Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-03-04

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

All times shown according to UTC.

Time Nick Message
00:05 snave joined #salt
00:11 rgarcia_ joined #salt
00:12 Gareth w 4
00:12 Gareth erm
00:12 themadcanudist dave_den, so i used the pyrenderer… it's the first .sls to be sucked in, but it appears to still fail on highstate
00:12 themadcanudist i can execute the bootstrap state independently and it works as expected
00:14 yomilk joined #salt
00:17 tr_h joined #salt
00:18 dyim joined #salt
00:20 vxitch joined #salt
00:21 vxitch hello! i'm getting a conflicting ID error for iptables.set_policy . here is my sls, can someone please tell me how to achieve what i'm trying to do but with proper YAML? http://hastebin.com/qaqilotute.sm
00:22 faldridge joined #salt
00:22 bhosmer joined #salt
00:24 tr_h joined #salt
00:28 Gareth vxitch: probably need to have a default_drop_input and a default_drop_forward.
00:29 vxitch so the best and cleanest way to go about this is different IDs for every iptables rule? ay...
00:30 nextdoorwarren joined #salt
00:30 Gareth vxitch: believe so.
00:31 themadcanudist anyone have any insight in to using a pydsl renderer to bootstrap mako installation before the remaining states are rendered (via highstate)? I want to bootstrap the minion with mako… Currently it throws an exception that it's missing mako.
00:31 vxitch thanks Gareth
00:31 hunter joined #salt
00:31 Gareth vxitch: np
00:32 hunter joined #salt
00:35 elfixit joined #salt
00:36 rojem joined #salt
00:39 bemehow joined #salt
00:40 vxitch okay, well, now i'm getting No Top file or external nodes data matches found
00:40 vxitch but only when executing state.highstate on my minion that ISN'T also the master
00:40 vxitch if i execute highstate on the hostname of the master, it works
00:40 fauxtoniche joined #salt
00:41 vxitch my top file has a '*' entry and a few grains entries, but it doesn't specify this hostname directly
00:41 vxitch do i have to have the hostname in the top file in order to be able to target it? am i limited to targetting only what's in my top file?
00:42 vejdmn joined #salt
00:47 taion809 joined #salt
00:52 DaveQB joined #salt
00:55 jeremyBass1 with the change of env to saltenv is the sls   'env:production': also changed so it'd be   'saltenv:production':?
00:56 Ryan_Lane joined #salt
00:59 dyim joined #salt
01:00 joshk left #salt
01:01 steveoliver joined #salt
01:06 tr_h joined #salt
01:09 rgarcia_ joined #salt
01:10 quickdry21 joined #salt
01:11 rgbkrk joined #salt
01:11 lzhang joined #salt
01:14 baniir joined #salt
01:14 baniir is it possible to pass pillar data to pkg state's debconf option
01:15 pydanny joined #salt
01:19 tr_h joined #salt
01:22 pdayton joined #salt
01:22 helderco joined #salt
01:22 xzarth joined #salt
01:23 yomilk joined #salt
01:27 ipalreadytaken joined #salt
01:27 yomilk joined #salt
01:31 rainmanj_ joined #salt
01:31 ajw0100 joined #salt
01:35 sgviking joined #salt
01:47 NotreDev joined #salt
01:48 TheWinnerFjk joined #salt
01:48 TheWinnerFjk left #salt
01:51 lzhang Perhaps when using pkg.latest errors shouldn't show up in the minion log when the versions are up to date.
01:52 raizyr joined #salt
01:53 pydanny joined #salt
01:53 kintel_ joined #salt
01:54 manfred baniir: unfortunately no, but what I have been doing is using a file.managed directive to create one a file on the system, then use that file to pass to the debconf option
01:57 joehh 2014.1 packages now up for wheezy on debian.saltstack.com rest coming soon
01:59 Gareth joehh: hooray!
01:59 fxhp joehh - woot
01:59 * Gareth gives joehh a cookie
02:00 dave_den baniir: use the debconfmod state module:  http://docs.saltstack.com/ref/states/all/salt.states.debconfmod.html#module-salt.states.debconfmod
02:05 jdenning joined #salt
02:06 baniir manfred dave_den: thanks
02:07 ipalreadytaken joined #salt
02:08 fxhp 3~dave_den - how does that work?  does it generate a preseed?
02:08 fxhp or does it build a state from an existing preseed?
02:10 fxhp dave_den *
02:10 ipalread_ joined #salt
02:10 blee joined #salt
02:11 zaz\ joined #salt
02:12 vejdmn joined #salt
02:17 danielbachhuber joined #salt
02:21 fllr joined #salt
02:29 errr joined #salt
02:30 rgbkrk joined #salt
02:36 dave_den fxhp: i'm not sure what you mean
02:36 juicer2 joined #salt
02:37 justlooks joined #salt
02:37 timoguin joined #salt
02:39 dave_den fxhp: if you look at the docs for the debconfmod, you can either set the debconf values from a dic, or you can use a file
02:41 rgarcia_ joined #salt
02:41 justlooks hi, i want to build web ui to control salt master action, how can i do it?
02:42 dave_den justlooks: https://github.com/saltstack/salt-api
02:44 justlooks dave_den: salt api must installed with salt master?
02:44 dave_den justlooks: read the documentation and it will tell you
02:44 Networkn3rd joined #salt
02:46 sgviking joined #salt
02:48 fxhp dave_den - I've only used debconf during provisioning (via preseed) so I guess I don't understand how/what the debconfmod state is used for
02:49 fxhp justlooks - https://github.com/saltstack/halite | might be worth looking at too
02:50 dave_den fxhp: debconf is debian (and derivative) method for setting configuration options for a package being installed - not just at os install time. this mean you can set your preferences for the application being installed before you run apt-get install package
02:51 dave_den usually when you are ssh'ed into a server and you install a package, if it needs configuration options set it would prompt you
02:53 fxhp dave_den - ahh yes, like postfix prompts
02:53 dave_den yep, exactly
02:53 fxhp Ok, so debconf could be used to give answers to the installer?
02:54 dave_den that's precisely what it's for
02:54 dave_den https://wiki.debian.org/debconf
02:54 fxhp I would just assume to accept the defaults of a package and "clobber" it to the correct config, with config management.  I assume in some ways that will not always work?
02:55 dave_den most of the time that will work just fine, sometimes it may not
02:56 dave_den so in the case where it does not, or if you want to make sure the program is installed initially as you want, even if you set the config files under salt management, you may need to use debconf
02:57 fxhp Interesting
02:57 fxhp going to try this on a test box
02:57 fxhp dpkg-reconfigure debconf
02:58 malinoff joined #salt
02:59 dave_den fxhp: one particular use for debconf for package installs is java, which forces you to accept a license before it will be installed.
02:59 quickdry21 joined #salt
03:00 Furao joined #salt
03:00 fxhp wow that is neat, dpkg-reconfigure debconf allows me to reconfigure debconf
03:01 fxhp I was able to switch to text-only
03:01 fxhp then to editor
03:01 fxhp then back to dialog
03:01 fxhp dave_den - yeah! the oracle version of java does that
03:04 xl1 joined #salt
03:06 joehh Gareth, fxhp: thanks - now onto squeeze
03:09 joehoyle joined #salt
03:12 ravibhure joined #salt
03:12 justlooks fxhp:  halite can not clone use git why? https://gist.github.com/justlooks/9339635
03:14 justlooks fxhp: ok ,i find why failed
03:14 joehoyle hey, about 20% of the time when a minion does a highstate, it doesn't return anything to the master - any idea why that would be?
03:17 jalbretsen joined #salt
03:19 sroegner_ joined #salt
03:23 rostam joined #salt
03:24 dstanek_afk joined #salt
03:25 blast_hardcheese joined #salt
03:25 justlooks fxhp: can halite run under python2.6?
03:26 fxhp joehh - precise will get lots of testers
03:26 fxhp : )
03:26 fxhp justlooks - I've not installed or tested halite
03:37 Mr_Pepper joined #salt
03:37 vbabiy joined #salt
03:38 ajw0100 joined #salt
03:38 raizyr joined #salt
03:39 NotreDev joined #salt
03:41 joehh fxhp: true
03:47 fxhp joehh - where do the PPAs show up ? I don't see them in the launchpad area
03:48 joehh Udner https://launchpad.net/~saltstack
03:48 dyim joined #salt
03:49 joehh The salt stable PPA, though you might get a preview at the salt-depends
03:52 forrest joined #salt
03:56 zain_ joined #salt
03:57 mgw joined #salt
04:08 middleman_ joined #salt
04:10 ipalreadytaken joined #salt
04:13 aaroneous joined #salt
04:15 APLU joined #salt
04:19 HeadAIX joined #salt
04:20 jeremyfelt joined #salt
04:28 APLU joined #salt
04:28 aaroneous joined #salt
04:40 Linz joined #salt
04:41 echos joined #salt
04:41 smcquay joined #salt
04:44 Linz joined #salt
04:46 middleman_ joined #salt
04:52 n8n joined #salt
04:55 [diecast] joined #salt
04:59 dyim joined #salt
05:02 liwen joined #salt
05:07 liwen joined #salt
05:13 Ryan_Lane joined #salt
05:19 sroegner_ joined #salt
05:20 kintel joined #salt
05:27 fllr joined #salt
05:33 mgw1 joined #salt
05:38 kermit joined #salt
05:42 pydanny joined #salt
05:51 zain_ joined #salt
05:52 karlgrz joined #salt
06:01 ChaosPsyke joined #salt
06:10 lzhang joined #salt
06:10 ipalreadytaken joined #salt
06:15 viod joined #salt
06:17 schimmy joined #salt
06:21 boxwh joined #salt
06:31 justlooks how to know my salt version?
06:31 blast_hardcheese joined #salt
06:32 malinoff justlooks, salt --version
06:32 malinoff Or, specific salt master: salt-master --version
06:32 malinoff salt minion: salt-minion --version
06:32 malinoff Btw, in linux, <command> --help is always useful
06:33 justlooks malinoff: it's say salt-minion 2014.1.0 ,i have no idea about 2014.1.0
06:33 malinoff justlooks, it's new salt versioning
06:33 malinoff Forget about 0.17 stuff
06:33 malinoff 0.18 is now 2014.1.0
06:33 malinoff http://docs.saltstack.com/topics/releases/2014.1.0.html
06:34 justlooks it is GA? or just a test version?
06:34 joehh packages for debian squeeze now making there way to debian.saltstack.com
06:35 malinoff justlooks, it's stable release
06:35 malinoff Well, stable is salt terms :)
06:35 malinoff in*
06:39 justlooks malinoff:  it seems i can not get  " salt 'saltslave' grains.ls " output use salt-master 0.17 and salt-minion 2014.1.0 ,there is any protocol incompatible like  between salt 0.16 and 0.17?
06:39 malinoff justlooks, I do not know exactly, but it is strongly recommended to have a master and a minion with the same versions
06:41 justlooks malinoff: let me try it
06:46 kintel joined #salt
06:47 tr_h joined #salt
06:58 zemm joined #salt
06:58 boxwh My linux is ubuntu 12.04,then i have installed salt with version 0.17.5.Is it ok?Or should download the newest version to use?
06:58 justlooks if issue "test.ping" command on master,but one of my minion not return ,when i ctrl+c on master ,the job still will run in minion ?
07:00 Furao_ joined #salt
07:04 malinoff justlooks, also, about the new versioning: https://groups.google.com/forum/#!topic/salt-announce/s1oYmaeN00I
07:06 SkyRocknRoll joined #salt
07:06 SkyRocknRoll joined #salt
07:07 ndrei joined #salt
07:10 viod joined #salt
07:12 schimmy joined #salt
07:12 Shenril joined #salt
07:13 viod joined #salt
07:13 favadi joined #salt
07:13 borgstrom joined #salt
07:15 schimmy1 joined #salt
07:18 vxitch joined #salt
07:19 Furao joined #salt
07:20 sroegner_ joined #salt
07:20 fllr joined #salt
07:33 Furao_ joined #salt
07:37 johtso joined #salt
07:38 jtang1 joined #salt
07:40 ml_1 joined #salt
07:50 fllr joined #salt
07:51 n8n joined #salt
07:54 fllr joined #salt
07:55 Kenzor joined #salt
08:06 martoss joined #salt
08:06 balboah joined #salt
08:09 ajw0100 joined #salt
08:13 harobed_ joined #salt
08:14 harobed_ joined #salt
08:15 harobed_ joined #salt
08:16 harobed_ joined #salt
08:16 harobed_ joined #salt
08:16 slav0nic joined #salt
08:20 kadel joined #salt
08:23 lzhang joined #salt
08:26 zooz joined #salt
08:28 zooz_ joined #salt
08:29 Ztyx joined #salt
08:37 topochan joined #salt
08:45 ndrei joined #salt
08:51 yomilk joined #salt
08:52 giantlock_ joined #salt
08:52 tr_h joined #salt
08:52 fllr joined #salt
08:55 yomilk_ joined #salt
09:02 liwen joined #salt
09:07 babilen Hello, I am using the users formular (i.e. https://github.com/saltstack-formulas/users-formula ) and have to install a specific package to enable a user to use it. What would be the best way to achieve that? It seems to be the wrong approach to try to extend the generated state for that user (as it doesn't work)
09:11 Ztyx left #salt
09:12 babilen http://paste.debian.net/85197/ is what I am playing with right now, but that results in "Extension name foo_user in sls user.foo is not a dictionary" - I am sure I simply overlooked something obvious/simple, but another pair of eyes would be appreciated.
