Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-09-03

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

All times shown according to UTC.

Time Nick Message
00:03 mordonez_ joined #salt
00:03 DaveQB Upgraded to 2014.1 and no worries the first few days, now remote minions are saying there is no top.sls file etc. The master's minion runs highstate fine though. What's the best way to troubleshoot that?
00:04 aparsons_ joined #salt
00:04 ropes joined #salt
00:05 gzcwnk there is a bug where remote minions dont respond
00:05 gzcwnk its killing my setup
00:08 DaveQB gzcwnk: Oh. My minions respond fine though, I mean I can run cmd.run and test.* modules
00:08 gzcwnk k
00:09 DaveQB gzcwnk: It is the builtin fileserver that is the issue for me
00:09 gzcwnk OK...so you are getting the top.sls from a salt master?
00:10 DaveQB gzcwnk: no, that's the problem I believe
00:10 DaveQB gzcwnk: I am getting "No Top file or external nodes data matches found"
00:10 gzcwnk ok so you are trying to and its failing?
00:10 DaveQB gzcwnk: Yes,
00:11 DaveQB gzcwnk: I have tried running the minion in debug mode but it is not revealing much more into (I wish it timestamped things too as a side topic)
00:11 gzcwnk ive not seen that and I get a lot of files this way
00:11 DaveQB gzcwnk: Been using salt for a few years in production at my work, never had this problem or anything like it
00:12 DaveQB gzcwnk: Yeah it is an odd one. I have checked the file_roots is correct and matches the file system
00:12 gzcwnk is this self compiled? or an rpm or something?
00:13 DaveQB gzcwnk: PPA deb packages
00:13 DaveQB gzcwnk: I might try downgrading back to 0.17.5 and see if this issue is still there
00:14 gzcwnk i was wondering that, ive had issues with 1.4 and 1.5 fixed in 1.7 rpms
00:15 gzcwnk a memory leak
00:15 yomilk joined #salt
00:16 dude051 joined #salt
00:17 DaveQB gzcwnk:  1.4, 1.5??
00:18 gzcwnk 2014.1.4 and 2014.1.5
00:18 snuffeluffegus joined #salt
00:18 gzcwnk rpm packages
00:19 gzcwnk im curently on salt-2014.1.7-3.el6.noarch
00:27 DaveQB gzcwnk:  Oh I see
00:29 littleidea joined #salt
00:29 jalaziz joined #salt
00:30 aparsons joined #salt
00:31 bhosmer_ joined #salt
00:33 yomilk joined #salt
00:36 oz_akan joined #salt
00:38 blackhelmet joined #salt
00:39 n8n joined #salt
00:41 blackhelmet Hi all. Question about states and windows installers. I want to create a state where if a specific application is not installed then the MSI will be copied from the salt fs and then installed.
00:41 blackhelmet This surely isn't complicated but I'm getting completely muddled between understanding the difference in defining a state and defining corrective action to get into a state
00:41 blackhelmet Anyone done something similar?
00:42 mosen Havent tried on windows
00:42 mosen the state module should take care of the corrective action
00:42 manfred blackhelmet: first thing you should do, make a module that installs the .msi or checks if it is installed
00:43 manfred then write your state, it should first check if it is installed... and if it isn't, then call the module to install it
00:43 blackhelmet You've nailed the hazy part
00:43 blackhelmet What is the difference in state and a module?
00:43 blackhelmet Example
00:43 blackhelmet In one state I have sun-java
00:43 manfred so
00:43 manfred a module is a thing you call from the command line
00:43 blackhelmet The formula from the github repo seems to do it all at once
00:43 manfred salt \* pkg.install <pacakge>
00:44 manfred a state is a thing you put in a state file, and define how the server should look
00:44 manfred something:
00:44 manfred pkg.installed:
00:44 manfred - name: something.msi
00:44 thayne joined #salt
00:44 manfred or whatever
00:44 mosen I would say, an execution module performs an action, and a state figures out which actions are needed to get to the end goal :)
00:44 manfred that will make sure that that msi is installed
00:44 manfred blackhelmet: https://github.com/saltstack/salt/blob/develop/salt/modules/win_pkg.py
00:44 blackhelmet I know pkg is a builtin, but pretending it weren't. IN that example is pkg a 'module'?
00:45 manfred depends on how you are calling it
00:45 manfred if you do salt \* pkg.install ... it is a module
00:45 manfred pkg.installed is a state though
00:45 manfred blackhelmet: https://github.com/saltstack/salt/blob/develop/salt/modules/win_pkg.py#L483
00:45 manfred module
00:45 manfred https://github.com/saltstack/salt/blob/develop/salt/states/pkg.py#L385
00:45 blackhelmet Thanks manfred, I feel a bit closer
00:45 manfred state
00:45 blackhelmet Looking at your link now
00:45 blackhelmet ok
00:46 blackhelmet A 'module' can have both state and actions then
00:46 manfred you have execution modules, and state modules, execution modules are called from teh commandline, state modules are called from state files
00:47 blackhelmet Oh, I get it now
00:47 manfred state modules usually use execution modules to A) see if the state is already present and done or B) configure the minion so it is in that state
00:47 blackhelmet I thought the two links were same file, diff positions
00:47 manfred check here
00:47 blackhelmet They are completely different though
00:47 aparsons joined #salt
00:47 manfred yeah
00:47 blackhelmet This is where I was messing it all up. I thought a 'module' was like a class with state and execution all in a package
00:47 manfred blackhelmet: here https://github.com/saltstack/salt/blob/develop/salt/states/pkg.py#L815
00:47 manfred this is where the pkg state, uses the meta pkg.install module to install your package
00:48 manfred execution module*
00:48 blackhelmet Got it
00:48 blackhelmet Seriously, thank you
00:48 manfred np :)
00:48 blackhelmet Such a minor thing but brought mounds of clairty
00:48 blackhelmet **clarity
00:48 manfred indeed, i like the way it works, it makes it inifitely easier to expand anything
00:48 manfred and be able to reuse stuff
00:49 blackhelmet Another one, once you call highstate is state constantly enforced or do you need to call it again each time you want to enforce state?
00:49 manfred each time you make a change, the highstate needs to be called, or you can do just individual states using the state.sls <state> instead of state.highstate
00:49 blackhelmet Interesting
00:50 blackhelmet So if I wanted to deploy a war everywhere, it would be better to use the execution modules than to call highstate or something similar
00:50 manfred it is not like chef or puppet where it checks in (granted you could do that with the scheduler...)
00:50 blackhelmet See, now that I get modules this is all making more sense
00:50 manfred blackhelmet: depends
00:50 blarghmatey joined #salt
00:50 blackhelmet I'm thinking workflow here. 1) remove from ELB 2) update war 3) unit test 4) join ELB
00:50 manfred blackhelmet: you could use a highstate to update and keep track of the exact version... then you set a pillar to tell the minion which .war to grab and use... and you just have to update the pillar data, then put the server into a highstate, and all of the servers will update
00:51 blackhelmet In that scenario it seems better to do commands than state
00:51 manfred blackhelmet: you can do that in a highstate
00:51 manfred nah
00:51 druonysus joined #salt
00:51 manfred hold on
00:51 blackhelmet I like your approach better
00:51 blackhelmet Feels self documenting since you have pillar data
00:52 blarghmatey I am running into an issue with the mongodb_user state where it is not actually creating the user. It throws a ValueError: dictionary update sequence element #0 has length 1; 2 is required
00:52 blarghmatey The state and full output is here: http://dpaste.com/2FWG3SK
00:52 manfred blackhelmet: do you have health checks on your lb?
00:52 manfred blackhelmet: http://paste.gtmanfred.com/bRzN/
00:52 blackhelmet Yes.
00:52 manfred remove your health check, update, place your health check file back
00:53 blackhelmet nice
00:53 manfred order: 1 so it gets removed first, order: last to make sure it gets put back
00:53 manfred salt \* state.highstate -b 2
00:53 manfred batch the highstate across your environment two servers at a time
00:53 blackhelmet What is -b?
00:53 blackhelmet nice
00:53 blackhelmet or salt 'web*' -b
00:53 blackhelmet That is stellar
00:53 manfred yeah
00:53 blarghmatey If I have to I can do this with a cmd.run that calls out to mongo, but I'd rather not have to do that.
00:54 blackhelmet manfred thank you repeatedly. I'm off to experiment a bit more
00:54 manfred blarghmatey: ... that is really weird... it shouldn't be doing that
00:54 manfred blackhelmet: no problem :) good luck!
00:55 blarghmatey manfred: I did see a github issue with the same error, but it didn't appear to have any resolutino.
00:55 blarghmatey s/resolutino/resolution/
00:55 manfred hold on
00:55 manfred it shouldn't do a dict()
00:56 blarghmatey I'm running 2014.1.10
00:56 manfred oh, interesting
00:56 manfred it should be constructing an ordered dictionary
00:56 manfred blarghmatey: do you only have 1 user currently?
00:57 blarghmatey I created the admin user
00:57 elfixit joined #salt
00:57 blarghmatey That's the only user present in the mongo instance.
00:57 manfred if you do salt-call mongodb.user_list
00:57 manfred what does it return?
00:57 manfred one user?
00:57 manfred yeah, that is the error
00:58 manfred hrm, i don't know enough about mongo to fix it, but it looks like python is acting up
00:58 malinoff joined #salt
01:00 mosen does it depend on an installed package, or just uses cmd.run ?
01:00 jkaye joined #salt
01:00 blarghmatey manfred: The admin user that I created is only present in the admin database and the user I am trying to create is destined for the salt database
01:01 bhosmer joined #salt
01:01 vbabiy joined #salt
01:02 TTimo joined #salt
01:03 manfred i just don't have an environment to test in, i would make a comment on that bug, so it isn't lost in the shuffle, just to make sure that no one forgets about it
01:06 blarghmatey Thanks, will do
01:06 lnxnut joined #salt
01:11 otter768 joined #salt
01:16 vbabiy joined #salt
01:27 melinath joined #salt
01:30 pled76 joined #salt
01:36 TyrfingMjolnir joined #salt
01:46 Alan_S joined #salt
01:53 shaggy_surfer joined #salt
01:53 ekristen joined #salt
01:59 tmh1999 joined #salt
02:02 pled76 joined #salt
02:05 pled76_ joined #salt
02:05 CeBe1 joined #salt
02:05 pled76_ joined #salt
02:08 ajolo joined #salt
02:17 thayne joined #salt
02:18 bhosmer joined #salt
02:21 rallytime joined #salt
02:24 kelseelynn joined #salt
02:25 rome joined #salt
02:25 TheThing joined #salt
02:26 TTimo joined #salt
02:27 Gareth rallytime: I have another commit to push but I'll hold off until you're done with the jenkins changes.
