Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-10-07

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

All times shown according to UTC.

Time Nick Message
00:06 efixit joined #salt
00:12 mofomikes joined #salt
00:17 CheKoLyN joined #salt
00:19 lemao joined #salt
00:37 zwe joined #salt
00:51 cro joined #salt
00:53 bdf what changed in 0.17 that would cause me to suddenly get this? Failed to open log file, do you have permission to write to /var/log/salt/master
01:07 forrest bdf, nothing as far as I know, are you unable to access the log as the root user?
01:07 Drawsmcgraw joined #salt
01:09 diegows joined #salt
01:10 bdf sorry, should have clarified, I have my normal user in the client_acl
01:10 bdf it works fine as root
01:10 forrest ahh
01:11 __number5__ http://xkcd.com/149 :)
01:25 andrej yawn
01:32 cro joined #salt
01:34 brianhicks joined #salt
01:36 jbunting joined #salt
01:51 sibsibsib_ joined #salt
02:03 mofomikes joined #salt
02:04 mgw joined #salt
02:06 anuvrat joined #salt
02:14 vipul joined #salt
02:22 talso joined #salt
02:27 malinoff joined #salt
02:35 cro joined #salt
02:38 talso joined #salt
02:57 malinoff Hi salt guys
02:57 Drawsmcgraw Anyone using something other than 'pam' for the external auth?
03:01 goodwill Drawsmcgraw: in what sense? like oath?
03:02 Drawsmcgraw I'm just curious what other people use for the external auth. I've seen 'pam' and 'ldap' so far
03:02 Drawsmcgraw The longer version is -> I'm trying to understand how I can get a PKI login to work for Halite. I'm looking at how/which external auth would be best for that.
03:03 Drawsmcgraw goodwill: ping, btw
03:04 Drawsmcgraw Ah. I may have found my answer: https://github.com/saltstack/salt/tree/develop/salt/auth
03:17 robinsmidsrod joined #salt
03:29 Ryan_Lane joined #salt
03:37 MTecknology goodwill: pong (was out hunting for the weekend)
03:37 goodwill MTecknology: delicious human ;)
03:37 goodwill MTecknology: PM allowed?
03:38 MTecknology sure
03:38 Pate_ joined #salt
03:39 mua joined #salt
03:53 jalbretsen joined #salt
04:02 malinoff joined #salt
04:27 cshuman joined #salt
04:28 Pate_ joined #salt
04:31 Oreh joined #salt
04:35 anuvrat joined #salt
04:36 cshuman joined #salt
04:37 cshuman_ joined #salt
04:39 stanlee joined #salt
04:39 stanlee
04:41 cshuman joined #salt
04:52 reinsle joined #salt
04:53 reinsle Hi *
04:54 malinoff Hi
04:54 Eugennerz rawr
05:01 mahul joined #salt
05:01 reinsle Is it possible using pillar.get not only to recieve a value of an top-level pillar element but recieve an value of an sublevel element?
05:04 jcockhren for any of the built-in returners (let's say redis, for example), if I set it in the master file, that is the same as setting it in config file for all of the minions right? Does this mean that any configiration needed for that returner should also be in the master?
05:05 joehh reinsle: I think you can use : to go down levels
05:05 joehh https://salt.readthedocs.org/en/latest/ref/modules/all/salt.modules.pillar.html#salt.modules.pillar.get
05:08 reinsle joehh: but this gives me None for an set Value
05:09 reinsle joehh: as example, i've this structure: http://pastebin.com/zvfC6Rp3
05:09 reinsle i will recieve value2
05:10 joehh will = want to?
05:10 reinsle so i think  pillar.get('toplevel1:key2') should work
05:10 reinsle but it dont
05:10 joehh I'll try
05:11 * reinsle uses the newest (0.17.0) salt version
05:13 reinsle salt '*' pillar.data toplevel1 shows me the data structure
05:14 awanderingvagabo joined #salt
05:19 joehh I'm seeing the same thing
05:19 reinsle joehh: strange?
05:20 joehh I think so
05:20 joehh I'd say it is worth an issue
05:20 mmilano joined #salt
05:20 reinsle should i open a issue?
05:21 joehh I think so - do you know how?
05:22 joehh I would say doesn't seem to work in 0.17.0
05:22 reinsle joehh: jip, its not the first one
05:23 winux joined #salt
05:24 winux hello.. i am using rbenv.installed to install versions of ruby, but running rbenv on the command line after running the state file shows that no specific versions of ruby have been installed. Anyone know what could be going on?
05:25 reinsle https://github.com/saltstack/salt/issues/7645
05:28 reinsle joehh: thanks for your support :-)
05:29 jkleckner joined #salt
05:30 cshuman joined #salt
05:31 jkleckner joined #salt
05:36 MK_FG joined #salt
05:39 middleman_ joined #salt
05:40 middleman_ joined #salt
05:43 nineteen1ightd joined #salt
05:46 cshuman joined #salt
05:50 isomorphic joined #salt
05:53 gildegoma_ joined #salt
06:08 winux exit
06:09 Lue_4911 joined #salt
06:13 Katafalkas joined #salt
06:21 yogi__ joined #salt
06:23 yogi__ upgraded to 0.17 salt stack on ubuntu 12.04 . Is it just me but I cannot find salt-ssh
06:23 TheCodeAssassin joined #salt
06:24 scristian joined #salt
06:25 markm joined #salt
06:25 sixninetynine joined #salt
06:26 markm joined #salt
06:30 ml_1 joined #salt
06:31 markm joined #salt
06:39 markm joined #salt
06:42 mmilano left #salt
06:55 SpX joined #salt
07:02 balboah joined #salt
07:02 pentabular joined #salt
07:06 ml_1 joined #salt
07:09 pentabular joined #salt
07:11 [M7] joined #salt
07:11 joehh yogi__: it is in the salt-ssh package
07:11 joehh reinsle: sorry got called away for meeting, hoepfully someone can track the issue down
07:11 joehh if it is a bug, i hear there is a 0.17.1 coming shortly
07:12 reinsle joehh: hope that to :-)
07:13 adepasquale joined #salt
07:25 eculver joined #salt
07:33 rmt2 joined #salt
07:45 pqpm joined #salt
07:48 zooz joined #salt
07:48 AshAsh joined #salt
07:50 kallek joined #salt
07:50 AshAsh hello :) Struggling to get a cmd.run command that launches a linux process to run the process in the background. Am I missing something? cmd.run 'java - jar xxx.jar &'
07:50 ronc joined #salt
07:52 rmt2 AshAsh, it's probably not passed to a shell.  You would have to do 'bash -c "java -jar xxx.jar &"' probably.  There might also be an argument to cmd.run that'll put it into the background.
07:56 kallek joined #salt
07:59 AshAsh ah okay cheers, will give it a go
07:59 qba73 joined #salt
08:07 Damoun joined #salt
08:08 bhosmer joined #salt
08:36 MrTango joined #salt
08:41 tomspur joined #salt
08:45 yota joined #salt
08:48 adepasquale1 joined #salt
08:59 EvaSDK hum someone using salt from ubuntu ppa ?
