Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-06-18

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

All times shown according to UTC.

Time Nick Message
00:00 redondos joined #salt
00:03 shaggy_surfer joined #salt
00:05 rgarcia_ joined #salt
00:11 acabrera joined #salt
00:13 CheKoLyN joined #salt
00:17 napper joined #salt
00:18 bejer joined #salt
00:21 ifnull I built PyZMQ: 14.3.1 and ZMQ: 3.2.4 from source on Ubuntu 14.04 to test if ZMQ 4 was the cause of this issue seen here https://github.com/saltstack/salt/issues/12248
00:21 forrest any luck ifnull ?
00:21 ifnull nope. same problem
00:21 forrest Lame
00:21 forrest why you gotta break it?
00:21 ifnull lol
00:22 ifnull that is what my coworker said
00:22 forrest Are you going to update the issue notating that?
00:22 ifnull yeah
00:22 forrest cool
00:22 Ryan_Lane joined #salt
00:23 ifnull so far we are only seeing this issue on the boxes running Ubuntu 14. All Ubuntu 12 boxes still work.
00:23 forrest which release of salt?
00:25 forrest ifnull, I've got about half a dozen sandboxes running 14.04 with 2014.1.3, and I am not encountering that issue
00:26 ifnull maybe an issue with 2014.1.4? i’ll give 2014.1.3 a try
00:26 forrest ahh I see your version report now
00:26 forrest ok cool, let me know, I'll add a comment to the issue
00:29 ifnull https://github.com/saltstack/salt/issues/12248#issuecomment-46382972
00:29 forrest yep I see, I'm commenting now
00:30 forrest https://github.com/saltstack/salt/issues/12248#issuecomment-46383026
00:30 forrest ifnull, the installed software I have is identical to what you had yesterday
00:32 forrest ifnull, are you going to spin up a 2014.1.3 instance tonight?
00:32 ifnull yeah
00:32 ifnull im doing it now
00:32 forrest cool
00:32 ifnull im using vagrant
00:33 forrest yea I saw that from your last comment
00:33 forrest was reading back through them over the past few minutes
00:34 ifnull which also mean that i’m using the stable salt-bootstrap
00:35 forrest I don't *think* that should matter
00:36 ifnull on a side note, does anyone know when we can expect to see this commit in stable? https://github.com/saltstack/salt-bootstrap/commit/18f42b648f3f4af4a8d12208a54c4df92b88c965
00:37 ifnull preseeding is currently broken for vagrant users since the vagrant plugin no longer allows you to specify a different url for bootstrap
00:37 forrest ifnull, I'd just ask s0undt3ch
00:37 forrest he's in IRC, just not in this channel
00:37 forrest if you ping him and let him know that it would probably be good
00:40 forrest ugh looking at the bootstrap issues, we need to go back and clean those up too :(
00:43 forrest ifnull, I'm gonna head out, let me know what happens
00:43 forrest or just update the issue
00:43 forrest I'll see it
00:44 ifnull yeah. i rely heavily on bootstrap at the  moment. we are however looking at docker for bootstraping since it has been added to vagrant
00:47 Joseph ifnull: docker is annoying
00:47 Joseph lol
00:47 ifnull im having trouble getting into it
00:47 ifnull :)
00:47 Joseph i have general idea of actually how it works now
00:47 Joseph and i "get" the idea
00:48 Joseph but because its essentially just modifications to special isolated filesystem it really doesn't have the concept of a process space and memory
00:48 Joseph kinda drives me crazy
00:48 Joseph however it could just be me not liking something new
00:49 Joseph but trying to get a hold of stdout and debug why an application keeps on dying in a container ...me no likey
00:49 Joseph and then don't get me started about "one entry point!!"
00:49 Joseph the hoops yuou have to jump through to get multiple processes to run in a single container
00:49 Joseph blah
00:50 ifnull good to know
00:50 oz_akan_ joined #salt
00:50 Joseph ifnull: in practice you basically need to use supervisord because a container dies the moment that a command "completes".
00:51 Joseph supervisord acts as the entry point which happens to start other processes etc
00:51 Joseph and never "stops"
00:52 Joseph ifnull: personally a stripped down VM + vagrant sounds much easier to use but hey maybe i just don't get it and need to accept that docker containers are so much more light weight that they get VMs but
00:52 Ryan_Lane is there any way to tell the service state to use different commands for start/stop/restart?
00:52 Joseph though the one thing i have noticed is that once you actually start using containers you actually end up needing to add so many more packages that the size gets close to a small VM real quick
00:53 matrix3000 joined #salt
00:53 Joseph Ryan_Lane: wouldn't that be based on the actual state module implementation for the linux distro?
00:54 Joseph at the end of the day "service.running" is going to do a service status and service start
00:54 otter768 joined #salt
00:54 * Ryan_Lane nods
00:55 Ryan_Lane puppet has an ability to pass in start/stop/restart/has_restart, etc
00:55 Ryan_Lane so that you can call things differently, in case you need to
00:55 Joseph Ryan_Lane: sounds like a git PR :). Just make it a keyword argument with a default value of none
00:55 Ryan_Lane has_restart allows you to use stop/start, rather than restart, for instance
00:55 Ryan_Lane yeah, I don't have a really strong need for it just yet
00:56 Ryan_Lane in this one case I'm going to modify the init script
00:56 dangra joined #salt
00:57 ajprog_laptop joined #salt
00:57 ajprog_laptop joined #salt
00:57 rgarcia_ joined #salt
01:03 rgarcia_ joined #salt
01:04 mgw joined #salt
01:07 piv joined #salt
01:12 dccc joined #salt
01:14 malinoff joined #salt
01:17 schimmy joined #salt
01:29 yomilk joined #salt
01:30 yomilk joined #salt
01:31 logix812 joined #salt
01:37 mateoconfeugo joined #salt
01:45 piv joined #salt
01:47 ilbot3 joined #salt
01:47 Topic for #salt is now Welcome to #salt | 2014.1.4 is the latest | SaltStack trainings coming up in SLC/London: http://www.saltstack.com/training | Please be patient when asking questions as we are volunteers and may not have immediate answers | Channel logs are available at http://irclog.perlgeek.de/salt/
01:49 piv joined #salt
01:51 therealGent joined #salt
01:51 therealGent hello. Is there a way to have a command to run in a state file only once?
01:52 JasonSwindle joined #salt
01:58 shaggy_surfer joined #salt
02:00 jalbretsen joined #salt
02:01 travisfischer joined #salt
02:02 huleboer joined #salt
02:14 oz_akan_ joined #salt
02:17 Networkn3rd joined #salt
02:18 tonthon joined #salt
02:21 dsolsona_ joined #salt
02:22 londo__ joined #salt
02:28 Networkn3rd joined #salt
02:28 jamescarr joined #salt
02:32 pugs left #salt
02:32 tonthon joined #salt
02:34 mgw joined #salt
02:39 manfred like, only once forever? or once per state run?
02:39 therealGent manfred, one forever
02:40 manfred depends on what the command does, in helium, you can add an unless or onlyif to any state. But you will need to do something else like watch the state and then touch a file or set a grain that is checked on the next run to know not to run the next time.
02:41 oz_akan__ joined #salt
02:42 n8n joined #salt
02:53 n8n joined #salt
02:53 MatthewsFace joined #salt
02:57 lionel joined #salt
03:02 logix812 joined #salt
03:03 MTecknology joined #salt
03:03 joehoyle joined #salt
03:07 rallytime joined #salt
03:08 aberdine joined #salt
03:13 bemehow joined #salt
03:15 thayne joined #salt
03:16 catpiggest joined #salt
03:17 [diecast] joined #salt
03:17 schimmy joined #salt
03:20 huleboer joined #salt
03:23 ndrei joined #salt
03:25 Guest___ joined #salt
03:38 joehoyle joined #salt
03:39 bemehow_ joined #salt
03:42 dimeshake joined #salt
03:43 mosen joined #salt
03:47 n8n joined #salt
03:50 oz_akan_ joined #salt
03:54 ipalreadytaken joined #salt
03:55 jamescarr joined #salt
04:00 jamescarr joined #salt
04:04 joehh ekristen: new packages making their way onto launchpad now
04:04 joehh hopefully they fix the issue, please let me know if it doesn't
04:05 garthk joined #salt
04:08 mprymek joined #salt
04:12 thayne joined #salt
04:16 mprymek joined #salt
04:20 joehillen joined #salt
04:20 bhosmer joined #salt
04:32 Guest___ joined #salt
04:32 n8n joined #salt
04:33 mateoconfeugo joined #salt
04:35 rawzone joined #salt
04:36 joehoyle joined #salt
04:37 mateoconfeugo joined #salt
04:49 felskrone joined #salt
04:51 jalbretsen joined #salt
04:51 oz_akan_ joined #salt
04:51 anuvrat joined #salt
04:57 lionel joined #salt
05:00 mikber joined #salt
05:21 yomilk joined #salt
05:23 lionel joined #salt
05:29 lionel joined #salt
05:36 lionel joined #salt
05:36 n8n joined #salt
05:44 lionel joined #salt
05:49 lionel joined #salt
05:52 oz_akan_ joined #salt
05:57 travisfischer joined #salt
05:57 mprymek joined #salt
05:59 jamescarr joined #salt
06:00 yomilk joined #salt
06:01 Katafalkas joined #salt
06:02 ndrei joined #salt
06:02 thayne joined #salt
06:02 lionel joined #salt
06:07 yomilk joined #salt
06:19 Katafalkas joined #salt
06:19 lionel joined #salt
06:19 jcsp joined #salt
06:22 mephx joined #salt
06:23 picker joined #salt
06:23 yomilk joined #salt
06:23 travisfischer joined #salt
06:23 Guest___ joined #salt
06:23 garthk joined #salt
06:23 mosen joined #salt
06:23 dsolsona_ joined #salt
06:23 timc3 joined #salt
06:23 ksalman joined #salt
06:23 davet1 joined #salt
06:23 ekristen joined #salt
06:23 rigor789|away joined #salt
06:23 workingcats joined #salt
06:23 dcolish joined #salt
06:23 zz_cro joined #salt
06:23 sroegner joined #salt
06:23 yidhra joined #salt
06:23 wigit joined #salt
06:23 ilako joined #salt
06:24 t0rrant joined #salt
06:24 harkx joined #salt
06:25 lionel joined #salt
06:28 m1crofarmer joined #salt
06:30 pmcg joined #salt
06:30 linjan joined #salt
06:31 lionel joined #salt
06:32 googolhash joined #salt
06:32 schimmy joined #salt
06:37 Katafalkas joined #salt
06:37 lionel joined #salt
06:40 greyhatpython joined #salt
06:43 lionel joined #salt
06:49 lionel joined #salt
06:52 oz_akan_ joined #salt
06:53 jalaziz joined #salt
06:53 yomilk joined #salt
06:55 lionel joined #salt
06:58 slav0nic joined #salt
07:01 pressureman joined #salt
07:08 zain_ joined #salt
07:12 pmcg joined #salt
07:16 briner joined #salt
07:19 ml_1 joined #salt
07:19 yomilk joined #salt
07:19 yomilk joined #salt
07:19 briner I've seen that salt can use python as a renderer. But is it an interactive way to play with it. Could I launch a salt-python, where I get the __pillar__ and also the __grains__. That way it would be more fun to learn
07:28 chiui joined #salt
07:31 Katafalkas joined #salt
07:32 stoffell joined #salt
07:34 markm joined #salt
07:52 jdmf joined #salt
07:53 oz_akan_ joined #salt
07:57 Katafalkas joined #salt
07:57 Lomithrani joined #salt
08:01 alanpearce joined #salt
08:06 vu joined #salt
08:06 harkx joined #salt
08:08 darkelda joined #salt
08:09 rjc joined #salt
08:10 mrTango joined #salt
08:11 stoffell joined #salt
08:15 jY is there anyway to do math in a template? or a state off a grain item
08:18 babilen jY: You can do that easily with Python. What exactly are you trying to do?
08:18 jY trying to set a template val for 90% of my memory on the server
08:19 babilen template == ?
08:20 babilen jY: Do you mean configuration file?
08:21 rawzone_ joined #salt
08:21 jY ya trying to get 90% of my memory into a var or such.. so i can use it in a template
08:21 babilen jY: Please be precise! You need the value of 90% RAM in a configuration file that you manage via salt's file.managed state and render that with template=jinja ?
08:22 babilen Is that correct?
