Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-01-06

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

All times shown according to UTC.

Time Nick Message
00:01 sroegner_ forresta: you still around?
00:04 linjan joined #salt
00:04 forresta yea what's up
00:05 sroegner_ forresta: i am looking for a way to make a piece of data (stuff i have in a pillar right now) part of the formular
00:07 rojem joined #salt
00:07 forresta ok
00:07 sroegner_ forresta: but i cannot just do something like import_yaml right?
00:07 forresta what do you mean?
00:08 sroegner_ forresta: of something that isn't highstate data
00:08 forresta so you have data that you want to move from pillar to somewhere else, but it isn't part of your settings, or your map?
00:08 sroegner_ https://github.com/accumulo/accumulo-formula/blob/master/pillar.example
00:09 sroegner_ look at the bottom - all the stuff under accumulo-site-profiles
00:09 sroegner_ maybe a map
00:10 sroegner_ i am just somewhat blocked right now
00:10 forresta what's wrong with it being in pillar since it's related to the appropriate thing?
00:10 sroegner_ you see the keys? like 512MB, 1GB, 2GB and so on?
00:11 forresta yea
00:11 sroegner_ these are to be exposed - but it doesn't make much sense to have a newbie carry the entire structure in their pillar
00:11 sroegner_ so i want to give the user a way of picking one of the profiles
00:12 sroegner_ but the data needs to be somewhere
00:12 forresta I feel like where the data is is fine
00:12 forresta this is simply an example pillar
00:12 forresta users can pick and choose what they want to have in there
00:12 forresta so maybe they only need 1GB or whatever
00:12 forresta but your example showing multiple options is good to have
00:12 sroegner_ but my idea of formulas is that they should work without pillar data
00:13 forresta right
00:13 forresta does this one not?
00:13 sroegner_ no
00:13 sroegner_ i have to have one for the resulting xml file
00:13 sroegner_ one profile that is
00:13 sroegner_ i can easily define a default profile key
00:14 sroegner_ but this is now why i want the profiles part of the formula
00:14 sroegner_ i could also just put them back into the template where they go anyway
00:14 sroegner_ :-)
00:14 forresta hah
00:14 forresta honestly I'd just set a default value: https://github.com/gravyboat/fail2ban-formula/blob/master/fail2ban/config.sls
00:14 sroegner_ make the switch there
00:14 forresta like I do there on lines 11-13
00:15 forresta but that formula is pretty complex
00:15 forresta yours that is
00:15 sroegner_ the service itself is pretty complex
00:16 sroegner_ which is not to say that theformula couldn't be a little lighter :-)
00:16 forresta heh yea
00:16 forresta yea I honestly don't have a good 'suggestion'
00:16 forresta since it's required data
00:16 forresta other than moving it somewhere else, which is kinda meh, or setting a default, which is also kind of meh
00:17 sroegner_ np
00:17 sroegner_ helped me anyway
00:17 sroegner_ thx
00:17 forresta I'll try to do some thinking on it, might be good to get whiteinge involved on Monday as wel
00:17 forresta yea np
00:17 taion809 joined #salt
00:23 fllr joined #salt
00:23 kermit joined #salt
00:24 mgw joined #salt
00:32 linx_0x joined #salt
00:33 jcockhren beach: it's secure as long as you keep your private key private
00:33 jcockhren beach: also, I suggested to add your public key as a deployment key. hence it'll only have read access
00:33 beach do you know how you got things to work with a passphrase? does it have something to do with the ssh-agent
00:34 beach everything is read only as root - no other user can read
00:34 beach thanks for the help with this... it's no in the perfect state, but at least now I'm moving on to the bigger picture
00:35 jcockhren beach: it's a ssh-agent thing
00:36 jcockhren beach: if you're worried about security, then don't run salt-master as root
00:36 beach okay... I was getting all kinds of 'cannot connect to agent messages' so that's something I need to drill into
00:37 jcockhren http://docs.saltstack.com/topics/nonroot.html
00:37 ruqui joined #salt
00:38 beach it's still running as root on minion though right?
00:38 sgviking joined #salt
00:39 beach the main security component is what i may endup storing in pillar/ or file scripts
00:39 ruqui hi! I have a FreeBSD minion that I need to install a package of a different (later) version than the one already installed ... how can I do this?
00:41 beach are you talking about specifying the version in a states file?
00:42 ruqui sort of  ... according to the docs that feature is not available in FreeBSD pkg .... is there any workaround?
00:44 beach couldn't tell you off the top of my head, but I'd guess that you'd be looking at writing some code
00:45 rager joined #salt
00:56 chuyskywalker forresta: I opened a ticket with the folks that made the vagrant box I was using and seeing bad results with (https://github.com/2creatives/vagrant-centos/issues/9), tried a few of their other boxes, went back to the trouble one and *poof* the problem was gone. Aiieee...
00:59 holys joined #salt
00:59 blarghmatey_ joined #salt
01:06 mgw joined #salt
01:24 drawsmcgraw joined #salt
01:30 jdenning joined #salt
01:33 rojem joined #salt
01:34 jfzhu_us joined #salt
01:35 nahamu http://blog.shalman.org/getting-started-with-saltstack-in-the-joyent-cloud/
01:39 yano joined #salt
01:55 rojem joined #salt
01:58 rojem joined #salt
02:06 rojem joined #salt
02:09 chuyskywalker Let's say I'm making a salt state for my dev team. I want the salt state to spin provision all our git repos for them -- but I need to get each person's personal SSH private key into the machine being managed (i'm using vagrant and masterless). How would I go about structuring this? IE: Per user data that is required for the state, but not include in my vagrant+salt repo.
02:13 jcockhren chuyskywalker: pillar
02:13 beach_ joined #salt
02:13 jcockhren chuyskywalker: actually let me explain.
02:14 beach_ does anyone have a clue what the cache TTL is for gitfs? I push content and I have no idea when it will get picked up. I can't find anything in the docs.
02:14 chuyskywalker I've not really read much on Pillars. I'm happy to read a good source if you've got one (instead off all that typin')
02:15 Teknix joined #salt
02:15 jcockhren http://docs.saltstack.com/topics/pillar/
02:16 jcockhren :)
02:20 drawsmcgraw beach_:  I use gitfs and my favorite way of ensuring the newest states are around is clobbering the local cache & restarting the master
02:20 drawsmcgraw i.e. 'rm -r /var/cache/salt/master/gitfs && service salt-master restart'
02:20 beach_ yeah, I've been doing that... it's a little abrasive though don't you think.
02:20 goki joined #salt
02:20 beach_ not predictive and automagic
02:27 chuyskywalker Hrm, I don't think pillar is what I want. Here's a more concrete example (https://github.com/chuyskywal) in the synced/files/ssh directory there is a bb-deploy (and .pub) which have been .gitignored. If I were to checkout this repo in the future, salt will fail because the files are missing. Should I just let it be and have it fail like that, or is there a better way to detect that the user needs to provide some post-checkout f
02:38 clearclaw I have a custom module in files/base/_modules/foo.py and on the master I can run salt-run foo.method bar succesfully, but foo.method:\n  - name: bar in an SLS gives "not avilable", as does the same on a minion with salt-call.  What am I missing?
02:38 forresta chuyskywalker, Weird
02:39 forresta chuyskywalker, It's odd it just started working :\
02:39 chuyskywalker forresta: ^shrug^ like I said on the gh issue: "heisenbug" :|
02:39 forresta hah
02:44 clearclaw I can also do `salt target foo.method bar` successfully.  WTF can't I do that in an SLS or directly on the minion?
02:45 clearclaw (and what's a good way to debug?)
02:46 drawsmcgraw beach_: I agree it's definitely brutal but it's the best I've come up with so far :/
02:46 beach_ cool deal. thanks!
02:50 drawsmcgraw Anyone using user.present -> I'm trying to find a way to generate the password hash to give to the 'password' parameter
02:50 drawsmcgraw Having a hard time finding it....
02:52 jfzhu_us drawsmcgraw: From (https://gist.github.com/UtahDave/3785738) python -c "import crypt; print crypt.crypt('password', '\$6\$SALTsalt\$')"
02:52 drawsmcgraw jfzhu_us: Sweet, thanks!
02:52 jfzhu_us where \%6\ is sha1sum and SALTsalt is a random string that you salt with and password is the password that you want to be encrypted
02:55 xl1 joined #salt
03:02 lahwran joined #salt
03:04 lahwran joined #salt
03:05 gamingrobot joined #salt
03:06 mjulian joined #salt
03:09 lahwran_ joined #salt
03:11 oz_akan_ joined #salt
03:20 rojem joined #salt
03:36 rojem joined #salt
03:46 rojem joined #salt
03:56 rojem joined #salt
04:02 scooby2_ joined #salt
04:10 favadi joined #salt
04:27 kedo39 joined #salt
04:28 bhosmer joined #salt
04:31 ajw0100 joined #salt
04:36 favadi joined #salt
04:40 jdenning joined #salt
04:41 oz_akan_ joined #salt
04:41 rojem joined #salt
04:42 rojem joined #salt
04:57 clearclaw What I've been missing is the difference between execution and state modules as discussed here: http://stackoverflow.com/a/20788572
04:57 clearclaw Argghhhhh!
04:57 * clearclaw used to have more hair.
04:59 akoumjian joined #salt
05:01 Linz joined #salt
05:09 forresta clearclaw, lol
05:14 abele_ joined #salt
05:18 forresta JordanRinke, you around?
