Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-05-13

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

All times shown according to UTC.

Time Nick Message
00:00 solidsnack joined #salt
00:11 Micromus joined #salt
00:18 c10b10 joined #salt
00:21 otter768 joined #salt
00:22 baweaver joined #salt
00:22 timoguin joined #salt
00:23 SneakyPhil joined #salt
00:23 SneakyPhil Q: Is it possible to prepend text to a file based on a pillar?
00:24 SneakyPhil http://pastebin.com/Jv3Frsai
00:26 iggy the "template: jinja" syntax is when you want to actually prepend templated files
00:26 hemphill joined #salt
00:27 iggy if you just want a string in there, just use - text: |\n{{ pillar['foo']|indent(8) }}
00:27 iggy or something along those lines
00:27 iggy don't remember exactly what args indent takes
00:28 babilen ... now wouldn't it be nice if one could easily use decorators for states? (e.g. wrap file.managed with a "prepend a salt header" function)
00:29 iggy pyobjects has a format that works kind of like that
00:30 paolo guys i'm trying to use salt-ssh but have some problem. for example, salt-ssh server1 test.ping gives permission denied and asks a password
00:30 paolo but ssh server1 works fine
00:30 iggy probably trying to ssh as different users
00:30 paolo so i guess the public key is deployed correctly
00:30 paolo as root
00:31 iggy <-- not a salt-ssh user... was merely a shot in the dark
00:31 ek6 paolo: when in doubt call out your private key   priv in roster file
00:32 SneakyPhil iggy: I've got the text showing up in the error I'm getting so I am progressing
00:33 SneakyPhil now I've gotta figure out why I'm getting this yaml error when the text appears
00:41 julez joined #salt
00:42 paolo ek6: could you have a look at my roster file? http://pastebin.com/WLkzjtcL
00:42 paolo i have this error now No hosts found with target server1 of type glob
00:42 ek6 where is that file
00:43 ek6 is it at /etc/salt/roster   ?
00:43 paolo yes
00:44 ek6 run with -l debug
00:45 ek6 in the spew look for Terminal Command  and try to execute that by hand
00:45 paolo http://pastebin.com/9rKkEJaq
00:45 bhosmer joined #salt
00:45 paolo salt-ssh 2014.7.5 (Helium)
00:46 ek6 yeah put host line in
00:46 ek6 well no..that wont help..
00:46 ek6 wont hurt but..
00:47 ek6 cant imagine why but salt-ssh -l debug 'server1' test.ping
00:50 ek6 oh duh...your format is wrong...sorry...been a long day
00:50 paolo wrong?
00:50 paolo ^wrong question.. :), where?
00:51 ek6 lose the - on user priv and host lines
00:51 ek6 although that shouldnt keep the match from happening..so im still missing something dirt obvious in my attempt to help
00:51 solidsnack joined #salt
00:53 paolo ek6: thanks for help, really appreciated.
00:53 paolo it works
00:54 joehoyle joined #salt
00:59 Nazca joined #salt
01:00 awpti Hey folks, probably a simple question, but I'm having a hard time finding an answer: When a given file is updated (by salt), how can I have salt also reload/restart the service related to that file? eg, nginx config updated, reload nginx. I know I can do a cmd.run, but I can't find any examples that live inside of an sls file.
01:01 manfred awpti: check out onchanges or watch
01:01 manfred awpti: http://docs.saltstack.com/en/latest/ref/states/requisites.html#onchanges
01:01 solidsnack joined #salt
01:01 manfred http://docs.saltstack.com/en/latest/ref/states/requisites.html#watch
01:02 rideh joined #salt
01:03 babilen awpti: You would define a watch_in: on the service in the file.managed state for that file. (I prefer that to listing all files in the respective service.running state) -- You might want to use listen_in if all you want is to ensure that the service is restarted at least once after changes have been made.
01:04 babilen (as opposed to "as soon as possible and potentially multiple times)
01:10 rhettg joined #salt
01:11 rhettg anybody here know anything about salt and ubuntu packaging?
01:11 rhettg i'm running 12.04.5, and today when creating some new instances, salt-minion 2014.7.5 installed, and it's missing the actual binary
01:12 rhettg all my other instances are running 2014.7.1
01:12 rhettg the contents of the package appear to be completely different
01:12 __number5__ rhettg: how do you install salt on your ubuntu?
01:12 rhettg using apt-get (in a user-data script at aws boot time)
01:12 rhettg from the saltstack ppa
01:13 babilen rhettg: Could you paste "dpkg -L salt-minion" to, say, http://refheap.com ?
01:13 __number5__ I think there's only 2015.5.0+ds-1precise1 on it now
01:13 __number5__ since 2015.5 released
01:13 babilen has it been pushed already?
01:13 __number5__ yep
01:13 rhettg i've got "deb http://ppa.launchpad.net/saltstack/salt2014-7/ubuntu precise main" configured in my sources list
01:13 __number5__ super quick this time
01:14 rhettg https://gist.github.com/rhettg/c637f747e9bb45bbe5f3
01:14 __number5__ since ppa only hold one version for each ubuntu release, there is no 2014.7.x for precise any more
01:14 babilen __number5__: Checked yesterday/today and it wasn't must have changed in the last couple of hours
01:15 rhettg sounds like the deb must be missing some rather important parts
01:16 rhettg this happened on 2015.5 as well, but I figured I just wasnt ready to upgrade that far
01:16 __number5__ babilen: from https://launchpad.net/~saltstack/+archive/ubuntu/salt 20 hours ago
01:16 rhettg but dpkg -L lists the same limited contents
01:16 babilen __number5__: Yeah
01:17 __number5__ rhettg: what do you have if `dpkg -L salt-common`
01:17 rhettg a lot more python, but no salt-minion. lemme see
01:17 babilen rhettg: Would you mind installing debsums and running "debsums salt-minion" ?
01:18 Nazca__ joined #salt
01:18 manfred rhettg: you will have to update your other things too, launchpad only allows us to keep 1 version of a package at a time unfortunately
01:19 rhettg __number5__: https://gist.github.com/rhettg/707dac6383bd13946e2c
01:19 babilen rhettg: I'd also be quite interested in "COLUMS=200 dpkg -l 'salt*'" along with the debsums output.
01:20 rhettg whoa, well that's weird:
01:20 rhettg https://gist.github.com/rhettg/2776d5e9c4d4d7cc989e
01:21 rhettg must have just partially installed or something
01:21 babilen rhettg: Okay, that .. well .. explains a bit more. If you could paste /var/log/apt/history.log along with /var/log/dpkg.log we could probably figure out what happened there
01:21 babilen *phew*
01:21 rhettg huh, still didn't install
01:22 babilen When you do what and what happens?
01:22 rhettg https://gist.github.com/rhettg/ec1f29e037f23a3dc8ff
01:23 babilen rhettg: And it is being shown as "rc" in dpkg -l again?
01:23 rhettg rc, yeah
01:24 babilen "dpkg --configure -a" does something?
01:24 cromark joined #salt
01:25 rhettg this seems notable from the apt history.log
01:25 rhettg https://gist.github.com/rhettg/32f975378e14b7d9be8f
01:25 rhettg some recursive thing with libzmq
01:25 joehh rhettg: very odd
01:26 MatthewsFace joined #salt
01:26 rhettg i saw a note in the docs about zmq changes in packaging
01:26 babilen That explains why the package was gone, but I can't quite make sense of why it *stays* that way.
01:26 rhettg "ZeroMQ 4 is available by default for Ubuntu 14.04 and newer. However, for Ubuntu 12.04 LTS, starting with Salt version 2014.7.5, ZeroMQ 4 is included with the Salt installation package and nothing additional needs to be done."
01:27 babilen And it is being installed when you install salt-minion (cf. https://gist.github.com/rhettg/ec1f29e037f23a3dc8ff#file-gistfile1-txt-L27 )
01:29 babilen Could you run "grep -A20 -e "Package: salt-minion" /var/lib/dpkg/status ?
01:29 babilen joehh: That's a new one, isn't it?
01:30 rhettg babilen: https://gist.github.com/rhettg/4c33afa7b739187143f2
01:30 joehh yes
01:30 babilen rhettg: "apt-cache policy salt-minion" ?
01:30 __number5__ rhettg: what's the output of `apt-cache policy salt-minion`?
01:31 __number5__ babilen: you beat me
01:31 babilen "Status: deinstall" is interesting (despite the explicit "apt-get install salt-minion")
01:31 rhettg https://gist.github.com/rhettg/1ed8754b976f17c33e25
01:31 rhettg would it be possible that i'm installing something else that uses zeromq and coudl cause a conflict?
01:32 babilen If there had been a conflict apt-get would have complained bitterly
01:32 rhettg yeah that's what I would expect
01:32 babilen This is something different. Something interesting.
01:32 babilen (and it is way too late here in GMT land to investigate that)
01:32 rhettg i certainly have other things using zeromq and msgpack, the python libs run out of a virtualenv though
01:33 joehh the original package should have more filess - https://www.refheap.com/100984
01:33 rhettg though they wouldn't have installed till later anyway
01:33 rhettg yeah my systems with 2014.7.1 look way different 'dpkg -L'
01:34 rhettg Well, sounds like filing an issue on github might be the next step?
01:34 babilen rhettg: The package installs fine, yet it is not installed. At least the system doesn't, for some reason, install it (or rather keep that state). You should have the .deb in /var/cache/apt/archives/ -- Could you try installing it manually with "dpkg -i /var/cache/apt/archives/salt-minion.....deb" (complete as appropriate)
01:35 babilen rhettg: I am not convinced that this is a salt bug.
01:35 rhettg the minion deb is like 30k
01:35 rogst joined #salt
01:36 joehh I get an md5sum of d85831f5241bf584f7ed7fb84b2888da
01:36 joehh for salt-minion_2014.7.5+ds-1precise1_all.deb
01:36 joehh is that the same as yours?
01:36 rhettg i get d85831f5241bf584f7ed7fb84b2888da
01:36 babilen rhettg: Along with md5sum what does "dpkg -V /path/to/the/package.deb" get you?
01:37 noway_ joined #salt
01:37 rhettg is that a valid option?
01:37 joehh I get an error for that one babilen
01:38 __number5__ http://ppa.launchpad.net/saltstack/salt2014-7/ubuntu/ <-- this
01:39 babilen joehh: yeah, that should have been "dpkg -V salt-minion", but that doesn't tell us much.
01:39 __number5__ by default salt stable ppa is: http://ppa.launchpad.net/saltstack/salt/ubuntu/
01:39 babilen It is the same md5sum anyway, so genuine.
01:39 babilen rhettg: How does the dpkg -i behave?
01:40 babilen But this system state is so wonderfully buggy that it would be a shame to fix it without figuring out what it is. Unfortunately it is a little too late for me
01:40 rhettg https://gist.github.com/rhettg/ee4bdf35b9dd5e0272e1
01:40 rhettg well the good news is I can reproduce it
01:41 babilen :)
01:41 rhettg an yeah, i kinda need to bail too
01:41 babilen rhettg: Okay, "apt-get -f install" probably results in the same "not installed" state, doesn't it?
01:41 joehh That last one may point to something
01:45 __number5__ rhettg: do a snapshot of your machine, save it for later :P
01:45 rhettg yes
01:45 moloney joined #salt
01:45 joehh I think the old pyzmq packages used to include the actual libzmq, that is now in the libzmq3 package proper
01:45 rhettg python-zmq is not going to be installed
01:45 babilen rhettg: What do you get for "apt-cache policy  python-zmq" ?
01:45 babilen and "apt-cache policy libzmq3"
01:45 joehh maybe there needs to be a conflicts or similar relationship between the old and new zmq packages
01:45 rhettg https://gist.github.com/rhettg/a6f57eec21c851bfcc4d
01:45 rhettg libzmq3: https://gist.github.com/rhettg/4a3d67354a75386230a0
01:45 joehh and what happens if you do an apt-get install python-zmq
01:45 rhettg unmet dependencies on libzmq3
01:45 joehh and then apt-get install libzmq3?
01:45 rhettg but libzmq3 won't install because
01:47 ilbot3 joined #salt
01:47 Topic for #salt is now Welcome to #salt | 2015.5.0 is the latest | Please use https://gist.github.com for code, don't paste directly into the channel | Please be patient when asking questions as we are volunteers and may not have immediate answers | Channel logs are available at http://irclog.perlgeek.de/salt/
01:47 rhettg oh, it woke up, connected to master
01:47 rhettg and something must have caused it to remove itself
01:47 rhettg lol
01:48 rhettg wow, ok
01:48 dalexander joined #salt
01:48 rhettg here is the transcript: https://gist.github.com/rhettg/f788432602404c13379c
01:48 babilen joehh: Could it be that the Ubuntu package(s) need a versioned dependency and Conflicts on the zmq packages?
01:48 rhettg it must be because i'm actually asking to install zmq, and it must conflict
01:49 joehh I think so
01:49 terratoma left #salt
01:49 rhettg probably take me a few to track down which
01:50 rhettg thanks for your help everyone... the wife is gonna kill me if I don't stop
01:50 joehh I think the zmq packages might also need something similar
01:50 joehh no worries - thanks for helping isolate the issue
01:51 rhettg i'll report back if I figure anything else out
01:51 dendazen joined #salt
01:51 babilen rhettg: She shall have you back. Ta for providing the information you provided.
