Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-03-27

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

All times shown according to UTC.

Time Nick Message
00:01 iggy or salt -C '* and not G@saltversion:2014.7.2'
00:01 ahammond iggy that looks kinda perfect.
00:01 iggy test first...
00:02 bhosmer__ joined #salt
00:03 forrest joined #salt
00:03 ccarney_ROCC joined #salt
00:07 Singularo joined #salt
00:08 Guest89 joined #salt
00:11 meylor joined #salt
00:14 HexOffender joined #salt
00:16 sunkist joined #salt
00:18 pdayton joined #salt
00:24 zwi joined #salt
00:30 scoates joined #salt
00:32 murrdoc joined #salt
00:32 hackel joined #salt
00:37 otter768 joined #salt
00:39 JDiPierro joined #salt
00:41 hackel Is there a way to make salt automatically refresh pillar data every time it is run?
00:41 hackel I assume there was some reason for designing it this way, but I sure can't see it.
00:43 iggy hackel: it does when you run a highstate
00:44 hackel Ahh, okay.  Good to know.
00:45 desposo joined #salt
00:47 desposo1 joined #salt
00:47 iggy production upgraded to 2015.2 without any hiccups
00:48 iggy \o/
00:56 mosen yay
01:00 penguinpowernz joined #salt
01:01 Nazca__ joined #salt
01:03 renoirb joined #salt
01:07 ahammond iggy 2015.2 is released?
01:07 ahammond also, you win, sir!
01:10 aqua^mac joined #salt
01:18 kusams joined #salt
01:24 murrdoc joined #salt
01:27 drawsmcgraw left #salt
01:29 __number5__ TIL: salt.loader magic: https://github.com/saltstack/salt/issues/19459
01:33 CheKoLyN joined #salt
01:35 kevops joined #salt
01:38 murrdoc joined #salt
01:40 primechuck joined #salt
01:41 fxhp joined #salt
01:45 murrdoc joined #salt
01:52 Vynce joined #salt
01:55 dyasny_ joined #salt
01:56 otter768 joined #salt
02:01 jespada joined #salt
02:02 scoates pkg.installed + sources  Ã¢â‚¬Â¦ doesn't do updates?
02:03 eichiro joined #salt
02:05 desposo joined #salt
02:13 bfoxwell joined #salt
02:14 iromli joined #salt
02:15 catpiggest joined #salt
02:17 admgre joined #salt
02:21 glyf joined #salt
02:24 evle joined #salt
02:26 \ask joined #salt
02:27 ITChap joined #salt
02:28 ITChap Hi everyone
02:28 donmichelangelo joined #salt
02:29 ITChap I have a strange issue with a SmartOS minion: package installation seems to work but not for a specific package (nsd)
02:30 ITChap Outside of nsd other packages can be managed without issue
02:45 germs_ joined #salt
02:45 germs_1 joined #salt
02:45 germs_1 left #salt
02:50 dalexander joined #salt
02:54 Vynce joined #salt
02:54 dyasny_ joined #salt
02:58 aqua^mac joined #salt
02:59 Vynce joined #salt
02:59 aqua^mac joined #salt
03:00 Vynce left #salt
03:01 jespada joined #salt
03:01 aqua^mac_ joined #salt
03:02 aqua^mac joined #salt
03:04 ITChap joined #salt
03:04 aqua^ma__ joined #salt
03:11 seev joined #salt
03:20 jerematic joined #salt
03:21 djslocker joined #salt
03:21 favadi joined #salt
03:22 Vynce joined #salt
03:23 Vynce so, basic question, probably, but i'm having trouble figuring out how to ask it of the internet. I've got a config file on a minion.  i want to make sure that it contains a certain line; if not, iw ant to add it.  can this be part of state?  would i do this with like, if not (file.contains etc.) { file.line etc. } ? where can i look for more about this sort of thing?
03:28 seev cmd.run with an "echo 'whatever' > file' and an onlyif with a grep
03:33 aqua^mac joined #salt
03:33 CheKoLyN joined #salt
03:35 aqua^mac_ joined #salt
03:43 badon joined #salt
03:44 Vynce @seev: can i put that in a a state file so that it's part of state.highstate?
03:45 Furao joined #salt
04:10 bhosmer_ joined #salt
04:21 glyf joined #salt
04:21 jerematic joined #salt
04:35 smkelly joined #salt
04:44 evle joined #salt
04:47 Guest89 joined #salt
04:47 davet joined #salt
04:50 jespada joined #salt
04:53 mosen joined #salt
04:57 dyasny joined #salt
05:01 armguy joined #salt
05:02 I3olle joined #salt
05:05 TyrfingMjolnir joined #salt
05:14 aw110f joined #salt
05:17 malinoff joined #salt
05:20 tkharju joined #salt
05:33 Guest89 joined #salt
05:36 chandankumar joined #salt
05:37 Furao who got the idea to put cornflakes in those burgers?
05:37 Furao … wrong window
05:37 chandankumar Hello
05:37 chandankumar Tomorrow we are hosting a python sprint in pune (India)
05:37 Furao too far!
05:38 chandankumar I am including salt project in Hackathon
05:38 chandankumar http://www.meetup.com/PythonPune/events/221206592/
05:44 MK_FG joined #salt
06:10 jerematic joined #salt
06:19 viq joined #salt
06:26 stanchan joined #salt
06:36 dopesong joined #salt
06:38 dopesong_ joined #salt
06:39 jespada joined #salt
06:42 krelo joined #salt
06:54 colttt joined #salt
06:56 refnode_ joined #salt
06:57 AndreasLutro joined #salt
07:03 malinoff joined #salt
07:18 ajw0100 joined #salt
07:24 dopesong joined #salt
07:31 JayFK joined #salt
07:34 flyboy joined #salt
07:35 KermitTheFragger joined #salt
07:40 joshin joined #salt
07:53 refnode_1 joined #salt
07:53 dopesong_ joined #salt
07:58 eseyman joined #salt
07:59 jerematic joined #salt
07:59 __gotcha joined #salt
07:59 AndreasLutro what's the purpose of salt.states.ssh_auth_present's name argument?
08:00 pf_moore joined #salt
08:02 Hipikat AndreasLutro: it's the actual 'public key' string
08:03 AndreasLutro Hipikat: but I'm providing the public key as a file via the source argument
08:03 TyrfingMjolnir joined #salt
08:04 Hipikat AndreasLutro: hmm, then you can probably name it something useful and descriptive and it'll get ignored :)
08:04 AndreasLutro hmm... hope so
08:10 Auroch joined #salt
08:11 stoogenmeyer joined #salt
08:11 MasterNayru joined #salt
08:14 Romlok joined #salt
08:14 jacksontj joined #salt
08:15 dthorman joined #salt
08:15 notnotpeter joined #salt
08:16 bluenemo joined #salt
08:16 bluenemo joined #salt
08:16 lb1a joined #salt
08:18 VSpike Is it possible in Salt on Windows to get a python prompt from the bundled python?
