Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-06-26

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

All times shown according to UTC.

Time Nick Message
00:01 shaggy_surfer where are the version names for salt… I am reading a doc and it states "New in version Helium"…  I am using 2014.1.3 and want to use the module salt.modules.blockdev but it says it's unavailable when I run it.
00:02 shaggy_surfer Do I have to upgrade to latest stable: 2014.1.5 to get that?
00:02 alainv shaggy_surfer: No, the docs default to the development branch. You'd have to grab that from git
00:02 dstokes shaggy_surfer: you'll have to use develop branch
00:03 dstokes alainv: beat me to it ;)
00:03 alainv ;)
00:03 shaggy_surfer I don't want to use develop branch… I am using the repo version for ubuntu from the ppa
00:03 shaggy_surfer I like to use stable as I am using this for production
00:03 dstokes shaggy_surfer: you won't be able to use block till next release then (this or next month)
00:03 alainv that remains my #1 pet peeve with the project so far, the official docs have such better pagerank but are a constant "haha, not for you!" for production use
00:04 shaggy_surfer ok…. how do I get a list of version numbers and their project names?
00:04 dstokes hydrogen is the current stable release
00:04 seventy3 joined #salt
00:04 Guest7481 joined #salt
00:04 alainv shaggy_surfer: I have switched to using Dash with a local copy of the docs for Stable. It's almost as good :/
00:05 BrendanGilmore joined #salt
00:05 shaggy_surfer how can I make sure I am reading the relative doc's for the release?
00:05 dvogt joined #salt
00:05 shaggy_surfer I am on 2014.1.3 and want to reference the documentation for that point release….
00:05 alainv The sidebar
00:06 alainv These docs are for Salt's development version: d550494.  -  Docs for previous releases are available on salt.rtfd.org.   - Latest Salt release: 2014.1.5
00:06 joehh shaggy_surfer: either that or the salt-doc package from your packages...
00:06 shaggy_surfer is the latest considered stable?
00:06 shaggy_surfer 2014.1.5?
00:06 alainv yes
00:07 alainv http://salt.readthedocs.org/en/v2014.1.4/ should have effectively the same functionality, there shouldn't be substantial differences in the docs
00:07 shaggy_surfer I could upgrade from 2014.1.3 to 2014.1.5 using ppa right, just apt-get install packagename on ubuntu right… do master first, then minions
00:07 yomilk joined #salt
00:07 alainv if they have salt installed I think they should already have the PPA, if you used the bootstrap script
00:07 Shenril joined #salt
00:07 shaggy_surfer correct they do
00:08 dvogt joined #salt
00:08 alainv hah nice now the "docs for previous releases" link is just an endless loop to /en/latest/
00:08 shaggy_surfer I read you are supposed to move up the master first and then the minions last
00:08 shaggy_surfer hehe
00:08 kivihtin joined #salt
00:09 alainv shaggy_surfer: for minor versions i don't -believe- it should break either way, but yes, master first
00:09 shaggy_surfer do you push salt version w/ salt on the minions?  :P
00:10 whiteinge stevednd: would you please file a ticket to support minion-local file paths? it would be stupid-easy to add
00:11 whiteinge stevednd: in the meantime, we could whip up a quick wrapper for dstokes' suggestion that you could use. the trick/distinction is renderer modules vs. execution modules. we just need to wrap the jinja renderer module with an execution module
00:11 whiteinge gimme a few and i'll pastebin something
00:11 shaggy_surfer How do I find out what the version name is for the releases?  like where did you find what version Helium is?
00:13 sunkist joined #salt
00:17 alainv shaggy_surfer: also not very well exposed. would be found in mailing lists and blogs
00:18 aquinas joined #salt
00:19 shaggy_surfer thanks for your answers… I will turn down the noise.
00:23 BrendanGilmore joined #salt
00:27 bmatt so let's say I have a module that runs "read"
00:28 bmatt calls cmd.run_all("read") or whatever
00:28 bmatt will it ever time out?
00:28 mastrolinux joined #salt
00:29 MZAWeb joined #salt
00:30 krow joined #salt
00:39 whiteinge stevednd: http://paste.fedoraproject.org/113292/37431441/
00:42 Guest7481 joined #salt
00:43 otter768 joined #salt
00:45 stevednd whiteinge: thanks, I'll take a look at that after I eat. Can you tell me when includes run? Is it before the other states, or where they appear in the file? The reason I ask is because you can't use requires on includes, so I couldn't say wait to run include x until after include y runs
00:46 bhosmer joined #salt
00:46 kivihtin joined #salt
00:46 ipalreadytaken joined #salt
00:47 whiteinge stevednd: the includes are parsed sequentially when the yaml is getting parsed into a data structure so the final data structure should be in the same order as the includes. i.e., the contents of the included file will appear where the include was in that file
00:48 oz_akan__ joined #salt
00:48 whiteinge stevednd: you can do sls-level includes
00:48 whiteinge s/includes/requires
00:48 whiteinge stevednd: http://docs.saltstack.com/en/latest/ref/states/requisites.html#require-an-entire-sls-file
00:55 blast_hardcheese joined #salt
00:58 bhosmer joined #salt
01:00 jonbrefe joined #salt
01:09 vejdmn joined #salt
01:10 Guest7481 joined #salt
01:14 vejdmn joined #salt
01:14 Guest0852 joined #salt
01:18 tkharju1 joined #salt
01:19 dvogt joined #salt
01:20 mgw joined #salt
01:22 Guest0852 joined #salt
01:24 DaveQB joined #salt
01:26 sunkist does anyone know thatch's birthday?
01:26 taterbase joined #salt
01:27 bhosmer_ joined #salt
01:28 mastrolinux joined #salt
01:30 bmatt is there a convenient way to interact with salt in a python interpreter?
01:30 bmatt so I can run python, import my module, and call its methods directly?
01:38 joehh sunkist: maybe try during us business hours over on #salt-devel
01:38 shaggy_surfer how do I get to the 2014.1.3 docs?  The Docs for Previous release loops to latest….
01:41 jnials joined #salt
01:47 yomilk joined #salt
01:48 dvogt joined #salt
01:49 sunkist Thanks joehh.
01:49 joehh shaggy_surfer: you could download the salt-docs package and view them locally
01:53 Guest0852 joined #salt
01:57 Guest0852 joined #salt
01:58 stevednd whiteinge: does the builtin saltstack 'include:' actually inject the sls into the current statefile? I didn't think it did. I know you can require an sls in a state, but I was talking more like include: - somesls - require: -othersls
01:58 stevednd something like that isn't valid, is it?
02:02 yomilk joined #salt
02:05 Guest0852 joined #salt
02:06 bmatt you can do
02:07 bmatt include:\n  foo.sls\n\nbar:\n  state.function:\n    - require:\n      - sls: foo
02:07 bmatt er, sorry, just "include: foo"
02:08 bmatt it's a pattern I find incredibly useful when, for example, sometimes /var/www/ is a file.directory and sometimes /var/www/ is a mount.mounted
02:08 bmatt I can parameterize the /var/www/ state inside wwwdir.sls, and just require the lot of it without caring about type
02:09 mgarfias joined #salt
02:09 jerrcs joined #salt
02:10 jcsp joined #salt
02:12 CeBe2 joined #salt
02:12 CeBe3 joined #salt
02:13 shaggy_surfer what's a good module to list block devices prior to mount?  example on linux I run :  lsblk —fs
02:14 shaggy_surfer I see a blockdev module in the recent 2014.1.5 but that doesn't list
02:14 bmatt I don't think such a thing exists
02:14 bmatt sounds like a good feature! submit a pull request =D
02:14 shaggy_surfer bummer.  I am using ec2 and want to use salt commands to remotely check if a ebs is attached
02:14 shaggy_surfer like a status
02:15 bmatt ¯\(°_o)/¯
02:15 manfred ¯\(°_o)/¯
02:16 MZAWeb joined #salt
02:16 shaggy_surfer i used the extfs module to make the file system on the block device after mount
02:16 shaggy_surfer I upgraded and looks like there is a module called blockdev that will allow me to adjust the readahead of a volume also
02:17 ipmb joined #salt
02:18 shaggy_surfer only thing lacking at this point is a way to query the linux sub-system and loop and wait to do those 2 things until the ebs volume is attached.  I tried using something like this:  ATTACH_STATUS=$(sudo salt ${HOSTNAME} cmd.run lsblk --fs | grep ${DEVICE_NAME})
02:19 shaggy_surfer but the issue is that ATTACH_STATUS has what salt returns which is the hostname and volume name, not the actual output from the system
02:19 shaggy_surfer I was trying to cheat and use the salt '*' cmd.run , but looks like I see now why I shouldn't do that.
02:22 bmatt eschew alteration. embrace idempotence
02:23 BrendanGilmore joined #salt
02:25 kballou joined #salt
02:28 mastrolinux joined #salt
02:29 oncallsucks joined #salt
02:29 ipmb joined #salt
02:29 ndrei joined #salt
02:30 yomilk joined #salt
02:41 lowfive joined #salt
02:46 dvogt joined #salt
02:46 Blacklite can i call highstate with a specific service? e.g. i want to just update this service and not go through everything just to call highstate
02:47 Eugene Yes, it's called state.sls, rather than highstate
02:49 Guest0852 joined #salt
02:51 mgw joined #salt
02:53 lowfive joined #salt
02:55 dude051 joined #salt
02:55 jcsp joined #salt
02:56 malinoff joined #salt
02:57 Blacklite ok
02:57 dude051 joined #salt
02:58 Blacklite so i just want to apply the rules from a particular role
02:58 Blacklite how would i go about doing that
02:58 rudolfsteiner joined #salt
02:59 mgw is there something other than an entry in file_roots: needed to add another env?
02:59 manfred there is not
02:59 manfred well, something for pillars maybe
03:00 mgw manfred: pillars are not my concern at the moment
03:00 mgw I'm getting a key error
03:00 manfred unfortunately I am about to go get some sleep or i would try and help
03:01 mgw https://gist.github.com/mgwilliams/5bb5b0e7a3c835c147ca
03:01 mgw manfred: ok, thanks
03:01 manfred good luck
03:01 mgw thanks
03:02 rudolfsteiner hi list, i have salt-master salt 0.17.5 and salt-minions 2014.1.4 and salt-minion 2014.1.5
03:02 rudolfsteiner when I run: salt '*' state.highstate
03:03 mgw rudolfsteiner: you don't want that
03:03 mgw your master should be no older than your minions
03:03 rudolfsteiner mgw: I have Ubuntu 14.04 on master, I install master from pip
03:04 mgw rudolfsteiner: from pip?
03:04 mgw or apt?
03:04 mgw 0.17.5 is very old
03:04 mgw on ubuntu 14.04 I'd recommend installing from saltstack's ppa
03:05 rudolfsteiner mgw: If I use the ppa it doesn´t work with centos images running different version of minions
03:06 rudolfsteiner mgw: do I have to use same version on master and minions?
03:06 mgw what version do you have on centos?
03:06 mgw not exactly
03:06 mgw but I'm not sure whether 2014.1 is backwards compatible to 0.17.5…
03:06 mgw I think it is
03:08 mgw what version do you have on centos?
03:08 rudolfsteiner mgw: salt-minion-2014.1.4-1
03:08 rudolfsteiner mgw: is the version that salt-cloud install from repos on cents 6.5
03:08 mgw that should be fine with 2014.1.5 that's on ppa
03:09 rudolfsteiner mgw: I read this: https://github.com/saltstack/salt/issues/13535
03:09 bmatt so my grain needs to call my custom module; how can I import a custom module?
03:09 bmatt do they still exist in the salt.modules namespace?
03:10 manfred rudolfsteiner: http://salt.readthedocs.org/en/latest/faq.html#can-i-run-different-versions-of-salt-on-my-master-and-minion
03:10 manfred http://docs.saltstack.com/en/latest/faq.html#can-i-run-different-versions-of-salt-on-my-master-and-minion
03:11 mgw rudolfsteiner: The latest Trusty packages have that fixed though
03:12 rudolfsteiner mgw: I´ll try! I need to have salt running tomorrow for a demo in one of my clients :)
03:13 manfred rudolfsteiner: read te wiki page i gave you on how versions are allowed to work together
03:13 manfred docs page
03:14 rudolfsteiner manfred: thanks a lot, I´ll try it I need to have salt running tomorrow for a demo in one of my clients :)
03:16 bhosmer joined #salt
03:17 aw110f joined #salt
03:22 yomilk joined #salt
03:23 yomilk joined #salt
03:28 mastrolinux joined #salt
03:31 thayne joined #salt
03:37 stevednd bmatt: yes, I know that that is possible, but it's not what I'm looking for. I'm talking about having two separate includes, and having one require that the other runs successfully before attempting to run
03:38 stevednd something like this https://gist.github.com/dnd/f54398254bc7e9102af3
03:38 stevednd which afaik is not possible currently with salt
03:39 manfred stevednd: that is not possible at this time
03:40 stevednd manfred: any options other than sticking with jinja includes, and having to know explicit state names in the include?
03:40 manfred stevednd: i do not believe so
03:41 manfred i havent dug into how includes works a lot yet
03:41 manfred i will take a look tomorrow or this weekend and see if I can find a way to do what you are looking for
03:41 aw110f joined #salt
03:42 stevednd cool, thanks. I seem to to have found the current include functionality lacking lately. If I can manage to get this project behind me I would probably try to get some of it added.
03:43 manfred yar, i will let you know when i get a chance to start looking, and then what I find or PR if I get a chance to work on it
03:44 stevednd sounds good
03:45 manfred still working on this cluster that blew up last saturday, but we figured out what happened today, so now i should have time tomorrow :P
03:46 deares joined #salt
03:50 joehh rudolfsteiner: let me know if you have issues with the trusty packages
03:50 joehh should work with earlier minions
03:50 rudolfsteiner joehh: it´s working!
03:51 rudolfsteiner joehh: know I´m trying to put sls files in github :)
03:51 joehh good to hear
03:58 BrendanGilmore joined #salt
04:18 ramishra joined #salt
04:19 kermit joined #salt
04:19 mosen joined #salt
04:20 jeffrey4l joined #salt
04:24 jcsp joined #salt
04:25 anuvrat joined #salt
04:27 ramishra joined #salt
04:28 mastrolinux joined #salt
04:48 otter768 joined #salt
04:49 felskrone joined #salt
04:50 ramteid joined #salt
04:50 ipalreadytaken joined #salt
04:57 __alex joined #salt
04:58 pdayton joined #salt
05:00 dvogt joined #salt
05:02 davet joined #salt
05:02 davet1 joined #salt
05:03 xmj so
05:03 antonw joined #salt
05:03 xmj speaking of packages- when's .6 due? :)
05:04 xmj also --- can someone please fix the topic, 2014.1.4 isn't the Latest.
