Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-04-16

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

All times shown according to UTC.

Time Nick Message
00:05 litwol_ Hello
00:05 litwol Is it possible for me to use grains as a value in command line cmd.run ?
00:05 litwol for example i want to mkdir [dir name] where dir name is minion's host ID
00:06 litwol strictly using CLI cmd.run
00:07 litwol ugh
00:07 litwol i can use 'hostname' in cmd.run
00:07 litwol hah!
00:07 litwol too much salt and i forgot my basics :-p
00:07 hal58th That works. I would like to still know that answer though
00:12 pahko joined #salt
00:14 nafg joined #salt
00:15 subsignal joined #salt
00:16 hal58th just for fun litwol, I mostly solved your problem.
00:16 hal58th sudo salt '*' cmd.run 'echo $(salt-call grains.get host)'
00:16 murrdoc u crazy
00:16 murrdoc and bravo
00:17 hal58th it's fun to wrap commands in other commands
00:19 ek6 hal58th: thats right up there with 'its fun to use regex'
00:25 Vynce hey, regex is fun
00:30 pahko joined #salt
00:30 elfixit joined #salt
00:31 danemacmillan joined #salt
00:33 Eureka__ joined #salt
00:33 ALLmight_ joined #salt
00:33 boredatwork joined #salt
00:33 primechuck joined #salt
00:34 debian112 joined #salt
00:35 Karlthane joined #salt
00:36 SaveTheRbtz joined #salt
00:37 johngrasty joined #salt
00:37 DaveQB joined #salt
00:37 stooj joined #salt
00:37 shoemonkey joined #salt
00:38 flebel joined #salt
00:38 cberndt joined #salt
00:41 aquassaut joined #salt
00:43 hax404 joined #salt
00:43 dojobonobo joined #salt
00:44 dojobonobo hello.. I’ve ran into a problem with YAML data in pillar.. I can’t get newlines in my multi-line strings
00:45 dojobonobo test: |
00:45 dojobonobo line1
00:45 dojobonobo line2
00:45 dojobonobo line3
00:45 dojobonobo renders as: line1 line2 line3
00:45 dojobonobo vs: line1\nline2\nline3
00:45 dojobonobo I tried > instead of | but both produce the same output
00:46 shoemonkey joined #salt
00:46 dojobonobo pyyaml generates the correct strign so IDK what salt is doing to disable \n’s
00:48 Daemonik joined #salt
00:50 Vynce when you say "renders as" where are you trying this?
00:51 cheus joined #salt
00:54 MatthewsFace joined #salt
00:55 pahko joined #salt
00:55 dojobonobo @Vynce here is the sls file that renders the data from pillar
00:55 dojobonobo https://gist.github.com/anonymous/313b47f571911e15ca39
00:55 dojobonobo so I guess because I’m passing it in as contents that it strips the newlines
00:55 Vynce my question was meant to be "ohw are you looking at what this renders?"
00:56 dojobonobo highstate
00:56 dojobonobo like if I cat out the file it doesn’t have newlines
00:56 dojobonobo after doing a highstate
00:56 dendazen joined #salt
00:56 yomilk joined #salt
00:57 fusionx86 joined #salt
00:58 dojobonobo pretty sure if I use `source` vs `contents` for file.manage and just pass the data in as `context` and render it with jinja it will work
00:59 dojobonobo kinda sucks that I have to create a seperate template file that all it does is render a variable.. maybe I can just loop over the variable with .split(‘\n’) in the sls file and do the | and everythign all over again
01:00 hal58th dojobonoobo or whatever your name is. why don't you just use contents_pillar?
01:01 hal58th It even says "This is useful when the pillar value contains newlines, as referencing a pillar variable using a jinja/mako template can result in YAML formatting issues due to the newlines causing indentation mismatches."
01:01 RandalSchwartz joined #salt
01:02 RandalSchwartz why is it so hard to google for salt docs and then navigate?
01:02 RandalSchwartz the highest seo for "salt highstate" returns http://docs.saltstack.com/en/latest/ref/states/highstate.html
01:02 RandalSchwartz which has broken prev/next/up links.
01:02 dojobonobo @hal58th I did not know of this! I was able to loop over it and get the result I wanted but contents_pillar appears like what I was looking for.. thank you1
01:03 hal58th the docs are kinda broken right now RandalSchwa.
01:03 robawt RandalSchwartz: it's come up a few times today that the current 'latest' doc is broken
01:03 hal58th welcom dojo
01:03 dojobonobo yeah I had to use dash.. live docs seem broke
01:03 RandalSchwartz I'm trying to remember how to run a single state in test mode
01:03 RandalSchwartz I think it's state.sls $state Test=true
01:03 RandalSchwartz or something like that
01:04 dojobonobo wow.. you are thee randal schwartz? of floss weekly?
01:04 RandalSchwartz yes I am
01:04 robawt RandalSchwartz: that's correct
01:04 dojobonobo I love your show :)
01:04 robawt salt/salt-call state.sls mystate test=True
01:04 RandalSchwartz ahh test=True
01:04 RandalSchwartz the capitalization never makes sense to me
01:05 robawt i like to think of it as a Python True
01:05 robawt :)
01:05 RandalSchwartz I don't know Python
01:05 litwol What happened to salt docs?
01:05 RandalSchwartz don't want to waste time learning
01:05 dojobonobo it’s Python True.. not perl ;)
01:05 robawt RandalSchwartz: your knowledge of Perl makes up for it
01:05 litwol ah. google has wrong URLs cached.
01:05 hal58th I've never heard of Floss Weekly. Time to watch an episode
01:05 robawt hal58th: it's pretty good, not just saying that because RandalSchwartz is here
01:06 robawt it's on my dogcatcher podcast player
01:06 hal58th just PM me the truth :P
01:08 dojobonobo yeah.. saltstack has gotten pretty decent coverage on floss.. think Thomas was on twice
01:08 RandalSchwartz yes
01:09 RandalSchwartz once for early salt, then for saltcloud
01:09 baweaver joined #salt
01:11 glyf joined #salt
01:12 hal58th hmmm, I'll watch that for sure. Thomas is great to listen too.
01:16 RandalSchwartz yeah he was a great speaker
01:16 RandalSchwartz convinced me to use salt :)
01:16 RandalSchwartz which would be fine if it weren't for that python aspect. :)
01:17 mgw joined #salt
01:17 dojobonobo how’s the perl config mgmt space?
01:17 iggy you take that back, right meow
01:17 RandalSchwartz config mgmt came after perl died. :)
01:18 dojobonobo you heard it folks.. straight from the perl man himself
01:19 robawt i clearly remember RandalSchwartz stating that he was doing this stuff in Perl way back when ;)
01:21 ek6 so if i do an salt-bootstrap install     git v2015.2   am i getting the latest off branch or the old tag?
01:21 enarciso joined #salt
01:22 ek6 bah nm
01:25 hal58th git develop :)
01:26 hal58th Btw, this show is pretty good RandalSchwartz. Watching the Online Abuse episode now. You have a new fan
01:28 hal58th To bad I can't use the built in search on that website
01:29 baweaver joined #salt
01:30 pahko joined #salt
01:30 dojobonobo they had pieter hintjens on from 0mq in case you are interested http://twit.tv/show/floss-weekly/195
01:30 dojobonobo he mentions salt towards the end
01:35 ckao joined #salt
01:36 buhman joined #salt
01:37 buhman can salt states be applied within a chroot?
01:37 robawt considering what makes up a chroot, yes
01:38 robawt you may want to consider whether the host machine will run the salt minion or you'll run a second instance
01:38 buhman ideally I'd avoid a second instance if I can
01:39 robawt yeah just run it from within the host instance then
01:40 buhman O.o I guess I don't understand how
01:40 CheKoLyN joined #salt
01:40 buhman is there some sugar that lets me say 'do this state, but enter this chroot first'?
01:40 robawt a chroot simply marks a specific directory tree as the root filesystem
01:40 seev joined #salt
01:40 robawt the filesystem still exists, just nothing outside of that FS exists for the targeted process
01:41 robawt well the directory is still there, so your minion can add/remove like it normally does
01:41 buhman O.o
01:41 buhman so in particular
01:42 buhman how about applying 'foo: pkg.installed' to a chroot
01:45 robawt ah I see.
01:45 robawt let's see if we can specify basedir
01:45 buhman well, that's a package manager thing ;p
01:45 buhman it would be useful to /actually/ enter the chroot I think
01:47 ilbot3 joined #salt
01:47 Topic for #salt is now Welcome to #salt | 2014.7.2 is the latest | Please use https://gist.github.com for code, don't paste directly into the channel | Please be patient when asking questions as we are volunteers and may not have immediate answers | Channel logs are available at http://irclog.perlgeek.de/salt/
01:58 druonysuse joined #salt
02:00 pipeep So when I run salt-ssh from my local machine, it always takes about ~20 seconds to complete, and I think this is because it's scp-ing over salt-thin every time.
02:00 pipeep Is there some way to get it to cache that?
02:01 pipeep (and running it with -l trace, and looking at the network graph seems to confirm this)
02:04 RandalSchwartz left #salt
02:04 ITChap joined #salt
02:10 pahko joined #salt
02:10 quickdry21 joined #salt
02:10 MaliutaLap joined #salt
02:11 MaliutaLap left #salt
02:20 p66kumar joined #salt
02:20 echo joined #salt
02:21 pdayton joined #salt
02:22 pipeep Also, I don't know if this is a bug or not, but the lxc module imports salt.utils.cloud, which in turn imports msgpack, which causes issues on salt-ssh, because msgpack isn't there.
02:28 dendazen joined #salt
02:34 Vynce sure sounds like a bug, unless lxc is marked as salt-ssh safe
02:35 Vynce er, unsafe
02:38 evle joined #salt
02:38 michelangelo joined #salt
02:39 favadi joined #salt
02:39 pipeep Vynce, it seems like `cloud` is only used for the cloud_init-type functions. https://github.com/saltstack/salt/blob/develop/salt/modules/lxc.py
02:40 Vynce hah.  missed RandalSchwartz. and i'm a perl fan.
02:42 capricorn_1 joined #salt
02:44 shoemonkey joined #salt
02:46 robawt joined #salt
02:49 runnyspot joined #salt
02:49 runnyspot joined #salt
02:53 theologian joined #salt
02:54 pahko joined #salt
02:55 RabidCicada joined #salt
03:00 viq joined #salt
03:02 fusionx86 left #salt
03:11 murrdoc joined #salt
03:19 p66kumar joined #salt
03:21 Hercules99 joined #salt
03:26 writtenoff joined #salt
03:44 Furao joined #salt
03:51 bhosmer_ joined #salt
03:51 zer0def joined #salt
03:52 murrdoc joined #salt
03:53 shoemonkey joined #salt
03:58 radu joined #salt
04:01 aw110f joined #salt
04:01 pahko joined #salt
04:07 runnyspot joined #salt
04:07 runnyspot joined #salt
04:07 aparsons joined #salt
04:08 TyrfingMjolnir joined #salt
04:08 pdayton joined #salt
04:17 pahko joined #salt
04:20 iggy ek6: vNNNN.N are tags NNNN.N are branch HEADs
04:20 SeeDickCode joined #salt
04:21 aparsons joined #salt
04:23 iggy pipeep: there's a bug open about it already
04:24 pipeep iggy, great, thanks!
