Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-01-27

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

All times shown according to UTC.

Time Nick Message
00:00 UtahDave joined #salt
00:02 Ahlee just got the ping about that myself forrest
00:02 forrest did you get any creds Ahlee?
00:02 Ahlee nup. Just prompted to create on
00:02 Ahlee e
00:02 forrest but only via contacting the organizer right?
00:02 forrest that seems like a lot of work
00:02 forrest for them to do
00:03 Ahlee dunno. too lazy to create yet another web account
00:03 Ahlee i should really straighten all this out, given i still haven't opened the email from HR with flight/hotel info
00:03 forrest lol
00:03 forrest are you serious?
00:03 forrest what if you flew out today?
00:03 Ahlee they've never let me down before
00:03 Ahlee i know i fly out tomorow
00:03 forrest HR not letting you down?
00:03 forrest woah
00:04 Ahlee now you got me all worried
00:04 forrest I've been burned many a time :P
00:05 Ahlee though, with what's descending on us right now
00:05 Ahlee the likelihood of my flight being on time tomorrow is slim
00:05 Ahlee leave ORD at 5:55, arrive SLC 4:07
00:05 forrest yea terminalmage got stuck in Denver
00:05 Ahlee nope, that's not right
00:05 Ahlee artive 8:35
00:05 forrest I was gonna say
00:05 forrest oh you live in chicago as well?
00:05 Ahlee aye
00:05 Ahlee I'm actually in Chicago, though
00:05 Ahlee not like poser terminalmage
00:06 forrest ahh gotcha
00:06 forrest not the burbs
00:06 Ahlee yeah
00:06 Ahlee where you?
00:06 forrest Arizona
00:07 UtahDave joined #salt
00:07 forrest a toasty 70F here right now, almost had to wear my shorts today
00:07 Ahlee huh. shoulda opened this sooner. Seems i'm leaving thursday afternoon at 4
00:07 Ahlee oh well
00:07 forrest that shouldn't be a problem to make
00:07 Ahlee breakout and labs i can bounce on
00:07 JasonSwindle forrest:  kldshfksdhfkjsdhfjkshdfkjhksjadf
00:07 forrest trax is right there
00:07 JasonSwindle this is so bothersome
00:07 forrest JasonSwindle, did you talk to your debug duck yet?
00:08 JasonSwindle works in one SLS, fails in the other
00:08 forrest or take a walk
00:08 forrest walk = solution almost always for me
00:08 JasonSwindle same frackin code
00:08 Ahlee i love the debug duck
00:08 forrest Ahlee, yea me too
00:08 Ahlee I'm totally going to buy one for my desk when I transition off ops later this year.
00:08 forrest I won the one I have at some event at a museum of contemporary art
00:08 forrest so it's purple striped
00:09 Ahlee ha
00:09 forrest had to give it inquisitive eyebrows saying 'oh really?'
00:10 JasonSwindle One I get this, I am debuging with gin
00:11 Ahlee i'm looking forward to trying some local SLC breweries
00:11 JasonSwindle Me too. I love to see how locals make different brews
00:12 Ahlee alright, flight info straightened out
00:12 Ahlee now to verify i have a hotel
00:12 forrest Ahlee, still leaving at 4?
00:13 forrest it shouldn't be too bad, trax is right there by the venue, so you should only miss one session max.
00:13 Ahlee yeah, i'm not gonna bump
00:13 Ahlee judging off the agenda it's just a breakout session form 2-
00:13 forrest 2:15 is when it ends
00:14 forrest but I wouldn't risk getting to the airport with that little time
00:14 JasonSwindle Yep, that is why I bumped to the next day
00:15 forrest well, your talk goes till 2:15 as wel
00:15 forrest *well
00:15 forrest so that was just dumb of whoever scheduled the flight
00:17 JasonSwindle UtahDave:  Welcome?
00:17 andrej does salt have an easy method to override a value defined via an os grain to something else based on a host name?
00:17 Ahlee eh, if i don't take that one it looks like i'd be changing flights, and forget that
00:17 forrest he's away JasonSwindle
00:17 forrest no help for you
00:17 forrest seriously though, did you break the grains down yet?
00:17 JasonSwindle I see green :P forrest
00:17 JasonSwindle I did
00:17 forrest and confirm that a single item returns correctly?
00:17 JasonSwindle Yep
00:17 forrest ok so what happened after that
00:17 JasonSwindle and got it worked in the compound
00:17 JasonSwindle and with grains
00:17 munderwo joined #salt
00:17 JasonSwindle but in ONE sls
00:18 forrest oh that second one you updated your gist with?
00:18 JasonSwindle the same one I have been trying to debug for a frackin week
00:18 JasonSwindle Let me update with the new code
00:18 forrest tell me again why you don't just set this in pillar?
00:18 forrest instead of reusing the same chunk over and over again
00:19 JasonSwindle the JINJA set?
00:19 forrest yea
00:19 cast will saltconf presentations be available afterwards on youtube?
00:19 munderwo Hi all. Im attempting to install salt into an ubuntu 12.04 image. I've added the ppa "ppa:saltstack/salt" and done an apt-get update but when I try to install salt-master it says that "Depends: python-support but it is not installable" Do I need to install this prior to salt-master?
00:20 cast [i'd like to go but its 8000 miles]
00:20 bejer joined #salt
00:21 forrest munderwo, no you should be ok and it will pull in all the deps, joehh would be your man regarding the ubuntu packages.
00:21 munderwo forrest: thanks, thats what I thought.
00:21 forrest cast, I believe they are going to record some of the sessions. However if for some reason the sessions aren't available as video, I'm sure the slides will be (if they aren't I'll harass the presenters for them).
00:22 cast excellent :)
00:22 forrest munderwo, Do you remember off-hand if debian 7 = ubuntu 12?
00:23 forrest I never remember since I don't usually use them
00:23 munderwo hmm… nope sorry forrest
00:23 Fetch joined #salt
00:23 forrest munderwo, no worries. So you are just following through http://docs.saltstack.com/topics/installation/ubuntu.html I assume?
00:23 munderwo I know ubuntu 12.04 came out 2 years aho?
00:23 munderwo yup
00:24 munderwo the only complicating thing is that I'm trying to do it in a docker container :)
00:24 forrest ahh
00:24 Fetch I'm trying to use Salt 0.17.2 under FreeBSD 10, using pkgng. Is there any combination of module and provider that will work for that? https://github.com/saltstack/salt/issues/8738 implies that it might not be the case, but tracking develop branch isn't an option for me
00:25 forrest Fetch, I'm not sure. If you hang around for a while some of the freebsd users may show up though
00:26 Fetch thanks
00:26 * Fetch activates the lurk
00:26 forrest Fetch, yea np, it's more common they are around during normal US business hours monday-friday
00:27 forrest so you might try then (or on the mailing list) as well
00:27 cedwards Fetch: what kind of issue are you having?
00:27 forrest there we go
00:27 * cedwards should set a highligth for FreeBSD in this channel
00:27 forrest You should
00:28 Fetch cedwards: I've tried pkg.installed and pkgng.install, but I cannot get a combination that installs apache22 package. One minute and I'll pastebin some combinations if you need
00:28 cedwards have you set pkgng in the providers section of the minion config?
00:29 Fetch yes
00:29 Fetch I've also tried it without
00:31 JasonSwindle forrest:  https://gist.github.com/JasonSwindle/6b3b41644a7d91edfcbd
00:32 JasonSwindle This works
00:32 forrest yea but then you're hard coding it
00:32 JasonSwindle not really
00:32 taion809 joined #salt
00:32 JasonSwindle I am using grain[]
00:33 forrest JasonSwindle, did you try creating another sls that uses the same format that causes an issue for the that state?
00:33 cedwards Fetch: hmm. I don't have any 10 machines yet so i haven't run into this.
00:33 forrest and I meant for the [beaver] section, it just has an IP for redis_url
00:33 Fetch cedwards: http://pastebin.com/4GkRkJxk
00:34 JasonSwindle forrest:  working on that
00:34 JasonSwindle I use this in a mix of loops and one-offs
00:34 JasonSwindle refactoring with the cleaner code
00:35 JasonSwindle and also rebuilding nodes
00:35 JasonSwindle also, this error only seems to happen on a '*' HS
00:35 JasonSwindle vs a targeted node HS
00:36 cedwards Fetch: just to cover our bases, you restarted the minion after defining pkgng in providers?
00:36 Fetch I'm only runing it in salt-call --local mode
00:37 Fetch I guess I shouldn't've been trying to use "install" as a state in the file?
00:38 cedwards your state looked ok, in the using pkg: - installed
00:39 Fetch oh, got a better result pulling pkgng out of the providers (but still not good)
00:39 Fetch cedwards: well I also tried to set state of "pkgng" which is a different concept than "pkg", and tried to set state "install" which is a function and I should've used "installed"
00:40 Fetch but I'm brand-new at Salt, just concept differences that I missed
00:41 cedwards Fetch: any reason you're still on 0.17.2? 0.17.4 should be in the ports tree.
00:42 Fetch installing from package...this is part of an amazon cloudformation script, so I'm striving to avoid building ports
00:42 Fetch when the package updates (shortly I'm sure), I'll use it instead
00:43 munderwo joehh: hey have you got a second about ubuntu not installing salt correctly?
00:43 munderwo exit
00:44 cedwards I should setup a FreeBSD 10 box to help test.
00:45 cedwards i must have missed the 10 release announcement
00:45 Fetch when I have a valid pkg.conf in /usr/local/etc, my issue was that freebsdpkg was passing -L to avoid updating repos (should've passed -U). Since it's deprecated to use freebsdpkg (in 0.17.4), I'm going to figure out how to specify refresh=true for my salt state and call it good
00:46 Fetch cedwards: thanks for the help
00:46 yomilk joined #salt
00:46 forrest munderwo, joehh lives in Australia
00:46 forrest so he might be busy at work
00:47 munderwo forrest: ahh right.. I'm an Ozzie living in SF. Actually public holiday today. Australia day.
00:47 forrest munderwo, oh ok
00:48 munderwo forrest: I think its an issue with docker and salt. As I can install salt fine on the host system...
00:49 forrest munderwo, yea docker seems to be weird sometimes, I've seen others have issues as well, but not that one specifically
00:49 forrest if you can figure out whether it's a docker or salt issue, and if it's salt, create an issue and we can see if someone knows what is up
00:49 forrest as in salt's interaction inside docker.
00:50 munderwo forrest: Ahh cheers.. I imagine its some kind of arcane interaction between them :)
00:50 forrest heh, yea it's always crazy it seems.
00:52 JasonSwindle forrest:  state '*' state.highstate fails
00:52 forrest I'd hope so :P
00:52 JasonSwindle state '*web*' state.highstate works
00:53 forrest is that the only machine being targeted by the highstate?
00:53 JasonSwindle yes, but I only need to edit the map file and I could have many
00:53 JasonSwindle but yes, the second HS only picks up one node
00:53 forrest well, it's interesting that ti fails on a highstate even if that's the only machine
00:54 cedwards Fetch: I'm building a FreeBSD 10 VM now and I'll see if I can replicate your issue
00:54 JasonSwindle forrest:  I am stumped
00:54 JasonSwindle unless it is some type of rendering I just cannot get salt to show
00:54 forrest yea, I don't see why it would treat the pillar data differetly.
00:55 forrest or that render specifically
00:55 kalloc joined #salt
00:56 JasonSwindle forrest:        - "env: DB_HOST={{ salt['publish.publish']( postgres_compound_string, 'network.ip_addrs', 'eth0', expr_form='compound').values()[0][0] }}"
00:56 JasonSwindle fails on that on '*' HS
00:57 JasonSwindle works on targeted HS
00:58 pydanny joined #salt
00:59 Gordonz joined #salt
01:02 JasonSwindle forrest:  see why I am so bothered?
01:03 forrest lol yea
01:03 forrest Did you already create an issue on it?
01:03 JasonSwindle Yep
01:03 JasonSwindle waiting on Tom
01:04 forrest probably won't be fixed till after the conference
01:04 Fetch cedwards: as you wish...my issue itself is resolved to my satisfaction and any PR I could send would target a deprecated module (for pkgng case)
01:04 JasonSwindle and I am to open source this code by then
01:04 Fetch cedwards: (so that you don't think I'm waiting :)
01:05 UtahDave joined #salt
01:05 forrest JasonSwindle, I'd just do so with the caveat that it only works right now when you directly target, versus highstate
01:06 JasonSwindle yea….. lame. but yea "|
01:06 JasonSwindle :|
01:06 cedwards Fetch: what solution did you come up with? I'm getting a traceback trying to apply a state on my VM
01:06 forrest better than breaking your face on it all day.
01:07 JasonSwindle which I have been
01:07 JasonSwindle along with packing
01:07 Fetch cedwards: the freebsdpkg.py provider adds -L as an argument, that should be changed to -U
01:07 forrest yea finishing laundry now, actually about time to get some food...
01:08 munderwo forrest: worked it out. the default docker image for ubuntu doesn't come with universe enabled. grrr.
