking at sync_modules, thanks a lot
21:06 forrest JesseC, all good, basically epel mirrors are not updated for some reason
21:07 JesseC forrest: I didn't look through the script, I'm checking it now. I've never talked to any of the EPEL guys.
21:07 forrest Yea I hit them up, no response yet, they aren't as lively as over here it seems
21:07 JesseC forrest: Is 0.16 dev or stable though?
21:07 JesseC that's what I was trying to find out before I left
21:07 forrest It should be the official release
21:07 forrest http://dl.fedoraproject.org/pub/epel/6/x86_64/repoview/salt.html
21:07 JesseC but I don't see version info in the repo
21:08 forrest if it's there, that means it should be available
21:08 JesseC ohh
21:08 mianos is there a state version of sync_modules?
21:08 forrest yea but if you hit the mirror url: http://download.fedoraproject.org/pub/epel/6/x86_64,
21:08 forrest minus that ,
21:08 forrest the verison will be 15
21:09 JesseC I see this salt-0.16.0-1.el6.noarch.rpm
21:09 forrest on the mirror?
21:09 forrest which one did you get pushed to? The 4 I hit all had 15
21:09 JesseC nm, yeah you're right
21:09 JesseC odd
21:10 forrest Yea I don't know enough about the epel mirror sync to know whose responsibility it is to do the sync, but it's been in the repovioew for 20 days, so it should be on the mirrors
21:10 JesseC it was a while back too, the mirrors should all be updated
21:10 forrest Yep exactly
21:10 oz_akan_ joined #salt
21:10 forrest I'll keep you updated on what the epel guys say
21:11 forrest you could always do a localupdate for the time being, ghetto, but an option
21:11 forrest Well, if the epel guys ever say anything
21:11 JesseC maybe I'll use salt to self update to the latest version
21:11 JesseC then run my other items
21:11 JesseC :)
21:11 forrest hah that would be funny
21:13 JesseC would be good training to use grains
21:13 forrest for sure
21:13 JesseC see if the version is > than 0.16
21:13 forrest I still want it updated in epel
21:13 JesseC if it's not, pull from the main epel and do a local update
21:13 forrest Yea that would be cool
21:15 rlb joined #salt
21:16 JesseC forrest: I'm reading through this http://fedoraproject.org/wiki/Infrastructure/Mirroring
21:16 JesseC forrest: it seems the public mirrors might be responsible for themselves
21:17 racooper probably. the utexas mirror hasn't updated yet and that's the one I keep getting (Internet2)
21:17 forrest JesseC, Yea that's why I was asking for clarification over in #epel, because if it was just one mirror that was problematic, I could see it
21:21 nliadm why does salt-call need to load its modules 4 times?
21:23 forrest UtahDave, 0.16 should be considered the latest stable release right?
21:24 forrest Here we go, I found a mirror that has it: http://mirror.pnl.gov/epel/6/x86_64/
21:24 forrest ugh, who is only updating their epel mirrors once a month, jeez
21:25 JesseC I'm trying to just use download.fedoraproject.org but it errors out on me
21:27 forrest try dl.fedoraproject.org
21:27 JesseC yeah just did, it still redirects you to a random mirror
21:28 forrest bummer
21:29 backjlack joined #salt
21:30 JesseC forrest: I wonder if they all only update once a month
21:30 JesseC forrest: that would make sense why none of them have it yet, beings though it was added less than a month ago
21:31 forrest JesseC, I was thinking the same thing, but that mirror I linked has dates throughout July
21:31 forrest it seems odd that you'd only do the rsync once a month
21:32 forrest defeats the whole purpose of the mirror
21:32 troyready joined #salt
21:33 forrest are you behind a corp firewall JesseC?
21:33 JesseC nah
21:34 forrest can you see what happens if you do yum list salt-0.16.0-1.el6.noarch.rpm ?