02:29 rallytime thanks Gareth! much appreciated
02:29 rallytime i'm working on setting up a nightly cloud test run, but in order to do so, I've got to change some of the base salt-jenkins files, which is breaking presently..... I'll let ya know when I've got it figured out
02:31 melinath joined #salt
02:32 AdamSewell joined #salt
02:34 pled76 joined #salt
02:48 kelseelynn left #salt
02:48 pled76_ joined #salt
02:48 tmh1999 joined #salt
02:50 xcbt joined #salt
02:56 melinath joined #salt
02:57 jaimed joined #salt
02:59 jalaziz joined #salt
03:01 jkaye joined #salt
03:03 tedski too kludgy? http://pastie.org/9523037
03:04 blarghmatey So, trying to use the v2014.7 tag via the bootstrap script, when I get to a file that I am trying to render using Jinja I get an error saying that The renderer yaml_jinja is unavailable
03:04 tedski managing hosts file entries on dhcp hosts so grains['fqdn_ip4'] returns properly
03:04 blarghmatey On 2014.1.10 I didn't get that issue.
03:10 rangertaha How do run runners without setting up a dev environment?
03:11 rangertaha salt-run nmap.run 192.168.0.0/16
03:12 rangertaha 'salt-run nmap.run 192.168.0.0/16' results in 'Function 'nmap.run' is unavailable'
03:13 rangertaha I put the runners in /etc/salt/_runners and updated the cofig file.
03:13 rangertaha Then restarted the service.
03:15 rallytime Gareth that should be all patched up now. You're good to push your other commit.
03:19 Gareth rallytime: coolness.
03:20 Gareth rallytime: I found some brokeness so it might be a bit.
03:20 rallytime no worries. :)
03:25 cybint joined #salt
03:25 smcquay joined #salt
03:32 NotreDev joined #salt
03:34 miles32 left #salt
03:36 oz_akan joined #salt
03:36 oz_akan joined #salt
03:37 oz_akan joined #salt
03:46 n8n joined #salt
03:47 smcquay joined #salt
03:48 icebourg joined #salt
03:51 rypeck joined #salt
03:55 icebourg joined #salt
03:56 mosen joined #salt
03:57 Nexpro1 joined #salt
03:58 ajw0100_ joined #salt
03:59 tkharju1 joined #salt
03:59 melinath joined #salt
04:04 kelseelynn joined #salt
04:04 icebourg joined #salt
04:06 bhosmer joined #salt
04:09 vbabiy joined #salt
04:17 melinath joined #salt
04:24 yomilk joined #salt
04:26 vbabiy joined #salt
04:29 ajolo joined #salt
04:30 jalaziz joined #salt
04:33 skullone https://www.youtube.com/watch?v=fvQvFdWIA-M
04:33 skullone holeee geezus, doom is awesome now
04:35 Nexpro joined #salt
04:38 oz_akan joined #salt
04:47 ramishra joined #salt
04:48 jalbretsen joined #salt
04:48 TTimo joined #salt
04:49 felskrone joined #salt
05:01 ramishra joined #salt
05:08 ramishra joined #salt
05:09 kermit joined #salt
05:14 ninkotech joined #salt
05:17 NotreDev joined #salt
05:31 jcockhren joined #salt
05:31 pdayton joined #salt
05:39 oz_akan joined #salt
05:42 ramishra joined #salt
05:43 pdayton joined #salt
05:45 ramishra joined #salt
05:55 bhosmer joined #salt
05:59 xcbt joined #salt
05:59 bhosmer_ joined #salt
06:01 Ryan_Lane joined #salt
06:07 lcavassa joined #salt
06:10 yomilk joined #salt
06:17 jdmf joined #salt
06:19 dccc_ joined #salt
06:21 TheThing joined #salt
06:21 duncanmv joined #salt
06:29 colttt joined #salt
06:38 middleman_ joined #salt
06:39 oz_akan joined #salt
06:40 mariusv joined #salt
06:43 mariusv joined #salt
06:44 kingel joined #salt
06:47 Alan_S joined #salt
06:48 Sweetshark joined #salt
06:51 TTimo joined #salt
06:51 jhauser joined #salt
06:53 tomspur joined #salt
06:53 tomspur joined #salt
07:05 z3uS joined #salt
07:07 sctsang joined #salt
07:08 oz_akan joined #salt
07:12 __number5__ joined #salt
07:14 chiui joined #salt
07:15 sectionme joined #salt
07:25 scott_w joined #salt
07:31 viq joined #salt
07:33 aparsons joined #salt
07:34 j-saturne joined #salt
07:35 occupant joined #salt
07:37 n8n joined #salt
07:43 ianmcshane joined #salt
07:44 bhosmer joined #salt
07:44 aparsons joined #salt
07:51 darkelda joined #salt
07:52 spookah joined #salt
07:54 ramishra joined #salt
07:56 favadi joined #salt
08:00 adam__ joined #salt
08:01 adam__ Hi all, could someone pls point me to some tutorial on creating custom salt state modules?
08:02 mosen did you see the manual topic on states? hangon ill grab the link
08:03 mosen http://docs.saltstack.com/en/latest/ref/states/writing.html
08:06 adam__ thanks a lot!
08:07 mosen no probs
08:09 eliasp joined #salt
08:09 oz_akan joined #salt
08:09 xcbt joined #salt
08:09 agend_ joined #salt
08:21 davet joined #salt
08:28 adam__ Can I use custom states locally on a minion?
08:29 adam__ I created one under /srv/salt/_states/custom_state.py as shown in the example, along with a state file in /srv/salt/custom_state.sls,  but "sudo salt-call --local state.sls custom_state" says the state is unavailable
08:30 adam__ It processes the state file, but does not seem to load the state module.
08:31 agend_ joined #salt
08:33 j-saturne joined #salt
08:34 xintron If I have nested folders to search for a state file, will all matches run or only the first match?
08:35 xintron /foo/state.sls /bar/state.sls; If running state.sls will the first folder (foo/state.sls) run or both?
08:40 ramishra joined #salt
08:42 adam__ Ok, after running "sudo salt-call --local saltutil.sync_states", my custom state module is working.
08:49 jeremyBass1 joined #salt
08:53 TTimo joined #salt
08:54 intellix joined #salt
08:55 Outlander joined #salt
09:00 workingcats joined #salt
09:03 ianmcshane joined #salt
09:04 NotreDev joined #salt
09:05 linjan joined #salt
09:08 N-Mi joined #salt
09:12 alanpearce joined #salt
09:15 vvoody joined #salt
09:15 mordonez_ joined #salt
09:20 darkelda joined #salt
09:21 kingel_ joined #salt
09:24 oyvjel joined #salt
09:25 mordonez__ joined #salt
09:33 bhosmer joined #salt
09:34 giantlock joined #salt
09:35 yomilk joined #salt
09:36 occupant joined #salt
09:41 jayfk joined #salt
09:41 ramishra joined #salt
09:42 jayfk How to check if something is in a list? {% if "foo" in grains['roles'] %} leads to compile errors
09:43 MrTango joined #salt
09:46 adam__ if grains['roles'] is a list object, you could use   grains['roles'].count('foo') > 0
09:46 mordonez joined #salt
09:47 babilen jayfk: Are you sure the former doesn't work? Which error do you get exactly?
09:47 ramishra joined #salt
09:49 mordonez_ joined #salt
09:51 babilen It should work, jinja supports containment checks with in. (e.g. {{ 1 in [1, 2, 3] }} )
09:52 mordonez__ joined #salt
09:53 jayfk babilen: I'm getting a unrelated error "Data failed to compile: ---------- No matching sls found for 'system.base' in env 'base'"
09:54 babilen Do you have system/base.sls or system/base/init.sls ?
09:54 TTimo joined #salt
09:54 jayfk system/base/init.sls
09:55 babilen jayfk: And it is available in file_roots and it renders without a problem? Could you paste that file to, say, http://refheap.com ?
09:55 babilen (assuming you want help debugging this problem)
09:59 jayfk babilen: sure, system/base/init.sls is here: https://dpaste.de/a52C, uwsgi/init.sls is here: https://dpaste.de/DREY
10:00 jayfk babilen: all other includes seem to work, the error occurs only when adding the last if-statement
10:00 babilen (brb in a second)
10:05 babilen jayfk: It might be a problem in *any* of the included SLS files. Could you run your master in debug mode (i.e. stop the service and then "salt-master -ldebug") and check for rendering errors?
10:06 babilen And could you make that 'django' rather than "django" ?
10:07 jayfk babilen: nevermind, got a error in the uwsgi include
10:07 ramishra joined #salt
10:07 babilen I don't see an error in there, what was it?
10:08 jayfk don't know exactly what, but if I include a other file it runs :)
10:08 babilen jayfk: Well, try "uwsgi" rather than "uwgsi" ;)
10:09 martoss joined #salt
10:10 jayfk jayfk: maaaan, those typos always get you! :D
10:10 eliasp I spent once more than a day debugging an apache issue… in the end it was Adress instead of Address
10:11 babilen *sigh*
10:11 oz_akan joined #salt
10:14 kingel joined #salt
10:24 istram joined #salt
10:26 kingel_ joined #salt
10:27 bhosmer joined #salt
10:36 Outlander joined #salt
10:42 felskrone1 joined #salt
10:47 giannello joined #salt
10:50 TyrfingMjolnir joined #salt
10:50 blarghmatey joined #salt
10:55 davidone left #salt
10:57 xcbt joined #salt
10:57 ramishra joined #salt
11:03 ZombieTwiglet joined #salt
11:04 akoumjian_ joined #salt
11:04 abele joined #salt
11:04 manytrees_ joined #salt
11:05 copelco joined #salt
11:17 intellix joined #salt
11:22 bhosmer_ joined #salt
11:22 mechanicalduck joined #salt
11:37 jaimed joined #salt
11:37 occupant joined #salt
11:40 TTimo joined #salt
11:45 j-saturne joined #salt
11:50 diegows joined #salt
11:51 Nexpro joined #salt
11:55 oz_akan joined #salt
11:56 fredvd joined #salt
11:57 viq joined #salt
11:57 viq joined #salt
11:57 bhosmer joined #salt
11:57 tmh1999 joined #salt
12:05 MrTango joined #salt
12:05 ianmcshane joined #salt
12:08 mechanicalduck joined #salt
12:09 blarghmatey joined #salt
12:10 adam__ Is there a default location for custom execution modules, somthing similar to /srv/salt/_states ?
