Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-06-04

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

All times shown according to UTC.

Time Nick Message
00:00 manfred usnig pythons socket module
00:01 manfred ldlework: https://github.com/gtmanfred/salt/blob/develop/salt/grains/core.py#L1090
00:01 Joseph ldlework: socket.getfqdn() from python
00:03 nvmme joined #salt
00:05 forrest whiteinge, http://getbootstrap.com/css/#tables-contextual-classes
00:05 forrest something like that
00:05 forrest except for code blocks
00:09 * whiteinge nods
00:12 ldlework I'm just confused because it lists the fqdn and host grains as one thing, but the 'localhost' to the thing inside of /etc/hostname
00:14 rjc joined #salt
00:18 n8n joined #salt
00:19 sunkist_ joined #salt
00:21 sunkist_ joined #salt
00:27 sunkist I am trying to bootstrap my entire lan with as few steps as possible, and I'm using salt to do so.
00:28 sunkist My first step has been to put an unattended install CD into my baremetal host, and get it running vanilla ubuntu.
00:29 sunkist Next, I salt the baremetal host, getting a virtualization layer installed.
00:30 sunkist I don't know if I want to use the baremetal machine as my salt server -- I'd like to use a VM instead.
00:30 ldlework I'm running different salt versions!
00:30 ldlework mann..
00:31 forrest hah
00:31 forrest sunkist, so why not run your salt server on a vm? There is no reason that it has to be a physical host
00:31 forrest that also allows you to dynamically scale based on your needs without going to the DC.
00:31 sunkist forrest, I think that's a good idea -- but it's a chicken and egg issue.
00:32 forrest how so?
00:32 sunkist I need salt to configure the virtualization host, I need the virtualization host to host the salt sever.
00:32 Dester__2 left #salt
00:32 sunkist s/sever/server
00:32 forrest so install the salt minion, then use that to configure the master portion, and have all minions point to the vm master
00:33 ldlework vm?
00:33 ldlework you mean docker container.
00:34 sunkist forrest, should I use a single set of states for that, or break it out into a bootstrap set and a production set?
00:34 sunkist ldlework, docker is okay... if you're into that sort of thing.
00:35 ldlework sunkist: yes, I am definitely into getting the same thing far far cheaper.
00:35 forrest sunkist, I'd probably just write my states as I would normally (very modular), and have that be one of them
00:35 garthk joined #salt
00:36 to_json joined #salt
00:37 otter768 joined #salt
00:41 falican joined #salt
00:42 elfixit joined #salt
00:43 ipalreadytaken joined #salt
00:46 alunduil joined #salt
00:48 falican I've got an issue with adding users to a group. I have some users controlled by user.present and a group controlled by group.present with a members setting listing the users I want in it. Example: http://pastebin.com/h0yR3Req
00:48 falican The group isn't being filled with the members I am specifying
00:49 sunkist ldlework, how would you recommend going about learning docker?
00:53 taion809 joined #salt
00:58 nvmme joined #salt
01:00 Joseph falican: do you ahve an error?
01:10 Shenril joined #salt
01:12 falican Joseph, no errors
01:13 jalbretsen joined #salt
01:14 falican just spits out http://pastebin.com/8T3qZr3G
01:15 Joseph falican: and what result were you expecting?
01:15 falican User rory_hart being a member of group sudo
01:15 falican which it isn't
01:16 Joseph but that result is saying the group is present. I don't think its commenting on whether rory_hart is part of the group or not
01:17 Joseph falican: if you want a user to be part of a group, then that should be handled in the user.present block
01:17 falican Agreed. but if I logged into the box and run groups rory_hart it shows only the rory_hart group.
01:17 Joseph would you first need to create the group before you can add the user into it
01:17 Joseph ?
01:17 Joseph would you not
01:18 falican either the group needs to exist or the user needs to exist depending on which state is managing the group membership
01:19 falican Joseph, the parameters of group.present has members, addusers, delusers implies you can manage group membership through this state.
01:20 Joseph http://salt.readthedocs.org/en/v2014.1.1/ref/states/all/salt.states.group.html?highlight=group#module-salt.states.group
01:20 Joseph i am looking at this...what documentation are you referring to?
01:20 falican If you cannot then I will work around it but it will be annoying because there is a number of places that specify user groups
01:20 falican http://docs.saltstack.com/en/latest/ref/states/all/salt.states.user.html
01:20 falican oop
01:21 falican http://docs.saltstack.com/en/latest/ref/states/all/salt.states.group.html this one
01:21 Joseph what version of saltstack are you on
01:32 falican err
01:32 falican Joseph, 2014.1.4-2precise1
01:33 Joseph I think the documentation you are looking at is from the develop branch
01:33 Joseph its possible this rev of the state module does not exist in your version
01:34 Joseph here is the doc for the groups in the release you are on http://salt.readthedocs.org/en/v2014.1.4/ref/states/all/salt.states.group.html?highlight=group#module-salt.states.group
01:35 falican ah kk thanks
01:35 falican *sigh*
01:35 Joseph but that said...doing it with the existing functionality shouldn't be that bad
01:35 Joseph you just need to handle the requisites a bit careful i would think
01:38 oz_akan__ joined #salt
01:38 falican yeah just need to do a bunch of loops through each list of groups to check if a user is a member
01:38 falican not as neat as I would like
01:39 Joseph really? Why's that? Why wouldn't you just ensure the group is present and then make sure the user.present always tries to add the user to the group. Its not going to hurt anything if you try to add someone to a group they are already a part of.
01:44 falican I mean that the pillar data for what groups a user belongs to comes from a few different places so building the group list for a user means looping over data in a few different places.
01:44 Joseph falican: have you looked at overstate?
01:44 Joseph falican: you can set the pillar data with one "run" and then iterate through it doing another run
01:45 falican Joseph, no I haven't. Thanks will take a look :)
01:45 Joseph http://docs.saltstack.com/en/latest/topics/tutorials/states_pt5.html
01:46 Furao joined #salt
01:59 littleidea joined #salt
02:07 jalbretsen1 joined #salt
02:10 arapaho joined #salt
02:13 to_json joined #salt
02:23 mapu joined #salt
02:36 jalbretsen joined #salt
02:36 bhosmer joined #salt
02:37 thayne joined #salt
02:44 bhosmer joined #salt
03:00 littleidea joined #salt
03:13 acabrera joined #salt
03:16 catpigger joined #salt
03:36 jalaziz joined #salt
03:58 ckao joined #salt
03:59 andredie_ joined #salt
04:06 l0x3py joined #salt
04:24 stevednd manfred: yeah, using the named replacement stuff in yaml has saved me a lot of repetition in other places where I have had to use it
04:24 jrcresawn joined #salt
04:24 stevednd whiteinge: yes, got it working successfully. I updated the gist https://gist.github.com/dnd/72f2da2354d8abe39636
04:25 bhosmer joined #salt
04:26 acabrera joined #salt
04:28 ipalreadytaken joined #salt
04:29 greyhatpython joined #salt
04:29 greyhatpython joined #salt
04:34 scoates joined #salt
04:41 oz_akan_ joined #salt
04:42 saravanans joined #salt
04:45 stanchan joined #salt
04:46 felskrone joined #salt
04:50 scoates joined #salt
04:52 thayne joined #salt
04:58 malinoff joined #salt
04:59 scoates joined #salt
05:02 rjc joined #salt
05:14 ajw0100 joined #salt
05:30 ldlework I have 4 serves. On two of them I have the version, salt 2014.1.0-3766-g30b7764 and on the others, salt-call 2014.1.0-6927-g5533838 (Hydrogen)
05:30 ldlework I cannot seem to upgrade the first two to the version the others are running
05:30 * ldlework is scratching his head
05:35 dane joined #salt
05:42 __number5__ ldlework: how do you upgrade the minions? salt-bootstrap?
05:42 oz_akan_ joined #salt
05:42 ldlework I tried running bootstrap, and I also tried upgrading with apt
05:50 ldlework __number5__: ^
05:50 __number5__ apt probably won't work, none of those versions are in ubuntu ppa
05:53 n8n joined #salt
05:53 ldlework __number5__: why doesn't running bootstrap update the systems directly to the latest version
05:55 shaggy_surfer joined #salt
05:56 __number5__ ldlework: are u using `sudo sh install_salt.sh git develop`?
05:56 __number5__ without `git develop` it will try to install latest release
06:08 dane I'm having problems running docker containers
06:09 dane Master and slave are on centos 6.5 with salt version 2014.1.4
06:10 dane I have a simple state:
06:10 dane http://pastebin.com/Rwicw890
06:10 CeBe joined #salt
06:12 dane It runs, but none of the ports are forwarded and i have this strange error constantly
06:12 dane http://pastebin.com/6gT9zzNL
06:12 dane ValueError: zero length field name in format
06:13 q4brk joined #salt
06:14 bhosmer joined #salt
06:19 Ivo joined #salt
06:20 __number5__ dane: python2.6 don't support '{!r}' format string. you might want to upgrade to python2.7 first
06:21 bhosmer joined #salt
06:22 emning left #salt
06:23 dane do people generally use python 2.7? I'm new to salt
06:23 jcockhren yeah
06:24 dane Looks like ill have to compile it... any salt state to make this easier?
06:25 dane Do I upgrade the master or slaves, or both?
06:26 jcockhren both
06:28 dane Will that mean I have to reinstall salt also?
06:28 dane I just used the bootstrap method
06:28 __number5__ just bootstrap again :)
06:30 dane How would you use salt cloud to bring up a new instance? Run a python update and install the minon via a state?
06:38 dane I guess I would have to make an AMI with with it installed ready
06:39 Husio once more with my git pull problem, maybe someone will have the answer today:
06:39 Husio I have problem with managing git repository through salt: I have two machines, both doing the same job: fetching the branch production, and since few weeks, one of the machines does not pull latest branch repo anymore
06:40 Husio both repos are clean, I can do git pull by hand and it works perfectly, but for some reason, when done by salt, it's getting old version on one of the machines
06:42 shaggy_surfer joined #salt
06:43 oz_akan_ joined #salt
06:46 hardwire joined #salt
06:47 MrTango joined #salt
06:50 q4brk joined #salt
06:50 flebel joined #salt
06:50 __number5__ dane, I found building AMI via packer and masterless salt provision works better for me
06:53 dane you dont happen to use centos and could provide the states by any chance?