04:24 iggy search for "salt-ssh msgpack" in the issues
04:24 iggy should get it
04:25 clintberry joined #salt
04:26 pipeep iggy, I don't see any specifically about lxc
04:26 pipeep But yeah, msgpack seems to be a longstanding issue
04:26 pipeep (in general)
04:27 bluenemo joined #salt
04:29 Aidan_ joined #salt
04:32 aparsons joined #salt
04:33 sunkist joined #salt
04:34 p66kumar joined #salt
04:40 runnyspot joined #salt
04:40 runnyspot joined #salt
04:40 aw110f joined #salt
04:43 murrdoc joined #salt
04:45 iggy yeah, it's not specific to lxc
04:46 shoemonkey joined #salt
04:49 aw110f_ joined #salt
04:50 pahko joined #salt
04:53 rdas joined #salt
04:53 __number5__ is there anyway to make sure git.latest using remote tags instead of local tags? always_fetch: True seems not working
04:56 stanchan joined #salt
04:57 favadi left #salt
04:57 iggy I think the key is _not_ to have remote and local tags that differ
04:57 iggy (in general... and not just in this case)
05:04 ajw0100 joined #salt
05:04 __number5__ iggy: actually it had issue because we re-used the same tag, our jenkins build tag lastSuccessful each build, local is older than remote
05:05 bmcorser joined #salt
05:06 __number5__ so when git.latest run, it found local have lastSuccessful tag then it just stopped and not doing fetch at all
05:06 Furao 2014.7.4 in ppa
05:08 tmh_ joined #salt
05:08 sxar_ joined #salt
05:13 dopesong joined #salt
05:20 wwwbukolaycom joined #salt
05:23 yggdrasi1 joined #salt
05:23 echo joined #salt
05:24 * wwwbukolaycom hi
05:27 I3olle joined #salt
05:31 yomilk joined #salt
05:37 tomasfej1 joined #salt
05:42 SpeedFuse joined #salt
05:45 favadi joined #salt
05:50 ITChap joined #salt
05:52 iromli joined #salt
05:54 krelo joined #salt
05:59 jalbretsen joined #salt
06:06 colttt joined #salt
06:10 AndreasLutro joined #salt
06:15 Karlthane joined #salt
06:21 dopesong joined #salt
06:21 wendall911 left #salt
06:21 pdayton joined #salt
06:21 dopesong joined #salt
06:22 ajw0100_ joined #salt
06:22 dopesong_ joined #salt
06:23 aravind joined #salt
06:24 flyboy joined #salt
06:24 tmh_ joined #salt
06:24 yomilk joined #salt
06:25 yomilk joined #salt
06:26 pahko joined #salt
06:27 I3olle joined #salt
06:29 malinoff joined #salt
06:32 SpX joined #salt
06:33 yomilk joined #salt
06:33 robawt joined #salt
06:34 rdas joined #salt
06:39 markm joined #salt
06:40 elfixit joined #salt
06:41 I3olle joined #salt
06:44 stoogenmeyer__ joined #salt
06:45 ndrei joined #salt
06:46 shoemonkey joined #salt
06:50 keimlink joined #salt
06:51 antonn __number5__ hey, are u still here?
06:53 antonn __number5__ u went and didn't gave me a hint about my orchestration question yesterday =(
06:55 __number5__ antonn: sorry, what was that again?
06:55 Romlok joined #salt
06:56 __number5__ oh, don't mix different environments in orchestration
07:00 otter768 joined #salt
07:02 hebz0rl joined #salt
07:02 wvds-nl joined #salt
07:03 hebz0rl joined #salt
07:05 N-Mi_ joined #salt
07:06 tomasfej1 joined #salt
07:07 ITChap joined #salt
07:09 kawa2014 joined #salt
07:12 fredvd joined #salt
07:15 eseyman joined #salt
07:16 HexOffender joined #salt
07:19 AndreasLutro joined #salt
07:20 nafg joined #salt
07:21 dRiN joined #salt
07:24 jhauser joined #salt
07:28 rdas joined #salt
07:29 seev joined #salt
07:30 Karlthane joined #salt
07:32 lb1a joined #salt
07:34 yomismo joined #salt
07:34 yomismo Hi, anyone can help me? https://serverfault.com/questions/682906/salt-order-execution-of-states
07:35 soren joined #salt
07:37 mccricardo joined #salt
07:38 mccricardo good morning everyone!
07:38 mccricardo noob question: in Salt is there an equivalent to Chef's 'knife bootstrap' command ?
07:39 pahko joined #salt
07:40 AndreasLutro mccricardo: not really, but you could set up a sls state that installs the salt minion and then run that via salt-ssh
07:42 mccricardo hum...
07:42 mccricardo sounds like a plan
07:43 mccricardo I just started looking into Salt, so many things are still new to me
07:43 AndreasLutro personally I'd just write a bash script that installs the minion
07:43 hubricode joined #salt
07:44 mccricardo yeah, I just created a simple python script to do just that
07:45 Grokzen joined #salt
07:45 mccricardo I was just wondering if there a Saltish way to achieve that
07:45 faust joined #salt
07:46 yggdrasi1 joined #salt
07:52 Not joined #salt
07:53 PI-Lloyd joined #salt
07:54 lothiraldan joined #salt
07:58 favadi left #salt
07:58 chiui joined #salt
07:58 malinoff mccricardo, https://github.com/saltstack/salt-bootstrap
07:58 pahko joined #salt
07:58 malinoff mccricardo, although it is not exactly 'knife bootstrap'
07:59 mccricardo yeah
07:59 Not https://www.shodan.io/
07:59 mccricardo I've written a simple python script that ssh's into whatever server I want to install a salt-minion on a runs salt-bootstrap
07:59 seev joined #salt
08:01 ktosiek joined #salt
08:01 CeBe joined #salt
08:06 kawa2014 joined #salt
08:08 leon__ joined #salt
08:10 kawa2014 joined #salt
08:18 ggoZ joined #salt
08:22 Xevian joined #salt
08:22 eliasp iggy: ha! finally managed to get around the firewall at my new workplace… back to my beloved Quassel IRC  - no more webchat.freenode.org :)
08:22 jrenner joined #salt
08:24 phx joined #salt
08:27 glyf joined #salt
08:33 favadi joined #salt
08:34 xintron joined #salt
08:35 tkharju joined #salt
08:43 joshin joined #salt
08:43 yomilk joined #salt
08:45 denys joined #salt
08:46 mccricardo left #salt
08:47 shoemonkey joined #salt
08:48 zz_Cidan joined #salt
08:49 Cidan joined #salt
08:54 KermitTheFragger joined #salt
08:58 yomilk joined #salt
09:01 otter768 joined #salt
09:02 dalexander joined #salt
09:05 yomilk joined #salt
09:09 Andre-B joined #salt
09:09 pahko joined #salt
09:09 Cidan joined #salt
09:14 ujjain joined #salt
09:15 monkey661 left #salt
09:19 ujjain joined #salt
09:26 |borat| Hi! Anyone from dev around?
09:28 huddy joined #salt
09:28 mrchris joined #salt
09:29 mrchris Hi Saltees.   I need to create a user within salt that uses the uid and guid values that exist in our LDAP system.  What's the best way of doing this with salt?
09:30 faust mrchris: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.user.html
09:31 mrchris faust: What's the best way of accessing LDAP and getting the uid and guid values from it?
09:31 faust mrchris: with http://docs.saltstack.com/en/latest/ref/pillar/all/salt.pillar.pillar_ldap.html
09:32 mrchris faust: thanks!
09:39 Furao joined #salt
09:41 cberndt joined #salt
09:41 iromli joined #salt
09:49 ChoOo7 joined #salt
09:49 ChoOo7 Hello :)
09:50 Auroch joined #salt
09:53 ChoOo7 I have some difficulties to run this simple case : http://docs.saltstack.com/en/latest/topics/tutorials/states_pt2.html, can someone try to help me to understand what i'm missing ?
09:54 mccricardo joined #salt
09:54 ChoOo7 The message is "Comment:   No Top file or external nodes data matches found", like if my top.sls is not accessible on the minion.but on the master, I have configurated the base file_roots... and my  top.sls is very minimal
09:55 soren joined #salt
09:55 ChoOo7 Which advice can you give to me in order to debug the situation ?
09:57 lothiraldan joined #salt
09:57 babilen ChoOo7: Welcome. I take it that you have installed your master and a minion. Where do you keep your states and top.sls files and how did you configure your master ? Could you paste relevant information to http://refheap.com ?
09:59 ChoOo7 https://www.refheap.com/99680
10:00 ChoOo7 Thanks for you help. More information could be usefull ?
10:01 zer0def is there a way to tell supervisord.running to force a service restart only when it's config file changes?
10:01 lothiraldan joined #salt
10:02 zer0def naturally, the supervisord.running state should otherwise just make sure that the service is running.
10:03 AndreasLutro zer0def: add a watch_in: { supervisord: name_of_service } to the config file's state
10:03 chiui joined #salt
10:03 AndreasLutro similar to how you'd do with service states
10:04 zer0def AndreasLutro: so no need to define restart=True in supervisord's then?
10:04 zer0def supervisord's running state then*
10:05 AndreasLutro zer0def: you can, I assume if you leave it as False it means reload instead of restart
10:05 AndreasLutro just like the service states
10:05 zer0def let me just double check
10:07 ChoOo7 Is there any interesting information in my copy paste ? which other information can I provide ,
10:07 ChoOo7 Is there any interesting information in my copy paste ? which other information can I provide ?
10:07 AndreasLutro ChoOo7: I don't know the answer to this but [DEBUG   ] Fetching file ** attempting ** 'salt://srv/salt/top.sls'
10:08 AndreasLutro as far as I know that means it's trying to look in /srv/salt/srv/salt/top.sls on the master
10:08 AndreasLutro which seems wrong...
10:08 Romlok try removing the last / on the file_root?
10:08 ChoOo7 Humm indeed, I'm going to look into this
10:08 babilen ChoOo7: Sorry, missed your reply (might want to highlight me). Will have a look now.
10:08 Romlok I have just /srv/salt, rather than /srv/salt/
10:09 babilen ChoOo7: A file root of /srv/salt is the default so you don't need that at all and the trailing slash *might* throw things off.
10:10 pahko joined #salt
10:10 ChoOo7 I think the probleme is state_top: /srv/salt/top.sls
10:11 AndreasLutro probably!
10:11 babilen ChoOo7: I don't see that at all in your paste, but that would *definitely* be problematic
10:11 ChoOo7 Thanks, it's working like a charm now :)
10:11 babilen \o/
10:12 babilen (paste the entire config the next time around)
10:12 bmcorser joined #salt
10:13 ChoOo7 :)
10:15 lothiraldan joined #salt
10:17 zer0def AndreasLutro: ok, added a watch on the file, apparently the terminology follows that of supervisord's manual, which gets a bit quirky compared to regular services
10:17 SpX joined #salt
10:17 babilen Ah, lovely: https://github.com/saltstack/salt/issues/22721
10:17 TyrfingMjolnir joined #salt
10:18 AndreasLutro oh? okay
10:19 zer0def AndreasLutro: basically supervisord.running makes sure that a particular process is started and the update param makes sure that if a process is already running, then it has re-read the changed config file
10:28 Jimlad_ joined #salt
10:35 mrchris faust: once I have the LDAP map setup in pillar_ldap.yaml, how do I get it to execute? I'm not seeing the expected values in my pillar dic
10:35 mrchris dict
10:44 CeBe joined #salt
10:48 shoemonkey joined #salt
10:51 favadi left #salt
10:56 babilen mrchris: Did you restart your master?
10:56 mrchris babilen: yes
10:56 lothiraldan joined #salt
10:58 babilen How did you configure it? Could you run the master in debug mode and/or check its logs?
10:59 drico joined #salt
10:59 lothiraldan_ joined #salt
10:59 drico does anyone uses this formula https://github.com/saltstack-formulas/php-formula ?
10:59 echo joined #salt
11:00 babilen drico: People use it, yeajh
11:01 wwwBUKOLAYcom joined #salt
11:01 drico I have a strange bug, when trying to use it with my pillar Rendering SLS "base:php.ng" failed: Jinja variable 'odict' is undefined and :
11:01 drico {% set php = salt['pillar.get']('php:ng', {    <======================
11:01 drico and I have defined php:, ng: in the pillar
11:01 drico any clues ?