12:16 martoss joined #salt
12:17 j-saturne joined #salt
12:23 KevinMGranger /srv/salt/_modules I believe
12:25 miqui joined #salt
12:25 babilen exactly
12:26 babilen Well, $file_roots/_modules/
12:26 KevinMGranger ^ that
12:27 dccc joined #salt
12:29 kingel joined #salt
12:35 ramishra joined #salt
12:35 TheRealBill joined #salt
12:36 TheRealBill joined #salt
12:39 CeBe joined #salt
12:40 vejdmn joined #salt
12:41 rome joined #salt
12:43 elfixit joined #salt
12:43 rome joined #salt
12:45 nitti joined #salt
12:45 dpfand0 joined #salt
12:46 rome joined #salt
12:46 mechanicalduck_ joined #salt
12:47 dpfand0 hi there! whats the best way to manage the ufw firewall with saltstack? there doesn't seem to be an builtin module. use cmd.run, just use iptables or use a third party module?
12:48 PI-Lloyd joined #salt
12:48 slav0nic joined #salt
12:50 PI-Lloyd Hey guys, having a bit of a weird problem. I have a salt master running in vagrant for development purposes, and I have a minion running on a LXC container on a different system. Everything was working fine up until a couple of hours ago whereby state.highstate stopped working. All other salt commands work fine from both minion  and master, just not a highstate. Keep getting timeout errors
12:50 PI-Lloyd minion and master are same version, minion and master both been rebuilt.. iptraf shows connections opening during highstate but nothing happens
12:51 dpfand0 left #salt
12:52 PI-Lloyd cmd.run from master works, manage.up works, fetching pillar data from master on minion works.. as said everything else works just a highstate times out constantly. Anyone have any suggestions on what to check, or have any ideas as to what is going on?
12:53 mechanicalduck joined #salt
12:54 csa_ babilen: why would you prefer $file_roots/_modules/ over $file_roots/modules/ (or .../_states over .../states)?
12:55 * csa_
12:55 babilen csa_: http://docs.saltstack.com/en/latest/ref/modules/ → "Modules placed in _modules/ will be synced to the minions …"
12:56 csa_ aaah! thx
12:57 babilen So you end up with _states, _modules, _grains, ... for your custom implementations and salt will sync them to your minions.
12:58 csa_ Cool!
12:58 jkaye joined #salt
12:59 AdamSewell joined #salt
12:59 mechanicalduck_ joined #salt
13:02 cpowell joined #salt
13:02 mordonez__ joined #salt
13:02 bhosmer joined #salt
13:03 cpowell joined #salt
13:05 ianmcshane joined #salt
13:05 mechanicalduck joined #salt
13:07 Ove_ http://www.beredskapsboden.se/se/articleimage.php?id=11029&amp;image=1944368&amp;fullwidth=0&amp;fullheight=0
13:07 jayfk is it possible to install a package with build dependency as in "apt-get build-dep foo"?
13:07 Ove_ Err
13:07 Ove_ Wrong url
13:07 Ove_ http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.groupadd.html
13:07 Ove_ Can this be used in a sls?
13:08 mordonez___ joined #salt
13:09 PI-Lloyd Ove_: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.group.html
13:10 mordonez____ joined #salt
13:12 babilen jayfk: That's not supported by aptpkg yet, but you can use cmd.run for that.
13:14 mordonez_____ joined #salt
13:14 babilen Ove_: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.group.html are the corresponding states that are implemented on top of that module
13:14 jayfk babilen: thanks
13:15 Ove_ babilen: Thank you!
13:16 thayne joined #salt
13:16 nitti joined #salt
13:17 hobakill is chocolatey state-able, or only a module?
13:17 babilen Ove_: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.module.html#salt.states.module.run might also be of interest (but you probably want to use group.present) -- Also: http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.jinja.html#calling-salt-functions
13:21 dude051 joined #salt
13:23 to_json joined #salt
13:24 lcavassa joined #salt
13:25 crane hi, i'm looking for a way to distrubute private keys on a secure way. anyone currently doing this? what is about using a cvs? do your repos contains sensitive datas like keys?
13:25 giannello crane, you can use pillars
13:25 giannello and store them in a separate git repo
13:26 babilen crane: Do it via pillars: http://docs.saltstack.com/en/latest/topics/pillar/ and reference the pillar value in file.managed with contents_pillar (cf. http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html#salt.states.file.managed )
13:27 crane giannello: babilen pillars can also still deal with templates?
13:27 babilen crane: I would recommend to *only* store that pillar on the master and edit it there. We use local git checkouts (with GitFS) to ensure that we can still keep track of who made which changes)
13:27 babilen crane: I don't quite follow. You mentioned templates for the fir time just now .. What are you trying to do?
13:27 babilen *for
13:29 crane babilen: sorry for that missunderstanding. i would like to deploy a tor service on several hosts. now each host is gaining his own config. that is the easy part. but also there should be private keys shipped for each server
13:29 babilen crane: And how do you do that?
13:29 crane and the last part is where i'm not sure how to deal with sensitive data like private keys
13:30 crane babilen: depending on the hostname in grains i use a config template
13:30 babilen Yes, I answered that question, but I am not sure how "templates" are related to this.
13:30 babilen So the private key is part of a configuration file?
13:31 crane babilen: nope :) a file stored beside but has a dependency on the servier tor
13:32 babilen And file.managed with content_pillar pointing to the pillar that holds the private key is not what you are looking for?
13:32 viq crane: this is one approach I saw before for that garthwaite.org/virtually-secure-with-openvpn-pillars-and-salt.html
13:32 TTimo joined #salt
13:33 oz_akan joined #salt
13:33 crane babilen: that was my question. is that possible to combine pillars (never used them before) with templates but i guess that is the answer
13:33 Lloyd_ joined #salt
13:33 crane viq: thx! i will read it
13:34 babilen crane: Why would you have to combine them? All you want is to populate a file with a private key. You use a file.managed state with content_pillar set to a suitable pillar and then you maintain the key in that pillar.
13:35 * babilen recommends to read http://docs.saltstack.com/en/latest/topics/pillar/ before continuing this as it will hopefully become clearer.
13:35 ramishra joined #salt
13:35 crane babilen: i never used pillars before and was not aware of this function. but i'm with you that i will now go on and read the docs to get a clue ;)
13:35 babilen crane: If you want actual help then paste something specific to, say, http://refheap.com so that we can comment on that. I still don't understand how "templates" are related to that (nor what you refer to exactly)
13:35 viq crane: pillar is a place to store a variable. You can use said variable in a template
13:36 babilen The point is that templates aren't needed if you simply want to populate the complete file with the value of a pillar.
13:37 cpowell joined #salt
13:37 crane babilen: viq: i will dig into it and start questening later if there is left anything open :)
13:38 babilen sure, have fun
13:38 occupant joined #salt
13:41 ianmcshane joined #salt
13:41 babilen crane: https://www.refheap.com/89746 is what I am thinking of
13:42 TyrfingMjolnir joined #salt
13:42 babilen (and you would, naturally, target pillars with different key values to various minions if you want them to use different keys)
13:46 slav0nic joined #salt
13:46 slav0nic joined #salt
13:49 mpanetta joined #salt
13:53 AdamSewell joined #salt
13:54 TheThing joined #salt
13:55 smcquay joined #salt
13:56 marnom anyone using salt-cloud with digital ocean? I can't seem to find my client_id as mentioned here: https://www.digitalocean.com/community/tutorials/automated-provisioning-of-digitalocean-cloud-servers-with-salt-cloud-on-ubuntu-12-04
13:57 marnom it was working before, but then digitalocean didn't have APIv2 yet... so maybe thats why I can't get it to work now?
13:57 manfred marnom: which version of salt are you on?
13:57 manfred it is working in develop right now
13:57 crane babilen: that looks pretty easy. nice. i will try this in a few hours :)
13:58 rome joined #salt
13:58 marnom manfred: I'm on the Helium RC
13:58 ajprog_laptop joined #salt
13:58 quickdry21 joined #salt
14:00 marnom manfred: maybe I need to define my provider differently? I tried only entering api_key and commenting client_key, but that doesn't work... any suggestions perhaps?
14:01 cron0 joined #salt
14:04 jkaye joined #salt
14:07 ramishra joined #salt
14:09 vejdmn joined #salt
14:11 manfred hrm? sorry been busy this morning
14:11 manfred let me clean my DO provider
14:11 manfred i have client_key and api_key
14:12 marnom manfred: I found my client_key now, but keep getting an AttributeError
14:12 marnom --list-images gives me a traceback with AttributeError: 'Response' object has no attribute 'getcode'
14:12 vejdmn joined #salt
14:12 manfred ¯\(°_o)/¯
14:13 manfred it is working here
14:13 marnom manfred: ok thanks for checking, then it must be a local issue
14:13 TheThing joined #salt
14:14 giantlock joined #salt
14:17 rallytime joined #salt
14:20 penguin_dan joined #salt
14:23 berserk joined #salt
14:23 hobakill while we're on cloud stuff - i get an AWS error telling me:                 Network interfaces and an instance-level security groups may not be specified on the same request ... which seems weird since i'm following the directions to a T
14:24 giannello joined #salt
14:25 peters-tx joined #salt
14:25 j-saturne joined #salt
14:26 metaphore joined #salt
14:26 jkaye joined #salt
14:33 CeBe joined #salt
14:36 PI-Lloyd joined #salt
14:39 vejdmn joined #salt
14:40 hobakill this is what i'm following: http://docs.saltstack.com/en/latest/topics/cloud/aws.html
14:40 hobakill :q
14:41 hobakill sorry
14:41 mordonez_____ joined #salt
14:44 hobakill https://github.com/saltstack/salt/issues/14963
14:46 icebourg joined #salt
14:47 cmthornton joined #salt
14:50 snuffeluffegus joined #salt
14:52 LukinLW joined #salt
14:52 CatPlusPlus joined #salt
14:52 hobakill are VPC launches supported? man i can't even launch a basic box using salt-cloud :(
14:53 martoss joined #salt
14:59 wendall911 joined #salt
15:00 intr1nsic joined #salt
15:03 j-saturne joined #salt
15:04 intr1nsic joined #salt
15:08 econnell joined #salt
15:09 intr1nsic joined #salt
15:10 brooksgarrett hobakill: They are, I use them extensively
15:10 brooksgarrett The secret is you have to specify the subnet id in your profile
15:10 intr1nsic joined #salt
15:10 brooksgarrett That will get it to launch in the correct VPC implicitly
15:11 brooksgarrett Don't specify the AZ
15:11 felskrone joined #salt
15:12 brooksgarrett Also, for security group I used the sg id instead of the 'name'
15:12 j-saturne joined #salt
15:13 intr1nsic joined #salt
15:14 SheetiS joined #salt
15:16 mikepea_ joined #salt
15:17 hobakill brooksgarrett: i definitely have the subnetId in the profile.