09:15 lessismore joined #salt
09:16 ninkotech joined #salt
09:20 carlos joined #salt
09:20 sroegner_ joined #salt
09:22 tmwsiy__ joined #salt
09:23 flebel joined #salt
09:25 seksifemale joined #salt
09:25 Ryan_Lane joined #salt
09:27 Ryan_Lane1 joined #salt
09:32 generj joined #salt
09:32 Ryan_Lane joined #salt
09:35 ndrei joined #salt
09:37 zooz joined #salt
09:39 CeBe joined #salt
09:43 millz0r joined #salt
09:50 helderco joined #salt
09:51 ndrei joined #salt
09:52 fllr joined #salt
09:53 generj So I'm on  Salt: 2014.1.0-2476-gb522333
09:53 generj Following this doc: http://docs.saltstack.com/topics/cloud/azure.html
09:54 generj sudo salt-cloud -p azure-ubuntu newinstance
09:54 generj WindowsAzureConflictError: Conflict (Conflict)
09:54 generj Error: There was a profile error: Failed to deploy VM
09:55 generj [ERROR   ] Error creating newinstance on Azure
09:55 generj [WARNING ] /usr/lib/python2.7/dist-packages/salt/cloud/clouds/msazure.py:491: DeprecationWarning: BaseException.message has been deprecated as of Python 2.6
09:55 generj vm_['name'], exc.message
09:55 generj So… what… is that about?
09:57 Vivi-1 generj: try with python 2.6
09:57 Vivi-1 not 2.7
09:57 Vivi-1 the error you get is clear
09:58 Vivi-1 BaseException.message has been deprecated as of Python 2.6
09:58 generj my python is Python 2.7.3
09:58 Vivi-1 then, go to 2.6
09:59 generj Vivi-1: I don't quite understand how the warning is critical
09:59 babilen Vivi-1: That is simply a Warning, it shouldn't cause the actual error
09:59 generj Salt-cloud internally uses some deprecated syntax that still works in python 2.7.3, its just a warning
10:01 generj babilen: here's a full pastebin
10:01 generj http://pastebin.com/0CZXVwjj
10:03 jpaetzel joined #salt
10:10 babilen generj: Sorry, I have no idea about salt-cloud and Azure - And I can also see nothing obvious in your paste
10:15 johtso joined #salt
10:24 MrTango joined #salt
10:25 Shenril joined #salt
10:28 Kruno joined #salt
10:28 Kruno I have a strange problem - one of my minions has error [DEBUG   ] Fetching file from saltenv 'base', ** attempting ** 'salt://top.sls'
10:28 Kruno other minions get the file without problems
10:29 darrend hi guys, we're running into this inode problem with master's job cache: https://github.com/saltstack/salt/issues/3663  Anyone else seen this recently?  We're on 17.4
10:37 topochan joined #salt
10:40 ndrei joined #salt
10:43 Ztyx joined #salt
10:45 Ztyx My minion need to cache a certain file. Do you think it's best practice to store it in /var/cache/salt/minion/{{mycompany}}, or shall I put it in /var/cache/{{mycompany}}?
10:52 fllr joined #salt
10:54 fllr joined #salt
10:56 babilen What would be the most idiomatic way to require the installation of a certain package?
10:59 sgflt joined #salt
10:59 lionel joined #salt
11:00 elfixit joined #salt
11:11 Pate joined #salt
11:12 generj joined #salt
11:12 generj babilen: so it turns out my error/paste is because I tried spinning up a vm with the name 'newinstance' which was taken
11:12 generj and instead of showing me a graceful error message salt-cloud crashes
11:12 generj spinning up a vm with some other name works
11:17 ravibhure joined #salt
11:17 Kenzor joined #salt
11:21 sroegner_ joined #salt
11:25 babilen generj: Oh dear
11:27 bhosmer joined #salt
11:30 Pate joined #salt
11:34 Ryan_Lane joined #salt
11:36 Pate joined #salt
11:38 fllr joined #salt
11:40 _fllr_ joined #salt
11:41 jrdx joined #salt
11:45 Pate joined #salt
11:46 colttt joined #salt
11:46 colttt hello..
11:46 bhosmer joined #salt
11:47 colttt i have a short question: i has a plaintext.conf file but for every server i must change two lines in the config file..(for example i must set the hostname an dIP-adress of the server).. how can i do that?
11:48 malinoff colttt, file.blockreplace
11:48 malinoff or file.replace
11:51 viq or file.managed with templating
11:52 viq if you want the file to be completely managed by salt, instead of it just reaching out to an existing file and changing some lines
11:52 colttt thnaks!
11:52 fllr joined #salt
11:56 Pate joined #salt
12:09 aedocw joined #salt
12:11 Pate joined #salt
12:12 ndrei joined #salt
12:21 Pate joined #salt
12:23 Kenzor joined #salt
12:25 topochan joined #salt
12:31 raizyr joined #salt
12:33 baniir joined #salt
12:34 JordanRinke joined #salt
12:35 Pate joined #salt
12:41 Kenzor joined #salt
12:46 Pate joined #salt
12:48 joehh \join #libcloud
12:49 joehh joined #salt
12:52 ndrei joined #salt
12:52 ericholscher joined #salt
12:52 Pate joined #salt
12:52 fllr joined #salt
12:57 yomilk joined #salt
12:57 sashka_ua joined #salt
13:03 faldridge joined #salt
13:11 blee joined #salt
13:11 generj joined #salt
13:14 helderco joined #salt
13:16 generj1 joined #salt
13:22 sroegner_ joined #salt
13:29 NV hrm, is it kosher to use salt modules in a pillar sls file?
13:29 NV I was using it on 0.17.5 without issues, but in 2014.1 it seems that the salt dictionary in jinja is now an empty list
13:29 GoKage joined #salt
13:30 NV (I'm aware that the modules would run in the context of the master)
13:42 sroegner_ joined #salt
13:42 generj joined #salt
13:44 diegows joined #salt
13:50 akitada joined #salt
13:50 jslatts joined #salt
13:52 puppet NV: Neo-vortex?
13:52 fllr joined #salt
13:53 mattmtl joined #salt
13:54 NV puppet: <.<
13:54 NV >.>
13:54 NV heyia :)
13:54 NV I see you've found the salty light iriche
13:55 puppet NV: we are trying to start using it more then with salt '*' cmd.run 'puppet agent -t' ;)
13:56 NV luls
13:56 quickdry21 joined #salt
13:56 NV I was a puppetd --test --verbose fan myself :P
13:56 puppet NV: moved over fully to salt?
13:56 NV (We're also moving from a rather old version of puppet too)
13:56 NV not entirely yet sadly
13:56 NV new stuff is 100% salt
13:56 NV slowly moving old stuff from puppet to salt
13:57 NV but at the same time looking at excuses to drop old RHEL5 boxes and do a clean saltified rebuild instead of just swapping out puppet for salt too :P
13:58 juicer2 joined #salt
13:58 generj joined #salt
13:58 puppet NV: ah, we have started with cobbler now and looking into moving over new infra to salt
13:59 puppet NV: but it is the time issue, we could take cobbler, use old puppet config and BAM done
13:59 puppet NV: but yeah :/
14:00 rgbkrk joined #salt
14:13 ericholscher left #salt
14:13 Ponycorn joined #salt
14:15 Ponycorn Hi guys, i'm discovering Salt and i'd like to know if there is a way to edit a file on a minion when you apply a new state (concrete example: add 'myuser' at the end of 'AllowUsers' in sshd_config when i add 'myuser' to my minion) ?
14:16 xt yes.
14:16 Ponycorn That's a good news. How ? :)
14:18 xt storing users in pillars, and use the pillar variables for a file.append-state that uses the pillar variables is one way
14:18 xt or even file.sed
14:18 xt just generating the whole sshd_config using jinja template + pillar is another
14:18 xt you might want to read pillar tutorial / walkthrough
14:18 puppet Ponycorn: I would recomend to have a LDAP server for centralised userhandling
14:19 vegardx_ Unless you're talking about hundreds of users I'd say it makes sense to do it in salt rather than LDAP.
14:20 Ponycorn xt: thx i'll look at pillars and file.append-state/sed
14:20 puppet vegardx_: I am thinking on the point in that some system you use users in different places, WLAN/ssh/web/network/etc etc
14:21 Ponycorn puppet: it's not how we are working right now but i'll keep that idea for later (and i need 10 users max so far)
14:21 Ponycorn thx guys
14:21 gnugnu8 joined #salt
14:21 HeadAIX joined #salt
14:29 mpanetta joined #salt
14:30 Mr_N Speaking of pillar
14:31 Mr_N Anyone heard of a good reason why refresh_pillar / sync_all wouldn't refresh pillars; nor report anything on master or minion, both in ldebug?
14:32 oznt joined #salt
14:33 rgbkrk joined #salt
14:35 oznt should I worry about my PR in github breaking build in jenkins?
14:35 oznt Mike already merged it, and I only did documentation changes, so it's quite weired
14:36 arthurlutz joined #salt
14:37 arthurlutz [blog] Second Salt Meetup builds the French community http://www.logilab.org/blogentry/226418
14:37 Networkn3rd joined #salt
14:41 balltongu joined #salt
14:41 akoumjian joined #salt
14:42 HeadAIX joined #salt
14:43 Ztyx joined #salt
14:44 colttt left #salt
14:44 Furao et moi je fais un talk sur salt dans un PyCon dans 2 semaine :)
14:45 kiorky Furao: #salt-fr ? ;)
14:45 razor Thanks to whoever worked on the new 2014.1.0 debian packages. Much appreciated :)
14:45 helderco joined #salt
14:48 Furao kiorky: je ne suis pas français :P
14:48 kiorky Furao: ben fallait pas parler francais dans ce cas :)
14:48 HeadAIX joined #salt
14:49 Furao http://www.pycon.my/mini-pycon-my-2014/speakers
14:50 toastedpenguin joined #salt
14:52 fllr joined #salt
14:53 generj joined #salt
14:55 GradysGhost joined #salt
14:57 gnugnu8 joined #salt
14:57 budrose joined #salt
14:57 generj joined #salt
15:01 ndrei joined #salt
15:01 tomasfejfar joined #salt
15:02 jforest hey, I have a question, will OLD salt-minion clients be able to work with a new (hydrogen) salt-master?
15:02 jforest aka: salt-minion = 0.17.5, salt-master = 2014.1.0
15:03 jforest I have just found that my new minions (2014.1.0) don't work with master 0.17.5
15:03 themadcanudist joined #salt
15:04 tomasfejfar jforest: Would not surprise me - I had plenty of problems with .17 minions and .15 master
15:04 themadcanudist dave_den: Are you around?
15:04 vejdmn joined #salt
15:04 tomasfejfar Is there a quickstart to get salt-api working? I need to issue state.sls on saltmaster from remote server...
15:04 jforest tomasfejfar: any concrete info?
15:05 jforest I'm unwilling to update my master until I know!
15:05 generj joined #salt
15:05 tomasfejfar jforest: Nope. Just didn't work. Updating minions fixed it :D
15:05 tomasfejfar sry. updating master
15:05 jforest I guess I'll hold off until I get more info
15:05 tomasfejfar practically what I do now is keep versions in order (all the same)
15:05 tomasfejfar installed from repo - not using bootstrap
15:06 jforest tomasfejfar: I would like to, but 0.17.5 pkgrepo doesn't work on centos6
15:06 jforest so I am forced to use hydrogen for centos6
15:06 tomasfejfar But you can use bootstrap with forced version
15:07 tomasfejfar My stack is all deb7, so I did not have such problems
15:07 Mr_N Anyone here use win_repo?
15:08 jeremyfelt joined #salt
15:11 generj joined #salt
15:12 themadcanudist anyone know if it's possible to leverage the pydsl to bootstrap the mako templating language so that when I run a highstate it will install mako first before parsing my mako templates? Currently it does not work?
15:12 jergerber joined #salt
15:13 tmwsiy_ joined #salt
15:13 jeremyBass1 joined #salt
15:14 danielbachhuber joined #salt
15:16 rgbkrk joined #salt
15:17 kintel joined #salt
15:19 HeadAIX joined #salt
15:22 Ztyx joined #salt
15:24 rostam joined #salt
15:28 kintel_ joined #salt
15:30 ipmb joined #salt
15:36 Ztyx joined #salt
15:36 alunduil joined #salt
15:37 pydanny joined #salt
15:37 pdayton joined #salt
15:38 jmlowe joined #salt
15:39 jmlowe So how about updating the ppa for raring and saucy to 2014.1.0?
15:40 HeadAIX joined #salt
15:42 timoguin themadcanudist, pretty sure that's possible. you can define functions to run: https://github.com/saltstack/salt-bootstrap#adding-support-for-other-operating-systems
15:42 themadcanudist timoguin: checking...
15:43 timoguin it looks like those are distro specific functions, but maybe you can add a post function to install mako
15:43 themadcanudist timoguin: thank you!
15:45 themadcanudist oh wait
15:45 fllr joined #salt
15:45 themadcanudist timoguin, this is good info… but I didn't mean this process
15:45 themadcanudist I meant from an actual highstate
15:45 Mr_N I'm beginning to think my desire to use win_repo may be seated deeply in insanity
15:45 forrest joined #salt
15:48 timoguin themadcanudist, ah, in that case, i'd just make a state to install mako and run it before all our mako-templated states
15:48 timoguin *your
15:48 * themadcanudist nods - i have done that
15:48 themadcanudist but can I make it a part of the high state?
15:49 Mr_N I'd think some sort of require stanza would do it
15:49 Mr_N but I'm bad at things.