01:09 forrest munderwo, ahhh gotcha
01:09 forrest nice
01:11 Fetch cedwards: as far as I can tell, the issue should exist in 0.17.4, but not on the develop tree (because pkgng module provides pkg state)
01:11 puppet joined #salt
01:12 sroegner_ joined #salt
01:12 rojem joined #salt
01:18 toofer joined #salt
01:21 xzarth_ joined #salt
01:25 jimallman joined #salt
01:26 xmltok joined #salt
01:28 JasonSwindle forrest:  in fact, It will fail on a SLS that DOES work
01:28 JasonSwindle no matter what I put in
01:29 JasonSwindle going to keep digging
01:32 JasonSwindle like there is a cache messing with me
01:32 * JasonSwindle sigh
01:43 dangra joined #salt
01:44 toofer joined #salt
01:51 freeminds_ joined #salt
01:55 kalloc joined #salt
01:56 carmony joined #salt
01:56 kasanop joined #salt
01:57 ede joined #salt
01:57 Voziv joined #salt
01:59 bastion2202 joined #salt
01:59 davet joined #salt
02:00 tsantero joined #salt
02:00 MTecknology joined #salt
02:00 Hipikat joined #salt
02:00 MTecknology joined #salt
02:00 jssjr joined #salt
02:00 Ahlee joined #salt
02:00 hvn joined #salt
02:02 xmj joined #salt
02:04 freeminds joined #salt
02:07 cedwards joined #salt
02:09 ccase joined #salt
02:10 quickdry21 joined #salt
02:13 scuwolf joined #salt
02:16 robinsmidsrod joined #salt
02:19 munderwo hmm… is there an automated way I can get a salt-minion to connect to a salt-master automatically? how do people normally get that step working correctly for multiple images? do you just use the same key? so the minion looks like the same minion that was accepted?
02:20 bastion2202 Mine question again : I am able to get the IP of eth0 using : salt 'centos6.local' mine.get 'san.local' network.ip_addrs
02:21 bastion2202 but i cannot fin a way to get it using a for loop. I get : ['192.168.2.3']
02:26 mgw joined #salt
02:34 cachedout_home joined #salt
02:38 toofer munderwo, I don't have the link at my fingertips, but I just ran across a document that spoke of using a pre-shared key for authentication.
02:38 toofer A minion will still be identified by its hostname though.
02:39 munderwo toofer: thanks! I think I found the same one. thats good to know. wasn't sure how it got its "minion name"
02:39 munderwo toofer: I'm playing around with using salt on configure docker containers. I also ran across the salt docker documentation.. do you know anything about that?
02:40 toofer no, I'm pretty new myself.
02:42 yomilk joined #salt
02:44 cast i've setup a simple /etc/salt/minion file with just the master: and master_finger: variables, when each minion starts it generates its own set of keys
02:44 toofer In the walkthrough (http://docs.saltstack.com/topics/tutorials/walkthrough.html) I found this section about specifying or overriding the ID of a minion also. It states the following:
02:44 toofer Overriding the id
02:44 toofer The minion id can be manually specified using the id parameter in the minion config file. If this configuration value is specified, it will override all other sources for the id.
02:44 cast [did that on EC2 night before with 100 minions :)]
02:44 KinyobiWan1 left #salt
02:44 rojem joined #salt
02:46 ilbot3 joined #salt
02:46 Topic for #salt is now Welcome to #salt - SaltConf Jan 28-30, 2014! http://saltconf.com (registration extended! Hurry!) | 0.17.4 is the latest | Please be patient when asking questions as we are volunteers and may not have immediate answers - Channel logs are available at http://irclog.perlgeek.de/salt/
02:49 Dinde joined #salt
02:49 zhifan joined #salt
02:49 zhifan hi, need help.
02:49 cast yeah...was a bit of a weak point when i ran salt-key -A --yes
02:50 zhifan I like salt since it is so convenient.  but some team members think it is not safe.  since no logging for a given operations.
02:51 zhifan so do we have a config so that  every command will be executed like  "  ssh non-root@host.  sudo su - ;  command"
02:51 zhifan so do we have a config so that  every command will be executed like  "  ssh non-root@host.  sudo   command",  so every command will be logged in client machine
02:52 cast i was pondering logging via the event framework [i'm just starting too...so may be a bit naive]
02:53 toofer zhifan, I'm not sure it will give you exactly what you want, but you can enable debug mode on the minion
02:53 toofer In my near default minion config file I find it around line 401.
02:53 toofer 403 #log_level: warning
02:53 toofer Uncomment and change that line to debug and that may do what you want.
02:54 toofer When I run a salt command I also run it with the -v option (aliased in my .bashrc so it will always run this way). I think there is a way to also get back verbose job info. But I'm not that far in yet.
02:54 toofer Does that help zhifan?
02:55 zhifan_ joined #salt
02:55 kalloc joined #salt
02:56 toofer cast, what do you mean logging via the event framework? I think you are talking about a more advanced option in SaltStack called a 'returner' or there are also 'outputters'.
02:56 zhifan_ thank both of you.   ideally,  we can log which non-root user run this command.
02:56 toofer One of those is used to be able to output/update things like a DB, or an external application.
02:57 zhifan_ anyway,  debug should be helpful.  since it should be able to answer when,  but not who.
02:57 toofer Ah, I think what you are looking for is just in the sudoers config then. If you're on linux then you should be able to check /var/log/secure I think for that info.
02:58 zhifan_ yes, exactly
02:58 zhifan_ but currently salt use root directly,  so no info in /var/log/secure at all
02:58 cast neat @ returners
02:59 toofer oh, so just grant the salt command to your sudoer users. Would that work?
02:59 toofer Ideally no one should login as 'root' directly.
02:59 zhifan_ yes
03:00 zhifan_ are you join in this webchar in browser ?
03:00 zhifan_ it seems not stable
03:01 zhifan_ do we have some desktop tools for this char?  if so, maybe we can store the message in disk
03:01 sroegner_ joined #salt
03:01 robinsmidsrod joined #salt
03:01 mgw joined #salt
03:02 cast toofer: well, roughly we want to know who called what and what the results are with regards to commands from the salt master, not sure about the who but the what and the answer was captured by hooking into the event system on the salt master, though as mentioned - am new so that may be quite naive
03:02 forrest zhifan_, it's just IRC, there are tons of clients out there.
03:03 zhifan_ thanks,  got a xchar.
03:03 zhifan_ xchat
03:04 Gordonz joined #salt
03:06 toofer cast, so you want the results of every salt command logged somewhere then?
03:07 cast the results hold some value, but a bit less than the commands
03:08 cast but yeah, thats the jist of it
03:10 toofer I think you want a returner then. The returner could theoretically just be a call to the 'logger' command to log the command locally or to syslog server.
03:10 toofer I've never tried it though. I think this would be a great question to post on the forum though cast.
03:11 toofer here is the link to the group in case you don't have it yet.
03:11 toofer https://groups.google.com/forum/#!forum/salt-users
03:11 cast cheers, will check out returners and maybe make a post
03:11 toofer Cool
03:13 cast do returners have visibility of of, say, that someone called status.uptime? or just the results of the command?
03:15 snuffeluffegus joined #salt
03:15 toofer I think just the results of the command. A returner from what I know is a triggered event on the minion to update some other app/log/etc before actually returning the output back to the console output on the master.
03:16 dangra joined #salt
03:16 cast gtg, be back in a while :)
03:16 snuffeluffegus joined #salt
03:17 toofer http://docs.saltstack.com/ref/returners/all/salt.returners.syslog_return.html?highlight=returner#salt.returners.syslog_return
03:17 toofer that will log your results at least I think.
03:18 toofer Then the sudoer info could be logged on your master? A bit hokey and there is likely a better answer, but that's a start at least.
03:18 oz_akan_ joined #salt
03:19 DaveQB joined #salt
03:21 cast had a bit of trouble finding wallet....hopefully wont be using sudo on the master, will take advantage of the (quite nice) client ACL system
03:22 Nexpro1 joined #salt
03:23 forrest yea finding a wallet is always a bummer, lots of different shapes/sizes, not to mention considering traditional versus one of those plastic shells, or a money clip...
03:24 rojem joined #salt
03:28 Voziv joined #salt
03:30 tsantero joined #salt
03:30 robinsmidsrod joined #salt
03:33 ConceitedCode joined #salt
03:34 xmltok joined #salt
03:38 mgw joined #salt
03:41 Ahlee anybody got a clever way to pull the minion's master? I thought it was in pillars, but i'm not seeing it
03:41 Furao joined #salt
03:41 toofer salt-call grains master
03:41 toofer that's from the minion
03:42 toofer wait that command is wrong.
03:42 Ahlee no problem
03:42 Ahlee thanks
03:42 Ahlee grains.item master
03:42 Ahlee will do swimmingly, thanks
03:42 toofer no worries.
03:48 pydanny joined #salt
03:51 patrek joined #salt
03:52 robinsmidsrod joined #salt
03:53 yomilk joined #salt
03:55 justlooks joined #salt
03:55 kalloc joined #salt
03:59 robinsmidsrod joined #salt
04:19 ajw0100 I have a custom module I can call with salt-call no problem however I can't call it with salt from the master. Has anyone seen this?
04:20 sharoonthomas joined #salt
04:20 freeminds_ joined #salt
04:24 Furao ajw0100: you have a different config for the running minion than the salt-call command
04:24 Furao restart your minion to apply new configuration
04:26 Gordonz joined #salt
04:30 ubuntu__ joined #salt
04:33 ajw0100 Furao thanks looks like that did the trick. so even though the custom module gets synced down to the minion when i run state.highstate, i still have to restart the minion before it knows about it?
04:39 cast forrest: oh...no i mean i lost it so i couldn't go out to buy groceries. found, and food acquired :)
04:50 sroegner_ joined #salt
04:54 Furao ajw0100: you need to run saltutil.sync_all (or sync_modules)
04:54 Furao OR you had a different path in your minion file that to where the modules are
04:55 Furao so it needed a restart to apply
05:07 yomilk_ joined #salt
05:11 xmltok joined #salt
05:11 NotreDev joined #salt
05:18 forrest cast, ahh gotcha
05:26 munderwo joined #salt
05:34 ajw0100 Furao I am just putting the custom modules in _modules in the master roots dir and calling state.highstate, I thought that would be enough according to this: http://docs.saltstack.com/ref/modules/#modules-are-easy-to-write. I still need to run saltutil.sync_all though?
05:35 ajw0100 I can see that the module is being synced to /var/cache/salt/minion/extmods/modules/ when I run state.highstate
05:35 jeddi joined #salt
05:36 ajw0100 it just can't be called via salt '*' module.func until I restart the master
05:36 ajw0100 restart the minion* i mean
05:37 pass_by_value joined #salt
05:38 Furao restart the master?
05:38 Furao this is weird
05:39 ajw0100 i meant to say minion. sorry
05:42 ajw0100 so basically this works: salt '*' state.highstate && salt '*' saltutil.refresh_modules && salt '*' module.func
05:42 ajw0100 so i guess i'll go with that
05:42 rojem joined #salt
05:43 ajw0100 i just thought state.highstate would take care of refreshing the modules for me
05:46 yomilk joined #salt
05:50 pass_by_value left #salt
05:50 ConceitedCode joined #salt
05:51 ravibhure joined #salt
05:53 sgviking joined #salt
05:54 kalloc joined #salt
05:55 n8n joined #salt
05:57 kalloc joined #salt
06:09 matanya joined #salt
06:12 Psyke joined #salt
06:27 Gordonz joined #salt
06:29 ConceitedCode joined #salt
06:39 sroegner_ joined #salt
06:39 salt_ joined #salt
06:42 salt_ what format should be the package would be?
06:43 ChoHag joined #salt
06:55 strgcloud joined #salt
06:59 fllr joined #salt
07:04 sharoonthomas joined #salt
07:07 leonardinius joined #salt
07:16 Psyke Morning
07:17 yomilk joined #salt
07:18 tomspur joined #salt
07:20 crane hey ho, is it possible to install a mysql server with several users? my salt is complaining about my mysql_user section and i guess it's because i wrote two sections (for each user) also i would like to deploy straight after bringing the mysql server up a database. has someone an example for me?
07:25 cnelsonsic joined #salt
07:26 ConceitedCode joined #salt
07:34 ckao joined #salt
07:34 asp joined #salt
07:35 asp how to configure file to install many items
07:35 asp how to configure file to install many items on minion
07:37 kalloc joined #salt
07:38 Furao asp: changing your nick won't help, as long you don't read the doc you'll ask these questions
07:39 asp thanks furao for the concern, but i didn't get to know how to package
07:40 Furao use your distro packaging system
07:40 Furao or use tar ball and automate deployment using salt
07:42 kalloc_ joined #salt
07:45 drdran joined #salt
07:46 fllr joined #salt
07:47 yomilk joined #salt
07:53 jeddi joined #salt
07:54 Psyke joined #salt
07:57 agend joined #salt
08:02 gammalget joined #salt
08:08 balboah joined #salt
08:13 Shenril joined #salt
08:19 linjan joined #salt
08:19 gurubert1 joined #salt
08:21 iknite joined #salt
08:23 sharoonthomas joined #salt
08:27 njpatel joined #salt
08:27 Gordonz joined #salt
08:28 sroegner_ joined #salt
08:31 mlister2006 joined #salt
08:32 giantlock_ joined #salt
08:35 NV terminalmage: logged a few bug reports for the freebsdports module/state and one for pkgng ftr :)
08:35 helminthe joined #salt
08:35 zooz joined #salt
08:35 NV quite useful state that freebsdports one :D
08:36 aleszoulek joined #salt
08:36 Psyke joined #salt
08:44 fxhp joined #salt
08:45 babilen Hello - I am trying to find articles that describe the design of salt deployments for multiple customers (as in: each customer has a set of boxes with different services and customer data should only be available to them). Can you think of one? What are "best practices" to keep in mind when starting to use salt in such an environment?