15:17 berserk_ joined #salt
15:17 hobakill brooksgarrett: and i have the SecgroupID in there as well.
15:17 brooksgarrett What version?
15:18 brooksgarrett Of salt-cloud?
15:18 hobakill brooksgarrett: 2014.1.10
15:18 brooksgarrett I'm on same here
15:18 hobakill brooksgarrett: do you have a sanitized version i could look at ?
15:18 brooksgarrett Sure
15:18 brooksgarrett Give me 5
15:18 hobakill brooksgarrett: sure. thanks.
15:21 berserk_ joined #salt
15:21 NotreDev joined #salt
15:23 ckao joined #salt
15:23 TTimo can I use a file:// style of URL in the source for file.managed? trying to manage a file source that isn't under the salt/ directory
15:23 TTimo (I'm asking because I tried and it doesn't seem to work)
15:24 TTimo 'unable to determine upstream hash'
15:25 conan_the_destro joined #salt
15:26 ramishra joined #salt
15:28 ajolo joined #salt
15:28 cron0 left #salt
15:31 joehillen joined #salt
15:31 intr1nsic joined #salt
15:33 kingel joined #salt
15:35 SpeeR joined #salt
15:36 blarghmatey joined #salt
15:36 giannello joined #salt
15:39 occupant joined #salt
15:40 NotreDev does someone have a best practice for using dockerio.built / installed / running? updating my dockerfile doesn’t ultimate have any effect on the latter two parts of the pipeline. and, when a docker container is installed or running, i can’t seem to replace it with a fresh container.
15:40 rome joined #salt
15:41 kingel joined #salt
15:42 TTimo NotreDev: afaik there's flags to force a rebuild/replace
15:42 jalbretsen joined #salt
15:42 pdayton joined #salt
15:42 NotreDev TTimo: there /is/ a flag? huh.
15:43 TTimo I remember having a similar problem, where it wouldn't rebuild and replace the running container
15:43 TTimo but it's pretty vague :)
15:43 NotreDev is it documented on salt or on docker-py?
15:43 TTimo yeah it's part of the dockerio state I mean
15:43 NotreDev i didn’t see anything about it on http://docs.saltstack.com/en/latest/ref/states/all/salt.states.dockerio.html
15:44 ericof joined #salt
15:44 troyready joined #salt
15:44 NotreDev the module definitely has some reload states
15:44 NotreDev or restart states
15:45 tligda joined #salt
15:46 kingel joined #salt
15:56 ramishra joined #salt
16:00 rome joined #salt
16:01 thehaven joined #salt
16:01 jeremyBass2 joined #salt
16:03 jut_ joined #salt
16:03 estherbester_ joined #salt
16:03 rcsheets_ joined #salt
16:03 neilf___ joined #salt
16:03 pass_by_value_ joined #salt
16:03 mordonez______ joined #salt
16:04 codekobe___ joined #salt
16:04 Hazelesque_ joined #salt
16:05 mikjaer_ joined #salt
16:05 madduck_ joined #salt
16:05 madduck_ joined #salt
16:05 gamingrobot_ joined #salt
16:05 __alex_ joined #salt
16:05 eclectic joined #salt
16:06 simonmcc____ joined #salt
16:06 cyberjames joined #salt
16:06 _Flusher joined #salt
16:06 babilen_ joined #salt
16:06 txmoose_ joined #salt
16:06 sirtaj_ joined #salt
16:06 rome joined #salt
16:06 bezaban_ joined #salt
16:07 hillna__ joined #salt
16:07 esogas` joined #salt
16:07 goal_ joined #salt
16:07 peno_ joined #salt
16:07 ikanobori_ joined #salt
16:08 EWDurbin_ joined #salt
16:08 arapaho_ joined #salt
16:08 mr_chris- joined #salt
16:08 eqe joined #salt
16:08 goki__ joined #salt
16:08 goodwill_ joined #salt
16:08 asmaps joined #salt
16:08 pwiebe__ joined #salt
16:08 jgelens joined #salt
16:08 basepi_ joined #salt
16:08 pwiebe__ joined #salt
16:08 dratone joined #salt
16:08 Kelsar joined #salt
16:08 gldnspud_ joined #salt
16:08 mechanicalduck joined #salt
16:09 seblu joined #salt
16:09 Emantor joined #salt
16:09 bnikolau1 joined #salt
16:09 mschiff joined #salt
16:09 mschiff joined #salt
16:09 chutzpah joined #salt
16:09 whiteinge joined #salt
16:09 fxdgear_ joined #salt
16:09 izibi joined #salt
16:09 xsteadfastx joined #salt
16:09 Nazzy joined #salt
16:09 nliadm joined #salt
16:09 rypeck joined #salt
16:09 Jarus joined #salt
16:09 ujjain joined #salt
16:09 darkelda joined #salt
16:09 darkelda joined #salt
16:09 CaptTofu__ joined #salt
16:09 redondos joined #salt
16:10 intr1nsic joined #salt
16:10 wiqd_ joined #salt
16:10 Ssquidly joined #salt
16:10 modafinil_ joined #salt
16:10 munhitsu_ joined #salt
16:10 imanc_ joined #salt
16:10 KyleG joined #salt
16:10 KyleG joined #salt
16:10 akoumjian_ joined #salt
16:10 cb joined #salt
16:10 che-arne joined #salt
16:11 Karunamon joined #salt
16:11 vukcrni joined #salt
16:11 skarn joined #salt
16:12 APLU joined #salt
16:12 linjan joined #salt
16:12 kaptk2 joined #salt
16:13 estherbester joined #salt
16:14 eliasp joined #salt
16:15 geekmush joined #salt
16:15 seblu joined #salt
16:16 codysoyland joined #salt
16:17 lcavassa joined #salt
16:21 racooper joined #salt
16:21 aparsons joined #salt
16:21 iggy Martin Hoefling?
16:21 jeffrubic joined #salt
16:23 hobakill many thanks to brooksgarrett for the AWS help. here's a hint, don't camel case in your config file. :(
16:24 iggy my hint for this week is read the docs that match the version of salt you are running
16:24 hobakill iggy: i'd agree. tho my version supposedly matches the dox version
16:26 metaphore joined #salt
16:27 aparsons_ joined #salt
16:28 iggy I thought so too until I fought per gitfs options for 2 days only to find out they are only in the unreleased version
16:28 iggy something else bit me too
16:28 ropes anyone have tips on debugging a "Data failed to compile" issue? AFIK the top.sls is correct and what it's claiming to be missing is present, but not much to go on as to where to start looking next..
16:28 ericof joined #salt
16:28 iggy ropes: try a yaml checker?
16:28 iggy usually that means I missed a leading -
16:28 oz_akan joined #salt
16:30 ropes iggy: is there a way to see the jinja generated yaml of an sls file? your guess may very well be correct.
16:32 NV joined #salt
16:33 oz_akan joined #salt
16:35 davet joined #salt
16:36 iggy state.show_sls might help
16:36 iggy it doesn't actually show you anything, but it seems to do some basic parsing and shows some debug output
16:36 schimmy joined #salt
16:36 iggy (i.e. here it complains about too many functions declared, etc.)
16:37 iggy ropes: ^
16:38 schimmy1 joined #salt
16:39 martoss joined #salt
16:41 ropes iggy: thanks for the tip!
16:42 timoguin ropes: state.show_low_sls is a good one too
16:42 timoguin that'll show the data once the final order of execution is determined
16:43 forrest joined #salt
16:43 mechanicalduck joined #salt
16:43 * iggy takes notes
16:44 kingel joined #salt
16:44 iggy if I'm using the salt-formula and have a salt/config/foo.sls in our main state repo, should that work, or should I rename salt in our main repo?
16:45 mik3 joined #salt
16:45 iggy salt is complaining about not being able to find some states, and that's the only thing I can think of
16:45 ropes state.show_top is necessary for the top file since if it's like ours and has too many functions in it.  just to contribute something back.
16:45 ropes timoguin: that's nice and handy!
16:46 Matt__ joined #salt
16:48 oz_akan joined #salt
16:49 ramishra joined #salt
16:49 ramishra joined #salt
16:50 brunolambert left #salt
16:50 kingel joined #salt
16:51 ramishra joined #salt
16:53 dcolish joined #salt
16:54 bhosmer_ joined #salt
16:55 aparsons joined #salt
17:00 ericof joined #salt
17:00 sectionme joined #salt
17:01 xcbt joined #salt
17:02 conan_the_destro joined #salt
17:03 metaphore joined #salt
17:03 MaZ- joined #salt
17:04 vejdmn1 joined #salt
17:07 vejdmn joined #salt
17:10 dude^2 joined #salt
17:10 spookah joined #salt
17:11 dude051 joined #salt
17:14 Ryan_Lane joined #salt
17:16 rap424 joined #salt
17:17 iggy any hints where to look to figure out why some state files in a formula aren't showing up?
17:18 forrest iggy, can you elaborate on the issue?