21:34 forrest see if it hits the other mirrors
21:34 JesseC sure
21:35 JesseC No matching Packages to list
21:35 forrest ok
21:36 pdayton joined #salt
21:36 JesseC now I want to start checking all of my packages I use there vs the main repo
21:36 Lue_4911 joined #salt
21:36 forrest Yea thankfully since I'm on RHEL we don't pull anything from EPEL, just using it for my test env
21:37 UtahDave forrest: Yeah, 0.16.0 is the latest release.
21:37 JesseC I only use it for a few items not available on the CentOS repo, but that's abou tit
21:37 forrest UtahDave, well some of the EPEL mirrors aren't updated, which is a bummer.
21:37 forrest Yea Jesse, I just had to use it for some extra items I had.
21:38 forrest UtahDave, have you seen that happen before? Do you have to contact each org asking them to actually update within a decent timeframe?
21:40 UtahDave forrest: I don't remember ever seeing them so out of date. herlo does our epel/yum packaging.  I'll see what he knows about it.
21:40 UtahDave We're working on getting our own SaltStack repo up, so that should help once that's ready.
21:41 forrest UtahDave, awesome he was going to be my next stop. I've only been able to find one EPEL mirror that actually had the latest.
21:41 forrest He updated it in the actual repo
21:41 forrest the mirrors just aren't up
21:42 JasonSwindle UtahDave:  awesome.
21:42 racooper is there a way to "blacklist" a particular mirror from epel?
21:44 JesseC racooper: http://www.techchorus.net/how-configure-yum-exclude-mirror
21:47 racooper thanks. didn't make any difference tho, the next mirror it picked up doesn't have .16 either
21:48 forrest http://mirror.pnl.gov/epel/6/x86_64 has it
21:48 forrest but I get a 401 for some reason
21:50 racooper yeah it doesn't make sense...pnl.gov has it updated Jul 14, and utexas has updates dated July 28, but not salt.
21:50 auser joined #salt
21:51 SEJeff_work racooper, Always use the source: http://dl.fedoraproject.org/pub/epel/6/x86_64/
21:51 SEJeff_work problem solved :)
21:52 JesseC SEJeff_work: you can't use the bootstrap script if you do that
21:52 forrest That's what JesseC and I were discussing earlier racooper
21:53 forrest the fact that utexas has others, but not salt
21:53 SEJeff_work Ah gotcha
21:53 JesseC SEJeff_work: because boostrap sets up epel and pulls from the mirrors
21:53 SEJeff_work Gotcha
21:53 SEJeff_work So sed the bootstrap script and put it on a webserver :)
21:53 SEJeff_work This is Linux, it is all open to change
21:54 racooper and using an I2 mirror...means I get avg 9ms pingtime instead of 57ms :P
21:55 cewood joined #salt
21:55 mgw1 joined #salt
21:56 forrest It's not that SEJeff_work, the bootstrap simply installs the epel repo, so it would have to be modified, unless there's an RPM that uses the primary repo.
21:57 SEJeff_work Understood. That is a pain
21:57 forrest Yea, I would have already fixed it if it was just adding the repo file :P
21:57 forrest Oh, JesseC, can you run a repoquery -i python-jinja2 ?
21:58 forrest I'm curious as to where you downloaded python-jinja2 from
21:58 forrest on cent
21:58 JasonSwindle use bootstrap + git
21:59 JasonSwindle curl --insecure -L http://bootstrap.saltstack.org | sudo sh -s -- git v0.16.1
21:59 JasonSwindle or
21:59 JasonSwindle curl -L http://bootstrap.saltstack.org | sudo sh -s -- git v0.16.1
21:59 forrest Oh good suggestion JasonSwindle, thanks.
22:00 JasonSwindle It is how I always do it… I never trust Ubuntu's packages
22:02 SEJeff_work I never trust Corey^WUbuntu's packages either :D
22:05 racooper https://admin.fedoraproject.org/updates/salt-0.16.0-1.el6
22:05 Gifflen_ joined #salt
22:06 racooper looks like salt 0.16 was pushed to epel el6 stable only today
22:06 forrest Nice catch racooper.