08:46 fllr joined #salt
08:46 babilen While the docs are great I am a bit struggling with wrapping my head around the big picture and am sure that there are a number of established idioms and typical setups that I am simply not aware of. I'd rather read a bit more of the documentation than to code myself into a corner early on. :)
08:51 slav0nic joined #salt
08:54 sharoonthomas left #salt
09:04 iknite_ joined #salt
09:07 Furao joined #salt
09:07 harobed joined #salt
09:08 Gordonz joined #salt
09:09 iknite joined #salt
09:12 drdran joined #salt
09:17 MaikDerstappen joined #salt
09:21 yomilk joined #salt
09:23 aleszoulek joined #salt
09:23 NV babilen: keep all your customer data in pillar and then no issues? the salt states themselves shouldn't necessairly contain any specific data - just templated generic stuff with the specifics pulled from pillar
09:23 NV either that or one master per customer
09:26 MTecknology I keep a lot of specifics in states, but any repetition or anything like user stuff is in a pillar
09:27 MTecknology I've slowly started reworking salt to accomodate windows, centos, and debian boxes
09:27 MTecknology eventually I'll get to the point where pillars are much more heavily used
09:46 iknite joined #salt
09:46 fllr joined #salt
09:50 babilen NV: Yeah, that was my impression too, but it would simply help to read an explanation of a slightly larger setup to get a feeling for reasonable approaches. :)
09:52 malinoff babilen, I build a deb/rpm package for each project. So my states are simply - ensure a project's repository is present and install the latest package from it
09:54 nocturn What's the best way with salt to clean up old configs?  for example, how do I clean up vhost files when sites are removed?  I use a pillar with sites to geneatae configs.  But I would like an easy way to remove them
10:00 iknite joined #salt
10:00 NV nocturn: https://github.com/saltstack/salt/issues/7924
10:01 NV if your vhosts were able to be applied using file.recursive you would be able to use clean successfully
10:02 NV but i don't think that would work if you need to create N files with contents derived from pillar
10:08 nocturn NV, thanks, but that's indeed not the case for me
10:11 yota joined #salt
10:14 iknite joined #salt
10:14 che-arne joined #salt
10:16 sroegner_ joined #salt
10:17 NV nocturn: you could potentially combine all your vhosts into one file using a jinja template and a for loop which could solve the issue from the other way around :P
10:29 lessismore joined #salt
10:42 yomilk joined #salt
10:46 fllr joined #salt
10:54 Shenril joined #salt
10:55 jfxura joined #salt
10:59 iknite joined #salt
10:59 jfxura i am trying to write a state which deletes certain lines beginning with "vc" from a securetty file, i tried file.sed with options set to -r -e "/^vc/d" but i get errors like in http://pastebin.com/UbSCPzpG
11:01 jeddi jfxura: can you try escaping the quotes?  (speculating)
11:02 * jeddi didn't even know there was a sed primitive .. though I note that it's been replaced with 'replace()' .. which 'has identical params to replace()' according to the documentation (I guess they mean identical params to sed())
11:04 jfxura ya but how can i use replace() to delete the entire line
11:06 jfxura btw.. i fixed that file.sed state with an ugly looking hack, this works only if i set options to -r -e '/^vc/d' -e
11:06 jfxura with an orphan '-e' at the end
11:07 jfxura jeddi: i understand file.sed is deprecated, but can i delete lines entirely using file.replace() ?
11:08 jfxura i was only able to replace the matched strings and ending up with blank lines.. is there someway i can delete those lines too?
11:08 diegows joined #salt
11:08 jfxura using file.replace i mean
11:12 Iwirada joined #salt
11:14 iknite joined #salt
11:15 saltconf joined #salt
11:16 dangra joined #salt
11:18 stackfile joined #salt
11:26 SkyRocknRoll joined #salt
11:28 scott_w joined #salt
11:36 slav0nic joined #salt
11:46 fllr joined #salt
11:47 clemenko joined #salt
11:50 gadams999 joined #salt
11:51 clemenko joined #salt
11:58 beardo__ joined #salt
11:59 beardo__ joined #salt
12:03 Ryan_Lane joined #salt
12:05 sroegner_ joined #salt
12:06 Whissi joined #salt
12:07 beardo__ joined #salt
12:10 Mua_ joined #salt
12:15 krak3n` joined #salt
12:25 Mua joined #salt
12:27 alice|wl joined #salt
12:27 alice|wl hello
12:28 alice|wl I d like to use the lxc module but cant figure out how to call modules (other than state) in sls files
12:29 alice|wl I tried like this: http://pastie.org/8671751
12:32 bastion2202 joined #salt
12:35 sroegner_ joined #salt
12:37 Furao alice|wl: use state module.run (or module.wait)
12:38 elithrar joined #salt
12:39 babilen I'd like to use a different user for salt on the minions; Looks as if that is easily changed in the minion config, but what else needs to be done for it to work?
12:41 alice|wl Furao: thankyou :)
12:41 jrdx joined #salt
12:46 fllr joined #salt
12:47 gadams999 left #salt
12:49 viq babilen: a lot of things minion needs to do require root privileges
12:54 agend joined #salt
12:54 elfixit1 joined #salt
12:56 babilen viq: I take it that you are implying that "It has to run as root as acquiring the necessary privileges is not easily implemented for other users" ?
12:57 babilen viq: It could, for example, be feasible to grant the right to run certain commands via sudo without a password or there might be other ways to achieve this, but then it might also not be possible/easy
12:58 babilen viq: Background to this is, among other things, that we would like logs and commits (etckeeper) for changes to /etc to be easily identifiable as salt rather than a general "root" user.
12:58 viq babilen: yeah, it would require quite a bit of work to change that
12:58 babilen okay
12:58 alice|wl next thing is I have to generate the minion keys for the lxc container and autoaccept them on master. Maybe somebody has an example how to set this up?
13:00 kalloc joined #salt
13:03 kalloc_ joined #salt
13:03 swa joined #salt
13:03 anuvrat joined #salt
13:04 swa Corey aside, who's gonna be at Fosdem this week-end ? :-)
13:04 viq swa: me!
13:04 viq :P
13:04 swa viq: o/
13:06 taion809 joined #salt
13:06 viq babilen: I'm using a slightly different approach: salt pulls things _only_ from a git repo, so for salt to change any files, do any changes (well, via states, not via ad-hoc commands) they have to be in git in the first place
13:07 viq swa: and hopefully also to CentOS Dojo, Configuration Management Camp and Infrastructure.Next ;)
13:08 yota does anyone know the changelog for the last version?
13:08 gadams999 joined #salt
13:09 swa viq: i'm very excited about this year's fosdem.. Wietse Venema from Postfix is here, and hopefully the salt team :)
13:10 viq Yeah :)
13:10 Gordonz joined #salt
13:12 ChoHag joined #salt
13:16 oz_akan_ joined #salt
13:17 oz_akan_ joined #salt
13:18 Leyria joined #salt
13:19 babilen viq: Okay, that sounds sensible. As I'm quite new to salt and simply try to wrap my head around it: How does that fit together with pillars and running states? I mean package installations (for example) would still trigger changes in, say, /etc
13:19 blee joined #salt
13:20 viq babilen: yes, they would. But I don't think you'd know as such either way who had done those changes from just automated etckeeper
13:21 babilen viq: True - How does your setup work. Is there any documentation (could be long, I don't mind) that can read about that?
13:22 farra joined #salt
13:22 ConceitedCode_ joined #salt
13:23 NotreDev joined #salt
13:27 viq babilen: http://docs.saltstack.com/topics/tutorials/gitfs.html has most of what you need
13:28 ConceitedCode_ joined #salt
13:28 viq babilen: only trick is that salt master refreshes it's git cache every minute or so, so if you just commited the files and want it to see them _now_ you need "salt-run fileserver.update"
13:28 Ahlee i need to swing pillars into gitfs, then figure out how to get /etc/salt/master
13:28 Ahlee short of post receive hooks, which would suck
13:29 viq What do you mean, how to get /etc/salt/master ?
13:29 Ahlee the master config
13:29 viq I still don't know what you mean
13:29 Ahlee where you define number of workers, open_mode, pillar_roots, gitfs_roots, etc
13:30 viq You want to set master config from salt, or you want to know what to put in master config so that it sees git?
13:30 Ahlee what version of salt master?
13:30 Gordonz joined #salt
13:30 Ahlee no, i want to find a way to have salt's master config file in git :)  gitfs won't work as you can't tell it easily how to find the repo
13:30 Ahlee so it's two phased
13:31 viq My master is running debian and it just updated itself to 0.17.5, I've started in low 0.17.x with it
13:31 viq ah
13:31 Gordonz joined #salt
13:31 viq Ahlee: well, there's https://github.com/saltstack-formulas/salt-formula/ that sets a bunch of salt settings from pillars
13:31 Ahlee unless i multistage a master setup, but that sounds more painful than my manual 'run this script'
13:32 viq Ahlee: so you could just set an initial "here's git to look at" and then set stuff via pillars
13:32 viq that are in git ;)
13:32 Ahlee That's not bad.
13:32 juicer2 joined #salt
13:33 viq Mind you, I don't know how complete that formula is, I just looked at it from a somewhat high level a couple days ago
13:33 harobed joined #salt
13:34 harobed joined #salt
13:35 NotreDev joined #salt
13:35 harobed joined #salt
13:36 harobed joined #salt
13:36 babilen viq: Thanks!
13:37 harobed joined #salt
13:41 ConceitedCode_ joined #salt
13:42 tessellare joined #salt
13:44 AdamSewell joined #salt
13:44 AdamSewell joined #salt
13:45 sroegner_ joined #salt
13:46 fllr joined #salt
13:47 wkf joined #salt
13:48 brianhicks joined #salt
13:53 wkf joined #salt
13:55 Mua_ joined #salt
13:57 Ahlee viq: yeah, i pulled it down and sliced and diced it a bit, adding a monit wrapper etc
13:57 Ahlee but other than that, pretty stock
14:06 timoguin joined #salt
14:12 harobed_ joined #salt
14:14 ccase joined #salt
14:15 iri- joined #salt
14:26 Shenril joined #salt
14:27 fllr joined #salt
14:29 _fllr_ joined #salt
14:30 anuvrat joined #salt
14:32 imaginarysteve joined #salt
14:43 Beastcraft joined #salt
14:46 fllr joined #salt
14:50 mpanetta joined #salt
14:53 anuvrat joined #salt
14:55 Voziv What is the master configuration file for halite?
14:56 timoguin Voziv, the normal master config file: /etc/salt/master
14:56 Voziv ah I see
14:58 NotreDev joined #salt
14:59 che-arne joined #salt
14:59 rlarkin joined #salt
15:03 ConceitedCode joined #salt
15:04 jimallman joined #salt
15:04 vejdmn joined #salt
15:05 yota do you know why the documentation said to use __salt__ and not "import salt; salt.xxx"
15:06 timoguin yota, where specifically?
15:06 yota http://docs.saltstack.com/ref/modules/index.html
15:06 timoguin for custom modules and states, that dictionary is already made available
15:06 yota Salt modules can be cross called by accessing the value in the __salt__ dict:
15:06 timoguin http://docs.saltstack.com/topics/development/dunder_dictionaries.html
15:06 timoguin there are other dictionaries that are made available too
15:07 yota hey, it's a doc fighting :)
15:07 timoguin no sense in doing unnecessary imports
15:07 yota no sense, I can see a value: testing the module w/o  launching salt
15:07 yota no sense? I can see a value: testing the module w/o  launching salt
15:08 yota pure Python, so you call the file directly
15:08 timoguin perhaps. i suppose you can do that if you need to.
15:08 GradysGhost joined #salt
15:08 yota btw thanks for the pointer
15:08 timoguin i would trust the dunder dictionaries to change less frequently than salt internals though.
15:08 yota good point
15:09 Voziv How do I get halite to show me minions attached to my master?
15:10 timoguin Voziv, if you go to the minion tab it should do that automatically
15:11 timoguin pretty sure it runs 'salt-run manage.up' in the background
15:11 timoguin but then i think that shows minions that are up, not all the minions that have keys accepted
15:12 Voziv timoguin: When I run that command I get 3 servers. But on halite it doesn't show anything. Is there a group that my user should be part of?
15:12 Voziv (I run the command as root)
15:13 timoguin It's been a while since I tried Halite, and I don't use it now.