07:02 __number5__ dane: sorry, I only use ubuntu
07:03 dane ubuntu comes with python 2.7 though? why do you need packer?
07:03 dane On the saltstack website it says Python 2.6 or 2.7 are the requirements. Are some states 2.7 dependant?
07:04 __number5__ dane: but http://www.packer.io/docs/provisioners/salt-masterless.html will install salt into any Linux OS it supports
07:05 __number5__ dane, yes I guess some newer states like docker is not strictly checked, you can file a github issue for that
07:05 dane I was just wondering what sort of things you would add to the image instead of just using a standard ubuntu image
07:06 __number5__ dane, I use packer mainly because I want to cut the provision time shorter by putting most packages into the AMI
07:07 __number5__ for example, a newer version of ruby and setup postgres db etc.
07:07 dane so all your common states are already there at the start
07:08 __number5__ yes
07:08 dane ok, that makes sense
07:09 dane I'll try python 2.7 and if it fixes things I'll file a bug. I'd prefer to use standard images if possible
07:09 ml_1 joined #salt
07:11 shaggy_surfer joined #salt
07:14 linjan joined #salt
07:23 agronholm centos STILL doesn't provide python 2.7??
07:26 chiui joined #salt
07:26 malinoff agronholm, http://dev.centos.org/centos/6/SCL/docs/#tabl-RHSCL-Components
07:27 malinoff agronholm, It's not provided in the standard repos
07:27 krow joined #salt
07:36 slav0nic joined #salt
07:37 zlh joined #salt
07:38 nvmme joined #salt
07:38 zlh how can i get minion real-time feedback?
07:40 malinoff zlh, events
07:41 zlh thank you :-)
07:43 oz_akan_ joined #salt
07:46 toddnni joined #salt
07:49 ghartz joined #salt
07:51 alanpearce joined #salt
07:59 bbnix joined #salt
08:01 xmj moin
08:01 xmj i'm wondering
08:01 xmj did anyone get the salt bind state to work with timelib?
08:02 xmj i'm seeing something along the lines of https://dpaste.de/KsGm/raw
08:03 bhosmer joined #salt
08:11 younqcass joined #salt
08:17 jdmf joined #salt
08:19 darkelda joined #salt
08:20 jeffrey4l joined #salt
08:24 druonysus joined #salt
08:26 xmj more specifically
08:26 xmj how do i create and import custom filters with salt's jinja templates?
08:31 TamCore Is there a built-in function to get the installed version of a specific package?
08:37 thayne joined #salt
08:43 jeffrey4l joined #salt
08:43 davidnknight joined #salt
08:44 jeffrey4l Is anybody know when the Helium will be released?
08:45 jgelens I have the following match statement:   'not G@virtual_:ec2 and G@virtual_subtype:Xen PV DomU':
08:45 jgelens however virtual_subtype doesn't match, because "Xen PV DomU" has spaces.
08:45 jgelens adding quotes doesn't help
08:45 jgelens any idea?
08:48 sinh joined #salt
08:48 perfectsine joined #salt
08:48 slav0nic joined #salt
08:48 slav0nic joined #salt
08:48 joehh joined #salt
08:48 dccc joined #salt
08:49 ghartz joined #salt
08:49 jnials joined #salt
08:49 l0x3py joined #salt
08:50 jeffrey4l joined #salt
08:51 younqcass joined #salt
08:52 pkimber joined #salt
08:54 kedo39 joined #salt
08:56 jpaetzel joined #salt
08:59 agronholm joined #salt
09:02 davidnknight joined #salt
09:03 bstromski joined #salt
09:04 notbmatt joined #salt
09:04 bstromski hey guys, does salt allow regex in an if statement inside an sls module?
09:05 agronholm bstromski: salt doesn't have if statements, does it?
09:06 bstromski ex.:  {% if grains['os'] == 'RedHat' %}
09:06 agronholm that comes from your templating engine of choice
09:06 agronholm in this case, jinja2
09:06 agronholm so consult its documentation
09:06 jeddi joined #salt
09:07 bstromski would i be able to do something like {% grains['nodename'] == '$1' %}
09:07 alanpear_ joined #salt
09:07 bstromski ahh ok
09:08 agronholm but salt does allow regex matching based on grains
09:08 agronholm iirc
09:08 bstromski agronholm : excellent, ill do a bit more digging around ; just wanted to make sure i want wasting my time :)
09:08 bstromski thanks!
09:09 georgemarshall joined #salt
09:11 the_lalelu joined #salt
09:13 jeddi I'm after an approach, ideally a best practice, for dealing with this hard-to-describe problem.  I want my apache2 service to restart when certain files (say anything under /etc/apache2) are modified.  This is easy enough.  On some hosts/minions, however, I'm installing wordpress (by including app.wordpress in the host.hostname.sls).  Iff a minion has wordpress installed, I then want mya pache service to be watching files under, say, /
09:13 jeddi var/lib/wordpress for changes *as well*, and restarting the apache2 service.
09:14 Eugene joined #salt
09:17 ggoZ joined #salt
09:22 faulkner joined #salt
09:25 mwalker joined #salt
09:26 mwalker does anyone have any guidance for troubleshooting a returner issue? on the client I'm running with -l debug, but all the log is telling me is "Returning information for job..." immediately followed by a "The return failed for job". Is there a way to get a more verbose logging mode?
09:27 N-Mi joined #salt
09:27 N-Mi joined #salt
09:28 bbnix joined #salt
09:30 malinoff joined #salt
09:33 druonysus joined #salt
09:33 druonysus joined #salt
09:38 Shenril joined #salt
09:42 mwalker ah; found my problem; I had corrupted state; refreshed it by calling `salt '*' state.clear_cache` fixed things
09:42 workingcats joined #salt
09:45 oz_akan_ joined #salt
09:47 oz_akan__ joined #salt
09:48 krow joined #salt
09:52 bhosmer joined #salt
09:52 berto- joined #salt
09:53 chamunks joined #salt
09:54 georgemarshall joined #salt
09:54 [vaelen] joined #salt
09:54 chutzpah joined #salt
09:56 pdayton joined #salt
10:01 alekibango_ joined #salt
10:08 Jahkeup joined #salt
10:15 bhosmer joined #salt
10:17 alanpearce joined #salt
10:21 orbit_darren joined #salt
10:24 hvn joined #salt
10:25 hvn hi all, does anyone know how to pass None to a salt function throught CLI? I'm passing to file.get_managed but it reconise it as a string 'None'
10:31 chiui joined #salt
10:33 ndrei joined #salt
10:36 jY hvn: show the command you are trying to run
10:37 hvn salt-call --local file.get_managed /tmp/hvntest jinja salt://wordpress/nrp_type: md5, hsum: a8243103a45f9d23cfaa238e198f1c21}' root root '755' base None None
10:37 alanpearce joined #salt
10:38 oz_akan_ joined #salt
10:39 nvmme joined #salt
10:50 andrej joined #salt
10:57 younqcass joined #salt
10:58 younqcass left #salt
11:16 thoughtjacked joined #salt
11:22 diegows joined #salt
11:26 bhosmer_ joined #salt
11:27 thoughtjacked joined #salt
11:29 tristianc|Alt joined #salt
11:30 joehh hvn: maybe try the suggestion at http://docs.saltstack.com/en/latest/topics/windows/windows-specific-behaviour.html#specifying-the-none-group
11:31 malinoff joined #salt
11:35 giantlock joined #salt
11:39 oz_akan_ joined #salt
11:41 saru11 joined #salt
11:42 tristianc|Mobile joined #salt
11:43 tristianc|Mobile left #salt
11:44 alfborge joined #salt
11:45 homelinen joined #salt
11:45 alfborge Hi guys, I'm planning on writing an artifactory file server backend.  Does anyone know if something like this already exists?
11:46 Furao alfborge: no but i got a client who will help you on that
11:46 Lomithrani joined #salt
11:46 alfborge My requirement is basically that we have binary artifacts (mostly java war files) that should be deployed to our minions.  We don't want our minions to access artifactory directly.
11:47 alfborge Furao: As in they also want something similar or as in they have something similar?
11:47 Furao alfborge: exact same requirements
11:47 Furao they want the same
11:47 Furao so far they used user/pass + http request
11:48 jaycedars joined #salt
11:48 Furao we can contribute and test
11:49 alfborge Do you agree that a file server backend is the way to go on this?
11:49 Furao i got like 8 python devs
11:49 saru11 hello guys, I would have a question regarding troubleshooting of custom execution module, when I run it directly, it works but when I want to call it from a state it throws an exception... please check the details here: http://pastebin.com/Y9ecLpYP
11:49 Furao i think it's the most "natural" way to do it in term of salt architecture
11:49 Furao BUT https is a lot faster than salt://
11:49 Furao and war are big sometimes
11:50 saru11 how can I troubleshoot that? is there a way to check that exception?!? I turned on verbose debugging on the master but there's nothing additional in the log file :-(
11:50 alfborge Yes, it might be that the zeromq channel isn't ideal for this...
11:50 Furao but that way you can have source: artifactory://bleh
11:50 Furao without know checksum first
11:51 xmj so
11:51 Furao source: artifactory://bleh-{{ pillar['version'] }}
11:51 xmj now that there's a bit more activity here.
11:51 xmj does anyone use salt's bind-formula?
11:51 alfborge Furao: Yes, it does simplify things from the minions side
11:51 Furao those formulas should be considered as example and to learn
11:52 xmj Furao: out of curiosity (we also use artifactory for stuff):
11:52 Furao let's create a github project and hack that :P
11:53 xmj why doesn't your client just package the jars with fpm into RPMs, set up a RPM repo and let pkg.installed: with that custom repo do the rest?