05:19 anuvrat joined #salt
05:19 forresta JordanRinke, Anyways when you see this, https://github.com/saltstack-formulas/fail2ban-formula now exists
05:34 elfixit joined #salt
05:36 trevorj joined #salt
05:42 micko joined #salt
05:54 micko joined #salt
06:00 micko joined #salt
06:08 grep_away joined #salt
06:09 taion809 joined #salt
06:11 zz_cro joined #salt
06:11 troyready joined #salt
06:16 bhosmer joined #salt
06:17 ravibhure joined #salt
06:45 middleman_ joined #salt
06:54 pranav joined #salt
06:56 pranav hi anybody here could help me installing pymongo package in salt minion environment of windows
07:06 rmt joined #salt
07:08 Heartsbane joined #salt
07:12 redmin joined #salt
07:17 srage joined #salt
07:18 srage_ joined #salt
07:23 alekibango joined #salt
07:30 ravibhure joined #salt
07:35 sroegner_ joined #salt
07:37 pranav left #salt
07:40 pranav___ joined #salt
07:45 fllr joined #salt
07:46 _fllr_ joined #salt
07:48 1JTABPSAQ joined #salt
07:48 dpn` joined #salt
07:48 abele joined #salt
07:48 [vaelen] joined #salt
07:48 nliadm joined #salt
07:48 crane joined #salt
07:48 Teknix joined #salt
07:48 Kyle joined #salt
07:48 seanz joined #salt
07:48 srijan4 joined #salt
07:48 modafinil joined #salt
07:48 Jahkeup joined #salt
07:48 dpac|away joined #salt
07:48 echos joined #salt
07:48 markizano joined #salt
07:49 ConceitedCode joined #salt
07:51 Shenril joined #salt
08:04 bhosmer joined #salt
08:08 ConceitedCode joined #salt
08:09 matanya joined #salt
08:09 matanya joined #salt
08:12 tonthon joined #salt
08:12 cym3try joined #salt
08:13 redmin joined #salt
08:17 ConceitedCode joined #salt
08:21 harobed_ joined #salt
08:23 ConceitedCode joined #salt
08:33 troyready joined #salt
08:34 kadel joined #salt
08:39 pears joined #salt
08:41 helminthe joined #salt
08:44 oz_akan_ joined #salt
08:47 fllr joined #salt
08:49 bhosmer joined #salt
08:59 lynxman joined #salt
09:07 pears left #salt
09:11 njpatel joined #salt
09:11 fatbox joined #salt
09:14 jpcw joined #salt
09:14 redmin joined #salt
09:15 elfixit joined #salt
09:22 ckao joined #salt
09:23 sroegner_ joined #salt
09:32 srage joined #salt
09:35 srage_ joined #salt
09:37 abk joined #salt
09:43 ravibhure joined #salt
09:45 oz_akan_ joined #salt
09:47 fllr joined #salt
09:51 troyready joined #salt
09:51 gasbakid joined #salt
09:51 srage joined #salt
09:51 jcockhren joined #salt
09:53 linjan joined #salt
09:55 svx joined #salt
09:56 bhosmer joined #salt
10:09 smkelly joined #salt
10:10 namtab joined #salt
10:12 che-arne joined #salt
10:13 whiskybar joined #salt
10:14 redmin joined #salt
10:16 psyl0n joined #salt
10:16 helminthe joined #salt
10:29 helminthe1 joined #salt
10:30 tinuva joined #salt
10:33 helminthe joined #salt
10:35 techdragon joined #salt
10:36 gasbakid_ joined #salt
10:37 helminthe1 joined #salt
10:38 viraptor joined #salt
10:40 gasbakid joined #salt
10:45 redmin joined #salt
10:46 oz_akan_ joined #salt
10:47 fllr joined #salt
10:49 fllr joined #salt
10:49 Linz joined #salt
10:58 ronc_ joined #salt
10:58 ronc_ hey guys. What could cause this: State raid.present found in sls raid is unavailable
11:01 ronc_ version is 0.17.4. Docs state 0.17.4 should have mdadm/raid: http://salt.readthedocs.org/en/v0.17.4/ref/states/all/salt.states.mdadm.html?highlight=mdadm#salt.states.mdadm
11:02 Furao ronc_: https://github.com/saltstack/salt/blob/0.17/salt/states/mdadm.py#L31
11:07 krak3n` joined #salt
11:09 derelm joined #salt
11:10 mortis_ anyone know if salt will have some sort of scheduled releases soon? i mean ...so i know when to expect the next release
11:12 sroegner_ joined #salt
11:14 MrTango joined #salt
11:15 krak3n` mortis_ hydrogen is 1 month late according to the GitHib issue tracker and still has loads of open issues: https://github.com/saltstack/salt/issues?milestone=39&state=open
11:15 mortis_ oooh lookie here :)
11:15 mortis_ thanks krak3n`
11:15 krak3n` np :)
11:19 xerxas joined #salt
11:20 cym3try joined #salt
11:22 ravibhure joined #salt
11:24 Shish joined #salt
11:24 psyl0n joined #salt
11:24 psyl0n joined #salt
11:28 TomasNunez joined #salt
11:42 thoht hi
11:42 thoht when i launch state.sls xxx.xxx on a minion, i got a timetout
11:43 thoht Waited 60 seconds
11:43 thoht if i run a simple command like cmd.run "ls" it works
11:43 thoht why state.sls is not working ?
11:44 thoht i m trying to synch 17M
11:45 ravibhure joined #salt
11:45 redmin joined #salt
11:47 oz_akan_ joined #salt
11:47 fllr joined #salt
11:48 njpatel joined #salt
11:49 cnf joined #salt
11:51 TomasNunez joined #salt
11:53 bhosmer joined #salt
11:56 Furao thoht: salt -t $highvalue
12:03 ronc_ Furao: thanks.
12:06 thoht Furao: what does that mean ? even with a simple file, i got the timeout
12:07 Nazca joined #salt
12:07 Nazca joined #salt
12:08 thoht salt.exceptions.SaltReqTimeoutError: Waited 60 seconds
12:09 sroegner_ joined #salt
12:12 techdragon joined #salt
12:13 Furao thoht: does it reply to test.ping ?
12:15 thoht True
12:17 thoht Furao: http://pastebin.com/TGLdAbmK
12:17 thoht this is the traceback
12:23 Furao thoht: you should run both master and minion with debug log level
12:23 Furao to see what is going on
12:24 Furao but you can use salt-run jobs.lookup_jid 20140106131607619572
12:24 Furao on the master to see the output
12:24 Furao for some reason the minion finish process the job
12:24 Furao but your client don't timeout
12:25 logix812 joined #salt
12:28 Linz joined #salt
12:29 sulky joined #salt
12:44 gadams999 joined #salt
12:46 redmin joined #salt
12:47 techdragon joined #salt
12:47 fllr joined #salt
12:47 oz_akan_ joined #salt
12:48 thoht Furao: http://pastebin.com/z9gZMvT5 tracekback of master
12:49 Furao it look normal
12:49 thoht on minion, i can see: [INFO    ] File /etc/init.d/tomcat1 is in the correct state
12:49 Furao left #salt
12:49 thoht so the command was run as well
12:49 Furao joined #salt
12:49 th3reverend joined #salt
12:52 thoht Furao: http://pastebin.com/hB3cazdp
12:52 thoht we can see on both side that the command was returned as well
12:52 th3reverend1 joined #salt
12:52 th3reverend1 left #salt
12:52 thoht i don t understand why the output of the command is salt.exceptions.SaltReqTimeoutError: Waited 60 seconds
12:55 Furao thoht: does it need really take more than 60 sec?
12:56 thoht no it was 2 seconds
12:56 msciciel__ joined #salt
12:57 msciciel__ does salt provides something to generate configuration file for something (eg. nagios) based on data from others minions ?
12:58 Furao msciciel__: I did implemented that
12:58 Furao but for shinken, a nagios replacement
12:58 Furao using salt mine
12:58 msciciel__ salt mine is a clue ? :)
12:58 Furao salt mine is a module
12:59 Furao minion send data to master
12:59 Furao and the minion that render config file to monitor use that data
13:00 blee joined #salt
13:00 toastedpenguin left #salt
13:00 thoht Furao: now it says:  The function "state.sls" is running as PID 31191 <== this pid don t exist
13:00 Furao thoht: it exists in the minion not on the master
13:00 Furao salt-run jobs.list_jobs
13:01 thoht Furao: i did  salt-run jobs.active and it returned {}
13:02 msciciel__ Furao: have you got any publice example of using salt mine ? i'm search google but i cannot find something for beginner
13:02 thoht ok i did re-run it and it successfully finished
13:02 thoht but it was damn long
13:02 Furao msciciel__: last time I checked there was no doc or just not enough to easily understand it
13:03 msciciel__ Furao: i think that nothing changed in this topic, documention says nothing how to use in practice
13:04 Furao http://salt.readthedocs.org/en/latest/topics/mine/index.html
13:04 Furao http://docs.saltstack.com/ref/modules/all/salt.modules.mine.html#module-salt.modules.mine
13:05 Furao in my case I just use {% for minion_data in salt['mine.get']('*', 'bleh') %}{# build config for shinken #}{% endfor %}
13:06 thoht i do want to use file.directory for synch /usr/local/tomcat for instance
13:06 thoht but it synched only the directory without the subdirs
13:06 msciciel__ Furao: ok it's one part a how did you configure miinon to provide data ?
13:06 thoht anyway to tell salt to synch in recursive ?
13:07 msciciel__ Furao: documentation mention to use minion pillar or via minion config file
13:07 Furao msciciel__: part of
13:07 Furao the other is that I wrote my own salt module that the minion call to send it's data to master
13:07 gadams9991 joined #salt
13:08 jfzhu_us joined #salt
13:11 Furao well I'm back now from vacation and ready to take some new salt consulting contracts :) we can assist you with this (had been doing salt consulting for more than a year)
13:11 Furao thoht: pas besoin de mettre un h
13:12 cym3try joined #salt
13:12 nebuchadnezzar joined #salt
13:12 Furao thoht: recurse: True ?
13:12 Furao oh no this is for permission
13:13 msciciel__ Furao: so in pillar i need to define module which will be used with mine, and create simple salt module which return list of checks from minion (or use grains) , sounds easier now
13:13 Furao thot: i guess you can use cp.list_master to build a list and for each of them use file.managed
13:14 Furao msciciel__: kind of
13:14 Furao msciciel__: in my case, I had to add more logic to cover all cases
13:14 Furao such as which IP to connect to the host? internal, external?
13:14 Furao and some per-service options
13:15 msciciel__ Furao: i'm trying to setup passive monitoring so i need only to have list of checks from minions
13:15 Furao ok!
13:16 msciciel__ Furao: thanks to You i know what i need to accomplish my task :)
13:16 Furao we have too much nrpe checks and we're switching to passive for some of them
13:16 Furao here https://github.com/bclermont/states
13:16 Furao there is some examples of nrpe check for some states
13:17 yannj joined #salt
13:17 msciciel__ Furao: we have own daemon for checking machine and for migration process i need something that will fetch list of checks from it and provides data for salt to generate icinga config
13:17 Furao but this is so old, today it's totally different, i use monitoring checks during states testing (not just production)
13:18 gasbakid joined #salt
13:19 gadams999 joined #salt
13:19 juice joined #salt
13:19 yannj Hello there, is there any way to have a reactor event dispatched when grains are updated, I am trying to read externally grains from a django app, and react with it
13:20 Furao yannj: I wrote a django application as a inventory system for salt and use ext_grains module that connect to it's tastypie REST API to feed pillars to minion
13:21 Furao and in some condition I send events to the bus that trigger reactor and perform some actions
13:21 Furao but there is no built-in event for grains change
13:22 Furao I never understood why there is a grains.set function
13:22 jfzhu_us joined #salt
13:23 Furao grains should be build at minion startup
13:24 thoht Furao: yes  i can use cp.list_master but i though it was possible to use file.directory
13:26 thoht Furao: so no other choice to use cp.get_dir ?