15:13 timoguin but that sounds like permissions
15:13 timoguin https://github.com/saltstack/halite#installation-quickstart
15:13 opapo joined #salt
15:13 timoguin you have to grant runner permissions to whatever user halite is running as
15:14 Voziv alright thanks
15:22 imaginarysteve joined #salt
15:30 iri- mount.mounted just gave me this error "Comment:   mount: /dev/mapper/vg0-spacefortangrepos already mounted or /var/run/tang busy mount: according to mtab, /dev/mapper/vg0-spacefortangrepos is already mounted on /run/tang"
15:30 rojem joined #salt
15:30 iri- it's already mounted.. is that a bug with mount.mounted?
15:32 cnelsonsic joined #salt
15:32 mgw joined #salt
15:32 ekristen joined #salt
15:33 ekristen I have startup_states: sls
15:33 ekristen and sls_list: - stateone - statetwo
15:34 ekristen they are not running on startup though
15:41 fllr joined #salt
15:43 mgw joined #salt
15:44 ekristen sroegner_, dave_den you here?
15:44 sroegner_ whatsup?
15:45 ekristen sroegner_: do you use startup states?
15:46 pdayton joined #salt
15:46 sroegner_ not yet
15:46 timoguin iri-, it looks like mount.mounted is working correctly based on that output. it's just giving you a warning that it's already mounted, right?
15:46 Ray____ joined #salt
15:46 pydanny joined #salt
15:47 iri- timoguin: it was an error
15:47 iri- timoguin: with red text
15:47 timoguin iri-, ah, that probably should be fixed then. mind opening an issue?
15:48 ekristen so startup_states: highstate works, but startup_states: sls doesn’t seem to work
15:48 timoguin that state should come back successful in that case, i think
15:48 iri- timoguin: that's what I expected.
15:48 sroegner_ ekristen: you having trouble with that? I was thinking about trying them this week
15:48 AdamSewell When i first started with salt, I used the Ubuntu PPA repo. It has kept salt up to date, 0.17.5 as of today, current but it has not updated salt-cloud (0.8.9). How can I update salt-cloud without messing up my master?
15:49 ekristen sroegner_: yeah, not sure if it is because I’m using salt environments or not
15:49 ekristen sroegner_: it guess it could be a bug in 0.17.5 too, as everything had been installed as 0.17.4 before now
15:50 forrest joined #salt
15:52 ekristen its really annoying
15:52 xmltok joined #salt
15:52 ConceitedCode_ joined #salt
15:53 timoguin AdamSewell, I install salt-cloud from pip.
15:53 linjan joined #salt
15:54 quickdry21 joined #salt
15:55 ConceitedCode1 joined #salt
15:56 [diecast] joined #salt
15:57 honestly haven't touched salt in two months
15:57 honestly coworker was like "hey could you install strace for me"
15:57 honestly > salt 'bb' pkg.install strace
15:58 honestly works
15:58 honestly :D
16:01 forrest heh
16:02 munderwo joined #salt
16:02 Guest56241 joined #salt
16:04 ekristen anyone using startup_states?
16:05 cachedout joined #salt
16:06 nahamu 2014.1.0rc3 Esky build for SmartOS is up.
16:10 quickdry21_ joined #salt
16:10 forrest rc3?
16:11 CheKoLyN joined #salt
16:11 forrest nahamu, I thought the only 'official' release was still rc1? rc3 isn't ready yet I thought
16:13 Psyke joined #salt
16:13 Furao joined #salt
16:14 jtang1 joined #salt
16:16 forrest I thought you'd appreciate that airline jab Corey :P
16:16 GradysGhost I'm having an issue where, even though I throw 'skip_verify: False' into my pkg.installed function, yum is still failing on the gpg check when it's applied
16:16 jalbretsen joined #salt
16:17 forrest GradysGhost, yea because you want skip_verify set to true
16:17 forrest if you want it to skip the check
16:17 GradysGhost Erp. Yes, the problem is that it's set to True and it's still failing
16:17 GradysGhost by bad
16:17 forrest by default it's already set to false
16:17 forrest oh ok
16:17 GradysGhost sorr
16:17 GradysGhost y
16:17 sroegner_ forrest: always a winner that one
16:17 GradysGhost ugh
16:17 * GradysGhost clearly needs a coffee refill
16:17 forrest no worries, what does your output look like on a -l debug?
16:17 forrest sroegner_, heh
16:17 GradysGhost Lemme check. It does log the gpg failure in the minion log without.
16:18 forrest weird...
16:18 forrest if that doesn't show anything, I'd say try a manual install, and see if something fails after the gpg check
16:18 kermit joined #salt
16:18 GradysGhost No, manual installs with yum --nogpgcheck succeed
16:18 forrest interesting..
16:19 forrest Is there seriously no way with this new org layout on github to have everything on a single page any longer?
16:20 nahamu forrest: dunno... it was tagged in git, so I built it.
16:20 forrest nahamu, ok, I think you probably shouldn't use that
16:20 forrest and should instead build RC1
16:20 forrest but you'd have to double check with the devs
16:20 nahamu rc1 build is already there.
16:20 forrest oh ok
16:21 forrest GradysGhost, I don't get it, the skip_verify is simply passing --nogpgcheck..
16:21 nahamu I skipped rc2 since rc3 already existed by the time I'd noticed there were new tags.
16:21 forrest gotcha
16:21 GradysGhost forrest: Debug produces no special output. http://www.hastebin.com/malinoduku.vbs
16:21 GradysGhost cphalo is the Cloud Passage agent, btw
16:22 GradysGhost From an external repo, hence the gpg failure
16:22 jergerber joined #salt
16:23 GradysGhost And yes, I checked the salt code on the minion, and it extends yum.YumBase and then passes "not skip_verify" as the "gpgcheck" option to its conf. So the salt code does look correct.
16:23 forrest GradysGhost, can you try to use the module real quick to see if it fails there as well?: http://docs.saltstack.com/ref/modules/all/salt.modules.yumpkg.html#salt.modules.yumpkg.install
16:23 forrest with a debug on there too
16:23 forrest see if we get the same thing
16:23 GradysGhost You're saying replace pkg.installed with yumpkg.installed, thereby bypassing the system translation part?
16:26 forrest when you run pkg.install from the command line it will use yumpkg
16:26 GradysGhost Maybe I'm doing it wrong. If I update my state from 'pkg.installed' to 'yumpkg.installed' I get 'State yumpkg.installed found in sls shared.cloud_passage.package is unavailable'
16:26 forrest yumpkg.installed won't work
16:26 forrest so I mean straight off the command line
16:26 GradysGhost If I run on the terminal:  salt 'mytarget*' yumpkg.install name=cphalos skip_verify=True
16:26 GradysGhost I get yumpkg.install is not available
16:26 forrest right, just use pkg.install
16:27 madduck joined #salt
16:27 madduck joined #salt
16:27 GradysGhost K, empty output
16:27 forrest -l debug?
16:27 * GradysGhost checks minion log
16:27 forrest blank output is lam
16:27 forrest e*
16:27 jeremyBass joined #salt
16:28 GradysGhost Well now it complains about the package not being available. Let me see if it actually got the repo itself installed.
16:28 forrest interesting
16:29 jankowiak joined #salt
16:29 * GradysGhost haz a stupid today
16:29 GradysGhost Wrong package name.
16:29 forrest oh no s huh
16:29 GradysGhost And still "Install failed: Public key for cphalo-2.7.9-1.x86_64.rpm is not installed"
16:30 bemehow joined #salt
16:30 forrest ok, can you crank logging to debug on the minion and restart the service?
16:30 forrest then try again
16:30 forrest because I have no idea why that would be failing.
16:30 forrest what release is that anyways?
16:31 GradysGhost That's been done already. I added -ldebug to my init script command options and restarted a while back. There is no additional debug output from that run.
16:31 GradysGhost salt-minion --version report 0.17.4
16:31 forrest hmm
16:32 GradysGhost I'll throw some extra output in the code and see if the value's getting swapped around somewhere.
16:32 forrest ok
16:33 andresr joined #salt
16:37 forrest GradysGhost, I just have a hard time seeing this as not having an issue created for it already for that release.
16:38 GradysGhost Well, I'm wondering if it's not a bug in your code
16:38 forrest GradysGhost, You know... is your repo file set to ignore the gpg key check for that?
16:38 GradysGhost But maybe a bug in the yum package
16:38 GradysGhost Hrm. No, I don't think it is. Will yum only ignore gpg if you put it in the repo file?
16:39 forrest I don't believe so, but I'm not 100% sure, it might be worth a shot to update, then try again with Salt
16:39 forrest maybe it's somehow doing a gpg check against the repo before, and failing out?
16:39 forrest and then doing a crappy job of loggin
16:39 forrest *g
16:40 GradysGhost Well, I turned it off, and it installed without even trying to check the gpg key
16:40 GradysGhost So I guess that's one solution.
16:40 forrest it's not really a solution though
16:40 GradysGhost I need to see what's wrong with Cloud Passage's repo, honestly
16:40 forrest they should have gpg keys
16:40 GradysGhost Why I don't have their keys\
16:41 forrest Regardless, it's still worth investigating. If you can confirm that another package from a repo other than that one fails to install only via salt with the gpg key missing, will you create an issue? That seems problematic.
16:42 GradysGhost I will if I see it again, but I've reviewed the code myself and it looks like you're handling it right.
16:42 UtahDave joined #salt
16:42 forrest GradysGhost, yea we can ask terminalmage if he ends up being around today, since he's much more familiar with the yum code.
16:43 sarkis joined #salt
16:43 sarkis hello salt community
16:44 * GradysGhost speaks for the entire community when he says, "Hello."
16:44 sarkis hah
16:45 sarkis gonna give salt another go, last i used it was ~0.10 and had some major leakage due to that 0mq bug
16:45 timoguin GradysGhost, you don't speak for me!
16:45 sarkis i'm mainly a puppet user and have been playing with ansible lately, but i think ansible is a bit too basic
16:45 MTecknology sarkis: salt keeps growing like it's a teen in a groth spurt. I think it's starting to finish with puberty.
16:46 * GradysGhost definitely speaks for timoguin. This is how this works.
16:46 sarkis MTecknology: feels like it, i'd be willing to help out with bugs / features to give back since it's "free" :)
16:46 sarkis thats bugfixes, i hope i don't add to the bugs :P
16:47 forrest sarkis, cool! This week might be a little light on the IRC since it's saltconf, so some of the regulars will be pretty busy.
16:47 sarkis i just noticed that
16:47 MTecknology I've found that salt is surprisingly easy to extend and add modules for. It's one of the things that makes me love it so much. When you run into a bug in core, though... that's a scary land of magic and master programming voodoo that's far beyond me.
16:47 sarkis wheres it at this year?
16:47 forrest salt lake city
16:47 sarkis oh nice
16:47 MTecknology sarkis: this is the first ever
16:47 forrest MTecknology, yea it helps that the contribs repo has some examples too
16:47 sarkis seems like your biggest offices are out there?
16:48 forrest their only office is out there :P
16:48 MTecknology I'd like to work for salt, but I'm definitely not a good enough developer.
16:48 GradysGhost I know a few folks who are going out to Saltconf (JasonSwindle is one of them).
16:48 sarkis oh i thought i saw something for Raleigh NC
16:48 GradysGhost I'm too busy to go...
16:48 MTecknology I'm still on the registration. I'll be there in spirit and tears.
16:48 sarkis now that i think of it, that might be ansible guys
16:49 forrest GradysGhost, are you a racker as well?
16:49 GradysGhost yerp
16:49 MTecknology ansible seems pretty solid and well thought out, but I'm pretty far from a fan of it.
16:49 sarkis any good snow in the mountains of SLC? i'm a snowboarder in los angeles, couldn't be more out of place
16:49 TheRealBill_here joined #salt
16:49 forrest for remote execution ansible is pretty great.
16:49 forrest sarkis, yea there is plenty of snow
16:49 sarkis MTecknology: i just think i am not the target audience of ansible
16:50 forrest I was considering staying through the weekend to go ski/snowboard.
16:50 sarkis i kept seeing 'ansible apeals to IT' and after using it, i can see why :P, almost 0 programming experience required
16:50 sarkis i guess salt maybe the same can be said if you just use yaml
16:50 toastedpenguin joined #salt
16:50 forrest you still need some jinja in there, but it isn't too bad
16:50 sarkis damn now i wish i knew about SaltConf sooner ;P
16:50 forrest hah
16:51 toastedpenguin left #salt
16:51 Ryan_Lane2 joined #salt
16:51 sarkis chances are my employer wouldn't buy me in but i'd just do out of pocket and get some snowboarding in too
16:51 colinbits joined #salt
16:51 forrest yea I'm flying out there this afternoon, hopefully the weather will be nice in SLC
16:51 MTecknology forrest: can I let him in on my rant? I'll keep it short.
16:51 toastedpenguin1 joined #salt
16:51 forrest MTecknology, no
16:51 forrest no more ranting
16:51 MTecknology ok
16:51 sarkis lol
16:52 sarkis pm me
16:52 forrest oh good, I looked at the temperature and saw 1 degree for SLC
16:52 forrest then had to realize it was still set to C
16:52 sarkis hah
16:54 forrest sarkis, here is a good video if you're not that familiar with salt: http://www.youtube.com/watch?v=yphLKSjnSU8, it's a bit out of date, but not TOO bad.
16:55 forrest Tom just goes through all sorts of stuff there.
