Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-02-25

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

All times shown according to UTC.

Time Nick Message
00:10 bujordan joined #salt
00:11 shaggy_surfer joined #salt
00:12 bujordan joined #salt
00:14 brianfeister joined #salt
00:14 shaggy_surfer joined #salt
00:16 subsignal joined #salt
00:20 bhosmer_ joined #salt
00:21 knite_ joined #salt
00:24 baweaver joined #salt
00:30 knite_ joined #salt
00:40 kevinquinnyo1 joined #salt
00:42 flowstate joined #salt
01:00 Muchoz joined #salt
01:06 nZac joined #salt
01:06 iceyao joined #salt
01:11 iceyao_ joined #salt
01:13 bujordan joined #salt
01:14 iceyao joined #salt
01:15 bujordan joined #salt
01:21 bujordan joined #salt
01:21 joe_n joined #salt
01:24 justanotheruser joined #salt
01:32 jfred joined #salt
01:36 Shirkdog joined #salt
01:40 flowstate joined #salt
01:42 no joined #salt
01:58 malinoff joined #salt
02:01 iceyao_ joined #salt
02:03 brianfeister joined #salt
02:07 nZac joined #salt
02:08 bhosmer_ joined #salt
02:09 michelangelo joined #salt
02:14 jim__ joined #salt
02:15 rem5 joined #salt
02:16 ajw0100 joined #salt
02:26 brianfeister joined #salt
02:27 ashmckenzie joined #salt
02:33 justanotheruser joined #salt
02:35 joe_n joined #salt
02:36 Muchoz_ joined #salt
02:37 andrew_v joined #salt
02:37 iceyao joined #salt
02:38 grumm_servire joined #salt
02:40 mgresser joined #salt
02:41 flowstate joined #salt
02:46 MikaT joined #salt
02:46 malinoff joined #salt
02:51 Alysum joined #salt
02:53 Alysum Hello I'd like to know more about the glusterfs salt state. Is there a full example illustrating what https://docs.saltstack.com/en/latest/ref/states/all/salt.states.glusterfs.html can do ?
02:53 ninjada joined #salt
03:01 fmn joined #salt
03:02 bhosmer_ joined #salt
03:04 hightekvagabond joined #salt
03:07 ageorgop joined #salt
03:08 nZac joined #salt
03:10 Alysum guess everyone is asleep now :)
03:10 hightekvagabond joined #salt
03:13 evle joined #salt
03:16 jaybocc2 joined #salt
03:16 fracklen joined #salt
03:25 cyborg-one joined #salt
03:33 favadi joined #salt
03:34 racooper joined #salt
03:34 Kraln joined #salt
03:38 bujordan joined #salt
03:41 flowstate joined #salt
03:45 paolo joined #salt
03:57 bhosmer_ joined #salt
03:58 quasiben joined #salt
04:09 nZac joined #salt
04:17 jaybocc2 joined #salt
04:17 jfg joined #salt
04:19 ageorgop joined #salt
04:19 totzky joined #salt
04:21 stooj joined #salt
04:25 jeffspeff joined #salt
04:28 zmalone joined #salt
04:29 totzky joined #salt
04:30 nZac joined #salt
04:33 totzky joined #salt
04:38 anmol joined #salt
04:41 flowstate joined #salt
04:46 totzky joined #salt
04:47 iceyao_ joined #salt
04:50 bujordan joined #salt
04:51 bhosmer_ joined #salt
04:51 totzky joined #salt
04:56 zmalone joined #salt
04:59 viq joined #salt
05:05 fracklen joined #salt
05:07 joe_n joined #salt
05:17 cyborg-one joined #salt
05:22 flowstate joined #salt
05:26 GreatSnoopy joined #salt
05:27 writtenoff joined #salt
05:29 tobstone joined #salt
05:39 rdas joined #salt
05:41 flowstate joined #salt
05:43 favadi joined #salt
05:45 bhosmer_ joined #salt
05:47 onlyanegg joined #salt
05:55 malinoff joined #salt
06:00 bujordan joined #salt
06:02 nZac joined #salt
06:02 cberndt joined #salt
06:04 ninjada_ joined #salt
06:05 kermit joined #salt
06:14 stooj joined #salt
06:15 cpowell joined #salt
06:16 ninjada joined #salt
06:17 sauvin joined #salt
06:18 jaybocc2 joined #salt
06:18 ajw0100 joined #salt
06:19 bujordan joined #salt
06:25 onlyanegg joined #salt
06:28 antpa joined #salt
06:32 kshlm joined #salt
06:32 bujordan joined #salt
06:32 ajmath_ joined #salt
06:35 jhauser joined #salt
06:35 dimeshake joined #salt
06:38 kawa2014 joined #salt
06:39 bhosmer_ joined #salt
06:41 flowstate joined #salt
06:43 kawa2014 joined #salt
06:45 cpowell joined #salt
06:47 favadi joined #salt
06:51 stooj joined #salt
06:53 neilf__ joined #salt
06:54 fracklen joined #salt
06:58 felskrone joined #salt
07:00 akhter joined #salt
07:01 CeBe joined #salt
07:03 nZac joined #salt
07:09 ywa_ joined #salt
07:10 ninjada joined #salt
07:15 netcho joined #salt
07:22 flowstate joined #salt
07:31 slav0nic joined #salt
07:34 bhosmer joined #salt
07:34 denys joined #salt
07:36 jaybocc2 joined #salt
07:40 flowstate joined #salt
07:43 elsmo joined #salt
07:44 antpa joined #salt
07:44 Glasswalker joined #salt
07:46 bernieke left #salt
07:49 cswang_ joined #salt
08:03 deniszh joined #salt
08:04 nZac joined #salt
08:11 MadHatter42 joined #salt
08:12 fracklen joined #salt
08:13 malinoff joined #salt
08:13 fracklen joined #salt
08:14 favadi joined #salt
08:14 colttt joined #salt
08:16 xfs joined #salt
08:16 Miouge joined #salt
08:17 RobertChen117 joined #salt
08:18 RobertChen117 how to do this in jinja?  if bash version < 4.1.2-8 (  bash-4.1.2-8.el6.x86_64.rpm ),  then upgrade it to bash-4.1.2-15.el6.x86_64.rpm)
08:18 tweakism joined #salt
08:20 colttt joined #salt
08:22 ninjada joined #salt
08:27 chiui_ joined #salt
08:28 bhosmer joined #salt
08:30 _JZ_ joined #salt
08:36 phx joined #salt
08:36 kim0 joined #salt
08:38 josuebrunel joined #salt
08:38 cberndt joined #salt
08:39 dariusjs joined #salt
08:39 RobertChen117 joined #salt
08:40 JohnnyRun joined #salt
08:40 flowstate joined #salt
08:40 JohnnyRun joined #salt
08:43 potens joined #salt
08:43 ninjada joined #salt
08:46 keimlink joined #salt
08:50 akhter joined #salt
08:50 elsmo joined #salt
08:52 bujordan joined #salt
08:53 jimcroft joined #salt
08:54 viq joined #salt
08:55 tweakism joined #salt
08:57 totzky joined #salt
08:58 LondonAppDev joined #salt
08:58 babilen RobertChen117: The more salty way would be to simply specify exactly which version you want (rather than to react to arbitrary versions you have), but pkg.version would come in handy here, I guess.
08:58 bujordan joined #salt
09:00 bujordan joined #salt
09:01 RobertChen117 the thing is for exactly which version, it can upgrade to it. But for newer than it, we do not want downgrade.
09:01 RobertChen117 @babilen
09:02 Rumbles joined #salt
09:03 notnotpe_ joined #salt
09:04 blue joined #salt
09:05 nZac joined #salt
09:14 rmnuvg joined #salt
09:16 babilen RobertChen117: This might be weird in RedHat-Land, but wouldn't you want to upgrade to the newest version that is available for your release?
09:17 ronnix joined #salt
09:18 ronnix joined #salt
09:19 Xevian joined #salt
09:21 bujordan joined #salt
09:22 bhosmer joined #salt
09:23 bujordan joined #salt
09:25 Chuck_0326 joined #salt
09:26 onlyanegg joined #salt
09:26 RobertChen117 latest version probably is not tested yet, so should not be pushed to production
09:26 RobertChen117 babilen
09:27 pooogles joined #salt
09:28 babilen RobertChen117: Oh, absolutely. I'm just surprised that you'd find versions in the repository (that would be installed on upgrade) that haven't been tested yet.
09:28 babilen Either way, pkg.version allows you to check the available version
09:31 GreatSnoopy joined #salt
09:40 flowstate joined #salt
09:41 antpa joined #salt
09:43 s_kunk joined #salt
09:46 totzky joined #salt
09:49 N-Mi joined #salt
09:51 antpa joined #salt
09:56 losh joined #salt
09:57 bujordan joined #salt
10:03 dgutu joined #salt
10:06 nZac joined #salt
10:08 kawa2014 joined #salt
10:10 ninjada joined #salt
10:11 joe_n joined #salt
10:14 antpa joined #salt
10:16 bhosmer joined #salt
10:20 dkrae joined #salt
10:21 RobertChen117 joined #salt
10:24 ronnix joined #salt
10:24 antpa joined #salt
10:26 anthpa joined #salt
10:31 fracklen_ joined #salt
10:39 stooj joined #salt
10:39 amcorreia joined #salt
10:42 dmaiocchi joined #salt
10:44 netcho_ joined #salt
10:45 sfxandy joined #salt
10:46 sfxandy hi everyone.  i know you can use external Pillar for targetting (via Pillar top.sls) your filesystem Pillar, but is it not possible to do lookups of external Pillar data from within filesystem Pillar?