11:53 xmj i'm sure artifactory would be interesting- just not *as* easy&feasible
11:53 Furao we have this https://archive.robotinfra.com/
11:53 Furao we mirror almost everything
11:53 Furao even salt itself
11:53 alfborge Furao: Might be we should create an issue on this first and spec it out a bit.
11:54 alfborge xmj: artifactory is actually a custom repo already
11:54 xmj gotcha
11:54 Furao I wrote a testing framework for salt formula, and when it runs for 19 hours (on SSD) it download so much files that when pypi or github is down test failed
11:54 xmj Furao: about bind- the formula seems broken, so i'm trying to find someone who has (successfully) used it
11:54 mschiff How can I run a command *before* a change will be made using file.replace?
11:55 xmj mschiff: write a cmd.run and use it through require_in ?
11:57 mschiff xmj: ah! seems what I was looking for thx ;)
11:57 mschiff ^like
11:57 alfborge xmj: One of the things that makes us want to manage these files through the salt fileserver is that some of our minions are behind firewalls we don't control.  We're not responsible for maintaining the OS (some minions are windows).
11:57 xmj wwwwwah that sounds horrible :)
11:58 xmj does that client pay well?
11:58 alfborge :)
11:58 alfborge They don't pay us to maintain their os, they pay us to maintain our application. :)
11:59 alfborge We use salt to manage Continous Delivery
11:59 xmj i charge extra for mental torture :)
12:01 jaimed joined #salt
12:04 zoresvit joined #salt
12:05 zoresvit Hi. When I use salt --out yaml ... I always get som extra debugging strings in the output that are extremely painful to parse. Even if I use -l quiet. Is there any way to avoid this output in salt CLI?
12:06 zoresvit Like this: salt cli avoid errorsNo minions matched the target. No command was sent, no jid was assigned. {}
12:06 zoresvit Noone needs explanation, just leave the empty dict :)
12:11 zoresvit Anyone? :)
12:16 alfborge Furao: I'm going to start looking at a simple implementation of the fileserver now.  Just a quick question...  Do you guys need an artifactory fileserver or a maven fileserver?
12:16 alfborge Maybe we could even get away with a simple http proxy fileserver.
12:16 giantlock joined #salt
12:17 alfborge The latter would be way more general but it wouldn't handle maven concepts like LATEST and CURRENT etc.
12:18 alfborge Thus these would have to be handled in a different step.
12:18 Furao I don't know about maven
12:18 alfborge A maven fileserver would handle these concepts (and you could then use maven coordinates).
12:18 Furao i know they used it but i don't think it's in the formulas, I need to look
12:19 alfborge Artifactory has a private API that could also be used (but that requires an artifactory licence).
12:20 alfborge That API gives a few nice things with regards to build management and release management through artifactory, however we don't currently use that.  Do you guys?
12:21 alanpearce joined #salt
12:22 Furao neither us
12:22 Furao no way we will buy an artifactory license, not that we're cheap
12:23 Furao but i don't want to be lock to a "license" in any way
12:23 alfborge :)
12:24 ksk joined #salt
12:25 alfborge So, I'm going to flesh out a simple spec for myself and start hacking.  Did you have an opinion on whether to assume maven or make it more general?
12:27 Lomithrani joined #salt
12:28 alfborge I'm going for the maven one for now.  Creating a general http one will be trivial later.
12:29 alfborge I'm aiming for something like this: salt://maven/repo/groupId/artifact-{{ pillar['version'] }}.war
12:29 Outlander joined #salt
12:29 alfborge Or maybe it would just be salt://repoId/groupId/artifactId-{{ pillar['version'] }}.war
12:31 alfborge Where repoId would be configured in the master config file.
12:31 Furao using salt:// prefix ?
12:32 Ivo joined #salt
12:36 ndrei joined #salt
12:36 mschiff xmj: do you have an example somewhere? cmd.run is run alyways so I assume I have to use cmd.wait ... but with "cmd.wait" the command is never run...
12:37 alfborge If I create a file server backend and add it to my fileserver_backends configuration, then the files from that backend would be available under salt://.  Or am I missing something?
12:38 alfborge If we want to use maven:// or some other scheme then I suspect the changes are bigger.
12:39 alfborge But I'm not intimate with the source code, so if anyone knows please speak up. :)
12:39 saimn joined #salt
12:40 jslatts joined #salt
12:40 Furao yes you're right
12:41 Furao to had contributed to gitfs, i'm not sure I would do it that way. but it might be simplier
12:41 Furao yes
12:42 ndrei joined #salt
12:49 xmj mschiff: do you want to only run it once?
12:49 gmoro joined #salt
12:50 mschiff xmj: yes, only once and only if the file is going to be changed
12:50 xmj ah hm.
12:50 mschiff or anytime, but only if the file going to be changed
12:50 xmj i have no clue
12:50 xmj let me know if you find out :)
12:52 jrdx joined #salt
12:54 mschiff xmj: ok ;)
12:54 Outlander_ joined #salt
12:55 giannello joined #salt
12:59 ndrei joined #salt
13:00 ccase joined #salt
13:02 * viq whines seeing how much work is to be done on the gitlab formula
13:03 superted666 joined #salt
13:03 bhosmer joined #salt
13:04 superted666 Hopefully simple questions guys... how do i add a variable to a command using salt? I'm keen to run 'command x' and output that to a local file as part of my module, trouble is i want the filename to inlcude the hostname of the client it's run on
13:04 bhosmer_ joined #salt
13:04 viq For one make sure it works with and depends on other formulas: git, nginx, mysql, postgres, redis
13:04 viq superted666: {{ grains['id'] }} (or fqdn instead of id)
13:05 superted666 oooh, will give it a shot now thanks
13:05 Ahlee by module, do you mean something you've written and dropped in _modules?
13:06 Ahlee or what viq said
13:06 superted666 A module we've written Ahlee
13:06 superted666 Just trying out the new syntax now
13:07 viq superted666: that's not exactly new, that's the jinja magic
13:07 racooper joined #salt
13:07 viq There are other renderers available too
13:07 _mel_ joined #salt
13:07 viq You certainly could use that in an sls file... I know nothing about modules.
13:07 Ahlee well, you have the entire python environment at your disposal then, so you could always import socket; socket.getfqdn()
13:07 viq And python for that matter, so we're clear :P
13:08 viq Ahlee: I think salt exports... dunno, was it __salt or something like that, where you can access the stuff salt knows about
13:08 Ahlee Yeah, you get the dunders
13:09 Ahlee *shrug* both'll work
13:09 Ahlee we moved away from _modules due to minions losing them constantly
13:10 viq oh?
13:10 viq How did they lose them, and any idea why?
13:10 Ahlee Nope.  They exist in the cache directory on the minion, but calling module.foo returns not found
13:11 viq huh
13:14 mgw joined #salt
13:21 resmike joined #salt
13:22 darkelda joined #salt
13:23 mgw joined #salt
13:24 oz_akan_ joined #salt
13:25 alfborge left #salt
13:28 jaycedars joined #salt
13:28 mattmtl joined #salt
13:32 acabrera joined #salt
13:34 netzmonster joined #salt
13:34 Lomithrani Hi guys , I encounter this error "Failed to authenticate, is this user permitted to execute commands?"  while trying to salt '*' test.ping
13:35 Lomithrani any idea what may have happened ? (worked like a charm yesterday)
13:35 scoates joined #salt
13:35 Lomithrani I'm was both yesterday and today logged as root
13:35 alanpearce joined #salt
13:36 mgw joined #salt
13:36 dude051 joined #salt
13:39 Lomithrani ok it worked once , right know and then the minute after same error ...
13:42 viq ok, WTF? http://pbot.rmdir.de/kJ2A3f_bFttGyZSTdYdv6A
13:43 viq Any ideas why salt doesn't see the formulas?
13:44 to_json joined #salt
13:44 N-Mi joined #salt
13:45 viq erm, yeah, doesn't matter whether I try to call it via gitlab or gitlab-formula.gitlab
13:47 alunduil joined #salt
13:51 Xiao joined #salt
13:53 worstadm1n joined #salt
13:53 worstadm1n Can I call group.add from my init.sls?
13:53 vejdmn joined #salt
13:54 quickdry21 joined #salt
13:55 viq no, you can call group.present
13:56 worstadm1n viq: And what if I wanted to do something like group.members wheel 'myself'
13:57 worstadm1n Or Im sorry
13:57 worstadm1n rather Id like to remove a group
13:58 worstadm1n Ive already seen the -members option
13:58 viq worstadm1n: have you looked at http://docs.saltstack.com/en/latest/ref/states/all/salt.states.group.html
13:58 worstadm1n ah thanks
14:02 Ahlee windows minion, UnpackValueError: Data is not enough.
14:04 tyler-baker joined #salt
14:12 robinsmidsrod joined #salt
14:13 GradysGhost joined #salt
14:14 BugeyeD joined #salt
14:16 fusionx86 joined #salt
14:16 worstadm1n I cant figure out how to set purge for a user in my init.sls
14:16 funzo joined #salt
14:16 Lomithrani is there a way to monitor whats going on while doing the state.highstate ?
14:17 Lomithrani I just launched my minion with salt-minion debug & , and it doesnt seem to do anything since the key got accepted
14:22 chiui joined #salt
14:23 ajprog_laptop joined #salt
14:23 wendall911 joined #salt
14:26 alunduil joined #salt
14:26 N-Mi joined #salt
14:26 N-Mi joined #salt
14:27 Ahlee Lomithrani: target the host from the master and run highstate, debug will logg some data
14:27 Ahlee you can go finer detail with salt-minion -l trace
14:27 Lomithrani the job is a highstate acuatlly
14:27 Ahlee but -l debug will tell you it's doing items
14:28 Lomithrani its been running for like 15 minutes
14:28 Lomithrani so I'm wondering whats going on because installing jre and jdk shouldnt be taht long
14:28 Ahlee hop on the master and lookup the jid with salt-run jobs.lookup_jid <jid>
14:28 Lomithrani nothing with that
14:28 mgw This has been discussed before, and I forget the conclusion. I'm working on a runner that can take variable number of kwargs, but I get dunder vars in my kwargs along with whatever is passed from the runner cli. Is that expected behavior? If so, is the normal workaround simply to filter out those vars starting with '__'?