05:04 bhosmer joined #salt
05:12 lionel joined #salt
05:25 maestropandy1608 joined #salt
05:25 maestropandy1608 Hi all
05:26 maestropandy1608 am new to salt stack, i want to install on windows server 2012, can any one help me please
05:26 malinoff maestropandy1608, http://docs.saltstack.com/en/latest/topics/installation/index.html
05:27 maestropandy16 joined #salt
05:28 mastrolinux joined #salt
05:33 maestropandy16 for windows server 2012 which package want to download from http://docs.saltstack.com/en/latest/topics/installation/windows.html
05:33 maestropandy16 pls help me yar
05:33 maestropandy16 i am having confusion to think which package suits for it
05:34 malinoff masterkorp, the latest, of course, 2014.1.5
05:35 maestropandy16 https://docs.saltstack.com/downloads/Salt-Minion-2014.1.5-AMD64-Setup.exe this only right for 64 bit
05:35 malinoff maestropandy16, I don't know, I have no windows minions. I guess you can try without any problems
05:35 bhosmer joined #salt
05:35 maestropandy16 thanks malinoff
05:38 Hell_Lap joined #salt
05:48 tristianc|Alt joined #salt
05:48 garthk joined #salt
05:50 anuvrat joined #salt
05:55 Shenril joined #salt
06:08 ajw0100 joined #salt
06:18 maestropandy16 how to install mysgit for windows 7
06:18 maestropandy16 while onstalling salt stack on windows
06:19 malinoff masterkorp, this channel is not google
06:22 thayne joined #salt
06:26 picker joined #salt
06:29 kiorky joined #salt
06:32 allanparsons joined #salt
06:38 TyrfingMjolnir joined #salt
06:44 jalaziz joined #salt
06:46 jalaziz_ joined #salt
06:49 xmj chances are you'll want to go to bed anyway.
06:51 chiui joined #salt
06:53 Hell_Fire joined #salt
06:53 bhosmer joined #salt
06:59 Kenzor joined #salt
07:11 slav0nic joined #salt
07:13 jdmf joined #salt
07:15 yomilk joined #salt
07:15 ml_1 joined #salt
07:16 eofs any tips on how to profile minions? e.g. how long it took to run highstate etc
07:16 martoss joined #salt
07:17 martoss1 joined #salt
07:20 alanpearce joined #salt
07:25 analogbyte joined #salt
07:28 ph8 joined #salt
07:37 analogbyte joined #salt
07:43 sander^work joined #salt
07:45 sander^work Whats the package name in ubuntu?
07:45 ndrei joined #salt
07:50 darkelda joined #salt
07:53 luette joined #salt
07:54 Lomithrani1 joined #salt
07:57 z3uS joined #salt
08:04 jeddi joined #salt
08:05 joehh sander^work: salt?
08:05 joehh do you mean salt-minion
08:05 joehh and salt-master?
08:05 joehh the ppa is ppa:saltstack/salt
08:07 tkharju2 joined #salt
08:08 ndrei joined #salt
08:08 philipsd6 We have a centralized uid/gid database that we need to enforce when creating users/groups. What's the best way to do this?
08:08 philipsd6 My idea is that we create a new module to manage the uid/gid api, and a new state module without uid/gid parameters, that calls salt's user.present state internally.
08:08 philipsd6 Ideally, our state module would replace the user state because we don't want any state to bypass our uid/gid management.
08:09 philipsd6 Is this a feasible plan?
08:11 CeBe joined #salt
08:12 N-Mi joined #salt
08:12 N-Mi joined #salt
08:18 elfixit joined #salt
08:19 CeBe1 joined #salt
08:20 GnuLxUsr joined #salt
08:24 CeBe joined #salt
08:28 jhauser joined #salt
08:32 jeffrey4l left #salt
08:41 Striki joined #salt
08:42 phx joined #salt
08:42 bhosmer joined #salt
08:42 phx hello
08:52 giantlock joined #salt
08:52 seventy3 joined #salt
09:24 phx what does exactly SLS stand for, what does it abbreviate?
09:24 malinoff phx, http://docs.saltstack.com/en/latest/topics/tutorials/starting_states.html
09:27 phx ty
09:28 eofs or Simply Less Stress ;)
09:28 malinoff Not with jinja
09:34 ipalreadytaken joined #salt
09:37 phx just doing my first steps, so doing nothing but learning as trodding the path
09:39 malinoff phx, http://docs.saltstack.com/en/latest/topics/tutorials/walkthrough.html read this
09:41 masterkorp malinoff: hello, good morning, were those 2 messages for me ?
09:43 malinoff masterkorp, no, I would have write your nick
09:45 masterkorp malinoff: http://pastie.org/private/0oap8mbzwy4jwog8k3w
09:45 masterkorp in this context =
09:46 masterkorp ?
09:46 malinoff Oh, sorry, misspelled the other guy's nick :)
09:46 masterkorp malinoff: ok, no problem, i thank you for helping others
09:54 phx i'm checking the walkthrough that malinoff pointed me to, however when based on that, i'm trying to make a state that makes ssh (both server and client) installed, i'm getting the following error: http://pastebin.com/Ka2hkhXK (both init.sls and error message included). could someone please take a look and help me what am i doing wrong here?
09:55 malinoff phx, there should be 2 spaces
09:55 malinoff or 4
09:55 phx so indent by multiples of 2 space?
09:55 phx s
09:56 malinoff I guess. Never tried more that 4 spaces
09:56 malinoff than*
09:56 phx seems to work, thanks
09:59 phx malinoff, and how do i specify the name of a service to ensure it's running? package name is openssh-server, service name is sshd
10:00 malinoff phx, you can explicitly use name argument - http://docs.saltstack.com/en/latest/topics/tutorials/starting_states.html#getting-to-know-the-default-yaml-jinja check out the example
10:13 dober joined #salt
10:13 dober left #salt
10:13 phx brb belly calls
10:18 thayne joined #salt
10:19 picker joined #salt
10:19 hsmade joined #salt
10:22 hsmade Hi all, why is it that with custom grains, I get a variable number of matches when filtering with -G 'xxx:*y*'  but I always get the full list with grains.item xxx ?
10:23 CeBe joined #salt
10:26 dober joined #salt
10:27 Hell_Fire hrmm, my version of state.keystone doesn't have endpoint_present
10:28 Hell_Fire ffff
10:28 CeBe joined #salt
10:31 bhosmer joined #salt
10:37 yomilk joined #salt
10:40 yidhra joined #salt
10:43 ghartz joined #salt
10:51 ekristen joined #salt
10:53 vu joined #salt
10:55 CeBe1 joined #salt
11:00 bhosmer joined #salt
11:04 picker joined #salt
11:13 sander^work I'm getting GPG error in apt-get update now. The following signatures couldn't be verified because the public key is not available...
11:14 logix812 joined #salt
11:14 tkharju3 joined #salt
11:14 sander^work No worries.. made it.-
11:24 Hell_Fire_ joined #salt
11:24 phx malinoff, thanks. seems like i got it to work, however i still have to understand what config mistakes did i make
11:25 ndrei joined #salt
11:25 martoss joined #salt
11:25 jrdx joined #salt
11:26 bhosmer joined #salt
11:29 ml_1 joined #salt
11:35 ninkotech joined #salt
11:37 xmj phx: yay, more freebsd users turning to salt
11:44 phx xmj, actually i'm working with centos6 here at work right now
11:44 xmj dont we all
11:44 rudolfsteiner joined #salt
11:45 xmj phx: fiddling to get a bhyve c6 vm use salt to deploy OpenStack.  :)
11:49 viq xmj: no, not all, mostly debian here, and I'm thinking about moving to a place where it's pretty much exclusively ubuntu server :(
11:50 martoss1 joined #salt
11:52 robinsmidsrod which is the better way to write your sls file for elasticsearch on ubuntu using ppa?
11:53 jaimed joined #salt
11:54 Lomithrani joined #salt
11:56 xmj viq: sigh
11:56 xmj viq: I want to find a place that's exclusively BSD.
11:59 babilen #freebsd?
11:59 xmj babilen: one that pays.
11:59 babilen heh
11:59 xmj babilen: (/whois is your friend ;)
12:00 viq xmj: I'm still tempted though. Here I'm maintaining a bunch of old and _old_ machines, bunch of them for various clients, and I'm slowly trying to bring the team into this decade, with using salt and logging into central place and so on. New place would be mostly puppet and OpenStack internals
12:00 xmj yikes
12:01 viq yikes to which part? ;)
12:01 blarghmatey joined #salt
12:01 xmj OpenStack
12:02 viq Why?
12:05 diegows joined #salt
12:05 happytux joined #salt
12:05 sxar joined #salt
12:06 * viq tries to figure out what would be a good way to target states to assign grains to machines
12:06 viq match: list with 30+ machines in it seems... ugly
12:07 sander^work What is the oldest debian distro salt works with?
12:08 viq sander^work: 6
12:08 viq you maybe probably could get it to work on 5, but you'd most likely have to compile your own packages
12:09 sander^work viq, whats the oldest one I could get it to work with.. when compiling myself?
12:11 viq sander^work: no idea, sorry. You'd need to compile python, zeromq and probably bunch of other stuff to get it to work.
12:12 sander^work python-zmq (>= 13.1.0)  dctrl-tools  msgpack-python  python-m2crypto salt-common salt-minion
12:12 viq To make it more interesting, I've seen OpenVZ guests that run debian 6 that don't ever make any connections to the master. The thing I found they have in common is the OpenVZ host is running debian 5, but I did not debug further to see what's really happening there
12:13 blarghmatey joined #salt
12:14 vbabiy joined #salt
12:15 phx a question about environments. is there a way to specify custom, say, on-demand environments? like when someone tries to get something done, he clones a given environment to be his own sandbox, than messes with the configuration there, after that's done, merging the required changes back to the proper environment
12:15 malinoff joined #salt
12:15 alanpearce joined #salt
12:15 viq phx: git branches with gitfs sound like exactly that (cue babilen ranting about gitfs ;)
12:17 babilen A wild babilen appears!
12:17 sxar joined #salt
12:17 viq ;)
12:17 phx mhm, i'll mess with that later
12:18 babilen phx: Just make sure to use a separate repository for top.sls and you should be fine. I personally find it braindead to conflate git branches and environments, but meh
12:18 babilen Oh man ... My little "Let's save time by modelling my environment in docker containers that I manage with vagrant" idea turned into a day project. :(
12:19 phx and another question. let's say a set of states is applied to a given node, which involves a set of packages being installed. is there a way to remove packages which (or their deps) aren't referenced by states?
12:19 babilen (/me can't stand virtualbox)
12:19 phx that is, enforce that only those packages are installed which are actually required
12:19 malinoff phx, you can always spawn cmd.run
12:20 phx malinoff, i would prefer not writing the logic for that
12:20 bhosmer joined #salt
12:20 babilen phx: I'm not aware of such a method. It shouldn't be too hard to purge everything but the essential set of packages *before* installing anything else though.
12:20 viq phx: are you up to specifying _ALL_ packages that a machine is to have?
12:20 malinoff phx, or, which is better, use http://docs.saltstack.com/en/latest/ref/states/all/salt.states.module.html#module-salt.states.module
12:21 phx viq, i'd count on dependencies
12:21 phx the point would be to disallow ninja manual installation of shit
12:21 viq phx: still, probably every run it would try to remove packages, and fail because they are depended upon
12:22 babilen Do you, by chance, use docker in vagrant to model salt master/minion setups and would be willing to share that setup. (I got pretty far, but ...)
12:22 viq phx: how about instead some monitoring tool for that?
12:23 phx viq, would be another component, i personally don't think this is a job for a monitoring tool
12:24 babilen phx: Well, you certainly wouldn't want to remove, say, base-files, coreutils, mount, ... on a Debian system. My recommendation would therefore be to purge everything but essential packages (aptitude purge '!~E') and take it from there ... note that you will miss many tools that you would normally expect to be present.
12:24 viq phx: sounds somewhat like 'golden images' && 'immutable infrastructure'
12:25 phx babilen, that's why i'd like to define a baseline for each environment
12:26 viq How would you guys go to assigning grains to machines via states? match: list in top.sls ? list of ifs in state itself?
12:26 babilen phx: Yeah, the baseline obviously is the set of essential packages. If you'd like more, I'd also install requires, important and standard packages (i.e. "aptitude install ~pstandard ~prequired ~pimportant)
12:26 viq Some other way?
12:27 babilen viq: I'd solve that via pillars and populate http://docs.saltstack.com/en/latest/ref/states/all/salt.states.grains.html#salt.states.grains.present from there
12:28 viq babilen: same question: how would you assign that in pillars?
12:28 babilen viq: Unless you have some other invariant that you can match on there is little you can do but to enumerate the hosts *somewhere* and I'd argue that pillars is the right place for that.
12:29 viq Yeah, I probably would need to enumerate them at some point. So, top.sls, or ifs inside the pillar file?
12:30 viq I guess ifs sound better...
12:30 babilen I would probably write that pillar in Python and generate suitable dictionaries for each host based on some more expressive datastructure.
12:31 CeBe joined #salt
12:31 viq 1) I don't know python enough for that. 2) I doubt rest of my team knows python enough either
12:31 babilen pity, it would allow you to get around a horrible Jinja mess
12:32 viq Though maybe it could be the time to introduce mako?
12:32 babilen I mean you could simply maintain sets of minions that should be assigned a given grain (or a dictionary of grains to which you add minions) and return a suitable dictionary
12:32 phx another question. policy says leavers' accounts have to be removed. i can propagate user account to unix systems, tried that and works, however how would i ensure that, let's say, accounts above uid10000 not mentioned in the users state are removed?
12:32 viq Or maybe time to look into reclass...
12:32 babilen It's really quite trivial Python code .. If you have some minutes I wouldn't mind writing the basic idea of that down
12:32 viq babilen: yeah, in no hurry here
12:34 viq phx: I'm not sure you can be that specific... Though if you start with all accounts specified in salt, you just modify user.present into user.absent once person leaves
12:35 phx maybe i can do that
12:35 phx think we're going to mark leaves in our identity management stuff, so i can generate the users.absent part
12:41 tkharju4 joined #salt
12:42 babilen viq: Would you prefer to enumerate all grains for each minion or list all minions for each grain?
12:42 phx http://docs.saltstack.com/en/latest/ref/states/top.html says if i have multiple environments then the top.sls should reference them. but in case of multiple environments, where should be the place of the top.sls?
12:44 viq babilen: I think list minions for grain
12:44 viq phx: that's what babilen was talking about. If you have top.sls in each branch, they will get merged, so it could be less confusing to have an external one
12:45 viq babilen: currently I am thinking of just a couple grains, and have a bunch of minions
12:45 babilen phx: You also want to be able to still merge between those branches (standard git workflow) and if you have different top.sls files in each it'll get ugly.