16:55 toastedpenguin2 joined #salt
16:55 sarkis ah i remember thomas from archlinux puppet days haha
16:56 cachedout It was pretty warm this weekend here in SLC, actually. We'll see a pretty solid storm coming in Wed/Thurs but mostly in the mountains. Skiing/snowboarding this weekend will be solid, for those who enjoy that sort of thing. :]
16:56 sarkis all the puppet pull requests to get archlinux support which didn't go anywhere and i think salt was born :P
16:56 toastedpenguin2 left #salt
16:56 MTecknology he's an incredibly brilliant guy that hires incredibly brilliant people
16:56 forrest man MTecknology, what a brown noser... :P
16:56 sarkis lol
16:58 MTecknology forrest: nah... I've just worked with a few of them including thomas to try to debug some weird stuff and I end up (with their direction) pointing them to the section of code that breaks and explain how it breaks and all of a sudden they have a patch that fixes it and I don't even get what they changed, even though I'm seeing the patch.
16:59 msciciel_ joined #salt
16:59 Ryan_Lane2 joined #salt
16:59 CheKoLyN Hello All, Do you know If reactors work over salt-ssh ?
16:59 bretep I'm setting up a new salt environement, migrating away from Chef. Does anyone have a good process maintaining muliple environements with salt? dev,staging, prod. I'm interested in learing how you merge or promote config from dev->staging->prod. At my last company we just had one main branch, 3 mastersers 1 master for each environment and then pulled in changes to each master when it was time to upgrade.
16:59 bretep In this new environment we have 1 salt master managing all environments
16:59 MTecknology CheKoLyN: some of it will, but it will be limited since there are (i believe) fewer events to work with
17:00 forrest MTecknology, I'm just messing with you
17:00 MTecknology forrest: I know, but I wanted to back it up
17:00 forrest gotcha
17:01 UtahDave CheKoLyN: Also, when using Salt-ssh there isn't the normal Salt event bus to send messages back to the master.
17:01 anuvrat joined #salt
17:01 UtahDave MTecknology: It's true, though. Working here at SaltStack is like working on a Graduate CS degree every day.
17:01 MTecknology bretep: there's a bunch of stuff for working with testing environments like that. They're supposed to be pretty solid and you can keep all of your configs in git and just use different branches. I've never done anything like that, though.
17:02 UtahDave bretep: Look here at gitfs   http://docs.saltstack.com/topics/tutorials/gitfs.html?highlight=gitfs
17:02 Gareth morning
17:02 UtahDave morning, Gareth!
17:03 Gareth UtahDave: hey :) hows it going?  All set for SaltConf?
17:03 UtahDave Pretty much!  We're just doing some final preparations and checks.
17:03 Gareth nice.
17:03 Gareth sorry to be missing it.
17:03 UtahDave Yeah, wish you could have made it!   How's SCALE coming along?
17:03 CheKoLyN Well, I"m actually reacting to a salt-cloud event; but I wanted to modify other systems with salt-ssh
17:04 CheKoLyN I think it might be possible
17:04 Gareth UtahDave: coming along good.  less than a month away.  still the calm before the storm :)
17:07 MTecknology UtahDave: any chance of recordings being available for people that signed up?
17:07 UtahDave CheKoLyN: It looks like Salt-SSH not hooked directly into the reactor yet. You can probably use state.sls runner from the reactor to call out to salt-ssh.  checking now..
17:07 UtahDave Gareth: cool.
17:08 MTecknology I know there was talk about it before but didn't really follow it.
17:08 UtahDave MTecknology: Yeah, we're going to be recording all the talks.
17:08 MTecknology :D
17:08 forrest Nice
17:08 timoguin yay
17:08 UtahDave I'm not sure exactly when they'll be available, but we'll get them out as soon as we can.
17:08 mpanetta joined #salt
17:08 MTecknology eventually is perfectly fine by most of us
17:09 timoguin As someone who hasn't contributed any money, I will be very upset if they're not up within 4 hours!
17:09 timoguin I am very entitled!
17:09 MTecknology eventually in salt usually means next hour or so, but I think we can all handle a little wait time :)
17:09 UtahDave timoguin: are you speaking at a conf soon about Salt? pytennessee or fosdem or something?
17:09 timoguin yea, PyTN
17:09 UtahDave cool!
17:10 forrest that's the only reason he's been hanging out, that free knowledge, what a cheapskate.
17:10 timoguin haha
17:10 MTecknology I want to read every page of the salt docs and do something to practice all of it
17:10 timoguin I'm going to be heading up a sprint the whole weekend too.
17:10 forrest there's a lot of pages man
17:10 forrest nice
17:10 MTecknology in timoguin's defense... he has actually helped out in here
17:10 UtahDave that's awesome, timoguin.  I wish I could be there!
17:10 MTecknology not entirely cheap..
17:10 forrest yea I know
17:11 MTecknology he's kinda has pretty lips too
17:11 * MTecknology winks at timoguin
17:11 * timoguin blushes
17:11 timoguin just kidding, i don't blush
17:12 yomilk joined #salt
17:12 MTecknology ooooh.... this channel makes me happy; I feel like it's a good representation of the company and what they offer. It takes a pretty awesome company to be able to foster a community like this. ... even if some of them are *hack* rhel users
17:13 ConceitedCode joined #salt
17:15 Iwirada left #salt
17:15 forrest woah there buddy
17:16 MTecknology woah?
17:16 forrest there's nothing wrong with rhel based distros
17:16 MTecknology :P
17:16 forrest or at least, no more than any other distro
17:16 timoguin pretty sure he's making fun of himself. ;-)
17:16 Nazzy [17:07:20] <MTecknology> UtahDave: any chance of recordings being available for people that signed up? <-- how about for those of us too poor to get to these awesome cons :p
17:17 Nazzy forrest, nothing right with them either *hides*
17:17 forrest Nazzy, Don't worry I'm sure someone will pirate them!
17:17 timoguin Nazzy, they'll publish the talks online afterwards
17:17 forrest timoguin, I think he wasn't sure if they'd be free or not
17:17 timoguin ah
17:17 forrest some conferences require that you log in with your creds to watch videos
17:18 timoguin I think that's silly.
17:18 Nexpro joined #salt
17:19 Nazzy that was, indeed, what I was querying ... I know far too many sites that paywall interesting stuff
17:19 forrest Nazzy, are you trying to discount this RHEL cert I have? It's a piece of paper, and it garners respect! :P
17:20 mpanetta Hrm... I thought ubuntu 13.10 had the latest salt (0.17.4) in its repos?
17:20 forrest mpanetta, I thought that it did
17:20 forrest pretty sure joehh built that
17:21 mpanetta The cloud images must have the wrong repos set up...  It installed 0.16.0!
17:21 Nazzy I have this cert that says my penguin of choice rules the world ... ignore that it's written in crayon ¬_¬
17:23 ashtonian joined #salt
17:23 Nazzy to be honest, my only gripe about RHEL and Fedora based distributions is the mess they make in /etc
17:24 mpanetta Ehh?  What mess?
17:24 UtahDave Nazzy: I'm pretty sure we're going to make all the recordings available. I'll have to check
17:25 Nazzy I primarily interact with 3 distributions on a regular basis... Gentoo, Ubuntu and CentOS ... compare the network configs on all three :p
17:27 troyready joined #salt
17:27 timoguin I miss Gentoo. That's what I learned on, but haven't really used it much since Ubuntu 4
17:28 timoguin the stack of free installer CDs got me
17:28 Nazzy *considers dates* it's way easier to install gentoo these days
17:28 Nazzy although I do kinda miss stage 1 installs, they're a fantastic way of learning the system
17:28 timoguin yea I've done it since.
17:28 timoguin I really loved Portage.
17:28 gadams999 joined #salt
17:29 Nazzy indeed
17:29 Nazzy when you have a binhost as well, the main drawback of portage goes away
17:30 * timoguin nods
17:30 timoguin Thinking back to my e17 overlay days and my huge CRT monitor. :)
17:31 timoguin I'll have to spin up some Gentoo servers again eventually.
17:32 Nazzy with salt *nod*
17:33 timoguin of course. looks like a Gentoo AMI might be the easiest way to get going with it again...
17:33 timoguin things these days...
17:35 Nazzy yay for cloudy penguins
17:36 wkf joined #salt
17:37 Ryan_Lane joined #salt
17:37 MedicalJaneParis joined #salt
17:38 Nazzy at this rate we're gonna have to retag as 2014.2 lol
17:39 bretep While gitfs looks great! How do my developers develop on a dev branch without a top.sls file
17:39 bretep What if they want to add a state to the top.sls file that is not ready for production?
17:39 bretep Why would all the top.sls files be merged?
17:40 rojem joined #salt
17:40 * bretep is trying to understand this logic.
17:40 gadams999 left #salt
17:40 amckinley joined #salt
17:41 GradysGhost bretep: Maybe I'm misunderstanding your concerns, but the top.sls file defines the state of a server by making reference to lots of states defined elsewhere.
17:42 bretep Yes
17:42 GradysGhost If you want to test changes to your salt config, you should probably store it in a VCS and have separate environments where you can test different versions/commits of the salt config.
17:42 GradysGhost That way you don't have to test salt in prod.
17:42 bretep I'm referencing http://docs.saltstack.com/topics/tutorials/gitfs.html?highlight=gitfs#branches-environments-and-top-sls-files
17:43 Nazzy bretep, I follow what you're getting at, though I'm not sure how to better phrase it
17:43 mgw joined #salt
17:43 Nazzy bretep, proper solution is to fork and use pull request style behaviours
17:44 davet joined #salt
17:44 bretep I'm talking for local development testing
17:44 GradysGhost Reading that, I understand the concern. Why would you want to merge different versions of the top file?
17:44 joehillen joined #salt
17:44 Nazzy bretep, basically have a dev repository instead of a dev branch
17:44 bretep I'll send an example
17:45 UtahDave bretep: Yes, all the sls files get merged, but each environment is still kept separate
17:45 timoguin bretep, I generally test states by themselves before putting them into top.sls, using state.sls
17:45 timoguin I believe you can have a top file in each environment if you want, but those top files can only target the environment they're in.
17:46 jcockhren timoguin: e18 is out now. ;)
17:46 timoguin jcockhren, oh lordy....
17:47 gnu-man joined #salt
17:47 toofer joined #salt
17:48 bretep So an example top.sls http://pastebin.com/pypVQ3BE
17:48 bretep Well that wasn't valid yaml
17:48 bretep lol
17:48 viq Huh, what the hell. Same invocation of RVM worked for me in vagrant but doesn't on a "real" machine
17:49 robbyt joined #salt
17:49 bretep Say I am in dev and want to remove example1 and add an example3 to the base
17:49 mpanetta Is salt 0.17.5 the RC that was floating around as 2014.1.x?
17:50 bretep since gitfs merges top.sls in dev, staging and master... I'd be screwed in production
17:50 forrest mpanetta, I don't believe so
17:50 mpanetta Ah ok, just wondering.
17:51 Nazzy I have noticed a few commits getting picked on to the 0.17 branch, so my guess is that it's a bug fix release only
17:51 bretep UtahDave does that make more sense?
17:51 timoguin bretep, I don't use the same hostnames for my dev servers as my prod servers. I'm using web001.foo.dev for dev web servers, for example.
17:51 bretep I do as well.
17:51 timoguin So the section that matches prod servers doesn't get touched.
17:52 bretep but http://docs.saltstack.com/topics/tutorials/gitfs.html?highlight=gitfs#branches-environments-and-top-sls-files says that the top.sls from each branch is merged into master
17:52 bretep and every environment gets it.
17:52 Kenzor joined #salt
17:52 bretep it's not specific to any environment
17:52 bretep that's the problem I'm having
17:53 jslatts joined #salt
17:53 Nazzy is there a reason you can't define a separate entry?
17:53 bretep meaning '*web*dev*'
17:53 Nazzy http://pastebin.com/EzpSZu30
17:53 bretep ?
17:53 Nazzy yea
17:54 bretep then what is the point of having separate environments?
17:54 bretep Nazzy, I'm trying to understand the logic of merging the top.sls from all the branches into one.
17:55 bretep Nothing else is merged, which si perfect!
17:55 bretep s/si /is /
17:55 Nazzy on that I can't answer, since I don't use them and I don't actually define my states this way, but my understanding of the point of environments is so that you can have overlaps of states
17:56 timoguin bretep, maybe something more like this: http://pastebin.com/UpdZEUqM
17:57 timoguin I still haven't really dove into using multiple environments with gitfs though
17:58 bretep Seems like I shouldn't have to basically designate environments twice...
17:58 UtahDave bretep: The top file only assigns the minions to specific environments
17:58 minaguib joined #salt
17:59 timoguin bretep, in the top file you posted, there is only one environment specified: base
17:59 bretep Correct
17:59 bretep With gitfs environments are defined by branches
17:59 bretep http://docs.saltstack.com/topics/tutorials/gitfs.html?highlight=gitfs#branches-environments-and-top-sls-files
17:59 wkf joined #salt
18:00 diegows joined #salt
18:00 timoguin they still need to be defined in the top.sls, otherwise your minions are all in one environment
18:00 bretep but each branch need's their own top.sls file.