15:49 naiyte joined #salt
15:49 timoguin themadcanudist, sure, i'd just put it at the top of my top.sls, apply it to all hosts in the base environment
15:49 timoguin that should do it
15:50 timoguin it gets evaluated top-down
15:50 themadcanudist timoguin, again, i've done that
15:50 themadcanudist but i get exceptions
15:50 themadcanudist it's like it tries to render all the tempaltes first (with mako) before it applies the initial mako state install
15:51 seapasulli joined #salt
15:52 timoguin ah, well that's phooey. i guess it tries to render and order everything first
15:53 themadcanudist yeah
15:53 jslatts is compound matching broken in latest version of salt?
15:53 themadcanudist dave_den was saying that if I use the pydsl, I may be able to do this using: http://docs.saltstack.com/ref/renderers/all/salt.renderers.pydsl.html#render-time-state-execution
15:53 themadcanudist but i'm a bit confused as to how?
15:53 jslatts i'm getting "Failed matching available minions with compound pattern"
15:55 conan_the_destro joined #salt
15:56 frasergraham joined #salt
15:57 timoguin themadcanudist, yea I'm really not sure. I'd have to see an actual implementation of that, I think.
15:57 themadcanudist same
15:57 mpanetta Hey guys, quick question... Will a salt 0.17.5 master not work with a 2014.1.0-1precise1 minion?
15:58 mpanetta Because salt-cloud is installing that minion version but after it is done salt test.ping fails.
15:58 timoguin mpanetta, master version needs to be greater than the minion
15:58 techdragon joined #salt
15:58 mpanetta Ah ok
15:59 jmlowe That would have been a nice addition to the release notes
16:00 mpanetta Yeah...
16:00 mpanetta For some reason I though that 0.17.x was forward compatible with all versions after it (unless otherwise noted)
16:00 forrest mpanetta, jmlowe, yea we don't know WHY it isn't compatible, and I forgot to update the release notes last night. No one seemed to be aware of this.
16:00 timoguin agreed. there've been quite a few people in here trying that.
16:00 timoguin forrest, hasn't it been a thing for a while now that the master needs to be greater than the minion?
16:01 faldridge joined #salt
16:01 forrest timoguin, no, that's why it doesn't make sense
16:01 timoguin at least i seem to remember that being the recommendation since around version 10 or so
16:01 forrest timoguin, it was the recommendation
16:01 forrest but NOT required
16:01 timoguin gotcha
16:01 jmlowe forrest: wouldn't be such an issue if the ppa kept up with the releases
16:01 mpanetta Ive been running a 0.15 master with 0.16 minions fine for months
16:01 forrest only with 0.16->0.17 because of the zmq/security fixes
16:01 helderco joined #salt
16:01 forrest jmlowe, you'd have to ask joehh about the status of the ubuntu package
16:02 forrest mpanetta, yea that works fine
16:02 jmlowe there is a precise package, but no quantal, raring, or saucy package
16:02 forrest I don't know what is up with this release from 0.17.5 -> 2014.1
16:02 forrest jmlowe, again, a question for joehh :\ There might have been issues building it, he's usually pretty damn fast
16:02 mpanetta oddness...  It is putting a crimp in my upgrade plans ;)
16:02 mpanetta I will just try to upgrade the master
16:02 forrest this reminds me that I need to mess with the opensuse build system this weekend, oh goodie
16:02 forrest mpanetta, you aren't running multi-master are you?
16:02 mpanetta forrest: Nope
16:02 forrest ok good
16:03 forrest there's a bug for that
16:03 dave_den mpanetta: are you still running 0.16 minions?
16:03 mpanetta dave_den: We are going to do rolling upgrades of all the production minions
16:03 forrest timoguin, did you create an issue for getting the release notes updated?
16:03 mpanetta Our production salt server is still 0.16.x
16:03 mpanetta The test one is 0.17.5
16:03 faldridge joined #salt
16:03 forrest mpanetta, straight from 0.16 -> 2014.1? Ballsy
16:03 mpanetta forrest: We are just going to destroy the old minions ;)
16:04 forrest mpanetta, heh
16:04 jeremyfelt joined #salt
16:04 mpanetta They have no state anyway
16:04 timoguin forrest, no, will do now
16:04 * mpanetta loves cloudy things
16:04 dave_den yeah, 0.16 -> 0.17 master is the only time when you need your minions to be on 0.17 before the master is upgraded
16:05 mpanetta forrest: Actually the original plan was to upgrade everything to 0.17.5, but I was not aware that 2014 was out...
16:05 mpanetta I may still end up doing that.  Not 100% sure yet.
16:05 jslatts joined #salt
16:05 mpanetta Are the 2014.x things the new way versioning is going to be done?
16:06 timoguin yes
16:06 mpanetta Ok
16:06 jalbretsen joined #salt
16:07 mpanetta What version would you guys suggest I upgrade to? 0.17.5 or 2014.x?
16:08 themadcanudist dave_den: Hey dude, just wondering if you're available for a quick follow up question from your pydsl bootstrap recommendation?
16:08 forrest timoguin, great
16:08 forrest mpanetta, I've heard fewer issues on 0.17.5
16:08 dave_den themadcanudist: yeah, let me handle some morning biz real quick. back in about 30
16:08 forrest mpanetta, but a lot of people also didn't go to 0.17.5
16:08 forrest because it was such a fast swing through revisions
16:08 mpanetta Ahh
16:09 mpanetta I've been using it for a bit now.
16:09 forrest I have it on my VMs at home
16:09 timoguin themadcanudist, i think i understand it now.
16:09 timoguin forrest, https://github.com/saltstack/salt/issues/10932
16:09 forrest but that's not a 'real' production environment
16:09 forrest timoguin, so demanding :P
16:09 timoguin haha
16:09 timoguin UPDATE NOW
16:09 timoguin themadcanudist, https://gist.github.com/timoguin/9349413
16:09 mpanetta Yeah it is just a test env here...  But it still manages about 40 instances...
16:10 timoguin I think if you put that into an SLS that you put at the top of your top.sls, it'll execute at render time
16:10 forrest that reminds me guys, I updated the pkg state docs last night wit hthe 'names' option, this was an option I didn't know existed till saltconf
16:10 forrest http://docs.saltstack.com/ref/states/all/salt.states.pkg.html#salt.states.pkg.installed
16:10 forrest so if you'd rather install all your packages in a single package manager command, instead of per call, you can use that. Very handy.
16:10 forrest *the
16:11 ipalreadytaken joined #salt
16:11 themadcanudist dave_den: timoguin, thanks
16:11 themadcanudist checking
16:11 timoguin lemme know if that works. i haven't tested it
16:11 themadcanudist yeah, i already did that
16:11 themadcanudist it doesn't seem to execute before everything
16:11 themadcanudist it's first in the list of my base includes
16:11 themadcanudist same problem
16:11 themadcanudist tries to render all the mako templates before mako is installed
16:11 themadcanudist during a highstate
16:12 themadcanudist i'll show you mine, hold
16:13 mpanetta forrest: Ok I am going to try to force the minions to 0.17.5 then I think.
16:13 forrest mpanetta, sounds good
16:13 openxmark joined #salt
16:13 timoguin forrest, so that can reduce a shitload of 'apt-get install' calls into one?
16:13 forrest timoguin, exactly
16:13 themadcanudist timoguin: i addeda  comment… basically the same deal, you just need to ensure your requirements are met
16:13 themadcanudist but yeah, still have the same issue
16:13 mpanetta I may still put the master at 2014.x..  Are there any known interaction issues with such?
16:13 forrest timoguin, basically just adds all those names to a single call.
16:13 timoguin yea
16:13 forrest mpanetta, not that we're aware of
16:13 themadcanudist OH SHIT, i think i know why… sec
16:13 forrest timoguin, it's really awesome
16:13 mpanetta forrest: Ok sweet
16:14 mpanetta Thanks!
16:14 forrest np, let us know how it goes so we can help out others
16:15 themadcanudist timoguin: nevermind
16:18 vxitch why would state.show_top not return anything for a host?
16:18 vxitch /host/minion/
16:19 vxitch show_highstate is empty as well, and state.highstate fails with "Comment: No Top file or external nodes data matches found"
16:20 Gordonz joined #salt
16:20 JaredRo Looking for ideas: I'd like to do a salt.highstate, and later, be able to go to a given minion and be able to figure out when the highstate happened, and where the source code came from for the sls files (git repo, branch, commit).
16:21 jmlowe vxitch: is your master 0.17.* and your minion 2014.1.0?
16:22 Gordonz joined #salt
16:22 yesman joined #salt
16:22 vxitch oh..yes
16:22 jmlowe vxitch: that would do it
16:22 vxitch wth? i thought i installed it from the same repo as my master
16:23 jmlowe vxitch: if you are using the ubuntu ppa, it only has precise packages; quantal, raring, and saucy are still back at 0.17.5
16:23 vxitch i'm on RHEL 6.5
16:23 vxitch RHEL, RHEL Server Optional, and EPEL enabled
16:24 jmlowe vxitch: hmm, all of my epel based minions are updating to 2014.1.0, my master is ubuntu
16:25 vxitch oh oh oh, okay. 2014.1.0 is the new version
16:25 vxitch i just installed the master a week or two ago, if that. wasn't expecting it to be out of date already
16:25 bemehow joined #salt
16:25 vxitch and i forgot to schedule updates
16:25 vxitch alright, my master is behind, thank you
16:25 ajprog_laptop joined #salt
16:26 bdf has anyone submitted a port PR for 2014.1.0 for freebsd yet?
16:27 zach_ has anyone used salt-cloud with XenServer?
16:27 [diecast] joined #salt
16:27 vxitch any way to see changes in docs between 0.17.5 and 2014.1.0??
16:28 hunter joined #salt
16:29 tomasfejfar vxitch: git diff branches?
16:29 tomasfejfar is it possible to somehow "inject" a param to state.sls call that can be used inside the sls file?
16:31 scoates :stare: Comment:   The following packages failed to install/update: salt-minion=0.17.5-1~bpo70+1~dst.1.
16:31 forrest bdf, cedwards would be the man to ask that question to
16:31 scoates were the old wheezy packages removed when 2014.1 was added?
16:31 rojem joined #salt
16:32 forrest scoates, the way the ppa works is there can only be one set of packages in there at a time I believe. Yes I know it's stupid, no we can't fix it :P
16:32 scoates this… broke everything.
16:32 forrest scoates, yea I'm trying to remember if they are somewhere else...
16:32 forrest timoguin, do you remember where the old packages are from the ppa
16:33 scoates can I have 2014.1 minions talk to a 0.17.5 master, safely?
16:33 forrest scoates, nope :(
16:33 jmlowe scoates: some things will work, states don't
16:33 forrest scoates, we don't know why yet, but for some reason 2014.1 minions and a 0.17.5 master causes issues. There's no reason it should either
16:33 scoates ok… how about 0.17.5 minions talk to a 2014.1 master?
16:33 ndrei joined #salt
16:33 forrest scoates, that should work fine
16:34 forrest scoates, mpanetta is testing it now
16:34 Gareth I tested that last night but no go.  Seems it has be 2014.1 across the board.
16:34 Gareth has to be.
16:34 forrest Gareth, are you serious??
16:34 forrest that sucks
16:34 scoates I hate to be a complainer, but… this breaks my entire deployment. I can't spin up new nodes right now.
16:34 forrest mpanetta, looks like I was wrong, Gareth is saying it doesn't work.
16:34 mpanetta Ah crap
16:34 mpanetta I see
16:34 scoates my only option seems to be: drop everything, and test + deploy 2014.1 ?
16:35 forrest scoates, yea I understand completely.
16:35 forrest scoates, well, what release are you on?
16:35 mpanetta Whatever happened to version compatibility? :P
16:35 Ahlee masters have always supposed to move first
16:35 scoates forrest: 0.17.5 everywhere
16:35 forrest mpanetta, I don't know man, it's been good for all of 0.17 and prior to the 0.16->0.17 security issue
16:35 forrest I have no idea why this release is screwing that up.
16:35 Gareth forrest: Yeah.  Upgraded by master last night, minions would connect, test.ping would work but highstates wouldn't run.
16:35 jmlowe scoates: you can use 0.17.5 master to pkg.upgrade to get all of your minions up to 2014.1.0 then do your master
16:35 Gareth forrest: highstates or state.sls runs.
16:36 forrest jmlowe, yea but 2014.1 has issues, specifically multi-master setups
16:36 forrest Gareth, that sucks
16:36 Ahlee oh?
16:36 Mr_N I haven't gotten multi master to work the way I want even _without_ 2014 -.-
16:36 Mr_N this bodes well ><
16:36 Ahlee heh.
16:37 Mr_N dumped the last half a week or so on failing at win_repo, too.  BLUGH.
16:37 scoates 11:37AM: too early to start drinking? Perhaps not.
16:37 Mr_N do not need an update right in the middle of this ;_;
16:37 Gareth forrest: nod.  definitely not ideal :)
16:37 dave_den themadcanudist: back
16:37 themadcanudist dave_den: cool
16:38 dave_den so it's not working?
16:38 themadcanudist hrm, actually, give me a sec. I may have figured it out
16:38 scoates perhaps a secondary ppa could be set up with salt-old ?
16:39 forrest Gareth are you on ubuntu?
16:40 Gareth forrest: debian
16:40 forrest Gareth, were you able to pull the 0.17.5 package?
16:40 forrest I can't remember if it's stored somewhere else once the PPA upgrades.
16:40 Gareth let me see.
16:40 scoates I'm going to put the 0.17.5 debs into our private repo for now. this probably breaks a lot of people, though…
16:40 Mr_N Ok.  Even copying the fucking EXAMPLE firefox repository sls file doesn't work the way the documentation says.  Time to stop feeling bad about making an issue about this -.-
16:41 forrest scoates, yea I know, I will try to make a PR on the release notes when I go home about the compatibility issue. This is a major problem.