11:01 babilen Oh, yeah ... .ng
11:02 otter768 joined #salt
11:02 babilen I'm not particularly fond of .ng formulas, but lets see .. Could you paste your pillar, top.sls and other applicable information (such as the entire error) to http://refheap.com and I'll take a look at the .ng crapola
11:03 lothiraldan joined #salt
11:04 Furao what are “.ng” formula?
11:05 babilen "new generation" formulas that are typically overly complicated and use a style that I wouldn't consider state of the art
11:06 Jimlad joined #salt
11:06 mrchris babilen: This is what I did : http://hastebin.com/xitoticibu.sm
11:06 denys joined #salt
11:06 mrchris babilen: i mean this http://hastebin.com/kidaretule.sm
11:07 babilen mrchris: Could you run "salt '*' saltutil.refresh_pillar" followed by "salt 'chrisr-Latitude-E6530' pillar.get test-users" ?
11:08 mrchris babilen: first command returns True, but second commands returns nothing for the single machine, just "chrisr-Latitude-E6530:"
11:08 drico babilen, https://www.refheap.com/2d922421460983fbe21b91479
11:09 mrchris babilen: Have this error in master log file : [salt.pillar      ][CRITICAL] Specified ext_pillar interface pillar_ldap is unavailable
11:09 babilen mrchris: And "neptune" is resolvable like that? Anything in the master's debug log?
11:09 babilen Ah, which version of salt is that?
11:10 mrchris babilen: 2014.7.2
11:11 pahko joined #salt
11:11 tomasfej2 joined #salt
11:11 babilen mrchris: You probably have to install python-ldap (or the package that provides the Python 'ldap' module). I don't necessarily see anything wrong with your configuration.
11:11 babilen Restart the master if you haven't done so already
11:11 babilen (after installing that package)
11:14 mrchris babilen: getting slightly closer now after installying package and restarting master :  error message is :    [salt.pillar      ][ERROR   ] Failed to load ext_pillar pillar_ldap: 'search_order'
11:15 mrchris + Traceback
11:16 tomasfej1 joined #salt
11:16 illern_ joined #salt
11:18 babilen May I see the traceback?
11:21 babilen mrchris: Looks as if you have to define "'search_order'" in your options. This does not seem to be documented though :-/
11:24 mosen joined #salt
11:25 babilen https://github.com/KrisSaxton/salt-ldap contains some additional information. The source is essentially iterating through elements in "search order" and is then using the configuration for those from the configuration. I'm not entirely sure how similar those modules are. Could you open a bug on https://github.com/saltstack/salt/issues with that error? It *should* be documented and people can't really be expected to read the source to track this down.
11:25 Furao joined #salt
11:25 babilen 8sigh*
11:29 _ale1_ left #salt
11:29 MAbeeTT joined #salt
11:30 babilen https://github.com/saltstack/salt/pull/2158/files is the original pull request from which bits of documentation were removed in https://github.com/saltstack/salt/pull/2263
11:31 babilen mrchris: Please file a bug, I'd like this to be clarified and I don't have the time to do more digging
11:31 AirOnSkin Hey guys. I'd like to do something and I don't know how. It's about HTTPD virtual host configuration via Pillar & a template file...
11:31 AirOnSkin I have the following pillar data: http://hastebin.com/ohevicipik.sm
11:32 babilen mrchris: fwiw, I guess that "search_order: - test-users" in your config might work (I would, however, rename that to test_ldap)
11:32 babilen err, test_users
11:32 AirOnSkin I know how to write a for-loop with .iteritems so the template file will be created for each virtual host in the list (vhost_1, vhost_2, ...)
11:32 Andre-B joined #salt
11:33 AirOnSkin What I don't know is, how I can access the values for ip & domain inside the files...
11:33 I3olle joined #salt
11:33 AirOnSkin Of course on the first run in the for-loop, the template file shall contain the ip and domain defined in vhost_1, in the second the ones for vhost_2
11:34 babilen You want to manage k files for k vhosts and generate those file.managed states from your pillar data, correct?
11:34 dopesong joined #salt
11:34 AirOnSkin babilen: yes, correct. the number of vhosts may differ from host to host
11:36 babilen AirOnSkin: Look at http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html#salt.states.file.managed context and defaults and read https://groups.google.com/forum/#!topic/salt-users/Es3DJCMzbu8
11:36 babilen That way you can pass those attributes "into" the rendered file
11:36 pahko joined #salt
11:36 babilen You might just want to pass the entire dictionary though rather than each attribute individually
11:38 AirOnSkin babilen: Ok, never used file.managed like that before. Will read into that. Thanks!
11:40 aqua^lsn joined #salt
11:45 AirOnSkin babilen: Just to clarify: I define a new variable in my file.managed state (which is inside a for-loop) and access that variable inside the template file (vhost.conf). The variable I define in file.managed may very well be a Pillar value, correct?
11:51 lothiraldan joined #salt
11:51 ltsampros joined #salt
11:52 ltsampros Hey all
11:52 ltsampros is there a way for salt-cloud API to become "async"?
11:52 ltsampros let's say via the job system ?
11:52 babilen AirOnSkin: You would generate various file.managed states (one for each vhost - that is what you want, isn't it?). You do so by iterating over entries in 'virtual_host' and you have access to each dictionary containing 'domain' and 'ip' entries.
11:54 CeBe1 joined #salt
11:55 ujjain- joined #salt
11:55 babilen AirOnSkin: You would either pass the entire dictionary or individual values via "context: - foo: the_thing_you_want_to_pass" into the template and access this data via {{ foo }} in there. If you were to pass the entire dictionary (i.e. you iterate over "for vhost in salt['pillar.get']('virtual_hosts')") you could pass it as "context: - vhost_config" and access individual values as {{ vhost_config.domain }} in the template itself
11:55 ujjain- joined #salt
11:56 babilen AirOnSkin: Well, it is probably better to pass it as "defaults" rather than "context", but that is the general idea
11:57 babilen That way you can use the same template for every vhost and dynamically create those files with appropriate data
11:57 AirOnSkin babilen: Great, I get that. Will try it out in a few ...
11:58 babilen Have fun
11:58 AirOnSkin Yeah, that's exactly how I want it :)
11:58 amcorreia joined #salt
12:15 zer0def uh, can root directories for pillars and the state tree be relative?
12:20 zer0def nevermind, works now; had some issue with relative paths earlier for some reason.
12:29 cmcmacken joined #salt
12:32 pahko joined #salt
12:39 cheus joined #salt
12:43 aquassaut joined #salt
12:49 dendazen joined #salt
12:49 shoemonkey joined #salt
12:49 wilfried joined #salt
12:50 wilfried joined #salt
12:50 wilfried left #salt
12:50 bluenemo joined #salt
12:50 bluenemo joined #salt
12:52 elfixit joined #salt
12:52 wilf1rst joined #salt
12:56 mapu joined #salt
12:59 glyf joined #salt
13:00 evle1 joined #salt
13:02 yggdrasi1 joined #salt
13:02 otter768 joined #salt
13:04 cpowell joined #salt
13:07 wilf1rst joined #salt
13:08 SpeedFuse joined #salt
13:11 favadi joined #salt
13:12 favadi left #salt
13:16 shoemonkey joined #salt
13:22 fusionx86 joined #salt
13:26 catpig joined #salt
13:28 Auroch joined #salt
13:29 shoemonkey joined #salt
13:36 primechuck joined #salt
13:39 cheus joined #salt
13:42 yggdrasi1 joined #salt
13:42 murrdoc joined #salt
13:48 Tecnico1931 joined #salt
13:48 patto joined #salt
13:50 Shawn joined #salt
13:52 timoguin joined #salt
13:53 Guest47329 Hi all
13:53 Guest47329 sorry i missed out something
13:54 rm_jorge joined #salt
13:55 yggdrasi1 joined #salt
13:55 davisj mport salt.config
13:55 davisj master_opts = salt.config.client_config('/etc/salt/master')
13:55 davisj minion_opts = salt.config.minion_config('/etc/salt/minion')
13:55 davisj >>> pprint(master_opts['nodegroups'])
13:55 andrew_v joined #salt
13:56 davisj ...paste fail...
13:56 Guest47329 Guys M new to salt stack
13:56 N-Mi_ joined #salt
13:56 Guest47329 I know that its a powerful tool and we can do a lot with configration management & remote command execution
13:57 Guest47329 I need to know, can we use salt to take a system & softwer inventory
13:57 Guest47329 if its possible pls tell me how
13:59 kaptk2 joined #salt
13:59 Guest47329 hellooooooooooooo
13:59 Guest47329 left #salt
13:59 Romlok salt will let you run...
13:59 Romlok oh well
14:00 drawsmcgraw Wow... short time window :/
14:00 manfred 3impacient5me
14:00 manfred impatient*
14:01 murrdoc HALP
14:01 murrdoc I COULD READ DOCS
14:01 murrdoc BUT HAALP
14:01 jacksontj Guest43385: "Please be patient when asking questions as we are volunteers and may not have immediate answers" ;)
14:01 fyb3r joined #salt
14:01 * murrdoc shakes head
14:01 manfred jacksontj:  wrong guest
14:02 sathya joined #salt
14:02 sathya Hi all
14:02 manfred yo!
14:02 murrdoc i am thinking of developing a salt-minion:// source tag
14:02 sathya can some one tell me how to use salt to take system & server inventory ?
14:02 murrdoc just like the salt://
14:02 debian112 joined #salt
14:02 jacksontj manfred: DOH :(
14:03 murrdoc instead of salt master being the fileserver
14:03 jacksontj too many guests ;)
14:03 sathya my question might be stupid, but M new to slat and my company needs it
14:03 murrdoc the minion has the files
14:03 murrdoc thoughts
14:03 nich0s Guest43385: Please tell your guest friend to chill. :)
14:03 jacksontj Guest43385: sorry about that-- wrong one ;)
14:03 soren joined #salt
14:04 rhodgin joined #salt
14:04 manfred jacksontj:  for you http://i.imgur.com/61QXhcd.png
14:05 jacksontj lol
14:05 jacksontj thanks
14:05 sathya anybody
14:05 manfred sathya:  what do you mean by inventory?
14:05 iggy mccricardo: there's salt-bootstrap and the saltify salt-cloud provider
14:05 mccricardo hum...
14:05 mccricardo let me look into that
14:06 masterkorp Do you guys configure salt-cloud on the master on your machine ?
14:07 ndrei joined #salt
14:07 manfred masterkorp:  i configure it for the master
14:07 sathya manfred : what are the softwares running on a server & when was it last patch, so kind of GUI stuff
14:07 manfred and in pillar data
14:07 manfred so that i can use salt cloud modules
14:07 sathya which people can understand just by looking at it
14:07 sathya does it make sence
14:07 sathya ?
14:07 manfred sathya:  you can do all that with just regular execution modules
14:07 berserk joined #salt
14:08 masterkorp manfred: so you run salt-cloud from the master ?
14:08 manfred you would have to check the logs, but you would probably have to write your own modules to do exactly what you want
14:08 manfred masterkorp:  yup
14:08 sathya yes, but I am unable to get the output in the way i want it
14:08 masterkorp manfred: and have a IAM role for the salt-cloud user
14:08 masterkorp ?
14:08 masterkorp what permissions do you give to it ?
14:08 sathya for eg: we need software invontory in excel formate , which I am unable to get it
14:09 giannello joined #salt
14:09 manfred sathya:  you could write your own returner to return it in whatever format
14:09 berserk joined #salt
14:09 manfred masterkorp:  i only use openstack
14:09 manfred what is IAM?