08:59 EvaSDK it looks like 0.17 hasn't made it to the ppa yet :(
09:13 aleszoulek joined #salt
09:24 joehh EvaSDK: I guess they are using the debian packages from debian.saltstack.com...
09:29 EvaSDK oh I see, thanks for pointing it out
09:30 az87c_ joined #salt
09:30 az87c joined #salt
09:31 geak joined #salt
09:32 EvaSDK hum any plan on getting 0.17 to an ubuntu release then ? wheezy is middly usable for raring
09:32 bud_ joined #salt
09:33 bud_ Has anyone used the lxc module from 0.17?
09:39 giantlock joined #salt
09:39 joehh EvaSDK: I believe someone is working on it, but I haven't heard from them since the weekend
09:40 EvaSDK k
09:42 NC left #salt
09:45 joehh EvaSDK: what does middly usable mean?
09:45 joehh are there obvious problems or something else?
09:45 EvaSDK well there are notable differences in the some of the dependencies
09:46 EvaSDK like libc
09:46 joehh yeah - ok
09:46 EvaSDK or python3.2 vs 3.3 but that shouldn't be a problem for now since salt is python2 only afaik
09:47 joehh I'm looking to simplify the debian/ubuntu packaging shortly
09:47 joehh sysvinit v upstart is the other one
10:16 oc I have a pillar, where I define (exim) as: disable_ipv6: 'true'
10:17 oc The file defaults configure like so:  disable_ipv6: {{cfg.get('disable_ipv6', 'false')}}
10:17 oc and the template like: disable_ipv6='{{ disable_ipv6 }}'
10:17 oc seems like Salt/python/yaml is too smart
10:18 oc and converts it to disable_ipv6='True'
10:18 oc how can I enforce that salt doesn't parse a boolean?
10:25 oc Second problem: after upgrading to v0.16.4 I get these errors when trying to restart / refresh a service: TypeError: __init__() got an unexpected keyword argument 'with_communicate'
10:25 oc ie. https://gist.github.com/oc/6865687
10:28 scott_w joined #salt
10:33 arapaho joined #salt
10:36 jrhorn424 joined #salt
10:38 joehh oc: how have you upgraded?
10:38 joehh packages, git, pypi, or something else?
10:39 oc joehh: packages
10:39 oc yum
10:39 oc sorry apt :)
10:39 oc this env = deb 7 boxes
10:39 joehh thought yum was a little odd
10:41 felixhummel joined #salt
10:41 joehh just reading source now
10:43 joehh is that gist from a 0.16.4 install?
10:43 joehh just the line numbers from the stack trace don't seem to line up... (could be me though)
10:43 oc hang on
10:46 oc humm, the actual server runs 0.17, don't know how that happened :~
10:46 oc Version: 0.17.0-1~bpo70+1~dst.1
10:46 joehh 0.17.0 was released to debian.saltstack.com through the week
10:47 joehh if you want to stick to 0.16.4 there are instructions at http://debian.saltstack.com/ on staying with a release stream
10:47 joehh but I'll now look at 0.17.0 :)
10:47 oc ok, someone might have been in on the server and done a manual update :~ that I don't know of; upgrading all to 0.17 now :)
10:47 oc joehh: well the master is still at 0.16.4
10:47 oc joehh: upgrading the master now
10:47 oc and will see if the slave starts the services properly
10:48 bhosmer joined #salt
10:48 joehh that could be it, but I doubt it (I could be wrong though...)
10:49 bud_ I receive a "TypeError: list indices must be integer, not str" when trying to create an LXC container on Debian Wheezy.
10:49 joehh oc: lines matching now!
10:49 oc all clear on the western front (after upgrading the master to .17) :)
10:49 joehh oc: that is nice to hear
10:49 oc sorry for the pebkac
10:49 joehh master must have more to do with things than I thought!
10:49 joehh oc: no worries
10:50 joehh :)
10:50 joehh bud_: 0.17.0?
10:51 bud_ https://gist.github.com/valentinbud/6865962
10:51 bud_ joehh: yes 0.17.0 on Debian Wheezy,
10:52 donatello joined #salt
10:53 donatello hi all. i've been experimenting with 0.16.4 and see that 0.17 is the current stable release. are there any release notes and info about upgrading backwards compatibility, etc?
10:57 joehh bud_: looking at the source, it looks like a pretty clear bug. Can you file an issue?
10:57 gildegoma joined #salt
10:57 bud_ joehh: of course, will do that in a couple of seconds. Can you tell me please where is the bug?
10:58 joehh donatello: https://salt.readthedocs.org/en/latest/topics/releases/0.17.0.html
10:59 donatello thank you joehh
10:59 oc joehh: thanks for maintaining the deb repos btw! I used to build these myself :)
11:00 joehh https://github.com/saltstack/salt/blob/develop/salt/modules/lxc.py#L346
11:00 joehh bud_: ^^ a list is created, no data added, then treated as a dict
11:01 joehh oc: no worries - need them in dayjob and if someone does it properly, it makes it easier for everyone
11:02 joehh donatello: not sure about backwards compatibility, someone here (oc) just had some issues with minion running 0.17.0 and master running 0.16.4
11:03 pengyao joined #salt
11:03 joehh donatello: if I was running in production and didn't need salt-ssh, I would stick on 0.16.4 for the moment
11:03 MrTango joined #salt
11:03 bud_ Ok. How could I fix the bug?
11:03 bud_ I have filed an issue - https://github.com/saltstack/salt/issues/7652.
11:04 oc joehh: tell me if you need a euro mirror site; i'll setup one
11:05 pengyao left #salt
11:05 donatello joehh: i want to start salt in production, to manage something like a 100 systems. I was hit by this bug - https://github.com/saltstack/salt/issues/7409 - which is fixed in the new version. any specifics about that bugs i might run into?
11:06 donatello i would be running same version of salt in the master and the minions
11:06 oc donatello: probably plenty of bugs, but 100 servers is quite overcomable with salt and staying on the current track
11:07 joehh bud_: I think they're alright with the bandwidth - I think they got some sort of unlimited bandwidth deal with their vm host for that mahcine
11:07 pengyao joined #salt
11:07 ninkotech ... so salt-ssh allows you to run salt commands even without minion running?
11:07 joehh they = saltstack
11:08 jslatts joined #salt
11:08 joehh donatello: you should be fine, just test upgrades on small subset first...
11:09 joehh most painful issue I am aware of is when debian packages prevented upgrading salt from within salt
11:09 joehh if you ran upgrade on many minions at once, it could require lots of manual interaction
11:09 joehh ninkotech: yes
11:10 pengyao left #salt
11:10 donatello joehh: oc: I am just starting off too, so I dont have a lot of existing code running with salt. i'll do some tests and see how it goes
11:11 joehh If you are just starting off, then probably go with 0.17.0
11:12 joehh bud_: whoops, meant oc before
11:13 ninkotech joehh: thanks,  this feature looks promising to me
11:13 bud_ joehh: thought so :).
11:13 joehh bud_: I do you mean the lxc side or the pull request side (I don't know anything about lxc containers)
11:14 bud_ joehh: changing list to list_ on line 345 of https://github.com/saltstack/salt/blob/develop/salt/modules/lxc.py#L346 works.