01:51 joehh email joehealy@gmail.com too
01:57 solidsnack joined #salt
01:59 saurabhs left #salt
02:00 murrdoc file.copy stinks
02:00 murrdoc it always has changes
02:01 murrdoc cant do a stupid
02:01 murrdoc only execute this state x if state y has changes
02:01 murrdoc because state y has file.copy
02:04 ajw0100 joined #salt
02:05 kusams joined #salt
02:07 c10b10 joined #salt
02:09 VR-Jack well, use_vt is cool in salt-call. not so cool with salt.
02:14 Tyrm joined #salt
02:14 theologian joined #salt
02:22 otter768 joined #salt
02:22 jab416171 joined #salt
02:27 solidsnack joined #salt
02:28 solidsnack joined #salt
02:30 julez joined #salt
02:30 writtenoff joined #salt
02:35 psojourner joined #salt
02:40 theologian joined #salt
02:40 stanchan joined #salt
02:42 evle joined #salt
02:44 Brew joined #salt
02:50 amcorreia joined #salt
02:50 thayne joined #salt
02:53 michelangelo joined #salt
03:08 viderbit joined #salt
03:21 favadi joined #salt
03:24 viderbit joined #salt
03:27 adelcast joined #salt
03:28 evle joined #salt
03:36 Nazca joined #salt
03:41 smcquay joined #salt
03:41 joren joined #salt
03:42 solidsnack joined #salt
03:52 Furao joined #salt
03:56 c10b10 joined #salt
03:59 bhosmer joined #salt
04:20 Furao joined #salt
04:22 otter768 joined #salt
04:24 alexhayes joined #salt
04:28 fxhp joined #salt
04:31 joeto joined #salt
04:36 Garo_ Now when 2015.5.0 is out has the previous 2014.7.5 debian packages dissapeared from the ppa repo? https://launchpad.net/~saltstack/+archive/ubuntu/salt/+packages can't find them anymore with apt-cache showpkg
04:37 bluenemo joined #salt
04:37 bluenemo joined #salt
04:37 manfred Garo_: not yet
04:38 manfred oh maybe
04:38 manfred yeah
04:38 manfred if the 2015.5.0 is in the repos
04:38 manfred the 2014.7.5 will be gone
04:38 Garo_ yeah it's in the repos
04:38 Garo_ oh why :(
04:38 manfred because launchpad only lets one version be in the repo
04:39 Garo_ ok, didn't know that. guess I need to switch to use bootstrap so that I can anchor my version so that I can decide when I upgrade my systems
04:40 __number5__ Garo_: you can use 2014.7.x from this one https://launchpad.net/~saltstack/+archive/ubuntu/salt2014-7
04:40 manfred i use git tags
04:40 linjan joined #salt
04:40 solidsnack joined #salt
04:40 Garo_ __number5__: thanks
04:41 TyrfingMjolnir joined #salt
04:45 iggy Garo_: sadly, ppa's use reprepro which is utter shit and only allows one version per package
04:46 iggy but yeah, I switched to useing salt-bootstrap with git tags some time ago because the packages were constantly causing more trouble than they were worth
04:46 TyrfingMjolnir joined #salt
04:47 iggy some day I hope to start building my own and having time to thoroughly test them before upgrades... lol, not really
04:52 Garo_ iggy: thanks, that's what I'm going to do (bootstrap git revision). at some point I need to mirror github locally because can't afford downtime when github decides to be under ddos ;)
04:53 iggy lol, yeah
04:57 c10b10 joined #salt
05:03 __number5__ All our code, including salt states is on github... a mate said github enterprise is the solution, but my boss won't be happy about that
05:06 iggy gitlab isn't too bad
05:07 seev I edit the code on the saltmaster and push it to github for accounting purposes only
05:07 seev github could burn and I wouldn't care
05:08 iggy states/pillars aren't the problem
05:08 iggy it's installing new minions
05:09 tzero joined #salt
05:11 linjan joined #salt
05:13 pdayton joined #salt
05:13 seev they come packaged as an AMI, pull their hostname/ID from an API endpoint, and then run highstate to self-assemble like autobots
05:13 seev it's all automagical
05:14 pdx6_ joined #salt
05:14 lietu joined #salt
05:14 seev I should say, they come packaged as part of an autoscaling group that designates their environment and function, which kicks off the rest
05:14 iggy but if you are using stock images, and installing salt-bootstrap from git at spin up time...
05:14 funzo_ joined #salt
05:15 seev my setup is all pretty customized and shoestringed together for a very specific use
05:15 thayne_ joined #salt
05:15 aravind1 joined #salt
05:16 TOoSmOotH_ joined #salt
05:16 cwyse_ joined #salt
05:17 writteno2f joined #salt
05:17 eliasp_ joined #salt
05:17 Lyfe_ joined #salt
05:18 schristensen_ joined #salt
05:18 buxy_bak joined #salt
05:18 ventris_ joined #salt
05:18 mitsuhiko_ joined #salt
05:19 cztanu1 joined #salt
05:19 FineTralfazz joined #salt
05:25 Nazca__ joined #salt
05:27 jalaziz joined #salt
05:27 cptsupermrkt joined #salt
05:28 linjan joined #salt
05:33 marcinkuzminski joined #salt
05:34 SaltyCharles joined #salt
05:42 lb1a joined #salt
05:44 danemacmillan joined #salt
05:44 dunz0r joined #salt
05:44 Ssquidly joined #salt
05:45 kupric joined #salt
05:49 zer0def joined #salt
05:51 Furao joined #salt
05:55 cztanu joined #salt
05:56 mosen joined #salt
05:59 JlRd joined #salt
06:03 colttt joined #salt
06:03 Tyrm joined #salt
06:06 Furao joined #salt
06:15 AndreasLutro joined #salt
06:15 wendall911 left #salt
06:22 rofl____ how secure would it be to keep the salt master open to the whole world?
06:22 rofl____ has there been any security issues?
06:23 markm joined #salt
06:23 Garo_ I'm testing my multi-master setup. A minion which started when its primary master is down eventually (after seven iterations of SaltReqTimeoutError) connected to a secondary master, but a minion which has been running and the primary master was turned of without touching the minion doesn't seem to notice that the master is down. Is this behavior a feature? Does the minion ping the master periodically to verify that it's up?
06:23 otter768 joined #salt
06:24 rdas joined #salt
06:24 Garo_ ah, to answer myself: I found out about ping_interval config
06:25 __number5__ rofl____: NEVER open your salt-master to the world
06:27 rofl____ __number5__: you would go for something like salt-syndic or smth then?
06:27 rofl____ for ex. multiple cloud provider infrastcures
06:28 toanju joined #salt
06:30 KermitTheFragger joined #salt
06:31 cromark joined #salt
06:34 iggy rofl____: mine's been open to the outside world for about 9 months now without problem... I doubt it'll stay that way, but...
06:35 rofl____ its kinda sketchy i agree
06:35 rofl____ salt syndic seems maybe the way to go
06:35 rofl____ or deploy something like openvpn on all nodes
06:36 ndrei joined #salt
06:38 flyboy joined #salt
06:38 soren joined #salt
06:39 fbergroth joined #salt
06:39 teogop joined #salt
06:40 dimeshake joined #salt
06:41 joeto joined #salt
06:42 cowyn joined #salt
06:42 c10b10 joined #salt
06:45 chiui joined #salt
06:46 iggy ipsec, etc... lot's of ways to handle it
06:47 iggy throw it on alternate ports and ship it
06:47 Garo_ I'm sure I have seen a state module which installs salt via the bootstrap script so that I can have an .sls which installs salt-master, but I can't find it anywhere. Am I dreaming?
06:49 kusams joined #salt
06:52 iggy salt-formula?
06:53 iggy it doesn't use salt-bootstrap
06:53 iggy salitfy salt-cloud provider?
06:53 viq joined #salt
06:54 favadi joined #salt
06:56 cowyn joined #salt
06:58 cowyn joined #salt
06:59 malinoff joined #salt
07:03 kawa2014 joined #salt
07:03 armguy joined #salt
07:03 evidence is it not possible to access utils in state files?  salt.utils.gen_mac() doesn't seem to be exposed
07:04 __number5__ Garo_: bootstrap can install salt-master with '-M'
07:05 Garo_ __number5__: yeah but I'd like to create a state which enforces the correct version of salt-master and I'd like to use bootstrap script with git revision instead of Pkg.installed("salt-master")
07:09 JayFK joined #salt
07:11 __number5__ Garo_: https://www.refheap.com/101016
07:14 Garo_ __number5__: thanks, that suits well
07:19 kusams joined #salt
07:20 mccricardo joined #salt
07:20 dkrae joined #salt
07:32 cromark joined #salt
07:33 yggdrasi1 joined #salt
07:34 markm joined #salt
07:39 supersheep joined #salt
07:46 txomon|home joined #salt
07:51 favadi joined #salt
07:57 linjan joined #salt
07:59 jalaziz joined #salt
08:02 favadi left #salt
08:04 photex- left #salt
08:04 bluenemo joined #salt
08:10 markm joined #salt
08:11 Xevian joined #salt
08:13 scristian joined #salt
08:13 jalaziz joined #salt
08:14 julez joined #salt
08:17 al joined #salt
08:20 kusams joined #salt
08:24 txomon|fon joined #salt
08:24 otter768 joined #salt
08:24 keimlink joined #salt
08:27 solidsnack joined #salt
08:29 PI-Lloyd why o why is gitfs such a pain in the ass? :(
08:31 N-Mi joined #salt
08:31 N-Mi joined #salt
08:31 c10b10 joined #salt
08:33 julez joined #salt
08:34 kawa2014 joined #salt
08:36 AndreasLutro joined #salt
08:37 losh joined #salt
08:37 julez_ joined #salt
08:39 mccricardo joined #salt
08:39 mccricardo left #salt
08:42 ksj I'm just reading through the release notes for Lithium, beacon looks brilliant, but I'm wondering about the inotify beacon - are hashes of files stored on the master?
08:43 ksj ideally I'd like the hashes all stored on the master to verify integrity of files - similar to how aide works, or any of those programs, but done over the network
08:43 VSpike Not sure who handles the website updates, but http://docs.saltstack.com/en/latest/topics/installation/windows.html is missing 2015.5.0 download links. Luckily it's pretty easy to figure out the URLs :)
08:44 julez joined #salt
08:47 PI-Lloyd based on these docs - http://docs.saltstack.com/en/latest/ref/pillar/all/salt.pillar.git_pillar.html#module-salt.pillar.git_pillar - I should be able to use gitfs for ext_pillar and set the subdir of the pillar data.... however it does not work as shown here - https://bpaste.net/show/38daede56ee7 - Hopefully it's something I'm doing wrong and not a bug in salt. Anyone have any input?
08:49 PI-Lloyd this is on 2015.5.0 btw
08:49 supersheep joined #salt
08:50 mage__ any idea for https://dpaste.de/Avzw ?
08:52 julez joined #salt
08:55 monkey66 joined #salt
08:56 ksj mage__: I'm no expert, but why do you have the initial pkgs: (line 10). Delete line 10 and decrement all the following lines by two spaces and it should work as far as I can see
08:57 favadi joined #salt
08:57 PI-Lloyd ^^
09:00 mage__ ksj: works, thanks :)
09:07 che-arne joined #salt
09:15 c10b10 joined #salt
09:21 Furao joined #salt
09:22 CryptoMer joined #salt
09:23 markm joined #salt
09:28 CeBe joined #salt
09:32 antani joined #salt
09:43 peters-tx joined #salt
09:47 toanju joined #salt
09:57 rubenb Hi, where should one download the Windows Minion installer?
09:57 mattiasr joined #salt
09:58 matthew-parlette joined #salt
09:58 c10b10 joined #salt
09:59 markm joined #salt
10:00 evle joined #salt
10:00 rubenb I can only find the old (2014.7) windows installer on http://docs.saltstack.com/en/latest/topics/installation/windows.html
10:02 bhosmer joined #salt
10:04 fredvd joined #salt
10:04 pelzi__ trim the download url, and you get a directory listing with newer installers
10:05 pelzi__ http://docs.saltstack.com/downloads/
10:06 leapyleigh joined #salt
10:07 leapyleigh hello
10:09 kusams joined #salt
10:20 ndrei joined #salt
10:36 ilbot3 joined #salt
10:36 Topic for #salt is now Welcome to #salt | 2015.5.0 is the latest | Please use https://gist.github.com for code, don't paste directly into the channel | 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/
10:38 markm joined #salt
10:38 cromark joined #salt
10:38 forze joined #salt
10:41 pcdummy joined #salt
10:41 pcdummy joined #salt
10:42 zajk joined #salt
10:42 superseb joined #salt
10:42 favadi joined #salt
10:42 zajk left #salt
10:43 zajk joined #salt
10:43 rubenb pelzi__: Thanks.