14:09 masterkorp i wish i could say that
14:09 timoguin joined #salt
14:09 sathya can you share me some examples, so that i can get some idea on how to write it
14:09 Furao joined #salt
14:09 masterkorp manfred: its basically role for multiple users of the same account
14:09 manfred oh yeah, i don't do that
14:10 manfred sathya:  if you go look at the repository, there are a bunch of returners and execution modules in there that you can base one off of
14:10 manfred sathya:  https://github.com/saltstack/salt/tree/develop/salt/returners
14:10 sathya alrite manfred, let me take a look at it
14:10 sathya thanks
14:10 manfred https://github.com/saltstack/salt/tree/develop/salt/modules
14:14 pdayton joined #salt
14:15 iggy no, salt can't do that
14:15 giannello anyone is distributing custom grains over gitfs?
14:16 iggy giannello: sure, put them in _grains in the git repo
14:16 giannello it seems it's not working
14:16 iggy saltutil.sync_all
14:17 iggy WORKSFORME ;)
14:17 giannello already done - it's not caching any custom grain
14:17 giannello might be because I have a subdirectory as the root of my gitfs?
14:17 iggy be more specific
14:18 giannello my git repo has /formulas/ and /pillars/ as subdirectories
14:18 giannello and gitfs configuration contains "root: formulas"
14:18 iggy but you have /formulas/_grains/ ?
14:18 giannello so I have my grains in /formulas/_grains -
14:19 iggy I would consider that a bug (or at least something that needs to be better documented
14:19 giannello well - let's test
14:19 manfred giannello: does it work if you move the _grains to / instead? if so, yeah i bet it just hasn't added the root: logic to the gitfs stuff and should be reported
14:20 jmdcal joined #salt
14:20 glyf joined #salt
14:20 giannello testing right now
14:21 awerner_ joined #salt
14:22 awerner_ joined #salt
14:22 giannello manfred, nope, still not working
14:22 bhosmer joined #salt
14:22 giannello also tried removing the root from the gitfs config - no luck
14:22 manfred kk, yeah i would report it. make sure to include salt —versions-report
14:23 giannello (y)
14:23 manfred oh, did you also clear out your /var/cache/salt/master/gitfs ? sometimes I have problems where i did something dumb and it causes my gitfs cache not to update
14:24 giannello let's try
14:24 wilf1rst joined #salt
14:25 giannello nope, still no luck
14:25 iggy in newer versions theres some specific commands to do that... handy as I was in the habit for a long time of rm -rf'ing /var/cache/salt/
14:25 giannello the custom grain is OK, as it's correctly loaded on my dev vm
14:25 manfred heh yeah
14:25 awerner_ joined #salt
14:25 giannello (using the same structure as the git repo)
14:26 manfred you could do what I used to do
14:26 iggy giannello: fwiw, I would use separate repos for states and pillars for a myriad of reasons
14:26 manfred have a git.latest that pulls it down to /srv/
14:26 manfred and then add a webhook to post-recieve git hook to hit the salt api to kick off that state :P
14:27 TheoSLC joined #salt
14:27 iggy ^
14:27 iggy that's on our TODO list
14:27 manfred until it is fixed at least*
14:27 giannello iggy, we're planning to split - this repo grew too quickly :D
14:27 iggy mostly because we have a ton of gitfs remotes and the fileserver update timer being tied to the scheduler timer is ridiculous
14:28 shoemonkey joined #salt
14:28 iggy when our git servers are slow, it takes more than a minute to sync all our repos
14:29 manfred iggy:  http://i.imgur.com/6Ga4u5v.jpg
14:29 TheoSLC joined #salt
14:30 iggy <3 TFS
14:30 HappySlappy joined #salt
14:32 giannello oh - well
14:32 giannello no idea what happened
14:32 giannello but they are there
14:32 giannello now
14:33 murrdoc never got the hang of thursdays
14:33 giannello I should have tried removing the gitfs cache earlier :D
14:33 dharper_ii joined #salt
14:33 scoates Salt is really saving my sanity, today. Thanks everyone.
14:34 giannello let us know as soon as you start hitting strange bugs ;)
14:35 Jimlad_ joined #salt
14:37 manfred heh
14:40 perfectsine joined #salt
14:42 bmac2 joined #salt
14:44 numkem_ joined #salt
14:48 randomuser joined #salt
14:49 wendall911 joined #salt
14:50 Grokzen joined #salt
14:50 Furao joined #salt
14:53 glyf joined #salt
14:55 clintberry joined #salt
14:56 StDiluted joined #salt
14:56 malinoff joined #salt
14:58 scbunn joined #salt
15:01 murrdoc is there a 'sanitize' method in salt.utils for cleaning up name parameters ?
15:02 bhosmer joined #salt
15:02 murrdoc i will look at code, hoping for an easy win is all
15:03 otter768 joined #salt
15:05 murrdoc def sanitize_win_path_string(winpath):
15:06 raygunsix joined #salt
15:07 lictor36 joined #salt
15:08 subsignal joined #salt
15:08 ndrei joined #salt
15:13 hasues joined #salt
15:15 mpanetta joined #salt
15:16 manfred lol, the ): in that function definition turned into a frowny face icon.  That makes me oddly happy
15:16 zircote joined #salt
15:17 I3olle joined #salt
15:18 hasues left #salt
15:18 esogas_` joined #salt
15:19 drawsmcgraw Anyone ever seen Salt calls take a long time
15:19 drawsmcgraw Like, a minute for a test.ping to return?
15:19 drawsmcgraw Not even across the network. Just a `salt-call test.ping` on the minion
15:20 desposo joined #salt
15:21 vauxia joined #salt
15:21 schristensen joined #salt
15:21 zircote joined #salt
15:22 clintberry joined #salt
15:23 mortis_ on what platform
15:23 drawsmcgraw CentOS 6
15:23 drawsmcgraw I have lots of other minions, also Cent, that behave just fine.
15:24 mortis_ i think we experienced something like that when salt filled our inodes up
15:24 mortis_ df -i
15:24 mortis_ ?
15:24 mortis_ ended up deleting jobcache
15:24 drawsmcgraw good try, but looks like there's plenty of space there...
15:24 iggy I've seen that with remote calls when certain minions were down... but not sure about via salt-call
15:25 drawsmcgraw 1% in use for each mount...
15:25 mortis_ weird, restart minion?
15:25 drawsmcgraw restarted the box several times
15:25 jalbretsen joined #salt
15:25 mortis_ old zmqversion?
15:25 drawsmcgraw hrm...
15:25 drawsmcgraw What's that incantation again?
15:25 mortis_ wild guesses here
15:25 mortis_ there is some salt version run thingie
15:25 mortis_ that lists all
15:26 drawsmcgraw yeah that's what I'm looking for :)
15:26 drawsmcgraw But
15:26 mortis_ im on ipad :x
15:26 drawsmcgraw oh goodness
15:26 drawsmcgraw Also, when I start the minion in the foreground, it hangs on 'reading config file /etc/salt/minion.d/master.conf'
15:26 drawsmcgraw I mean, it gets there eventually, but that's the last line in the output I see for upwards of a minute before it resumes
15:26 mortis_ is the master up? :D
15:27 drawsmcgraw heh. Yes, the master's up
15:27 masterkorp I am trying to create an instance with salt-cloud
15:27 masterkorp [ERROR   ] There was a profile error: 'str' object has no attribute 'copy'
15:27 manfred drawsmcgraw:  what version of zmq do you have on that?
15:27 masterkorp any ideas ?
15:27 manfred masterkorp:  do a debug, sounds like someone did a .copy on something in there
15:27 drawsmcgraw manfred: Do you have the versions_list incantation handy?
15:27 manfred salt-call —versions-report
15:27 manfred ?
15:27 manfred i just want
15:27 mortis_ thats the one
15:27 manfred python -c 'import zmq; print zmq.__version__'
15:28 drawsmcgraw manfred: ZMQ 3.2.5
15:28 manfred that one should be fine
15:28 drawsmcgraw I'm becoming convinced that it's not a Salt issue...
15:28 manfred the real problems with zmq were/are with 2.x.x
15:28 drawsmcgraw Rather the underlying OS
15:28 drawsmcgraw ah
15:29 mortis_ --log-level=garbage ?
15:29 manfred drawsmcgraw:  https://github.com/saltstack/salt/issues/891
15:29 manfred but yeah, probably not that
15:30 drawsmcgraw manfred: Interesting read, but yeah, not that
15:31 dopesong joined #salt
15:31 masterkorp manfred: https://www.zerobin.net/?d6cd5c246f1ee163#7XUUoIJ3l1Q3qkxRagZxDmAuebmWr2PJ9Az2X/IYXT4=
15:32 masterkorp manfred: not much more information
15:32 manfred that is all i needed
15:32 manfred just that stack trace
15:32 masterkorp manfred: https://www.zerobin.net/?9370e1b1c4bfeb65#B4M1NpOfLo/aO0LhqC4I7sb2+T07NpwoIXamfyTs/r8=
15:33 drawsmcgraw So... I ran a test.ping with -l trace and this is what came out after a long (> 1 minute) pause: https://www.zerobin.net/?9b4721f9d677c8b7#j4wzMFVuXXEgLE86AgF+5gQqRO3RK6DW8TzSLBlVerU=
15:33 drawsmcgraw afk for a moment...
15:33 masterkorp manfred: what does tell me ?
15:33 manfred gimme a minute
15:33 manfred what version of salt are you on?
15:34 Brew joined #salt
15:34 manfred masterkorp:  ^^
15:34 scbunn joined #salt
15:34 bhosmer joined #salt
15:35 masterkorp manfred: salt 2014.7.2 (Helium)
15:36 Brew joined #salt
15:38 manfred masterkorp:  has this config worked in the past?
15:38 manfred can you sanatize it and past it?
15:38 manfred paste*
15:38 masterkorp no
15:39 masterkorp i am trying for the first time
15:39 manfred aight, please sanatize it and paste it, your provider config
15:39 masterkorp manfred: https://www.zerobin.net/?1592a3726157f8bf#gYTcO606uK0q7lXmSJ3Z1U3D/4yEY1neBWxjZ/5stf4=
15:39 masterkorp oh the provider seems to work
15:39 Furao joined #salt
15:39 masterkorp just not the profile
15:40 manfred ahh, then sanatize the profile please
15:40 masterkorp I can query AWS just fine
15:40 masterkorp salt-cloud -Q give me acuratie information
15:40 seev joined #salt
15:40 manfred can you paste the profile so I can see it?
15:41 lothiraldan joined #salt
15:41 p66kumar joined #salt
15:41 masterkorp manfred: https://www.zerobin.net/?c5f40ac19403cd49#+zWqaNFEkrl30ocse0rxZv93w4UY4XZkA8aq59FNKpo=
15:43 manfred that all looks right
15:45 manfred just for shits and giggles, can you paste your sanatized provider?
15:46 manfred and also the exact command you are running to build the server
15:46 masterkorp manfred: https://www.zerobin.net/?1592a3726157f8bf#gYTcO606uK0q7lXmSJ3Z1U3D/4yEY1neBWxjZ/5stf4=
15:46 masterkorp that is my provider
15:46 masterkorp just that
15:47 mrchris joined #salt
15:47 manfred ok
15:47 masterkorp any idea ?