11:14 joehh ninkotech: requires python2.6 on other end
11:14 joehh bud_: yeah, that looks like it
11:15 joehh can you do a pull request?
11:16 joehh either that or just mention it in the issue
11:16 bud_ Ok.
11:25 ninkotech joeh, ty, thats ok
11:35 jbunting joined #salt
11:42 geak joined #salt
11:52 zwe joined #salt
11:58 anuvrat joined #salt
12:05 diegows joined #salt
12:05 brianhicks joined #salt
12:13 jbunting joined #salt
12:37 copelco joined #salt
12:37 timoguin joined #salt
12:38 Drawsmcgraw joined #salt
12:43 reinsle joehh: ping
12:44 reinsle joehh: my error using pillar.get is already in 0.16.4, i backported to 0.16.4, and it don't work
12:50 joehh reinsle: hi
12:50 zooz joined #salt
12:50 joehh did you see it worked for me in a template file
12:50 joehh and also the other issue in which it seemed to work
12:53 p3rror joined #salt
12:55 mgw joined #salt
12:55 jY any news on the salt 0.17 debs for ubuntu?
12:55 blee joined #salt
12:56 blee good monday friends.
12:59 joehh jY: no - I believe Corey is working on them
12:59 joehh blee: tuesday here in 1 minute!
13:00 joehh or now :)
13:00 joehh reinsle: it has just worked for me
13:00 joehh what happens if you restart yuor minion
13:02 joehh reinsle: It seems pillar data is cached - if I change the pillar data then I keep getting the old data
13:02 joehh until I restart the minion
13:02 joehh I suspect there are other ways to get it to update though
13:02 elfixit joined #salt
13:06 bhosmer joined #salt
13:06 mua joined #salt
13:07 bhosmer joined #salt
13:08 ipmb joined #salt
13:08 oz_akan_ joined #salt
13:11 LucasCozy joined #salt
13:12 juicer2 joined #salt
13:13 Gifflen joined #salt
13:15 tseNkiN joined #salt
13:16 Kholloway joined #salt
13:17 racooper joined #salt
13:21 timoguin joined #salt
13:23 oc hmm, all salt pillars have disappeared after upgrading to .17... (i.e. salt '*' pillar.data
13:23 oc shows empty on all servers
13:25 oc 2013-10-07 15:24:38,861 [salt.loader                              ][CRITICAL] Failed to load grains defined in grain file opts.opts in function <function opts at 0x294f7d0>, error:
13:25 oc AttributeError: 'str' object has no attribute 'get'
13:26 oc https://gist.github.com/oc/6867986
13:27 oc seems like __opts__ or __pillar__ has strangely become a string
13:27 oc in opts.py
13:29 p3rror joined #salt
13:33 joehh oc: odd, pillar.data is working for me on 0.17.0
13:33 oc I have a huuuge pillar tree, so bisecting now :)
13:35 toastedpenguin joined #salt
13:36 micah_chatt joined #salt
13:36 jkleckner joined #salt
13:36 oc sigh
13:37 oc I need to add some CI / test suite soon
13:37 bcc joined #salt
13:37 cro1 joined #salt
13:38 bcc I need my minion to get its id/hostname from a bash script.. is this possible?
13:41 geak_ joined #salt
13:42 cro1 joined #salt
13:43 faldridge joined #salt
13:49 bcc is it possible? :/
13:50 anuvrat joined #salt
13:52 cetex bcc: you can set it in the minion cfg, so you could bash-script the minion config.
13:52 kaptk2 joined #salt
13:52 bcc cetex, thanks.. can you give me an example?
13:53 Ram_ joined #salt
13:53 Ram_ Hi Guys
13:53 Ram_ How are you
13:53 Ram_ i need help from u guys
13:54 Ram_ i need to write code in saltstack
13:54 Ram_ can you help me on that
13:54 Ram_ like configuration management
13:55 cetex bcc: no, not really. but i'm quite certain that you have an option for "minion id" in the minion cfg.
13:55 brianhicks joined #salt
13:55 cetex bcc: assuming you haven't already set it you could: "echo id: my.minion.com >> /etc/salt/minion" or something
13:56 bcc yeah could do.
13:59 oc GAHHHHHHHHHH$%@W
13:59 oc is there some decent docs on how to debug pillar data
13:59 oc salt 'lamp*' pillar.raw
13:59 oc gives me nothing
14:00 terminalmage joined #salt
14:00 anteaya joined #salt
14:03 StDiluted joined #salt
14:07 joehh oc: have you tried restarting a minion
14:07 mofomikes joined #salt
14:07 oc joehh: yup
14:07 joehh I found once I restarted then my issues went away
14:07 joehh ugghhhghghg
14:07 oc restarted all
14:07 mofomikes morning ppl
14:10 Drawsmcgraw oc: pillar.data ?
14:10 oc Drawsmcgraw: not showing any pillar data
14:10 Drawsmcgraw wow srsly? Zero? Interesting
14:10 oc zero :~
14:11 Drawsmcgraw You say pillar.data, and *nothing* comes back?
14:11 Drawsmcgraw just the name of the minion with some dashes under it?
14:11 Drawsmcgraw And I'm guessing there's nothing in the Minion log...
14:11 jslatts joined #salt
14:11 oc I do get some interesting error messages in the log
14:11 oc gist soon
14:11 oc need to isolate it
14:12 oc Drawsmcgraw: yes
14:13 Drawsmcgraw k
14:18 oc https://gist.github.com/oc/6868768
14:18 oc started in debug mode (salt-minion -l debug)
14:18 oc there is obviously something wrong
14:18 oc but I'm unable to pinpoint it
14:19 oc my minion cfg is quite simple, just: 'master: mgmt.int.lowercase.no'
14:21 Drawsmcgraw oc: so... does the minion even run after that stacktrace?
14:22 oc yeah
14:22 oc it doesn't exit
14:22 oc at least
14:22 oc it responds on salt '*' grains.list
14:22 oc but pillar.data remains empty
14:23 oc root@mgmt:/srv/salt# salt 'mgmt*' pillar.data
14:23 oc mgmt.int.lowercase.no:
14:23 oc <nothing>
14:23 Drawsmcgraw odd
14:24 Drawsmcgraw Definitely something about the 'critical' error. I just fired up my minion and it doesn't throw anything like that
14:24 Drawsmcgraw I think the absence of Pillar data is a symptom of the actual problem.
