Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-09-07

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

All times shown according to UTC.

Time Nick Message
00:01 littleidea joined #salt
00:18 mwillhite joined #salt
00:19 baniir joined #salt
00:27 mgw joined #salt
00:40 logix812 joined #salt
00:48 jaequery joined #salt
01:00 vmdsch joined #salt
01:00 dthom91 joined #salt
01:02 oz_akan_ joined #salt
01:03 littleidea joined #salt
01:22 Katafalkas joined #salt
01:52 jaequery joined #salt
01:52 lemao joined #salt
01:58 littleidea joined #salt
02:03 littleidea joined #salt
02:04 StDiluted joined #salt
02:22 Katafalkas joined #salt
02:40 oz_akan_ joined #salt
02:58 rich42 left #salt
02:59 dthom91 joined #salt
03:00 forrest joined #salt
03:15 anb__ joined #salt
03:23 Katafalkas joined #salt
03:24 jacksontj joined #salt
03:34 racooper joined #salt
03:34 cewood joined #salt
03:54 tyler-baker joined #salt
03:54 tyler-baker joined #salt
03:55 wolfpack_ joined #salt
03:55 jaequery joined #salt
03:59 sep_ joined #salt
04:02 forrest Does anyone have a good quick way to review the docs after you build them so you can confirm formatting: http://docs.saltstack.com/topics/hacking.html#editing-and-previewing-the-documentation
04:02 forrest I mean I can spin up a web server or something
04:04 benspaulding You should be able to simply open the index file in your browser.
04:05 lineman60 joined #salt
04:05 forrest this isn't a local build
04:05 forrest it's out on a VM in a subdirectory
04:09 cedwards forrest: use the python SimpleHTTPServer module maybe?
04:09 forrest yea that's an option as well, I ended up just tarring it up and copying it to my local box
04:20 forrest I'm surprised no one has written some states to spin everything up for a dev environment
04:23 Katafalkas joined #salt
04:24 ggoZ joined #salt
04:35 cewood joined #salt
04:49 Ryan_Lane joined #salt
04:51 vbabiy I can I use use pillar data from one sls file in another?
04:54 forrest hey s0undt3ch, did you ever resolve https://github.com/saltstack/salt-bootstrap/issues/69 ?
04:57 StDiluted joined #salt
04:57 cce_ joined #salt
04:57 djinni`_ joined #salt
04:58 jcristau_ joined #salt
04:59 gadams joined #salt
04:59 gadams joined #salt
04:59 shadowsun joined #salt
05:00 Boohbah_ joined #salt
05:00 Ahlee joined #salt
05:00 cowmix joined #salt
05:00 Valdo joined #salt
05:01 Daviey joined #salt
05:01 defunctzombie joined #salt
05:05 NV joined #salt
05:16 mgw joined #salt
05:23 Katafalkas joined #salt
05:58 sparc_ joined #salt
06:07 dthom91 joined #salt
06:14 gpkvt joined #salt
06:17 gpkvt Hi. Got a problem with my jinja template. I'm reading the IP-Address from eth0 with bind: {{ salt['grains.get']('ip_interfaces:eth0') }} and put it into my config file like this bindnetaddr: {{ bind }}. It almost works, the only problem is that the IP is not set as 10.0.0.1 but ['10.0.0.1']. How do I avoid those brackets and single quotes?
06:20 joehh can you do {{ bind[0] }}
06:20 joehh ?
06:24 Katafalkas joined #salt
06:24 gpkvt gna, thanks.
06:24 gpkvt works fine, i go to bed now
06:24 Furao joined #salt
06:46 aantony joined #salt
06:56 neilf joined #salt
06:57 ninkotech joined #salt
06:58 UtahDave joined #salt
07:06 faldridge joined #salt
07:06 UtahDave joined #salt
07:07 crazysim joined #salt
07:13 littleidea joined #salt
07:16 anb__ joined #salt
07:18 ninkotech joined #salt
07:24 Katafalkas joined #salt
07:28 balboah_ joined #salt
07:28 gildegoma joined #salt
07:29 \ask UtahDave: I didn't follow-up earlier so I just wanted to say thanks -- I got our stuff to work nicely with watch_in.
