Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-02-26

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

All times shown according to UTC.

Time Nick Message
00:00 vlcn bretep, well the issue is that I won't want to allow all minions access to cmd.run, unless I am missing something here
00:00 vlcn how would I actually target a specific sls?
00:01 bretep You change the * to *coolserverthatcandothisfunction*
00:01 bretep I'll create a quick mock... hold on
00:02 vlcn isn't that the other way around?  I need *all* minions to have the ability to call that whereas setting it like you just said would mean only that one server would be able to use it
00:02 vlcn I'm thinking I may need to convert it to a runner?
00:07 * bretep is reading python code to see if this is doable, but not documentd
00:07 bretep s/documentd/documented/
00:07 alunduil joined #salt
00:09 sroegner joined #salt
00:09 seblu hi
00:09 timoguin joined #salt
00:09 seblu is there a way to debug salt minion
00:10 seblu I mean I get: the directory 'salt://binutils/' does not exist on the salt fileserver in environment 'base'
00:10 vlcn salt-minion -l debug
00:10 bretep salt-call -l all
00:10 bretep all will give you trace output
00:12 Guest88308 joined #salt
00:13 Guest88308 helllo
00:13 Guest88308 some on available there..?
00:13 Guest88308 i just want to ask so0mething
00:13 robawt Guest88308: just ask away
00:14 hunter joined #salt
00:15 Guest88308 yeah ahh... it just regading on face book anyway..
00:16 bretep vlcn: You are correct. The only check is to see if the minion is allowed to make a query. You might be able to do this: https://gist.github.com/bretep/a3fb4f9570d3b7f82dbd
00:16 bretep I'd try it
00:17 Guest88308 I'm using a company computer and i cant open any website and it says this page is blocked
00:18 xmltok joined #salt
00:18 fllr joined #salt
00:19 mgw joined #salt
00:20 krow joined #salt
00:20 Guest88308 hello
00:23 vlcn bretep, thanks -- I'll give that a go.
00:24 n8n joined #salt
00:25 baniir joined #salt
00:25 Vye bretep: Maybe provide your feedback in that issue I linked you. I agree with you. I'd rather use highstate and a single top file vs one for each env.
00:25 Vye bretep: What they did works well for merging environments but not very well if you want to isolate them.
00:26 FarrisG_ joined #salt
00:27 Vye by default salt does merging between environments which is a no-go for me.
00:29 bretep Vye: Same here
00:30 seblu debug is not more helpful, on master or on minion
00:30 bretep I'll see if I can solve the issue and do a pull request tonight. Make it a master config option
00:30 seblu it only tolds me that this directory not exists
00:30 seblu but it was, I don't remember changing something about that
00:30 seblu is there a way to list file in a salt:// protocol?
00:31 cachedout joined #salt
00:31 Vye bretep: take a look at how ext_pillar does it and this is my PR for winrepo: https://github.com/saltstack/salt/pull/10756
00:32 sgflt joined #salt
00:34 bretep Vye: You might find the subfilters interesting: https://gist.github.com/bretep/9221008
00:34 bretep This is where include files would be wonderful. You can includ underneath the subfilter then all your sls files would look the same
00:34 schimmy joined #salt
00:37 bretep Vye: Like this https://gist.github.com/bretep/9221008
00:37 n8n joined #salt
00:37 bretep Here you would just diff dev.sls and prod.sls and you can see the differences, the name of the file is the enviornment, everything else would be the same
00:38 Vye bretep: interesting approach.
00:38 bretep subfilter already works, it was merged a month or so ago
00:39 bretep https://github.com/saltstack/salt/commit/3f8f99f7ae9750477fa2f0db457b2c630ee541e6
00:39 bretep I've got to go
00:39 Vye bretep: ttyl
00:43 schimmy1 joined #salt
00:43 danielbachhuber joined #salt
00:45 snuffeluffegus joined #salt
00:45 Networkn3rd joined #salt
00:46 pydanny joined #salt
00:47 kiorky joined #salt
00:48 taion809 joined #salt
00:52 Networkn3rd joined #salt
00:53 baniir joined #salt
00:56 pydanny joined #salt
01:00 NotreDev joined #salt
01:07 elfixit joined #salt
01:09 HeadAIX joined #salt
01:18 aleszoulek joined #salt
01:20 bemehow joined #salt
01:22 xzarth_ joined #salt
01:23 epicserve joined #salt
01:24 epicserve Trying salt for the first time and bootstraping the minion just stalls. http://cl.ly/image/1R080i2Z0p3d
01:25 pdayton joined #salt
01:27 christopherl joined #salt
01:34 epicserve nvm ... tried another one liner and it worked.
01:41 epicserve well crapola
01:41 epicserve Got this after trying to install apache, https://www.dropbox.com/s/a5oaez06xttdfrx/Screenshot%202014-02-25%2019.41.29.png
01:42 KyleG Package name is probably something like apache2
01:42 epicserve I'm just following this ... http://docs.saltstack.com/topics/tutorials/quickstart.html
01:42 KyleG epicserve: try naming it apache2 instead, I bet it'll work
01:42 epicserve So far it's 0 for 2
01:43 epicserve The Quickstart guide should be updated.
01:43 KyleG The package name differs by OS
01:43 KyleG I don't think that was specifically written for Ubuntu
01:43 KyleG It's more of a general idea, the assumption being that if you're getting into config management, you already have prior working knowledge of how the *nix OS works
01:44 GradysGhost joined #salt
01:44 paramah joined #salt
01:44 epicserve what is a *nix OS?
01:45 epicserve JK
01:45 l0x3py joined #salt
01:46 epicserve KyleG: I'm new to Salt, and I have no idea if salt has the ability to generically specify a package without the version.
01:47 KyleG it does
01:47 KyleG But the package name for ubuntu is "apache2" not "apache"
01:47 KyleG I looked it up
01:47 KyleG It'll attempt to install whatever the latest version available in the repo is
01:47 KyleG unless you specify a version
01:47 KyleG Specifying a version is actually an optional kwarg
01:48 Furao joined #salt
01:49 epicserve I'm just saying it might be helpful to add a note like, "Other OSes might name the package different (e.g. Ubuntu Server)."
01:53 n8n joined #salt
01:54 xmltok joined #salt
01:55 xl1 joined #salt
01:55 KyleG epicserve: I'm sure they would love help with the docs
01:56 KyleG http://docs.saltstack.com/topics/hacking.html
01:56 epicserve KyleG: Good point ... I should probably make a PR rather than complaining. :)
01:56 KyleG :P
01:58 pydanny joined #salt
02:04 lionel joined #salt
02:06 malinoff joined #salt
02:10 baniir joined #salt
02:13 LowWalker joined #salt
02:14 mgw joined #salt
02:23 snuffeluffegus joined #salt
02:27 zzzirk joined #salt
02:28 Guest28864 joined #salt
02:29 schimmy joined #salt
02:31 schimmy1 joined #salt
02:38 schimmy joined #salt
02:40 ravibhure joined #salt
02:45 ravibhure left #salt
02:45 schimmy joined #salt
02:50 schimmy joined #salt
02:50 forrest joined #salt
02:55 Ryan_Lane joined #salt
03:02 marsdominion joined #salt
03:04 marsdominion joined #salt
03:11 ravibhure joined #salt
03:11 ravibhure left #salt
03:12 NotreDev joined #salt
03:17 srage joined #salt
03:17 epicserve left #salt
03:20 temp joined #salt
03:20 temp question about pillars and include. Is there any way to include a file from a subdirectory?
03:21 temp it seems that "include: -file" will include from /pillars
03:21 blee joined #salt
03:23 temp ok, nevermind ... it's not really intuitive but "include: - subdir/file" seems to work
03:23 ravibhure1 joined #salt
03:23 rostam joined #salt
03:25 ravibhure1 left #salt
03:26 Linz joined #salt
03:30 yomilk joined #salt
03:32 seapasulli joined #salt
03:34 hunter joined #salt
03:35 krow joined #salt
03:36 hunter joined #salt
03:39 dave_den temp: - include:\n  - subdir.file
03:42 pydanny joined #salt
03:49 alunduil joined #salt
03:49 krow joined #salt
03:50 smcquay joined #salt
03:57 baniir joined #salt
03:58 smcquay joined #salt
03:59 baniir I'm getting "State ports.installed found in sls ... is unavailable" when trying to run highstate with salt version 2014.1.0
04:00 xmltok joined #salt
04:01 jeremyfelt joined #salt
04:04 baniir any thoughts on what could be going wrong?
04:05 malinoff joined #salt
04:12 manfred did you try checking the minion log file to see if it gave you something more specific?
04:15 schimmy joined #salt
04:17 Linz joined #salt
04:18 schristensen joined #salt
04:18 schimmy1 joined #salt
04:20 baniir i didn't. thanks for the tip. it looks like salt is expecting bash be installed while it wasn't. i'm testing now
04:22 Gifflen joined #salt
04:22 MedicalJaneParis joined #salt
04:24 ipalreadytaken joined #salt
04:24 Ryan_Lane joined #salt
04:25 christopherl left #salt
04:27 borgstrom joined #salt
04:33 baniir i needed bash
04:39 krow joined #salt
04:40 skynet joined #salt
04:40 gt3 joined #salt
04:41 srage joined #salt
04:42 schimmy joined #salt
04:44 srage joined #salt
04:46 JohnD__ joined #salt
04:47 baniir is there a way to use pillar data on a masterless salt run
04:48 jcockhren concerning: https://github.com/saltstack/salt/issues/10682
04:48 jcockhren Would it make the most sense to expose part of it as a runner?
04:48 gt3 left #salt
04:50 Furao joined #salt
04:50 thayne joined #salt
04:53 hardwire I'm recommending an advanced message queue system built via salt be called.. brine.
04:53 hardwire heh
04:53 zdong joined #salt
04:55 Math` joined #salt
04:56 dave_den baniir: yes, i thin you should be able to define pillar_roots with a masterless setup
04:56 dave_den jcockhren: seems like a runner module would make sense to have for that
04:57 dave_den hardwire: clever name. i like it. wondering about the details of the messaging system
04:58 dave_den dave_den: good job, keep it up
04:58 dave_den hey thanks, dave_den!
04:58 * hardwire shakes head
04:58 dave_den heh
04:59 hardwire dave_den: it'd be nice to be able to fire off events to a master in a fire-and-forget fashion
04:59 hardwire basically.. zmqmail :)
04:59 dave_den does event.fire_master not fill that job?
04:59 hardwire it times out if the master is awol
05:00 dave_den ah
05:00 hardwire I haven't tested --async that way
05:01 dave_den how will it be built around salt?
05:01 hardwire it could be build somewhat easily as a 3rd party spool handler for events
05:02 hardwire as an integrated component.. no idea.  I'm still making my way through Salt Air.
05:03 krow joined #salt
05:05 joehillen joined #salt
05:08 hardwire I'm at a point where I need to stop thinking about this as a way to run lots of jobs in parallel until I figure out how a everything works
05:08 hardwire I've got some simple top file stuffs going on.. some simple file serving.
05:09 malinoff hardwire, are you a programmer?
05:09 hardwire yup
05:09 hardwire python for 1 billion years.
05:09 malinoff hardwire, so the best way is to dive into the source code
05:09 malinoff Makes things clear
05:11 smcquay joined #salt
05:11 hardwire that's scary.
05:11 hardwire I'm still trying to figure out why things exist.