08:19 defenestratexp joined #salt
08:20 VSpike I'm getting this error when trying to download a couple of things from the Windows repo, both coming from an https:// URL
08:20 VSpike https://bpaste.net/show/9daac19c502a
08:21 VSpike I checked, and I can download the file in IE on Windows, so the built-in certs are OK. Not sure what certificate store salt/python uses though
08:22 Terminus- VSpike: why not just run python? you get a prompt when you do.
08:23 aravind joined #salt
08:23 ajw0100 joined #salt
08:23 freelock joined #salt
08:23 flebel joined #salt
08:23 Edgan joined #salt
08:24 sk_0 joined #salt
08:24 NightMonkey joined #salt
08:24 kormoc joined #salt
08:25 AndreasLutro apparently you *need* the ssh pubkey in the state, or salt won't be able to remove it - kinda fail to see the point of the source argument then!
08:27 VSpike Ah https://github.com/saltstack/salt/issues/16773
08:28 VSpike Terminus-: There is no python, I don't think
08:28 jespada joined #salt
08:28 VSpike Other than the python DLL and libs in C:\salt\salt-2014.7.2.win-amd64
08:30 surajd joined #salt
08:32 ekle joined #salt
08:32 Andre-B joined #salt
08:33 ekle is it normal that salt-run manage.status blocks when there are offline minions ?
08:33 Terminus- VSpike: ah... i see what you mean. just unpacked the installer.
08:37 VSpike Is github quite timeout-y this morning, or is just me?
08:37 wincyj joined #salt
08:42 krelo joined #salt
08:42 babilen VSpike: it is
08:44 ThomasJ very much so :\
08:45 Romlok halite is telling me I need pyOpenSSL installed, but it *is* installed :(
08:48 ThomasJ github under ddos https://status.github.com/
08:52 davet1 joined #salt
09:01 Xevian joined #salt
09:06 ckao joined #salt
09:07 clmsy joined #salt
09:11 aqua^mac joined #salt
09:11 Podgrade joined #salt
09:16 Romlok apparently pyopenssl has a new dependency in its chain, which the halite docs don't account for
09:19 babilen Shouldn't the package manager pull that in?
09:21 Romlok apparently not if you use pip to install it
09:21 Romlok it depends on the "cryptography" package, which fails to install
09:21 Romlok but that doesn't seem to stop pyopenssl itself installing
09:22 Romlok cryptography fails to *compile*, I should say
09:22 Romlok halite is up and running now, if I install python-openssl using apt
09:22 Romlok though for some reason it's not using TLS D:
09:23 babilen Sometimes I wonder why every single programming language needs a crappy package manager ;)
09:23 Romlok because there's no accepted cross-platform standard
09:23 Romlok so we need to come up with our own standard
09:23 BtbN So they continue to break distro packages, because people get told to "sudo easy_install"
09:24 linjan joined #salt
09:24 refnode joined #salt
09:25 VSpike don't even get me started on ruby gems / bundler
09:26 BtbN Those at least work without root permissions
09:26 BtbN And setup some virtualenv.
09:26 babilen curl ... |sh ;)
09:26 babilen Anyway, I am procrastinating.
09:28 glyf joined #salt
09:28 fredvd joined #salt
09:30 huddy joined #salt
09:31 Romlok okay, so apparently cherrypy was ignoring the tls setting
09:31 Romlok when I switch to paste it's using TLS
09:32 VSpike Well, this is odd. I've update my branch of the win repo to use http instead of http because of https://github.com/saltstack/salt/issues/20502
09:33 VSpike After a few attempts I got salt-run -l debug winrepo.update_git_repos to work
09:33 clmsy you are trying to use halite ?
09:33 babilen Does anybody know what happened to https://github.com/saltstack-formulas/riemann-formula ? Is it just gone due to GH's ddos or has it been removed?
09:34 VSpike Then I did salt-run winrepo.genrepo and salt -G 'os:Windows' pkg.refresh_db, as well as a salt-call pkg.refresh_db on the minion I'm testing on
09:34 VSpike But when I try install the package on the minion, it's still attempting to use https
09:34 clmsy but isn't halite development completely stopped ?
09:34 clmsy yeah its completely stopped
09:34 clmsy last commit
09:35 clmsy 25 oct 2014
09:35 VSpike can I just delete C:\salt\var\cache ?
09:35 clmsy so yeah Romlok i wouldn't bet on it
09:35 babilen VSpike: The presumed equivalent on Linux is unproblematic
09:36 VSpike nope, that doesn't fix it
09:36 * babilen wants GH back :(
09:37 Romlok so if halite is abandoned, are there any other salt GUIs?
09:37 clmsy bet on this
09:37 clmsy https://github.com/tinyclues/saltpad
09:37 Romlok though Oct '14 isn't _that_ long ago
09:37 clmsy it seems more active
09:37 babilen saltpad is definitely more active
09:37 clmsy its _that_ long ago
09:38 clmsy for the oss scene
09:38 babilen halite is dead and Saltstack seem to be working one something "enterprisey" that won't be open sourced
09:38 jespada joined #salt
09:38 otter768 joined #salt
09:38 Romlok oh lawd
09:38 Romlok salt seemed to be the only one that wasn't buried under enterprise money
09:39 clmsy you have to go for enterprise money
09:39 clmsy its totally understandable
09:39 Romlok "have to", nupe
09:39 Romlok but it is understandable
09:40 clmsy ok.. "have to" is a bit strong in that context
09:40 ktosiek joined #salt
09:41 germs_ joined #salt
09:42 The_Loeki joined #salt
09:42 AndreasLutro I want to set up a gitignored top.sls-like file where users can put in whatever to test on a local vm - how do I set this up so that salt registers it? an include call in top.sls? you can't have multiple top.sls files right?
09:42 VSpike My winrepo.p file has not been modified since dec 4th :/
09:42 refnode_ joined #salt
09:43 VSpike wt and indeed f
09:44 Grokzen joined #salt
09:44 Romlok "Error! HTTP Fetch Docs Failed!"
09:45 VSpike oh god. insufficient caffeine error
09:46 The_Loeki howdy all :) any SaltStack devs around?
09:46 Romlok oh well, I guess my little adventure with halite is at an end
09:47 JlRd joined #salt
09:47 The_Loeki I seem to remember HAlite is sort-of deprecated anyways ;)
09:48 jerematic joined #salt
09:48 Romlok apparently so
09:48 Romlok but that's the one that I read about when I was researching salt
09:48 Terminus- ugh... controlling whitespace in jinja is so damned annoying.
09:48 VSpike Got an email from MS touting salt enterprise on Azure, saying they'd demoed salt + azure at salt conf. Given the general flakiness of Windows support on salt, I wonder what they demoed. Was it all Linux stuff?