10:46 sfxandy or is there a workaround/method of doing so?
10:47 antpa joined #salt
10:50 akhter joined #salt
10:50 tweakism there's an option, I think it's ext_pillar_first, soemthing like that
10:50 tweakism you would need to do that to reference ext_pillar from filesystem pillar; but then you won't be abe to refeence filesystem pillar from ext_pillar
10:51 sfxandy well that directive is enabled (ext_pillar_first = True)
10:52 sfxandy but when I try and do a lookup of external Pillar from a filesystem Pillar, the lookup returns nothing
10:55 tweakism dunno, pretty sure it worked for me when I tried the same thing
10:55 tweakism but I am not currently using that config
10:55 sfxandy got an example?
10:55 tweakism negative
10:56 sfxandy which backend did you use as your external Pillar...
10:56 tweakism file_tree
10:56 tweakism although it shouldn't matter
10:57 sfxandy i hear you
11:07 nZac joined #salt
11:10 bhosmer joined #salt
11:12 Netwizard joined #salt
11:13 bujordan joined #salt
11:16 jaybocc2 joined #salt
11:24 potens joined #salt
11:30 joe_n joined #salt
11:49 daemonkeeper Can I somehow include all sls in a directory without making a init.sls referencing all of them?
11:52 ronnix joined #salt
11:53 fracklen joined #salt
11:53 fracklen joined #salt
12:05 bhosmer joined #salt
12:05 jaybocc2 joined #salt
12:08 nZac joined #salt
12:12 wych joined #salt
12:16 dkrae joined #salt
12:18 iceyao joined #salt
12:18 hvn joined #salt
12:18 hvn joined #salt
12:25 dkrae joined #salt
12:29 kawa2014 joined #salt
12:41 txmoose joined #salt
12:42 netcho_ joined #salt
12:44 armyriad joined #salt
12:46 txmoose morning all.  Got a quick head scratcher... I have 3 ubuntu VMs on the same hypervisor, all have the same saltstack/salt ppa installed, but 2 of them have salt minion 2015.5.3 and one has 2015.8.0.  I'm not 100% positive how the community repos are handled with ubuntu, but is there any standard repo out there I can count on to get the latest version and only one version from?
12:46 txmoose Has anyone else noticed getting different versions from the same repo?
12:46 tweakism if repo.saltstack.com supports ubuntu, use that
12:46 tweakism those only have 1 version at a time of each package
12:47 txmoose thanks tweakism I'll give that  swing
12:47 tweakism and, it does support ubuntu
12:47 antpa joined #salt
12:47 shiriru joined #salt
12:47 tweakism txmoose: you could see the behavior you got if you had a repo w/ multiple versions, and something installed only on certain hosts that depends on the old version
12:48 txmoose tweakism: I just relized that the one vm with the newer version also has the master on it
12:48 txmoose but that's the only difference
12:49 tweakism neither of those versions are the newest, though
12:49 txmoose yea, my arch machines get 2015.8.7
12:50 txmoose but I figure fight one battle at a time
12:50 txmoose let's get the Ubuntu machines in sync, then we'll get all versions in sync
12:50 txmoose also, FWIW:  my master is only 2015.8.0, so getting things in sync with that is probably the best route
12:51 tweakism yes, master should always be >= minion version
12:51 tweakism minion > master is not supported
12:52 txmoose yep, and I think it's easier to pin packages with pacman than with anything else... but I might be partial to arch
12:53 tweakism I have no arch experience
12:53 txmoose oh, actually, repo.saltstack.com has 2015.8.7.... derp... OK, cool, I'll roll with that, then
12:53 tweakism but it's pretty trivially easy in debian too
12:53 tweakism update your master first
12:53 antpa joined #salt
12:53 txmoose I'm sure it is, I've just never done it with apt, where as I have with arch.  Arch does rolling releases, so sometimes things get updated out of sync and things break, so you kinda have to.  Sometimes.
12:54 tweakism you should probably test for regressions too before allowing automated jobs to keep running
12:55 txmoose I don't have any automated jobs running yet.  These are proof of concept things I'm working on standing up right now.  Just getting it started tonight.
12:55 tweakism ah, kk
12:55 armyriad joined #salt
12:55 txmoose But that's a good point, and I'll put it in my notes.
12:55 txmoose Thanks :D
12:55 tweakism then def. start w/ the newest version :)
12:56 txmoose That's the idea.  Which was mind boggling when ubuntu repos would hand out different versions to devices that were updating from the same location... literally the same hypervisor.  Was just.... I couldn't process it.
12:56 txmoose Lesson learned:  Don't ever trust ubuntu repos.  Just use the project official repos.
12:58 tweakism heh, I generally prefer distro repos, but for certain things they tend to be older than I want and there are decent project repos available, and salt is one of those things.
12:58 AndreasLutro considering debian/ubuntu repos are just plain files served over http, I'd say you jumped to a conclusion too fast there
12:58 tweakism PPAs, though, and other random repos, I definitely like to avoid.
12:59 bhosmer joined #salt
12:59 txmoose I've never had issues with PPAs before, but then again, I've never really had to care about versioning.  I'm learning as I go.
12:59 tweakism they're just maintained by random people
13:00 tweakism if you know and trust the maintainer, it's a little better
13:00 txmoose I'm actually trying to learn to use states right now, too, which is daunting in and of itself.  I'm trying to get a more or less basic mail server kicking and I want to use salt to manage it
13:00 atmosx joined #salt
13:00 txmoose I have the configs, so I think it's just a matter of getting the right packages managed by salt, then having my salt master push the configs to the mail server(s).
13:02 netcho_ joined #salt
13:02 rem5 joined #salt
13:03 antpa joined #salt
13:06 txmoose Woo!  Versions agree now.  Thanks for the help, tweakism
13:07 ninjada joined #salt
13:09 nZac joined #salt
13:14 bujordan joined #salt
13:16 flowstate joined #salt
13:17 antpa joined #salt
13:20 rotbeard joined #salt
13:24 akhter joined #salt
13:24 antpa joined #salt
13:25 pooogles joined #salt
13:28 onlyanegg joined #salt
13:28 kawa2014 joined #salt
13:30 renaissancedev joined #salt
13:32 antpa joined #salt
13:34 is_null joined #salt
13:35 dmaiocchi joined #salt
13:39 DammitJim joined #salt
13:44 AndreasLutro does anyone know practically how a compromised salt minion would gain access to states/files not associated with it?
13:48 tweakism it could change its grains
13:48 tweakism so if you define roles by grains, it could get data for whichever role it wants
13:49 AndreasLutro I'm not targetting anything important with grains
13:49 antpa joined #salt
13:49 tweakism I'm not sure abuot files, I was under the impression that all minions have access to all files, so sensitive ones should be in pillars
13:53 bhosmer joined #salt
13:54 subsignal joined #salt
13:57 AndreasLutro found it: salt-call cp.cache_dir salt://path/to/dir
13:57 kevinquinnyo1 joined #salt
13:57 subsignal joined #salt
13:57 ywa_ joined #salt
13:59 lompik joined #salt
13:59 ronnix joined #salt
14:01 gh34 joined #salt
14:01 racooper joined #salt
14:04 DammitJim how do you guys normally deal with setting up and "grouping" your servers
14:04 DammitJim do you configure grains
14:04 DammitJim or do you configure pillars?
14:04 babilen I rely on a mixture of hostnames/minion IDs and pillar data
14:04 antpa joined #salt
14:04 tweakism I do pillars, because I don't want what I just described to be possible.
14:04 tweakism and pillar data is targeted by minion ID which for me is the FQDN
14:04 DammitJim I don't like the grains idea for some reason (spread out configuration)
14:05 tweakism but grains is a lot easier in some ways
14:05 babilen (grains for "this is Debian version FOO, this is Ubuntu ....", this is a Xen Hypervisor, ....)
14:05 DammitJim babilen, yeah, so more of a "os thing/hardware"?
14:05 babilen tweakism: The only reasons why grains are easier are: 1. They are available everywhere and 2. You can write them easily from within salt
14:06 DammitJim pillars for what the purpose of the minion is
14:06 babilen DammitJim: Yeah, grains for all that "what kind of host in terms of hardware and the real world is this?" stuff and IDs and pillars for semantic on top of that
14:06 DammitJim babilen, how do you write them easily from within salt?
14:06 tweakism babilen: I was thinking more because you can set them at deploy-time without having to separately configure stuff on the master
14:06 DammitJim have a minion template or something?
14:07 babilen DammitJim: grains.present -- the equivalent is missing for pillars (no real surprise there)
14:07 DammitJim grains.present pushes that config to the minion if it doesn't exist?
14:07 babilen It writes it on the minion, yeah
14:08 babilen grains are entirely distributed and under the control of the minion
14:08 DanyC joined #salt
14:08 babilen (which doesn't make sense for information you want to manage in a central and secure manner)
14:08 babilen tweakism: You can just as well ensure that the data is present in pillars
14:09 tweakism babilen: you can, and as I said, I do.
14:09 babilen I mean .. in the best of all worlds you would read that info from an external pillar that is being managed by a software layer on top that you use for installing/configuring/provisioning your boxes. The "click here to make it a db server" way ;)
14:10 babilen God .. I hate backwards incompatible changes
14:10 tweakism sigh, my setup is so incredibly less clean.
14:10 tweakism next release, I can start using salt-cloud instead.
14:11 nZac joined #salt
14:11 babilen I *nearly* removed all nagios nrpe commands (would have loved to see my monitoring) because somebody decided that "nagios:nrpe:nrpe_commands" is a better name than "nagios:nrpe:checks" (for no other reason)
14:11 babilen fuckers
14:11 tweakism shouldn't it support the old way for 2 versions?
