Time |
Nick |
Message |
00:03 |
TheCodeAssassin |
hehe oke UtahDave, ill be back later. its 2 am here |
00:03 |
TheCodeAssassin |
im a bit tired also hehe |
00:04 |
UtahDave |
ok, night, man |
00:10 |
|
teebes joined #salt |
00:12 |
|
cxz joined #salt |
00:15 |
|
mwillhite joined #salt |
00:19 |
|
gatorall1 joined #salt |
00:24 |
|
stanchan joined #salt |
00:26 |
|
krissaxton joined #salt |
00:34 |
|
kstaken joined #salt |
00:35 |
|
kermit joined #salt |
00:38 |
|
SEJeff joined #salt |
00:39 |
|
thehyperadvisor joined #salt |
00:39 |
thehyperadvisor |
hi there. just wondering if someone knows how to change the default PATH for salt when trying to run remote excutions |
00:43 |
|
DavidEzek joined #salt |
00:43 |
|
whit joined #salt |
00:44 |
DavidEzek |
Hi, I'm having an issue with the CLI throwing "salt.exceptions.SaltReqTimeoutError: Waited 60 seconds... |
00:44 |
DavidEzek |
http://pastebin.com/VsFCi6RG |
00:44 |
DavidEzek |
Looking through the code, the timeout does not appear to be configurable. |
00:45 |
|
auser joined #salt |
00:47 |
|
Gifflen joined #salt |
00:48 |
DavidEzek |
Using Salt 0.15.3 and ZMQ 2.1.11. |
00:48 |
DavidEzek |
Should I open an issue for this? The only related issue I see is https://github.com/saltstack/salt/issues/5730. |
00:49 |
gatorall1 |
can I have salt's generated minion/master keys be used for http authentication with ssl? |
00:51 |
|
avienu joined #salt |
00:52 |
|
oz_akan_ joined #salt |
00:55 |
|
brianhicks joined #salt |
00:56 |
|
dthom91 joined #salt |
01:09 |
|
auser joined #salt |
01:12 |
|
ammmir joined #salt |
01:13 |
|
renothing joined #salt |
01:14 |
|
juanlittledevil left #salt |
01:16 |
|
Superdawg joined #salt |
01:16 |
|
Nexpro1 joined #salt |
01:20 |
|
renothing joined #salt |
01:20 |
|
UtahDave joined #salt |
01:24 |
|
oz_akan_ joined #salt |
01:24 |
|
cxz joined #salt |
01:27 |
|
krissaxton joined #salt |
01:30 |
|
liuyq joined #salt |
01:35 |
|
druonysus joined #salt |
01:37 |
|
benno joined #salt |
01:47 |
|
liuyq joined #salt |
01:52 |
|
falican joined #salt |
02:04 |
|
sixninetynine joined #salt |
02:05 |
|
Furao joined #salt |
02:11 |
|
mwillhite joined #salt |
02:12 |
mwillhite |
Hi - brand new to salt…how can I debug a "Recursive requisite found" error…? |
02:12 |
mwillhite |
or any error for that matter…other than the state I don't get a lot of feedback, I do have the verbose messaging turned on in the Vagrantfile |
02:13 |
mwillhite |
the real problem I'm trying to fix is a failed gem install, but I don't get any info as to why its failing |
02:13 |
|
rberger joined #salt |
02:14 |
|
zian joined #salt |
02:17 |
|
mgw joined #salt |
02:18 |
zian |
hi guys. when attempting to set up a masterless salt config on an Ubuntu vagrant machine, I'm running the following: salt-call --local state.highstate -l debug but am receiving the following error: No Top file or external nodes data matches found. my state tree looks like this on the vagrant machine: /srv/salt/roots/salt/top.sls |
02:18 |
|
mgw joined #salt |
02:19 |
cedwards |
mwillhite: sounds like an error in your .sls. can you post a sanitized version of what you're applying? |
02:19 |
zian |
when I move top.sls (along with all the other sls files it uses) up one directory in the tree, everything works fine |
02:19 |
zian |
it's like salt-call can't find top |
02:19 |
zian |
my top.sls is an almost verbatim copy of the nginx example in the salt docs, with the addition of the vim package |
02:20 |
zian |
base: '*': - nginx.nginx - requirements.essential |
02:20 |
cedwards |
zian: i would expect it should work one directory up. unless you update file_roots to include your /srv/salt/roots path |
02:20 |
zian |
could you explain the difference between roots and file_roots? and where does one set the file_roots path |
02:20 |
mwillhite |
cedwards: take a look at this http://pastebin.com/rHtgdGNJ |
02:20 |
cedwards |
zian: file_roots is defined in the master.conf and defaults to /srv/salt/ |
02:21 |
mwillhite |
this is what I would *like* to do |
02:21 |
terminalmage |
eightyeight: added this earlier today: http://docs.saltstack.com/ref/states/providers.html#setting-a-provider-in-the-minion-config-file |
02:21 |
mwillhite |
minus the runas in the bundler state |
02:21 |
mwillhite |
without the runas…the ruby installs correctly, but bundler fails to install |
02:21 |
mwillhite |
I can see that the gemset is created as well |
02:23 |
mwillhite |
when I use the runas call I get the recursive requisite error |
02:23 |
mwillhite |
I thought maybe the gem wasn't installing with the correct permissions or something… |
02:24 |
|
xl1 joined #salt |
02:27 |
cedwards |
mwillhite: there's nothing really standing out to me in your paste, but i could be missing it |
02:27 |
mwillhite |
any ideas as to why the gem wouldn't install? |
02:27 |
mwillhite |
the only message I get is Gem failed to install |
02:27 |
mwillhite |
or something like that |
02:27 |
mwillhite |
yet I can install it just fine when I ssh into the box |
02:27 |
mwillhite |
under root of course…because rvm is installed as root |
02:27 |
|
krissaxton joined #salt |
02:31 |
cedwards |
i'm not very familiar with the rvm/gem modules, sorry |
02:31 |
cedwards |
if nobody else pops up you might try the mailing list |
02:34 |
UtahDave |
mwillhite: Your user_central stanza is requiring itself |
02:34 |
|
oz_akan_ joined #salt |
02:34 |
mwillhite |
ah! in user-central-db right? |
02:34 |
UtahDave |
zian: your top.sls should go here by default: /srv/salt/top.sls |
02:35 |
UtahDave |
Salt will not recurse down to try to find it in another location |
02:36 |
mwillhite |
UtahDave actually I'm not seeing it :/ |
02:36 |
|
malinoff joined #salt |
02:36 |
zian |
UtahDave: is there a way to tell salt I'd like to change the location? in many of the sample state stacks I've studied on git the srv/salt directory usually contains the directories salt and pillar, then within salt the top.sls file and all the sls files it uses |
02:37 |
UtahDave |
zian: yeah, those are examples. |
02:37 |
UtahDave |
zian: there's the file_roots: config option in /etc/salt/master |
02:38 |
UtahDave |
but you can only have one top.sls in your base environment. If you're going to use more than one state from those examples, then you have to merge the top file examples |
02:38 |
UtahDave |
mwillhite: So is that not a complete sls you pasted? |
02:38 |
mwillhite |
thats the whole thing |
02:39 |
mwillhite |
oh, minus one stanza |
02:39 |
mwillhite |
it isn't related to the others though |
02:39 |
mwillhite |
I mean…it never gets that far because bundler needs to come first |
02:40 |
mwillhite |
here is the whole thing: http://pastebin.com/XkAHH3bT |
02:41 |
|
whit joined #salt |
02:44 |
mwillhite |
but…removing any notion of a user from the packages/states…I still can't get this gem to install and I have no idea why |
02:48 |
jacksontj |
UtahDave: is there an internal mechanism to schedule a job from within the master code? |
02:52 |
UtahDave |
jacksontj: I think you have to add jobs to the scheduler through the master config. Seems like a great idea to enable that programmaticly though |
02:53 |
jacksontj |
yea, i was thinking for the gitfs module it should have a cleanup job which removes hash cache entries which no longer exist |
02:56 |
mwillhite |
UtahDave any guesses as to why Bundler wouldn't install with RVM? |
02:57 |
UtahDave |
jacksontj: yeah. how does gitfs handle things like that/ |
02:57 |
jacksontj |
the gitfs module doesnt ;) |
02:58 |
* UtahDave |
reading mwillhite's pastebin |
02:58 |
jacksontj |
so the hash cache (currently) grows unbounded as far as i can tell |
02:58 |
jacksontj |
assuming you change/add/remove files |
02:58 |
jacksontj |
just orphans |
02:59 |
UtahDave |
lovely! |
02:59 |
jacksontj |
i was wanting to add something similar to the roots |
02:59 |
UtahDave |
mwillhite: Hey, I'm not sure. I know nothing about the ruby world. |
02:59 |
jacksontj |
half way through implementation i realized that it doesn't clean up |
03:00 |
mwillhite |
alright…well thanks for looking :P |
03:00 |
UtahDave |
mwillhite: I have to go put my girls to bed now. I'll see what I can figure out a little later tonight. |
03:00 |
mwillhite |
thanks I appreciate it |
03:01 |
mwillhite |
I'll be online tomorrow if you don't see me in here later…I'm -500 so its getting a little late :) |
03:11 |
|
auser joined #salt |
03:16 |
|
whit joined #salt |
03:16 |
UtahDave |
mwillhite: what OS are you on? |
03:16 |
mwillhite |
I'm using vagrant with precise64 |
03:17 |
mwillhite |
currently I'm running it to see if it'll install the gem outside of the gemset |
03:17 |
mwillhite |
if so that might be okay for now |
03:17 |
mwillhite |
this is just a development box with a bunch of apps that work together |
03:18 |
UtahDave |
ok, so first off user-central doesn't seem to be a package in the default ubuntu repositories |
03:18 |
mwillhite |
it is not, that is the first app I'm trying to get running |