12:47 bhosmer joined #salt
12:47 ndrei joined #salt
12:48 Shish joined #salt
12:48 phx that reads, 1) it doesn't matter where i put the top file in case of multiple envs 2) if i have multiples they get merged
12:49 joehh sander^work: the key dependency to go back to debian 5 is python2.6
12:49 viq phx: kinda, yeah, though not sure what would happen if base (master) wouldn't have one
12:49 viq And not sure if it even matters
12:49 phx sounds like to avoid 2, i'd just make a separate pseudo-env for the tops file, which's only purpose of existence is to avoid confusion
12:49 tkharju joined #salt
12:50 joehh there have been reports of someone running back that far, but everything had to be compiled from scratch
12:51 pressureman joined #salt
12:51 babilen viq: http://paste.debian.net/106843/ (untested and ugly, but you get the idea)
12:52 mprymek joined #salt
12:52 bensons_ hey, stupid question but how can i cmd.wait based on a package installation? there is no stateful argument to the pkg state, right?
12:52 viq babilen: what would it look like if I wanted to set eg backupserver:backup1.example.com ?
12:53 mprymek hi, is there a way to let salt-master listen on _both_ ipv4 and ipv6?
12:53 viq bensons_: I think there is, it would run the command only once after the package got installed (or after each time it got updated if you used pkg.latest)
12:53 bensons_ ah no more, had a typo...
12:53 bensons_ viq: yes, my fault, sorry...
12:53 mprymek if I use ipv6: True, I must set interface also...
12:53 babilen viq: http://paste.debian.net/106844/
12:55 viq babilen: doesn't that set empty grain 'backupserver' on 'backup1.example.com' ?
12:57 babilen viq: exactly
12:57 babilen Ah, that is not what you want ?!
12:57 * babilen chuckles
12:57 babilen I need better specifications!
12:57 viq babilen: and I want to set a grain backupserver with value backup1.example.com on nodes node1 node2 node3
12:58 BrendanGilmore joined #salt
12:59 babilen Ah, yeah, naturally. Yeah, the code there is not able to handle that. You would need an additional level in the dictionary to enumerate the values of the grains.
13:00 kivihtin joined #salt
13:00 blarghmatey joined #salt
13:01 non7top joined #salt
13:03 brandon joined #salt
13:04 sxar joined #salt
13:06 sxar joined #salt
13:08 sxar_ joined #salt
13:10 sxar__ joined #salt
13:10 racooper joined #salt
13:11 miqui joined #salt
13:11 viq hm, can I get a jinja variable from a pillar? Would it be something like {{ {{ salt['pillar.get']('backupserver') }} }} ?
13:11 ndrei_ joined #salt
13:12 phx when i specify environments in file_roots, do i also need to specify them accordingly in pillar_roots?
13:12 babilen viq: No need for the nested {{ }}
13:12 sxar joined #salt
13:13 luette1 joined #salt
13:13 babilen viq: http://paste.debian.net/106851/ is one way .. but for more development I'll charge some beers ;)
13:15 viq babilen: thanks :)
13:15 sxar joined #salt
13:15 rojem joined #salt
13:16 viq babilen: but that would just get me the value of grain, I want to put there contents of jinja variable (which is where I put the backup server's ssh key) that is named the same as the value of that grain
13:16 babilen viq: In fact I don't really like that, I would personally prefer nested dictionaries ..
13:17 babilen ...
13:17 sxar joined #salt
13:17 babilen You are just like every other customer ;)
13:17 viq (re: no need for nested {{ }} )
13:17 phx top.sls also can be a python script, right?
13:18 babilen "Here, I'm done with your project. It does what you wanted to do" ... "yeah, but we actually want it to do something else. Please write it again"
13:18 stevednd viq: {% set yourvar = salt['pillar.get']('backupserver') -%}
13:18 phx speaking of states' top.sls
13:18 viq babilen: http://blog.thingsdesigner.com/uploads/id/tree_swing_development_requirements.jpg  :P
13:19 babilen viq: Oh, that's funny. A colleague just hung that poster in our office
13:20 sander^work joehh, thanks!
13:20 viq stevednd: I was thinking of putting ssh public keys of backup servers in jinja variables, and then putting them on server depending on grain which says which backupserver they should be backed up by
13:22 babilen viq: Wait, so you don't really need the whole "set arbitrary grains in pillars", but you simply want a mapping of hosts → backupsserver ?
13:23 blarghmatey joined #salt
13:23 stevednd viq: are your pubkeys in pillars?
13:23 viq babilen: well, that's the first use of such. I also want to set (in grains? in pillars? Somewhere) upgrade policy for hosts
13:23 stevednd if they aren't, I would suggest putting them there
13:23 JoeHazzers joined #salt
13:23 viq stevednd: not yet, I guess could be
13:24 stevednd then your jinja shouldn't end up as ugly
13:24 dude051 joined #salt
13:25 JoeHazzers how can i compose a file based on other states, like having a "logger" and then iterating over all declared "loggers" to compose a file?
13:25 blarghmatey joined #salt
13:25 viq {{ salt['pillar.get']('grains[backupserver]') }} ?
13:27 stevednd iterate salt['grains.get']('backupservers') and just add an entry for salt['pillar.get']('backupservers:keys', {}).get(backupserver_grain_value)
13:27 rudolfsteiner joined #salt
13:27 stevednd viq: if you only have one, and that is exactly where you stored it, then your way would work
13:28 viq stevednd: I have multitple, but a host will have only one assigned
13:29 sxar joined #salt
13:29 babilen viq: Okay, I would want to write more robust Python code for that ...
13:29 stevednd viq: yeah, that should be fine then, assuming that you store the entire pillar path in the grain. I would probably just store the server name there, and in the statefile add the rest of the pillar path
13:29 mgw joined #salt
13:30 viq mhmm
13:31 scott_w joined #salt
13:32 oz_akan_ joined #salt
13:33 joehh sander^work: some very old notes on installing (old) salt on lenny
13:33 joehh https://gist.github.com/justone/3210047#file-salt_install-sh
13:33 viq hm, backupserver:keys:backupserver1.example.com holding public ssh key for backupserver1, and then in ssh_auth.present use -name: {{ salt['pillar.get']('backupservers:keys' .... erm, what here?
13:33 joehh not sure how they would go today (and haven't tried them myself either)
13:34 joehh you would want to do zmq 3.2+ now
13:34 jdmf joined #salt
13:35 stevednd salt['pillar.get']('backupservers:keys', {}).get(backupserver_grain_value)
13:35 brandon joined #salt
13:36 stevednd viq: you can just substitute grains['backupserver'] or whatever you call it for backupserver_grain_value
13:36 viq and the backupserver_grain_value will get the value, or do I have to define/reference it somehow?
13:36 viq ah, thank you
13:37 viq so salt['pillar.get']('backupservers:keys', {}).get(grains['backupserver']
13:37 to_json joined #salt
13:38 stevednd I was using backupserver_grain_value before because I was iterating over the grains before, so it would have been a variable. Since you only have a single value, you're good with just sticking the grain call there
13:38 viq sweet, thank you
13:38 stevednd np
13:42 non7top hello all, how can I make this code work: {% if 0 == salt['cmd.retcode'](template=jinja, "test -f /etc/nginx/{{ dir }}") %}
13:43 zain_ joined #salt
13:46 ekristen joined #salt
13:47 ggoZ joined #salt
13:48 mapu joined #salt
13:49 stevednd non7top: if this is already in jinja why are you templating it? ("test -f /etc/nginx/" + dir) should work just fine
13:49 vejdmn joined #salt
13:50 tyler-baker joined #salt
13:50 indomyas joined #salt
13:51 indomyas hey there my windows minion wont send a key to my ubuntu server, any suggestion why?
13:51 non7top stevednd: thank you, that worked as expected.
13:52 stevednd non7top: np
13:52 kaptk2 joined #salt
13:52 aquinas joined #salt
13:53 indomyas ping works, c++ installed
13:53 luette joined #salt
13:54 non7top salt-minion -l debug
13:57 JoeHazzers what would happen if i had two pillar files included on a minion, one of which contains packages: x: y and the other packages: a: b
13:57 JoeHazzers would they merge, or would one of them be inaccessible?
13:58 nkuttler JoeHazzers: what do you mean? can you paste an actual example? (or, just try it?)
13:59 nkuttler JoeHazzers: hm, you might also want to look into pillars, and grains, if you haven't yet
13:59 aquinas joined #salt
13:59 JoeHazzers https://gist.github.com/JoeHazzers/b2e55fe16724cf1cf54d
13:59 nkuttler JoeHazzers: but you can include from multiple state files, if that's what you mean, sure
14:00 nkuttler sure
14:00 quickdry21 joined #salt
14:00 JoeHazzers the expected result from pillar would be to have both packages.httpd and packages.mysql available, no?
14:00 pdayton joined #salt
14:00 Sr_Guzman joined #salt
14:01 JoeHazzers as for the other situation
14:01 emilisto_ left #salt
14:02 nkuttler JoeHazzers: yes, but your state syntax looks wrong to me
14:02 nkuttler JoeHazzers: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkg.html
14:02 JoeHazzers that was meant to be pillar...
14:02 nkuttler doh! :)
14:03 nkuttler well, then everything depends on your states
14:03 JoeHazzers what i'm asking is does pillar merge or just pick whatever's first?
14:03 nkuttler ok, sorry
14:04 JoeHazzers as for my earlier issue
14:05 JoeHazzers since for example, an apache.sls might contain packages to install, configuration files, and services, isn't it reasonable that i might want a "logged file" in there as well?
14:05 JoeHazzers for which, i need to compose a file later on
14:05 indomyas thanks for the help, i set up everything in the guide to install firefox on my minion, but it says "unable to cache file '...' from saltenv base
14:06 JoeHazzers https://github.com/saltstack/salt/issues/7302
14:06 JoeHazzers > closed november 2013
14:06 JoeHazzers this is crippling.
14:08 tyler-baker joined #salt
14:10 rojem joined #salt
14:17 Rojematic joined #salt
14:18 JoeHazzers i think you can pretty much see my use case here: https://gist.github.com/JoeHazzers/82f8279d1de8a839f465
14:18 aquinas joined #salt
14:20 rr123 joined #salt
14:20 rr123 left #salt
14:27 smcquay joined #salt
14:29 dangra joined #salt
14:29 phx i'm getting some error when i'm trying to distribute an /etc/motd to a box in an environment, could someone please take a look: http://pastebin.com/mEQGce9Z (sls files, tops, files, command, error message, etc)
14:31 JoeHazzers take out the - in front of file.managed
14:31 JoeHazzers it's meant to be a dictionary, not a list.
14:33 ajprog_laptop joined #salt
14:37 phx thanks
14:37 jslatts joined #salt
14:37 phx and regarding the command, if i say ``states.sls defaults'' it fails to find the state, however i define in the top.sls that this state for this box is in the dev environment, why isn't that taken into account?
14:38 felskrone joined #salt
14:41 march_rabbit joined #salt
14:41 march_rabbit hi all
14:42 mgw joined #salt
14:42 march_rabbit i have a question: is it possible to find out IP address of some Salt minion from master side?
14:42 march_rabbit let's say i know its ID
14:42 bmatt get the ipv4 grain
14:42 JoeHazzers phx, http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.sls
14:42 DaveQB joined #salt
14:42 JoeHazzers the documentation shows the default env for that command is "base"
14:44 march_rabbit bmatt, thanks. is it information from minion itself or from master underlying transport?
14:44 bmatt grain info is intrinsic to the minion
14:44 march_rabbit i want to know external IP of a minion in case of any NATs between minion and master
14:44 bmatt oh, yeah, no, a minion couldn't know that
14:44 march_rabbit yea
14:45 phx JoeHazzers, thanks. and how can i push all applied states for a given node?
14:45 bmatt a master couldn't either
14:45 march_rabbit i think it's a way to zeromq?
14:45 JoeHazzers phx, do you mean state.highstate?
14:46 phx JoeHazzers, thanks. sorry for my noobish questions, i'm still learning salt
14:46 millz0r joined #salt
14:46 TyrfingMjolnir_ joined #salt
14:48 ggoZ So whats going on in the salt planet nowadays?
14:55 abe_music joined #salt
14:56 aquinas joined #salt
14:57 phx and how do i put salt.modules.timezone.set_zone(x) into a sls file?
14:58 racooper to those who were trying to help with my gitfs pillar issue yesterday...it's fixed.  I totally missed the pillar gitfs cache; it had cached the wrong URI so my changes in master config were never having any affect.
14:59 sander^work joehh, awsome, i'll try it out :-)
14:59 dvogt joined #salt
15:00 felskrone joined #salt
15:00 JoeHazzers phx, http://docs.saltstack.com/en/latest/ref/states/all/salt.states.timezone.html
15:00 phx thanks
15:02 aquinas joined #salt
15:06 will___ joined #salt
15:06 ramishra joined #salt
15:09 ndrei joined #salt
15:09 ndrei_ joined #salt
15:09 phx node states are cached on the master node, right? i've reinstalled my test minion node, issued state.highstate and nothing happened
15:09 phx also i'm sure to got a new key for itself, however salt-key -L still lists its old key, and nothing of the reinstalled one
15:10 conan_the_destro joined #salt
15:10 LordMike91 joined #salt
15:11 jalbretsen joined #salt
15:13 JoeHazzers If it's got a new key under the same name/id, delete the old key first.
15:13 sxar joined #salt
15:13 aquinas__ joined #salt
15:14 aquinas__ joined #salt
15:15 rojem joined #salt
15:18 malinoff joined #salt
15:19 laubosslink joined #salt
15:19 mapu Good morning. I am trying to creat an if clause based on a grain ‘role’. There are multiple values for ‘role’. My if statement is {% if grains[‘role’] == ‘build’%}
15:20 mapu but it does not seem to pick the role up a one of the multiple values
15:20 mapu I beleive I have the wrong comparison?
15:20 rallytime joined #salt
15:21 UtahDave joined #salt
15:21 Lomithrani joined #salt
15:23 mapu laubosslink: trying that now
15:27 nicksloan I need to execute a query to create an extension in postgres, and apparently the postgres_extension.present state isn't released yet. How should I execute any random SQL query?
15:27 nicksloan I noticed there is a salt.modules.postgres.psql_query
15:27 mapu laubosslink: that did it. silly brain fart on my part
15:27 Karunamon So here's an interesting one.. Say you include a state file that has a syntax error somewhere in it. Would that error cause salt to stop parsing the top level file that does the include and act as if it doesn't exist?
15:27 nicksloan can I use that from an sls?
15:27 mapu iteration is my friend.
15:28 Karunamon In my case, it looks like: top.sls -> commonpackages (commonpackages.sls) -> includes a whole bunch of individual sls files relating to package installation
15:28 Karunamon except now i'm getting errors saying 'commonpackages' doesn't exist.
