Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-04-21

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

All times shown according to UTC.

Time Nick Message
00:01 ninkotech joined #salt
00:02 TyrfingMjolnir joined #salt
00:15 bhosmer joined #salt
00:21 jslatts joined #salt
00:30 jeffg_ joined #salt
00:33 schimmy joined #salt
00:36 Ryan_Lane joined #salt
00:36 jeremyfelt joined #salt
00:55 gw joined #salt
00:55 gw Hi,... is there a recommended way how to use formulas with multiple environments?
00:56 gw as far as I understand it, gifts would match branches into environments, but most formulas I have seen have only a master branch, and therefore they won't be seen in e.g. a 'dev' environment
01:08 dancat having an issue with salt and vagrant
01:09 dancat the issue is the same as yesterday. I am installing salt minion on a vm then packaging the vagrant box
01:09 dancat but the problem is that once the box is created my local minion file is not getting copied over to the vm
01:10 dancat so I manually copied the minion file to the location /etc/salt/minionm
01:10 dancat now when I try to provision it is still failing to recognize that the provision is intending to be masterless
01:10 dancat and I am not sure why
01:12 doanerock joined #salt
01:13 dancat considering my /etc/salt/minion has one line file_client: local
01:19 dancat embarrasingly I think it is my mistake
01:29 xzarth joined #salt
01:45 Luke__ joined #salt
01:45 pb__ joined #salt
01:51 elfixit joined #salt
01:53 joehoyle joined #salt
01:56 sroegner joined #salt
02:11 Guest37620 joined #salt
02:30 malinoff joined #salt
03:01 joehoyle joined #salt
03:04 yomilk joined #salt
03:12 UtahDave joined #salt
03:13 catpigger joined #salt
03:15 alekibango joined #salt
03:16 jeremyfelt joined #salt
03:26 NotreDev joined #salt
03:37 Tekni joined #salt
03:37 xl1 joined #salt
03:39 ajw0100 joined #salt
03:49 mgw joined #salt
04:08 Ryan_Lane1 joined #salt
04:12 ajolo joined #salt
04:13 DaveQB joined #salt
04:20 acabrera joined #salt
04:32 joehoyle joined #salt
04:32 ajolo joined #salt
04:39 ajolo_ joined #salt
04:40 cro joined #salt
04:54 cro joined #salt
05:03 jalaziz joined #salt
05:05 joonas joined #salt
05:09 ajolo__ joined #salt
05:10 ajolo joined #salt
05:10 yomilk joined #salt
05:18 jeremyBass joined #salt
05:18 jeremyBass left #salt
05:24 arthabaska joined #salt
05:33 joehoyle joined #salt
05:34 joonas left #salt
05:35 yomilk joined #salt
05:37 mohanS joined #salt
05:49 ravibhure joined #salt
05:57 andabata joined #salt
06:06 honst joined #salt
06:13 jeddi joined #salt
06:24 anuvrat joined #salt
06:28 arthabaska joined #salt
06:32 joehoyle joined #salt
06:38 stanchan joined #salt
06:38 mr_chris joined #salt
06:46 l0x3py joined #salt
06:54 Nazca joined #salt
06:54 Nazca joined #salt
06:59 astol joined #salt
07:08 mnaser joined #salt
07:19 slav0nic joined #salt
07:19 slav0nic joined #salt
07:40 otsarev_ joined #salt
07:47 Madkinder joined #salt
07:49 Madkinder Hi. I wanted to know how do I declare inter-node state dependency in sls?
07:52 Madkinder Consider an example when I have two salt minions: 1 for mysql server and 1 for application server. The sls for app server creates mysql database and user for its operation. Since salt would sync the states on both nodes in parallel, how can I make sure that MySQL has been already set up and started on minion-1 before I try to create the database on minion-2?
08:00 Eugene Madkinder - You would invoke the highstate calls sequentially rather than in parallel
08:00 Madkinder isn't there any way to specify it somewhere?
08:00 Eugene Not between minions like that(AFAIK), no.
08:01 Madkinder I'm concerned about sharing this knowledge between several admins
08:02 Madkinder I've found the very same question asked on the mailing list: https://groups.google.com/forum/#!topic/salt-users/iMdOXaqXonc
08:03 Madkinder though it has been asked two years ago. hasn't anything changed since then?
08:03 MrTango joined #salt
08:05 TyrfingMjolnir joined #salt
08:34 anuvrat joined #salt
08:42 TyrfingMjolnir joined #salt
08:51 seamuz joined #salt
08:51 giantlock joined #salt
09:05 mgw joined #salt
09:11 yomilk joined #salt
09:17 dramagods joined #salt
09:31 Kenzor joined #salt
09:57 shoma joined #salt
09:59 Madkinder what would be the execution order of three states, if state A requires state B, that is in turn being watched by state C?
09:59 Madkinder Here's an example sls: http://pastebin.com/WaAA0YNQ
09:59 Daviey joined #salt
10:10 gw Madkinder, with circular dependencies you'll always have a problem, but for simpler cases, you might get away by using the overstate system (not sure no experience with it but it sounds like that's what you want)
10:13 gw could also be mistaken (not sure whether I understood the question)
10:15 yomilk joined #salt
10:16 yomilk_ joined #salt
10:24 cowyn joined #salt
10:39 gw joined #salt
10:39 gw joined #salt
10:40 andrein__ joined #salt
10:40 t0rrant joined #salt
10:42 vortec joined #salt
10:50 jcsp joined #salt
11:11 bhosmer joined #salt
11:12 diegows joined #salt
11:14 yomilk joined #salt
11:17 fsniper joined #salt
11:17 fsniper Hello!
11:17 fsniper is it possible to refer a state from another environment? like: prod: host1: state*dev*
11:37 nocturn joined #salt
11:40 Kenzor joined #salt
12:06 yomilk_ joined #salt
12:08 shoma joined #salt
12:16 Leech joined #salt
12:22 funzo joined #salt
12:22 sroegner joined #salt
12:24 ckao joined #salt
12:31 jslatts joined #salt
12:32 jslatts joined #salt
12:33 faldridge joined #salt
12:36 doanerock joined #salt
12:40 ndrei joined #salt
12:52 ipmb joined #salt
12:56 happytux joined #salt
13:10 jslatts joined #salt
13:13 Kenzor joined #salt
13:14 mpanetta joined #salt
13:21 chiui joined #salt
13:23 joehoyle joined #salt
13:23 racooper joined #salt
13:24 HeadAIX joined #salt
13:27 mapu joined #salt
13:27 vejdmn joined #salt
13:28 mapu joined #salt
13:28 quickdry21 joined #salt
13:29 tyler-baker joined #salt
13:40 joehh1 packages for saucy now available on launchpad, trusty coming tomorrow
13:42 jnials joined #salt
13:46 jtang1 joined #salt
13:47 pdayton joined #salt
13:50 jchen nice
13:51 Psi-Jack_ joined #salt
13:53 yomilk joined #salt
13:54 undefinednull joined #salt
13:54 danielbachhuber joined #salt
14:01 sroegner joined #salt
14:06 ajprog_laptop joined #salt
14:07 AdamSewell joined #salt
14:07 AdamSewell joined #salt
14:08 tessellare joined #salt
14:09 otsarev_ Hi guys
14:09 otsarev_ Any comments? https://groups.google.com/forum/#!topic/salt-users/8XmvQHGDBT8
14:11 anuvrat joined #salt
14:11 thedodd joined #salt
14:14 doddstack joined #salt
14:14 JasonSwindle joined #salt
14:14 doddstack joined #salt
14:14 oc joined #salt
14:15 debian112 joined #salt
14:15 doddstack joined #salt
14:16 debian112 I wonder if it is possible to salt-call a "id declaration" directly?
14:16 Madkinder joined #salt
14:16 debian112 salt-call state.sls module.id declaration
14:16 debian112 ?