08:22 jY yes
08:22 babilen jY: Okay, unfortunately you cannot render files with Python (I am not sure why, but so far only jinja, mako, and wempy are supported)
08:23 jY ya.. if i can't figure out how to do it in jinja.. i'll just make a grain for it
08:23 jY just trying not to do that
08:23 jhauser joined #salt
08:23 babilen That means that you have to compute that value somewhere else. I would use a pillar written in Python for that in which you do the computation based on grains
08:24 babilen An external grain would also make sense, but you probably want to set other values in the pillar anyway
08:24 jY ok thanks
08:26 babilen A pillar is essentially just a Python dictionary. And you could implement something like: http://paste.debian.net/105589/
08:28 harkx joined #salt
08:30 babilen jY: grains['mem_total'] is what you are looking for btw
08:31 younqcass joined #salt
08:31 jY ya i know that grain.. its why i was wondering if jinja could do match
08:31 jY math*
08:31 jY but doesn't look like it can
08:33 babilen jinja is extremely limited in what it can do and I would just use Python whenever jinja gets awkward
08:33 babilen I personally prefer some nicely written Python code over some horribly unreadable jinja hack :)
08:34 jY ya i'll go the pillar or grain route
08:34 Lomithrani can I set a variable in a state file in order to use it many time in the file ?
08:35 ashb jinja can do math http://jinja.pocoo.org/docs/templates/#math
08:35 babilen jY: I'd argue that "90% RAM" is not a very good external grain ... You will probably want to set additional values in the pillar anyway and it is trivial to do the computation there ..
08:35 babilen ashb: Oh! So it can then :)
08:35 ashb not saying its a good idea, just that it can
08:35 babilen ta!
08:37 Lomithrani doing math in jinja is quite a struggle if you want to have the right of decimals and such
08:37 davet joined #salt
08:37 Lomithrani last time I did it I ended up with :  {{ ("%f" % ((((grains['id']|replace("Cassandra", "")|float)-1)*(2**127)) / number_of_cassandra)).rstrip('0').rstrip('.')   }}
08:37 Lomithrani (and thats like a very simple calculation ...)
08:40 Lomithrani so is it possible to set a variable in a state file ? can I add someting in grains for example ?   something like  :grains['myVar']="hey" and then use {grains['myVar']
08:42 babilen Lomithrani: I'd argue once again that you would want to write that in Python not jinja
08:42 malinoff Lomithrani, {% set name = value %}
08:43 malinoff jY, or you can use mako which will give you access to the python right in the file
08:43 babilen Lomithrani: But setting variables is discussed in http://jinja.pocoo.org/docs/templates/#assignments
08:44 Lomithrani thanks , was lost in the salt doc ^
08:44 babilen malinoff: I decided against mako. I would have much preferred it to be the default templating engine in salt, but I'd rather have jinja + py/pydsl than having to maintain things in two different markup languages + py + ...
08:44 antonw joined #salt
08:45 Katafalk_ joined #salt
08:45 malinoff babilen, that's why my suggestion is to drop jinja at all and use mako everywhere
08:45 babilen +1000
08:46 babilen A start would be to actually use #!jinja|yaml everywhere and not rely on the default. That way "legacy" jinja code (of which there is a lot) would continue running even if the default renderer is changed
08:47 babilen jinja was a mistake IMHO and will continue to bite the salt community
08:47 malinoff yep
08:48 malinoff I heard that salt guys used jinja as default, because it is very similar to the Django template engine. When I said that salt is used much more by administrators than Django developers, I had no answer :)
08:49 babilen jinja is meant for people who can't code nor want to learn it
08:49 babilen It's easy and dumbed down
08:49 malinoff It's not really easy, lol
08:49 malinoff of course, you can write simple thing with it
08:49 babilen There's that too
08:49 malinoff but when you try to make a complex thing, jinja is hell
08:50 babilen I mean the whitespace handling is just nuts .. Either way: The people that are likely to use salt are also people who could easily handle mako (or even nice Python DSLs)
08:50 malinoff "Oh, goddamn, fuck those whitespaces, I'll write this in python"
08:50 babilen hehe
08:50 babilen The only thing I need now is "file.managed: ... - template: py"
08:51 babilen I would make rendering configuration files with, for example, ConfigParser much easier
08:51 kermit joined #salt
08:51 babilen But I think it is to late. salt is stuck with jinja now
08:51 babilen Which is a shame
08:51 malinoff yeah, too many places to change
08:51 malinoff examples, code snippets, formulas, etc
08:52 TyrfingMjolnir joined #salt
08:53 malinoff Also, I really don't get the idea about spending so many time on implementing RAET, when you can just take amqp where everything you need is present
08:53 babilen Is there a way to debug the compuration of grains? grains['virtual'] is wrong across our infrastructure and I don't really understand why as the Code looks quite alright
08:54 malinoff babilen, print('here'); print('there')
08:54 oz_akan_ joined #salt
08:54 malinoff or even log.debug('here'); log.debug('there')
08:54 babilen malinoff: So I essentially copy the upstream grain into _grains on the master and then tinker with that and see that output in "salt-minion -ldebug" ?
08:55 malinoff babilen, guess so
08:55 malinoff I never tried :)
08:55 babilen neither have I
08:56 malinoff I don't use salt since 0.17.1, so I can't really say what can be done and what can't right now
08:56 oz_akan__ joined #salt
08:56 babilen It's just that virtual should be "xen" or "kvm" and not "physical" for a bunch of boxes and I can't quite figure out why the grain is not correct (the code looks alright, but that function is quite complex (like a lot of functions in the codebase)
08:56 alanpear_ joined #salt
08:57 malinoff I can't imagine why do you have to know is this a physical machine, or a virtual :)
08:57 babilen malinoff: Because certain states should only be targeted to xen dom0 (and similar things)
08:58 malinoff Why? What is the difference?
08:58 babilen malinoff: The difference is that software to manage domUs is needed on the dom0 for example
08:58 jab416171 joined #salt
08:59 babilen Either way the grain is wrong and I'd like to know why
09:00 malinoff I don't get it. Do you want to manage a XEN virtual machine, or XEN supervisor system?
09:01 babilen malinoff: I want to install software to manage guest boxes only on the supervisors
09:01 alanpearce joined #salt
09:02 malinoff babilen, and supervisor is a physical machine, right?
09:02 mprymek joined #salt
09:02 babilen malinoff: A physical machine that has been booted with a XEN kernel
09:02 babilen (for xen)
09:02 babilen kvm behaves differently in that case
09:03 malinoff babilen, alright, so you want to target minions that were started on that supervisor?
09:03 babilen No
09:04 babilen I want to target all xen dom0 (xen supervisor) and install managexenguests on them and kvm supervisors and install managekvmguests on those
09:04 babilen (and so on)
09:05 babilen For that I want the "virtual" grain to be "xen" and "virtual_subtype" to be "Xen Dom0" -- That is, unfortunatly not the case and I don't quite understand why (as the code seems to do the right thing)
09:05 malinoff grains['virtual'] defines whether the minion is running within a hypervisor (e.g. XEN)
09:05 rtucker joined #salt
09:06 malinoff If you want to manage a supervisor (which is a physical machine), of course grains['virtual'] will be 'physical'
09:08 dmorrow joined #salt
09:08 babilen No, it should be "xen" and "virtual_subtype" should be "Xen Dom0" for supervisors and "Xen [PV|HVM] DomU" for guests
09:11 malinoff I don't know much about XEN, but I thought it is not like ESXi, you can freely install it on a linux system, and that linux will be supervisor, and XEN will be hypervisor. If you have minion installed on the supervisor, it will have 'physical' grain, and minions that started within the hypervisor will have 'XEN' grain
09:11 Flusher joined #salt
09:12 malinoff Gotta go, hope you will figure out what's going on
09:13 chiui joined #salt
09:17 TyrfingMjolnir joined #salt
09:20 ndrei joined #salt
09:22 ggoZ joined #salt
09:23 kupric joined #salt
09:29 marnom joined #salt
09:30 CeBe joined #salt
09:33 Katafalkas joined #salt
09:40 giantlock joined #salt
09:42 chokosabe joined #salt
09:43 chokosabe Hi all
09:43 chokosabe getting a problem when running salt-call --local state.highstate -l debug
09:43 chokosabe File "/usr/lib/python2.7/dist-packages/salt/payload.py", line 95, in loads     return msgpack.loads(msg, use_list=True)   File "_unpacker.pyx", line 119, in msgpack._unpacker.unpackb (msgpack/_unpacker.cpp:119) msgpack.exceptions.UnpackValueError
09:43 chokosabe paste.bin is here:
09:44 chokosabe http://pastebin.com/EQ884c2W
09:44 bhosmer joined #salt
09:44 chokosabe Initially thought it was a msgpack issue but looks like its not
09:52 Sypher_ joined #salt
09:56 oz_akan_ joined #salt
10:03 chokosabe Any ideas from anyone at all?
10:06 babilen chokosabe: Which version of salt is that?
10:07 dsolsona joined #salt
10:07 chokosabe babilen: straight form ubuntu 14.04
10:07 chokosabe babilen: x64 build
10:08 babilen Okay, which version of salt ships with Ubuntu these days?
10:08 babilen You could run "salt --versions-report"
10:08 jY 0.17.5 i think
10:09 babilen That's mighty old
10:09 chokosabe babilen: Package: salt-common Status: install ok installed Priority: extra Section: admin Installed-Size: 4171 Maintainer: Ubuntu Developers <ubuntu-devel-discuss@lists.ubuntu.com> Architecture: all Source: salt Version: 0.17.5+ds-1 Depends: python (>= 2.7), python (<< 2.8), python:any (>= 2.7.1-0ubuntu2), python-jinja2, python-yaml, python-pkg-resources Recommends: lsb-release, python-mako, python-git Suggests: salt-doc (= 0.17.5+d
10:10 babilen It might not work with the msgpack version in Ubuntu....
10:11 chokosabe babilen:            Salt: 0.17.5          Python: 2.7.6 (default, Mar 22 2014, 22:59:56)          Jinja2: 2.7.2        M2Crypto: 0.21.1  msgpack-python: 0.3.0    msgpack-pure: Not Installed        pycrypto: 2.6.1          PyYAML: 3.10           PyZMQ: 14.0.1             ZMQ: 4.0.4
10:11 chokosabe babilen: its showing msgpack-pure as not being installed
10:13 chokosabe babilen: Do you think that is needed? There doesn't seem to be an ubuntu package for it, there is a pypi python package though
10:13 babilen chokosabe: Well, I would recommend to install salt not from the Ubuntu repositories, but the current version as detailed in http://docs.saltstack.com/en/latest/topics/installation/ubuntu.html
10:14 tinuva joined #salt
10:16 chokosabe babilen: Perfect! Thanks, that seems to have worked
10:17 babilen chokosabe: Perfect (that is still a bug in the version that is being shipped in Ubuntu though)
10:19 chokosabe babilen: My problem was that I couldn't find anyone else that had ahad the same problem. But this is an Ubuntu bug rather than a salt bug...
10:21 babilen yeah
10:21 Katafalk_ joined #salt
10:22 Katafalk_ joined #salt
10:34 Katafalkas joined #salt
10:34 bhosmer joined #salt
10:39 ay left #salt
10:41 ekristen joehh: fantastic
10:41 ekristen joehh: I’ll test today
10:45 Fa1lure joined #salt
10:57 oz_akan_ joined #salt
10:58 rawzone joined #salt
11:04 ndrei joined #salt
11:06 bezaban joined #salt
11:06 xmj viq: so, of course, we use MySQL as gitlab's DB engine.
11:06 xmj which isn't supported by the state as-is.
11:07 xmj Someone *cough* should make it.
11:08 viq xmj: https://github.com/viq/gitlab-formula/pull/6/files
11:08 viq Yeah, I'm a slacker...
11:08 MZAWeb_ joined #salt
11:08 younqcass joined #salt
11:08 xmj what can i do to make you integrate that one?
11:09 mprymek joined #salt
11:09 xmj also it seems to be missing -at-least-one- dependency on mysql in packages.sls
11:10 xmj MTecknology: ping
11:11 DuckAssassin joined #salt
11:11 viq xmj: I'm working on updating https://github.com/viq/gitlab-formula-1
11:11 viq I want to make it depend on formulas (which proved to be near impossible for redis)
11:12 viq I also would like to make a pillar accept one of mysql, postgresql and possibly postgres, and set proper values in files depending on that (instead of having to put mysql2 in pillar as is in that pull request)
11:13 xmj postgresql,postgres -- uh what?
11:13 xmj viq: assume i have two-three hours right now to contribute to that one.
11:13 xmj how do i do it?
11:14 viq xmj: ok, forget postgres. Let me think for a moment.
