Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-11-14

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

All times shown according to UTC.

Time Nick Message
00:00 pears the minion by default will connect to "salt" on port 4505
00:00 honestly either works - you could also put the ip in /etc/hosts with the "salt" hostname
00:00 jacksontj joined #salt
00:00 pears so you can either set that in your hosts file or DNS, or tell the minion to connect somewhere else in its config file
00:01 jmpf joined #salt
00:03 jmpf getting a weird error in my logs (salt master isnt working)  -- http://pastie.org/8478820
00:03 jmpf but df is showing plenty of space
00:03 pears did you check inodes too?
00:04 jmpf now - the dir. it's complaining about -- def. has quite a few directories in it
00:04 pears can you make a directory manually in that location?
00:04 jmpf pears: no I haven't
00:04 jmpf lemme chk
00:05 jmpf hrm... actually some of these directories don't exist anymore
00:08 jmpf can I just blow out these directories for now? kill the keys and let them re-sync?
00:09 jslatts joined #salt
00:11 Guest48444 @pears you meant port 4506, yes?
00:12 Guest48444 is there a command I can telnet into the server to test connectivity?
00:12 pears 4505 is the main connection, 4506 is used for job data I think
00:13 Guest48444 when I telnet, it just says Trying xxx ...
00:13 pears if you do "telnet salt 4505" that will more or less do what the minion does by default
00:13 pears if that hangs you have connectivity problems (firewall perhaps)
00:13 Guest48444 doe
00:14 jdenning joined #salt
00:18 jdenning UtahDave - just got the thumbs up from my bosses to attend SaltConf in January - I know we missed the deadline for speaker proposals, but was wondering if you knew if they still had any spots to fill or if there were already plenty of proposals?  I'd be happy to give a talk on our git workflow or our custom event-driven stuff. (I run DevOps at FlightStats, Inc.)
00:18 UtahDave jdenning: Cool, that will be awesome to have you here!
00:19 honestly UtahDave: I maded you a PR, did you see it?
00:22 cocoy joined #salt
00:23 UtahDave which one, honestly?
00:24 honestly https://github.com/saltstack/salt/pull/8496
00:24 jmpf yeh - it was totally inode cnt
00:24 honestly just a simply docs update for something that bit me in the ass
00:24 jmpf so -- the jobs directory was freaking huge -- is there some sort of cleanup idiom that we should be doing ?
00:24 honestly simple*
00:25 UtahDave cool, honestly. Thanks!
00:25 UtahDave jmpf: how many nodes and how many jobs?  Salt cleans that out every 24 hours
00:25 UtahDave by default
00:26 jmpf UtahDave: after I decided to just nuke it was @ /dev/xvda1                         9830400 9502380    328020   97% /
00:26 jmpf it's taking a while to delete
00:26 jmpf hrm
00:26 jmpf ok
00:26 UtahDave honestly: hey would mind line wrapping that comment in your pull req?
00:26 UtahDave how many nodes are you running, jmpf?
00:27 jmpf UtahDave: minions??
00:28 UtahDave how many servers are you managing with Salt?
00:28 UtahDave Usually people have job cache issues when they have a large node count and also are running jobs on a cron every 30 minutes or something
00:29 jmpf there's prob. north of 70 right now
00:29 jmpf I took down the master while I'm rming this
00:29 honestly UtahDave: Did that... let's see what happens...
00:29 UtahDave did you push it?
00:29 UtahDave ah, just appeared
00:30 honestly UtahDave: I merged in the develop branch after doing that commit, so might have to rebase and move the commit around a bit
00:30 honestly hmm, looks fine
00:30 UtahDave yeah, it's usually better to just rebase
00:31 honestly let's see
00:34 honestly rebase successful, but let me delete the PR
00:34 honestly and remake it
00:34 jslatts joined #salt
00:34 honestly oh, it just worked
00:35 honestly awesome
00:35 honestly one commit, one change, everything good (:
00:35 * honestly is starting to become somewhat competent at git
00:36 UtahDave :)  it takes a while.
00:40 Linz joined #salt
00:41 Guest48444 I had to open ports in aws - thanks all
00:42 honestly UtahDave: PR is fine now?
00:42 UtahDave Yeah, I think it looks great, honestly
00:42 honestly (:
00:42 honestly bedtime!
00:44 UtahDave thanks for the help, honestly!
00:44 honestly happy to help (:
00:47 halfss joined #salt
00:47 anti_ joined #salt
00:48 zandy joined #salt
00:48 anti_ joined #salt
00:49 diegows joined #salt
00:50 diegows is there a way to create a dependency between minions?
00:50 diegows I could do it with conditionals, recipes and mine system, but may be there is something in salt to do it
00:51 UtahDave diegows: have you looked at the overstate?
00:51 xmltok joined #salt
00:51 diegows let me see
00:52 hvn joined #salt
00:52 diegows UtahDave, looks like what I need... thanks!
00:52 UtahDave you're welcome!
00:53 lynxman joined #salt
00:54 pipps_ joined #salt
01:05 steveoliver how can i find all the salt modules available?
01:05 UtahDave salt '<minion>' sys.doc
01:05 alunduil joined #salt
01:06 steveoliver ok, i'm on 0.17.1, and see no tls.
01:06 UtahDave steveoliver: you have to make sure your system has all the deps required for the module
01:06 steveoliver ok, thx
01:07 UtahDave so for the tls module you have to have pyopenssl installed, then restart the salt-minion service
01:07 steveoliver ah - restart minion, too
01:07 steveoliver that did it.  thanks.
01:08 UtahDave you're welcome!
01:09 zandy joined #salt
01:10 NotreDev joined #salt
01:14 isomorphic joined #salt
01:18 halfss__ joined #salt
01:19 bhosmer joined #salt
01:19 NotreDev didn't it use to be the case that you couldn't have two keys on a yaml dict that had methods attached? e.g. {'www-data': {'user.present': [{...}],[{...}]}, 'group.present': {...} }}
01:20 NotreDev as in, i'm surprised this works: http://pastebin.com/sGuKqQFX
01:20 UtahDave NotreDev: that has always worked.
01:20 ajw0100 joined #salt
01:21 NotreDev UtahDave: do you have any idea what i might have been thinking of?
01:21 UtahDave you can't have 2 instances that use the same state function
01:21 NotreDev :)
01:21 NotreDev ah that's it
01:21 UtahDave like user.present and user.blabla
01:22 NotreDev on a lot of these states, i use `- names:`, but i can't find it in the docs. perhaps this was from older days? is this still supported?
01:23 Gifflen joined #salt
01:23 Furao left #salt
01:26 jacksontj joined #salt
01:28 * diegows has a talk scheduled for tomorrow, not enough time to test overstate ... but.. just ... a try :)
01:30 zandy joined #salt
01:37 NotreDev i have a pillar of users that includes their groups. if i want to invert this data (list the groups with users), i can do so easily in python. I can't do so very easily in jinja2. how should i go about this?
01:38 lesnail joined #salt
01:52 gkze joined #salt
02:00 fragamus joined #salt
02:06 fragamus_ joined #salt
02:07 redondos_ joined #salt
02:10 Ash__ joined #salt
02:15 pentabular joined #salt
02:16 juicer2 joined #salt
02:18 oz_akan_ joined #salt
02:22 junedm joined #salt
02:22 pentabular left #salt
02:23 anti joined #salt
02:23 gkze joined #salt
02:24 anti joined #salt
02:25 bhosmer joined #salt
02:26 franky joined #salt
02:26 ajw0100 joined #salt
02:36 anitak joined #salt
02:38 pentabular1 joined #salt
02:38 hvn left #salt
02:42 xl1 joined #salt
02:48 pentabular1 left #salt
02:48 Ryan_Lane joined #salt
02:53 AdamSewell joined #salt
02:54 AdamSewell in sly files, what is the proper spacing? is it all spaces or will tabs work as well?
02:55 AdamSewell sls*
02:57 jalbretsen joined #salt
02:59 bhosmer joined #salt
03:09 pipps_ joined #salt
03:12 gkze joined #salt
03:16 joehh AdamSewell: I believe it is spaces
03:17 AdamSewell joehh, yea i found that. i seem to be having some issues with a sls file even though
03:17 AdamSewell it is with spaces
03:19 cnelsonsic joined #salt
03:19 AdamSewell joehh, see http://pastie.org/8479105
03:22 ajw0100 joined #salt
03:26 joehh I've got http://pastie.org/8479109 as an example
03:26 joehh looks like an extra space and I've used names rather thanpkgs
03:27 joehh AdamSewell: ^^
03:28 AdamSewell joehh, extra space where?
03:28 AdamSewell also, does it have to lead off with packages:
03:28 AdamSewell ?
03:29 joehh no, i had copmanyname-packages:
03:29 joehh your - php5 linus up with the space inbetween - and pkgs
03:29 jimallman joined #salt
03:30 joehh my - is indented one more space
03:30 joehh try that
03:30 joehh also my - for names is under the g rahter than the k
03:31 AdamSewell ok, odd
03:31 AdamSewell i got it
03:31 AdamSewell had to add extra spaces on each line
03:32 joehh I often use a variant on this bit of code to help track down dodgy sls files
03:32 joehh http://pastie.org/private/yba2njnlkq8hpbfnkazila
03:32 joehh yeah - it can be finicky like that
03:33 AdamSewell thanks for sharing,
03:36 mapu joined #salt
03:38 Baribal_ joined #salt
03:43 thelorax123 joined #salt
03:45 joehh no worries
03:51 steveoliver http://paste2.org/eHaZ9Ph8  <— Why is the second {{ sitename }} instance being marked, and not the first?
03:52 Ryan_Lane joined #salt
03:59 mwillhite joined #salt
04:11 vipul joined #salt
04:11 jkleckner joined #salt
04:12 steveoliver http://paste2.org/wKjyNMhX <— Seems it's the last occurrence of the var that has issues.
04:14 NotreDev joined #salt
04:14 NotreDev what exactly does salt.states.iptables.set_policy do?
04:16 forrest NotreDev, hmm you know I'm not really sure, let's see what the code says
04:16 forrest how did that thing go for your users earlier NotreDev?
04:16 NotreDev yeah, could do that. i read the manpages for iptables, and had no inkling
04:16 NotreDev forrest excellent actaully
04:16 NotreDev *actually
04:16 forrest what did you end up doing?
04:17 NotreDev so i have states/access/core.sls which are core users that should be on every machine
04:17 forrest ohhh I see what the policy is for, it must be for the DEFAULT policy for each chain
04:17 NotreDev in addition i have states/access/www-data.sls which is included by my states/nginx/init.sls, etc
04:17 forrest I'll make an issue on that to get it updated.
04:18 forrest ok
04:18 NotreDev so adding users is conditional on them being required.
04:18 NotreDev the big test was whether an sls file would be discoverable if it wasn't automatically included in the init.sls file
04:19 NotreDev breaking up init.sls files is very helpful for building out state files!
04:19 forrest so I guess I'm confused
04:19 forrest how did you add the extra users?
04:19 forrest depending on the type of machine?
04:20 forrest you added them to an sls and then require that sls?
04:22 forrest Is the Chris Jones that made this commit around? https://github.com/saltstack/salt/commit/3756f12a10f022fb1f72c6a502be36bf23488bca#diff-617e242e036e50e1bc111067953b04b9
04:22 forrest his user no longer seems to be active on github, so if you're in here, can you please respond?
04:25 ravibhure joined #salt
04:27 vu joined #salt
04:29 forrest ok NotreDev I put in a pull request to update the docs to explain its' for the default policy for the chain in iptables.
04:29 forrest it's
04:29 ravibhure joined #salt
04:29 NotreDev forrest: sorry, i didn't make my users dependent on the role of the machine, but rather i'm matching for role in the state's top.sls file. a particular state can then require a user. so i've gotten around my error
04:30 forrest ahhh ok that makes more sense
04:31 NotreDev when i try and execute the iptables.append command it fails. here is what test=True prints. a couple of these options look funny. http://pastebin.com/hPXC8bF5
04:31 NotreDev can't figure out where the system would log that error
04:31 forrest what does your state look like?
04:32 forrest and are the rules already on the server you're applying this to?
04:33 NotreDev forrest, fresh set of rules. i'll paste the state
04:33 forrest ok
04:33 forrest Are you in another part of the world? Or are you just working on this at home
04:34 forrest or somewhere outside the states I should say
04:34 NotreDev forrest: austin texas :) http://pastebin.com/SUjTHz59
04:34 forrest rackspace?
04:34 forrest I can't remember
04:34 forrest from earlier
04:34 NotreDev haha, no. but their recruiter called me again today
04:34 forrest hah
04:35 forrest They dodged my calls after I told them I didn't want to move to Texas :P
04:35 NotreDev haha
04:36 forrest ok
04:36 forrest so when this runs, it didn't apply anything on the server you're running this on?
04:37 forrest so nginx is installed, the nginx.jinja2 file is there, and the service is running?
04:37 NotreDev eyp
04:38 NotreDev yep; the only thing that doesn't execute is the iptables addition
04:38 forrest can you try to modify the existing - connstat:
04:38 forrest and make it JUST NEW
04:38 forrest and then add another line with a second one for ESTABLISHED?