14:24 Drawsmcgraw What the *actual* problem is.... I'm afraid I can't say
14:25 oc File "/usr/lib/python2.7/dist-packages/salt/grains/opts.py", line 12, in opts
14:25 terminalmage pillar data is compiled on the master so any errors in compiling it would be in the master log
14:25 oc seems like __opts__ or __pillar__ evals to str
14:25 oc and i have no idea how
14:27 micah_chatt joined #salt
14:31 JaredR joined #salt
14:32 oc when printing __pillar__ in the opts.py context (i.e. print dir(__pillar__), i get: pillar == ['__add__', '__class__', '__contains__', '__delattr__', '__doc__', '__eq__', '__format__', '__ge__', '__getattribute__', '__getitem__', '__getnewargs__', '__getslice__', '__gt__', '__hash__', '__init__', '__le__', '__le
14:32 oc n__', '__lt__', '__mod__', '__mul__', '__ne__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__rmod__', '__rmul__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', '_formatter_field_name_split', '_formatter_pa
14:32 teskew joined #salt
14:32 oc rser', 'capitalize', 'center', 'count', 'decode', 'encode', 'endswith', 'expandtabs', 'find', 'format', 'index', 'isalnum', 'isalpha', 'isdigit', 'islower', 'isspace', 'istitle', 'isupper', 'join', 'ljust', 'lower', 'lstrip', 'part
14:32 oc ups
14:32 bhosmer joined #salt
14:32 oc didn't plan on pasting all that :)
14:32 jkleckner joined #salt
14:33 teebes joined #salt
14:34 blee_ joined #salt
14:35 bhosmer_ joined #salt
14:36 oc i.e. a string
14:36 oc and not a dict
14:38 yidhra joined #salt
14:40 oc how do you structure your pillar folders?
14:41 geak joined #salt
14:43 woswasi joined #salt
14:44 woswasi left #salt
14:44 matanya joined #salt
14:45 pdayton joined #salt
14:51 mmilano joined #salt
14:54 jslatts joined #salt
14:54 StDiluted joined #salt
14:55 cshuman joined #salt
14:56 cshuman joined #salt
14:57 jbunting joined #salt
14:58 aptiko I need to run a shell command from a sls file; not a cmd.run, but something that I will use to set a template variable according to the results of the shell command. I tried {% set myvar = salt['cmdmod.run']('mycommand') %} but it says 'dict object' has no attribute 'cmdmod.run'.
15:02 torandu joined #salt
15:02 bhosmer joined #salt
15:03 tyler-baker joined #salt
15:08 aptiko Probably a documentation error; http://docs.saltstack.com/ref/modules/all/salt.modules.cmdmod.html should probably be cmd, not cmdmod.
15:08 kenbolton joined #salt
15:08 kenbolton left #salt
15:10 noxeor joined #salt
15:11 cshuman joined #salt
15:11 noxeor Hi there; Totally low-level noob Salt question.  Is it free to use?  I am interested in it for remote execution and config management for 100-200 platforms.
15:11 StDiluted yes, it is free to use
15:12 StDiluted they charge for support if you need it
15:12 noxeor Awesome.  Looking around the page it wasn't totally clear...
15:12 StDiluted however, I've gotten all the support I could ever need here
15:12 noxeor excellent.  thanks for the prompt response!
15:12 StDiluted yeah I had the same question initially
15:12 jalbretsen joined #salt
15:12 noxeor I'm comparing Salt, puppet, CFEngine, and Chef.  I like Salt the best but the cost was a question. :)
15:13 StDiluted in my opinion, as a sysadmin who knows just enough ruby to maybe understand and write a simple program, salt is much easier to learn
15:13 StDiluted and I was up and running with salt in hours, where it took me a couple days to get anything going in chef when I tried it
15:14 noxeor I don't know much Ruby but am proficient in perl
15:14 StDiluted puppet is sort of along the same lines as chef
15:14 StDiluted salt is written in python
15:14 SunSparc joined #salt
15:14 jrb28 joined #salt
15:14 StDiluted but generally speaking you wont need python to use it well
15:14 StDiluted it's all YAML
15:14 noxeor familiar with python but have not sunk my teeth in, if I get to dive deeper into python that's just a bonus for me ;)
15:15 noxeor Thanks; Very helpful.  I'll push for Salt but the bosses will decide.... sigh
15:17 toastedpenguin joined #salt
15:17 StDiluted no problem
15:20 geak_ joined #salt
15:22 abe_music joined #salt
15:25 taylorgumgum joined #salt
15:27 geak joined #salt
15:30 jumperswitch joined #salt
15:31 torandu joined #salt
15:31 fitzdsl joined #salt
15:31 fitzdsl Hello guys. I'm working with TheForeman community to get Foreman more aware of other configuration management systems. Historicaly Foreman was puppet focused. We just started the implementation of a Chef to Foreman connector. But obviously it needs some re-design of Foreman to be less puppet oriented. So we are looking of people with good knowledge of others CM to be able to design a generic Report processor and Facts/Ohai/whatsever processor. If anyone
15:32 timoguin theforeman supporting salt would be A++
15:32 timoguin fitzdsl, your message cut off at "If anyone"
15:32 fitzdsl If anyone is interrested to jump in the discussion, feel free to add some comments here : http://projects.theforeman.org/projects/foreman/wiki/CMS-Report
15:33 mmilano joined #salt
15:33 timoguin fitzdsl, you might want to post that to the mailing list
15:35 DredTiger joined #salt
15:35 jumperswitch joined #salt
15:37 matanya joined #salt
15:39 druonysus joined #salt
15:42 jumperswitch_ joined #salt
15:44 pdayton joined #salt
15:46 jumperswitch joined #salt
15:47 pentabular joined #salt
15:49 aptiko Is there any equialent to Puppet's "notify"?
15:49 balboah what does Puppet's notify do?
15:50 pnl notifications, i guess
15:51 timoguin good guess
15:51 aptiko balboah: I want salt to display something while processing an sls file.
15:51 aptiko balboah: for debugging.
15:51 aptiko I want to see the value of a template variable.
15:55 jkleckner joined #salt
15:55 jeffasinger joined #salt
15:55 cro joined #salt
15:56 pipps1 joined #salt
15:57 oc I found the error btw; seems like I cannot do empty node groups in the top salt state
15:58 StDiluted aptiko: salt -l debug command
15:58 StDiluted that's a minus L
16:00 aptiko If I have a template variable holding a string that contains newlines, can I somehow write it to a file, e.g. using file.managed?
16:01 jeffasinger aptiko: yeah, pass it in the context to file.managed
16:02 aptiko jeffasinger: Ah, so I need a template file. I was wondering whether I can do it directly (with the "contents" of file.managed), but if that's not possible I'll do it with a file template.
16:06 jeffasinger aptiko: I'm far from an expert in salt, I just know that using a template file would work, there may well be another option, perhaps a custom template engine
16:07 aptiko Hmm, there _are_ ways to do it. One is "- contents: |" and below that {{ gpg_pub_key.replace("\n", "\n        ") }}
16:07 aptiko It's a hack, but it works.
16:07 jcockhren Soooo
16:07 jcockhren fluent+elasticsearch+salt is doable
16:08 opapo joined #salt
16:08 jcockhren my setup uses syslog returner for salt
16:10 jcockhren ideally, I'd want to use the redis returner. however, the returned values are just a list of key:value pairs.
16:11 seanz joined #salt
16:11 jcockhren if the key:value pairs were submitted as a jason value of some key, then elastic would munch it up nicely
16:11 jcockhren also, I could've just been doing it wrong
16:12 jcockhren I wonder how far we can push this for log management
16:13 jcockhren currently the returners just forward the results from ran modules/states
16:13 m_george left #salt
16:13 jcockhren no hardbeat information
16:13 jcockhren heartbeat*
16:14 ldlework Jason is a terrible serialization format.