22:07 racooper basically, if I understand from the zabbix packager correctly, a new release goes to testing for 14 days unless it's tested and reported back to shorten that time. after 14 days it's released to stable.
22:07 SEJeff_work also you have to remember a lot of these stay in epel-testing for 2-3 weeks
22:08 SEJeff_work bingo
22:08 SEJeff_work racooper, In koji, the fedora package build system, you have to give a package "karma" to say it worked
22:08 SEJeff_work that will get it out of epel-testing early
22:08 forrest Good to know SEJeff_work, I've messed with koji a bit but not for epel
22:09 SEJeff_work forrest, yup
22:09 SEJeff_work you can do yum install --enablerepo=epel-testing
22:09 bostonian joined #salt
22:09 forrest UtahDave, you don't have to bother herlo, racooper found it.
22:10 SEJeff_work forrest, Perhaps add a flag to the bootstrap script to pull it out of epel-testing
22:10 forrest I'm trying to keep my test env as close to what production would look like as possible, so I'll use the one liner.
22:10 forrest that's a good idea SEJeff_work
22:10 awesomess3 joined #salt
22:10 SEJeff_work forrest, If you don't want to do that yourself, open an issue and I'll get to it if someone doesn't beat me to it. File an issue so it doesn't get forgotten please
22:10 forrest I
22:11 forrest * will file the issue regardless, I don't know if I'll get to it, it won't be till the weekend at the earliest
22:11 SEJeff_work ditto
22:11 SEJeff_work I've got in-laws in town until monday
22:11 SEJeff_work so I'm a bit busy until then and don't want to forget
22:11 SEJeff_work It would be a trivial addition
22:11 forrest for sure
22:11 bostonian Salt stack uses hostnames by default... but can it use private IP addresses?
22:11 forrest yea
22:12 forrest My salt master is only recognized via the IP
22:12 SEJeff_work I use hostnames for quasi-dr
22:12 SEJeff_work YOu can copy the pki directory on the master to a new master and update dns
22:12 SEJeff_work your minions cache the master key and won't connect to a master with a different key
22:13 bostonian forrest: Where can I change it?
22:13 SEJeff_work so no man in the middle worries
22:13 SEJeff_work bostonian, /etc/salt/minion look for "master:"
22:13 forrest bostonian, the salt master? /etc/salt/minion
22:13 SEJeff_work defaults to the hostname salt
22:13 forrest yup
22:13 awesomess3 left #salt
22:14 bostonian forrest: ahh, how silly of me! I was just there. thanks :)
22:15 forrest bostonian, yea no problem, I don't remember if it's clarified in the docs whether you can use an IP or not
22:16 blee joined #salt
22:18 forrest SEJeff_work, https://github.com/saltstack/salt-bootstrap/issues/190
22:18 rsimpkins UtahDave: I figured out the issue with that memory problem. It isn't salt using it. It is the slab allocator. Something is causing the dentry cache to balloon to monster sizes. It is caching more dentry objects than I have on my system! :)
22:19 rsimpkins 100,000 dentry objects on a system with about 8,000 directories.
22:19 SEJeff_work rsimpkins, With slabtop
22:20 forrest SEJeff_work, you might want to modify your comment to ensure we cover salt-api and such as well
22:20 forrest and cloud
22:21 JasonSwindle UtahDave:  I see 0.16.2 just dropped
22:21 JasonSwindle https://github.com/saltstack/salt/blob/v0.16.2/doc/topics/releases/0.16.2.rst
22:21 oz_akan_ joined #salt
22:21 JasonSwindle Is everything moving away from runas?
22:21 hazzadous joined #salt
22:21 SEJeff_work forrest, Feel free to comment :)
22:21 JasonSwindle oz_akan_:  !
22:21 JasonSwindle Welcome!
22:23 forrest SEJeff_work, done, I will try to get to it tonight.
22:24 SEJeff_work terminalmage, Did you do the release notes for this release?