10:50 cowyn joined #salt
10:54 leapyleigh hi, would like to be able to run a executable on a windows minion using cmd state module within an sls file. having probs. I have tried escaping the backslash to no avail...
10:54 leapyleigh https://gist.github.com/4752861d69392dffe9a3.git
10:55 zer0def joined #salt
10:55 programmerq joined #salt
10:55 Rory joined #salt
10:57 bhosmer joined #salt
10:59 pcdummy joined #salt
10:59 pcdummy joined #salt
11:00 giantlock joined #salt
11:01 pcdummy joined #salt
11:02 mr-op5 joined #salt
11:04 denys joined #salt
11:10 kusams joined #salt
11:11 markm_ joined #salt
11:11 dopesong joined #salt
11:13 PI-Lloyd has the salt-api changed in 2015.5 ? One of our jenkins jobs is failing on the response from the api call.
11:15 johtso joined #salt
11:16 johtso Are there any tools for analysing salt states? Would be quite useful to have something that showed you a tree of triggers so you could identify states that are running unnecessarily and what's causing them to run, and maybe with execution time in there too.
11:18 PI-Lloyd state.show_lowstate might be of some use johtso
11:18 teogop_ joined #salt
11:19 PI-Lloyd other show_ functions are available - http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.show_highstate
11:21 linjan joined #salt
11:22 johtso PI-Lloyd, does that basically show a tree of all your states?
11:23 johtso that's a bit unwieldy really, I just want to pinpoint problem points
11:23 johtso the simplest case being, I've just run highstate, and I want to work out why some states are being executed when I immediately do another highstate
11:24 johtso the state execution output doesn't tell you why it executed in the first place
11:25 c10b10 joined #salt
11:29 ndrei joined #salt
11:29 teogop joined #salt
11:39 c10b10 joined #salt
11:41 PI-Lloyd Wouldn't know where else to look other than the sls files themselves. Only other thing i can think of would be to look at the debug output of a highstate, but I'm not sure that will show you want you want to see either
11:41 c10b10_ joined #salt
11:41 PI-Lloyd there might be something buried in http://docs.saltstack.com/en/latest/topics/troubleshooting/ that may be of some use
11:43 mage_ I have a cmd.run: in an SLS state, how can I execute something after it finishes successfully ?
11:46 PI-Lloyd mage_: a 'require' on the other state should do it
11:46 Furao joined #salt
11:48 c10b10 joined #salt
11:49 mage_ ok :) another question: what's the best approach to install packages ? for ex let's say I have a DB server and a WEBAPP server, and some packages should only be isntalled on DB server, same for WEBAPP server
11:49 mage_ should I put those packages in the pillar and state with a jinja template ?
11:50 PI-Lloyd I would split them down into 'roles', so a role of 'WEBAPP' and a role of 'DBSERVER' which you have seperate states for, each state has the packages required for each system
11:51 PI-Lloyd if the two roles share some packages you could put those in a 'common' state and include them in the roles
11:51 mage_ make sense, thanks
11:52 c10b10_ joined #salt
11:52 joeto joined #salt
11:52 bhosmer joined #salt
11:52 amcorreia joined #salt
11:53 PI-Lloyd That's just the way I do it, everyone has a different approach, there isn't really a 'best' method afaik
11:57 leapyleigh This is how I did it:
11:57 leapyleigh https://gist.github.com/8767fd3a20ba0a3711fb.git
11:58 leapyleigh (answering my previous question)
12:05 tmclaugh[work] joined #salt
12:12 mapu joined #salt
12:13 fredvd joined #salt
12:16 c10b10_ joined #salt
12:17 dRiN joined #salt
12:26 otter768 joined #salt
12:26 zajk I've made a rkhunter formula, so if anyone needs it, it's currently at https://github.com/zajk/rkhunter-formula
12:28 colonD joined #salt
12:29 mage_ if I have 10 venvs to built on each machine is there of possibility to built them in parallel ?
12:30 Furao joined #salt
12:32 c10b10 joined #salt
12:34 joehh iggy: if you have any problems with the debian/ubuntu packages, please let me know - I'm also happy to help you learn to build your own
12:35 cowyn joined #salt
12:36 dendazen joined #salt
12:38 LotR joehh: now that jessie is released, will the sid/testing packages get updated more frequently?
12:40 subsignal joined #salt
12:40 c10b10 joined #salt
12:43 antani joined #salt
12:44 cmcmacken joined #salt
12:45 linjan joined #salt
12:46 julez joined #salt
12:47 zipkid is it possible to 'package' a custom module (the kind you put in '_modules') in a salt formula?
12:52 JDiPierro joined #salt
12:52 ferbla joined #salt
12:56 toastedpenguin joined #salt
12:57 c10b10 joined #salt
12:57 joyC joined #salt
12:58 joyC Hi guys! Anyone faced "AttributeError: 'TimedProc' object has no attribute 'stdout'"?
12:59 colonD joined #salt
13:01 toastedpenguin joined #salt
13:08 primechuck joined #salt
13:09 jdesilet joined #salt
13:10 babilen joyC: Do you?
13:11 hasues joined #salt
13:11 hasues left #salt
13:12 antani joined #salt
13:14 timoguin joined #salt
13:14 joyC Yep
13:15 joyC This error happens occasionally
13:17 babilen joyC: Could you paste the entire error to, say, http://refheap.com ?
13:18 Furao joined #salt
13:19 Furao joined #salt
13:23 inad922 joined #salt
13:24 subsigna_ joined #salt
13:25 markm joined #salt
13:25 c10b10_ joined #salt
13:25 racooper joined #salt
13:25 ndrei joined #salt
13:25 joyC Sure. https://www.refheap.com/101044
13:28 Deevolution joined #salt
13:29 Tecnico1931 joined #salt
13:31 evle joined #salt
13:33 bhosmer joined #salt
13:33 debian112 joined #salt
13:34 fredvd joined #salt
13:34 dyasny joined #salt
13:35 c10b10 joined #salt
13:36 oyvindmo joined #salt
13:38 ponpanderer joined #salt
13:39 mapu joined #salt
13:40 ksj is there a state for downloading a file, or should I do it with a wget?
13:41 ksj ....of course, just change the source of a file.managed....ignore me
13:41 Tyrm joined #salt
13:41 pdayton joined #salt
13:42 c10b10 joined #salt
13:44 thayne_ joined #salt
13:46 _JZ_ joined #salt
13:46 c10b10 joined #salt
13:53 bhosmer_ joined #salt
13:53 huddy joined #salt
13:55 zajk joined #salt
13:57 kaptk2 joined #salt
14:00 kusams joined #salt
14:02 primechuck joined #salt
14:02 rideh joined #salt
14:03 andrew_v joined #salt
14:03 Sacro Eugh, "from salt import __version__ as saltversion" no longer works
14:03 chiui joined #salt
14:09 hasues joined #salt
14:10 hasues Are there going to be any more training opportunities after June?
14:10 hasues http://saltstack.com/training/
14:10 hasues Does anyone know?
14:11 timoguin joined #salt
14:13 joyC joined #salt
14:14 emaninpa joined #salt
14:14 aawerner joined #salt
14:14 iggy joehh: my problem was that 2014.1.11+ was a total stinker of a release, so not necessarily with the packaging, but the releases themselves were terrible
14:16 iggy joehh: that said a little more documentation on how to stick to old repos (i.e. the 2014-7 repo or whatever it's called)
14:17 iggy joyC: did you look on the issue tracker? that sounds familiar
14:17 drawsmcgraw hasues: I'm sure there will be more training and it just hasn't hit the list yet
14:18 iggy ^
14:18 drawsmcgraw You can always send an email or ask for an on-site training if you have enough people.
14:19 JayFK joined #salt
14:19 thayne_ joined #salt
14:21 txomon|home iggy, do you use openssh-formula?
14:21 iggy no
14:21 txomon|home humm ok
14:21 hasues drawsmcgraw: Okay, I just needed to make sure it would happen so I could ask for budgeting to allow it.
14:21 txomon|home I was thinking on extending a little more the openssh.auth state
14:22 hasues drawsmcgraw: Onsite would be a possibility should this get "adopted"
14:22 txomon|home https://github.com/txomon/openssh-formula
14:22 txomon|home now it is aware of the sshd authorized keys path
14:22 iggy openssh is installed by default on my machines... doesn't really seem useful to do anything beyond that
14:23 drawsmcgraw hasues: I was lucky. I asked if they could do an on-site and they already had a similar request from someone nearby so my company only had to send the few of us that were interested.
14:23 txomon|home iggy, managing authorized files
14:23 hasues drawsmcgraw: nifty.
14:24 joyC iggy: no. haven't found anything so far
14:24 iggy yeah, that's already handled as well
14:25 txomon|home pff ok
14:25 iggy joyC: might check the irc archives then
14:26 iggy txomon|home: why do you ask?
14:27 txomon|home iggy, because in the servers we just have root with all the ssh keys of the authorized ones in /etc/ssh/auth/root
14:27 otter768 joined #salt
14:27 txomon|home and now, the formula doesn't support anything like that
14:27 gladiatr joined #salt
14:27 txomon|home so I changed that, but I still don't like the design of the formula itself, so I was looking for opinions on it
14:29 iggy txomon|home: remind me in about 30 minutes and I'll look at it
14:29 iggy off to work
14:29 txomon|home ok
14:30 MyTallest joined #salt
14:31 bezaban joined #salt
14:32 philipsd6 joined #salt
14:32 dendazen joined #salt
14:34 markm joined #salt
14:35 Brew joined #salt
14:35 julez joined #salt
14:36 c10b10 joined #salt
14:37 wytek joined #salt
14:37 wytek hello
14:37 wytek i have a file managed by salt
14:37 JayFK joined #salt
14:37 babilen so do we all!
14:37 wytek but when i run saltstack the file don't get create, there is no trace of it in debug
14:37 wytek is a bit weird
14:37 peno joined #salt
14:37 babilen Why not?
14:37 wytek any clue on why this can happen?
14:38 babilen wytek: Why do you think that salt should manage that file? Could you paste your top.sls, the state file in which you manage the file, your commands and their output to, say, http://refheap.com ?
14:39 hasues left #salt
14:39 wytek the file is inside a mongodb class
14:39 wytek other file are managed correctly
14:39 wytek this one should include the content of a pillar
14:39 wytek i use contents_pillar
14:39 wytek and when i check with get.pillar and the key
14:40 wytek i can see the pillar populated in the minion
14:40 wytek again, the file is working correctly
14:40 wytek i can paste this info, but not now unfortunately, i was wondering if you have a clue (cache for example?)
14:43 mpanetta joined #salt
14:44 bobby_ joined #salt
14:44 bobby_ hello, just getting started with salt
14:44 bobby_ what should be my first step?
14:46 rojem joined #salt
14:47 c10b10_ joined #salt
14:49 heaumer joined #salt
14:49 stephanbuys joined #salt
14:50 PI-Lloyd bobby_: http://docs.saltstack.com/en/latest/
14:51 stephanbuys hi all, I'm thinking of implementing something analogous to docker-compose using saltstack modules, I've got some sls files that currently do the job ok, but I'm wondering what the right way to do this is? I guess a "docker.compose_up" state is an option or perhaps a runner?
14:54 fyb3r joined #salt
14:54 c10b10 joined #salt
14:55 numkem who should I shake my fist at for not having 2015.0 in centos repos?
14:56 numkem epel, nvm my question
14:59 mage_ any idea for this: https://gist.github.com/silenius/d87fa6b23da455861245 ?
14:59 iggy wytek: it could literally be hundreds of things wrong... seeing the aformentioned output would help us immensely
14:59 wytek ok, i'll paste it later then!
15:00 drawsmcgraw bobby_:  Have a particular problem to solve?
15:01 chiui joined #salt
15:02 iggy mage_: where is gem installed? is it in the path listed there?
15:02 iggy can you run that command by hand?
15:03 mage_ no, the virtual env (rbenv) should be loaded first .. but that what that ruby: at line 61 is supposed to do ..
15:04 colonD joined #salt
15:07 jalbretsen joined #salt
15:07 jalaziz joined #salt
15:08 VR-Jack joined #salt
15:08 mapu joined #salt
15:11 iggy well, it doesn't look like it's trying to run some custom "gem", it looks like it's trying to run one that's in the path
15:11 mage_ mmh
15:12 iggy maybe set gem_bin?
15:12 mage_ let's try .. :)
15:12 iggy oh, the docs say use ruby instead
15:12 iggy I got nothing
15:12 mage_ yes
15:12 murrdoc joined #salt
15:13 iggy thankfully, I've never had to do ruby stuff
15:13 mage_ :)
15:13 linjan joined #salt
15:18 VR-Jack hmm, does weel have to run on master or can it run on a minion?