04:38 forrest because the example only uses a single item for the connstat
04:38 forrest other than that, your state looks identical.
04:39 NotreDev fails the same.
04:40 forrest what about with just a single connstat?
04:40 forrest so - connstat: NEW
04:41 NotreDev yep on connstat. is `--__env__ base` actually being used as an argument? dunno how to strace inside salt :)
04:41 forrest I honsetly have no idea, I doubt it. Seems to me like that's more for tracking purposes saying which environment and state file it came from
04:41 forrest it still failed on a single connstat?
04:41 forrest Are you running this with debug?
04:43 NotreDev debug?
04:43 NotreDev http://docs.saltstack.com/topics/troubleshooting/index.html
04:43 NotreDev heh
04:44 forrest just run your state again with -l debug
04:49 NotreDev interesting: http://pastebin.com/vzU0jyWi
04:50 NotreDev cmd.run might be smarter here
04:52 forrest that's ..... weird
04:52 MTecknology forrest: care if I pm you?
04:52 forrest basepi, if you're around can you review this real quick: https://github.com/gravyboat/salt/commit/139e3b3bb9b6085701f2aa411f63b4a87caefa62 I don't want to make assumptions.
04:53 forrest MTecknology, you can't PM me. I've tried pm'ing you before and it fails because I'm not registered.
04:53 MTecknology y u no reg?
04:53 forrest because some guy that only logs in once every 2 months has my user
04:53 forrest I'm waiting for it to expire
04:53 forrest I almost had him last time
04:53 forrest but he logged in right at 9 weeks
04:54 MTecknology k... I'm not -R now
04:54 forrest yea go for it then
04:59 AdamSewell Could anyone explain why I'm getting this error "State file.mkdir found in sls nginx is unavailable" when running: http://pastie.org/8479244
05:00 anuvrat joined #salt
05:02 forrest mkdir doesn't exist?
05:02 forrest in file
05:02 forrest you want file.directory
05:02 AdamSewell forrest, i'm trying to create the directory
05:02 forrest yea
05:02 forrest use file.directory
05:02 AdamSewell wouldn't mkdir be what i want?
05:02 AdamSewell oh
05:03 forrest http://docs.saltstack.com/ref/states/all/salt.states.file.html#salt.states.file.directory
05:03 AdamSewell so what is file.mkdir for then?
05:03 forrest I don't know, I don't see file.mkdir in the docs
05:03 AdamSewell http://docs.saltstack.com/ref/modules/all/salt.modules.file.html
05:03 forrest ok that's the modules AdamSewell
05:03 AdamSewell oh
05:03 AdamSewell oops
05:03 forrest modules and states aren't the same thing
05:04 forrest it happens all the time :P
05:04 AdamSewell i'm learning!
05:05 forrest It's like you just started using Salt or something AdamSewell, JEEZ!
05:05 AdamSewell yea, kinda am
05:06 troyready joined #salt
05:06 forrest I'm joking with you
05:06 AdamSewell forrest, i appreciate all the help
05:06 forrest if that wasn't clearly conveyed
05:06 AdamSewell i know :)
05:06 forrest ok good, sometimes I forget text is bad at that stuff
05:06 forrest and np man
05:08 reinsle joined #salt
05:08 AdamSewell forrest, can salt create mysql databases on remote servers that aren't minions?
05:09 AdamSewell i'm guessing it'd have to be a minion thouggh
05:09 forrest yea it's gonna need to be a minion
05:10 AdamSewell hem, alright
05:17 dvogt joined #salt
05:23 UtahDave joined #salt
05:29 forrest Hey UtahDave, if you have a minute can you review this? https://github.com/gravyboat/salt/commit/139e3b3bb9b6085701f2aa411f63b4a87caefa62 I don't want to put words in the team's mouth
05:30 UtahDave sure!
05:30 forrest thanks
05:33 UtahDave So has anyone said that backwards compatibility isn't going to happen in the future?
05:33 forrest no
05:33 forrest that's why I put in there that you guys were gonna continue to try for it
05:33 forrest it's just an issue with 0.17
05:34 UtahDave Yeah, we're still going to work to ensure backwards compat, but 0.17.1 was a special case because of the security issue
05:34 forrest right
05:34 forrest That's why I wanted to add it to the FAQ that it's specifically that version
05:35 UtahDave so i wouldn't say you can no longer run different versions of Salt, just that there's a break right there where you can't
05:39 forrest UtahDave, I think this is a bit better
05:39 forrest https://github.com/gravyboat/salt/commit/1b9849049a34c47129efa7526850c9c23ee08089
05:39 NotreDev joined #salt
05:40 UtahDave forrest: yeah, that's definitely an improvement. I'd like to have Tom chime in on it too
05:41 forrest Yea sounds good, I made a note in the actual pull request asking that someone doesn't just merge it and it gets reviewed
05:41 forrest I'm no legal guy, I just know I'll forget what the reason is a month down the road and want it documented somewhere I remember to look :P
05:42 UtahDave for sure!
05:43 * Gareth sends in another pull request
05:43 forrest I added a note to the pull request to summon Tom in
05:45 UtahDave Gareth: thanks! that gnomedesktop pull req was cool
05:46 Bosse joined #salt
05:46 Gareth UtahDave: this one is some fixes to that.  Found some issues with it immediately after Tom merged it :)  had to rework how it was working.
05:47 UtahDave oh, yeah, Gareth. I was going to ask. Is there a __virtual__ function in it?
05:47 Gareth there is.
05:47 UtahDave cool
05:49 dcmorton joined #salt
05:49 druonysus joined #salt
05:49 druonysus joined #salt
05:51 UtahDave Gareth: can you add  the    import pwd  to the top try/except block?
05:52 UtahDave the pwd module doesn't exist on Windows
05:52 UtahDave that will cause a stacktrace on windows
05:52 forrest psssh
05:52 forrest windows
05:52 forrest :P
05:53 AdamSewell what's the correct statement to  restart a service after a config file change? refresh: true?
05:53 UtahDave AdamSewell: no
05:54 UtahDave AdamSewell: you'll want to have a service.running section for that service that has a watch:\n  - file: theconfig
05:55 AdamSewell UtahDave, this is what i have now. http://pastie.org/8479294
05:55 Gareth UtahDave: ahh good call. lemme fix that.
05:56 AdamSewell so, i need to move the nginx.conf under the service.running statement?
05:56 UtahDave let me modify that for you
05:56 Ryan_Lane joined #salt
05:57 UtahDave http://pastie.org/8479298
05:58 AdamSewell oh ok, i see
06:00 AdamSewell this will only reload if the file changed correct?
06:00 UtahDave yep
06:00 IJNX joined #salt
06:00 AdamSewell excellnt
06:01 UtahDave I added both those config files there for completeness, but you don't have to have both if you don't want them
06:01 AdamSewell excellent*
06:01 AdamSewell yea, i want both :)
06:01 AdamSewell a question though
06:01 UtahDave sure
06:01 Gareth UtahDave: done.
06:01 UtahDave thanks, Gareth!
06:02 AdamSewell is it possible to have two different nginx.conf files say one for ssl and one for non-ssl sites and then some how choose which one to send?
06:02 Gareth UtahDave: no worries :)
06:02 Gareth bbl.
06:02 AdamSewell i'm not sure how i'd determine that though to be honest
06:03 UtahDave AdamSewell: yeah, there are a few ways to do that.
06:03 UtahDave AdamSewell: you could use jinja to implement your logic and use jinja variables for the source file
06:03 AdamSewell UtahDave, that's what I did for the server name
06:04 UtahDave AdamSewell: you could also have multiple implementations and have your matching in your top.sls determine which gets applied
06:04 AdamSewell the top file would be based on id though right?
06:06 AdamSewell actually, i know how i could determine this
06:07 AdamSewell is it possible to store the profile that was used to create a server (like from a cloud provider)?
06:07 UtahDave the top file matches on id by default, but you can also match on grains, pillars and other things
06:07 AdamSewell for example, cloud-salt --profile tier_1 new domain.com
06:07 AdamSewell i'd like to keep that tier_1 piece
06:07 UtahDave you could stick that in the minion's grains
06:08 UtahDave in fact, you could put that in your salt-cloud profile so it happened automatically
06:08 AdamSewell how would i do that?
06:12 UtahDave let me find an example
06:14 AdamSewell ok, thanks UtahDave
06:14 UtahDave try something like this: http://pastebin.com/xasaDbM3
06:14 UtahDave I added two grains there
06:20 AdamSewell so now, in my nginx init sls, i should test for that specific grain some how and determine which config to use, correct?
06:21 UtahDave yeah, you could check for that grain in your top.sls.  In your nginx init sls when you're in jinja you have that "grains" dict available
06:23 AdamSewell i'm not sure i follow the last part of that sentence. how do you get an sls into ninja?
06:25 middleman_ joined #salt
06:26 \ask joined #salt
06:28 UtahDave {{ grains['image'] }}
06:29 luminous joined #salt
06:30 AdamSewell UtahDave, i think i understand. thanks for the help.
06:30 luminous joined #salt
06:30 UtahDave you're welcome!  Let me know how it goes
06:36 druonysus joined #salt
06:38 carlos joined #salt
06:39 anuvrat joined #salt
06:41 NotreDev joined #salt
06:43 NotreDev the virtualenvs that salt creates don't have ./bin/activate. what was the reason behind this/
06:43 shuaiming joined #salt
06:44 NotreDev never mind. i don't know what i was looking at
06:44 UtahDave NotreDev: ok, that seemed quite odd to me!  :)
06:47 NotreDev for some reason the activate script is not giving executable permissions
06:49 NotreDev i guess i've been using virtualenvwrapper for took long. forgot how vanilla virtualenv works :)
07:01 bhosmer joined #salt
07:02 UtahDave :)
07:10 anti joined #salt
07:12 anti Hi - I've been looking at Salt's event system, specifically event.iter_events.  In testing I see duplicate events for things such as 'test.ping' from the same minion.  I can't determine why and/or if this is functioning as expected.
07:13 anti Has anyone encountered that before?
07:16 anti joined #salt
07:16 malinoff joined #salt
07:17 juasiepo joined #salt
07:17 UtahDave anti: what version of Salt are you using?
07:17 malinoff Hi all
07:17 UtahDave there has been a lot of work on the event system lately
07:17 UtahDave hey, malinoff!
07:18 hvn joined #salt
07:19 malinoff UtahDave, I have a suggestion as to why salt spends so much time in yaml parser. Gonna check that right now
07:20 UtahDave malinoff: cool
07:29 scalability-junk joined #salt
07:31 slav0nic joined #salt
07:31 slav0nic joined #salt
07:32 hvn left #salt
07:33 malinoff UtahDave, I suggested that salt goes through already downloaded yaml files and renders them one-by-one. I'm wrong - I see the same result on just installed minion
07:35 IJNX joined #salt
07:42 UtahDave malinoff: huh
07:43 pipps joined #salt
07:44 jkleckner joined #salt
07:44 pipps_ joined #salt
07:50 malinoff UtahDave, yeah. I'm a little upset.
07:50 giantlock_ joined #salt
07:52 druonysus joined #salt
07:52 druonysus joined #salt
07:54 pipps joined #salt
07:54 thelorax123 joined #salt
07:56 scott_w joined #salt
07:59 harobed joined #salt
08:02 mastrolinux joined #salt
08:02 mastrolinux Hi, how is it possible to use the output of a cmd.run as input (variable) of another state?
08:04 matanya joined #salt
08:04 timbobsteve joined #salt
08:05 UtahDave mastrolinux: use jinja to execute the cmd.run
08:06 xoJIog joined #salt
08:06 scott_w joined #salt
08:06 mastrolinux yes I was thinking about setting a variable with set in jinja but I am not able to do it, do you have links with examples?
08:09 mastrolinux UtahDave: it does not solve the problem, my problem is not to run cmd.run state, the problem is to retrieve the output, basically I did not found a way in salt to interact with input/output and store it somewhere
08:10 malinoff mastrolinux, {% set result = __salt__['cmd.run'](args) %}
08:10 malinoff result will be a dictionary
08:10 UtahDave mastrolinux: in your own python application?  or like malinoff showed
08:10 ollins joined #salt
08:10 mastrolinux malinoff: ah ok, that's nice, thanks a lo
08:11 mastrolinux thanks both I'll let you know if it works fine
08:12 xoJIog Hi, guys! I cannot get salt-ssh to work with nodegroups, or just plain list of hosts
08:12 honestly __salt__['cmd.run'] <- shouldn't salt do it instead of __salt__?
08:13 xoJIog salt-ssh  -L 'host1,host2' test.ping
08:14 xoJIog or salt-ssh  -N 'group1' test.ping
08:14 malinoff honestly, http://docs.saltstack.com/topics/development/dunder_dictionaries.html
08:15 honestly yeah
08:15 honestly but salt and grains works too
08:15 honestly what's the difference?