15:28 timoguin nicksloan: look at the module.run/module.wait states
15:29 nicksloan timoguin: will do. Thanks.
15:29 nicksloan Also, on rtfd, is there a way to specify docs for a particular version?
15:30 n8n joined #salt
15:32 n8n joined #salt
15:33 LordMike91 Is there a way to prevent 3-4 apt-get update's in a salt.highstate? :P
15:34 ecdhe joined #salt
15:35 LordMike91 I think I'm going to simplify my apt-sources
15:35 Karunamon ah, not a syntax error but a missing file
15:36 Karunamon wonder if I should report that one as a bug? The error message you get if you do that is kind of misleading
15:37 felskrone joined #salt
15:38 ramishra joined #salt
15:39 TyrfingMjolnir joined #salt
15:39 mgw joined #salt
15:41 TyrfingMjolnir joined #salt
15:41 bemehow_ joined #salt
15:45 sxar joined #salt
15:47 sxar joined #salt
15:47 LordMike91 Does anyone have debian repositories configured through salt?
15:47 ml_1 joined #salt
15:49 thayne joined #salt
15:50 sxar joined #salt
15:53 sxar joined #salt
15:53 sxar joined #salt
15:54 sxar joined #salt
15:56 sxar joined #salt
15:56 tkharju1 joined #salt
15:57 Vye joined #salt
15:58 felskrone joined #salt
15:58 tkharju2 joined #salt
15:59 UtahDave laubosslink: I'm not sure what you mean.  Are you wanting to query another vm?
15:59 Lomithrani joined #salt
16:00 bhosmer_ joined #salt
16:01 _ale_ joined #salt
16:01 zain_ joined #salt
16:03 kermit joined #salt
16:03 ajolo joined #salt
16:04 rallytime joined #salt
16:04 babilen LordMike91: I am one of those people
16:04 babilen .. so: yes
16:05 ipalreadytaken joined #salt
16:06 tligda joined #salt
16:08 tkharju3 joined #salt
16:09 scoates joined #salt
16:09 LordMike91 Can it really be, that each pkrepo state invokes 2 apt-get updates?
16:10 UtahDave laubosslink: You can use the "peer interface" to allow a minion to run a command on another one for live data, you can also use the Salt Mine to cache the results of a query that are then available to the other minions.
16:10 LordMike91 babilen?
16:10 tkharju4 joined #salt
16:11 KyleG joined #salt
16:11 KyleG joined #salt
16:11 babilen LordMike91: That might be the case, I can neither verify nor falsify that right now
16:12 LordMike91 Currently it's nuts.. I have 9 pkrepo's, meaning I'd be doing 18 apt-get updates for a highstate... Checking to see if it's only on a change in a moment
16:12 LordMike91 But how do you define the repos - without a key I presume?
16:13 meteorfox joined #salt
16:14 tkharju joined #salt
16:16 kballou joined #salt
16:17 bhosmer_ joined #salt
16:18 ipalreadytaken joined #salt
16:22 ckao joined #salt
16:23 joehillen joined #salt
16:24 Ryan_Lane joined #salt
16:26 shaggy_surfer joined #salt
16:27 LordMike91 Oh great, I fixed my issue.. On every run apt-get was run twice for each pkgrepo. Turns out I had a file.recurse on /etc/apt/sources.list.d/ with clean: True
16:28 ramishra joined #salt
16:29 E1NS joined #salt
16:30 dude051 joined #salt
16:30 schimmy joined #salt
16:32 schimmy1 joined #salt
16:34 possibilities joined #salt
16:34 nickg i just upgraded to 2014.1.5 and for some reason now salt is not happy with my top.sls file.   On the minion debug I see: TypeError encountered executing state.highstate: 'bool' object is not iterable. See debug log for more info.
16:35 viq nickg: https://github.com/saltstack/salt/issues/13535
16:35 ramteid joined #salt
16:35 viq nickg: specifically https://github.com/saltstack/salt/issues/13535#issuecomment-46736029
16:35 ksalman Salt-master reports that state i deployed on a minion finished. But on the minion side if i then do a 'salt-call state.highstate' or anything else, it complains that function is running as PID XXXX ... with jid XXXXXXX, and it seems to get stuck there until i restart salt-minion. This is on Windows by the way
16:35 viq nickg: either apply that patch, or update your minions to 2014.1.5
16:37 tkharju joined #salt
16:37 bemehow joined #salt
16:39 sxar joined #salt
16:39 diegows joined #salt
16:39 alanpearce joined #salt
16:40 nickg viq thanks.  i just installed the salt-minion this morning via ubuntu's apt-get.. i guess the minion repo is out of date
16:41 yetAnotherZero joined #salt
16:41 schimmy joined #salt
16:42 thayne joined #salt
16:43 juanlittledevil left #salt
16:45 davet joined #salt
16:45 allanparsons joined #salt
16:46 davet1 joined #salt
16:51 TyrfingMjolnir joined #salt
16:51 jslatts joined #salt
16:52 wendall911 joined #salt
16:52 forrest joined #salt
16:54 schimmy joined #salt
16:55 racooper trying to get quotas working from state via module.run. I can't figure out the proper syntax. https://gist.github.com/racooper/970be82b500a5c044afb
16:58 ramishra joined #salt
16:59 kermit joined #salt
17:05 tmanthemaster joined #salt
17:05 shaggy_surfer joined #salt
17:06 tmanthemaster left #salt
17:06 druonysus joined #salt
17:10 mapu I need to ensure that several directories are absent- /tmp/pip-*  - is there a way to use wildcards to address these directories?
17:14 dstokes anybody knowledgeable on the ssl implementation in salt-api? getting a weird tls error when trying to connect w/ rest_cherrypy
17:14 forrest mapu, while file.recurse can ensure the directories are absent, I wouldn't use it like that. I'd suggest just to do a cmd.run.
17:15 forrest mapu, did you already try doing a file.absent with the wildcard matching?
17:15 mapu Yes- I had this:
17:15 mapu file.absent
17:15 mapu whops
17:16 mapu had /tmp/pip-*:
17:16 forrest ahh, no dice?
17:16 mapu file.absent
17:16 mapu no
17:18 racooper you said the pip-* were directories, or are they files?
17:18 forrest mapu, yea I'd just use cmd.run then, I don't think file.absent was intended for wildcard matching, or if it is, I've never seen an example work, because at that point salt would be managing multiple items
17:19 mapu Ok- thanks. I tend to try things the “salty” way first, then drop to cmd.run if I can’t get it going
17:20 forrest mapu, yea totally understandable
17:20 jaimed joined #salt
17:20 forrest it's pretty rare that I have to resort to using cmd.run, but in cases like this it makes sense.
17:21 JoeHazzers don't forget to put an 'unless' argument on it ;3
17:24 ekristen joined #salt
17:24 mapu excellent- that worked.
17:24 forrest cool
17:24 dstokes ssl error w/ salt-api seems like it might be related to cherrypy, and by extension, pyopenssl..
17:24 forrest dstokes, noooooooo
17:24 Ryan_Lane the CSS on readthedocs for salt is all messed up now
17:25 Ryan_Lane (please, please let's kill readthedocs)
17:25 dstokes forrest: heading to oftc (irc) net to configm
17:25 dstokes confirm*
17:25 forrest Ryan_Lane, I thought we already killed read the docs
17:25 Ryan_Lane nope
17:25 forrest dstokes, cool
17:25 forrest dstokes, let Whiteinge know as well if that's the case would you?
17:25 Ryan_Lane the one plus of readthedocs is that you can see the docs specific to a version of salt, rather than always seeing the unreleased version's docs
17:25 dstokes forrest: will do
17:26 forrest Ryan_Lane, yea I believe whiteinge was working on fixing that though
17:26 forrest so the main docs had multiple versions available
17:26 * Ryan_Lane nods
17:26 forrest that reminds me, what is your opinion on the new changes to the code boxes?
17:26 forrest a coworker of mine made the change, I need to go through and find all the screw ups now though :\
17:26 ksalman can one not have jinja in the sls files for winrepo?
17:27 Ryan_Lane I kind of like it. you need to force refresh the page to see it properly, though
17:27 forrest Ryan_Lane, yea caching
17:27 Ryan_Lane otherwise the part that says "python" slightly sits over the code
17:27 Ryan_Lane I like it, though
17:27 forrest Ryan_Lane, cool
17:27 Ryan_Lane there isn't enough visual distinction in the docs, for the most part
17:27 Ryan_Lane and this helps
17:27 forrest yea that was the whole point when I talked to our front end guy
17:27 forrest he didn't like it when he was looking at the docs
17:28 forrest ksalman, I don't know. Did you try it and it didn't work?
17:28 Ryan_Lane the docs have a *lot* of text, so making it easier to distinguish things helps
17:29 Ryan_Lane it would also be really nice for the docs to have sections (with headings)
17:29 forrest Ryan_Lane, what do you mean?
17:29 Ryan_Lane for instance: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.s3.html the configuration and depends sections blend in with the functions
17:30 forrest oh yea I think that's just a failing of how that section was written
17:30 n8n joined #salt
17:30 ksalman forrest: yea
17:30 ksalman found character '%' that cannot start any token
17:30 Ryan_Lane there's not really any good guidelines for developers when writing the docs
17:30 forrest ksalman, ahh, in that case I'd ask UtahDave when he shows up, maybe there is a hack, or something you have to do
17:30 Ryan_Lane no style guide, no review, no quick reference for the syntax or macros
17:30 ksalman https://gist.github.com/anonymous/daf0a66a943b92526fd3
17:30 forrest Ryan_Lane, Seth wrote a page for that
17:31 Ryan_Lane oh?
17:31 ksalman okay
17:31 forrest Ryan_Lane, yea I'm trying to remember where it's at.
17:31 forrest I thought I made an issue to get it linked on the contributing docs
17:31 forrest hmm
17:34 bhosmer joined #salt
17:34 forrest Ryan_Lane, http://docs.saltstack.com/en/latest/topics/development/conventions/documentation.html
17:34 forrest there we go
17:35 Ryan_Lane sweet
17:35 forrest yea it needs to be linked in the contributing doc badly though
17:35 Ryan_Lane yeah
17:35 forrest I thought I made an issue for it
17:35 pdayton joined #salt
17:35 forrest but I must not have
17:37 jslatts joined #salt
17:37 * timoguin really needs to put this english degree to use on some docs
17:37 timoguin so much too do that it's hard to focus on the small tasks
17:38 forrest timoguin, look at issue 12446
17:38 forrest and just pick something
17:38 forrest lol
17:38 timoguin yes yes i know
17:39 timoguin i get overwhelmed thinking about overall structural improvements
17:39 timoguin it's a problem
17:40 Ahlee anybody got a clever way to query all non-stock grains on all minions?
17:40 luminous joined #salt
17:40 forrest Ahlee, for loop?
17:40 Ahlee i guess i can just get a list of stock grains, get a list of all grains, then just purge the stock
17:40 forrest timoguin, yea analysis paralysis
17:41 forrest timoguin, we'd need like a weeklong sprint to get the structure fixed.
17:41 n8n joined #salt
17:41 timoguin yea agreed
17:42 luminous hello! I need a little assistance understanding http://salt-zh.readthedocs.org/en/latest/ref/states/all/salt.states.module.html#module-salt.states.module - I'd like to run a salt module (tar) if another state (the source file) changed - which of these three should I be using?
17:44 Ryan_Lane forrest: ^^ see? :)
17:44 luminous yea, that isn't a great/clear doc..
17:44 Ryan_Lane well, it's a site that's deprecated
17:44 forrest ugh
17:45 forrest read the docs needs to diaf
17:45 luminous what's wrong with rtd?
17:45 forrest well, I have my own beef with them
17:45 forrest but that doesn't apply here
17:45 Ryan_Lane luminous: the correct site is docs.saltstack.org
17:45 forrest specifically in the context of salt
17:45 luminous does docs.saltstack.org have versions now?
17:45 luminous all versions..
17:45 forrest Ryan_Lane, that redirects to the .com now as a heads up
17:45 Ryan_Lane ah
17:46 Ryan_Lane luminous: it doesn't, unfortunately
17:46 luminous so then that's useless to me :)
17:46 luminous I don't run the latest 'stable'
17:46 forrest YOU'RE USELEss :'(
17:46 luminous hah
17:46 luminous do you really run the latest in production forrest ?
17:47 forrest luminous, no
17:47 luminous hah
17:47 luminous so what do you use for docs?
17:47 forrest but the last few releases have been good about having 'added in' whatever version
17:47 forrest I use docs.saltstack.org
17:47 forrest or .com
17:47 forrest whatever
17:47 luminous how do you look up the docs for a specific version?
17:47 forrest I don't
17:47 luminous hah
17:47 luminous ooook
17:47 forrest I just go look at the docs for that state
17:48 Ryan_Lane I also use docs.ss.o and try to rely on docs having "added in"
17:48 luminous well either way.. back to the original question
17:48 luminous Ryan_Lane: what do you do for attributes that have changed..?
17:48 Ryan_Lane luminous: you want to use wait
17:48 forrest yep wait for sure
17:49 luminous that's what I thought, but I'm a bit confused without an example there
17:49 luminous is this the same as cmd.wait?
17:49 Ryan_Lane luminous: I'm actually using the develop branch, so the docs are accurate
17:49 nmadhok joined #salt
17:49 * Ryan_Lane doesn't use master/minion
17:49 nmadhok hello
17:49 nmadhok Hello @UtahDave
17:49 forrest nmadhok, I don't think he's around this morning
17:50 timoguin luminous: yea it's similar to cmd.wait
17:50 nmadhok whiteinge?
17:50 forrest maybe
17:50 timoguin forrest?
17:50 nmadhok i have a merge conflict with my pull request
17:50 forrest yes
17:50 nmadhok how do i fix that?
17:50 luminous nmadhok: learn git rebase :)
17:51 forrest nmadhok, are there actually line conflicts?
17:51 luminous Note that this function actually does nothing -- however, if the watch is satisfied, then mod_watch (defined at the bottom of this file) will be run. In this case, mod_watch is an alias for run(). <<< what does this mean?
17:51 forrest luminous, basically it's just run, but with watch tacked on
17:52 luminous ok, I think I can handle that
17:52 luminous thanks for clarifying!
17:52 timoguin that's a very develerop-y line
17:52 viq meh, ssh_auth.present should have a force flag, to remove all other keys
17:53 shaggy_surfer joined #salt
17:53 n8n joined #salt
17:54 AdamSewell joined #salt
17:54 AdamSewell joined #salt
17:54 viq https://github.com/saltstack/salt/issues/13340
17:54 acabrera_ joined #salt
17:54 luminous timoguin: It's terribly confusing :P
17:55 Ryan_Lane viq: :)
17:55 Ryan_Lane I think both absent and present should have a flag that ensures the state of the entire file
17:55 Ryan_Lane I don't know why every config management system handles ssh keys this way
17:55 luminous do you pass in the args to the module through **kwargs? or through name?