03:18 |
mwillhite |
I have a user-central formula (which is what I pasted) |
03:19 |
mwillhite |
its a submodule under /vagrant/apps/ |
03:19 |
UtahDave |
mwillhite: Ok, lines 4 through 6 are going to fail, then. |
03:19 |
UtahDave |
can you see why? |
03:19 |
mwillhite |
does that make a call to apt-get or something? |
03:20 |
mwillhite |
I'm not really clear about how that works |
03:20 |
UtahDave |
yep. pkg.installed uses the OS package manager |
03:20 |
mwillhite |
ok |
03:20 |
UtahDave |
essentially on Ubuntu that's going to translate to sudo apt-get install user-central |
03:20 |
mwillhite |
thats okay…I can use the gem installation as a dependency |
03:20 |
mwillhite |
gotcha, that makes a lot of sense |
03:20 |
UtahDave |
and fail because user-central doesn't exist |
03:21 |
mwillhite |
so…then for example, if I want to make user-central-db require user-central-gems…what format would I use? |
03:21 |
mwillhite |
I assume I can't use 'pkg: user-central-gems' |
03:22 |
UtahDave |
mwillhite: also lines 12 and 13: |
03:22 |
UtahDave |
166.78.146.78 |
03:22 |
UtahDave |
ooops wrong paste |
03:22 |
UtahDave |
- require: - rvm: ruby-1.9.2-p290- require: - rvm: ruby-1.9.2-p290 |
03:22 |
UtahDave |
where is your rvm call that installs ruby-1.9.2-p290? |
03:22 |
mwillhite |
I'll paste in my rvm-formula |
03:23 |
UtahDave |
I thought you said that was your entire sls. lol |
03:23 |
mwillhite |
http://pastebin.com/KdyD0kek |
03:23 |
mwillhite |
well yeah for user-central, I have others :) |
03:24 |
UtahDave |
bam, there's your errors right there. Just like I thought originally! |
03:24 |
UtahDave |
mwillhite: see how after - bash on line 16 that there's another indentation level? |
03:24 |
mwillhite |
in the rvm sls? |
03:25 |
UtahDave |
yeah, your last pastebin |
03:25 |
mwillhite |
yes… |
03:25 |
UtahDave |
line 16 |
03:25 |
mwillhite |
- bash |
03:25 |
UtahDave |
- names can only take a list. |
03:25 |
UtahDave |
- coreutils is indented another level |
03:25 |
mwillhite |
oh, no thats a screwup in pastebin |
03:26 |
mwillhite |
this this out: http://cl.ly/image/1A020T272J16 |
03:26 |
UtahDave |
also, use two spaces for all indentations in yaml |
03:26 |
UtahDave |
are there any tabs in there? make sure they're all 2 spaces for each indent |
03:26 |
mwillhite |
hm…I copied this file…out of the web :( |
03:26 |
mwillhite |
I'll check it out |
03:26 |
UtahDave |
also, one thing that will speed up your sls |
03:27 |
mwillhite |
but the thing is…the rvm ruby installs just fine… |
03:27 |
UtahDave |
- names is really handy because each item below it gets that state applied to it separately. It's a shorthand notation |
03:27 |
mwillhite |
oh cool |
03:27 |
UtahDave |
but if you replace - names with - pkgs it will install each of those packages in one apt-get call instead of an apt-get call for each package |
03:28 |
|
NotreDev joined #salt |
03:28 |
mwillhite |
ah |
03:28 |
|
krissaxton joined #salt |
03:28 |
NotreDev |
howdy. is there an easy way to figure out non-obvious dependencies? |
03:28 |
NotreDev |
more of a theoretical problem than anything else... |
03:29 |
mwillhite |
so…bundler installs outside of the gemset just find (when I don't specify a gemset in the - ruby function) |
03:29 |
jacksontj |
NotreDev: try it and see what breaks ;) |
03:29 |
mwillhite |
but then my cmd doesn't work…probably because its entering the gemset (which is clean) when it tries to execute |
03:29 |
jacksontj |
do you mean of you app or of salt? or of a state file? |
03:30 |
UtahDave |
mwillhite: is a gemset kind of like a virtualenv? |
03:31 |
mwillhite |
oh sorry, its an rvm thing…it just isolates your gem installs so that you don't have dependency conflicts between separate applications |
03:32 |
mwillhite |
it literally just creates a subdirectory and stores your gems in there, then when you are in the directory with the referenced gemset (via .rvmrc) it looks at that directory for all of your gems |
03:34 |
UtahDave |
ok |
03:34 |
UtahDave |
back in a sec |
03:34 |
jacksontj |
UtahDave: looks like the update() method is called reguarly |
03:34 |
jacksontj |
so i can just use that :D |
03:35 |
|
jkleckner joined #salt |
03:36 |
UtahDave |
jacksontj: ah, cool. |
03:36 |
UtahDave |
mwillhite: OK, so what errors are you getting now? |
03:37 |
mwillhite |
I'm running again with putting it back in the gemset…its take a while to install all the rubies, so I'll get back to you |
03:37 |
mwillhite |
I fixed the whitespace stuff |
03:37 |
UtahDave |
ok, good |
03:41 |
mwillhite |
I have a feeling that the rvm.gemset_preset function doesn't quite work the way I'm expecting it to, so maybe bundler is trying to install into a gemset that doesn't exist |
03:41 |
mwillhite |
after this runs (and likely fails) I'll ssh into the box and see what gemsets exist |
03:42 |
UtahDave |
ok |
03:45 |
shadowsun |
yo. |
03:45 |
* shadowsun |
de-idles and looks around |
03:45 |
UtahDave |
yo, shadowsun |
03:46 |
shadowsun |
How goes |
03:47 |
jacksontj |
UtahDave: https://github.com/saltstack/salt/pull/6848 |
03:47 |
jacksontj |
:) |
03:47 |
|
StDiluted joined #salt |
03:50 |
jacksontj |
actually-- i can do one better, one sec |
03:51 |
UtahDave |
nice, jacksontj |
03:51 |
UtahDave |
good, shadowsun, you? |
03:56 |
|
xl1 joined #salt |
04:02 |
mwillhite |
UtahDave it failed, and as I suspected it was creating the gemset differently than I had thought…after reviewing I realized there is no way it could have done what I was expecting it to do |
04:02 |
|
sixninetynine joined #salt |
04:02 |
mwillhite |
so…made a naming adjustment and am running again |
04:03 |
NotreDev |
where could i find an example of salt using [json + yaml] for rendering? |
04:03 |
|
nonuby joined #salt |
04:04 |
UtahDave |
ok, let me know how it goes, mwillhite |
04:04 |
UtahDave |
NotreDev: both json and yaml? |
04:04 |
NotreDev |
nah, just json |
04:05 |
NotreDev |
i'm looking here http://docs.saltstack.com/ref/renderers/index.html#composing-renderers |
04:06 |
nonuby |
let say as part of a state/sls we add a kvm instance which requires the minion to do an assortment of things including a long running op such as downloading an iso from a mirror, as state are not executed concurrently this then blocks any other changes, how would one handle this? (experimenting with doing something like http://hveem.no/saltvirt) |
04:06 |
|
tyler-baker joined #salt |
04:07 |
UtahDave |
NotreDev: I don't have any examples handy. |
04:08 |
NotreDev |
I can't seem to find any in documentation, but I'm still searching through the google results. Any reason why it'd be a bad idea to use JSON vs YAML? (aside from all the examples currently being in YAML) |
04:08 |
UtahDave |
nonuby: regular executions are run concurrently. You could just kick off an execution to do the download separatley |
04:09 |
UtahDave |
Is there a reason you want to use json? |
04:09 |
UtahDave |
(I'm not trying to be contrarian.) :) |
04:10 |
nonuby |
So as part of my layering on top, i execute say a curl via remote execution, in order to ensure all pre-requistes are satisifed for states operation to run quickly |
04:10 |
nonuby |
hmm interesting |
04:11 |
NotreDev |
UtahDave: thought it might be cool to store it in MongoDB… pet project per se |
04:11 |
NotreDev |
json -> json seems interesting |
04:12 |
UtahDave |
NotreDev: json is valid yaml |
04:12 |
NotreDev |
well, json can be converted to yaml, but it's semantically different |
04:12 |
NotreDev |
but if something is written in yaml, it can't necessarily be converted to json |
04:14 |
UtahDave |
yaml parsers will load json just fine |
04:14 |
UtahDave |
but anyway, let me see if I can find you an example |
04:14 |
|
renothing joined #salt |
04:16 |
jesusaurus |
json is a subset of yaml |
04:17 |
jesusaurus |
or more precisely, yaml is a superset of json because ingy expanded on json when designing yaml |
04:19 |
UtahDave |
NotreDev: OK, this should work: http://pastebin.com/YTRWVH1c |
04:20 |
UtahDave |
NotreDev: I haven't tested this, but I think that will get you going |
04:20 |
UtahDave |
jesusaurus: Thanks for the clarification! I knew it was something like that. |
04:20 |
jesusaurus |
actually, it looks like it wasnt an actual superset until yaml 1.2 |
04:21 |
jesusaurus |
http://www.yaml.org/spec/1.2/spec.html |
04:21 |
jesusaurus |
paragraph2, sentence1 |
04:22 |
UtahDave |
NotreDev: Have you tried that example yet? |
04:22 |
NotreDev |
UtahDave: taking a look at it now |
04:22 |
NotreDev |
thanks for providing it. |
04:23 |
UtahDave |
you're welcome! I'm interested in how your mongodb experiment works out. |
04:24 |
UtahDave |
It would be cool to have a mongo_fs as a storage backend like gitfs or the new hgfs |
04:24 |
UtahDave |
storing all the states in mongo would be quite interesting |
04:24 |
|
kiorky joined #salt |
04:24 |
|
twinshadow joined #salt |
04:25 |
|
crazysim joined #salt |
04:25 |
|
retr0h joined #salt |