09:48 jhauser joined #salt
09:49 Romlok the first iPhone was flaky as hell during it's public demo
09:49 Romlok but it worked perfectly because they stuck to a strict script
09:49 ITChap joined #salt
09:51 The_Loeki Who could forget Win98's on-stage BSOD when demoed by Gates :)
09:55 matthew-1arlette joined #salt
09:56 ekle how can i make module.run only run if a file does not exit ?
09:59 Romlok hm, is there a way to have salt automatically remove files which _were_ managed, but are no longer in any state?
09:59 Romlok or what is common practice for removing old config files?
10:03 babilen Romlok: There is no common practice. To which state do you want to revert them to and on what kind of system?
10:04 Romlok on Debian. I created a state to configure halite on the master, and now I want it gone :)
10:04 Romlok I'm currently writing "no-halite.sls"
10:05 Romlok would be nice if there were an auto-undo for states
10:06 Romlok though the ticket I found explains it's not that simple
10:12 VSpike Aw, cmon. I updated my win repo to use http instead off https to get around that bug, and the flipping python download servers 301 http requests to https
10:16 babilen heh
10:17 babilen Romlok: So, which state do you want to revert to?
10:17 matthew-1arlette joined #salt
10:24 ITChap joined #salt
10:30 favadi joined #salt
10:32 giantlock joined #salt
10:33 GnuLxUsr joined #salt
10:39 Romlok babilen: I wanted the server in the state it was in before I added the halite state, effectively
10:40 Romlok so I wrote a new state which removed the config file and pip-uninstalled
10:40 babilen Romlok: Then you have no other option than to create a snapshot of the server and revert to that
10:40 Romlok then did a `salt salt-master state.sls no-halite`
10:41 babilen That is not the state it was in before
10:41 Romlok no, but close enough for my means
10:41 Romlok but I gotta go; moving offices
10:41 refnode_1 joined #salt
10:47 aquassaut joined #salt
10:54 ITChap does someone use salt on SmartOS ?
10:54 ITChap any issue with pkgin ?
11:02 surajd joined #salt
11:02 flebel joined #salt
11:04 CeBe joined #salt
11:06 williamthekid joined #salt
11:09 ninkotech joined #salt
11:12 Andre-B_ joined #salt
11:13 __gotcha joined #salt
11:20 peters-tx0 joined #salt
11:26 bhosmer_ joined #salt
11:32 denys joined #salt
11:36 jerematic joined #salt
11:39 otter768 joined #salt
11:40 Hell_Fire joined #salt
11:41 favadi left #salt
11:42 linjan joined #salt
11:42 patrek joined #salt
11:50 zip_kid joined #salt
11:50 dkrae joined #salt
11:54 sjohnsen joined #salt
11:55 patrek joined #salt
11:57 erjohnso joined #salt
12:11 fredvd joined #salt
12:15 surajd joined #salt
12:19 \ask joined #salt
12:20 jeffspeff joined #salt
12:22 aphor left #salt
12:23 thayne joined #salt
12:24 wincyj joined #salt
12:24 dyasny joined #salt
12:27 \ask joined #salt
12:28 Deevolution joined #salt
12:31 aphor joined #salt
12:34 jhauser joined #salt
12:34 SheetiS joined #salt
12:34 ktosiek joined #salt
12:34 aphor If I need to write grains or states in python, and import stuff that needs to be pip installed, I'm not sure how to make that work.
12:36 aphor If python sls files need to run, and then return the highstate data, the import will fail, then the dependencies get fed back to salt, and then the pip state to get the package would get installed.
12:37 jerematic joined #salt
12:38 \ask joined #salt
12:38 zwi joined #salt
12:39 aphor Same with grains.
12:43 jerematic joined #salt
12:43 * aphor switching boxes
12:46 AndreasLutro salt-call is taking longer than salt-ssh, that's odd
12:47 VSpike Can anyone give me an example of how you'd modify an AWS cloud profile to add an extra EBS volume to an instance?
12:47 edrocks joined #salt
12:48 aphor joined #salt
12:48 jhauser joined #salt
12:48 markm joined #salt
12:49 amcorreia joined #salt
12:51 jdesilet joined #salt
12:55 subsignal joined #salt
12:57 dendazen joined #salt
12:58 Deevolution Anyone have any tips on how to improve the performance of the Salt-Mine?  Seeing it drastically slow down when querying a couple of hundred machine's results.
12:58 dendazen Is it possible to run single state from top.sls?
12:58 dendazen by id or any other way?
12:59 VSpike dendazen: salt 'foo' state.sls stuff.state1
12:59 VSpike dendazen: doesn't work if it has includes though, IIRC
12:59 dendazen what is ‘foo’ in that case?
13:00 VSpike hostname
13:00 dendazen i have for the host in top.sls something like this:
13:00 dendazen spcwlk1.sec1.*':
13:00 dendazen - users.velio.jenkins
13:00 dendazen oh ok
13:00 Ahlee There a way to query the "base" interface?  grains.item ip_interfaces doesn't seem 'sorted'
13:00 dendazen can i run it from within minion?
13:01 dendazen with salt-call?
13:01 aphor Ahlee: what exactly do you mean by the "base" interface?
13:01 JDiPierro joined #salt
13:02 Ahlee aphor: for example, this host has four interfaces, the one with the default route i call the base
13:02 Ahlee bond0.101 or similar
13:02 evle joined #salt
13:02 Ahlee i'll look at hte logic for how facter returns it and write a module for it
13:03 JDiPierro joined #salt
13:04 aphor Ahlee: so you actually want to know the interface of the default route?
13:04 Ahlee aphor: yeah, that'd fit my needs
13:04 VSpike dendazen: if you just want to run those machines, you can do salt 'spcwlk1.sec1.*' state.highstate
13:04 Ahlee i think just by convention here first interface is, which is why our facter checks always just worked, whereas the dictionary grains.ip_interfaces returns isn't sorted (of course, no sorted dicts)
13:04 VSpike but that will apply all rules from top.sls that match those machines
13:05 VSpike if you just want to run a single state on one or more machines, use the state.sls
13:05 dendazen I do not want higshtate
13:05 Ahlee dendazen: state.sls users.velio.jenkins
13:05 dendazen it will apply lots fo stuff from the base to that host
13:05 dendazen which i want to avoid
13:05 Ahlee state.sls runs single states
13:05 VSpike yep so salt 'spcwlk1.sec1.*' state.sls users.velio.jenkins
13:05 dendazen Thanks
13:06 Ahlee or salt-call state.sls users.velio.jenkins if youw ant to call from minion
13:06 aphor Ahlee: can you get what you need from salt.modules.network.default_route ?
13:06 ekristen joined #salt
13:07 dendazen Great, It worked. Thanks guys.
13:07 Ahlee Can I get to those within a state?  I need to modify samba config to have a specific interfaces = {{ int with default route }} lo
13:09 calve joined #salt
13:09 calve hi all
13:10 aphor Ahlee: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.module.html
13:10 * aphor commute time
13:10 calve is it allowed to use a jinja macro as a state name (inside a yml file) ?