17:55 jslatts joined #salt
17:55 Ryan_Lane luminous: name
17:56 Ryan_Lane myargs=myval (from the cli)
17:56 Ryan_Lane myarg: myval (from states)
17:56 Ryan_Lane s/myargs/myarg/
17:57 luminous so something like: name: archive.tar arg: val a2: v2?
17:57 dstokes whiteinge: i'm getting nothing in oftc #cherrypy regarding ssl error (in salt-api). what version of c-py you using when testing ssl?
17:57 luminous that doesnt' seem right..
18:00 ramishra joined #salt
18:02 forrest viq, I actually wrote a mod to the users formula to do that
18:03 forrest except it's for authorized keys
18:03 forrest so it handles it with a managed file
18:03 forrest it's a shit hakc
18:03 forrest *hack
18:03 forrest but whatever
18:04 viq forrest: oh?
18:05 forrest yea
18:05 felskrone joined #salt
18:06 allanparsons joined #salt
18:07 kivihtin joined #salt
18:07 viq forrest: got it somewhere viewable?
18:07 dvogt joined #salt
18:07 forrest I don't, hang on a minute
18:07 ajolo joined #salt
18:07 aquinas joined #salt
18:09 tkharju1 joined #salt
18:10 nickg i'm having difficultiy configuring salt pkgrepo to use the datastax repository.  the only error I receive is State pkg.repo found in sls cassandra.repo is unavailable.  is there anyway to get more debug?
18:11 forrest viq, https://gist.github.com/gravyboat/638b69b90c010dbdf929
18:11 ipalreadytaken joined #salt
18:11 forrest viq, a lot of it is similar to the existing users one, it's just the ssh auth that is different
18:11 forrest nickg, did you laready check the apt/yum logs on the box?
18:12 nickg forrest: i dont see any output from apt on the system
18:12 forrest viq, you'll probably have to remove the indents for the jinja as a heads up
18:12 forrest I just haven't yet in my copy
18:12 forrest nickg, so has the repo been added in lists.d?
18:13 aquinas joined #salt
18:13 luminous is there a way to get a list of modules that salt 'sees' ?
18:13 viq forrest: thank you
18:13 forrest viq, yea np
18:13 nickg forrest: well, i assumed salt would create those settings for me. i guess i can configure salt to create those manually
18:13 forrest viq, let me know if you improve on it
18:14 forrest nickg, no it should if you're using pkg.repo, I'm just asking if it DID create them
18:14 forrest to confirm it's getting that far you know?
18:14 TyrfingMjolnir joined #salt
18:15 nickg forrest: ive found the issue
18:15 nickg thanks
18:15 forrest nickg, np, what was it?
18:16 nickg forrest: pkg.repo.managed != pkgrepo.managed
18:16 viq forrest: ok
18:16 forrest nickg, ahh yea that will do it
18:16 viq though for now it's time to call this a day
18:16 forrest viq, I updated the gist, so you'll see it if you refresh no
18:16 forrest *w
18:16 nickg forrest: yeah the error said pkg.repo i just didn't make the correlation
18:16 forrest nickg, gotcha
18:16 viq closed browser already :P
18:16 forrest lol ok
18:16 forrest did you fork it?
18:17 viq Not yet, want me to?
18:17 forrest no I don't care, just wanted to make sure you don't lose track ofi t
18:17 forrest *of it
18:17 forrest I won't delete it or anything
18:17 viq OK, thanks
18:17 viq Time to head out then
18:17 forrest yea np
18:17 viq thanks and cya
18:17 forrest have a good one
18:20 UtahDave smcquay: did whiteinge get you sorted out yesterday?
18:21 brendangilmor joined #salt
18:21 smcquay nobody can truly help me.
18:22 Ahlee UtahDave: ever had complaints of nssd freaking out on network sever?
18:22 taterbase joined #salt
18:22 Ahlee I've got ~100k log entries after what i can only assume is the minion lost communication with the master
18:23 sctsang joined #salt
18:23 ipalreadytaken joined #salt
18:24 UtahDave Ahlee: Hm. I haven't heard anything regarding that.
18:24 luminous is there a way to get a list of modules that salt 'sees' ?
18:25 Ahlee it's popping up occasionally across a variety of servers, first one lost it's ability to auth with AD, another couldn't access external databases
18:25 smcquay UtahDave: I double checked the master config; it's got nothing exotic, minons are likewise configged with the auth flood fix. still seeing timeouts during salt calls on the master. I try now and I can't really get anything to run; when I c-c it just says exiting gracefully.
18:25 UtahDave luminous: salt 'minion-id' sys.doc
18:25 UtahDave Ahlee: ah, are these on windows servers?
18:25 Ahlee this one no idea, just has hundreds and hundres of nssm http://chidjira02:8080/browse/CDELIVERY-397, i'll keep digging
18:25 Ahlee UtahDave: correct, sorry, nssm, not nssd
18:26 UtahDave Ahlee: yes, we did have an issue with that.  I just committed a fix yesterday.
18:26 tkharju2 joined #salt
18:26 Ahlee nice!
18:26 Ahlee Bug fix, so in next release?
18:26 Ahlee or for helium
18:26 Rebus There a way to manage paackage alternatives via salt, e.g. set default editor to vim-full or defualt pager to most ?
18:26 UtahDave https://github.com/saltstack/salt/pull/13738
18:27 UtahDave Ahlee: It will be in the next bugfix release and also helium.  We're looking to cut 2014.1.6 bugfix release soon.  hopefully in the next couple of days
18:28 Rebus nvm found the states module for alternatives
18:28 Ahlee UtahDave: Thanks.  Will hold off on upgrading until we see it
18:28 UtahDave ok, cool.
18:29 matrix3000 joined #salt
18:29 Ahlee UtahDave: in the mean time, do you know off hand if nssm is configurable to only try restarting N times before giving up?
18:29 smcquay UtahDave: I am clearing out our salt cache and trying to roll it up the hill again.
18:29 Ahlee i'd rather have a dead salt minion than nssm/salt-minion eatting 100% CPU trying to talk back to master
18:29 UtahDave Yeah, there are some arguments for that
18:30 ajolo_ joined #salt
18:30 UtahDave Ahlee: Yes, exactly. I'm trying to figure out the best way to manage that correctly with nssm.  Can't have this blowing up like that
18:32 maboum joined #salt
18:37 repl1cant joined #salt
18:40 austin987 joined #salt
18:40 alanpearce joined #salt
18:40 yomilk joined #salt
18:42 cliffstah joined #salt
18:43 martoss joined #salt
18:44 diegows is it possible to override the value of one pillar in one sls in ther sls
18:44 diegows for example
18:44 diegows pillar1.sls I have attr=value1
18:45 diegows then pillar2.sls attr=value2
18:45 diegows In top.sls I have
18:45 diegows match_condition:
18:45 diegows - pillar1
18:45 diegows - pillar2
18:46 diegows but the value of attr is value value1 (the only difference between both sls is that pillar2 has only one variable
18:46 diegows :(
18:46 ndrei joined #salt
18:47 martoss joined #salt
18:47 austin987 joined #salt
18:47 maboum joined #salt
18:47 matrix3000 joined #salt
18:47 tkharju2 joined #salt
18:47 sctsang joined #salt
18:47 taterbase joined #salt
18:47 aquinas joined #salt
18:47 allanparsons joined #salt
18:47 felskrone joined #salt
18:47 jslatts joined #salt
18:47 AdamSewell joined #salt
18:47 n8n joined #salt
18:47 shaggy_surfer joined #salt
18:47 luminous joined #salt
18:47 pdayton joined #salt
18:47 druonysus joined #salt
18:47 schimmy joined #salt
18:47 forrest joined #salt
18:47 wendall911 joined #salt
18:47 thayne joined #salt
18:47 yetAnotherZero joined #salt
18:47 sxar joined #salt
18:47 bemehow joined #salt
18:47 ramteid joined #salt
18:47 possibilities joined #salt
18:47 dude051 joined #salt
18:47 E1NS joined #salt
18:47 kballou joined #salt
18:47 meteorfox joined #salt
18:47 KyleG joined #salt
18:47 scoates joined #salt
18:47 tligda joined #salt
18:47 rallytime joined #salt
18:47 _ale_ joined #salt
18:47 Vye joined #salt
18:47 ecdhe joined #salt
18:47 jalbretsen joined #salt
18:47 LordMike91 joined #salt
18:47 conan_the_destro joined #salt
18:47 abe_music joined #salt
18:47 dangra joined #salt
18:47 quickdry21 joined #salt
18:47 kaptk2 joined #salt
18:47 vejdmn joined #salt
18:47 mapu joined #salt
18:47 to_json joined #salt
18:47 brandon joined #salt
18:47 scott_w joined #salt
18:47 JoeHazzers joined #salt
18:47 racooper joined #salt
18:47 ninkotech joined #salt
18:47 jrdx joined #salt
18:47 logix812 joined #salt
18:47 vu joined #salt
18:47 dober joined #salt
18:47 phx joined #salt
18:47 Striki joined #salt
18:47 jhauser joined #salt
18:47 GnuLxUsr joined #salt
18:47 jeddi joined #salt
18:47 z3uS joined #salt
18:47 sander^work joined #salt
18:47 analogbyte joined #salt
18:47 ph8 joined #salt
18:47 jalaziz_ joined #salt
18:47 kiorky joined #salt
18:47 garthk joined #salt
18:47 __alex joined #salt
18:47 jcsp joined #salt
18:47 jerrcs joined #salt
18:47 jnials joined #salt
18:47 blast_hardcheese joined #salt
18:47 sdebot joined #salt
18:47 Blacklite joined #salt
18:47 mrchrisadams joined #salt
18:47 gamingrobot joined #salt
18:47 gldnspud joined #salt
18:47 octarine joined #salt
18:47 Hollinski joined #salt
18:47 londo_ joined #salt
18:47 tinuva joined #salt
18:47 gothix joined #salt
18:47 borgstrom joined #salt
18:47 abele joined #salt
18:47 masterkorp joined #salt
18:47 rmnuvg joined #salt
18:47 geekmush joined #salt
18:47 chamunks joined #salt
18:47 herlo joined #salt
18:47 zartoosh joined #salt
18:47 googolhash joined #salt
18:47 perfectsine joined #salt
18:47 andabata joined #salt
18:47 yano joined #salt
18:47 toastedpenguin joined #salt
18:47 UForgotten joined #salt
18:47 zz_keyvan joined #salt
18:47 Tekni joined #salt
18:47 pfallenop joined #salt
18:47 jpaetzel joined #salt
18:47 davidone joined #salt
18:47 rjc joined #salt
18:47 johanek joined #salt
18:47 xmj joined #salt
18:47 ampex joined #salt
18:47 cruatta joined #salt
18:47 sashka_ua joined #salt
18:47 alikins joined #salt
18:47 Rebus joined #salt
18:47 Ch3LL_ joined #salt
18:47 MK_FG joined #salt
18:47 kedo39 joined #salt
18:47 hotbox joined #salt
18:47 InAnimaTe joined #salt
18:47 Heartsbane joined #salt
18:47 cwright joined #salt
18:47 benturner joined #salt
18:47 Sp00n joined #salt
18:47 artificialexit joined #salt
18:47 joehh joined #salt
18:47 cyrusdavid joined #salt
18:47 JPaul joined #salt
18:47 Ahlee joined #salt
18:47 zirpu joined #salt
18:47 baffle joined #salt
18:47 Sway|Away joined #salt
18:47 alekibango joined #salt
18:47 techdragon joined #salt
18:47 rogst joined #salt
18:47 gfa joined #salt
18:47 APLU joined #salt
18:47 rglen joined #salt
18:47 penguin_dan joined #salt
18:47 redondos joined #salt
18:47 flebel joined #salt
18:47 msciciel_ joined #salt
18:47 totte joined #salt
18:47 Hazelesque joined #salt
18:47 ahale joined #salt
18:47 \ask joined #salt
18:47 chitown__ joined #salt
18:47 thehaven joined #salt
18:47 tedski joined #salt
18:47 crazysim joined #salt
18:47 funzo joined #salt
18:47 dstokes joined #salt
18:47 djinni` joined #salt
18:47 terinjokes joined #salt
18:47 individuwill joined #salt
18:47 dcolish joined #salt
18:47 Corey joined #salt
18:47 [M7] joined #salt
18:47 twoflowers joined #salt
18:47 icebourg joined #salt
18:47 babilen joined #salt
18:47 agronholm joined #salt
18:47 gwmngilfen|afk joined #salt
18:47 berto- joined #salt
18:47 Dattas joined #salt
18:47 trevorj joined #salt
18:47 bezaban joined #salt
18:47 Kelsar joined #salt
18:47 Teknix joined #salt
18:47 svx joined #salt
18:47 bretep joined #salt
18:47 retr0h joined #salt
18:47 dancat joined #salt
18:47 eculver joined #salt
18:47 twiedenbein joined #salt
18:47 platforms joined #salt
18:47 logandg joined #salt
18:47 aarontc joined #salt
18:47 scarcry joined #salt
18:47 jpcw joined #salt
18:47 Bosch[] joined #salt
18:47 Guest17816 joined #salt
18:47 toddnni joined #salt
18:47 Jahkeup joined #salt
18:47 codekobe joined #salt
18:47 rlarkin joined #salt
18:47 dpac|away joined #salt
18:47 NV joined #salt
18:47 [vaelen] joined #salt
18:47 dccc joined #salt
18:47 Doqnach joined #salt
18:47 sindreij joined #salt
18:47 patrek joined #salt
18:47 schristensen joined #salt
18:47 nkuttler joined #salt
18:47 scalability-junk joined #salt
18:47 cro joined #salt
18:47 Whissi joined #salt
18:47 zemm joined #salt
18:47 anteaya joined #salt
18:47 packeteer joined #salt
18:47 eofs joined #salt
18:47 Yoda-BZH joined #salt
18:47 cofeineSunshine joined #salt
18:47 JordanTesting joined #salt
18:47 programmerq joined #salt
18:47 intr1nsic joined #salt
18:47 iMil joined #salt
18:47 errr joined #salt
18:47 drogoh joined #salt
18:47 sgate1_ joined #salt
18:47 jamesog joined #salt
18:47 georgemarshall joined #salt
18:47 rcsheets joined #salt
18:47 thunderbolt joined #salt
18:47 mihait joined #salt
18:47 terminalmage joined #salt
18:47 sag47 joined #salt
18:47 huleboer joined #salt
18:47 nicksloan joined #salt
18:47 Aletheuo joined #salt
18:47 vnico joined #salt
18:47 jmccree joined #salt
18:47 jesusaurus joined #salt
18:47 vandemar joined #salt
18:47 marcinkuzminski joined #salt
18:47 Fa1lure joined #salt
18:47 jab416171 joined #salt
18:47 pmcg joined #salt
18:47 t0rrant joined #salt
18:47 ilako joined #salt
18:47 wigit joined #salt
18:47 rigor789|away joined #salt
18:47 mephx joined #salt