04:25 |
|
retr0h joined #salt |
04:28 |
|
krissaxton joined #salt |
04:36 |
nonuby |
http://docs.saltstack.com/ref/states/writing.html#return-data - i presume I can return more keys should I wish if I write a custom state module? |
04:40 |
|
mwillhite joined #salt |
04:42 |
|
Jahkeup joined #salt |
04:43 |
mwillhite |
UtahDave that did it…love and hate the stupid mistakes ;) |
04:45 |
mwillhite |
but now I've got something else…although the log says the gem was installed I can't find it :) |
04:45 |
|
lazyguru joined #salt |
04:45 |
mwillhite |
thats a problem for tomorrow though, thanks for the help |
04:49 |
|
Katafalkas joined #salt |
04:50 |
|
isomorphic joined #salt |
04:56 |
|
xl1 joined #salt |
05:02 |
|
auser joined #salt |
05:04 |
UtahDave |
nonuby: no, it has to be that data structure. |
05:05 |
UtahDave |
mwillhite: :) ok, I'm glad you're making progress. Sorry I'm not more help with the ruby stuff |
05:08 |
|
jalbretsen joined #salt |
05:08 |
mwillhite |
one last question UtahDave…rvm seems to installed under the vagrant user |
05:08 |
UtahDave |
what user is Salt running as? |
05:09 |
mwillhite |
if I want to run my cmd under vagrant, I just need to add - runas: vagrant? |
05:09 |
UtahDave |
yeah, I think so |
05:09 |
mwillhite |
I don't know…I'm not doing any user management… |
05:09 |
mwillhite |
oh sorry, I forget to separate vagrant from salt sometimes |
05:09 |
mwillhite |
I'm learning them both at the sametime |
05:10 |
UtahDave |
Yeah. My problem is that I never use vagrant or Salty Vagrant. I need to sit down one of these days and get that all figured out |
05:11 |
mwillhite |
this setup is going to be fantastic once I get it all going |
05:11 |
mwillhite |
this company has a ton of disjointed apps that all depend on each other and are really difficult to setup |
05:11 |
mwillhite |
so once this is put together, they should be able to onboard new developers in less than 15 minutes |
05:11 |
Jahkeup |
UtahDave: salty-vagrant is the best ;) |
05:11 |
mwillhite |
er…at least get their dev environments setup |
05:12 |
UtahDave |
mwillhite: that will be awesome! |
05:12 |
akoumjian |
Yeah UtahDave , the best. :-) |
05:12 |
akoumjian |
Actually, I am waiting to get a not busy 2 hours so I can finally fix all the open issues and get the next release out. |
05:12 |
UtahDave |
:) Some incredibly brilliant developer must have put salt-vagrant together. |
05:13 |
* akoumjian |
would feel guilty if he had a spare moment |
05:13 |
UtahDave |
nice, akoumjian |
05:13 |
Jahkeup |
heh |
05:14 |
akoumjian |
I've been holding back because I thought Mitch would have merged it into core by now, and I could just update from a Vagrant fork. But that doesn't look to be happening any time, well, ever. |
05:14 |
UtahDave |
lol, same here. I've been up until 2 or 3 am the last 5 nights straight. I'm forcing myself to go to be at 11:30 tonight |
05:14 |
UtahDave |
akoumjian: why's that? |
05:14 |
akoumjian |
UtahDave: Who knows. I guess Mitch is as busy as we are. But he could have just merged it in. |
05:15 |
akoumjian |
My next free two hours after that is going into spm. |
05:15 |
* EugeneKay |
puts a hot water bottle in UtahDave's bed |
05:16 |
UtahDave |
first some blankies, now a hot water bottle! thanks, EugeneKay :) |
05:17 |
UtahDave |
that's too bad he hasn't even given you any indication of his thoughts on it. Oh, well |
05:19 |
|
druonysus joined #salt |
05:29 |
|
krissaxton joined #salt |
05:30 |
|
renothing joined #salt |
05:30 |
|
mike25ro joined #salt |
05:35 |
|
suodla joined #salt |
05:37 |
|
david_a joined #salt |
05:39 |
|
kleinishere joined #salt |
05:43 |
|
Katafalkas joined #salt |
05:44 |
|
kleinish_ joined #salt |
05:47 |
|
mattbarto joined #salt |
05:49 |
|
cxz joined #salt |
05:55 |
|
ds_shadof joined #salt |
05:55 |
ds_shadof |
when i ask salt to run command on specific node, does salt send it to all nodes like mcollective? |
05:58 |
UtahDave |
ds_shadof: It's a pub/sub interface. the minions all listen on the pub and then take action if the command is for them |
06:03 |
|
cxz joined #salt |
06:03 |
|
kleinishere joined #salt |
06:08 |
|
Fengyu joined #salt |
06:12 |
|
fengyu joined #salt |
06:14 |
|
rberger joined #salt |
06:18 |
|
agend joined #salt |
06:20 |
|
Furao joined #salt |
06:21 |
|
malinoff joined #salt |
06:21 |
|
Furao left #salt |
06:24 |
|
matanya joined #salt |
06:24 |
|
kleinish_ joined #salt |
06:30 |
|
krissaxton joined #salt |
06:47 |
|
vaxholm_ joined #salt |
06:52 |
|
malinoff joined #salt |
06:54 |
|
balboah_ joined #salt |
07:01 |
|
ml_1 joined #salt |
07:09 |
|
druonysuse joined #salt |
07:09 |
|
druonysuse joined #salt |
07:12 |
|
tyler-baker joined #salt |
07:15 |
|
linjan_ joined #salt |
07:16 |
|
gatoralli joined #salt |
07:17 |
|
gatorall1 joined #salt |
07:35 |
|
arapaho joined #salt |
07:40 |
|
suodla_ joined #salt |
07:44 |
|
david_a joined #salt |
07:51 |
|
ricicle joined #salt |
07:52 |
|
az87c joined #salt |
07:54 |
|
Katafalkas joined #salt |
07:58 |
|
scott_w joined #salt |
08:01 |
|
Xeago joined #salt |
08:02 |
|
Lue_4911 joined #salt |
08:16 |
|
zooz joined #salt |
08:21 |
|
gildegoma joined #salt |
08:37 |
s1acky |
I have a stupid question. What if someone will hack my DNS server or find other way to replace my master server. Will minions listen to new server? |
08:41 |
EugeneKay |
No. |
08:42 |
EugeneKay |
The masters and minions both have keys to identify themselves to the other side |
08:42 |
jcristau |
they cache the server's key |
08:44 |
drags |
is there any way to ban a command from the master, even as root? (shy of editing source).. for example 'system.shutdown' |
08:50 |
|
ricicle joined #salt |
08:55 |
xmj |
Access Control Levels |
08:56 |
|
ricicle joined #salt |
09:02 |
|
it_dude joined #salt |
09:06 |
|
bemehow joined #salt |
09:14 |
|
Nexpro1 joined #salt |
09:20 |
|
xl2 joined #salt |
09:20 |
|
lazyguru joined #salt |
09:23 |
|
Furao joined #salt |
09:25 |
|
bhosmer joined #salt |
09:29 |
|
renothing joined #salt |
09:29 |
|
xl1 joined #salt |
09:30 |
|
waverider joined #salt |
09:30 |
|
waverider joined #salt |
09:42 |
|
fredvd joined #salt |
09:49 |
|
agend joined #salt |
09:58 |
|
ricicle joined #salt |
09:58 |
|
Xeago joined #salt |
10:03 |
|
felixhummel joined #salt |
10:35 |
|
tuxIO joined #salt |
10:38 |
|
lemao joined #salt |
10:48 |
|
NotreDev joined #salt |
10:49 |
|
platforms joined #salt |
10:50 |
|
zwevans joined #salt |
10:50 |
|
coolj joined #salt |
10:51 |
|
codysoyland joined #salt |
10:51 |
|
robinsmidsrod joined #salt |
11:02 |
|
giantlock joined #salt |
11:06 |
|
TJ1980 joined #salt |
11:07 |
|
MrTango joined #salt |
11:13 |
|
tuxIO_ joined #salt |
11:20 |
|
tuxIO joined #salt |
11:21 |
|
waverider_ joined #salt |
11:25 |
|
TJ1980 joined #salt |
11:26 |
|
Nexpro1 joined #salt |
11:27 |
|
carlos joined #salt |
11:35 |
|
aat joined #salt |
11:47 |
|
ahale_ joined #salt |
11:47 |
|
TJ1980 joined #salt |
11:50 |
|
renothing joined #salt |
11:56 |
|
aat joined #salt |
11:57 |
|
LucasCozy joined #salt |
11:59 |
|
blee joined #salt |
12:09 |
|
TJ1980 joined #salt |
12:20 |
|
teebes joined #salt |
12:23 |
|
oz_akan_ joined #salt |
12:25 |
|
oz_akan_ joined #salt |
12:31 |
|
fredvd joined #salt |
12:34 |
|
ricicle joined #salt |
12:39 |
|
LotusBleu joined #salt |
12:39 |
LotusBleu |
hi |
12:41 |
LotusBleu |
I'm failled to install a Salt-Minion-0.16.2-AMD64-Setup on a Win2008R2: the service is paused and I cannot start it (Microsoft Visual C++ 2008 redistribuable is installed as seggested in the manual) |
12:47 |
LotusBleu |
install is working with Salt-Minion-0.16.2-AMD64-Setup NOT Salt-Minion-0.16.3-AMD64-Setup |
12:47 |
|
fredvd joined #salt |
13:03 |
|
danielbachhuber joined #salt |
13:05 |
|
kenbolton joined #salt |
13:07 |
|
tyler-baker joined #salt |
13:10 |
|
mgw joined #salt |
13:11 |
|
blee joined #salt |
13:13 |
|
jkleckner joined #salt |
13:14 |
|
ml_11 joined #salt |
13:15 |
gamingrobot |
have you tried running the salt minion via the command line? |
13:16 |
|
whit joined #salt |
13:16 |
s1acky |
no. i like clie) |
13:16 |
s1acky |
*cli |
13:17 |
s1acky |
ops. my mistake. read it wrong. I have to sleep more) |
13:22 |
|
kenbolton joined #salt |
13:22 |
|
mgw joined #salt |
13:23 |
|
aat joined #salt |
13:24 |
|
gldnspud joined #salt |
13:25 |
|
mwillhite joined #salt |
13:25 |
|
juicer2 joined #salt |
13:25 |
|
racooper joined #salt |
13:26 |
|
cedwards joined #salt |
13:27 |
|
ipmb joined #salt |
13:29 |
|
carlos joined #salt |
13:30 |
|
toastedpenguin joined #salt |
13:31 |
gamingrobot |
I would like to help patch some of these bugs, but salt is so large I dont even know where to start |
13:33 |
|
aat joined #salt |
13:34 |
joehh |
LotusBleu: I'm pretty sure 0.16.3 has worked for us, though it may have been as an upgrade to 0.16.2 |
13:34 |
joehh |
does it work if you install .2 first then .3 afterwards? |
13:34 |
|
titubear joined #salt |
13:36 |
|
anteaya joined #salt |