13:26 yannj My question is stupid, but would it be interesting to have the grains as an attribute of salt start event
13:28 Furao yannj: if you trigger the event on the minion itself, you can use grains value as parameter, yes
13:29 yannj Furao : I mean at this time there is an event triggered automatically when a minion start. Wouldn't it be a good idea to have the grains sent at the same time?
13:31 Furao just use http://docs.saltstack.com/ref/modules/all/salt.modules.event.html#module-salt.modules.event and pass grains value
13:31 Furao you can use module.run in a .sls file
13:34 gasbakid joined #salt
13:39 thoht Furao: the thing is that i want to copy the directory but with same permissions
13:39 thoht in the directory there is symlink
13:39 thoht they are not kept
13:40 thoht i m going to do a tar at the END ?
13:41 ronc_ guys, my top file has exploded. Any ideas how to reorganize?: https://gist.github.com/roncohen/3eb249b5749ed12bd602
13:41 cnelsonsic joined #salt
13:44 canci joined #salt
13:46 Furao thoht: same permission than?
13:46 thoht Furao: the directory source contains symlinks
13:46 Furao thoht: use {# #} instead of #
13:47 thoht when i used cp.get_dir, instead of keeping the symlink, it copied the folders
13:47 thoht Furao: what do you mean ?
13:47 redmin joined #salt
13:47 fllr joined #salt
13:47 bhosmer joined #salt
13:47 Furao thoht: sorry this was for ronc_
13:48 Furao ronc_: use  {# #} instead of #
13:48 thoht Furao: using  cmd.run "cd /usr/local/tomcat1 ; tar xvf websc_tomcat1.tgz" at the end
13:48 thoht too bad
13:48 thoht very desapoint to not be able to use the file.directory module in top.sls
13:50 Furao keep top.sls simple don't put too much logic there
13:50 Furao just refer to other .sls
13:50 thoht Furao: yes to other .sls, but same at the end
13:50 thoht i can t use .sls files
13:50 thoht for directies
13:50 thoht folders
13:51 ronc_ Furao: was that for me? i only have one comment :)
13:56 brianhicks joined #salt
13:58 wkf_ joined #salt
13:58 fllr joined #salt
13:58 colinbits joined #salt
14:00 _fllr_ joined #salt
14:01 ekristen joined #salt
14:01 grep_awesome joined #salt
14:05 rgbkrk joined #salt
14:05 redmin joined #salt
14:11 timoguin joined #salt
14:12 juicer2 joined #salt
14:13 oz_akan_ joined #salt
14:13 pdayton joined #salt
14:16 JasonSwindle joined #salt
14:17 dw joined #salt
14:17 oz_akan_ joined #salt
14:17 Mua joined #salt
14:18 Gifflen joined #salt
14:19 Furao joined #salt
14:19 racooper joined #salt
14:23 jeffrubic joined #salt
14:28 Mua_ joined #salt
14:29 techdragon joined #salt
14:29 Gifflen joined #salt
14:33 mattikus joined #salt
14:36 jslatts joined #salt
14:36 gasbakid_ joined #salt
14:38 backjlack joined #salt
14:42 Linz joined #salt
14:42 wkf_ joined #salt
14:43 pdayton joined #salt
14:43 quickdry21_ joined #salt
14:44 toastedpenguin joined #salt
14:45 vejdmn joined #salt
14:46 mpanetta joined #salt
14:47 fllr joined #salt
14:50 vejdmn joined #salt
14:52 fishpen0 joined #salt
14:56 tvs_ joined #salt
14:58 tvs_ hi, I need to delete a package in FreeBSD if a specific version is installed .... how can I put this in a state file?
14:59 Gifflen joined #salt
14:59 tvs_ I know I can get the version of a  package with pkg.version, but I don't know how to use it iwithin an 'if' condition
15:00 th3reverend joined #salt
15:01 th3reverend left #salt
15:01 ravibhure joined #salt
15:01 timoguin {% if __salt__['pkg.version']('apacheblahblah') equals "12345" %}
15:01 krak3n` :o
15:02 timoguin tvs_, and then use pkg.removed inside that, closing it with a {% endif %}
15:02 timoguin should do the trick
15:03 tvs_ thank you timoguin !!!
15:03 Whissi joined #salt
15:03 timoguin welcome!
15:04 gadams999 joined #salt
15:06 Linz joined #salt
15:07 gadams9991 joined #salt
15:08 bhosmer joined #salt
15:11 btorch joined #salt
15:11 favadi joined #salt
15:13 jdenning joined #salt
15:13 tyler-baker joined #salt
15:14 davidfischer joined #salt
15:18 TheRealBill joined #salt
15:18 sroegner_ joined #salt
15:19 ravibhure joined #salt
15:20 davidfischer joined #salt
15:21 robinsmidsrod joined #salt
15:23 taion809 joined #salt
15:23 diegows joined #salt
15:27 ndrei joined #salt
15:28 kasanop joined #salt
15:31 mattmtl joined #salt
15:32 che-arne joined #salt
15:32 higgs001 joined #salt
15:35 danielbachhuber joined #salt
15:36 fishpen0 joined #salt
15:37 jrdx joined #salt
15:38 cachedout joined #salt
15:38 forresta joined #salt
15:41 aleszoulek joined #salt
15:41 TheRealBill joined #salt
15:44 vejdmn joined #salt
15:45 fishpen0 joined #salt
15:51 tmmt joined #salt
15:53 blee joined #salt
15:55 bhosmer_ joined #salt
15:58 jcsp joined #salt
15:59 jalbretsen joined #salt
16:00 Marion joined #salt
16:00 bhosmer_ joined #salt
16:01 svx joined #salt
16:03 svx joined #salt
16:03 UtahDave joined #salt
16:07 kaptk2 joined #salt
16:11 opapo joined #salt
16:13 forresta timoguin, are you around?
16:14 timoguin yerp
16:14 forresta ok, don't use that fail2ban formula I started working on, use this one instead: https://github.com/saltstack-formulas/fail2ban-formula
16:16 forresta it's much more complete
16:16 timoguin ah! messin' stuff up for me! ;-)
16:16 rojem joined #salt
16:16 forresta hah
16:16 timoguin i didn't think there was an official one
16:17 forresta kedo39 only committed it a few days ago
16:17 forresta so when it breaks you can blame him :P
16:17 timoguin I so will.
16:18 JasonSwindle Anyone using SaltStack 2014.1.0 RC1 and SaltCloud?
16:18 JasonSwindle I am having problems and not sure if me or me.....
16:18 forresta not yet JasonSwindle, was too lazy to set it up this weekend.
16:18 forresta I'm gonna go with "It's you"
16:18 forresta :P
16:18 forresta since that seems to be the norm
16:18 JasonSwindle Yeah……… I hope it is me.
16:19 timoguin are you talking about the new cloud module/state?
16:19 JasonSwindle Let me get logged in so I can get the error / data.
16:19 JasonSwindle I am just trying to get SaltCloud to build and deploy Salt.
16:20 timoguin well, with the RC, salt-cloud has been merged into salt, so it's no longer a separate project
16:20 timoguin are you using the salt-cloud command?
16:21 JasonSwindle Yep, I saw that.
16:21 JasonSwindle Yep
16:21 JasonSwindle and it builds the nodes.
16:21 JasonSwindle but salt bootstrap fails.
16:21 timoguin gimme dem errors
16:21 davet joined #salt
16:21 UtahDave JasonSwindle: what's the error you getting?
16:22 JasonSwindle Working on that now…… need to get into my Master and rebuild the nodes and keep the tmp
16:22 gadams999 joined #salt
16:23 forresta the bootstrap fails?
16:23 forresta that seems odd
16:23 JasonSwindle I agree, very.
16:23 forresta usually that piece is pretty solid
16:23 JasonSwindle Yep
16:23 nahamu UtahDave: I have some extra files I use for generating the SmartOS esky builds and I was thinking I should get them into the salt repo under pkg/smartos/esky
16:23 nahamu https://github.com/nshalman/salt/commit/e94f42449a18922e77332e807bc0f02aa9a08c04
16:24 nahamu In theory that commit should be good for both the develop branch and the 2014.1 branch (it's just some new files in a new subdirectory and nothing else)
16:25 nahamu If I wanted to get it into both branches should I open two pull requests, a single pull request against one of those branches? what's the current procedure?
16:26 forresta UtahDave, looks like issue 274 on bootstrap is a dupe of this: https://github.com/saltstack/salt-bootstrap/issues/69
16:27 JasonSwindle forresta:  UtahDave Building nodes now.
16:27 Gareth 'lo
16:27 JasonSwindle Howdy Gareth
16:27 Gareth JasonSwindle: howdy. hows it going?
16:28 JasonSwindle Goes great.  Got the SaltConf is x days away panic going.  :)
16:30 JasonSwindle UtahDave:  I get "[ERROR   ] Failed to deploy and start Salt on Cloud VM dev-iad3-lb-n001.example.com"
16:30 JasonSwindle Working on getting more information on what the heck that means....
16:30 timoguin that's not a very helpful message.
16:30 timoguin JasonSwindle, try running with -l debug
16:31 JasonSwindle I also kept tmp on the node
16:31 JasonSwindle I am hoping I can dig something out on that, too
16:35 UtahDave nahamu: Just do a pull request against the develop branch and make a note in the comment that this would be good to have cherry picked to 2014.1 branch
16:35 analogbyte are there any problems using the service state for salt-master/salt-minion and watching the configuration? Do they restart themselves without issue?
16:36 abk left #salt
16:36 JasonSwindle timoguin:  Running again with Debug……. /tmp was empty
16:36 UtahDave analogbyte: usually that works just fine.  I would test it on a couple minions first because there were a couple instances where a distro's init system didn't play well
16:38 nahamu UtahDave: thanks!
16:38 UtahDave Thank you, nahamu!  You've done a LOT of awesome work with SmartOS and Salt
16:38 nahamu UtahDave: Scratching my own itch. :)
16:39 analogbyte UtahDave: thanks for the info, I'll report back if something doesn't work ;)
16:40 JesseC joined #salt
16:40 nahamu Done. https://github.com/saltstack/salt/pull/9584
16:40 UtahDave thanks, nahamu!