14:11 tweakism or was this some 3rd party formula or something
14:12 babilen Yes, wouldn't that have been the way to do it?
14:12 penguin_dan joined #salt
14:12 babilen Note that this was the nagios-formula and not something published by saltstack itself, but still ...
14:13 bhosmer joined #salt
14:16 DammitJim someone used the F word in here!!!
14:16 jaybocc2 joined #salt
14:17 DammitJim babilen, sorry to single you out, but is that some state file with a ton of elif statements for each server name?
14:19 rotbeard joined #salt
14:20 babilen "that" ?
14:20 TTimo hu weird, so I have two entries in my top.sls pillar file, one with '*' and one with 'master' .. my minim called 'master' only does the 'master' part and not the '*' part .. is that intentional ?
14:20 TTimo s/minim/minion/
14:20 TTimo probably some simple syntax thing that's escaping me
14:20 babilen DammitJim: I typically do that in Python and use globbing or keep it in a database
14:21 sfxandy hi everyone.  i know you can use external Pillar for targetting (via Pillar top.sls) your filesystem Pillar, but is it not possible to do lookups of external Pillar data from within filesystem Pillar?
14:21 babilen sfxandy: Look into pillarstack
14:21 DammitJim holy crap
14:21 DammitJim what do you mean do it in python and use globbing?
14:22 DammitJim what should I search online for?
14:23 sfxandy babilen, ah ok thats in the development branch of Salt...
14:23 TTimo the docs seem to indicate both '*' and 'master' should be available :-(
14:25 tweakism sfxandy: looks like you can use it on older versions easily
14:25 sfxandy yeah just drop in the stack.py file into the external modules dir
14:25 sfxandy jsut getting my head around how it works!?!?!
14:25 babilen I don't think that you find it anywhere .. The idea is essentially that I have a dictionary with globs as keys and settings/pillar data as value and then just (v for k,v in the_setting_thing if fnmatch.fnmatch(grains['id'], k)) or so
14:25 zsoftich4 joined #salt
14:26 babilen DammitJim: ^
14:26 babilen You would normally want to use a sorted dictionary and then merge all the resulting dictionaries "smartly" with dictupdate.update
14:28 baoboa joined #salt
14:31 ywa_ Hi guys, I'm learning Salt and playing with states and top file. How can I avoid salt to return failed status for minion that don't match any state when i do a "salt '*' state.highstate" ?
14:31 andrew_v joined #salt
14:32 jaybocc2 joined #salt
14:33 flowstate joined #salt
14:36 zzzirk joined #salt
14:37 numkem joined #salt
14:44 tampakrap joined #salt
14:44 jimcroft joined #salt
14:45 rotbeard joined #salt
14:47 zmalone joined #salt
14:52 _Cyclone_ joined #salt
14:54 MadHatter42 joined #salt
14:59 pooogles joined #salt
15:00 kawa2014 joined #salt
15:04 PsionTheory joined #salt
15:05 KennethWilke joined #salt
15:07 dmaiocchi joined #salt
15:08 rory joined #salt
15:08 flowstate joined #salt
15:08 Eureka70_ joined #salt
15:08 rory Hi. Is it possible, as part of a Salt state, to run a curl command, and take action depending on the HTTP status code of the response?
15:09 rory ie if 200, continue, but if 403, perform some action and then continue
15:09 hasues joined #salt
15:09 hasues left #salt
15:10 AndreasLutro rory: just write a shell script and use cmd.script
15:10 rory That feels filthy
15:11 rory OK so let me rephrase my question
15:11 rory Can I take different action in a salt state file depending on the output of a script
15:11 rory I feel that putting the curl command in a script just moves the problem
15:11 AndreasLutro look into the stateful arg
15:11 nZac joined #salt
15:12 AndreasLutro I'm not sure it'll work for what you want though
15:12 rory To be clear, it is the futur actions of the salt state I need to change. The actions can't be done within the script
15:12 rory In ansible I could catch the output of the curl command and perform different actions based on it
15:13 rotbeard joined #salt
15:14 AndreasLutro if you don't care about purity you could always just do {% set myvar = salt['cmd.run']('curl whatever') %} and do jinja logic on the string
15:16 bujordan joined #salt
15:16 fracklen joined #salt
15:16 Miouge joined #salt
15:16 rory This is being thrown out within a year, I'm happy with that
15:16 rory ty AndreasLutro I'll look into that approach
15:17 toastedpenguin joined #salt
15:18 toastedpenguin joined #salt
15:22 ronp_usa joined #salt
15:22 malinoff joined #salt
15:22 toastedpenguin joined #salt
15:23 pezus hi. i've made a custom grain 'id' on salt and want that to be templated. it contains an integer value. i templated it with {{ grains('id') }} but when running a highstate it gets resolved in [id] instead of just id
15:23 pezus anyone knows what's up with that?
15:24 toastedpenguin joined #salt
15:24 gizmodo_ joined #salt
15:24 subsigna_ joined #salt
15:25 JohnnRun joined #salt
15:25 rubendv joined #salt
15:25 rubendv joined #salt
15:26 antpa joined #salt
15:27 hoonetorg joined #salt
15:28 onlyanegg joined #salt
15:28 rm_jorge joined #salt
15:29 stooj joined #salt
15:31 iceyao_ joined #salt
15:32 rvankleeck joined #salt
15:33 morissette joined #salt
15:34 mavhq joined #salt
15:35 antpa joined #salt
15:36 rvankleeck Having some difficulty with the MD5 hash of a file on download. If I download the file using wget on the minion, and then run `salt-call file.check_hash` against it, it returns true. But the state says that the hash does not match
15:37 rvankleeck any ideas on how I could either skip the hash check, or force it to accept it?
15:37 edrocks joined #salt
15:40 flowstate joined #salt
15:42 ronnix joined #salt
15:46 hairytoez joined #salt
15:47 drawsmcgraw joined #salt
15:48 dkrae joined #salt
15:48 nZac joined #salt
15:52 Tanta joined #salt
15:54 MikaT joined #salt
15:54 Jarus joined #salt
15:54 deniszh joined #salt
15:55 onlyanegg joined #salt
15:56 freeaks joined #salt
15:56 freeaks hi there
16:03 ksalman Are there really no 2015.8.x packages for Debian? I am looking here http://debian.saltstack.com/debian/dists/
16:04 freeaks i'm new to salt, i'm trying to create a whitelist for shorewall that salt will deploy on all his minion. here's what i got so far: http://paste.debian.net/403642/
16:05 freeaks i'm having trouble figuring the proper way to use jinja template in my shorewall/rules file
16:06 mavhq joined #salt
16:06 freeaks could someone help me a bit on this ?
16:07 rvankleeck freeaks: are you trying to add a line for each of these IPs in whitelist?
16:07 freeaks yes
16:08 rvankleeck I would use a for loop for that
16:08 rvankleeck and put the whitelist in pillar data
16:08 freeaks i'm very new to salt, the simplier the better
16:08 antpa joined #salt
16:09 freeaks i'm still struggling with all the different concepts and names, grains, pillars etc..
16:09 freeaks rvankleeck, what i'm written so far look ok to you ?
16:10 freeaks *i've
16:11 rvankleeck freeaks: https://bpaste.net/show/6e26baca07e6
16:11 rvankleeck that’s how I would do it
16:12 rvankleeck however, you dould use that {% for ip in %} block around your current setup to make it work, I think
16:12 freeaks thanks a lot for the example :)
16:13 gtmanfred ksalman: they exist on http://repo.saltstack.com/#debian
16:13 ksalman huh i guess they moved the repo
16:13 ksalman thanks
16:14 ksalman awww..this repo is only for Debian 8 =(
16:15 ksalman i am looking for 7 (jessie)
16:15 cpowell joined #salt
16:15 tweakism no, they definitely have a jessie repo on repo.saltstack.com
16:15 tweakism but, 8 is jessie
16:15 ksalman whops i meant wheezy
16:16 cpowell joined #salt
16:16 rvankleeck joined #salt
16:16 ksalman http://repo.saltstack.com/apt/debian/
16:16 ksalman does not have 7
16:17 pfallenop joined #salt
16:17 mavhq joined #salt
16:17 ksalman i guess i could install it via curl but packages would've been better
16:17 ksalman the bootstrap script rather
16:17 tweakism there should be links there to a different repo that has wheezy packages
16:18 kahiru joined #salt
16:18 ksalman https://docs.saltstack.com/en/latest/topics/installation/debian.html
16:18 ksalman for wheezy it points be back to debian.saltstack.com, which does not have anyting after 2015.5.3
16:18 hightekvagabond joined #salt
16:19 shiriru joined #salt
16:19 zmalone ksalman: I've never seen that subdomain before, I believe https://repo.saltstack.com/#debian is current
16:19 tweakism welp; I guess that's life on old-stable; sorry.
16:20 zmalone which doesn't contain wheezy, I understand
16:21 ksalman It's odd that they aren't making packages for wheezy anymore. It's still widely used =)
16:21 ksalman oh well
16:21 tweakism I pretty much agree, but I don't know the reasoning.
16:22 blue left #salt
16:23 Heartsbane So I have developed this real love/hate relationship with Docker. My latest problem is Docker will not handle concurrent connections throught docker.py so if I have a state being executed on a host and another state comes along to execute something else docker just returns "Host is busy". Does anyone know of a way to queue up salt requests if a host is busy, or to initiate a sleep/wait?