13:37 |
LotusBleu |
just tried joehh : yes it's works installing .3 over .2 |
13:38 |
|
Gifflen joined #salt |
13:38 |
LotusBleu |
I'm wordering why I missing on windows de grains.item virtual like on linux ? |
13:43 |
|
StDiluted joined #salt |
13:46 |
|
benno joined #salt |
13:46 |
|
cron0 joined #salt |
13:46 |
|
swa joined #salt |
13:49 |
|
Kholloway joined #salt |
13:51 |
|
mgw joined #salt |
13:52 |
LotusBleu |
I'm missing some grains items on windows: pythonversion, saktpath, saltversion, server_id, virtual |
13:59 |
|
bemehow joined #salt |
14:03 |
joehh |
I've got them on 0.16.0 (the only windows minions I have handy) |
14:04 |
|
mannyt joined #salt |
14:05 |
|
mannyt joined #salt |
14:08 |
|
kenbolton joined #salt |
14:09 |
|
carlos joined #salt |
14:12 |
|
ipmb joined #salt |
14:16 |
|
waverider joined #salt |
14:19 |
|
m_george|away joined #salt |
14:21 |
|
opapo joined #salt |
14:23 |
|
test123321 joined #salt |
14:23 |
MTecknology |
there's test.ping for seeing if a host responds, but isn't there something else for seeing if the host is down? |
14:25 |
|
wrale joined #salt |
14:25 |
tempspace |
Is there a way to completely activate a virtualenv before a pip install rather than using the pip from the virtualenv? |
14:25 |
|
kermit joined #salt |
14:25 |
wrale |
anyone using cobbler with salt? |
14:25 |
MTecknology |
cobbler: http://salt.readthedocs.org/en/latest/ref/pillar/all/salt.pillar.cobbler.html |
14:26 |
|
gatoralli joined #salt |
14:26 |
|
gatorall1 joined #salt |
14:26 |
wrale |
fantastic.. thanks |
14:26 |
|
iPaulo joined #salt |
14:28 |
|
mannyt_ joined #salt |
14:32 |
|
opapo joined #salt |
14:35 |
|
Jahkeup joined #salt |
14:38 |
|
micah_chatt joined #salt |
14:42 |
mwillhite |
I'm trying to get a version of ruby installed under rvm with salty-vagrant. I'm trying to install the ruby as the vagrant user, but it keeps failing: https://gist.github.com/miwillhite/6307517 |
14:42 |
mwillhite |
works fine when I don't specify runas…anyone have any ideas? |
14:44 |
mwillhite |
maybe because the original (not rvm) ruby, I am installing in root…… |
14:47 |
|
UtahDave joined #salt |
14:50 |
|
whiskybar joined #salt |
14:50 |
|
ckao joined #salt |
14:53 |
|
mgw joined #salt |
14:56 |
tempspace |
Does anybody know if there's a way to fully activate a virtualenv before installing requirements.txt or a pip? I have a pip that is trying to install to /opt and in a fully activated virtualenv, / is chroot'd to the virtualenv directory |
15:03 |
akoumjian |
tempspace: You can always just reference the virtualenved pip location directly. /opt/myenv/bin/pip install -r requirements.txt |
15:03 |
akoumjian |
I believe either the pip or virtualenv state modules will do this for you |
15:03 |
akoumjian |
By specifying "bin_env" or some similar option |
15:04 |
|
aat joined #salt |
15:04 |
tempspace |
akoumjian: I did that, it didn't work |
15:04 |
tempspace |
akoumjian: It appears to just run pip from the virtualenv directory when you do that |
15:05 |
tempspace |
rather than activating the virtualenv |
15:05 |
tempspace |
same with specifiying requirements: to the virtualenv state |
15:05 |
akoumjian |
tempspace: "activating" a virtualenv, just aliases the virtualenv's python directories as preferential to your system python paths |
15:06 |
akoumjian |
tempspace: I do this all the time. It should work to just to call the binary directly, it will install it into the virtualenv's packages directory |
15:06 |
tempspace |
akoumjian: I do it with lots of pips as well, but this one is problematic |
15:07 |
akoumjian |
Then later, if you want to run something from inside your virtualenv, you need to run it via the python in the virtualenv as well "/opt/myenv/python mycommand.py" |
15:07 |
tempspace |
akoumjian: https://groups.google.com/forum/#!topic/salt-users/dvyhuMly0jY |
15:07 |
|
DredTiger joined #salt |
15:08 |
tempspace |
akoumjian: If I manually activate the virtualenv and pip install, it creates opt as a directory within my virtualenv |
15:09 |
akoumjian |
It looks like you've got a simple permissions error |
15:10 |
tempspace |
akoumjian: That's not really the problem |
15:10 |
akoumjian |
so you're saying it's confusing that it tries to install in opt at all |
15:10 |
tempspace |
akoumjian: If I install pip as the user I'm specifying in the state within the virtualenv, it installs no problem |
15:10 |
tempspace |
it's only a problem with the problem that salt is doing on it |
15:12 |
Ahlee |
Salt's external_auth is only checked at start, isn't it |
15:12 |
|
forrest joined #salt |
15:12 |
|
jalbretsen joined #salt |
15:13 |
|
mako_reactor joined #salt |
15:16 |
akoumjian |
tempspace: So it seems like it's using the correct user, but not the correct bin_env? That's my best guess |
15:16 |
mwillhite |
ok still failing to install ruby with rvm with a user other than root |
15:17 |
tempspace |
akoumjian: I ended up working around it by using a cmd.script call that runs a shell script that activates the virtualenv and then runs pip install |
15:17 |
|
chrisgilmerproj joined #salt |
15:17 |
mwillhite |
here are the relevant sls files (in whole) and the error: https://gist.github.com/miwillhite/6307517 |
15:17 |
* akoumjian |
nods |
15:17 |
mwillhite |
any help will be greatly appreciated!! |
15:17 |
mwillhite |
using vagrant/salt |
15:17 |
mwillhite |
er salty-vagrant rather |
15:17 |
|
wyrd1 joined #salt |
15:17 |
akoumjian |
tempspace: I have one where I just do a cmd.run which does /var/www/env/bin/pip install … and runas 'user' |
15:18 |
mwillhite |
it works when I don't specify the user, but then I can't run my commands outside of root and the gems don't seem to get placed correctly |
15:18 |
akoumjian |
mwillhite: Have you gotten it to work installing it manually by hand first? |
15:19 |
mwillhite |
no I haven't tried that |
15:19 |
mwillhite |
I'll give that a shot right now |
15:19 |
tempspace |
mwillhite: It looks like your apt is broken |
15:19 |
tempspace |
mwillhite: Looks like this might be the problem 'Note: root's PATH should usually contain /usr/local/sbin, /usr/sbin and /sbi' |
15:21 |
mwillhite |
tempspace when I echo $PATH from root I have everything except /sbi |
15:21 |
LotusBleu |
joehh, minion .3 is working on 2012 |
15:21 |
|
titubear joined #salt |
15:22 |
tempspace |
akoumjian: I think I found the issue, I looked at the source of the pip I'm installing |
15:22 |
|
StDiluted joined #salt |
15:22 |
tempspace |
akoumjian: base_path = '%s/opt/newrelic_plugin_agent' % os.getenv('VIRTUAL_ENV', '') |
15:22 |
akoumjian |
tempspace: yeah, that's silly |
15:22 |
akoumjian |
tempspace: Time to file a bug with newrelic :-) |
15:24 |
LotusBleu |
joehh, and still missing same grains on 2012 |
15:24 |
|
kaptk2 joined #salt |
15:25 |
|
Jahkeup left #salt |
15:25 |
|
Jahkeup joined #salt |
15:25 |
mwillhite |
akoumjian it worked installing it locally /home/vagrant/.rvm/src/ruby-1.9.2-p290 |
15:25 |
mwillhite |
manually, I mean |
15:26 |
akoumjian |
mwillhite: Now your only task is to get salt to exactly replicate the steps you took to do it manually. |
15:26 |
mwillhite |
so…just run a cmd instead of using state.rvm |
15:28 |
StDiluted |
morning |
15:29 |
|
abe_music joined #salt |
15:29 |
mwillhite |
I just don't get why it fails under the user…but is fine as root |
15:32 |
|
jacksontj joined #salt |
15:39 |
|
MrTango joined #salt |
15:40 |
|
Ryan_Lane joined #salt |
15:43 |
|
littleidea joined #salt |
15:45 |
|
teebes joined #salt |
15:47 |
|
pdayton joined #salt |
15:47 |
|
devinus joined #salt |
15:55 |
|
Jahkeup joined #salt |
16:01 |
|
lazyguru joined #salt |
16:01 |
|
teskew joined #salt |
16:04 |
|
SunSparc joined #salt |
16:07 |
tempspace |
ls |
16:07 |
tempspace |
rm -rf * |
16:07 |
tempspace |
dang, isn't working |
16:07 |
tempspace |
</fail> |
16:08 |
abe_music |
are there any util methods available to help me parse and interpret the results of a highstate? don't want reinvent the wheel here if salt already provides some useful tool(s) |
16:11 |
|
troyready joined #salt |
16:19 |
jacob_s |
The output is in yaml abe |
16:20 |
SEJeff_work |
abe_music, Via the api or stdout? |
16:21 |
abe_music |
SEJeff_work, stdout…i know i can set it to return in yaml and i can load that up to a dictionary, but was wondering if there was something that would help me interpret those results to show to a user |
16:21 |
SEJeff_work |
abe_music, nope |
16:22 |
SEJeff_work |
Why not use a returner and setup a web frontent to view highstate runs stored in the job cache? |
16:22 |
|
issackelly left #salt |
16:23 |
abe_music |
SEJeff_work: cool, that's fine…that's a good idea. we already have the front end that's allowing users to manage clusters using salt, so a returner is probably the best way to go about it |
16:23 |
abe_music |
right now i'm saving all the logs of a highstate for historical reasons, and if there are issues i need to pinpoint those issues for the user in a clear way so they know what went wrong |