16:40 JasonSwindle timoguin:  I have the log…. and it looks like Salt is having problems logging in.  But using the same command / password and I get in.
16:41 redmin joined #salt
16:42 bbinet hi
16:42 JasonSwindle Howdy
16:42 timoguin JasonSwindle, so are you specifying the user/pass in the cloud config? I've only ever used keys with salt-cloud
16:42 bbinet is it possible to configure salt authentication so that a non-root user can issue salt-key -L ?
16:42 timoguin I have salt running as the "salt" user, and it has the default key that all my VMs have when they're launched
16:43 timoguin bbinet, yes, it is. i've done that simply by making my main user a member of the "salt" group.
16:43 timoguin and setting permissions appropriately for /etc/salt, /var/cache/salt, /var/log/salt
16:43 bbinet timoguin: thanks, I will try that
16:44 JasonSwindle timoguin:  I am not.  It is trying to use root (which is correct) and is trying to use the node's new password (which is also correct)
16:44 JasonSwindle https://gist.github.com/JasonSwindle/a51dcda7d586acc2bfa0
16:44 JasonSwindle UtahDave:  ^
16:45 JasonSwindle UtahDave:  This map file which works in all of 0.16.x and 0.17.x FYI
16:46 timoguin its new password? where are you setting that in your config?
16:46 JasonSwindle I added on         log_level: debug script: bootstrap-salt script_args: -c /tmp/ for debugging.
16:47 fllr joined #salt
16:48 bbinet timoguin: hum, but if I use the salt group workaround, I guess the user will also have the right to issue a "salt -A" command
16:48 bbinet which I'd like to avoid
16:49 timoguin bbinet, hmm, yea. i'm not sure what the best practice is for that.
16:50 bbinet I'd like to allow only listing the current salt keys like "salt-key -L" do
16:50 JasonSwindle UtahDave:  Does this help any?
16:50 JasonSwindle What more can I supply? :)
16:50 EugeneKay THe beer!
16:51 UtahDave bbinet: You could add that user to your sudoers file and only allow them to execute  "salt-key -L"
16:52 UtahDave JasonSwindle: looking right now
16:52 JasonSwindle kk
16:53 bbinet UtahDave: yes, I was thinking about that
16:53 UtahDave JasonSwindle: try:    sudo salt-cloud -U
16:53 bbinet but wondering if something was directly possible with salt authentication
16:53 bbinet http://docs.saltstack.com/topics/eauth/index.html
16:53 JasonSwindle That is the salt bootstrap update, correct?
16:53 Guest51501 joined #salt
16:53 UtahDave JasonSwindle: Yeah.
16:54 JasonSwindle UtahDave:  I can run that again.  I did it on Saturday.
16:54 UtahDave JasonSwindle: And if I remember correctly, I think the decision was made to not make the bootstrap script a hard dependency for Salt.
16:54 bbinet anyway, thanks UtahDave, I'll use /etc/sudoers :)
16:54 JasonSwindle UtahDave:  Done, let me try again.  Any other things to try before I run it?
16:55 UtahDave bbinet: You're welcome. I think that's your best bet.  I'm pretty sure Salt authentication doesn't get that granular for salt-key
16:55 UtahDave JasonSwindle: give it a spin!  :)
16:56 JasonSwindle Wiping nodes, and building again
16:56 nhubbard joined #salt
16:57 UtahDave back in a few.  meeting
16:57 forresta UtahDave, You should just announce when you are back, and the standard assumption is you are in a meeting :P
17:03 blee_ joined #salt
17:03 BRYANT__ joined #salt
17:05 ndrei joined #salt
17:05 rojem joined #salt
17:06 alunduil joined #salt
17:07 xmltok joined #salt
17:08 rlarkin joined #salt
17:12 bhosmer joined #salt
17:14 tvs_ I have this error when trying to execute 'salt-call state-highstate' from a FreeBSD minion: local:     Data failed to compile: ----------     No function declared in state "cmd:run" in sls perlmods
17:15 jimallman joined #salt
17:15 scoates I'm writing a custom _states/ state to help deploy some of our apps. I'd like to reference/render a template in there. Can someone point me at docs on how to do this (including: where do I put the jinja file)?
17:16 tvs_ here's the relevant definition: http://pastebin.com/8JgSfEqm
17:17 forresta tvs_, can you re-run that with -l debug to see if it gives you more details?
17:18 scoates tvs_: cmd.run, not cmd:run on line 7?
17:18 forresta nice catch scoates
17:19 tvs_ right!!!!!!!!
17:19 tvs_ thanks scoates!
17:19 scoates sure.
17:19 forresta scoates regarding your question, do you mean a jinja template or something?
17:20 scoates forresta: yeah. I want to render a jinja template from pythonland (and I need to know how to reference the file)
17:21 forresta Well, there's this, but it only applies to state files, not custom states: https://salt.readthedocs.org/en/latest/ref/states/vars.html
17:21 forresta I don't know if I've seen an actual state that does that, you might want to look through the states dir over at the main repo to see.
17:21 scoates yeah. it's kind of hard to search for this because results are almost all .sls, not .py
17:22 KyleG joined #salt
17:22 KyleG joined #salt
17:23 forresta scoates, yea you might hav to manually look through: https://github.com/saltstack/salt/tree/develop/salt/states
17:24 forresta because you actually want to the use the template itself, in your state right?
17:25 scoates yep
17:25 scoates looking there now.
17:26 scoates I think I can probably do something like `__salt__['file.managed']…`
17:26 forresta Hmm, if you find one that does such a thing let me know, I've never seen something like that...
17:26 troyready joined #salt
17:26 derelm joined #salt
17:26 scoates just not sure where I should put the actual .jinja… I guess in a generic [roots]/templates
17:26 forresta yeah I have no idea :\
17:27 JasonSwindle I had JINJA in top.sls….. but that seems to have stopped working in 0.17.x
17:27 JasonSwindle But that may of just been the way I was doing it.
17:28 scoates I'll try it, forresta. Will paste if it works. (-:
17:28 forresta awesome
17:29 ajw0100 joined #salt
17:31 gmoro joined #salt
17:31 Linz joined #salt
17:40 Veticus joined #salt
17:47 jdenning joined #salt
17:47 jms2 joined #salt
17:48 ndrei_ joined #salt
17:48 scoates do I need to do something specific to make __salt__ available to my custom states?
17:50 UtahDave scoates: Nope.  The Salt loader adds __salt__ in
17:51 gadams999 left #salt
17:51 scoates UtahDave: I thought so. I'm seeing `KeyError: 'file.managed'` when doing `__salt__['file.managed'](…)` … am I barking up the wrong tree?
17:51 UtahDave forresta: ANNOUNCEMENT:  I am back.    :)
17:51 forresta heh
17:52 UtahDave scoates: Usually that means there was some kind of traceback when the minion started up and wasn't able to load the file module.
17:52 anuvrat joined #salt
17:52 scoates hmm. I just did a successful sync_all.
17:53 scoates or at least I thought it was successful heh
17:53 UtahDave scoates: can you start your minion in a terminal and see if there are any tracebacks?
17:53 scoates http://paste.roguecoders.com/p/6cf869c372da300c9d5ce3db88425012.txt
17:53 scoates yep, sure.
17:53 UtahDave JasonSwindle: How's everything going?
17:55 gadams999 joined #salt
17:55 stoffell joined #salt
17:55 scoates UtahDave: http://paste.roguecoders.com/p/42e172989f8a42fe394414b42e6666a8.txt
17:55 scoates no traceback on startup
17:56 UtahDave scoates: can you pastebin your state?
17:57 scoates yep. sec.
17:59 scoates UtahDave: I think this is everything: http://paste.roguecoders.com/p/64b4675edc232b6b456ec97e1c641742.txt
18:00 druonysus joined #salt
18:02 fishpen0 joined #salt
18:02 bhosmer joined #salt
18:03 mattmtl I am having problems using the API with the module pkg... I am doing this: local.cmd(machine1, 'pkg.install', name='python-devel') --- returns --> {machine1: {}}
18:03 mattmtl But on machine1, yum list installed | grep mako is empty
18:03 mattmtl Also, when doing it using the CLI, the package will install
18:03 darless joined #salt
18:05 JasonSwindle UtahDave:  [ERROR   ] Failed to deploy and start Salt on Cloud VM dev-iad3-util-n001.example.com
18:05 JasonSwindle It still failed even after the salt-cloud -u
18:05 darless joined #salt
18:06 UtahDave scoates: can you try pprinting    __salt__   ?
18:06 scoates yep
18:07 UtahDave JasonSwindle: OK, could you open an issue on that?  Make sure to mention you're using the RC
18:07 darless joined #salt
18:08 scoates oh…
18:08 scoates UtahDave: is file.managed not exported, somehow?
18:08 JasonSwindle UtahDave:  Sure.
18:08 scoates -> http://paste.roguecoders.com/p/cd8f9036d06117607e7c7400439299c4.txt
18:09 JasonSwindle cachedout:  Got one more issue to submit…..
18:09 cachedout Sup?
18:09 UtahDave scoates: Ahh. got it.
18:09 JasonSwindle Having Salt-Cloud / Salt Bootsrap issues.
18:09 JasonSwindle Creating issue on it now.
18:09 UtahDave scoates: __salt__ has all of Salt execution modules, not states.
18:09 Ahlee_ Here's a new one for me.  SaltMaster was under load from attempting to handle responses from ~200 minions, this poped up on a few minions during that run http://pastebin.com/gPby9ciZ
18:09 UtahDave file.managed is a state, not an execution module
18:09 scoates ah
18:09 scoates that makes more sense.
18:10 cachedout JasonSwindle: OK. Sounds good.
18:10 UtahDave Ahlee_: what version of salt?
18:10 danielbachhuber joined #salt
18:11 scoates UtahDave: I'll try with __salt__['file.manage_file'] . thanks.
18:11 jacksontj joined #salt
18:13 UtahDave you're welcome, scoates
18:16 mattmtl is there a way to easily debug clients API calls?
18:16 Voziv joined #salt
18:18 fishpen0 joined #salt
18:19 UtahDave sorry, mattmtl. I'm not ignoring you, I'm just not very knowledgable about using the api.  whiteinge, do you know what's causing mattmtl this problem?
18:20 JasonSwindle UtahDave:  Does this help? https://github.com/saltstack/salt/issues/9587
18:20 JasonSwindle cachedout:  https://github.com/saltstack/salt/issues/9587 Do you see if I need to add any more?