11:14 xmj please do
11:15 xmj (also, check PMs :)
11:15 CeBe joined #salt
11:16 viq :)
11:17 viq So, high level first
11:17 techdragon joined #salt
11:18 marcinkuzminski joined #salt
11:18 viq Ideal world: user puts in pillar db_engine: mysql or postgresql. Formula tries to depend on appropriate formula, and sets appropriate variables inside.
11:19 xmj bingo
11:19 xmj * sets appropriate variables for package dependencies
11:19 xmj * puts appropriate variables in generic config files
11:20 viq Further in: that requires ensuring proper mysql|postgresql databases are created. I guess that can be done by having gitlab.(mysql|postgresql) states, that get selected depneding on pillar
11:20 xmj * puts db-specific config files in place.
11:21 Outlander joined #salt
11:21 viq The variety between config file for postgres and mysql is not that big. Mainly encoding, and adapter - and here's first annoying thing, for mysql the adapter is called mysql2
11:21 TyrfingMjolnir joined #salt
11:21 viq The pull request mostly does that
11:22 viq Another thing to keep track of, the 'bundle install' commands are different depending on db_engine value, again pull request takes care of that.
11:24 viq I *think* that the (mysql|postgresql)-formulas have support for creating databases, so this formula wouldn't even have to do that - but that would need to be documented in readme
11:25 xmj Ok
11:25 xmj Do I want to fork one of your repos (which?), address the open issues that are related (which? :), test it and submit a Pull request to you (as upstream) ?
11:26 viq Ugh, it's a bit of a mess at the moment... Let me have a look to see what would be best
11:26 JasonSwindle joined #salt
11:27 xmj right, well, let's unmess it
11:27 xmj (check your PM for the reason)
11:27 jY is it possible to write tests around state files or do i pretty much have to use something like vagrant to verify on a vm?
11:28 viq Yeah, saw. Just continuing conversation in the public, because I see no reason not to, someone might chime in ;)
11:28 viq jY: look up salt-kitchen
11:28 jY thansk
11:29 bhosmer joined #salt
11:29 briner hello, I try to use the renderer py for the .sls as describe here (http://salt.readthedocs.org/en/latest/ref/renderers/#multiple-renderers). I've put it under /srv/salt/base/mako.sls, and the salt spit errors:root@unixpriv3:/srv/salt/salt/base# salt lunidebrepos1.unige.ch state.sls mako
11:29 briner lunidebrepos1.unige.ch:
11:29 briner Data failed to compile:
11:29 briner ----------
11:29 briner No matching sls found for 'mako' in env 'base'
11:29 briner
11:29 tristianc|Phone joined #salt
11:29 briner does some of you have experience on using py for the renderer of state sls ?
11:30 xmj viq: exactly
11:30 babilen briner: You just use "#!py" in the first line
11:30 briner cat mako.sls
11:30 viq xmj: I guess ignore viq/gitlab-formula ; viq/gitlab-formula-1 is a fork of saltstack-formulas/gitlab-formula to make it easier for pull requests and such. I guess "best" approach could be to grab what I have at viq/gitlab-formula-1 and run with it, making a pull request at saltstack-formulas/gitlab-formula. I am notoriously bad at doing things in any reasonable amount of time, so let's keep it out of my hands.
11:30 briner #!py
11:30 briner def run():
11:30 briner '''
11:30 briner Install the python-mako package
11:30 briner '''
11:30 briner return {'include': ['python'],
11:30 briner 'python-mako': {'pkg': ['installed']}}
11:30 babilen (why do you call it mako.sls and why is it in base/mako.sls ?
11:31 viq briner: paste it online somewhere, please
11:31 babilen briner: And please use a pastebin such as http://paste.debian.net if you want to paste more than two lines :)
11:31 xmj viq: alright
11:31 xmj let's see how far i get during the next three hours.
11:31 * xmj has a challenge
11:31 xmj WOOOHOOO
11:31 viq :D
11:31 briner babilen, yep... this is what I copy and paste from http://salt.readthedocs.org/en/latest/ref/renderers/#multiple-renderers.
11:31 briner babilen, Ok, I'll use pastebin
11:31 chiui joined #salt
11:32 babilen briner: Why is it in base/mako.sls ? It can't find that sls .. there is no problem with its content
11:32 babilen briner: That would be "- base.mako" in your top file
11:33 babilen (I mean it depends on your master config as I have no idea how you configured your file_roots and so on)
11:34 babilen I mean you might very well have "file_roots: base: - /srv/salt/base" in there
11:34 briner babilen, I'm jut using this: http://salt.readthedocs.org/en/latest/ref/renderers/#multiple-renderers. I dunno what is mako...is it also a renderer.
11:34 babilen briner: Why did you create that file in /srv/salt/base/mako.sls and not in /srv/salt/mako.sls ?
11:35 briner babilen, as I'm specifying the state I'm interested, I do not have the obligation to put it in the top.sls
11:35 TyrfingMjolnir joined #salt
11:35 babilen briner: You would have to run " salt lunidebrepos1.unige.ch state.sls base.mako" *unless* you configured your file_roots as detailed above.
11:36 babilen briner: So, please answer my question(s)
11:36 briner babilen, I'm a bit a newbie... I put all the config in /srv/salt/base for the state
11:36 babilen briner: What do you have in your master config?
11:37 viq xmj: I won't be paying as much attention to this window, so hilight or PM me if you want my attention sooner
11:37 vandemar joined #salt
11:38 briner babilen, http://paste.debian.net/105621/
11:38 xmj viq: understood
11:39 babilen briner: Ah, so you *do* use multiple environments
11:39 briner and I know that salt & pillar are in /srv/salt/salt & /srv/salt/pillar respectively
11:39 briner At that time I only use the base one.
11:39 TyrfingMjolnir joined #salt
11:40 babilen briner: Wait, you want that to be "/srv/salt/salt/base/mako.sls" and not "/srv/salt/base/mako.sls"
11:40 babilen briner: Why do you use salt/salt ?
11:40 jrdx joined #salt
11:41 briner babilen, no... in the chat to be more "normal" I say that the file is in /srv/salt/base/mako when actually it is on /srv/salt/salt/base/mako.sls
11:41 babilen briner: Seriously, I would *strongly* advise you to simplify that. Just use /srv/salt/base or /srv/salt for the base environment and don't define the others.
11:42 babilen briner: So you change information you present here in order to fit your expectations when this is *exactly* the error source? What else did you change?
11:42 briner it works fine if I do not use a shebang !#py and use a standard YAML sls insteaed
11:42 babilen It is #! not !#
11:42 briner babilen, sorry: #!py
11:43 babilen briner: I have no way to help you if you don't provide actual information about your setup.
11:43 babilen (nor am i interested in trying to guess which parts you changed and which actually reflect the current situation)
11:43 briner babilen, what will you need ?
11:44 babilen Some actual data, commands and configuration.
11:44 babilen But now I'll have lunch, you had your chance!
11:44 briner babilen, have a good lunch
11:45 TyrfingMjolnir joined #salt
11:47 Lomithrani joined #salt
11:47 rawzone_ joined #salt
11:50 dsolsona joined #salt
11:50 jamescarr joined #salt
11:58 oz_akan_ joined #salt
12:00 xmj anyone know if i can make an - unless: statement conditional on a pillar?
12:01 xmj e.g. run psql derp foo bar 'select * from baz;' if the pillar provides 'postgresql', mysql derp foo bar otherwise?
12:02 viq xmj: wrap whole cmd in {% if %}  ?
12:02 viq Or just the - name
12:03 jas-_ joined #salt
12:03 xmj ah
12:03 xmj ofc
12:04 cofeineSunshine {% set mongos = salt['pillar.get']('mongos') %}
12:04 cofeineSunshine {% if mongos %}
12:04 cofeineSunshine in state i use this
12:04 cofeineSunshine {% endif %}
12:04 cofeineSunshine in state
12:06 alanpearce joined #salt
12:10 xmj right
12:14 Lomithrani hey how is that possible : "No such file or directory" http://pastebin.com/CSx021JF when I clearly stated http://pastebin.com/SsbRD7WT that it requires the said file , shouldnt it have either said prerequisite wasnt done or find it ? what did I do wrong  ? :(
12:16 Lomithrani http://pastebin.com/cNQdQMDR
12:18 viq Lomithrani: do you have /opt/CaSsAndra_CluSter_01 ?
12:18 martin3k joined #salt
12:18 martin3k hi #salt, i got a problem using the mysql module
12:18 Lomithrani no why should have ?  its just an argument
12:19 viq oh, right
12:19 Lomithrani might be that its not the file that it doesnt find but the "." ?
12:19 viq Lomithrani: try changing the name to absolute path?
12:19 TyrfingMjolnir joined #salt
12:19 robinsmidsrod anyone got any idea what I'm doing wrong trying to get pillar to work? https://gist.github.com/robinsmidsrod/dfca8d96c84bd674ff7b
12:19 martin3k this call: http://paste.debian.net/105633/
12:20 Lomithrani viq: isnt the cwd salt argument here to precise the path ?
12:20 martin3k produces this output: http://paste.debian.net/105634/
12:20 viq robinsmidsrod: can you paste your pillar?
12:20 rawzone joined #salt
12:20 Lomithrani viq: would you suggest me to set cwd : /   and to put the absolute path like . /opt/change_name.sh Cassandra_cluster_01 ?
12:21 viq Lomithrani: dunno about cwd, but I would see if . likes/dislikes relative path
12:21 robinsmidsrod viq: I've updated the gist, just refresh it
12:21 martin3k i just changed the password hash to paste the file
12:22 jamescarr_ joined #salt
12:22 viq robinsmidsrod: try putting the email address in quotes
12:23 robinsmidsrod viq: single or double quotes?
12:23 viq robinsmidsrod: I don't think it matters, I think usually single quotes are used
12:24 robinsmidsrod viq: didn't seem to help, but when I run state.show_highstate I get another error: TypeError encountered executing state.show_highstate: 'bool' object is not iterable. See debug log for more info.  Possibly a missing arguments issue:  ArgSpec(args=['queue'], varargs=None, keywords='kwargs', defaults=(False,))
12:24 babilen Yeah, exactly. It would break if you have quoted strings (e.g. abc."defghi".xyz@example.com ) as part of the mail address
12:25 babilen (but who would do that?)
12:25 blarghmatey joined #salt
12:25 robinsmidsrod viq: the really odd thing is that it breaks on one ubuntu lucid machine, but works on another
12:26 babilen I am rather surprised that this (pillar?) breaks on syncing the modules. Are you sure that alias.sls is the one to blame here?
12:27 robinsmidsrod babilen: no, I'm not sure that is the error, but I'm having problems tracking down what exactly is causing the error
12:27 babilen robinsmidsrod: Yeah, okay .. It actually seems to break when calculating top_matches
12:27 jamescarr joined #salt
12:27 robinsmidsrod babilen: yeah, that's sorta where I've gotten to as well
12:28 babilen robinsmidsrod: Is this your pillar? Does it work if you s/myemail@mydomain.com/foo ?
12:29 robinsmidsrod babilen: still get the error with that modification
12:30 babilen empty file?
12:30 babilen master/minion version involved here btw?
12:32 jamescarr_ joined #salt
12:32 robinsmidsrod babilen: indeed, the broken host is on 2014.1.4, while the working ones are on 2014.1.5
12:32 robinsmidsrod maybe I should update :)
12:33 babilen Your master is at 2014.1.5 and the non-working minion on 2014.1.4 ?
12:33 babilen (working minions on 2014.1.5 naturally)
12:34 xmj I might just make it.
12:34 xmj terrific
12:34 babilen w00t!
12:34 Fa1lure which version of salt is codename Helium?
12:34 babilen unreleased as of now
12:34 robinsmidsrod babilen: upgrading the failing minion to 2014.1.5 solved the issue
12:35 ipmb joined #salt
12:35 babilen robinsmidsrod: And your master is on .5 ?
12:35 robinsmidsrod babilen: yup
12:36 TyrfingMjolnir joined #salt
12:36 robinsmidsrod it was released today, wasn't it?
12:36 babilen Why are there no .5 release notes. *again*
12:36 * babilen checks github
12:39 xmj viq: so
12:39 xmj viq: testing https://github.com/xmj/gitlab-formula-1/commit/49a40448e08743070076a049a4185e71f1d4e34e
12:39 xmj viq: let me know if you have any ideas.
12:40 rawzone_ joined #salt
12:43 viq xmj: looking
12:43 analogbyte joined #salt
12:43 viq xmj: you commited your .swp file ;)
12:45 viq xmj: what do we want users to set in pillar as db_engine? mysql, or mysql2 ?