08:16 UtahDave xoJIog: do the docs say you can use nodegroups with salt-ssh?   I think you can only use rosters
08:17 xoJIog UtahDave: from salt-ssh --help:
08:17 xoJIog -N, --nodegroup     Instead of using shell globs to evaluate the target                         use one of the predefined nodegroups to identify a                         list of targets.
08:17 xoJIog and for list:     -L, --list          Instead of using shell globs to evaluate the target                         servers, take a comma or space delimited list of                         servers.
08:19 UtahDave xoJIog: Hm. if it's not working, could you open an issue on that?  We're hoping to release 0.17.2 tomorrow morning.
08:19 UtahDave I'd like to get that fixed for the release
08:20 pentabular joined #salt
08:20 xoJIog ok, I'll do that
08:20 xoJIog thanks
08:20 UtahDave thanks, xoJIog!
08:20 pentabular left #salt
08:35 ze- joined #salt
08:37 cedwards1 joined #salt
08:40 tempspace_ joined #salt
08:41 zach__ joined #salt
08:41 ksalman_ joined #salt
08:44 middleman_ joined #salt
08:44 halfss__ joined #salt
08:44 mr_chris joined #salt
08:44 davet joined #salt
08:44 fishpen0 joined #salt
08:44 Damoun joined #salt
08:44 yota joined #salt
08:44 BogdanR joined #salt
08:44 whitepaws joined #salt
08:44 johnsocp joined #salt
08:44 rhand joined #salt
08:44 seanz joined #salt
08:44 Sypher joined #salt
08:44 throwanexception joined #salt
08:44 minaguib joined #salt
08:44 cce joined #salt
08:44 cyp joined #salt
08:44 Ahlee joined #salt
08:44 andersb joined #salt
08:44 pt|Zool joined #salt
08:44 Kraln joined #salt
08:44 tru_tru joined #salt
08:44 ewong| joined #salt
08:45 ajw0100 joined #salt
08:45 scalability-junk joined #salt
08:45 xoJIog joined #salt
08:45 Kraln- joined #salt
08:46 xl2 joined #salt
08:46 juicer21 joined #salt
08:46 Koma_ joined #salt
08:46 mike251 joined #salt
08:47 mastrolinux joined #salt
08:47 reinsle joined #salt
08:47 Vye joined #salt
08:47 philipsd6 joined #salt
08:47 Zuru joined #salt
08:47 heewa joined #salt
08:47 ioni joined #salt
08:47 schristensen joined #salt
08:47 VSpike joined #salt
08:47 mephx_ joined #salt
08:47 errr joined #salt
08:47 jesusaurus joined #salt
08:47 fxdgear joined #salt
08:47 _mzito___ joined #salt
08:47 cowmix joined #salt
08:48 mike251 hey guys.... is there a way to group states? for example i have a pillar group: web .... and a state that matches that pillar... and for that state i want to install 3 states: apache mysql and php (i already have the states) - i need to trigger all 3 states depending on that pillar WEB group. IS it possible?
08:50 schristensen joined #salt
08:50 timbobsteve mike251: just create the state and use "include" to include the other required states
08:51 timbobsteve so web.sls just "include"s apache2 + mysql + php
08:52 mike251 timbobsteve: YES you are awesome... i am an idiot... i just read yesterday about includes... pfff thanks buddy!
08:53 kadel joined #salt
08:59 honestly /8[C/
09:00 halfss joined #salt
09:18 halfss__ Hi, how could i import multi var by "from"?
09:18 halfss__ it's work "{% from 'lib.sls' import port with context %}",but i want "{% from 'lib.sls' import * with context %}"
09:19 mike25ro halfss__: i also want to know that :)
09:19 halfss__ mike251: look this: http://docs.saltstack.com/ref/states/include.html
09:20 NotreDev to use the supervisord state, do i need to ensure the supervisor daemon is already running?
09:20 mike25ro ok...
09:22 timbobsteve NotreDev: It depends on whether or not the module uses supervisorctl command to modify configuration, or whether it just creates configs. I'd assume that it would probably need both to do all tasks
09:24 timbobsteve NotreDev: answer is yes... it uses supervisorctl to add/remove/restart/stop etc... so supervisor will need to be running and the salt user will need to be able to run supervisorctl as an admin
09:24 NotreDev timbobsteve: i don't understand what you're getting at. i'm trying to declare that a supervisor service is running. looking at the output it's calling supervisorctl -c, but falling over because the daemon isn't running
09:24 NotreDev ok
09:25 NotreDev so a running supervisorctl service is a prereq
09:29 timbobsteve NotreDev: yes... which was what I was getting at when I said "the module uses supervisorctl command"
09:29 timbobsteve supervisorctl connects to the running supervisord socket file to talk with the daemon... if it's not running, there is no socket to connect to ;-)
09:30 NotreDev timbobsteve: for the longest time i assumed it would launch a supervisor daemon if need be
09:31 timbobsteve NotreDev: yeh, it would make sense, but I think, given that server.{running,start,stop} is another module, it makes sense to declare service.running in your state and make it a pre-req for your supervisor state declaration
09:32 NotreDev timbobsteve: i've never seen a supervisor install provide a service wrapper
09:34 NotreDev i noticed there is a yum package of supervisor in EPEL.
09:34 timbobsteve timbobsteve: I'm the opposite, I use the supervisor package provided with Ubuntu and I've never /not/ had one :P
09:36 NotreDev timbobsteve: perhaps then i shouldn't be installing supervisor in it's own virtualenv
09:37 timbobsteve NotreDev: I wondered that too... doing it in Virtualenv gives you the latest/greatest version, but the packaged versions always work out of the box... so it's a trade-off. There are init scripts for the PIP installed version though, so you could pip install it and deploy an init script with Salt if you do want to go down that route
09:38 balboah joined #salt
09:38 timbobsteve init scripts are out on the great Interwebs somewhere.. let me see if I can rustle up a link
09:39 timbobsteve NotreDev: https://github.com/Supervisor/initscripts <-- lots of different versions depending on your particular flavour of *nix
09:40 NotreDev thanks
09:44 pcarrier joined #salt
09:45 ggoZ joined #salt
09:45 patrek joined #salt
09:46 mastrolinux joined #salt
09:53 phillip joined #salt
09:55 phillip Can someone point me at the best resources for using saltstack to manage a bunch of mac minis (os x)
09:56 malinoff phillip, http://docs.saltstack.com
09:56 phillip haha, yeah i guess i deserved that
09:56 emilisto_ have there been any changes in what's a valid ID declaration from v0.16 to 0.17?
09:57 phillip I'm having some interesting issues with 20-70 zombie processes being started on my minions
09:57 emilisto_ I'm using spaces in my ID declarations, and some of my states broke during the upgrade
09:58 mastrolinux last question, than I stop boring you: {% set gitversion = __salt__['cmd.run'](name='git log | head -1', cwd='{{ pillar["home"] }}/newclient', user='{{ pillar["user"] }}') %} does not work are there errors?
10:00 BbT0n joined #salt
10:04 fr0de joined #salt
10:04 fr0de Hi all
10:04 malinoff fr0de, Hi
10:05 malinoff mastrolinux, what do you get in gitversion?
10:07 fr0de Just a quick question: if we have a server running, say Unicorn, and we set up salt to manage its config (state), how do we get new versions of the actual webapp onto the server? what's the best way to deploy new code, in other words? I have been reading docs and googling, but haven't seen a clear answer (or my search skills are poor ;) ..?
10:07 mastrolinux it does not matter is just an example of the command I want to store the commit hash
10:07 mastrolinux malinoff: even if I change the command name with ls the gitversion variable is not set
10:08 malinoff mastrolinux, try to {% set test = __salt__['cmd.run']('echo test') %} Does it work?
10:09 mastrolinux malinoff, nope
10:09 mastrolinux Undefined jinja variable; line 353 in template
10:10 nocturn joined #salt
10:10 malinoff mastrolinux, Can you pastebin the whole file?
10:10 mastrolinux doing
10:11 mastrolinux malinoff: http://pastebin.com/0WNyk9KU
10:12 malinoff mastrolinux, try this: http://pastebin.com/gYvPR0nC
10:13 mastrolinux this means there is no more correlation with the command previously executed
10:13 mastrolinux and the cmd.run used as a test varbiable
10:13 mastrolinux needs to be done after the newclient_git recipe
10:13 mastrolinux but I'll try
10:14 viq morfternoon
10:14 mastrolinux malinoff, still not working
10:14 Balu_ joined #salt
10:15 mastrolinux I am able to reproduce on different salt servers
10:15 mastrolinux my versions: http://pastebin.com/V9E5NMfV
10:16 elfixit joined #salt
10:16 baffle fr0de: I've been contemplating the same thing; I know there was some talk about getting functionality ala Capistrano as a module to salt, but I've not heard about anyone doing it.
10:17 mastrolinux I thin the only one way to do it is to use pydsl maybe but this is now needed for us, I already move a tons of fabric script to salt
10:17 baffle fr0de: So I think you have to use some third party tools for that for now. Capistrano is ofcourse one of the more well known code deployment tools, but hopefully there are better alternatives.
10:17 baffle fr0de: Hopefully one that isn't written in Ruby. \o/
10:18 malinoff mastrolinux, http://pastebin.com/Y64brDte
10:18 fr0de baffle: yeah - we're using puppet at work now, so getting to switch to salt would necessitate a deploy tool being available or at least easy to create
10:18 malinoff mastrolinux, that's how you should do that
10:19 malinoff just `salt` instead of `__salt__`
10:19 baffle fr0de: Are you using Puppet for code deployment as well?
10:20 fr0de baffle: no, we've got a hackish in-house-developed tool that makes use of mcollective
10:20 mastrolinux malinoff, now it works outside the state but as I said I need it inside the state to be dependent from the previous state
10:20 fr0de but it's not very elegant
10:20 baffle fr0de: Well, since you allready have the remote execution possibilities with the message queue of salt, it might be somewhat easy to port..
10:21 fr0de baffle: what I would like is the elegance of salt (everything is data!) rather than a bunch of dodgy routines that rely on the not-overly-reliable activemq/mcollective
10:22 malinoff mastrolinux, you can't put that right under a stateid, because this will cause formatting error.
10:22 malinoff mastrolinux, http://pastebin.com/dRSaseFS
10:22 backjlack joined #salt
10:22 MK_FG joined #salt
10:22 MK_FG joined #salt
10:23 fr0de baffle: but I think you're right, we'd need to build our own (and hopefully share it back with community)
10:23 mastrolinux malinoff, that's cool indeed
10:25 dangerousbeans joined #salt
10:25 mastrolinux malinoff, but wait salt['cmd.run'] accepts just the command, not the other parameters like cwd, user and group
10:27 malinoff mastrolinux, {{ salt['cmd.run']('echo abc', cwd='/home/user') }}
10:28 mastrolinux yes I was doing the same, some steps forward
10:28 mastrolinux now the problem is '{{ pillar["home"] }}/newclient' is not expanded in jinja set itself
10:28 mastrolinux so this must be tricky to do
10:30 malinoff What do you mean by saying 'not expanded in jinja set itself'?
10:30 thelorax123 joined #salt
10:31 hvn joined #salt
10:31 hvn left #salt
10:31 aib_233 joined #salt
10:33 graingert joined #salt
10:33 graingert I'm having an issue with the npm salt state, let me just wait for a provision and paste the error
10:34 graingert https://gist.github.com/graingert/7464711
10:35 alunduil joined #salt
10:36 MK_FG joined #salt
10:36 MK_FG joined #salt
10:38 bhosmer joined #salt
10:38 aib_233 joined #salt
10:43 IJNX_ joined #salt
10:45 baffle Hmm, I'm having some problems building debs, dpkg-buildpackage errors out with "ImportError: No module named ext.pygmentplugin". python-mako [0.8.1] is installed, and /usr/share/pyshared/mako/ext/pygmentplugin.py exists.
10:52 aib_233 hi. does saltstack provide functionality related to tc / qos on linux servers?
10:52 zandy joined #salt
10:55 graingert aha https://github.com/isaacs/npm/commit/f33163c5723809f64ffbbe56c05502d9ab449ad5
10:57 thelorax123 joined #salt
10:58 Furao joined #salt
11:06 diegows joined #salt
11:10 joehh baffle: is this the right channel? or are you trying to build salt debs? if so, which os/release?
11:17 blee joined #salt
11:19 baffle joehh: I'm trying to build salt debs. Ubuntu/Precise.
11:19 juasiepo joined #salt
11:19 joehh cool - what is your aim?
11:19 joehh are you adding a patch?
11:19 joehh just trying to do it?
11:19 joehh ... something else?
11:20 baffle joehh: I think I've been hit by some bugs in the current version in the repo, that I think are fixed in master/head.
11:20 baffle joehh: So I need to build some new debs and distribute to see if it works. :)
11:20 joehh ok, so you want to build master/head as a deb
11:21 baffle It's a bug in building the docs it seems.. This seems very releavnt: https://groups.google.com/forum/#!msg/salt-users/CMpN0oJN53w/coHxP5qMK0oJ
11:21 joehh :) the docs have been painful over time...