16:24 |
abe_music |
using celery to handle all the long-running salt and salt-cloud tasks |
16:28 |
LotusBleu |
joehh, I got on 2012 : osrelease: post2008Server |
16:28 |
forrest |
Is anyone else watching the puppetconf 2013 stream? |
16:33 |
jacob_s |
anything good? |
16:34 |
|
kstaken joined #salt |
16:36 |
|
Kholloway joined #salt |
16:37 |
|
carmony joined #salt |
16:38 |
|
arapaho joined #salt |
16:41 |
forrest |
I don't know I'm just listening to it. They're streaming it off a potato, so none of the slides are clear enough to read. |
16:43 |
z0rkito |
when writting a custom returner is there anyway to get which state file a change is coming from? |
16:43 |
SEJeff_work |
The id |
16:43 |
SEJeff_work |
or the name attribute |
16:43 |
SEJeff_work |
It is a giant struct / dictionary |
16:43 |
z0rkito |
yeah i'm not finding it in that dict. |
16:44 |
z0rkito |
i want to know the exact sls file that the command is coming from. |
16:44 |
SEJeff_work |
not possible |
16:44 |
z0rkito |
thats what i was finding, just wanted to verify. thanks. |
16:46 |
z0rkito |
is there a reason that data is not in the return? |
16:47 |
z0rkito |
would be useful for troubleshooting issues to know exactly where something is coming from. |
16:47 |
|
whit joined #salt |
16:47 |
|
alekibango joined #salt |
16:51 |
|
racooper joined #salt |
16:51 |
ipmb |
I have this in one of my states, but it looks like it fails on the first run (before postgres is installed) https://dpaste.de/9O6I8/#L3 |
16:52 |
ipmb |
I'm not sure how to solve it because it requires the postgresql service, but the rendering happens before that |
16:54 |
|
littleidea joined #salt |
16:57 |
|
rberger joined #salt |
16:59 |
|
carmony joined #salt |
17:04 |
|
KennethWilke joined #salt |
17:04 |
KennethWilke |
howdy |
17:09 |
|
pdayton left #salt |
17:10 |
|
benno joined #salt |
17:11 |
UtahDave |
howdy! |
17:13 |
|
littleidea joined #salt |
17:13 |
jacob_s |
I just tried to run a salt command through emacs built in eshell and it bricked emacs. |
17:14 |
|
jacksontj joined #salt |
17:14 |
UtahDave |
jacob_s: That's a feature! :) |
17:14 |
jacob_s |
good to know salt favors vim. |
17:15 |
UtahDave |
mwillhite: It's possible there's a bug in the ruby stuff |
17:15 |
|
cnelsonsic joined #salt |
17:16 |
mwillhite |
yeah UtahDave I've been struggling with it too long…going to see if I can ditch rvm altogether for this project |
17:17 |
|
whit joined #salt |
17:19 |
|
mgw joined #salt |
17:20 |
|
teebes joined #salt |
17:22 |
SunSparc |
I just set up two machines (with Vagrant/VirtualBox) and installed salt master on one, salt minion on the other. They are both running, but the minion does not seem to be able to see the master. |
17:22 |
SunSparc |
The /var/log/salt files are all blank. |
17:22 |
Ahlee |
anybody got a clever way of blocking until all hosts return? I.e., currently I hit the API to kick something off, then i poll once a second until all the nodes have completed |
17:22 |
SunSparc |
I can ping the machines from inside each of them. |
17:23 |
UtahDave |
SunSparc: Is the "master" config option set in /etc/salt/minion? |
17:24 |
SunSparc |
UtahDave: Yes, it is set to the hostname of the master, like so: master: salt-master |
17:24 |
Ahlee |
are ports 4505 and 4506 open? |
17:25 |
SunSparc |
Ahlee: They are not specifically blocked... I will have to check with netstat... |
17:25 |
|
alunduil joined #salt |
17:26 |
Ahlee |
SunSparc: i'd shut down the minion and launch it into the foreground with -l debug as well |
17:27 |
SunSparc |
Ahlee: Good idea. Looks like I can telnet from the minion to the master, but not the other way. Hmm.... |
17:37 |
SunSparc |
Ah, ok, it was trying to connect to opendns (who like to hijack unknown dns requests). So I specified the IP and hostname in /etc/hosts and now it is connecting. |
17:37 |
SunSparc |
Thanks Ahlee and UtahDave. |
17:38 |
|
teebes joined #salt |
17:39 |
|
JasonSwindle joined #salt |
17:39 |
JasonSwindle |
Howdy all |
17:39 |
KennethWilke |
JasonSwindle: howdy! |
17:40 |
JasonSwindle |
I have an odd question. |
17:40 |
|
devinus joined #salt |
17:40 |
JasonSwindle |
I am trying to remove repeats in my code, AKA going DRY. But, I am not sure I can do what I want to do. |
17:41 |
|
mattbarto joined #salt |
17:41 |
JasonSwindle |
I have JINJA that I use in about 6 to 7 files, that is always the same. |
17:41 |
JasonSwindle |
https://gist.github.com/anonymous/803fc5e85b8cac6836f7 |
17:41 |
JasonSwindle |
I would love to move this to pillar or somewhere so I only have to update it once and not 7 times. |
17:42 |
JasonSwindle |
Can you have JINJA as a value in pillar? |
17:42 |
KennethWilke |
this pillar sls files can be jinja |
17:42 |
|
giantlock joined #salt |
17:43 |
minaguib |
JasonSwindle: Does this fit the bill ? http://jinja.pocoo.org/docs/templates/#import |
17:43 |
JasonSwindle |
Let me look |
17:43 |
KennethWilke |
JasonSwindle, i used a jinja import for https://github.com/saltstack-formulas/salt-formula/tree/master/salt |
17:43 |
KennethWilke |
see the master.sls and package-map.jinja |
17:43 |
|
whit joined #salt |
17:44 |
JasonSwindle |
KennethWilke: Looking |
17:44 |
minaguib |
I think reading the jinja docs end-to-end should be required for anything past the "hello world" salt examples |
17:45 |
JasonSwindle |
The JINJA docs have helped a lot |
17:45 |
KennethWilke |
i certainly came to salt with some familiarity with jinja, but i would much prefer if mako was the more common templating language |
17:45 |
JasonSwindle |
KennethWilke: I think that will work! |
17:46 |
|
dthom91 joined #salt |
17:46 |
|
auser joined #salt |
17:46 |
KennethWilke |
JasonSwindle: it better! |
17:46 |
KennethWilke |
fyi, in the future once 0.17.0 has hit all the major distros i'll be using http://docs.saltstack.com/ref/modules/all/salt.modules.grains.html#salt.modules.grains.filter_by |
17:47 |
KennethWilke |
whiteinge made this handy function that'll make those files even cleaner, letting you cleanly declare defaults in-state while allowing pillar or other dict merging |
17:48 |
JasonSwindle |
Neat |
17:48 |
JasonSwindle |
Wow, that is going to be handy |
17:48 |
KennethWilke |
yeah its quite nice and clean |
17:49 |
KennethWilke |
my workflow with salt continues to improve too, you should contribute to the salt-formulas org :p |
17:50 |
whiteinge |
KennethWilke: i've been thinking about our conversation on how best to handle that for earlier salt versions. how about having a stand-alone execution module "formula-util" as a fallback instead of trying to recreate that functionality in jinja? |
17:50 |
whiteinge |
you would still have to distribute it to minions via the _modules directory |
17:50 |
whiteinge |
but it wouldn't have the danger of upgrading the oft-used grains.py module |
17:51 |
|
TJ1980 joined #salt |
17:51 |
whiteinge |
then the jinja would just have to check one or the other |
17:52 |
KennethWilke |
hmm, i dunno i need to let that stew over a bit more |
17:52 |
KennethWilke |
lol |
17:52 |
KennethWilke |
seems like it wouldn't be too long til 0.17.0 comes out and is packaged for the major distros, solving that problem |
17:54 |
KennethWilke |
gotta run to a meeting, be back in a few |
17:59 |
|
whit joined #salt |
18:00 |
Linuturk |
so, I'm having trouble integrating the salt-forumla repo into my salt setup |
18:01 |
Linuturk |
I've got it setup in the master file as a gitfs |
18:01 |
|
druonysus joined #salt |
18:01 |
|
druonysus joined #salt |
18:01 |
Linuturk |
I try to apply the default salt.master state against my master (including the appropriate pillar data in /srv/pillar/salt.sls from the example file) |
18:01 |
Linuturk |
and, I'm not seeing anything actually applied |
18:12 |
Linuturk |
http://pastebin.com/m7JDtJ9i << configs and pillar data |
18:13 |
|
m_george left #salt |
18:13 |
Linuturk |
am I crazy to think this should work? |
18:13 |
MTecknology |
*grumble* I'm fighting with salt-minion running fine on some boxes, but salt-master not being able to talk to them and the firewall not being at fault |
18:13 |
MTecknology |
always fun |
18:16 |
|
jacksontj joined #salt |
18:17 |
|
matanya joined #salt |
18:22 |
MTecknology |
I see this on the minion - http://dpaste.com/1353505/ The master apparently isn't getting the command to the minion |
18:26 |
|
grunge joined #salt |
18:26 |
|
kenbolton joined #salt |
18:27 |
|
littleidea joined #salt |
18:27 |
|
druonysus joined #salt |
18:28 |
zach |
Hey folks, can anyone help me out with this traceback? https://p.6core.net/p/TaT9MAlAGNRsOvuJ7EPLngQT -- Here's the perl.sls state file: https://p.6core.net/p/e3PtbPzJ2KZgDz6zGJc4uKBb -- taken from: https://github.com/saltstack/salt/issues/2130 |
18:29 |
mwillhite |
zach looks like it should be perl_mods instead? …could be talking out of my ass here |
18:30 |
zach |
mwillhite: I can try it. Not sure what will come of it, one moment. |
18:30 |