12:46 rawzone joined #salt
12:46 xmj mysql2 looks good
12:46 xmj derp. i love gitignore!
12:47 viq mhmm. I'm just thinking of "why I have to set mysql2 when I want to use mysql?" but it makes things easier internally I guess
12:47 xmj exactly
12:47 xmj derp
12:48 JasonSwindle joined #salt
12:48 xmj there was this gitrepo, hang on
12:48 babilen ffs
12:48 xmj viq: github.com/Tinuva/gitlab-formula
12:48 xmj which had good ideas.
12:48 babilen https://github.com/saltstack/salt/issues/11877 seems to break xen detection :-/
12:48 xmj :q
12:49 babilen Just upgraded all minions to .5 and grains['virtual'] content changed
12:49 viq xmj: yeah, haven't looked in much detail at it yet though
12:50 viq babilen: hah, and that was my (well, our) bug ;)
12:51 blarghmatey joined #salt
12:51 xmj fair enough
12:51 babilen viq:  if os.path.isfile('/proc/self/status'):
12:51 babilen Should be "elif ..."
12:51 viq babilen: I just reported, didn't touch code at all
12:51 babilen Oh, where is that report?
12:51 viq babilen: https://github.com/saltstack/salt/issues/11877
12:52 * babilen reloads the page
12:52 viq just == only
12:52 babilen Ah, so you are to blame?!
12:52 viq "all i did was report, I did not touch the code"
12:52 viq I guess, kinda :P
12:52 babilen how dare you get other people to write broken code!
12:53 to_json joined #salt
12:54 bhosmer joined #salt
12:54 viq :P
12:55 martin3k how do i reset or override the mysql root passwd?
13:00 ndrei joined #salt
13:01 rawzone_ joined #salt
13:02 xmj you'll find this on stackoverflow, i'm sure.
13:05 rawzone joined #salt
13:06 xmj i must be doing something right. state.highstate test=True takes forever \o/
13:09 oz_akan_ joined #salt
13:11 viq :D
13:12 babilen https://github.com/saltstack/salt/issues/13534 trying to come up with a fix .. It is clear what the issue is, but I have no OpenVZ boxes around to test that also.
13:12 MTecknology xmj: pong
13:13 MTecknology belated pong?
13:13 viq babilen: obviously I have some, you poke me with a diff I can run it on a couple
13:15 babilen viq: yeah, just testing variations here ...
13:15 xmj MTecknology: belated pong; we're mutilating viq and your gitlab state to make it run on centos 6.5 with mysql
13:15 MTecknology oh
13:15 xmj yes
13:15 MTecknology I have it using an external mysql db
13:16 xmj succeeded: 6; failed: 37.
13:16 xmj my bad
13:16 MTecknology I still need to build a formula to install gitlab-ci, I just haven't had time :(
13:16 xmj derp
13:16 racooper joined #salt
13:17 joehh babilen, xmj if come up with an appropriate patch, I'm happy to release updated packages including it
13:17 joehh otherwise, hassle saltstack for a .6...
13:18 xmj joehh: don't worry.
13:18 babilen joehh: It's just that they messed up the logic in there.
13:18 xmj i'm good at nagging people.
13:18 xmj need to get it working locally first ...
13:18 MTecknology I have a MASSIVE list of stuff that needs to happen before I'll have time for that... Replacing a netware server with linux servers and rebuilding the logic to how drives get assigned during windows logons and such is just a small line item on my to do... :(
13:18 viq MTecknology: sounds "fun"
13:18 rawzone_ joined #salt
13:18 viq MTecknology: total OT, what do you use to keep track of tasks and priorities?
13:19 xmj vim /etc/motd
13:19 MTecknology My OT pay (yay hourly!!) is ~1/3 of my take home so far this year
13:19 MTecknology viq: vim
13:20 xmj OT pay = ??
13:20 viq overtime ?
13:20 MTecknology overtime
13:20 xmj ah
13:20 viq MTecknology: ah, ok. I'm playing some with taskwarrior, quite liking it
13:21 xmj How do I make salt explicitly list *which* prerequisites failed?
13:21 MTecknology Actually, I'm consolidating all of my list items from many written notes right now
13:22 xmj That is one damn good idea.
13:22 xmj (i'm too scared to implement it.)
13:22 diegows joined #salt
13:23 viq xmj: the consolidated todo list?
13:23 * viq recommends http://tasktools.org/
13:24 xmj sudo vim /etc/motd. but there's too much stuff to put in.
13:24 viq I kinda started making a port for taskd for OpenBSD, though user management there is... somewhat complicated
13:27 ecdhe joined #salt
13:27 martin3k how do i enable the salt-minion mysql support?
13:28 viq martin3k: install the package listed in requirements, restart minion
13:29 manfred or just reload the modules
13:29 manfred shouldn't require a full restart
13:31 xmj State git.latest found in sls gitlab.gitlab-shell is unavailable
13:31 xmj Joy.
13:33 viq *blink*
13:33 xmj damn mangled containers. shouldn't have tried to run it on an old one where /home/git/ was already existing (root:root)
13:33 * xmj blinks
13:33 Katafalkas joined #salt
13:34 martin3k viq: well its just the python mysql module, this is installed but i cant manage to set the initial user and password
13:34 felskrone joined #salt
13:34 xmj does anyone know how well salt supports bhyve, these days?
13:34 martin3k http://paste.debian.net/105644/ here is the .sls just in case i am missing something
13:34 JasonSwindle joined #salt
13:36 viq xmj: for controlling, or running in?
13:36 jacksoncage joined #salt
13:36 toastedpenguin joined #salt
13:36 malinoff joined #salt
13:37 jv115 joined #salt
13:37 Katafalkas joined #salt
13:37 xmj yes
13:37 jesusaurus joined #salt
13:37 rawzone joined #salt
13:37 xmj the former would be more interesting to me personally. the detection works, so the latter should be OK
13:38 Katafalkas joined #salt
13:38 tinuva joined #salt
13:40 tinuva_ joined #salt
13:40 viq Just a guess, but the latter should probably look a lot like any other QEMU
13:41 babilen viq: That code is ugly .. why would you compile the RE *again* for every single line :(
13:41 viq babilen: what code?
13:41 Katafalkas joined #salt
13:41 babilen viq: The fix for your bug
13:42 TyrfingMjolnir joined #salt
13:42 viq babilen: oh. I didn't touch any code in relation to that ;)
13:43 viq xmj: hm, http://lists.freebsd.org/pipermail/freebsd-current/2014-March/048996.html
13:43 rawzone joined #salt
13:43 arcleo joined #salt
13:44 xmj viq: the vmrc script allan talks about is a bit of a ... pile of dogpoo :)
13:44 mgw joined #salt
13:45 viq heh :P
13:45 xmj it does too much. it can do all the provisioning of new vms, way beyond start/stop/status you'd usually find
13:45 xmj will have to find out if by now there's a good, cleaned up fork
13:46 miqui joined #salt
13:47 robinsmidsrod is there a good recipe on setting up an openvpn star topology with client certificate authentication?
13:47 robinsmidsrod I have one already, but I think maybe I'm going about the certification generation in a bit too hard way
13:48 jacksoncage hi folks, I'm having issues with dockerio state module with starting a container from salt state and mounting volumes (docker -v), http://serverfault.com/questions/606099/salt-cloud-docker-state-handle-volumes
13:48 viq robinsmidsrod: as an Idea, I found http://garthwaite.org/virtually-secure-with-openvpn-pillars-and-salt.html
13:50 xmj 17 success, 27 fail. getting better!!
13:50 robinsmidsrod viq: thanks - looking...
13:51 zain_ joined #salt
13:51 maxskew joined #salt
13:52 rawzone joined #salt
13:54 MTecknology http://p.lustfield.net/414 <-- current To Do list; probably lacks a lot of stuff, but at least I'm not holding it inside my brain anymore
13:54 rallytime joined #salt
13:54 xmj Failed to create user gitlab@None  << how do i specify which host the user should connect from ('localhost' in this case) ?
13:55 xmj is it as simple as modifying the user to "gitlab@localhost" ?
13:56 viq DB? I guess
13:56 younqcass joined #salt
13:56 xmj - host: pillar.get(...)
13:56 Networkn3rd joined #salt
13:56 xmj learn all the things
13:58 cro joined #salt
14:00 xmj odd return message: These values could not be changed: {'home': '/home/git'}
14:00 TheThing joined #salt
14:01 quickdry21 joined #salt
14:04 JasonSwi_ joined #salt
14:04 * viq thinks it's time to go eat
14:05 xmj mind if i shoot you a question before?
14:08 londo_ joined #salt
14:11 jaimed joined #salt
14:12 aquinas_ joined #salt
14:13 peters-tx joined #salt
14:13 JasonSwindle joined #salt
14:14 londo__ joined #salt
14:19 oz_akan_ joined #salt
14:22 JasonSwindle joined #salt
14:23 jmccree joined #salt
14:26 babilen viq: https://github.com/saltstack/salt/pull/13536 + https://github.com/saltstack/salt/pull/13537
14:27 [diecast] left #salt
14:27 darkelda joined #salt
14:29 Katafalk_ joined #salt
14:31 oz_akan_ joined #salt
14:31 will joined #salt
14:32 babilen joehh: ^^^ (would like to hear some comments on that though. It is a fairly easy fix and they work on Xen and KVM hosts as they should, but I can't test against OpenVZ right now)
14:33 viq babilen: I'll test in an hour or so, and now I should really go eat :P
14:34 babilen You still haven't?
14:34 babilen go go go go!
14:34 danielbachhuber joined #salt
14:35 racooper trying to get gitfs working.  I have the master config set up, the git repo is available, but...when I try to run a test highstate against one host, debug log doesn't show any attempt to contact the remote git server.
14:35 martin3k did anyone encountered this yet?  /etc/salt/master - expected '<document start>', but found '<block mapping start>'
14:35 martin3k in "<string>", line 370, column 1:
14:35 martin3k pillar_roots:
14:37 mgw joined #salt
14:37 babilen racooper: Mind stopping the master and running "salt-master -ldebug" as the user you run the master as? Also paste your master configuration file along with that output to http://paste.debian.net
14:39 vejdmn joined #salt
14:41 CheKoLyN joined #salt
14:43 babilen basepi: Sorry, as a fairly new committer I am not sure if PR's should target develop (and would be backported from there) for bug fixes in 2014.1 or should be based on 2014.1 and would then be cherry-picked into develop.
14:44 babilen I don't mind either way, but that is a bit unclear
14:45 vejdmn joined #salt
14:48 manfred babilen: doesn't matter the order, but usually we fix it in develop and then request to mark it as backported
14:48 manfred or cherry picked
14:48 babilen manfred: Perfect, I'll do that in the future.
14:49 shel3over joined #salt
14:56 thayne joined #salt
14:57 younqcass joined #salt
14:57 mateoconfeugo joined #salt
15:00 xerxas joined #salt
15:00 jnials joined #salt
15:01 scoates is there a way for me to get just the rendered object that *would* be stated? I mean instead of `salt-call state.sls statename`, I'd like to see the object that statename renders to.
15:03 manfred scoates: salt-call state.show_sls statename
15:03 manfred or state.show_low_sls
15:03 scoates thanks!
15:03 scott_w joined #salt
15:03 ajprog_laptop joined #salt
15:03 vejdmn joined #salt
15:04 jergerber joined #salt
15:05 kermit joined #salt
15:09 pdayton joined #salt
15:09 vejdmn joined #salt
15:11 racooper babilen,  prefer gist but here's master log, master conf and output from command: https://gist.github.com/racooper/180a5188d3c2d7864328
15:13 manfred fs            ][DEBUG   ] Fetching from https://git@github.tamu.edu:racooper/ocss-salt-stat
15:13 manfred it loaded it fine
15:14 manfred where is your top.sls file located?
15:14 racooper it's at the root level of the git repo
15:15 JasonSwi_ joined #salt
15:15 thaddeusmt joined #salt
15:16 manfred i see your pillar top.sls...
15:16 xmj viq: https://github.com/viq/gitlab-formula-1/pull/1
15:16 racooper I'm wondering if it's an auth error...apparently our local github doesn't allow anon access...
15:16 xmj viq: review please
15:16 xmj (once you're done eating, yadda yadda)
15:17 manfred racooper: that would be my guess.  Try setting up an ssh key and include the key/configure you /root/.ssh/config to use that sshkey when pulling the information down
15:17 thaddeusmt hello everyone, I am trying to use the 'environments' to deploy different gitfs state branches to different servers (dev, test, prod, etc), and I am having trouble. could anyone help answer a few questions please?