18:01 Nazzy the branches define which states are in the environment, the top.sls defines what minions are in the environments (and what states they need)
18:01 timoguin the top.sls is what tells each minion what environment it belongs to
18:01 bretep I think I get it now. in my top.sls file for dev, I'll only target dev servers
18:02 bretep then the master will combine all the top.sls file
18:02 bretep s
18:02 bretep The documentation is very confusing.
18:04 bretep That wouldn't be a great workflow either...
18:04 bretep Editing should only be done in dev branch, everthing else should be merge
18:05 Ryan_Lane2 joined #salt
18:05 timoguin That's why I'm planning on using just one top file in the master branch.
18:06 timoguin and it defines matches for every environment
18:06 Voziv Is there a reason why state.highstate always returns with a status of 2?
18:07 Nazzy it's why I said the proper solution is to fork to a dev repo and do pull requests :p
18:08 Ahlee huh, i did not relaize this was a multi track conference
18:08 bretep timoguin, that's what we'll be doing
18:09 bretep one top.sls
18:09 Nazzy Ahlee, makes it way easier to mix later if do multi track, was a pain when you were on reel to reel but much easier with digital mixers </bad joke>
18:09 matanya joined #salt
18:10 forrest Ahlee, lol, I feel like you are just now opening all the emails
18:10 Ahlee true, but we're only sending me to the conference based on the info that was up on saltconf, guess I should have asked
18:10 Ahlee forrest: well, i am :)
18:13 SpeeR joined #salt
18:13 Xiao__ joined #salt
18:14 KyleG joined #salt
18:14 KyleG joined #salt
18:14 gadams999 joined #salt
18:16 viq Bah, any pointers to figure out why same state using RVM worked in vagrant but doesn't see the ruby on a "real" machine?
18:18 jeremyfelt joined #salt
18:20 forrest viq, is 'because it is RVM' not a legit reason?
18:20 jcockhren forrest++
18:21 fridiculous joined #salt
18:22 Nazzy people**2
18:22 Nazzy power to the people yo
18:22 MTecknology groups = [g.gr_gid for g in grp.getgrall() if user in g.gr_mem]  <-- I'm trying to understand exactly what this does...
18:23 Nazzy MTecknology, I was saw that exactly what that does is violate readability :p
18:23 Nazzy but it's not actually that bad
18:23 viq forrest: eh, and I thought it would be a much easier way to get ruby since it needs to be compiled anyway...
18:24 Nazzy for every g in grp.getgrall(), if user is in g.gr_mem it adds g.gr_gid to the list that is assigned to groups
18:26 MTecknology it seems like there should be a MUCH more efficient way to do that... for every group, go through every user, see if they're a member of the group
18:26 Nazzy that's how I read it at any rate ... it's a rather unfortunate bit of code
18:27 Nazzy no, user is static, I'm assuming it's assigned somewhere above
18:27 forrest MTecknology, it doesn't seem to go through every user, just checks if the user is in that group.
18:27 Nazzy basically it's getting a list of gid's a given user is a member of
18:27 forrest yep
18:28 MTecknology ah
18:28 MTecknology don't groups contain that information?
18:28 Nazzy apparently not
18:29 aleszoulek joined #salt
18:29 troyready joined #salt
18:30 alunduil joined #salt
18:30 fridiculous this is embarrassing, but i am still having issues running upstart conf through service with salt
18:30 MTecknology I disagree...
18:30 Nazzy in /etc/group it only lists the members of a group, /etc/passwd gives the primary group ... supplementary groups are figured like that
18:30 MTecknology I just tried to run it. It seems that grp.getgrall() does indeed have users in it
18:31 MTecknology Is this collecting that, though? If it is, I'm missing it
18:32 Nazzy this is getting the list of groups a user is in, not the users that are in a group
18:32 davet joined #salt
18:32 MTecknology is there any reason to do it that way instead of just getting members of groups?
18:33 MTecknology I'm a novice here, but I did track down a massive headache to this section
18:33 MTecknology well... to this exact line
18:35 Nazzy the members of a group is different ... this is like `id -G`
18:35 viq Ah, no, when starting afresh on test one doesn't work either
18:35 xmltok joined #salt
18:38 MTecknology Nazzy: doing a lookup on every user to find the groups there in is causing me ~3hr state.highstate execution time; if I can use just what's returned from the group membership, I'd probably drop that down to ~15min. That's why I'm trying to understand this bit to see if it can be reduced.
18:39 steveoliver it's 'require_in', not 'requisite_ins', right? (2nd and 4th paragraph)  Ã¢â‚¬â€ http://docs.saltstack.com/ref/states/requisites.html
18:39 Nazzy steveoliver, bad phrasing, but accurate
18:40 MTecknology steveoliver: I believe you are correct
18:40 * steveoliver would like to issue a PR :)
18:40 steveoliver also, got a question — can i include states from other states that aren't explicitly included in the top file?
18:40 Nazzy 'requisite_ins' is referring to a generic term for require_in, watch_in, in plural
18:41 steveoliver mmmm
18:41 MTecknology oh
18:41 MTecknology PR to make that more clear?
18:41 steveoliver thx nazzy that makes sense…  just needs a little clarification and/or styling
18:41 steveoliver MTecknology: whatever it takes to get an edit in that makes that a little more clear
18:41 Nazzy possibly better parsed, but much less English, form might be <requisite>_in's
18:42 steveoliver where should/can i create an issue?
18:42 steveoliver to flesh it out other than here/now?
18:42 Nazzy github issue :)
18:42 steveoliver sure, cool
18:42 MTecknology github.com/saltstack/salt
18:42 Nazzy yup, there ^
18:42 steveoliver so, previous Q: "can i include states from other states that aren't explicitly included in the top file?"
18:43 MTecknology for g in grp.getgrall() \n\t print g.gr_mem()  <-- this is exactly my issue... now... to find some sort of resolution
18:43 MTecknology hrm......
18:44 Nazzy steveoliver: so top.sls includes a.sls, and a.sls includes b.sls, but b.sls isn't in top.sls?
18:44 steveoliver yes
18:44 steveoliver nazzy^
18:44 Nazzy that's fine, yes
18:44 steveoliver k, cool…  thought so...
18:44 forrest steveoliver, can you open an issue regarding those docs? and that being confusing?
18:44 fllr Hey guys. How can I download a s3 directory using a salt state?
18:44 steveoliver just wondered real quick if /that/ was "the issue" i'm having...
18:44 steveoliver forrest, i will...
18:44 forrest thanks, I'll probably snag it today before I fly out, but just in case I forget.
18:45 Nazzy MTecknology, where's that nasty list comp living?
18:45 steveoliver you have an idea for text, or you want me to suggest?
18:45 steveoliver forrest^
18:45 n8n joined #salt
18:46 MTecknology Nazzy: https://github.com/saltstack/salt/blob/develop/salt/modules/groupadd.py#L104
18:46 forrest steveoliver, Well, it's trying to explain those ideas, but the page is just too confusing really. I'll probably make a comment just saying (require_in) or something along those lines, and then propose a full page rewrite to reduce complexity.
18:46 steveoliver forrest: k
18:47 forrest steveoliver, if you have something to propose, go for it!
18:47 steveoliver not now, (not today), but maybe...
18:47 forrest ok
18:48 andresr joined #salt
18:48 [diecast] joined #salt
18:48 Nazzy MTecknology, I meant the ugly list comprehension you pasted before
18:49 gadams999 left #salt
18:49 Nazzy that for loop is puzzling in other ways ... breaking out to a function for that little code is horrible from a performance stand point
18:50 MTecknology Nazzy: which one?
18:50 MTecknology this? -->        groups = [g.gr_gid for g in grp.getgrall() if user in g.gr_mem]
18:50 Nazzy yea
18:50 MTecknology sorry... I gave you the wrong link - https://github.com/saltstack/salt/blob/develop/salt/utils/verify.py#L196
18:51 steveoliver forrest: https://github.com/saltstack/salt/issues/9974
18:51 forrest great!
18:51 Nazzy that specific line is taking a long time to run? it shouldn't be enumerating users at all, only groups
18:52 MTecknology yup, it takes ~12min to run
18:52 MTecknology If I do strace, I see a lookup being run on every single user
18:52 Nazzy how long does `id -G` take?
18:52 MTecknology https://github.com/saltstack/salt/issues/9743
18:53 MTecknology id -G   itself is instant
18:54 fridiculous i have a challenge for people
18:54 ekristen is there a way to have a file.managed tell a service to restart?
18:55 MTecknology Nazzy: I'm working with 180 remote servers and most are connected via T1's (1.5/1.5 mbit) and they authenticate with LDAP (libnss-ldapd & libpam-ldapd). The AD server has ~300 groups & 40,000 users. I'm in a bit of a unique situation... I bet I'm one of the very few people that would every realize this issue.
18:55 fridiculous ekristen i'm interested in a very similar question
18:55 fridiculous i'm getting the following error
18:55 fridiculous ----------           ID: kibana-init     Function: service.running       Result: False      Comment: Service kibana-init is already enabled, and is dead      Changes:
18:56 forrest ekristen, yes, use watch_in
18:56 fridiculous from http://pastebin.com/FwM5bxRT
18:56 forrest http://docs.saltstack.com/ref/states/requisites.html#watch-in
18:56 MTecknology Nazzy: running just 'g = grp.getgrall()' takes about 15 seconds.
18:56 MTecknology brb
18:57 fridiculous forrest, would that solve my problem as well?
18:58 forrest fridiculous, I don't think so.
18:58 Nazzy MTecknology, I don't /think/ that's actually salt, I'd hazard that basepi is right, it's the underlying library being stupid
18:58 forrest fridiculous, yours is confusing me because you have - restart set to True
18:58 ekristen forrest: hrm ok I’ll have to play with it
18:58 forrest as well as having - reload
18:58 Nazzy ooh, I do have a box that has ldap configured like this
18:59 forrest fridiculous, also your watch statement should not be watching the file name like that, it should be - file: kibana-init .
19:00 forrest you're referencing the ID to watch, not the name of the file itself, does that make sense?
19:02 forrest I've gotta head out, I'll try to be back on later.
19:05 Nazzy let's see ... loads lib_nss stuffs, loads a shed load of other ELF SO's, reads ldap.conf, does uname and resolv.conf read twice, tries and fails on nscd unix socket ... really, sock_raw? ¬_¬
19:06 fridiculous hmm
19:06 fridiculous i'll change it and try it
19:07 Nazzy getpeername on ldap ... connect.. this is way messier when local ldap cache isn't available :p
19:07 vejdmn joined #salt
19:11 Nazzy and there we are ... a single query write, multiple answer reads... that list comp doesn't seem responsible for user enumeration
19:13 MTecknology Nazzy: k... no- you're brilliant and I have no idea what the crap is going on
19:17 MTecknology do you have any ideas about fixes?
19:17 Nazzy I'm exploring currently... I suspect user.present is your issue
19:18 MTecknology Nazzy: there is also this same 15 minute delay when I run salt-call or start salt-minion
19:19 MTecknology I'd be a bit less concerned if it only happened once, but that's still causing me some massive headaches
19:19 MTecknology also...   salt-call state.sls sys.files.centernumber test=True   <-- this actually made the change on the disk...
19:20 cedwards looks like i just got three more people from work to register last minute for saltconf
19:21 kainswor joined #salt
19:21 MTecknology cedwards: nice! Are they required to go to your talk?
19:22 fridiculous forrest, i made the changes you suggested but i'm still getting " Service kibana-init is already enabled, and is dead"
19:23 cedwards MTecknology: or else! :)
19:23 fridiculous http://pastebin.com/hQA6xKfQ
19:24 MTecknology cedwards: Have a drink of pepsi for me! (I'm not drinking booze) :D
19:24 MTecknology I'll have to look forward to meeting you next year.
19:26 cedwards too bad you couldn't make it last minute. real bummer.
19:27 Corey I have arrived in Salt Lake City.
19:28 Corey UtahDave was not at the airport with a sign and a dozen roses.
19:28 Corey So far I'm unimpressed with the speaker treatment.
19:29 MTecknology HAHA!
19:29 cedwards lol
19:30 Corey Is anyone other than Rhett at the Marriott yet?
19:30 whiteinge Corey: woot! welcome! or possibly go eff yourself. i haven't seen the memo on how we're supposed to treat speakers yet so i'll hit both extremes...
19:31 timoguin ah community
19:31 whiteinge Corey: s0undt3ch is here, and terminalmage too
19:32 Corey whiteinge: I'm in the lobby.
19:33 whiteinge they're both at lunch along with a few guys from the office
19:34 NotreDev joined #salt
19:35 ajw0100 joined #salt
19:37 Corey One of our salesweasels just arrived, so I think I'm set.
19:39 MTecknology Corey: you didn't forget to tell them ahead of time when you'll show up, did you?
19:39 MTecknology also... holy crap these headphones are amazing
19:39 MTecknology expensive, but amazing
19:39 Corey MTecknology: Oh, I'm fine at just barging in. It's fun.
19:39 Corey MTecknology: Which headphones?
19:40 MTecknology Corey: I meant the fact that they weren't waiting for you.
19:40 MTecknology Bose QC15
19:41 Nazzy MTecknology, ASN.1 DER is evil and should diaf ¬_¬
19:41 MTecknology Nazzy: hm?