15:18 VR-Jack wheel.* even
15:20 N-Mi joined #salt
15:21 denys joined #salt
15:22 Parabola joined #salt
15:23 Parabola new user, it looks like the epel version is outdated Version     : 2014.7.5
15:23 sdm24 joined #salt
15:23 Parabola is this going to be an issue for a new implementation? ie should i do this outside of package management
15:24 TomJepp joined #salt
15:25 VR-Jack a couple days isn't outdated
15:25 IBMan joined #salt
15:25 TomJepp joined #salt
15:25 Parabola 7/5/2014 is how i'm interpreting that version. its 2015 now mate
15:25 iggy Parabola: epel should get 2015.5 soon enough (it's in epel testing afaik)
15:26 VR-Jack Parabola: it's the last version before the one that just released
15:26 iggy Parabola: what VR-Jack means is 2015.5 was just released on Monday (prior to that 2014.7 was the most recent release)
15:26 whytewolf Parabola: 2014.7.5 was released last month.
15:26 Parabola understood, my concern isnt that i must be on the newest, its more that if there are huge changes between the two versions i dont want to learn the outdated one
15:27 iggy they aren't huge
15:27 Parabola eg, when i started learning laravel it was with 4x, shortly after 5x was released
15:27 whytewolf the .5 is a patch version not part of the date
15:27 mage_ iggy: I used https://dpaste.de/oFhL .. a little bit ugly, but works
15:27 iggy but yeah, if you're starting fresh, just install salt from epel testing for the next couple weeks
15:27 Parabola and it wasn't a total "startover" but it was substantial :)
15:27 Parabola awesome, thanks gents i appreciate it
15:28 iggy Salt doesn't have turnover like web frameworks (well, hasn't yet anyway)
15:28 VR-Jack the new code has support for some stuff like sdb, but the core of salt in docs seems to be the same
15:28 Parabola awesome, you guys rock
15:28 iggy and the docs are all version annotated
15:28 iggy so you can see what features were added when
15:29 iggy ...for the most part :(
15:29 VR-Jack yeah, but I don't think I've run into a single "Lithium" in docs. Everything is for the next version
15:29 Parabola on page 20 of 1,760 of the manual :X this is going to take a while
15:29 iggy VR-Jack: lithium was changed to 2015.2 when it was branched in December
15:29 iggy (which has now been changed to 2015.5 ;)
15:30 VR-Jack iggy: even so. Just haven't had issues really
15:30 VR-Jack The feature list must be in sdb and stuff I don't use. heh
15:30 iggy VR-Jack: you'll see them more towards the end of the dev cycle when people are trying to rush in features and forget to add version annotations... people see stuff in the docs and try to use it... boom
15:31 iggy BEACONS!!!!!!
15:31 iggy that was all that anybody talked about at the conf
15:31 Parabola its awesome that my new employer (startup) is at a place where i can take the 'cycles' to implement config enforcement
15:32 VR-Jack Haven't read up on those. The main thing in ber that I like is some of the new module functions in file.* and stuff.
15:32 iggy oh my flying spaghetti monster... when will people quit trying to give clever names to security vulnerabilities
15:32 Parabola iggy never
15:33 Parabola people finding these things like to feel clever
15:33 iggy venom... nobody leaves the floppy drive enabled in kvm... stupid fucks
15:33 VR-Jack so the question is, can I run wheel.key.gen_accept from a minion or do I need to pop it through reactor?
15:33 iggy sensationalist pricks
15:33 iggy "I want my name in the news"
15:34 txomon|home joined #salt
15:34 murrdoc VR-Jack:  i did the reactor
15:34 VR-Jack guessing that wheel is designed for master only
15:35 Parabola iggy iirc thats been patched by xen and qemu already, not sure about kvm
15:36 iggy well, like I said, nobody leaves the floppy enabled
15:36 VR-Jack murrdoc: it's annoying. trying to handle the libvirt installs and bootstraps in a highstate.
15:36 CeBe1 joined #salt
15:37 mirko iggy: don't forget the logos!
15:37 mirko every security issue needs a logo nowadays!
15:38 iggy but now I have to look like a twat researching the problem by the name... oh wait, they actually have the CVE in the article
15:38 iggy good, don't have to give some security prick extra hits
15:39 iggy neat... 2015 3456
15:39 iggy easy to type
15:39 murrdoc iggy:  what are u on about
15:39 murrdoc VR-Jack:  what about saltify
15:39 Parabola ugh, so if zeroMQ4 is needed, and should be done prior to the salt install, why is it BELOW the install instructions in the documentation?
15:40 * Parabola grumbles
15:40 iggy murrdoc: https://marc.info/?l=oss-security&amp;m=143151580308576&amp;w=2  security vuln in kvm (qemu, xen, vbox, etc)
15:40 Rory left #salt
15:41 murrdoc wait floppy drive
15:42 iggy which every hosting company worth a damn already disables
15:42 iggy (already check GCE and DO)
15:42 murrdoc does it affect gce ?
15:42 iggy *checked
15:42 iggy no
15:42 murrdoc cool
15:42 iggy [    4.792146] floppy0: no floppy controllers found
15:42 iggy that's on gce
15:43 murrdoc whee
15:44 twork joined #salt
15:44 iggy and I'm pretty sure GCE runs their instances inside containers as well, so even less worry
15:45 iggy (which I also think any company worth a damn should also be doing)
15:45 conan_the_destro joined #salt
15:46 dfinn joined #salt
15:48 wendall911 joined #salt
15:49 mirko i don't get why this is xen specific
15:52 PI-Lloyd iggy you are correct with the GCE containers comment. Everything they do is containerized ;)
15:53 iggy mirko: it's not, that's just where the bug was first noticed
15:53 txomon|home iggy if you can have a look at the openssh formula =)
15:53 iggy anything that uses qemu (and indeed vbox "borrowed" some qemu code years ago and is vulnerable)
15:53 txomon|home btw, I sent a PR for supporting token expansion https://github.com/saltstack/salt/pull/23665
15:54 bhosmer_ joined #salt
15:54 iggy what's the url to your fork?
15:55 txomon|home iggy, same but under txomon/
15:55 txomon|home github.com/txomon/openssh-formula
15:55 txomon|home (IIRC)
15:55 txomon|home need to leave now!
15:58 VR-Jack hmmm. looks like my best option is to have minion event.send a key request, have reactor create/accept key and send the other key back to minion via event.fire, minion collects dict and continues on with bootstrap
15:58 kunersdorf joined #salt
15:58 VR-Jack presuming that event.fire can get called while a state.sls is still running.
15:59 iggy sure
16:00 ajw0100 joined #salt
16:00 murrdoc this is all bad
16:01 murrdoc why cant u pre seed the minion
16:01 murrdoc master cant do shit on a minion that isnt auth'ed
16:01 VR-Jack murrdoc: that is to preseed the minion. a minion is bootstapping another minion
16:02 OneAngryDBA joined #salt
16:02 penguin_dan joined #salt
16:02 iggy open_mode ftw
16:03 checho hi all...silly question here before breaking something.... can I use grain properties in the SLS files?   I.E.: I want to create a directory in my managed system using the minion hostname.... can I use something like /home/{{network.get_hostname}}  as my file.directory: entry?
16:03 iggy /home/{{ grains['nodename'] }}
16:04 checho iggy: tks!  grains...
16:05 MrThrax joined #salt
16:06 ek6 murrdoc: knowing im going to get blasted for ek6's bad idea of the day number 32 but......  For an event I ran an open_mode master which had the sole job of getting the minion up to speed and then passing in preseeded keys and different master pointer for final provisioning..
16:06 ek6 did this after people finding 100 different ways to mangle 'applying' a preseeded key...
16:09 murrdoc :)
16:09 murrdoc iggy said open_mode
16:10 TomJepp joined #salt
16:10 txomon joined #salt
16:10 murrdoc i use salt-api to get a preseded key and drop it on the minion
16:11 ek6 yeah but i couldnt I still think cant get master_sign.pub through that mechanic so its a non-starter for me
16:11 iggy we have a private network, if anyone gets into it far enough to install a fake minion, we're already fucked six ways to sunday
16:12 ek6 yeah...thats really it iggy...they have to be 'inside' to talk to the open_mode server to get the preseed that they can take and use 'outside' in rough terms
16:12 txomon iggy: thanks for the review, however, I was referring to the syntax the auth part now has, with the aim of improving it
16:12 ek6 murrdoc: I keep forgetting to file a FR to get them to supply signing pub in that key/tar api response
16:13 txomon I feel it's horrible way to specify keys (which indeed need to be told to be list values and not keys)
16:14 writtenoff joined #salt
16:18 JDiPierro joined #salt
16:20 solidsnack joined #salt
16:21 solidsnack joined #salt
16:23 KyleG joined #salt
16:23 KyleG joined #salt
16:23 solidsnack joined #salt
16:23 sdm24 Does anyone know if it is possible to not have the highstate run on a new minion, after the master accepts its key?
16:24 WildPikachu racooper, is only your states stored in git, nothing like reactor sls files?
16:24 KyleG joined #salt
16:24 KyleG joined #salt
16:24 racooper I don't use reactor so I don't really have any experience with that. right now just my states and and pillars in separate repos.
16:24 JayFK joined #salt
16:25 WildPikachu thanks racooper :)
16:25 sdm24 When my coworkers make new VMs, sometimes they don't follow my instructions, so the minion ID is left to the default name of the VM template, and I have no way of figuring out what actual server the minion is. I don't want to run the highstate after accepting the key, because I don't want to potentially break a new server
16:25 JayFK joined #salt
16:26 aparsons joined #salt
16:26 kawa2014 joined #salt
16:27 Parabola would be awesome if there were an autocomplete for the salt functions
16:28 otter768 joined #salt
16:28 SaltyCharles joined #salt
16:29 subsignal joined #salt
16:29 Parabola s/functions/commands
16:32 tkharju joined #salt
16:33 ageorgop joined #salt
16:33 VR-Jack sdm24: does highstate run automatically? I didn't think it did by default
16:34 sdm24 Mine does, and I don't have reactors set up. As part of the highstate, I install postfix on Linux servers. Unfortunately, I accepted the key of one of our mail servers, which removed the outward mail client to install postifx....
16:35 primechuck joined #salt
16:35 sdm24 Does refesh.pillar run the highstate or something? Because I did add it to the top pillar
16:37 smcquay joined #salt
16:37 sdm24 I do have the highstate on a schedule, and I might've just gotten lucky and had it run soon after I accepted the key..
16:38 VR-Jack sdm24: yeah. I don't have mine scheduled. highstate on key accept would totally screw up my orchestrate
16:38 conan_the_destro joined #salt
16:39 sdm24 I looked through the jobs cache and couldn't find a scheduled highstate running, but now looking  I think no scheduled highstates are in the cache
16:40 hackel joined #salt
16:41 solidsnack joined #salt
16:41 sdm24 although it probably did just run on schedule. Lesson learned though: I updated my state top.sls file to 'G@kernel:Linux and not mail*', so it won't install postfix on mailserver
16:41 jpic hi all, are there any plans to get rid of this error "state foo is unavailable" ?
16:41 iggy lol, "I've got the same vulnerability designation on my luggage!"
16:42 iggy spaceballs, the gift that keeps on giving
16:42 jpic i hate it so much, it can always be anything, nothing in logs
16:42 sdm24 comb the log files!
16:42 iggy Parabola: there is
16:42 soren_ joined #salt
16:42 ageorgop joined #salt
16:42 iggy jpic: there's usually more to the error message than that (if not, upgrade)
16:43 Parabola iggy salt 'host' <tab> doesn't work, i'll look into it :)
16:43 markm joined #salt
16:43 jpic iggy: yep, sometimes there's more in the logs
16:44 jpic and we've already upgraded
16:44 VR-Jack sdm24: My whole setup is being designed around the master pillar defining everything. That way I don't have issues with unknown minion.
16:44 iggy Parabola: https://github.com/saltstack/salt/blob/develop/pkg/salt.bash some distros don't install it
16:44 Parabola oh awesome
16:45 TomJepp joined #salt
16:45 Sketch iggy: i thought it was overblown as well, until i saw this comment about it:
16:45 Sketch "For many of the affected virtualization products, a virtual floppy drive is added to new virtual machines by default. And on Xen and QEMU, even if the administrator explicitly disables the virtual floppy drive, an unrelated bug causes the vulnerable FDC code to remain active and exploitable by attackers."
16:45 UtahDave joined #salt
16:45 hal58th_ joined #salt
16:46 sdm24 VR-Jack: when a new minion talks to the master, it uses the VM template id. I have a setup state on the template to change and configure everything, but people don't use it. After accepting the generic minion, I can then change the ID and restart the minion. But the problem I'm waiting for is if the scheduled highstate runs and the minion gets files that it shouldn;t or something
16:46 sdm24 basically I'm just paranoid
16:47 ek6 sdm24:  can you use startup_states   ?  or would they just screw that up as well
16:47 spookah joined #salt
16:47 iggy Sketch: that's the drive, most places have the entire controller disabled
16:48 iggy Sketch: at least the ones I've looked at today (GCE and DO)
16:48 iggy jpic: what version?
16:48 sdm24 ek6: That could possibly be one fix, but I would still have to accept the minion before it runs
16:48 sdm24 ek6: Or can I use startup_states to run local files saved on the minion, so I don't need to accept the minion first?
16:49 timoguin joined #salt
16:49 iggy if you are using salt-cloud, doesn't it do all the key stuff for you?