15:48 manfred so, your profile has the provider set to ec2, but the actual provider is named aws-micro, and it has it's provider set to aws
15:48 masterkorp oh
15:48 manfred can you change the profile name to something other than aws, like my-aws
15:48 manfred and then set the provider in the profile to aws-micro
15:48 manfred and run salt-cloud -p my-aws <servername>
15:49 mrchris Hi All. I've using pillar_ldap and it runs an LDAP query and gets some valid data, but it actually returns just {}.  You can see it all in the logs.  What is happening here?  http://hastebin.com/piteyoviya.vhdl
15:49 zwi joined #salt
15:49 masterkorp manfred: oh, i tough i would call salt-cloud like salt-cloud -p <profile-name> <server-name>
15:49 manfred yes
15:50 manfred i said change the profile name to my-aws
15:50 manfred the one you have here https://www.zerobin.net/?c5f40ac19403cd49#+zWqaNFEkrl30ocse0rxZv93w4UY4XZkA8aq59FNKpo=
15:50 manfred first line, my-aws:
15:50 masterkorp why ?
15:50 manfred and change the provider in that file to aws-micro
15:50 manfred because it conflicts with the drivername
15:50 manfred which i think is causing your problem
15:50 masterkorp oh
15:50 masterkorp that makes sense
15:51 masterkorp manfred: no the problem persists
15:51 manfred did you change the provider in the profile as well to be aws-micro?
15:52 masterkorp hmm, maybe I am not understanding your point very well, can you paste me example configs so i can follow your toughts ? If that is not asking too much of your good will.
15:53 manfred https://www.zerobin.net/?c5f40ac19403cd49#+zWqaNFEkrl30ocse0rxZv93w4UY4XZkA8aq59FNKpo=
15:53 manfred in that file, you have provider: ec2
15:53 masterkorp manfred | can you change the profile name to something other than
15:53 manfred but your provider configured in https://www.zerobin.net/?1592a3726157f8bf#gYTcO606uK0q7lXmSJ3Z1U3D/4yEY1neBWxjZ/5stf4=
15:53 masterkorp | aws, like my-aws
15:53 manfred is named aws-micro
15:53 masterkorp oh here you mean provider ?
15:53 manfred no, both
15:54 manfred in the profile file.  Change the first line to my-aws
15:54 masterkorp done
15:54 manfred and then on the provider line, change it to provider: aws-micro
15:54 manfred (in the profile
15:54 manfred the provider file should still have the provider set to aws
15:54 * iggy thinks a pastebin would be a bit faster
15:54 manfred meh
15:55 zircote joined #salt
15:55 OneAngryDBA joined #salt
15:56 masterkorp manfred: https://www.zerobin.net/?860284e0a5a0b598#b5uF7KeMAs+BYb9SxEcKgm31T7m3e5TumfLCoOao51w=
15:56 OneAngryDBA joined #salt
15:56 manfred show me your profile
15:56 masterkorp i think the provider on really needs to be value of ec2
15:56 manfred no
15:56 manfred it is fine to be aws
15:56 manfred it will use the libcloud_aws driver
15:57 masterkorp now. not even salt-cloud -Q will work
15:57 masterkorp ok, just a second
15:58 enarciso joined #salt
15:58 manfred masterkorp:  http://ix.io/hV2
15:58 manfred like that
15:59 manfred bah
15:59 manfred not like that
15:59 masterkorp manfred: https://www.zerobin.net/?bf7e83027f76008f#9oGSy/+QgPMH7F8KQagh8skAFeaFtPJt8z+ZOr5XGs8=
15:59 manfred masterkorp:  http://ix.io/hV3
15:59 * manfred headdesks
15:59 manfred masterkorp:  in the profile file
15:59 manfred change
15:59 manfred the provdier
15:59 manfred to aws-micro
16:00 manfred http://ix.io/hV3
16:00 masterkorp i've tried that
16:00 timoguin joined #salt
16:00 manfred well it can't be aws
16:00 manfred because that is the driver
16:00 manfred and isn't configured
16:00 masterkorp ok, now it makes sense
16:00 manfred also
16:00 masterkorp manfred: thanks for you patience
16:00 manfred you made the wrong one names my-aws
16:00 theologian joined #salt
16:01 manfred http://i.imgur.com/qYCH6ot.gif
16:01 manfred <3
16:02 denys joined #salt
16:03 masterkorp manfred: question: on your paste. the provider.conf is really the profile.conf right ?
16:03 manfred no
16:03 dramagods joined #salt
16:03 manfred actually
16:03 manfred apparently yes
16:03 manfred i have been reading them wrong dammit
16:04 manfred hold on
16:04 manfred let me fix it
16:04 masterkorp manfred: https://www.zerobin.net/?384716a3e76278d8#RSk8eCEy5krJGiidVYH+HNSqrPLn49XtCxqE5V0k4wI=
16:04 masterkorp this is the error
16:04 manfred but that is definitely your problem,
16:04 manfred yeah
16:04 manfred gimme a second, i can fix it
16:05 manfred http://ix.io/hV4
16:05 manfred that one should be all in the right order now
16:05 manfred that would explain why there was confusion when I was trying to explain it
16:06 manfred the provider: aws was throwing me off in the profile, because that would be directly referencing a driver, instead of your configured provider, which is why provider has to be my-aws instead of aws
16:06 manfred otherwise it causes conflicts when looking up provider/driver information
16:07 masterkorp i still get
16:07 masterkorp Error:
16:07 masterkorp Profile aws-micro is not defined
16:07 murrdoc http://ix.io/ is too l33t
16:07 murrdoc i like it
16:07 manfred it shouldn't give you that
16:08 masterkorp manfred: https://www.zerobin.net/?0552bf6f50e5fb19#KJ1BWtLZjbiaN4mqQHJbAPkHW6iSez9KuBK9sJOJnP8=
16:08 manfred murrdoc:  i like it, tail -c +0 *.conf | curl -F 'f:1=<-' ix.io
16:08 manfred masterkorp:  change the provider in the profile configuration to my-aws
16:08 manfred aws-micro:
16:08 manfred provider: aws
16:09 masterkorp will do
16:09 SeeDickCode joined #salt
16:09 manfred that should be my-aws
16:09 manfred otherwise the alias isn't configured in that providers profiles in the lookups
16:09 shoemonkey joined #salt
16:09 masterkorp ok, this now starts to make sense
16:09 manfred i wish that providers in the provider config were actually called drivers:
16:09 manfred would make it make more sense
16:10 SeeDickCode joined #salt
16:10 masterkorp Error:
16:10 masterkorp Profile my-aws is not defined
16:10 masterkorp oh its to aws
16:10 manfred yeah
16:11 manfred put it back to ec2
16:11 manfred that was my bad
16:11 manfred i had it confused
16:11 manfred cause there is a driver named 'aws'
16:11 manfred which caused all the problems
16:11 masterkorp i am more confused than ever
16:11 masterkorp let me paste you the files and output
16:12 iggy anybody else running 2015.2? Have you run into issues with the new functionality of salt installing all the packages up front?
16:12 steffo joined #salt
16:12 jespada joined #salt
16:12 iggy <3 babilen
16:12 manfred https://github.com/saltstack/salt/blob/develop/salt/cloud/clouds/libcloud_aws.py#L85
16:12 ek6 iggy: yes i am, and no but im pretty careful iggy with grouping my states together
16:12 babilen iggy: yw :)
16:13 masterkorp manfred: https://www.zerobin.net/?a198812e11154df0#B2X3BCbXlqLdxk3jYYbpZRXicKyjn6Jb85KNicVlP60=
16:13 manfred iggy:  i am, what is the problem?
16:13 SeeDickCode joined #salt
16:13 manfred masterkorp:  that should be good
16:13 iggy so 2015.2 added this new feature of salt going through and finding all the packages to install, and installing them all up front
16:13 manfred gross
16:13 iggy unfortunately it's doing that before my pkgrepo is setup
16:13 masterkorp manfred: but it gives me that error
16:13 manfred yeah
16:13 ek6 iggy: you just have to consider build dependencies and install order
16:13 manfred masterkorp:  what error now?
16:14 masterkorp [ERROR   ] There was a profile error: 'str' object has no attribute 'copy
16:14 masterkorp same as in the begining
16:14 iggy so... it's installing old crappy distro packages instead of my shiny new mongodb.org packages
16:14 babilen iggy: Do you have a requisite on the pkgrepo?
16:14 iggy ek6: you mean I need a require: ?
16:14 KyleG joined #salt
16:14 KyleG joined #salt
16:14 manfred masterkorp:  i would open a bug, cause I have no idea now, but your thing should be all good, and the provdier shouldn't conflict
16:14 manfred iggy:  that is gross… i don't want to have to do a require. i like my top down order
16:15 iggy same
16:15 manfred i am using 2015.2 and haven't run into that problem
16:15 manfred i have don't currently have any requires
16:15 MatthewsFace joined #salt
16:15 ek6 iggy: well that certainly would do it...but i just ended up doing basically 3 steps.... prep  which does things like your talking setting up all my private repos etc....install...post install configure
16:16 manfred iggy:  https://github.com/bighops/taplist-salt/blob/master/repos/init.sls repos
16:16 iggy manfred: well, it's showing up mainly on a highstate of a completely fresh system (haven't noticed it otherwise either
16:16 manfred iggy:  redis30u comes from ius https://github.com/bighops/taplist-salt/blob/master/redis/init.sls
16:16 manfred repos before redis in top.sls
16:16 manfred https://github.com/bighops/taplist-salt/blob/master/top.sls
16:16 ek6 depends on what your installing of course manfred...try and have openstack install in whatever order salt wishes and just watch the mid air explosion
16:16 manfred iggy:  yeah, i did it on a completely fresh system
16:16 manfred my thing would fail if the ius repo wasn't enabled
16:17 ek6 iggy: for me the speed increase was worth the refactor though..no doubt about it
16:17 faust joined #salt
16:18 desposo joined #salt
16:18 iggy sadly, this is in a formula, I can't expect people to jump through a bunch of hoops to use it
16:20 ek6 bah...well you can always just hand toggle it off in the formula right?  I figured you had to count on folks having that enabled going forward for those
16:20 iggy ENOPARSE
16:21 sjwoodr joined #salt
16:21 sjwoodr Hey guys.  Is there some known issue with 2014.7.4 and the "'fun' and 'state' are invalid keyword arguments"  warnings I keep getting now ?
16:22 iggy sjwoodr: safe to ignore... someone backported stuff from devel they _really_ shouldn't have
16:22 axion joined #salt
16:23 sjwoodr ok.  even this:  http://pastebin.com/HAx7yfM3  ... generates that   :)
16:23 axion left #salt
16:23 sjwoodr and im clearly not using 'fun' or 'state' ..
16:23 ek6 wow iggy how politically correct of you
16:23 dalexander joined #salt
16:24 iggy sjwoodr: yeah, it's... safe to ignore
16:24 sjwoodr iggy: great, thanks.  :)
16:25 iggy I think the problem isn't with that actual state, it's the internal call/event that gets fired because of it
16:25 iggy but yeah... ignore it or scream on one of the bugs about it (I like that idea)
16:26 iggy 2015.2 is much more chatty (in both good and bad ways)
16:26 ek6 iggy: i know you use the bootstrap to install, have you found it safe to rerun the bootstrap to UPGRADE or do you do a bunch of removes before rerunning the bootstrap?
16:26 FeatherKing joined #salt
16:26 iggy just install on top
16:26 iggy but then we trash minions fairly often, so I'm not sure I'd call it fool proof
16:27 sjwoodr one thing i've seen is that my salt-master is 2015.2.0rc2 but its creating minions with 2014.7.4
16:27 ek6 alright just wanted to confirm before three week from now and you giving me 'well what did you do that for?'  more talking about the master
16:27 sjwoodr it = salt-cloud
16:28 masterkorp manfred: https://github.com/saltstack/salt/issues/22749
16:28 masterkorp any ideas would be welcome
16:28 iggy sjwoodr: you'd have to force salt-cloud to install anything other than the "latest release"
16:28 FeatherKing does anyone know when "tcp_keepalive: True" became valid in the master config
16:28 FeatherKing when = what salt version
16:28 MatthewsFace joined #salt
16:31 iggy FeatherKing: it's a minion config option
16:32 FeatherKing http://salt.readthedocs.org/en/latest/ref/configuration/examples.html
16:32 iggy yes, it's in the minion section
16:32 FeatherKing ah i see both are on the page
16:32 tomh- joined #salt
16:32 p66kumar joined #salt
16:32 FeatherKing iggy: do you know what minion version that is valid?