13:10 teskew1 joined #salt
13:10 calve I get have the following error, which does not much helps : Rendering SLS 'base:api.crons' failed: Unknown yaml render error; line 11
13:11 calve line 11 is successfully generated using a jinja macro
13:11 Ahlee thanks aphor
13:13 N-Mi_ joined #salt
13:14 calve okay got it ; my macro output a colon, which is interpreted by yaml, even if it is quoted
13:17 overyander joined #salt
13:17 numkem joined #salt
13:24 cpowell joined #salt
13:25 ndrei joined #salt
13:27 ITChap joined #salt
13:27 jeffspeff joined #salt
13:35 jespada joined #salt
13:35 furball365 joined #salt
13:35 cmcmacken joined #salt
13:37 furball365 does anyone know how to get the Salt Mine example from the documentation working? I just get network.ip_addrs: - eth0
13:38 bhosmer joined #salt
13:40 furball365 i'm running 2015.2 and i'm trying to get a set of addresses for minions which support a particular role
13:40 otter768 joined #salt
13:41 furball365 i notice in the example on http://docs.saltstack.com/en/latest/topics/mine/ that the mine_interval is set to 5 rather than the default of 60 used higher up the page. Any reason?
13:42 ShibMcNe I assume it is just an example value
13:47 timoguin joined #salt
13:48 favadi joined #salt
13:49 drawsmcgraw joined #salt
13:49 frankS2 joined #salt
13:49 frankS2 Hi, is there any way to import a postgresql database with salt?
13:49 frankS2 one-time
13:49 Tecnico1931 joined #salt
13:49 furball365 ok.. no idea why you would add it unless there was a reason. Examples in the docs are usually assumed to be best cases.
13:49 frankS2 Im looking at this: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.postgres.html
13:50 JDiPierro joined #salt
13:50 BtbN Can't you just launch the import command?
13:50 BtbN As shell command
13:51 frankS2 BtbN: Yeah ofcourse :) but if there already were support for it
13:52 ITChap looks like a bad idea to automatically re-import a db ...
13:52 AirOnSkin When I setup my  Salt master and some minions I chose to install the stable packages. Is it possible to easily switch to dev now? Or do I have to do a full uninstall & install?
13:53 frankS2 ITChap: this is for vagrant, and im going to destroy the VM after using it
13:53 ITChap frankS2: Oh ok in this context it make sense
13:54 ITChap frankS2: You probably will have to go for the cmd state
13:54 frankS2 allright, thanks
13:56 furball365 bueller?
13:59 jespada joined #salt
14:00 BET joined #salt
14:03 I3olle_ joined #salt
14:04 aphor joined #salt
14:04 andrew_v joined #salt
14:04 BradleyET joined #salt
14:04 BET left #salt
14:05 markm joined #salt
14:05 mpanetta joined #salt
14:05 stanchan joined #salt
14:06 __gotcha joined #salt
14:06 mpanetta joined #salt
14:09 dopesong joined #salt
14:11 primechuck joined #salt
14:11 BradET joined #salt
14:12 drawsmcgraw1 joined #salt
14:12 redzaku joined #salt
14:15 StDiluted joined #salt
14:15 carnedepassaro joined #salt
14:17 mens_ joined #salt
14:18 aphor joined #salt
14:19 chandankumar joined #salt
14:23 BradleyET I think I would like to use one of the results of an execution module in another module. For example this command returns a whole bunch of stuff:
14:23 BradleyET salt 'mywebserver' apache.fullversion
14:23 BradleyET But I only want a part of the results returned:
14:23 BradleyET compiled_with: - HTTPD_ROOT="/etc/httpd"
14:23 BradleyET I think I would like to use one of the results of an execution module in another module. For example this command returns a whole bunch of stuff:
14:23 BradleyET salt 'mywebserver' apache.fullversion
14:23 BradleyET But I only want a part of the results returned:
14:23 BradleyET compiled_with: - HTTPD_ROOT="/etc/httpd"
14:23 BradleyET I'm thinking that if I can get my hands on that "etc/httpd" value then I know where to go in the filesystem to make further changes. Obviously I could hard-code that value or parameterize that value somewhere but if I can have the system tell me what it is isn't that better? I feel like I'm trying to invent an anti-pattern here.
14:23 kaptk2 joined #salt
14:24 kusams joined #salt
14:24 BradleyET Sorry - my IRC skills are now what they used to be
14:26 SheetiS joined #salt
14:26 ndrei joined #salt
14:26 VSpike the salt minion installer for Windows never seems actually to start the service. v annoying
14:26 stanchan_ joined #salt
14:26 kusams Hi can some one help...salt bin file are getting stored in /usr/local/bin but salt-minion init script is looking at /usr/bin it happening from last two days...i am not sure it python location issue or salt bootstrap issue
14:28 Norbell_ joined #salt
14:32 drawsmcgraw joined #salt
14:32 dude051 joined #salt
14:38 chandankumar joined #salt
14:38 aparsons joined #salt
14:39 favadi joined #salt
14:44 aparsons joined #salt
14:46 timoguin joined #salt
14:48 BradleyET joined #salt
14:48 bastion1704 joined #salt
14:48 debian112 joined #salt
14:50 BradleyET joined #salt
14:53 Brew joined #salt
14:56 jespada joined #salt
15:01 iggy VSpike: file a bug
15:03 hackel joined #salt
15:04 redzaku joined #salt
15:06 scbunn joined #salt
15:06 jdesilet joined #salt
15:07 carnedepassaro joined #salt
15:14 Norbell_ joined #salt
15:14 favadi left #salt
15:18 Morbus joined #salt
15:20 Grokzen joined #salt
15:23 TyrfingMjolnir joined #salt
15:25 Norbell_ joined #salt
15:26 JDiPierro joined #salt
15:27 JDiPierro I'm using jinja's json filter but it doesn't seem to be converting to json... anyone have this issue before?
15:27 desposo joined #salt
15:29 Norbell_ joined #salt
15:41 otter768 joined #salt
15:45 TyrfingMjolnir joined #salt
15:46 __gotcha joined #salt
15:46 hackel joined #salt
15:46 wincyj joined #salt
15:49 racooper joined #salt
15:51 v0rtex I have a masterless machine someone set up with file_roots and pillar_roots in /etc/salt/minion. If I connect it to my master does anyone know if salt-call --local will still work as it has with those paths so I don't break their process?
15:51 MTecknology viq: Did you ever happen to look at gitlab-ci?
15:52 iggy ahammond: no, we're running from git (I was expecting the release by now, but you work with what you got I guess)
15:57 yuhl_work__ joined #salt
15:57 swa_work joined #salt
16:00 giantlock joined #salt
16:00 SheetiS iggy: It's only Feb 55th, so give it time.