17:19 ipmb joined #salt
17:20 kermit joined #salt
17:21 iggy in gitfs_remotes, I have the salt-formula (along with some others) and then our main salt state repo, trying to assign salt.<anything> to hosts in our top.sls gives errors about no matching sls found when doing a highstate
17:23 forrest iggy, did you move the directory so it considers everything salt?
17:23 forrest you should just need to reference directories inside of the remote location
17:24 iggy no follow
17:24 forrest iggy, so let's say my remote is awesome-formula, inside of that is the awesome subdirectory
17:24 forrest inside of that, init.sls
17:24 forrest to reference it, I would use salt 'minion' state.sls awesome
17:25 forrest iggy, also, you made sure to restart the master after adding those remote refs right?
17:25 iggy fwiw, a lot of stuff is working (other formulas, our states, etc.), just not this
17:25 forrest iggy, okay, can you paste the relevant section for the specific gitfs item that isn't working from your master conf?
17:26 iggy well, I'm using salt-call state.highstate, but yeah, the entries in the top file should be right
17:26 iggy and yeah, restarted
17:26 chiui joined #salt
17:26 drawks when using an include in a state is the name space relative or absolute?
17:26 forrest drawks, absolute from the file root
17:27 drawks ah k
17:27 iggy https://github.com/saltstack-formulas/salt-formula.git is the formula we're using
17:27 forrest drawks, just reference it like you were applying the state.
17:27 forrest iggy, okay, so you do salt 'minion' state.sls salt.minion and it fails?
17:27 drawks thats a bit of a bummer, is there no way to use relative references?
17:27 forrest iggy, can you add -l debug?
17:27 iggy in the top file, I'm trying "- salt.minion" "- salt.master"
17:28 forrest iggy, let's skip using the top for now, try running the states individually
17:28 drawks i mean it would make it simpler for doing standard layouts of complex state trees
17:28 drawks oh well
17:28 xcbt joined #salt
17:28 shaggy_surfer joined #salt
17:28 thayne joined #salt
17:29 iggy salt '*salt*' state.sls salt.minion
17:29 iggy prints nothing
17:29 forrest iggy, but no traceback? Add -t 60 -l debug
17:29 forrest drawks, seems the opposite to me, but different strokes for different folks I guess!
17:31 drawks I'm just imagining maybe a standard layout where each class or role would have an init that would include say, packages; services; configs; other
17:31 drawks but maybe I'
17:31 drawks just need more time to see how this ends up shaking out
17:31 brooksgarrett Why not use base and then extend?
17:32 forrest drawks, so a few people are doing that, personally I'm not a fan, but it's up to the individual
17:32 forrest drawks, I prefer to go from the init, and branch from there
17:32 murrdoc joined #salt
17:32 drawks I'm still kicking the tires over here :)
17:33 aparsons joined #salt
17:33 drawks I'm probably the first person who's opened 3 issues on the github project before actually having gotten salt to do anything useful ;)
17:33 forrest drawks, for sure..
17:33 forrest drawks, haha, probably not :P
17:33 forrest drawks, here's the cool thing you can do if you do it from the init and branch out, let's say you have a service state, and a config state.
17:34 madphoenix joined #salt
17:34 forrest drawks, the config includes the init (which installs the packages), and the service, but the service only requires the init.
17:34 iggy forrest: no traceback, just the same error about no match sls found, also running the master with -l debug and not seeing anything useful in there
17:34 metaphore left #salt
17:34 forrest drawks, so you could have two unique configs with your custom configuration, versus the default, but the service still works for both
17:35 madphoenix hi all, i think i've hit a bug with salt 2014.1.10 finding a module, would somebody mind looking at my gist?  https://gist.github.com/brandentimm/5377de8d2a2969d2b28f
17:35 forrest iggy, Hmm, I'm not sure then..
17:35 madphoenix salt is clearly finding es_logstash_kibana/init.sls, but still reporting  no matching sls found
17:36 drawks forrest: I'm not sure I follow exactly, but in time I hope things become more clear with a bit of focus
17:36 druonysus joined #salt
17:36 xcbt joined #salt
17:37 forrest drawks, just think of it like this, you have two users, yourself, and Mike, the developer who loves to change things manually because he sucks. You know how you want nginx configured, so you have a config.sls that configures it that way, Mike on the other hand needs a VM spun up, and HOW DARE YOU try to configure his nginx for him, he wants to do it! So you can run your config state, which includes the service, which includes the init, but Mike can just incl
17:37 forrest ude the service state, which includes the init. Now you're both happy
17:38 forrest so you gain modularity that way
17:39 forrest madphoenix, can you try to run salt 'minion' state.sls es_logstash_kibana ?
17:39 madphoenix yep, i get the same thing
17:39 forrest madphoenix, is that local on the system?
17:40 madphoenix i've tried through our master which uses gitfs, and with local clones of the git repository at /srv/salt
17:40 madphoenix both give the same result
17:40 forrest iggy, what release of salt are you running
17:40 forrest iggy, and just to be safe, it is the same release between master and minion right?
17:40 occupant joined #salt
17:40 madphoenix i've also run it with --log-leve=debug, and it's clearly reading and rendering the YAML for that .sls
17:41 forrest madphoenix, hmm, can you gist that state?
17:43 iggy forrest: 2014.1.10 same everywhere
17:43 melinath joined #salt
17:45 CeBe joined #salt
17:45 forrest iggy, hmm, ok let's thing this through then, what makes that repo different from your other gitfs repos?
17:48 iggy a typo apparently
17:49 iggy had salt-formulas instead of saltstack-formulas
17:49 xcbt joined #salt
17:49 forrest iggy, oh ok
17:49 * iggy slaps himself around with a giant speckeled trout
17:49 forrest iggy, as a heads up you should fork the repo
17:49 forrest iggy, don't directly link against the main one, could be a security concern, or if changes occur something might break
17:51 iggy I know
17:52 forrest ok cool
17:52 iggy I've got about half of them, but I don't have the ability to create repos in our company org, so I link to the saltstack-formulas spot and then have someone else clone later
17:53 forrest iggy, awesome
17:53 madphoenix forrest: https://gist.github.com/brandentimm/a3a179079c749643c986
17:54 forrest madphoenix, hmm, you included java, but I don't see that state from your initial gist
17:54 forrest just oracle-java
17:55 madphoenix ah, good point - that is a mistake
17:55 madphoenix if that's the cause, it's a poor error message, no?
17:55 godber joined #salt
17:55 skyler When I require the mysql formula I get a "The following requisites were not found:" message. This is not the case with the apache formula. Everything for mysql installs correctly.
17:55 forrest madphoenix, yes it is
17:56 forrest madphoenix, I can't remember if any of that was fixed in the latest release.
17:56 forrest madphoenix, I think it has something to do with how the yaml is treated? I can't remember exactly, either way it is not great.
17:57 felskrone joined #salt
17:57 forrest skyler, so you can call mysql, but requiring it doesn't work?
17:57 brooksgarrett Hey all, anyone using windows?
17:58 forrest brooksgarrett, there are some guys that do use it in here, I don't know if they are around though
17:58 wangofett brooksgarrett: at least a few of us
17:58 brooksgarrett I seem to remember when installing the minion on windows (building the golden image) there was some step so it would gen a new key and id once the image is provisioned
17:58 brooksgarrett Can't recall that magic step though
17:59 wangofett brooksgarrett: afaik it just does that by itself
17:59 brooksgarrett So it will detect that sysprep ran and create a new salt-key?
17:59 brooksgarrett Sweet
18:00 wangofett at least it did when I installed 2014.1.10 here a day or two ago
18:00 brooksgarrett I'll give it a shot before I get any further in building this image.
18:00 brooksgarrett Thanks!
18:00 wangofett are you using theinstaller?
18:00 skyler forrest: Yes, I can include mysql, and it works just fine. However, another pkg that has a require: sls: mysql says it was not found.
18:00 forrest skyler, you need to both do an include, AND require
18:00 forrest skyler, you can't just put a require
18:01 aparsons joined #salt
18:01 skyler forrest: I am doing both. So there are two entries, one that is the include, and the other is a pkg which I have require the mysql sls.
18:01 forrest skyler, ahh ok, can you gist the state?
18:03 xcbt joined #salt
18:03 madphoenix forrest, thanks for the help, that did it
18:03 forrest madphoenix, awesome, basepi are you around?
18:03 spookah joined #salt
18:03 brooksgarrett wangofett: yes.
18:03 spookah joined #salt
18:04 basepi forrest: I suppose that depends on your question.  ;)
18:04 brooksgarrett I installed, didn't start, and sysprepped
18:04 skyler forrest: https://gist.github.com/skylerberg/5206d7de0993cd618998
18:04 forrest basepi, hah, do you ever remember if anyone fixed the errors when a state you have that includes a state which is missing actually reports back the missing state is the problem, and not the state doing the include?
18:05 forrest basepi, I can't remember.
18:05 forrest basepi, so you do an include, that include is missing, it says the main state couldn't be found, not the actual missing state in the include
18:05 basepi forrest: Hrm, I want to say someone was working on that, but I'm not sure if it got fixed or not
18:05 occupant joined #salt
18:05 basepi Yes, I know the issue.  I am unsure whether it's been fixed or not
18:06 forrest basepi, Yeah I felt like someone was working to resolve it, or had, but couldn't remember either
18:06 forrest basepi, ok thanks
18:06 basepi Sorry I'm not more help.   =)
18:06 jalaziz joined #salt
18:07 dude051 joined #salt
18:07 forrest basepi, no worries, the fact you also think someone was working on it makes me believe I'm not QUITE as crazy
18:07 basepi Heh
18:07 linjan joined #salt
18:07 ajprog_laptop joined #salt
18:07 mik3 don't be confused, you're nuts kid
18:07 forrest mik3, heh
18:07 mik3 :)
18:08 forrest skyler, what does the directory structure for the mysql stuff look like? /srv/salt/mysql/...
18:08 UtahDave joined #salt
18:08 forrest oh Dave is hear, everyone ask him your questions so he can't get any work done
18:08 skyler forrest: I am using gitfs, git://github.com/saltstack-formulas/mysql-formula.git
18:09 Gareth morning
18:09 forrest skyler, interesting...