16:49 sdm24 We are not
16:49 ek6 sdm24: sure..my bad..i thought your issue was that after accepting it was running highstate right away
16:49 sdm24 No worries, I'm basically complaining about 2 different things now haha
16:50 ek6 sdm24:  well again that depends on what you can feed..you can certainly do a short duration file_client local   startup_states blah  have those execute and have the last thing it executes is to remove the file_client local and point up to real master
16:50 sdm24 ooh smart
16:51 theologian joined #salt
16:51 hal58th_ sdm24, ek6 took the words out of mouth. Seems a pain, but if you really really need it, go for it
16:52 sdm24 Thanks for the tips guys
16:52 murrdoc joined #salt
16:52 ek6 dont know about smart..i would say clever....of course clever in this channel is often used as a pejorative so ymmv
16:52 jpic iggy: 2014.7, just upgraded to 2015.05, running the test suite, seems pretty backward compatible ;)
16:53 Edgan joined #salt
16:53 jpic still got empty errors: http://dpaste.com/182G2TY
16:53 tzero joined #salt
16:54 ageorgop joined #salt
16:54 tzero joined #salt
16:54 Parabola this product is amazeballs
16:54 Parabola that is all
16:57 murrdoc go Parabola
16:57 Parabola go where?
16:57 Tyrm_ joined #salt
16:57 Parabola :( i dont wanna leave
16:57 murrdoc like you go
16:57 murrdoc like go Team
16:57 smcquay joined #salt
16:58 murrdoc go go go Parabola !!1 go go go!
16:58 Parabola oh hehe
17:03 sandah joined #salt
17:04 MatthewsFace joined #salt
17:08 Matthews_ joined #salt
17:13 markm joined #salt
17:13 thayne_ joined #salt
17:14 iggy jpic: oh, that means you are missing some dependency (the jenkins_plugins state isn't loading)
17:15 jpic thanks for the clue ;)
17:15 iggy and that looks like something custom, so you're on your own fixing that
17:15 iggy but usually it's a binary missing or some python lib missing
17:16 jpic what do you mean the state isn't loading ? did it raise an exception with a traceback or something ?
17:16 jpic maybe saltstack is swallowing ImportErrors ?
17:16 iggy it does, yes
17:16 jpic i'd love this to end :D
17:16 iggy otherwise, you see errors from every module under the sun
17:17 iggy (i.e. it makes no sense to throw an error about the yumpkg module not loading on Debian)
17:17 murrdoc saltstack will swallow import errors
17:17 murrdoc but it will log a 'so and so module doesnt exists' error in the log file
17:17 jpic if a module is available and raises an exception, it sounds wrong to report that the module is unavailable
17:17 drawsmcgraw So... many people here have experience/opinions on Sensu?
17:17 jpic it's *not* unavailable, it's *failing*, imho that's a lot different
17:18 murrdoc pretty much if u are writing python in salt, you are using absolute import, and there fore should use try/catch around all includes in your code
17:18 iggy ^
17:18 iggy it's up to you to write useful error messages in your own code
17:18 jpic while that makes sense, it doesn't change that imho, there is a different between a module that's not there, and a module that's failing.
17:19 jpic maybe i'm the only one here with that opinion ?
17:19 iggy it does... a failing module isn't "there"
17:19 iggy see the example I gave earlier
17:19 iggy (i.e. it makes no sense to throw an error about the yumpkg module not loading on Debian)
17:20 iggy and it makes no sense to throw errors about raid admin utils not existing on VMs
17:20 jpic well we have those: http://dpaste.com/16Q5ED0
17:20 nzero joined #salt
17:20 forrest joined #salt
17:21 iggy there's a bug about that already
17:21 iggy last I looked it was fixed in devel, waiting on backport
17:21 jpic not complaining about the bug, what i mean is that in this case, i can haz traceback (thank god)
17:22 murrdoc joined #salt
17:22 cpowell joined #salt
17:22 jpic maybe we could have tracebacks with -l trace at least ? would that make sense ?
17:22 yorjo joined #salt
17:23 iggy file a feature request
17:23 murrdoc there is a trace log
17:23 murrdoc i forget what it is
17:23 murrdoc might even be -l trace
17:23 iggy it is
17:24 jpic yeah -l trace is an option
17:24 iggy doubt it'll show you tracebacks though
17:25 jpic well it would make sense if it did wouldn't it ?
17:26 iggy jpic: looks at all the things in salt/{states,modules}/ and think about how many of those you use on a daily basis... now imagine _every_ other one of them filling your logs with garbage
17:26 dalexander joined #salt
17:26 Tyrm joined #salt
17:27 fbergroth joined #salt
17:27 jpic well i'd rather have useful info amongst a lot of garbage than nothing useful xD
17:27 cpowell joined #salt
17:27 jpic but i guess that's why i like -l trace
17:27 Ahlee where do we find these people
17:28 prwilson joined #salt
17:28 VR-Jack the Internet?
17:28 tmclaugh[work] joined #salt
17:28 jpic keep in mind that if you're doing IoC then swallowing ImportErrors **in all case** is probably not the best practice in Python
17:29 jpic Ahlee: i'm sorry ?
17:29 Ahlee jpic: sorry, wrong channel.
17:29 iggy that's 663 things possibly spewing things at you
17:29 Rory joined #salt
17:29 iggy (in devel anyway)
17:30 Gareth morning morning.
17:30 jpic iggy: i don't understand
17:31 mohae joined #salt
17:31 jpic exceptions should happen only in exceptionnal error cases, it's not a substitute for the goto statement
17:31 cpowell joined #salt
17:31 jpic if something is excepted to fail, then it should not be loaded at all
17:31 timoguin joined #salt
17:31 iggy it's expected to fail _if_ the requirements aren't met
17:32 iggy which is probably going to be a lot of different modules for most people
17:33 jpic i'm not convinced it's a good practice to expect exceptions as a normal part of a program's flow, why would module be loaded if they are expected to fail ?
17:33 iggy well, prior to the LazyLoader, salt tried to load everything it could and just left out whatever failed
17:34 murrdoc lazyloader was in 2014.7.2 /
17:34 murrdoc or something
17:34 jpic imho there's room for improvement
17:35 jpic if only module which are not expected to fail are loaded, then you'll only have relevant exceptions.
17:35 murrdoc but salt will quietly fail if you write bad python
17:35 iggy as I said, open a feature request
17:35 murrdoc write better code is what i think people will come back wit
17:35 murrdoc with*
17:35 iggy how is salt supposed to know what's expected to succeed?
17:35 iggy you want to keep that in a list for each host you have?
17:35 iggy sounds... tedious
17:36 jpic there has got to be a way to improve error handling in salt
17:36 iggy it's actually not too bad... once you read/understand the messages
17:36 jpic zfs module for example: if i don't have the zfs kernel module, then why load the zfs module ?
17:37 jpic well i guess your python is much better than mine then ;)
17:37 iggy that's a bug
17:37 iggy you should stop hyping that as an example of salt being bad
17:37 jpic what ?
17:37 whiteinge PI-Lloyd: what command are you running through salt-api that changed in 2015.2?
17:37 iggy the zfs thing... it's a bug, it's been fixed already (just hasn't made it into every branch)
17:38 jpic ok so we can't talk about the zfs thing because it's related to a bug in salt
17:38 jpic let's use another example, dmidecode / GPU grain
17:39 jpic if dmidecode is not there, salt logs that it isn't and that it's can't load any GPU grain. For me, that's perfect
17:39 pheer1 joined #salt
17:39 bluenemo joined #salt
17:39 iggy right, because someone specifically put that in the module
17:39 ageorgop joined #salt
17:39 iggy so take your jenkins modules, make them log errors properly, problems solved
17:40 iggy I don't think it's a generic problem that needs to be solved
17:40 jpic ok, so exception swallowing is a feature of salt, got it
17:40 iggy go back to my yumpkg on Debian example... nobody wants to see all those errors for stuff they aren't ever going to use
17:41 jpic ok, so loading yumpkg on debian would cause an exception to blow up at some point. Question: why load it at all ? "just in case" ?
17:42 jpic if somebody runs a yumpkg state on Debian, then it should blow up, because that's a bug in the user code right ?
17:42 iggy that's how module loading works... it tries to load everything it can (*historically)
17:42 iggy then they'd get an error like you got, saying yumpkg wasn't available
17:43 cpowell joined #salt
17:43 jpic i can understand that of course, perhaps it'd be a bit better to parse sls first and then load desired modules ? in which case we could let them blow up without swallowing i guess
17:43 rojem joined #salt
17:44 nzero jpic, i dunno man. your argument feels a bit like asking why bash lets you type in the yumpkg command on Debian. why doesn't bash know that doesnt make sense and stop you from doing that
17:44 jpic nzero: it's not going to try yumpkg unless i actually type it
17:45 jpic if i type yumpkg on debian, it'll blow up
17:45 jpic if i don't, it won't try ...
17:45 jpic i'm specifically loading this module from my code, so attempting to load it is perfectly legit imho
17:46 jpic but attempting to load modules which i'm not calling from my code sounds like a good solution for prototyping, but maybe not the best option on the long run
17:46 rhettg joined #salt
17:47 jpic nzero: it's a bit more like salt tries yumpkg on debian even though i didn't type it, hence ImportError swallowing
17:48 nzero jpic, no, you did tell it to use yumpkg. you just didnt tell it not to limit it to the right OS
17:48 WildPikachu can I use any of the builtin modules, like  'aliases' to set system aliases in my statefile?
17:49 jpic nzero: in which case it's my mistake ?
17:49 markm joined #salt
17:49 jopecko joined #salt
17:49 jpic "oh no, i'm calling yumpkg on debian, i'll just wrap it in the right if statement and everything is all right"
17:51 jpic nzero: if i'm calling a module that won't work then it's my mistake, i guess even here an exception would be all right if i haven't yet added an ERROR log and graceful exit.
17:52 jpic nzero: but in the case of saltstack, it's loading **all** modules, even those which i haven't called (be the call correct or a failure), so i don't think it's quite the same
17:52 jpic but i'm confident that there will be a module loading refactor at some point
17:52 Parabola where can i find information on automatic enforcement of a specific state(s) on specific node(s) ?
17:53 twork can i get a quick rtfm hint? what's the distinction between state and highstate? or... between 'state.highstate' and 'state.sls'? again, not asking for explanation, just a url, and googling for these is leading me everywhere.
17:53 Parabola eg, if a config file is changed, i want it to automatically be put back
17:53 murrdoc rtfm saltstack beacon
17:53 Parabola me or twork?
17:53 jpic twork: you'll like that one http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html
17:54 murrdoc twork:  state.highstate is a state.sls run which figures out the states to run by looiking at top.sls
17:54 Parabola i read through the states page, i didnt see what i thought i should be looking for
17:54 murrdoc u parabola
17:54 nzero jpic, you're venting in a non-constructive way. there is no way to respond to everything you are saying. you may have valid points. hard to say at this point because i'm not even sure what you root issue is. just seems like you hate saltstack
17:54 Parabola kk
17:54 Parabola yep thats it thanks!
17:54 ek6 parabola: look at inotify to trigger a reactor that would refresh your file back to the way you like it...
17:55 Parabola exactly it, thanks again gents :)
17:55 jpic nzero: .... really ... i'm trying to prepare a potential pr ...
17:55 bhosmer_ joined #salt
17:55 murrdoc we need a tip jar
17:55 murrdoc like we should have issues in saltstack-formulas
17:55 seev twork state.sls is for a single state execution, highstate assembles all states
17:55 ek6 although now that i think about it would that work....or would you go into a inotify reactor deathspin as your own state changes the file?
17:55 jpic ok guys, this has become a bit too emotionnal for me, seeya later, sorry for the misunderstandings, keep up the great work !
17:55 murrdoc and get people to solve an issue in the saltstack formulas area
17:55 Ryan_Lane basepi: the 2015.5 release notes don't mention anything about new modules and states :(
17:55 murrdoc everytime they are happy with the service provided
17:55 Ryan_Lane it doesn't cover any of the cloud additions either
17:56 basepi Hrm, we should definitely get those added.
17:56 Ryan_Lane the AWS support in salt (especially this version) is *really* good. better than terraform, ansible, chef, etc.
17:56 Ryan_Lane we should be banging the drum
17:56 basepi Feature release notes have the same problems as dot release notes, except compounded exponentially
17:56 basepi Something like 14000 commits since 2014.7
17:57 Ryan_Lane another good reason for smaller, faster releases ;)
17:57 basepi Agreed. That's the plan. =)
17:57 nzero jpic, good luck
17:57 basepi Would you mind doing a quick write-up on your AWS work? You know it much better than I.
17:57 basepi Ryan_Lane: ^
17:57 Ryan_Lane really, though, this release is massive and has lots of awesome things, but the release notes are a bit underwhelming
17:57 * basepi nods
17:57 murrdoc u need to use jira basepi
17:57 murrdoc :D
17:57 Ryan_Lane releases are basically a time for marketing
17:58 basepi Definitely.
17:58 seev send me a packet of free t-shirts, I will be your brand ambassador'
17:58 basepi We have no t-shirts at this moment, which is completely untenable.
