Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-01-23

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

All times shown according to UTC.

Time Nick Message
00:00 sav joined #salt
00:06 yomilk joined #salt
00:09 UtahDave signull: the docs are built out of the "docs" directory in the main salt git repo.
00:09 timoguin left #salt
00:10 timoguin joined #salt
00:11 jonbrefe joined #salt
00:16 otter768 joined #salt
00:18 signull just made the pull request
00:19 oldmantaiter joined #salt
00:23 UtahDave thanks, signull!
00:23 ccarney_ROCC left #salt
00:24 tafa2 joined #salt
00:27 PovAddict joined #salt
00:30 I3olle joined #salt
00:34 bluenemo_ joined #salt
00:35 bluenemo_ joined #salt
00:41 __number5__ iggy: I know how you feel, 2014.7.1 is in my staging env yesterday :P
00:57 jhauser joined #salt
00:57 murrdoc joined #salt
01:08 jY is it possible to run masterless with external pillars?
01:12 stoogenmeyer__ joined #salt
01:14 yomilk joined #salt
01:18 jms joined #salt
01:19 Guest18311 Is there a way to get a job ID back when you are building a map file and don't want to wait?
01:20 aqua^mac joined #salt
01:22 quickdry21 joined #salt
01:24 Ryan_Lane joined #salt
01:25 druonysus joined #salt
01:25 druonysus joined #salt
01:33 murrdoc joined #salt
01:37 tmh_ joined #salt
01:41 druonysus joined #salt
01:41 druonysus joined #salt
01:50 Tahm joined #salt
01:51 nitti joined #salt
01:52 Furao joined #salt
01:57 monkey66 joined #salt
02:04 CaptainMagnus joined #salt
02:05 PovAddict I'm trying to decide if salt would be overkill for my purposes :x
02:05 KyleG never
02:05 PovAddict :)
02:05 PovAddict I have just one server
02:06 CeBe joined #salt
02:07 murrdoc yes it is
02:07 murrdoc adopt it now
02:08 murrdoc so when u ahev 100 servers
02:08 murrdoc u thank your self
02:08 murrdoc its like learning vim
02:08 PovAddict 'yes it is' sounds like 'yes it is overkill'
02:08 PovAddict :P
02:08 PovAddict considering my original question
02:08 murrdoc it sucks when u start
02:08 murrdoc but then u hate when its not around
02:08 murrdoc anyhow
02:08 murrdoc /pontification
02:09 murrdoc original question ?
02:09 PovAddict <PovAddict> I'm trying to decide if salt would be overkill for my purposes :x
02:09 PovAddict <murrdoc> yes it is
02:09 PovAddict doesn't sound right :P
02:09 murrdoc its totally overkill
02:10 murrdoc but u should do it
02:10 PovAddict oh :D
02:10 PovAddict puppet and chef were other options
02:11 murrdoc do u like managing versions of ruby
02:11 PovAddict I don't know anything about ruby other than the very basics of the language
02:11 PovAddict python was a big reason why I decided to try salt :)
02:12 murrdoc then sir
02:12 murrdoc that is your answer
02:13 murrdoc in features chef is pretty good , puppets got a better 'external node classifier', salts got the best swiss army knife of stuff , and ansible is nice if u like to do stuff ssh stuyle
02:13 murrdoc google ryan lane puppet salt ansible
02:13 murrdoc for more intelligent answer
02:17 GabLeRoux joined #salt
02:20 GabLeRoux joined #salt
02:20 bhosmer joined #salt
02:21 MugginsM joined #salt
02:24 hvn joined #salt
02:24 hvn joined #salt
02:36 favadi joined #salt
02:38 bhosmer joined #salt
02:45 TheThing joined #salt
02:46 twellspring joined #salt
02:47 aquinas joined #salt
02:51 PovAddict http://docs.saltstack.com/en/latest/topics/tutorials/starting_states.html#default-data-yaml I think the docs may be wrong here
02:52 PovAddict "The second and fourth lines are the start of the State Declarations, so they are using the pkg and service states respectively." but the second and fourth lines are "pkg.installed: []" and "- require:" respectively
02:52 rocket joined #salt
02:56 PovAddict yep definitely wrong
03:02 murrdoc looks fine to me
03:02 murrdoc at most it needs to be
03:03 murrdoc apache: pkg.installed: name: apache service.running: - require: - pkg:apache
03:03 murrdoc with appropiate \t \ns
03:03 PovAddict I just tracked down the commit that changed the example code
03:03 PovAddict and I was right, and it was already fixed, just apparently the fix was not deployed to the website yet
03:05 mosen joined #salt
03:11 vbabiy joined #salt
03:12 brendanashworth joined #salt
03:12 wincyj joined #salt
03:13 brendanashworth left #salt
03:13 renoirb joined #salt