18:21 UtahDave JasonSwindle: Yeah, that's perfect. Thanks!
18:21 UtahDave obviously, RC bugs are the high priority right now, so we'll get right on that.
18:21 JasonSwindle Thank you.   This sadly is a road-block for me on using the RC.
18:22 Corey UtahDave: Is 2014.1.0rc1 still the current RC?
18:22 JasonSwindle Corey:  Going by the tags, it seems so
18:22 JasonSwindle https://github.com/saltstack/salt/tree/v2014.1.0rc1
18:22 UtahDave Corey: Yes, it is.
18:23 ndrei joined #salt
18:25 Corey Goody.
18:26 * Gareth shakes his head at Corey and gives a disapproving look
18:26 forresta lol
18:26 mattmtl Oh, I get this on my minion:  [ERROR   ] No package sources passed to pkg.install.
18:27 mattmtl how come the result from my cmd is {}?
18:29 Gareth UtahDave: are the RC issues tagged somewhere?
18:29 blee joined #salt
18:29 thrashr888 joined #salt
18:30 JasonSwindle That is a great question Gareth.
18:30 JasonSwindle That way I do not dup other issues.
18:30 JasonSwindle Or etc.
18:30 blee joined #salt
18:31 Thiggy joined #salt
18:32 jslatts joined #salt
18:33 troyready joined #salt
18:33 william_20111 joined #salt
18:38 harkx I have a lot of systems with some specific config (postfix, apache config files, etc..), is it common to use a specific init.sls for specific minions ? (like: all dev machines get the dev-core config but each also gets its own specifics) - like this: http://paste.debian.net/74612/
18:39 forresta harkx, some people do that, it's more common to have your 'core' however, then a single item say for dev, where the files use pillar and grain data to ensure the machine gets the correct info
18:40 harkx hm, are there any resources into practical usage of this? (as in, how far can I go to tune a specific machine)
18:40 JasonSwindle harkx:  It would be better to use include:
18:41 JasonSwindle and pillar, keep it DRY/ :)
18:41 gadams9991 joined #salt
18:41 amckinley joined #salt
18:43 harkx ok, thanks a lot ! I will look further into pillars, grains and includes.. can't wait to meet some hands-on experienced people at cfgmanagement in a couple of weeks :p
18:43 jcockhren lesson learned... don't code while under the influence of anything. sheesh
18:44 forresta jcockhren, are you writing crappy code again/
18:44 harkx jcockhren,  does that include caffeine? :)
18:44 JasonSwindle jcockhren:  ouchy?
18:44 jcockhren harkx: I don't count caffeine. ;)
18:45 forresta Really? The guys drinking 4 cups of coffee and a red bull always seem to be the worst
18:45 jcockhren forresta: luckily, I knew better than to push
18:46 forresta lol
18:46 Ahlee_ UtahDave: 0.17.2
18:48 khaije1 joined #salt
18:48 scoates what is sfn?
18:48 foxx joined #salt
18:48 khaije1 does Salt support setting roles by AD groups (or similar)?
18:55 JasonSwindle joined #salt
18:56 forresta scoates, http://en.wikipedia.org/wiki/8.3_filename maybe?
18:57 scoates looks like "source file name" … but unclear.
18:58 scoates https://github.com/saltstack/salt/blob/develop/salt/modules/file.py#L2453
18:58 forresta shouldn't that be source
18:58 nhubbard is it fair to say that salt has better integration with RHEL then with Debian?
18:59 forresta I don't think so
18:59 forresta nhubbard, what makes you say that?
18:59 scoates source and sfn are different… ?
18:59 forresta sorry that was directed at nhubbard
18:59 scoates file.managed sends sfn=''
19:00 nhubbard I'm just starting with trying to set some stuff up to manage some machines in salt, and I'm noticing there is support for managing ip/interfaces but that isn't available on debian
19:01 forresta for networking? http://docs.saltstack.com/ref/states/all/salt.states.network.html
19:02 foxx joined #salt
19:02 Gareth nhubbard: Networking support for Debian/Ubuntu machines is coming in the next release.
19:03 nhubbard Gareth: yeah I find the debian_ip module in the latest docs, but it's not in the current release. Are there other things like this? Where RHEL has support but Debian doesn't or the other way around
19:04 forresta Gareth, good to know
19:05 Gareth nhubbard: that was the one thing I noticed, so I added the debian_ip module :) I haven't noticed anything else major, but always looking for something new to contribute so if you find something that you can't do on Debian I'd be curious to hear it.
19:06 psyl0n joined #salt
19:06 Gareth forresta: I'm assuming this based on what modules I saw in the RC tag :)
19:06 forresta gotcha
19:06 nhubbard Gareth: will do, I have a home lab that I'm wanting to setup with salt to learn more about salt, and have some centos and some debian machines and right now just starting with networking and going to add upon that, but I got hung up on it fairly quick so wasn't sure if there were other things I was going to run into or not
19:07 forresta yea the spread is pretty good, I think the other issue some people have encountered, is there's no simple firewall or whatever it's called on debian
19:08 redbeard2 joined #salt
19:08 nhubbard yeah I've gotten around that by implementing the startup script that rhel uses in debian
19:08 Gareth I was thinking about the iptables module/state the other day.  was curious how it worked.
19:08 Ahlee_ ok, i'm now seeing that ZMQError: Operation cannot be accomplished in current state in every environment
19:08 redbeard2 left #salt
19:11 khaije1 left #salt
19:11 wkf_ joined #salt
19:12 gasbakid_ joined #salt
19:13 ndrei joined #salt
19:14 UtahDave forresta: ANN: I'm back from lunch!   ;)
19:15 UtahDave JasonSwindle: Hey, did you install libcloud?
19:15 JasonSwindle Yes, sir
19:15 JasonSwindle python-libcloud :)
19:15 scoates UtahDave: any idea how I can pass the context to file.manage_file ?
19:16 scoates or maybe I need to pre-render the template before sending it to file.manage_file?
19:16 scoates this is quite confusing. /-:
19:17 UtahDave scoates: Yeah, I'm thinking you might want to use file.get_managed()
19:17 tessellare joined #salt
19:17 Gareth forresta: looks like there is now ufw, never used it myself.
19:18 scoates ah. and then pass that to file.manage_file as kwargs?
19:19 djinni` joined #salt
19:19 quickdry21_ joined #salt
19:21 UtahDave scoates: Let's back up a second.  What are you trying to accomplish?
19:22 scoates I'm trying to write a custom state so that we can more easily manage our python + nginx configs. The state should render the vhosts file from a template.
19:23 jdenning joined #salt
19:23 scoates maybe I should just rework this as a "use" inheritance…?
19:25 UtahDave scoates: yeah, maybe
19:25 scoates is there a man page for exactly how "use" works? it's pretty hard to search for that. (-:
19:25 scoates e.g. https://dl.dropboxusercontent.com/1/view/7zg9c2ctlkmshbe/Captured/TVS2n.png
19:26 UtahDave lol, nice.
19:26 scoates aha. found something. https://salt.readthedocs.org/en/latest/ref/states/extend.html will read.
19:27 redmin joined #salt
19:27 DerekRBN joined #salt
19:27 redmin joined #salt
19:28 DerekRBN How does salt check if a service is running? Is it the pid or a status call? anyone know off hand?
19:28 Gareth nhubbard: fyi, there is a package called iptables-persistent that looks interesting.
19:28 vejdmn joined #salt
19:28 nhubbard Gareth: thanks, I'll check it out, from a quick look, it seems to be what I have done, so a package would be nice
19:29 gadams999 joined #salt
19:29 UtahDave DerekRBN: it can do either. I think it's pid, by default
19:29 linx_0x from the master settings: "The file server works on environments passed to the master" .. how can I pass an environment to the master?
19:30 DerekRBN So if my pid file doesnt match the name of the service that might be why it's not starting it?
19:30 UtahDave DerekRBN: http://docs.saltstack.com/ref/states/all/salt.states.service.html#salt.states.service.running
19:30 UtahDave DerekRBN: Yeah, I think you have to pass the - sig   option so it can find it in the output of   ps
19:30 DerekRBN Derp sig is what im looking for
19:31 abe_music joined #salt
19:32 DerekRBN Thank you!!
19:32 UtahDave did that work?
19:33 abe_music is there a way to tell salt-cloud to bootstrap minions in a map file? we tried launching a 200 node cluster, 183 started, and then something on the EC2 side failed causing salt-cloud to fail with 183 machines not doing anything :(
19:34 DerekRBN Yes sig was what I needed. my spawn-fcgi process is called php-cgi
19:34 abe_music i would like to kick off the bootstrap and have the minions come up for those 183 instead of terminating them (wasting $90 or so)
19:34 tvs_ joined #salt
19:34 tvs_ I got this error: [ERROR   ] Invalid input: PERL_AUTOINSTALL='--defaultdeps' PERL_EXTUTILS_AUTOINSTALL='--defaultdeps' PERL_MM_USE_DEFAULT=1 PERL_MM_NONINTERACTIVE=1, must be a dict or string - yaml represented dict
19:34 DerekRBN when i put sig: php-cgi. did the trick. Finally moving everything to nginx
19:34 UtahDave abe_music: have you tried the saltify driver for salt cloud?
19:35 JasonSwindle Anyone doing Hadoop at SaltConf 2014?
19:35 tvs_ I have defined this in /srv/pillar/perl.sls: http://pastebin.com/SZ5PJicy
19:35 JasonSwindle UtahDave:  Updated Issue, FYI.
19:36 cym3try joined #salt
19:36 linx_0x never mind my last question makes no sense :)
19:36 abe_music UtahDave: i haven't
19:36 tvs_ and I have this in /srv/salt/perl.sls:http://pastebin.com/2NSmWczt
19:37 abe_music i'll take a look
19:38 tvs_ is it ok the definition I have in /srv/pillar/perl.sls ?
19:43 tyler-baker joined #salt
19:44 danielbachhuber joined #salt
19:50 tollmanz joined #salt
19:53 jacksontj joined #salt
19:57 JasonSwindle joined #salt
19:59 ccase_ joined #salt
20:03 tyler-baker joined #salt
20:05 djinni` joined #salt
20:07 jacksontj joined #salt
20:13 schristensen joined #salt
20:16 forresta abe_music!
20:16 forresta what's happening with stackd.io?
20:16 abe_music forresta!!! what's up?!
20:16 forresta It's monday
20:17 forresta so all the things are broken of course
20:17 abe_music it's going...slowly...we ran into an issue with launching larg-ish clusters with overstate.