15:17 Katafalkas joined #salt
15:17 vejdmn joined #salt
15:18 JasonSwi_ joined #salt
15:18 racooper was just thinking that.  it's disconcerting when there is no error message when attempting the gitfs access, though.
15:21 racooper using git+ssh protocol I get: "ssh: Could not resolve hostname github.tamu.edu:racooper: Name or service not known"
15:22 thaddeusmt basically, I want to make sure what I am trying to do will even work with environments. I want to have a 'base' or 'prod' set of salt states in the gitfs repo, and have separate 'dev' and 'test' branches containing states that may be a little different from the prod ones, but may also be identical. when a state change is finalized on dev it get promoted to test and prod by merging.
15:22 oz_akan_ joined #salt
15:22 racooper using the URI "git+ssh://git@github.tamu.edu:racooper/ocss-salt-states.git" which is what is provided by the local github.
15:23 thaddeusmt it seemed like this was the exact use case, but I get these errors: Rendering SLS "prod:backups.cron-backup-mysql" failed: Conflicting ID "dev:backups.cron-backup-mysql"
15:23 thaddeusmt it seems like I can't have states with the same names in two separate branches?
15:24 manfred racooper: having never set it up for ssh, i have no idea, but i was under the impression that there were known problems with some ssh stuff and one of the gitfs backends
15:26 timoguin thaddeusmt: that was my block with using gitfs branches too. having unique ids for states across branches seemed to completely defeat the purpose.
15:26 TheThing joined #salt
15:26 racooper ah. replaced the colon in the path between the host and the user with a / and no errors in the log this time. and it worked! Thanks manfred
15:26 manfred nice
15:26 manfred yeah, that would do it heh
15:27 thaddeusmt timoguin: so it's working how it's supposed to? it's just not working the way I am expecting? hmmm. it's really not very useful this way
15:28 timoguin i'm not really sure, as there aren't many examples of multi-environment gitfs out in the wild
15:28 timoguin so i'm just not using multiple environments atm.
15:28 jslatts joined #salt
15:29 manfred thaddeusmt: it is working for me in develop :/
15:30 timoguin manfred: do you have an example in a repo anywhere? i'd like to give it another go soon too
15:30 manfred https://github.com/gtmanfred/salt-states
15:30 manfred i just pushed a dev environment
15:30 manfred with no changes to it
15:31 manfred and everything worked fine when doing a highstate on master/base and dev
15:32 timoguin the top.sls merging makes a normal git merge workflow difficult
15:32 viq xmj: looking
15:32 manfred lemme check that
15:32 manfred if i just keep the top.sls on the master
15:32 timoguin you couldn't actually merge dev to master without having to do a dummy merge on the top.sls
15:32 thaddeusmt manfred: thanks for posting those, I'll take a look. so if you do 'salt 'dev' state.highstate' the dev branch states are run on dev? with no errors?
15:32 jalbretsen joined #salt
15:33 rgarcia_ joined #salt
15:33 manfred salt dev\* state.highstate saltenv=dev
15:33 timoguin i've seen discussion on the mailing list about keeping top.sls separate from gitfs for that reason
15:35 joehh babilen: subtle - I'll wait 24 hours or so (partly for any other issues) and assuming no probs with openvz I'll push up a .5-2
15:36 Katafalkas joined #salt
15:36 xmj :)
15:38 scoates "2014.1.4 is the latest" -> # apt-cache show salt-common | grep Version -> Version: 2014.1.5+ds-1~bpo70+1
15:38 dsolsona joined #salt
15:38 scoates is the topic just out of date?
15:38 manfred it is
15:38 scoates ok, good. thanks.
15:38 manfred kind of
15:38 scoates heh
15:38 manfred 2014.1.5 should be announced today
15:39 scoates ah. ok.
15:39 scoates coincidence
15:39 rawzone_ joined #salt
15:42 scoates bummer my git changes aren't in yet. /-: https://github.com/saltstack/salt/blob/v2014.1.5/salt/modules/git.py
15:43 manfred if you made big changes since january, they probably won't be in until helium
15:43 manfred which should be soonish...
15:43 manfred for rc1
15:44 rawzone joined #salt
15:44 scoates yep. I see. thanks.
15:47 manfred thaddeusmt: the thing that is worth noting, is that you should also have seperate nameing schemes.  The top file should not match '*something.com' for both base and dev, dev should be something like '*dev.something.com' so that it matches your testing environment
15:47 cliffstah joined #salt
15:47 kballou joined #salt
15:49 brandon_ joined #salt
15:49 rawzone joined #salt
15:53 viq xmj: just one comment so far
15:54 viq babilen: still want me to test this against OpenVZ ?
15:56 joehh viq: that would be great
15:56 viq OK
15:56 travisfischer joined #salt
15:57 thaddeusmt manfred: thanks for the tips, still playing with it but you have giving me hope. I didn't realize you had to specify a 'saltenv' parameter, I thought you could target via the topfile base/dev yaml structure
15:57 thaddeusmt *given
15:57 manfred you can, but if you don't give the saltenv variable, then it will attempt to apply states from both environments if the names or something matches
15:58 tligda joined #salt
15:58 timoguin you can also set the minion's environment via config file or pillar
15:58 m1crofarmer joined #salt
15:59 rojem joined #salt
15:59 viq joehh: yeah, with 13536 and 13537 applies I still see: virtual: openvzve
16:00 viq Though I'll test on hosts as well
16:00 * xmj goes looking
16:01 basepi babilen: looks like you got your question answered.  most pull reqs should be against develop, just ping me to cherry-pick it to 2014.1
16:01 xmj viq: ah
16:01 xmj viq: bookmarked, will do that $later
16:01 xmj (today, if possible
16:01 xmj )
16:02 vnico joined #salt
16:02 viq xmj: cool, thanks
16:02 vnico hi, how can I get the list of minions affected by a cmd_async() request?
16:03 kermit joined #salt
16:04 viq joehh: yeah, virtual: openvzhn works too
16:09 jimklo joined #salt
16:14 harkx joined #salt
16:17 CeBe joined #salt
16:18 KyleG joined #salt
16:18 KyleG joined #salt
16:25 teskew joined #salt
16:26 maxskew joined #salt
16:26 mateocon_ joined #salt
16:26 edwx joined #salt
16:28 edwx try this site www.SoccerTips4Sure.com very profesional, nice earnings
16:28 ipalreadytaken joined #salt
16:29 oz_akan_ joined #salt
16:31 CeBe1 joined #salt
16:31 carmony man, I wanted to thank edwx for his great soccer insight...
16:31 vejdmn joined #salt
16:38 masterkorp Hello everyone i am writing a nginx formula to abstract the sites configuration
16:39 masterkorp so pillar data will be converted to a site configuration file
16:39 kmshultz joined #salt
16:41 chiui joined #salt
16:41 masterkorp i would like to specify pillar data like
16:42 forrest joined #salt
16:43 troyready joined #salt
16:43 bemehow joined #salt
16:43 joehillen joined #salt
16:44 masterkorp http://pastie.org/9302582
16:44 masterkorp and generate the file
16:45 ipmb joined #salt
16:45 millz0r joined #salt
16:45 masterkorp i can do a simple for key, value .iteritems to do it
16:46 shaggy_surfer joined #salt
16:46 possibilities joined #salt
16:49 thayne joined #salt
16:50 Ryan_Lane joined #salt
16:50 bemehow_ joined #salt
16:50 possibilities joined #salt
16:52 schimmy joined #salt
16:53 shaggy_surfer joined #salt
16:54 schimmy1 joined #salt
16:59 Aletheuo joined #salt
17:00 chrisjones joined #salt
17:00 tristianc|Phone1 joined #salt
17:02 HeadAIX joined #salt
17:02 aw110f joined #salt
17:12 HeadAIX joined #salt
17:13 marnom_ joined #salt
17:13 younqcass joined #salt
17:16 thedodd joined #salt
17:17 racooper ok, got gitfs working for state tree.  now trying to get it working for pillars. using the same URI syntax as with states, I get an error "does not appear to be a git repository".
17:18 jrb28 joined #salt
17:18 shaggy_surfer joined #salt
17:19 TyrfingMjolnir joined #salt
17:20 n8n joined #salt
17:22 babilen basepi: Will do, I will summon you for juicy cherries whenever applicable
17:22 rawzone joined #salt
17:22 racooper command line "git clone" works to the same repo, but "git fetch" gives the same fatal error.
17:23 babilen basepi: I am always itching to refactor code (i.e. break up some of the mega-functions into smaller ones) whenever I have to fix an issue in salt. Would such "no-op" commits be appreciated? Is there a style guide somewhere?
17:26 basepi babilen: so, there are trade-offs.  Churn for churn's sake is not always good, makes it much harder to cherry-pick fixes and whatnot.  So in general, I would recommend not fixing it just to fix it, but if you're already in the file for another fix, you're welcome to throw in some style fixes as well.  We do have a style guide here:  http://docs.saltstack.com/en/latest/topics/development/conventions/style.html
17:26 forrest babilen, you should write tests :P
17:26 ajolo joined #salt
17:28 babilen basepi: Yeah, I see what you mean. But, for example, the _virtual(..) in salt/grains/core.py that I touched today just (IMHO) begs to be split into _is_openvz(), _is_xen(), _is_.... and I simply wondered if I should work on things like that.
17:28 babilen Maybe it's me and I have a lower tolerance for "long functions" than other people
17:28 bhosmer joined #salt
17:29 babilen forrest: That too although I really don't know how I would test _virtual() grains without the actual hardware/setup.
17:30 rojem joined #salt
17:30 basepi babilen: those kinds of fixes would probably be welcome.  some of that grains stuff is gnarly long
17:30 babilen yeah
17:31 basepi and if it helps to more easily tell which checks are for which grain and OS and whatnot, I can see it being useful.
17:31 babilen Okay, I'll trust my instincts then. Just thought I'd ask before I invest some of the scarce time into something that is then being shrugged off :)
17:31 oz_akan_ joined #salt
17:32 druonysus joined #salt
17:32 druonysus joined #salt
17:32 forrest PR from babilen *close without reading*
17:36 TyrfingMjolnir joined #salt
17:36 viq forrest: you stop that, or I'll subscribe you again to all the discussions about gitlab-formula! ;)
17:36 forrest nooooooooo
17:37 piv joined #salt
17:37 forrest That reminds me, am I just being stupid here? https://github.com/saltstack-formulas/nginx-formula/pull/32
17:37 viq And there was quite a bit of activity today thanks to xmj ;)
17:37 forrest I don't get why we'd want an apache package in the nginx formula
17:38 xmj viq: I blame nicotine patches, awesome tea and coffee supply and a deadline.
17:38 forrest xmj, you ever think about an e-cig?
17:38 Gareth morning
17:38 forrest hey Gareth
17:39 xmj forrest: tried them before.
17:39 Gareth hey forrest, hows it going?
17:39 forrest xmj, ahh ok
17:39 forrest Gareth, pretty good, just ate a waffle, you?
17:39 xmj forrest: nicotine patches + lucky strikes for _those days_
17:39 viq forrest: http://pbot.rmdir.de/obyaOPwgPCjYKv_dim0oiQ
17:39 forrest xmj, lol
17:39 xmj now someone help me figure out how to start ubuntu inside of freebsd 11's bhyve.
17:39 Gareth forrest: doing well. :) No waffle though :(
17:39 vu This is broken: https://media.readthedocs.org/pdf/salt/latest/salt.pdf
17:39 forrest Gareth, clearly you should get waffles
17:40 Gareth I should.  I should get up, walk out of this meeting and go find some waffles!
17:40 babilen forrest: heh
17:40 forrest vu, oh lame, whiteinge https://media.readthedocs.org/pdf/salt/latest/salt.pdf appears to be busted.
17:41 forrest Gareth, you should buy a waffle iron for your desk
17:41 forrest charge people to use it
17:41 viq forrest: so yeah, apparently, at least on centos, httpd-tools contains htdigest. Not sure if plain nginx contains a tool to generate htpasswd
17:41 forrest that's pretty stupid...
17:41 Gareth forrest: convert my desk into a waffle bar.  not a bad idea.
17:41 viq Actually I can check that, give me a sec
17:42 forrest Gareth, exactly
17:42 viq forrest: yeah, nginx packages (on centos) does not have those tools
17:42 forrest :(
17:42 viq So in fact it does appear to make sense
17:43 xmj it does
17:43 xmj because the PR basically removes a previously existing inconsistency across distros.