14:17 timoguin debian112: there's no way to do that now, but i think a feature request for that functionality was put in a week ago
14:17 timoguin actually i think it may be in the develop branch already
14:18 ndrei joined #salt
14:18 Madkinder left #salt
14:18 JasonSwindle joined #salt
14:25 debian112 timoguin: Nice!
14:26 icarus joined #salt
14:27 pdayton joined #salt
14:34 Ahlee I know it's possible, but i'm blanking on what it's called when one minion tells another minion to execute a state - can somebody job my memory?
14:34 calvinhp_mac joined #salt
14:35 jrdx joined #salt
14:35 calvinhp_mac is the apache state something that is standard, or is it something deprecated?  I'm trying to use apache.config in a state file, but the minion is telling me that they state is unavailable for apache.config
14:35 kermit joined #salt
14:36 alunduil joined #salt
14:38 Ahlee peer communication is what I was looking for
14:38 calvinhp_mac or is apache.config just too new
14:38 Ahlee calvinhp_mac: looks like apache is a default state
14:39 Ahlee looks like first commit was 3/14
14:39 mgw joined #salt
14:40 Ahlee so you'll have to run develop to get it, as it's not in hydrogen
14:40 calvinhp_mac I wish the salt docs wouldn't default to the latest development branch
14:40 calvinhp_mac can be confusing for folks trying out new stuff
14:41 Ahlee Agreed.
14:42 KennethWilke joined #salt
14:42 * calvinhp_mac goes to rewrite his apache conf back to a managed.file
14:44 opapo joined #salt
14:44 timoguin also agreed there.
14:44 timoguin there are versionadded directives on most things, but they can be easy to miss
14:45 wendall911 joined #salt
14:45 Ahlee Still waiting for the enterprise release, too.
14:47 GradysGhost joined #salt
14:48 JasonSwindle left #salt
14:49 jeremyBass1 joined #salt
14:49 schristensen joined #salt
14:51 calvinhp_mac heh, apache's state is missing the version added, that would have been a good tip off
14:52 smcquay joined #salt
14:53 repl1cant joined #salt
14:54 mway joined #salt
14:54 ZombieFeynman joined #salt
14:54 timoguin calvinhp_mac: can you open an issue for that please?
14:55 ipmb joined #salt
14:56 calvinhp_mac sure
14:56 joehoyle_ joined #salt
14:57 mateoconfeugo joined #salt
15:00 Gordonz joined #salt
15:01 Gordonz joined #salt
15:01 calvinhp_mac done
15:01 calvinhp_mac https://github.com/saltstack/salt/issues/12162
15:02 mgw joined #salt
15:02 cnelsonsic joined #salt
15:03 Psi-Jack_ joined #salt
15:05 conan_the_destro joined #salt
15:05 kballou joined #salt
15:06 bastion2202 joined #salt
15:08 jergerber joined #salt
15:08 toastedpenguin joined #salt
15:09 NullWagon joined #salt
15:09 manfred calvinhp_mac: that is my fault, i forget to add those when I do new states/modules :/
15:10 jalbretsen joined #salt
15:12 Guest3775 joined #salt
15:13 Networkn3rd joined #salt
15:17 ZombieFeynman joined #salt
15:18 bhosmer joined #salt
15:21 Zuru joined #salt
15:23 eriko joined #salt
15:23 calvinhp_mac love the quick service around this joint :-)
15:24 Gareth morning
15:24 zhalas joined #salt
15:27 jtang1 joined #salt
15:27 zhalas hi, I'm pretty new to Salt so please excuse me my ignorance
15:27 ravibhure joined #salt
15:28 zhalas I would like to create a state, that is a composition of states already defined by state modules
15:29 zhalas let's say I have a standard django project and I would like to define a state that will create an user, http server configuration and a databse
15:30 zhalas so it's just a composition of existing features
15:31 zhalas I probably could express this using jinja templates, but I would prefer if it was a proper state having the same syntax as any other state
15:31 zhalas is there a good way to do that?
15:32 timoguin zhalas: you'll basically be writing a state that wraps other states. so you can call them inside your functions
15:32 zhalas timoguin: so you suggest writing a state module that calls other state modules?
15:33 srage joined #salt
15:33 srage_ joined #salt
15:34 timoguin zhalas: pretty much. you could call other states or other modules
15:38 zhalas timoguin: so I would have to express the dependencies in Python, which I don't like beceause I would loose yaml readability
15:38 zhalas is there a way to do that without sacrificing yaml readability?
15:39 smcquay joined #salt
15:39 meteorfox joined #salt
15:40 srage joined #salt
15:42 kballou joined #salt
15:42 timoguin zhalas: oh sure. i thought you were wanting to write a custom state.
15:43 timoguin there are a number of blogs floating around showing ways to manage django apps
15:44 zhalas timoguin: django was just an example, I want to have a custom state that composes multiple predefined states
15:45 zhalas and I would like to keep the definition clean
15:45 timoguin zhalas: that's what SLS files are for. you have multiple states per SLS
15:46 timoguin i'm not sure exactly what kind of functionality you're looking for
15:47 zhalas timoguin: yes, I understand, but I want to have a state that wraps a few other states and I can instantiate it in a parametrized way
15:47 timoguin zhalas: it sounds like what you're wanting is a custom state, which you'd have to implement in python
15:49 zhalas timoguin: I thought so, but I was wondering if there is no other way I could use which would be more clean
15:49 timoguin zhalas: you have a variety of renderers available in writing the states (jinja, mako, pydsl, pyobjects)
15:49 timoguin but other than that you're in custom state territory
15:49 zhalas from the functional standpoint what I want is equivalent to expanding simple jinja yaml template with a single paramter
15:50 zhalas but I would like it to behave like a regular state
15:51 tligda joined #salt
15:52 ZombieFeynman joined #salt
15:54 fragamus joined #salt
15:56 [diecast] joined #salt
16:00 ekristen joined #salt
16:02 gothix joined #salt
16:03 faldridg_ joined #salt
16:04 srage joined #salt
16:05 gothix anyone know the name of the salt ldap module
16:05 harobed_ joined #salt
16:06 jeremyfelt joined #salt
16:06 timoguin gothix: salt/modules/ldapmod.py
16:07 timoguin it's loaded on the minion as "ldap" if the ldap python module is present.
16:07 haroldjones joined #salt
16:07 gothix timoguin, thanks im getting a wierd error Authentication module threw an exception: Failed to connect to LDAP, module not loaded
16:07 haroldjones joined #salt
16:09 gothix its present on both the minion and the master
16:09 timoguin gothix: sounds like the module isn't loading because python-ldap isn't there
16:09 timoguin or is it?
16:09 haroldjones joined #salt
16:10 gothix no its not thanks
16:12 chrisjones joined #salt
16:17 nlb joined #salt
16:18 cruatta joined #salt
16:18 joehoyle joined #salt
16:18 bhosmer joined #salt
16:19 manfred it shouldn't get that far if you don't have python-ldap installed :/ https://github.com/saltstack/salt/blob/develop/salt/modules/ldapmod.py#L63
16:20 KyleG joined #salt
16:20 KyleG joined #salt
16:20 gothix it has to be set in the minion config ??
16:21 gothix i was setting it on the master
16:21 gothix i wouldent think that it would need to be set on the minion
16:21 timoguin actually that import line will only fail if python-ldap isn't installed
16:22 timoguin since all it does it try to import the module
16:22 timoguin the module isn't clear as to where the config needs to be set. i assume it could be via pillar or minion config. maybe others.