11:23 baffle The complete error is here: http://paste2.org/MgdHCc1c
11:24 joehh that looks new to me
11:24 joehh I suspect that there is a new dependency
11:24 joehh also, the debian dir in the repo is a little outdated
11:24 baffle Hmm, am I not using the correct repo?
11:24 joehh how much experience do you have building debs (just wanting to aim help at the right level)
11:25 joehh ?
11:25 baffle joehh: I've been doing it on and off since 99/2000.
11:25 joehh cool
11:25 joehh so I would grab the debian dir from the launchpad build
11:26 baffle (Not as a maintainer, just to backport or make customer ugly packages)
11:26 joehh dget -ux http://ppa.launchpad.net/saltstack/salt/ubuntu/pool/main/s/salt/salt_0.17.1-2precise.dsc
11:26 baffle Yeah, I did that first.
11:26 baffle (It's not this one)
11:26 joehh and still had the problem
11:26 joehh ?
11:27 scott_w joined #salt
11:27 joehh I'm just cloning down the git repo to give it a go
11:28 baffle Yeah.
11:28 joehh I suspect you're giving me a heads up for building the debian packages
11:29 baffle Hmm, I do notice that I have a newer version of python-mako than what is in base. Standard base has 0.5, I have 0.8.1 due to the Ubuntu Cloud repos being enabled.
11:29 joehh can't comment on that - python-mako hasn't traditionally caused any probs
11:31 baffle Downgrading just to check.
11:31 baffle No, that was not it, same error.
11:33 timbobsteve left #salt
11:34 joehh its very odd - only reference I can find to pygments is in css
11:37 baffle _static/pygments.css doesn't exist; Maybe it is something that should have been created by that Python module it says it doesn't find? :)
11:37 baffle joehh: Did it build for you?
11:38 baffle joehh: Just entering doc/ and doing "make html" triggers it.
11:38 joehh haven't tried yet - still looking at code
11:38 joehh cool
11:39 joehh I'll try that
11:39 ipmb joined #salt
11:41 joehh I know head was in a state of flux a few days ago with integration of salt cloud
11:41 joehh don't know if this is related
11:42 joehh building now - my buildmachine is not quick...
11:45 joehh works for me on sid...
11:45 baffle So it probably is something local then.
11:46 baffle Wich repo are you using? I'm using https://github.com/saltstack/salt.git
11:46 joehh same one
11:47 baffle Sphinx version 1.1.3?
11:47 baffle python-sphinx
11:47 joehh 1.1.3+dfsg-8
11:48 baffle Hmm, when I'm doing "make html", I get this:
11:48 baffle /usr/lib/python2.7/dist-packages/pygments/plugin.py:39: UserWarning: Module mako was already imported from /dev/null, but /usr/lib/python2.7/dist-packages is being added to sys.path import pkg_resources
11:49 joehh I've seen that with pydap...
11:49 joehh are you using virtualenvs
11:49 joehh ?
11:49 joehh https://github.com/saltstack/salt/issues/5718
11:50 joehh ?
11:50 baffle Not here, no. Unless the build system uses it. I've just done a checkout + build-dep on salt.
11:51 baffle joehh: Hmm, should pygmentplugin be in python-sphinx? Do you get any output with: dpkg -L python-sphinx | grep -i pygmentplugin
11:52 baffle Nah, that would be strange.
11:54 joehh no
11:54 joehh not in sphinx source package either
11:54 joehh I'm now lookingin pygmnets source package
11:56 joehh I wonder if it is not mako related?
11:57 joehh root@unstable-builder:/tmp/d/mako-0.9.0# find ./ -type f | grep -i pygmen
11:57 joehh ./mako/ext/pygmentplugin.py
11:57 joehh ./doc/_static/pygments.css
11:57 joehh interesting...
11:58 joehh I've got python-mako 0.9.0-1 installed
12:00 joehh 0.7.0-1 also has ext.pygment or whatever it was
12:00 joehh what does dpkg -L python-mako | grep pygm give you?
12:07 baffle /usr/share/pyshared/mako/ext/pygmentplugin.py
12:07 baffle /usr/lib/python2.7/dist-packages/mako/ext/pygmentplugin.py
12:08 harobed joined #salt
12:09 harobed joined #salt
12:10 joehh so it should be there...
12:11 baffle Yeah.
12:11 baffle (Just did ultradirty hack and ignored building docs for now)
12:12 joehh You wouldn't believe the number of times I've been tempted to do that... though only ever done it once for raring
12:12 joehh :)
12:13 giantlock_ joined #salt
12:17 baffle Gah, and now salt-minion won't start..
12:17 baffle http://paste2.org/kk8gc5t9
12:18 harobed joined #salt
12:19 harobed joined #salt
12:19 blee baffle, im shooting in the dark, but do you ahve customs grains? is the syntax borked in those files?
12:19 baffle blee: Not on that minion. :)
12:20 joehh :(
12:20 harobed joined #salt
12:20 baffle blee: I have a few grains in /etc/salt/grains on some other machines, but not this one.
12:21 joehh unfortunately I reckon you're best to wait for 0.17.2. They're working feverishly to finish it - hopefully today...
12:21 mkoistin_ joined #salt
12:21 blee oh man
12:21 blee sounds srs
12:21 joehh Hassle UtahDave or basepi when they show up around US west cost morning
12:21 joehh 2-3 hours from now
12:22 mkoistin_ Does salt-master have a setting which would require a user to type in a password to run?
12:23 baffle UtahDave seems to be around *always*. :)
12:24 joehh true :) except now :(
12:30 Iwirada joined #salt
12:32 IJNX joined #salt
12:34 bhosmer joined #salt
12:36 graingert techhat
12:37 baffle What I'm trying to get fixed is this strange error I'm suddenly having after an upgrade of salt, where variables doesn't seem to be available to jinja templates when running a state:
12:37 baffle http://paste2.org/BZd67v9h
12:37 baffle joehh: This is *strange* right?
12:41 joehh not familiar enough, but malinoff and mastrolinux seemed to have a discussion about something similar recently
12:42 sterut joined #salt
12:43 sterut How would one generate custom grains serverside? That is to say, instead of in the _grains folder which, as I understand, is ran from the minion, run something locally on the salt master?
12:43 sterut Or is that even possible?
12:44 dangerousbeans Hihihhihi, does someone have an example of using grains to build a hosts file? <HOPEFUL FACE>
12:44 wunki joined #salt
12:45 wunki cedwards1: you seem to be a big user of Salt stack on FreeBSD. Did you ever experience Salt locking up your box?
12:46 aleszoulek joined #salt
12:49 dangerousbeans er I meant Pillars
12:50 baffle dangerousbeans: Not really, but I'm doing ugly things by abusing the publish.publish API from minions to get information about "roles". :-)
12:50 viq sterut: how about pillars?
12:50 dangerousbeans baffle: thanks I'll have a look at this :V
12:51 baffle dangerousbeans: Not sure if it is a good idea, tho'. And maybe mine is better, but I haven't looked into it. :)
12:51 baffle (Mine caches grains serverside, right?)
12:51 sterut viq: I am thinking more minion specific. Specifically taking a grain on a minion (say: server id) and using another service to then suplement the info with that ID.
12:51 malinoff baffle, can you pastebin the whole statefile?
12:52 dangerousbeans this is what I'm doing: http://pastebin.com/M5Q2KXBt
12:52 dangerousbeans but I have no idea what I'm doing
12:52 baffle malinoff: Yes
12:52 dangerousbeans I wish salt had erb templates :P
12:52 dangerousbeans </rubytroll>
12:53 malinoff dangerousbeans, there is no place for ruby in python's world
12:53 dangerousbeans malinoff: that can be viewed the other way around too :P
12:53 baffle malinoff: http://paste2.org/Jk0bxPVz
12:54 malinoff baffle, oh god
12:54 baffle malinoff: I know.
12:54 malinoff baffle, my 1st question, why don't you use pillars?
12:55 baffle malinoff: I do.
12:55 malinoff Hm
12:55 joehh dangerousbeans: maybe replace .get('all', []) with .items()
12:55 dangerousbeans joshh: key trying that shizzle
12:55 malinoff Is "openstack/config.sls"  pillar file?
12:55 baffle malinoff: openstack/config.sls that is included at the top pulls everything out of pillars.
12:55 sterut dangerousbeans: What are you trying to accomplish? Wouldn't central configuration of DNS be better handled by a DNS server?
12:55 malinoff baffle, got it
12:55 joehh pillars is a good way to construct hosts files. There may be better ways (I don't know them), but pillars can do the job
12:56 malinoff baffle, But i'd rather use {{ pillar['config']['debug'] }} and so on right in nova.conf
12:56 malinoff But anyway
12:56 dangerousbeans sterut: we have a 10 machine app, but each machine needs to know where it's memcache, database etc is
12:56 malinoff That could be the way
12:56 baffle malinoff: It's based on something someone else wrote, I've just extended it. *My* shit looks different. :)
12:56 dangerousbeans which we do by using standardised hostnames 'database', 'memcache' etc
12:56 nocturn joined #salt
12:56 dangerousbeans and just mapping them to different places
12:57 dangerousbeans as in our code it just wants to connect to 'database' or whatever
12:57 malinoff baffle, okay...
12:57 dangerousbeans if we used DNS we'd need to be specific
12:58 dangerousbeans web_database, web_cache, backend_cache etc
12:58 malinoff baffle, can you comment the line that uses rabbit_vhost ? What kind of error will you receive?
12:59 jslatts joined #salt
12:59 baffle malinoff: It will error on the line above. {{ rabbit_vhost }} is the last {{ }} in the template, it seems to "fill it out" from the bottom up.
13:00 baffle malinoff: Also, if I insert $RANDOM other variable from the export: list, it will error as well. :)
13:00 baffle As you can see from the first paste, state.show_sls does say that the correct data is exported. :-/
13:01 malinoff baffle, what version of salt do you use?
13:02 baffle malinoff: 0.17.1-2precise
13:03 AviMarcus joined #salt
13:03 sterut dangerousbeans: Can you paste an example of your pillar structure?
13:04 malinoff baffle, there are many jinja fixes in development branch. I heard that salt 0.17.2 will be released tomorrow or so
13:04 malinoff So I'd rather wait that version
13:04 malinoff Dunno what's the problem right now
13:04 joeTSUNAMI joined #salt
13:05 baffle malinoff: It worked with 0.17.0.1-1precise. I think it also worked with 0.17.1-1precise, but failed with 0.17.1-2precise. Might be 0.17.0 -> 0.17.1 that broke it tho.
13:05 baffle Sigh.
13:05 baffle :)
13:05 malinoff baffle, yeah, salt is buggy enough :)
13:06 Brew joined #salt
13:07 graingert joehh: heya
13:08 graingert can we get a hotfix on the ppa for https://github.com/saltstack/salt/commit/4292ad334eabaa7492765338748d47bcd69a03d4 ?
13:08 foxx joined #salt
13:12 harobed joined #salt
13:16 joehh graingert: I won't (Corey might, but I doubt it), two reasons, 0.17.2 is too close, but more importantly you can drop that patched file into _modules and sync it out
13:17 joehh graingert: hope that helps
13:17 graingert ah I see
13:19 joehh baffle: as far as I remember, the only change in 0.17.1-2 was related to salt-ssh, so I think it was 0.17.0 -> 0.17.1
13:20 wunki is it normal that salt opens this many sockets: https://gist.github.com/wunki/c3094943174b848fea21
13:20 blee joined #salt
13:23 giantlock_ joined #salt
13:25 anti joined #salt
13:26 anti joined #salt
13:26 joeTSUNAMI joined #salt
13:29 graingert joehh: cool, thanks for that
13:29 oz_akan_ joined #salt
13:29 graingert joehh: I hear 0.17.2 is coming out soonish
13:30 oz_akan_ joined #salt
13:31 mwillhite joined #salt
13:33 joeTSUNAMI1 joined #salt
13:34 dangerousbeans joined #salt
13:34 ipmb joined #salt
13:34 dangerousbeans sterut: hey sorry was eating
13:34 dangerousbeans sterut: base:
13:34 dangerousbeans '*':
13:34 dangerousbeans - addresses_rackspace
13:34 dangerousbeans and
13:34 sterut send it on pastebin dangerousbeans
13:34 dangerousbeans machines:
13:34 dangerousbeans master:               10.180.72.128
13:34 dangerousbeans web_production_alpha: 10.180.73.233
13:35 dangerousbeans # web_production_beta:
13:35 dangerousbeans # gateway:
13:35 dangerousbeans classifier:           10.180.72.126
13:35 dangerousbeans counter:              10.180.74.7
13:35 dangerousbeans database:             10.179.198.128
13:35 dangerousbeans learner-1:            10.180.69.253
13:35 dangerousbeans # learner-2:
13:35 dangerousbeans messenger:            162.13.124.199
13:35 scott_w joined #salt
13:35 dangerousbeans oh yeah oopse forgot this wasn't basedamp S:
13:35 dangerousbeans camp*
13:36 jrdx joined #salt
13:38 DredTiger joined #salt
13:39 sterut dangerousbeans: Can you send it on pastebin?