17:43 logix812 joined #salt
17:43 forrest Yea I added another comment
17:43 forrest Didn't research it this morning when I was laying in bed after getting the email
17:43 ScrumpyJack joined #salt
17:48 piv joined #salt
17:50 g-man joined #salt
17:52 Katafalkas joined #salt
17:52 piv_ joined #salt
17:52 cheus State modules question: If no changes will be made, is it a good practice to add a comment to that effect or leave it blank. I've seen both practices in place.
17:52 masterkorp any ideas ?
17:53 manfred cheus: You should have a comment
17:53 cheus manfred, Danke
17:54 rojem joined #salt
17:56 yomilk joined #salt
17:56 linjan joined #salt
18:00 jslatts joined #salt
18:00 ml_1 joined #salt
18:00 Rojematic joined #salt
18:04 rojem joined #salt
18:04 gq45uaethdj26jw7 joined #salt
18:04 druonysus joined #salt
18:04 druonysus joined #salt
18:05 gq45uaethdj26jw7 anyone ever have an issue where SCP is not installed by default on the image you load up with salt-cloud? cant scp the deploy script over. is there a way to install a package on the minion BEFORE the deploy script is loaded?
18:06 gq45uaethdj26jw7 having this issue on multiple os's with multiple providers
18:06 ccase joined #salt
18:09 piv joined #salt
18:10 JeroenH_ joined #salt
18:11 chiui joined #salt
18:13 piv_ joined #salt
18:15 jdmf joined #salt
18:17 tristianc|Mobile joined #salt
18:18 fragamus joined #salt
18:19 tristianc|Phone joined #salt
18:20 KyleG1 joined #salt
18:21 JasonSwindle joined #salt
18:22 JasonSwindle joined #salt
18:24 tristianc|Phone joined #salt
18:34 Franz_ joined #salt
18:36 Franz_ the following jinja works in a template just as I'd like it to
18:36 Franz_ ldapIps="{% for minion, grains in salt['mine.get']('*', 'grains.item').items() %} {% if "ldap" in grains["roles"] %}{{ grains.fqdn_ip4[0] }}{% endif %}{% endfor %}
18:36 Franz_ but if i put it in a pillar, all hell breaks loose
18:37 Franz_ i guess if i understood exactly what it was doing, maybe i'd understand why
18:38 Franz_ so does the '*' in the mine.get command imply iterate over all minions as might be expected?
18:42 rjc joined #salt
18:43 smcquay joined #salt
18:44 oz_akan_ joined #salt
18:45 shel3over joined #salt
18:45 __alex joined #salt
18:47 lionel joined #salt
18:48 therealGent joined #salt
18:49 ramteid joined #salt
18:53 jslatts joined #salt
18:58 n8n joined #salt
19:00 Joseph joined #salt
19:02 gq45uaethdj26jw6 joined #salt
19:03 rojem joined #salt
19:03 ifnull joined #salt
19:05 yetAnotherZero joined #salt
19:05 JasonSwindle Anyone ever dealt with salt \* test.ping, and getting 3 dup minion returns?
19:08 bmatt yeah, i ran into that
19:08 bmatt turns out there were multiple salt-minion processes running, each with their own keys
19:08 bmatt I have no earthly idea how it happened, and I haven't seen it since
19:08 marnom_ JasonSwindle: yeah same here... really annoying.. I suspect it happens because of the way I deploy my VM's and switching IP's... but it really shouldn't do this
19:08 bmatt dupe IDs, different keys, multiple processes
19:08 pssblts joined #salt
19:09 marnom_ I end up having to manually killall -9 salt-minion and restart the service... (a service restart alone doesn't solve it for me)
19:09 marnom_ bmatt: I don't see different keys on my minions afaik (they are pre-seeded in a clean image..)
19:09 mgarfias so I get this error: "Rendering SLS "test:apps.server" failed: Jinja variable 'dict object' has no attribute 'apps'; line 11" but when i use pillar.get apps on the cmd line, i get back good data.  why would the master not see the pillar data?
19:09 rojem joined #salt
19:09 twobitsprite somehow I broke my salt master... I tried enabling the git backend, and when I couldn't get it to work, I disabled it and now the master can't find any states in the file_root...
19:10 marnom_ mgarfias: is it allowed to access the pillar data? You could reference it with a default value so it falls back to that instead..
19:10 twobitsprite running with -l debug doesn't seem to give anything interesting, no errors or even any mention of the file_root directory
19:10 mgarfias yes, minion/node is allowed explicitly to access that pillar data
19:10 twobitsprite I restarted the service and even the system itself
19:10 JasonSwindle Now I have 5 dup minions
19:11 JasonSwindle the same one.
19:11 phool joined #salt
19:12 Joseph twobitsprite: there are multiple configuration properties in /etc/salt/master for gitfs. You'll need to disable them all to get back to the local file system. Can you do a gisthub of your master conf?
19:13 twobitsprite Joseph: I reverted my master conf to a pervious version before I enabled it, I don't think that's the problem
19:14 Joseph and does the master conf use the default root /srv/salt or do you have a non default location?
19:14 JasonSwindle marnom_: Not a think is moving or changing....but I keep getting growing minion dups
19:14 twobitsprite default
19:14 twobitsprite http://pastie.org/9302975
19:14 forrest ifnull, weird that you're still seeing that error on 2014.1.3
19:15 ifnull yeah
19:15 ifnull im out of ideas
19:15 timoguin twobitsprite: i'd try clearing out /var/cache/salt first
19:15 twobitsprite timoguin: did that
19:15 Joseph twobitsprite: and you are seeing this with highstate and state.sls or b oth?
19:15 twobitsprite Joseph: both
19:15 Joseph timoguin: darn...that was a good suggestion!!
19:16 Joseph twobitsprite: so you cleared out the cache on both the master and minion and then restarted both services?
19:16 twobitsprite it doesn't even generate any state cache files
19:16 mgarfias ok, weird
19:16 mgarfias pillar.get apps has data
19:16 mgarfias pillar.items doesn't have an apps key
19:16 mgarfias why would those be out of sync?
19:17 twobitsprite hadn't tried it on the minion, but did just now. no joy
19:17 bhosmer_ joined #salt
19:18 Joseph what release are you running?
19:18 mgarfias 2014.1.4
19:19 twobitsprite 2014.1.0+ds-1~bpo60+1
19:19 Joseph okay so gisthub one of the SLS in /srv/salt
19:19 Joseph oh what OS are you one? are they virtualized?
19:19 twobitsprite Joseph: you talking to me or mgarfias ?
19:19 Joseph twobitsprite: to you
19:20 twobitsprite ok, so the SLS files haven't changed and I've tried several of them that used to work
19:20 twobitsprite and the master is on debian squeeze with the minion on RHEL6
19:21 Joseph twobitsprite: so put the master and minion in foreground with debug log level and then run a state.sls
19:21 Joseph gisthub the log output for both
19:21 Ahlee so i have a minion with two grains set
19:21 twobitsprite Joseph: I have a lot of minions, running the master in debug spews a ton of messages
19:21 Joseph what's your size?
19:22 Ahlee {% if grains['foo'] == 'bar' and grains['baz'] == 'qux' %}
19:22 Joseph 1000 or more?
19:22 twobitsprite ~450 minions
19:22 Ahlee yet it's not rendering
19:22 Ahlee i bet 'and' isn't valid jinja, is it
19:22 Joseph Ahlee: not sure but i do know jinja is evil sooo....
19:22 twobitsprite Ahlee: I'm pretty sure it is,
19:22 Ahlee ain't that the truth.
19:22 ipalreadytaken joined #salt
19:23 jimklo joined #salt
19:23 twobitsprite Ahlee: yeah, I have "and"s in my templates
19:23 mprymek joined #salt
19:23 Ahlee ok, thoughts on next steps? I only want this state to execute if both grains are configured
19:23 twobitsprite Ahlee: however, I use "grains.foo" instead of "grains['foo']"... not sure if that makes a difference
19:23 Joseph twobitsprite: could you either firewall off all but one minion temporarily oin the master?
19:24 forrest Has anyone used http://docs.saltstack.com/en/latest/ref/states/all/salt.states.ssh_auth.html#salt.states.ssh_auth.absent but somehow matched it on a wildcard to just remove all authorized keys?
19:24 twobitsprite Joseph: I suppose I could do that... give me a few minutes
19:25 Ahlee i hope my boss never adds up the amount of time i've spent on salt.
19:25 forrest Ahlee, why?
19:25 forrest it benefits you guys
19:25 brandon_ joined #salt
19:25 Ahlee forrest: but would a cost analysis show positive flow?
19:25 Ahlee i somehow doubt it.
19:26 forrest *shrug*
19:26 marnom_ JasonSwindle: sorry I don't know the solution either I'm still looking into it myself
19:26 JasonSwindle It is odd.....  still digging in myself
19:26 Joseph Ahlee: any technology can be used in such a way that the costs outweigh the benefits
19:27 Joseph Ahlee: i am really surprised though that you are facing this. compared to other configuration management tools, salt was the least costly to get up and running and doing something useful for me.
19:27 Ahlee yes yes, was a low blow as i'm frustrated to again be spending hours trying to debug stpuid jinja issues
19:27 Joseph Ahlee: yes jinja really is evil
19:27 Joseph Aside from some very very specific use cases its not worth the trouble
19:27 Joseph I fell into that trap as well
19:27 jchen wait what? jinja is awesome
19:28 bensons_ better than nothing at least
19:28 timoguin I've started learning flask and feel like i have a much better understanding of jinja after a few tutorials
19:28 Joseph jchen: i banish thee from all polite comapny
19:28 Joseph aslo thems fighting words
19:28 Joseph jchen: j/k
19:28 Ahlee foo.bar not working either, it appears
19:28 jchen Joseph: ? i feel like the problem is not jinja but rather your understanding of jinja
19:28 timoguin but it's not easy to debug with salt when you have multiple machines and file server updates in the mix
19:29 jchen also, you can just write your thing in python if you're doing logic
19:29 Joseph jchen: could be. My understanding of jinja is rudimentary. I know just enough about jinja to having to use it in salt.
19:29 Joseph to hate
19:29 timoguin Ahlee: I'm assuming you've made sure the minion actually has those grain values set?
19:29 Franz_ joseph: i'm using jinja to mine.get minion grains.  Is there an alternative way?
19:29 Ahlee timoguin: indeed
19:29 bhosmer_ joined #salt
19:29 Ahlee grains.item return correctly
19:29 timoguin try grains.get
19:29 Ahlee and they're strings
19:29 gothix joined #salt
19:29 gothix joined #salt
19:30 Joseph jchen: exactly....aside from iterating through a grain or pillar or a simplistic if condition, i'll always go to a custom module any day.
19:30 Ahlee after having that fun of learning how to deal with integers, too
19:30 fllr joined #salt
19:30 Ahlee if grains.get('foo') == bar ?
19:30 Joseph jchen: beyond some extremely simplistic use cases, i find using jinja torturous to use and a nightmare to maintain. And god help me if i encounter a macro in a state file.
19:30 timoguin Ahlee: just try grains.get at the cli first
19:30 xzarth_ joined #salt
19:31 fllr Hey guys. I'm getting a weird error: https://gist.github.com/felipellrocha/48f1a45f654134e31970 Jinja cannot find the variable 'pillar', so that state fails to execute. What could be the cause of it?
19:31 Joseph Franz_: could you elaborate on your use case?
19:31 Ahlee so one is returning True.  I wonder if it's being converted to a boolean somehow
19:32 Ahlee yep.
19:32 Joseph fllr: looks like the pillar variable that you are trying to use in your state file doesn't actually exist
19:32 fllr Joseph: It's there, though...
19:33 fllr Joseph: It works for some machines, but not others, I should add...
19:33 timoguin Ahlee: so it's not well documented, but as far as i know grains.item is rendered on the master and grains.get is rendered on the minion.
19:33 Joseph fllr: possible that your pillar top is not applying the pilalr data to the nodes that you think it is
19:33 timoguin i'm not completely sure of the distinction
19:33 timoguin but i've seen grains.item return correctly and not grains.get
19:33 timoguin usually when grains haven't been sync'd
19:33 Joseph timoguin: reaaaaaally? i never would have guessed that. I thought grain data was all genereated on the minion
19:33 Franz_ joseph: trying to get the ip of the second ldap master (role:ldap-master2) so that ldap-master1 can use it to set up replication
19:33 Franz_ using this: {% for minion, grains in salt['mine.get']('*', 'grains.item').items() %} {% if "ldap-master2" in grains["roles"] %}{{ grains.fqdn_ip4[0] }}{% endif %}{% endfor %}
19:34 Franz_ that's in a template
19:34 TheThing uhh
19:34 Franz_ it works is unwieldy
19:34 Joseph Franz_: of course because its jinja...it always "works" but its ugly and painful to maintain
19:34 Joseph or i should say jinja in salt
19:34 Joseph maybe jinja outside of salt is amazing
19:35 Joseph Franz_: have you programmed in python ever?