05:11 malinoff A bit :)
05:12 Gifflen joined #salt
05:12 Math` joined #salt
05:13 Gifflen joined #salt
05:13 borgstrom joined #salt
05:13 anuvrat joined #salt
05:14 malinoff http://postimg.org/image/p4tevnwgh/
05:14 hardwire It is impossible to learn about salt and watch Gurren Legann at the same time
05:14 Math` joined #salt
05:15 schimmy joined #salt
05:16 hardwire one thing I tried and completely failed to get working was firing off master events from python (0.17.5)
05:16 hardwire I got it to a point where I needed to understand sending events needed a node id for hashing and utilizing the hashed socket name.
05:17 hardwire I watched traffic go nowhere via tcpdump however
05:17 xmltok joined #salt
05:17 hardwire so I'm doing something silly.
05:17 malinoff >tcpdump
05:17 hardwire it looks like at some point salt-minion became possibly salt-minions.
05:17 malinoff Oh god :D
05:18 hardwire in a multi master environment it would be groovy to have a queue solution
05:18 hardwire especially if the plan going forward is to support multiple salt-minions per machine as needed
05:19 malinoff hardwire, why do you need this?
05:19 hardwire I need to send events off to the master. salt-call is fine but it is a bit slow for a handsome throughput.
05:19 hardwire due to it auditing the system first via the os probe
05:20 Gifflen joined #salt
05:20 malinoff What kinds of events?
05:21 hardwire my use case is to send back results from events happening on the minion machine
05:21 hardwire outside of salt
05:21 hardwire for instance.. "I need a dm-crypt authorization"
05:21 hardwire or.. "You should find somebody to eat my database"
05:23 malinoff And event.fire_master is not suitable for you?
05:23 gnugnu6 joined #salt
05:23 hardwire if I could get it to work from python I think it'd be an amazing start for me
05:23 malinoff Hm
05:23 hardwire salt-call is too slow due to otherstuffs that are being done.
05:23 malinoff I'm 100% sure it worked for me
05:23 hardwire I'm 100% sure it did too.
05:24 hardwire do you have some example minion code?  What's on the side doesn't work since it will attempt to hash an id of '' (which is a really boring and common md5sum)
05:24 hardwire the client socket contains a hash of the client id vs ''
05:24 hardwire once that was resolved however.. I saw nothing headed toward the master.
05:25 thayne joined #salt
05:25 hardwire I spent a bit in event.py
05:25 NotreDev joined #salt
05:26 Math` joined #salt
05:35 fllr joined #salt
05:36 xDamox joined #salt
05:41 Math` joined #salt
05:42 glen_ joined #salt
05:42 hardwire even added fancy print statements before the zmq publish event :)
05:42 hardwire faaancy
05:44 glen_ Hi guys, I've set up a salt master to be my mac and now I'm tying to set up a minion on vshpere. I've generated keys and placed them following the documetation and had a working set up using vagrant for minions. When running salt-minion test.ping on the minion I get an error that the master has rejected the public key.
05:44 malinoff hardwire, I have no idea what is 'hash' you're talking about
05:44 glen_ I've deleted and regenerated them and stil get this error
05:44 malinoff http://docs.saltstack.com/topics/event/index.html#firing-events - there is nothing about it
05:48 hardwire malinoff: the md5sum.  one sec
05:49 hardwire https://github.com/saltstack/salt/blob/develop/salt/utils/event.py#L141
05:49 hardwire nanaaa
05:50 hardwire so my client code has to look a lot different than the client side code in the docs
05:50 hardwire I connect to 'minion' not 'master' as the node which is fed to load_uri.
05:51 malinoff hardwire, do you have 'id' in your minion config?
05:51 hardwire it finds the socket... in my case: /var/run/salt/minion/minion_event_6c95257d5ac497fc9f9b64997d6ccfe9_pull.ipc
05:51 hardwire yeh.
05:51 hardwire no.
05:51 hardwire but.. yes..
05:52 thayne joined #salt
05:52 malinoff hardwire, is it empty? :)
05:52 hardwire if it were empty that hash would be d41d8cd98f00b204e9800998ecf8427e
05:52 hardwire the most boring md5sum in the universe.
05:52 mgw joined #salt
05:53 malinoff I still do not understand what's the problem, sorry :(
05:53 hardwire that's fine
05:54 hardwire https://gist.github.com/whardier/a9817d6089c3aab6844d
05:54 hardwire the python is the corrected version of the client code I made
05:54 hardwire the strace shows it connecting properly to the ipc socket and sending the protobuf
05:55 hardwire nothing is received on the master
05:55 hardwire I'm not sure how to address to the master.
05:56 dmalinovsky joined #salt
05:56 dmalinovsky joined #salt
05:57 hardwire like I'm having a difficult time following what exactly salt-call is doing right.
05:57 dmalinovsky hardwire, Are you saying, everything is ok - salt finds the socked, it sends the payload, but actually there is nothing on the interface?
05:58 hardwire yes.. there's a bit of a headslap here however.
05:58 hardwire https://github.com/saltstack/salt/blob/develop/salt/utils/event.py#L654
05:58 dmalinovsky hardwire, what is "headslap" ?:)
05:58 hardwire it looks like fire_master does some witchery that I wasn't aware of.
05:58 hardwire dmalinovsky: when you firmly slap your head with your palm.
05:58 hardwire palmface?
05:58 hardwire :)
05:59 dmalinovsky facepalm? :)
05:59 hardwire no.. I put my face down hard.. left my palm where it was.
05:59 dmalinovsky Ah
05:59 dmalinovsky Ok
05:59 dmalinovsky :D
05:59 malinoff joined #salt
06:00 hardwire it looks like fire_event was not what I needed at all if I wanted to push something up to the master.
06:00 malinoff hardwire, I think, this is the only option you have
06:00 hardwire and it looks like I'll need to do something else in order to create a situation where a StateFire class knows all about the authentication requirements.
06:01 ssshi joined #salt
06:02 clintberry joined #salt
06:02 hardwire looks like the event that heads up to the master is an encrypted payload before it even hits the socket
06:02 hardwire kudos on security
06:03 hardwire the salt-minion daemon doesn't appear to be at all responsible for that aspect of this payload.
06:04 manfred hardwire: facedesk?
06:05 hardwire qwertycheek.
06:06 hardwire hmm.. qwertynoggin.
06:06 hardwire yeh.
06:06 hardwire I like that more.
06:06 alunduil joined #salt
06:06 manfred all are strong candidates
06:07 hardwire now to see about setting an environment variable before doing a pkg.install
06:08 hardwire I need to install some apt packages but use a different path so that they don't get automatically started or added to a runlevel
06:09 CeBe joined #salt
06:10 leonardinius joined #salt
06:13 leonardinius_ joined #salt
06:17 xmltok joined #salt
06:19 ravibhure joined #salt
06:20 ravibhure left #salt
06:21 schimmy joined #salt
06:23 renoirb joined #salt
06:24 bretep Does anyone off the top of their head know how to get the environment from a pillar or grain when rendering a template?
06:25 ndrei joined #salt
06:29 yes456 joined #salt
06:31 yes456 where i can find salt deb pkg for 2014.1.0?
06:33 manfred http://docs.saltstack.com/topics/installation/debian.html
06:33 anuvrat joined #salt
06:34 manfred doesn't look like it has been updated yet
06:35 yes456 all doc updated for as per 2014.1.0 not pkg repo it quite *** ;(
06:36 malinoff yes456, use readthedocs (https://salt.readthedocs.org/en/v0.17.4/), Luke!
06:36 manfred yes456: you could always build from pip
06:37 manfred or python setup.py install from a git checkout on the v2014.1.0 tag
06:38 schimmy joined #salt
06:38 tomspur joined #salt
06:38 tomspur joined #salt
06:38 yes456 manfred: thanks for readthedocs link. but for prdouction use it easy install from repo
06:42 joehh yes456: working on it right now
06:42 joehh debian or ubuntu?
06:42 yes456 ubuntu
06:42 yes456 joehh: ubuntu lucid
06:43 manfred good luck with that
06:43 joehh ok - will be a few hours then - I'm doing debian first, then ubuntu - I'll start with lucid and precise though
06:44 _fllr_ joined #salt
06:44 yes456 thank u guys :)
06:46 srfwii joined #salt
06:49 Hipikat joined #salt
06:54 krow joined #salt
06:58 n8n joined #salt
06:58 anuvrat joined #salt
07:00 millz0r joined #salt
07:09 NotreDev joined #salt
07:10 kiorky /b 38
07:12 terminalmage yes456: debian package should be done soon
07:12 terminalmage our debian packager is working on it
07:13 terminalmage night folks
07:13 yes456 terminalmage: sure thanks waiting ubuntu lucid
07:18 bretep Awesome! Can't wait for the ubuntu package. :) It will be nice to replace my custom one
07:20 leonardinius joined #salt
07:27 cewood joined #salt
07:36 bretep Is this where I'm going to look for the latest build of salt in the upcoming day for Ubuntu? https://launchpad.net/~saltstack/+archive/salt?field.series_filter=precise
07:37 Alpha1131- joined #salt
07:37 Alpha1131- left #salt
07:37 MedicalJaneParis joined #salt
07:42 mephx joined #salt
07:46 ml_1 joined #salt
07:48 Kenzor joined #salt
07:49 Iwirada joined #salt
07:50 fllr joined #salt
07:51 AndChat|359961 joined #salt
07:51 balboah joined #salt
07:52 _fllr_ joined #salt
07:53 carlos joined #salt
07:54 techdragon joined #salt
08:01 lessismore joined #salt
08:09 skynet joined #salt
08:10 slav0nic joined #salt
08:17 xmltok joined #salt
08:18 harobed joined #salt
08:22 fllr joined #salt
08:33 jrdx joined #salt
08:34 jrdx left #salt
08:34 Ztyx joined #salt
08:35 giantlock_ joined #salt
08:36 Ztyx I'm thinking about implementing my own custom formatter. Where do I put it? In "_output"? "_formatter"?
08:36 krow joined #salt
08:36 zooz joined #salt
08:37 johtso joined #salt
08:39 krow1 joined #salt
08:41 astol joined #salt
08:44 sroegner joined #salt
08:46 yomilk joined #salt
08:56 anuvrat joined #salt
08:56 sabayonuser_ joined #salt
08:58 schimmy joined #salt
08:59 leonardinius__ joined #salt
09:05 sfello joined #salt
09:07 miguitas joined #salt
09:08 kadel joined #salt
09:08 ambientsound joined #salt
09:12 joehh bretep: yes
09:21 nonuby joined #salt
09:21 krow joined #salt
09:22 fllr joined #salt
09:27 mband joined #salt
09:31 astol joined #salt
09:33 Ponycorn joined #salt
09:34 zzzirk joined #salt
09:37 gammalget joined #salt
09:42 Ymage joined #salt
09:52 fllr joined #salt
09:54 _fllr_ joined #salt
09:56 xmj joined #salt
09:56 xmj moin
10:01 anuvrat joined #salt
10:02 tr_h joined #salt
10:04 MrTango joined #salt
10:12 che-arne joined #salt
10:22 aleszoulek joined #salt
10:26 johtso joined #salt
10:29 ndrei joined #salt
10:29 astol joined #salt
10:33 sroegner joined #salt
10:36 johtso Is this the right place to look to see when the new salt version is up on the ubuntu PPA? https://launchpad.net/~saltstack/+archive/salt
10:37 leonardinius joined #salt
10:45 faldridge joined #salt
10:50 david_a joined #salt
10:52 fllr joined #salt
10:52 joehh johtso: yes - not up yet
10:53 joehh working on debian now, then doing ubuntu
10:53 johtso okay, cheers
10:53 johtso looking forward to it :)
10:53 joehh me too :)
10:56 johtso Couldn't believe my luck. Found a fix that I needed that had only just missed out of the last release, and got a bit despondent, but then saw a brand new release was shipping!