19:41 Nazzy I'm trying to decode what the ldap query it sent actually means
19:41 MTecknology oooh
19:46 MTecknology Nazzy: if I can help, lemme know. not sure I have much to offer, though
19:47 Ryan_Lane2 joined #salt
19:52 zooz joined #salt
19:54 UtahDave Corey: Welcome!  I thought I sensed a change in the Force.
19:54 fridiculous hey UtahDave
19:54 Nazzy MTecknology, I can't work out what it's doing ... my knowledge of LDAP line protocol just isn't good enough to apply what I know about ASN.1 and ABNF to strace's horrific output
19:54 pass_by_value joined #salt
19:55 ipmb joined #salt
19:55 fridiculous @UtahDave i can't for the life of me get upstart to work with this salt state
19:56 ipmb what's the preferred/recommended method for configuring a returner?
19:56 ipmb some use pillars others use minion conf
19:56 UtahDave fridiculous: Ubuntu?  can you pastebin your state?
19:56 gadams999 joined #salt
19:56 gadams999 left #salt
19:57 fridiculous http://pastebin.com/hQA6xKfQ
19:57 fridiculous ubuntu and i get the following result
19:57 fridiculous ID: kibana-init     Function: service.running       Result: False      Comment: Service kibana-init is already enabled, and is dead      Changes:
19:58 ipmb fridiculous: sounds like the init file is bad
19:58 Nazzy MTecknology, it's definitely only sending one query, and I don't think it's actually sending a bind request... it appears to be just doing "tell me everything you know", the query isn't long enough for anything more sophisticated
19:58 ipmb check the init file and verify you can run it manually
19:58 fridiculous i've tried many combinations of reload, reload_modules, restart, enable
19:58 ipmb service kibana start
19:58 fridiculous so that's the funny thing
19:59 fridiculous the service doesn't get listed
19:59 ipmb yeah, that's upstart for you
19:59 fridiculous it only shows up if i do sudo salt service
19:59 ipmb can you paste /etc/init/kibana-init.conf
19:59 ipmb upstart doesn't do you any favors troubleshooting a bad script
19:59 ipmb it just pretends it doesn't exist
19:59 fridiculous yeah
20:00 Nazzy MTecknology, I would be sorely tempted to fire up a profiler and try that run profiled instead of straced, that'll tell you exactly where it's taking that time
20:00 timoguin fridiculous, i've had some issues with elasticsearch as well.
20:00 timoguin haven't fixed them yet
20:00 fridiculous which issues
20:00 linjan joined #salt
20:00 fridiculous i get elasticsearch working
20:00 KyleG1 joined #salt
20:01 fridiculous just deamonizing with upstart has been a headache
20:01 fridiculous at least the current headache
20:01 timoguin yea that's what i'm talking about.
20:01 UtahDave fridiculous: I'm terrible with init systems, but do you have to symlink something to /etc/init.d/kibana-init   or something?
20:01 timoguin logstash is a pain to daemonize as well
20:01 timoguin because I'm also not good with init stuff
20:01 fridiculous http://pastebin.com/ygjYnape
20:01 fridiculous yay, i'm not the only one!
20:02 timoguin I'm used to packages that just work.
20:02 ipmb fridiculous: ok, we can fix this up :)
20:02 timoguin But a lot of new stuff doesn't even use init out of the box
20:02 fridiculous i've commented stuff out to experiment with different run levels
20:02 tessellare fridiculous: I had something similar happen with another app. It ended up I was trying to start the service with an account that didn't have correct permissions to the folders it needed, so it would fail.
20:02 ipmb fridiculous: give me a minute
20:03 fridiculous awesome
20:04 ipmb fridiculous: https://dpaste.de/jfru
20:04 ipmb that should be all you need
20:04 ipmb oops, remove the comment
20:04 fridiculous haha
20:05 fridiculous gotchya
20:05 ipmb this https://dpaste.de/cYjt
20:05 fridiculous thanks, let me give it a shot
20:06 timoguin well that looks easy...
20:07 ipmb fridiculous, timoguin: http://lincolnloop.com/blog/joy-upstart/
20:07 fridiculous eh.....
20:07 fridiculous nope
20:07 fridiculous same thing
20:07 ipmb is that your *actual* file on the server?
20:07 ipmb or your template?
20:07 fridiculous that's a good question
20:08 fridiculous yes
20:08 xmltok im noticing in pydsl when i do an include() it doesn't get inserted into the highstate where I think it would. by going through the code it appears that this may be the case more than just pydsl. has anyone else noticed this problem?
20:08 fridiculous and it works if i just call the exec file
20:08 fridiculous exec line*
20:08 ipmb fridiculous: you have the template code on the exec line?
20:08 fridiculous let me show you
20:09 ipmb ok, are you running it as the "ubuntu" user when you test?
20:09 ipmb when you run it manually does it run forever or fork a daemon?
20:09 fridiculous http://pastebin.com/Y05y5xNT
20:10 ipmb you are running it as "vagrant"
20:10 ipmb does the user "ubuntu" exist?
20:10 fridiculous no vagrant isn't the user
20:10 fridiculous it's just the machine
20:10 fridiculous actualyl that's a good point
20:10 fridiculous haha
20:10 fridiculous let me try root
20:10 ipmb root is the default
20:10 ipmb just comment out setuid ubuntu
20:11 ipmb fridiculous: have you ever used nginx?
20:11 fridiculous MOFO
20:11 fridiculous it worked
20:11 fridiculous root was the issue
20:11 fridiculous or sorry, ubuntu
20:11 ipmb you probably don't want to run it as root in production
20:12 ipmb are you using a flask wrapper to serve kibana?
20:12 fridiculous change permissions based on environment using jinja
20:12 fridiculous yes
20:12 fridiculous yes i am haha
20:12 ipmb nginx is probably a better option
20:13 ipmb and you won't have to manage the daemon
20:13 ipmb and it will run as www-data
20:13 fridiculous hmm
20:14 fridiculous ok i'll give it a shot. any other reason to prefer nginx that to a simple http server / flask, if upstart is already working?
20:15 ipmb fridiculous: https://dpaste.de/eWrq
20:15 n8n joined #salt
20:15 ipmb it's simpler to configure and install
20:16 ipmb lighter-weight and more performant
20:16 mgw joined #salt
20:16 ipmb but this is probably low-traffic and if you've already done the work...
20:16 jaustinpage joined #salt
20:17 ipmb you don't even need flask if you want to go super lo-fi :)
20:17 fridiculous that's the http.conf correct
20:17 rojem joined #salt
20:18 ipmb fridiculous: that's kibana.conf in /etc/nginx/sites-enabled
20:18 ipmb cd /path/to/kibana && python -m SimpleHTTPServer
20:18 alice|wl python -m SimpleHTTPServer 8080     besides that nginx is a pretty simple webserver :)
20:18 mgw1 joined #salt
20:18 fridiculous thing about simplehttpserver is that it's a pain to switch application directory
20:19 ipmb not in upstart
20:19 ipmb http://upstart.ubuntu.com/wiki/Stanzas#chdir
20:20 ipmb add another line:
20:20 ipmb chdir /path/to/kibana
20:20 ipmb to you upstart conf
20:21 fridiculous with nginx, how am i managing multiple domains?
20:22 fridiculous say i want kibana / halite to run on the same machine
20:22 ipmb https://dpaste.de/eWrq#L7
20:22 MTecknology Nazzy: eh.... new tools for me
20:22 ipmb add another file to /etc/nginx/sites-enabled
20:22 fridiculous i see it, but i don't understand it :D
20:22 ipmb change the server_name
20:22 ipmb this is for one server
20:22 fridiculous each file in /etc/nginx/sites-enabled is a seperate server
20:23 ipmb not in the process sense...
20:23 ipmb in the end, they are all just concatenated and act as your config
20:23 ipmb you just need to define multiple "server {}" blocks
20:24 ipmb typical way to do that is as separate files in sites-enabled
20:24 fridiculous so 1 nginx process runs multiple domains
20:24 ipmb correct
20:24 fridiculous ::high five::
20:24 ipmb add the config and reload the server
20:24 ipmb sudo service nginx reload
20:24 ipmb salt can do that for you
20:25 fridiculous yeah, that's a breeze
20:25 fridiculous the salt stuff, that is :D
20:25 ipmb "nginx -t" is your friend as you are setting it up
20:25 ipmb it will test your config and tell you if you've done something wrong
20:26 ipmb set it up manually first and then move it to salt
20:26 blee joined #salt
20:26 ipmb so you are only troubleshooting one thing at a time :)
20:28 kalloc joined #salt
20:28 mgw1 Shouldn't `salt-run -d xmodule.method` show docs for that method only? It's a bit annoying that it shows docs for every method of every runner.
20:28 fridiculous thanks ipmb
20:29 fridiculous you've been super helpful
20:29 ipmb np :)
20:31 mgw1 In fact, man for salt-run say that it displays docs only for the passed runner. I'll open an issue and maybe get a chance to look at the code for it.
20:35 andresr joined #salt
20:38 UtahDave mgw1: yeah, it should work like you're expecting
20:39 MTecknology Nazzy: Do I want python-profiler?
20:40 wolfpackmars2 joined #salt
20:41 Nazzy MTecknology, I'm trying to get my browser to behave long enough to find the link lol
20:52 Nazzy I'm beginning to think my browser has gained the ability to literally cry
20:52 Nazzy this might, possibly, have something to do with the fact I have 400 open tabs
20:55 KyleG joined #salt
20:55 KyleG joined #salt
20:56 kermit joined #salt
20:56 mgw joined #salt
20:57 mgw UtahDave, thanks… I opened a ticket for tracking. I can take a look at the code for it after SaltConf, or maybe during the sprint.
20:57 diecast joined #salt
20:58 Nazzy MTecknology, there's a profiler that is supposed to be really good: http://packages.python.org/line_profiler/
20:59 Nazzy MTecknology, you'll probably need cProfile to find where the offending code is
21:00 Nazzy MTecknology, docs for cProfile: http://docs.python.org/2/library/profile.html
21:01 bemehow joined #salt
21:01 Nazzy MTecknology, there's also RunSnakeRun that I've just come across and looks helpful to understanding what cProfile is telling you
21:01 UtahDave mgw: cool, thanks!  I don't imagine it should be too dificult to get it working correctly.
21:02 n8n joined #salt
21:02 alunduil joined #salt
21:05 Ryan_Lane2 joined #salt
21:06 mgw UtahDave, yeah, should be easy
21:07 jeffmendoza joined #salt
21:08 cachedout_home joined #salt
21:09 NotreDev joined #salt
21:09 jtang1 joined #salt
21:10 cewood joined #salt
21:10 n8n joined #salt
21:17 mpanetta Hrm, weird error.  I am getting     Specified SLS common.ssh_auth in environment marconi-test2-ord is not available on the salt master   However the file exists and is accessable...
21:18 mpanetta And is in my file_roots
21:18 hunter joined #salt
21:19 Ryan_Lane2 joined #salt
21:20 DavidL joined #salt
21:20 JordanRinke joined #salt
21:20 Iskondi joined #salt
21:21 JordanRinke Morning
21:21 mpanetta And it is readable..
21:21 Iskondi hello everyone
21:21 UtahDave mpanetta: are your environments configured correctly?
21:21 UtahDave hey, JordanRinke and Iskondi!
21:22 Iskondi Hiya UtahDave
21:22 blee joined #salt
21:22 Iskondi I was wondering if someone might know what I'm running into. I'm trying to install packages on an RHEL box but I keep getting an odd error.
21:22 chadhs_ joined #salt
21:22 Iskondi If I run the same state against a CentOS box it works
21:23 Iskondi RHEL 6.5 and CentOS 6.5
21:23 mpanetta UtahDave: They should be, if they aren't I can't find the issue. :(
21:23 mpanetta They weren't when I first tried to run the command, maybe something got cached on the minions?
21:24 Iskondi At the end of the error I see: PluginYumExit: Could not contact any CDS load balancers: rhui2-cds01.us-west-2.aws.ce.redhat.com, rhui2-cds02.us-west-2.aws.ce.redhat.com.
21:24 Iskondi However, if I run the yum update through a cmd.run it'll work...
21:26 chadhs_ joined #salt
21:26 mpanetta UtahDave: Trying to run a saltutil.sync_all to see if that will help...
21:27 mpanetta hrm, nope
21:28 hunter Iskondi:  have no idea.
21:29 Iskondi Thanks Hunter. I'm at a loss myself, it's quite odd...
21:30 hunter I suspect soemthign in the plugin isn't paying full attention to internal web pathing/routing used by redhat. Which doesn't really exist for centos and works fine with the commnad line.
21:30 mpanetta UtahDave: If it helps any I am moving my formula from a 0.16.4 master to a 0.17.5 master, I copied all the configs, so they should be the same.
21:31 Iskondi Hunter: that fits. Is this something I should file as a bug against pkg.installed?
21:32 Nazzy mpanetta, did you update the slaves to 0.17.1+ yet?
21:32 hunter I would - but it'll most likely need a dev that knows that state AND a proper RH account.