16:33 iggy a _long_ time ago
16:33 FeatherKing ok, having a problem with my wan minions
16:33 iggy but note, there's a bug fixed in 2015.2/devel that actually made it work both ways
16:33 FeatherKing i think the firewall might be trashing the connection
16:33 FeatherKing without telling anyone
16:33 FeatherKing like the bug could be a master side setting?
16:34 iggy I don't remember what side it was on
16:34 FeatherKing ok
16:34 FeatherKing i run into this every so often where the minions show established connections on 4505 but no one is home
16:34 iggy sounds like the one
16:34 FeatherKing if you ever restart the minion in debug it fixes the issue with a new connection
16:35 FeatherKing is it worth pushing the tcp keepalive to the minions or just wait for 2015.2
16:35 bluenemo_ joined #salt
16:36 iggy https://github.com/cachedout/salt/commit/54f3f16c9fd9cc252fc9695eab8bb5efe3c4213f
16:37 FeatherKing so even if i set the keepalive it will probably only work on 4505
16:37 FeatherKing until i get that fix
16:37 iggy si
16:37 iggy at least afaik
16:37 FeatherKing seems like it
16:38 FeatherKing ill probably set it then
16:38 iggy your description matches what Mike was saying at saltconf
16:38 iggy the minion keepalive defaults to on
16:38 FeatherKing yeah i vaguely remember hearing it at saltconf
16:38 FeatherKing oh i didnt know that
16:39 FeatherKing does 4506 always exist
16:39 FeatherKing should it
16:39 iggy Default True, enabled
16:39 iggy I'm not sure tbh
16:39 FeatherKing ok
16:39 FeatherKing in my head i thought maybe it just opened it when it needed to send return data
16:40 FeatherKing couldnt remember
16:40 iggy I know we've had 0 problems with minions dissappearing, becoming unresponsive, etc. since we went to 2015.2
16:40 kawa2014 joined #salt
16:40 FeatherKing nice
16:40 FeatherKing apologize if im breaking a irc rule, but is there a date for 2015.2 yet
16:40 iggy to the point that I've pushed "Play with RAET" _way
16:40 FeatherKing seeing as its not '2' anymore
16:40 iggy _ down my TODO list
16:40 FeatherKing ah nice
16:40 FeatherKing it was on my todo as well
16:40 FeatherKing because of this
16:41 iggy I... can't even get into versions/release dates/etc
16:41 stanchan joined #salt
16:41 iggy it stresses me out
16:41 speedlight joined #salt
16:41 speedlight joined #salt
16:41 FeatherKing ha np
16:41 FeatherKing we have almost no local minions
16:41 FeatherKing they are split 1600 total, 800 in two data centers
16:42 FeatherKing so we have this pretty often
16:42 manfred Barium
16:42 manfred Lithium
16:42 FeatherKing dumb firewalls trashing the connection or whatever
16:42 manfred Neon
16:42 iggy we have a long way to go till Barium
16:42 manfred heh
16:42 FeatherKing stress level rising
16:42 iggy lol
16:42 manfred that was the goal <3
16:43 FeatherKing ok well thx for the info, ill probably just keep an eye on it until 2015.2
16:43 iggy usually a couple test.ping's would get things talking again
16:44 FeatherKing i was thinking of setting a cron for now
16:44 FeatherKing every couple minutes maybe
16:44 manfred FeatherKing:  have you seen salts scheduler?
16:44 iggy something about the minion realizing that it wasn't talking back to the master right, so it would, re-set-up the connection
16:44 jeffspeff joined #salt
16:44 manfred you could do it with that instead of cron
16:44 manfred http://docs.saltstack.com/en/latest/topics/jobs/schedule.html
16:45 FeatherKing i remember skimming over it once
16:46 Rory_ joined #salt
16:46 FeatherKing does it run on all minions by default or am i missing the targeting
16:47 spookah joined #salt
16:48 masterkorp https://github.com/saltstack/salt/issues/22749
16:48 masterkorp I've think I've added the most information possible
16:48 manfred FeatherKing:  it can be configured on the master to tell the minions to do it?
16:49 manfred i do not believe it has targeting
16:49 FeatherKing manfred: yeah i was just thinking throw it on the master for a test.ping every 5 minutes or something
16:49 manfred i think it uses the runner to run it on all minions
16:49 FeatherKing ah
16:49 FeatherKing thats what i was wondering if it was all or nothing
16:49 FeatherKing seems to be all
16:49 FeatherKing if you put it on the master side
16:50 pdayton joined #salt
16:50 manfred actually, i have no idea now
16:50 ja-s joined #salt
16:50 FeatherKing heh
16:50 manfred might be using runners if you configure it on the maser
16:50 manfred and execution modules if configured on the minion
16:50 Gareth FeatherKing: what are you trying to do with the salt scheduler?
16:50 manfred since there is no targeting
16:50 ndrei joined #salt
16:50 manfred Gareth:  hit all his minions with test.ping every 5 minutes
16:50 FeatherKing Gareth: well a temporary test.ping as a keepalive until 2015.2
16:51 manfred he was going to do it with a cron, i thought you could do it with the scheduler from the master
16:51 gerl1ng joined #salt
16:52 Gareth scheduler on the master is only for runners, but you could have a schedule job on all minions that runs test.ping fairly easily.
16:53 FeatherKing only downside is then im touching 1600 minions where salt is already kind of broken
16:54 FeatherKing but i have to touch them anyway at this point to restart the service
16:54 manfred might just do the cronjob then :P
16:54 FeatherKing im not discounting the scheduler
16:54 FeatherKing just thinking out loud
16:54 Gareth Or toss the schedule item in pillar, then target pillar to the minions you want it run.
16:54 manfred ahh yes
16:54 manfred forgot it could be in pillars
16:54 FeatherKing better
16:55 mgw joined #salt
16:55 FeatherKing that i will probably do
16:55 FeatherKing because i dont want to bury my master
16:55 FeatherKing seems like one data center is worse
16:55 Gareth Or...using the schedule execution modules to add the schedule job :)
16:55 Gareth ...if you're running a release that has the schedule execution module.
16:55 manfred what is this madness!
16:55 FeatherKing 2014.7
16:56 iggy salt.states.schedule
16:56 Gareth FeatherKing: Yup.  Available in there.
16:56 Gareth Yup.  state module would work too.
16:56 manfred i need to start reading the pull requests again
16:56 writtenoff joined #salt
16:56 manfred I am falling behind!
16:56 FeatherKing great ty all. i hope for a speedy 2015.2 since i think that will actually solve my problem
16:56 Gareth manfred: You should have come to my talk at SaltConf :)
16:56 manfred salt '*' schedule.add job1 function='test.ping' seconds=3600
16:57 manfred Gareth: when was it?
16:57 iggy I got in trouble when I was following all of salt
16:57 Gareth second day...11am iirc.
16:57 manfred hrm, i don't remember what I was at, but I should have
16:57 manfred saltconf is all a blur to me now
16:57 iggy apparently me telling every other pull request to add version annotations to their code was frowned upon
16:57 manfred is your video up already?
16:57 manfred iggy:  lol
16:58 iggy all the videos are up afaik
16:58 manfred iggy:  sounds like the day I got in trouble for deleting terrible aur packages for archlinux because they should be doing them better
16:58 manfred iggy:  i had techhat poke rhett about putting mine up
16:58 manfred so i am pretty sure there are still more not up yet
16:58 Gareth manfred: Don't see it up there yet.
16:58 manfred some of them are still private
16:59 iggy I got an email or FB notice or something that said the videos were up.... guess that should have been "some of the videos are up"
16:59 manfred yeah
17:00 jasonrm joined #salt
17:02 shoemonkey joined #salt
17:02 bhosmer_ joined #salt
17:03 hal58th_ joined #salt
17:03 robawt i'm supposed to complain to rhett about the lack of Salt shirts at Pycon
17:03 hal58th_1 joined #salt
17:03 robawt I am now forced to go to work half naked due to the lack of Salt clothing
17:04 manfred heh
17:04 manfred yeah
17:04 manfred UtahDave promised me that he wouldn't let the lack of shirts go unpunished
17:04 hal58th_1 If you want the salt Tech videos, I have the link.
17:04 manfred and wouldn't happen again
17:04 otter768 joined #salt
17:04 manfred hal58th:  hrm?
17:04 * Gareth covers his eyes to avoid seeing a half naked robawt
17:05 * murrdoc whips out camera , turns off flash
17:06 hal58th_1 if only I can remember the command to pm someone in irssi
17:06 Gareth /query
17:07 murrdoc /msg <personname> sup fool
17:07 hal58th_1 eh, I wanted a private window.
17:07 murrdoc it does that
17:07 murrdoc whenever the person replies
17:07 robawt isn't it /msg? or is that cool?
17:08 manfred it is both
17:08 manfred some clients don't open the extra window with /msg though
17:08 iggy carrier pigeon or gtfo
17:09 manfred https://www.ietf.org/rfc/rfc1149.txt
17:09 aparsons joined #salt
17:14 timoguin joined #salt
17:14 Heartsbane so if I want to test a single state file in a directory the syntax is: salt <target> state.sls directory.file
17:15 Heartsbane right?
17:15 drawsmcgraw Heartsbane: Should be, yes
17:15 * Heartsbane wonders why it is giving me the finger.
17:16 mattiasw joined #salt
17:17 hal58th_1 Heartsbane, you could even go a step further and add test=True to the end.
17:18 Heartsbane I did
17:19 Heartsbane Anyway lunch ... it will occur to me after foodz
17:19 ajw0100 joined #salt
17:19 iggy Heartsbane: that would be salt://directory/file.sls
17:19 iggy (not a local file... but there are options to do that too)
17:20 matthew-parlette joined #salt
17:29 murrdoc joined #salt
17:30 7JTAAF5KW joined #salt
17:31 rap424 joined #salt
17:31 seev I also use salt-call state.sls <state> on the minion
17:32 seev sometimes just to trigger dynamic state updates
17:33 Edgan joined #salt
17:33 iggy add -l debug to that and you have one of my primary debugging tools
17:34 drawsmcgraw Trying to use salt-cloud against an OpenStack cluster, when suddenly: http://dpaste.com/08HS1YG
17:35 bhosmer__ joined #salt
17:35 drawsmcgraw I *know* I've seen this before. Apparently I never wrote down how I fixed it :/
17:36 drawsmcgraw And my configs were copied over from a functional setup, so I'm certain it's nothing in the configs...
17:39 zircote joined #salt
17:40 drawsmcgraw Looks like I may have an older version of simplejson installed....
17:40 subsignal joined #salt
17:42 pdayton left #salt
17:42 timoguin drawsmcgraw: yea looks like 3.1.0 is the first version with that exception
17:43 drawsmcgraw timoguin: Yeah I have a CentOS 6.5 box with simplejson 2.0.9 installed
17:43 drawsmcgraw odd, I know I've run salt-cloud on Centos 6.5 no problem in the past....
17:43 shoemonkey joined #salt
17:43 timoguin maybe it got installed with pip before?
17:43 drawsmcgraw So... I guess, pip install simplejson, then yum remove python-simplejson?