17:58 * basepi yells at Rhett
17:58 seev I've implemented Salt in two tech startups so far, it's spreading
17:58 Ryan_Lane hell, I don't even know what work we have included in 2015.5 :)
17:58 basepi seev: awesome!
17:59 hemphill joined #salt
17:59 Ryan_Lane because we push everything into develop way, way before it's released
17:59 seev I even see it popping up in job postings around here, along with Chef, Puppet, Ansible
17:59 Ryan_Lane then our changes just get included with the release
17:59 twork much thanks guys. just what i needed. rtfming i go.
18:00 Ryan_Lane hm. I wonder what branch the pagerduty stuff went into
18:00 ek6 so can anyone verify that inotify beacon on a file tripping a reactor to reset that file doesnt again trip the inotify beacon causing a loop?
18:00 Ryan_Lane that state/execution module is probably press-release worthy with the PD folks :)
18:00 toastedpenguin the hostname of a minion changed prior to the key being accepted on the master, the previous hostname is showing up in the list of unaccepted keys, if I reject this key how do I get the minion to resend?
18:00 basepi 1617 files changed, 137833 insertions(+), 32295 deletions(-)
18:00 basepi ^ diff from 2015.5 to v2014.7.5
18:01 markm joined #salt
18:01 Ryan_Lane basepi: there's also no mention of sdb ;)
18:01 Ryan_Lane is there some reason that's being kept secret?
18:01 Ryan_Lane there's no docs either
18:01 solidsnack joined #salt
18:01 VR-Jack No one is sure what it is or what it does. :P
18:01 bastion1704 joined #salt
18:02 Ryan_Lane it's just a key/value that can be used from any part of the system
18:02 Ryan_Lane including states
18:02 bastion1704 Hello using ubuntu can we force the version of salt-minion ?
18:02 VR-Jack hehe. yeah. I figured as much.
18:02 Ryan_Lane abstraction for a key/value DB, that is
18:02 VR-Jack the long needed variable state system
18:02 Ryan_Lane really need to add a dynamodb sdb module :)
18:03 bradturber joined #salt
18:03 tmclaugh[work] joined #salt
18:03 linjan joined #salt
18:04 bastion1704 I do not undersant why the deb repo only have the latest. we cannot add new machines anymore because of this limitation
18:05 bradturber is it valid to use source: salt://dir/files/filename paths with archive.extracted? I am getting "failed to download... errors
18:08 monkey66 joined #salt
18:11 baweaver joined #salt
18:13 Fiber^ joined #salt
18:14 druonysus joined #salt
18:14 druonysus joined #salt
18:16 timoguin joined #salt
18:20 doug joined #salt
18:21 doug it looks like i've got to set aliases.file within each minion's config file -- is there a way to do that on the master instead?
18:21 doug in order for alias.present to find /etc/mail/aliases
18:21 denys joined #salt
18:23 jespada joined #salt
18:24 rojem joined #salt
18:26 evidence is it not possible to access utils in state files?  salt.utils.gen_mac() doesn't seem to be exposed to the lazy loader
18:26 bastion1704 can 2015.5.0 talk to old salt-minion ?
18:26 evidence tried salt['util.gen_mac']() and saltutils['gen_mac']()
18:27 evidence bastion1704: yes, i have my master on 2015.5.0 and most minions still 2014.7.5
18:27 evidence always upgrade the master first
18:27 dendazen with salt how would i disable service so it goes away from chkconfig?
18:28 dendazen would that be included in service.running: - enable: False?
18:28 Nicop06 joined #salt
18:28 evidence dendazen: service.dead with enable: False
18:28 dendazen or that will just make sure that the process isn’t running?
18:28 dendazen oh, thanks
18:28 dendazen That’s what i need
18:28 evidence service.running will start it but make sure it's disabled in chkconfig
18:29 evidence (with enable: false)
18:29 otter768 joined #salt
18:29 dendazen well i want opposite if it running, then stop it
18:30 dendazen so i need service.running false + service.dead
18:30 evidence i might be missing something.. what is the desired state?
18:30 dendazen to make sure both goals get achieved.
18:30 lictor36 joined #salt
18:30 evidence service.dead makes sure the service is not running, adding enable: false to that makes sure it's also not set to start via the system init scripts
18:31 dendazen oh ok, thanks.
18:32 evidence though there are some small issue with how salt deals with return code from status checks currently, you'll see errors at every state run because status commands return 1 when a service isn't running, so salt reports it as an error in the minion log
18:32 evidence not a huge deal though
18:32 markm joined #salt
18:33 dendazen hmm, not too good.
18:33 dendazen can i  bypass that somehow?
18:33 evidence write a patch for it? :)
18:33 evidence it's just because salt does a service status check before it takes any action, and after the service has stopped, most OSes return 1 on not running
18:34 rackermanfred joined #salt
18:36 dendazen should be an easy evaluation, no?
18:36 dendazen service not running? No. Don’t evaluate exit status
18:37 evidence yeah, probably needs to parse the return string in some cases.  FBSD returns a 1 if the service doesn't exist and if the server isn't running.  you probably do want to error in the no exist case
18:37 evidence some return different codes for them though, so easier in those cases
18:38 evidence srsly though, has anyone tried to use any of the functions under util in the states?  I guess they might not be exposed?  salt['util.gen_mac']() would be the syntax followed for modules and such.. maybe utils are exposed another way?
18:40 baweaver joined #salt
18:41 neogenix joined #salt
18:42 doug left #salt
18:42 dendazen i did both of those in the same state:
18:42 dendazen https://gist.github.com/anonymous/148490d4163bb6e29865
18:43 dendazen and it gives me an error.
18:43 dendazen ID 'rsyslog' in SLS 'services/syslog-ng/disable_rsyslog' contains multiple state declarations of the same type
18:43 evidence you can have 2 service types in the same state
18:43 hunmaat joined #salt
18:43 evidence err can't
18:43 toastedpenguin is there any manual way to get a minion to register with the salt master?
18:43 evidence rsyslog_dead and rsyslog_alive, break em up
18:44 evidence toastedpenguin - salt-call state.highstate --master={hostname} is a way to bootstrap
18:44 evidence assuming the master has rules to write a proper minion file, you can boot strap all hosts w that oneliner
18:44 dendazen break them up by different state sls files?
18:44 evidence dendazen: no, just under different names
18:45 dendazen oh ok.
18:45 toastedpenguin I'll try that, I rejected a minion's request, deleted the keys from the minion and recreated them on the minion but it has yet to attempt to re-register
18:45 evidence restarting the service should do it also
18:45 evidence service salt-minion restart
18:45 neogenix So, I'm using the schedule.add, and that seems to work, but then the schedule.list doesn't show the jobs, nor can I find them anywhere on the filesystem. Anyone have any ideas?
18:45 toastedpenguin I did that and nothing
18:45 evidence if the minion already has a proper conf file, you can drop --master from the salt-call also
18:45 dendazen can it be under rsyslog: both of them though?
18:46 dendazen like rsyslog: …
18:46 dendazen rsyslog …
18:46 evidence dendzen: no, you can only use a service once per name
18:46 evidence toastedpenguin: did you delete the rejected key from the master?  salt-key
18:47 dendazen so i should use -name in the state then explicitly?
18:47 toastedpenguin evidence: no, where are the rejected keys stored
18:48 evidence toastedpenguin: run salt-key
18:48 evidence they are in /etc/salt/pki on Linux though
18:48 evidence /etc/salt/pki/master/minions_rejected/
18:48 tomh- joined #salt
18:49 xcyclist joined #salt
18:49 evidence dendazen: check your gist
18:49 toastedpenguin yeah I did delete it....I used salt-key -d minion
18:49 evidence you might just be testing things.. but you shutting the service down and enabling at boot, then starting the service and disabling it at boot
18:51 kitplumm_ joined #salt
18:51 evidence alright no utils functions outside of pyobjects i guess.. gl gais
18:52 dingo in archive.extracted, for tar, does its '- name:' indicate the -C /dest/path tar option? Or should I secify my own -C?
18:53 bastion1704 joined #salt
18:56 giantlock joined #salt
18:56 markm joined #salt
18:58 nickdew joined #salt
18:58 spookah joined #salt
18:59 nickdew Hi, I've had a pull request open for salt-vim for over two months. I was wondering if someone could look at it and approve it?
18:59 hackel joined #salt
19:00 dendazen Thank you
19:03 soren_ joined #salt
19:05 WildPikachu is there a way I can require a file.directory for a file  I want to create inside it?
19:06 forrest WildPikachu: Yes, just create the file.directory, then use a - requires: \n  - file: file_directory
19:06 WildPikachu ah!
19:06 WildPikachu so to require a file or directory uses the same syntax forrest ?
19:07 murrdoc nickdew:  thats not us
19:07 murrdoc thats the mainline salt dev
19:07 murrdoc we are just people who use salt and help others because we like salt
19:07 murrdoc and saltstack-formulas
19:07 murrdoc your best option is to old school *bump* the pull requets
19:07 forrest WildPikachu: Yes, any time you have something that is part of 'file' it simply requires file, that's why you cannot have two file stanzas within the same ID. Also take a look at 'makedirs': http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html#salt.states.file.managed
19:07 nickdew murrdoc: Cool. Thanks.
19:08 forrest whiteinge are you around? nickdew has some salt-vim stuff he's been waiting to have reviewed
19:08 tmclaugh[work] joined #salt
19:08 forrest nickdew: I just pinged the maintainer, we'll see if he has time to respond today.
19:08 nickdew forrest:Thanks!
19:09 WildPikachu forrest, its actually for .ssh so it needs perms of 700 , so I decided to use file.directory to do that
19:09 forrest nickdew: You got it, if he doesn't respond he'll probably see it anyways and check out the PR.
19:09 forrest WildPikachu: Makes sense.
19:10 nickdew forrest:I did also comment on the pull request, so hopefully.
19:10 dendazen https://gist.github.com/anonymous/7a0cacd6b2db447dd5b6 hmm service still running after i ran that state.
19:11 forrest I saw. I figure most of the devs have so many repos being watched that sometimes PRs slip through the cracks :)
19:11 forrest dendazen: Looks like you have two IDs which turn off, then turns on rsyslog?
19:11 ajw0100 joined #salt
19:11 nickdew forrest:I totally understand. I'm no stranger to signal to noise ratios.
19:12 dendazen oh never mind i have supervised which starts it up, sorry forgot about that.
19:12 forrest nickdew: For sure, just wanted to make sure you didn't think you were being ignored on purpose. Thankfully that's pretty rare within the salt project(s)
19:12 spookah joined #salt
19:13 dirkraster joined #salt
19:13 nickdew forrest:Thanks for your help! And, I didn't think that.
19:13 forrest I'm glad you mentioned that one actually, as I kept forgetting to install the jinja vim plugin
19:13 forrest Cool!
19:14 baweaver joined #salt
19:17 izibi joined #salt
19:18 hax404 joined #salt
19:18 nzero joined #salt
19:20 forrest basepi: Are you around?
19:21 WildPikachu would file.append append the same text multiple time if for instance I use it in my top.sls and highstate a minion multiple times?
19:21 dirkraster is anyone aware of a salt state (or example) that uses salt.modules.win_service to manage the creation/deletion of windows services?
19:22 forrest nickdew: Thanks for bringing that repo to my attention, I've just gone through and commented on all the existing open issues as well to try and get some of those closed/addressed if we can. Looks like there hasn't been a ton of maintenance. I'll talk to the salt guys and see if we can get that repo managed more.
19:22 solidsnack joined #salt
19:22 forrest WildPikachu: I know that it used to.
19:22 * basepi hides
19:22 forrest WildPikachu: I'd double check, there might have been stuff added in the docs to address that.
19:22 forrest basepi: Is Seth around today?
19:22 basepi I think he's working from home. whiteinge?
19:23 forrest he didn't respond, wasn't sure if he was on a job or not
19:23 WildPikachu forrest, reading that in the docs right now, it doesn't mention it as far as I can see, I"m trying to find out how I can write the fqdn to /etc/mailname :)
19:23 forrest WildPikachu: Can't you just make the whole thing a managed file?
19:23 WildPikachu forrest, but I get the fqdn from a grain :)
19:23 forrest basepi: If he's around I'm sure he'll see.
19:23 forrest WildPikachu: okay? So pop the grain into the managed file :)
19:23 nickdew forrest:I just got all the GH notifications. If any of the issues are still present, I'll see what I can do to help.
19:24 forrest nickdew: Sounds good, most of them are just readme updates and a few other things
19:24 WildPikachu forrest, using contents: right?
19:24 forrest nickdew: My vim fu isn't good enough to address that one from back in 2013.
19:24 forrest WildPikachu: what? Nope, just straight jinja
19:24 forrest WildPikachu: hang on a sec.
19:24 WildPikachu forrest, contents_grains? :)
19:25 bradthurber joined #salt
19:25 forrest WildPikachu: lol no, look at this: https://github.com/gravyboat/hungryadmin-sls/blob/master/salt/hungryadmin/files/hungryadmin.conf
19:25 forrest see how I use pillar.get?