07:29 zach joined #salt
07:29 UtahDave \ask: Ah, good to hear!
07:40 Vivek joined #salt
07:43 zach joined #salt
07:58 zach joined #salt
08:08 tuxIO joined #salt
08:25 Katafalkas joined #salt
08:32 zach joined #salt
08:33 bhosmer joined #salt
08:34 sfello joined #salt
08:55 jY joined #salt
09:11 hjubal joined #salt
09:11 hjubal joined #salt
09:25 Katafalkas joined #salt
09:34 sparc_ joined #salt
09:56 MrTango joined #salt
10:04 linjan joined #salt
10:16 claudep joined #salt
10:17 claudep hi, is it not possible to mix file.sed and file.append for the same file?
10:17 bhosmer joined #salt
10:19 claudep http://dpaste.com/1369278/
10:20 Eugene You have multiple file.seds is the problem
10:21 Eugene Have you considered file.managed instead?
10:21 claudep strange, because i also have multiple file.seds elsewhere and got no errors
10:22 Eugene That's my readin of the error. I'ved never used .sed or .append, because .managed works great ;-)
10:22 claudep but with file.managed, you "freeze" the entire file
10:22 Eugene That's one way to look at it
10:23 Eugene If you need to change a setting on a per-machine basis, put that into Pillar, then modify your state to set it appropriately ;-)
10:24 claudep my concern is rather that when you update the package, you might easily miss general additions/removals (or even comment updates)
10:25 claudep of course, when I can separate modifications in a different file (à la conf.d), it's always better
10:25 Katafalkas joined #salt
10:28 Eugene Depends upon your packaging system.
10:28 Eugene rpm will create new configs(If you've updated the running one) as .rpmnew most of the time
10:31 anb__ joined #salt
10:48 Katafalkas joined #salt
10:53 anb__ Salt cannot preserve link and file mode when recursive copy a directory. Am I missed some options?
10:57 shiin left #salt
11:07 s0undt3ch faulkner: bah, nope, not yet
11:37 sfello joined #salt
11:38 isomorphic joined #salt
11:38 felixhummel joined #salt
11:50 lemao joined #salt
11:53 dthom91 joined #salt
11:57 gildegoma joined #salt
12:00 Katafalkas joined #salt
12:01 tuxIO joined #salt
12:04 scooby2 joined #salt
12:05 p3rror joined #salt
12:06 claudep joined #salt
12:12 zooz joined #salt
12:47 sgviking joined #salt
13:04 Drexx joined #salt
13:10 felixhummel joined #salt
13:24 sfello joined #salt
13:30 lemao_ joined #salt
13:31 Psi-Jack joined #salt
13:33 inver joined #salt
13:40 cewood joined #salt
13:51 efixit joined #salt
13:52 Psi-Jack joined #salt
13:56 premera joined #salt
13:59 tuxIO_ joined #salt
14:03 jslatts joined #salt
14:07 premera does anybody have a users (as in linux account user) management example where I define users and their properties globally but pick and choose what users are materialized on what minion ?
14:08 zzzirk joined #salt
14:09 whit joined #salt
14:12 isomorphic joined #salt
14:14 kleinishere joined #salt
14:20 copelco premera: we have a basic setup like that in our project template, https://github.com/caktus/django-project-template/blob/master/conf/pillar/top.sls#L4
14:20 premera copelco: thanks! will check it out
14:21 copelco the state that does most of the work is here: https://github.com/caktus/margarita/blob/master/users/devs.sls
14:24 tuxIO joined #salt
14:25 zooz joined #salt
14:25 premera copelco: from what I understand in devs.sls you define all available users, how then do you declare which of those users go into production and which go into staging minions ?