16:41 scoates forrest: again, I'm sorry if this sounds like complaining, but this isn't even a release problem. I haven't upgraded. I wouldn't have read the release nodes yet.
16:41 scoates this breaks *old* deployments.
16:42 forrest scoates, You aren't gonna hear any argument from me man.
16:42 Gareth forrest: manually yes.  but not through apt.  looks like they're not in the package list anymore.  just the 2014.1 release.
16:42 forrest and I understand the complaining
16:42 scoates if you guys need these, I have them
16:42 forrest Gareth, yea that's the standard
16:42 scoates salt-common_0.17.5-1~bpo70+1~dst.1_all.deb salt-minion_0.17.5-1~bpo70+1~dst.1_all.deb
16:42 forrest scoates, I might need them this weekend depending...
16:42 dave_den forrest: what's the difference between 'names' and 'pkgs' for pkg.installed?
16:43 forrest dave_den, names does all the installs in a single command
16:43 scoates I have this, too, actually: salt-master_0.17.5-1~bpo70+1~dst.1_all.deb
16:43 forrest pkgs does each one as it's own command
16:43 forrest dave_den, was I not very clear on the docs? It was getting late when I wrote those, might need a review/rewrite on the description :P
16:43 cedwards bdf: I've submitted a PR. I'm just waiting for the commit.
16:43 forrest cedwards, have you upgraded any of your machines?
16:43 forrest are you seeing the compatibility issue on freebsd as well?
16:44 davidone I'm reading this: http://docs.saltstack.com/ref/modules/all/salt.modules.nginx.html
16:44 dave_den pkgs should do it all in one command as well, right?
16:44 davidone but it seems that nginx.status cli command isn't working
16:44 davidone (i'm using salt-* 2014.1.0+ds-1~bpo70+1 on wheezy)
16:45 cedwards forrest: my systems are running 2014.1.0 based on my patch.. which issue?
16:45 davidone all the other commands (nginx.version, nginx.signal and so on) work successfully
16:45 forrest cedwards, compatibility problems between 0.17.5 and 2014.1
16:45 forrest if either master or minion isn't the same, states can't be run
16:46 forrest dave_den, pkgs does each command as it's own call
16:46 rgbkrk joined #salt
16:46 forrest so yum install 1, yum install 2, versus yum install 1 2
16:46 yesman does pkg.installed rely on pkg.install? I've got a pkg.installed state failing while calling salt-call pkg.install works with the same arguments
16:46 davidone ok, found the issue
16:46 scoates if it would help, I could set up a very-makeshift debian wheezy apt repo for 0.17.5 (if this works)
16:47 forrest yesman, what's the error?
16:47 davidone status is not like service nginx status
16:47 dave_den forrest: for apt, pkgs does it all in one. not sure about rpm
16:47 davidone it's retrieving /status uri
16:47 forrest dave_den, really? Hmm.
16:47 yesman forrest: sudo salt-call state.single pkg.installed name=nodejs fromrepo=epel
16:47 yesman fails
16:48 forrest with what error?
16:48 yesman Package 'nodejs' not found (possible matches: nodejs.i686, nodejs)
16:48 dave_den forrest: yeah, just confirmed again on 0.17.5.  we should make pkgs work the same way for yum if it doesn't already
16:48 yesman sudo salt-call pkg.install name=nodejs fromrepo=epel works
16:48 forrest dave_den,  agreed, I thought the whole point of name was to do it all in one command, maybe I have it the other way around?
16:48 forrest dave_den, it's been a while since I opened the issue.
16:49 xmltok joined #salt
16:50 forrest yesman, does it still fail in a state?
16:50 vxitch how can i target the master without using the hostname?
16:50 forrest vxitch, IP address
16:50 vxitch ah, there's no built in thing that hits the master?
16:50 yesman forrest, yeah that's what led me to debug it down to this call. I think it's failing when querying for available packages
16:50 analogbyte basepi: you commented on issue #9741 that states are not namespaced by environment... why is that? I just tried to replace my base env with a legacy one via gitfs to be able to work on master (=base) without breaking salt in the meantime...
16:50 dave_den from what i understand, 'names' should cause a state to be created for each item of 'names', all with the same args. that does not seem like what you mean to intend for pkg.installed..
16:50 yesman but I wanted to confirm this was a bug before trying to fix it
16:51 Mr_N Hmm...  Is there really no way to insert Jinja or some other ability to do dynamic content into win_repo sls files?
16:52 forrest dave_den, whiteinge, can you clarify 'names' when used in pkg.installed? I updated the docs last night (http://docs.saltstack.com/ref/states/all/salt.states.pkg.html#salt.states.pkg.installed ) and want to make sure they are correct.
16:52 scoates forrest: if this helps http://files.seancoates.com/salt-0.17.5-debs/
16:53 scoates those are Wheezy
16:53 forrest scoates, great thanks
16:53 scoates and FWIW, putting them into our custom/private apt mirror seems to have worked
16:53 gmoro joined #salt
16:54 taharqa joined #salt
16:54 dave_den forrest: this is the behavior i imagine for using 'names' https://gist.github.com/dlanderson/9b60a0b307593282d5fa
16:54 davidone ok, found why salt '*' nginx.status cannot be used
16:55 forrest dave_den, yea, each as a single command, I might have switched them up last night
16:56 Gareth hm. may have jumped the gun.  re-testing a 2014.1 master talking to a 0.17.5 minion.  seems to be working.
16:56 racooper joined #salt
16:56 taharqa left #salt
16:56 forrest Gareth, :O
16:56 forrest I already told s0undt3ch that we should test that stuff.
16:57 yesman forrest, sorry to keep buggin ya, I'm gonna go ahead and fix the state stuff unless this is expected behavior. I think it's a bug in the way available packages are listed when enablerepo is used
16:58 jforest Gareth: that would be very awesome to know
16:58 Gareth well thats very odd.  was just to able to upgrade it on the 0.17.5 minion too, via a salt call.
16:59 yomilk joined #salt
16:59 jcockhren Gareth: twilight zone
16:59 Gareth jcockhren: seriously.  very strange.
16:59 rainmanjam joined #salt
16:59 timoguin forrest, is there an open issue to figure out why a 2014.1 minion can't talk to a 0.17 master?
16:59 timoguin i'm adding to the release notes
17:00 forrest timoguin, I just spoke with s0undt3ch and he says that is expected behavior
17:00 timoguin it still needs to be in the notes though, yea? seems to have caught a lot of people up
17:00 amckinley joined #salt
17:01 forrest timoguin, yea of course
17:01 forrest timoguin, it also needs to be in the FAQ
17:01 forrest http://docs.saltstack.com/faq.html#can-i-run-different-versions-of-salt-on-my-master-and-minion is not thorough enough from my initial write up.
17:02 scoates fwiw, jperras just noticed this: deb http://debian.saltstack.com/debian squeeze-saltstack-017 main
17:02 jperras yeah
17:02 forrest timoguin, I'm creating an issue nw.
17:02 forrest *now
17:03 forrest So is anyone encountering actual issues with a 2014.1 MASTER talking to a 0.17.5 minion? Other than Gareth's false report and lies :P
17:03 jperras separate archives for each release stream wasn't very obvious until I actually started looking at the archives
17:07 tomasfejfar can i do something like `salt-call state.sls some-state param=10`?
17:08 KyleG joined #salt
17:08 KyleG joined #salt
17:08 mgw joined #salt
17:09 jperras tomasfejfar: yes
17:10 jperras you don't need the state.sls, iirc
17:10 jperras can just do `salt-call some.state param=10`
17:10 baoboa joined #salt
17:11 tomasfejfar jperras : and how would i access the param then?
17:11 hunter joined #salt
17:11 jperras tomasfejfar: the param gets sent to the callable as a kwarg
17:11 racooper good morning.  I'm having an odd issue with an Oracle Linux server after updating to minion v2014.1.0 from EPEL.  pkg.list_upgrades is no longer showing any previous updates as available, only one that came available after the minion update.
17:12 carlos joined #salt
17:12 racooper pastebin of salt-call pkg.list_upgrades and a yum check-update: https://gist.github.com/racooper/9350774
17:12 tomasfejfar jperras: let's say "name=MyProject" - now I want to checkout git://myproject into /var/www/gitproject - how would I write the state?
17:12 jperras if you have salt-call foo.bar param=10, that sends kwargs of {'param': 10} to the bar() callable in the foo state
17:12 terminalmage timoguin: it's already being added to the release notes, hold off
17:12 tomasfejfar jperras: But I need it for state.sls :)
17:13 yesman tomasfejfar: sounds like a good use of grains there, which you can also set locally
17:13 timoguin terminalmage, I already sent a PR
17:13 dave_den tomasfejfar: salt-call state.sls yourstate pillar="{'yourkey': 'yourvalue'}"
17:14 dave_den tomasfejfar: in your state sls file, access it as {{ pillar['yourkey'] }}
17:14 dave_den (jinja)
17:14 tomasfejfar dave_den: That's it!!! Thank you so much!
17:14 jperras right
17:14 terminalmage timoguin: it will probably not be merged, then.
17:14 jperras that's in the docs, tomasfejfar
17:14 tomasfejfar dave_den: and the pillar data get's merged or replaced
17:14 timoguin terminalmage, phooey!
17:15 jperras tomasfejfar: https://github.com/saltstack/salt/blob/develop/salt/modules/state.py#L334
17:15 dave_den merged in 0.17.5 i believe. i have not tested it
17:15 terminalmage timoguin: basepi is already working on it along with other additions to the release notes
17:15 terminalmage timoguin: sorry
17:15 timoguin np
17:15 tomasfejfar jperras: I was looking at http://salt.readthedocs.org/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.sls
17:15 terminalmage timoguin: thanks for the eagerness though!
17:16 jperras tomasfejfar: that's probably a bit out of date. That example was added on
17:16 jperras 2014-02-26
17:16 jperras [ https://github.com/saltstack/salt/commit/01907556c4d39bb5dc769871a7e2d9e44c4d4640 ]
17:17 tomasfejfar jperras: I supposed it's builded regularly :(
17:17 basepi timoguin: i merged yours, i'll rebase.  =)
17:17 jperras tomasfejfar: well to be fair, that was only like a week ago
17:17 jperras :)
17:18 tomasfejfar :S
17:18 jperras I just have a habit of reading docs in code for that exact reason
17:18 tomasfejfar :D
17:18 jdenning joined #salt
17:18 c0deli0n joined #salt
17:18 c0deli0n Hi all
17:19 forrest hello
17:19 c0deli0n Just ran a package update for salt-master / minions
17:19 fllr Guys, I'm getting this error after upgrading to 2014.1.0: [ERROR   ] Unable to process requirements file salt://django/dev-requirements. Error: pip requirements file 'salt://django/dev-requirements' not found
17:19 c0deli0n and something seems to have broken on the postgresql states
17:19 forrest fllr, are both your master and minions 2014.1?
17:19 forrest same for you c0deli0n, what release?
17:19 fllr forrest: this is a masterless setup
17:20 forrest fllr, awesome, is the file missing?
17:20 forrest fllr, if not can you run it with debug and see if you get anything else?
17:20 Ztyx joined #salt
17:20 c0deli0n 2014.1.0-1.el6
17:20 fllr forrest: It's not, it used to work fine previously. I double checked that so many times now. Lol.
17:21 fllr Oh, yeah... I also get this eror
17:21 fllr error*
17:21 fllr forrest: [ERROR   ] Unable to cache file 'salt://django/dev-requirements' from saltenv '/opt/mmi/virtualenvs/mutualmind'.
17:21 c0deli0n psql: fe_sendauth: no password supplied
17:21 c0deli0n thats the error
17:21 dave_den fllr: you can try removing your /var/cache/salt dir
17:21 terminalmage forrest: if this was due to the protocol difference, the minion wouldn't have gotten the command and wouldn't have tried to run the installation, and wouldn't have generated that error
17:21 terminalmage forrest: talking about the pip issue referenced above
17:21 c0deli0n I am supplying the "user" as postgres and it is passwordless
17:22 fllr dave_den: I've actually just created this vm, but i'll try that...
17:22 fllr dave_den: Same thing...
17:23 c0deli0n I also put a host    all             postgres         127.0.0.1/32           trust in my hba file
17:23 c0deli0n not a fan of this .. but just for trial
17:23 c0deli0n but still getting the same error
17:23 c0deli0n it worked till yesterday without any issues
17:23 forrest terminalmage, gotcha
17:23 c0deli0n so not sure if something changed on salt or postgres
17:24 fllr forrest: dave_den: this is the interesting part of -l debug: https://gist.github.com/felipellrocha/887d9630552036c73bdb
17:24 terminalmage fllr: that error suggests you're using "/opt/mmi/virtualenvs/mutualmind" as the environment
17:24 terminalmage fllr: can you post your SLS file for this state?
17:25 fllr terminalmage: I'm using that as the enviroment for my django install, not for salt... yeah... gimme a sec
17:25 fllr terminalmage: https://gist.github.com/felipellrocha/3bc59aab9666e1aea6ef
17:26 terminalmage fllr: change that "env" line to "bin_env"
17:26 scoates yep. deprecated: http://docs.saltstack.com/ref/states/all/salt.states.pip.html
17:27 fllr aaaah
17:27 terminalmage "env" is used for salt fileserver environments, but has been deprecated in favor of "saltenv" to reduce confusion like this
17:28 terminalmage bin_env is used to point pip at a virtualenv
17:28 terminalmage which I'm guessing is what you want to do
17:29 jtang1 joined #salt
17:30 c0deli0n anyone run into a postgres issue lately? haven't see a bug reported yet ..