20:17 abe_music lol
20:17 forresta you friggin slackers
20:17 abe_music i think we have a grasp on that, but now we've encountered another potentially large issue this morning with salt-cloud bailing out when EC2 gets a little flaky
20:17 forresta how dare your free software not work immediately!
20:18 abe_music yeah i know right?
20:18 forresta you guys are still planning to open source it right?
20:18 abe_music absolutely
20:19 forresta cool
20:19 admineric joined #salt
20:20 shinylasers joined #salt
20:21 quickdry21_ joined #salt
20:21 tvs_ hi, anyone could help me with this error: [ERROR   ] Invalid input: PERL_AUTOINSTALL='--defaultdeps' PERL_EXTUTILS_AUTOINSTALL='--defaultdeps' PERL_MM_USE_DEFAULT=1 PERL_MM_NONINTERACTIVE=1, must be a dict or string - yaml represented dict
20:21 mattmtl found some kind of a reason... the CLI is using yumpkg5.py and the client's API command is using yumpkg.py... if I used bootstrap to install salt minion, how possible I ended up having yumpkg.py in my libs?
20:22 tvs_ I've tried to set that environments variables via pillar to install some perl modules using CPAN and I got that error
20:23 tvs_ pillar config: http://pastebin.com/SZ5PJicy   ... state config: http://pastebin.com/2NSmWczt
20:24 abe_music forresta: we'll probably need to add some retry logic to salt cloud...do you know who i should talk to about that?
20:25 forresta abe_music, soundtech, aka redbeard2. He isn't in this channel, but if you send him a message directly he usually responds.
20:25 abe_music forresta: ah yes, i remember now...thanks!
20:25 forresta np
20:26 gadams999 joined #salt
20:27 vejdmn joined #salt
20:27 dfinn joined #salt
20:29 dfinn does this look OK as a way to detect OS version and use different config files based on that?  I'm getting some errors from this about the first line
20:29 dfinn http://pastebin.com/Mr69PggX
20:30 UtahDave mattmtl: could you open an issue on that?  They should both be using the same package provider on a minion
20:30 forresta dfinn, might help if you close those ifs :P
20:31 forresta {% if grains['osmajorrelease'][0] == '5'} should be {% if grains['osmajorrelease'][0] == '5' %}
20:31 UtahDave forresta: you know s0undt3ch and redbeard2 are two different people, right?
20:31 forresta god not soundtech, techhat
20:31 forresta sorry abe_music, techhat
20:31 UtahDave lol.  OK.  just making sure.
20:31 forresta UtahDave, yes, apparently I cannot think on a monday however :P
20:32 abe_music forresta: got it...redbeard and i are chatting now :)
20:32 forresta cool
20:32 Vye_ Is it possible to run an if statement on a dynamic pillar key when iterating through values? malformed example:  {% for instance in pillar['app_instances'] %} {% if salt['pillar.get']('app_instances:{{ instance }}:version' %} do someting {% else %} do something else
20:32 forresta phew, UtahDave saving me some embarrassment there
20:36 dfinn thanks forresta, that has now lead me to more/different problems ;)
20:36 forresta huzzah!
20:38 cewood joined #salt
20:39 Gareth UtahDave: are you sure about that? have you seen them both in the same room at the same time?
20:40 UtahDave lol, Gareth. You're entirely right. They have never been in the same room at the same time.
20:40 ajw0100 joined #salt
20:41 cym3try joined #salt
20:43 Gareth UtahDave: :)
20:43 dfinn here's the new issue that I'm not seeing a problem with
20:43 dfinn http://pastebin.com/qWPTzDyG
20:44 dfinn it seems to like the user user creation and cmd.run for the nslcd user
20:44 dfinn *to not like
20:48 vejdmn joined #salt
20:49 Vye_ Nevermind, got it to work with: salt['pillar.get']('app_instances:' ~ instance ~ ':version') %}
20:53 octarine joined #salt
20:55 UtahDave dfinn: cmd.run has 3 spaces before it.
20:56 dfinn thank you!  i had actually just found that and then once I fixed that it didn't seem to like the cmd.run syntax so I changed it slightly and now it's working
20:56 dfinn cmd.run:
20:56 dfinn - name: '/usr/sbin/authconfig --enableldapauth --update'
20:59 UtahDave great!
20:59 jacksontj is there a way in the salt master to send an event/job to all master worker processes?
21:01 mattmtl UtahDave: https://github.com/saltstack/salt/issues/9594
21:02 dfinn1 joined #salt
21:02 UtahDave mattmtl: thanks!  Hey, could you add the output of   salt 'minion_name' test.versions_report   ?
21:03 UtahDave jacksontj: I'm not sure exactly what you mean?  Do you want all the processes to execute a Salt command?
21:03 jacksontj i want to run some scheduled task in all the master's worker processes
21:03 jacksontj the Mworker processes
21:04 jacksontj working on another performance "bug" we've run into ;)
21:04 jacksontj right now the master on every pillar request loads all the minion modules per request
21:04 jacksontj which is a LOT of disk activity
21:04 jacksontj so i'm trying to set up some more async-ish method for updating the modules on the master side
21:05 UtahDave Hm. Let me query the big brains
21:05 jacksontj but there are some other tasks that would be better scheduled instead of every request (such as update aes)
21:05 jacksontj in the master server main process we have _clear_old_jobs which is basically a job scheduler now-adays
21:06 nliadm I've got a minion that hangs at '[DEBUG   ] Loaded minion key: /etc/salt/pki/minion.pem'
21:07 nliadm oh, strace claims it's a timeout
21:08 nliadm poll([{fd=28, events=POLLIN}], 1, 0)    = 0 (Timeout)
21:08 nliadm poll([{fd=28, events=POLLIN}], 1, 0)    = 0 (Timeout)
21:08 nliadm poll([{fd=28, events=POLLIN}], 1, 0)    = 0 (Timeout)
21:08 nliadm poll([{fd=28, events=POLLIN}], 1, 60000
21:11 rojem joined #salt
21:11 UtahDave jacksontj: tom's looking at the exact details, but that probably can be generated just a start time.
21:11 GMan joined #salt
21:11 jacksontj problem will be we'll need to update the modules at some interval in case they change on disk
21:11 jacksontj we'll want to cache- not just load :/
21:11 jacksontj so i can do a class global (per worker process) cache easy
21:12 jacksontj i was hoping to have some event system or something so i could add it to the master scheduler and just push it
21:12 jacksontj i might just add one though, we have zmq all over, whats one more ;)
21:12 jacksontj lol
21:13 Gifflen joined #salt
21:13 oz_akan_ joined #salt
21:13 vejdmn joined #salt
21:14 rojem joined #salt
21:14 UtahDave jacksontj: So those aren't actually the minion's modules, in pillar the modules refer to the master's modules
21:14 geraintjones joined #salt
21:14 jacksontj yea, modules just loaded on the master
21:14 jacksontj sorry-- confusing the terminology ;)
21:15 mgw joined #salt
21:15 geraintjones Hi, would anyone be willing to try and help me figure out why this is so damn slow : https://gist.github.com/kiwigeraint/8289872/raw/1057be32890264f3bbde1a28f232da6ef782a214/gistfile1.txt
21:15 jacksontj the master's pillar loader loads up the modules on pillar load
21:15 UtahDave So this is actually very easy.  Tom's fixing it right now
21:15 jacksontj line 79 of the __init.py pillar file
21:15 jacksontj ok, i'm curious to see what he's got then ;)
21:16 jacksontj to me it looks like  you are doing the whole loader thing for all modules on every request-- which we "want" to do so we can have the most up to date modules
21:16 jacksontj but we dont want becaue its a lot of work ;)
21:17 UtahDave jacksontj: ok, Tom just pushed a commit that should fix this.  Can you test it?
21:17 jacksontj sure, let me find the commit
21:18 jacksontj how does that update the modules loaded?
21:18 KoFFiE joined #salt
21:19 jacksontj if i understand right someone has to call gen_modules every interval so that it says up to date
21:19 jacksontj otherwise if you change a module on the master the pillar stuff won't see the change until a bounce
21:19 linx_0x how can I declare more than 1 function per state: ie. /etc/hosts: file.managed ... file.append ... ?
21:22 jdenning joined #salt
21:25 forresta linx_0x, I'm not sure if you can combine file.managed and file.append, but something like this might be worth a shot: https://gist.github.com/gravyboat/8290064
21:25 forresta graned the 'name' value might conflict, I honestly don't know.
21:25 forresta and I don't have anything to test it on at work
21:25 forresta so you'll have to play with it
21:25 Thiggy joined #salt
21:25 dave_den you can't
21:25 forresta dave_den, awesome, good to know
21:26 forresta I will delete that gist then.
21:26 dave_den you need a unique ID declaration and module combination as of right now
21:26 forresta so there you go linx_0x
21:26 linx_0x ok
21:26 linx_0x thanks
21:27 gadams999 left #salt
21:27 dumol joined #salt
21:28 dave_den i don't think there's anything that would have prevented multiple functions of the same module under one ID declaration, but changing the behavior now would break backward compatibility
21:28 vejdmn joined #salt
21:29 ConceitedCode joined #salt
21:29 dumol hi all. how do I use pkg.set_selections in a state file? salt 0.17.4 here. thanks!
21:30 jacksontj UtahDave: ah i see, that module access doesn't include user space modules (ones in _modules)
21:30 gadams9991 joined #salt
21:31 sschwartz joined #salt
21:31 oz_akan_ joined #salt
21:32 jacksontj UtahDave: https://github.com/saltstack/salt/pull/9595
21:32 jacksontj missed one thing though ;)
21:32 MZAWeb joined #salt
21:32 sschwartz So, can anyone tell me how to fix the: DeprecationWarning: 'OutputOptionsWithTextMixIn' has been deprecated. Please start using 'OutputOptionsMixIn'; your code should not need any further changes." error? I tried installing salt-cloud from the development pull, and got a "setup.py" missing, so I went with "yum install salt-cloud."
21:33 forresta sschwartz, what version of salt are you on, versus the version of salt-cloud?
21:34 sschwartz 0.17.2 for salt, runtime error for salt-cloud, but I was grabbing 0.8.11.
21:34 UtahDave jacksontj: I'll make sure Tom sees that.