16:23 gothix i am settinhg it up in the end of the master config and i can auth with pam but not ldap
16:23 timoguin so this module doesn't have much to do with authentication. looks like it's just for sending commands to an LDAP server
16:24 timoguin specifically, just LDAP searches
16:25 timoguin for auth, there's this module: https://github.com/saltstack/salt/blob/develop/salt/auth/ldap.py
16:25 gothix i was figuring i could use it to to authenticate who can do what on the master
16:25 timoguin that's what the auth modules are for. i've only used pam for external auth
16:26 gothix so ill need the linked module then
16:27 TheRealBill joined #salt
16:27 pass_by_value joined #salt
16:27 pass_by_value left #salt
16:29 jtang1 joined #salt
16:29 UtahDave joined #salt
16:29 linuxlewis joined #salt
16:34 srage joined #salt
16:36 rstone left #salt
16:36 srage joined #salt
16:37 linuxlew_ joined #salt
16:41 CeBe joined #salt
16:43 sarlalian joined #salt
16:46 jcockhren when using: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.rvm.html#salt.states.rvm.installed
16:46 stanchan joined #salt
16:46 jcockhren in order to use the installed ruby, would the 'name' setting be 'system'?
16:50 faldridge joined #salt
16:51 schimmy joined #salt
16:56 haroldjones joined #salt
16:56 Ahlee UtahDave: Issue 6859 / 6240, winrepo templating
16:56 Ahlee How likely is that to make it into Helium?
16:57 ajolo joined #salt
16:59 joehoyle joined #salt
17:03 ajolo_ joined #salt
17:05 mgw joined #salt
17:05 HeadAIX joined #salt
17:08 Heartsbane So if I need to modify one line of a config file with salt ... I could use module file.replace but I need to replace the whole line
17:09 Heartsbane and I don't know what the string of what I will be replacing because it is always changing
17:09 Heartsbane I just know what the line will begin with
17:09 Heartsbane Does anyone have any suggestions
17:10 timoguin write a regex that matches the appropriate beginning of the line and replaces all the way to the end of hte line?
17:10 timoguin something like ^linebeginning.*$
17:11 Ryan_Lane joined #salt
17:13 Heartsbane ya I was hoping for nice salt way .. but that may work
17:14 aw110f joined #salt
17:14 haroldjones joined #salt
17:14 Ryan_Lane manfred: excellent!
17:15 Ryan_Lane looks like it was addressed in a different PR, though?
17:16 manfred yeah
17:16 manfred Ryan_Lane: whiteinge wanted to go with an event state, because it gives you better control over the requires and stuff
17:16 * Ryan_Lane nods
17:18 KennethWilke joined #salt
17:18 whiteinge manfred: kudos on your pull req, btw! i haven't seen many diffs to state.py that are that cleanly done  :-)
17:18 manfred :)
17:18 manfred thanks
17:18 * whiteinge fears state.py
17:18 Gareth "There be monsters"
17:19 whiteinge ^^
17:19 Heartsbane whiteinge: MAN UP
17:19 manfred i am still trying to grok it, i want to figure out how salt makes the list into a dictionary in lowstate data, i think it is just in the compile_* functions, but haven't dug that far, which was why I had the parser for the list in - event:
17:19 * Heartsbane give whiteinge five across the eyes.
17:21 UtahDave Ahlee: regarding the winrepo templating,  I started on it a while back, but haven't had the time to dedicate to it to finish it
17:22 mateoconfeugo joined #salt
17:23 whiteinge UtahDave: i'll sell you some of my story points
17:23 UtahDave ??
17:25 Ryan_Lane UtahDave: is there any way to have salt-call return a non-0 response for running state.highstate?
17:25 calvinhp_mac joined #salt
17:25 whiteinge more story points == more time for stuffs
17:25 UtahDave whiteinge: oh, yeah. I've missed all that.
17:26 UtahDave Ryan_Lane: Yeah, I think there's an option you can pass in
17:26 Ryan_Lane it seems it always return 0 right now, but salt-call gives back a list of successes, failures, and changes. it seems it should know if a highstate fails.
17:26 UtahDave add --retcode-passthrough
17:26 Ryan_Lane ah. excellent. is that undocumented?
17:26 UtahDave no, it's in   salt-call --help
17:27 Ryan_Lane ah. why did I not notice that before? thanks
17:28 UtahDave no problem!
17:30 Ahlee UtahDave: so slim to none? :)
17:30 srage joined #salt
17:33 UtahDave Ahlee: looks like there's a couple weeks until the feature freeze. I may be able to get it in
17:33 Ahlee Fair enough.  Good luck.
17:33 zain joined #salt
17:37 Gareth UtahDave: feature freeze for 2014.2?
17:38 UtahDave Yeah, it's going to freeze with the first RC of 2014.2
17:38 Gareth awesome.
17:38 rojem joined #salt
17:48 jaimed joined #salt
17:51 elfixit joined #salt
17:52 druonysus joined #salt
17:52 druonysus joined #salt
17:53 ajolo_ joined #salt
17:55 gothix using rhel 6 and salt 0.17 &  salt-api 0.8.2 involking salt-api throughs and exception "AttributeError: 'module' object has no attribute 'logging'
17:55 gothix "
17:56 whiteinge gothix: the RPM in EPEL is crazy out of date.
17:56 whiteinge if you don't mind grabbing an individual RPM you can use this one instead:
17:56 whiteinge https://github.com/saltstack/salt-api/releases/download/v0.8.4.1/salt-api-0.8.4.1-1.el6.noarch.rpm
17:56 ajw0100 joined #salt
17:57 racooper the one in epel s 2014.1.0...
17:57 gothix whiteinge, i seen your git hub comment about it and figured it was fixed - Thanks for the link
17:57 ndrei joined #salt
17:57 whiteinge gothix: sure. sorry for the trouble
17:57 * whiteinge shakes his fist at packaging
17:59 bhosmer joined #salt
18:00 mapu joined #salt
18:01 yusuket joined #salt
18:02 racooper 2014.1.3 is pending testing right now, tho docs.saltstack.com show the latest version as 2014.1.1.
18:02 jtang1 joined #salt
18:03 yomilk joined #salt
18:03 whiteinge racooper: docs will be updated once the official announcement goes out...which should be soon as most packages are ready
18:04 racooper ah, wasn't sure how that timing worked.  I was surprised to see testing packages newer than docs, that just doesn't ever happen!
18:05 ajolo__ joined #salt
18:06 CaptTofu_ joined #salt
18:07 chrisjones joined #salt
18:08 haroldjones joined #salt
18:10 anuvrat joined #salt
18:11 bastion2202 joined #salt
18:19 austin987 joined #salt
18:21 elfixit joined #salt
18:24 srage joined #salt
18:26 fragamus joined #salt
18:29 thayne joined #salt
18:29 ajw0100 joined #salt
18:31 bawnzai joined #salt
18:37 UtahDave joined #salt
18:39 druonysuse joined #salt
18:39 druonysuse joined #salt
18:40 meteorfox|afk joined #salt
18:40 CeBe hi, not sure I understood correctly. Is it save to have the salt master-minion comunication over public internet or do I need some VPN or internal network for that?
18:41 ajprog_laptop anyone using salt on Ubuntu 14.04 yet?
18:41 CeBe it is not mentioned explicitly in the docs but it sounds like comunication is encrypted
18:41 jergerber I'm doing a `bootstrap-salt.sh git v2014.1.3`, and salt gets installed and started, but I don't get /etc/ufw/applications.d/salt*, or /etc/init.d//etc/init.d/salt-minion... is there anything I do to get those files with a git install?
18:41 [diecast] joined #salt
18:42 ajprog_laptop I am trying to use pkgrepo and I am getting an error -- CommandExecutionError: Error: aptsources.sourceslist python module not found
18:42 CeBe Ah well, needed to read a bit further. found it :) http://docs.saltstack.com/en/latest/ref/configuration/index.html#key-management
18:42 timoguin CeBe: transport is encrypted with AES keys, but i'd still recommend a VPN layer.
18:42 CeBe thanks timoguin
18:43 CeBe will first set it up without VPN and try to add that extra security later
18:44 linuxlewis joined #salt
18:44 bawnzai left #salt
18:47 rawtaz holy cow
18:47 rawtaz it's the CeBe
18:47 KennethWilke joined #salt
18:48 CeBe rawtaz: holy pig, it's rawtaz!
18:48 rawtaz i am!