17:31 fllr terminalmage: Yep. It worked! :) And it makes sense...
17:31 terminalmage fllr: sweet!
17:32 pdayton joined #salt
17:36 c0deli0n salt job throws me the following error: AttributeError: 'NoneType' object has no attribute 'version'
17:36 c0deli0n I found a thread where someone had an issue since they were using a different port
17:37 c0deli0n I am too, but I have passed the postgres.port in pillars
17:37 c0deli0n and salt is picking it up
17:37 wendall911 joined #salt
17:37 martoss joined #salt
17:38 c0deli0n from the logs: /usr/bin/psql --no-align --no-readline --username postgres --host localhost --port 50432 --dbname postgres
17:38 c0deli0n so ruling that out
17:38 UtahDave joined #salt
17:39 forrest Welcome to the party UtahDave! I hope you brought the chips.
17:39 UtahDave Mmm. chips.
17:39 forrest with a bucket of salsa
17:40 UtahDave OK, now I'm in for sure.
17:40 schimmy joined #salt
17:41 timoguin no double dipping
17:43 jcockhren ok. we got to do something about the varing output given to returners and logging
17:43 jcockhren for example... always have 'retcode'
17:44 jcockhren present.
17:44 jcockhren and 'success'
17:45 hunter joined #salt
17:45 juicer2 left #salt
17:46 Ztyx1 joined #salt
17:48 runnyspo1 joined #salt
17:49 chrisjones joined #salt
17:49 schimmy1 joined #salt
17:51 cro_ joined #salt
17:51 cb_ joined #salt
17:53 nicksloan joined #salt
17:55 btorch is it possible to specify the version of salt to grab from the repo ?
17:56 harobed__ joined #salt
17:57 Xe joined #salt
17:57 Ztyx joined #salt
17:57 harobed__ joined #salt
17:58 ml_1 joined #salt
17:59 Tinuva joined #salt
17:59 Nazzy joined #salt
17:59 Nazzy joined #salt
17:59 mfournier joined #salt
18:00 austin987 joined #salt
18:00 Gareth chips...chips...chips....chips...
18:00 torrancew joined #salt
18:01 millz0r left #salt
18:02 btorch I ask about the version cause I just poped a new box but that grabbed 2014.1.0-1precise1 and that seems to have broken something
18:02 jperras btorch: check http://debian.saltstack.com for versioned repos
18:02 Linz joined #salt
18:03 jperras under "Release Streams"
18:03 Linz joined #salt
18:05 forrest btorch, masterless minion?
18:06 forrest btorch, you can specify via the bootstrap, but only for specific releases from git
18:06 forrest there are some examples of doing so over here: http://docs.saltstack.com/topics/tutorials/salt_bootstrap.html
18:06 c0deli0n ok, I figured out my issue. I was passing the host as "localhost" explicitly in the state, which for some reason it wasn't liking.
18:06 forrest c0deli0n, interesting
18:08 rdorgueil joined #salt
18:10 che-arne joined #salt
18:10 Chrisje joined #salt
18:11 Ztyx joined #salt
18:12 btorch forrest: well I'm turning debug to check it out .. the master is 0.17.5 ... and the minion used to be that as well
18:12 johtso joined #salt
18:12 forrest btorch, that is a problem, you can't run a 0.17.5 master and a 2014.1 minion
18:12 forrest doesn't work
18:12 Ryan_Lane joined #salt
18:13 kaptk2 joined #salt
18:14 Thiggy joined #salt
18:14 btorch yeah I guess I'll try what jperras mentioned since I need some sort of stability
18:14 forrest cool
18:15 btorch jperras: do you know if that is ok for ubuntu though ?
18:15 bhosmer joined #salt
18:18 rdorgueil joined #salt
18:21 Gareth fxhp: answering your question from #scale12x....Yes.  I am :)
18:22 Ztyx joined #salt
18:23 [diecast] joined #salt
18:23 ckao joined #salt
18:24 smcquay joined #salt
18:27 kermit joined #salt
18:27 nicksloan left #salt
18:29 TheRealBill joined #salt
18:30 lzhang joined #salt
18:31 Zuru joined #salt
18:32 bhosmer joined #salt
18:33 fxhp Gareth: awesome, I had lots of fun, send my thanks to the whole team that ran SCaLEx12 this year
18:33 fxhp You guys are great
18:33 Gareth fxhp: will do. :) Thanks for the compliments.
18:35 btorch anyone tried running these pkgs http://debian.saltstack.com/debian on ubuntu ?
18:37 saurabhs joined #salt
18:37 JordanRinke Morning
18:38 Nazca joined #salt
18:38 Nazca joined #salt
18:39 chutzpah is the 2014.1.0 dependency on libcloud optional or is it a hard dep? I just got bug filed suggesting it's optional, but it's listed in the requirements.txt as a hard dep.
18:40 mpanetta Um... Does "script_args: '-- git v0.17.5'" not work anymore?
18:41 tomasfejfar Is there an option to make salt-run return in different format? It does not seem like it is :( I'm trying to use salt --async + wait 10s + salt-run jobs.lookup_jid
18:42 tomasfejfar I tried salt -t5 state... but it does not exit after 5s, but hangs until minion returns (and waits for minions that are down)
18:42 tomasfejfar I am not sure if that's a bug or feature
18:42 timoguin chutzpah, I'm pretty sure it's only required for the cloud modules/states.
18:42 redondos joined #salt
18:42 redondos joined #salt
18:43 tomasfejfar (i have timeout: 600 in my master config)
18:44 budrose joined #salt
18:45 mpanetta This is bad... I can't force the minion version anymore...
18:45 tomasfejfar mpanetta: It should - try chaing v0.17.5 to 0.17.5 or somthing like that
18:46 mpanetta Hmm, ok.  For 0.16.4 I had to have the v there...
18:46 tomasfejfar mpanetta: should=it'm not sure :D
18:46 mpanetta haha ok :)
18:48 tomasfejfar mpanetta: yes, it's the other way around - you need the "v" :( sorry, hehe
18:48 mpanetta Hrm
18:48 Vesperatus joined #salt
18:48 mpanetta well with the V it failed.
18:48 timoguin I think v0.17.5 will installed from that tag. is you do just 0.17, it will install the latest in that branch
18:48 mpanetta I am trying again without, I started before you said that and it isn't done yet heh.
18:49 mpanetta timoguin: When I tried with v0.17.5 the script failed.
18:49 mpanetta Let me see if I can get more info
18:50 tr_h joined #salt
18:50 tomasfejfar timoguin: yes, that's it, now i remember - I tried 0.17.5 (and the branch was not found)
18:50 Vesperatus Good day.
18:50 mpanetta Weird
18:50 mpanetta without the V it is cloning the repo fine...
18:50 mpanetta seems to be so far anyway
18:51 btorch man this blows
18:51 tomasfejfar mpanetta: i might have been right? wow :) that's unexpected - i was convinced I got it wrong :D
18:52 btorch how stable is this 2014.1.0-1precise1 ?
18:52 Vesperatus I have a question regarding two commands that should not run at the same time but the order does not matter.
18:52 mpanetta tomasfejfar: Well we shall see, it is taking a lot longer this time, hopefully it won't fail!
18:52 mpanetta Nope it failed...
18:52 mpanetta error: pathspec '0.17.5' did not match any file(s) known to git.
18:53 mpanetta Let me try with the v and see what errors I get
18:53 tomasfejfar that's my error mpanetta
18:53 timoguin mpanetta, i haven't read all the scroll-back, but are you talking about using script_args with salt-cloud?
18:53 mpanetta timoguin: Yep
18:53 tomasfejfar Vesperatus: make them required (one in other) - if they are on the same minion
18:54 dangra joined #salt
18:54 mpanetta Here is what one of my profiles looks like that uses it:
18:54 mpanetta https://gist.github.com/anonymous/eb90cc8248e9151f648e
18:54 conan_the_destro joined #salt
18:54 mpanetta I used it fine with 0.16.4
18:54 mpanetta But now I am having issues for some reason :(
18:54 tomasfejfar Vesperatus: that way order does matter, but it won't run simultaneously
18:54 timoguin hmm, I'm using the Ubuntu PPA with my salt-cloud setup, not passing and script_args
18:54 timoguin But I have a local setup with Vagrant where I'm successfully installing 0.17 from git
18:55 mpanetta Well, let me see if I can find the error, last time it errord I did not have enough scrollback
18:56 snuffeluffegus joined #salt
18:57 Vesperatus Trying to figure how to send you a private message..
18:58 mpanetta Vesperatus: /msg <nick>
19:01 Ztyx joined #salt
19:01 mpanetta Crap it failed again... Ok lets see why...
19:01 Linz joined #salt
19:02 mpanetta https://gist.github.com/anonymous/b52259cd5acf5509524c <-- tomasfejfar, timoguin that is the error...
19:03 n8n joined #salt
19:03 mgw joined #salt
19:03 themadcanudist dave_den: I solved the bootstrap issue
19:03 tomasfejfar mpanetta: that's too late... " *  WARN: Upstart does not appear to know about salt-minion" --> error was before
19:03 themadcanudist It was what I thought. You can't have mako as the default renderer in master.conf.. which is annoying, but workable.
19:03 dave_den themadcanudist: cool, what was the issue?  i tested it over here and it worked as well
19:04 dave_den ah
19:04 themadcanudist once you turn that off
19:04 themadcanudist then in your bootstrap.sls you make that a pydsl, with the appropriate options, voila
19:04 mpanetta tomasfejfar: I don't see any other errors.  Maybe it is because the git install of salt does not install an upstart control file?
19:04 themadcanudist semi-annoying, but it works
19:04 themadcanudist dave_den: thank you !
19:04 btorch I still see 0.17.5-1 on the ppa repo but apt-get specifying that version doesn't seem to work .. anyone know a repo for ubuntu pkgs that would allow that ?
19:05 dave_den themadcanudist: i suppose you could always have your pydsl state file that installs mako set the minion configuration options in /etc/salt/minion.d/default_renderer.conf once mako is installed...
19:05 dave_den that doesn't help on the initial highstate though
19:06 * themadcanudist nods
19:06 tomasfejfar mpanetta : maybe, no idea though
19:07 timoguin mpanetta, i'm testing real quick on my VPS
19:07 mpanetta This is not good...  Is there any way to tell the bootstrap to use a specific PPA package version then?
19:07 mpanetta timoguin: Thank you!
19:07 btorch mpanetta: hehe I've been trying that but doesn't like it
19:08 seapasulli joined #salt
19:08 mpanetta btorch: Ah crap hah
19:09 timoguin mpanetta, okay, i just successfully bootstraped from v0.17.5 on ubuntu precise
19:09 timoguin creating gist
19:09 mpanetta Oh cool, can't wait to see it!
19:10 btorch I was just now reading over the binary-amd64/Packages on ppa repo and only has the new one
19:10 timoguin mpanetta, https://gist.github.com/timoguin/9353429
19:10 btorch timoguin: what about those that don't use the bootstrap script ?
19:10 timoguin i ddid my script_args differently
19:10 mpanetta Ah!  Ok trying now.
19:10 mpanetta Thank you!
19:11 timoguin btorch, i've been using what's in the PPA. that might be an issue now if it automatically tries to install 2014.1 though
19:11 btorch yeah it does
19:12 timoguin it'd be nice to able to pass a different PPA to the bootstrap script
19:12 mpanetta +1!
19:12 timoguin ppas are pretty limited
19:15 vxitch i'm getting strange py errors when calling highstate --> http://hastebin.com/watolanasa.rb
19:15 vxitch what do they mean? I don't have any { } in my sls, here's the sls http://hastebin.com/nohimaheli.sm
19:16 vxitch well, the relevant part of the sls
19:16 jcsp left #salt
19:16 vejdmn joined #salt
19:16 mpanetta timoguin:         Executing the command '/tmp/.saltcloud/deploy.sh -c /tmp/.saltcloud -M git v0.17.5' failed  :(
19:17 mpanetta I have no idea what I am doing wrong :(
19:17 mpanetta Here is one of my profiles
19:17 mpanetta https://gist.github.com/anonymous/0c4e8dde7a15ebd91d65
19:18 mpanetta timoguin: what version of salt-cloud are you running?
19:18 mpanetta Mine is salt-cloud 0.8.11
19:24 ajw0100 joined #salt
19:26 tr_h joined #salt
19:26 timoguin mpanetta, same
19:26 timoguin and ubuntu 12.04 on master and minions
19:26 mpanetta Crap
19:26 mpanetta Why isn't it working for me...
19:27 timoguin that's what i'm wondering
19:28 johtso Oh wow.. pkg. upgrade "some-package" ignores the argument and does a bare apt-get upgrade?
19:29 johtso that was unexpected, should be fun trying to sort out this mess :(
19:29 quickdry21 joined #salt
19:30 mpanetta timoguin: Was the system you bootstrapped on a fresh system with no prior salt install?
19:30 timoguin mpanetta, yes
19:30 timoguin the base image on digital ocean
19:30 mpanetta Hmm, this is so weird
19:30 thayne joined #salt
19:30 timoguin so the bootsrap installs git and deps too
19:31 johtso think that's a lesson learnt, running remote commands using salt can be dangerous, as you get no "are you sure you're not being an idiot?" prompts
19:31 timoguin mpanetta, i'd try remoting into the new machine and running that command manually
19:32 mpanetta timoguin: Which?  The bootstrap?
19:32 timoguin yes
19:32 mpanetta Ok
19:32 mpanetta BTW, what does your install say here:
19:32 mpanetta https://gist.github.com/anonymous/ce76c51103d7d0899962
19:32 timoguin salt's error output there isn't too helpful
19:32 mpanetta It does not seem to install any upstart things...