16:28 voxxit joined #salt
16:29 MikaT joined #salt
16:31 cro_ joined #salt
16:31 mavhq joined #salt
16:34 llua joined #salt
16:34 rvankleeck Is there a way to use archive.extracted with a file local to the minion?
16:36 DammitJim rvankleeck, probably
16:36 DammitJim the file won't be copied over if it is already there, I think
16:37 rvankleeck DammitJim: I’m having difficulty getting archive.extracted to work on a large file (as it says the hash is different, but it checks out when i download it manually). So I’m trying to download it using cmd.run and then extract it locally.
16:38 rvankleeck but archive.extracted needs a URL for the source (e.g. salt://, http://, etc.)
16:38 jimklo joined #salt
16:38 DammitJim the hash?
16:38 DammitJim can you import the file from the minion to the master?
16:40 jfindlay rvankleeck: did you file https://github.com/saltstack/salt/issues/31466?
16:40 saltstackbot [#31466]title: archive.extracted fails on large tar.gz | ```...
16:41 rvankleeck jfindlay: indeed
16:41 rvankleeck trying to find workarounds right now
16:42 murrdoc joined #salt
16:43 jaybocc2 joined #salt
16:43 tligda joined #salt
16:44 amcorreia joined #salt
16:45 freeaks i have another question, in my  'for loop, here: IPs_WHITELIST={% for ip in my_variable[whitelist] %}ip,{% endfor %}
16:46 freeaks if we're at the last line, how can i avoid the last ip,  <--- the last comma
16:46 jimklo joined #salt
16:47 freeaks how to detect the loop is processing the last line, so i don't add the last comma
16:47 andrew_v joined #salt
16:47 tweakism can you do  ','.join(pv6-irc.snoonet.org)  in jinja?
16:47 tweakism er
16:47 tweakism ','.join(my_variable[whitelist])  rather
16:48 freeaks tweakism, so the full line would look like:
16:48 freeaks for loop, here: IPs_WHITELIST={% for ip in my_variable[whitelist] %}ip','.join(my_variable[whitelist]){% endfor %}
16:48 freeaks ?
16:48 tweakism no
16:48 freeaks ;)
16:49 tweakism if it works, and I am not sure it does
16:49 tweakism it would be
16:49 tweakism IPs_WHITELIST={{ ','.join(my_variable[whitelist]) }}
16:49 freeaks ok so i have two lines
16:50 freeaks 1st one is:
16:50 freeaks for loop, here: IPs_WHITELIST={% for ip in my_variable[whitelist] %}ip,{% endfor %}
16:50 freeaks and 2nd one is:
16:50 tweakism in python, ','.join(['a','b','c']) returns 'a,b,c'
16:50 freeaks IPs_WHITELIST={{ ','.join(my_variable[whitelist]) }}
16:50 freeaks (i forgot to remove the comma just after 'ip')
16:51 freeaks (on the 1st line)
16:51 tweakism ok
16:51 freeaks ok nice, thanks for the help , i will try that
16:51 tweakism you realize they're both intended to do the same thing, right?
16:52 freeaks what do you mean ?
16:52 freeaks one is a loop, the other one  is just concatenate to an existing string
16:52 tweakism you wouldn't need *both* lines
16:52 jimklo_ joined #salt
16:52 freeaks ah ? then i guess i missed something ;)
16:53 tweakism the purpose is to take the items in my_variable[whitelist] and make a string out of them, with a , between each one, right?
16:53 freeaks yes, but not after the last item
16:53 tweakism right.
16:53 tweakism so, in python, ','.join(['a','b','c']) returns 'a,b,c'
16:54 KennethWilke left #salt
16:54 tweakism so just that one statement alone, outputs what you want
16:54 tweakism if it works in jinja
16:54 freeaks i will try it
16:54 alemeno22 joined #salt
16:55 alemeno22 joined #salt
16:56 tweakism it looks like the jinja way to do it is:  IPs_WHITELIST={{ my_variable.whitelist|join(',') }}
16:58 mr_chris joined #salt
16:58 mr_chris Does Salt Stack have an official method for encrypting pillar data yet?
16:59 mr_chris Is it https://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.gpg.html ?
16:59 jaybocc2 joined #salt
16:59 cberndt joined #salt
17:01 evle joined #salt
17:04 mavhq joined #salt
17:04 ageorgop joined #salt
17:06 mavhq joined #salt
17:09 jimklo joined #salt
17:11 toastedpenguin joined #salt
17:14 mavhq joined #salt
17:17 malinoff joined #salt
17:17 bujordan joined #salt
17:18 writtenoff joined #salt
17:18 rvankleeck joined #salt
17:18 mavhq joined #salt
17:20 jfindlay rvankleeck: how much memory do you have on that system?
17:21 rvankleeck jfindlay: it was 1G before, now it is 15.25G
17:22 jfindlay hm, 1G for a minion to extract an arbitrarily sized file should be enough
17:26 flowstate joined #salt
17:29 rvankleeck jfindlay: I think it had to do with the fact that it was trying to load the entier ~300M file into memory in order to pull it
17:29 rvankleeck rather than just downloading it
17:30 rvankleeck looked like it was pulling into memory to then write it out
17:32 Lionel_Debroux_ joined #salt
17:36 wych joined #salt
17:38 khaije1 joined #salt
17:39 khaije1 is the function signature for a local runner the same as the analogous salt-call function call?
17:39 garphy joined #salt
17:41 sfz- joined #salt
17:42 flowstate joined #salt
17:46 impi joined #salt
17:49 gcorey joined #salt
17:57 ecdhe joined #salt
17:59 jfindlay right
18:00 jfindlay khaije1: what do you mean by a local runner?
18:00 Fiber^ joined #salt
18:01 rvankleeck joined #salt
18:01 khaije1 jfindlay: Hi, that'd be the runner formerly known as 'cmd' if I'm understanding this correctly.
18:03 jfindlay what is it called?  I know I've tried to find a cmd runner before
18:03 ecdhe joined #salt
18:04 jfindlay but if you're talking about cmd.run as an exec module function vs whatever.run as a runner, they're going to be different pyton functions in the codebase
18:04 jfindlay but will probably share the same or very similar signature
18:04 hal58th joined #salt
18:08 Eugene Is there anything I can do to speed up salt-key and/or the bash-completion that uses it?
18:08 Eugene tab-completing minion names is painful
18:09 mavhq joined #salt
18:11 pjs joined #salt
18:12 mavhq joined #salt
18:12 shaggy_surfer joined #salt
18:13 khaije1 jfindlay: I got the idea of there being a "local" runner from the description here: https://docs.saltstack.com/en/2015.5/topics/reactor/#calling-execution-modules-on-minions
18:14 shaggy_surfer joined #salt
18:15 khaije1 though the list of runners does not include it... could you characterize how these things relate to each other?
18:15 armyriad joined #salt
18:15 nZac joined #salt
18:16 jfindlay so, that's not a runner.  A runner is like an execution module that runs on the master
18:18 khaije1 jfindlay: I think I'm starting to get the hang out it. I was thrown by the notion that the reactor "mapping" occurs on the master.
18:18 wendall911 joined #salt
18:19 jfindlay I'm not familiar with the mechanism of how that works, but it looks like it goes through the client system to run commands on the minion
18:19 jfindlay what specific arg from the cmd.run execution module are you thinking of using?
18:19 Netwizard joined #salt
18:19 flowstate joined #salt
18:20 khaije1 There is also a runner.state (https://docs.saltstack.com/en/develop/ref/runners/all/salt.runners.state.html#module-salt.runners.state) so if I want to run a state on the master in response to an event is there a reason to chose either local.state.single or returner.? (can't quite figure what this option looks like)
18:20 khaije1 sending an email :)
18:20 khaije1 a demo of the "monitoring/alerting" use case
18:21 garphy joined #salt
18:24 sjorge joined #salt
18:24 sjorge joined #salt
18:24 khaije1 After communicating about the question I feel a bit more organized in my thinking enough to continue work. I'll check back if I get stumped again :)
18:24 khaije1 thanks jfindlay :)
18:25 jfindlay sure
18:27 bujordan joined #salt
18:29 rvankleeck joined #salt
18:30 rory joined #salt
18:31 baweaver joined #salt
18:33 antpa joined #salt
18:34 mavhq joined #salt
18:35 DammitJim can I easily call salt highstate for a group of minions depending on the states that they have assigned to them?
18:35 colegatron joined #salt
18:42 cprior joined #salt
18:42 antpa joined #salt
18:43 murrdoc joined #salt
18:44 metalseargolid joined #salt
18:45 metalseargolid left #salt
18:45 GreatSnoopy joined #salt
18:45 akhter Hey guys, is there something wrong with the pillars on 2015.8.7?  I created a pillar like the following and it's not showing up on the list in pillar.items.  https://gist.github.com/AkhterAli/9b82118567f8eabe9f71
18:45 jhauser joined #salt
18:46 metalseargolid joined #salt
18:46 josuebrunel joined #salt
18:48 forrest joined #salt
18:48 andrew_v joined #salt
18:49 fracklen joined #salt
18:51 disbound joined #salt
18:52 PredatorVI joined #salt
18:52 colegatron joined #salt
18:53 MatthewsFace joined #salt
18:53 Miouge joined #salt
18:53 baweaver joined #salt
18:55 mavhq joined #salt
18:55 DammitJim silly question
18:55 DammitJim brb
18:55 murrdoc tease
18:55 PredatorVI I'm a bit rusty with my Python and want to debug the salt orchestration process.  Is there any documentation on doing this?  I've been running the 'salt-run' process in the python debugger but I get into the async stuff and can't figure out how to get into the guts of state.orchestrate.  I'm currently in the bowels of mixins.py.