22:25 terminalmage SEJeff_work: yes
22:25 chrism_ so I need to match on a grain via glob within a state file, but i can't find any docs on that in particular
22:25 chrism_ like within an if statement
22:25 SEJeff_work terminalmage, Because they are good. I really like that you linked to the bugs. Please continue doing the release notes. These are the best I've seen
22:25 terminalmage I have them ready to push
22:25 terminalmage SEJeff_work: thanks!
22:25 chrism_ i see how to glob on a grain from teh CLI but not within a state file.  Any idea where i'd find that?
22:26 SEJeff_work chrism_, {% if grains['os'] = 'Fedora' %} ...
22:26 terminalmage I also linked to the state and module docs
22:26 SEJeff_work terminalmage, I appoint you the new release notes person!
22:26 terminalmage hehehe
22:26 chrism_ SEJeff_work: How can I match for something like grains['cpu_model'] = '^Intel'
22:26 chrism_ er ^Intel*
22:27 terminalmage chrism_: there is a grains_pcre matcher
22:27 SEJeff_work chrism_, It is jinja. I'm not sure that you can
22:27 terminalmage oh
22:27 SEJeff_work terminalmage, inside a state
22:27 terminalmage jinja
22:27 terminalmage hrmm
22:27 chrism_ aaah
22:27 chrism_ uh hrm
22:27 terminalmage you can use str member functions
22:27 chrism_ well that makes things interesting
22:27 chrism_ i'm not sure what that is?
22:27 SEJeff_work chrism_, {% if "Intel" in grains['cpu_model'] %} would work
22:28 terminalmage {% if grains.get('cpu_model', '').startswith('Intel') %}
22:28 terminalmage pretty sure that works
22:28 SEJeff_work touche, that would do the trick!
22:28 chrism_ lol awesome
22:28 chrism_ doing so now, tyvm!
22:28 SEJeff_work but grains['cpu_model'] would always exist unless there was massive brokenness
22:28 terminalmage no prob
22:28 SEJeff_work however, the .get for defensive programming is still not bad form
22:28 terminalmage yeah, i just have a habit of using get
22:28 chrism_ aah yeah that's so it doesnt' return null right?
22:29 SEJeff_work chrism_, Thats so it doesn't throw a KeyError
22:29 devinus joined #salt
22:29 SEJeff_work grains['bogus']
22:29 chrism_ gotcha
22:29 SEJeff_work would blow up with a KeyError
22:29 devinus joined #salt
22:30 SEJeff_work whiteinge, Mind kicking off a build for the 0.16.2 release notes on docs. and rtfd?
22:31 whiteinge sure
22:31 SEJeff_work danke
22:31 terminalmage whiteinge: I was gonna take care of that when it was time, since you told me how
22:31 whiteinge doh!
22:32 SEJeff_work My bad
22:32 terminalmage but if you started it, that's fine
22:32 SEJeff_work I'll hush
22:32 terminalmage ha
22:32 JasonSwindle terminalmage:  Maybe you knwo
22:32 whiteinge yeah, running...
22:32 JasonSwindle *know
22:32 JasonSwindle Is salt moving away from runas and to user?