18:47 dimeshake joined #salt
18:47 bejer joined #salt
18:47 ldlework joined #salt
18:47 mackstick joined #salt
18:47 Flusher joined #salt
18:47 canci joined #salt
18:47 bigl0af joined #salt
18:47 dwfreed joined #salt
18:47 eightyeight joined #salt
18:47 lz-dylan joined #salt
18:47 rnts joined #salt
18:47 bdf_ joined #salt
18:47 Valdo joined #salt
18:47 manfred joined #salt
18:47 shano joined #salt
18:47 mariusv joined #salt
18:47 balltongu_ joined #salt
18:47 ze- joined #salt
18:47 tmmt joined #salt
18:47 Sway joined #salt
18:47 hillna_ joined #salt
18:47 delkins joined #salt
18:47 AlcariTheMad joined #salt
18:47 etw joined #salt
18:47 whitepaws joined #salt
18:47 _blackjid joined #salt
18:47 neilf_ joined #salt
18:47 Sypher joined #salt
18:47 agh joined #salt
18:47 chutzpah joined #salt
18:47 CaptTofu joined #salt
18:47 codekoala joined #salt
18:47 kaictl joined #salt
18:47 jY joined #salt
18:47 copelco joined #salt
18:47 Eugene joined #salt
18:47 madduck joined #salt
18:47 xenoxaos joined #salt
18:47 toddejohnson joined #salt
18:47 stevednd joined #salt
18:47 bfwg joined #salt
18:47 cedwards joined #salt
18:47 jacksontj joined #salt
18:47 Voziv joined #salt
18:47 twobitsprite joined #salt
18:47 pviktori joined #salt
18:47 lazybear joined #salt
18:47 ifur joined #salt
18:47 bernieke joined #salt
18:47 moos3 joined #salt
18:47 melinath joined #salt
18:47 stotch joined #salt
18:47 hardwire joined #salt
18:47 mirko joined #salt
18:47 kula joined #salt
18:47 nickg joined #salt
18:47 bensons_ joined #salt
18:47 Jarus joined #salt
18:47 eliasp joined #salt
18:47 jbub joined #salt
18:47 cb joined #salt
18:47 Nazzy_ joined #salt
18:47 Sacro joined #salt
18:47 xintron joined #salt
18:47 hopthrisC joined #salt
18:47 hoodow joined #salt
18:47 rpotucek joined #salt
18:47 nadley joined #salt
18:47 drags joined #salt
18:47 steveoliver joined #salt
18:47 timoguin joined #salt
18:47 lude joined #salt
18:47 brewmaster joined #salt
18:47 uzomg joined #salt
18:47 btorch joined #salt
18:47 cwyse joined #salt
18:47 Dinde joined #salt
18:47 wm-bot4 joined #salt
18:47 jcristau joined #salt
18:47 mortis__ joined #salt
18:47 Ymage_ joined #salt
18:47 lyddonb_ joined #salt
18:47 jjasinski joined #salt
18:47 ashb joined #salt
18:47 arapaho joined #salt
18:47 snoozer joined #salt
18:47 Ssquidly joined #salt
18:47 xt joined #salt
18:47 fxhp joined #salt
18:47 Schmidt joined #salt
18:47 jcockhren joined #salt
18:47 st0newa11 joined #salt
18:47 amontalban joined #salt
18:47 tru_tru joined #salt
18:47 maber joined #salt
18:47 darrend joined #salt
18:47 oeuftete joined #salt
18:47 dotplus joined #salt
18:47 rockey joined #salt
18:47 ksk joined #salt
18:47 kuffs joined #salt
18:47 al joined #salt
18:47 devx joined #salt
18:47 v0rtex joined #salt
18:47 esogas joined #salt
18:47 ange joined #salt
18:47 micko joined #salt
18:47 robawt joined #salt
18:47 nahamu joined #salt
18:47 svs joined #salt
18:47 Daviey joined #salt
18:47 LordOfLA joined #salt
18:47 Xiao_ joined #salt
18:47 juice joined #salt
18:47 Spark joined #salt
18:47 crane joined #salt
18:47 mpoole joined #salt
18:47 dstanek joined #salt
18:47 andredieb joined #salt
18:47 jperras joined #salt
18:47 TaiSHi joined #salt
18:47 whyzgeek joined #salt
18:47 crashmag joined #salt
18:47 arnoldB joined #salt
18:47 jeblair joined #salt
18:47 DenkBrettl joined #salt
18:47 lynxman joined #salt
18:47 n0arch joined #salt
18:47 lahwran joined #salt
18:47 peno joined #salt
18:47 mfournier joined #salt
18:47 kamal_ joined #salt
18:47 Hipikat joined #salt
18:47 hhenkel joined #salt
18:47 mikkn joined #salt
18:47 MaZ- joined #salt
18:47 jakubek joined #salt
18:47 dh joined #salt
18:47 txmoose joined #salt
18:47 ede joined #salt
18:47 majoh joined #salt
18:47 tempspace joined #salt
18:47 jgelens joined #salt
18:47 mschiff joined #salt
18:47 fivethreeo joined #salt
18:47 ikanobori joined #salt
18:47 Heggan joined #salt
18:47 kevinbrolly joined #salt
18:47 Ixan joined #salt
18:47 supplicant joined #salt
18:47 honestly joined #salt
18:47 oc joined #salt
18:47 EntropyWorks joined #salt
18:47 ThomasJ|d joined #salt
18:47 fejjerai joined #salt
18:47 eclectic joined #salt
18:47 vlcn joined #salt
18:47 Twiglet joined #salt
18:47 seb` joined #salt
18:47 nlb joined #salt
18:47 rhand joined #salt
18:47 djaykay joined #salt
18:47 utahcon joined #salt
18:47 rofl____ joined #salt
18:47 TamCore joined #salt
18:47 dean|away joined #salt
18:47 mortis joined #salt
18:47 Damoun joined #salt
18:47 martoss1 joined #salt
18:47 ndrei__ joined #salt
18:47 kossy_ joined #salt
18:47 keekz joined #salt
18:48 eofs joined #salt
18:48 monokrome joined #salt
18:48 ashb joined #salt
18:48 fivethreeo joined #salt
18:48 Heggan joined #salt
18:48 dh joined #salt
18:48 kevinbrolly joined #salt
18:48 Hipikat joined #salt
18:48 ksk joined #salt
18:48 ikanobori joined #salt
18:48 bensons joined #salt
18:48 jeblair joined #salt
18:48 tru_tru joined #salt
18:48 Ixan joined #salt
18:48 Ymage joined #salt
18:48 johanek joined #salt
18:48 jperras joined #salt
18:48 rhand joined #salt
18:48 TaiSHi joined #salt
18:48 micko joined #salt
18:49 martoss joined #salt
18:49 Ixan joined #salt
18:49 jjasinski joined #salt
18:49 ajolo__ joined #salt
18:49 speed145a joined #salt
18:50 Hell_Fire joined #salt
18:50 nlb joined #salt
18:50 nlb joined #salt
18:50 djaykay joined #salt
18:51 kermit joined #salt
18:51 oncallsu1ks joined #salt
18:51 ndrei joined #salt
18:51 claytron_ joined #salt
18:51 erjohnso_ joined #salt
18:51 emostar_ joined #salt
18:51 hvn_ joined #salt
18:51 ksalman_ joined #salt
18:52 the_lale1u joined #salt
18:52 yidhra_ joined #salt
18:52 nineteen1ightd joined #salt
18:52 forrest whiteinge, if you have time today can you review: https://github.com/saltstack-formulas/users-formula/pull/39 ? I don't think it should be blindly merged as it changes how the users formula works for part of it.
18:55 happytux joined #salt
18:56 ecdhe forrest, are there any PRs that should be blindly merged?
18:57 forrest ecdhe, yea I'll do it sometimes if it's just a readme update or something, I'll skim
18:57 ecdhe makes sense!
18:58 ecdhe forrest, you were one of the first to help me with salt back in November '13 when I was first learning it.
18:58 forrest ecdhe, cool! Still enjoying it?
18:58 ecdhe Absolutely!
18:58 forrest that's good :D
18:58 ecdhe I've moved from google the docs whenever I need to do something to typing out states from memory.
18:59 forrest haha nice
18:59 renoirb joined #salt
18:59 ecdhe s/google/googling  (but actually duckduckgo-ing)
18:59 vbabiy joined #salt
19:00 herlo running a masterless salt-call state.highstate, I want to specify my 'dev' environment and not my 'prod' environment. How do I do this?
19:00 ramishra joined #salt
19:00 tkharju2 joined #salt
19:01 forrest herlo, are you not using a local top file?
19:01 herlo forrest: yes
19:01 ahammond joined #salt
19:01 forrest oh you're not doing --local
19:01 herlo it's in the environment directory
19:02 herlo forrest: well, I am, just from the config file.
19:02 forrest oh, hmm
19:02 ecdhe forrest, I hope to make the transition this year to being able to help folks on the mailing list / irc.
19:02 herlo forrest: salt-call --local state.highstate
19:02 forrest herlo, it isn't picking up the local top file?
19:02 forrest ecdhe, that would be great! You know we're doing a documentation sprint next month right?
19:02 forrest ecdhe, I don't remember what part of the world you're located in
19:03 herlo forrest: is that not the same as salt-call state.highstate if I modify /etc/salt/minion to say file_clients: local?
19:03 herlo forrest: it is, but I have two environments, dev and prod
19:03 herlo forrest: right now, I don't have one in prod.
19:03 forrest herlo, that's why I'm confused, it should be picking it up from the top if you have multiple environments right? Especially if it's local
19:03 dober joined #salt
19:03 herlo forrest: let me restate
19:04 ecdhe forrest, I'm in the midwest.  A docs sprint sounds interesting!
19:04 dude^2 joined #salt
19:04 forrest ecdhe, https://www.eventbrite.com/e/saltstack-documentation-sprint-tickets-12010895913
19:04 herlo forrest: I have /srv/salt/dev/top.sls and /srv/salt/prod/top.sls (well, not yet but I want to), which gets invoked when I run salt-call --local state.highstate?
19:04 lionel joined #salt
19:04 forrest we'll have a virtual hangout one if you want to join us.
19:04 hhenkel joined #salt
19:04 herlo forrest: I do not have a /srv/salt/top.sls
19:04 herlo forrest: and I don't plan to have one.
19:05 forrest herlo, I see, I'm not sure then
19:05 herlo forrest: even if I did, I have a 'dev' and a 'prod' environment inside each top.sls...
19:05 forrest Yea
19:05 herlo forrest: yeah, now you understand my dilemma. :)
19:05 forrest herlo, yea lol
19:05 herlo forrest: :)
19:05 forrest if you figure it out can you document it?
19:05 forrest or at least create an issue
19:05 forrest so one of us will
19:05 * herlo waits for UtahDave or whiteinge to answer my question. :)
19:05 forrest psssh, just test it, duh
19:05 herlo forrest: ??
19:06 herlo how would I test it?
19:06 forrest herlo, see which one runs when you do the highstate
19:06 herlo I mean, I get that I can figure it out, but how would I know why?
19:06 forrest ba dum tish
19:06 herlo I'd have to dig through a bunch of code.
19:06 forrest Yea
19:06 ml_1 joined #salt
19:06 herlo and I figure someone wrote this, and should know the answer
19:06 forrest well, you could strace the process too :D
19:06 herlo forrest: meh
19:06 herlo for now, I can just make the top.sls the same in both branches
19:06 forrest fair enough
19:07 herlo forrest: speaking of which, you might find my approach to git branching interesting for salt
19:07 forrest I'd still like to know the answer when you figure it out
19:07 herlo forrest: sure, I think I can handle that, IF I figure it out.
19:07 forrest hah
19:07 herlo forrest: I do not use gitfs, but I do use just one git tree for managing prod and dev environments. :)
19:08 herlo and my config, too
19:08 forrest so they're on different branches?
19:08 herlo forrest: yes.
19:08 herlo it gives me flexibility when I merge from dev to prod but still gives me separation of environments that I want.
19:09 forrest fair enough
19:09 forrest we used to do something like that when I used puppet
19:09 troyready joined #salt
19:09 herlo essentially, the 'dev' branch maps to /srv/salt/dev and the 'prod' branch maps to /srv/salt/prod. the config branch maps to /etc/salt/config, though I might have to something with that if I decide to go to a master solutoin.
19:09 herlo forrest: yeah, environments are nice in puppet. I can't say the same yet about salt...
19:10 forrest interesting
19:10 forrest Yea it could use some work
19:10 herlo well, there's probably a way to do it, but I've not found the docs for it yet...
19:11 herlo forrest: http://docs.saltstack.com/en/latest/ref/configuration/minion.html#environment
19:11 herlo maybe?
19:12 forrest herlo, ahh yea that would probably work
19:12 forrest since you have no top
19:12 forrest like a CRAZY PERSON
19:12 herlo I have a top
19:12 blast_hardcheese joined #salt
19:12 herlo two, in fact.
19:12 forrest but not in the standard location I mean
19:12 herlo standard?
19:12 herlo the location I have it is perfectly acceptable
19:12 forrest /srv/salt/top.sls
19:13 forrest yea but with how you're doing it there isn't a way to 'force' one via the top
19:13 forrest at least I don't think so
19:13 * goodwill hugs herlo
19:13 herlo http://docs.saltstack.com/en/latest/ref/states/top.html
19:13 herlo goodwill: !!!
19:13 forrest so I think that the config file is the way to go
19:13 * herlo belly bumps goodwill!!
19:13 goodwill belly bump!!!!
19:13 herlo forrest: well, so the thing is, I want to specify it at runtime.
19:13 MindDrive joined #salt
19:13 herlo forrest: I'm wondering if environment=dev will do it, though.
19:13 forrest it might
19:15 freelock joined #salt
19:17 happytux_ joined #salt
19:18 happytux joined #salt
19:18 schimmy joined #salt
19:19 marnom joined #salt
19:19 ipalreadytaken joined #salt
19:19 jaimed joined #salt
19:19 davet1 joined #salt
19:19 ckao joined #salt
19:19 UtahDave joined #salt
19:19 smcquay joined #salt
19:19 oz_akan_ joined #salt
19:19 pressureman joined #salt
19:19 CeBe2 joined #salt
19:19 notpeter_ joined #salt
19:19 johtso joined #salt
19:19 jasonrm joined #salt
19:19 gmoro joined #salt
19:19 mattikus joined #salt
19:19 philipsd6 joined #salt
19:19 torrancew joined #salt
19:19 faulkner joined #salt
19:19 basepi joined #salt
19:19 jeremyBass joined #salt
19:19 seblu joined #salt
19:19 MTecknology joined #salt
19:19 jforest joined #salt
19:19 herlo forrest: explain this to me: http://ur1.ca/hm0ix :)
19:19 herlo oh, haha, not in the /srv/salt/prod directory. I bet that is it...