19:33 mpanetta Yeah I was looking at the install output
19:33 vxitch can someone please give me an idea of what this means? http://hastebin.com/watolanasa.rb
19:35 timoguin mpanetta, my output looks the same as yours for that section. the order it's operating on each one is slightly different though.
19:35 timoguin but it's operating on all the same files
19:36 yesman sorry if this is an obvious question but is there a reason pkg.installed and pkg.install use completely different code paths for installing a package? at east on yum
19:37 rallytime joined #salt
19:37 timoguin vxitch, that looks like it might be a bug in the iptables state
19:39 timoguin yesman, what do you mean by code paths?
19:39 ml_1 joined #salt
19:39 timoguin i'm guessing you're seeing the difference between calling a state module vs. an execution module
19:39 yesman meaning if I call pkg.install, it uses yumpkg.install
19:39 yesman but if I call pkg.installed, it never uses that function call
19:40 timoguin what does pkg.installed use?
19:40 yesman and in my case, is failing to install packages when an extra repo is enabled
19:40 yesman pkg.installed is using states/pkg.py:installed
19:41 jmlowe left #salt
19:41 yesman I posted this a bit earler, basically:  sudo salt-call pkg.install name=nodejs enablerepo=epel works and sudo salt-call state.single pkg.installed name=nodejs enablerepo=epel fails
19:41 Gareth vxitch: that looks fun :)
19:42 johtso Is there a way to make test=True the default?
19:42 johtso when running commands from the master
19:42 yesman so yumpkg.py's install function works with enabled repos, but states/pkg.py is failing to install the package, despite an exact match being found: Comment: Package 'nodejs' not found (possible matches: nodejs.i686, nodejs)
19:42 timoguin yesman, hmm, the state should be calling the same functions as the exec module
19:43 vxitch timoguin, Gareth: I'm looking through /usr/lib/python2.6/site-packages/salt/states/iptables.py and at line 493 i'm not finding anything out of the ordinary. the whole flush method looks just fine actually.
19:43 yesman right? I've been knee deep in pdb calls all day trying to see where the two diverge
19:43 yesman states/pkg.py:installed never calls modules/yumpkg.py:isntall
19:43 Gareth vxitch: nod. im looking at the code now too.  looks fine.
19:44 Gareth vxitch: entirely possible its not at line 493 though.
19:44 yesman johtso: you should be able to specify `test: True` in your /etc/salt/master
19:44 helderco joined #salt
19:44 johtso nice
19:44 jeffro joined #salt
19:44 Ryan_Lane joined #salt
19:45 Gordonz_ joined #salt
19:46 timoguin yesman, yea, it's really weird. because the state directly calls the pkg.install module
19:47 mpanetta_ joined #salt
19:47 Gordonz__ joined #salt
19:47 mgw joined #salt
19:48 yesman timoguin, this is also only happening with enablerepo/fromrepo btw, but once I got into the code I realized the two separate code paths for installing packages
19:49 Gordonz joined #salt
19:50 vxitch is there a way to see the compiled sls?
19:50 kermit joined #salt
19:50 vxitch as in, the actual string that python uses after YAML/jinja is parsed
19:51 timoguin vxitch, yea, salt 'foo' state.show_sls state
19:51 mpanetta timoguin: Hmm ok.  I will log in to the machine and manually run the bootstrap and see what it does
19:52 vxitch timoguin: thanks
19:52 res0nat0r joined #salt
19:53 vxitch timoguin: ah no, that doesn't spit out the string. it prints the pretty and verbose output salt usually gives back
19:54 mpanetta timoguin: Yeah, same output, same error when I run it by hand.
19:56 mpanetta There are upstart files in the pkg dir though
19:56 mpanetta bleh
19:56 seanz Here's an easy question: How do I use the 'os' module in a salt state?
20:03 xmltok joined #salt
20:03 andrej If I have iptables in centos set-up, I have one service (iptables), but two packages (iptables & iptables-ipv6).  How would I formulate something like this in the sls yaml-lingo?
20:04 Ryan_Lane joined #salt
20:05 yesman andrej, pkg.installed: pkgs: -iptables -iptables-ipv6 should work
20:05 mpanetta Uh, I found the issue, for some reason it started the master instead of the minion???
20:06 andrej Thanks yesman ... I'll give it a shot
20:06 ajw0100 joined #salt
20:06 yesman andrej: http://docs.saltstack.com/ref/states/all/salt.states.pkg.html#salt.states.pkg.installed specifically pkgs
20:06 yesman if you specify a list to pkgs, it will override name and install those pacakges listed
20:08 vxitch Gareth: any suggestions on how I can try and resolve this?
20:09 themadcanudist guys, can you define global namespace variables in salt, like pillar[] ?
20:09 themadcanudist from any formula?
20:09 themadcanudist without it being a pillar data
20:09 yesman themadcanudist: all data defined in pillar files will get glommed into a global
20:09 themadcanudist can you create a new varialbe that is "like" pillar
20:10 timoguin mpanetta, oh whoops... that -M in my gist is the install the master. i wasn't paying attention.
20:10 Dewara joined #salt
20:10 yesman you can make it a point that ever file_name.sls has a top level file_name: key at the top and everything falls under it
20:10 yesman vxitch: can you output raw instead? http://docs.saltstack.com/ref/output/all/index.html
20:11 Dewara Hello all....have a quick question about pillars .....
20:11 yesman themadcanudist, just re-read your question, my answer doesnt address this, sorry, misread it
20:11 Ztyx joined #salt
20:11 vxitch yesman: that's what i was looking for! thanks. how do i use the raw output mod?
20:12 timoguin vxitch, --output=raw
20:12 mpanetta timoguin: Ah!  And I just figured out another issue with my conf, somehow the minion was getting pinted to the wrong paster
20:12 zooz joined #salt
20:12 vxitch timoguin: thanks, just found taht in man
20:12 mpanetta master...
20:12 johtso yesman: adding test: True to the config didn't do what I thought it would do
20:12 Dewara I am setting a management entry group for state trees  using a pillar....now I have a need of using that pillar entry in another pillar for setting passwords
20:13 yesman johtso if your master is running a minion as well, set it in your minion config instead perhaps?
20:13 Dewara Are you able to use information from one pillar into another in that manner ....
20:13 johtso salt commands aren't run in test mode by default
20:13 johtso no minion on my master
20:13 yesman johtso: did you restart the master too?
20:13 johtso yep
20:14 yesman johtso: sorry I messed up that answer, http://docs.saltstack.com/ref/configuration/examples.html
20:14 yesman I searched for test: True and thought it was for the master
20:14 yesman but it must be set on all minions
20:14 joehoyle joined #salt
20:14 yesman look under Module configuration in the minion config near the bottom
20:15 vxitch yesman: here's the raw of show_sls http://hastebin.com/yicahugugu.md
20:16 tr_h joined #salt
20:17 andrej http://pastebin.com/xGb4nGBK seems to work as desired. Is it efficient to write it like this?
20:17 TheRealBill joined #salt
20:18 yesman andrej: that looks pretty good to me, not very repetitive at all
20:18 yesman vxitch I'm afraid I forgot the original reason you wanted to see the raw compiled state
20:18 Mr_N Yet another lesson to never ever ever expose BMC's to the public internet, a new flag you can pass to supermicro BMC's which will allow ANY PASSWORD to work successfully. (and cannot necessarily be disabled in a config file anywhere)
20:18 * Mr_N grumbles a bit.
20:18 ajw0100 joined #salt
20:19 Mr_N Rather; not a new flag, a flag I've newly learned the existence of.
20:19 Eugene Shit, I've known better than to put any sort of management interface on the public internet for YEARS
20:19 Eugene Defense in depth beats security by obscurity any day
20:19 Mr_N I still consider this to be a rather entertaining discovery.  "Null password hashing".
20:20 Mr_N Yup.  That's something I'd _ever_ want.
20:21 hunter MR_N: all of those things (except maybe IBM's) fall under the category of it worked once on a test machine on a devs desk - SHIP IT.
20:22 Mr_N I'd actually be somewhat surprised if dell's IDRAC was this porous.
20:22 Mr_N But I'd also be unsurprised to find that I'm entirely wrong.
20:22 Ztyx joined #salt
20:23 hunter in any event I too follow Eugene's general practice - this is what private unroutable nets are for.
20:24 Mr_N I don't disagree.
20:24 hunter I myself like old cheap switches for this but could be convinced of seperate vlan's on the primary switches - if the primaries were being "upgraded" less often.
20:25 Mr_N Yum Vlans.  A guaranteed way to ensure that I'll spend the better part of a few hours reading documentation before running a single command -.-
20:27 tr_h joined #salt
20:28 quantumriff joined #salt
20:28 icarus joined #salt
20:28 andrej Cool - thanks for your support & feedback yesman
20:29 yesman andrej: no worries man, always good when things can get resolved simply!
20:29 quantumriff I have a couple of questions about the salt reactor.. I have it set to accept any key that starts with 'test-' in the server name right now..  That is pretty slick.  However, I would like to run some scripts just once on a newly provisioned machine. (setup authentication)
20:30 quantumriff is there an event state that i could monitor for the key being accepted?
20:32 quantumriff is that just event 'salt/key' with 'act': 'accept' and 'result': 'true'?
20:37 Mr_N What.
20:38 Mr_N [ERROR   ] An un-handled exception was caught by salt's global exception handler:
20:38 Mr_N This'll be a fun afternoon -.-
20:39 Mr_N OH GOOD, the minion I updated to 2014.1
20:40 quantumriff secondly, in reactor, I see lots of examples of running cmd's, but can I copy files down to the server in the reactor sls file?
20:40 alunduil joined #salt
20:44 helderco joined #salt
20:44 rgbkrk joined #salt
20:45 cewood joined #salt
20:47 yesman if anyone's still following along, looks like the culprit is in _preflight_check which is only run with pkg.installed, not pkg.install
20:47 Eugene Brainfart. What's the salt invocation I want to copy a given file from master to minion? Just once via CLI(authorized_keys update)
20:47 mikkn joined #salt
20:47 Eugene Oh heh, there's salt-cp
20:48 defunctzombie_zz joined #salt
20:48 Mr_N I was going to suggest some contrived file.managed thing
20:48 Mr_N that's probably way saner.
20:48 kevinbrolly joined #salt
20:48 Gordonz joined #salt
20:48 bejer joined #salt
20:48 Mr_N So I'm probably way late to the party, but apparently 2014.1 broke everything?
20:49 Eugene Yup, didn't even involve a salt://; worked straight out of my homedir.
20:49 Mr_N And on the off chance this time of asking is any more useful; anyone here got win_repo working even a little?
20:49 Mr_N I can't even get the example in the docs to work.
20:50 quantumriff Mr_N: I found my master updated itself the other day, and I had some weird errors, CPU utilization through the roof, etc.. I had to roll it back.
20:51 Mr_N https://github.com/saltstack/salt/issues/10732 is what I'm following.
20:51 Mr_N I relied pretty heavily on multimastering, so.... things are quite broke.  Luckily I only pushed to my test minion.
20:51 mpanetta timoguin: Thank you again for the help.  I think I have it all working now.
20:51 timoguin mpanetta, good to hear
20:52 jtang1 joined #salt
20:52 mpanetta timoguin: Is there a link to the options one can pass to the bootstrap with script_args?
20:54 timoguin mpanetta, there doesn't seem to be. the bootstrap script itself doesn't seem to have that documented aside from what's in the README
20:54 mpanetta Hmm ok.
20:55 Ryan_Lane joined #salt
20:55 Mr_N I can work around the multimaster thing, I haven't been able to get a single pkg.install to work to a windows system :/
20:55 Mr_N which is a pain.
20:55 timoguin mpanetta, https://github.com/saltstack/salt-bootstrap/blob/develop/bootstrap-salt.sh#L179
20:55 timoguin that might be the best place to look
20:56 mpanetta timoguin: Thank you very much
20:56 lzhang joined #salt
20:56 mpanetta I am bookmarking this.
20:59 quantumriff or besides the reactor.. is there any other good way to do some initial setup on machines that only needs to be done once?
20:59 ndrei joined #salt
20:59 kballou joined #salt
21:00 Mr_N Oh hell... quantumriff ; how'd you roll back? Unless I'm insane, it seems like they nuked the old versions in yum
21:00 UtahDave left #salt
21:00 quantumriff I ended up uninstalling the RPM's from yum, and used the bootstrap scripts
21:00 Mr_N I may have a copy in my spacewalk
21:00 Mr_N Ow.  That's... painful.
21:01 Mr_N Any reasoning for why they nuked the old packages you know of?
21:01 quantumriff wget -O install_salt.sh http://bootstrap.saltstack.org; sh install_salt.sh -P -M git v0.17.5
21:01 Mr_N Yeah; I just, when I can, try to keep everything package-managed
21:01 quantumriff my problem was that machine was still running scientific linux, a distro the script doesn't recognize
21:01 quantumriff I had to change the system to say its centos
21:01 Mr_N Heh.  We run SL as well.
21:02 quantumriff Mr_N: I wish I could say I did it nicely, but that was our only SL 6.x box, the rest are OEL or CentOS, so I just migrated it to CentOS..
21:03 Mr_N We have in the hundreds of SL boxes.
21:03 Mr_N ><
21:04 mgw joined #salt
21:04 quantumriff Mr_N.. I was sleepy, and about 4 beers in late last night.. but it shouldn't be too hard to update the script to also look for SL.. just submit your patch back on github when your done.. save the next guy :)
21:05 Mr_N Oh; I'm sure you're right.  I may even do this.  But as I said, I just tend to like to do things contiguously through yum if possible.