22:32 mikedawson joined #salt
22:32 terminalmage I just wanted to make sure that the release notes were up by the time tom announces it on the mailing list
22:32 JasonSwindle I ask because of 0.16.2 and pip moving away from runas
22:32 SEJeff_work Then the answer is yes :)
22:32 JasonSwindle https://github.com/saltstack/salt/blob/v0.16.2/doc/topics/releases/0.16.2.rst#pip
22:32 forrest SEJeff_work, is there a location in git to query for the latest stable release? I'm thinking about the logic here, and trying to think how we'll determine whether it's the 'latest' or not
22:33 SEJeff_work forrest, Sure, git tag -l
22:33 terminalmage JasonSwindle: runas is supposed to be for the cmd module
22:33 JasonSwindle ah ok
22:33 terminalmage elsewhere, we want to be consistent
22:33 JasonSwindle because I see it is also used on git state
22:33 forrest SEJeff_work, ahh ok
22:33 terminalmage and "user" is the most common usage
22:34 JasonSwindle I know because I had to change it because I had it as user and had to change it to runas
22:34 terminalmage JasonSwindle: ahh ok, well we'll probably deprecate that as well then
22:34 terminalmage weird
22:34 JasonSwindle d;oh
22:34 terminalmage ok
22:34 terminalmage sorry man
22:34 terminalmage 'user' is more common, and we want to remain consistent
22:34 terminalmage in fact, we might even deprecate it in the cmd module eventually as well
22:35 JasonSwindle https://github.com/saltstack/salt/blob/develop/salt/states/git.py#L37
22:35 SEJeff_work forrest, $ git tag -l | sort -t. -k2,3 -n | tail -n1
22:35 SEJeff_work v0.16.2
22:35 JasonSwindle No biggie
22:35 SEJeff_work That should do the trick for you I think
22:35 JasonSwindle It looks easy to fix; so easy I could do it
22:35 terminalmage user will still work, it's just deprecated
22:36 JasonSwindle When I used `user`, it chowned everything to root on second highstate
22:36 whiteinge SEJeff_work, terminalmage: docs are built
22:36 JasonSwindle but when I used runas in git; worked great
22:36 terminalmage sweet
22:36 terminalmage JasonSwindle: no I meant in pip
22:36 JasonSwindle oh
22:36 terminalmage sorry for the confusion
22:36 SEJeff_work thanks seth!
22:36 JasonSwindle ah yes
22:37 JasonSwindle terminalmage:  do you work on Salt-Cloud?
22:37 terminalmage JasonSwindle: nah, I haven't touched any of that code
22:38 JasonSwindle ah ok.
22:38 UtahDave JasonSwindle: redbeard2 is the main dev on salt-cloud
22:39 JasonSwindle Ok.  I am loving it.  We are able to deploy our stack in 15 minutes or less with it
22:40 JasonSwindle Is GIT_FS going to get love?
22:40 JasonSwindle It is a wonderful idea; but lacking some items
22:41 terminalmage there are issues with GitPython
22:41 terminalmage for one, it wraps the git cli and it does a bunch of asserts on odd things
22:42 terminalmage like number of lines in output
22:42 JasonSwindle ah ok
22:42 JasonSwindle because for me to even touch it, I need submodule support and tags
22:42 terminalmage which, if your git remote is behind SSH and the remote box has an /etc/issue, will mess with that and make it fail
22:42 JasonSwindle even more lovely
22:43 terminalmage yeah
22:43 UtahDave JasonSwindle: are there any gitfs issues you're following? You might comment on one of them to bump the interest
22:44 JasonSwindle Ah well, you guys will get around to it in time.  I am in no rush.
22:44 JasonSwindle No issues; but lack of what I need
22:44 terminalmage http://docs.saltstack.com/topics/releases/0.16.2.html
22:44 JasonSwindle but, I can wait because I got my work flow down well
22:45 JasonSwindle terminalmage:  I dig the links to the issue
22:45 JasonSwindle +1
22:45 terminalmage that's the live copy of the docs with working links for the functions
22:45 JasonSwindle http://docs.saltstack.com/topics/releases/0.16.2.html#user-group-management < - doc issue
22:45 terminalmage JasonSwindle: thanks
22:45 JasonSwindle the last bullet
22:45 terminalmage JasonSwindle: d'oh!
22:45 terminalmage thanks
22:45 JasonSwindle np
22:45 JasonSwindle I will be doing doc work all tomorrow
22:48 forrest SEJeff_work, cool.
22:48 JasonSwindle I may work on a troubleshooting for saltstack if I have the time
22:49 forrest JasonSwindle, are you gonna be updating the OS specific docs?