18:48 rawtaz i thought you were an ansible guy
18:48 CeBe rawtaz: finally got some time to try saltstack
18:48 rawtaz nice
18:48 CeBe rawtaz: I am neither salt nor ansible yet. need to get on that
18:48 rawtaz i see
18:52 linuxlew_ joined #salt
18:56 nicksloan left #salt
18:56 pydanny joined #salt
18:57 haroldjones joined #salt
18:57 doanerock joined #salt
19:01 TheRhino04 joined #salt
19:01 TheRhino04 Hello
19:01 TheRhino04 just curious, are returners available for reactor states?
19:02 Ryan_Lane TheRhino04: hm. I think they are not, but they really should be.
19:02 * Ryan_Lane checks
19:02 whiteinge they are
19:02 Ryan_Lane oh. neat
19:02 Ryan_Lane does that mean they are also available for states now too?
19:03 manfred they are available as a state.
19:03 Ryan_Lane returners?
19:03 whiteinge the "cmd" reactor wraps LocalClient() which supports the "ret" kwarg
19:03 manfred oh no
19:03 manfred not returners
19:03 manfred sorry
19:03 whiteinge er. maybe I didn't understand the OP question
19:04 manfred Ryan_Lane: i was looking into it, but didn't add them like we discussed the other day
19:04 whiteinge TheRhino04: did you want to call a returner directly from the reactor or did you want to use a returner for commands sent out to minions via the reactor?
19:05 Ryan_Lane hm: https://github.com/saltstack/salt/issues/3137
19:05 TheRhino04 my ideal
19:05 TheRhino04 I wanted to call the wheel.key.gen_accept
19:05 TheRhino04 when the reactor was triggered
19:05 Ryan_Lane looks like states themselves can use a returner. I guess reactors don't use states the same way, though
19:05 TheRhino04 and save the return value, the private key that was generated, into a custom returner
19:06 TheRhino04 does that seem possible?
19:08 Ryan_Lane TheRhino04: hm.
19:08 whiteinge TheRhino04: ah, i see. sorry, no. you'd have to write a small runner to wrap the wheel.key.gen_accept call in order to do something with that return
19:08 manfred Ryan_Lane: is it for all states, or just the module state?
19:09 Ryan_Lane manfred: just the module state, it seems
19:09 manfred yar
19:09 Ryan_Lane that's not what I wanted...
19:09 * Ryan_Lane reopens his ticket
19:09 Ryan_Lane hm. you can't reopen tickets, can you?
19:09 manfred Ryan_Lane: tag me in it, I am was looking into it, but having groked how we render states enough to ddd it in
19:10 druonysuse joined #salt
19:10 druonysuse joined #salt
19:11 Ryan_Lane manfred: what's your github username?
19:11 manfred gtmanfred
19:12 calvinhp_mac I'm seeing a strange issue that may just be "normal" behavior…  I have multiple environments setup like base, testing, acceptance
19:13 Ryan_Lane anyone know how to reopen this issue? https://github.com/saltstack/salt/issues/3137
19:13 whiteinge reopened
19:13 calvinhp_mac right now the base env version of the file is being sent over to the minion
19:14 Ryan_Lane whiteinge: thanks
19:14 calvinhp_mac but the file exists in the testing tree with changes
19:14 whiteinge there wasn't a reopen button at the bottom? i didn't know that was a perms-required thing :-/
19:14 whiteinge Ryan_Lane: have you seen https://github.com/felskrone/salt-eventsd ?
19:14 Ryan_Lane yeah, it's a perms related thing
19:14 calvinhp_mac state.show_top shows the state only in base though, do I have to specify it in the testing environment to have it picked up there too or should it merge them when figuring out what file to send?
19:14 whiteinge boo :(
19:17 Ryan_Lane whiteinge: that's pretty interesting
19:17 Ryan_Lane could be used as an audit database too
19:17 danielbachhuber joined #salt
19:17 calvinhp_mac seems like I have misunderstood the environments setup
19:17 whiteinge i totally love that idea. i want it in core :)
19:17 calvinhp_mac sounds like each environment needs the full definitiion if you are overriding an existing file from base
19:18 sandbender1512 basepi you around?
19:18 basepi sandbender1512: depends.....  ;)
19:19 srage joined #salt
19:19 sandbender1512 haha... know off the top of your head if there's an open issue for 'REQUIRE SSL' for mysql_grants state and/or if there are plans to add already?
19:19 sandbender1512 (if not, I'm gonna add and submit a pull request, hence the question)
19:20 basepi sandbender1512: open issue:  https://github.com/saltstack/salt/issues/5778
19:20 sandbender1512 thx I'll check it out :)
19:20 basepi as far as I can tell, no one is working on it, so if you want to take it that would be great!
19:21 sandbender1512 kk will do... I'll probably hit it up first thing tomorrow morning
19:21 manfred Ryan_Lane: whiteinge: the events stuff took priority, cause I saw how to do it more easily, once i figure out how to do it, I would like to at least have a set of patches to add returners for states, at least to talk about how easy or terrible it would be to add
19:22 manfred but I am in the middle of moving to a new apartment, so it will probably be on hold until thursday or friday
19:22 whiteinge Ryan_Lane: i mentioned it because i've been debating the place for returners since i saw that project.
19:23 jcsp joined #salt
19:23 jalaziz joined #salt
19:24 whiteinge not that there isn't a place for them, but returners have been around since the beginning
19:24 whiteinge and salt's event system has been changing older features little-by-little since it's introduction
19:26 manfred i really think the power for returners comes in with monitoring states.  and monitoring statistics over time. stored in a database.
19:27 whiteinge manfred: no disagreement there. but seealso:: https://github.com/felskrone/salt-eventsd  :-)
19:29 manfred yeah, was looking at that on friday. i like that idea too.
19:32 calvinhp_mac hmm…  doesn't this https://github.com/saltstack/salt/blob/develop/salt/config.py#L2061 conflict with https://github.com/saltstack/salt/blob/develop/salt/utils/parsers.py#L652
19:32 calvinhp_mac one wants to read the users .salt config file and one wants to create a directory called .salt dir to log into
19:35 timoguin whiteinge: i'd love to see that in core as well. it makes more sense to me
19:40 Ryan_Lane whiteinge: well, true, but returners are pretty useful. they can be used without a master, for instance
19:41 mway joined #salt
19:44 haroldjones joined #salt
19:45 mapu joined #salt
19:45 googolhash joined #salt
19:46 bmcorser joined #salt
19:47 bmcorser hi i have issues with salty-vagrant master/minion set up
19:47 zhalas left #salt
19:47 bmcorser i simply cannot get the pair to provision and talk to each other
19:48 jcockhren bmcorser: use a private network
19:48 bmcorser either bootstrapping fails or preseeding fails
19:48 bmcorser i do
19:48 bmcorser jcockhren: the issue is mainly "pre-seeding" the master
19:49 jcockhren so basically, you're saying, yiu can't get the master to accept the seeded minion key so provisioning can happen?
19:51 bmcorser i had that issue at one time, but now using the latest bootstrap-salt shell script and installing vagrant from source, the key doesn't even get seeded
19:51 bmcorser ie i cant see it when i run salt-key
19:53 timoguin i've been noticing more people having issues with salty-vagrant lately.
19:53 bmcorser yes
19:53 timoguin wonder if there are some bugs there...
19:53 whiteinge Ryan_Lane: good distinction
19:53 * akoumjian guiltily been MIA from salty-vagrant development for months now
19:54 bmcorser the docs on the vagrant site aren't too in depth (i am a vagrant noob and ruby sceptic)
19:54 whiteinge Ryan_Lane: have you played with the minion-only event bus at all?