19:35 bhosmer_ joined #salt
19:35 Franz_ nope, but i'm willing ot lean
19:35 timoguin Joseph: go through the Flask Mega Tutorial if you want. it's a create-a-twitter-clone tutorial
19:35 timoguin gave me a much better understanding of jinja
19:35 Franz_ er, learn joseph
19:36 Ahlee flask mega tutorial, check
19:36 Joseph Franz_: pardon?
19:36 Ahlee thanks timoguin
19:36 timoguin i'm still finishing it up
19:36 tubie joined #salt
19:36 tubie howdy,'
19:36 Joseph timoguin: that looks itneresting....i'll definitely give it a try
19:36 Franz_ joseph: sorry, i haven't programmed in python but am willing to learn
19:37 Franz_ do you think that would be a better approach to getting that info versus jinja in salt?
19:37 Joseph Franz_: so this is just me and i suspect that some would disagree but whenever jinja gets semi complicated i would opt for a custom state or execution module
19:37 fllr Joseph: It looks like the pillar data is there... I'll show you why, let me paste it...
19:39 timoguin Joseph: looking at the grains.py module i may be talking out of my ass about it rendering on the master...
19:39 timoguin but regardless there is a difference between grains.item and grains.get that isn't well clarified
19:39 Joseph timoguin: i am definitely no expert on the salt internals...i had just assumed grains was on minion and pillar was on master and the two shall never cross over
19:40 piv joined #salt
19:40 Joseph timoguin: i may start hating you for showing me this tutorial i have a sinking feeling i am about lose sleep and any remaining social life i had left over b ecause i am going to become obsessed with this tutorial
19:40 fllr Joseph: https://gist.github.com/felipellrocha/49aef5e0637c13461d48#file-gistfile1-txt-L97
19:40 Joseph fllr: looking....
19:41 timoguin Joseph: lol i have definitely lost sleep
19:41 timoguin but it's been fun
19:41 timoguin he has lots of other great tutorials too
19:41 fllr Joseph: I had to redact the pillar data, but it compiled all right... but when it gets to the bottom... When I actually get to use the data, the thing is gone...
19:41 Franz_ joseph: i think you've surpassed my knowledge with that "custom state or execution module" statement - does that relate to writing a tiny python program to do what i'm trying to do?
19:41 Joseph timoguin: it looks extremely interesting. And of all programming languages out there, python is the only one i genuinely love using.
19:42 Joseph Franz_: in essence yea...http://docs.saltstack.com/en/latest/ref/states/writing.html
19:42 Joseph salt is built to be extensible so it will distribute the modules on your behalf and then let you use them in your states or do remote execution with them
19:43 Joseph Franz_: in my experience whatever you are doing jinja becomes waaaaay simpler once its just python code
19:43 Joseph Franz_: now there's alternative template renders that salt supports which you can look at to....for example pyobjects
19:44 Franz_ joseph: thanks - i'll do some research
19:45 Joseph Franz_: http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.pyobjects.html
19:45 Joseph that allows you to template in pure python
19:45 KyleG joined #salt
19:45 KyleG joined #salt
19:46 twobitsprite Joseph: ok, I got the firewall set up... do you want me to try running a state or just paste the startup debug logs?
19:46 Joseph state...because thats what you said isn't working
19:46 whiteinge forrest: thanks for the PDF heads-up. it's choking on unicode chars :(
19:46 Joseph twobitsprite: for both minion and master
19:46 forrest whiteinge, laaaaaaaaaaaaaaaaaaame, and it was really all vu
19:47 twobitsprite crap, nm, my firewall isn't working afterall, one sec
19:47 Joseph fllr: have you cleared the salt cache and freshed pillar data?
19:47 Joseph twobitsprite: ah yes iptables...the gift that keeps on giving
19:47 twobitsprite heh, indeed...
19:47 * whiteinge shakes his fist at vu
19:48 twobitsprite Joseph: mind double-checking my rule? iptables -A INPUT -s 10.1.10.36/32 -p tcp -m tcp --dport 4505 -j ACCEPT
19:48 twobitsprite Aame for 4506. That's on a DROP policy, so it should filter out all others, right?
19:48 Joseph twobitsprite: i'll try but iptables is where dragons live and i don't like getting eaten or flamed
19:48 Joseph twobitsprite: in theory!!
19:48 twobitsprite haha
19:48 forrest twobitsprite, did you put that at the end of the chain?
19:50 Joseph twobitsprite: can you run "iptables -L" and put the output in a gist?
19:50 twobitsprite https://gist.github.com/anonymous/eebba307f051a04c328e
19:50 taterbase joined #salt
19:50 fllr Joseph: No. All I did was run highstate twice...
19:51 ggoZ joined #salt
19:51 Joseph fllr: its possible you have a cahcing problem on some nodes but not others....you could try clearing all cached data and then execing the sync pillar command
19:51 gq45uaethdj26jw6 joined #salt
19:52 piv joined #salt
19:52 forrest whiteinge, good changes to that apache pillar
19:52 twobitsprite might just be established connections getting through the RELATED,ESTABLISHED rule... let me reboot it real quick to clear any old connections
19:52 dsolsona joined #salt
19:52 whiteinge forrest: thanks. i should've updated that a good while ago. didn't realize it would cause so much trouble...
19:52 Joseph twobitsprite: oh yea that could definjitely be it
19:52 forrest whiteinge, yea me neither
19:53 diegows joined #salt
19:53 forrest pretty good article by techhat over on infoq http://www.infoq.com/articles/saltstack-configuration-management
19:53 fllr Joseph: This is my state file, btw: https://gist.github.com/felipellrocha/a5ca47a98f1174c6acb0
19:53 fllr Joseph: Ummm, k... I'll try that...
19:55 JasonSwindle joined #salt
19:55 Bosch[] joined #salt
20:01 jimklo joined #salt
20:06 HeadAIX joined #salt
20:06 Joseph forrest: this part cracked me up "Many of our users will use SaltStack to tell Puppet manifests what to do because SaltStack is highly efficient remote execution at its core"
20:06 Joseph its funny cause its true
20:11 forrest yea
20:12 Joseph forrest: you know i was thinking you could turn puppet's scale problems into a drinking game. Do we think it'll seize up at 100, 250, 500......
20:12 forrest sieze up at 2, as soon as the master is in one datacenter, and the other servers are in another
20:12 Joseph forrest: ohhh good one
20:13 JasonSwindle 2014-06-18 15:11:52,925 [salt.master      ][ERROR   ] An extra return was detected from minion xyz, please verify the minion, this could be a replay attack
20:13 druonysuse joined #salt
20:13 forrest JasonSwindle, are you running two of the same minion?
20:13 JasonSwindle How do I deal with this when it is not an attack?
20:13 JasonSwindle nope
20:13 JasonSwindle Or should not be
20:13 forrest JasonSwindle, https://github.com/saltstack/salt/issues/3996
20:14 JasonSwindle looking
20:14 mgw joined #salt
20:14 JasonSwindle I am not even going that far, just doing test.ping
20:14 viq JasonSwindle: restart minion / make sure only one minion is running
20:15 JasonSwindle Oh I have restarted it.  Deleted the key.
20:15 JasonSwindle What is funny is, I will have 3 dups, and it will grow to 5
20:16 viq Huh. Can't say then, whenever I had that issue, I stopped minion, made sure no other processes were running, started minion, and issue was no more.
20:16 forrest JasonSwindle, did you stop the service entirely?
20:16 JasonSwindle Yep
20:17 forrest and there are no hanging services?
20:17 forrest or running processes?
20:17 JasonSwindle Digging into that
20:17 JasonSwindle This box is a Syndic, Master, and Minion
20:17 JasonSwindle lol
20:19 JasonSwindle forrest: viq salt-syndic looked to not dying corrctly
20:19 possibilities joined #salt
20:19 JasonSwindle killing off
20:22 JasonSwindle viq: forrest it was the salt-syndics
20:24 twobitsprite Joseph: not sure if you're still around, but I got the firewall working and here are the outputs: https://gist.github.com/anonymous/d86c3190448a852c6079
20:24 Joseph twobitsprite: looking
20:24 twobitsprite thanks
20:26 twobitsprite brb
20:26 patrek_ joined #salt
20:26 Joseph twobitsprite: are there a lot of files in your job cache?
20:28 rojem joined #salt
20:28 n8n joined #salt
20:29 Joseph twobitsprite: its possible you are exhausting file descriptors due to the size of the job cache which may prevent successfully serviing up a file.. For example i see this warning in the logs "The number of accepted minion keys(423) should be lower than 1/4 of the max open files soft setting(1024). Please consider raising this value."
20:32 kballou joined #salt
20:37 jaewint joined #salt
20:37 smcquay joined #salt
20:41 shaggy_surfer joined #salt
20:44 twobitsprite Joseph: arg, yeah... I keep thinking I fix that but it doesn't stick...
20:44 Joseph twobitsprite: you should eitehr disable job cache or change the returner to be a remote storage like a mysql db etc
20:44 twobitsprite Joseph: as a general rule, or just for testing?
20:45 Joseph twobitsprite: as a general rule for large scale salt deployments
20:45 twobitsprite how do I disable the job cache?
20:45 Joseph twobitsprite: you aren't the first one that has dealt with fds being exhausted. If you have a couple hundred minions and you run manu jobs, you can end up with millions of files on the salt master
20:45 twobitsprite just comment out the cachedifr?
20:45 Joseph twobitsprite: its a setting in the salt master conf
20:46 Joseph job_cache: False
20:46 twobitsprite yep, just found it
20:47 Joseph twobitsprite: that said....you might want to check with the google group or a salt dev about this. this type of fd exhaustion is expected with a couple thousands nodes not a couple hundred.
20:47 Joseph twobitsprite: if you are seeing fd exhaustion with 500 nodes, that may in fact be some type of bug or set up problem
20:48 Joseph off to a meeting...good luck
20:49 rojem joined #salt
20:52 oz_akan__ joined #salt
20:56 gildegoma joined #salt
21:00 quickdry21 joined #salt
21:01 oncallsucks Question. I have an init.sls where I want to check if a post is listening and if its not do NONE of the other things
21:01 oncallsucks how can I go about this?
21:03 rojem joined #salt
21:04 ckao joined #salt
21:05 felskrone joined #salt
21:05 jY anyone use kitchen-salt for testing.. getting an error "Message: Failed to complete #converge action: [No such file or directory]"
21:06 JeroenH_ joined #salt
21:12 mattmtl I decided to move my custom modules, grains to another directory... I did changed my configuration as like this: http://paste.ofcode.org/GdR6c89f3DxyFjyEEEiNY5 but they seem not been seen anymore
21:12 mattmtl is there some kind of cache that I should clear?
21:13 nicksloan joined #salt
21:13 nicksloan will salt.states.mount remount a filesystem if the options are different?
21:15 timoguin mattmtl: /var/cache/salt on master and minion(s)
21:15 timoguin you also should just be able to do a saltutil.sync_all
21:18 gothix Looks like we dont have a snmp module
21:18 diegows joined #salt
21:19 viq oncallsucks: I would probably start with cmd.run with netstat and grep and then require that
21:19 mattmtl timoguin: it is by doing a  saltutil.sync_all that I see that my custom modules are not discovered
21:20 timoguin then clear cache, restart master, etc.
21:20 timoguin should pick it up
21:20 oncallsucks viq: Do I have to require in every other module?
21:22 viq oncallsucks: or chain them, but it would need to be at the top of the chain for other things to not happen, yeah
21:22 oncallsucks I see - how do I chain? Is there a link to this?
21:22 viq And you would get a bunch of failed states in the report with it
21:22 oncallsucks Ah I see- Im trying to avoid noisy output
21:22 viq oncallsucks: http://docs.saltstack.com/en/latest/ref/states/requisites.html
21:23 oncallsucks Oh so I can just require chain
21:23 oncallsucks How can I supress the long errors?
21:23 mattmtl timogin: deleted /var/cache/salt and restarted master without success
21:24 oncallsucks I prefer prereq to require
21:24 viq oncallsucks: not sure. You could also probably use some jinja with http://docs.saltstack.com/en/latest/ref/states/all/salt.states.saltmod.html with netstat from docs.saltstack.com/en/latest/ref/modules/all/salt.modules.network.html
21:27 oncallsucks viq: Pretty much everything supports require: right?