10:56 joehh I'll be messaging here and the mailing list (google group) when they are up
11:03 bhosmer joined #salt
11:03 Sypher|NL joined #salt
11:15 nn0101 joined #salt
11:15 nn0101 joined #salt
11:17 sroegner joined #salt
11:18 think-free joined #salt
11:20 think-free Hi, is there a way to write a log message from a sls ?
11:22 Furao joined #salt
11:22 nn0101 joined #salt
11:24 vbabiy joined #salt
11:31 ravibhure1 joined #salt
11:32 ravibhure1 left #salt
11:39 Ztyx left #salt
11:40 krow joined #salt
11:42 astol joined #salt
11:48 ndrei joined #salt
11:49 xmj sure- at the worst, cmd.run with logger
11:51 ckao joined #salt
11:52 fllr joined #salt
12:05 TheRealBill joined #salt
12:05 sabayonuser_ joined #salt
12:06 zzzirk joined #salt
12:08 nonuby joined #salt
12:09 TheRealBill joined #salt
12:09 bhosmer joined #salt
12:12 leonardinius joined #salt
12:14 think-free xmj : well, sure .. but no buildin function ?
12:17 justBob joined #salt
12:17 xmj did you find nothing on google?
12:23 krow joined #salt
12:28 MrTango joined #salt
12:28 viq think-free: that's actually an interesting question. Seems you can specify an returner per execution, but I haven't found any mention of being able to specify a returner in a state...
12:29 ndrei joined #salt
12:30 think-free xmj : nope
12:37 diegows joined #salt
12:39 krak3n` joined #salt
12:41 _viq joined #salt
12:41 _viq test
12:46 nonuby joined #salt
12:46 nonuby joined #salt
12:48 xt no ubuntu packages for latest salt?
12:48 yomilk joined #salt
12:48 johtso joined #salt
12:50 viq xt: in the works from what I hear
12:52 fllr joined #salt
12:53 xmj how different are ubuntu packages from debian ones, anyway?
12:56 elfixit joined #salt
12:56 gnugnu6 joined #salt
12:57 jrdx joined #salt
12:58 sroegner joined #salt
12:59 markm joined #salt
13:03 godber joined #salt
13:16 ChaosPsyke joined #salt
13:23 fragamus joined #salt
13:23 baniir joined #salt
13:23 jaimed joined #salt
13:24 krow joined #salt
13:25 Katafalkas joined #salt
13:26 Katafalkas What is a good way to use salt-cloud from within salt-runner ?
13:26 ndrei joined #salt
13:33 dwiden joined #salt
13:33 dwiden I know its early,  but is anyone here?
13:34 ipmb joined #salt
13:36 xmj 15:33:41 < dwiden> I know its early
13:37 Math` joined #salt
13:37 xmj you have a really California-centric world view
13:38 dwiden I've had to adopt that world view, I talk to a lot of mozilla devs and they're on CA time.  Plus its still a little early for the east coast
13:38 HeadAIX joined #salt
13:39 vlcn morning guys
13:40 sfello joined #salt
13:40 dwiden Anyway, do you happen to be familiar with installing software on windows minions using the windows software repository?
13:40 dwiden good morning
13:48 djaime joined #salt
13:49 xmj dwiden: rm -rf / && install a real OS
13:49 xmj in the absence of rm, format c: (:
13:50 dwiden I wish I wasn't using windows, but its necessary
13:50 jcsp xmj: you're being unhelpful.  It would be better if you could keep quiet when you don't have anything useful to add.
13:50 dwiden I'm trying to get Visual Studio on it now, I'm gonna have to get IDA on eventually
13:51 Furao joined #salt
13:52 fllr joined #salt
13:54 xmj left #salt
13:59 dwiden found a hack to do Visual Studio.  how do I put "run this command as long as you have done this file.recurse" state into a cmd.run state?
13:59 jcsp left #salt
14:00 viq dwiden: require/watch
14:00 viq dwiden: also may want to make this cmd.wait so you don't run this every time
14:00 dwiden normally you would do: require: pkg: Python, or something along those lines.  Does this change since its a file.recurse, or can I just use pkg?
14:01 viq require: - file: your_recursive_state's_name
14:01 viq pkg is for pkg, for file you use file :P
14:03 dwiden viq: okay thank you.  so using watch will only run the command if the files have changed?
14:03 dwiden also, is file.recurse designed to work with a large directory?  I need to transfer about 400 MB
14:05 viq dwiden: watch with cmd.wait yes, will trigger only (and every time) the watched state changes. And I heard comments about people copying gigabytes over salt:// protocol, so I guess it should work, but I never used it for anything large so can't comment personally
14:05 dwiden viq: okay great, thank you very much for the help
14:06 sroegner joined #salt
14:09 dwiden viq: I have one more question.  if the named directory in file.recurse does not exist on the minion, it will be created, right?  so if I say "name: C:\temp\VisualStudio", it should just make those for me?
14:11 vlcn how would I go about converting this sls https://gist.github.com/kelchm/9229997 into a runner module?
14:11 viq dwiden: it should, though on unix if the intermediate temp wouldn't exist it would create it with weird permissions
14:11 viq dwiden: IIRC there are actually flags/settings for this in the state
14:12 dwiden gotcha, since I'm on a windows minion I'm sure this will be an adventure
14:12 viq vlcn: runner? Not an execution module?
14:13 vlcn whichever makes more sense I guess ;)
14:13 _david_a joined #salt
14:14 viq And how... I guess look at some of existing modules ;) Something that comes to mind that isn't overly complex is https://github.com/saltstack/salt/blob/develop/salt/modules/freebsdjail.py but there should be others as good examples too
14:16 racooper joined #salt
14:18 juicer2 joined #salt
14:18 Pinkanil joined #salt
14:19 fllr joined #salt
14:19 mpanetta joined #salt
14:20 johtso joined #salt
14:20 _fllr_ joined #salt
14:20 Pinkanil_David_p joined #salt
14:21 Pinkanil_David_p Good Morning
14:21 Pinkanil_David_p Having a bit of an issue
14:21 Pinkanil_David_p lets say, I have a role called "pg_master"
14:21 Pinkanil_David_p and one of my state files has code that should only be executed on that role.
14:22 Pinkanil_David_p This is what I wrote as an if condition, but for some reason isn't registering and executing
14:22 Pinkanil_David_p {% if grains['roles'] == 'pg_master' %}
14:23 psyl0n joined #salt
14:23 Pinkanil_David_p I can't figure out where that might be wrong, could anyone check and help me out with that?
14:23 krow joined #salt
14:23 viq Pinkanil_David_p: paste that state somewhere
14:25 faldridge joined #salt
14:25 oznt joined #salt
14:25 Pinkanil_David_p viq: If I might ask, what would you be looking for? If it helps, the state works, everything executes in it on all the right places except for what is inside that if statement on the computer with the role pg_master
14:26 johtso Pinkanil_David_p, and grains.items shows the expected grain in place?
14:26 mpanetta joined #salt
14:26 oznt hi everyone, salt-key -L shows my salt minions, but I am not seeing any output when doing salt '*' test.ping, do you happen to have an idea?
14:28 faldridge joined #salt
14:28 timoguin joined #salt
14:28 Pinkanil_David_p johtso: yes, it does. grains.get roles shows "    - pg_master" for that id
14:28 Pinkanil_David_p among other roles
14:28 bryano__ joined #salt
14:29 Pinkanil_David_p also `salt -G "roles:pg_master"` works to get that particular nodes information in general
14:30 johtso Pinkanil_David_p, well, it's a list, so I think you want {% if 'pg_master' in grains['roles'] %}
14:31 eofs joined #salt
14:33 seapasulli joined #salt
14:35 mgw joined #salt
14:35 vejdmn joined #salt
14:36 Pinkanil_David_p johtso: thanks, trying it now
14:38 Pinkanil_David_p johtso: many many thanks, it works!
14:41 Gordonz joined #salt
14:41 quickdry21 joined #salt
14:41 johtso no problem :)
14:42 Gordonz joined #salt
14:43 rgbkrk joined #salt
14:44 thayne joined #salt
14:45 mattmtl joined #salt
14:45 blee joined #salt
14:48 pdayton joined #salt
14:51 Networkn3rd joined #salt
14:52 fllr joined #salt
14:53 oraqol Hello, I'd appreciate some help with salt.  Is there a way to schedule highstate on the master in batches by grain?
14:54 jaimed joined #salt
14:54 viq oraqol: http://docs.saltstack.com/topics/jobs/schedule.html ?
14:54 jeddi joined #salt
14:56 zzzirk joined #salt
14:56 budrose joined #salt
15:00 oraqol viq: so does this mean that I can just place the batch flag in the function: parameter
15:00 oraqol so it would look like this? - function: -G "grain:match" -b 20 state.highstate
15:00 DaveQB joined #salt
15:03 oraqol or would it be - function: highstate args: - '-G "grain:match" -b 20'
15:05 linjan joined #salt
15:06 GradysGhost joined #salt
15:08 oraqol I hate to be a pain, but I've already seen the link you shared out, but it doesn't specify how to use -b and -G with the scheduler
15:10 viq oraqol: well, for grains that should be simple, as you can put it in pillar, and you can give out pillars by grain
15:10 viq Batch, I'm not sure about yet
15:11 oraqol oh ok
15:14 viq oraqol: I wonder if adding batch or batch-size to the settings there would work
15:15 y0j joined #salt
15:15 oraqol I'm looking into it, as of now I'm using conditional statements that match the grain in the sls file
15:15 ahtik joined #salt
15:16 oraqol so I can just run highstate across * and configs distribute according to that.  How would using pillars be more beneficial than my current settings?
15:17 viq Not sure if they would. But what I meant is that you can put the schedule in a pillar, and that way vary it by grain
15:18 david_a joined #salt
15:19 sabayonuser_ joined #salt
15:19 djaime joined #salt
15:20 HeadAIX joined #salt
15:20 ahtik joined #salt
15:21 oraqol Ok, I will look into it, thanks viq :]
15:21 pass_by_value joined #salt
15:23 krow joined #salt
15:23 rostam joined #salt
15:25 joehh joined #salt
15:26 ahtik joined #salt
15:28 Gifflen joined #salt
15:28 bt joined #salt
15:30 kaptk2 joined #salt
15:30 ndrei joined #salt
15:31 rojem joined #salt
15:31 ahtik joined #salt
15:33 Nazca joined #salt
15:34 SkyRocknRoll__ joined #salt
15:35 fllr joined #salt
15:36 psyl0n joined #salt
15:36 psyl0n joined #salt
15:36 Linuturk so, I'm comparing rackspace product support between Ansible and Salt. I noticed that Salt seems to have compute support, but nothing else. Ansible has support for a wide range of Rackspace products, including load balancers, networks, queues, etc etc. Ansible has the advantage having a Rackspace employee contributing. If we wanted to add similar support to Salt, where would we start?
15:37 kermit joined #salt
15:37 ahtik joined #salt
15:38 pydanny joined #salt
15:42 ahtik joined #salt
15:43 forrest joined #salt
15:43 jforest anyone seeing issues on centos6.5 with pkgrepo.managed
15:44 alunduil joined #salt
15:46 jforest specifically, on centos6.5 pkgrepo.managed just fails
15:46 jforest even though on 0.17.5 on centos5 it works great
15:46 jforest on 0.17.5 on centos6.5 fails
15:47 Eugene TBQH I just file.managed in my /etc/yum.repos.d/mycompany.repo file
15:47 forrest jforest, https://github.com/saltstack/salt/issues/10261 ?