19:20 bhosmer_ joined #salt
19:20 herlo and I missed a colon! :(
19:21 herlo dammit! That fails! :(
19:22 herlo forrest: http://ur1.ca/hm0jk
19:23 herlo why salt, why do you have to be so troublesome. :/
19:23 schimmy joined #salt
19:23 nebuchadnezzar joined #salt
19:23 chiui joined #salt
19:28 luminous UtahDave: thanks for that sys.doc snippet!
19:28 luminous that's really sweet
19:29 herlo forrest: http://ur1.ca/hm0kk <-- looks like this now, but still doesn't do what I want. :(
19:31 luminous is there an 'unless' for module.run?
19:31 mprymek joined #salt
19:31 xzarth_ joined #salt
19:32 matrix3000 joined #salt
19:33 herlo forrest: looks like there's already a bug about this: https://github.com/saltstack/salt/issues/9186
19:37 herlo seems https://github.com/saltstack/salt/issues/6313 is even closer to my issue.
19:41 luminous is there a way to prevent module.run from running, just as with cmd.rung?
19:42 tkharju3 joined #salt
19:42 luminous module.run doesn't seem to care about unless
19:43 whiteinge dstokes: i saw that first-hand yesterday. i'm not sure yet if the disconnect is in cherrypy's ssl adapter or in pyopenssl. i'll dive into that today. soon as I can find two spare minutes :-P
19:47 bhosmer_ joined #salt
19:49 racooper UtahDave or anyone who might know...I'm trying to set up a state using module.run to set quotas for users.  not sure of the syntax.  here's what I have so far, with the error I"m getting: https://gist.github.com/racooper/970be82b500a5c044afb
19:50 dstokes whiteinge: looks like salt hardcodes CP config to use python builtin ssl, so don't think it's pyopenssl. going to resort to nginx ssl termination for now (will also allow me to use basic auth for /hook)
19:53 unixman9000 joined #salt
19:53 jslatts joined #salt
19:56 matrix3000 joined #salt
19:58 chrisjones joined #salt
20:00 mapu Having a devil of a time getting raid working using saltstack to create a raid volume. Using the syntax given in the docs, I am see no success. I have my creation stanze here: http://pastebin.com/mELB0Bup
20:00 mapu any advice would be helpful
20:01 whiteinge dstokes: sounds like a good plan. i'll keep you updated
20:01 dstokes whiteinge: thx
20:01 ramishra joined #salt
20:01 * herlo actually thinks whiteinge may have been helpful today...
20:01 herlo the next test will tell us for sure, however.
20:03 possibilities joined #salt
20:03 dstokes whiteinge: there aren't currently any sysv / upstart init scripts for the salt-api right?
20:04 UtahDave racooper: I think you need to change '- func: ' to '- name: '
20:04 dstokes nvm, just found them
20:04 ndrei joined #salt
20:05 ndrei_ joined #salt
20:06 whiteinge dstokes: sounds like you're good but just in case: we have init scripts for sysv, upstart, and systemd. they're each in different subdirs though
20:06 dstokes whiteinge: found em in the repo, but looks like they're not being installed on setup or bootstrap. looking for that code
20:06 dstokes (only have master and minion scripts)
20:07 whiteinge dstokes: setup.py doesn't install init scripts or etc configs. only the distro-level packages will do that.
20:08 herlo whiteinge: you failed me, again!
20:08 herlo http://ur1.ca/hm0ul
20:09 herlo whiteinge: this might be the last time I trust you!
20:09 dstokes whiteinge: pretty sure salt-bootstrap is only installing master, minion, syndic. probably need a pr there
20:09 dstokes creating..
20:10 herlo UtahDave: salt-call state.top file:///srv/salt/dev/dev.sls result is the same as http://ur1.ca/hm0ul
20:12 maxskew joined #salt
20:15 UtahDave herlo: try    salt-call state.top  dev/dev.sls
20:15 Ahlee Is there any difference between reactor running cmd.saltutil.sync_modules and just running saltutil.sync_modules?
20:15 herlo UtahDave: that's what I did the first time. :)
20:16 Ahlee When reactor runs, I get https://gist.github.com/jalons/30ef32e756cb388d876c, if I just run saltutil.sync_modules, I get no error
20:16 nineteeneightd joined #salt
20:16 UtahDave herlo: true.  so   /srv/salt/dev/dev.sls  exists, right?
20:16 herlo yes
20:16 whiteinge Ahlee: both should use 99% the same code paths
20:16 herlo # ls -l /srv/salt/dev/dev.sls
20:17 herlo -rw-r--r--. 1 root root 25 Jun 26 20:05 /srv/salt/dev/dev.sls
20:17 herlo UtahDave: ^^
20:17 UtahDave gotcha.  Ok let me test something here real quick
20:17 Ahlee whiteinge: that's what I figured, guess I'm a 1%er ;)
20:17 racooper UtahDave,  that fixed the inital problem. new error: The following arguments are missing: device". I tried adding  "- device: /home" but then it says "Module fuction quota.set threw an exception".
20:17 whiteinge Ahlee: is this that same traceback you were seeing via the reactor last week (-ish?)
20:17 Ahlee whiteinge: yep.
20:17 Ahlee Actually scheduled time to dive in
20:18 Ahlee so, diving.
20:18 whiteinge Ahlee: hm. wonder why the mongo_future returner is looking for a top.sls file
20:18 Ahlee whiteinge: I am, as well.
20:19 Ahlee just realized this host doesn't have pymongo, either.  adding, restarting, retesting
20:19 Ahlee wondering if this is just a very strange 'i can't connect' error
20:20 UtahDave herlo: can you pastebin your file_roots option from your master config?
20:21 herlo sure
20:22 herlo UtahDave: http://ur1.ca/hm0xt
20:22 MTecknology joined #salt
20:25 UtahDave ok, herlo.   You need a base environment to house your top file.  http://docs.saltstack.com/en/latest/ref/configuration/master.html#file-roots
20:25 herlo why?
20:25 UtahDave you can have both your dev and prod top files in that base environment and use state.top to pick which one you want to use
20:25 herlo UtahDave: hang on, I have an example of where that's not necessary.
20:25 UtahDave because the way you have it set up now, the minion doesn't know where to get it's files from.
20:27 herlo UtahDave: so I guess the question there is, why did it work with state.highstate? I mean it wasn't the way I wanted, but it worked without a base section.
20:28 herlo UtahDave: there's an example at the bottom of this document, but now that I read it closer, it doesn't explicity say that base isn't required.
20:28 herlo http://docs.saltstack.com/en/latest/ref/states/top.html
20:29 shaggy_surfer joined #salt
20:30 mateoconfeugo joined #salt
20:30 mgw joined #salt
20:31 herlo UtahDave: so if I did something like /srv/salt/top.sls and put just base: in there, that's all I'd need?
20:33 UtahDave herlo: well, that would work to allow /srv/salt/top.sls to match on which minions get which environments for a regular highstate
20:33 herlo UtahDave: I never plan to do that...
20:33 UtahDave if you want to toggle which to run at the cli you would still need  /srv/salt/devtop.sls  and /srv/salt/prodtop.sls
20:34 UtahDave then you could execute   salt 'minion' state.top dev.sls
20:34 Ryan_Lane joined #salt
20:35 herlo I'm not sure I understand why it can't be separate like I have it already.
20:35 herlo why would I need to put it in those directories?
20:35 herlo UtahDave: ^^
20:35 herlo the whole point of this is so I don't have to use /srv/salt/
20:36 herlo If I have to put a base: section in my /srv/salt/dev/dev.sls and /srv/salt/prod/prod.sls just to get around this, I'm okay with that, I guess.
20:36 UtahDave because Salt won't know which environment to look in.  The /srv/salt/top.sls   tells the minion to look in   /srv/salt/dev   or /srv/salt/prod for its files
20:38 herlo wha?
20:39 herlo UtahDave: but when I do a state.highstate I don't have a requirement of the base environment or even a /srv/salt/top.sls
20:39 kermit joined #salt
20:39 UtahDave did you set the environment in the minion config?
20:40 herlo yes, I did, but that had no bearing
20:40 herlo environment: prod
20:41 herlo UtahDave: tried also to pass environment=prod on the state.highstate at the cli, still nothing was different.
20:41 herlo state.top is likely what I want.
20:41 herlo but it shouldn't require a base environment if state.highstate doesn't
20:43 racooper UtahDave,  that fixed the initial problem. new error: The following arguments are missing: device". I tried adding  "- device: /home" but then it says "Module fuction quota.set threw an exception". saying "- opt: /home" gave the "argument missing" error too.
20:43 Shish joined #salt
20:43 herlo UtahDave: so I added /srv/salt/top.sls
20:44 Ahlee if i update a reactor state file, I do not need to restart the master for it to take effect, correct?
20:44 herlo in it I added a base: section that just had '*' and the test module I am using besides common.
20:44 herlo UtahDave: I've not gotten a different result yet.
20:44 UtahDave Ahlee: restart not required for modifying the reactor file.  just if you modify the master conf
20:44 Ahlee ty
20:44 herlo UtahDave: I don't want to maintain a top.sls in /srv/salt because it screws with my git branching scheme...
20:45 UtahDave how about making    /srv/salt/base   ?
20:45 herlo I can put one there that just sits there...
20:45 herlo UtahDave: base can just be prod for all I care.
20:45 herlo but again, it seems arbtrarily silly to require me to name an environment 'base'
20:45 * herlo goes and changes things around a bit...
20:46 UtahDave herlo: yes, base is required.
20:48 herlo UtahDave: still no workie: http://paste.fedoraproject.org/113601/38156771/
20:49 herlo UtahDave: I'd even be happy to change the dir to /srv/salt/base if you like, though I don't think it's going to change anything.
20:49 CheKoLyN joined #salt
20:51 herlo UtahDave: did the change to /srv/salt/base, still same error.
20:51 herlo and the prod.sls to top.sls
20:51 herlo even ran salt-call state.top base/top.sls and with full path, no dice.
20:52 herlo UtahDave: when I run /srv/salt/top.sls (even though it's not defined anywhere), it works.
20:52 herlo http://ur1.ca/hm15x
20:53 UtahDave are you restarting the salt-master after modifying the master config?
20:53 herlo masterless
20:53 herlo /etc/salt/minion
20:53 UtahDave ah, ok
20:53 herlo UtahDave: thus the salt-call
20:53 herlo not salt '*'
20:53 herlo :)
20:54 UtahDave gotcha.   Do you have file_client: local in your minion config?
20:55 herlo yes
20:55 UtahDave ok, let me configure that here and test.
20:55 herlo okay
20:56 UtahDave just a minute to get it going
20:56 jnials joined #salt
20:59 stevednd If I'm writing a module that is more or less a wrapper around a command line app, would the best course of action be to just use cmd.run_all inside the methods, and then pick and choose what I want to return from my module?
21:02 ramishra joined #salt
21:09 oz_akan__ joined #salt
21:10 bmatt hey so, I've got a Deep Magic type question
21:10 Ryan_Lane1 joined #salt
21:10 bmatt I need to import a custom execution module into a custom grain
21:11 bmatt and then I need to run it
21:12 UtahDave bmatt: here's how we do it in the core grains module:  https://github.com/saltstack/salt/blob/develop/salt/grains/core.py#L36
21:13 tedski UtahDave: is the salt.modules namespace populated with the custom modules?
21:13 UtahDave tedski: nope.  Your importing relatively straight from the modules directory
21:13 UtahDave The grains don't have access to the loaded modules.  Chicken and egg problem
21:14 tedski yeah, bmatt needs to import from a custom module
21:14 bmatt ^
21:14 tedski err s/from//
21:14 UtahDave the Salt loader needs system info from grains
21:14 UtahDave bmatt: yeah, so just import your custom module from wherever it is.  It's just a python file
21:14 bmatt from _modules?
21:14 UtahDave herlo: still testing
21:15 herlo UtahDave: okay
21:15 kermit joined #salt
21:15 UtahDave bmatt: Hm. You probably need to find the direct path to the directory where synced modules reside.
21:15 bmatt oof
21:15 bmatt that sounds brittle
21:15 UtahDave herlo: just want to make sure you know I'm still working on that
21:17 herlo :)
21:17 matrix3000 joined #salt
21:18 jnials joined #salt
21:19 shaggy_surfer joined #salt
21:21 gildegoma joined #salt
21:22 mgw joined #salt
21:31 Linuturk joined #salt
21:32 herlo UtahDave: any progress?
21:32 UtahDave herlo: yep!  i got one variation working. now testing a second
21:33 herlo okay, awesome!
21:34 UtahDave Yeah, I got two different but similar scenarios working.  Let me pastebin what I have
21:35 herlo cool
21:40 kermit joined #salt
21:43 druonysus joined #salt
21:43 kermit joined #salt
21:45 ninkotech__ joined #salt
21:45 jnials joined #salt
21:45 alekibango_ joined #salt
21:45 UtahDave ok, herlo.  here's one scenario.
21:45 UtahDave https://gist.github.com/UtahDave/5234e366c7e032b2af56
21:46 * herlo looks
21:46 jrdx joined #salt
21:46 UtahDave /srv/salt/herlo  is the base environment  and /srv/salt/herlo/prod    and /srv/salt/herlo/dev  are subdirectories
21:49 UtahDave herlo: I also tested making  /srv/salt/herlo/base  and putting all three top files in that directory, and that worked, too
21:49 kermit joined #salt
21:49 herlo UtahDave: I guess that could work.
21:50 herlo the former will absolutely not work if I'm required to have a top.sls at one directory up the tree.
21:50 UtahDave if the base,prod, and dev directories are all siblings, and all the top files are in the base environment then this works, too
21:50 UtahDave [boucha@dasalt ~]$ sudo salt-call state.top prodtop.sls
21:50 herlo but putting everything in one dir could work...
21:51 herlo UtahDave: so can you enlighten me on the logic behind requiring something named base in the hierarchy? Seems awfully arbitrary.
21:51 allanparsons is there a way to install a develop branch from the PPA?
21:51 herlo UtahDave: and thinking further on it now, having everything in one directory won't work...
21:52 herlo because I want to be able to separate out environments by directory. I don't want to have stuff in production that isn't read to be there.
21:52 * herlo thinks base and dev will be the way to go, probably.