18:56 DammitJim ok, question
18:56 DammitJim how do I get key/value pairs out of an array?
18:57 nZac joined #salt
18:58 murrdoc ask nice
18:58 murrdoc not me
18:58 saltycharles joined #salt
18:58 murrdoc the array
18:59 tweakism DammitJim: that does seem kindof silly, since arrays aren't key->value :)
19:00 tweakism DammitJim: are you wanting keys like 0, 1, 2, etc.?
19:02 fracklen_ joined #salt
19:02 xfs joined #salt
19:03 DammitJim tweakism, yes, a silly question
19:03 DammitJim so, whatever this would be: a:1,b:2,c:3
19:03 DammitJim what do you call that? a list?
19:04 DammitJim man, I don't know my python!!!
19:04 jfindlay that would be called a dictionary, most likely
19:04 jfindlay DammitJim: for item in dict.items(): print('key: {}'.format(item[0]), 'value: {}'.format(item[1]))
19:05 rvankleeck joined #salt
19:05 fracklen joined #salt
19:05 jfindlay or `for k in data: print('key: {}, value: {}'.format(k, data[k]))`
19:06 PredatorVI any pointers on debugging state.orchestrate per question above?  How can I set a breakpoint to stop inside the state.orchestrate module?
19:06 jfindlay PredatorVI: debugging async code is difficult :)
19:07 jfindlay you might try cachedout or jacksontj
19:07 tweakism jfindlay: I assum you can for k, v in mydict.items(): print('key: {}, val: {}'.format(k, v))
19:07 tweakism in jinja?
19:07 DammitJim dictionary!
19:07 DammitJim woohoo
19:07 DammitJim but how do I do that in jinja
19:07 jfindlay tweakism: yeah, even better :)
19:07 DammitJim I guess it's the same
19:07 jfindlay should be the same
19:07 bujordan joined #salt
19:08 jfindlay akhter: if you can't get it to work, try filing an issue and I'll try to reproduce
19:09 jfindlay tweakism: were you asking earlier about documentation bugs?
19:09 tweakism yeah, yesterdayish
19:10 jfindlay tweakism: well, this is not documentation really, but it is easy and does need to get done: https://github.com/saltstack/salt/wiki/December-2015-Sprint-Beginner-Bug-List
19:10 bonzibuddy joined #salt
19:10 bonzibuddy hey folks
19:10 jaybocc2 joined #salt
19:10 bonzibuddy jinja templating -> is there a way to get a list of all minions matching a pattern
19:10 baweaver joined #salt
19:11 akhter jfindlay: [ERROR   ] Got a bad pillar from master, type str, expecting dict:  That's the error I'm getting.
19:11 jfindlay tweakism: also, we need to standardize the reference documentation for module functions.  We haven't decided on the formatting yet, but when we do and put it out for crowdsource, I'll let you know
19:11 tweakism word.
19:11 flowstate joined #salt
19:11 ajw0100 joined #salt
19:12 jfindlay mainly it has to do with the rst formatting
19:12 fracklen_ joined #salt
19:12 tweakism I can probably do a bunch of these __virtual__ ones tomorrow
19:12 ronnix joined #salt
19:14 jfindlay tweakism: nice
19:14 fracklen joined #salt
19:15 tweakism I would start now, but I was just about to go to bed :)  I keep a pretty random schedule.
19:17 jim__ joined #salt
19:18 babilen joined #salt
19:19 flowstate joined #salt
19:19 urda joined #salt
19:19 jfindlay no problem
19:19 urda can someone explain to me: Whats the difference between matching on ID for minions vs using the "environments"
19:20 urda for example, it seems easier to just match on minions with 'dev*' as dev machines and 'prod*' as prod machines instead of using hard coded environments
19:21 rvankleeck urda: environments themselves help with things like the gitfs backend, where you can have salt look at a repo and use a different branch per environment
19:22 dlam joined #salt
19:22 rvankleeck IMO: environments are there for being able to roll code through them and not have to copy/past/reformat a bunch of stuff
19:22 urda Ah ok that makes sense. Is it an "acceptable" practice to instead of having different branches, just have a single branch and use salt pillars to build out the differences? I'm using variables from a salt pillar to setup nginx urls on dev and a qa minion, that way I can keep a single design in one spot
19:22 urda so all dev, qa, prod machines get nginx. the difference is say the serverurl
19:22 urda thus the salt pillar
19:24 rvankleeck urda: if the only difference comes down to variables, I would say that is “acceptable” (i.e. makes sense to me). If you ever need to change the underlyings states/formulas, though, it would be better to use individual environments
19:24 quasiben joined #salt
19:24 urda makes sense
19:25 morissette joined #salt
19:29 fracklen joined #salt
19:30 tweakism urda: the advice I've seen given is to not use environments until you have the specific need for that capability
19:30 tweakism and correct me if I'm wrong $channel, but there seem be a lot of places where environments don't work quite right
19:30 tweakism just from flipping through the bug reports
19:31 mavhq joined #salt
19:31 denys joined #salt
19:31 Eugene Any scenario involving "environments" that I've seen is better served by having a separate master
19:32 Eugene For actual separation
19:32 urda see we were wanting to control the whole shebang from two masters who are together
19:35 DammitJim I want to use pillar information on a jinja template
19:35 DammitJim but what should I use in my jinja template in case that the pillar info doesn't exist (because it's not needed)?
19:37 rvankleeck joined #salt
19:37 bonzibuddy DammitJim you canmatch on grains and things of the like without the pillar IIRC
19:38 bonzibuddy ie if grains[id] == "foo"
19:38 jgelens joined #salt
19:38 bonzibuddy im trying to figure out how to get a list of all minions within a jinja template right now
19:38 bonzibuddy i feel like there should be a way without invoking the pillar or mine
19:39 DammitJim ok, I just have to check if the pillar exists, then
19:39 DammitJim thanks
19:39 DammitJim it's silly
19:42 Miouge joined #salt
19:43 babilen bonzibuddy: test.ping, but I'd use te mine
19:43 babilen *the
19:49 fracklen joined #salt
19:49 cpowell joined #salt
19:51 bujordan joined #salt
19:51 jfindlay for anyone who cares, the instructions at https://repo.saltstack.com/ have been revised to give more clear instructions on how to pin to a specific release
19:53 felskrone joined #salt
19:54 srinivas joined #salt
19:55 babilen +20
19:55 shaggy_surfer joined #salt
19:55 bujordan joined #salt
19:56 fracklen joined #salt
20:00 bujordan can someone help me figure out why the package python-mysqldb isn't getting installed? I have it listed in require for mysql but I'm getting KeyError: 'mysql_database.present'
20:01 bujordan here's my sls files: https://gist.github.com/bujordan/ed400a7a82bb74c41727
20:01 ajw0100 joined #salt
20:01 AirOnSkin joined #salt
20:01 bujordan this is on a fresh vagrant of ubuntu 12. manually running sudo salt minion1 pkg.install python-mysqldb fixes things
20:02 ageorgop joined #salt
20:02 Netwizard joined #salt
20:04 cyborg-one joined #salt
20:05 dlam anyone know a good way to handle a shared git repo on a remote server?  i wanna be able to have different users to say 'git pull' the latest code from it
20:05 dlam but its tricky cuz of different users/ different ssh keys etc.
20:06 bujordan dlam - here's an overview of the options: http://git-scm.com/book/en/v2/Git-on-the-Server-The-Protocols
20:06 dlam oh i think i also wanna make it work with manual 'git pull's  too  (not only with salt-call state.sls update_code) blah blah
20:06 jmedinar joined #salt
20:07 bujordan if you have less than a few dozen users it's usually not to difficult to drop their ssh key on the server and put them in an appropriate unix group with proper permissions
20:07 Edgan bujordan: Regular salt output where it lists the status of the individual states would be helpful.
20:08 dlam bujordan: ohhh ok, i guess maybe i should just put everyones key in github  (thats where the repo is)    maybe thats better idea
20:08 bujordan Edgan: sure one sec and I'll update the gist. (gotta uninstall python-mysql and re-run)
20:09 jhauser joined #salt
20:09 bujordan dlam: if it's on github why not just use their access mechanisms? does everyone in the group not have a github account?
20:09 Tanta dlam: use bitbucket or github
20:09 Edgan bujordan: On the surface, installing the package via salt or doing it manually should be equal
20:09 Tanta running your own gitlab with salt is an option or something but why bother
20:10 jmedinar Question... if I try to match my minions as follows it works
20:10 jmedinar # salt -C '* and G@os_family:RedHat and E@(_prd_)' test.ping
20:10 jmedinar _prd_isa2_db_hou2_yz_blc_:
20:10 jmedinar True
20:10 jmedinar _prd_pm_app_wa1_yz_blc_:
20:10 jmedinar True
20:10 jmedinar _prd_dmms_app_wa4_gz_blc_:
20:10 jmedinar True
20:10 bujordan Edgan: that's what I thought too - I've tried several different variations of require: and none work so it's probably just me doing something silly (i'm still new to salt)
20:10 khaije1 is there a way to quickly diff files between two minions using salt?
20:10 jmedinar but if I try to check like this is not working
20:10 dlam bujordan: yeah everyone has a github account,  maybe i am noob and need to - file.directory set permissions so everyone can 'git pull' on it
20:10 jmedinar # salt -C '* and G@os_family:RedHat and E@(_prd_) and E#(_pm_)' test.ping
20:10 jmedinar No minions matched the target. No command was sent, no jid was assigned.