14:28 worstadm1n Does matter what I do group.present - members: -user
14:29 worstadm1n Doesnt add users to the group
14:29 Lomithrani job is still in jobs.active
14:29 mgw basepi: I think you may have been involved in that discussion
14:29 Ahlee Then kill the job and start it again
14:29 GradysGhost joined #salt
14:29 Ahlee if the minion was started with debug before you ran highstate, if it was doing any work it'd log the work
14:30 Ahlee if it's not, it's not doing anything and the job is likely stuck
14:30 Lomithrani mm it didnt show anything when I pinged it though
14:30 Lomithrani but the pinged work
14:31 Lomithrani I think my minion is a bit broken
14:31 Ahlee do you have two minions running
14:31 Ahlee and the one without logging is servicing the requests
14:31 jalbretsen joined #salt
14:31 Lomithrani I only target one minion
14:31 Lomithrani sudo salt App02 cmd.run "echo test" is doing fine
14:32 Lomithrani I get "App02:  test"
14:32 Lomithrani but nothing shows up on my minion monitor
14:32 Ahlee I meant on the host.
14:32 Ahlee on your app02, do you have more than one minion running
14:32 Lomithrani no
14:32 Lomithrani one zone per minion
14:32 Lomithrani (I'm on smartos)
14:33 Ahlee i don't know what that is
14:33 Ahlee but, you have a minion_id as App02
14:33 Ahlee so App02's minion is currently running in the foreground with salt-minion -l debug
14:33 Lomithrani well I minion per IP / VM
14:33 Ahlee and from your master, when you salt App02 test.ping
14:33 Ahlee you're not getting info logged to your foregrounded minion
14:34 Ahlee if not, the minion that's responding to App02 isn't the foregrounded process, and you have a key error
14:34 Ahlee well, not key error, but confusion in the keys
14:34 worstadm1n http://pastebin.com/aUZYVgE9
14:34 worstadm1n Has anyone run across this?
14:34 Ahlee I would shut the minion down, salt-key -y -d App02 on your master, rm -rf /etc/salt/pki/ on App02
14:34 Lomithrani ok I rerunned minion and this time minion appears to be running foreground
14:34 Ahlee start the salt minion again, accept hte key if not open mode, and try again
14:35 Lomithrani I'll try again with my highstate
14:35 gothix_ is there a way to have two base trees i need one file_roots for amazon and one for in house but they are going to bve similar files in them
14:36 thayne joined #salt
14:36 micah_chatt joined #salt
14:36 gothix joined #salt
14:36 gothix joined #salt
14:37 Ahlee gothix: http://docs.saltstack.com/en/latest/ref/file_server/file_roots.html#directory-overlay
14:38 CheKolyN joined #salt
14:38 gothix so it will use directory overlay
14:39 worstadm1n Any suggestions on creating this group with members?
14:39 worstadm1n Salt is not complaining about my state file, however there are no members in the group
14:42 Lomithrani Ahlee: thanks now at least I have a feedback on my state.highstate. Which drive me to my second question : did anyone encountered a problem with pkg.installed with sun-jre6 ?
14:42 worstadm1n I cant find any documentation anywhere on why this isnt working
14:44 thedodd joined #salt
14:44 AdamSewell joined #salt
14:44 AdamSewell joined #salt
14:45 Lomithrani worstadm1n: never worked on it but are you sure you arent supposed to use the gid and system aswell ?
14:45 Lomithrani http://docs.saltstack.com/en/latest/ref/states/all/salt.states.group.html
14:45 worstadm1n Lomithrani: Would it not throw an error if I didnt?
14:46 worstadm1n Even when added it still fails to add members
14:46 worstadm1n No logs in /var/log/salt/minion
14:46 worstadm1n or rather relevant log entries
14:46 worstadm1n just a successful run
14:49 xmj does anyone here use salt to deploy BIND servers?
14:52 jslatts joined #salt
14:53 kaptk2 joined #salt
14:54 conan_the_destro joined #salt
14:55 ajolo joined #salt
14:58 philip joined #salt
14:59 sunkist joined #salt
14:59 Ahlee Lomithrani: does sun-jre require acknowledging a license agreement?
14:59 Lomithrani mmm probably
14:59 mattmtl joined #salt
14:59 Lomithrani is there a way to acknowledge it automatically ?
15:00 Ahlee install from outside your minion, all salt is doing is using your OS's package management to run the equivalent install (yum for rhel, apt for debian, zypper for suse)
15:00 Ahlee dunno.
15:00 Ahlee if there's a quiet method
15:00 Ahlee wrap cmd.run
15:02 Lomithrani yeah trying a cmd.run "pkgin in sun-jre6" atm , seemed just a lot less elegant that pkg.installed
15:02 Ahlee functional > elegant
15:02 alanpearce joined #salt
15:02 Ahlee gotta figure out why it's not working before making it elegant
15:03 Ahlee think i found a regression in 0.17.5 on https://github.com/saltstack/salt/issues/6601
15:04 younqcass joined #salt
15:05 stevednd whiteinge: I further updated the gist to show how to include hash blocks since they require a double indent
15:05 malinoff joined #salt
15:07 Ahlee huh. Removing both salt and salt-minion cleared it up
15:07 Ahlee lovelu
15:13 jeffrey4l joined #salt
15:13 JasonSwindle joined #salt
15:13 picker joined #salt
15:15 picker hi guys,  salt.master    ][ERROR   ] An extra return was detected from minion WIN-IAHETR4G4TT, please verify the minion, this could be a replay attack
15:16 picker Seems to be due to same minion id on instances
15:16 viq minion ids are supposed to be unique
15:16 picker Is there any way we can map it directly to ip address?
15:17 viq salt 'something' network.ip_addrs
15:17 viq might work
15:20 picker I guess , have to set it "id" value in minion, and network.ip_addrs will work over there?
15:24 picker viq: I was right it won't work in minion
15:25 viq picker: I'm not sure what you mean
15:27 picker viq: Was trying to change the "id" in minion config with network.ipaddrs
15:28 viq picker: id is in /etc/salt/minion_id
15:30 ndrei joined #salt
15:31 picker viq:Gotcha, thanks. Deleting it solved the problem
15:31 netzmonster joined #salt
15:34 jcsp joined #salt
15:37 conan_the_destro joined #salt
15:38 basepi mgw: I don't recall that discussion.  Seems like a bug to me, but again, I haven't been involved.
15:38 philip Hey guys, I've got a local vagrant instance with a masterless minion but it's trying to get ec2 metadata via 169.254.169.254 and just hanging since it cant resolve outside ec2, any way to turn that check off entirely?
15:43 timoguin philip: do you have custom EC2 grains?
15:44 philip well I have them in a "deployed" environment, and my masterless minion config doesn't have that env as part of it's file roots. Would it still "find" the _grains folder  though?
15:44 mateoconfeugo joined #salt
15:46 worstadm1n I'm authenticating users using ldap and trying to set them as members in my salt state file (given they dont exist according to /etc/passwd) - will I run into problems?
15:46 timoguin philip: i'm not sure. it sounds like it's loading it though...
15:46 kermit joined #salt
15:46 philip timoguin yeah seems to be the case, I'll keep digging
15:47 philip yeah, that's it... I wonder what the right way would be to have custom grains only run on matching environments
15:48 jhulten joined #salt
15:48 GradysGhost joined #salt
15:49 timoguin philip: check out the __virtual__ function in other grains and modules
15:49 timoguin have your custom grain do a check for the necessary libraries or environments, and not load if they don't have them
15:50 philip sounds good. It does seem a bit strange that states in given environment wont be called if the environment is not present, but _grains will
15:51 timoguin actually looks like grains might not have that function...
15:52 timoguin they tend to just return empty looks like: https://github.com/saltstack/salt-contrib/blob/master/grains/ec2_info.py#L67
15:53 philip yeah, just not very ideal to have to wait for an http request to fail when I'd rather the custom grain not load at all, but I'll take a look at this
15:54 philip thanks timoguin
15:55 Linuturk left #salt
15:56 thayne joined #salt
15:58 UtahDave joined #salt
15:58 darkelda joined #salt
15:58 jimklo joined #salt
16:00 tligda joined #salt
16:00 kballou joined #salt
16:00 toddejohnson joined #salt
16:00 redondos joined #salt
16:01 redondos joined #salt
16:01 mpanetta joined #salt
16:02 jeremyBass joined #salt
16:05 worstadm1n How can I reference the minion name from a state file is there a way to do this?
16:05 UtahDave worstadm1n: {{ grains['id'] }}
16:07 worstadm1n UtahDave: So if I want a managed file to reference that grain
16:07 worstadm1n can I sneak it into the file?
16:07 repl1cant joined #salt
16:07 worstadm1n I want the file to have a different value per minion
16:07 UtahDave yep!
16:07 Lomithrani http://pastebin.com/bVu3sx7N  I have dozens of those that keep popping during a state.highstate , any idea why ?
16:07 worstadm1n Sweet!
16:08 Lomithrani been like 10 minutes that I have 1 every 2 or 3 seconds
16:08 UtahDave worstadm1n: Totally easy.  You can either use {{ grain['id'] }}   in your state file to set a value, or you can put it directly in a templated file you're managing
16:08 UtahDave Lomithrani: That's the master checking with the minions to see if the minions are still running the highstate or not
16:09 Lomithrani so If I have only that in the minion log , there something wrong right ?
16:09 UtahDave Lomithrani: it happens every 5 seconds, by default.  You can modify that timeframe with the -t cli option
16:09 UtahDave Lomithrani: not necesarily.
16:10 UtahDave Lomithrani: Whenever you run a command, such as a highstate, the master will check in with the minions every 5 seconds if the minion hasn't returned yet
16:12 troyready joined #salt
16:12 Lomithrani /opt/local/bin/pkgin -f -y in sun-jdk6     well that's the latest command my minion has tried to run , shouldn't take 20 minutes right ?