21:05 vxitch yesman: i wanted to see the raw output because that state is caused an unmatched { error from python, which makes no sense
21:07 ipalreadytaken joined #salt
21:07 vejdmn joined #salt
21:10 yesman vxitch: gotcha, what does the whole error look like? did you figure it out?
21:10 rgbkrk joined #salt
21:12 GradysGhost Is the search feature on docs.saltstack.com down atm?\
21:12 schimmy1 joined #salt
21:14 vxitch didnt figure it out yet, heres the error
21:14 vu_ joined #salt
21:14 vxitch NO_STORE don't store history (fc -l) command
21:14 vxitch http://hastebin.com/watolanasa.rb
21:15 vxitch ignore that NO_STORE, hit paste by accident
21:18 yesman vxitch: can i see your sls file?
21:18 yesman in full
21:18 vxitch sure, h.o
21:18 vu_ joined #salt
21:18 yesman what you are seeing is not a json parse error, it's a python string formatting error somewhere
21:19 yesman >>> "{ foo".format('oops') Traceback (most recent call last):   File "<stdin>", line 1, in <module> ValueError: unmatched '{' in format
21:20 vxitch http://hastebin.com/lolexeluli.sm
21:20 vxitch yesman: yeah, i got that. i checked the file referenced and the entire method the line that throws the error resides in is clean
21:21 vxitch so it's somewhere else. and i'm trying to find someone with salt dev expertise to take a crack at it before i dive into the snake pit
21:21 yesman hmmm, can you put in a breakpoint and take a look at the args and kwargs there?
21:21 vxitch how do i do that?
21:21 ajw0100 joined #salt
21:21 lzhang joined #salt
21:21 * vxitch is not a python dev, but likes it
21:22 yesman I'd put in an import pdb; pdb.set_trace() in state.py line 1370, right above ret =
21:22 yesman what platform are you on?
21:22 vxitch RHEL 6.5
21:22 Mr_N yesman: I feel somewhat dirty by comparison to your strategy, I was about to suggest printf debugging -.-
21:22 Mr_N s/printf/print
21:22 DaveQB joined #salt
21:22 pydanny joined #salt
21:23 yesman haha noooooo, pdb is the place to be
21:23 Mr_N I've never been able to integrate it really tightly into my coding workflow, TBH
21:23 jtang1 joined #salt
21:23 Mr_N despite liking a lot of what it offers.
21:23 yesman I used to just print stuff all day too, but one day I just sat down and decided to give it a try
21:23 yesman it's life changing
21:24 Mr_N I've used it once or twice; back in my masters when I was working on multi thousand line python
21:24 yesman vxitch: your salt install should be at /usr/lib/<python_version>/site-packages/salt
21:24 Mr_N but since then I just really haven't needed that size of a hammer ><
21:24 faldridge joined #salt
21:24 yesman Mr_n I find it most useful when there's metaprogramming going on, like in this codebase
21:25 yesman it's like entering bullettime and you get to walk around the code
21:25 Mr_N Fair-nuff.  My complaint was always a rather weak one; mostly that if I wanted a dedicated debugger, I want it to behave like gdb.
21:25 yesman vxitch, let me know if they are indeed there
21:25 yesman vxitch I'll walk you through how I'd debug this
21:26 vxitch yesman: thanks, yes
21:26 vxitch i have a meeting right now, should only be 20-30 mins, ill be back though
21:26 vxitch will you be around?
21:26 yesman cool, shoot me an email, philip@readability.com
21:26 vxitch will do, thanks :)
21:26 BrendanGilmore joined #salt
21:27 gnugnu8 joined #salt
21:29 HeadAIX joined #salt
21:32 philip joined #salt
21:32 jared_ joined #salt
21:32 johtso joined #salt
21:33 joehoyle joined #salt
21:35 defunctzombie joined #salt
21:35 Guest8810 i just updated my minion to 2014.1.0 and im running with file_client: local. When i run highstate I am no longer getting the summary, has anyone else experienced this?
21:35 helderco joined #salt
21:37 Mr_N Guest8810: 2014.1 broke _a lot_.
21:37 Mr_N including, in my case, multimastering.  A rollback was necessary.
21:38 philip I decided to forge on instead of rolling back, pull requests incoming :-P
21:39 Guest8810 sigh...
21:40 Guest8810 i was thinking of pinning our version, are the old deb files hosted anywhere that can be easily installed with apt?
21:41 Mr_N At least for yum; it appears they pulled the old versions off, which made my life even more fun.
21:41 Guest8810 yeah im right there with you at the moment
21:43 saurabhs joined #salt
21:53 druonysus joined #salt
21:53 druonysus joined #salt
21:59 quantumriff Mr_N: it sure would be nice to keep the last revision..
21:59 philip for anyone on yum based distros, this one might be biting you if you are using epel or another extra repo that isn't enabled by default https://github.com/saltstack/salt/issues/10944
21:59 rojem joined #salt
21:59 quantumriff in my base i have a '*' that matches everything.. how do I exclude some minions named 'test-*'?
22:00 lzhang joined #salt
22:00 johtso joined #salt
22:00 vuvu joined #salt
22:00 vu_ joined #salt
22:02 philip quantumriff http://docs.saltstack.com/topics/targeting/compound.html
22:02 philip compound matchers support logical and,or operators
22:03 philip '* and not test-*"
22:03 jeffro joined #salt
22:03 philip make sure you specify it's a compound matcher and not grain in your topfile
22:04 forrest philip, quantumriff as a heads up, the top file documentation contains details on how to match that way now: http://docs.saltstack.com/ref/states/top.html#other-ways-of-targeting-minions
22:04 forrest I added it a while back
22:04 quantumriff thanks
22:04 quantumriff both of you..
22:04 joehoyle joined #salt
22:05 philip thanks forrest, just a heads up I was the guy earlier with the yum package problem, figured it out
22:05 forrest philip, which one?
22:05 forrest sorry there were a LOT of problems this morning.
22:05 brian joined #salt
22:06 brian Hey guys
22:06 philip forrest no worries, that pkg.installed was behaving differently than pkg.install documented here https://github.com/saltstack/salt/issues/10944
22:06 brian I'm trying to make state.highstate run on some other servers when it runs on a specific server
22:06 forrest philip, oh right right
22:06 brian But when I try to make it target a list of servers to run the command, it doesn't do anything
22:06 forrest philip, very cool
22:06 brian It works fine if I just specify one, but i need to trigger updates on two servers
22:06 brian http://pastebin.com/qXJchTNr
22:07 brian I have reactor configuration in place, just need to figure out correct syntax for "tgt
22:07 philip brian maybe your highstate calls are trying to be run while the other one isnt done
22:07 patrek joined #salt
22:07 philip which would cause an error as a highstate is already underway
22:07 brian There are no errors occurring
22:07 philip anything in your master log when this happens? I'd make your trigger do something else, like just a simple echo, to see if it's firing
22:08 brian It is firing
22:08 brian It just doesn't work when I try to target a list of servers
22:08 brian If I set it to one server it works fine
22:08 brian If you look at the if statement, you'll see that there is no way it will fire twice
22:08 brian Because it doesn't match the redis rule, and learner-load,, which is another target, is excluded in the if statement logic
22:09 brian Basically any server that starts with learnerator-load should trigger updates on redis and learnerator-load
22:09 brian if state.highstate is called
22:09 brian It's a syntax issue with my "tgt" there is no example of targeting a list of servers so I'm clueless on how to do it
22:10 forrest brian, did you already try - tgt: 'redis, learnerator-load'
22:10 brian yes
22:10 forrest what happened there?
22:10 philip ah gotcha, can that be a compound matcher instead? like 'redis and learnerator-load' with a type of compound?
22:11 brian Why? That complicates configuration, this is very simple config
22:12 brian forrest: nothing happens
22:12 brian I run salt-call state.highstate on learnerator-load-3 for example
22:12 brian And it doesn't trigger state.highstate to be ran on redis, and learnerator-load, it doesn't do anything
22:12 brian No log output either...
22:13 forrest brian, hmm I thought the matching logic was the same as a normal highstate run, I don't have any instances here, can you try a few things with just salt targeting for salt 'redis, learner-load' test.ping
22:13 brian That won't work
22:13 forrest brian, and see if that returns whether the list is incorrectly formed
22:13 brian You have to specify list type
22:13 forrest yea I know, I'm talking about variants thogh
22:13 forrest *though
22:13 forrest to get the targeting itself figured out
22:13 brian [root@redis ~]# salt -L 'redis, learnerator-load' test.ping
22:13 brian I know how it works
22:14 forrest ok
22:14 brian Well, it doesn't work with that space
22:14 brian But I don't even have that in my config file
22:14 philip seems like 6 of one half dozen of the other for compound vs list, it'd be nice to know _if_ that works to see wether this is a matcher bug on list in reactor specifically
22:15 brian Hm, I do have a space in my config, however....
22:15 brian It would at least target the minion "redis"
22:15 brian And it does not work
22:15 yomilk joined #salt
22:16 philip tgit
22:16 philip oops, disregard cat on keyboard
22:16 brian ok so how do i workaround this
22:17 brian Because it's not gonna work obviously
22:17 philip so you said with a single server match in the reactor file, it does fire
22:17 philip correct?
22:17 brian yes it works great with a single server, let me confirm
22:18 forrest brian, as dumb as it sounds what about this: https://gist.github.com/gravyboat/9356977
22:18 brian ok it actually doesn't work at all now
22:18 brian I don't know what's wrong with it at all
22:18 brian I just want this to work
22:19 philip ok so what was the original that _was_ working, we can work up from there
22:19 brian That gave an error forrest
22:19 brian I have no idea
22:19 forrest brian, hmm ok, let me see if we can find what expr_form supports
22:19 jperras I've got a custom state that I'm writing, and want to get debug/info logs to print when I highstate (or just state.sls) the remote minion. Is there anything I need to enable/configure for this? Adding `-l debug` to the salt <host> state.sls <state> doesn't yield any of my log messages that should be invoked.
22:20 brian why don't I get any logging
22:20 brian this is so annoying
22:20 forrest jperras, did you turn on debug logging on the minion?
22:20 philip brian what's your logging configured to?
22:20 philip in your master and minion configs
22:20 jperras forrest: will that send back debugging information to the master?
22:20 jperras or only have it present on the minion?
22:20 forrest jperras, no, it would log the data on the minion
22:20 * jperras nods
22:21 jperras okay, that should be good enough
22:21 renoirb joined #salt
22:21 brian it just says "Gathering reactors" in the log
22:21 brian Suggesting that the reactor isn't doing anything
22:22 jperras I was wondering if there was some way (similar to the mine) that logs from the minions could be collected on the master
22:22 jperras guess I could always go with syslog-ng to capture the output & aggregate as well
22:22 eeffff joined #salt
22:22 yano joined #salt
22:22 forrest jperras, hmm, I am not aware of that. Maybe remote logging with syslog
22:23 forrest jperras, yea exactly
22:23 jperras cool, thanks a bunch
22:23 forrest basepi, are you around?
22:23 forrest jperras, np
22:23 brian I really need to get this working :(
22:23 jperras figured I'd ask first, since most of the time when I wonder if salt does something, it does ;)
22:23 brian It's not even that complicated of a configuration
22:23 basepi forrest: definitely not around.
22:23 basepi ;)
22:23 snuffeluffegus joined #salt
22:23 forrest basepi, do you remember off hand how to target multiple machines with the reactor for a highstate call?
22:24 forrest http://docs.saltstack.com/topics/reactor/#mapping-events-to-reactor-sls-files
22:24 philip brian if it worked at some point, it's likely not this configuration but something else that's causing your reactor states not to fire
22:24 forrest similar to that 'simple reactor sls'
22:24 philip especially if it's not working with a single machine target
22:24 forrest how do you do two tgts?
22:24 bhosmer joined #salt
22:24 brian well how do I do one target, because at this point I'm confused about how to do anything with this
22:24 brian I get no errors
22:24 forrest one target is like the example I linked
22:24 brian I mean, how am I supposed to know what the problem is?
22:25 philip brian up your log level to debug on both ends and watch the output
22:25 philip (minion and master)
22:25 brian salt stack works great when it works
22:25 brian But when it doesn't work, I get nothing in the debug log that is useful
22:26 brian I already ran debug, it gave me nothing useful at all
22:26 philip brian also, this is open source dude, and we are here to help, you gotta get your hands dirty sometimes
22:26 brian It just said "gathering reactors"
22:26 basepi forrest: it's globbing by default, so you can use stars.  But if you look further down, you can also define `expr_form` to define alternate matchers, such as grain or compound.
22:26 brian No one has told me how to troubleshoot issues with it though
22:26 philip brian: are you version controlled? could you roll back your commits to a working example?
22:26 forrest basepi, right but what about in the case where I have two very specific machine names I want to match, and I don't want to use a grain or a compound matcher?
22:26 forrest basepi, I actually just want 2 unique targets.
22:26 basepi use list
22:27 bemehow joined #salt
22:27 basepi `expr_form: list`
22:27 forrest http://pastebin.com/qXJchTNr
22:27 basepi and then `tgt: minion1,minion2`
22:27 forrest ok
22:27 forrest so no []
22:27 forrest brian can you give that a shot?
22:27 forrest I feel like you did already
22:27 basepi no, no brackets
22:27 forrest cool
22:27 forrest thanks
22:28 eeffff I have a few cmd.run states that use an 'unless' check to ensure that files do *not* exist. after the 2014.1 update, my logs are showing '[ERROR   ] Command '[ -f /foo/bar/lol.py ]' failed with return code: 1' on highstate -- am I doing this wrong?