14:26 isomorphic joined #salt
14:26 xinkeT joined #salt
14:32 copelco in our case we setup the same users in both environments, but you could adjust the pillar to have a staging.devs and remove the devs under the '*' target
14:32 copelco like we do with env https://github.com/caktus/django-project-template/blob/master/conf/pillar/top.sls#L6
14:34 premera yeah, but what I am looking for is to define all users in one place but pick and choose what user goes where, e.gI have bob, alice and erik, I want to put bob and alice on production and bob end erik on staging without repeating bob definition in both environments
14:36 premera for production I would like to say in production.env pillar: allowed_users: [bob, alice] and in staging.env allowed_users: [bob, erik]
14:36 anteaya joined #salt
14:37 premera bob, alice, erik would be probbaly define in pillar base.devs but not materialized unless "allowed_users" directive is used in mininon
14:37 copelco ah i see
14:38 copelco sounds doable
14:38 copelco you could just check to see if the user name is allowed_users while looping through the pillar
14:39 premera ok, will try that, was hoping for more declarative approach, thanks
14:39 copelco yeah, there probably is :)
14:40 premera :-)
14:43 bhosmer joined #salt
14:48 joehh ~.
15:00 linjan joined #salt
15:06 ckao joined #salt
15:12 littleidea joined #salt
15:14 robertkeizer joined #salt
15:17 TheCodeAssassin joined #salt
15:20 tuxIO_ joined #salt
15:32 mgw joined #salt
15:36 claudep joined #salt
15:37 msheiny joined #salt
15:42 claudep left #salt
15:52 marcinkuzminski joined #salt
15:52 mgw joined #salt
15:55 Lue_4911 joined #salt
15:55 mwillhite joined #salt
15:59 tuxIO joined #salt
16:12 Katafalkas joined #salt
16:13 Katafalkas joined #salt
16:20 Lue_4911 joined #salt
16:23 tuxIO_ joined #salt
16:41 mgw joined #salt
16:42 kenbolton joined #salt
16:42 StDiluted joined #salt
16:49 dthom91 joined #salt
16:50 zzzirk joined #salt
16:52 zzzirk joined #salt
16:53 superflit joined #salt
16:54 jpeach joined #salt
17:01 honestly using mako
17:02 tuxIO joined #salt
17:02 honestly "no module named mako.exception"
17:02 honestly uhhh.
17:02 lineman60 joined #salt
17:02 honestly mako.exceptions*
17:02 honestly do I have to install mako first?
17:03 honestly it appears I would have to
17:10 TheCodeAssassin joined #salt
17:14 honestly how do I check the sls render result?
17:30 honestly okay...
17:31 honestly why does this not retrieve the master value: pillar.get('users_nfs:master','')
17:31 honestly but this does: pillar.get('users_nfs', {}).get('master','')
17:31 qba73 joined #salt
17:32 honestly that's in a mako template, it works from salt-call
17:32 robertkeizer joined #salt
17:34 honestly salt 0.16.3 fwiw
17:44 Katafalkas joined #salt
17:53 Jahkeup_ joined #salt
17:57 Jahkeup joined #salt
18:02 TheCodeAssassin joined #salt
18:08 dthom91 left #salt
18:10 m_george|away joined #salt
18:14 Jahkeup_ joined #salt
18:17 dave_den honestly: in your salt states, 'pillar' is a python dict.
18:18 m_george left #salt
18:18 dave_den when you run salt-call pillar.get, that is using the pillar module and running the get() function
18:19 honestly okay
18:20 honestly so I should just do pillar['users_nfs']['master'] in the sls?
18:20 dave_den honestly: you can either use set users_nfs = pillar.get('users_nfs', {}) and then use users_nfs.get('master', '')
18:20 Katafalkas joined #salt
18:20 dave_den yeah, if you are sure the master key will be in the dict
18:24 gadams joined #salt
18:24 gadams joined #salt
18:24 sfello joined #salt
18:31 honestly yeah
18:32 honestly_ joined #salt
18:41 rjc joined #salt
18:54 lineman60 joined #salt
18:59 forrest joined #salt
19:00 gadams joined #salt
19:00 gadams joined #salt
19:02 Jahkeup joined #salt
19:11 tuxIO joined #salt
19:15 felixhummel joined #salt
19:15 [vaelen] joined #salt
19:20 isomorphic joined #salt
19:21 CheKoLyN joined #salt
19:25 mgw joined #salt
19:29 copelco is there a disadvantage to installing salt-minion through the apt ppa as opposed to the bootstrap script?