20:10 jmedinar ERROR: No return received
20:10 jmedinar why is not finding the _pm_ string in the minion name ?
20:11 Edgan bujordan: the default is in order, so it shouldn't matter that require is there or not
20:11 Edgan bujordan: As long as the package comes first in the .sls
20:12 Edgan bujordan: I consider putting "failhard: True" in /etc/salt/master a must. It makes it stop on the first error it hits, instead of continuing.
20:13 bujordan Edgan: seems like that's a good idea especially when developing. I have another problem with creating a salt user for mysql but one problem at a time.
20:13 Edgan bujordan: Also for mysql there is https://github.com/saltstack-formulas/mysql-formula   I generally only use them as a reference, but the public formulas in https://github.com/saltstack-formulas/ can be useful.
20:14 bujordan Edgan: when you say default salt output - do you mean master with trace turned on? or using salt-call locally?
20:14 Edgan bujordan: You did an state.apply, I normally always use a state.highstate
20:14 AirOnSkin joined #salt
20:15 wendall911 joined #salt
20:15 jim__ joined #salt
20:15 Edgan bujordan: normal output, http://fpaste.org/329448/56431327/
20:15 bujordan Edgan: yeah I tried for quite a while getting the mysql-formula to work. I've based mine heavily off the formula, just removed the stuff that's not needed for Ubuntu
20:16 bujordan Edgan: got it - that's in the gist (https://gist.github.com/bujordan/ed400a7a82bb74c41727) under error.log (Sorry guess I should've called that console output)
20:16 Edgan bujordan: I also prefer to break it out into different sls files. init(includes only, good for putting if statements around certain states), users, pkgs, files, and services. I do them in that order.
20:17 bujordan Edgan: state.highstate returns the same thing - KeyError 'mysql_database.present'
20:17 Edgan bujordan: There should be more than just that error if you haven't changed the output option
20:17 Edgan bujordan: The default is very verbose
20:18 bujordan Edgan: normally I get something like what you put in your example, but only after I install the package (or disable the database.sls file)
20:19 Edgan bujordan: even afterward it should still mention each thing and if it was clean or not
20:19 bujordan Edgan: I just tried moving the install block to the very top of the init file before anything else - same error
20:19 rvankleeck joined #salt
20:19 jmedinar got it
20:20 jmedinar # salt -C 'E@(_prd_??_pm_)' test.ping
20:20 jmedinar _prd_pm_app_wa1_yz_blc_:
20:20 jmedinar True
20:20 jmedinar thanks anyways :)
20:20 bujordan Edgan: yeah it seems like this should be handled better rather than barfing an exception and dying
20:21 Edgan bujordan: I would also replace all the sets in database.sls with a map.jinja
20:21 bujordan Edgan: I found a discussion of this in the irc logs and someone claimed it should be fixed in 8.5 http://irclog.perlgeek.de/salt/2015-09-23. I'm on salt 2015.8.5 (Beryllium)
20:21 Edgan bujordan: and not import database.sls into init, only include it
20:22 Edgan bujordan: what version are you on?
20:22 Edgan bujordan: I have a vagrant up, going to test your state on my side
20:24 Edgan bujordan: it is ubuntu 14.04
20:24 bujordan Edgan: salt 2015.8.5 (Beryllium) - awesome thanks. it's ubuntu 12 - precise64
20:25 jimcroft joined #salt
20:25 bujordan Edgan: the datbase.sls file is almost verbatim from the mysql-formula but open to suggestions if it fixes things (and don't mind sending a pull request)
20:26 Edgan bujordan: I would look at other formulas. If that is how the mysql formula is doing it, it is less than optimal.
20:26 baweaver joined #salt
20:27 Edgan bujordan: What are you doing in minion.d/mysql.conf that you need to restart the minion? salt mine stuff?
20:29 bujordan Edgan: I thought the restart was needed for the minion to pick up mysql.conf if there were changes
20:30 Edgan bujordan: yes, but what are you doing in that file
20:30 fracklen joined #salt
20:30 bujordan Edgan: it's been a bit of a challenge finding optimal ways of doing things since salt is so flexible- the forumlas looked like a good place but I guess they're all not created equal
20:31 Tanta most formulas are overly generic and verbose
20:31 bujordan Edgan: so this is probably sub-optimal. mysql.conf just does mysql.default_file: '/etc/mysql/salt.cnf'
20:31 Tanta good starting point if you're doing something from 0
20:31 flowstate joined #salt
20:32 bujordan Tanta: they seem like a really good idea -and I like that they're supposed to "just work". but that's not been my experience thus far. Then when you start using them, it seems all the generic stuff is just extra stuff that could break
20:33 hightekvagabond joined #salt
20:33 Edgan bujordan: I get http://fpaste.org/329457/14564324/, but then my pillar doesn't include a user/pass for mysql
20:35 bujordan Edgan: my pillar file is on the gist - it's called mysql.sls. I also updated it with the mysql.conf + salt.cnf.jinja
20:36 Edgan bujordan: I now get http://fpaste.org/329461/64325981/  after putting all your pillar info in
20:37 Edgan bujordan: wonder if this is a version of python-mysqldb or 12.04 issue
20:37 Edgan bujordan: 1.2.3-2ubuntu1 on 14.04
20:37 bujordan hum. I can upgrade to 14 and see if that works. I have the image local already. let me give it a go.
20:38 bujordan Edgan: so.. try re-running now that you have the salt error. second run usually fixes it for me (after I install the python pkg of course)
20:38 AndreasLutro it's probably not an issue with python-mysqldb
20:38 Edgan bujordan: same error every time
20:39 Edgan bujordan: looks like it is basically the same version of 12.04 and 14.04
20:39 Edgan bujordan: and earlier when I was talking about standard output, I wanted to see something like, Name: python-mysqldb - Function: pkg.installed - Result: Clean
20:40 Edgan bujordan: Which says it checked to see if python-mysqldb is installed, and it already was, so did nothing
20:40 Rebus joined #salt
20:41 bujordan Edgan: the only way I can get to that output is to manually install the package first.
20:41 Edgan bujordan: just works for me on 14.04 and 2015.8.3
20:41 Edgan bujordan: but I am also doing a state.highstate
20:41 bujordan Edgan: what I can do is comment out the db stuff then I think I can get the output you're wanting. I have a vagrant up in process right now on trusty64
20:42 Edgan bujordan: I created a mysql folder in my salt directory and included mysql for the host in top.sls
20:42 bujordan Edgan: since the only thing I have in my top file is mysql I didn't think there would be a difference between highstate and state.apply mysql no?
20:42 davromaniak joined #salt
20:42 davromaniak joined #salt
20:42 Edgan bujordan: Honestly salt is buggy enough never assume. I would test the highstate directly
20:43 bujordan 'k.. just got the same error on trusty64 and vagrant up (so fresh install and going to highstate)
20:43 fracklen joined #salt
20:43 Rebus Hi guys, I would like to test some new patches for the dockng module recently merged into salt:2015.8. I have followed https://docs.saltstack.com/en/latest/topics/development/hacking.html and been successful in running the develop version of salt, but unfortuantely when trying to run with 2015.8 I get a AttributeError: 'module' object has no attribute 'BASE_THORIUM_ROOTS_DIR' when starting the salt master. That problem has been fixed in
20:43 bujordan Edgan: haha yeah. config management is still a bit of black magic. on that note - is Ubuntu the most popular in use with salt?
20:44 AndreasLutro bujordan: don't assume that states get ran in the order you define them in your sls
20:44 AndreasLutro in your states, salt is probably trying to add the mysql users/databases before installing python-mysql
20:44 bujordan AndreasLutro: I did see mention of that briefly in the docs - hence why I explicitly put the require: on mysql (and tried it several different ways)
20:44 AndreasLutro either specify order: 0 on the python-mysqldb state, or set up proper requires that specify the order of things
20:45 Edgan AndreasLutro: Unless he changed the default, and ignoring requires changing the order, in order is a same bet. This isn't puppet. :)
20:45 AndreasLutro you don't have *any* requires in database.sls
20:45 cpowell joined #salt
20:45 AndreasLutro Edgan: you'd be surprised. I'm mostly thinking about the include statement
20:46 bujordan AndreasLutro: so this brings up a good question - maybe I'm not understanding a few things. the top line in init.sls - {% from 'mysql/database.sls' import db_states with context %}
20:46 Edgan bujordan: That is only for importing variables from database.sls
20:46 Tanta that makes the pillar dict db_states available in the importing state
20:46 Edgan bujordan: variables are local name space only, per file, unless you import them
20:47 AndreasLutro Tanta: not a pillar dict, it's just a jinja variable
20:47 Edgan bujordan: your include later is what actually tells it to run database.sls
20:47 Tanta oh, whoops
20:47 bujordan right - doesn't execute anything. that happens at line 105 in init.sls no?
20:47 Tanta I always do a bunch of {% set local_var = salt['pillar.get']('whatever') %} in my states
20:47 AndreasLutro bujordan: that use of db_states seems extremely unorthodox
20:47 Edgan bujordan: AndreasLutro may have a point about your include, but then my standard operating procedure is to do all imports in the init.sls, and nothing else. So it would workaround his question about includes.
20:48 bujordan the last part of init.sls which says if db_states length > 0 then require under mysql service         {{ requisites('mysql_database', db_states) }}
20:48 AndreasLutro I would drop it entirely
20:48 Edgan Tanta: Use a map.jinja, far cleaner
20:48 AndreasLutro what are you even trying to do? reload mysql every time you add a database? that's not necessary
20:48 bujordan AndreasLutro: the only reason I used it was because that's how they did it in mysql-formula ;)
20:48 Tanta more elegant but more obfuscated
20:48 AndreasLutro right... formulas are overcomplicated and buggy
20:49 bujordan AndreasLutro: no just load databases from my pillar - seemed cleaner and less code to loop through.