15:47 Psi-Jack if [ 'apps', 'web' ] in salt['grains.get']('roles', [])   -- Is this correct to check if apps or webs is in the roles grain?
15:47 Psi-Jack Specifically: OR, not and.
15:47 jforest forrest: yah, it's http
15:48 jforest actually my centos5 boxes may be 0.17.4, let me check that
15:48 danielbachhuber joined #salt
15:48 ahtik joined #salt
15:49 jforest forrest: yes, I'm mistaken, my centos5 machines are 0.17.4, and my centos6 is 0.17.5
15:50 forrest jforest, ahh
15:50 jforest however my ubuntu 10 are 0.17.5
15:50 jforest and they work fine
15:50 jforest so for me, at least, it's centos6 issues
15:50 Psi-Jack CentOS 5 has an outdated zeromq.
15:50 jforest centos5 works fine
15:50 jforest centos6 fials
15:50 jforest fails
15:51 jforest ubuntu10, fine, ubuntu12, fine
15:52 xzarth joined #salt
15:52 lionel_ joined #salt
15:53 sabayonuser_ joined #salt
15:54 ahtik joined #salt
15:54 forrest jforest, I don't see why it would be treated differently on centos6 versus 5, when you run it with -l debug on that cent6 machine, what error do you get?
15:55 mgw joined #salt
15:55 ndrei joined #salt
15:55 jforest forrest: it doesn't provide much of use
15:55 forrest jforest, did you enable debug loggin on the minion then? That would be the next step
15:56 jforest [ERROR   ] Failed to configure repo 'percona-repo': ''
15:56 jforest yes, I did debug and then trace
15:56 jforest nothing
15:56 jforest [INFO    ] Fetching file ** skipped **, latest already in cache 'salt://repositories/percona.sls'
15:56 jforest [DEBUG   ] Jinja search path: '['/var/cache/salt/minion/files/base']'
15:56 jforest [DEBUG   ] Rendered data from file: /var/cache/salt/minion/files/base/repositories/percona.sls:
15:56 jforest percona-repo:
15:56 jforest pkgrepo.managed:
15:57 jforest - humanname: percona
15:57 jforest - baseurl: http://repo.percona.com/centos/$releasever/os/$basearch/
15:57 jforest - enabled: 1
15:57 forrest can you use pastebin or gist? Otherwise you'll get kicked for spam
15:57 jforest - priority: 5
15:57 jforest - gpgcheck: '0'
15:57 jforest - key_url: http://www.percona.com/downloads/RPM-GPG-KEY-percona
15:57 jforest [DEBUG   ] Results of YAML rendering:
15:57 jforest OrderedDict([('percona-repo', OrderedDict([('pkgrepo.managed', [OrderedDict([('humanname', 'percona')]), OrderedDict([('baseurl', 'http://repo.percona.com/centos/$releasever/os/$basearch/')]), OrderedDict([('enabled', 1)]), OrderedDict([('priority', 5)]), OrderedDict([('gpgcheck', '0')]), OrderedDict([('key_url', 'http://www.percona.com/downloads/RPM-GPG-KEY-percona')])])]))])
15:57 Mr_N ohgod.
15:57 jforest [INFO    ] Executing state pkgrepo.managed for percona-repo
15:57 jforest [ERROR   ] Failed to configu
15:57 jforest yup, sorry
15:58 jforest http://pastebin.com/Pz1RqEVa
15:59 ahtik joined #salt
15:59 akoumjian joined #salt
16:01 nyx joined #salt
16:02 Guest23992 I have a pull-request I'm considering making: it allows specification of master key (pub/priv) when creating masters with salt-cloud (in the interest of multi-master setups mainly), I'm asking here because I want to be 100% sure I'm not re-creating functionality already handled
16:03 abe_music joined #salt
16:03 jdenning joined #salt
16:04 ahtik joined #salt
16:04 Guest23992 the diff:  https://gist.github.com/nyx/8ed344254250c7b25c79
16:05 y0j joined #salt
16:10 ahtik joined #salt
16:10 tyler-baker joined #salt
16:11 renoirb Hi all, I just seen that 2014.1 is out.
16:11 renoirb Is it normal that the deb repositories doesn't show?
16:11 forrest renoirb, it may not have been built yet
16:11 renoirb oh, yeah. makes sense :)
16:13 frasergraham joined #salt
16:13 smcquay joined #salt
16:14 dave_den joined #salt
16:14 Mr_N At the risk of a rather simple question; I wonder if there are any downsides I'm not seeing to just always populating anything requiring information about the master from pillar['salt-master']
16:15 jalbretsen joined #salt
16:15 ahtik joined #salt
16:17 bizboo joined #salt
16:20 ahtik joined #salt
16:23 Gifflen joined #salt
16:23 krow joined #salt
16:23 Gifflen joined #salt
16:24 hunter joined #salt
16:25 NotreDev joined #salt
16:25 jmccree Anyone have any clues what could be causing this error "No Top file or external nodes data matches found"?
16:25 jmccree It's only happening for one minion (redhat) I just added.
16:26 srage joined #salt
16:27 Gifflen joined #salt
16:28 Gareth morning.
16:29 budrose joined #salt
16:32 ndrei joined #salt
16:33 sfello joined #salt
16:33 jmccree this is a head scratcher
16:34 Ztyx1 joined #salt
16:36 Ztyx1 I'd like to create users using a state, preferably using some kind of loop. I could "names:" for this, but in this case I'd like to supply "fullname", too. How can I loop over items with multiple properties (in this case username and fullname)?
16:36 srage joined #salt
16:36 Mr_N Ztyx1: Have you looked at pillar yet?
16:37 Mr_N UtahDave has some good examples in his Gists of how to do user management
16:38 jeremyfelt joined #salt
16:38 Gifflen joined #salt
16:39 dave_den jmccree: that just means you are not matching the minion ID in your top files for applying statefiles to that minion
16:39 forrest Ztyx1, https://gist.github.com/UtahDave/3785738
16:39 jmccree dave_dev, I have a '*': in my base/top.sls
16:40 forrest jmccree, can you apply individual states to that machine?
16:40 jeremyBass left #salt
16:40 dave_den jmccree: what are you calling when you get that error
16:41 oraqol left #salt
16:41 Ztyx1 Mr_N: Yes, but I wasn't sure how to get the hierarchy right. Looking at UtahDave's gist I understand he uses dummy keys "user1", "user2"…that explains things. Thanks
16:41 Ztyx1 forrest: Thanks!
16:41 jmccree dave_den, highstate. forrest good idea checking
16:41 forrest Ztyx1, yea np.
16:42 Gifflen joined #salt
16:43 jmccree forrest, ah nope. And "salt-call cp.list_master" returns nothing on the minion.
16:43 forrest dun dund dunnn
16:43 forrest jmccree, is the minion properly hoooked up to the master? Do you see the IP in netstat from the master?
16:44 seapasulli_ joined #salt
16:45 FL1SK joined #salt
16:45 blee_ joined #salt
16:46 jmccree forrest, ip shows up on netstat, test.ping works, and running minion as debug I can see activity when I run a state from master.
16:46 jmccree But it appears like the file server isn't working for some reason.
16:48 forrest jmccree, can you hit the file server from the minion?
16:49 psyl0n joined #salt
16:54 Ztyx1 I'd like to look at a generated template on the master as seen from a minion. Is that possible?
16:54 jmccree forrest, yep.
16:55 krow joined #salt
16:56 ipalreadytaken joined #salt
16:58 jmccree forrest, I've got 10+ ubuntu minions working fine. But I just added two redhat ones and they both show this issue.
16:59 jmccree gonna try running master in debug
17:00 mjulian joined #salt
17:00 mjulian joined #salt
17:00 MedicalJaneParis joined #salt
17:00 Ztyx1 Figured it out; state.show_low_sls.
17:00 pdayton joined #salt
17:02 pass_by_value joined #salt
17:04 bemehow joined #salt
17:05 patrek_ on centos, mysql installs without a root password. I have a state that sets a value for the root password. But if I apply my state 2 twice, it will fail the second time because it now needs a password. Is there a check I can do to use connection_pass only if needed?
17:09 jeremyBass joined #salt
17:09 xmltok joined #salt
17:11 Mr_N Either my head is full of bricks, or Jinja is erroring on a line that is a comment.
17:12 Mr_N #{% for master in salt_masters %}    <======================
17:12 Iwirada left #salt
17:13 R_Shackleford joined #salt
17:13 Mr_N Oh; I am dull.  Of course it'd try to populate the comments.
17:13 Mr_N if it's a jinja var.  blah.
17:13 ajw0100 joined #salt
17:13 Mr_N Wonder how I make a "comment comment..."
17:13 Mr_N ><
17:13 Ztyx1 left #salt
17:13 mgw joined #salt
17:14 patrek_ Mr_N, you want to put your jinja in comment so it will not be interpreted, or do you want to write comments in the resulting file?
17:15 mjulian joined #salt
17:15 mjulian joined #salt
17:16 Gifflen joined #salt
17:17 Mr_N patrek_: I had been attempting to put jinja in a comment that wouldn't get intepreted, but I was rather shortbus in how I chose to do it.
17:17 dave_den {# this is a jinja comment #}
17:17 Mr_N ><
17:17 dave_den #{% this is jinja code behind a yaml comment %}
17:17 Mr_N dave_den: Indeed; I was just dense and didn't realize this distinction would need to exist beforehand :(
17:18 Mr_N Mostly beating myself up over this I feel like it should have been obvious to me that it'd behave in this fasion.
17:18 dave_den yeah, two different renderers
17:18 ahtik joined #salt
17:19 berto- joined #salt
17:20 jslatts joined #salt
17:20 Gifflen joined #salt
17:23 jmccree dave_den, yeah I've got minion and master with debug all.. still no clue what's wrong :/
17:24 Gifflen_ joined #salt
17:24 pydanny joined #salt
17:24 Gifflen_ joined #salt
17:26 jab416171 joined #salt
17:27 dave_den is your minion pinned to a particular environment?
17:28 mgw joined #salt
17:30 joehillen joined #salt
17:30 KyleG joined #salt
17:30 KyleG joined #salt
17:31 jmccree nope
17:33 smcquay joined #salt
17:33 UtahDave joined #salt
17:35 dave_den there should be something in the debug/all logs from the master and minion..
17:36 Mr_N Interesting.  I have a command that when run via salt-call in the minion completes successfully, but run on the master returns nothing
17:36 krow joined #salt
17:36 gcbirzan joined #salt
17:36 MedicalJaneParis joined #salt
17:37 Mr_N Oh; but now after running it once manually on the minion, it now runs on the master.
17:37 Mr_N This is a recurring problem I've noticed, sometimes the salt command from the master to a distinct minion returns silently; but without any sort of meaningful success.
17:37 dave_den Mr_N: make sure you use a long enough timeout on the master.  -t 300 for five minute
17:38 ipalread_ joined #salt
17:38 Mr_N dave_den: That'd do it. I had set it to 10, so I thought that would be sufficient, or rather, that it'd yell a lot harder if I violated it.
17:38 Mr_N Thanks for the pointer.
17:39 ipalreadytaken joined #salt
17:42 ipalrea__ joined #salt
17:42 Gordonz joined #salt
17:42 [diecast] joined #salt
17:43 Gordonz joined #salt
17:47 mgw1 joined #salt
17:51 psyl0n joined #salt
17:52 * viq waits for 2014.1 to hit repos so he can verify and possibly complain about wrong timestamps sent to logstash
17:52 ndrei joined #salt
17:53 jmccree dave_den, comparing working minions vs not, this part never happens: "[INFO    ] Fetching file ** done ** 'top.sls'"
17:54 mjulian joined #salt
17:54 leonardinius joined #salt
17:57 schimmy joined #salt
17:58 gcbirzan So, um. I'm trying to get salt-minion to work, but the version I got from the bootstrap seems to be totally broken
17:59 gcbirzan In particular, https://github.com/saltstack/salt/commits/develop/salt/config.py is missing a couple of fixes that make it... well, work. It dies on empty lines in /etc/hosts, tries to run WIndows code on Linux...