22:49 JasonSwindle not sure; I may hit up Salt-Cloud the most.  This is where we had the most pain
22:50 JasonSwindle Our stack is heavy on it
22:51 jefferai joined #salt
22:52 avienu joined #salt
22:52 forrest JasonSwindle, ok cool nevermind then
22:53 jslatts joined #salt
22:53 JasonSwindle UtahDave:  I was never able to get the salt-cloud map to work
22:53 UtahDave JasonSwindle: really?  Hm.  I've used it a ton.
22:53 UtahDave can you pastebin what you were trying and what the error was?
22:54 JasonSwindle sure
22:54 JasonSwindle There was no "error", but I could not Highstate.  Digging up gist
22:55 JasonSwindle UtahDave:  https://gist.github.com/JasonSwindle/e3503cddb2d47e64397a
22:55 trigoman joined #salt
22:56 JasonSwindle file1 is what you gave me
22:56 JasonSwindle The minions got 0.16.1; and I could match via grains
22:56 JasonSwindle but HS would not even fire off
22:56 trigoman akoumjian: I'm using salty-vagrant and I have a few concerns. When I do vagrant up, I get two things that I don't know if I should worry about. 1) salt-minion not found 2) salt-call not found. Then it hangs at Bootstrapping Salt
22:57 JasonSwindle trigoman:  that is odd
22:57 JasonSwindle Do you have to VagrantFile?
22:57 terminalmage JasonSwindle: I fixed that issue in the release notes and the docs are rebuilding now
22:58 terminalmage thanks for catching that
22:58 JasonSwindle trigoman:  Mine for review; https://gist.github.com/JasonSwindle/4552855c7e7c35069a5e
22:58 JasonSwindle terminalmage:  Sweet
22:58 kstaken joined #salt
22:58 terminalmage ahh crap, I found another issue. The rabbitmq link near the bottom is also busted
22:58 terminalmage d'oh
22:59 JasonSwindle terminalmage:  I think I found a way to make docs better as well.
22:59 JasonSwindle If you are not using the Bootstrap icons; why self host the bootstrap CSS
22:59 JasonSwindle http://netdna.bootstrapcdn.com/twitter-bootstrap/2.3.0/css/bootstrap-combined.no-icons.min.css <-
22:59 JasonSwindle and your CSS is raw and not minify
23:00 terminalmage whiteinge: ^^^
23:00 JasonSwindle and you can have jquery host you js for you
23:00 trigoman JasonSwindle: I'm using this as an example: https://github.com/ollin/example-vagrant-tomcat
23:01 JasonSwindle trigoman:  is your minion file in salt/minion.conf?
23:01 trigoman Also, is there anything special that I have to do to get salt to work behind a proxy?
23:02 JasonSwindle terminalmage:  http://code.jquery.com/jquery-1.10.1.min.js <---
23:02 trigoman JasonSwindle: Yes, the salt/minion.conf file is there.
23:02 JasonSwindle what is in it?
23:03 trigoman file_client: local
23:03 JasonSwindle add master: localhost to the first line
23:03 JasonSwindle master: localhost
23:04 JasonSwindle see if that helps
23:05 trigoman It still says that salt-minion and salt-call were not found.
23:05 trigoman Hanging at Bootstrapping Salt
23:05 JasonSwindle odd
23:05 dthom91 joined #salt
23:05 JasonSwindle I do wonder if it is the proxy issue
23:05 alunduil joined #salt
23:05 JasonSwindle I am not proxyed nor have faced that
23:05 trigoman It is usually an issue for me.
23:06 trigoman Proxies ruin everything.
23:06 JasonSwindle terminalmage:  Are you the correct person to bug about docs.saltstack.com ?
23:06 UtahDave JasonSwindle: you can edit the docs yourself in the main Salt repo here: https://github.com/saltstack/salt/tree/develop/doc
23:07 UtahDave JasonSwindle: Hey, so I'm looking at your map file. I think that looks fine.  But I find it very strange that you can execute commands against the minions, but not a highstate.