21:34 jacksontj actually, i'm going to update slightly ;)
21:34 jacksontj but yea
21:34 forresta hmm, 0.17.2 should work with 0.7.11
21:34 forresta err 0.8.11
21:34 UtahDave ok, let me know when you're ready, jacksontj
21:34 druonysuse joined #salt
21:34 joehillen joined #salt
21:34 dave_den dumol: you can use the "module" state module to run any execution module from a state: http://docs.saltstack.com/ref/states/all/salt.states.module.html
21:34 druonysuse joined #salt
21:35 jacksontj UtahDave: https://github.com/saltstack/salt/pull/9596
21:35 sschwartz Indeed; that was what I thought.  Is there a patch for the "OutputOptions"? I'm a bit loath to go digging in the code and doing a search-and-replace, since I am not a Python hacker.
21:35 jacksontj tom's too fast
21:35 jacksontj already merged my other one ;)
21:35 UtahDave jacksontj: ah, he merged that pull req
21:35 UtahDave yeah
21:35 jacksontj lol
21:35 jacksontj this other one is just a style thing ;)
21:36 dumol dave_den: interesting, i'll read up on that. 10x
21:36 joehillen [QUESTION] I'm  trying to use a cloned (soon to be a submodule) formula, but I don't know what to add to file_roots.
21:37 UtahDave jacksontj: there you go.  So in your initial testing, does this help?
21:37 sschwartz The error came up when I tried to do a command-line salt-cloud invocation, if that helps make things clearer.
21:37 jacksontj UtahDave: i'll load it up :)
21:37 jacksontj lets see :D
21:38 sschwartz Ah-*hah*! reinstalling from the git repo managed to get me a new, shiny error message that looks like it's my fault. Thank you for letting me make a fool of myself in public. ;)
21:38 tollmanz joined #salt
21:40 vejdmn joined #salt
21:41 tollmanz joined #salt
21:42 oz_akan_ joined #salt
21:43 UtahDave Ah, glad you got it, sschwartz
21:43 SaltedMan joined #salt
21:45 troyready joined #salt
21:45 quickdry21_ joined #salt
21:46 linx_0x how can I have a file require itself before appending to it?
21:48 Gifflen joined #salt
21:48 godber joined #salt
21:48 dave_den you mean having your file.append require file.managed?
21:49 SaltedMan good morning/afternoon/evening/night to everyone
21:50 forresta sschwartz, yea np, I do it all the time :P
21:50 SaltedMan I'm in front of a bottle half full of Laphroaig and I want to understand better pillars, or get drunk, which ever comes first
21:50 linx_0x yes
21:50 linx_0x exactly
21:50 dave_den linx_0x: just use a new ID declaration.
21:51 dave_den . /etc/hosts/: file.managed...
21:51 dave_den append_/etc/hosts:  file.append:  - name: /etc/hosts/
21:52 mgw joined #salt
21:52 SaltedMan I have a question about pillars. I have created some salt tree state in a git backend and have a git pillar configured. The minions set them selves to highstate as commanded and read data off the pillar as needed via the master
21:52 jacksontj UtahDave: looks like a ~35-40% cpu reduction in our staging environment
21:52 dave_den and require file: /etc/hosts in the append_/etc/hosts state
21:52 SaltedMan now I would like to have a production pillar, what i don't get is how do I associate it with minions
21:53 SaltedMan I created a prod branch on the pillar
21:53 UtahDave jacksontj: nice!!
21:53 tempspace Is there a bug in 0.17.4 that prevents this from working correctly any longer? {{ salt['network.ip_addrs'](salt['pillar.get']('INTERFACE', 'eth0')) | first }}
21:54 tempspace gives me "Jinja variable No first item, sequence was empty."
21:54 jacksontj i'm very excited about getting this next build out, 3 major performance fixes for us big installs :)
21:54 UtahDave jacksontj: Tom apologizes for not catching that before!
21:54 druonysuse joined #salt
21:54 druonysuse joined #salt
21:54 jacksontj jacksontj: no problem ;) we just noticed it ourselves
21:54 jacksontj lol
21:54 jacksontj so we have the inifinite recursion fix, the listdir/stat fix, and this one
21:55 jacksontj guess my master is going to be really bored after this release
21:55 UtahDave Yeah, some of those problems don't appear until you really stress the system
21:56 jacksontj we think there might be another one (just by loooking at the CPU usage graphs)
21:56 psyl0n joined #salt
21:56 jacksontj but its hard to tell since we now have 3 fixes we need to push
21:56 UtahDave right
21:56 jacksontj so we're going to make this build, push it out then re-look
21:56 UtahDave cool!
21:56 jacksontj i also need to get some time to finish up the event-driven module sync stuff
21:56 jacksontj that will save our install a lot of work
21:57 forresta basepi, just don't ask JasonSwindle to test on the develop branch for the pip/virtualenv stuff, his code is broken everywhere :P
21:57 JasonSwindle forresta:  :P
21:57 JasonSwindle I am good at being the end user....
21:57 basepi ha!
21:57 jacksontj oh, and i forgot-- we are all on the new versioning scheme now too
21:59 gadams9991 left #salt
21:59 JasonSwindle basepi:  I will know if #4648 is fixed once I can move past https://github.com/saltstack/salt/issues/9587
21:59 rofl____ joined #salt
22:00 basepi so much copy pasta!  ;)  gist.github.com anyone?
22:00 basepi it's fine, just had to rib you a little.
22:00 JasonSwindle :p
22:01 JasonSwindle I am trying to open source my SaltStack deployment for SaltConf and my talk....
22:01 JasonSwindle and trying to use 2014.1.0 :)
22:01 dumol in regards to running pkg.set_selection with module.run, it seems i can't get it right and debug output doesn't help much. can someone take a look at https://gist.github.com/dumol/327e001cd54e53452755 . thanks!
22:02 basepi JasonSwindle: s0undt3ch probably went to bed or something.  How dare he sleep!  o.O
22:02 forresta basepi, he's not in here any longer
22:02 basepi forresta: i know.  =)
22:02 forresta oh ok
22:02 basepi my tab-completion didn't work, had to actually type it out!
22:02 forresta lol
22:03 timoguin speaking of tab-completion
22:03 basepi JasonSwindle: but ya, we're definitely working on figuring out what's going on with that one.
22:03 oz_akan_ joined #salt
22:03 timoguin the bash completion for salt is so beautiful
22:03 timoguin and makes me feel warm and happy
22:03 basepi timoguin: hahaha, woot!  i had a sinking feeling, thought you were about to tell me it was broken and horrible
22:03 JasonSwindle basepi: I figure as much.  Let me know what you need me to do.  The master and minions can be rebuilt at any time
22:03 EWDurbin joined #salt
22:03 timoguin haha, nope!
22:03 basepi JasonSwindle: for sure, thanks.
22:04 eclectic joined #salt
22:04 forresta on the note of bash completion
22:04 forresta has anyone talked to our friends at redhat about selinux policy
22:05 UtahDave jacksontj: looks like our integration tests are indicating a need for a couple small fixes.
22:09 blarghmatey joined #salt
22:12 forresta dumol, I'm pretty sure the issue lies with your selection
22:12 kossy joined #salt
22:12 dumol I've also included the command that works
22:12 Gifflen joined #salt
22:12 forresta sigh
22:12 dumol in a comment below the paste
22:12 forresta I commented
22:12 forresta if you refresh
22:13 forresta but the damn formatting is all screwed up
22:13 forresta basically I'm saying do - selection: \n  - hold: firefox
22:13 forresta try it like that
22:13 forresta where hold is a level below selection perhaps
22:13 timoguin joined #salt
22:14 forresta github should really save indenting in comments
22:14 forresta I don't get why they don't
22:14 dumol i think i've tried that already and got: Module function pkg.set_selections threw an exception
22:14 elithrar joined #salt
22:14 forresta I'm not sure then
22:15 dumol forresta: for indeting use ``` in the lines before and after the code
22:16 dleedyna joined #salt
22:16 forresta yea I don't want to have to do that though...
22:16 forresta it's a pain in the ass.
22:17 UtahDave hey, JasonSwindle did ya'll remove Ubuntu 13.04 images?
22:17 Vye joined #salt
22:17 forresta so trying it like that didn't work dumol?
22:17 dumol forresta: sorry, your code threw an exception too. thanks anyway!
22:17 forresta :\
22:18 forresta basepi, are you familiar with the module.run items when you have something that requires multiple arguments?
22:18 forresta or multiple levels of arguments
22:18 forresta such as trying to use: http://docs.saltstack.com/ref/modules/all/salt.modules.apt.html#salt.modules.apt.set_selections the selection portion
22:19 JasonSwindle UtahDave:  From public cloud?
22:19 JasonSwindle I don't think so....
22:19 JasonSwindle Let me look and find you the ID of it
22:19 basepi forresta: if by "familiar" you mean "done something similar before", the answer would still be no.  ;)  i haven't messed with it that much.  that said, i would think you could do JSON syntax (supported by YAML) to make sure the data structures ended up correct
22:19 UtahDave it's not appearing in the web gui nor in the api.
22:20 UtahDave I'm chatting with a support rep right now, too
22:20 forresta basepi, oh that's a good idea..
22:20 JasonSwindle Odd, I see it in my….. let me message you.
22:20 dave_den dumol: selection is a hash, not a list
22:20 rojem joined #salt
22:20 dave_den - selection:\n    hold:\n      - firefix
22:21 UtahDave JasonSwindle: Ah, 13.04 is EOL
22:21 dumol dave_den: let's see... :)
22:23 dumol dave_den: i'm not sure i'm writing this correctly, can you please add a comment at https://gist.github.com/dumol/327e001cd54e53452755 ? thanks
22:24 dave_den dumol: done.
22:25 cdcalef joined #salt
22:25 dumol dave_den: thanks, that's what i thought. it executes successfully, but doesn't really hold the package as it should
22:26 dave_den have you confirmed that your hold is set in dpkg?
22:26 ccase joined #salt
22:27 dumol i can set the hold using command line: salt '*' pkg.set_selections selection='{"hold": ["firefox"]}'
22:27 dumol but not using the sls
22:27 forresta dave_den, ahh thanks for adding that, was off
22:27 forresta *I was off
22:27 vejdmn joined #salt
22:28 dumol i hope we are  getting close :)
22:28 JasonSwindle basepi:  Updated an issue as Mr. Sound wanted.
22:29 basepi JasonSwindle: coolio
22:29 dave_den dumol: you have the set-selection depending on the package install of firefox. are you sure the package is being installed on your state run?