18:00 gcbirzan In particular, I seem to have 0.17.5-1~bpo60+1~dst.1 which is the only version provided by the repo it added. Is this supposed to go away (I see those bugs were fixed two days ago), or should I just get a(working)nother version?
18:01 Vye New release. woooo!
18:01 Math` joined #salt
18:01 gcbirzan Actually, funnily enough, on my workstation, I get a similar version... Great.
18:02 schimmy joined #salt
18:02 gcbirzan Or, let me ask differently. What version of salt should I be using?
18:04 forrest jmccree, sorry was in a meeting, that is odd. Did you figure it out?
18:05 jmccree forrest, not yet. It appears that the _file_list cmd is sent and received fine, but the master never get's the file _serve_file command. And the code doesn't contain any debug for that failure case.
18:07 dave_den what versions are you using on the minion and master?
18:07 dave_den jmccree: salt —versions-report      on minion:    salt-call —versions-report
18:08 jmccree dave_den, thinking that may be the issue.
18:08 Gifflen joined #salt
18:08 jmccree master, Salt: 0.17.5. minion, Salt: 2014.1.0
18:08 whiskybar joined #salt
18:09 jrdx joined #salt
18:10 dave_den as a general rule your master should always be the same version or newer than your minions. i'm not aware of what would cause a 2014.1.0 to not talk with a 0.17.5 minion though
18:11 dave_den rather a 2014.1.0 minion
18:14 ndrei joined #salt
18:15 seapasulli joined #salt
18:15 jmccree dave4925, that was it. switched from epel-testing to epel and things work now
18:16 jmccree dave_den, forrest : Thanks for all the help. switching from epel-testing to epel version corrected things.
18:16 forrest were you on 2014.1.0?
18:16 jmccree yep
18:17 forrest odd
18:17 forrest and these were cent 6.5 machines?
18:17 jmccree 6.4
18:18 hunter joined #salt
18:19 jmccree I've got one client that requires centos. Everything else I do is ubuntu, so I'm not the most familiar with centos.
18:20 hunter_ joined #salt
18:20 Ryan_Lane1 joined #salt
18:20 rojem joined #salt
18:21 MedicalJaneParis joined #salt
18:21 quickdry21 joined #salt
18:21 schimmy joined #salt
18:23 zooz joined #salt
18:24 Gifflen joined #salt
18:24 zdong is there anyone who has sucess to set up custom exection module on remote git server ?
18:25 zdong i have a custom module. I put it on local master /srv/salt/_modules and the master picked it up after restart.
18:26 zdong however, if i put it on remote git repo, under _modules in file roots, the master does not seem to pick it up.
18:26 timoguin zdong, i'm pretty sure it polls the remote repo and updates. might just be taking a minute
18:27 zdong i waited and it did not work. what trouble-shooting steps would you suggest to check ?
18:28 krow joined #salt
18:28 timoguin you can try deleting the master's cache at /var/cache/salt/master/gitfs and restarting
18:28 zdong i am trying it now.
18:29 Ryan_Lane1 joined #salt
18:33 bhosmer joined #salt
18:33 NotreDev joined #salt
18:33 zdong timoguin, i purged /var/cache/salt/master and restarted salt master. that did not work.
18:34 zdong during restarting, i looked at "ps" output and saw it wasing polling git repo.
18:35 zdong but for whatever reason, it did not load anyting from _modules.
18:35 timoguin zdong, that's surprising.
18:35 schimmy joined #salt
18:35 timoguin try changing into /var/cache/salt/master/gitfs/<HASH> and doing a 'git log --all'
18:35 timoguin jsut to make sure it has all yoru commit history
18:36 Ryan_Lane1 joined #salt
18:38 zdong i did the commit several days back. now "git log --all" has a lot of other people's commit .
18:38 zdong is there anyway to see what gitfs has downloaded ?
18:38 hunter_ joined #salt
18:39 timoguin yea, try a 'git checkout master' in that directly
18:39 timoguin *directory
18:39 saurabhs joined #salt
18:39 timoguin assuming your trying to see the master branch
18:40 timoguin if you're using multiple environments with branches, you'll want to checkout the appropriate branch
18:40 zdong yeah. git checkout master and i saw _modules/mymodule.py is there.
18:40 krow joined #salt
18:41 timoguin so it's just not showing up on the minion?
18:41 Gifflen joined #salt
18:42 dave_den zdong: does running this on a minion sync it?  salt-call saltutil.sync_all
18:42 timoguin try this: salt 'minion' saltutil.refresh_modules
18:42 timoguin damn you, dave_den!
18:42 timoguin ;-)
18:43 zdong it is not on minion. i am not sure about if it is on master or not, how can i check ?
18:43 timoguin zdong, either of the commands we just posted with sync new stuff down to the minions
18:43 dave_den zdong: is this an execution module for minions, or is this an extension module for the master?
18:44 zdong timoguin, dave_den, i ran saltutil.sync_modules command. will run saltutil.sync_all now.
18:44 zdong it is an execution module for minion.
18:45 mgw joined #salt
18:46 ipalreadytaken joined #salt
18:46 hardwire I guess I didn't realize there was a version schema change.
18:46 zdong ran both saltutil.sync_all and saltutil.refresh_modules. not helping.
18:46 jgelens joined #salt
18:47 hardwire and now the question about debian 2014.1.0 packages makes a lot more sense
18:48 timoguin zdong, salt '*' sys.list_modules should show it if it's available on the minion
18:49 timoguin if it's not there, maybe try restarting the minion.
18:49 zdong ran sys.list_modules and it is not listed.
18:49 dave_den zdong: how are you testing to see if the module is being synced to the minion?
18:50 timoguin if it's still not there after that, your minion might be pulling from a different environment. or there may be an issue with your module.
18:50 zdong restarted minion. no luck.
18:50 timoguin other than that, i'm out of ideas
18:50 dave_den there may be a syntax error or similar in your module that is preventing the minion from loading it
18:50 zdong i tested my module with local /srv/salt/_modules and it works.
18:50 yomilk joined #salt
18:50 razor joined #salt
18:51 zdong i tested my module with salt-call sys.doc nsutil   (name of the module is nsutil.py)
18:51 dave_den zdong: how are your file_roots and gitfs configured on the master?
18:51 forrest jmccree, ok I'll try to remember to spin up a 6.4 machine with 2014.1 tonight to see if I can dup the problem.
18:51 leonardinius hello guys. I try to proviosion salt-master using salt-cloud and `make_master` and would like to make my new ec2 master to use gitfs. Is there any easy way how to automatically install gitpython and gitdb (for gitfs support). At the moment I get non-working master (no states, because of gitfs fails without gitpython and gitdb).
18:51 zdong there is nothing in /etc/salt/master (all comment out).
18:52 dave_den zdong: also, are you seeing your states from gitfs when you run:  salt 'yourminion' cp.list_master
18:52 Gifflen joined #salt
18:52 zdong in /etc/salt/master.d, there are fileserver_backend.conf .
18:53 zdong ran salt 'yourminion' cp.list_master and it listed all my custom states.
18:53 gfa hello, how can i launch an state from an state, doing cmd.run salt-call -l debug state.sls foobar doesnt' work
18:54 zdong file_roots is not defined anywhere.  i think it will use default.
18:54 timoguin gfa, not sure what your use case is, but that might be better as a simple include
18:54 dave_den gfa: you want to simply include the state file.  http://docs.saltstack.com/ref/states/highstate.html#include-declaration
18:54 zdong fileserver_backend: - git
18:54 jmccree forrest, for what it's worth, it was repeatable on both 6.4 ec2 instances.
18:55 ipalreadytaken joined #salt
18:55 leonardinius do everybody install their masters+gitfs manually or what is the solution to bootstrap master on the cloud?
18:55 forrest jmccree, gotcha
18:55 gfa timoguin: dave_den i can't use include, the second state has a cp.get_file_str, the file is generated on the first state file
18:55 zdong the fact that i can see _modules in /var/cache/salt/master/gitfs puzzled me.
18:55 Katafalkas joined #salt
18:55 kermit joined #salt
18:56 gfa if file don't exists it will fail
18:56 Gifflen_ joined #salt
18:56 dave_den gfa: use requisites or orders to arrange dependency or order of state runs
18:56 gfa and the second state is a call to event.fire_master module, so i can't require on the first
18:56 chrisjon_ joined #salt
18:56 ipalreadytaken joined #salt
18:57 clearclaw joined #salt
18:57 gfa i didn't know orders, does it work with event.fire_master ?
18:57 dave_den gfa: http://docs.saltstack.com/ref/states/ordering.html
18:57 zdong i think this might be a bug for loading custom moudle from git repo, unless someone has validated this is working.
18:58 * jnials facepalms upon realizing you have to use reclass parameter substitution in reclass files instead of using pillar.get.
18:58 mephx joined #salt
18:58 n8n joined #salt
18:58 timoguin zdong, I can test it out later today
18:59 zdong timoguin, thank you. this will definitely help me. i have spent countless hours trying to search around google and salt documents.
18:59 gfa dave_den: i've tried with require and/or require_in but cp.get_file_str bails if the file it not there (parse time), but file gets created earlier run time
18:59 travisfischer joined #salt
19:01 travisfischer Is there a way to match in a top file on the absence of a grain key value. Something like "not G@mygrain:*" or "not G@mygrain"?
19:02 gfa dave_den: http://paste.debian.net/84217
19:02 krow joined #salt
19:03 ipmb joined #salt
19:03 nyx joined #salt
19:03 n8n joined #salt
19:03 leonardinius Anyone, how to automatically provision salt-master instance with gitfs dependecies on? (e.g. git, GitPython, gitdb)
19:04 leonardinius e.g. make_master: True, plus custom cloud script?
19:05 leonardinius How could one define/pass custom script used to bootstrap instance?
19:05 ipalreadytaken joined #salt
19:07 dave_den gfa: http://docs.saltstack.com/ref/renderers/all/salt.renderers.pydsl.html#render-time-state-execution
19:09 dave_den travisfischer: you have to include all minions first:  '* and not G@mygrain:blah'
19:09 ml_1 joined #salt
19:09 dave_den travisfischer: http://docs.saltstack.com/topics/targeting/compound.html
19:09 gfa dave_den: can i do it with jinja? or do i need to use pydsl?
19:10 leonardinius leonardinius, found it in the code. I could either use absolute path to the script OR define deploy_scripts_search_path configuration entry
19:10 dave_den gfa: you will have to use pydsl
19:10 gfa :(
19:10 gfa thanks
19:10 ipalreadytaken joined #salt
19:11 MrTango joined #salt
19:11 travisfischer dave_den: yah, I'm looking at those docs. I'm specifically wanting to target the minions that are lacking the presence of a grain key not a specific grain value as is shown the docs
19:11 travisfischer If 'mygrain' is a grain that is present on the machine I don't want this block to apply
19:13 pdayton joined #salt
19:14 gfa dave_den: can i use pyobject?
19:14 dave_den travisfischer: g@mygrain:* will work i think
19:14 dave_den G@mygrain:*
19:15 travisfischer testing now.  thanks for the confirmation
19:15 seapasulli joined #salt
19:15 ipalreadytaken joined #salt
19:15 TheRealBill joined #salt
19:15 cofeineSunshine joined #salt
19:15 dave_den gfa: probably. give it a test
19:20 hunter_ joined #salt
19:22 Corey Okay, how horrible was that FOSDEM talk on Salt?