19:25 forrest You can just use grains.get
19:25 forrest easy
19:25 WildPikachu forrest, one sec, I"ll show you what I came up with, I think its even more awesome :)
19:26 nickdew forrest:Yeah, I'm not sure about that one either.
19:26 forrest http://docs.saltstack.com/en/latest/ref/states/vars.html#grains would be the example.
19:26 forrest nickdew: Vimscript booo
19:26 supersheep joined #salt
19:26 nickdew Vimscript isn't that bad... blows php out of the water. :D
19:27 JlRd joined #salt
19:27 forrest heh, I'm just not willing to spend time learning it
19:27 Ludo- ES_HEAP_SIZE={{ grains['mem_total'] / 2 }}
19:27 WildPikachu http://paste.linuxassist.net/view/beb274ce  <= forrest :)
19:27 Antiarc joined #salt
19:28 Ludo- how can I say to salt to get a int instead of a float out of it?
19:28 forrest basepi: Anyways, how do we discuss what gets 'watched' by the maintainer crew now for PRs and such? Salt-vim needs a little love.
19:28 nickdew I mess with it out of necessity more or less.
19:30 Ludo- more a jinja question I guest
19:30 Ludo- guess
19:30 basepi forrest: I'll bring it up, thanks for pointing that out.
19:31 forrest WildPikachu: If it's just one line, yeah you could just use contents/contents_grains. I'm not sure what the contents_grains format looks like, probably just fqdn going off the contents_pillar example. I've created https://github.com/saltstack/salt/issues/23683 to get more details added.
19:31 forrest basepi: Awesome thanks a bunch! I made comments on all the existing issues so we can try to get those sorted out.
19:32 forrest just the one PR waiting which looks good as well.
19:32 WildPikachu forrest, it does actually say it behaves exactly like contents_pillars :)
19:33 WildPikachu "same as contents_pillar, but with grains"
19:33 WildPikachu I do suppose it could be expanded
19:33 forrest Ludo-: use int: http://jinja.pocoo.org/docs/dev/templates/#int, so your example would be ES_HEAP_SIZE={{ (grains['mem_total'] / 2)|int }} I think
19:34 forrest WildPikachu: There's no reason not to provide a more detailed example. If it saves someone 5 minutes, or one question in the IRC it's worth it to me.
19:34 WildPikachu thanks forrest
19:35 Springwater joined #salt
19:36 keimlink joined #salt
19:37 Springwater I'm trying to share state files between my dev and qa environments but I haven't been able to figure out how to do this. I'm trying to keep the state files in the base environment and reference them in the top files for the different environments but it's not working. Is this the wrong approach?
19:37 forrest WildPikachu: np
19:38 keimlink joined #salt
19:38 forrest Springwater: Could you please gist.github.com your top file so we can see an example? I don't see why your approach would be wrong.
19:38 forrest basepi: Thanks for merging that PR in!
19:39 basepi =)
19:40 ckao joined #salt
19:41 Springwater Here's what I'm doing on my top.sls: https://gist.github.com/anonymous/eae30445506e140e7042
19:42 babilen Hooray for Venom :(
19:42 Whissi joined #salt
19:44 Springwater We have our environments split up into git branches
19:44 forrest Springwater: Refresh, I've added a comment to your gist
19:44 Springwater where the top file is in the master
19:45 Springwater there's a branch for dev, another for qa, etc.
19:46 Springwater I put the test sls file in the master with the top file but can't seem to reference it from the dev environment
19:47 forrest Springwater: What do you mean by 'reference it from the dev environment'?
19:47 forrest all servers are joined to the master, certain ones have dev. Did you set the file roots for dev in the master conf?
19:49 gulti joined #salt
19:49 Springwater So in the top.sls under dev: I'm trying to make a state file apply to certain dev nodes. That state file doesn't exist in the dev branch
19:49 baweaver joined #salt
19:49 Springwater I'll have to look at how the file roots are set up
19:51 forrest okay
19:51 Springwater Looks like we have fileserver_backend: -git
19:51 Springwater So file roots aren't set
19:51 hybridpollo joined #salt
19:51 gulti hi...can someone help me with this watch on an nginx service      - watch:       - file: /etc/nginx/conf.d/*    says the following requisites was not found  file: /etc/nginx/conf.d/* what else should i add?
19:52 Antiarc You need a salt state for those files
19:52 Emantor joined #salt
19:52 MatthewsFace joined #salt
19:53 Antiarc /etc/nginx/conf.d: file.directory, - recurse: true, - watch_in: service: nginx
19:53 Antiarc That will manage your conf.d directory and restart nginx when salt changes things in it
19:53 markm joined #salt
19:53 bastiandg joined #salt
19:54 forrest gulti: Think of it like this, when you tell salt to 'watch a file' it is looking for an ID. It is not watching the file(s) directly, but the ID associated with those files, so whatever you did when you put my_nginx_config:\n  file.managed:
19:54 forrest you would want to change that to watch the file my_nginx_config
19:54 forrest does that make sense? I can gist an example if not
19:54 gulti ah thanks...i tried /etc/nginx/conf.d/ without the recurse true and it gave the same error.
19:55 scoates joined #salt
19:55 solidsnack joined #salt
19:55 forrest gulti: Right, because that's probably not an ID you have set up somewhere correct?
19:55 timoguin Any of you guys had an issue with GitFS suddenly not pulling down one of your remotes? I'm not getting any errors but one of my main repos is show "gitfs received 0 objects" in the master log.
19:55 forrest gulti: If you want to gist your state, we can go over it
19:55 timoguin Just upgraded to 2014.7.5.
19:55 Antiarc salt isn't going to watch that directory for external changes, it'll just keep it synced to whatever source you set, then execute the watch_in when it makes a change
19:55 bhosmer__ joined #salt
19:55 iggy timoguin: there's an issue
19:55 gulti yes...i want it to only restart when it updates a file
19:55 forrest anything similar to this timoguin: https://github.com/saltstack/salt/issues/23677
19:56 timoguin I don't think I'm seeing an SSL error anywhere.
19:56 forrest iggy: Is there another issue somewhere?
19:56 timoguin This one happens to be my main states repo with my top.sls.
19:57 iggy yes
19:57 timoguin But all my formulas and pillar repos are fine.
19:57 * timoguin searches
19:58 gulti Antiarc:     Function: file.directory       Result: False      Comment: "recurse" must be formed as a list of strings      Started: 14:57:14.849238     Duration: 1.165 ms      Changes:
19:58 iggy https://github.com/saltstack/salt/issues/22987
19:58 iggy timoguin: ^
19:58 mpanetta joined #salt
19:58 Antiarc gulti: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html#salt.states.file.directory
19:59 forrest Antiarc: Was watch updated to recursively watch all files if you set it to a directory?
19:59 timoguin Thanks, iggy
19:59 timoguin Looks like the same issue
19:59 Antiarc forrest: If you're using file.directory to manage a directory then any changes in it will trigger watches
19:59 iggy ^5
20:00 forrest sweet, wasn't aware of that.
20:00 ek6 parabola: you out there?
20:00 timoguin And to think I decided to deploy the v2014.7.5 tag instead of my usual 2014.7 branch
20:01 iggy timoguin: don't you hate when that happens... (happened to us at the tail end of 2014.1.x too
20:01 julez joined #salt
20:01 ek6 parabola: you absolutely dont want to do what i suggested with inotify beacon..you will death spin  https://github.com/saltstack/salt/issues/23685
20:02 ek6 and if you use the default inotify beacon timer i guess the good news you dont spin for long because it goes boom
20:03 bradthurber I've been struggling all day with minions not pulling down files referenced as salt://...filename. Master is on 2105.5.0 and minions on 2014.7.5. On minion debug log I just see
20:03 bradthurber Fetching file from saltenv 'base', ** attempting ** 'salt://oracle-client/files/oracle-instantclient12.1-basic-12.1.0.2.0-1.x86_64.rpm'
20:04 timoguin iggy: thanks again. upgrading to the branch head fixed it.
20:04 ek6 but I do encourage anyone if they have a better suggested workaround then mine ..which is as many here would expect..horrible and gross
20:05 iggy bradthurber: did it work before?
20:08 Chadk joined #salt
20:08 bradthurber iggy: I believe so. Up until now, I've had master/minions on bootstrap develop branch. Files are part of a formula. I'm also getting *lots* of 'Minion did not return. [No response]' whenever I ask the minions to do anything more than test.ping
20:08 spookah joined #salt
20:09 bradthurber this is using gitfs - and I'm using pygit2.
20:10 iggy you're running -devel
20:10 iggy ... tried 2015.5 branch instead?
20:10 gulti thanks Antiarc...  /etc/nginx/conf.d/: file.directory:     - watch_in: service: nginx was enough
20:11 Chadk So I'm finding it a little odd that the Salt sales team is the first one I've tried to be in touch with, that you actually have to chase up on, as opposed to bothers you once a week.
20:11 iggy gulti: you know that's not going to do anything if the _contents_ of /etc/nginx/conf.d changes right?
20:11 bradthurber iggy: not at the moment - I'm running 2105.5 on master and 2014.7.5 on minions.  I could put 2105.5 on a minion I guess so they match
20:11 Antiarc you want to add a source for the directory, so salt can manage the files in it
20:12 iggy bradthurber: oh, I thought you said you were running devel (the master branch of saltstack)
20:12 bradthurber iggy: well I was in a different environment when I thought this formula was working. Sorry for confusing the issue
20:13 bradthurber first "develop" environment was local. This new environment is on AWS
20:13 iggy bradthurber: it might be worth it to test
20:14 gulti Antiarc: I want to have different state files...one with the definitions for basic nginx setup...and a second state file will definitions unique for the app. so i dont have anything in there for source
20:14 iggy also, make sure the files you expect to see show up in salt-call cp.list_master
20:14 Antiarc gulti: You'll just want to use a watch_in per state file then
20:14 spookah joined #salt
20:15 gulti not the directory you mean
20:15 Antiarc Right
20:15 Antiarc Directory will only work if you're letting that state manage the whole directory
20:15 forrest Chadk: Are you complaining a sales team isn't harassing you? ;)
20:15 bradthurber iggy: yes, files do show up in salt-call cp.list_master
20:16 UtahDave lol, Chadk. sorry about that!  Want to pm me? I can make sure you get to the right person
20:16 iggy Chadk: you asked for it...
20:17 forrest Yeah now you have to deal with that TERRIBLE UtahDave guy :P
20:17 Chadk forrest, in this case, yes.
20:17 Chadk Because the delay in getting back has made me look at Ansible again
20:17 Chadk :(
20:17 forrest Chadk: Haha, talk to Dave then, he'll hook you up :)
20:17 UtahDave Yeah, I'll get sorted out.
20:17 Chadk :)
20:19 iggy UtahDave: while you're here... someone mentioned 2015.5 isn't on http://docs.saltstack.com/en/latest/topics/installation/windows.html
20:21 iggy and somebody named obestwalter was asking about Windows automated testing, but they appear to be gone now, so... nvm
20:21 UtahDave iggy: yeah, obestwalter emailed me directly.   thanks for pointing him my way!
20:22 iggy I love to throw you SSE guys under the bus ;)
20:23 UtahDave iggy: thanks pointing out the missing link, too.  I think the installers are actually in the downloads directory, just no links to them in the docs
20:23 iggy rubenb: ^
20:23 iggy VSpike: ^
20:24 forrest yeah it's totally there UtahDave: http://docs.saltstack.com/downloads/Salt-Minion-2015.5.0-x86-Setup.exe
20:24 nzero joined #salt
20:25 Chadk I don't see it on there.
20:25 bradthurber iggy: upgrading minion from 2014.7.5 to 2015.5.0 solved the issue state:// files not pulling down that you were discussing with me. Thank you
20:25 forrest Chadk: no link, but the direct link works
20:25 Chadk Heck, based on the docs, I had no idea that 2015.5.0 was the latest stable.
20:25 UtahDave http://docs.saltstack.com/downloads/
20:25 Chadk I thought it was 2014.7
20:25 iggy it's in the /topic
20:26 UtahDave it was JUST released.
20:26 Chadk Yeah, that's how I realized.
20:26 iggy 2015.2 4lyfe
20:26 solidsnack To setup a new mountpoint for an mounted drive, it stands to reason one would use `prereq` to unmount it if mounted. However, I see this error:   Warnings: '__prerequired__' is an invalid keyword argument for 'mount.unmounted'.
20:26 UtahDave iggy: you still enjoying your Salt 2015.2 bicep tattoo?   lol
20:26 iggy *neck tattoo
20:27 iggy in old english
20:27 forrest classy
20:27 UtahDave ah, yeah
20:27 iggy solidsnack: what version of salt?
20:28 Emantor joined #salt
20:28 Chadk So I just upgraded from 2014.7 to 2015.5. And the vsphere profile I spent hours getting to work broke :(
20:28 solidsnack aptitude show salt-master | fgrep Version:
20:28 solidsnack Version: 2015.5.0+ds-1trusty1
20:28 Chadk solidsnack, shush, none of that linux magic!