16:12 mpanetta joined #salt
16:14 Lomithrani (job is still active but minion doesnt show anything appart from the master checking if it has finish)
16:14 stanchan joined #salt
16:17 to_json1 joined #salt
16:18 joehillen joined #salt
16:18 alanpearce joined #salt
16:19 UtahDave Lomithrani: The installer may be sitting there waiting for some manual input, or something.
16:19 oz_akan_ joined #salt
16:20 worstadm1n UtahDave: So I put {{ grain['id'] }} in my managed file but it just showed up on the minion as  ServerName {{ grains['id'] }} - Im guessing I forgot some boilerplate to make sure this gets interpreted?
16:21 bemehow joined #salt
16:21 viq worstadm1n: - template: jinja
16:21 viq in the file.managed block
16:22 worstadm1n Many thanks
16:24 UtahDave yep, thanks, viq!
16:25 abe_music joined #salt
16:25 KyleG joined #salt
16:25 KyleG joined #salt
16:26 philipsd6 Using salt-ssh, is it possible to stagger runs? I know I can use max_procs to limit the number of simultaneous ssh connections, but I don't want just want to limit the connections, I want to control the rate at which they initiate connections. Is that possible?
16:27 philipsd6 i.e. for salt-ssh runs that might take 10 minutes, I want to run 5 max procs, but have them start one at a time after 1 minute each.
16:27 Lomithrani UtahDave: Might have been that yes . I don't have the issue anymore , the job finished really quickly but doesn't return any information , (I have checked with the jid aswell ...)
16:28 ndrei joined #salt
16:29 forrest joined #salt
16:30 jalaziz joined #salt
16:31 Lomithrani http://pastebin.com/RY9FtcT2  , does anyone sees something really wrong with that ?
16:32 Lomithrani (because the job executing it has nothing , no error nor good thing)
16:33 pdayton joined #salt
16:33 forrest Lomithrani, did you already run the state with -l debug
16:33 forrest *?
16:34 Lomithrani from the master , no , but the minion is in -l debug
16:35 forrest are you running the job from the master or minion?
16:35 Lomithrani from the master to the minion
16:35 Lomithrani sudo salt App02 state.highstate -l debug
16:36 Lomithrani didnt give much info jid .. found all minions
16:36 Lomithrani thats all
16:36 Lomithrani http://pastebin.com/hMG6WBB9
16:37 forrest ok, so I'd say first run it with -l debug from the master, you might also want to increase the timeout in case a long running command is the issue. If it still just drops out, check to make sure the job is active or not, then query the job id itself to see what the status is
16:37 forrest I assume there's nothing in the minion log
16:37 thayne joined #salt
16:37 gothix how do i set a default in the map.jinga file
16:37 Lomithrani not much no , a sudo salt App03 cmd.run "echo test" works just fine
16:37 gothix forget it found it
16:38 Lomithrani so I figured out it would be in my sls file the problem and not a minion or master config
16:39 Lomithrani ]Command 'state.highstate' failed with return code: 127
16:39 Lomithrani lol well thats normal
16:39 teskew joined #salt
16:39 Lomithrani a bunch of thing happens when I highstate (like a lot) but I don't get any result form the job
16:40 Lomithrani very long log minion side though
16:41 forrest seems like you need to look at increasing your timeout then. Or you need to start running states individually to determine where the error is
16:41 Lomithrani ok I'll do that then
16:42 bhosmer_ joined #salt
16:43 bhosmer joined #salt
16:44 viq UtahDave: :)
16:44 shaggy_surfer joined #salt
16:46 forrest good change on the redis formula viq, though I can't say I'm happy with how the init is structured, though that isn't your fault.
16:47 viq Yeah, just first look at it
16:47 viq Since it looks like my gitlab formula will need to depend on like 6 others...
16:48 viq git, redis, mysql, postgresql, nginx, postfix, epel in case of redhat...
16:48 chadhs joined #salt
16:48 forrest yea always fun when that happens
16:48 notbmatt kind of annoying that tomcat7 ships with a ROOT war that can't be deleted by the tomcat7 user
16:48 viq though on redhat I need first to figure out ruby
16:48 viq So small steps
16:48 forrest viq, just enable the epel repo, and install the latest version
16:48 forrest or a more recent version
16:48 forrest there should be one available
16:48 notbmatt all my tomcat-formula uses have to include a "delete ROOT" watch state :(
16:49 viq Also I'm not sure I'm happy with nginx formula, but as I said, small steps, first I'll get it well supported on debian, and they I'll dive into CentOS
16:49 viq And then I'll probably try a BSD, and that will make it interesting for the formulas this depends on
16:50 forrest Once again this is an 'old' formula
16:50 forrest that has the structure I friggin hate
16:51 notbmatt any idea why salt-call state.highstate would occasionally return YAML instead of the prettified output I usually get?
16:52 shaggy_surfer joined #salt
16:52 viq Also for me annoying about the redis formula is that with default settings it compiles from source
16:53 * timoguin nods
16:53 timoguin should default to distro packages and have source deployment as options
16:54 viq Yeah
16:54 viq Also seems rather ubuntu-centric
16:54 travisfischer joined #salt
16:55 bhosmer joined #salt
16:55 viq But now it's time to go eat something ;)
16:55 utz joined #salt
16:56 jaycedars joined #salt
16:58 forrest timoguin, bleh source deployments
16:58 forrest so ghetto
16:58 netzmonster joined #salt
17:00 mgarfias_ trying to run highstate on a minion, and i get a stacktrace.  how the heck to i debug this?
17:00 hvn joehh: thanks but that does not work. That example is try to pass a None string, but in my case, salt treats None without quote as a string already
17:01 chrisjones joined #salt
17:01 gothix_ joined #salt
17:01 forrest mgarfias_, well, I'd say start by confirming that things like test.ping, and state.sls work
17:03 ml_1 joined #salt
17:03 mgarfias_ well, shit
17:03 mgarfias_ Failed to authenticate, is this user permitted to execute commands?
17:04 kermit joined #salt
17:07 forrest are you running as root?
17:07 mgarfias_ yup
17:07 nvmme joined #salt
17:07 stemmetje joined #salt
17:07 forrest weird, might be worth stracing to see where the error occurs on what file
17:08 stemmetje Any chance I could get some help diagnosing a problem with my salt master?
17:09 forrest stemmetje, just explain what error you're getting and what you've done so far and if someone can help they usually will
17:09 stemmetje It's running very, very slowly and consuming a lot of CPU.
17:10 stemmetje When I run it in the foreground on debug, it gets stuck here for a very long time: [DEBUG   ] Configuration file path: /etc/salt/master
17:11 stemmetje It eventually gets past that but still runs very slowly, most jobs timeout.
17:11 stemmetje I'm running version 2014.1.4.
17:12 CheKolyN joined #salt
17:14 shaggy_surfer joined #salt
17:15 chadhs joined #salt
17:16 bhosmer joined #salt
17:16 bhosmer_ joined #salt
17:16 schimmy joined #salt
17:17 stemmetje I don't see any errors in the logs.
17:17 stemmetje Except for the timeouts: "did not return in time"
17:18 schimmy1 joined #salt
17:19 notbmatt I wonder if the process is busy-waiting for a file handle or something similar
17:19 stemmetje How do I check for that?
17:19 Ryan_Lane joined #salt
17:19 notbmatt not sure
17:19 notbmatt does python have something like gdb? :-\
17:19 cruatta anybody have any idea why in a file.uncomment with this regex: '^exclude=' would uncomment this: # i.e.    exclude=kernel* php*
17:20 cruatta that commented string is on one line...
17:20 notbmatt cruatta: from http://salt.readthedocs.org/en/v2014.1.4/ref/states/all/salt.states.file.html#salt.states.file.uncomment, "A leading ^ character will be stripped for convenience"
17:20 troyready joined #salt
17:21 cruatta o_O
17:22 forrest notbmatt, https://wiki.python.org/moin/DebuggingWithGdb
17:22 cruatta strange. either way, the line that is getting uncommented begins with i.e.
17:22 cruatta so this behavior is unexpected
17:23 notbmatt cruatta: because file.uncomment strips a leading carat, it matches on exclude=
17:23 cruatta what about this: The regex will be searched for from the beginning of the line, ignoring leading spaces (we prepend '^[ t]*')
17:24 cruatta basically it sounds like the developers are including a ^ for you by default
17:24 cruatta am i wrong?
17:24 notbmatt yep
17:24 notbmatt nope
17:24 cruatta so then my regex should be right, with our without the ^ and the line that is getting uncommented is not supposed to be uncommented
17:25 cruatta with or without*
17:25 jgr68_ joined #salt
17:25 notbmatt ^[ t]*exclude=* will match that line, unfortunately
17:25 cruatta [ t]* will only match whitespace no?
17:25 cruatta wait
17:25 cruatta no
17:25 cruatta no it will
17:27 nineteeneightd joined #salt
17:27 Gareth ahoy hoy.
17:28 ipmb joined #salt
17:28 torbock joined #salt
17:29 alanpearce joined #salt
17:30 torbock hi all, i was wondering if anyone else has had security concerns about running a minion process as root that can be remotely controlled? Personally, I feel adequately secure as I am with running an ssh server... i was just wondering if anyone has had similar concerns?
17:30 forrest hey Gareth
17:30 forrest I really don't appreciate that recruiters send email directly to my personal email
17:30 Gareth forrest: hey. hows it going?
17:31 forrest Gareth, good, one step closer to Friday
17:31 notbmatt torbock: my concerns are mitigated if you own/control the DNS record for the master, and by the PKI in place
17:31 forrest torbock, this is something that we have discussed more times than I can remember, as with any such tool there are always concerns regarding how secure the machine itself is, as well as steps to take, overall it just depends on the user.
17:32 forrest how about you Gareth?