16:14 brimpa joined #salt
16:14 jcockhren noted
16:14 ldlework :)
16:15 abe_music joined #salt
16:16 bud_ Can someone tell me what the following does - return __salt__['config.option']('lxc.profile', {}).get(profile, {}) ?
16:17 alunduil joined #salt
16:19 jcockhren maybe I'll do a returner for elasticsearch hmmmmm
16:20 jcockhren what would everyone/anyone here be interested in logging (for searching and diagnostics) relative to salt usage?
16:23 CheKoLyN joined #salt
16:25 jcockhren hmmm
16:25 jcockhren maybe salt-cp calls and mine?
16:27 timoguin i'll probably be doing mine and spitting everything to logstash on intervals
16:27 timoguin here soon
16:27 mofomikes is there a way to use pillar data in the argument of a remote command execution?
16:28 thekabal joined #salt
16:28 ldlework mofomikes: sure, use jinja2 to render the value into your file
16:28 ldlework {{pillar['key']}}
16:28 ldlework at it simplest
16:28 thekabal left #salt
16:29 mofomikes and that works with remote command executions? not just in state files?
16:29 mofomikes (sorry im new to sal)
16:29 mofomikes salt
16:29 ldlework mofomikes: oh you mean like on the commandline? (you can do remote command execution in the states as well)
16:30 cf_psmallwood joined #salt
16:30 mofomikes yeah i meant from command line
16:30 mofomikes my goal is to control svn checkouts across multiple minons
16:30 druonysus joined #salt
16:30 cf_psmallwood Hello. How do I check whether a given/specific minion matches a new sls/state that I have built?
16:30 jcockhren timoguin: yeah. I was thinking of going the logstash route. For now, it's too big for me. I think maybe a http returner could be interesting since many of logstash and the like have HTTP interfaces
16:30 Lue_4911 joined #salt
16:31 ldlework mofomikes: well then you'll need to store the value first, say into an shell variable
16:31 ldlework mofomikes: seeing as how the pilliar is output as yaml/json, etc you might need some finangling
16:31 mofomikes idlework: ok thats reasonable
16:31 ldlework unless there's a way to get a single value out of the pillar from the commandline (there probably is)
16:31 mofomikes yeah there is
16:31 mofomikes i'll pull the pillar data then pass it back using a shell var i guess
16:32 bhosmer joined #salt
16:37 mr_chris cf_psmallwood, You mean matches in the top.sls?
16:37 cf_psmallwood Rephrasing slightly, I've made an sls file for setting the /etc/security/limits.conf to a set of user/value combinations. How do I check to see if a specific minion matches that configuration? salt 'minion' state.show_sls sec_limits dev seems right, but I'm not sure.
16:38 sashka_ua guys, not sure how I can add more returners by defualt, like I want all output to be logged in mongo. I can use mongo_return, but how to enable it persistently?
16:38 jcockhren sashka_ua: ext_job_cache in the master
16:38 jcockhren config
16:38 bud_ Is anyone using the lxc module from 0.17.0?
16:38 sashka_ua let me check
16:38 sashka_ua jcockhren: ty, gonna check this
16:39 jcockhren I think --out=json should be smarter
16:39 jcockhren maybe not include newlines
16:40 jcockhren and spaces
16:40 sashka_ua jcockhren: yes, thank you. could not find that obvious option :)
16:40 cachedout joined #salt
16:40 Ryan_Lane joined #salt
16:40 sashka_ua especially that name of the option doesnt reflect its meaning :)
16:41 jcockhren ha yeah
16:42 mr_chris cf_psmallwood, That seems right to me.
16:43 cachedout Could somebody with ops please bump the channel topic to reflect the 0.17 release.
16:43 cf_psmallwood mr_chris> Thanks. Unfortunately, when I run that against a known-match minion, and a known-NOT-match minion, the results are identical. Is there a different way to check whether a given file matches what my sls says it should be set to be?
16:46 KyleG joined #salt
16:46 KyleG joined #salt
16:46 mr_chris cf_psmallwood, I'm not very clear on what you're trying to do. Are you trying to verify that you state is actually setting /etc/security/limits.conf to what you want or are you trying to verify that a specific sls is getting applied to a specific host?
16:46 bud_ s0undt3ch: I see you're the last person to commit on the LXC module.
16:47 bud_ Are you around to ask you a question.
16:48 cf_psmallwood mr_chris> Neither. I have built an sls, and prior to deployment, I'd like to be able to determine whether a given minion matches the state I defined in that sls. Basically, I want a command that will say "Yes, this minion's file matches that setting", prior to sending it out.
16:48 jcockhren cf_psmallwood: yo. Instead of checking for appends in the limits.conf, I'd instead file.manage s file with the settings you want in /etc/security/limits.d directory
16:48 jcockhren much easier
16:50 bud_ The LXC module docs state that you can configure profiles in pillar using lxc:
16:50 bud_ dictionary.
16:50 bud_ I have followed the docs but it doesn't work.
16:51 cf_psmallwood jcockhren> That seems.. very elegant! Unfortunately, I'm stepping into an already existing environment, and the reason I'd like to query/check the current setting (in limits.conf) is that the production environment might have lower and/or higher values than my choices. In some cases, that could be very bad, so I don't want to deploy there.
16:51 bud_ The pillar now contains a lxc dictionary with 2 profile defined under it.
16:52 troyready joined #salt
16:52 mgw bud_: I wrote the lxc module, I'll try to answer your questions.
16:53 LarsN I really need to sit down and really wrap my head around Pillars.
16:53 cf_psmallwood jcockren> The output/check result will also help justify the setting, as in "100 of the 400 minions report that this matches or exceeds the requested setting, so we should be safe to deploy it"
16:53 bud_ Hello mgw. Thank you!
16:53 jcockhren cf_psmallwood: so it sounds like you're asking for a way to return that file?
16:53 bud_ First of all I think it's broken :|.
16:53 xmltok joined #salt
16:54 LarsN And also likely wrap my head around syndic
16:54 bud_ mgw: here is an issue I have encountered - https://github.com/saltstack/salt/issues/7652
16:54 mgw bud_: Unfortunately, I had some updates that did not make 0.17
16:54 cf_psmallwood jcockhren> I was more hopeful that there was a way for salt to validate whether the settings were present (grep?) and matching (awk/sed?), but yes, I could manually make do with a grab of the file
16:54 cowmix joined #salt
16:55 luminous if salt.client.Caller executes as if it were salt-call -  https://github.com/saltstack/salt/blob/develop/salt/client/__init__.py#L1376 - why am I getting stuff like: Loaded minion key: /etc/salt/pki/minion/minion.pem
16:55 cowmix what is the best Centos 6 base to do SaltVagrant work?
16:55 LarsN cf_psmallwood: one of the features I  would like to see is similar to cfengine's "append if no such line"
16:55 cowmix SaltyVagrant
16:56 mgw bud_: I had used "def list" unintentionally; soundtech fixed it for me, but failed to make that update.
16:56 * cf_psmallwood nods to LarsN
16:57 luminous LarsN: isn't there a file.append? does it not work that way?