21:29 jliljenq joined #salt
21:32 tligda joined #salt
21:33 forrest oncallsucks, pretty much
21:34 forrest oncallsucks, here's an example where I do that (across states as well) https://github.com/gravyboat/hungryadmin-sls/blob/master/salt/hungryadmin/app.sls
21:35 oncallsucks ah perfect, thanks
21:35 forrest yea np
21:35 forrest keep in mind that the stuff that I have inside of 'include' at the top will run before the stuff below
21:35 forrest I just use a lot of requires because of old puppet habits, and it makes it easier for me to see the relationships
21:37 jliljenq hey all, does anyone know if it's possible to have salt-call route its output in realtime to stdout on masterless minions? would be nice to see the usual progress output for things like wget and scp
21:37 Ahlee iterate over all directories in /home.... thoughts?
21:38 jnials joined #salt
21:38 Ahlee i want to populate .pydistutils for any user that's logged into a server, but only on servers that user is actually using
21:39 fridder joined #salt
21:44 mgarfias is there any dox concerning adding an openstack network to my cloud config?  specifically i'd like to boostrap images that do not hae public IPs, and rely on my network for internode communication
21:45 gothix how would i check for a service to have something  runnim
21:46 gothix running in a state and act on it
21:46 TheThing joined #salt
21:47 gothix i need to configure snmp and a user - since i can't find a module how would i ensure the user doesnt esist on reruns of the state
21:47 gothix exist
21:49 gothix :wq
21:57 mgw joined #salt
22:02 ifnull joined #salt
22:03 TyrfingMjolnir joined #salt
22:05 n8n_ joined #salt
22:06 oz_akan_ joined #salt
22:06 jv115 joined #salt
22:06 huleboer joined #salt
22:09 jv115 joined #salt
22:16 rojem joined #salt
22:17 dsolsona joined #salt
22:17 n8n joined #salt
22:18 jv115 joined #salt
22:19 Joseph joined #salt
22:22 Joseph when i run a state.sls, i get this error 9d554158-4328-42f7-b658-49e0e2b1611a:     Data failed to compile: ----------     Illegal requisite "[OrderedDict([('file', 'virtdriver')])]", please check your syntax.
22:22 Joseph here is the stae file https://gist.github.com/jaloren/00e3e5a9e3ea9bfdfce9
22:23 Joseph I don't understand why it thinks that's invalid
22:25 APLU joined #salt
22:27 alunduil joined #salt
22:27 manfred Joseph: you are in the same state, it should be file: /etc/sysconfig/libvirtd iirc
22:27 manfred state-id *
22:27 alunduil left #salt
22:27 Joseph i am following the example here
22:27 Joseph http://docs.saltstack.com/en/latest/topics/tutorials/cloud_controller.html
22:27 Joseph isn't it doing the exact same thing where the requisites are referring to the same state id?
22:28 forrest https://gist.github.com/jaloren/00e3e5a9e3ea9bfdfce9#file-gistfile1-txt-L18
22:28 manfred hrm... it is
22:28 forrest should be virtdriver.
22:28 pssblts joined #salt
22:28 manfred oh right
22:28 Joseph forrest: ah probably
22:29 forrest also
22:29 forrest watch is indented too far
22:29 Joseph forrest: i hate state files that use the resource as the state id so i pay when i change it :)
22:29 forrest https://gist.github.com/jaloren/00e3e5a9e3ea9bfdfce9#file-gistfile1-txt-L19
22:29 Joseph forrest: yep again
22:29 forrest yea I would split it up even more than you did.
22:30 Joseph forrest: yea i agree
22:30 n8n_ joined #salt
22:32 AdamSewell joined #salt
22:32 forrest Joseph, also, I assume line 11 is referencing something that is actually included?
22:32 Joseph forrest: yes its truncated
22:33 forrest cool
22:33 forrest Joseph, if you didn't already fix it, I added a comment to your gist
22:35 Joseph forrest: okay i'll look
22:36 druonysuse joined #salt
22:36 druonysuse joined #salt
22:38 jnials joined #salt
22:40 Luke_ joined #salt
22:42 Joseph job management question.....i have the JID, the jobs.active returns an empty dict...however when i do lookup on the jid nothing is returned
22:42 Joseph why would that be?
22:48 smcquay joined #salt
22:50 Kraln joined #salt
22:51 nebuchad` joined #salt
22:51 kermit joined #salt
22:52 Joseph hmmm i find the jid in the jobs directory but there's literally nothign else there
22:52 Joseph how could nothing be rewturned
22:52 Joseph isn't that odd?
22:57 forrest Eugene, http://saltdocsprint.eventbrite.com
23:00 rallytime joined #salt
23:01 possibilities joined #salt
23:07 Outlander joined #salt
23:12 n8n joined #salt
23:18 mosen joined #salt
23:19 ifnull joined #salt
23:21 yomilk joined #salt
23:21 ifnull forrest: i believe i have a solution/workaround for the zmq issue
23:21 forrest ifnull, oh?
23:21 forrest throw computer out of window?
23:22 ifnull lol
23:22 forrest become lumberjack?
23:22 ifnull in all my tests i am using vagrant/virtualbox with cpu=2 mem=2048
23:22 forrest ok
23:22 ifnull i switch cpus to 1 and everything works
23:22 ifnull consistantly
23:22 forrest interesting
23:22 ifnull i have reprovisions 5 times
23:23 ifnull and i have changed it back to 2 cpus and reproduced the problem
23:23 forrest yea that would make sense why I can't dup it
23:23 forrest the lxcs are running on machines with 24 cores
23:24 forrest ifnull, can you change it to 4 cores?
23:24 forrest and see if you can duplicate the problem?
23:24 ifnull i’ll give it a try
23:25 forrest if it only happens on machines with 2 cores, I'm not quite as worried
23:25 forrest still dumb
23:25 forrest but at least it isn't production breaking
23:25 ifnull yeah
23:25 ifnull may be virtualbox
23:25 forrest exactl
23:25 forrest *y
23:26 forrest the real problem is that the error message can be returned for multiple issues
23:26 forrest so for all we know, these other guys are inode problems
23:27 forrest we need more info from danielfrg
23:27 ifnull yeah
23:27 ifnull I think 4ba20946ae4ad061e6851d2868ddf734ef94e45c was to address this.
23:27 ifnull but is really isnt an auth error
23:28 forrest you mean the inode error
23:28 forrest yea
23:28 forrest that's what I'm saying
23:28 forrest It would be good if it was more explicit (even saying both of those errors seems more confusing to me when thinking about newbies)
23:29 ifnull yeah
23:30 ifnull ok. i was able to reproduct the problem with 4 cores
23:30 ifnull *reproduce
23:30 forrest what about 3 cores
23:30 forrest lol if it's just when there is an odd number of cores
23:30 forrest err even
23:38 ifnull sure .. why not … i’ll give it a try
23:38 bensons_ stupid question, is it best practice to run a state.higstate periodically for all minions (all 5 minutes or so) and if so, how? should i use the minion scheduler for that?
23:38 jnials joined #salt
23:38 bensons_ highstate that is :)
23:38 nineteeneightd joined #salt
23:38 ndrei joined #salt
23:39 manfred bensons_: depends on what you want to do
23:39 ifnull bensons_: i run as needed.
23:39 manfred i run highstates when I need them
23:39 manfred and use salt-api + webhooks to tell the reactor to kick of a highstate on certain minions when doing a git push
23:40 forrest I only run as needed
23:40 forrest just depends on your infrastructure and setup
23:40 bensons_ hm
23:40 timc3_ joined #salt
23:41 forrest Usually the box is built, then the highstate applied, and then from there I only touch them when I have updates to make or new code to deploy
23:41 forrest if you had some sort of system where you needed every machine to know about every other machine, then it would make sense to run a highstate all the time
23:41 forrest or on an interval
23:41 forrest as you build systems or whatever.
23:41 forrest manfred, did you make those webhooks public
23:42 forrest *?
23:42 forrest I'm just curious
23:42 manfred i have not yet
23:42 forrest are you planning to? Seems like a cool use of the reactor
23:42 manfred i will eventually P
23:42 manfred :P
23:42 forrest lol
23:42 manfred i need to sanatize them and finish them
23:42 bensons_ well basically git hooks are fine when i do changes, but what about the state of services.. for instance if i reboot a host, it will not run a highstate - ok usually the os should take care to kick up the services, but what if its custom services that are just managed by salt?
23:42 forrest Eventually^TM
23:42 manfred forrest: you will be the first to know when I push them my git repo
23:42 manfred
23:42 forrest bensons_, you should write an init script for those services.
23:42 mgw joined #salt
23:43 forrest bensons_, then you can manage them with salt, and tell them to start on reboot/boot
23:43 manfred bensons_: startup_states: highstate
23:43 forrest manfred, sweet
23:43 forrest yea that's an option as well
23:43 forrest http://docs.saltstack.com/en/latest/ref/states/startup.html
23:43 forrest you should still write init scripts if possible though :P
23:43 bensons_ ah thats neat, good
23:43 steveoliver ubuntu 12.04: why does apt-get update && apt-get upgrade not bring me to 2014.1.4 (stuck on 0.17.4) ?
23:43 blarghmatey joined #salt
23:44 forrest steveoliver, what release is it bringing you to?
23:44 forrest is it just saying 0.17.4 is the latest?
23:44 steveoliver 0.17.4
23:44 forrest that's... odd
23:44 forrest I know joehh has more recent releases out there
23:44 forrest can you try an apt-get install ?
23:44 manfred 2014.1.5 should be available for 14.04
23:44 forrest to see if it catches it
23:44 forrest he's on 12.04
23:44 manfred sorry
23:44 manfred 12.04
23:44 forrest JEEZ
23:44 manfred should still be available
23:44 forrest this manfred guy...
23:44 steveoliver oh
23:44 manfred for all of the supported ubuntus
23:44 forrest :D
23:44 steveoliver had to apt-get install again
23:44 forrest ahh ok
23:44 steveoliver (weird) ?
23:45 forrest yea that is odd
23:45 forrest but so is apt
23:45 forrest so what are you gonna do *shrug*
23:45 manfred ¯\(°_o)/¯
23:45 forrest manfred's favorite macro right there
23:45 manfred that is a lie
23:45 manfred ????
23:45 manfred that one is
23:45 forrest hah
23:46 manfred http://www.fileformat.info/info/unicode/char/1F638/index.htm
23:46 steveoliver manfred :)
23:46 steveoliver i like the shrug :)
23:46 manfred the shrug is pretty awesome
23:46 steveoliver wacked out eyes and all
23:46 mosen hiya saltines
23:47 steveoliver thx guys
23:47 TheThing lol, mosen o/
23:47 mosen halite is early days hey?
23:47 manfred (╯°□°)╯︵ ┻━┻
23:47 forrest how did we get the nickname saltines?
23:47 forrest so we're a stale shitty cracker? :(
23:48 mosen we are crackers who use salt
23:48 forrest mosen, it has been around for a while, and works pretty well from what I last heard once you get it up and running, haven't played with it in about 8 months though
23:48 forrest that seems a bit discriminatory
23:49 mosen sorry, doesnt have the same connotation here
23:49 forrest ahh ok
23:50 forrest Cracker = Something that’s great, like the best bargain for the day being the “cracker of the day.”
23:51 mosen I had an idea for a minion monitoring system, using socket.io to show minion events in realtime
23:51 mosen but might be a duplication of halite kinda
23:51 Shenril joined #salt
23:51 Joseph question about the cloud controller http://docs.saltstack.com/en/latest/topics/tutorials/cloud_controller.html
23:52 Joseph for setting up the bridge, doesn't that state block need the "use" keyword argument
23:52 Joseph ??
23:52 kuyaedz joined #salt
23:54 troyready joined #salt
23:54 forrest sorry Joseph you're messing with stuff I haven't touched so I'm not sure
23:55 Joseph forrest: no worries i haven't messed with this stuff eitehr. That's how i managed to get my salt minion to completely brick the networking and cut off all access to the node. I had to access the node out of management facility to fix it. :)
23:55 Joseph forrest: however in my defense i was just using the example provided verbatim ha!
23:56 kuyaedz joined #salt
23:56 Joseph forrest: also for the record boo to not documenting keyword arguments for network.managed
23:57 tristianc|Mobile joined #salt
23:58 joehh steveoliver: 2014.1.5 should be available - which ppa are you using and maybe you need to do an apt-get install salt-*
23:59 forrest joehh, he was able to get it with an apt-get install, apt-get upgrade just didn't work, sorry to have tagged you :\
23:59 forrest Joseph, see if there is an issue, if not make one so we can get it documented.
23:59 tristianc|Phone joined #salt

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