21:32 Nazzy mpanetta, cause 0.17.1 is a backward compatibility breaker
21:32 mpanetta Nazzy: recreating them from scratch
21:32 mpanetta this is a test env
21:32 mpanetta so I can do that ;)
21:33 [diecast] joined #salt
21:33 Iskondi Hunter: gotcha, ok, perhaps I'll ask internally, here at my company, if someone can help me figure out where it may be breaking so I can provide more detail.
21:33 hunter I would file it on pkg now-  give the salt people a chance - they've been nicely responsive to me so far.
21:34 Nazzy mpanetta, kiki, figured it was worth checking
21:34 Iskondi Ok, cool.
21:35 mpanetta Nazzy: No problem :)
21:36 Nazzy oooooh, wait... is it this: https://github.com/saltstack/salt/issues/4610 ?
21:37 UtahDave mpanetta: can you pastebin your top.sls? Or at least the relevant portion?
21:37 Nazzy (long shot, just remembered that files with dots suck :p)
21:38 mpanetta UtahDave: I think I found it...
21:38 mpanetta We have a file in the env that is attempting to include common.ssh_auth, but that file is only in base env.
21:39 mpanetta Was there something that changed between 0.17 and 0.16 that would affect such things?
21:40 UtahDave mpanetta: possibly.  There's a format to allow for including a file from another environment.  It looks similar to an option being passed in a web uri
21:41 mpanetta I copied the file and that fixed parts of it, so that seems to be the issue.
21:41 mpanetta I will have to change the include
21:41 UtahDave i'm looking for that syntax
21:42 mpanetta UtahDave: Thank you :)
21:43 UtahDave mpanetta:    include: \n  env_name: my_sls
21:44 n8n joined #salt
21:45 mpanetta thanks!
21:46 UtahDave and actually here's the docs.  Look at the second example:  http://docs.saltstack.com/ref/states/include.html
21:46 Ryan_Lane joined #salt
21:47 mpanetta Awesome, thank you UtahDave.  Everything works now.
21:47 UtahDave cool!
21:47 mpanetta I have a few formula to clean up, but otherwise I think things are ok.
21:47 david_a joined #salt
21:47 UtahDave good to hear.
21:50 carmony UtahDave: if I didn't have to give a presentation at SaltConf to get ready for, I actually might have had my secret project finished by then :P
21:50 UtahDave :)  nice.
21:50 carmony but I'm going to have to stop work on it until my slides are done :P
21:51 UtahDave carmony: I'm dying to share what you're doing with a few of our saltstack devs who like what you're working on.
21:52 carmony UtahDave: you can share with the saltstack folks :P You guys are too busy with awesome salt stuff to do anything with the idea.
21:53 UtahDave too true!
21:54 Ryan_Lane2 carmony: I'll share my project if you'll share yours
21:54 Ryan_Lane2 ;)
21:55 DerekRBN joined #salt
21:55 ipmb glad to see I'm not the only slacker still working on his slides
21:55 felixhummel joined #salt
21:55 * cedwards isn't done with his slides either
21:55 * Ryan_Lane2 finished his slides last night
21:57 cedwards i find my presentations are never quite done until i'm in front of the room
22:02 timoguin they never really are
22:02 timoguin http://slid.es/ is a pretty nice site for those of you who don't know it
22:02 timoguin gives you WSIWYG for the reveal.js framework
22:03 timoguin and you can export everything to HTML. is niiice.
22:03 analogbyte joined #salt
22:04 carmony UtahDave: you'll have to let me know what they think
22:07 DerekRBN Anyone have a recommended way of updating salt on ubuntu? dpkg says im at 0.17.5 but a salt-minion --version says 0.15.3
22:08 cedwards DerekRBN: have you restarted the minion since upgrading the package?
22:08 DerekRBN Yes
22:08 cedwards ..or since 0.15.3? :)
22:08 DerekRBN I even tried to remove the package
22:09 hunter UtahDave: Thanks for the syntax for env: includes
22:10 UtahDave carmony: they thought it was really cool!
22:10 cedwards DerekRBN: do you have any other pending upgrades? i wonder if you've upgraded part but not all of Salt
22:11 mgw joined #salt
22:12 gldnspud joined #salt
22:12 DerekRBN Nothing that I believe should be keeping salt back: linux-headers-virtual linux-image-virtual linux-virtual mysql-client-5.5
22:12 aleszoulek joined #salt
22:13 jeremyfelt joined #salt
22:13 cedwards hmm
22:13 nkuttler DerekRBN: what do you mean by dpkg says?
22:14 UtahDave hunter: no problem!  did that help you, too?
22:14 DerekRBN dpkg -l shows what packages are installed and the version
22:14 hunter not yet. But I want to restructure and could see a use.
22:14 DerekRBN might be wrong about that >.<
22:15 nkuttler DerekRBN: can you paste the output of dpkg -l | grep salt ?
22:15 nkuttler DerekRBN: while you're at it, which salt-minion, and dpkg -L salt-minion | grep salt-minion
22:16 gldnspud joined #salt
22:17 ekristen whats the best way to set grains on a host?
22:17 DerekRBN http://pastebin.com/eHtjMhNd
22:18 nkuttler ekristen: the grains execution module?
22:18 BrendanGilmore joined #salt
22:18 nkuttler DerekRBN: is /usr/local/bin/salt-minion the 0.15 version? that doesn't look like a .deb file
22:18 cedwards DerekRBN: the 'which salt-minion' looks suspicious to me.. I wouldn't expect the .deb to install to /usr/local
22:19 nkuttler DerekRBN: should be /usr/bin/salt-minion
22:19 nkuttler (unless ubuntu really messed that up, heh)
22:19 DerekRBN ls -la `which salt-minion` lrwxrwxrwx 1 root root 20 Jan 27 22:05 /usr/local/bin/salt-minion -> /usr/bin/salt-minion
22:20 nkuttler DerekRBN: dpkg -L salt-minion
22:20 DerekRBN http://pastebin.com/2St9Q3R1
22:21 DerekRBN Good look on that capitol L.
22:22 nkuttler DerekRBN: um, depending on your python path i guess your /usr/local stuff could mess up things..
22:22 nkuttler i'd get rid of that first.. still odd though
22:23 DerekRBN This is more trouble then it's worth. Thanks for the help. I think I'll just bring this up as amazon linux. those minions upgraded with no issues
22:25 nkuttler :D
22:25 toofer Hey, is anyone here going to the conference tomorrow?
22:26 yomilk joined #salt
22:26 cedwards toofer: should be quite a few. i'll be there.
22:26 toofer Did you get the email about the scheduling?
22:26 alunduil joined #salt
22:27 cedwards i've received a number of emails from them the last couple days.
22:27 cedwards which one are you referring to?
22:27 toofer My question that I'm on the sched.org site because it seemed I needed to lock in the sessions that I want to attend, but there is no where to save my schedule.
22:27 scooby2 joined #salt
22:27 cedwards i don't think you have to save. i've quit and logged back in and my selections were still there.
22:28 toofer Hmm... okay. I ask because I hit the "Hide sidebar" and that fully cleared my schedule for me.
22:29 cedwards man sched is slow..
22:30 toofer hmm... I emailed it to myself and now it appears to save. /me shrugs
22:30 cedwards i think i noticed it cleared mine when i did the same earlier.
22:31 cedwards yeah, hiding the sidebar looks like it clears it, but if you go back to another view it should be all there
22:31 toofer Cool, thanks for confirming.
22:33 bemehow joined #salt
22:33 joehh mpanetta: from http://packages.ubuntu.com/search?keywords=salt-common&amp;searchon=names&amp;suite=all&amp;section=all
22:34 joehh you can see that saucy has only 0.16.0 directly in the ubuntu release
22:34 joehh newer packages are in the ppa
22:34 mpanetta joehh: Thanks, I switched to PPA, since I figured it was more up to date anyway :)
22:35 joehh trusty has 0.17.5 - and should do so for the LTS release
22:35 joehh good to hear
22:36 kermit joined #salt
22:38 taion809 joined #salt
22:42 darless joined #salt
22:44 darless So in the pillar top file. Is there a predefined order that things are matched? Is there a way to define an order?
22:44 ConceitedCode joined #salt
22:45 UtahDave darless: starts at the top and works its way down
22:46 hunter 0.17.5? when did that hit?
22:47 cedwards hunter: this morning
22:47 jcsp1 joined #salt
22:47 hunter dang
22:47 gldnspud i'm wondering that too.  :)   i hope there's progress on https://github.com/saltstack/salt/issues/9638 soon
22:48 aleszoulek joined #salt
22:48 gldnspud there's also https://github.com/saltstack/salt/tree/v0.17.4 but no https://github.com/saltstack/salt/tree/v0.17.5  :(
22:51 kermit joined #salt
22:52 Mua joined #salt
22:53 gldnspud re salt-cloud -- any advice for those who are tracking salt 0.17.x but who are also using salt-cloud?  how to get them to coexist?  perhaps just install salt from 'develop' branch into a virtualenv, and symlink its salt-cloud script to /usr/local/sbin?
22:53 gldnspud i'll give that a try for now but open to something simpler if it's a known problem that's already been solved
22:55 darless UtahDave: that's what I assumed. Just got very confused trying to figure out why the pillar data was not looking how I was expecting. One sec I'll paste an example
22:55 darless http://pastebin.com/Uc6NAdSW
22:55 dlam joined #salt
22:56 rawzone joined #salt
22:58 higgs001 joined #salt
22:58 toofer joined #salt
23:02 KyleG salt+PagerDuty Support?
23:02 KyleG What's that for
23:04 Ryan_Lane21 joined #salt
23:05 rawzone joined #salt
23:05 zsoftich joined #salt
23:06 Gordonz joined #salt
23:06 strgcloud joined #salt
23:07 toofer_ joined #salt
23:08 toofer joined #salt
23:13 darless UtahDave: After some more testing it looks like if I use a different keyword besides 'roles' it works as expected
23:13 bastion2202 joined #salt
23:15 bretep I'd like to use the environment targeting on the minion size vs top.sls https://gist.github.com/bretep/8ef9fb357fa840320c54
23:17 bretep I know it's not a prefered method, but would the yaml I posted work?
23:19 david_a joined #salt
23:20 timoguin joined #salt
23:23 fr00 joined #salt
23:24 cstelma I'm having some trouble getting multi-master to work.  Everything seems to work fine as long as I don't try to run more than one master at the same time.
23:25 cstelma If I do, it looks like the minions swap which master they're reading commands from, until they hit some limit, and then they swap back
23:25 cstelma things go worse if I interleave commands.
23:26 cachedout joined #salt
23:26 cstelma I'm just running "salt '*' test.ping" from 2 hosts which are both masters and both minions
23:26 cstelma if I run that about 3 times on one host with both masters active, it looks like things time out
23:28 dlam where can i find documentation about passing a pillar variable sat the command line to salt-call?
23:29 kainswor lol 17.5 comes out the day before saltconf
23:30 kainswor I was going to have to rant about the file.managed screw up the whole time otherwise ;)
23:31 kainswor I wish I had time to deploy it across our DC's before going though
23:33 bemehow_ joined #salt
23:33 rawzone joined #salt
23:34 cachedout joined #salt
23:39 pydanny joined #salt
23:41 xmltok it appears that the iptables state requires the iptables parameters in kwargs dict, yet the order of iptables parameters is important
23:43 sinh_ joined #salt
23:44 jeremyfelt joined #salt
23:44 MTecknology Nazzy: k... back to this issue
23:45 cachedout joined #salt
23:45 terminal1 joined #salt
23:45 cyp__ joined #salt
23:45 nineteeneightd joined #salt
23:46 elithrar_ joined #salt
23:46 _fllr_ joined #salt
23:46 darless_ joined #salt
23:46 harobed__ joined #salt
23:47 MTecknology Nazzy: So... I'm confused. This profiler stuff seems to be for profiling python stuff, but as you said, it seems that salt and python are not to blame...
23:47 Mua_ joined #salt
23:48 dwfreed_ joined #salt
23:48 jeffrubi` joined #salt
23:48 yota_ joined #salt
23:48 eliasp_ joined #salt
23:48 minaguib_ joined #salt
23:50 smkelly joined #salt
23:50 nhubbard joined #salt
23:53 jcsp joined #salt
23:54 alunduil joined #salt
23:54 jtang1 joined #salt
23:55 Nazzy MTecknology, correct ... my thought was to identify exactly which function call is actually causing the user enumeration
23:55 MTecknology the function call within grp.getgrall() ?
23:56 MTecknology or... one of those....
23:56 chitown joined #salt
23:57 Ryan_Lane2 joined #salt
23:57 Nazzy I don't think it's grp.getrall() ... I suspect pwd, but my test suggests neither grp, pwd or spwd actually do a call like that
23:58 fr00 While this is not the typical Salt question, I figured someone here might have some good points. I'm trying to decide upon how to handle my base images for my cloud. I'll have as few base images as possible, one per each OS; and will use Salt to configure them automatically once i create instances. I would be glad to hear if any of you handle base images for a cloud and which software you are using.
23:58 Nazzy it could be a difference in the ldap config, but knowing what function really triggers it will show where to dig
23:58 fr00 I'm thinking about eventually going for Vagrant, build and manage the base images locally. Then provision them to the cloud using a cloud provider.

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