18:09 forrest Gareth, hey Gareth
18:09 Gareth hey forrest
18:09 Gareth UtahDave: what is the meaning of life?
18:09 UtahDave 42
18:09 Gareth UtahDave: how much wood could a woodchuck chuck if a woodchuck could chuck wood?
18:09 UtahDave Hey, everyone!  I almost said good morning, but it's 12:09 pm here.
18:09 skyler forrest: Also, I am using git://github.com/saltstack-formulas/apache-formula.git, and require: sls: apache works fine. It is mysql in particular that does not work.
18:10 UtahDave 42
18:10 forrest skyler, I'm pretty confused then, you can apply the mysql state fine, but when it's a require, it does't seem to work? Just to double check, the master and minion are both the same release correct?
18:10 forrest skyler, yeah that's why I am confused :\
18:10 forrest UtahDave, yeah only Gareth can get away with mid day mornings it seems
18:10 UtahDave :) yeah
18:11 Gareth forrest: morning is a relative term :)
18:11 skyler forrest: both master and minion where made using he bootstrap script, I will check to make sure they are the same though. The include for mysql seems to work, but then the require fails.
18:11 matt__ joined #salt
18:12 forrest skyler, very strange.
18:12 forrest skyler, your gist looks fine to me :\
18:12 shaggy_surfer joined #salt
18:13 skyler forrest: yeah, they are both 2014.1.10 (Hydrogen).
18:13 forrest skyler, I'm not sure then, maybe I'm missing something.
18:14 forrest UtahDave, do you see something I'm missing here? https://gist.github.com/skylerberg/5206d7de0993cd618998
18:15 skyler plot twist: if I require: "- sls: mysql.server" instead of "- sls: mysql" it works.
18:16 forrest but there's an init :\
18:18 aparsons joined #salt
18:19 UtahDave forrest: what's the name of the directory?
18:19 UtahDave i have to run to a meeting. back in a sec
18:21 CeBe joined #salt
18:22 jkaye joined #salt
18:22 diegows joined #salt
18:23 skyler okay, I have figured out exactly how to break it: require an include which ONLY has other includes in it.
18:23 forrest skyler, ?
18:23 forrest UtahDave, I believe skyler is just using the mysql-formula stuff, so just the mysql dir
18:24 forrest skyler, that would be funny
18:24 rome joined #salt
18:24 aparsons joined #salt
18:25 skyler Seriously, if a.sls includes b.sls and requires b.sls, and b.sls includes c.sls, then a.sls will fail.
18:25 skyler if b.sls includes c.sls and has another entry, a will succeed.
18:29 sectionme joined #salt
18:30 ajolo joined #salt
18:33 forrest weird
18:34 UtahDave skyler: what version of Salt?
18:34 xcbt joined #salt
18:34 kermit joined #salt
18:35 TTimo joined #salt
18:35 skyler UtahDave: 2014.1.10 (Hydrogen)
18:36 UtahDave ok, could you open an issue on that? Please include as much info as you can so we can reproduce it.
18:36 xcbt_ joined #salt
18:37 skyler UtahDave: will do
18:37 UtahDave thanks!
18:39 ianmcshane joined #salt
18:40 jkaye joined #salt
18:41 rome joined #salt
18:44 Gareth bah.  so close to having all automated tests pass.
18:45 xcbt joined #salt
18:47 n8n joined #salt
18:49 vbabiy joined #salt
18:53 skyler UtahDave: https://github.com/saltstack/salt/issues/15481
18:53 UtahDave perfect. Thanks, skyler!
18:54 aparsons joined #salt
18:54 ianmcshane joined #salt
18:55 murrdoc joined #salt
18:56 skyler UtahDave: There is a problem with that issue, just a second, I will fix it.
18:56 tedski too kludgy? http://pastie.org/9523037  i'm trying to manage hosts file entries on dhcp hosts so grains['fqdn_ip4'] returns an accurate value
18:56 UtahDave cool
18:57 skyler UtahDave: fixed. I accidentally pasted the wrong file in for c.sls.
18:58 QuinnyPig cp.get_file is awesome, in case anyone was keeping score at home.
19:00 hobakill thanks basepi for updating my issue #14430. will give it a try later and update the ticket.
19:00 smcquay joined #salt
19:01 nw0428 joined #salt
19:01 basepi hobakill: sounds great.
19:01 nw0428 I am having issues with yumpk.installed and versioning
19:01 matt__ joined #salt
19:02 nw0428 Basically regardless of how specific the version I use is I cant get it to work
19:02 forrest nw0428, are you installing via the command line? Or in a state?
19:02 nw0428 in a state
19:02 nw0428 forrest in a state
19:03 forrest nw0428, http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkg.html
19:03 nw0428 hive:   pkg.installed: - version: 0.13.0
19:03 forrest nw0428, and that isn't working?
19:03 nw0428 I read the docs
19:03 nw0428 Nope
19:03 forrest are you getting an error? Or is it just not installing the right version
19:03 nw0428 the output is  Comment: The following packages failed to install/update: hive=0.13.0
19:04 kingel joined #salt
19:04 murrdoc joined #salt
19:04 nw0428 it makes me think it is trying to do yum install hive=0.13.0
19:04 forrest nw0428, what does the actual version look like when you do a yum install on the box? Maybe the format is slightly different?
19:05 p2 joined #salt
19:05 nw0428 the actual version is Version     : 0.13.0.2.1.0.0
19:05 nw0428 and I tried using that as well
19:05 nw0428 and even 0.13.0.2.1.0.0.el6.noarch
19:06 bhosmer_ joined #salt
19:06 forrest nw0428, hmm, maybe install it, then see what an rpm -qi says about the version value, and try that?
19:07 nw0428 it says Version     : 0.13.0.2.1.0.0
19:09 forrest nw0428, in that case I'd track down via the logs (or salt-call commands on the minion) what it's actually trying to install and create an issue if it's a salt problem. Which release of salt are you on by the way?
19:09 forrest also, that versioning is hilarious
19:09 saltydog joined #salt
19:09 nw0428 1.7
19:09 nw0428 the versioning is hilarious
19:10 cmthornton This doesn't solve your actual problem nw048, but if you want an immediate solution, I'd just `cmd.run: yum install hive <version>` with a `require_in: pkg: hive` then for the `pkg: hive` state I'd just install GNU hello or some dummy package
19:10 forrest nw0428, the master and minion are the same release right?
19:10 forrest cmthornton, that's a good suggestion.
19:10 martoss joined #salt
19:11 nw0428 @forrest they are
19:11 nw0428 And thanks for the suggestion @cmthornton
19:12 cmthornton you're welcome
19:12 saltydog Afternoon salt masters/experts and the like. Question for you all. Does the custom module directory support subdirectories for relative import? for example: /srv/salt/_modules/filesystem/__init__.py and /srv/salt/_modules/filesystem/files.py
19:12 saltydog with the apropriate __virtualname__ and __virtual__ declarations of course
19:13 forrest nw0428, yeah if you can give the salt-call thing a shot on the minion to see what exactly is happening that would be great.
19:14 nw0428 yeah
19:14 nw0428 I think I figured it out
19:14 nw0428 it is running: yum -y   downgrade "hive-0.13.0.2.1.0.0"
19:14 saltydog I've been trying to get this to work for the past hour or so and i've come up with nothing. Granted im not a salt expert, so I thought i'd ask you all.
19:14 nw0428 and yum is saying:
19:15 nw0428 Only Upgrade available on package: hive-0.13.0.2.1.0.0-92.el6.noarch
19:15 quup left #salt
19:17 nw0428 @forrest any idea what to do?
19:19 forrest nw0428, hmm, was hive already installed?
19:19 nw0428 Yeah
19:20 forrest if it's the same version, it should simply be saying that it's already installed..
19:20 nw0428 but the state should be idempotent
19:20 nw0428 and not fail
19:20 forrest right
19:20 forrest I wonder if it has to do with being such a long version number?
19:20 nw0428 I have tried almost every length imaginable
19:20 forrest nw0428, I'd say go with cmthornton's suggestion, and create an issue detailing that, and the output that you're seeing where it's trying to downgrade, but that's the same version as the current release.
19:21 nw0428 Do you know where the yumpkg source is? Also, where should I put the issue?
19:21 forrest nw0428, https://github.com/saltstack/salt/issues/new
19:22 forrest nw0428, this is the module: https://github.com/saltstack/salt/blob/develop/salt/modules/yumpkg.py
19:22 forrest nw0428, and here's the pkg state: https://github.com/saltstack/salt/blob/develop/salt/states/pkg.py
19:22 jkaye joined #salt
19:23 nw0428 @forrest thanks
19:24 forrest nw0428, np, sorry I don't have a solution :\
19:25 nw0428 You were super helpful anyway =)
19:25 forrest haha
19:25 iggy saltydog: I would think so, since at some level it's either using python funtionality or trying to emulate it as closely as possible... but I'm no expert myself
19:26 saltydog Yeah thanks Iggy. I was operating under the exact same impression as well.
19:26 metaphore joined #salt
19:27 forrest saltydog, sorry missed your question, I'm actually not sure, I've only seen people use straight up .py files.
19:27 forrest saltydog, https://github.com/saltstack/salt-contrib/tree/master/modules
19:28 TheThing joined #salt
19:28 saltydog forrest, thanks for the reply. Yeah we are using the .py files as they stand today, and that works. Problem is that its getting unwieldily large in _modules and the ability to break things down by component/service/thing would benefit us greatly.
19:30 jkaye joined #salt
19:30 forrest saltydog, understandable. Yeah I'm not sure. If you test it out and it works, could you either make a PR against these docs: http://docs.saltstack.com/en/latest/ref/modules/
19:30 bhosmer_ joined #salt
19:30 forrest saltydog, or create an issue detailing your example, and someone else will update them
19:31 saltydog Yeah for sure. I'll keep plugging away and let the community know what i've come up with.
19:32 TheThing_ joined #salt
19:32 forrest saltydog, awesome, it might also be a good idea that if it doesn't work, to create an issue to see if support can be added.
19:32 forrest or just add it :P
19:33 saltydog Right. Im actually going through the salt.loader code now to see how its broken out. Trying to see if I can simply add and submit upstream
19:33 forrest cool
19:35 jergerber joined #salt
19:36 brooksgarrett All, in a state module, is there a way to check for a condition and abort (or wait) if that condition is present?