mwillhite |
well it looks like that was a typo |
18:30 |
mwillhite |
because perl_modules doesn't exist in that snippet |
18:30 |
zach |
same issue |
18:31 |
zach |
perl_modules is a pillar |
18:31 |
whiteinge |
Linuturk: sometimes gitfs can be a little hard to troubleshoot. try running ``cp.list_master`` on one of your minions to see if the formula .sls files are available |
18:32 |
zach |
Here is the pillar.items output: https://p.6core.net/p/A19p84Iq1LAA5cRYxcovBpeP |
18:33 |
zach |
I see. |
18:33 |
zach |
I'm blind |
18:33 |
zach |
;-) |
18:33 |
zach |
perl_mods not perl_modules *headdesk* |
18:34 |
|
qba73 joined #salt |
18:34 |
Linuturk |
whiteinge: found out i was missing the gitpython module |
18:34 |
whiteinge |
bah. that needs to be way more obvious than it is |
18:36 |
zach |
Now my only issue si this: "No function declared in state "cmd" in sls System.Packages.perl" - but I have cmd: - run blah already in place |
18:38 |
JasonSwindle |
Linuturk: You got it? |
18:40 |
Linuturk |
JasonSwindle: jaja |
18:40 |
JasonSwindle |
Linuturk: NVM. :) |
18:41 |
JasonSwindle |
Goood gooood |
18:43 |
|
zooz joined #salt |
18:45 |
|
jacksontj joined #salt |
18:47 |
zach |
fixed my issue - https://p.6core.net/p/KdAAPmytmpXzks5VAt0zaI0O |
18:57 |
SEJeff_work |
Corey, Can you do me a flavor please and join #graphite? |
18:59 |
|
dthom91 joined #salt |
19:02 |
auser |
hey all |
19:07 |
|
m_george|away joined #salt |
19:09 |
|
devinus joined #salt |
19:11 |
|
wrale left #salt |
19:11 |
JasonSwindle |
Hey auser |
19:11 |
|
jkleckner joined #salt |
19:17 |
|
mattbarto joined #salt |
19:17 |
|
zwevans joined #salt |
19:18 |
|
devinus joined #salt |
19:18 |
drags |
in the Salt ACL system, is there anyway to restrict certain commands instead of allow? (default allow + exlicity deny) |
19:18 |
drags |
also is there anyway to apply an ACL to 'root' on the salt master? |
19:19 |
|
gildegoma joined #salt |
19:19 |
|
m_george left #salt |
19:20 |
UtahDave |
drags: yeah, I think there's an acl black_list. |
19:22 |
|
jacksontj joined #salt |
19:23 |
|
jpeach joined #salt |
19:24 |
|
saurabhs joined #salt |
19:24 |
saurabhs |
Hi in minion log, what this WARNING refers to, am I missing some configuration or something: |
19:24 |
saurabhs |
[WARNING ] The function 'grains()' defined in '/usr/lib/python2.7/dist-packages/salt/loader.py' is not yet using the new 'default_path' argument to `salt.config.load_config()`. As such, the 'SALT_MINION_CONFIG' environment variable will be ignored |
19:25 |
UtahDave |
saurabhs: that's a minor bug that's been fixed. |
19:25 |
|
ipmb joined #salt |
19:27 |
|
Katafalkas joined #salt |
19:27 |
|
bhosmer joined #salt |
19:27 |
saurabhs |
cool. One more questions though. I am running my salt master with worker_threads:4 . So the file descriptor limit that (Default 1024) does it apply to all the processes for salt master or that limit is applicable to individual salt master process? |
19:28 |
|
it_dude joined #salt |
19:28 |
saurabhs |
I am seeing collectively all these processs of salt master are consuming more than 1300 FDs, although individual process only uses max of 700 |
19:29 |
drags |
huzzah, client_acl_blacklist: { modules: - system.poweroff } |
19:30 |
UtahDave |
saurabhs: I'm really not sure. basepi do you know the answer to saurabhs' question about the descriptor limit? |
19:30 |
drags |
most of my fleet is ephemeral VMs.. if they were to all shutdown at once it'd knock me offline for days :) |
19:31 |
mwillhite |
what does this mean: Unable to determine upstream hash of source file /vagrant/apps/user_central/config/database.yml.example |
19:32 |
mwillhite |
calling file.managed |
19:33 |
titubear |
how do I diagnose a require failure? -require: cmd: blah |
19:33 |
titubear |
is very much not working. and it takes me an entire reload to test it |
19:33 |
titubear |
from scratch |
19:33 |
titubear |
and I will blow up at some point from this :D |
19:37 |
titubear |
can a command just not require another command? |
19:38 |
MTecknology |
What are ports 4505 and 4506 for? |
19:38 |
MTecknology |
One is one thing, one is another thing? |
19:38 |
MTecknology |
My firewall guy wants to be able to label what they're each for specifically |
19:39 |
SEJeff_work |
MTecknology, It is in the master config file. 4505 == the publish port. 4506 = the return port |
19:40 |
SEJeff_work |
The master publishes commands and the minions return the results to the master |
19:40 |
MTecknology |
thanks :D |
19:40 |
SEJeff_work |
more or less |
19:40 |
SEJeff_work |
np |
19:40 |
Linuturk |
so, I'm using the salt-forumla repository to keep my salt minions and master correct. I noticed that if an update touches my master that requres a service restart, the process kinda borks itself. Is this a known issue where a master updating and restarting it's own service fails to do so? |
19:42 |
KennethWilke |
Linuturk: i have that issue as well outside of on ubuntu and fedora boxes |
19:42 |
KennethWilke |
https://github.com/saltstack/salt/issues/6724 |
19:42 |
|
devinus joined #salt |
19:42 |
KennethWilke |
and this wasn't a master restarting it's own minion or master process, this was a master failing to restart other masters |
19:43 |
Linuturk |
gotcha |
19:43 |
KennethWilke |
which was what i went to after seeing the issue you're describing |
19:43 |
|
mmilano joined #salt |
19:44 |
KennethWilke |
i can certainly see why it might be difficult for a master or minion to restart itself and still return data, but in my experience it's worked sometimes |
19:52 |
|
aboe joined #salt |
19:52 |
|
littleidea joined #salt |
19:53 |
|
dthom91 joined #salt |
19:53 |
aboe |
whiteinge, can you take a look at this https://github.com/aboe76/puppet-formula.git |
19:57 |
ipmb |
what do you guys call a set of salt states, modules, and pillars? |
19:57 |
ipmb |
Is there one name that encompasses all the stuff I drop on my master server? |
19:57 |
|
teebes joined #salt |
19:57 |
* whiteinge |
looks |
19:59 |
basepi |
UtahDave saurabhs: sorry, didn't see the mention. i'm not very up-to-speed on salt's fd limits, or how it does it per-process, sorry |
19:59 |
basepi |
not much help |
20:01 |
|
dbruns joined #salt |
20:02 |
whiteinge |
aboe: wow! looks really great |
20:02 |
aboe |
tested it in docker with ubuntu 12.04 only the package-map.jinja I couldn't test |
20:03 |
aboe |
ubuntu 12.04 doesn't have salt 0.17 yet... |
20:03 |
* whiteinge |
nods |
20:03 |
aboe |
maybe something to add to saltstack-formula's ?? |
20:04 |
whiteinge |
you can copy the grains.py module out of develop and put it into your /srv/salt/_modules directory if you don't want to wait 0.17 |
20:04 |
whiteinge |
aboe: definitely something to add! |
20:04 |
whiteinge |
hang on, i'll get it set up |
20:04 |
aboe |
thanks for the suggestion will get grains.py from my arch set up. |
20:05 |
|
welby_away joined #salt |
20:06 |
aboe |
whiteinge, only for those who can't live without puppet LOL... |
20:06 |
whiteinge |
using salt is a great way to manage puppet :-) |
20:07 |
aboe |
whiteinge, lol, one thing is missing, puppetdb formula... |
20:08 |
|
david_a joined #salt |
20:09 |
whiteinge |
aboe: you're part of the formulas organization now. i *think* you should be able to transfer the repo to the org |
20:09 |
aboe |
will transfer it. |
20:09 |
titubear |
cmd.wait doesn't work... require doesn't work... gah... need to make this stateful I guess |
20:12 |
|
pdayton joined #salt |
20:14 |
aboe |
whiteinge, It keeps saying I don't have admin rights to salt-stack-formulas |
20:14 |
|
carmony joined #salt |
20:14 |
whiteinge |
hm. ok |
20:14 |
whiteinge |
we may have to do this another way |
20:14 |
whiteinge |
sec.. |
20:15 |
UtahDave |
titubear: can you pastebin what your sls and the error you're getting? |
20:15 |
StDiluted |
so can anyone suggest what the best way to add lines to the check_mk config file would be when i highstate a machine, (not the check_mk host) |
20:16 |
UtahDave |
StDiluted: file.append? not sure exactly what you're trying to do |
20:16 |
titubear |
UtahDave: it's really been about 15 different attempts. I think the cmd.wait just doesn't work unless you make the command before it stateful |
20:16 |
titubear |
UtahDave: yup, just worked. :D |
20:17 |
aboe |
whiteinge, I see the issue I'm a contributor to the organization. |
20:17 |
titubear |
imho cmd.run: blah -require: -cmd: prior-command |
20:17 |
aboe |
with only pull permissions. |
20:17 |
titubear |
should wait for prior-command in a stateful way |
20:17 |
titubear |
otherwise why require it |
20:18 |
titubear |
stateful meaning meaning just a clean exit |
20:19 |
drags |
is reloading the master not recommended? when I try to run a command after reloading (and waiting ~60s for things to flush): for the first 10 minutes or so nothing returns and the salt binary fails with: salt.exceptions.SaltReqTimeoutError: Waited 60 seconds |
20:20 |
UtahDave |