16:57 mgw s0undt3ch: What can we do to get a bug fix into 0.17 branch?
16:57 LarsN cf_psmallwood: I "think" you might be able to hack something together with a sufficiently complext jinja2 template.  It's also possible there's a process and I'm just dumb.
16:57 bud_ mgw: I have changed to list_ in the code as stated in the issue and it works now.
16:57 mgw yeah, that's what has to change. I have that change in my local rep
16:57 cf_psmallwood I think jcockhren has an elegant end-state for me to head towards, and it will be (much) easier once I get there. A file in limits.d that I can check presence for. But for now, I'm trying to find the non-ideal way through to there. ;)
16:57 bitz joined #salt
16:57 bud_ The problem is that the profile thing doesn't work.
16:57 LarsN luminous: It might, I admit I hadn't seen file.append before.  I'll need to check whether or not it "detects" the existance before trying to append.
16:57 luminous I think it does
16:58 LarsN would hate to have:  include *Allthethings (appended 100 times)
16:58 bud_ I have 2 profiles defined in pillar under lxc.
16:58 mgw bud_ You're familiar with the _modules dir, right? So you don't have to hack the salt source tree.
16:58 LarsN ;:)
16:58 Ryan_Lane joined #salt
16:58 bud_ mgw: yes I am. Can I place a new lxc module there and it would override the base one?
16:58 mgw yes
16:58 mgw that's how I do my dev
16:59 bud_ Cool, thanks for your hint.
16:59 seanz joined #salt
16:59 * LarsN adds file.append to his ToDo List.
16:59 mgw bud_: can you gist your (redacted) pillar data?
16:59 jcockhren cf_psmallwood, LarsN: go ahead and write a limits.conf module ;)
17:00 bud_ mgw: https://gist.github.com/valentinbud/6871294
17:00 cf_psmallwood I'm only two weeks into using salt, but I wouldn't count it out in the long term. ;)
17:01 mgw try doing lxc.profile:
17:01 mgw instead of lxc: profile:
17:01 LarsN cf_psmallwood: I think it's a great tool to have in the toolbox.  We're primarily a Chef shop, but there are places Salt still fits in pretty well.
17:01 luminous question, isn't salt-call supposed to be usable without a master?
17:01 mgw I think either should work, but that's how I have it
17:01 mgw bud_: ^
17:02 luminous ah, right.. salt-call --local
17:03 bud_ mgw: trying it now.
17:06 taylorgumgum joined #salt
17:07 brimpa joined #salt
17:07 dthom91 joined #salt
17:07 shinylasers joined #salt
17:08 pdayton joined #salt
17:08 luminous anyone know how to run salt.client.Caller() as if it were salt-call --local?
17:09 bud_ mgw: https://gist.github.com/valentinbud/6871397 with details.
17:10 mgw bud_: remove line 4
17:10 az87c joined #salt
17:11 josephholsten joined #salt
17:11 az87c_ joined #salt
17:12 bud_ mgw: Same output.
17:12 mgw bud_: that's odd
17:12 bud_ Did a saltutil.refresh_pillar after modifying the pillar and a test via pilllar.item lxc.profile
17:13 mgw so your output from pillar.item was updated, I assume
17:13 mgw but it didn't use the profile?
17:13 amahon joined #salt
17:13 bud_ Yes it was updated. But it didn't use the profile.
17:14 bud_ The lxc-create command should have been lxc-create -n lxc211 -B (etc...) but it only has -n lxc211 no other stuff in it.
17:15 mgw bud_: I see it
17:16 bud_ Tried with profile='debian' (quoted), still the same output.
17:16 mgw bud_: It's not going to work, it's a bug :-(
17:17 mgw I'll submit a patch, and link you to the patched version.
17:17 bud_ mgw: that's great news :-), thanks!
17:18 mgw bud_: http://git.io/lZMHhA
17:19 mgw I'm not 100% sure that version will make it into salt, due to a slight API change
17:19 Drawsmcgraw joined #salt
17:19 mgw in the info() function
17:20 mgw but other than a change from s/ifaces/nics/ in the return from info, I think it will make it upstream
17:20 bud_ Cool, so this is your code that you've said you had it done but didn't make it in the release.
17:21 mgw yes
17:21 bud_ mgw: one more thing, with lxc: \n  profile: etc. The salt '*' config.option lxc.profile doesn't work, with lxc.profile it does.
17:22 mgw ok, I'll fix the docs then
17:22 mgw thanks
17:22 mgw Not sure why I put it in the docs that way
17:23 mgw let me know if that module I linked to fixes it for you
17:24 amckinley joined #salt
17:26 jumperswitch_ joined #salt
17:26 bud_ Thank you mgw :-). It does work.
17:27 jefimenko joined #salt
17:27 bhosmer joined #salt
17:34 xmltok does the top file need to be yaml or does it run through the render system? I'd like to put a lot of hostname based logic into the top file
17:35 xmltok it would also be good to be able to hit an external service to determine if additional states should be added
17:36 mofomikes joined #salt
17:37 seanz joined #salt
17:37 zooz joined #salt
17:38 jbunting joined #salt
17:42 StDiluted xmltok: explain what you mean by that?
17:49 faldridge joined #salt
17:53 Katafalkas joined #salt
17:56 xmltok stdiluted: well, i want to include states based on more things than just the hostname wildcards. So if I can use a #!py in the top file, I could then write python code to hit my database and find out other things to make state inclusion decisions on
17:57 mgw bud_: glad it's working, I'm working on getting the bug fixes into stable right now
18:02 mannyt joined #salt
18:02 pdayton joined #salt
18:03 josephholsten joined #salt
18:03 jslatts joined #salt
18:06 jdenning joined #salt
18:07 jkleckner joined #salt
18:08 cro joined #salt
18:09 seanz joined #salt
18:14 bhosmer joined #salt
18:15 mattmtl joined #salt
18:16 bhosmer joined #salt
18:18 mofomikes joined #salt
18:19 DredTiger joined #salt
18:24 Lue_4911 joined #salt
18:28 pdayton1 joined #salt
18:30 xuru joined #salt
18:31 amckinley joined #salt
18:32 xuru hey everyone…   Q: how would I manage deploying a django app to two or more environments (say, vagrant, qa, and production)?  Do I use grains?
18:35 cachedout xuru: A good place to start might be part 4 of the States tutorial, which describes just such a scenario.
18:37 xuru Yeah, I read that, but it was kind of confusing.  First it talked about having seperate directories for each, but then in the "practical example" they reference things like 'web*prod*' in the top.sls
18:41 pentabular joined #salt
18:41 cachedout Ah, yes. I found that same section somewhat confusion. I took it to mean that when they write something like 'web*prod*' that they're assuming a minion naming scheme that would match against a set of minions responsible for serving production code.
18:41 cachedout (But I could of course be wrong. I agree that the section is hard to follow.)
18:41 cro1 joined #salt
18:43 rgbkrk joined #salt
18:44 xuru so, say I want prod to use mysql, but vagrant to use sqlite.  The fall through model kind of falls apart then
18:46 xuru I guess what I'm asking is, can I somehow conditionally include different states based on the environment?