19:36 brooksgarrett Example, Server A is a backend processing server. I don't want to upgrade mid task. So if process X is running, send signal to kill graceful and wait for everything to exit before continuing with the highstate.
19:37 XenophonF joined #salt
19:37 forrest brooksgarrett, maybe unless and onlyif? http://docs.saltstack.com/en/latest/ref/states/requisites.html#altering-states But they aren't in the current release :\
19:38 XenophonF can I create a pillar sls with a filename something like pillar/websites/www.example.com.sls?
19:38 brooksgarrett Thanks forrest, at least I can plan forward
19:39 forrest brooksgarrett, np, I believe that unless and onlyif are in the current release for cmd.run: for the time being you could try to use
19:39 forrest though you'd have to double check, just in case someone didn't notate it :P
19:39 brooksgarrett I'll take a look
19:39 forrest cool
19:39 forrest it won't be as awesome as the stuff supported in the next release though, since you can use that wherever which is awesome
19:40 XenophonF alternatively I guess I can do something like pillar/www/example/com.sls and apply it as www.example.com in pillar/top.sls
19:41 TheThing_ joined #salt
19:41 forrest XenophonF, I've never tried to use a url like that for a pillar file, I'd say just try it!
19:42 thayne joined #salt
19:50 XenophonF left #salt
19:53 totte joined #salt
19:55 totte left #salt
19:56 rangertaha Where do I put custom runners? and how do I config it?
19:57 drawks is it possible to have requisites on inclusions?
19:57 drawks hrm, actually that wouldn't make sense
19:57 drawks never mind
20:02 ianmcshane joined #salt
20:03 jdmf joined #salt
20:04 rangertaha I made custom runner folder in /srv/salt/_runners and put this in the config file but its not working any advice?
20:05 vbabiy joined #salt
20:07 Ryan_Lane basepi: howdy. I had marked some more boto changes as needing backport. seems they're also blocking me from pushing in the rest of the VPC fixes needed
20:07 Ryan_Lane then all the modules will have full VPC support, which would be excellent for release
20:08 basepi Ryan_Lane: alright, I'll do some backporting this afternoon
20:08 ianmcshane joined #salt
20:09 schimmy joined #salt
20:09 Ryan_Lane basepi: thanks :)
20:09 Ryan_Lane sorry about missing those last time
20:09 basepi np
20:09 brooksgarrett Anyone have an idea about this error in Salt-cloud? [WARNING ] Please switch to the new providers configuration syntax
20:09 chrisjones joined #salt
20:10 brooksgarrett 0.8.11
20:10 cmthornton /etc/salt/cloud, cloud.providers, cloud.profile syntax changed recently because the "aws" provider is now "ec2"
20:11 brooksgarrett Right, my provider: ec2 line is there
20:11 brooksgarrett Not aws
20:11 schimmy1 joined #salt
20:12 cmthornton what's in /etc/salt/cloud? just `minion:` and `start_action` stuff? or `provider` stuff?
20:12 ianmcshane joined #salt
20:13 jalaziz joined #salt
20:13 ianmcshane joined #salt
20:15 cmthornton I had just setup salt-cloud a couple days ago, I believe I just had to move something from /etc/salt/cloud to /etc/salt/cloud.provider or profile to stop seeing that warning
20:16 rap424 joined #salt
20:16 drawks fg
20:16 ianmcshane joined #salt
20:16 drawks whoops wrong window
20:17 sectionme joined #salt
20:18 brooksgarrett I move all providers stuff to cloud.providers
20:18 brooksgarrett cloud is only minion stuff
20:18 manfred cmthornton: so, /etc/salt/cloud is just your extra stuff, providers used to go in there... now they go in /etc/salt/cloud.providers, or any .conf in /etc/salt/cloud.providers.d/
20:19 manfred brooksgarrett: false, make_master: exists to make the server a master
20:19 manfred so you can spin up your syndicate
20:19 midacts joined #salt
20:20 manfred http://ix.io/eaV this is the stuff that should go in /etc/salt/cloud
20:20 manfred the global stuff that doesn't pertain to one provider or profile
20:20 brooksgarrett Found it
20:20 brooksgarrett You can't have a '.' in your provider name
20:20 manfred http://docs.saltstack.com/en/latest/topics/cloud/misc.html
20:21 brooksgarrett I had stg.us.west: as the provider name
20:21 brooksgarrett Cause the parsing to explode
20:21 brooksgarrett changing to stg-us-west: works fine
20:23 rome joined #salt
20:23 aparsons joined #salt
20:26 martoss joined #salt
20:27 davidone joined #salt
20:27 justyns joined #salt
20:29 ianmcshane joined #salt
20:32 xcbt joined #salt
20:38 jdmf joined #salt
20:39 martoss joined #salt
20:41 rome joined #salt
20:46 viq joined #salt
20:48 jdmf joined #salt
20:50 jdmf joined #salt
20:50 thayne joined #salt
20:51 cmthornton joined #salt
20:55 kingel joined #salt
20:56 schimmy joined #salt
20:57 blarghmatey joined #salt
20:58 xcbt joined #salt
20:58 nkuttler i really didn't consider that i don't have the patience to autoconfigure my workstation with salt..
20:59 nkuttler i've written nice templates and everything, but am doing things by hand... lesson, only write config templates you'll use for many hosts..
20:59 nkuttler (and pull config from backups)
20:59 schimmy1 joined #salt
21:08 iggy trying to figure out why salt-call state.highstate keeps not running my states "Not Run:   2"... only thing I can find from googling is stuff about test=True which I'm definitely not using
21:10 Eugene Pastebin of the output would be helpful
21:10 SheetiS iggy: It is possible that you have a default configured to test=True.  Have you tried the call with test=False to see if it runs then?
21:11 CeBe joined #salt
21:11 iggy how would that be set?
21:11 SheetiS http://docs.saltstack.com/en/latest/ref/states/testing.html#default-test for default test in minion config.
21:13 iggy thanks
21:13 n8n joined #salt
21:13 iggy silly mistake
21:13 iggy just started here and there's already a huge looming deadline and I've got 20 things going on at once
21:14 iggy seem to keep missing the small things
21:14 SheetiS iggy: Understood.  We all have those crazy deadlines from time to time ;-)
21:17 basepi GAH!  Ryan_Lane can I just copy these files from develop to 2014.7?  So many merge conflicts, since we backported some things out of order, and all of these backports rely on each other.
21:17 micah_chatt joined #salt
21:17 Ryan_Lane basepi: heh
21:17 basepi I think develop should have all of the bugfixes, right?  Since we merge forward
21:18 Ryan_Lane yes
21:18 Ryan_Lane that's fine
21:18 basepi Alright, I'm just going to do that.
21:18 Ryan_Lane sounds good
21:19 Ryan_Lane those files will lose attribution in the branch
21:19 Ryan_Lane what happens then when you forward port my new changes?
21:19 Ryan_Lane we'll lose it in develop too, right?
21:20 sectionme joined #salt
21:21 aparsons joined #salt
21:21 jalaziz joined #salt
21:22 micah_chatt joined #salt
21:24 xcbt joined #salt
21:24 robawt anyone systematically installing CA certs?
21:26 basepi Ryan_Lane: Have you submitted new changes that aren't merged yet?
21:27 basepi I had Tom merge forward before I started this process.  So theoretically any fixes you made on 2014.7 are now in develop, and will be in these files I backported.
21:27 basepi Ryan_Lane: https://github.com/saltstack/salt/pull/15490
21:27 basepi And yes, the changes will unfortunately lose attribution in 2014.7.  But that's not the end of the world as far as I'm concerned.
21:29 Ryan_Lane basepi: well, hulu had also pushed in some changes
21:30 Ryan_Lane it's not easy tracking down all the changes
21:30 Ryan_Lane there was some other org that submitted code to boto_elb as well
21:30 basepi Well, this should get us up to date as of today.
21:30 kingel joined #salt
21:30 basepi Then we'll have to be careful to backport additional fixes to the modules
21:30 aparsons joined #salt
21:31 snuffeluffegus joined #salt
21:34 rome joined #salt
21:35 smcquay joined #salt
21:46 diegows joined #salt
21:47 jalaziz joined #salt
21:47 WarP|onwork joined #salt
21:52 melinath joined #salt
21:54 xcbt joined #salt
21:56 micah_chatt joined #salt
21:56 shaggy_surfer joined #salt
21:57 kballou joined #salt
21:58 nw0428 joined #salt
22:01 notbmatt joined #salt
22:01 berto- joined #salt
22:01 kingel joined #salt
22:02 deepz88 joined #salt
22:06 druonysuse joined #salt
22:08 rome joined #salt
22:08 kermit joined #salt
22:12 rjc joined #salt
22:14 geekmush1 joined #salt
22:14 orev joined #salt
22:15 mawburn joined #salt
22:16 mawburn left #salt
22:21 sectionme joined #salt
22:23 gzcwnk ee
22:23 iggy woo! my first salt (formula) pull request
22:23 basepi Ryan_Lane: merged:  https://github.com/saltstack/salt/pull/15490#issuecomment-54376568
22:23 gzcwnk anyone know if the minion's going to sleep fixed yet?
22:25 Ryan_Lane basepi: cool. thanks
22:27 mosen joined #salt
22:36 dude051 joined #salt
22:37 dude051 joined #salt
22:37 ericof joined #salt
22:44 justyns joined #salt
22:44 KevinMGranger What if you have something you want installed that is available as a package on some distros, but not on others? You could jinja the state based on OS grain, but what happens if you need to depend on it in another state?
22:45 nitti_ joined #salt
22:46 manfred KevinMGranger: use names and the depend on the state_id
22:46 forrest KevinMGranger, an OS check for that second item maybe?
22:46 aquinas_ joined #salt
22:47 metaphore joined #salt
22:47 sectionme joined #salt
22:47 manfred KevinMGranger: http://paste.gtmanfred.com/mNW/
22:50 xcbt joined #salt
22:50 KevinMGranger manfred: I mean something closer to this: https://gist.github.com/KevinMGranger/0ad0c0fc7434091fa7e8
22:51 KevinMGranger not packaged at all in other distros
22:51 KevinMGranger . How do I know what to depend on?