22:29 linx_0x joined #salt
22:29 dumol dave_den: ah, i see what you mean. let's test
22:31 dumol i've added the results in the same gist: https://gist.github.com/dumol/327e001cd54e53452755
22:32 dumol dave_den: sorry, it doesn't work even after installing the package it requiares
22:33 dave_den it's being executed, so if it's not behaving as intended you will need to set your debug (or trace, even) output on the minion and see what the logs say
22:35 dumol nothing relevant, sorry, i've added the debug output too
22:36 dave_den set it to trace
22:36 dave_den make sure you restart the minion
22:36 dumol sure, just a sec
22:37 dumol it looks the same
22:37 Vye Has anyone had any issues with salt 17.2 or 17.4 where gitfs cache doesn't contain all of the folders that exist in the same branch of the remote?
22:37 dave_den there should be a lot more output
22:38 dumol hhm, you're right
22:39 jfzhu_us joined #salt
22:39 blee_ joined #salt
22:39 vejdmn joined #salt
22:39 Vye What got my attention is that not all my formulas were being applied so after restarting the master I started checking the cache dir and found missing folders. I cleared the cache but the next time it was populated some folders were missing.
22:39 dave_den dumol: FYI, there are two log settings, log_level and log_level_logfile
22:40 dumol but i'm trying to find something relevant in the TRACE output and there doesn't seem to be anything relevant
22:40 dumol except [TRACE   ] Added pkg.set_selections to module
22:40 dumol what should i look for?
22:41 tyler-baker joined #salt
22:41 dave_den it's possible there's no debug code in that function.
22:41 dave_den i haven't looked
22:42 dave_den but running salt-call pkg.set_selections... from the command line works?
22:42 dumol it seems so, nothing more gets output'ed besides those INFO lines
22:43 dumol yep, that works
22:44 dave_den hrm, try indenting "hold" another two spaces so that the 'h' is under the 'l' in selection
22:44 dumol and there is INFO output from salt.loaded.int.module.cmdmod
22:45 dumol which shows everything working in proper order
22:45 dumol let's try that...
22:45 dumol that means indenting the following line two, i suppose
22:46 dumol too*
22:46 dave_den you can if you want
22:47 dumol hoo-ray!
22:47 dumol that worked :)
22:47 dave_den bleh. silly yaml
22:47 dumol let me double-check
22:47 dave_den ;)
22:47 forresta heh
22:47 forresta dumol can you leave this gist?
22:47 forresta I've starred it just in case I need an example again
22:47 dumol sure
22:48 dumol i'll paste the correct code
22:48 forresta great thanks
22:50 william_20111 joined #salt
22:50 dumol added the working code too
22:51 dumol dave_den: forrestaL thanks a bunch!
22:51 dave_den no prob. that module probably needs a bit of error checking at some point
22:51 dave_den i'm out for the day. see you guys tomorrow
22:52 forresta later
22:52 forresta dumol, always happy to lead you down the wrong path :P
22:52 dumol :)
22:56 bhosmer joined #salt
22:58 redmin joined #salt
23:01 dumol left #salt
23:01 mjulian joined #salt
23:01 mjulian joined #salt
23:01 linx_0x joined #salt
23:02 d10n joined #salt
23:02 linx_0x what permissions recommended on /srv/salt directory?
23:02 linx_0x ownership
23:02 UtahDave linx_0x: whatever the salt master is running as.
23:04 linx_0x it requires me to sudo when running the salt command: does this mean that it's root then?
23:04 jacksontj UtahDave: tom was saying that the tests were failing, that all settled out yet?
23:04 jacksontj wanted to wait on that getting settled before cherry picking
23:05 Gareth linx_0x: you're master and minion are likely running as root.
23:08 kedo39 is using block literals in the pillar yaml files still the best way to distribute stuff like keys and certs onto minions ?
23:09 UtahDave jacksontj: a remote dev restarted our jenkins server. I just kicked off a new round of tests.  We should know in 20 minutes or so
23:11 UtahDave back in a moment
23:11 alunduil joined #salt
23:13 UtahDave joined #salt
23:18 MindDrive joined #salt
23:19 jacksontj UtahDave: looks like the tag for the 2014.1.0 release in git is wrong-- so salt version isn't happy with the version number :/
23:20 jacksontj or something :/
23:20 jacksontj messing with that now
23:22 elithrar joined #salt
23:24 jacksontj is the version supposed to be 2014.1 or 2014.1.0
23:24 UtahDave not sure, lemme ask
23:25 william_20111 joined #salt
23:27 davet joined #salt
23:28 jacksontj the comments in this version file conflict with the actual tag names too
23:28 jacksontj says "<2-digit-year>.<month>.<bugfix>"
23:29 jacksontj and the version in the tag is currently 2014.1
23:29 Corey jacksontj: You are correct.
23:29 jacksontj sorry, which one is correct?
23:29 Corey jacksontj: That it's a mess.
23:29 jacksontj all i know is that they are inconsistent ;)
23:29 jacksontj oh, lol
23:29 jacksontj :)
23:29 Corey jacksontj: We've successfully created a versioning system so ridiculous even the engineers behind it can't figure it out.
23:30 jacksontj i can fix it, just need to know which it should be
23:30 jacksontj personally i preferred the 0.17.1 naming
23:30 jacksontj makes the compatibility thing a bit easier to remember
23:30 Corey jacksontj: Personally I prefer alcohol poisoning. This is *terrible*
23:30 jacksontj the year thing doesn't
23:30 Corey But nobody asked me.
23:30 forresta clearly, we should go 0.17.4 -> 17.5.0 right?
23:31 Corey forresta: No, that's Puppet.
23:31 forresta WHAT, ARE YOU SAYING I WOULD MAKE FUN OF THE PUPPET VERSIONING SCHEME??
23:31 forresta How dare you!
23:31 darless haha
23:31 Corey forresta: Salt making fun of the Puppet versioning scheme is throwing rocks whilst we live in a glass house.
23:31 brucewang lol
23:31 Corey On a major fault line.
23:31 forresta Corey I know, I'm just playing :D
23:31 Corey Whilst cavorting naked atop the piano.
23:32 jms2 joined #salt
23:32 brucewang Is 0.17.5 Soon(tm)?
23:32 redmin joined #salt
23:33 forresta brucewang, there is no 0.17.5, the versioning scheme has changed.
23:35 UtahDave forresta: actually, there might be a 0.17.5.  That's just another bugfix release on the 0.17 line
23:35 brucewang oh, I forgot we are already 2014 XD
23:35 forresta UtahDave, oh nice that would be very cool to see.
23:36 Corey forresta: Followed by a very quick 0.17.6.oops.mymistake
23:36 forresta UtahDave, you guys better be careful, or you're gonna get requests for 0.16 bug fixes :P
23:36 forresta Corey, lol, so bitter.
23:36 Corey forresta: Quite.
23:36 UtahDave jacksontj: there are a few tests still failing. investigating
23:38 jacksontj so for this new versioning scheme we have at least 3 variants in code/commits
23:38 jacksontj (1) 2014.1 (2) 2014.1.0 (3) 14.1.0 (4) 14.1
23:38 jacksontj don't know which one is correct
23:38 forresta one more jacksontj
23:38 forresta http://docs.saltstack.com/topics/releases/2014.01.0.html
23:38 forresta 2014.01.0
23:38 jacksontj :(
23:39 forresta it's ok
23:39 forresta it's nearly 6
23:39 forresta *5
23:39 forresta well, 4 for you
23:39 chitown like the codename :)
23:39 th3reverend joined #salt
23:39 th3reverend left #salt
23:40 vejdmn joined #salt
23:40 chitown has salt-api been merged into the the salt "core"?
23:43 jacksontj so, who is "the guy" who makes the call on the version number?
23:44 jacksontj i want to make a release but i cant since the versions are all screwey
23:44 rojem joined #salt
23:45 Gareth forresta: Corey...resident curmudgeon :)
23:45 xmltok joined #salt
23:45 forresta lol
23:45 __number5__ "the guy"
23:45 Corey Gareth: Hey, look! UtahDave is about to release a new versioning scheme! http://static.fjcdn.com/pictures/Drunk_c1435c_762659.jpg
23:46 Gareth Corey: so UtahDave is the drunk panda in this case?
23:46 Corey Gareth: Yes. I'm assuming adorable inebriation rather than malice at this point. :-)
23:46 Corey Gareth: We'll have to see if this is borne out by Salt 0.19.x vs Salt 2014.2.12.7.waitwhat.igotconfused.rc7-b2
23:47 Corey I should be curmudgeonly more often. IT's a great respite from working!
23:47 UtahDave jacksontj: are you getting a traceback when you change the version in version.py?
23:48 jacksontj not changing it, i have cherrypicked some commits so i get the git version string
23:48 * UtahDave hugs Corey
23:48 jacksontj v2014.1-1511-g381c0c8
23:48 jacksontj which it can't parse
23:49 jacksontj which lead me to look at the rest of this versioning stuff which is… wrong ;)
23:49 jacksontj or at least inconsistent
23:49 UtahDave jacksontj: what do you mean by it can't parse that?
23:50 jacksontj i get a valueerror unablet to parse version string
23:50 jacksontj regexes at the top can't cope with that version string
23:50 jacksontj i was going to clean up this versioning stuff, just need to know what the correct versioning  scheme is
23:50 UtahDave Oh, there's another commit you need to cherry pick
23:50 jacksontj the tag, commit message, docs, release notes, and comments all disagree
23:51 UtahDave let me find it.  s0undt3ch fixed that regex
23:52 jacksontj 77e74caf3a529cdd9ca28377d21f8f40b525ff38
23:52 jacksontj looks like
23:52 UtahDave jacksontj: https://github.com/saltstack/salt/pull/9537    is that commit found in your branch?
23:52 UtahDave yeah, that's the one
23:53 bhosmer joined #salt
23:54 honestly appropriate: http://i.imgur.com/cm9ocSS.jpg
23:54 jacksontj cool, looks like that got it
23:54 UtahDave cool
23:55 jacksontj the tag name should probably match the version names
23:55 jacksontj IMO the tag should be 2014.1.0 instead of v2014.1
23:55 Corey jacksontj: The v is long established, I'm not quite sure *why* though.
23:56 __number5__ the tag should always start with a letter
23:56 jacksontj err, sorry didn't mean to remove the v
23:56 __number5__ or else it will confuse lots of thing
23:56 jacksontj just add the .0 ;)
23:56 jacksontj and now im just adding to the confusion :(
23:56 Corey Time for a versionbot to lurk in here.
23:57 __number5__ !version
23:57 Corey Error: I am only a bot, please don't think I'm intelligent :)
23:57 __number5__ no repsonse
23:58 ccase_ joined #salt
23:59 jacksontj so this means the "correct" version is 2014.1.0

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