18:46 adepasquale joined #salt
18:47 ldlework xuru: my understanding is that you're supposed to use different "environments". Like the "base" environment is the default one.
18:47 ldlework http://docs.saltstack.com/ref/states/top.html#environments
18:47 ldlework xuru: and the answer is yes
18:48 xmltok joined #salt
18:48 xuru ah, ok.  Let me read though that and see if it's clearer then the tutorial.  thanks
18:52 cachedout That doc shows that I was mistaken in my assumption, FWIW. Apologies for that.
18:52 xuru np
18:56 Drawsmcgraw joined #salt
19:05 bhosmer joined #salt
19:07 amckinley joined #salt
19:12 jbunting joined #salt
19:13 s0undt3ch mgw: what's the issue?
19:13 s0undt3ch bud_: what's up?
19:14 LarsN Are any of you using any tools to lint test your state files?
19:15 mgw s0undt3ch: basepi took care of it, there were two bugs
19:15 s0undt3ch mgw: awesome!
19:15 mgw one was introduced by the mapping of list to list_
19:15 mgw the other was a plain ol' bug
19:15 mgw regarding lxc profiles
19:15 mgw It's all good now though
19:15 nahamu LarsN: https://github.com/saltstack/salt/issues/802 looks like it doesn't exist yet... :/
19:16 LarsN mgw: there are no bugs, only undocumented features. :)
19:16 LarsN nahamu: I've found tools that can at least lint test yaml in general.
19:16 LarsN nothing specific to salt though eh?
19:16 nahamu I have no idea other than that bug report.
19:16 LarsN nahamu: rgr.
19:16 nahamu I turned to google because I was curious too.
19:16 mgw LarsN: in this case it was a documented non-feature
19:16 LarsN mgw: heh
19:17 LarsN mgw: similar to tabs in your sls files.
19:17 nahamu https://gist.github.com/alecthomas/4497119
19:17 LarsN undocumnted non-feature.  ;)
19:17 jumperswitch_ joined #salt
19:17 LarsN nahamu: I might give that a try.  Anything that self-proclaims "very hacky" seems perfect :)
19:18 nahamu :)
19:18 LarsN in my test environment anyway (which qualifies as very hacky itself)
19:18 nahamu found that in a comment in that issue
19:20 jaequery joined #salt
19:25 giantlock joined #salt
19:26 lempa joined #salt
19:27 lempa joined #salt
19:27 gaoyang joined #salt
19:30 bhosmer joined #salt
19:32 aantony joined #salt
19:33 bhosmer joined #salt
19:33 Tekni joined #salt
19:46 mofomikes joined #salt
19:46 rgbkrk Contributions on GitHub -- should they be in master or develop?
19:46 rgbkrk I'm guessing develop, but thought I'd ask
19:46 balltongu joined #salt
19:47 rgbkrk HACKING.rst makes one comment about `git pull upstream develop`
19:47 kermit joined #salt
19:48 rgbkrk And submitted that tiny PR to develop. Cheers!
19:52 taylorgumgum joined #salt
19:54 zwe joined #salt
20:05 bogdanbu joined #salt
20:06 xmltok joined #salt
20:07 aantony rgbkrk: yeah develop
20:08 bogdanbu hi there ! i'm using salt 0.17 and i get this error "AttributeError: 'list' object has no attribute 'get'" in a file.managed section in a jinja template i guess
20:08 aantony bogdanbu: there is a list where a dict is expected.  any more info?
20:08 redondos joined #salt
20:09 bogdanbu in a second
20:09 bogdanbu to paste-it
20:10 bogdanbu http://pastie.org/8384872
20:11 bogdanbu the file.recurse part is working
20:11 redondos joined #salt
20:12 robbyt joined #salt
20:12 bogdanbu where is the list ? :)
20:14 toddejohnson joined #salt
20:14 logix812 joined #salt
20:17 mmilano joined #salt
20:17 a1j https://github.com/saltstack/salt/issues/7661
20:19 josephholsten joined #salt
20:19 xuru joined #salt
20:20 aantony bogdanbu: i'm not too famillair with this area, but one thing that looks, off:  are those ports supposed to be keys under context?
20:20 aantony bogdanbu: the indenting seems too shallow
20:20 shinylasers joined #salt
20:20 bogdanbu i saw that
20:21 bogdanbu and i corrected but the same error ...
20:22 bogdanbu should i empty some cache before retry ?
20:22 aantony i don't think there is any spacial caching for templating
20:23 aantony if you omit the context, i assume you get errors about missing template params?
20:23 bogdanbu http://pastie.org/8384898
20:24 bogdanbu with right indent..i guess
20:25 andrej Does terminalmage/Erik Jonhson hang out here? :)
20:25 bogdanbu hmm…i removed context but the same error appears
20:26 andrej I see him in /whois output
20:26 aantony bogdanbu: try this, just for kicks,  instead of 'context', name it 'defaults'
20:26 robertkeizer joined #salt
20:27 bogdanbu yeah no problem but .. it was defaults..i've changed to context :)
20:27 aantony "The general best practice is to place default values in defaults, with conditional overrides going into context, "
20:28 xmltok joined #salt
20:28 terminalmage andrej: what's up?
20:28 ckao joined #salt
20:28 aantony bogdanbu: what if you define defaults, and override them context, does that work?   maybe the bug is that templating requires that defaults be defined, even if context is supplied
20:29 andrej hi terminalmage ... just wondering about the "cannot reproduce" on the github submission of mine :)
20:29 bogdanbu ok, i'll try in a minute
20:29 andrej Granted, what I did was stupid; but I would still have expected some sort of error ;)
20:30 cwright Does SaltStack Inc produce/maintain the salt packages on launchpad.net?
20:31 bogdanbu aantony: sorry…but the same error
20:32 bogdanbu aantony: i have defaults: - and context: - … with the same keys: values
20:32 aantony bogdanbu: that does not sound good, then.   you appear to be doing it all by the book
20:32 luminous to anyone familiar with using salt.client from python: I'm trying to use salt.client.Caller() like salt-call --local, but am unclear how to specify that the client should run in masterless mode - is thispossible?
20:32 bogdanbu on 0.16.4 it was ok
20:33 avienu joined #salt
20:33 bogdanbu 0.17 is some kind of beta ?
20:34 luminous bogdanbu: stuff breaks
20:34 luminous salt-stack now has jenkins, which has helped tremendously.. but the team has a lot to cover and certainly misses stuff
20:34 bogdanbu i think this is a very basic feature …
20:34 luminous file a bug if you feel confident you have found on
20:34 luminous *one
20:35 bogdanbu I'm kind of new to stack…so i'm not so confident … yet :p
20:35 jaequery joined #salt
20:35 bogdanbu thanks !!!
20:36 josephholsten joined #salt
20:36 aantony bogdanbu: to be clear are you using the released 0.17 package, or the 0.17 branch from github?
20:36 bogdanbu the package
20:37 bogdanbu apt-get ...
20:37 timoguin left #salt
20:37 terminalmage andrej: I deal with a lot of issues on github, can you be more specific?