titubear: cmd.run with a require means that the function will be run EACH time, as long as the required state is succesfful |
20:20 |
UtahDave |
titubear: cmd.wait with a - watch means that it will only run when the watched item returns a change |
20:21 |
aboe |
whiteinge, https://help.github.com/articles/what-are-the-different-access-permissions |
20:21 |
titubear |
that doesn't seem to work that way |
20:21 |
titubear |
UtahDave: I appreciate it though |
20:22 |
titubear |
UtahDave: with a require it won't work if the required state isn't finished for example |
20:23 |
whiteinge |
aboe: ah, thank you |
20:23 |
StDiluted |
UtahDave: I have a monitor-ops1 computer, and I want to modify a text file on it, when i highstate some other server, say… web1. |
20:23 |
whiteinge |
aboe: try now |
20:24 |
aboe |
whiteinge, lol it already exists |
20:24 |
whiteinge |
doh! |
20:24 |
whiteinge |
aboe: try now (again) |
20:24 |
|
jkleckner joined #salt |
20:25 |
aboe |
whiteinge, woot, it works. |
20:26 |
whiteinge |
woot! |
20:26 |
whiteinge |
aboe: thanks! |
20:26 |
aboe |
this is a better permission layer. |
20:27 |
drags |
more info: reloading master after changing /etc/salt/master causes master process to apparently stop processing commands. anything run from master fails with exception like: salt.exceptions.SaltReqTimeoutError: Waited 60 seconds |
20:27 |
drags |
master then needs to be kill -9'd and restarted before operation can continue |
20:27 |
drags |
any thoughts? |
20:27 |
drags |
nothing obvious in master or minion logs |
20:28 |
drags |
this is also both for when adding config as well as commenting it out, so I don't believe the config changes are suspect themselves |
20:30 |
aboe |
whiteinge, no problem will keep building salt-formula's |
20:30 |
StDiluted |
UtahDave: does that make sense? |
20:31 |
StDiluted |
I want to modify a file on computer A when B is highstated |
20:33 |
UtahDave |
StDiluted: Yeah, you could have the highstate for B fire an event. Then the reactor could listen for that event and execute an sls that modifies that file on Computer A |
20:34 |
StDiluted |
ok. I haven't messed with reactor yet at all, so I guess maybe this is my opportunity |
20:37 |
jpeach |
what does the use_in requisite actually do? |
20:38 |
|
jimr_ joined #salt |
20:38 |
jpeach |
I'm wondering why the bond0 example at http://docs.saltstack.com/ref/states/all/salt.states.network.html uses it |
20:43 |
dave_den |
jpeach: use_in allows you to use the settings from one state in another. in that case, the settings for bon0 are also used to provide the settings for the eth2 and eth3 states so you don't have to repeat yourself |
20:43 |
jpeach |
hmm …. how is that different from "use" ? |
20:44 |
jpeach |
does a "use" always need a corresponding "use_in"? |
20:45 |
StDiluted |
anyone have some examples of reactor usage that are a little more fully explained than the official docs? |
20:46 |
dave_den |
no. it's like watch/watch_in and require/require_in. if you want, say, state2 to use the settings from state1 as defaults, you would use: state1 in state2. if you wanted state1 to make sure state2 was using its defaults, you would then say use_in: state2 in state1 |
20:47 |
dave_den |
in the network example, eth2 and eth3 could have each had use: bond0 |
20:48 |
dave_den |
instead of bond0 having use_in: eth2, eth3 |
20:49 |
jpeach |
dave_den: so it's analogous to push vs. pull? use_in pushed the settings into another state, and "use" pulls them into the current state? |
20:49 |
|
linjan_ joined #salt |
20:49 |
|
TJ1980 joined #salt |
20:51 |
dave_den |
jpeach: yeah, you can think of it that way. watch_in is very handy for something like apache or nginx virtualhosts file functions. you can simply define your vhost function to manage a file, then have that function watch_in: apache so that apache is automatically reloaded any time you add or change a managed vhost. they also come in handy with loops and meta functions |
20:51 |
|
ricicle joined #salt |
20:52 |
StDiluted |
watch_in is great |
20:52 |
dave_den |
where you wouldn't know the name of the function until it gets templated or compiled into highstate data |
20:52 |
StDiluted |
saves a lot of typing |
20:52 |
jpeach |
yeh, I can see how that would be useful |
20:52 |
jpeach |
thanks dave_den, that's much clearer |
20:52 |
dave_den |
np |
20:52 |
|
qba73 joined #salt |
20:54 |
|
druonysuse joined #salt |
20:54 |
|
mattbarto joined #salt |
20:54 |
|
juanlittledevil joined #salt |
20:55 |
|
whit joined #salt |
20:55 |
juanlittledevil |
hi guys! got a quick quesiton for ya. In the states.file.managed module. can I require a user? |
20:56 |
UtahDave |
juanlittledevil: yes, if you're defining that user in a state |
20:56 |
juanlittledevil |
UtahDave: I am in a way. I've got the users defined dynamically via pillars |
20:57 |
juanlittledevil |
UtahDave: so I can just pass it the name of the user I would assume. |
20:59 |
UtahDave |
- require:\n - user: joeblow |
20:59 |
juanlittledevil |
UtahDave: Got it! Thank you! |
21:03 |
|
DerekRBN joined #salt |
21:04 |
|
p3rror joined #salt |
21:04 |
UtahDave |
you're welcome! |
21:05 |
DerekRBN |
Hey all, about to deploy a large windows .net project to aws and im pretty unfamiliar with windows. I'm wondering how useful windows version of salt is. How different is it? |
21:07 |
|
mattbarto joined #salt |
21:07 |
UtahDave |
DerekRBN: It's not quite as full featured as on the *nix platforms. But you can install software, create and manage users, install server roles, get system status, disk usage, service status, restart |
21:07 |
UtahDave |
DerekRBN: quite a bit |
21:08 |
UtahDave |
DerekRBN: is there anything in particular that you really need to do? |
21:09 |
|
jaequery joined #salt |
21:09 |
|
jacksontj joined #salt |
21:09 |
DerekRBN |
Thanks for the info. I dont think it will be that useful in this application really its just IIS and .net so... theres not a bunch of services and the company isnt even in source control so it seems like it's going to all be manual anyways. Boss just asked me to look into it |
21:10 |
DerekRBN |
Can we do windows updates? |
21:10 |
UtahDave |
Well, you can definitely install the IIS server role, install software and deploy code |
21:11 |
UtahDave |
DerekRBN: there's not a module right now for windows updates, but you can always shell out with cmd.run and you can use powershell with cmd.run if you want, too |
21:11 |
gamingrobot |
Can you create services on windows with salt? |
21:11 |
DerekRBN |
Ok thanks for the info Dave |
21:13 |
UtahDave |
gamingrobot: I use nssm to create the services, so yeah, you could do that |
21:13 |
|
kenbolton joined #salt |
21:18 |
|
z3uS joined #salt |
21:23 |
|
nrub joined #salt |
21:23 |
StDiluted |
does this code: http://docs.saltstack.com/topics/event/index.html#listening-for-events go inside the sls file? |
21:24 |
UtahDave |
No, those go in a python file |
21:25 |
UtahDave |
StDiluted: you probably want to look in the reactor docs |
21:25 |
|
oz_akan_ joined #salt |
21:25 |
StDiluted |
UtahDave: I am. The docs suck |
21:25 |
StDiluted |
or at least |
21:25 |
StDiluted |
if you are dumb like me |
21:25 |
StDiluted |
they suck |
21:26 |
StDiluted |
the reactor docs make no mention of how to listen for an event |
21:26 |
StDiluted |
just that you should do so |
21:26 |
UtahDave |
you don't have to do anything. Salt listens for you |
21:26 |
nkuttler |
fwiw, i hate the new look of the docs |
21:26 |
nkuttler |
and where did the navbar go... |
21:26 |
StDiluted |
that page i pasted says how to listen for an event, but makes no mention of where that code goes |
21:27 |
UtahDave |
StDiluted: that's if you're writing your own python code |
21:27 |
UtahDave |
StDiluted: you don't have to do anything to listen. |
21:27 |
UtahDave |
Add the tag you're wanting to listen to in the "reactor" config stanza in your master config |
21:27 |
UtahDave |
then give it a list of sls files to execute when it sees that tag |
21:28 |
StDiluted |
ok. and where is any sort of documentation on what's in 'data' |
21:28 |
StDiluted |
I see that it has 'data' in it |
21:28 |
StDiluted |
and 'id' |
21:28 |
StDiluted |
what else is in there? |
21:29 |
UtahDave |
It depends on the event that was fired. You can print it out if you want |
21:38 |
|
dthom91 joined #salt |
21:38 |
|
KyleG joined #salt |
21:38 |
|
KyleG joined #salt |
21:40 |
|
qba73_ joined #salt |
21:47 |
|
ninkotech__ joined #salt |
21:47 |
|
Ryan_Lane joined #salt |
21:49 |
|
bhosmer joined #salt |
21:49 |
|
Fin1te joined #salt |
21:53 |
|
blee joined #salt |
21:53 |
|
jacksontj joined #salt |
21:57 |
|
VertigoRay joined #salt |
21:58 |
jpeach |
nice, adding a custom grain in python was pretty easy |
21:58 |
VertigoRay |
kula: Thanks again for the help. Figured I'd link yah what you helped me with. https://github.com/VertigoRay/salt-osx-grains Thanks again! |
22:01 |
kula |
VertigoRay: no problem |
22:03 |
Ahlee |