13:40 dangerousbeans sterut: kk
13:41 dangerousbeans sterut: http://pastebin.com/ZJjQTb7G
13:48 cron0 joined #salt
13:51 tty-1 joined #salt
13:52 joeTSUNAMI joined #salt
13:56 mwillhite joined #salt
13:57 micko joined #salt
14:03 derelm joined #salt
14:07 whiskybar joined #salt
14:07 juicer2 joined #salt
14:08 sgviking joined #salt
14:09 brianhicks joined #salt
14:11 sterut dangerousbeans: Hey, i'm back, someone grabbed me here at work. Anyway...{%- for host_name, host in pillar['machines'].iteritems() %}
14:11 micko joined #salt
14:11 dangerousbeans sterut: ohhh cool thanks
14:12 vejdmn joined #salt
14:13 micko joined #salt
14:14 scott_w joined #salt
14:14 racooper joined #salt
14:14 mpanetta joined #salt
14:15 Anb joined #salt
14:15 Anb joined #salt
14:15 toastedpenguin joined #salt
14:17 tdillio joined #salt
14:19 tdillio When I try to run a particular salt state, I get nothing back but the hostname. Other states work as expected and setting the log level to "all" doesn't show any error messages. How can I get more information about the failing state?
14:21 brutasse tdillio: make sure your salt versions match between the master and the minion
14:23 tdillio brutasse: Thanks for the suggestion, I just confirmed they do match and are version 0.17.1. I found the section of my init.sls file that was causing the state to silently fail, but I'm wondering if I can convince salt to tell me why it was failing.
14:25 scott_w joined #salt
14:26 brutasse tdillio: this only happened to me when I had different versions. Otherwise I pretty much always get proper errors
14:26 mpanetta joined #salt
14:26 ktenney joined #salt
14:28 mattmtl joined #salt
14:29 tdillio brutasse: I think this is an error on my part, for not reading the git salt state thoroughly.
14:30 _ikke_ tdillio: enable debug logging for the minion and watch the salt log for the minion
14:30 tdillio _ikke_: Thanks, that sounds like what I'm looking for.
14:31 tdillio _ikke_: I only added debugging on the master, I dunno why that didn't occur to me, not enough coffee yet :)
14:35 NotreDev joined #salt
14:38 maninacan joined #salt
14:40 HeadAIX joined #salt
14:41 tdillio Does salt require python >= 2.7?
14:42 Balu_ Hi all, I found an equivocal msg in salt-ssh. client/ssh/__init__.py
14:42 tdillio CentOS shipes wth 2.6.6, but I'm getting an expection: AttributeError: 'OrderedDict' object has no attribute 'append'. OrderedDict wasn't added to Python until version 2.7
14:42 Balu_ 173                     'Password for {0}@{1}:'.format(host, target['user'])
14:42 quickdry21 joined #salt
14:42 Balu_ usually user@host, am i right?
14:43 Kholloway joined #salt
14:44 mapu joined #salt
14:44 scott_w joined #salt
14:47 anti Hi - I've been looking at Salt's event system, specifically event.iter_events.  In testing I see duplicate events for things such as 'test.ping' from the same minion.  I can't determine why and/or if this is functioning as expected.
14:47 zandy joined #salt
14:49 mpanetta anti: I get duplicated output when the minion is running more then once.
14:49 micah_chatt joined #salt
14:49 AdamSewell joined #salt
14:49 AdamSewell joined #salt
14:52 scott_w joined #salt
14:52 anti hmm
14:53 mpanetta joined #salt
14:53 anti this is duplicate on all minions I have running
14:54 anti yeah, just one minion running
14:54 mpanetta Hmm
14:54 mpanetta Then I donno
14:54 mpanetta Ive not messed with the event system yet
14:55 anti yeah, it's pretty cool minus the duplication of events I'm seeing
14:55 anti may be linked to the master workers I raised in the config?  not sure
14:56 juasiepo joined #salt
14:57 diegows joined #salt
14:59 danielbachhuber joined #salt
15:01 mpanetta joined #salt
15:03 amahon joined #salt
15:06 scott_w joined #salt
15:06 Kholloway joined #salt
15:11 dangerousbeans how can I make a directory nested inside a bunch of also not existing directories in a way which is more DRY than this? http://pastebin.com/raw.php?i=9dY8SeCc
15:12 thelorax123 joined #salt
15:13 viq dangerousbeans: file.recurse?
15:14 joeTSUNAMI joined #salt
15:14 premera joined #salt
15:15 dangerousbeans so
15:15 dangerousbeans change to one entry and put: file.recurse:
15:15 dangerousbeans ?
15:16 anti hmmmm, maybe line 1258 in master.py is suspect to duplicate events I'm seeing in the event iterator
15:16 anti Am I like the only one using python to listen to events on the master? ;)
15:16 viq dangerousbeans: also, did you read the docs? have you seen makedirs ?
15:17 dangerousbeans viq: ah yeah I see it now
15:19 joeTSUNAMI joined #salt
15:21 cedwards1 wunki: i've never had Salt lock up my machine, no.
15:21 forrest joined #salt
15:22 tyler-baker joined #salt
15:22 tyler-baker joined #salt
15:22 cedwards joined #salt
15:22 aleszoulek joined #salt
15:22 lineman60 joined #salt
15:23 alunduil joined #salt
15:23 wunki cedwards: ah, that's good. My dedicated just keeps locking up but don't know if Salt stack is the culprit
15:24 wunki cedwards: all network connections just fail
15:24 scott_w joined #salt
15:25 NotreDev joined #salt
15:26 zandy joined #salt
15:26 wunki cedwards: if you could take a look, is it normal that salt opens this many sockets: https://gist.github.com/wunki/c3094943174b848fea21
15:27 forrest wunki, how many boxes are in your environment?
15:27 wunki forrest: only three I believe at that moment
15:27 forrest oh
15:34 networkpadawan joined #salt
15:35 pass_by_value joined #salt
15:36 anti so yeah, the # old dup event comments in master.py are suspect. anyone know the history there? i've modified some code but maybe i just need to get on github
15:37 pdayton joined #salt
15:37 jalbretsen joined #salt
15:38 JulianGindi joined #salt
15:38 pipps joined #salt
15:44 amahon joined #salt
15:45 pipps joined #salt
15:46 jkleckner joined #salt
15:46 ravibhure joined #salt
15:47 bezaban yay. Demonstrated the weeks works on rolling out salt to my boss (who has been home sick) and he was majorly impressed
15:48 forrest nice bezaban
15:48 bezaban full roll out in testing next week, then production :)
15:48 forrest did you get all your states written?
15:48 bezaban still a lot left to do there
15:48 forrest ahh ok, how are you handling your states for test versus your states for prod?
15:49 bezaban I'm doing compound matches on hostnames, but a lot of the states will be very similar
15:49 forrest ahh ok
15:49 anti bezaban: you deploying it with any sort of high-availability?
15:49 shinylasers joined #salt
15:50 bezaban mostly it's very generic stuff at first, ssh keys, user management, ntp config, ssh config
15:50 bezaban then there are a lot of projects we are going to be needing it for.  Centralized logging, log signing etc
15:50 kermit joined #salt
15:50 shinylasers joined #salt
15:50 bezaban anti: no HA atm, just a single (but virtualized and redundant) server
15:50 jkleckner1 joined #salt
15:51 bezaban the existing virtual infrastructure is all redundant
15:51 bezaban RAID60 SAN connected to 2x virtualization platforms and dual switches, firewalls etc
15:52 anti that helps.  I was looking into master clustering capabilities.  Didn't get too far as I'm trying to debug some stuff in the event bus right now
15:53 bezaban still vulnerable to configuration errors, but I should be spared for hardware issues
15:53 scott_w joined #salt
15:53 bezaban I'm still very new to salt, but got the green light to delve further into it, so very happy
15:54 teskew joined #salt
15:55 bezaban also, all the meta data in the hostnames proved very useful
15:57 scott_w joined #salt
15:59 NotreDev joined #salt
16:04 scott_w joined #salt
16:05 Iwirada left #salt
16:06 ipmb joined #salt
16:06 pipps_ joined #salt
16:08 rgbkrk joined #salt
16:10 mpanetta_ joined #salt
16:26 dangerousbeans has anyone got any experience of installing/using rvm with salt?
16:26 forrest I haven't personally dangerousbeans, but there's docs.saltstack.com/ref/states/all/salt.states.rvm.html
16:27 forrest it has a pretty thorough example
16:27 dangerousbeans yeah that all ran fine
16:27 dangerousbeans and rvm is installed okay
16:27 dangerousbeans but you're meant to be able to run inside the rvm shell, via this
16:27 dangerousbeans http://docs.saltstack.com/ref/modules/all/salt.modules.rvm.html
16:28 halfss joined #salt
16:28 dangerousbeans which is saying: rvm is not installed
16:28 forrest that's a module
16:28 forrest not a state
16:28 forrest or are you trying to do that off the command line after running the state?
16:31 dangerousbeans forrest: sorry wrong link, I'm trying to ensure a gemset exists, via the state functions, but also this fails:
16:31 dangerousbeans salt-call rvm.gemset_create 2.0.0-0 passenger
16:32 dangerousbeans Ruby  is not installed.
16:32 dangerousbeans WAIT, typo
16:32 dangerousbeans ignore the last one
16:32 dangerousbeans so this works: salt-call rvm.gemset_create 2.0.0-p0 passenger
16:32 dangerousbeans gemset created passenger=> /usr/local/rvm/gems/ruby-2.0.0-p0@passenger
16:32 dangerousbeans but inside my state dude it says rvm is not installed
16:33 dangerousbeans if I use: rvm.gemset_present:
16:33 ipmb joined #salt
16:33 jkleckner joined #salt
16:33 dangerousbeans http://pastebin.com/9TDnjH0f
16:34 forrest where's your rvm.installed?
16:35 dangerousbeans which rvm
16:35 forrest so in the example they always run an rvm.installed.
16:35 smccarthy joined #salt
16:36 dangerousbeans "/usr/local/rvm/bin/rvm"
16:36 dangerousbeans so it's in the right place
16:36 dangerousbeans all the users seem to be able to access it
16:36 basepi forrest: can you link that commit you wanted me to review again?  #salt is so busy I don't have the scrollback
16:36 toastedpenguin joined #salt
16:36 basepi i also have a mention by joehh but i don't know if it's still relevant.  =P
16:37 forrest basepi, you don't need to, UtahDave took a look at it, and Tom actually needs to review it
16:37 basepi forrest: cool.  yay for procrastination!  ;)
16:37 forrest heh
16:37 forrest well, it was way outside business hours
16:37 basepi ya, i was just heading to bed when my phone beeped, and so it didn't happen.  =)
16:38 forrest dangerousbeans, I think you should try to implement an rvm.installed for that just to see if it works
16:38 forrest basepi, yea I don't blame you
16:38 forrest once I lay down unless I'm on-call the phone doesn't get touched.
16:38 basepi yep
16:38 dangerousbeans forrest: kay
16:38 basepi i used to, but then i would end up online for another hour or two and be exhausted the next day
16:38 forrest yea screw that
16:39 pipps joined #salt
16:39 scott_w joined #salt
16:39 dangerousbeans forrest: okay new info: from root, if I run 'su nowa; rvm -v'
16:39 dangerousbeans it says it's not installed
16:39 dangerousbeans but if I run
16:40 dangerousbeans 'su nowa -l; rvm -v'
16:40 dangerousbeans it's fine
16:40 dangerousbeans -l makes the shell a login shell, which rvm uses to fire up all it's magic
16:40 forrest yea
16:40 forrest awesome
16:40 dangerousbeans is there some way I can use cmd.run in interactive mode?
16:40 dangerousbeans I mean login
16:41 forrest I'm honestly not sure, that seems like an issue anyways that it does that, basepi do you know if this was a bug identified at some point?
16:41 dangerousbeans without like cmd.run name: '/bin/bash -l -c "COMMAND HERE"'
16:41 basepi for rvm or an interactive shell?
16:41 Corey dangerousbeans: That sounds dangerous and wrong. What are you trying to do? :-)
16:41 forrest yea
16:41 basepi interactive shell there's an open feature request that hasn't gone anywhere
16:41 forrest for rvm basepi
16:42 basepi rvm there are a couple of open issues, but nothing very recent
16:42 forrest well, the issue is that his http://pastebin.com/9TDnjH0f gemset.present isn't working
16:42 forrest seems more like an rvm issue than anything
16:42 * Gareth waves
16:42 Corey Gareth!
16:42 dangerousbeans well, when rvm installs, it puts a lot of stuff into your .bashrc or something
16:42 dangerousbeans which isn't loaded unless the shell is a login shell
16:42 Corey Gareth: Don't tell me you're at work already! When I worked there folks would roll in at the crack of noon.
16:42 basepi dangerousbeans: ah, that could be a problem
16:43 forrest which is a bummer that's for sure
16:43 Gareth Corey: nope.  not at work yet :)
16:43 basepi dangerousbeans: a couple of these could be related:  https://github.com/saltstack/salt/search?q=rvm&amp;ref=cmdform&amp;state=open&amp;type=Issues
16:43 Corey dangerousbeans: This is why rvm saddens me. What I've done in the past is stuff it into a script that I can run idempotently, and hook that with Salt usually.