17:32 notbmatt torbock: in our implementation there are concerns about guessing hostname glob matches and auto-accepted keys leading to leakage of pillar data, but that's an edge-y case
17:32 utz1 joined #salt
17:33 davidnknight joined #salt
17:34 torbock yea, we've actually implemented masterless setup, using fabric and salt-call.... however we had to build in some targeting like salt provides, it works... we just cant use some of the call things outside of salt-call
17:34 torbock cool*
17:34 tdegrunt joined #salt
17:35 Ryan_Lane forrest, whiteinge: Ok. July 10th works for Lyft offices for the sprint
17:35 forrest ok, if whiteinge is cool with that I'll email my boss and ask
17:36 notbmatt is there a sprint happening? what's the theme?
17:36 mgarfias_ stupid question, reinstalling everything and salt-cloud isnt there.
17:36 stemmetje I got gdb working.
17:36 Ryan_Lane notbmatt: documentation sprint
17:36 forrest notbmatt, yes, documentation
17:36 stemmetje (gdb) py-list   24    # This follows symbolic links, so both islink() and isdir() can be true   25    # for the same path ono systems that support symlinks   26    def isfile(path):   27        """Test whether a path is a regular file"""   28        try:  >29            st = os.stat(path)   30        except os.error:   31            return False   32        return stat.S_ISREG(st.st_mode)   33   34
17:36 ndrei joined #salt
17:36 cliffstah super quick question, does salt-ssh require salt to be installed on the node still?  or does it work like Ansible, where the remote commands are simply bundled up and executed remotely?
17:36 forrest cliffstah, bundled and executed remotely
17:37 forrest no minion when you use salt-ssh
17:37 cliffstah I see, I tried to execute a remote command and got a "can't open file /tmp/.salt/salt-call'"
17:37 cliffstah I wondered if perhaps I still need salt installed remotely
17:40 kermit joined #salt
17:40 JasonSwindle joined #salt
17:42 Gareth forrest: doing well.  on call this week and working Sunday, so I'm waiting for Monday next week :-D
17:42 JasonSwindle joined #salt
17:42 forrest Gareth, booooooo
17:43 felskrone joined #salt
17:43 ramteid joined #salt
17:45 abe_music joined #salt
17:46 travisp joined #salt
17:46 forrest hmm, never heard of this Jeff Porcaro guy who is the engineering VP at salt now
17:46 whiteinge Ryan_Lane: woot!
17:46 whiteinge forrest: +1
17:46 forrest cool
17:48 Gareth forrest: whiteinge needed a new handler.
17:48 forrest heh
17:49 whiteinge https://www.youtube.com/watch?v=m4ILcqYegx4
17:50 travisp Does anyone know a way to run mysql_secure_installation (or achieve the same results) from salt?
17:51 ajprog_laptop joined #salt
17:51 Gareth whiteinge: so...you're the grasshopper or the bug that roars?
17:52 whiteinge to be completely technical i'm the one who is spends way too much time watching kids movies (partly against my will)
17:53 Gareth hah
17:53 forrest whiteinge, what you and cro watch together is your business.
17:54 jcsp joined #salt
17:54 stemmetje joined #salt
17:55 viq travisp: if non-interactive then cmd.run should work
17:55 whiteinge hehe
17:56 travisp viq: Can you run mysql_secure_installation non-interactively?
17:58 viq travisp: no idea :P
17:59 bemehow joined #salt
17:59 stemmetje Is there a way to clear out all jobs from the master?
17:59 travisp viq: Ah ok, The man page doesn't mention it so I don't think that's possible.
18:00 viq travisp: otherwise you could try to write a state that replicates what it does
18:01 rlarkin joined #salt
18:02 linjan_ joined #salt
18:03 JasonSwindle joined #salt
18:04 xcbt joined #salt
18:05 q4brk_ joined #salt
18:06 n8n joined #salt
18:07 worstadm1n If I want to pass a file to a managed file so Jinja can set somethign accordingly - how can I go about that?
18:08 dude051 joined #salt
18:13 worstadm1n Pillar is the answer
18:16 jhulten joined #salt
18:16 mgarfias_ so, fired up a new instance via salt-cloud, and the minion was not installed.
18:17 mgarfias_ why would this be so?
18:20 eliasp is there a reason why nodegroups (http://docs.saltstack.com/en/latest/topics/targeting/nodegroups.html) can only be specified in the master's config, but not via top.sls? I don't want to restart my master each time I'm changing a nodegroup… I could live with this situation if https://github.com/saltstack/salt/issues/11724 was already implemented, but the current situation is a bit… sub-optimal
18:21 schimmy joined #salt
18:21 eliasp ah… https://github.com/saltstack/salt/issues/5787 might have some more details… reading now
18:24 schimmy1 joined #salt
18:34 davidnknight joined #salt
18:37 jcsp1 joined #salt
18:39 toddnni joined #salt
18:41 smcquay joined #salt
18:42 mgarfias joined #salt
18:43 stanchan joined #salt
18:46 chutzpah joined #salt
18:46 scoates joined #salt
18:46 n8n joined #salt
18:47 ipalreadytaken joined #salt
18:47 whiteinge stevednd: thanks for updating that gist. are you sure that's working? the resulting data structure doesn't look like a valid salt state data structure (user and group appear under a single list item)
18:48 scalability-junk joined #salt
18:48 druonysuse joined #salt
18:50 scoates (repeat from yesterday) I'm trying to highstate a specific node (node_id) programmatically. I have: `saltclient = salt.client.LocalClient(); saltclient.cmd_async(node_id, 'state.highstate')` … but that doesn't seem to highstate (I don't see it in the minion log, and I don't see a file's timestamp change as it should/does on a manual highstate.) Any ideas? Code is running on my master node, as root.
18:52 toddnni joined #salt
18:53 xcbt joined #salt
18:54 whiteinge scoates: do you see anything in the jobs runner: salt-run jobs.list_jobs
18:54 scoates checking
18:56 scoates I do see it
18:57 scoates I'm running the minion, manually, in debug mode. I don't see the highstate in the minion log/output
18:59 schimmy joined #salt
19:00 whiteinge hm. your syntax is good and it's showing up in the job cache. does that node_id match in the top.sls file? do you see the minion execute things correctly if you run it manually at the CLI?
19:00 ndrei joined #salt
19:01 scoates yep. if I do "salt testminion.fictivevpn.com state.highstate", it highstates
19:01 scoates job's still sitting here like this, now, though: http://paste.roguecoders.com/p/ffbfa16471be185e0945a8ace027434b.txt
19:02 scoates (that's the one I created with saltclient.cmd_async)
19:04 worstadm1n How can I use salt to check for a listening port?
19:05 dude051 joined #salt
19:06 whiteinge !modules.network.netstat | worstadm1n
19:06 wm-bot4 worstadm1n: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.network.html#salt.modules.network.netstat
19:07 Furao joined #salt
19:08 worstadm1n whiteinge: But what about from a state file?
19:09 rlarkin getting in trouble sometimes with defined environments.  looks like I can get passed source:// being relative with ?saltenv= , but I still have states that won't run with state.sls only highstate
19:09 whiteinge worstadm1n: you could call that function via jinja or via the module.run state. what's your end-goal?
19:09 timoguin worstadm1n: you can use module.run to execute a module within an SLS
19:10 wt_ joined #salt
19:10 wt_ hey
19:10 wt_ any docker users here?
19:10 scoates … if I run it manually, here's what the second job looks like (*1961) this runs properly. the old job is still idling.
19:10 wt_ https://github.com/saltstack/salt/issues/13235 <-- I'd like someone to double check my description of an issue.
19:10 rlarkin wt_ I'm using LXC --a lot-- , but not docker.
19:11 whiteinge scoates: what do you see when you run salt-run jobs.lookup_jid 20140604145557453376
19:12 scoates whiteinge: http://paste.roguecoders.com/p/f2db57dc4876e7c2a4a983ff9c917206.txt
19:12 wt_ rlarkin,  do you the runner from the master or the lxc states module?
19:13 worstadm1n whiteinge: To make sure postgres is listening on 5432 (I would check pkg but it varies versions) and if its listening install a cron backup script
19:13 wt_ rlarkin, and how to do you manage images?
19:13 teo_ joined #salt
19:13 teo_ howdy,
19:14 teo_ how can I pull a file from a minion?
19:14 rlarkin I use lxc templates.  I'm not sure but I assume docker is calling a template an image?
19:14 rlarkin I should kept quiet sorry, I've used lxc for so long and haven't had time to look at docker yet.
19:15 notbmatt teo_: can you expound a bit? what do you mean by "pull a file"?
19:16 scarcry joined #salt
19:16 worstadm1n whiteinge: Looks like I can run it via module.run but Im curious how to parse and use the results
19:16 scoates `salt-run jobs.list_jobs` also shows `saltutil.find_job` jobs for that ID now, whiteinge. weird.
19:17 troyready joined #salt
19:17 whiteinge worstadm1n: i'd suggest running it via jinja to store the result as a variable, then to build the cron state based on the result of that
19:17 smcquay joined #salt
19:17 ZombieFeynman joined #salt
19:17 worstadm1n whiteinge: Any good articles that describe doing this?
19:18 whiteinge not this specific check. i can help you through it though
19:18 viq I need a bit of a hint how to debug this. Working with https://github.com/saltstack-formulas/gitlab-formula and when I do salt \* state.sls gitlab.ruby it works, but salt \* state.sls gitlab says "No matching sls found for 'gitlab' in env 'base'"
19:18 forrest is there an init? :P
19:19 whiteinge scoates: when you run a command at the CLI salt will automatically fire find_job periodically to see if it's still running. you shouldn't see that via the python api though
19:19 notbmatt viq: do you do any includes?
19:19 viq notbmatt: plenty
19:19 viq forrest: yes
19:19 worstadm1n whiteinge: Basically if I can get an idea of how to jinja module.run and iter through the results I can figure it out - but Im clueless as to how to do that
19:19 forrest viq, oh, what does the cache look like?
19:19 forrest for salt
19:19 scoates whiteinge: I see. So… less weird. (-: still confused why it's not highstating, though.