If you've defined an ext_job_cache, does /jobs/<jid> on the master no longer make sense/work? Or will a call to /jobs/ forward that request to the backend cache? |
22:03 |
Ahlee |
sorry, this is brokering through the API |
22:04 |
UtahDave |
kula++ |
22:05 |
UtahDave |
all the jobs functions still work the same |
22:05 |
Ahlee |
so master then should poke the cache, and then forward your results |
22:05 |
Ahlee |
so if i'm not getting that, i'm still doing it wrong |
22:06 |
UtahDave |
I'm pretty sure. I haven't used it through the API, but I'm pretty sure that's what's supposed to happen |
22:07 |
Ahlee |
k, i'll tear into this tomorrow |
22:07 |
|
whit joined #salt |
22:11 |
|
middleman_ joined #salt |
22:11 |
|
devinus joined #salt |
22:17 |
|
elbaschid joined #salt |
22:24 |
|
issackelly joined #salt |
22:26 |
StDiluted |
anyone help me with an event trigger traceback? |
22:28 |
StDiluted |
https://gist.github.com/dginther/6313592 |
22:29 |
StDiluted |
hm, can't recreate |
22:30 |
StDiluted |
however, the trigger doesnt seem to be firing when i run the event |
22:30 |
StDiluted |
if i hop on any minion, and do 'salt-call event.fire_master 'west-add-server' 'wato'' |
22:30 |
StDiluted |
i get back local: True |
22:30 |
StDiluted |
but i see no event fire on the master |
22:30 |
|
middleman_ joined #salt |
22:35 |
|
oz_akan_ joined #salt |
22:36 |
|
kenbolton joined #salt |
22:36 |
|
dthom91 joined #salt |
22:39 |
|
druonysuse joined #salt |
22:39 |
|
druonysuse joined #salt |
22:42 |
StDiluted |
is there a way to monitor events on the master |
22:42 |
StDiluted |
like, see them fire? |
22:43 |
|
mattbarto joined #salt |
22:43 |
auser |
yeah |
22:43 |
auser |
there is |
22:43 |
auser |
looking through notes, sec |
22:44 |
auser |
see the python script on http://docs.saltstack.com/topics/event/index.html |
22:44 |
StDiluted |
thanks auser |
22:44 |
StDiluted |
yeah |
22:44 |
auser |
np StDiluted |
22:44 |
StDiluted |
I have seen that |
22:44 |
StDiluted |
however |
22:44 |
StDiluted |
UtahDave assured me that salt listens to events automatically |
22:45 |
|
mwillhite joined #salt |
22:45 |
StDiluted |
ah, looks like it is firing but the command I'm running is failing |
22:45 |
StDiluted |
I get the traceback i pasted earlier |
22:45 |
StDiluted |
https://gist.github.com/dginther/6313592 |
22:45 |
auser |
StDiluted: https://gist.github.com/auser/5761268 |
22:45 |
auser |
that's what I use |
22:46 |
auser |
you don't have to set cmd.run like that |
22:47 |
StDiluted |
i want to run the command on a different server than fired the trigger |
22:47 |
StDiluted |
so server X fires the trigger, and then on monitor-ops1 a file is appended |
22:47 |
|
Damoun joined #salt |
22:47 |
auser |
I left you a comment |
22:48 |
StDiluted |
that's cool. That one is working |
22:48 |
StDiluted |
I will edit to make it cleaner |
22:48 |
auser |
k |
22:48 |
StDiluted |
the one that's failing is the wato.sls one |
22:48 |
auser |
oh the reactor? |
22:48 |
StDiluted |
yep |
22:48 |
StDiluted |
{'west-vpc-prod-mail-outbound1': {'cmd.run': [{'name': 'salt \'monitor-ops1\' file.append /etc/check_mk/conf.d/wato/vpc_west_stack/tmp.tmp \\ "This is a test. This is only a test."'}]}} |
22:48 |
StDiluted |
that's what I see in the log |
22:48 |
StDiluted |
that fails |
22:49 |
auser |
I think it's |
22:49 |
auser |
cmd.cmd.run: |
22:49 |
StDiluted |
works on the command line though |
22:49 |
StDiluted |
ah |
22:49 |
auser |
correct me if I'm wrong UtahDave |
22:52 |
UtahDave |
yes, cmd.cmd.run |
22:53 |
StDiluted |
still failed |
22:53 |
StDiluted |
added a comment |
22:54 |
|
jkleckner joined #salt |
22:54 |
|
druonysuse joined #salt |
22:54 |
|
druonysuse joined #salt |
22:54 |
StDiluted |
and then, until i restart the salt-master, no more events go through |
22:55 |
UtahDave |
StDiluted: did you modify your firing like auser commented? |
22:55 |
StDiluted |
yep |
22:55 |
UtahDave |
what's your reactor config look like? |
22:55 |
StDiluted |
https://gist.github.com/dginther/6313592 |
22:56 |
StDiluted |
it's under /srv/reactor/wato.sls |
22:56 |
UtahDave |
StDiluted: no, in your master config in the "reactor" section |
22:56 |
StDiluted |
oh |
22:56 |
StDiluted |
reactor: |
22:56 |
StDiluted |
- 'wato': |
22:56 |
StDiluted |
- /srv/reactor/wato.sls |
22:57 |
StDiluted |
and I'm just calling the event from the command line on one of the minions |
22:57 |
StDiluted |
alt-call event.fire_master 'west-add-server' 'wato' |
22:58 |
StDiluted |
er salt-call* |
22:58 |
jpeach |
basepi: what's the etiquette for closing github issues? |
22:58 |
UtahDave |
StDiluted: did you change it to cmd.cmd.run? |
22:58 |
StDiluted |
yes |
22:58 |
basepi |
jpeach: hehe, depends. if it's resolved, then hit close! =D |
22:58 |
|
middleman_ joined #salt |
22:59 |
StDiluted |
UtahDave: in my gist, second comment, you can see that it now says cmd.cmd.run |
22:59 |
jpeach |
basepi: roger that |
23:01 |
UtahDave |
StDiluted: Hey, can you cut out the templating while we're testing? |
23:01 |
StDiluted |
which templating? |
23:01 |
StDiluted |
{{ data['id'] }} ? |
23:01 |
VertigoRay |
How do I divide out a lab environment from production using the same master? |
23:02 |
UtahDave |
yes, just put some string there. |
23:02 |
StDiluted |
ok done |
23:02 |
StDiluted |
one sec |
23:02 |
UtahDave |
StDiluted: and create a new pastebin with the new reactor sls and the output |
23:03 |
UtahDave |
VertigoRay: you'll want to look up "environments" and "targeting" |
23:04 |
VertigoRay |
UtahDave: thanks for those keywords! |
23:05 |
StDiluted |
UtahDave: https://gist.github.com/dginther/6313846 |
23:06 |
|
tuxIO joined #salt |
23:06 |
UtahDave |
VertigoRay: here's a link to the environments: http://docs.saltstack.com/ref/file_server/file_roots.html?highlight=environment |
23:06 |
StDiluted |
the event fires, the correct sls gets executed, but there is an error somewhere in the command that gets created from that sls |
23:06 |
UtahDave |
VertigoRay: then you'll want to look at the top.sls docs to see how you map to those environments |
23:06 |
|
mwillhite joined #salt |
23:07 |
|
cewood joined #salt |
23:07 |
VertigoRay |
UtahDave: Thanks so much for getting me pointed in the right direction. Very helpful. |
23:07 |
UtahDave |
you're welcome, VertigoRay! |
23:07 |
UtahDave |
let me know if you have any questions as you go along |
23:07 |
UtahDave |
StDiluted: have you looked how the cmd.cmd.run examples are put together at the bottom of the reactor doc? |
23:08 |
|
middleman_ joined #salt |
23:08 |
StDiluted |
yeah, i can try that |
23:08 |
UtahDave |
StDiluted: look at the bottom of the page here: http://docs.saltstack.com/topics/reactor/index.html |
23:10 |
StDiluted |
I see that. so cmd.file.append? |
23:10 |
StDiluted |
tgt is self explanatory |
23:14 |
StDiluted |
the arg would be ? |
23:14 |
|
middleman_ joined #salt |
23:15 |
|
issackelly joined #salt |
23:15 |
StDiluted |
UtahDave: would it be cmd.file.append? |
23:15 |
UtahDave |
yes |
23:16 |
StDiluted |
https://gist.github.com/dginther/6313846 |
23:16 |
StDiluted |
last comment |
23:16 |
StDiluted |
still not working, though monitor ops is executing something |
23:16 |
StDiluted |
I think it's just syntax now |
23:17 |
UtahDave |
yeah. try removing the \ |
23:18 |
|
dthom91 joined #salt |
23:18 |
UtahDave |
and then if that doesn't work, try making it two arguments, instead of one long argument on the same line |
23:19 |
StDiluted |
ok |
23:20 |
StDiluted |
makes the file tmp.tmp and also a file called tmp.tmp "This is a test. This is only a test." |
23:20 |
StDiluted |
got it |
23:20 |
StDiluted |
teo arguments |
23:20 |
StDiluted |
two |
23:20 |
UtahDave |
nice! |
23:21 |
StDiluted |
thanks! |
23:23 |
UtahDave |
StDiluted: can you pastebin your final working version? |
23:24 |
|
cxz joined #salt |
23:25 |
|
chrisgilmerproj left #salt |
23:27 |
jesusaurus |
what are people doing to test their salt states? does anyone have a simple way to do unit tests or integration tests for their state tree? |
23:28 |
|
middleman_ joined #salt |
23:29 |
UtahDave |
jesusaurus: furao has some kind of framework he built to do that. Let me see if I can find his repo again |
23:30 |
bashcoder |
That would be good, because mine sometimes looks a lot like this: vagrant destroy ; vagrant up ; vagrant ssh ; lather ; rinse ; repeat :) |
23:30 |
|
mmilano left #salt |
23:32 |
jesusaurus |
yeah, right now im just pushing to staging, running highstate, and inspecting for errors. not very automatic |
23:32 |
|
teebes joined #salt |
23:36 |
UtahDave |
jesusaurus: Yeah, I'm not finding furao's stuff right now. |
23:37 |
UtahDave |
Here at SaltStack we're actually building out a Jenkins server that uses Salt cloud to spin up VM's of various OSes and executing unit and integration tests. |
23:37 |
UtahDave |
Back in a little while. |
23:41 |
|
middleman_ joined #salt |
23:44 |
|
aat joined #salt |
23:45 |
|
alunduil joined #salt |