16:44 dangerousbeans Corey: yeah rvm is crazy
16:44 dangerousbeans http://rvm.io/support/faq#what-shell-login-means-bash-l
16:44 Gareth rvm. *shudder*
16:45 dangerousbeans they put up some possible fixes
16:46 mkoistinen Is there a way to output the template rendering of the top.sls file for a specific host?
16:46 Corey dangerousbeans: I did some work with either basepi or utahdave (all salt engineers look alike to me) on the RVM module about a year or so ago.
16:47 Corey RVM isn't an environment manager; it's an atrocity.
16:47 Corey "Oh, run this curl script to install it, it's randomly hosted on some guy's github accout. It's probably fine..."
16:47 dangerousbeans agreed it's complex and problematic if setup wrong, luckily I've found an awesome system setup automation tool :P
16:48 dangerousbeans ...or I could ask the chef guys? <troll>
16:48 dangerousbeans >:D
16:48 forrest The Chef dudes are pretty nice
16:48 zandy joined #salt
16:48 Corey dangerousbeans: Actually, that's not a completely horrible idea. If your shop is seriously into Ruby in a big way, there are worse decisions. Like... shoehorning python in. :-)
16:49 dangerousbeans nooo we used chef and it's worse then leprosy
16:49 scott_w joined #salt
16:49 dangerousbeans salt is already 9000% better
16:51 dave_den mkoistinen: salt 'yourminion' state.show_top
16:54 noob joined #salt
16:54 Gifflen joined #salt
16:54 noob should salt run as its own user?
16:54 forrest most people run it as the root user. You could run it as it's own user though
16:55 Gifflen joined #salt
16:56 anti basepi: do you know what the # old dup event comments are about in master.py?  It looks like maybe something legacy was kept in there when tagging was added.  I had to comment out one of those lines to prevent duplicate data in the python event iterator.  Should I open up an issue/bug report?
16:56 pipps_ joined #salt
16:56 scofflaw left #salt
16:56 Guest86288 that would be safer, but salt tries to write into /var/run and maybe elsewhere owned by root - not bueno
16:57 heewa After running 'salt-call state.highstate', the exit code can be 0 (good) while there were errors with some of the states. Any way to find out if all states executed correctly or not?
16:57 heewa Like in a script?
17:00 basepi anti: nah, that's just because we switched to namespaced tags in 0.17, and you should look for those instead (they look like salt/job).  the old non-namespaced events are on a deprecation path for 3 releases from now
17:00 basepi (that's the word from Tom)
17:04 anti basepi: hmm.. I was going to raise a ticket for duplicate events displayed when doing verbatim the last salt.utils.event / event.iter_events() example.  I wanted to listen for all events, so i'm not specifying a tag.
17:04 anti This results in duplicate events.
17:06 derelm joined #salt
17:08 basepi anti: yep.  Tom says that technically these aren't duplicates, as they dont' have the same tag, but you're right that there is more than one event for each actual "event" in salt.  it's a known issue and those other events are on a deprecation path.  no need to create an issue, we're definitely aware.  =)
17:08 higgs001 joined #salt
17:09 anti cool, thanks!
17:11 anuvrat joined #salt
17:12 mkoistinen dave_den: Awesome, thank you!
17:13 tyler-baker joined #salt
17:13 tyler-baker joined #salt
17:13 Koma joined #salt
17:15 scott_w joined #salt
17:16 KyleG joined #salt
17:16 KyleG joined #salt
17:17 rgbkrk joined #salt
17:17 ipmb_ joined #salt
17:18 micah_chatt joined #salt
17:21 redondos joined #salt
17:21 NV joined #salt
17:24 ajw0100 joined #salt
17:25 dangerousbeans okay I'm back :D
17:25 dangerousbeans how do I do something along the lines of: 1) pass in BASH_ENV to bash when running a command
17:26 dangerousbeans 2) run a command in a login shell, instead of a lame normal one
17:26 dangerousbeans 3) prepend all commands with something like 'source /etc/rvm/ohgod/why/scary_rvm_profile.sh'
17:26 viq dangerousbeans: for 1, http://docs.saltstack.com/ref/states/all/salt.states.cmd.html
17:27 pass_by_value joined #salt
17:27 dangerousbeans viq: oh yeah the env thing
17:28 kaptk2 joined #salt
17:29 dangerousbeans syntax?   (in a state)    - env: BASH_ENV=scary_rvm
17:31 viq dangerousbeans: also there you can specify what shell to use
17:31 JulianGindi joined #salt
17:34 viq dangerousbeans: as for 3) I think that's on you ;)
17:37 dave_den dangerousbeans: when you use 'runas' it will su to a login shell for that user. even for root.
17:37 dave_den for 2)
17:38 pentabular joined #salt
17:40 dccc joined #salt
17:42 xmltok joined #salt
17:42 troyready joined #salt
17:42 ravibhure joined #salt
17:47 lemao joined #salt
17:49 jacksontj joined #salt
17:49 scott_w joined #salt
17:49 dvogt joined #salt
17:50 NotreDev joined #salt
17:50 networkpadawan joined #salt
17:51 dangerousbeans dave_den: I don't see 'runas' in the docs: http://docs.saltstack.com/ref/states/all/salt.states.cmd.html#salt.states.cmd.run
17:51 dangerousbeans dave_den: is it hidden?
17:52 dave_den oh, it's "user" now.
17:53 amckinley joined #salt
17:53 pentabular joined #salt
17:54 pentabular left #salt
17:55 jcookeman joined #salt
17:57 amahon joined #salt
17:57 jdenning joined #salt
17:59 ravibhure joined #salt
18:00 scott_w joined #salt
18:02 jhulten joined #salt
18:06 scott_w joined #salt
18:09 Guest8023 Now that salt cloud is merged in with salt, is there any way to update salt docs on rtd?
18:09 forrest I'm pretty sure that's a to-do
18:09 rgbkrk In an issue list somewhere?
18:10 jacksontj joined #salt
18:14 forrest_ joined #salt
18:18 ipmb joined #salt
18:23 tty-1 joined #salt
18:25 Ryan_Lane joined #salt
18:28 snuffeluffegus joined #salt
18:29 Ryan_Lane joined #salt
18:34 maninacan joined #salt
18:36 maninacan *new to salt* Has anyone found good personal uses for salt, such as setting up home systems? And if so, how would you approach configuring your salt states?
18:37 maninacan I'm debating whether there would be any need for minions.
18:37 jcockhren maninacan: salt is great for remote configuring
18:38 grooven joined #salt
18:38 jcockhren like you don't have to log into every box on your network to do install stuff and whatnot
18:38 [diecast] joined #salt
18:38 maninacan jcockhren: Cool. And to do that, I'd need minions, correct?
18:38 jcockhren also, you can leverage salt for scheduled backups, configuration verification, checks, etc
18:39 jcockhren maninacan: so yeah. you have have your main machine be a master and all other devices be minions
18:39 tty-1_ joined #salt
18:40 maninacan jcockhren: Ok, I'm seeing it now.  So before I can do any remote operations, I need to set up my remote machines as minions be it through shelliing in or in person?
18:40 grooven hi all! does any know if you can use "salt-ssh" via the Python client API ? current documentation says that salt.client.LocalClient must be executed on a Salt master... is there something comparable for utilizing SSH instead of master<->minion 0MQ channel?
18:41 jcockhren maninacan: yep
18:41 jcockhren or or
18:41 maninacan jcockhren: Great! Thank you much.
18:42 jcockhren if your home environment sits on ESX or the like. it's possible to spin it up from a master
18:42 jcockhren libvirt and all
18:43 maninacan More stuff I'll need to look into.
18:43 jcockhren maninacan: what are you building/have at home if you don't mind sharing
18:43 pentabular joined #salt
18:44 giantlock_ joined #salt
18:45 mapu joined #salt
18:46 Nexpro joined #salt
18:48 zandy joined #salt
18:52 thelorax1231 joined #salt
18:53 Vye On my W7 minion I'm installing packages via winrepo. However, the system doesn't reboot after the software is successfully installed, like I would expect it to.
18:54 Vye I see an example of the reboot flag documented here: http://docs.saltstack.com/ref/windows-package-manager.html but strangely not in http://docs.saltstack.com/ref/states/all/salt.states.pkg.html#module-salt.states.pkg
18:54 NotreDev i'm trying to serve up a vanilla sentry installation with a fresh sentry.conf.py file. if i do `sentry runserver 0.0.0.0:8000` and try to access it from another machine on the network, i get suspicious activity alerts, and internal service errors. how do i just deploy it?
18:55 jacksontj joined #salt
18:55 log0ymxm joined #salt
18:56 mwillhite joined #salt
18:58 ausmarton joined #salt
19:02 Brew joined #salt
19:03 jacksontj joined #salt
19:05 forrest joined #salt
19:05 Thiggy joined #salt
19:09 bemehow joined #salt
19:10 ipmb_ joined #salt
19:11 mkoistinen Why does salt-master run all the time on my system? Seems to me it should shut itself down after a short while of inactivity, then fire up again when you run a salt command.
19:11 jcockhren mkoistinen: nah. at all times, it needs to check the status of the minions
19:12 mkoistinen Why?
19:12 mkoistinen What does it do with that info?
19:12 mkoistinen I'm not running salt-mine
19:12 mkoistinen Even if I were, it only needs the latest state, which it can get shortly after firing up.
19:12 jcockhren mkoistinen: allows you to run states and check minion status, run schedulers etc
19:13 tty-1 joined #salt
19:13 jcockhren mkoistinen: salt-master is pretty low demand
19:13 mkoistinen I'm not sold.
19:14 jcockhren k. \o/ I give
19:14 jcockhren forrest: ? ^
19:14 jcockhren lol
19:14 pass_by_value joined #salt
19:15 mkoistinen Having salt-master listen on a port all the time when it is doesn't need to 99.9% of the time is just a security nightmare waiting to happen.
19:15 heewa Anyone have tips on debugging why my minion's not getting any pillar data? (empty result of 'salt-call pillar.items')
19:15 DredTiger joined #salt
19:15 jcockhren mkoistinen: why do you say that?
19:15 Gifflen joined #salt
19:16 jcockhren the minions can only communicate with their accepted by their master
19:16 jcockhren when they're*
19:16 anti joined #salt
19:17 joehh basepi: mention still relevant - now I know your phone beeps, I'll avoid mentioning you outside of your hours
19:17 mkoistinen True, and once they've accepted, they can run any arbitrary cmd given by the master. If the master is found to have a sec-vuln, it'll be game over for your whole network.
19:17 joehh thought you may have been there
19:17 nahamu is there a command line tool for watching events the way halite can?
19:17 mkoistinen And as far as "low resources", on my default installation, it spawns 74 processes.
19:17 basepi joehh: no worries, i silence it so it doesn't wake me up or anything
19:17 basepi joehh: what was your question?
19:17 joehh basepi: got that round the wrong way - mention not relevant
19:17 basepi oh, kk
19:17 nahamu In particular I want to note when minions go away and come back...
19:17 basepi =)
19:17 joehh email from tom cleared that up
19:18 mkoistinen and consumes ~ 70 MBs of RES.
19:18 joehh someone was keen to know when 0.17.2 would be available - I said ask you or utah dave :)
19:19 jcockhren mkoistinen: yes. that's applies to any orchestration software.
19:19 jcockhren mkoistinen: but you can also run salt as an unpriviledged user
19:19 jcockhren and utilize the acl features
19:19 sroegner joined #salt
19:20 mkoistinen jcockhren: Even if salt-master runs as 'nobody', if it is compromised, it can control my whole fleet of servers, so, it doesn't really matter what user it runs as.
19:21 mkoistinen And, just because 'any orchestration software' can wreak havoc if compromised should make it OK to leave the door wide open.
19:21 kaptk2 joined #salt
19:21 mkoistinen s/should/shouldn't/
19:21 pears if you don't want to use the master listening port model, use salt-ssh
19:21 jcockhren pears++
19:21 mkoistinen Thank you pears, I'll look into that!
19:21 foxx[cleeming] joined #salt
19:21 foxx[cleeming] joined #salt
19:22 ipmb joined #salt
19:24 monokrome joined #salt
19:24 Teknix joined #salt
19:24 mkoistinen f-yeah, thank you pear, this is precisely what I needed!
19:25 mkoistinen pears^^
19:28 xmltok joined #salt
19:29 heewa Ok, well, does anyone know how I can dive into the code where the pillar top file is supposed to be read? I've grepped around the source and can't seem to find it.
19:29 Ryan_Lane joined #salt
19:30 jslatts joined #salt
19:30 heewa The code is so implicit. So much happens as strings passed around that are lookups into tables of functions and defered execution and dictionaries of options. It's really hard to see where something actually happens, why, who caused it to happen, etc.