17:43 timoguin should do the trick
17:43 drawsmcgraw I do a `yum list` and Yum tells me it's installed
17:43 drawsmcgraw k, I'll try that route. What's the worst that could happen? :D
17:44 _ikke_ Is it possible to run salt commands as non-root? I've added myself to the salt group and chmodded the /etc/salt/master dir to 775, but a command like salt-key first tries to switch user, which it cannot do as normal user
17:44 mccricardo left #salt
17:44 druonysus joined #salt
17:46 drawsmcgraw woo! timoguin, just a 'pip install --upgrade simplejson' was sufficient. I tested out the import in a Python REPL and it seems it pulled in the later version
17:46 bhosmer_ joined #salt
17:47 drawsmcgraw I'd `yum remove python-simplejson` but Yum wants to remove allllll the related openstack packages with it so... I slowly walked away from that.
17:47 drawsmcgraw 'course, still can't launch an instance but... I'll get there
17:48 kunit joined #salt
17:51 _1_rakee joined #salt
17:59 HappySlappy joined #salt
18:00 jeffspeff joined #salt
18:01 OneAngryDBA joined #salt
18:03 timoguin joined #salt
18:03 niseak joined #salt
18:04 krelo joined #salt
18:05 _JZ_ joined #salt
18:06 ek6 going to have to hang periodic table in my cube just so I can parse 'will change in carbon' warning messages and how that relates to 2015.2 and how that relates (or doesnt) to gregorian calendar
18:08 Furao it’s 2015.4 and 2014.7 is still not yet stable
18:08 niseak left #salt
18:09 ek6 yeah...i just havent had the problems with 2014.7 that others did
18:10 Furao salt 1975.13 is finally stable with python 1.3 support
18:11 ek6 would have like the joke more if you had said lisp but still
18:12 iggy ek6: lol, I did just that (with little version numbers next to the ones we already know)
18:12 Furao cobol is more credible for 1975
18:12 ek6 iggy: post that somewhere..save me mindcycles
18:14 ek6 furao:  yeah but i can mentally picture thatch coding lisp I guess
18:15 awerner joined #salt
18:15 Furao Originally specified in 1958, Lisp is the second-oldest high-level programming
18:15 Furao ouch
18:16 Furao only Fortran is older (by one year)
18:16 iggy murrdoc: I'm thinking about switching to salt.utils.dictupdate.merge vs default_settings.formulaname.update(dict)
18:16 murrdoc what does it give you
18:16 murrdoc checking
18:16 OneAngryDBA joined #salt
18:16 iggy i.e. https://github.com/saltstack-formulas/salt-formula/blob/master/salt/map.jinja#L69
18:16 murrdoc no i knew where
18:16 iggy it recursively merges vs a top level overwrite
18:17 murrdoc oh word!
18:17 murrdoc checking in 2014
18:17 iggy I used update in the mongodb-formula and shot my own foot off
18:17 murrdoc is it in all versions
18:17 murrdoc https://github.com/saltstack/salt/blob/v2014.7.2/salt/utils/odict.py
18:17 iggy https://github.com/saltstack/salt/blob/2014.7/salt/utils/dictupdate.py
18:18 murrdoc <3 the documentation
18:18 murrdoc salt.utils.dictupdate.merge vs default_settings.formulaname.update(dict)
18:18 murrdoc i prefer the salt.utils.dictupdate.merge
18:19 iggy on 2014.7 it would probably act pretty similar to the way it does now, on later it would be "smarter"
18:19 murrdoc does salt.utils.dictupdate.merge  work ?
18:19 murrdoc like in jinja ?
18:19 iggy about to try ;)
18:19 wwwBUKOLAYcom joined #salt
18:19 murrdoc or is salt['utils']['dictupdate.merge'] ?
18:20 SpX joined #salt
18:21 iggy bah! hitting that lazyloader bug
18:21 iggy fml
18:21 murrdoc yeah
18:22 iggy well... shoot
18:23 murrdoc there has to be a way to access the salt.utils functions
18:23 murrdoc but then we d need a dict module
18:23 iggy then I'm going to have to do a lot of refactoring to squeeze this all into some sort of defaults/merge format
18:24 murrdoc what are u trying to do
18:24 iggy not overwrite the entire config with update... lol
18:25 murrdoc it only over writes the key it matches
18:25 iggy then I have some other problem
18:25 iggy because the config is not coming out right
18:26 iggy https://github.com/saltstack-formulas/mongodb-formula/blob/master/mongodb/map.jinja
18:26 iggy on line 9/10 I have that actually written wrong
18:26 iggy it should be update(distro_map) update(code_map)
18:27 iggy when you make those changes... it breaks saying something is None
18:27 murrdoc you could, *could* add a _modules dir to the formula
18:27 murrdoc with a lzdict module
18:27 murrdoc that all it does is call salt.utils.dictupdate.merge
18:28 iggy I'm unsure if that would fix my problem now though
18:29 murrdoc cos then u can use salt['lzdict.update'](defaults.mongodb, codemap)
18:29 murrdoc something like that ?
18:29 iggy I'll test it
18:35 primechuck joined #salt
18:37 rem5_ joined #salt
18:39 Vynce joined #salt
18:41 catpig joined #salt
18:41 denys joined #salt
18:41 Tecnico1931 joined #salt
18:42 Gareth Anyone seen a way to force an update of something like __opts__ following an event being fired?
18:42 bhosmer_ joined #salt
18:45 ekristen joined #salt
18:47 iggy there's a wheel module for it... but it's... destructive
18:50 iggy write something to /etc/salt/minion.d/myconfig.conf and reload the minion is probably your best bet
18:50 iggy a little clunky, but that's all I can come up with
18:51 faust joined #salt
18:52 Gareth Don't really want to write the config out though.
18:52 * Gareth looks at wheel
18:53 pepper joined #salt
18:53 pepper hello everybody.
18:56 cpowell_ joined #salt
18:58 pepper I have a doubt on how to launch remote a whole state to a minion.  I have an state with diferents modules prepared and the minion configured. But I don't know the correct syntax to execute the state file exactily or I'm doing anything wrong. Thank in advance. And sorry if I didn't explain me correctly
18:58 murrdoc check out saltstack modules state
19:01 racooper joined #salt
19:01 pepper sorry, how?t do you mean?
19:04 pepper The state file is on the repositoey file. I see the standard system command to launch on the minion as test.ping syntax. But I don't see anything in the documentation on how to launch a full state
19:05 otter768 joined #salt
19:06 pepper http://docs.saltstack.com/en/latest/topics/jobs/index.html#states
19:13 ggoZ joined #salt
19:23 robinsmidsrod joined #salt
19:29 ipmb joined #salt
19:30 hal58th_1 figure it out yet pepper?
19:31 hal58th joined #salt
19:31 conan_the_destro joined #salt
19:31 robinsmidsrod joined #salt
19:35 Not joined #salt
19:42 zwi joined #salt
19:43 shaggy_surfer joined #salt
19:43 ndrei joined #salt
19:44 tomasfej1 joined #salt
19:44 shaggy_surfer Hi all, dealing with an issue where the salt-minion process goes "defunct", it only happens when I run against a particular init script.  My question is, what is the salt-minion expecting back from the script?  Something in the script is hanging it.  It's only on the start of the service not the stop
19:45 pepper joined #salt
19:45 shaggy_surfer sorry should have mentioned I am running the cmd.run on the command line like so:
19:46 iggy there are issues (probably closed) about running salt under upstart/systemd if that's what you're talking about
19:46 pepper sorry I fall. Anybody replys me. please paste it again
19:47 shaggy_surfer hi iggy, I am using ubuntu 14.04.1 with the package version:  salt-minion 2014.7.0 (Helium)
19:47 shaggy_surfer it is using an init script located in /etc/init.d/ but linked to upstart yes
19:48 shaggy_surfer sudo salt -t 60 -v servername cmd.run '/etc/init.d/beanstalkd start'
19:48 shaggy_surfer that's the command I am running, but then the salt minion shows defunct and never returns and keeps logging in the log for the minion w/ new jid's
19:48 iggy sudo salt servername service.start beanstalkd
19:49 shaggy_surfer is that something new?  service.start
19:50 shaggy_surfer instead of cmd.run
19:50 iggy not in the slightest
19:50 racooper that's like, basic functionality for salt, I think it's been there since the beginning
19:50 N-Mi joined #salt
19:51 shaggy_surfer I ask, because I have never used it… I figured it was sufficient to run a cmd.run against what works in the local shell on the box
19:51 iggy it's been there long enough that it doesn't have version annotations... so at least a couple years
19:51 shaggy_surfer so I am doing it wrong?
19:51 iggy yes
19:51 iggy A. that's wrong even on the command line, you should use "service foo start"
19:51 racooper cmd.run should be the backup, if what you want to do isn't provided by an existing module
19:52 iggy B. you should use the modules available... they've all been written to deal with little oddities of running commands without proper environments setup, etc.
19:53 shaggy_surfer I tried that w/ the cmd.run also… sudo salt -t 60 -v servername cmd.run 'service beanstalkd start'
19:53 Sketch iggy: especially if you don't want to have to rewrite your scripts if you upgrade* to systemd
19:53 shaggy_surfer funny thing is the stop/status and others work, just not the start
19:54 markm joined #salt
19:55 shaggy_surfer what about for the services that haven't had a script migrated from the standard /etc/init.d to the upstart style?
19:55 ajw0100 joined #salt
19:57 iggy that's why you use service.start
19:58 iggy it hides all those cross-distro differences from you
19:58 shaggy_surfer perfect.  Thanks so much for the clarification guys.  much appreciated.
19:59 murrdoc is the python zmq versoin 2.2.0 requirement solid
19:59 iggy 3+
19:59 murrdoc is that doc'ed
19:59 thayne joined #salt
19:59 murrdoc got a co worker at work bitching about using ppas
20:01 iggy it's certainly mentioned in a ton of bugs
20:02 iggy ZeroMQ >= 3.2.0
20:02 iggy on the installation page
20:02 iggy http://docs.saltstack.com/en/latest/topics/installation/
20:02 iggy oh, you meant pyzmq
20:02 iggy and yes, it's mentioned on that same page
20:05 iggy sweet, they went 2.2.0.1 -> 13.0.0
20:05 murrdoc sweeeet
20:05 murrdoc they dropped the leading number on all pips
20:06 iggy wonder if that's a year
20:06 tmh_ joined #salt
20:08 shaggy_surfer getting the same defunct behavior when running the command:  sudo salt servername service.start 'beanstalkd'
20:08 shaggy_surfer it's weird, it's only on start too and I am running a -x on the script… it shows the script execute when I do a status/stop  but it doesn't execute when I do a start
20:09 bhosmer_ joined #salt
20:11 shaggy_surfer the logs just shows:  [INFO    ] Executing command 'service beanstalkd start' in directory '/root' and then Starting a new job with PID 9221
20:11 shaggy_surfer keeps looping w/ new pids
20:12 iggy look through the issue tracker
20:12 iggy someone has probably hit something similar before
20:13 mgw joined #salt
20:15 perfectsine joined #salt
20:15 denys joined #salt
20:17 rap424 joined #salt
20:20 speedlight joined #salt
20:20 aquassaut joined #salt
20:21 bhosmer_ joined #salt
20:23 bhosmer__ joined #salt
20:27 bhosmer_ joined #salt
20:30 bmac2 joined #salt
20:34 shaggy_surfer I checked the issue tracker and there are number of issues related… I did however find one #16153 about mysql and it seems that it's most likely the actual init script itself if I had to guess.
20:34 shaggy_surfer since it's the start-up, I just need to toy around w/ the script and find the return that is causing the problem
20:35 shaggy_surfer this is however the first time I have seen this and seems to only be w/ this beanstald packages init script.