19:30 vmdsch joined #salt
19:31 Eugene AFAIK the bootstrap script just adds the ppa if you're on Ubuntu
19:31 forrest copelco, not really, you can pull the latest release from git with the bootstrap where as you can't from the PPA.
19:31 forrest I mean the latest dev release.
19:31 copelco ah ok
19:31 copelco thanks
19:32 isomorphic joined #salt
19:50 p3rror joined #salt
19:55 vmdsch joined #salt
19:56 whit joined #salt
19:59 Katafalkas joined #salt
20:04 waverider joined #salt
20:08 obuisson joined #salt
20:09 obuisson hello all
20:10 obuisson i would like to know if it is possible to launch a VM with salt-cloud via a RESTFul API. I didn't find anything in the salt-api and salt-cloud docs
20:10 waverider left #salt
20:11 gadams joined #salt
20:11 gadams joined #salt
20:13 gadams joined #salt
20:13 gadams joined #salt
20:14 gadams joined #salt
20:14 gadams joined #salt
20:18 gadams joined #salt
20:18 gadams joined #salt
20:20 kspath identify kspath
20:21 zooz joined #salt
20:25 isomorphic joined #salt
20:25 mgw left #salt
20:28 superflit joined #salt
20:35 whit joined #salt
20:45 qba73 left #salt
20:46 mgw joined #salt
20:47 faldridge joined #salt
20:54 matanya joined #salt
20:56 maxleonca joined #salt
20:58 maxleonca Hello, I'm trying to set up a nginx config test on the state so reactor can pick up the response and if there is an error then notify of it, but cannot get the check to work.  I'm adding -Module: nginx.configtest under the nginx: pkg in the init file.
20:58 maxleonca Any idea on why it may be not working? I'm all across on 0.16.3
20:59 kenbolton joined #salt
21:01 littleidea joined #salt
21:21 jslatts joined #salt
21:21 forrest maxleonca, can you post your sls?
21:22 maxleonca sure
21:25 maxleonca nginx:   pkg:     - latest    service:     - running     - enable: True     - watch:       - pkg: nginx       - file: conf-dir       - file: ssl-dir  conf-dir:   file.recurse:     - name: /etc/nginx/conf.d     - source: salt://nginx/conf.d     - require:       - pkg: nginx       - pkg: openssl  ssl-dir:   file.recurse:     - name: /etc/nginx/conf.d/ssl     - source: salt://nginx/ssl     - include_empty: True
21:25 forrest can you post that on gist or pastebin or something please?
21:25 forrest formatting is all borked.
21:26 maxleonca Try this: http://pastebin.com/501e7UNz
21:26 forrest way better thanks
21:26 maxleonca thank you
21:27 forrest so do you usually do an include here?
21:27 forrest because you're saying that you have something else in the init?
21:28 maxleonca well the idea is to get the configtest module there.
21:28 maxleonca then catch the output via reactor.
21:29 maxleonca but the configtest fails right off the bat so....
21:29 forrest but I assume the configtest passes fine if you run it from the command line?
21:29 maxleonca correct
21:30 forrest what about if you do salt 'host' nginx.configtest?
21:30 maxleonca not a problem
21:30 forrest that's od
21:30 forrest *d
21:30 forrest are you always trying to catch the output from call with reactor? What if you just put the config test into that sls, what happens then?
21:30 lineman60 joined #salt
21:31 maxleonca nope, that is the plan, starting with the config standing alone
21:31 maxleonca no other parts are in the equation right now.
21:31 maxleonca what you see is what it is up to now.
21:31 forrest ok
21:32 forrest I wish there was an example of using config test inside an sls
21:32 maxleonca so do I, that's why I'm bugging the community :)
21:32 forrest underneath nginx, on the same indent level as package, but a line below can you try to do - configtest ?
21:33 maxleonca Well is worth the try.