16:00 iggy v0rtex: I would probably rename the existing /etc/salt/ to /etc/salt.local or something then pass -c when doing local calls
16:01 rojem joined #salt
16:01 iggy SheetiS: oh, that was in reference to a question someone asked last night about my rolling out 2015.2 in our production environment last night
16:01 swa_work joined #salt
16:03 v0rtex iggy: good idea...I'll probably do something like that later today
16:04 thedodd joined #salt
16:04 catpig joined #salt
16:05 SheetiS A little off topic but it looks like amazon released an update for amazon linux that has dbus-python packages for python26 and python27 now.  I might actually be able to update to 2015.03 and still use spacewalk.  I think I still need to provision via salt cloud off of 2014.09 and update my way to 2015.03 so that python26 is preinstalled and salt cloud works at all (since the epel salt packages that amazon linux uses are py26)
16:05 SheetiS but at least now I can slow-roll my migration away from Amazon Linux.
16:05 desposo joined #salt
16:06 desposo1 joined #salt
16:09 chandankumar joined #salt
16:09 cpowell for the most recent release of Salt, which is the preferred rest interface for the API?
16:10 cpowell cherrypy, tornado or wsgi?
16:10 KyleG joined #salt
16:10 KyleG joined #salt
16:11 iggy cherrypy is still (afaik) the "suggested" one... I would expect that to change
16:12 cpowell ok, thanks
16:14 nitay joined #salt
16:17 nitay I’m using salt with docker (via docker-formula) as soon as I added docker my networking seems to have haywire, I think something happened with docker network interfaces taking over localhost, i.e. this seems weird - https://gist.github.com/nitay/24047d97266b41c23d25
16:19 speedlight joined #salt
16:22 CheKoLyN joined #salt
16:25 hackel joined #salt
16:26 tzero joined #salt
16:27 _prime joined #salt
16:28 _prime_ joined #salt
16:29 _prime joined #salt
16:30 _prime_ joined #salt
16:30 lictor36 joined #salt
16:31 aparsons joined #salt
16:32 thayne joined #salt
16:36 spookah joined #salt
16:38 denys joined #salt
16:39 BradleyET ls
16:40 wendall911 joined #salt
16:44 dopesong_ joined #salt
16:46 jalbretsen joined #salt
16:49 bluenemo joined #salt
16:49 bluenemo joined #salt
16:55 wendall911 joined #salt
16:55 zwi joined #salt
16:58 Andre-B joined #salt
17:00 murrdoc joined #salt
17:02 scbunn joined #salt
17:03 forrest joined #salt
17:07 JDiPierro joined #salt
17:07 dendazen joined #salt
17:08 TyrfingMjolnir joined #salt
17:09 murrdoc joined #salt
17:09 murrdoc joined #salt
17:09 murrdoc joined #salt
17:09 ajw0100 joined #salt
17:15 prwilson joined #salt
17:16 irctc625 joined #salt
17:17 wincyj joined #salt
17:20 bmac2 FYI - salt isn't smart enough to catch it when you don't put the correct minion names in the command. Can we work on that?  :-)
17:23 moos3 joined #salt
17:24 kormoc joined #salt
17:24 totte joined #salt
17:29 dopesong joined #salt
17:30 Fiber^ joined #salt
17:32 mpanetta joined #salt
17:34 iggy you sound like the person that wanted salt to warn you when you were working on "live" systems
17:35 btorch joined #salt
17:35 gchao that sounds really useful ^^^
17:35 nkuttler huh? incorrect name -> no matches, no?
17:36 germs_1 joined #salt
17:38 iggy more like "salt 'dns*' file.remove /etc/named.conf" <- whoops
17:38 iggy salt is good, but damn...
17:38 iggy You can't anti-barney-fife the entire world
17:39 druonysus joined #salt
17:39 druonysus joined #salt
17:39 forrest lol
17:39 Ryan_Lane :D
17:40 ajw0100 joined #salt
17:42 otter768 joined #salt
17:43 davet joined #salt
17:44 murrdoc joined #salt
17:44 dalexander joined #salt
17:45 aw110f joined #salt
17:45 pdayton joined #salt
17:50 forrest These pushes directly to the formula repo are starting to piss me off
17:50 iggy amen
17:51 iggy especially to update copyright years when nothing changed
17:51 forrest The salt perms are seriously pissing me off.
17:51 forrest *github
17:51 forrest not salt
17:51 iggy github?
17:51 forrest Yeah
17:51 iggy right
17:51 forrest they are so goddamn limited
17:51 timoguin joined #salt
17:51 forrest sucks
17:51 forrest timoguin: what uuuup
17:52 forrest https://github.com/saltstack-formulas/apache-formula/commit/42426f3f4ab7b1b30f9e8ae17e9991ae4f6fdb41 Do you know what user 'Jonathan Ballet' is iggy?
17:52 forrest I don't see them on the contributor list
17:52 forrest shit is making me angry because there's no good way to find what their github username actually is, because github is stupid.
17:52 timoguin forrest: lunch break that's what up. byes!
17:52 forrest timoguin: hah, later
17:53 forrest It's a serious problem. What if someone just merges in a change to download some malicious package
17:53 timoguin also, yea, Github doesn't help you if people misconfigure their author config.
17:53 forrest fucking dumb.
17:53 forrest timoguin: Yeah exactly, 'oh you have the right ssh key? Fuck publishing the author properly!'
17:53 forrest look at the key, and tell me their username.
17:53 rojem joined #salt
17:53 timoguin Bitbucket has a way to map those unknown ones to actual accounts.
17:53 iggy I don't
17:53 timoguin And it will pick it up like an alias.
17:54 timoguin I mean, if you know who it is.
17:54 iggy https://github.com/multani
17:55 babilen Why can he push directly?
17:56 Auroch joined #salt
17:56 canci joined #salt
17:57 murrdoc hes in the org
17:57 murrdoc if u are in the org you can push directly
17:58 iggy nmadhok did a bunch of them recently to update copyright years
17:58 murrdoc not good
17:58 iggy don't you guys have like a list of things you're not supposed to do written somewhere?
17:59 iggy some secret cabal page somewhere?
17:59 babilen murrdoc: Ah, didn't see him listed on https://github.com/orgs/saltstack-formulas/people
18:00 babilen (still don't)
18:01 iggy gravyboat merged them in #48 and #49
18:01 iggy err forrest
18:02 forrest ?
18:02 fbettag joined #salt
18:02 forrest iggy: There were some direct merges onto the repo
18:02 forrest that weren't PRs
18:02 iggy but not those
18:02 forrest which ones? The copyright ones?
18:02 forrest I am referring to the one I linked earlier only
18:02 murrdoc joined #salt
18:03 otter768 joined #salt
18:03 iggy yes, you merged the one you linked earlier
18:03 forrest Really? Why does it show it as a direct commit
18:03 iggy because the username isn't associated maybe?
18:04 iggy I don't really see where it says it's a direct commit, so that's just speculation
18:04 forrest yeah you're right
18:04 forrest I'm just looking at the commit log, but yeah seems to just be an error because the username isn't associated
18:04 forrest fuck me man
18:04 bhosmer_ joined #salt
18:05 Deevolution iggy, thanks for logging the bug on the 2015.2.0RC1 cmd.run changes.  It saved me from more fruitless searching as to why my code broke.