20:28 Chadk :(
20:29 iggy Chadk: make sure you are using the latest libcloud
20:29 otter768 joined #salt
20:30 iggy I've noticed in the past, there's a sort of connection between the releases (both projects are evolving pretty fast, hopefully they get in sync at some point)
20:30 Chadk Humm. 0.14.1
20:30 iggy little on the old side
20:31 solidsnack Here is the full example: https://gist.github.com/solidsnack/87a609fe28bdcc6fb37f
20:31 Chadk Odd. I've done a full apt-get update/upgrade and have the saltstack ppa.
20:31 iggy you'll probably have to pip.install it
20:31 Chadk Oh.
20:31 Chadk 0.17.0 now
20:32 iggy but if that does fix it for you, I'd mention it to the ubuntu pkg mantainer as a possibility for adding to the ppa
20:33 sdm24 joined #salt
20:33 Chadk Nope, no luck. I think it may be a different issue though. It seems like if you clone a VM, you only have to specify a datastore. If you clone from a template, it somehow also expects a resource pool
20:33 Chadk Which is weird, because resource pools aren't required in vsphere.
20:34 UtahDave Chadk: you should be getting contacted shortly.  sorry about that!
20:34 Chadk <3
20:34 Chadk Much love
20:34 fxhp joined #salt
20:37 Gareth Wooo 2015.5.0!  I propose that the Beryllium be released tomorrow, so we can move onto the much easier to spell Boron release :)
20:38 murrdoc Bore on
20:38 murrdoc amirite ?
20:38 murrdoc no
20:38 murrdoc ok
20:38 UtahDave Gareth: I lose track of all of the names, lol
20:38 Chadk Is there any rationale for the fact that vsphere profiles only accept the object Ids for things(datastore-13, resgroup-11) rather than the pretty names for things?
20:39 amcorreia joined #salt
20:40 Gareth UtahDave: print this out and hang it in the office :) - http://www.wpclipart.com/science/atoms_molecules/periodic_tables/periodic_table_of_elements_BW.png
20:40 UtahDave Gareth: Oh, we've got a HUGE periodic table on our wall that Tom had custom printed.  I just forget to go look at it!   lol
20:41 Gareth UtahDave: haha
20:41 forrest Gareth: Nooo, don't give them any additional reason to use it! ;)
20:41 Gareth forrest: I think that point has blown past already :)
20:41 forrest Maybe... wait no, yes.
20:42 UtahDave Chadk: I'm sorry, I'm sure someone has asked you this already, but what version of Salt are you on?
20:42 Chadk 2015.5 now! \o/
20:42 forrest I'm looking forward to the masterless reactor
20:42 forrest been putting off updating my blog automation stuff and reworking some IDs.
20:43 UtahDave Chadk: Hm. regarding your question regarding the object Ids... I'm not sure.
20:43 Chadk Yeah, idk. The only reference I can find is: https://github.com/saltstack/salt/issues/14926
20:43 UtahDave Chadk: sometimes that depends on what the underlying library we're using requires
20:43 RickCH__ joined #salt
20:44 Chadk The steps you have to go through are a little obscure to say the least.
20:44 RickCH__ hello I am trying to use the salt-api to restart a service what is the format for it?
20:44 edrocks joined #salt
20:44 UtahDave Chadk: If that's a problem for vsphere users, please open an issue on github and we'll get the proper people looking at it.
20:44 giantlock joined #salt
20:45 markm joined #salt
20:46 Chadk Yeah, I'll have a look at it to make sure I'm not being a dumb-dumb.
20:46 Chadk I'm still super green with salt, and provisioning from a template was my first step. So uh, I'm still getting to terms with everything.
20:47 UtahDave Chadk: I haven't used the vsphere driver for quite a while. basically since it was first written.  I think Nitin has been making a lot of improvements to it lately
20:47 UtahDave RickCH__: Just a sec. Let me put togther an example for you
20:48 RickCH__ UtahDave Thanks
20:48 Chadk Actually, reading the code, it seems like it's intended that you have to use the MOR.
20:48 Chadk Yay reading the code.
20:51 Berty_ joined #salt
20:51 jdesilet joined #salt
20:52 djgerm joined #salt
20:53 treborTech joined #salt
20:54 Captain|2 joined #salt
20:55 jdesilet joined #salt
20:55 programm1 joined #salt
20:56 jab416171 joined #salt
20:56 solidsnack joined #salt
20:57 hackel joined #salt
20:59 UtahDave RickCH__: If you look here: http://docs.saltstack.com/en/latest/ref/netapi/all/salt.netapi.rest_cherrypy.html#authentication   you'll see an example of running a test.ping
20:59 nzero joined #salt
20:59 UtahDave swap out test.ping for service.restart
21:00 RickCH__ @UtahDave that I have running, but how do it pass it both a service and a server so I can start/stop/restart a service
21:00 UtahDave you change the target. just a sec
21:00 RickCH__ so the target has both the server and the service?
21:00 mirko is there a way to use salt-stack with "sudo" but without NOPASSWD?
21:01 UtahDave curl -sSk https://localhost:8000             -b ~/cookies.txt             -H 'Accept: application/x-yaml'             -d client=local             -d tgt='my_server_name'             -d fun=service.restart -d arg=['apache2']
21:01 ajw0100 joined #salt
21:01 mirko that way i could spread the user-password and the key to different machine which would make me feel more comfortable
21:02 RickCH__ perfect will try it out.
21:02 hal58th__ joined #salt
21:02 Captain|2 Hello, sorry in advance if I should have found this somewhere else, but I seem to have an issue where pkgrepo.managed clean_file does not actually remove existing entries from the named file.
21:03 hal58th_1 joined #salt
21:03 RickCH__ @utahDave thanks that was perfect
21:05 the_lale1u joined #salt
21:05 llb42_ joined #salt
21:05 litwol_ joined #salt
21:05 canci_ joined #salt
21:05 Auroch_ joined #salt
21:05 SheetiS joined #salt
21:06 Chadk Oh, I think I see why the vsphere provider is a little odd. It would appear that the the vmware module in Beryllium replaces it, as it uses pyVmomi instead of pysphere, which isn't maintained at all.
21:06 snaggleb_ joined #salt
21:06 lz-dylan_ joined #salt
21:06 oeuftete_ joined #salt
21:06 oblio_ joined #salt
21:06 aparsons joined #salt
21:07 pfalleno1 joined #salt
21:07 saifi_ joined #salt
21:07 sophomer1c joined #salt
21:07 mirko basepi: could you point me to the commit of where that made it upstream? i fail to find it... https://github.com/saltstack/salt/issues/23628#issuecomment-101812347
21:08 mirko ah, or is it just planned?
21:08 buxy_bak joined #salt
21:09 Anthrax joined #salt
21:09 Anthrax joined #salt
21:09 fxhp joined #salt
21:10 jdesilet joined #salt
21:10 cb joined #salt
21:10 chitown joined #salt
21:10 primechu_ joined #salt
21:10 wendall911 joined #salt
21:10 nadley joined #salt
21:10 MaZ- joined #salt
21:11 markm joined #salt
21:11 cwyse_ joined #salt
21:11 UtahDave joined #salt
21:11 dean joined #salt
21:12 wm-bot4 joined #salt
21:12 primechu_ joined #salt
21:14 Guest36286 joined #salt
21:20 drawsmcgraw joined #salt
21:23 djgerm joined #salt
21:27 murrdoc joined #salt
21:29 baweaver joined #salt
21:30 Nicop06 joined #salt
21:37 jab416171 joined #salt
21:38 basepi mirko: implemented: https://github.com/saltstack/salt/pull/23272
21:38 basepi And it's been merged forward into 2015.5 and develop
21:41 aparsons joined #salt
21:47 mapu joined #salt
21:48 checho Hi all.... I'm facing a weird issue...after upgrading master and minion to 2015.5 I'm receiving an "AttributeError: 'module' object has no attribute 'traverse_dict_and_list'" error in the minion
21:49 Tyrm_ joined #salt
21:50 RickCH__ Trying to call a highstate via salt-api and getting {"return": [{}]}  is that to be expected?
21:50 jab416171 joined #salt
21:51 JDiPierro joined #salt
21:51 murrdoc does the iptables implementation still stink
21:54 hemphill joined #salt
21:54 Chadk basepi, you alive for a quick PM?
21:55 iggy checho: can you gist the full error message?
21:56 checho Nevermind...I have to restart the minion host....
21:56 iggy murrdoc: on RH, I think
21:56 bhosmer joined #salt
21:57 SheetiS joined #salt
21:57 c10b10 joined #salt
22:00 jespada joined #salt
22:01 Tyrm joined #salt
22:04 iggy hmm, 2 different environments, same salt versions, same salt trees, one fails, one is okay
22:07 murrdoc iggy:  kk
22:11 baweaver joined #salt
22:12 nzero joined #salt
22:14 ageorgop joined #salt
22:15 Gareth murrdoc: depends on what release of RH, so long as your iptables has the --check option available iptables should be fine.
22:16 mirko iggy: i'd say it's the environment ;)
22:19 murrdoc i am on ubuntue
22:19 murrdoc ubuntu*
22:19 jespada joined #salt
22:19 Singularo joined #salt
22:19 murrdoc i want to put the salt master 'recommended' iptables
22:19 murrdoc was hoping to use salt to put them on there
22:20 murrdoc http://docs.saltstack.com/en/latest/topics/tutorials/firewall.html scroll to debian
22:21 Gareth Ubuntu iptables should work fine.
22:24 donmichelangelo joined #salt
22:30 jab416171 joined #salt
22:30 otter768 joined #salt
22:31 scoates joined #salt
22:32 mosen joined #salt
22:33 bfoxwell joined #salt
22:41 solidsnack joined #salt
22:42 kusams joined #salt
22:53 c10b10 joined #salt
22:55 iggy UtahDave: basepi: whiteinge: can one of you make a blank GCE-formula in saltstack-formulas?
22:55 UtahDave sure, just a sec.    just name it      GCE-formula?
22:55 basepi lowercase
22:55 iggy assuming that fits the format
22:55 iggy ^
22:56 iggy thought so
22:56 iggy google-cloud-engine-formula seems a bit long
22:56 * basepi is happy to have UtahDave sharing the op "love"
22:56 basepi Need to op some of you awesomesauce people
22:56 UtahDave there you go!  https://github.com/saltstack-formulas/gce-formula
22:56 iggy danke
22:57 murrdoc hate it
22:57 murrdoc so much
22:57 murrdoc (dont )
22:57 * basepi promptly ops murrdoc
22:57 basepi bwahahaha
22:57 iggy I think murrdoc should be made to start sharing in the formula "love"
22:58 ek6 oh my...  ducks
22:58 murrdoc -o  ?
22:58 Gareth Aww.  Was hoping to see murrdoc run a muck...drunk with power.
22:58 murrdoc i had o ?
22:58 murrdoc OH NO!
22:58 basepi briefly. =D
22:58 murrdoc I WANTS IT
22:58 murrdoc PRECIOUS
22:58 basepi hahahaha
22:59 * basepi is going to change his name, op EVERYONE, leave #salt, and live on a beach
22:59 murrdoc u can just op me
22:59 murrdoc nicest dick tator u know
22:59 basepi hahaha
23:00 murrdoc damnit autocorrect
23:00 basepi SURE it was
23:00 Gareth ....with a yo-yo named 'Wilson'.
23:01 ek6 gareth: long as i get to cut the string at the base...
23:01 murrdoc srsly tho
23:01 murrdoc i would like to be part of the saltstack-formula org
23:01 adelcast joined #salt
23:02 scoates joined #salt
23:03 baweaver joined #salt
23:03 forrest murrdoc: There really aren't any benefits, but you can ask whiteinge to add you.
23:05 fyb3r left #salt
23:07 UtahDave murrdoc: what's your github username?
23:09 mosen murrdoc: you hafta accept my PRs then
23:09 keimlink joined #salt
23:12 murrdoc or reject
23:12 murrdoc :stinks: will be my most used comment
23:19 drawsmcgraw1 joined #salt
23:21 brianfeister joined #salt
23:24 iggy everybody goes through the formulas in the morning, I'm _not_ a morning person
23:25 babilen Neither am I, but I'm in a different timezone
23:25 trevorj joined #salt
23:26 litwol joined #salt
23:27 jespada joined #salt
23:27 joshfu joined #salt
23:29 murrdoc iggy:  lets reject stuff with |default('True') in it
23:29 murrdoc pillar.get + defaults.yml should cover defaults
23:30 iggy lol, I'm not going to make someone trying to make a ~10 line change do the defaults.yaml conversion
23:30 murrdoc k make an issue ?
23:31 murrdoc lets get people to pick up issues and fix em
23:31 iggy yeah, feel free to do that
23:31 mapu joined #salt
23:32 JlRd joined #salt
23:34 Viper233 joined #salt
23:35 drawsmcgraw joined #salt
23:38 julez joined #salt
23:47 hasues joined #salt
23:47 hasues left #salt
23:47 dalexander joined #salt
23:49 dendazen joined #salt
23:50 drawsmcgraw joined #salt
23:53 keimlink_ joined #salt
23:55 crashmag joined #salt
23:57 rideh joined #salt
23:57 bhosmer joined #salt
23:58 JDiPierro joined #salt

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