20:49 jtylers joined #salt
20:49 bujordan AndreasLutro: I was about to just explicitly write out db.present nodes in a database.sls but then to maintain I have to edit both the pillar and the database.sls so. Is there an example you can point me to that would be better?
20:50 mavhq joined #salt
20:50 Rebus I resorted to storing all defaults in pillar a while ago, then merging node specific vars with state defaults and using those inside the state. That way all configuration vars reside in pillar and can be easyly displayed
20:51 bujordan Rebus: that sounds like what I'm wanting - do you have an example anywhere I can look at?
20:51 AndreasLutro bujordan: https://bpaste.net/show/afd80e752014 this is my mysql state cleaned up a bit
20:51 flowstate joined #salt
20:52 AndreasLutro forgot to remove the remove_conflicting_pkgs bit - you can ignore that
20:54 bujordan AndreasLutro: thanks that looks much simpler. For {{ root_pw }} - is there a variable declaration / import I'm not seeing?
20:55 AndreasLutro yeah I stripped it. it's just fetching from pillar
20:55 bujordan AndreasLutro: cool let me give that a shot.
20:55 Rebus bujordan: I have no online accessible version of my states folder but I started out with this https://github.com/Arabus/saltstack-node-specific-pillar-states and then just included a /srv/pillar/default folder with all the state default vars
20:56 akhter Hey guys, quick question.  If I set a grain on host 1 say 'project:1:environment:app' and on host 2 the grain was not set at all, not even the project grain.  How would I check if grain exists in jinja?
20:56 AndreasLutro I'd recommend against the managing of schema thing you're trying to do - salt is not a well suited tool for that
20:56 AndreasLutro akhter: same as with pillars, {% if salt['grains.get']('project:...') %}
20:56 akhter Thanks!
20:57 fracklen joined #salt
20:57 bujordan AndreasLutro: why do you say that? I was thinking I'd actually use salt to back db dumps up to S3? I would also only use it for initial db state - everything else would be managed by the app deploy process
20:58 Rebus Anyone have a hint how to fix the 2015.8 branch, to get it running? Apparently the issue was fixed in develop but ahs not been backported sofar
20:58 AndreasLutro I'd just use salt to make a cronjob that backs up to s3. as for the initial db state, why not just let the application handle it itself
20:59 AndreasLutro Rebus: depends on how much you want to mess up your system
20:59 bujordan AndreasLutro: is the salt s3 module *ahem* not as reliable in your experience, or some other reason for your recommendation?
21:00 AndreasLutro bujordan: you'd want database backups to work fine even if your salt master crashed overnight without you noticing, wouldn't you?
21:00 AndreasLutro besides uploading to s3 is such a simple console command if you install awscli
21:00 Rebus AndreasLutro: I have a virtualenv from where I run the current develop and wanted to try some emrges that ahve only been merged to 2015.8 so far
21:00 Edgan Rebus: what distro/version?
21:01 Rebus Edgan: gentoo
21:01 AndreasLutro Rebus: as long as you run in a virtualenv, just checkout a new branch based on 2015.8, then cherry-pick the commit you want
21:01 bujordan AndreasLutro: yep - good point. The only reason I was really even going with a master is because it's easy to run commands across a cluster and I don't already have a tool for that. I was thinking of actually doing masterless
21:01 Rebus Ah cherry-picking that was the thing i was missing
21:01 Tanta masterless is amazing
21:01 Edgan Rebus: Well then you should be advanced enough to know how to rebuild the package with newer code :)
21:02 Edgan Tanta: Masterless is useful, but it is really just effectively an installer. Real magic is writing salt code that you can reuse for masterless/ssh/master
21:02 Rebus Edgan: the problem is that 2015.8 os broken currently because od a bug in smartos that is fixed in develop and I was wondering how to fix it - cherry-picking seems just the way to do it
21:02 AndreasLutro bujordan: salt is indeed great for executing changes across clusters, but db backups can and should be done independently if you ask me
21:03 AndreasLutro Rebus: just be aware that you'll have to manually merge in or rebase new changes to the 2015.8 branch
21:03 Tanta I use autoscale + cloud init + Git + masterless to make a fully automatic architecture
21:03 bujordan my goal is a mostly automated restore from clean VM's - so the recommendation would be - salt to setup packages then roll my own db restore scripts, then whatever my standard app deploy process is?
21:04 Edgan Tanta: Say use the salt formula to make a salt master in vagrant with salt-call/masterless. Then use ssh to bootstrap a salt master in AWS, and then use the same code again to update your new salt master with itself.
21:04 Rebus AndreasLutro: I do not intend to use it for production, simply to test some recent changes to the dockerng module
21:04 Tanta the problem with salt master is that it requires an open network path from the minion to the master
21:04 bujordan Tanta: I assume that's on AWS then?
21:04 AndreasLutro bujordan: you can place the db scripts on the server with salt's file.managed, and set up cronjobs with salt's cron.present
21:04 AndreasLutro just don't let salt itself *do* the backup/restore
21:05 Edgan Tanta: I consider having to keep a copy of the pillars, or play games with them, on disk of every server a no go.
21:05 bujordan AndreasLutro: then for restore I'd just have salt pull a script from git and run it, which would fetch from S3 and do the necessary?
21:05 Tanta only store non-secure data in pillar
21:05 Edgan Tanta: And what do I do with database passwords?
21:05 AndreasLutro bujordan: maybe, or just make the script part of your states directory
21:05 Tanta the secure stuff goes into grains
21:06 Edgan Tanta: you mean the grains yml on disk?
21:06 Tanta eventually, yeah
21:06 Edgan Tanta: And how do you distribute those without writing scripts? this sounds like play games.
21:06 Tanta I'll show you
21:06 Edgan Tanta: you can do pillars securely
21:06 bujordan I thought pillars were preferred over grains for sensitive info? and IIRC, targeting by grains is less preferred?
21:06 AndreasLutro bujordan: rules don't apply in masterless
21:07 Rebus since the only trustworthy grain is the minionid I'd not do anything security related based on grains alone
21:07 Edgan bujordan: He is taking about using a yaml file on the disk of the instance to feed in grains to that instance. He isn't talking about the python script version of grains.
21:07 keimlink joined #salt
21:08 nihe joined #salt
21:08 Rebus But with masterless you need to store all the security stuff on the isntance anyway :-/
21:08 Matthews_ joined #salt
21:08 Tanta http://pastebin.com/7xN89nBQ
21:09 bujordan Rebus: I've heard some people using IAM with S3 buckets to control that but that seems a lot more work - at least for a small setup. Would be curious how others have solved that for masterless though.
21:10 Tanta IAM roles handle the security part
21:10 dimeshake hey guys - for a masterless setup, how do you control the pillar merge strategy? in master config you can set pillar_source_merging_strategy - but masterless.. does that parameter work in /etc/salt/minion as well?
21:10 Rebus so there is a grains file for every IAM role, securing access to variable sets via the aws api
21:11 AndreasLutro dimeshake: masterless just means the minion is its own master - /etc/salt/master configuration still applies
21:11 Tanta there is a grains file for each role and each environment, generally grouped together into common environment buckets that are shared across all servers
21:11 dimeshake AndreasLutro: ... duh. Thank you :)
21:12 Tanta I agree with you -- conceptually, pillar and grain are the same thing, just structured data, but one needs to be used to bootstrap the other
21:13 Rebus Sound like a feasible workaround for masterless, provided the same care is taken for state files
21:14 Tanta what good is a state file with no pillar or grain data?
21:15 Rebus depends on the state file, but it at least show which states are in use somewhere which is none of a nodes business if it doesn't need to know that state
21:15 bujordan Without access to IAM it seems going with a salt master is probably a better option (I'm probably not going on AWS initially)
21:15 Tanta I guess, the philosophy of Salt as I understood it was that states and pillars are shared across all minions anyway
21:15 ninjada joined #salt
21:16 Rebus pillars are not, at least not in master-minion mode
21:16 Rebus and with a little trickery you can create teh state list for top.sls from a nodes pillar
21:16 Rebus or a grains file in your case
21:18 Rebus Oh I forgot that in masterless mode you do not need to restrict the content of top.sls because you probably habe an instance specific top.sls anyway
21:18 Rebus *have,
21:18 Tanta I suppose it depends on how much overhead you want for separating different pieces of the Salt codebase
21:19 Tanta and how paranoid you are
21:20 fracklen joined #salt
21:20 Tanta I don't lose sleep over an application server knowing how MongoDB is configured on a database server, for instance
21:22 Rebus salt has made some very hard to understand choices considering how much information it shares, especially considering states. That workaound with the top.sls to stop spreading a top.sls with all node classifiers known to a salt-master is just one to fix that
21:22 rem5 joined #salt
21:23 Tanta I like Salt because it doesn't force any conventions down your throat and you can do with it whatever you please
21:23 Rebus Knowledge about a how setup is configured is very valuable when exploiting it which is why I try to limit my minions on a need-to-know level
21:23 Tanta it's a tool made for experts
21:24 andrej left #salt
21:25 zmalone I'm not sure I'd classify many of them as choices, Salt has grown organically, and it is pretty difficult to use Salt in a secure manner.