19:19 notbmatt viq: I struggled with this yesterday - in my case there was a syntax error in one of your included states
19:19 notbmatt er, one of my included states
19:19 forrest yea I've also had that happen before
19:19 notbmatt it fails to compile but does so silently
19:20 forrest it's quite annoying
19:20 viq forrest: as in 'ls -l /var/cache/salt/master' or what?
19:20 notbmatt I debugged it by using state.sls and invoking each included state in turn
19:20 whiteinge worstadm1n: i'll pastebin a basic example. sec...
19:20 forrest yea down in the gitlab cache dir just to make sure it exists.
19:20 doriftoshoes joined #salt
19:20 worstadm1n whiteinge: thanks
19:20 forrest honestly I think notbmatt is more likely correct here, run each one individually, see what happens
19:20 viq forrest: /var/cache/salt/master/gitfs/hash/master/gitlab/ has some stuff in it
19:21 forrest including the init?
19:21 TyrfingMjolnir joined #salt
19:21 honestly forrest: I can't say anything about this since I have not followed the discussion.
19:21 forrest whiteinge, did you ever write those docs on how to contribute to the docs?
19:21 viq $ ls -1 /var/cache/salt/master/gitfs/hash/master/gitlab/init*
19:21 viq /var/cache/salt/master/gitfs/hash/master/gitlab/init.sls.hash.blob_sha1
19:21 viq /var/cache/salt/master/gitfs/hash/master/gitlab/init.sls.hash.md5
19:21 forrest honestly, sorry I wasn't trying to tag you
19:22 forrest was using the word honestly
19:22 forrest a bad habit
19:22 honestly I know that (:
19:22 forrest ok
19:22 whiteinge forrest: http://docs.saltstack.com/en/latest/topics/development/conventions/documentation.html
19:22 viq forrest: OK, I'll try them one by one
19:22 forrest whiteinge, nice, we should link that here: http://docs.saltstack.com/en/latest/topics/development/contributing.html
19:22 whiteinge ah, yeah
19:23 thayne joined #salt
19:24 mgarfias CommandExecutionError: Error: aptsources.sourceslist python module not found.
19:25 mgarfias what pkg do i need to install for this?
19:26 scoates whiteinge: I'm firing the highstate at `salt/auth`; I wonder if it's firing before the minion is able to receive the event?
19:26 forrest mgarfias, pretty sure it's just python-apt
19:26 forrest http://www.elmundoenbits.com/2014/05/salt-python-module-not-found.html
19:26 forrest some guy was kind enough to write a blog post on it
19:27 mgarfias thx
19:28 alanpearce joined #salt
19:28 viq Oh, d'oh. If you're using gitfs it actually helps a lot to, you know, commit the changes you want salt to see.
19:28 mgarfias thx
19:28 viq notbmatt: you were right, typos, which I thought I fixed, but not commited
19:29 notbmatt yeah, I'm too lazy to actually create a github account so I can submit issues
19:29 notbmatt viq: would you mind? =D
19:29 viq hah :P
19:30 viq I am waiting for a run to finish, so I guess I have a couple minutes... ;)
19:30 xzarth_ joined #salt
19:30 forrest notbmatt, booooooooo
19:31 forrest viq, my favorite when using gitfs is to put everything in there, add it to the master conf, and then forgetting to restart the master
19:31 forrest and being confused why it's erroring
19:31 notbmatt forrest: I accept all derision. there's also the small matter of my employer refusing to grant permission to submit things discovered or built on company time
19:32 notbmatt that's what keeps me from getting over my laziness
19:32 viq forrest: I'm doing that in my vagrant setup at the moment via salt-formula, so that's taking care of that ;)
19:32 worstadm1n whiteinge: I thought this might be a good use case for an execution module that has just tcp listening ports that jinja can iter over?
19:33 whiteinge worstadm1n: you're definitely right about that. we were kicking around the idea of adding such a module here a couple weeks ago
19:34 worstadm1n Cool Ill write one up
19:34 whiteinge awesome :)
19:34 whiteinge scoates: ah, you could be right about that. are you accepting the minion key as part of that process?
19:35 scoates no; it's already all accepted
19:35 viq notbmatt: though now not sure, did you mean me submitting issues about typos I found, or submitting an issue for you?
19:35 davidnknight joined #salt
19:35 notbmatt submitting an issue regarding salt failing silently when included sls files have syntax errors
19:36 viq aaah
19:38 mateoconfeugo joined #salt
19:38 whiteinge scoates: using the auth event can be tricky because that event will fire somewhat frequently (like when refresshing pillar). are you triying to kick off a highstate when the minion first connects?
19:39 scoates whiteinge: when it connects, yes. That's not the same thing as minion start, though, unfortunately.
19:39 whiteinge scoates: so is this more of a one-time run kind of thing?
19:40 scoates whiteinge: yeah. trying to build a reactor script that will highstate a minion if we've flagged it as needing a highstate. my minions connect erratically (virtual machines on not-always-on connections)
19:41 whiteinge gotcha. do you have a second data point that you're checking (like a custom grain) to determine if the minion is in need of a highstate?
19:42 scoates whiteinge: we're just putting a file named the minion_id in a specific place
19:43 whiteinge that works
19:43 scoates `touch /var/salt/run/master/pile/testminion.fictivevpn.com`
19:43 saimn left #salt
19:43 scoates fwiw, my guess was right. if I `sleep(10)` first, it does pick up the highstate
19:43 whiteinge nice
19:44 whiteinge -ish  ;-)
19:44 scoates heh yes. ish (-:
19:44 scoates I think a sleep is fine.
19:44 scoates chance I'll miss events, but that's probably OK
19:45 whiteinge ooc, what's the distinction between the auth event and the minion start event for your use-case?
19:45 whiteinge oh, when the already-running minion daemon reconnects
19:46 stevednd whiteinge: yes, it works. I have it working in statefiles currently.  I'm not 100%, but I think yaml takes named blocks like that out of the structure. So salt doesn't even see that information
19:46 whiteinge scoates: have you looked into presence events at all?
19:47 scoates hmm. no? when were those introduced? We're on old salt for now… this will help move us forward.
19:47 whiteinge stevednd: very interesting. i'll experiment with it. it's a nice syntax
19:48 Ryan_Lane joined #salt
19:48 dusel joined #salt
19:48 whiteinge scoates: 2014.1
19:50 bemehow joined #salt
19:50 * scoates nods
19:50 scoates thanks. I'll look it up
19:51 vejdmn joined #salt
19:52 smcquay_ joined #salt
19:55 viq notbmatt: https://github.com/saltstack/salt/issues/13236
19:55 Heartsbane Just a general question, any time line for Salt 1.5 on Centos I have someone who constantly complaining that client_acls are broken
19:56 notbmatt viq: thanks sir
19:56 * Heartsbane blames basepi.
19:56 littleidea joined #salt
19:57 * basepi hides
19:58 basepi When you say "1.5" I assume you mean 2014.1.5, right?
19:58 mgarfias whats the best way to have a state look at the os rev and chose a set of packages based on that?
19:58 basepi Hopefully later this week.
19:58 basepi Or early next.
19:59 wt_ rlarkin, thanks for the info
19:59 matrix3000 joined #salt
20:00 viq mgarfias: if blocks based on grains, I guess
20:00 viq mgarfias: or a jinja map that amounts to the same thing. Maybe pillars.
20:01 ghartz_ joined #salt
20:01 n8n joined #salt
20:02 matrix3000 need a little bit of help does Salt have a repository of states like Chef has recipes and Puppet has modules, looking for something similar for Salt so that we don't have to build a bunch of custom modules
20:03 manfred matrix3000: salt-formulas and salt-contrib
20:03 to_json joined #salt
20:03 manfred https://github.com/saltstack-formulas
20:03 forrest matrix3000, hell yea we do https://github.com/saltstack-formulas, and https://github.com/saltstack/salt-contrib
20:03 manfred https://github.com/saltstack/salt-contrib
20:03 matrix3000 manfred: thanks
20:03 forrest nope I win manfred
20:03 manfred :)
20:03 manfred i got one first!
20:03 Heartsbane basepi: yes
20:03 forrest pssssssssh
20:03 to_json1 joined #salt
20:04 Heartsbane Sorry I got distracted by "moar werq asshattery"
20:04 Heartsbane thank you
20:04 Gareth manfred, forrest: fight to the death.  GO!
20:04 forrest Heartsbane, did you try back porting some of the acl stuff? Or was there an issue specifically resolved for 2014.1.5
20:04 * manfred stabs forrest in the neck with a shank
20:04 forrest :(
20:05 * viq drops an anvil on Gareth "you didn't specify whose death..."
20:05 Heartsbane forrest: specific issue I can get you the link if you would like it
20:05 forrest Heartsbane, nah
20:05 forrest Heartsbane, I was just curious regarding what was keeping you from just trying to backport it
20:06 forrest man poor Gareth, on-call, and anvils
20:06 Heartsbane it is compounded by the testing I need to do for the Salt for Windows stuff UtahDave committed to 1.4
20:06 londo__ joined #salt
20:08 Gareth viq: touche.
20:08 Gareth forrest: pretty sure being hit by an anvil is the best possible excuse for getting out of the oncall rotation. :)
20:09 forrest yea but then you can't do anything fun
20:09 forrest on-call is only 1 day
20:09 forrest err 1 week
20:09 notbmatt grar. I'm not convinced modules/debian_ip.py is correct
20:10 Gareth notbmatt: entirely possible.  it's beta and far from complete.
20:11 xcbt joined #salt
20:13 jlund joined #salt
20:14 Joseph joined #salt
20:17 Theo-SLC joined #salt
20:18 * Theo-SLC creeped out by the silence
20:19 to_json joined #salt
20:19 whiteinge heh
20:19 UtahDave Theo-SLC: I think this might fix that sentry issue you were having:  https://github.com/saltstack/salt/pull/13237
20:20 Theo-SLC :) yeah that my pull.  I'm deuscapturus.