19:54 Ryan_Lane whiteinge: also, I can very heavily customize what a returner does
19:54 * whiteinge has not
19:54 Ryan_Lane I haven't
19:54 akoumjian All kinds of people have been committing changes to it over the past months, so I'm not even sure what's in the latest Vagrant release
19:54 Ryan_Lane that also assumes you're running a minion
19:54 Ryan_Lane currently I'm using salt with no daemons
19:54 whiteinge ah, right
19:54 Ryan_Lane (at lyft anyway)
19:55 Ryan_Lane (wikimedia uses master/minion)
19:55 thedodd joined #salt
19:56 bmcorser random ruby question: whats the difference between {k: 'v'} and {'k' => 'v'}
19:56 jcockhren bmcorser: nothing
19:56 bmcorser are they different data types?
19:56 bmcorser ok
19:56 Ryan_Lane the former is the default syntax
19:56 Ryan_Lane err
19:56 Ryan_Lane the latter is
19:56 thayne joined #salt
19:57 bmcorser cool
19:57 bmcorser back to salty-vagrant .. it's not the run_highstate that seeds the keys is it?
19:58 manfred whiteinge: someone needs to get the reactor doing something on the minion :P
19:58 timoguin bmcorser: i don't think so.
19:58 bmcorser ok
19:58 timoguin i think it copies the keys over before trying to run highstate
19:59 manfred whiteinge: do you happen to know if this eventd works will listen to the minion event bus?
20:00 manfred looks like just the master
20:00 bmcorser ok timoguin
20:02 KyleG1 joined #salt
20:04 mway Anyone have any experience managing galera cluster with salt?  Specifically with having N identical nodes forming a cluster (1 node starting the cluster, N-1 joining) without using grains/matching to designate a single minion as the starting node, so that other minions can join the cluster without events/additional states?
20:06 sadburger joined #salt
20:08 KyleG joined #salt
20:11 DammitJim joined #salt
20:12 DammitJim can one use salt to deploy applications to multiple machines or is that a different tool?
20:13 srage joined #salt
20:13 Corey You can do that presuming a semi-sane deploy process, yes.
20:13 manfred that is one of the things it can do
20:13 Ryan_Lane DammitJim: I've been working on this for a while: https://github.com/trebuchet-deploy
20:14 Ryan_Lane you can, of course, just use salt's file server as well
20:16 bmcorser can i generate keys for use with salt using ssh-keygen?
20:17 meteorfox|afk joined #salt
20:19 Ryan_Lane are parameters passed into modules ordered or not ordered?
20:19 Ryan_Lane meaning, if I pass in a json string, will it get rendered into an ordereddict, or a dict?
20:20 Ryan_Lane hm. I guess it's a string....
20:20 Ryan_Lane how about if that data is being passed into a state?
20:20 StDiluted joined #salt
20:21 wendall911 joined #salt
20:21 DammitJim Ryan_Lane, what does that mean?
20:21 manfred bmcorser: you can do it with salt-key --gen-keys
20:21 Ryan_Lane well, say for instance I'm writing a state module
20:21 Ryan_Lane and it's using pillar data
20:21 StDiluted afternoon. I am working on commenting out a line in a file and simultaneously commenting another line. using file.comment and file.uncomment doesn’t seem to work in the same state ID, because they are ducplicate delcarations. What’s the best way to do this?
20:22 Ryan_Lane will that pillar data be ordered, or not ordered?
20:22 Ryan_Lane and how about if rather than a pillar, it's being passed into the state module via __opts__?
20:22 manfred StDiluted: can't have multiple <state>.* in the same state ID
20:22 manfred StDiluted: just have to make multiple state id's
20:23 StDiluted hrm. Tried that too.
20:23 ahammond I'm using cmd.run to run git fetch. Is there an easy way to differentiate between when fetch actually gets new commits and when it doesn't?
20:23 StDiluted ah never mind i found waht i did
20:24 ajw0100 joined #salt
20:26 manfred ahammond: git doesn't return anything but 0, even if it is up to date, so ... not really, maybe check and see if the git module has something for this
20:26 DammitJim can one efficiently do it in Windows?
20:27 ahammond manfred will do. thanks!
20:27 manfred ahammond: ideally, git would turn a 1 if it didn't actually fetch anything
20:27 StDiluted the line I’m trying to comment has a ${BLAHDEBLAH} in it, and doesn’t seem to be being discovered. Is there a way to escape that in my regex declaration?
20:27 manfred but it doesn't
20:27 manfred and neither -q nor -v does any good
20:27 Ryan_Lane ahammond: you should be able to check FETCH_HEAD
20:28 Ryan_Lane manfred: well, not necessarily, because fetching nothing is still a successful fetch ;)
20:28 StDiluted specifically:     - regex: "wrapper.java.command=${JAVA_COMMAND_LOCATION}"
20:29 manfred but unsuccessfull in the sense that there was nothing to fetch
20:29 StDiluted that’s not getting found
20:29 gadams999 joined #salt
20:29 manfred in a similar sense, | grep something that doesn't match anything returns a 1, even though it was successfull and didn't find anything
20:29 manfred ¯\(°_o)/¯
20:30 manfred would just be nice to have something differentiate :)
20:30 * Ryan_Lane nods
20:30 manfred ahammond: doesn't look like the git.fetch module will be much help either, cause it is just going to look at git's retcode as well https://github.com/saltstack/salt/blob/develop/salt/modules/git.py#L258
20:31 ahammond hmm, I can't be the first guy who wants to do this...
20:31 analogbyte joined #salt
20:31 ahammond the git.latest module uses a pull, which is a fetch / merge. I want to do a fetch / reset --hard (because sometimes people edit history in such a way that breaks merges)
20:32 manfred the problem is that git doesn't give a way to do it, we may be able to alter the git module to just match that all of them are up to date. but i would be supprised if they would be merged, might want to have a custom module
20:32 manfred /srv/salt/_modules
20:32 Ryan_Lane ahammond: basically you'll always need to do a fetch, then have something that requires it that checks FETCH_HEAD vs HEA
20:32 Ryan_Lane *HEAD
20:33 Ryan_Lane if they don't match, then you'll do a reset
20:33 ahammond that makes sense. so... how do I write a state that does that? use the unless: and then some script?
20:33 ahammond ugh.
20:34 ahammond how do I check FETCH_HEAD?
20:34 Ryan_Lane get the commit id of it
20:34 Ryan_Lane and the commit id of head, and compare them
20:35 Eugene man git-rev-parse
20:35 Eugene Oh, hah.
20:35 StDiluted ah, had to escape the $ and the { and the } ;)
20:36 Eugene I'd normally point you at the bot's !deploy factoid, but we don't have gitinfo here.
20:36 JasonSwindle joined #salt
20:37 TyrfingMjolnir joined #salt
20:37 opapo joined #salt
20:37 cruatta joined #salt
20:37 grep_awesome joined #salt
20:42 stevendgonzales joined #salt
20:44 stevendgonzales Not sure if any reviewers for saltstack formulas are out here
20:44 stevendgonzales but I have made a PR on the redis formula https://github.com/saltstack-formulas/redis-formula/pull/11
20:44 stevendgonzales Thanks!
20:45 doanerock why am I getting a error when I try to run  salt-call state.highstate --local    https://gist.github.com/doanerock/11155947
20:47 manfred what version of salt? it is working for me here on develop
20:47 Ryan_Lane stevendgonzales: I'm confused about that change
20:47 Ryan_Lane it's possible to set the version via pillars
20:47 stevendgonzales sure
20:47 stevendgonzales yes, however
20:47 Ryan_Lane so this means the formula will hardcode a version that may be insecure
20:47 stevendgonzales the download uri used in the recipe only links to beta, stable, old
20:48 stevendgonzales no
20:48 stevendgonzales there is a separate donload uri
20:48 wendall911 joined #salt
20:48 Ryan_Lane +{% set version = redis.get('version', '2.8.8') -%}
20:48 stevendgonzales so as it is before the change tar is downloaded form http://download.redis.io/redis-{{ version }}.tar.gz
20:49 stevendgonzales this particular link only allws the replacement of version with stable, beta, etc
20:49 Xiao joined #salt
20:49 stevendgonzales so you have a default checksum in the formula
20:49 Ryan_Lane ah, right. true.