23:07 JasonSwindle Oh boy, I am about to go crazy on that.   http://www.webpagetest.org/result/130801_XR_3EMF/
23:08 JasonSwindle UtahDave:  I agree
23:08 JasonSwindle I remove that one line out of the map; and boom.  It works
23:08 JasonSwindle In fact, the map is being use right now in staging
23:08 intchanter joined #salt
23:09 UtahDave so script_args is keeping a highstate from running?
23:09 UtahDave That just doesn't make any sense.  Once salt-cloud creates the server, salt is the only thing touching the minion itself.
23:09 jefimenko left #salt
23:09 JasonSwindle I agree
23:09 UtahDave I'm going to have to do some testing here to find out what's going on.
23:10 JasonSwindle OK
23:10 UtahDave what OS are you using on your vms?
23:10 trigoman How did you guys install salt?
23:10 trigoman I did a non-sudo install and I guess I'm missing a bunch of stuff.
23:10 JasonSwindle UtahDave:  12.04
23:11 JasonSwindle as root; I do the salt bootstrap pointing to the git tag
23:11 UtahDave JasonSwindle: So the second map in your gist is what you're using?
23:11 trigoman I'm running osx, using brew. Thought I didn' have to :(
23:11 JasonSwindle I was playing
23:11 JasonSwindle curl -L http://bootstrap.saltstack.org | sudo sh -s -- git v0.16.2
23:12 JasonSwindle That will do the minion
23:12 JasonSwindle curl -L http://bootstrap.saltstack.org | sudo sh -s -- -M -N git v0.16.2
23:12 Fin1te joined #salt
23:12 JasonSwindle That will do the master
23:13 UtahDave ah, right.
23:13 JasonSwindle UtahDave:  The second file was me playing; the first one is what you gave and it seemed to work
23:18 UtahDave ok
23:25 JasonSwindle This is ot; but anyone good at NGINX and uWSGI?  It seems like you need a PHD in them to tune them.
23:25 Joe630 if you need a phd to tune them, tuning them probably won
23:25 Joe630 'thelp you
23:26 Joe630 <-- myaxiom of kernel tuning
23:26 Joe630 applies to webservers, too
23:26 JasonSwindle true
23:27 Joe630 what is broken
23:28 JasonSwindle Not a thing; just trying to get the most out of them.
23:28 Joe630 dont do it man
23:28 Joe630 prematureoptimization makes guido cry
23:28 JasonSwindle in early stages of using them
23:28 JasonSwindle lol
23:29 oz_akan_ joined #salt
23:35 zooz joined #salt
23:38 JasonSwindle1 joined #salt
23:38 p3rror joined #salt
23:40 JasonSwindle1 Any luck?
23:41 JasonSwindle1 UtahDave:  Any luck?
23:41 Joe630 whats happening
23:41 UtahDave kinda
23:41 JasonSwindle joined #salt
23:43 cxz joined #salt
23:43 trigoman JasonSwindle: Thanks for your help! Cya1
23:44 JasonSwindle trigoman:  NP
23:44 JasonSwindle later
23:44 UtahDave JasonSwindle: So I've got it working, I think.  It just looks REALLY finnicky on how the yaml is put together.
23:45 JasonSwindle Is that a YAML issue; salt-cloud issue; or something salt-cloud needs to work on?
23:49 UtahDave OK, I got it working. Let me post a gist.
23:49 mikecb joined #salt
23:50 UtahDave JasonSwindle: this worked for me.  https://gist.github.com/UtahDave/c657dc428191f76f2f07
23:51 JasonSwindle Let me test that
23:51 JasonSwindle I don't see a different; but let me look
23:51 UtahDave I also ran it on the latest salt-cloud from git.
23:52 UtahDave But I ran into an issue where it never returned from the salt-cloud -m mymap.map -P    command
23:52 UtahDave just hung there.
23:53 aat joined #salt
23:54 JasonSwindle I will test it out, and let you know tomorrow.  I am off, later guys!
23:58 jschadlick left #salt

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