21:25 Rebus I do like the way it is configurable but sometimes I wisg it would offer at least the option to use a set of conventions for ease of use
21:26 Rebus one being using salt in a secure manner ;-)
21:26 AndreasLutro salt seems secure enough, except for distributing binary files securely in a distributed fashion (like gpg private keys)
21:26 zmalone It is far better then it used to be, but it is still very easy to shoot yourself in the foot
21:26 Rebus I use gpg keys from pillar - not so nice but it does the job
21:27 zmalone http://www.cryptofails.com/post/70059600123/saltstack-rsa-e-d-1 as an example
21:27 fracklen joined #salt
21:27 AndreasLutro with base64 encoding?
21:27 jim__ joined #salt
21:27 Rebus yes, deploying them is bit ugly
21:27 zmalone or CVE-2016-1866, or CVE-2015-8034
21:27 AndreasLutro I tried that but the yaml refuses to render
21:28 jimcroft joined #salt
21:29 slav0nic joined #salt
21:29 kim0 joined #salt
21:30 bujordan Hum. some of this is making me re-think just using salt master initially and maybe going masterless from the start.
21:30 Tanta it's better than hand-rolling every server
21:30 bujordan Tanta: been there, done that, hence why I'm here :)
21:30 Rebus I tried mixing salt with ansible a while ago, using ansible to deploy masterless salt on docker containers, which worked pretty nicely
21:31 AndreasLutro I think if I were going masterless, salt would not be my go-to option
21:31 zmalone Personally, the lack of comprehensive warnings about sticking pillar root inside file root was something that made my salt environment very insecure.  I've since fixed it though, and updated the salt docs.  (the initial decision to do that predated me)
21:31 bujordan AndreasLutro: what would you choose instead?
21:31 Tanta what other choices are there? puppet? chef? home-grown?
21:31 Tanta every other choice is worse than salt in large degrees
21:32 Rebus So far ansible seems a fairly good alternative
21:32 zmalone (if you put pillar root inside file root, any minion can bypass pillar targeting and request the raw pillar files)
21:32 AndreasLutro I don't know, I haven't investigated the others very much, but after digging deep into salt while maintaining an older puppet setup I'm pretty convinced puppet has an undeserved bad reputation here :p
21:33 Tanta puppet is great until you have to inherit someone else's codebase
21:33 Rebus for standalone puppet is also pretty ok
21:33 bujordan zmalone: good to know. that should be mentioned at the top of the getting started tutorial for pillars.
21:33 AndreasLutro Tanta: so basically the same as every other type of software?
21:33 Tanta I can understand anyone's Salt code in a few minutes
21:33 Tanta it's just YAML and state declarations
21:33 Rebus Tere is a warning in the state tutorial that tells you that the file_root is copied to every minion
21:34 bujordan AndreasLutro: so what would you go with instead of salt masterless?
21:34 Rebus putting pillar inside that folder thus seems an obvious bad choice to me
21:34 AndreasLutro Tanta: you haven't seen complex salt code yet then!
21:34 AndreasLutro bujordan: like I wrote above, I don't know
21:34 Tanta doubtful Andreas, I've been at it for a while
21:35 Rebus It mostly depends on how you write your states - i have a project using pure python states and the code is pretty awful to maintain
21:35 Tanta also the flexibility of the salt core states is unbeatable
21:35 zmalone bujordan: Double check permissions of every file salt touches too.  Some modules and states let you specify what permissions and ownership should be set to, others make assumptions.  Sometimes the assumptions are very dangerous.
21:35 Tanta you can't get the same thing in Chef or Puppet without downloading a patchwork of 3rd party crap
21:37 bujordan zmalone: thanks. Looks like I have file_root and pillar root separate - just from reading the docs that didn't stand out as much as it probably should have.
21:38 atmosx joined #salt
21:39 bujordan has anyone gone from a salt master to masterless and how much effort was it? It seems it wouldn't be too terrible if you have access to git + iam and a build server to push things to s3
21:42 akhter joined #salt
21:44 rem5 joined #salt
21:45 zsoftich4 joined #salt
21:46 ajw0100 joined #salt
21:47 armguy_ joined #salt
21:50 quasiben joined #salt
21:50 subsignal joined #salt
21:51 spaceSub joined #salt
21:53 jhauser joined #salt
21:54 _JZ_ joined #salt
21:59 MatthewsFace joined #salt
22:02 wangofett joined #salt
22:03 Rebus yuck that  whole smartos stuff is icky ... if id find the commit that stated the BASE_THORIUM_ROOTS_DIR shit I`d just remove it
22:03 totzky joined #salt
22:04 zmalone I missed the start of that conversation Rebus, but I'd expect smartos to break around 2015.8, a lot of non-linux platforms didn't deal with tornado in salt cleanly.
22:04 nahamu .oO(SmartOS stuff?)
22:05 nahamu Who is having trouble on SmartOS?
22:05 zmalone Someone's been complaining about it on salt-users, although I have not paid too much attention to it.
22:05 nahamu Ah, I'm not on the mailing list.
22:06 Rebus it seems i ahve problems with the smartos support in salt because it interferes from me starting tha saltmaster from the 2015.8 branhc :-(
22:06 Rebus I do not use smartos, it is just the broken support for it that breaks the branch :-(
22:07 nahamu Rebus: I'd check with sjorge... He's been hacking on a lot of the SmartOS related bits lately and may be able to help...
22:07 Rebus it seems he already has for the develop branch
22:07 nahamu Rebus: but you're saying that SmartOS specific code is breaking things on non-smartos machines?
22:08 Rebus but unfortuantely his fixes have only been backported to teh 2016.3 branch :-/
22:08 jfindlay the thorium stuff is also 2016.3 and newer
22:09 sjorge Rebus: a few simple bits went into 2015.8 to get the SmartOS package for the GZ to build again
22:09 sjorge but really so much stuff is not working in there
22:09 Rebus i have that error https://github.com/saltstack/salt/issues/30968 when running the 2015.8 branch from a virtualenv and starting the salt-master
22:09 saltstackbot [#30968]title: develop is currently broken smartos | ```...
22:09 Rebus mayhaps I did something else that kicket it
22:09 sjorge 2016.3 should be much better in that regards, especially network related
22:10 sjorge Rebus: yeah I only fixed that in develop
22:10 sjorge it made it back into 2016.3 too
22:10 Rebus I wanted to try some dockerng fixes that have recently been merged to 2015.8
22:11 Rebus And afaik they have not been merge into the newer versions yet
22:11 Rebus so either i backport the thorium stuff to 2015.8 or try merging the dockerng stuff to 2016.3
22:12 sjorge I'd summarise it as: 2015.8 -> it runs, 2016.3 -> it mostly works
22:12 mpanetta joined #salt
22:13 mpanetta joined #salt
22:16 flowstate joined #salt
22:19 Rebus Hmm then maybe it is not smartos itself but something that looks similar: http://pastebin.com/2k4FyX83  thats what I get with checked out remotes/upstream/2015.8
22:19 bhosmer joined #salt
22:21 sjorge I'm not near my dev box, but I'll have a look on my phone
22:21 sjorge maybe it's something easy
22:22 andrew_v joined #salt
22:23 Rebus thanks :-)
22:26 ZiLi0n joined #salt
22:26 sjorge Rebus: master or minion?
22:27 sjorge neiher have issues if building the SmartOS global zone package
22:28 ZiLi0n Hello everyone, I am getting following warning:  Warnings: 'name' is an invalid keyword argument. That shows when I run my module. Does it mean that mymodule.function style of calling a module from a state is being deprecated?
22:29 Rebus sjorge: I get the error when starting the master
22:29 sjorge let me see if I can get it in a virtualenv
22:30 Rebus maybe I am assuming something wrong but I though simply checking out upstream, switching to the 2015.8 branch and then installing and running from a virtualenv should do the trick
22:31 nidr0x joined #salt
22:32 sjorge Rebus: I cannot reporduce it, using https://docs.saltstack.com/en/latest/topics/development/hacking.html
22:33 sjorge for the virtual env steps
22:33 sjorge https://gist.github.com/sjorge/74e18f0665b4fc694e9b
22:33 Rebus hmm I had a different version installed in that virtualenv before maybe theres some leftovers... ill try with a clean one
22:33 srinivas joined #salt
22:35 sjorge anyway time for bed, it's late
22:36 sjorge feel free to poke me here or in #smartos tomorrow, not sure when I will be online though
22:36 Rebus thanks already ill poke at it a while longer
22:38 Rebus hmm jsut created a clean virtualenv, checkout out remotes/upstream/2015.8, same problem. remotes/upstream/2016.3 works though
22:43 Matthews_ joined #salt
22:44 zenlot joined #salt
22:45 hal58th_ joined #salt
22:46 MatthewsFace joined #salt
22:47 mavhq joined #salt
22:47 totzky joined #salt
22:48 andrew_v joined #salt
22:49 pppingme joined #salt
22:49 Rebus maybe it is some sort of system packa interference ill try the virtualenv without --system-site-packages to test it
23:02 aarontc joined #salt
23:02 totzky joined #salt
23:06 baweaver joined #salt
23:07 mavhq joined #salt
23:10 aarontc joined #salt
23:11 nidr0x joined #salt
23:14 bhosmer joined #salt
23:17 srinivas joined #salt
23:21 jim__ joined #salt
23:22 mavhq joined #salt
23:25 ageorgop joined #salt
23:35 Matthews_ joined #salt
23:36 ronrib joined #salt
23:37 bujordan joined #salt
23:37 jim__ joined #salt
23:39 shaggy_surfer joined #salt
23:39 gimpy2938 left #salt
23:40 lorengordon joined #salt
23:41 flowstate joined #salt
23:43 bujordan joined #salt
23:45 baweaver joined #salt
23:55 seblu joined #salt
23:57 jtylers joined #salt

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