18:05 nitay joined #salt
18:06 iggy so that's something you should add to formula secret cabal checklist notes
18:07 SheetiS joined #salt
18:07 iggy Deevolution: I wish they'd just undo that... but SECURITY!!!!
18:08 Deevolution iggy: I'll comment on the bug, but you're right that some warnings should be given.  It's a pretty radical change...
18:12 iggy not to mention that all the cmd.run docs are wrong now
18:12 iggy s/docs/examples/
18:13 g3cko joined #salt
18:15 markm_ joined #salt
18:17 Deevolution iggy: Oh, I noticed that.
18:20 calve joined #salt
18:22 babilen Deevolution: Do you happen to have a link?
18:22 iggy https://github.com/saltstack/salt/issues/22052
18:22 londo joined #salt
18:23 aparsons joined #salt
18:23 babilen ta
18:29 timoguin joined #salt
18:30 baweaver joined #salt
18:32 LeProvokateur joined #salt
18:33 timoguin babilen: because Github only uses SSH keys for auth. Git doesn't care about that config. It just has to be set.
18:33 timoguin When they push they auth as their github user. But that doesn't affect the commits themselves.
18:35 markm_ joined #salt
18:37 babilen timoguin: Well, it obviously wasn't the case that that user pushed himself, but rather that some of the commits got merged
18:38 timoguin He pushed to his own repo.
18:39 iggy it just looks weird because it's not tied to an actual github user
18:40 babilen I simply understood the initial conversation as "user FOO pushed directly to the repo" and was wondering how that happened
18:40 babilen There is no mystery left at this moment
18:41 timoguin You are a mystery!
18:41 iggy there are other cases of people pushing directly to repos... just not that particular case
18:41 babilen The great LICENSE push of 2015
18:42 iggy which is a dubious change to begin with... plus the fact that they were all pushed directly to master
18:44 nitti joined #salt
18:53 drawsmcgraw left #salt
18:54 forrest babilen: Some users still push. Users in the default saltstack-formulas group have write access.
18:54 forrest Just github permissions being shitty and lacking
18:54 forrest been a problem for several YEARS
18:55 timoguin YEAS
18:55 timoguin YEARS
18:58 murrdoc OMGWTFBBQBEARS
18:59 spookah joined #salt
19:02 babilen forrest: Yes, "there is no mystery left"
19:02 evidence so the master on 2014.7.2 seems to get this error every 10 minutes.. i'm not having any functional issues i can spot, but i can't find any info on what this might be signaling
19:02 evidence 2015-03-27 11:56:27,775 [salt.daemons.masterapi][ERROR   ] Exception Object not found - failed to find pack entry (7bc7a96094f010aef62a3e8ffa49e7395f97cf10) occurred in file server update
19:02 evidence 2015-03-27 11:56:27,790 [salt.master      ][ERROR   ] Exception  occurred in file server update for git_pillar module.
19:03 babilen Why did I say something?
19:03 forrest babilen: Well, the true mystery, why do people push to master :P
19:03 toastedpenguin joined #salt
19:08 elektrix_ say I have an Exim formula and a Postfix formula. Only one of those will be installed at a time. Either of them could have procmail specific settings. Wondering the best practice to organize this. I'm guessing have a separate procmail formula, then somehow include it in the Exim and Postfix formulas?
19:10 elektrix_ basically Exim and Postfix could each have configurations with or without procmail
19:11 jespada_ joined #salt
19:11 babilen evidence: I'd clear caches and maybe git fsck (also check what's up with that commit)
19:11 babilen evidence: yeah, that's exactly what you should do
19:13 baweaver joined #salt
19:13 iggy evidence: also, if you your git_pillar is on github, there's a salt issue open for handling github being down more gracefully
19:13 krelo joined #salt
19:16 krelo joined #salt
19:17 I3olle joined #salt
19:18 elektrix_ maybe explain a different way. Have postfix module, have a procmail module. need to have postfix formula determine if procmail has been applied to the host, and if so, make the appropriate changes to the postfix main.cf
19:18 elektrix_ erm %s/module/formula/
19:18 SheetiS joined #salt
19:20 iggy elektrix_: there's not a "standard" way of doing that right now
19:20 iggy for the most part formulas are fairly self contained _or_ have hard dependencies on other formulas
19:20 bluenemo_ joined #salt
19:21 elektrix_ ah...so can you point me to any common work arounds for this (i.e. any example formulas)?
19:21 iggy I can't even think of one off the top of my head that has an optional dependency on another formula
19:21 elektrix_ k
19:21 elektrix_ hrm...
19:21 NewToSalt joined #salt
19:22 iggy I think I would personally key off some file that the procmail formula laid down
19:22 badon joined #salt
19:22 elektrix_ ah, so check to see if a file exists, then if so, make change x?
19:23 NewToSalt Does anyone have the correct syntax for enabling RDP in an sls file?
19:23 iggy i.e. {% if salt['file.directory_exists']('/etc/procmail') %} ... do procmail specific postfix config ... {% else %} standard config {% endif %}
19:23 nitay joined #salt
19:24 elektrix_ thanks for the snippet!
19:24 iggy it's a little dirty, but less so than other options (imo)
19:25 elektrix_ that'll work...otherwise I was going to put something in a pillar that indicated "yes procmail is installed on this host", but then I'd have to maintain that in two places.
19:25 elektrix_ i.e. in the pillar and in the salt states
19:25 iggy NewToSalt: not a lot of salt+windows users in here, you could try the mailing list
19:26 speedlight joined #salt
19:26 speedlight joined #salt
19:27 giantlock joined #salt
19:27 iggy elektrix_: actually, the way I'd handle it would probably be different and less portable than the above, but we use grains['roles'] for everything... so I'd have a {% if 'procmail' in salt['grains.get']('roles') %} do procmail config {% endif %} in my postfix formula
19:28 denys joined #salt
19:29 elektrix_ thanks...I'm still new to salt (2 days in, coming from puppet), so I'll check out both ways
19:29 elektrix_ and day 1 was just reading doc
19:32 baweaver joined #salt
19:33 iggy I wish more people would spend an entire day reading docs (or longer)
19:35 elektrix_ yeah, found the best practices guide, got syntax highlighting working in vim. couldn't get pylint working though
19:35 elektrix_ map.jinja, store secrets in pillar...etc
19:35 markm_ joined #salt
19:36 babilen elektrix_: I don't really see a problem with an explicit procmail state in both formulas
19:37 jespada__ joined #salt
19:38 igorwidl left #salt
19:38 babilen elektrix_: https://github.com/saltstack-formulas/salt-formula/blob/master/salt/map.jinja / https://github.com/saltstack-formulas/influxdb-formula/blob/master/influxdb/map.jinja is a map.jinja style that we are currently trying to establish (after some talks during this years saltconf)
19:39 elektrix_ yeah, that's pretty much what I'm doing with map.jinja
19:40 elektrix_ anything can can be generalized in going in map.jinja (e.g. os/version/whatever specific stuff), and things that are unique to our org are going in pillar (e.g. ssl certs, host name, etc)
19:40 babilen elektrix_: The main part of that is the merging, that allows you to completely refrain from any direct pillar lookups in SLS files and keeps default values in one defined place
19:40 babilen elektrix_: Yes, and you would then merge the pillar into foo_settings to have everything in one place
19:42 iggy the salt-formula is starting down that path too
19:44 dalexander joined #salt
19:44 ajw0100 joined #salt
19:45 elektrix_ ahh...took me a sec at looking at it, so do the pillar call inside the map.jinja, and assign it to a var, then use that var in the salt state
19:45 ltsampros joined #salt
19:46 ltsampros hey ho
19:46 ltsampros trying to do some magic through salt.wheel.key
19:46 ltsampros from salt.wheel.key import gen_accept
19:47 ltsampros I then say gen_accept(minion_id) but it complains that: global name '__opts__' is not defined
19:47 ltsampros is that normal ?