19:22 Corey I haven't the nerve to watch it myself.
19:27 ipalreadytaken joined #salt
19:29 forrest Corey, haven't watched it yet
19:30 Corey Yeah, I need to get drunker first.
19:30 faulkner joined #salt
19:30 forrest Corey, Hmm, there is a puppet talk in your talk
19:30 Mr_N Lucky you.  I can't get drunker for another 3 hours; and by then the comments appearing in my state files will be REALLY self-hating.
19:31 Gifflen joined #salt
19:31 forrest Corey, since your talk was like 20 minutes
19:32 forrest the rest of the vid is a guy talking about puppet
19:32 Ahlee whiteinge: you around?
19:32 whiteinge Ahlee: yo!
19:33 Corey forrest: Ahh, okay.
19:33 Ahlee Got a sec to chat about the api and dev environments?
19:33 Corey forrest: the funny part is that video was uploaded by Saltstack.
19:33 Ahlee oh. god.
19:33 Ahlee i'm dumb.
19:33 Ahlee nevermind
19:33 Corey Ahlee: Hi.
19:33 NotreDev joined #salt
19:33 forrest Corey, yea
19:33 Ahlee my venv doesn't have salt, thus salt.log isn't available.
19:33 Ahlee Corey: Hey!
19:34 forrest whiteinge, you might want to let Tom know that the vid of Corey's talk from FOSDEM didn't get chopped properly and after the 18 minute mark or so it's all puppet
19:34 forrest whiteinge, or whoever edited the clip
19:34 cofeineSunshine joined #salt
19:34 Ryan_Lane1 :D
19:34 Ryan_Lane1 that's almost funny
19:35 Ryan_Lane1 want to know about salt? how about this puppet talk instead
19:35 whiteinge Ahlee: will you file an issue for that? we'll take a look.  ;-)
19:35 whiteinge forrest: thanks
19:35 * Ahlee cries
19:35 forrest whiteinge, yarg
19:36 Ahlee Salt issue #whatever, Ahlee's dumb.
19:36 mjulian joined #salt
19:36 mjulian joined #salt
19:37 hunter anyone here actively using reclass with salt?
19:38 fllr joined #salt
19:38 whiteinge forrest: trimmed!
19:39 ml_1 joined #salt
19:39 Katafalk_ joined #salt
19:39 whiteinge er. maybe
19:39 forrest whiteinge, sweet, looks like it is still showing up, might take some time to process.
19:39 MedicalJanePar-1 joined #salt
19:40 balboah_ joined #salt
19:40 zzzirk_ joined #salt
19:40 FarrisG_ joined #salt
19:41 jY- joined #salt
19:41 timoguin forrest, haha. i was skipping through the talk without audio and like, "why is there another guy talking about puppet now..."
19:41 schimmy1 joined #salt
19:41 whiteinge forrest: tom says thanks
19:41 ckao1 joined #salt
19:42 forrest timoguin, oh yea I did the same thing, I saw Corey's talk was short, and was like, why does this dude have a puppet logo in the corner
19:42 forrest whiteinge, yea of course, no free advertising, they gotta pay if they want to be in there! :P
19:42 valgrind joined #salt
19:42 jeremyfe_ joined #salt
19:43 think-free___ joined #salt
19:43 ndrei_ joined #salt
19:43 whiteinge i'd settle for letting us co-work out of their office. they have one slick cafeteria
19:43 hunter the puppet people?
19:43 whiteinge yeah
19:44 * timoguin imagines people falling down in cafeteria because the floor is slick
19:44 forrest whiteinge, hah, yea my buddy over at twitter was having a hard time, apparently they were out of dried fruit in their snack area for a few days
19:44 timoguin sounds horrible!
19:44 chrisjo__ joined #salt
19:44 Corey forrest: Did he walk out? I would have.
19:44 whiskybar1 joined #salt
19:44 forrest nah, he just complained about it
19:44 forrest jokingly
19:44 timoguin dried fruit ain't no joke
19:44 whiteinge ^^ this
19:45 forrest he eats there 3 times a day for free and has all the snacks he can handle, I'm pretty sure he can go a few days without some dried fruit
19:46 Mr_N It has been way too long since I've worked anywhere with free food...
19:46 krow joined #salt
19:46 whiteinge i almost invariably like puppetlabs employees. they seem to hire friendly people
19:46 whiteinge or maybe it's a portland thing
19:47 Mr_N I find devtool communities as a whole to be generally more friendly
19:47 dwfreed whiteinge: or both
19:47 dwfreed also hi :)
19:47 valgrind joined #salt
19:47 * whiteinge starts campaigning to move the Salt office to Portland
19:48 whiteinge dwfreed: hi! :D
19:48 Mr_N whiteinge: Apparently there's good coffee there :|
19:48 cro joined #salt
19:48 Mr_N (among far more legitimate reasons...)
19:48 whiteinge the beer is pretty solid too :)
19:48 jacksontj joined #salt
19:49 Mr_N See; far more legitimate reasons.
19:50 vlcn I need some guidance n the best way to accomplish something
19:50 Gareth whiteinge: just start slowing shipping furniture to Portland, eventually it will catch on :)
19:50 vlcn I have this sate: https://gist.github.com/kelchm/9229997 which works great from the master, but I want to be able to execute it on one specific minion from *all* other minions
19:50 whiteinge lol. devious. brilliant
19:51 Ahlee whiteinge: happen to remember off hand where salt-api looks up jobs?
19:51 whiteinge vlcn: you could use the Peer system so other minions can directly call that state. you could also use the Reactor and fire an event from the various minions to trigger that state run on the one minion
19:52 vlcn whiteinge, the issue is that I don't know of a way to do that without opening up *all* states
19:52 whiteinge Ahlee: it just wraps salt-run jobs.list_jobs, jobs.lookup_jid
19:52 Ahlee i was afraid you'd say that.
19:52 Ahlee Thanks.
19:52 vlcn reactor might indeed be the better way to do it so its just hardcoded to run on a single minion
19:52 whiteinge Ahlee: anything i cal help with?
19:53 whiteinge vlcn: yeah, you could hard-code the single minion in the reactor target. you could add validation there too to make sure the event came from an "approved" minion, etc
19:54 markm joined #salt
19:54 Ahlee whiteinge: eh, having issues with jobs runner calling out to ext cache
19:54 jrdx joined #salt
19:54 Ahlee not salt-api related.
19:54 Ahlee $ salt-run jobs.list_jobs
19:54 whiteinge ah, fun
19:54 Ahlee {}
19:54 Ahlee whee
19:54 Ahlee into the depths i die
19:54 Ahlee dive
19:55 juicer2 joined #salt
19:55 whiteinge Ahlee: did you ever get the ext job cache working "Correctly (TM)"?
19:55 Ahlee whiteinge: Nope.
19:55 davet joined #salt
19:55 Ahlee That's what I"m working on currently
19:55 l0x3py joined #salt
19:55 Ahlee so, more info coming.
19:55 TheRealBill joined #salt
19:56 Ahlee easy enough as it's all contained in salt/runners/jobs.py
19:56 lionel_ joined #salt
19:56 nineteeneightd joined #salt
19:57 racooper joined #salt
19:57 sroegner_ joined #salt
19:57 Nazca joined #salt
19:57 pdayton1 joined #salt
19:58 zain_ joined #salt
19:58 flebel joined #salt
19:58 che-arne joined #salt
19:58 jgelens joined #salt
19:58 jab416171 joined #salt
19:58 sabayonuser_ joined #salt
19:58 rostam joined #salt
19:58 borgstrom joined #salt
19:58 Xe joined #salt
19:58 Nazzy joined #salt
19:58 Chrisje joined #salt
19:58 Tinuva joined #salt
19:58 torrancew joined #salt
19:58 cb joined #salt
19:58 mfournier joined #salt
19:58 psyl0n joined #salt
19:58 vejdmn1 joined #salt
19:58 Kraln- joined #salt
19:59 cro_ joined #salt
19:59 Katafalkas joined #salt
19:59 jY joined #salt
19:59 pcarrier_ joined #salt
20:00 clearcla1 joined #salt
20:00 bryano joined #salt
20:00 forresta joined #salt
20:00 vbabiy_ joined #salt
20:01 kaptk2_ joined #salt
20:01 schristensen_ joined #salt
20:01 Vivi-Idl1 joined #salt
20:01 xzarth_ joined #salt
20:01 valgrind joined #salt
20:01 zooz_ joined #salt
20:01 mattmtl joined #salt
20:01 z3uS joined #salt
20:01 GradysGhost joined #salt
20:01 pass_by_value joined #salt
20:01 pcarrier_ joined #salt
20:02 rojem joined #salt
20:02 oznt joined #salt
20:02 ajw0100 joined #salt
20:02 troyready joined #salt
20:02 whiteinge Ahlee: did you ever have anyone verify/reproduce the situation you're seeing with ext cache?
20:03 whiskybar joined #salt
20:03 timoguin joined #salt
20:03 valgrind joined #salt
20:03 pmcg joined #salt
20:03 shadowsun joined #salt
20:03 aleszoul3k joined #salt
20:03 Ahlee whiteinge: No, I never posted it to the mailing list. I wanted to get this information gathered first
20:05 austin987 joined #salt
20:05 runnyspot joined #salt
20:05 Katafalkas joined #salt
20:05 cro joined #salt
20:05 che-arne joined #salt
20:05 jgelens joined #salt
20:05 jab416171 joined #salt
20:05 sabayonuser_ joined #salt
20:05 rostam joined #salt
20:05 borgstrom joined #salt
20:05 Xe joined #salt
20:05 Nazzy joined #salt
20:05 Chrisje joined #salt
20:05 Tinuva joined #salt
20:05 torrancew joined #salt
20:05 cb joined #salt
20:05 mfournier joined #salt
20:05 paramah joined #salt
20:06 MrTango joined #salt
20:07 sulky joined #salt
20:07 vbabiy joined #salt
20:07 flebel joined #salt
20:07 tharkun joined #salt
20:09 krow joined #salt
20:09 Networkn_ joined #salt
20:10 pmcg_ joined #salt
20:10 Mr_N Unknown yaml render error.  Thanks pillar.
20:11 Rusty_Shacklefor joined #salt
20:11 GradysGhost_ joined #salt
20:11 jacksontj_ joined #salt
20:11 ndrei joined #salt
20:11 _david_a joined #salt
20:11 Saint-N_ joined #salt
20:12 Ryan_Lane2 joined #salt
20:12 canci_ joined #salt
20:13 Ymage_ joined #salt
20:13 linjan_ joined #salt
20:13 FL1SK joined #salt
20:14 kvbik what would you check, if a command on one minion runs for 3 seconds and i you run it on all of them (65) it runs for about 2min? salt==0.17.4
20:14 kermit joined #salt
20:15 z3uS| joined #salt
20:16 mgw1 joined #salt
20:16 vlcn_ joined #salt
20:17 vlcn_ joined #salt
20:17 hunter_ joined #salt
20:18 hunter_ joined #salt
20:18 hunter joined #salt
20:20 sulky joined #salt
20:21 Rusty_Shacklefor joined #salt
20:23 ndrei joined #salt
20:23 krow joined #salt
20:24 srfwii joined #salt
20:25 yomilk joined #salt
20:25 vlcn joined #salt
20:27 TheRealBill joined #salt
20:28 vlcn for a reactor, how would I go about 'creating' a tag?