20:20 whiteinge and also handsome and a gentlemen, it would seem
20:21 whiteinge stevednd: confirmed that does indeed work. i'm looking in to whether that's unintended or not
20:21 sroegner joined #salt
20:21 UtahDave Theo-SLC: Ha!  sorry, I didn't realize that.    :)
20:21 UtahDave thanks for the fix!
20:22 mgarfias how do i remove a host from the master?
20:22 UtahDave salt-key -d 'minion-id'
20:22 mgarfias er, i have dropped an instance, but the master still tries to talk to it
20:22 mgarfias oh
20:22 mgarfias i'll stfu now
20:22 UtahDave :) nah, don't worry about it.
20:23 mgarfias i just assumed that was for key management only
20:23 UtahDave well, once the key is deleted, the master stop caring about it.  It doesn't delete things like logs and cached jobs, etc.
20:26 dccc joined #salt
20:27 mgw1 joined #salt
20:30 shaggy_surfer joined #salt
20:30 chrisjones joined #salt
20:34 mateoconfeugo joined #salt
20:35 to_json joined #salt
20:36 to_json joined #salt
20:36 davidnknight joined #salt
20:38 jY is there a way using salt-call and gitfs to reference a branch?
20:39 notbmatt salt-call state.sls foostate.bar saltenv='mybranch'
20:40 jY thanks
20:41 mgarfias is there anyway to push out a hosts file entry on a salt-cloud minion bootstrap?
20:42 CeBe joined #salt
20:43 manfred mgarfias: no, but you could specify master: <ipaddress> in the minion: directive in /etc/salt/cloud
20:43 manfred minion:
20:43 manfred master: <ipaddress>
20:43 manfred or whatever the domain is you actually want to point to
20:46 cheus__ joined #salt
20:46 ajprog_laptop joined #salt
20:49 wt_ Is it safe to restart the minion during a highstate run?
20:50 wt_ I'd like to push out a minion config and restart the minion if it changes.
20:52 mgarfias yeah, have a hostname there, just waiting for the A record to propagate
20:52 UtahDave wt_: Well, the highstate won't start running again, unless you have the minion set to run a highstate upon starting up
20:53 wt_ well, I do have the set. Is there any way to make sure to only restart the minion as the last highstate action?
20:53 wt_ I'd like to be able to do something similar for rebooting, actually.
20:56 mgw joined #salt
20:57 UtahDave wt_: one easy way is to have   - order: last    on the reboot state
20:57 jalaziz joined #salt
20:58 kevinjcash joined #salt
20:58 smcquay joined #salt
20:59 worstadm1n Hey UtahDave - If I want a file to exist if a cmd.run has a return value and not exist if it doesn't - whats the best way to do that?
21:00 wt_ UtahDave, is there something liek order: last - 1?
21:00 wt_ I imagine that the minion restart should come before the restart...although practically, it may not matter
21:00 jalbretsen HA HA!  Feature request made and done within 24 hours.  Way to go team!
21:02 smcquay_ joined #salt
21:03 UtahDave jalbretsen: :)  We do the best we can!  :)
21:04 UtahDave worstadm1n: I'd probably have the command you give to cmd.run create the file.
21:04 UtahDave wt_: Hm. not exactly, but you could put something like   - order: 1000000000
21:08 jdmf joined #salt
21:10 dancat joined #salt
21:11 bemehow joined #salt
21:12 oz_akan_ joined #salt
21:13 bemehow_ joined #salt
21:14 conan_the_destro joined #salt
21:15 ml_1 joined #salt
21:21 vbabiy joined #salt
21:22 wt_ UtahDave, thanks
21:25 jrdx joined #salt
21:25 UtahDave fivethreeo: Hey, are you around?
21:28 stevednd whiteinge: why wouldn't it be intended? That's how yaml is supposed to work. It's nothing to do with salt.
21:29 alanpearce joined #salt
21:30 kmshultz joined #salt
21:31 mgw1 joined #salt
21:33 sverrest joined #salt
21:34 cliffstah I'm getting an odd issue using salt-ssh - the tgz that is being uploaded when trying to execute a remote command is somehow corrupt (gzip: salt-thin.tgz: unexpected end of file) - is this a known / recent issue?
21:34 oz_akan_ joined #salt
21:34 matrix3000 how do you minimize the size of your top.sls file if we have lots of hosts? is there a way to have it look at other files?
21:36 ZombieFe_ joined #salt
21:36 davidnknight joined #salt
21:37 fivethreeo UtahDave: yes
21:37 Corey matrix3000: http://reclass.pantsfullofunix.net/ is one option.
21:37 shaggy_surfer joined #salt
21:38 n8n joined #salt
21:40 mgw joined #salt
21:42 jdmf joined #salt
21:42 n8n_ joined #salt
21:44 whiteinge stevednd: it is valid yaml, it is not a valid data structure for salt states as described here: http://docs.saltstack.com/en/latest/ref/states/highstate.html
21:45 whiteinge ...but salt's state compiler is compiling it to a valid data structure. thus the question
21:45 jdmf joined #salt
21:45 stevednd whiteinge: I'm not sure I follow. What part exactly shouldn't be considered valid?
21:46 viq matrix3000: http://docs.saltstack.com/en/latest/topics/targeting/index.html
21:46 whiteinge stevednd: this is the difference distilled: http://paste.fedoraproject.org/107289/40191839
21:47 jdmf joined #salt
21:48 smcquay joined #salt
21:49 taion809 joined #salt
21:50 matrix3000 left #salt
21:53 n8n joined #salt
21:57 vbabiy joined #salt
21:58 stevednd whiteinge: is that using yamllint?
21:58 stevednd .com
21:59 whiteinge stevednd: no, i made a simple state to explore the different data structures.
22:00 whiteinge don't worry about it. you've got it working and it's not likely to change. i'll let you know what i find out
22:00 stevednd using the parser here http://yaml-online-parser.appspot.com/ to output as python, is that giving the same structure you're seeing?
22:01 stevednd I think I understand what you're saying, how it effectively gives a an array of hashes
22:02 stevednd maybe there's something in the parsing, or interpretation that automatically merges all of those hashes
22:02 stevednd or conversely, blows them apart
22:02 whiteinge yeah, exactly right. salt's state compiler is translating both formats to the same format
22:03 whiteinge nice yaml parser, btw. thanks for the link
22:05 n8n joined #salt
22:07 UtahDave fivethreeo: can I send you a pm?
22:10 mgarfias is there anyway to kill a running job on a minion?  or do i wait?
22:11 mgarfias i have two minions running with jobs that appear to be stuck
22:11 mgarfias when i do a jobs.lookup_jid they show complete though
22:11 xzarth joined #salt
22:13 viq mgarfias: try jobs.active
22:13 viq mgarfias: there's also jobs.terminate or something like that
22:14 viq !help
22:14 wm-bot4 I'm a documentation bot. To control me, please use #salt-bot to avoid channel spam. See this URL for my commands: http://meta.wikimedia.org/wiki/WM-Bot
22:16 Jarus joined #salt
22:16 to_json1 joined #salt
22:18 notbmatt joined #salt
22:19 kedo39 joined #salt
22:21 ZombieFeynman joined #salt
22:24 mwalker left #salt
22:30 alanpearce joined #salt
22:34 kermit joined #salt
22:36 n8n joined #salt
22:37 davidnknight joined #salt
22:38 jbirdman joined #salt
22:39 jlund joined #salt
22:39 fragamus joined #salt
22:41 ZombieFeynman joined #salt
22:49 happytux joined #salt
22:55 bhosmer joined #salt
22:57 Outlander joined #salt
22:57 bhosmer_ joined #salt
22:59 jjasinski joined #salt
23:00 kedo39 joined #salt
23:02 travisfischer joined #salt
23:02 scoates joined #salt
23:05 londo__ joined #salt
23:06 happytux_ joined #salt
23:08 Comradephate joined #salt
23:10 godber1 joined #salt
23:11 cliffstah hey folks, I'm having a strange issue running salt-ssh from the develop branch, could someone verify it's an actual issue or whether it's likely just something local to me?
23:12 cliffstah https://gist.github.com/cliffrowley/752bd51d5b9d72889d42
23:12 ajolo joined #salt
23:13 forrest does the stable release work?
23:14 cliffstah it does, however I'd like to determine whether this is an issue with the way Homebrew is installing head - I already fixed 1 issue in the homebrew recipe, I'd really like to get it working
23:15 cliffstah if noone else is experiencing this issue then I'll start beating it out of homebrew ;-)
23:16 ml_1 joined #salt
23:17 forrest yea I don't know, haven't seen anyone else report that problem
23:17 forrest but not a lot of people are running salt-ssh straight from develop either
23:17 cliffstah hm ok, thanks.
23:17 cliffstah I started initially just to grab the Saltfile support
23:18 cliffstah but then, you know, one thing leads to another..
23:20 jjasinski Hi all, is there a way to issue "salt" or "salt-ssh" command to find grains.items and have it return the result in a json dict?
23:21 whiteinge --out=json
23:21 UtahDave jjasinski: just addd    --out json
23:21 UtahDave gah! whiteinge is too fast
23:21 jjasinski Great, thank you!
23:21 mgarfias both of you are :)
23:21 manfred 40 seconds is an eternity
23:22 jjasinski (Very neat - I should have know something like that would be built in! :-)
23:24 UtahDave jjasinski: yeah, that's what we call an "outputter" .   There are several
23:25 sroegner joined #salt
23:25 jjasinski Good to know the term for it - I'll check those out.
23:26 manfred https://github.com/saltstack/salt/tree/develop/salt/output
23:28 jcsp joined #salt
23:29 jjasinski (Looks like are a lot of output options - really cool. Will make it easy to integrate)
23:31 alanpearce joined #salt
23:33 whiteinge jjasinski: if you have existing systems or data it goes the other way too -- it's easy to get data into salt  :)
23:37 ldlework hi jjasinski
23:38 ldlework glad to hear you guys are adopting salt!
23:38 davidnknight joined #salt
23:38 ZombieFeynman joined #salt
23:48 ldlework :/
23:49 jalaziz joined #salt
23:51 mosen joined #salt
23:57 forrest he doesn't want to talk to you

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