20:49 Ryan_Lane that's not terribly helpful
20:49 stevendgonzales and when redis releases a new stable version, the formula breaks, because the stable bits dont match the cheksum
20:49 Heartsbane UtahDave: ping
20:50 stevendgonzales so, i changed the download uri http://download.redis.io/releases/redis-{{ version }}.tar.gz.
20:50 chrisjones joined #salt
20:50 stevendgonzales this uri allows the ownload by specifying a specific version
20:50 * Ryan_Lane nods
20:50 stevendgonzales thus, i put a default version of 2.8.8, and updated it to the corresponding checksum
20:50 manfred https://i.imgur.com/bm4IPHg.jpg
20:50 manfred Air cooled
20:51 stevendgonzales So now the user can ovveride the pillar data with a specific version number, even a specific build, and link tot he appropriate checksum
20:51 stevendgonzales So if i want t install 2.8.6 for some reason, i can
20:51 conan_the_destro joined #salt
20:51 stevendgonzales Ryan_Lane does that make more sense?
20:51 * Ryan_Lane nods
20:52 Ryan_Lane I don't think there's any really ideal for this when downloading from source, but this is indeed better than what's currently there
20:53 stevendgonzales yes, at least if you run the formula with the defaults it wont fail
20:53 UtahDave hey, Heartsbane!
20:53 stevendgonzales installing form package doesnt work very well on this formula either
20:53 Heartsbane UtahDave: are you going to be at the meetup tomorrow?
20:54 doanerock joined #salt
20:54 stevendgonzales i think the ubuntu version is 2.2 and their are some breaking config values in the 2.4 and 2.6 config version in the formula
20:54 UtahDave Heartsbane: yep!  are you coming?
20:54 Heartsbane I have a meeting with a vendor up there at 7:30PM so I was going to cut out early
20:54 stevendgonzales Ryan_Lan I weill try and clean it up as i use it
20:55 analogbyte joined #salt
20:55 Heartsbane But I was wondering if I could pick your brain about this deploy method
20:55 stevendgonzales but i figured small PRs are better
20:55 UtahDave yeah, defiinitely
20:55 Ryan_Lane stevendgonzales: yeah, I was going to push in a fix for that. I had a couple bad config values
20:55 TyrfingMjolnir joined #salt
20:55 Heartsbane UtahDave: so I was wondering if I could come up early
20:56 UtahDave Yeah, I'll be here!
20:57 Heartsbane because pillar='{derp_version:"6.0.12-15189"}' is harshing my mellow
20:57 Heartsbane where all versions of derp are a legacy app that should DIAF
20:57 UtahDave lol.  ok
20:57 schimmy joined #salt
21:05 faldridge joined #salt
21:07 srage joined #salt
21:08 bhosmer joined #salt
21:09 [diecast] joined #salt
21:09 stevendgonzales Ryan_Lane Has anybody talked about adding Vagrantfile to salt-formulas so that you can use the provisioner to test the formula when working on it?
21:10 Ryan_Lane dunno. I'm just another volunteer dev ;)
21:10 stevendgonzales ah cool
21:10 stevendgonzales well I like the idea!
21:11 Ryan_Lane I'd like to have a vagrant file in salt's normal repo too
21:12 timoguin yes, Vagrantfiles everywhere!
21:12 jrdx joined #salt
21:12 timoguin and .kitchen.yml
21:14 timoguin and then stick them on jenkins for CI
21:14 whiteinge manfred: salt-eventsd may Just Work (TM) with the minon event bus if you switch which socket it listens to. i would hope the events on the minion have the same basic format, but I have never looked at them to be certain.
21:14 manfred they have the same format, yeah you would just need to get the hash of the minion_id and it should be able to find the /var/run/salt/minion/*.pub.ipc
21:15 sm1ly joined #salt
21:15 sm1ly re2all. Detected conflicting IDs, SLS IDs need to be globally unique.
21:15 sm1ly The conflicting ID is "salt-minion" and is found in SLS "base:ejmysql" and SLS "base:network" what does he mean?
21:15 haroldjones joined #salt
21:15 joehoyle joined #salt
21:16 timoguin sm1ly: that means you have two states named salt-minion, one in ejmysql and one in network
21:16 TheGerm joined #salt
21:16 sm1ly timoguin, oooh, understand, my mistake. thx
21:16 sm1ly need some sleep)
21:17 kermit joined #salt
21:18 TheGerm question for everyone, have you had issue upgrading salt? i am trying to go from pip install to pkg install and it is not a fun process so far
21:18 whiteinge stevendgonzales: thanks for the pull req!
21:18 whiteinge stevendgonzales: +1 to a vagrant formula
21:19 thayne joined #salt
21:21 mway joined #salt
21:21 alunduil joined #salt
21:22 TheGerm my issue seems to be a path issue with the deb pkgs…. when running "which salt-minion" i get /usr/bin/salt-minion… when i try to run "salt-minion --version" i get the error "-bash: /usr/local/bin/salt-minion: No such file or directory"….. any thoughts?
21:23 sm1ly I got reconfigure system by salt. versions of solt correct and ==. but I cant use this minion yet. I try to reinstall. to delete key on master. and now I dont see key on master at all. nothing in log
21:24 TheGerm @ sm1ly make sure to remove the keys from the minion as well before trying to reconnect to the master
21:25 sm1ly TheGerm, doesnt help too
21:25 TheGerm what version are your running?
21:25 sm1ly salt-minion 2014.1.0
21:26 sm1ly its only one minion of 47 which brokes(
21:26 TheGerm hmmm… didy ou check your iptables or selinux?
21:26 TheGerm on that one minion?
21:27 ZombieFe_ joined #salt
21:27 sm1ly TheGerm, selinux disabled, iptables got right rules
21:27 TheGerm okay… can you ping the minion from the master?
21:27 sm1ly yeah
21:28 TheGerm minion to master?
21:28 yomilk joined #salt
21:28 sm1ly TheGerm, this minion is the part of my galera cluster. salt similary got web iface to control the cluster. and its work. dont work only minion(
21:28 sm1ly yes I can minion to maser
21:29 diogobaeder joined #salt
21:29 haroldjones joined #salt
21:30 timoguin whiteinge: not a vagrant formula, specifically, but distributing a Vagrantfile along with each formula
21:30 sm1ly TheGerm, and if I try salt-call state.highstate it works too) lol
21:31 TheGerm haha… well that doesnt help
21:31 whiteinge timoguin: ah, right. s/$/file/
21:31 whiteinge timoguin: +1 :)
21:33 whiteinge that root formula directory is fair-game for vagrant/travis/etc files. salt will ignore them
21:35 whiteinge if someone is interested in adding one to a couple repos and adding that to the formula conventions doc, we'd definitely accept that
21:38 shoma joined #salt
21:43 TheGerm anyone able to help with the path issue? from a pkg install
21:44 Gareth_ joined #salt
21:44 joehoyle joined #salt
21:45 otsarev_ joined #salt
21:45 whiteinge TheGerm: sounds like the /usr/local/bin/salt* scripts are still in place. if ``pip remove`` didn't get them you can nuke them yourself
21:46 TheGerm nice! thanks whiteinge… i will check that out
21:46 TheGerm nope not there
21:46 whiteinge dang
21:47 freelock joined #salt
21:47 whiteinge TheGerm: what is the she-bang at the top of /usr/bin/salt-minion set to?
21:48 TheGerm #!/usr/bin/python
21:48 kermit joined #salt
21:49 whiteinge are there any left over salt files in the python dir /usr/local/lib/python2.X/dist-packages/salt*  ?
21:49 whiteinge ^^ path may not be 100% accurate
21:49 whiteinge ...don't have a running ubuntu box to double-check with
21:51 NullWagon joined #salt
21:51 TheGerm checking
21:52 bastion2202 joined #salt
21:52 whiteinge a quick way to check what paths python could be looking in is: python -c 'import sys; print sys.path'
21:52 jslatts has anyone used S3 for a source for the file.managed state?