19:47 ltsampros even if I parse master opts and put in __opts__ the error is still there
19:47 iggy ltsampros: what context is this? (custom grain? custom module? standalone script?)
19:48 babilen elektrix_: Yeah, that's why I pointed that out and linked the salt, collect and influxdb formulas
19:48 ltsampros iggy: stand alone script
19:48 iggy whoops, I totally missed that you linked the salt-formula
19:48 Brew joined #salt
19:48 iggy ltsampros: then yeah, you have to set that up
19:48 thehaven joined #salt
19:48 iggy ltsampros: grep source for '__opts__ = ' for examples
19:48 ltsampros iggy: I set that up but still there
19:49 elektrix_ cool...coming from Puppet, I used a params file with which did the os/version level stuff, the set variables for hiera calls. nice that I can do that here too
19:49 iggy ltsampros: paste code I guess...
19:49 ltsampros iggy: I set it up like this __opts__ = salt.config.client_config('/etc/salt/master')
19:50 ltsampros iggy: will do.
19:50 elektrix_ to babilen and iggy...thanks!
19:52 ltsampros iggy: http://pastebin.com/BCLkWMK6
19:54 ltsampros iggy: I defined another function in same shell like: def a(): print __opts__ and it seems to work normally
19:57 iggy ltsampros: import salt.config instead of just salt?
19:57 iggy I know the standalone scripts I've written were pretty picky about imports
19:57 ltsampros naah
19:58 iggy (i.e. import salt.minion, salt.client worked while import salt didn't
19:58 ltsampros just tried that
19:58 ltsampros it doesn't seem to be the case
19:59 ltsampros I'm afraid that I'm just totally misusing the API
20:00 Vynce joined #salt
20:00 nexsja^ joined #salt
20:03 cpowell_ joined #salt
20:03 iggy the script api stuff is a bit of black magic... just keep poking at it until you get what you need
20:03 wicope joined #salt
20:04 TyrfingMjolnir joined #salt
20:04 ltsampros iggy: the problem seems to be that wheel loads key through loaders (I think)
20:05 chp1n joined #salt
20:05 ltsampros iggy: it just escapes me that calling that function directly cannot see my instance of __opts__ in the global namespace
20:06 ltsampros I though that python was looking first at local namespace, then global and then raised an error
20:06 baweaver joined #salt
20:06 iggy you have tried this in an actual script, not just in ipython right? (just to rule out ipython oddities)
20:07 ltsampros nope.
20:09 ltsampros iggy: well solved. you have to salt.wheel.key.__opts__
20:09 ltsampros do not understand completely why though
20:10 iggy yeah... that doesn't make sense
20:10 refnode_ joined #salt
20:14 zwi1 joined #salt
20:17 TyrfingMjolnir joined #salt
20:17 ndrei joined #salt
20:22 evilrob joined #salt
20:30 aphor joined #salt
20:32 TyrfingMjolnir joined #salt
20:32 ajw0100 joined #salt
20:36 baweaver joined #salt
20:39 aphor if I write a state in python, and it depends on an external python package, what is the best way to have salt pip install it?
20:41 robawt aphor: use the salt pip module to install it
20:41 robawt even better install virtualenv and install it into a virtualenv to not muck up system Python
20:41 iggy make sure you put reload_modules: True on the pip state
20:42 aphor thx.
20:43 aphor So if the #!py xls file runs, and the import fails, I guess I can handle the exception and generate a stub highstate return value, but will salt know to retry all the states that the pip depends on?
20:46 aphor xls... spellcheck... I meant sls
20:46 iggy no, you'll have to put your pip state before the python state
20:48 iggy so in your top file you could have somestate.prepip (your normal yaml state with "prereq:\n  pip.installed:\n    - reload_modules: True") and then somestate
20:48 iggy since salt processes top files top-down, it'll run the pip command then your state
20:48 aphor iggy: thanks a lot. I think I have enough to go and try that.
20:48 iggy at least that's one way to do it
20:49 iggy as with anything in salt... there are others
20:49 writtenoff joined #salt
20:49 aphor basically, I want to use https://github.com/aphor/data_size <-- this
20:49 aphor in a bunch of grains and states, but don't want to assume it's installed, and I need it to be installed early.
20:50 armguy got stuck updating an old box with salt on it and ran into the following error that is stumping me on upgrade. Ubuntu 12.04 http://pastebin.com/zVdqg022
20:50 edrocks joined #salt
20:51 edrocks joined #salt
20:51 aphor I guess I could just put it at the beginning of my base top.sls and dependencies would be there for debugging/didactic reasons.
20:54 aphor armguy: what happens if you tar up the pki directory, and wipe both the cache and pki directories, and restore the keys in the pki directory after the deb installs?
20:55 armguy let me try. Thanks for the idea. One of those days for me
20:55 mnguyen joined #salt
20:56 baweaver joined #salt
20:57 aphor armguy: I would consider the pki directory configuration files, and believe the deb should not fail to install unless a post install script decides it's not compatible.
20:59 aw110f joined #salt
21:00 racooper howdy. I'm having trouble with a repository state. error and sls is at https://gist.github.com/racooper/017c3ded4c85c85268cb
21:00 zwi joined #salt
21:02 armguy aphor ends now with a missing dir http://pastebin.com/hWBTUKje
21:02 baweaver joined #salt
21:04 forrest racooper: Can you try to drop 'name'
21:04 forrest I can't remember if humanname overwrites it
21:06 armguy seems related to this https://github.com/saltstack/salt/issues/21107
21:07 racooper I know it's worked as configured in the past, and doesn't throw an error on a server that already has this state applied. I'll test though.