21:54 herlo s/read/ready/
21:54 UtahDave gotcha
21:55 druonysuse joined #salt
21:55 herlo UtahDave: but without putting a top.sls into /srv/salt, which is 'before' the directory structure I want to use, it seems to fail still...
21:55 herlo UtahDave: this really seems like a fundamental thing I *should* be able to do .
21:56 herlo UtahDave: so I just discovered something interesting....
21:56 * herlo is verifying his findings....
22:01 ajprog_laptop joined #salt
22:01 DaveQB joined #salt
22:03 herlo UtahDave: okay, so here's what I found.
22:03 ramishra joined #salt
22:03 nickg can you match pillars on other pillars?
22:03 herlo UtahDave: http://ur1.ca/hm1mi
22:04 herlo UtahDave: if I'm in the /srv/salt/dev directory and run salt-call state.top dev.sls, it works. If I am, however, in /srv/salt/ and run salt-call state.top dev/dev.sls, it fails.
22:04 herlo UtahDave: in terms of a cron job, this is a non-issue, but it does seem mighty brittle to require me to be in the directory to run my states this way.
22:05 UtahDave it shouldn't matter at all.
22:05 UtahDave both of those calls look like they succeeded in your paste
22:06 UtahDave oh, in your situation never use dev/dev.sls
22:06 herlo right, I have in the past
22:06 UtahDave that path isn't referring to your system's filesystem
22:06 UtahDave it's all coming from Salt's file server   salt://
22:07 herlo UtahDave: http://ur1.ca/hm1n3
22:07 UtahDave yeah,  just use   dev.sls
22:07 herlo UtahDave: so if I specify file:///srv/salt/dev/dev.sls, that should work too?
22:07 UtahDave no
22:07 herlo or would it be salt://<something>
22:07 UtahDave not that either
22:07 * herlo is confuzled
22:08 UtahDave it's using your file_roots as the base of the path.
22:08 herlo not that I have a problem with my way, other than it's a bit brittle, like I said.
22:08 herlo UtahDave: ahh, so I could have something like
22:08 herlo base:
22:08 herlo - /srv/salt
22:08 herlo and then do the directories like I have them and I'd have to use dev/dev.sls
22:08 herlo if I had
22:08 herlo base:
22:08 herlo - /
22:08 herlo I could use 'srv/salt/dev/dev.sls'
22:09 herlo right?
22:09 herlo not that I'd do that...
22:09 UtahDave yes, exactly
22:09 UtahDave :0
22:09 UtahDave :)
22:09 herlo okay, so I think I see where I was misunderstanding this...
22:09 herlo and now I go back to my fun experiments :)
22:09 herlo thanks UtahDave
22:10 UtahDave Hey, I have to run to a meeting.
22:10 matrix3000 anyone releasing any good Salt books soon?
22:10 UtahDave catch you soon
22:10 UtahDave left #salt
22:10 yomilk joined #salt
22:10 herlo ttfn
22:10 matrix3000 other than a manual, something with more case studies and implementation best practices
22:11 forrest matrix3000, probably not 'soon' https://github.com/gravyboat/salt-book/tree/master/en
22:11 forrest but I'm working on it
22:12 forrest damned things like 'work' getting in the way
22:12 forrest sheesh
22:12 forrest implementation best practices would be a good idea to add
22:12 tkharju4 joined #salt
22:13 ajprog_laptop joined #salt
22:14 bhosmer joined #salt
22:15 matrix3000 Because Salt is so open to customization it's very easy to end up in a bad situation it seems.
22:15 matrix3000 Knowing best practices of similar implementations might help user acquisition
22:16 jalbretsen joined #salt
22:17 zach https://media.readthedocs.org/pdf/salt/latest/salt.pdf
22:17 zach that is just a single page...
22:18 nickg i need settings for conf files to per facility.  i have a subnet (/16) in each facility, which could be used as the determining factor. any recommendations on how to setup pillars without hard coding the subnet into each file?
22:21 bmatt well, the information mapping data to facility has to live somewhere
22:21 bmatt where do you want it to live?
22:22 nickg i'd prefer for it to be in a central location where i can find a value that can be checked in various files..
22:22 bmatt (rephrased, the information that gets used to map your settings to a particular facility identifier has to live somewhere)
22:22 forrest zach, Seth is aware of that issue
22:22 forrest whiteinge, did you ever figure out what was with the PDF not generating properly on build?
22:22 bmatt nickg: what kind of "central location"?
22:23 nickg bmatt i just want it in one location, then within each package it can use a single field maybe pillar['facility'] = 'FAC_AWS_EAST' or 'FAC_EQUINIX_ASH_1' etc
22:24 nickg maybe i just need to write a custom grain
22:24 herlo forrest: so I got it all working the way I want...
22:24 herlo http://paste.fedoraproject.org/113631/82127814/
22:24 herlo forrest: no base file_root required, either. :)
22:25 whiteinge forrest, zach: PDF isn't building due to bad unicode chars. god-willing, i'll have the replacement up something late today
22:26 forrest herlo, so what did you have to modify then
22:26 forrest whiteinge, lol
22:26 forrest whiteinge, WHY YOU BREAK THINGS
22:26 forrest seriously, unicode characters, sigh
22:26 whiteinge pdflatex is getting a tad dated
22:26 whiteinge i'm switching to use xelatex for the builds
22:27 gothix_ joined #salt
22:27 forrest whiteinge, what made you choose that over rst2pdf?
22:27 forrest I'm just curious
22:27 forrest since I know rst2pdf has sphinx integration
22:27 whiteinge latex uber alles!
22:28 forrest lol, fair enough
22:28 aquinas joined #salt
22:28 herlo forrest: my paste should explain it all
22:28 ipalreadytaken joined #salt
22:28 bhosmer joined #salt
22:28 whiteinge ^^ pfft
22:28 forrest herlo oh you actually just referenced the sls itself
22:29 forrest did you add that to the docs?
22:29 herlo forrest: no
22:29 forrest or file an issue with that example?
22:29 herlo forrest: because it's already there
22:29 herlo forrest: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.top
22:29 forrest where? Sorry I've been in a meeting
22:29 herlo :)
22:29 Shish joined #salt
22:29 forrest that should have examples using something like you have I think though
22:29 forrest reverse_top.sls isn't as clear as prod.sls and dev.sls
22:29 herlo forrest: feel free to add them. :)
22:30 herlo forrest: yeah, I'm sure it could be improved.
22:30 whiteinge forrest: fyi, herlo is a known troll. i'm happy to kickban him for you
22:30 herlo forrest: plus, there's some other interesting things you have to assume / know before you use it.
22:30 forrest whiteinge, no I have met herlo in person, he's a troll there too
22:30 * herlo kicks whiteinge first...
22:30 seventy3 joined #salt
22:30 * herlo kicks forrest in the shins.... twice
22:31 * herlo runs away, laughing! :D
22:31 * whiteinge heads back to his meeting
22:31 herlo haha
22:34 patarr anyone have an idea why salt is reporting a package is not installed, but I clearly see it is with rpm -qa
22:34 tkharju4 joined #salt
22:35 forrest matrix3000, I've got a best practices page: http://docs.saltstack.com/en/latest/topics/best_practices.html Do you mean regarding a specific implementation of software? If so I believe that's unrealistic to do, there are too many variables.
22:39 Hell_Lap joined #salt
22:39 Hell_Lap Woo, found a bug in modules.keystone :D
22:39 patarr ah. Hm. Seems like if you name your ID the same as the package in 'sources' it will bug out
22:40 manfred Hell_Lap: is it in the issue tracker? if not, please add it there :)
22:40 Hell_Lap looking it up now :)
22:40 mgw joined #salt
22:41 CeBe joined #salt
22:42 anotherZero joined #salt
22:45 Hell_Lap Doesn't look like it is, woo, I'll do up an issue and a patch :)
22:45 patarr I've noticed that yum will usually install the nearest package that resembles the name you gave it. Does salt not operate this way too?
22:45 manfred patarr: salt passes all the things to yum
22:46 manfred but it looks for it with repoquery
22:46 Outlander joined #salt
22:46 patarr manfred: I typed in "yum install esl-erlang-R16B03" in the system and it proceeded to install. But when I have that in salt, it will not. It tells me it cant find it.
22:47 manfred patarr: probably cause it can't find that exact thing with repoquery?
22:47 manfred patarr: it does a repoquery to see if the package is out there, then installs it if it is
22:47 patarr ah so repoquery means I need to specify exact packages.
22:47 manfred possibly
22:47 aw110f joined #salt
22:47 herlo dammit! This still is not working. One environment is overriding the other. :(
22:48 patarr aand now I have a caching issue.
22:49 obimod joined #salt
22:49 obimod can the pillar SLS files use other pillar SLS files to do the jinja?
22:50 obimod like {{ pillar.get( ... ) }} in pillar/base/nginx/init.sls
22:50 herlo forrest: so get this. I have the following: http://ur1.ca/hm1w5. Wanna guess which one gets run when I do 'salt-call state.top dev.sls'
22:50 herlo ??
22:50 obimod kind of a "why-don't-you-try-it" kind of question, but i figured I'd try to save time with a simple yes/no question : D
22:50 manfred obimod: should be able to
22:50 obimod manfred: sweet!
22:50 forrest herlo, I am guessing the second one
22:51 herlo forrest: well, it's run alphabetically.
22:51 patarr manfred: Specified full package name from the repo. http://pastebin.com/ge6gamnw
22:51 herlo so dev first, then prod...
22:51 forrest lol
22:51 * herlo doesn't want the other to run...
22:51 manfred obimod: https://github.com/saltstack/salt/issues/4326
22:51 forrest herlo, can you trye salt-call state.top dev.dev.sls?
22:51 forrest or some modification of that
22:51 herlo huh, but that would require the file_roots to change, no?
22:51 stevednd manfred: when was the functionality to access pillar data from a pillar added? Last time I tried that I believe it failed quite miserably.
22:51 manfred ¯\(°_o)/¯
22:51 herlo forrest: yup...
22:52 manfred stevednd: pillar.get() works
22:52 manfred stevednd: pillar[] doesn't?
22:52 manfred stevednd: according to https://github.com/saltstack/salt/issues/4326
22:52 stevednd that could very well be the case. I didn't try .get(), though I think I also tried salt['pillar.get']
22:52 stevednd might be mistaken though
22:53 manfred stevednd: https://github.com/saltstack/salt/issues/6955
22:53 herlo forrest: the problem with that is then I have to put the modules in /srv/salt and the top files in separate directories so they don't run both. Seems silly when I just want it to run the ONE dev.sls
22:53 forrest herlo, yea understandable
22:53 manfred stevednd: i have no idea no
22:53 herlo the main problem is that I will want to then merge my dev branch into my master branch (which represents prod). I wonder if there's a way in git to not allow merging a specific file.
22:54 obimod dangg
22:54 manfred stevednd: obimod https://docs.saltstack.com/en/latest/topics/pillar/#including-other-pillars
22:54 manfred oh that is includes
22:54 manfred nevermind
22:54 manfred not that one
22:55 herlo forrest: so I think if I could cherrypick commits, it would work, but that defeats the purpose of simple and fast.... :(
22:55 forrest yea
22:56 obimod manfred: stevednd: looks like neither works
22:56 manfred meh
22:56 herlo ugh!
22:57 obimod ^_^
22:59 herlo forrest: figured out something useful, actually.
22:59 herlo .gitattributes. :)
22:59 forrest lol
23:01 herlo forrest: it's pretty cool actually, still working out the details, though.
23:03 savvy-lizard joined #salt
23:03 ramishra joined #salt
23:09 Hell_Lap come on clone, hurry up
23:10 xmj_ joined #salt
23:14 KyleG basepi: Did you ever get anywhere on https://groups.google.com/forum/#!topic/salt-users/zEA224dbXvo ?
23:14 KyleG I'm experiencing the same shit...
23:15 KyleG I can't highstate this minion, I've tried deleting keys, restarting master/minion multiple times
23:15 KyleG It gets through and then randomly it'll deny/timeout on getting the key
23:15 KyleG and say "Did the master accept it?"
23:15 KyleG well no shit it did, you were just talking to it. I'm curious as to why the minion needs to decrypt the key so many times instead of just grabbing what it needs to do and doing it
23:18 basepi KyleG: Weird!  I'm not sure that an issue was created for that, would you mind creating one?
23:19 KyleG Sure, I'll enable some deeper debugging and get one out to ya
23:20 herlo forrest: basepi: that sure was fun. I just decided that instead of futzing around with environments, I'd futz around with the minion config file and that works. :) no cross-contamination of environments for me. :)
23:20 basepi herlo: woot!@
23:21 herlo basepi: it's a very bastardized way to do it
23:21 herlo basepi: this is what I ended up doing -- # salt-call -c /etc/salt/dev/ state.top dev.sls
23:21 basepi ew
23:21 herlo yeah
23:22 basepi =P
23:22 basepi well, at least it works?
23:22 herlo but otherwise, salt combines all of my environments into one...
23:22 herlo which really seems to be NOT what i want.
23:22 herlo I could see aggregating all the components in an environment, but not all the environments together.
23:24 ipalreadytaken joined #salt
23:25 bhosmer joined #salt
23:26 bhosmer_ joined #salt
23:27 schimmy1 joined #salt
23:28 joehh nickg: what os is your master? that patch should be working in the ubuntu master packages
23:31 Hell_Lap there we go, pull request is in :)
23:31 ahammond My opts.id is foo.bar.com I'd like to get just foo. Is there an easy way to do that in jinja?
23:33 mgw joined #salt
23:33 schimmy joined #salt
23:38 mgorman joined #salt
23:40 schimmy joined #salt
23:50 nickg joehh: ubuntu 12.04 LTS.  I upgraded the salt-master this morning and installed a new VM with the salt-minion at the same time.  The minion is 14.04
23:51 obimod manfred: stevednd: i ended up doing a { load_yaml ... } call in the data pillar file and an { from .. import ... with context } (like here: https://github.com/bechtoldt/httpd-formula/blob/master/httpd/init.sls#L1) and it worked!!! =DDD * dance *
23:52 KyleG basepi: I'm going to hold off on the bug report, my minion was running 2014.1.4 while my master was on 2014.1.0, so I downgraded the minion to 2014.1.0 and highstate works, I'll hold off until I see the issue on same version minion/master
23:54 joehh Now I understand, I haven't updated the packaging for precise or lucid (different workflow - so not as easy) so your master won't have got the patch
23:54 joehh I'll aim to do them over the weekend (esp since 12.04 is likely to be most common)
23:55 basepi KyleG: interesting!  I don't know of any known incompatibilities between those versions, but you should definitely always upgrade your master first.
23:55 KyleG yeah I didn't realize my co-worker had updated our package repo..
23:58 schimmy1 joined #salt
23:59 schimmy2 joined #salt

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