21:53 garthk joined #salt
21:53 jslatts http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html seems to indicate that it is possible
21:53 TyrfingMjolnir joined #salt
21:53 timoguin /usr/share/pyshared/salt is where ubuntu puts the salt-common files
21:54 TheGerm this folder /usr/lib/python2.7/dist-packages/salt is full with salt stuff…. is this leftover from pip?
21:55 whiteinge TheGerm: that's probably the folder installed via the .deb
21:55 whiteinge timoguin: ah, yes. that rings a bell. thanks
21:56 whiteinge there should be a /usr/local variant of that, yes?
21:56 whiteinge if installing via pip, i mean
21:57 timoguin yes, /usr/local/lib/python2.7/dist-packages for pip on ubuntu
21:58 TheGerm nothing there "/usr/local/lib/python2.7/dist-packages: No such file or directory"
21:58 TheGerm i have been pounding my head on this since yesterday haha
22:00 TheGerm also as of now the with the deb pkg is immediatly dies upon startup
22:01 srage joined #salt
22:01 yusuket joined #salt
22:04 ndrei joined #salt
22:05 TheGerm okay just tried uninstall all of salt….. and /usr/lib/python2.7/dist-packages/salt/ remained…. removed that folder and reinstalled via apt-get and still no luck same thing
22:05 Gareth joined #salt
22:07 stevendgonzales yes whiteinge
22:07 stevendgonzales i meant adding Vagrantfile to each formula
22:07 stevendgonzales that way you can add jenkins or travis jobs
22:07 stevendgonzales that run the build
22:07 whiteinge TheGerm: is there more than one result for this? : find /home -path '*salt/version.py'
22:08 whiteinge TheGerm: er. /
22:08 TheGerm yes
22:08 whiteinge stevendgonzales: yeah. (i'm a little slow today.) that would be killer
22:08 elfixit joined #salt
22:08 TheGerm here: /usr/share/pyshared/salt/version.py, /usr/lib/pymodules/python2.7/salt/version.py, /usr/lib/python2.7/site-packages/salt/version.py
22:08 whiteinge gah
22:09 TheGerm i can remove everything clear out those dir's and see whats happens
22:09 whiteinge TheGerm: yeah. also nuke any files named 'salt*' in those base directories
22:09 whiteinge e.g., /usr/share/pyshared/salt*
22:09 whiteinge etc
22:09 TheGerm okay… give me a couple
22:12 Ryan_Lane UtahDave: so, the codename vs the number releases is incredibly annoying
22:12 TheGerm okay removed everything, after uninstalling… then reinstalled
22:12 TheGerm same issue
22:12 TyrfingMjolnir joined #salt
22:12 TheGerm salt-minion --version: -bash: /usr/local/bin/salt-minion: No such file or directory
22:13 TheGerm which salt-minion: /usr/bin/salt-minion
22:13 UtahDave Ryan_Lane: Yeah, I can never remember which is the next one.
22:13 Ryan_Lane UtahDave: well, the biggest issue is that --version reports the number
22:13 TheGerm and process never lives on startup
22:13 Ryan_Lane and all of the docs say "available in <codename>"
22:14 Ryan_Lane UtahDave: is there an equivalent to this before helium: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkg.html#salt.states.pkg.uptodate ?
22:14 whiteinge TheGerm: sheesh. debian? which version did you install?
22:15 UtahDave I don't think so
22:15 TheGerm Ubuntu 12.04.4…. salt-minion-2014.1.1
22:15 whiteinge possible there's a problem with the package. though i'm still leaning toward some errant python file somewhere from the pip install
22:15 whiteinge TheGerm: installed via the PPA?
22:15 TheGerm i agree with you
22:15 TheGerm fresh machine everything works fine
22:15 whiteinge ah
22:16 TheGerm this is upgrading from pip on a current machine
22:16 TheGerm so remove pip install of salt… then install via pkg
22:17 whiteinge pip lays down -egg directories and -egginfo files in addition to the salt source files. can you remove the package and do a system-wide search for salt and just nuke everything?
22:17 TheGerm my steps for this on CentOS worked without issue… tried to copy on Ubuntu and of course had to give issues haha
22:17 whiteinge haha
22:17 whiteinge of course! ;)
22:17 TheGerm yea give me a second
22:19 TheGerm okay no traces of salt found on the system… reinstalling via apt-get
22:20 yomilk joined #salt
22:21 TheGerm same thing… process dies instantly…. and salt-minion --version looking in /usr/local/bin/
22:21 Eugene Ryan_Lane - I've been using `cmd.run "yum update -yq"`, but no, there is not.
22:21 Ryan_Lane UtahDave, whiteinge: can you provide an example for http://docs.saltstack.com/en/latest/ref/states/all/salt.states.module.html#module-salt.states.module that shows passing in kwargs?
22:21 Eugene One of the big reason I'm looking forward to He
22:22 CeBe how do I configure a state to run a command after it updaded a file managed by this state?
22:22 CeBe in this case run   postmap  command on the file it has created
22:22 jtang1 joined #salt
22:22 whiteinge Ryan_Lane: http://paste.fedoraproject.org/95770/98118943   like that?
22:23 whiteinge it's just - kwarg: val
22:23 whiteinge same indentation as the main args
22:24 andrej The latest salt update on ubuntu 12.04 seems to have broken things
22:24 andrej my minion can't communicate to the master - have there been reports along those lines already?
22:24 andrej Minion PULL socket URI: ipc:///var/run/salt/minion/minion_event_8a6de61b96fb02d932bc49f886927dcc944cc48b55b143ea84d16dee8877a076389177776c6bdc43ea3fea37b5fa8412740d18c6421b0b02b3e503e1c6e2c233_pull.ipc
22:24 Ryan_Lane whiteinge: can you update the docs?
22:24 andrej [ERROR   ] An un-handled exception was caught by salt's global exception handler:
22:24 andrej ZMQError: File name too long
22:25 ajolo joined #salt
22:25 Eugene andrej - 2014.1? Yes, you need to update the master falt.
22:25 Eugene This is in the release notes
22:25 andrej master falt?
22:25 Eugene first
22:25 * Eugene is sick and in bed, no alcohol to power his brain :-(
22:26 andrej heh
22:26 andrej poor thing - go to sleep!
22:26 manfred Eugene: that is a shame.  I am shaving with my straight razer, and a bottle of scotch int he other hand
22:26 Ryan_Lane whiteinge: also, what if I don't want to name the resource event.fire_master, but wanted to use: fire_master_event
22:27 Ryan_Lane I can't figure out the syntax for that
22:27 ajolo joined #salt
22:27 Eugene CeBe - cmd.run, with a require: for the file
22:27 Kraln joined #salt
22:28 andrej Hmmm ick
22:28 CeBe cmd.run  in a state will run that command once when I apply the state, or does it try to run it later when conditon changes?
22:28 Ryan_Lane whiteinge: ah: http://pastebin.com/GvEuxc6b
22:28 Ryan_Lane but, yeah, docs could use some help there. it was pretty difficult to figure out
22:28 andrej master & minion are on 2014.1.3-1precise1 - I'm still getting the error. And yes, both processes have been restarted
22:29 manfred CeBe: should have an unless or onlyif, to say only run the command if the condition is met
22:29 whiteinge Ryan_Lane: nice. yeah, i'll get an example in there
22:29 Eugene CeBe - it runs any time that state is invoked. postmap is idempotent, so that shouldn't be a problem.
22:29 analogbyte joined #salt
22:29 Ryan_Lane whiteinge: thanks
22:29 CeBe manfred, Eugene: cool, thanks!
22:30 rawtaz fancy that
22:30 doanerock joined #salt
22:31 CeBe rawtaz: currently moving my whole self-made server management scripts to salt.
22:31 CeBe getting much cleaner with salt :)
22:32 rawtaz nice
22:32 andrej So ... if master and minion are on the same latest release, and the minion still can't talk to the master w/ the same error
22:33 andrej where else do I look to resolve this?