22:28 jeddi joined #salt
22:28 brian Does salt-call not trigger reactor?
22:28 brian Please help what am I doing wrong
22:29 brian Actually, this should trigger when the minion starts
22:29 brian :(
22:29 philip eeffff what does you state look like
22:29 brian one second i will paste it
22:30 vu_ joined #salt
22:30 vuvu joined #salt
22:33 eeffff philip: http://pastebin.com/4paFmF6C
22:34 eeffff (I know I should be using the new archive state, but for sake of example…)
22:36 philip eeffff that command will return 1 if the file doesnt exist, which will likely be considered an error condition by the cmd runner
22:37 ndrei joined #salt
22:37 philip cd /tmp/ && touch present && [ -f present ]; echo $? && [ -f nope ]; echo $?
22:37 philip returns 0 1 for me for return codes
22:38 philip eefffff but im not sure how `unless` treats exit codes
22:38 eeffff philip that's exactly what's happening. prior to 2014, there were not errors being dropped in the logs.
22:39 millz0r joined #salt
22:39 eeffff the state still runs as expected, by the way
22:39 dyim joined #salt
22:40 philip eeffff: in states/cmd.py, def_run_check is defined and the comment says that unless succeeds if the result is 0
22:40 vuvu joined #salt
22:40 vu_ joined #salt
22:40 eeffff ah!
22:40 philip so I think unless _has_ to return 0 to not be considered an error condition
22:40 vuvu joined #salt
22:40 vu_ joined #salt
22:40 eeffff interesting. thanks for your help!
22:41 philip eeffff no problem
22:41 philip eeffff the line in question (I believe) https://github.com/saltstack/salt/blob/develop/salt/states/cmd.py#L258
22:42 yano joined #salt
22:42 philip brian have you managed to roll back to a working trigger?
22:45 dansalmo joined #salt
22:46 jperras I've noticed that when I call a state.sls <state>, the minions continue to use the cached versions of the state (even though the state contains changes on the master), while state.highstate does not
22:46 dpac|away joined #salt
22:47 KyleG joined #salt
22:47 KyleG joined #salt
22:47 philip jperras calling it via salt-call or salt
22:47 jperras via salt, on the master
22:47 ndrei joined #salt
22:47 jperras while tailing the minion logs, I see that it continues to use the previous version of the state, until I highstate
22:48 Linz_ joined #salt
22:48 jperras at which point it uses the latest version that existed when I highstate'ed
22:48 forrest jperras, correct, because it hasn't synced yet
22:49 forrest jperras, try http://docs.saltstack.com/ref/modules/all/salt.modules.saltutil.html#salt.modules.saltutil.sync_all
22:49 jperras ah, I was trying state.clear_cache
22:49 forrest I THINK that syncs states, I'm trying to remember..
22:49 forrest maybe it's only dynamic modules, hmm
22:49 forrest jperras, http://docs.saltstack.com/ref/modules/all/salt.modules.saltutil.html#salt.modules.saltutil.sync_states
22:49 forrest there we go
22:49 Gareth vxitch: did you figure our your iptables issue?
22:49 forrest jperras, but again, that's _states
22:49 jperras gotcha
22:50 forrest jperras, let me see if I can find it. So you highstate, and it doesn't get the latest, until another highstate?
22:50 jperras right – this is for changes in a custom _state/_module
22:50 forrest oh excellent then, that should resolve it
22:50 jperras awesome
22:51 forrest I thought you were saying normal states, and was a bit confused :P
22:51 philip gareth I think he had to step out
22:51 vxitch Gareth: no :(
22:51 vxitch i just got back
22:52 jperras forrest: thanks, I believe that worked!
22:53 forrest jperras, great!
22:53 Gareth vxitch: what does your state look like?  I ran a simply flush one and it ran fine.
22:53 vxitch http://hastebin.com/lolexeluli.sm
22:55 bhosmer joined #salt
22:55 vuvu joined #salt
22:55 vu_ joined #salt
22:56 Gareth vxitch: weird.  runs fine.  have you tried taking everything out instead of the flush section?  narrow it down to see which part is failing.
22:58 rojem joined #salt
22:58 smcquay joined #salt
22:58 gnugnu8_ joined #salt
22:58 ccase joined #salt
22:58 vxitch no i havent tried that yet
23:00 brian philip: I thought I changed it back to the working configuration, however it's still not working
23:00 brian Basically when the minion starts, I want to run state.highstate on other servers
23:01 kermit joined #salt
23:04 KyleG joined #salt
23:04 KyleG joined #salt
23:05 vxitch Gareth: i've removed everything except the flush, and the state still fails
23:06 Gareth same error?
23:06 vxitch yes
23:06 Gareth can you repaste the state
23:06 vxitch the one im running now? the stripped down one?
23:06 vxitch http://hastebin.com/migekufaxu.sm
23:06 vxitch thats my entire iptables/init.sls
23:08 Gareth vxitch: very strange.
23:08 vxitch thats all anyone ever says about me
23:09 Gareth hah
23:09 vuvu joined #salt
23:09 vu_ joined #salt
23:10 krow joined #salt
23:10 snuffeluffegus joined #salt
23:11 Gareth vxitch: how are you running it?  I'm testing it with salt 'host' state.sls state.sls
23:12 vxitch yup thats it. `salt -G 'role:master' state.sls iptables`
23:13 Gareth all machines running 2014.1?
23:13 vxitch mmhm
23:13 themadcanudist joined #salt
23:17 Gareth vxitch: same error right?
23:17 vxitch exactly the same :(
23:18 dansalmo joined #salt
23:19 dave_den joined #salt
23:19 kaictl joined #salt
23:21 dansalmo how do can I tell if salt-master is running and how do I restart it if it is?
23:21 vxitch ping?
23:21 dansalmo When I try "salt-master -l debug"
23:21 vxitch salt 'salt.domain' test.ping
23:22 dansalmo It says : The ports are not available to bind
23:22 vxitch if youre on the local machine, do `service salt-master status`
23:22 dansalmo salt-master start/running, process 3864
23:23 lzhang joined #salt
23:23 vxitch check that youre a privileged user when you try to run any salt commands, and that you arent blocking traffic to loopback/127.0.0.1
23:23 jslatts joined #salt
23:23 dansalmo I use sudo
23:23 vxitch if your firewall/iptables default policy is DROP, you must explicitly allow communication to the loopback int
23:23 vxitch what OS
23:24 dansalmo Ubuntu
23:24 vxitch mk
23:25 vxitch check firewall, check that youre elevated, check that salt master and minion are running as services, check that theyre compatible versions (seems the be the common issue this week), and check the logs (/var/log/salt/)
23:26 dansalmo I did "sudo ufw allow salt"
23:26 dansalmo I do not have any minions yet
23:26 themadcanudist left #salt
23:26 vxitch i dont know ufw, im RHEL on iptables
23:26 dansalmo I just did the install and changed the IP address
23:26 vxitch your salt master is also a minion
23:26 dansalmo Ok
23:26 vxitch check the service status of salt-minion
23:26 forrest dansalmo, if you are getting an error regarding the ports being unavailable to bind, that means it is usually already running
23:27 forrest vxitch, the master doesn't have to be a minion unless you install the salt minion
23:27 druonysus joined #salt
23:27 druonysus joined #salt
23:27 dansalmo I have only installed master
23:28 forrest dansalmo, yea you should check out the walkthrough most likely then: http://docs.saltstack.com/topics/tutorials/walkthrough.html
23:28 vxitch forrest: true, but then don't you need a minion to interact with which then sends the commands to the master?
23:28 dansalmo I am not sure if it restarted with the new IP configuration
23:28 forrest vxitch, correct, but he was just asking about the service and confirming it was running. Just didn't want confusion regarding what comes with salt-master
23:30 NV hey forrest - how legitimate is using salt modules in pillar sls files?
23:30 ndrei joined #salt
23:30 dansalmo sudo salt 192.168.1.6 test.ping   ->  No minions matched the target. No command was sent, no jid was assigned.
23:30 forrest NV, I've never done it
23:31 NV have you done it in 2014.1?
23:31 vxitch dansalmo: you need a minion, first off, and second just use '*' instead of the IP for now, to rule out targetting issues
23:31 NV I upgraded from 0.17 to 2014.1 and found salt in jinja is now what appears to be an empty list, not a dict
23:31 vxitch dansalmo: is your salt-minion running?
23:31 forrest NV, I haven't, my machines are still running 0.17.5 at home, haven't had a chance to upgrade yet this week.
23:32 NV mhmm fair enough
23:32 dansalmo No, I was going to install minion on another machine
23:32 NV but it is legitimate to do so? Or is it a "it works, but only coincidentally and was never intended to work"?
23:32 vxitch dansalmo: well then :P that's why you can't get anything
23:33 vxitch salt-minion can also run on the master machine
23:33 dansalmo But I was not sure master was running and the interface was configured
23:33 vxitch it doesn't have to, like it has been said
23:33 vxitch but it can
23:33 vxitch oh ok
23:33 forrest NV, do you have an example?
23:33 vxitch well it seems like it is if you can issue the command and get that response back
23:33 vxitch and if the service is running
23:34 forrest dansalmo, if you go through the walkthrough it will talk about configuring the salt minion and everything
23:34 NV forrest: http://pastie.org/8863394
23:35 forrest NV, Hmm, well none of the docs talk about doing that, honestly I'm not sure. What do you think about using salt modules in pillar files?
23:35 forrest terminalmage ^
23:35 forrest NV, I'd be curious to see what kind of performance hit that creates.
23:37 NV eww, unrelated on 2014.1 getting AttributeError: 'TagObject' object has no attribute 'tree' when trying to highstate a specific machine
23:37 NV in the server log
23:37 terminalmage forrest: ?????
23:38 forrest terminalmage, http://pastie.org/8863394
23:38 forrest NV wants to know if that is ok to do
23:38 NV in pillar
23:38 bhosmer joined #salt
23:38 terminalmage ¯\_(ツ)_/¯
23:39 terminalmage NV: have you tried it?
23:39 NV it worked in 0.17, fails hard in 2014.1
23:39 terminalmage jinja templating is allowed in pillar, it should work
23:39 NV also getting this http://pastie.org/8863400 in 2014.1
23:39 terminalmage NV: ok, throw together an example then and open an issue
23:40 forrest terminalmage, I mean more in a 'is this a good practice' sort of way, as opposed to functional
23:40 forrest and I feel like you have a macro for shoulder shrug guy.
23:40 terminalmage forrest: I do
23:40 terminalmage :D
23:40 forrest heh
23:40 terminalmage NV: please just open issues
23:41 dansalmo forest: thanks, I was following the walk through, but did not go past the master yet.  Will try installing minion
23:41 bemehow joined #salt
23:41 terminalmage I'm working on like 2 or 3 things at a time, and honestly I'm never going to look at something that someone just blurts out in IRC
23:41 terminalmage well, maybe not *never*
23:41 NV mhmm, just sanity checking that im not doing something retarded :P
23:41 terminalmage but, it's not going to be as likely
23:41 NV wasnt sure if it was intended that salt modules shouldn't work in pillar or not
23:41 terminalmage we do want to address issues if there are regressions, of course
23:41 NV mhmm true
23:42 terminalmage NV: no they should definitely work
23:42 terminalmage jinja works in pillar SLS files
23:42 NV alright, that's all i was after - opening an issue :)
23:42 terminalmage and salt execution functions are exposed in the jinja context
23:42 terminalmage make sure to include all tracebacks that you have observed, that definitely helps narrow things down
23:43 kintel_ joined #salt
23:44 terminalmage NV: also, in case you are not aware, if you're posting shell examples and tracebacks, put a triple-backtick line before and after, and it'll nicely format them with monospaced fonts in the github interface
23:44 NV yup :)
23:44 terminalmage cool :)
23:45 n8n joined #salt
23:45 budrose joined #salt
23:46 NV https://github.com/saltstack/salt/issues/10954 :D
23:46 NV now to work out why my gitfs is being retarded
23:47 goodwill terminalmage: hey big papa
23:47 terminalmage goodwill: sup
23:47 goodwill terminalmage: just saying hi
23:48 terminalmage NV: I made some changes to gitfs that made it into 2014.1, so it is possible that I introduced an issue
23:48 NV mhmm
23:48 terminalmage NV: be aware that the current state of the gitfs module in develop is quite different though
23:49 terminalmage NV: I'm wondering if you might be doing a "set salt" somewhere else in your jinja?
23:50 terminalmage because the "salt" variable in jinja should be a dict
23:50 NV that's what I thought...
23:50 NV I've grepped for 'salt' across my entire pillar
23:50 NV and nada
23:51 NV i do have a pillar that has 'salt:' in it at the top level
23:51 NV but that's yaml not jinja
23:51 NV and jinja happens before the yaml is parsed so i can't see it being that
23:51 NV (plus that contains a dict structure anyway...)
23:52 terminalmage NV: yeah... that should result in pillar['salt']
23:52 terminalmage should be a key
23:52 terminalmage and should not affect the salt variable in the jinja context
23:52 NV yeah - but even then pillar isn't populated yet at that stage is it?
23:54 NV updated with versions report btw
23:54 NV forgot to initially
23:54 terminalmage I don't think so. It's possible that pillar data already exists when new pillar data is requested, but I don't think existing pillar variables affect compilation of fresh pillar data
23:54 terminalmage NV: thanks! that should be standard information, I think :)
23:54 terminalmage we're trying to get GitHub to support templates
23:54 terminalmage like google code
23:55 terminalmage so we can pre-fill the issue box and ask for certain information
23:55 terminalmage because many people don't include it and we end up wasting time asking them
23:55 NV mhmm

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