20:29 Gordonz joined #salt
20:29 vlcn I understand where salt/auth and minion_start are coming from -- but what if I wanted something actually unique?
20:30 Gordonz joined #salt
20:30 timoguin vlcn, you just specify the tag you want when you fire the event.
20:31 pydanny joined #salt
20:31 tessellare joined #salt
20:32 timoguin vlcn,
20:32 timoguin e.g., salt '*' event.fire '{'foo': 'banana'}' 'super/duper/tag'
20:33 vlcn what about from a minion, though?
20:34 vlcn I'm not seeing the event when I use salt-call event.fire '{"data": "message to be sent in the event"}' 'tag' from the minion
20:34 timoguin if you want to see it on the master you use event.fire_master
20:34 sroegner joined #salt
20:34 abe_music joined #salt
20:34 vlcn aha, there it is!  Thanks timoguin.
20:39 pydanny joined #salt
20:40 Linz joined #salt
20:41 yml joined #salt
20:42 pydanny joined #salt
20:43 johtso joined #salt
20:44 AdamSewell joined #salt
20:48 cewood joined #salt
20:49 ipalreadytaken joined #salt
20:49 Mr_N So I'm having some trouble querying nested items with a pillar.get properly
20:50 Mr_N I can do a roughly equivilent query that works fine on stuff under master; but when I try to do so on some pillars I defined (which appear correctly in pillar.items), I get back null results
20:50 ipalreadytaken joined #salt
20:54 cetex joined #salt
20:56 bhosmer joined #salt
20:57 rojem joined #salt
21:01 ajw0100 joined #salt
21:02 Mr_N http://pastebin.com/k65M21eL  If anyone cares to take a peek, I would welcome another pair of eyes.
21:03 Mr_N since pillar.items is magic, and can see all, but when I try to dig down manually, I can't get anywhere in the custom pillars.
21:03 druonysus joined #salt
21:03 druonysus joined #salt
21:03 bhosmer_ joined #salt
21:05 zzzirk joined #salt
21:06 mgw joined #salt
21:06 zzzirk joined #salt
21:08 darrend joined #salt
21:08 moos3 joined #salt
21:09 moos3 question in top.sls when you do like 'web*qa*' is there way to do 172.30* or 172.16* instead of hostname ?
21:09 hunter Is there a way to ask jinja "here is a string, please interpolate the variables in it" based on the current loop you are in?
21:09 krow joined #salt
21:11 Mr_N moos3: Yes; specify in standard fasion, e.g.  '128.220.233.0/24':
21:11 HeadAIX joined #salt
21:11 moos3 oh sweet
21:12 ipalreadytaken joined #salt
21:13 schristensen joined #salt
21:14 timoguin Mr_N, do you have to specify '- match: subnet' for that?
21:15 Mr_N timoguin: I don't believe so?  At least, I haven't had any problems thus far.
21:15 Mr_N Here.  I should probably have validated that this was working more rigorously and not just "everything seems ok", let me flip a bit and see if things still work.
21:16 dave4925_u joined #salt
21:17 kamal joined #salt
21:17 chrisjon_ joined #salt
21:17 timoguin just curious. i know with other matchers like compound and PCRE you're supposed to specify the match type in top.sls
21:17 patrek_ joined #salt
21:17 patrek_ on centos, mysql installs without a root password. I have a state that sets a value for the root password. But if I apply my state 2 twice, it will fail the second time because it now needs a password. Is there a check I can do to use connection_pass only if needed?
21:17 timoguin at least according to the docs
21:17 dcolish_ joined #salt
21:17 Valda joined #salt
21:18 \ask_ joined #salt
21:18 philipsd7 joined #salt
21:19 Mr_N timoguin: You're right, in fact.  The state I wanted was proliferating through another channel, so I didn't notice a problem.
21:20 robawt1 joined #salt
21:20 SaveTheR0btz joined #salt
21:20 Mr_N I've written 99% of this with docs sitting right next to me due to the vary obvious "me not knowing salt worth a damn"; so I'm curious as to how I missed that.
21:20 rizumu joined #salt
21:20 Ryan_Lane joined #salt
21:20 Gordonz joined #salt
21:21 Guest34564 joined #salt
21:21 Mr_N Sigh.  That's what I get for trying to do anything from memory.  Sorry for misleading you partially moos3 X.x
21:22 timoguin Mr_N, it happens. I don't know nada either. :)
21:22 moos3 lol
21:22 cnelsonsic joined #salt
21:22 moos3 the issue is I need to tell my VPC nodes in AWS apart since we dont know what their hostnames will be when they come up
21:23 hunter In answer to my own question (I did google it, I just didn't read far enough) jinja supports the "format" filter which should do what I want.
21:23 smcquay_ joined #salt
21:23 forresta Mr_N, if you felt that the docs confused/mislead you at some point, could you please create an issue?
21:23 zain joined #salt
21:24 codysoyland joined #salt
21:24 nebuchad` joined #salt
21:24 saurabhs1 joined #salt
21:24 jdenning_ joined #salt
21:25 pdayton joined #salt
21:25 bryano_ joined #salt
21:25 think-free joined #salt
21:26 saurabhs1 joined #salt
21:26 Mr_N forresta: I'm always nervous to criticize, since I worry that due to jumping around the docs (I think I only read through the core guide once top to bottom) I'm missing something as it would be intended.
21:27 Mr_N In this case, it seems mostly in that the top file documentation utilizes the match ipcidr stuff, but I don't believe ever stresses it in text or makes a note of it
21:27 teskew joined #salt
21:27 forresta Mr_N, Yea you should be creating issues for that stuff
21:27 Mr_N and I just wasn't accustomed to those pcre tags being necessary.
21:27 forresta so we can add more examples
21:27 Mr_N I think I expected too much magic
21:27 forresta gotcha
21:27 darrend_ joined #salt
21:27 forresta clarification is always good though
21:27 Mr_N Allright; will do so :)  I mostly didn't want to accidentally offend.
21:27 shadowsu1 joined #salt
21:27 Vivi-Idle joined #salt
21:27 Rojematic joined #salt
21:27 forresta Mr_N, worst case scenario is someone says 'no it is here on the page, or explained this way'
21:28 Ymage joined #salt
21:28 forresta people have said a lot more offensive stuff than 'these docs are unclear!'
21:28 Mr_N Hah.  Indeed.  I do need to be far better at contributing; I just have been trained a bit shy/twitchy in that respect, from past interactions ><
21:29 jp_cognet joined #salt
21:29 forresta yea don't worry about it, worst case scenario someone goes back and fixes stuff, or comments on the PR
21:29 Mr_N I had "It's never their problem, it's always yours" beaten into me rather aggressively back when I was a student.
21:29 hunter I like hesitating
21:29 Mr_N and I never really grew out of it.
21:29 Mr_N (in terms of dealing with software/libraries)
21:29 hunter Some <ahem> channels are rather rude about such. This one is friendlier than most.
21:30 jp_cognet anybody know why I'd be getting this error: "Error parsing configuration file: /etc/salt/master - while parsing a block mapping" when setting auto_accept to True? am I missing something or do it wrong?
21:30 Mr_N hunter: I have some horror stories from when I first touched #python and #cslounge...
21:31 elithrar joined #salt
21:31 jp_cognet nevermind, got it
21:32 xmltok im trying to use a cmd.run based on an onlyif, the onlyif command exits 1 but salt sees that as a failure so it doesnt execute my state -- is that correct? does a 'true' to onlyif mean a 0 exit code?
21:32 forresta Mr_N, even if you don't do PRs, the issues are still useful
21:33 jeremyBass1 joined #salt
21:33 jeremyBass1 left #salt
21:33 hunter #python can be awesome in its flaminess at times. Why we ever expect new people to pick up the language if thats how they respond is mystery to me.
21:33 hunter Mr_N: the puppet channel has its days for sure
21:33 hunter postgres is fairly friendly
21:33 hunter and I won't touch online photography forums, newsgroups, irc channels in any way
21:33 hunter Too many asshats.
21:33 hunter But I digres
21:33 ryepup joined #salt
21:33 ryepup good afternoon
21:34 vlcn joined #salt
21:34 ryepup I'm having a problem trying to run salt on windows
21:34 trevorj joined #salt
21:34 jpaetzel_ joined #salt
21:34 forresta ryepup, feel free to just toss your question in here, if someone knows, they'll respond.
21:34 ryepup I installed Salt-Minion-2012.1.0-AMD64, did the key exchance with my salt master
21:34 luminous_ joined #salt
21:34 ryepup and when I do state.highstate no files get copied to my windows machine
21:35 baniir joined #salt
21:35 forresta ryepup, can you test.ping the windows machine from the master?
21:35 zach__ joined #salt
21:35 jforest joined #salt
21:35 ryepup forresta: yes
21:35 druonysuse joined #salt
21:35 alienresidents joined #salt
21:35 ryepup when I do state.highstate, C:\salt\var\cache\salt\minion\files\base gets created, but it's empty
21:35 mephx joined #salt
21:35 forresta ryepup, can you try to run your highstate, but add -l debug
21:35 forresta that will provide debug output with more details
21:35 backjlack joined #salt
21:36 svx joined #salt
21:36 mr_chris joined #salt
21:36 rojem joined #salt
21:36 schristensen joined #salt
21:37 bhosmer joined #salt
21:37 baniir how are people managing rc.conf files in freebsd
21:37 cofeineSunshine joined #salt
21:37 ryepup forresta: log at http://pastebin.com/jT6MHXkX
21:38 ryepup nothing jumped out
21:38 ryepup just seems like salt isn't copying top.sls from my master to my local cache, then quitting cause there's no top.sls
21:39 timoguin ryepup, what happens if you initiate highstate from the master?
21:39 timoguin or are you masterless?
21:39 codysoyland joined #salt
21:39 ryepup we've got a master
21:39 MTecknology joined #salt
21:39 dave_den xmltok: 1 is false, 0 is true
21:39 forresta timoguin, http://pastebin.com/jT6MHXkX is what happens
21:39 robawt joined #salt
21:39 ryepup we use salt for linux hosts, I'm trying to get it for win hosts too
21:40 forresta ryepup, you installed it as the admin user right?
21:40 timoguin forresta, yea i just read through it, but that's using salt-call
21:40 forresta timoguin, yea good point
21:40 forresta ryepup, timoguin has a good point, can you try salt 'windows_minion' state.highstate -l debug from the master server?
21:41 timoguin the only time i've seen that error happen is when i was testing masterless minions in a different environment that didn't have a top.sls
21:41 xmltok dave_den: got it, maybe the exit code should be in the doc to remove confusion -- or maybe my cold is making me dumb today :)
21:41 forresta timoguin, gotcha
21:41 ryepup forresta: same result
21:42 ryepup putting together a paste now
21:42 forresta xmltok, it's just the standard linux return values
21:44 forresta xmltok, or exit codes, whatever you want to consider them
21:45 ryepup forresta: timoguin: salt 'windows' state.highstate > http://pastebin.com/NfSZT8gx
21:45 bhosmer joined #salt
21:45 ajw0100 joined #salt
21:46 timoguin ryepup, are you using multiple environments, or just a single base environment?
21:46 ryepup single base environment
21:46 vlcn https://gist.github.com/kelchm/9239311
21:46 vlcn any idea what I'm missing here?
21:46 ryepup in top.sls we have a 'os:Windows' grain match that should be applying
21:46 ryepup but it seems like top.sls just never makes it down to the windows minion
21:47 blast_hardcheese joined #salt
21:48 timoguin ryepup, weird. maybe a bug. can you run single states?