21:33 forrest if we figure it out I'll file an issue on it then add an example via pull request :P
21:35 msheiny joined #salt
21:36 maxleonca no, it didn't fly, setting the config test under module at the same level as pkg or service.
21:37 forrest hmm
21:38 Ryan_Lane joined #salt
21:38 maxleonca hmmm, under service as module: nginx.configtest it ran without issues
21:38 maxleonca right between enalbe and watch
21:38 maxleonca enable, sorry
21:38 forrest cool
21:39 maxleonca I just need to check that indeed it ran the configtest
21:39 maxleonca Thanks for your assistence
21:40 forrest Glad I could provide you with incorrect solution :P
21:40 forrest Here's the relevant code by the way
21:40 forrest https://github.com/saltstack/salt/blob/develop/salt/modules/nginx.py#L43
21:40 forrest for the cmd line version at least.
21:41 forrest did you put nginx.configtest?
21:41 forrest between enable and watch
21:42 maxleonca http://pastebin.com/M0mNqZWu
21:42 forrest ok cool
21:43 maxleonca basically the same, but before watch, which not sure if it actually does what I want it to do.  But is a start and a bit of light on how integrate the module calls on the sls.
21:43 forrest agreed
21:43 forrest if you figure out more about it let me know, there's not a lot of clear documentation on how to integrate those modules into an sls
21:44 tuxIO_ joined #salt
21:44 maxleonca I'll do some debugging and I'll post  the results here.
21:45 forrest awesome thanks
21:45 maxleonca as a matter of fact checking as I type.
21:52 maxleonca OK, the syntax check worked but I don't see the check being executed, so scratch this.
21:54 bhosmer joined #salt
21:57 berto- joined #salt
21:59 tuxIO joined #salt
22:11 matanya joined #salt
22:11 zloidemon joined #salt
22:19 msheiny joined #salt
22:20 hannesstruss joined #salt
22:36 honestly_ I have this sls file: https://gist.github.com/duk3luk3/6480007
22:36 honestly_ can anyone tell me why salt doesn't think it's a proper sls file?
22:37 honestly_ "does not render to a dictionary"
22:40 forrest can you run it through with the debug option honestly? Then we can see what it does/doesn't parse.
22:43 honestly I did run it through the debug option
22:43 honestly "Data failed to compile: SLS nfs-users does not render to a dictionary"
22:43 honestly that's all the info I get
22:44 honestly this is the command I ran: salt-call -l debug state.sls nfs-users test=True
22:44 forrest gotcha
22:44 forrest can you try to comment out the exports item?
22:45 honestly why?
22:45 honestly same error
22:45 forrest I was just trying to see if it was having an issue interpreting the text for some reason
22:47 tuxIO_ joined #salt
22:47 honestly according to http://yaml-online-parser.appspot.com/ it does render to a dictionary...
22:48 yano joined #salt
22:48 honestly my sls has mako in it of course
22:48 honestly but the mako renders fine
22:48 honestly what I pasted is the render output salt-call -l debug prints
22:51 tuxIO joined #salt
22:54 robertkeizer joined #salt
22:58 jslatts joined #salt
23:05 UtahDave joined #salt
23:06 UtahDave left #salt
23:06 UtahDave joined #salt
23:09 honestly :|
23:10 honestly it's the #!mako
23:10 honestly it doesn't get removed automatically
23:10 honestly but if I don't put it in the file, guess what, it won't call the mako renderer
23:10 honestly that is... weird
23:11 forrest ahh ok
23:11 forrest good to know
23:11 honestly I'm on 0.16.3
23:11 forrest that is odd though
23:11 forrest the examples show that should work
23:15 honestly #!mako|yaml
23:15 honestly :|
23:16 maxleonca try with yaml_mako and see if it works
23:16 forrest lol
23:21 honestly mako|yaml works
23:22 forrest maybe because when you only ran it through with mako, it wasn't passing the input to yaml?
23:22 forrest that's the only reason I can think it wouldn't configure it as a dict.
23:23 honestly yeah
23:23 honestly it didn't parse it as yaml
23:39 jeddi joined #salt
23:46 jalbretsen joined #salt
23:54 premera joined #salt

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