19:30 anti joined #salt
19:31 bemehow joined #salt
19:31 anti joined #salt
19:33 mapu I;m having some difficulty ensuring I am running the correct versions of salt and salt-cloud. If I do a git clone https://github.com/saltstack/salt.git, and verify I am on the develop branch, when I do pip install -e ./salt, the --versions-report shows as 0.17.0-4570-g6d78c0e.
19:33 mapu salt-cloud is seen in this clone, and is installed.
19:34 mapu If I then do a git checkout tag/v0.17.2 , and do an install, the version shows as 0.17.2, but then salt cloud is not in the directory, and running salt-cloud fails
19:35 mapu So my question is: what is the right procedure for ensuring that I have the latest and compatible versions of both salt and salt-cloud?
19:38 ipmb joined #salt
19:38 heewa mapu: I heard someone else mention that problem, but a cursory search of the open github issues didn't turn anything up. But it does sound like a bug.
19:39 pass_by_value joined #salt
19:39 mapu Yeah- I don't mind the inevitable bugs that are inherent when running the latest 'develop' branch, but I need to keep my installation work so I can at least move forward in my work and testing.
19:41 NotreDev i'm at wits end with sentry. i have a fresh installation, and if i run `sentry runserver 127.0.0.1:9000` it's usable through my nginx proxy. if i run `sentry start` it fails (even if SENTRY_WEB_HOST='127.0.0.1 and SENTRY_WEB_PORT=9000). http://pastebin.com/ckvJSDLS
19:42 NotreDev sorry, meant to post that in #sentry
19:47 kermit joined #salt
19:48 zandy joined #salt
19:50 scott_w joined #salt
19:51 Ryan_Lane joined #salt
19:52 mkoistinen note to self: "salt '*' service.reload salt-minion" is a bad idea
19:53 kula everyone has to shoot their own foot at least once. i certainly did.
19:53 mkoistinen =(
19:54 mkoistinen Thankfully, there's always "ssh [minion] 'service salt-minion start'" =)
19:56 ckao joined #salt
19:57 mgw joined #salt
19:57 mkoistinen I think if I was a black-hat, I'd scan DNS records for hosts named 'salt' as priority targets.
19:58 mkoistinen And ports for 4505 and 4506
19:58 mgw basepi: do I understand the formula principles properly, that the base init.sls should only install and start the service, and other sls files should handle configuration?
20:00 mgw looking at the convenstions again, not sure where i got that idea, actually
20:02 EugeneKay mkoistinen - also check for records 'slug' ;-)
20:02 basepi mgw: not sure there's a set-in-stone convention for that
20:02 basepi mgw: whiteinge has actually been spearheading a lot of the formula work.  you might ask him
20:03 mkoistinen I presume that is another configuration management tool common configuration?
20:03 mkoistinen Or perhaps a common anti-configuration for salt?
20:03 EugeneKay .....no, it's a joke.
20:03 mkoistinen =)
20:04 EugeneKay (my Salt master is named "slug")
20:06 dave_den joined #salt
20:08 Linz joined #salt
20:12 mwmnj joined #salt
20:18 thelorax123 joined #salt
20:19 whiteinge mgw: i'd suggest having each .sls file in a formula do the least possible thing so that other formulas or people's regular state tree's can ``include`` that sls file in their own states without side-effects
20:19 higgs001 joined #salt
20:19 forrest sorry jcockhren what's up? I was at lunch
20:20 whiteinge if the formula you are writing has a package and a service and people will always want to run the service after installing the package (e.g., apache) then put them in the same .sls file
20:20 elfixit joined #salt
20:20 whiteinge (init.sls would be a good place for basic stuff like that)
20:20 whiteinge on the same token, if a package won't run without additional configuration it would make sense to include that in the same sls
20:21 mkoistinen I'm trying to run salt-ssh and once I connect, I get a Python traceback ending with "ValueError: No JSON object could be decoded". Anyone know what this is offhand before I go wading into the codebase?
20:22 whiteinge mgw: if something (like configuration) is an extra or nice-to-have on top of the base install then i'd suggest putting that into it's own config (e.g., the apache/debian_full.sls formula)
20:24 higgs001 joined #salt
20:24 mkoistinen Hmmm, nevermind, I got it working
20:24 forrest what was it mkoistinen?
20:25 mkoistinen I needed to tell my roster to log in as root (I had assumed that was the default, apparently not)
20:25 pdayton joined #salt
20:25 forrest ahh ok
20:27 jrdx joined #salt
20:31 smccarthy2 joined #salt
20:33 viq dafuq? " State file.replace found in sls elasticsearch is unavailable"
20:33 viq I thought it was "pure python" so should just work...
20:36 smccarthy joined #salt
20:40 diegows_ joined #salt
20:44 jacksontj joined #salt
20:46 mkoistinen How can I tell all my minions to no longer trust the salt-master?
20:46 EugeneKay rm /etc/salt/pki -rf
20:46 EugeneKay Actually, they'll just reconnect to the master. I'm not sure /how/ to force checking of the master, but I have to imagine there's a way
20:47 _ikke_ Yeah, minions seem to just trust the master they get to connect to
20:47 heewa mkoistinen: EugeneKay What if the master's public key on the minions was replaced with a bad one ('echo no > /etc/salt/pki/minion/master.pub' or something)?
20:48 EugeneKay Not a clue ;-)
20:48 mkoistinen Perhaps just to nuke the salt-master's private key AND set a gibberish config in the minion’s config file '?key
20:48 heewa Or you could change the master's key pair by deleting them & restarting master.
20:48 EugeneKay Never done that(yet)
20:48 zandy joined #salt
20:48 mkoistinen master_finger
20:49 mkoistinen I'll set master_finger on all the minions to '$', then delete the master keys. I hope I can still use salt-ssh after that though!
20:50 [1]VertigoRay joined #salt
20:52 mkoistinen Yes! OK, super. =)
20:56 VertigoRay joined #salt
20:58 Brew joined #salt
20:59 pentabular joined #salt
20:59 pentabular left #salt
21:01 cewood joined #salt
21:05 lineman60 joined #salt
21:10 zach Has anyone ran into an issue where salt gets blocked for whatever reason and spirals out of control spawning 20+ salt-minion processes?
21:10 zach salt-minion took down our infrastructure this morning and now I'm under the gun
21:10 mkoistinen Ouch
21:11 zach yeah, not so fun for an infrsatructure responsible for $70,000,000/yr
21:11 mkoistinen Have you already recovered from this and now you're doing a post-mortem?
21:12 zach yep
21:13 VertigoRay joined #salt
21:13 jcockhren zach: me
21:14 dave_den zach: https://github.com/saltstack/salt/issues/5729  and https://github.com/saltstack/salt/issues/8435
21:15 cron0 joined #salt
21:15 mkoistinen zach, I've switched to using salt-ssh which doesn't use agents. So far my experiences are rather positive and I feel a lot safer from a security standpoint. And without agents, it will possibly help your situation. (although, I'm not so sure how well this scales relative to the agent-ful version of salt)
21:16 pentabular1 joined #salt
21:17 forrest according to https://github.com/saltstack/salt/pull/8222 it looks like it might have been fixed.
21:17 zach great
21:17 heewa Does anyone know how pillar_roots gets populated or used? I see the file salt/salt/wheel/pillar_roots.py, but I can't see anyone importing from it.
21:18 zach Thank you guys
21:18 forrest zach, props to dave_den for finding the issues :P dave_den, would you agree taht pull request looks like it should fix it?
21:18 mpanetta joined #salt
21:19 zach Mine is a combo of those plus another issue
21:19 forrest :\
21:19 forrest always a bummer when that happens
21:19 dave_den that fix only applies to mine functions, so if his forks are from mine operations, then it should i think
21:19 zach The script in question somehow failed, which spawned a python shell. Did not notice it right away, somehow that caused saltstack to stack up and not execute other commands that ran after it
21:20 zach it's weird, but yes - the issue in those we've seen as well, I need to fix that
21:20 forrest oh good point dave_den, I don't know why I assumed zah was using mine
21:20 forrest *zach
21:22 pipps_ joined #salt
21:23 bemehow joined #salt
21:23 forrest hey basepi, how is https://github.com/saltstack/salt/issues/8435 fixed by the changes to the Mine itself?
21:23 zach It's a strange issue none the less
21:23 forrest maybe I am just missing how they are connected.
21:23 basepi forrest: it's not directly changes to the mine
21:24 basepi it's changes to keep track of processes spawned by the mine and make sure that we don't keep spawning if previous processes haven't finished.
21:24 forrest ok that makes sense, I was confused by the merge topic then
21:24 forrest versus the changes to the actual code :P
21:25 zach Ok great, that's exactly what happened to us
21:28 zach Has anyone put saltstack into a CI of sorts to build out RPMs?
21:29 forrest uhhh no I hadn't thought to do that, why would you use salt for that instead of using salt to manage an actual rpm build tool like koji?
21:29 zach I was just curious since we have jenkins, could just have it check out the source code to build the rpm and place it into our repo
21:30 bemehow joined #salt
21:31 zach Probably better to rely on than using the packages in epel
21:31 zach lesson 1 - don't take shortcuts ;)
21:33 mgw joined #salt
21:34 forrest well, jenkins is cool for building RPMs, except I don't think it supports MOCK right?
21:34 forrest so you can't assure a clean build environment
21:35 forrest unless someone wrote a plugin for it
21:35 Ryan_Lane joined #salt
21:35 forrest *ensure
21:36 zach that's true
21:36 zach I'll setup koji, I've been building all other RPMs manually, and now I need to start building salt RPMs to keep things current
21:37 forrest setting up koji sucks
21:37 forrest just as a heads up
21:37 forrest you know that as soon as the salt RPM gets built you can get it from epel-testing right?
21:37 zach Oh really? I can just enable epel-testing for salt-minion
21:38 Oscar joined #salt
21:38 zach That would make things infinately easier
21:38 forrest I know basepi said that the latest release of salt is in pypi, and I believe the RPMs are now being built by terminalmage?
21:38 basepi yes.  though they may not be in testing until Monday
21:38 forrest yea, I just don't know what the ETA is on getting it built for epel-testing
21:38 terminalmage yes but I probably won't get them built until the weekend
21:38 forrest ahh good to know basepi
21:38 terminalmage lot going on right now
21:38 forrest thanks terminalmage
21:38 forrest So zach, if you need a quick workaround
21:38 forrest you can just use mock
21:39 [diecast] joined #salt
21:39 zach will do, appreciate it
21:39 forrest zach, the specs/patches are here: https://github.com/saltstack/salt/tree/develop/pkg/rpm
21:39 forrest have you used mock before?
21:40 Oscar Hi there, I was wondering why i tried to use some saltstack modules but this almos always returns an error saying that the module is unavailable. I'm using salt master ver 0.17.1 & salt minion ver 0.17.1
21:40 zach yep, use mock to build my other rpms
21:40 forrest zach, awesome
21:40 zach mock and/or rpmbuild
21:40 zach mock*
21:40 zach my eyes are broken, sorry
21:40 forrest yea you could always go really ghetto and just rpmbuild it for the time being
21:40 Ryan_Lane joined #salt
21:40 zach stressful morning/day
21:40 forrest since mock can sometimes be annoying
21:40 forrest yep
21:41 forrest Oscar, are you trying to include the salt module in a state? Or just from the command line?
21:41 Oscar state
21:41 forrest ok that's why
21:41 forrest modules can't go in states
21:41 Oscar I'm trying to use archive module
21:41 Oscar Mmmm
21:42 forrest Yea I don't think there is an archive state, or at least not one I can find.
21:42 rgbkrk_ joined #salt
21:43 Oscar why I can use mysql_user, mysql_database and mysql_grant from states? these aren't modules?
21:43 sroegner there is one in salt-contrib
21:43 cdcalef joined #salt
21:43 forrest Oscar, those are states so you should be fine
21:43 forrest sroegner, for archive?
21:43 Oscar Ok.
21:43 sroegner yea
21:44 Oscar Ho if I need to uncompress a zip files, there exist a way using states?
21:45 forrest Oscar, I'm not sure outside of the one that sroegner suggested in salt-contrib, looks like this issue is still open: https://github.com/saltstack/salt/issues/4449
21:46 forrest here's the salt-contrib one if you want to add it https://github.com/saltstack/salt-contrib/blob/master/states/archive.py
21:47 Oscar Can I add needed states from contrib to my current deployed salt master-minion?
21:47 AdamSewell Does anyone here work with salt-cloud? In my profile for digital ocean, I've specified that I want private networking enabled but it continues to create them without private networking enabled.
21:47 sroegner Oscar: that is the interesting part
21:47 sroegner I tried archive and it worked
21:48 finisherr_ joined #salt
21:48 sroegner but couldn't bring myself to depending on salt-contrib as a whole
21:48 zandy joined #salt
21:48 Oscar sroegner: How I do that?
21:48 finisherr_ Do the hostnames of minions have to be resolvable in order for the master to send it commands?
21:48 pdayton joined #salt
21:49 Oscar sroegner: What files do I need? and where I must copy to?
21:49 sroegner Oscar: there is some description on the GH page
21:50 sroegner i am not sure about the state or future of salt-contrib
21:51 dave_den finisherr_: no, the master publishes commands over the ZMQ pub socket and the minions look for commands that apply to them. by default it is by minion ID glob.