21:49 ryepup timoguin: no
21:49 ryepup timoguin: get "No matching sls found" for ones that should apply
21:49 Mr_N hunter: Got dragged off to be educated on why I should "always never not be using FFT"
21:49 Mr_N but yes, #python is "special"
21:49 Mr_N Which is strange, since as a _language_ it has always seemed quite friendly...
21:50 timoguin smaller communities always tend to be better
21:50 timoguin more bigger = more troller
21:50 hunter mr_N: my understanding is the mailling lists are friendlier - but I'm not on them myself.
21:50 fauxtoniche joined #salt
21:50 hunter timoguin: thats true to a certain extent, but some are far more hateful than their size should imply.
21:51 timoguin indeed
21:51 hunter #photography on this irc net is one such - or at least was about 4 years ago
21:51 miguitas joined #salt
21:52 Mr_N Oh by the way forresta ; since the work day is almost over it'll have to wait till tomorrow, but I can certainly push an issue on my thoughts from doing a "dive into salt"
21:52 ndrei joined #salt
21:52 Mr_N I mention since another one came to mind; the pillar docs seem to imply that a .items does a refresh; and while it does, the refresh does not "persist" outside of the items call
21:53 Mr_N which might be a worthy distinction to make
21:53 ajw0100 joined #salt
21:53 timoguin ryepup, i'll have to play with it. not sure why it's not working for you.
21:53 ryepup timoguin: any more data I can gather?
21:53 tr_h joined #salt
21:54 ryepup before I make a github issue?
21:54 Mr_N Or, perhaps I've misunderstood it, since despite pillar.raw and pillar.items returning different things; saltutil.refresh_pillar doesn't cause them to converge
21:54 timoguin ryepup, you sure your windows minion isn't specifying an environment or anything?
21:56 baniir is it possible to create a file if it doesn't exist with file.append
21:56 timoguin baniir, what happens if you do file.append for a non-existent file?
21:56 timoguin i've done file.touch before file.append before
21:57 luminous joined #salt
21:57 baniir timoguin: file not found
21:57 ryepup timoguin: the error I get is ` No matching sls found for 'ServerManagement.windows' in env 'base'`, so I think it's using env 'base'
21:57 timoguin ah yea, seems like it.
21:58 ryepup going to try reinstalling and using some older windows binaries
21:58 timoguin gl
21:58 Mr_N Any reason why refresh_pillar wouldn't refresh pillars?
21:58 timoguin baniir, yea i guess you can't then. file.touch first will do the trick
22:00 Mr_N sync_all doesn't do it, refresh doesn't do it.
22:00 Mr_N Fuck it. I'm going home to drink.  I'll nag you all about this tomorrow ><
22:01 timoguin Mr_N, sounds like a good plan!
22:01 ndrei joined #salt
22:01 forresta Mr_N, sounds good!
22:02 Mr_N have a good evening all.
22:02 forresta later
22:06 yomilk joined #salt
22:07 eliasp I have the feeling, something's not right with my states… is there a way to locally validate them/syntax-check them on the master? running 0.17.5…
22:08 srfwii joined #salt
22:08 krow joined #salt
22:08 eliasp on the minion, I just run into this exception: http://dpaste.com/1663432/
22:09 renoirb Hey all.   Question about Jinja.  If i follow documentation here [1]  i should be able to use filters. I'd like to use the int() filter on the result of a multiplication.  Is it possible?
22:09 renoirb [1]: http://jinja.pocoo.org/docs/templates/#list-of-builtin-filters
22:09 renoirb Something like {% set max_clients = salt['grains.get']('mem_total', '1024') / 1024 * 150 %}
22:10 hunter joined #salt
22:10 kainswor @Mr_N I experienced similar issues
22:10 eliasp hmm, this exception is caused by something else, not by invoking "state.sls env=dev"
22:10 renoirb but when i surround the expression after "="  with int(...)   state fails saying it doesn't find variable with name 'int'.
22:12 n8n joined #salt
22:13 ryepup using Salt-Minion-0.17.5-2-AMD64-Setup.exe worked for me, so there's some regression with the latest windows salt minion installer.
22:14 elfixit joined #salt
22:18 ipmb joined #salt
22:20 clearclaw joined #salt
22:22 Ryan_Lane joined #salt
22:22 johtso joined #salt
22:22 allanparsons joined #salt
22:23 allanparsons hello - i was reading salt docs and was a little unclear on using GitFS + Multiple branches
22:23 allanparsons how does salt know what branch to read?
22:24 allanparsons "As stated above, when using the gitfs backend, branches will be mapped to environments using the branch name as identifier"
22:24 allanparsons where is the branchname defined in the /etc/salt/master config?
22:26 fauxtoniche joined #salt
22:28 smcquay joined #salt
22:29 harobed joined #salt
22:30 kula allanparsons: it isn't defined in the /etc/salt/master config. in the master config you point it at a git repo. in your top.sls file, an environment 'foo' will map to a git tree-ish thing (branch, tag) called 'foo'.
22:30 kula the special case is 'base' maps to 'master'
22:30 kula (that is, the 'base' environment maps to the 'master' branch in git)
22:30 harobed joined #salt
22:31 allanparsons ahhhh
22:31 allanparsons gotcha
22:31 allanparsons ok
22:31 kula this allows you to create arbitrary environments simply by creating a tag or branch, which i'm rather fond of.
22:32 allanparsons that makes sense
22:32 allanparsons though i think it'd be nicer if salt master had a config to force a branch
22:32 tr_h joined #salt
22:32 allanparsons IMO
22:33 pydanny joined #salt
22:34 kula i like the arbitrary bits of it. i can point my test salt-master at the same git repo, create an environment to do some work in simply by creating a branch, much around with new states, and then merge them into production or whatnot.
22:35 kula but i'll freely admit that maps to how my brain works :-)
22:35 allanparsons :)
22:35 allanparsons that makes sense
22:35 allanparsons it just isn't too clear in the /etc/salt/master config comments
22:35 kula i'd grant that
22:37 kula i've also lately been using the git state and git repos as a hacked-up half-arsed package manager
22:42 faldridge joined #salt
22:43 akoumjian joined #salt
22:45 ajw0100 joined #salt
22:50 alekibango joined #salt
22:51 scalability-junk joined #salt
22:52 lahwran joined #salt
22:56 tr_h joined #salt
22:59 elfixit joined #salt
23:00 kermit joined #salt
23:00 Ryan_Lane joined #salt
23:02 rojem joined #salt
23:07 kermit joined #salt
23:08 krow joined #salt
23:08 glen_ joined #salt
23:10 chutzpah is libcloud a _hard_ requirement for 2014.1.0?
23:19 bhosmer joined #salt
23:21 aleszoul3k joined #salt
23:23 seanz joined #salt
23:24 MindDrive_ left #salt
23:24 MindDrive_ joined #salt
23:24 krow joined #salt
23:24 seanz whiteinge: It was good to see you at SCaLE.
23:24 seanz terminalmage: It was good meeting you at SCaLE.
23:26 whiteinge seanz: you too!
23:26 seanz whiteinge: Is PyCon in your future?
23:27 akoumjian joined #salt
23:28 whiteinge likely not
23:28 seanz Oh man!
23:29 terminalmage seanz: I'm afraid I don't recognize your nick
23:29 akl joined #salt
23:29 seanz terminalmage: I'm one of the hundreds of people you talked to at SCaLE. :)
23:29 terminalmage Haha
23:29 seanz Really nobody of consequence. We had a conversation about salt's reactor system.
23:30 seanz I was working at the python booth.
23:30 terminalmage Yeah. Well, with few exceptions, everyone at SCaLE was awesome
23:30 seanz I may well be one of those exceptions.
23:31 terminalmage Ahh! That does narrow it down
23:31 seanz haha
23:31 terminalmage Think I know who you are now
23:32 Xe- joined #salt
23:32 terminalmage Nah, there were just a couple people that were really obtuse
23:33 terminalmage Asking vague questions and looking at me like I'm an idiot when I try to answer the question
23:34 seanz Fortunately, that doesn't sound like our conversation.
23:34 terminalmage "that's not what I meant"
23:34 terminalmage Well here's a tip...  SAY WHAT YOU MEAN
23:35 Corey Yeah, when I've given Salt talks before most of the questions are brilliant.
23:35 Corey But there's always ONE GUY...
23:35 Corey And it's not that the question is dumb, it's that the guy is taking personal offense over something. "Well why not do it my way instead?!"
23:37 terminalmage Corey;
23:37 Corey That's me!
23:37 terminalmage Dammit
23:38 terminalmage Typing on phone
23:39 davet joined #salt
23:40 terminalmage Corey: another good one is: "my particular use case is (some Rube Goldberg monstrosity) ", why should I use salt?
23:40 terminalmage Then you explain how salt can do that without a lot of the intermediate steps
23:40 Corey terminalmage: "...you shouldn't, because people will judge Salt by your crazy workflow."
23:41 terminalmage And instead of being happy, they get offended that they'd need to change their workflow
23:43 torrancew You mean I can't have a massive improvement over my process without changing my process?!? WHAT IS THIS MADNESS?!?
23:43 druonysus joined #salt
23:43 druonysus joined #salt
23:46 Corey terminalmage: "So I take one of Eucalyptus's cloud-in-a-backpack rigs and I strap it to my golden retriever, right? Next, I ..." <-- How deployments happen at these shops.
23:46 terminalmage LOL
23:47 allanparsons in salt is there a way that i can auto-accept keys from a particular subnet?
23:47 allanparsons i know i can do that in puppet
23:47 allanparsons and it'd be insanely useful in an Amazon VPC
23:47 allanparsons with auto-scaling groups.
23:47 leonardinius allanparsons, yes
23:47 leonardinius auto-accept: True I believe
23:47 Corey leonardinius: Not the question that was asked.
23:47 allanparsons yes
23:47 allanparsons i want to lock it down by subnet
23:48 allanparsons auto-accept poses too large a security risk
23:48 Corey allanparsons: Not as such. :-) YOu can autoaccept or not, but more granularity than that you're probably pushing it off into iptables.
23:48 allanparsons @Corey ok.
23:48 allanparsons @Corey I guess i can just lock it down by AWS Security Groups
23:49 platforms joined #salt
23:49 terminalmage Getting dinner, back later
23:49 leonardinius allanparsons, I'm not sure what your requirements and expectactions are, but please take a look into http://docs.saltstack.com/ref/configuration/master.html. It mentions auto_accept parameter (will auto-accept minion keys)
23:50 leonardinius Enyone - is out there some decent documentation /example on salt cloud AWS/EC2 VPC with network_interfaces and private ip addresses
23:50 leonardinius could not find any example in salt docs AT ALL
23:53 allanparsons @leonardinius  i want to auto-accept, but i want to restrict it to hosts of a particular subnet (say, 10.0.0.0/16)
23:56 leonardinius allanparsons, I see. One way to go is to still use auto_accept, just make salt-master listen on internal/VPC interface, other than 0.0.0.0
23:57 ashb joined #salt
23:57 allanparsons leonardinius right, but that address gets translated.
23:57 leonardinius allanparsons, or follow Corey advice ;)
23:57 allanparsons ec2 EIP gets NAT'd to an internal interface
23:57 ashb does the environment show up somewhere in the grains or pillar data?
23:57 allanparsons yeah, i'll end up doing that.
23:58 allanparsons i'm just saying, that was one feature i liked in puppet - the autosign.conf file
23:58 allanparsons i could add a subnet or hostnames or wildcards
23:59 allanparsons what i ended up doing w/ that was polling the EC2 API every 60 seconds for hosts and then dumping to that autosign.conf file.
23:59 allanparsons but even 60s is too long to wait for a host to come online (by that time, an autoscale event in my industry is almost over)

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