22:34 CeBe Eugene: same for restarting a service after applying state?
22:34 manfred andrej: taht is an incredibly long file name, how long is your minion_id ?
22:34 Eugene CeBe - yup. Relatively straightforward, really.
22:35 andrej the fqdn is 28
22:35 andrej and it worked till Thursday last week ;}
22:37 UtahDave andrej: what's the output when you start the minion in the cli?    sudo salt-minion -l debug
22:37 happytux joined #salt
22:37 possibilities joined #salt
22:37 andrej I'll pastebin it
22:38 calvinhp_mac joined #salt
22:40 kickerdog joined #salt
22:41 andrej http://pastebin.com/6bG2kgQk
22:41 jtang1 joined #salt
22:44 andrej back in a sec
22:44 UtahDave andrej: what os are you on?
22:44 andrej ubuntu 12.04
22:45 UtahDave this is  2014.1.3?
22:45 andrej yah
22:45 andrej updated this morning
22:45 UtahDave have you modified /etc/salt/master?
22:45 andrej first thing after the long weekend
22:45 andrej nope
22:46 UtahDave er,  how about /etc/salt/minion
22:46 andrej nope
22:46 andrej -rw-r--r-- 1 root root 19546 Apr  4 15:56 minion
22:46 UtahDave hm.  OK, can you open an issue on this on github?
22:46 andrej sure can
22:47 hhenkel UtahDave: 2014.1.3 is the latest?
22:47 meteorfox joined #salt
22:47 hhenkel UtahDave: If so, maybe you or someone else should update the topic...
22:47 UtahDave It hasn't been officially announced quite yet, but it's being packaged right now.  Looks like the ubuntu packages have already hit the ppa
22:51 andrej done
22:51 andrej kept it quite brief, just a title stating the obvious, and the paste in the text field
22:53 Ryan_Lane why does salt-call --local try to send an event into the minion socket?
22:54 mpanetta joined #salt
22:54 Ryan_Lane http://pastebin.com/Q9xyJLWF <-- this doesn't seem to make much sense if there's no master or minion daemon
22:55 UtahDave Ryan_Lane: there's a local minion event bus
22:55 Ryan_Lane right, but what if there's no minion daemon?
22:55 mpoole hey, so I'm writing a deployment system using salt and I wondered, if I'm using client.cmd_async to send 'deploy' commands to multiple servers, whats the best way to check a) the command succeeded b) the output?
22:55 mpoole client.get_cache_returns seems to get the message
22:55 srage joined #salt
22:55 mpoole but I think i'm dumbing out on a function that gives me the return code
22:55 Ryan_Lane that call happens twice and each time adds 3-4 seconds
22:56 mpanetta joined #salt
22:56 Ryan_Lane it adds almost 10 seconds to a highstate run and there's nothing to disable it
22:56 Ryan_Lane (unless I'm wrong and there is indeed something to disable it :) )
22:57 UtahDave Ryan_Lane: Hm. I'm not sure if it can be disabled, or if it's even necessary.  3 to 4 seconds is an awfully long time. I wonder what it's waiting for.
22:57 Ryan_Lane it may be waiting on a timeout
22:57 cewood joined #salt
22:58 Ryan_Lane you'd think at minimum it would be a non-blocking call :)
22:58 Ryan_Lane should I open an issue on this?
22:58 yusuket joined #salt
22:58 UtahDave Ryan_Lane: Yes, please do.  I'd love to get that sorted out'
22:59 UtahDave I've always felt like salt-call was slow to get started
22:59 cewood joined #salt
22:59 manfred yeah, just double checked, it doesn't create an event bus either
23:00 UtahDave maybe it's waiting for the event bus and timing out.
23:00 mway joined #salt
23:00 manfred that is what it sounds like mr. lane was observing
23:01 Ryan_Lane https://github.com/saltstack/salt/issues/12173
23:02 UtahDave cool, thanks, Ryan_Lane!
23:02 Ryan_Lane yw
23:02 sealibora joined #salt
23:03 Ryan_Lane I hope I'm not giving you too much work. I feel like I'm constantly bothering all of you :D
23:04 meteorfox joined #salt
23:04 UtahDave no, never a bother, man.  We love help!
23:04 rawtaz if you weren't someone else would have to do it
23:04 Ryan_Lane heh
23:06 UtahDave all right. time to head home.  Catch you all tomorrow!
23:06 manfred peace o/
23:09 jslatts joined #salt
23:12 meteorfo_ joined #salt
23:14 andrej https://github.com/saltstack/salt/issues/12172
23:14 andrej Just as a follow-up :)
23:16 elfixit joined #salt
23:17 sealibora joined #salt
23:18 mpoole hey, is there anything like get_cache_returns that will give me the return code?
23:18 doanerock I am even having trouble getting running salt-call "  root@ubuntu:/srv/salt# salt-call --version     ,   %salt-call %s"
23:23 shoma joined #salt
23:25 pjs "<@UtahDave> no, never a bother, man.  We love help!" <- MAJOR reason why Salt stands out amongst it's 'competitors' and OSS communities in general.
23:27 rawtaz yeah, that is respectful
23:27 andrej true that
23:29 rawtaz ansible doesnt seem bad either as a software, but it doesnt feel equally open
23:29 rawtaz which is sad
23:29 rawtaz (those two are what im looking at)
23:30 pdayton joined #salt
23:31 tyler-baker joined #salt
23:31 tyler-baker left #salt
23:31 sealibora joined #salt
23:33 Ryan_Lane doanerock: can you expand on your problem a bit?
23:34 doanerock Ryan_Lane,  sure https://gist.github.com/doanerock/11160183
23:34 doanerock does that help
23:38 jplaza joined #salt
23:39 andrej doanerock : which version of *buntu are you on, and what does dpkg say about your salt?
23:39 Ryan_Lane doanerock: weird. how do you have salt installed?
23:41 rawtaz DO you have salt installed? :-)
23:41 doanerock Ryan_Lane, I installed using aptitude it is ubuntu 12.04 LTS
23:42 joehh1 doanerock: can you do a dpkg -l | grep salt
23:42 doanerock i mean Description:Ubuntu 12.10
23:43 doanerock added to https://gist.github.com/doanerock/11160183
23:43 joehh1 wow - really old packages
23:43 andrej heh
23:43 Ryan_Lane indeed.
23:43 andrej geriatric, even
23:43 Ryan_Lane doanerock: you may want to use the PPA
23:44 Ryan_Lane doanerock: https://launchpad.net/~saltstack/+archive/salt
23:44 joehh1 apt-add-repository ppa:saltstack/salt
23:44 Gareth_ left #salt
23:45 bastion2202 joined #salt
23:45 joehh1 how did you get those packages? we should remove them from wherever they are hosted if we can
23:46 joehh1 ahhh - part of ubuntu
23:46 joehh1 http://packages.ubuntu.com/quantal/salt-common
23:46 joehh1 not sure we can remove them
23:47 joehh1 we have similar issues with saucy - fixed though in trusty
23:49 it_dude joined #salt
23:51 CeBe how do I configure a service and a package that have the same name? https://gist.github.com/cebe/032dd8ee8168dbe7bbe7
23:51 CeBe this gives me          Rendering SLS "base:mail.mx.postfix" failed: Conflicting ID "postfix"
23:52 sealibora joined #salt
23:55 Ryan_Lane CeBe: this should work: https://gist.github.com/ryan-lane/4ba08c8de48317cfc662
23:56 davet joined #salt
23:57 CeBe Ryan_Lane: nope: Rendering SLS "base:mail.mx.postfix" failed: Unknown yaml render error; line 3
23:57 kickerdog1 joined #salt
23:58 Ryan_Lane I always have a hard time remembering how the combined syntax works
23:59 Ryan_Lane CeBe: for sure this works: https://gist.github.com/ryan-lane/4ba08c8de48317cfc662
23:59 Gareth joined #salt

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