21:51 finisherr_ good news
21:51 finisherr_ dave_den: Thanks
21:51 dave_den np
21:52 forrest damn, I just got my hands on the Redhat satellite 6 documentation and all this crap is seriously married to puppet for config management, RHEL, WHY YOU NOT AGNOSTIC :(
21:52 finisherr_ Also, I
21:53 Oscar sroegner: there is a readme file, but no clue on how to copy it to my current salt installation, I don't know where are the files located in the master.
21:54 finisherr_ Also I am struggling between Puppet and Salt for OS X Desktop management. I know salt mostly caters to *nix servers, but I really need something a bit more full-featured then just Munki (although I will likely use this in conjunction). Puppet has providers to allows admins to install OS X packages right from the Puppetmaster. I was wondering if Salt has this type of functionality. I'm more of a Python guy, so I would very much like to use Salt. I've
21:54 finisherr_ gotten Puppet to work in some of the ways I want, but I like that Salt's ability to execute commands remotely on  many machines
21:54 pipps joined #salt
21:54 sroegner Oscar: i did that some time ago: I  believe the idea is to clone the repo side-by-side to your stuff and then run the relink script
21:55 sroegner I wouldn' do that in production
21:55 Oscar sroegner: my installation was made by a launchpad repo using a ppa
21:55 Oscar so I will try by following the instructions in the documentation to install modules
21:56 Oscar it's easy: http://docs.saltstack.com/ref/modules/
21:57 forrest finisherr, I know there are some modules for osx specifically, and the instructions for installing are here: http://docs.saltstack.com/topics/installation/osx.html I haven't tested it for doing package installations and such
21:57 dave_den finisherr_: salt's pkg provider for osx is homebrew
21:59 diegows_ joined #salt
21:59 finisherr_ dave_den: Yeah. I saw that. That is sweet. Except that I create custom packages all of the time that are specific for the needs of my organization, so to be able to deploy those packages with salt would be very cool.
21:59 cro finisherr_: you can get it from MacPorts now though also
21:59 pdayton1 joined #salt
21:59 ajw0100 joined #salt
22:00 jslatts joined #salt
22:00 dave_den finisherr_: it's definitely possible, but i don't think anyone has written a module to do install from things like dmg
22:01 dave_den you could write one :)
22:01 finisherr_ crap. well, i guess I could take a stab at that.
22:01 dave_den it's pretty easy to write modules
22:01 finisherr_ ha, you read my mind
22:02 finisherr_ so if I get one working should I just submit a pull request or something?
22:02 forrest finisherr_, yea I don't know what the rules are regarding unit tests for modules
22:03 forrest but pretty much, pull request
22:04 dave_den i guess this is how puppet installs an app from dpmg: https://github.com/puppetlabs/puppet/blob/master/lib/puppet/provider/package/appdmg.rb
22:08 Oscar sroegner: done I cloned salt-contrib and linked to my roots path and all just worked.!!! thanks
22:08 sroegner Osacr: cool
22:09 forrest sroegner, should update the docs with that :P
22:10 mo-man joined #salt
22:11 * mo-man slaps mo-man around a bit with a large trout
22:12 Ryan_Lane joined #salt
22:12 UtahDave joined #salt
22:15 _chjohnstwork joined #salt
22:16 z3uS joined #salt
22:16 mapu I'm curious why, when doing a pip install of the develop branch, that the salt version reported is v0.17.0 (0.17.0-4590-gb456fb9)?
22:17 mapu I;m trying to figure out how to keep salt-cloud and salt aligned so they play together nicely and so I can keep moving forward with salt.
22:17 thelorax123 joined #salt
22:18 mapu when i roll back to this version, I lose all pillar items
22:19 whiteinge Furao: ping
22:19 Furao whiteinge: i'm here
22:20 Furao for some reason (it's 6 am here)
22:20 whiteinge heh
22:20 mapu are there any steps I can use to get to salt v.0.17.2, and allow sallt-cloud to installm and work beyond the DictObject error?
22:20 UtahDave mapu: the develop version of Salt self identifies as the branch name, how many commits since the last tag and the first bunch of characters of the git commit hash
22:20 whiteinge Furao: i'd like to get your archive.extracted state into salt core. how do you feel about that?
22:21 mapu UtahDave: right. So - that I understand. Unfortunately, for some reason, when running that version- I have no pillar items. it;'s like they don't exist, so my entire configuration is broken.
22:22 mapu BUt- sallt-cloud will work
22:22 DanGarthwaite joined #salt
22:22 mapu if I do a git checkout tags/v.0.17.2
22:22 UtahDave mapu: do you get a stacktrace?
22:22 Furao whiteinge: no problem
22:22 Furao license is actually BSD in my repo
22:23 Furao (my updated repo)
22:23 whiteinge Furao: do you have a sec to send a pull request? (so you keep all the glory, fame, and adulation)
22:23 mapu then- my pillar items are there- but there is no salt-cloud. And on launching an ec2 provider profile, the launch breaks with a "Dict object expected, not "'testing has been created at 54.204.121.10'"."
22:24 mapu There is no stack trace
22:24 mapu I just get output showing my hosts
22:24 mapu separated by dashes
22:24 Furao whiteinge: I'll do that tomorrow, I was waked up by a client for an emergency few minutes ago
22:24 Furao whiteinge: patate@fastmail.cn
22:24 UtahDave mapu: if you use tags/v.0.17.2, then you should use salt cloud from its own repo
22:25 mapu Ok- I have an error in that scenario too
22:25 UtahDave mapu: in develop, Salt Cloud was merged into the main Salt repo
22:25 mapu one moment
22:25 mapu right- I saw that
22:25 UtahDave mapu: let me get redbeard in here to help
22:25 mapu what I am experiencing right now is I have no working scenario :(
22:26 whiteinge Furao: noted. thanks a lot
22:26 Furao or UtahDave got my personal private email
22:26 UtahDave Furao: Here, let me paste Furao's private email in here.  Just a second....
22:27 UtahDave lol,  just kidding!
22:28 Furao :)
22:28 mapu thanks UtahDave
22:29 diegows_ joined #salt
22:29 whiteinge UtahDave: need i remind you Furao is awake at the unholy hour of 6:00am working on an emergency? tsk tsk
22:30 UtahDave :)
22:30 schristensen joined #salt
22:31 Furao I wrote archive.extract state before salt.modules.archive exists
22:31 Furao last time I checked I couldn't use the module to do everything I did
22:34 bastion2202 joined #salt
22:34 whiteinge it's a very handy state. that workflow in vanilla salt is something of a pain
22:37 UtahDave mapu: Salt 0.16.4 will work with Salt Cloud 0.8.9
22:37 kermit joined #salt
22:38 ahammond joined #salt
22:38 mapu is there a version of salt-cloud that plays nicely with 0.17.x ? Just the merged version?
22:39 UtahDave let me ask
22:40 seventeenpoint2 joined #salt
22:40 mapu Ok. I have a substantial amount of time invested in using salt-cloud and salt that if I cannot currenty get the current version working together, I have to bail on the project for awhile till things stabilize.
22:41 seventeenpoint2 Greetings
22:41 DredTiger joined #salt
22:41 seventeenpoint2 I noticed 17.2 was released!
22:41 mapu I;m trying to sell it to my organization but this makes it difficult
22:41 seventeenpoint2 I'm not very familiar with how launchpad works though
22:41 seventeenpoint2 how long does it take between release and getting on ppas?
22:42 sroegner mapu: i am in a similar situation - am using salt:HEAD in a virtualenv to run salt-cloud out of there against a stock 0.17.1 master
22:42 Corey It is not yet released publically, though we appreciate your interest. :-)
22:42 mapu sroegner: yeah- that's what I was doing. Against HEAD though?
22:42 mapu By stock- pip or ppa ?
22:43 seventeenpoint2 Corey: what do you mean?  it says on docs.saltstack.org that the cur version is 17.2
22:43 sroegner EPEL :-)
22:43 nocturn joined #salt
22:46 pipps_ joined #salt
22:47 mafrosis joined #salt
22:48 zandy joined #salt
22:48 mapu Yeah- that confuses me as well- the "release" on the site says 0.17.2- but I can't get that AND salt cloud to work.
22:48 forrest Pssh this UtahDave guy, he's been MIA all day so now he's trying to answer all the questions so he can look good. We know your game!
22:49 mapu Hah
22:49 forrest mapu
22:49 UtahDave mapu: clone the salt-cloud github repo
22:49 forrest the docs are built from the dev branch
22:49 UtahDave mapu: git checkout b3b2ccf498d2f826a8ecb512c633528108bdfbc6
22:49 forrest so sometimes they have more up to date info than what's available
22:49 UtahDave mapu: python setup.py install --force
22:49 UtahDave that should work with Salt 0.17.x
22:49 mapu ok- let's try that
22:49 scott_w joined #salt
22:50 UtahDave mapu: that's the commit right before salt cloud was merged into Salt.  Hopefully that will work.
22:52 mapu [WARNING ] The cloud driver, 'ec2', configured under the 'tnr_ec2_east' cloud provider alias was not loaded since 'ec2.get_configured_provider()' could not be found. Removing it from the available providers list
22:52 mapu that;'s using that checkout
22:52 druonysus joined #salt
22:52 mapu quite the conundrum
22:54 cdcalef joined #salt
22:56 pass_by_value joined #salt
22:57 mwillhite joined #salt
22:57 pass_by_value left #salt
22:58 UtahDave mapu: can you open an issue on the pillar issue in develop?
22:58 mapu sure. I am rolling though the various install scenarios I have seen to verify tha they are not on my part
22:58 mapu and then will open an issue.
22:58 seventeenpoint2 in the pillar top file, can you target using PCRE?
22:59 Thiggy joined #salt
22:59 mapu right now- I think I have a fix- at least- a "get it to work fix"
22:59 mapu once I have my info and verify- I'll open an issue.
22:59 druonysus joined #salt
23:00 UtahDave thanks, mapu
23:01 sroegner using salt-cloud in ec2 i am trying to destroy an instance that has previous incarnations still listed as terminated - salt-cloud returns the current status as terminated when there actually is one running instance
23:02 sroegner is there a workaround for this?
23:02 foxx joined #salt
23:02 foxx joined #salt
23:03 UtahDave sroegner: there's an option in salt-cloud to have it rename the minion when you destroy it.  that way you can spin up a new instance with the original name immediately
23:04 sroegner UtahDave: excellent - I'll try that
23:04 halfss joined #salt
23:05 rgbkrk joined #salt
23:06 halfss joined #salt
23:10 Linz joined #salt
23:17 rgbkrk joined #salt
23:22 Nazca__ joined #salt
23:24 jslatts joined #salt
23:27 NV joined #salt
23:31 jslatts joined #salt
23:40 jslatts joined #salt
23:41 derelm joined #salt
23:42 exabytes18 joined #salt
23:43 exabytes18 Let's say I have sls A and sls B, sls A depends on sls B such that all states in sls B must be run before those in sls A. How do I represent this in salt?
23:43 pipps joined #salt
23:45 UtahDave exabytes18: you can have sls B include and require the entire sls A
23:45 exabytes18 How do I require the entire sls A?
23:47 UtahDave let me find the doc real quick
23:47 exabytes18 ah i got it, but it seems its a 0.16 feature
23:47 exabytes18 any way to do so with 0.15.3?
23:47 UtahDave http://docs.saltstack.com/ref/states/requisites.html#require-an-entire-sls-file
23:48 UtahDave exabytes18: in 0.15.3 you just would have to include sls A and then manually require each item in SLS A
23:48 zandy joined #salt
23:50 exabytes18 i so; doing so is problematic in terms of maintaining these sls files
23:51 exabytes18 do stateconf and include play nice, or does the combo break ordering?
23:51 UtahDave Yeah, that's why requiring the entire sls file was added
23:51 UtahDave I haven't used stateconf at all, so I'm not sure
23:52 Gareth anyone on a RH or have access to one? :)
23:53 exabytes18 utahdave: thanks for the info
23:53 Gareth ..with salt available.
23:53 UtahDave you're welcome, exabytes18. wish I had a better answer for you.
23:55 UtahDave Gareth: need to test something?
23:56 forrest joined #salt
23:56 Gareth UtahDave: wanted to see what the output of this looked like: salt '*' ip.get_interface eth0
23:56 UtahDave is a cent machine ok?
23:56 Gareth yeah. should work fine.
23:57 Guest87114 joined #salt
23:57 Benben joined #salt
23:58 Benben Hi there
23:58 UtahDave Gareth: Here's a cent5 machine: http://pastebin.com/8bU2hF2U
23:58 Gareth excellent.
23:59 UtahDave hey, Benben
23:59 halfss joined #salt
23:59 Benben I'm trying to run a master/minion conf for the first time, and i'm completely stuck with a "No Top file or external nodes data matches found" error. My salt-master is running on mac osx and my minion on a vagrant VM. It worked well when running in masterless mode, and i can run test.ping on my minion. I have absolute no idea where I should look from now.

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