20:35 bhosmer joined #salt
20:36 giantlock joined #salt
20:37 redzaku joined #salt
20:40 cpowell joined #salt
20:41 clintberry joined #salt
20:41 pdayton joined #salt
20:41 MatthewsFace joined #salt
20:41 seev I write all my own init scripts
20:42 seev it's the only way to get predictable behavior and sane return codes
20:43 timoguin joined #salt
20:43 nafg_ joined #salt
20:45 shaggy_surfer looks like it definitely has to do w/ return codes in the init script.  That's why I was asking the ? earlier about what the salt-minion is expecting?  At least that way I might be able to find the issue quicker
20:48 iggy ugh, multiple merges of dicts in a map.jinja gets messy quick
20:50 murrdoc yeah
20:51 murrdoc whats the use cases
20:51 bhosmer joined #salt
20:53 clintber_ joined #salt
20:55 JayFK joined #salt
20:57 HappySlappy joined #salt
20:59 hal58th I'm sure Salt or rather the pythong library is expecting a zero exit code shaggy_surfer
21:03 andrew_v joined #salt
21:04 shaggy_surfer yeah, looks like there is a "return 1" in there causing the problem… but the script doesn't behave properly w/o it
21:04 iggy murrdoc: merging defaults, os_family map, oscodename map, pillar foo:lookup, and pillar foo
21:05 markm joined #salt
21:05 murrdoc iggy:  i feel like formulas should dumb down
21:05 murrdoc only have the one os
21:06 murrdoc and tell people to use pillars to override stuff
21:06 iggy hah
21:06 iggy good luck with that
21:06 murrdoc and we should just provide the over ride pillars
21:06 otter768 joined #salt
21:06 iggy you might as well ditch formulas and just hand write everything (ryan would love that btw)
21:07 speedlight joined #salt
21:07 speedlight joined #salt
21:07 murrdoc nah
21:07 murrdoc i write my formulas to support trusty
21:07 murrdoc and over ride for precise
21:08 iggy I write mine for Debian
21:09 nafg joined #salt
21:09 iggy mongodb is kind of strange in that the upstream repos (the suggested means of installing) only provide packages for precise and trusty
21:09 iggy so everything in between is what I actually have to worry about
21:12 p66kumar joined #salt
21:12 pipeep joined #salt
21:15 seev you can fix that by upgrading to CentOS
21:15 murrdoc :D
21:17 iggy Yeah, or fbsd
21:17 iggy then all of the formulas would be equally useless
21:22 zmalone joined #salt
21:23 litwol Hi
21:23 litwol I forgot how to execute a state command through CLI
21:23 litwol without writing sls file
21:23 litwol i want to do salt \* pkg.installed ntp
21:24 murrdoc write the state file
21:24 murrdoc call it with state.sls :D
21:24 ajw0100 joined #salt
21:24 litwol 17:31 <litwol> without writing sls file
21:24 murrdoc i read it
21:24 litwol i know how to write sls files
21:24 iggy don't use state modules on the command line
21:25 timoguin joined #salt
21:26 StDiluted joined #salt
21:28 smcquay joined #salt
21:36 ingwaem joined #salt
21:36 ingwaem` joined #salt
21:36 mattias__ joined #salt
21:41 mattiaswolff joined #salt
21:45 ingwaem`` joined #salt
21:45 Vynce joined #salt
21:45 Vynce1 joined #salt
21:46 bhosmer joined #salt
21:47 adrianhannah left #salt
21:53 jcockhren hmmm
21:54 jcockhren it's looking like netapi isn't properly processing kwargs for the "/run" endpoint.
21:56 Chuck_ joined #salt
21:56 jcockhren for example given a master 2014.7+, the runner, jobs.list_jobs accepts the "ext_source" kwarg ONLY.
21:57 jcockhren as expected, running salt-run jobs.list_jobs something_else='blah' returns: "The following keyword arguments are not valid: {something_else: 'blah'}"
21:58 jcockhren however, when running through the netapi "/run" endpoint all kwargs are ignored
21:58 jcockhren even the excepted kwargs
21:58 jcockhren seems like a bug
22:00 iggy which one?
22:00 jhauser joined #salt
22:01 jcockhren iggy: in reference to netapi.rest_cherrypy
22:01 ndrei joined #salt
22:03 otter768 joined #salt
22:05 tmh_ joined #salt
22:11 echo joined #salt
22:11 yomilk joined #salt
22:12 Karlthane_ joined #salt
22:15 ndrei joined #salt
22:18 HappySlappy joined #salt
22:21 kaictl joined #salt
22:26 hal58th joined #salt
22:27 zonirunner joined #salt
22:27 hal58th_ joined #salt
22:30 amcorreia joined #salt
22:35 otter768 joined #salt
22:39 murrdoc can we call state functions inside custom states ?
22:39 Alexey1 joined #salt
22:40 murrdoc oh this is cool, I can do sate.high to run a state function!
22:40 murrdoc whee
22:41 zmalone left #salt
22:42 iggy murrdoc: https://gist.github.com/iggy/93c65ca26ea6c7eff144
22:42 iggy that was about the cleanest thing I could come up with
22:43 murrdoc is nice
22:43 murrdoc u are over 80 chars in line 17
22:43 iggy so
22:44 iggy I actually had it fixed, but I kept erasing everything and starting over...
22:47 murrdoc joined #salt
22:47 murrdoc what we need
22:48 murrdoc is a 'debug' state
22:48 murrdoc one where you can iterate over the variables and get them out
22:48 murrdoc similar to teh {{ show_full_context }}
22:48 murrdoc but with the ability to pass it a variable
22:50 iggy I'd like something like state.template_str test=True
22:51 murrdoc does that work
22:51 iggy with some -l debug magic thrown in
22:51 iggy to show the rendered output
22:51 murrdoc yeah
22:52 ndrei joined #salt
22:56 druonysuse joined #salt
22:56 druonysuse joined #salt
22:58 otter768 joined #salt
23:02 murrdoc show show_low_sls work iggy ?
23:03 iggy I never learned to read low state
23:03 iggy I usually just satl-call -l debug state.sls foo.bar test=True
23:04 jalbretsen joined #salt
23:06 Vynce what does "foo.bar" represent in that command?
23:07 Auroch joined #salt
23:08 jcockhren iggy: fwiw, It's similar seemingly to: https://github.com/saltstack/salt/issues/19996
23:08 jcockhren but it's not an issue of documentation
23:08 iggy Vynce: directory.slsfile (or /srv/salt/foo/bar.sls if that helps)
23:09 Vynce brtiful, iggy, thanks
23:10 Brew joined #salt
23:10 Vynce (point of note: naming the command "state.sls" where you have to literally type ".sls" to operate on a file named .sls but where you don't type the .sls is… perhaps not helping the learning curve.)
23:11 Vynce (not that this was a decision made by anyone in this room, i know.  just pointing out.)
23:11 iggy it makes more sense if you've been using salt for a while
23:11 iggy that's why they have state.apply in 2015.2
23:12 iggy (it will do a single state file if specified or a full highstate if no args)
23:12 Vynce s/highstate/apply will be a lot more sensical to new users.
23:14 Guest99117 joined #salt
23:14 iggy yeah... now we just need 2015.2 to come out so we can start telling people to use it ;)
23:15 murrdoc use state.apply
23:25 iggy murrdoc: you think that map.jinja I pasted earlier is commit worthy?
23:25 murrdoc i think so
23:26 iggy cool, I'll PR it and see what others think
23:26 mosen joined #salt
23:27 murrdoc its too bad I cant 'watch' an org
23:27 pahko joined #salt
23:27 murrdoc i d love to see PRs coming in
23:27 murrdoc even if  i acant merge them in
23:27 iggy There's only been like 5 in the last couple of days (since I got added)
23:27 iggy and 2 of those have been mine
23:28 iggy grains.filter_by gains a "base" argument in 2015.2
23:28 murrdoc ?
23:28 iggy should make some of this crazy merging a little easier
23:29 murrdoc when is the lazy loader bug getting fixed
23:29 murrdoc so i can use salt.util.dictupdate
23:30 primechuck Can cherrypy/netapi lowstate on a list of minions as a batch or concurrently?  It appears you can passed targets, but it only excecutes jobs one by one
23:32 iggy I think if we extended default.get/config.get (or made a new function like them) that could do all this look up config, pillar, defaults.yaml on it's own we'd save a lot of these headaches
23:32 iggy but then you run into... how much functionality do you bake into it by default
23:35 iggy could always throw it in _modules and not have to wait on upstream release
23:35 murrdoc if u throw it in _modules then anyone not using formulas as gitfs is stuck
23:35 murrdoc but thats a small population
23:35 iggy tough
23:35 iggy ;)
23:35 murrdoc those who use the 'recommended way'
23:35 murrdoc will be fine
23:36 sb_ joined #salt
23:36 iggy there are a few formulas that have _modules/_states already
23:36 murrdoc then do it
23:36 murrdoc make a _module that uses the salt.utils.dict in it
23:36 bfoxwell joined #salt
23:36 murrdoc put it in comments, I HAD TO DO THIS COS U BROKE IT LAZYLOADER
23:36 murrdoc and go from there
23:37 iggy lol
23:37 iggy I'm not even sure if that lookup would work anyway
23:37 iggy I think salt[''] only looks in modules
23:37 murrdoc with merges it will work fine
23:37 murrdoc yeah thast why u have to make the _module
23:38 cheus joined #salt
23:39 peters-tx joined #salt
23:40 ajw0100 joined #salt
23:40 iggy I'll look at it some more tonight
23:41 cromark joined #salt
23:46 evilrob can you get salt-cloud map files to be jinja templates?
23:46 jcockhren bah. figured it out. false alarm
23:47 Vynce is there a straightforward idiom for turning all my minions to test=True by default?
23:47 iggy evilrob: sadly, no
23:47 Vynce (and, i know it's possible for all minions to be test=True by default; is there a good argument against it, other than "if you need it, you have the wrong sysads"?)
23:47 iggy Vynce: put test: True in the config files
23:48 otter768 joined #salt
23:48 evilrob iggy: :(
23:48 iggy Vynce: I imagine it's fairly common in environments with multiple "salters"
23:48 Vynce OK… is there an idiom to do that, or do i have to go edit 40+ config files?
23:49 evilrob we've got multiple envioronments, each with a salt master.  I've got the profiles defined to pass the IP of the salt-master to the deploy script.  I was hoping not to have different profiles for each env.
23:49 cberndt joined #salt
23:50 iggy Vynce: /etc/salt/minion.d/00-test-true.conf\n  file.managed:\n    - contents: test: True
23:50 Vynce (like, is it just salt '*' cmd.run echo 'test=True > /etc/salt/minion.conf' or whatever? or is there a better way to manipulate minion configs en masse?)
23:50 evilrob I'll have to come up with something fancy.
23:50 Vynce ah, OK.
23:50 Vynce sweet.
23:51 iggy shhh, don't tell anyone I used a filename as a state_id
23:51 murrdoc so ghetto right now
23:51 Vynce i've been wondering about that.  it appears to be the expected thing, but it seems gross to me
23:52 murrdoc its gross
23:53 Vynce related to my docco work yesterday (yeah, i gave back, sorta…) — I have been unable to find clear docs for the markup used in the .rst files in salt source, e.g. ":doc:`something </some/path>" … is there a good reference for that?
23:53 iggy it's prone to clashes
23:54 murrdoc explicit id > implicit name
23:54 murrdoc imho
23:54 iggy Vynce: it's restructured text
23:54 murrdoc also its easier to use with listen_in
23:54 nafg joined #salt
23:54 Vynce I guessed it was, but the restructured text documentation i found didn't seem to cover all teh stuff salt uses. but it also looked like old docs.
23:55 Vynce if not, no bigs, i can look again or just file an issue instead of a pull request.
23:56 KyleG joined #salt
23:56 KyleG joined #salt
23:58 iggy maybe there might be some pydoc'isms in there too
23:58 iggy check the docs for readthedocs
23:58 iggy that's where they used to be hosted, so everything that works there, should work...
23:59 Vynce 'k, coo, thx
23:59 rhodgin joined #salt

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