20:38 terminalmage perhaps an issue number?
20:38 anveo joined #salt
20:38 andrej absolutely ... 7601
20:40 terminalmage andrej: I marked it as Cannot Reproduce before the issue was clarified
20:40 terminalmage I thought you had meant that, after you fixed it, it didn't work
20:40 terminalmage as in, it was completely broken
20:41 andrej Ah OK
20:41 ckao is there any documentation to describe how to use complex conditional statements with Pillar?  For example, I'm trying to do something like this:  {% if (grains['virtual'] == 'physical' && grains['mem_total'] == '1034144' && grains['num_cpus'] == '64') %}
20:41 andrej No, it's not totally broken. :)
20:41 terminalmage andrej: I'll update the tags on that issue
20:41 andrej Ta
20:43 luminous ckao: maybe use python for that?
20:43 oz_akan_ joined #salt
20:43 luminous ckao: or an ext_pillar?
20:44 ckao luminous:  i guess i'll need to learn python then =)
20:44 luminous eh, don't worry, it's easy
20:44 luminous and you'll be way more effective with salt
20:44 oz_akan_ joined #salt
20:44 ckao what's ext_pillar?
20:44 luminous or, rather, you'll have more freedom
20:45 luminous ckao: ext_pillar is a way for you to feed in pillar data from some external system/command/service/etc
20:45 luminous some external source
20:45 avienu joined #salt
20:47 ckao okay, thanks for the suggestions.  does python code live within the {% .. %} in sls files?
20:48 luminous ckao: I think that depends on the sls renderer you are using (default is jinja)
20:49 luminous but I think you may even be able to use some simple python in jinja, I forget
20:49 luminous I try to keep things simpler than complex conditionals dictate
20:50 knut_ joined #salt
20:50 ckao got it, thank you!
20:50 LarsN luminous: while Jinja is the default, I believe you still have to specify that that's what you're using.
20:51 luminous LarsN: if you are rendering a template, yes
20:51 LarsN or at least I remember having to specify ~ 0.15.x or it would just pri....
20:51 LarsN ahh....
20:51 luminous but not .sls in pillar
20:51 LarsN yeah that makes more sense.
20:51 luminous ;)
20:51 LarsN sorry just chalk it up to LarsN == Almost completely stupid.
20:52 knut_ left #salt
20:54 Katafalkas joined #salt
20:55 amahon joined #salt
20:58 xuru could someone point me to a good example of using environments?  I can't seem to find any on github from some reason...
21:02 jrb28_ joined #salt
21:02 matanya joined #salt
21:03 cro joined #salt
21:03 pipps joined #salt
21:06 jkleckner joined #salt
21:07 nineteeneightd joined #salt
21:09 matanya joined #salt
21:20 MohShami joined #salt
21:28 rgbkrk joined #salt
21:29 pentabular joined #salt
21:30 StDiluted_ joined #salt
21:31 sashka_u1 joined #salt
21:31 Chocobo_ joined #salt
21:31 arapaho_ joined #salt
21:32 ninkotech joined #salt
21:32 amckinley joined #salt
21:34 rawzone joined #salt
21:35 ClausA joined #salt
21:35 Sypher joined #salt
21:36 cshuman joined #salt
21:36 echos joined #salt
21:36 _0x4b joined #salt
21:36 Jarus joined #salt
21:36 rmt2 joined #salt
21:36 ede joined #salt
21:37 dthom91 joined #salt
21:38 woebtz joined #salt
21:39 uta joined #salt
21:41 budmang joined #salt
21:43 LucasCozy joined #salt
21:43 josephholsten joined #salt
21:46 FL1SK joined #salt
21:47 bogdanbu joined #salt
21:47 bogdanbu aantony: i think i found the problem :)
21:49 bogdanbu aantony: i used the same id for file.recurse: - … and file.managed: -…
21:49 bogdanbu aantony: i made two states with different ids and is ok …
21:52 amckinley joined #salt
21:55 zwe joined #salt
21:59 pentabular joined #salt
22:05 qba73 joined #salt
22:08 cdsrv joined #salt
22:10 zooz joined #salt
22:10 Marion joined #salt
22:10 mianos joined #salt
22:10 kermit joined #salt
22:14 SEJeff_work Why not semver?
22:14 SEJeff_work Just curious why salt is moving to the new version number setup
22:18 BRYANT__ joined #salt
22:19 DanGarthwaite joined #salt
22:21 DanGarthwaite I would like to pass a dict into a template with aggregated information from all minions.  So the aggregation step should occur on the master, but the file.managed rendering on a certain minion.  Why can't I get my head around where/how this is to be broken up?
22:21 alunduil joined #salt
22:22 DanGarthwaite Custom state or module code would be executed on the minion, right, so no aggregation without peer.publish shenanigans.
22:22 cwright DanGarthwaite: have you looked into using Salt Mine?
22:22 DanGarthwaite yes.  A lot.  :)  This is not for that.
22:23 DanGarthwaite I'm aggregating all the default gateways of all minions for use in a nagios template.
22:25 drags DanGarthwaite: I believe the Mine is for this.. write a custom module that just returns the data you're looking for, have all nodes send that data into the mine, then have the nagios host pull from the mine
22:25 cshuman_ joined #salt
22:25 DanGarthwaite It would work, but I want this on demand, not periodically.
22:25 jkleckner joined #salt
22:26 DanGarthwaite I'll do it and report back.
22:28 ccase joined #salt
22:36 josephholsten joined #salt
22:37 cdsrv "those aren't the data's you're looking for"
22:37 cdsrv ;)
22:37 DanGarthwaite SaltMine is still documented sparsely.
22:37 DanGarthwaite https://gist.github.com/dangarthwaite/b17dd245715c44207211
22:38 DanGarthwaite Is there a way to set the name of the key in mine_functions?
22:38 DanGarthwaite How do you specify more than one 'cmd.run' functions?
22:42 druonysuse joined #salt
22:46 anveo DanGarthwaite do you mean to want to run sequential commands?
22:46 DanGarthwaite Did you see the gist?
22:47 DanGarthwaite How do you recall the output of the third defined function?     salt lead mine.get lead ____________
22:47 DanGarthwaite cmd.run?
22:48 drags DanGarthwaite: yeah, you can't get argumented module calls from the mine (yet?)
22:48 drags you can wrap that in a custom module to give it a unique name
22:50 DanGarthwaite So like so: https://groups.google.com/d/topic/salt-users/N011ydFYU-A/discussion
22:50 dthom91 joined #salt
22:53 tyler-baker joined #salt
22:53 tyler-baker joined #salt
22:55 cbloss I there an issue with the salt-formula using 0.17? One of my minions keeps losing the pillar data
23:02 xmltok joined #salt
23:05 amckinley joined #salt
23:10 mmilano joined #salt
23:12 jslatts joined #salt
23:15 punal joined #salt
23:15 bhosmer joined #salt
23:16 amckinley joined #salt
23:19 StDiluted joined #salt
23:24 amahon joined #salt
23:25 druonysus joined #salt
23:36 troyready joined #salt
23:49 KyleG1 joined #salt
23:55 shinylasers joined #salt

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