22:52 forrest KevinMGranger, unless manfred has a more elegant selection, I'd just do another grain check
22:52 manfred rename your state
22:52 manfred so that you are always just requiring 'composer'
22:52 forrest manfred, oh you want them to be the same name?
22:52 forrest lol
22:52 forrest that's ghetto
22:52 forrest :P
22:52 forrest but it is a good idea
22:52 rome joined #salt
22:52 manfred but then he can just use the same require: statement
22:53 Outlander joined #salt
22:54 KevinMGranger don't I need to specify a state module type each time? That's my issue-- how do I know what state module type to use?
22:54 manfred KevinMGranger: this would totally work http://paste.gtmanfred.com/4FgRl/
22:54 KevinMGranger or would I just have to put it in a separate SLS (probably good practice anyway)
22:54 Ryan_Lane when I want to throw an error from an execution module, how should I go about it?
22:54 manfred just use require:\n - cmd: composer
22:54 KevinMGranger err hang on manfred , my example wasn't good enough
22:55 Ryan_Lane I want the CLI to return an exit code, but I want the error condition to propagate through to the state system
22:55 manfred everywhere you need it
22:55 Ryan_Lane is there an exception for this?
22:55 KevinMGranger updated https://gist.github.com/KevinMGranger/0ad0c0fc7434091fa7e8
22:56 KevinMGranger I could put the same OS family checks each time I need to require composer, but that seems really messy
22:56 manfred yo don't need to
22:56 manfred dude, use mine
22:56 manfred don't name the watch state composer
22:56 manfred name the get-composer state composer
22:57 manfred that way it will only run the first time, and then the watching state will run once, and that is it
22:57 manfred the next time, it will just pass because you have the unless
22:57 TheThing joined #salt
22:57 UtahDave Ryan_Lane: Yeah, you'll want to just thrown an error. Salt will propagate the error correctly.
22:57 Ryan_Lane UtahDave: is there a specific exception to use?
22:57 UtahDave There are several Salt errors you can throw.   SaltCommandError, etc
22:58 KevinMGranger manfred: check out the second file in my gist. What I'm saying is, how do I know how to require that composer is installed for a separate state? I'd need to pick between cmd: composer or pkg:composer
22:58 manfred KevinMGranger: if you use mine, and you just only ever have to require: cmd: composer, and no os_family check
22:58 Ryan_Lane UtahDave: these aren't documented in the developer docs! :(
22:58 manfred oh wait...
22:58 manfred meh
22:58 * Ryan_Lane opens a bug
22:58 manfred KevinMGranger: throw then in a seperate state, and require the sls
22:58 UtahDave Ryan_Lane: https://github.com/saltstack/salt/blob/develop/salt/exceptions.py
22:58 manfred right, forgot about package
22:58 Guest10538 joined #salt
22:58 manfred require: sls: composer
22:58 manfred done
22:58 KevinMGranger that does seem like the cleanest way to do it. Thanks!
23:00 manfred KevinMGranger: yeah, been a long day, I completely zoned out on the pkg vrs cmd
23:00 Ryan_Lane UtahDave: if I throw an CommandExecutionError, what return code will be used?
23:01 KevinMGranger it's cool. something something hump day
23:01 KevinMGranger It might be nice to have some sort of writeup on best practices for composite states like that. Maybe I'll write it and PR it
23:01 UtahDave Ryan_Lane: Hm. Actually, now that I think about it a moment more, maybe I'm wrong on that.
23:01 KevinMGranger thanks again for the help
23:01 manfred KevinMGranger: https://github.com/saltstack/salt/issues/15026
23:02 manfred that needs to get done is what needs to happen :P
23:02 kingel joined #salt
23:03 CeBe1 joined #salt
23:03 KevinMGranger oh I have to include them to use require: sls: ?
23:03 KevinMGranger dang
23:03 KevinMGranger eh won't be too bad if everything's separated properly
23:04 manfred KevinMGranger: you do not have to include them
23:04 manfred KevinMGranger: they will have to be in teh full state run though
23:04 manfred so if they aren't int eh same file, to do state.sls , you will have to include that other state so that it actually has the state to require
23:06 druonysuse joined #salt
23:08 smcquay joined #salt
23:11 jdmf joined #salt
23:12 blarghmatey joined #salt
23:13 yomilk joined #salt
23:15 KevinMGranger manfred: got it
23:15 Ryan_Lane basepi: https://github.com/saltstack/salt/pull/15504 <-- look at that, I'm following the docs for 2014.7 changes for once ;)
23:16 basepi Ryan_Lane: looks great!  As long as the diff looks right, I don't see any issues.
23:17 Ryan_Lane heh, yeah, the diff is fine.
23:17 Ryan_Lane not sure how easy it'll be to backport
23:17 mosen hi RL I think you mentioned in your blog you had something committed
23:18 nitti joined #salt
23:18 basepi Ryan_Lane: won't need to be backported.  You submitted it directly into 2014.7, so we'll just merge it forward into develop
23:18 mosen how possible is it to get execution modules pulled into salt, or not at all
23:18 basepi The joys of the new workflow.
23:18 Ryan_Lane err. I mean I don't know how easy it'll be to forward port
23:18 Ryan_Lane hopefully easy
23:18 Ryan_Lane it'll probably cause a merge conflict
23:18 basepi Shouldn't be a problem.  Currently 2014.7 and develop are exactly the same in the boto stuff, so it will be fine.
23:19 * Ryan_Lane nods
23:19 basepi If they diverge in the future, then fixes may be harder to merge forward.  But for now, we're good.
23:19 Ryan_Lane cool
23:21 iggy what's the schedule looking like for 2014.7?
23:21 xcbt joined #salt
23:21 ocdmw joined #salt
23:23 aquinas joined #salt
23:23 aquinas_ joined #salt
23:24 ocdmw Hi! Ever since upgrading my server from Ubuntu 13.10 to 14.04, I cannot seem to run salt commands from the master to the minions as any user other than root, although I have restarted the master and removed /var/cache/salt/master/*.key.  I use client_acl in the master config without any changes compared to how it worked correctly in the past.  The error I see is: "Failed to
23:24 ocdmw connect to the Master, is the Salt Master running?".  Doing anything in salt as root works fine.  Any thoughts or tips?
23:25 forrest ocdmw, are both master and minion running the same release of salt?
23:25 manfred iggy: still waiting
23:25 manfred iggy: just in rc1 right now still, lots of work going into it
23:25 ocdmw Yes, I did have to upgrade the master to follow a recent upgrade on the minions and the upgrade of the master didn't resolve this.
23:25 forrest manfred, start growing a beard, don't shave till release :P
23:26 forrest ocdmw, hmm, ok and both services were restarted after the upgrade?
23:26 manfred would be good if I could grow a beard :(
23:26 ocdmw forrest, yep
23:26 forrest heh
23:26 forrest ocdmw, ok, I'd say start the master and minion in debug mode, and see if anything pops up in the log
23:26 manfred ocdmw: check on the minion, is it getting this error https://github.com/saltstack/salt/issues/12248
23:27 ocdmw forrest and manfred, thanks!  Will follow these leads.
23:28 Gareth hm. grains are persistent right?
23:28 KevinMGranger Gareth: what do you mean?
23:29 Gareth set a grain value, restart a minion.  is the grain value still there?
23:29 rallytime joined #salt
23:29 KevinMGranger how are you setting it?
23:29 forrest Gareth, if you set it on the command line?
23:30 Gareth forrest: Nod.  CLI.
23:30 Gareth or....using __salt__['grains.setval'] inside code.
23:30 forrest Gareth, hmm, I don't think that actually writes it anywhere, I've always set it in a file
23:31 Gareth it seems to be persisting between restarts.
23:31 forrest Gareth, oh cool
23:31 Gareth hm. yeah. /etc/salt/grains.
23:32 to_json joined #salt
23:33 metaphore joined #salt
23:33 pdayton joined #salt
23:37 rap424 joined #salt
23:44 ocdmw manfred, the error I see is different from https://github.com/saltstack/salt/issues/12248.
23:45 ocdmw forrest, running both minion and master in debug mode shows nothing in the log when I try to issue commands as a non-root user who should be permitted by client_acl config on the master (and who worked before the OS upgrade to Ubuntu 14.04).
23:46 ocdmw So I infer the command is failing somewhere very early since it shows me a failure message without logging anything on master or minion.
23:46 forrest ocdmw, hmmm, that is interesting. I'm trying to think of anything that could affect it between releases.
23:47 sectionme joined #salt
23:49 ajprog_laptop joined #salt
23:52 ocdmw forrest, I also note what looks like a regression to a symptom I haven't seen for several months -- when I try to update a minion with state.highstate, all the right stuff gets done on the minion but the command returns on the master before displaying any information about the steps to bring the state up to date (which I am used to seeing in color based on success status)
23:52 ocdmw (that is, when I run state.highstate as root of course, since other users fail completely)
23:53 forrest ocdmw, oh that's a timeout issue, increase that with -t 60 or something
23:53 forrest should resolve that
23:53 ocdmw I am used to this lack of status info if I don't upgrade the master to match the minions.  Ahhhh forrest thanks yes
23:53 Ryan_Lane UtahDave: https://gist.github.com/ryan-lane/6473cbc59ca4dcef5b38
23:54 forrest ocdmw, I seriously don't get what is happening between two releases of ubuntu though, the users aren't hooked into AD or something weird are they?
23:54 ocdmw forrest, certainly not lol
23:54 forrest ocdmw, honestly I'd consider searching through the issues, and then opening an issue if you can't find anything
23:54 forrest it makes no sense it would just stop working
23:55 ocdmw Thanks for thinking about it, I'll do that.  I feel like I can't be alone since I'm not doing anything fancy on the machine as far as I know (admittedly, everyone thinks they are not doing anything fancy)
23:56 forrest ocdmw, haha
23:57 ocdmw forrest, the -t argument solved the lack of output.
23:58 forrest ocdmw, awesome
23:58 warmwaffles joined #salt
23:58 benharper joined #salt
23:58 warmwaffles Is there a way I can render a given sls file so that I can inspect my rendering fuckup?
23:59 forrest warmwaffles, not really unfortunatey. The best way is to past test=True
23:59 warmwaffles :( I was hoping salt dumped the rendered templates out somewhere

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