21:07 forrest armguy: https://github.com/saltstack/salt/issues/19124 might want to ask jfindlay for an update
21:08 armguy thanks, using the workaround got the update to finish. So that part is good at least will see if I can contribute anymore to the issue .. Thanks for the help guys
21:09 forrest racooper: in your /etc/yum.repos.d on the old box is there an 'ocss' file? Are there any weird perms issues or anything strange going on in the new box for /etc/yum.repos.d?
21:10 racooper on the old box is 'ocss.repo', and nothing strange on the new one, it's a cleanly-loaded CentOS 6 server.
21:10 baweaver joined #salt
21:11 racooper and name is required, it's what sets the [reponame] field and the filename.
21:13 racooper hrm...now I'm confused. I'll go over the docs some more.
21:15 matthew-parlette joined #salt
21:27 Hell_Fire joined #salt
21:30 dthorman joined #salt
21:31 _prime_ joined #salt
21:32 aphor armguy: your packages are missing.
21:38 zwi joined #salt
21:40 jerematic joined #salt
21:41 baweaver joined #salt
21:43 hal58th joined #salt
21:44 murrdoc joined #salt
21:51 druonysus joined #salt
21:52 dude051 joined #salt
21:53 davet joined #salt
21:56 dude051 joined #salt
21:56 MindDrive So I have a separate Salt configuration directory, /etc/salt.tds, to allow users to send commands to the Salt master without the need for sudo.  Now if I run 'salt-call -c /etc/salt.tds ...', it works just fine.  However, if I try to use the Salt API and do 'opts = salt.config.minion_config('/etc/salt.tds/minion')'  along with 'caller = salt.client.Caller(mopts=opts)', it complains that it :q
21:57 MindDrive er, that it can't decrypt the 'random minion token' and that the 'Salt Master server's public key did not authenticate'.
21:57 MindDrive I'm confused as to what I might be missing on the programmatic side to have this operate properly.
21:58 chiui joined #salt
22:00 iggy do you have the same pki dir contents for each? (or have pki dir set in /etc/salt.tds/minion)
22:01 druonysus joined #salt
22:02 catpig joined #salt
22:02 MindDrive the /etc/salt.tds/pki/minion directory mirrors /etc/salt/pki/minion, with the exception that the minion.{pub,pem} keys are actually different.
22:03 iggy is the minion_id different then?
22:03 MindDrive You mean in the config file?
22:06 MindDrive If so, yes.
22:09 MindDrive I get this feeling some other option isn't being set correctly, but I'm not sure what.
22:24 iggy sounds like you've covered everything that I would have thought to look at
22:33 MindDrive I'm using pdb to step through Caller() right now... eyes may glaze over at some point. :)
22:36 druonysus joined #salt
22:37 evidence iggy/babilen: sorry pulled away, thx for the ideas on the pillar/github related error.  there is no commit with that SHA, not sure wtf that hash is from, and i did try clearing the cache yesterday with no luck :(  git fsck looks clean on there
22:38 evidence our internal github hasn't been down recently, i assume if there was any lingering issue from the last time it was, clearing cache would have nuked it?
22:38 evidence the pillar is displaying everything properly also, so whatever it is, is not causing any apparent problems
22:39 evidence guess i'll try to dig into the code a bit and see wtf causes is
22:39 iggy I mentioned github (and the issue was opened) because they've had an ongoing ddos attack going for the last 3-4 days
22:39 evidence we run an internal gh
22:39 iggy yeah, so that wouldn't apply to you I guess ;)
22:41 jerematic joined #salt
22:42 JayFK joined #salt
22:48 aquassaut joined #salt
22:51 otter768 joined #salt
22:51 babilen evidence: You might also want to investigate different git providers. We've seen some bugs with signed commits before and I wouldn't be too surprised if some "more funky" things in git might trip up whatever povider you use at the moment
22:51 refnode joined #salt
22:56 markm joined #salt
22:56 thehaven joined #salt
23:06 timoguin joined #salt
23:07 murrdoc joined #salt
23:09 refnode_1 joined #salt
23:09 Guest89 joined #salt
23:10 sunkist joined #salt
23:10 refnode_2 joined #salt
23:11 evidence babilen: we are using the recommended pygit2, i guess that's an easy check to try gitpython or dulwich
23:13 evidence oop.. GitFS authentication parameter 'privkey' ..  is only supported by the following provider(s): pygit2
23:22 evidence ah here we go.. https://github.com/saltstack/salt/pull/20141 caused a regression in 2014.7.. https://github.com/saltstack/salt/pull/20854 fixes that, and will be in 2014.7.3
23:25 SaintAardvark ~.
23:25 * SaintAardvark is just glad that wasn't a password
23:25 JordanTesting lol
23:25 evidence or was it?
23:35 ccarney_ROCC joined #salt
23:36 ccarney_ROCC left #salt
23:38 _JZ_ joined #salt
23:38 MindDrive Well, found my problem... the m2crypto Python module is only installed under Python 2.6 (and my app uses 2.7).  Trying to get it built under 2.7 with our current Jenkins/fpm setup is proving to be a challenge.  Yay.
23:39 mpanetta joined #salt
23:41 baweaver joined #salt
23:41 forrest MindDrive: Isn't it just available via pip install on 2.7? Shows it is supported over at the cheese shop
23:42 jerematic joined #salt
23:43 MindDrive forrest: We use fpm to package our Python modules because they're managed via Puppet, so they need to be in RPM format.
23:43 refnode_ joined #salt
23:43 refnode_1 joined #salt
23:44 thayne joined #salt
23:46 Ahlee forrest: but then you'd need a development environment on all the servers
23:46 forrest Ahlee: Yeah true
23:46 forrest MindDrive: Are you having openssl errors?
23:47 forrest I'm finding quite a few pre-compiled RPMs just by googling python m2crypto rpm
23:47 MindDrive fpm isn't passing a compiler define through (__x86_64__), so yes, openssl gets unhappy.
23:48 baweaver joined #salt
23:48 forrest MindDrive: Did you install openssl-devel?
23:48 Ahlee There's a blurb in the build second about fedora_build or similar
23:49 forrest MindDrive: Should be able to just run fedora_setup.sh
23:49 forrest Ahlee: Yeah gotta use the fedora script
23:49 MindDrive forrest: I already know the issue, I'm looking for the solution.
23:49 forrest MindDrive: Cool
23:49 cberndt joined #salt
23:50 Ahlee How goes forrest? ltns
23:50 forrest hey Ahlee, things are going well, you?
23:50 Ahlee As well as can be expected still being at work at 10 to 7 on a Friday
23:51 forrest Ahlee: That sounds shitty.
23:51 forrest Should go home
23:51 bfoxwell joined #salt
23:52 Ahlee Miles to go before I sleep.
23:52 forrest Do those miles on Monday
23:52 forrest person health/sanity > work
23:53 Vynce joined #salt
23:53 Ahlee Markets don't wait for my health and sanity to open :)
23:53 Vynce1 joined #salt
23:53 forrest Ahlee: heh
23:54 glyf joined #salt

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