03:13 jsm joined #salt
03:15 murrdoc nmadhok isnt around
03:46 wolfpackmars2 joined #salt
03:48 wolfpackmars2 I want to run a particular state every hour for certain minions.  I'm thinking of setting up a cron job on the minion.  The reason I'm thinking of doing it this way is because I can configure this in the state file itself
03:48 MugginsM joined #salt
03:49 murrdoc http://docs.saltstack.com/en/latest/ref/states/all/salt.states.schedule.html
03:49 wolfpackmars2 The other option is to use salt's scheduler, either through config or pillar - but this would require either restarting the minion/master (whichever its set up on) or reinitializing the minion pillar
03:51 murrdoc no restarts needde
03:51 murrdoc saltutil.refresh_pillar should cover refreshing the minion pillars
03:52 murrdoc according to http://docs.saltstack.com/en/latest/topics/jobs/#scheduling-jobs
03:52 wolfpackmars2 is that different from this: http://docs.saltstack.com/en/latest/topics/jobs/schedule.html
03:53 murrdoc doesnt look like it
03:53 wolfpackmars2 I'm looking at the second paragraph at that link which says the schedule option is configured in the config files
03:53 murrdoc right
03:53 murrdoc oh you dont have it on right now
03:53 murrdoc got it
03:53 murrdoc my bad
03:54 murrdoc i ll keep quiet no
03:54 murrdoc now*
03:54 wolfpackmars2 no ... I'm thinking of a way to set it up as part of the state file
03:54 wolfpackmars2 no need for apologies murrdoc ... I wasn't clear at first
04:03 PovAddict why does the example apache state file create a user and a group? doesn't the distro package for apache do that already?
04:15 TheoSLC joined #salt
04:16 StDiluted joined #salt
04:18 yomilk joined #salt
04:21 krelo joined #salt
04:24 wolfpackmars2 which example are you referring to?
04:31 TyrfingMjolnir joined #salt
04:33 malinoff joined #salt
04:36 kermit joined #salt
04:48 otter768 joined #salt
04:51 echtish joined #salt
04:53 GabLeRoux joined #salt
04:54 Furao joined #salt
04:57 mdupont joined #salt
04:59 stoogenmeyer__ joined #salt
05:02 bdols joined #salt
05:05 Diaoul joined #salt
05:06 gaker_ joined #salt
05:08 gaker_ joined #salt
05:09 GabLeRoux joined #salt
05:14 teohhanhui joined #salt
05:18 teohhanhui UnicodeDecodeError when using pkgrepo.managed ppa - https://gist.github.com/teohhanhui/65e395d5595296f6bcc9
05:18 mosen joined #salt
05:18 aqua^mac joined #salt
05:18 yomilk_ joined #salt
05:21 teohhanhui Salt is v2014.7.1
05:22 dec so I'm just starting with Salt... I ran a simple SLS which just runs pkg.installed on 'nginx'.  It worked fine and installed nginx (and its dependencies, nginx-core and nginx-common) on my ubuntu minions.  I then manually apt-get remove'd the 'nginx' package and tried to run the SLS again, but it ran no action as it already thought nginx was installed.
05:23 dec it turns out that not until the 'nginx' package *and* it's dependencies nginx-core and nginx-common were removed, did pkg.installed attempt to reinstall anything.
05:23 dec this seems undesirable - if a package is missing and the desired state is that it's installed, surely the action should be to install it...
05:23 dec is this a bug, or the correct functionality?
05:24 GabLeRoux joined #salt
05:28 teohhanhui When I run `sudo apt-add-repository ppa:ondrej/php5` manually it works as expected
05:28 teohhanhui I'm doing `vagrant provision` from a Windows host if that's relevant
05:33 __number5__ pkgrepo ppa is broken in 2014.7.0, maybe still not fix
05:34 Ch3LL_ joined #salt
05:35 teohhanhui @__number5__ it works when I turn off "salt.verbose = true" in Vagrantfile (I'm guessing that's just log level for Salt?)
05:36 teohhanhui lol mention fail.. got the syntax mixed up
05:38 teohhanhui I see many Windows-related bugs in the issues, so I'm not sure if I should open yet another..
05:39 GabLeRoux joined #salt
05:44 teohhanhui will verify again, probably it has nothing to do with Windows anyway (it can't be, right?)
05:45 brendanashworth joined #salt
05:45 brendanashworth left #salt
05:46 echtish joined #salt
05:47 __number5__ what if you manually run `salt-call --local state.highstate`? will that work
05:47 __number5__ add -l debug to see what's the real errors behind
05:48 yomilk joined #salt
05:56 nafg_ joined #salt
05:56 malinoff joined #salt
06:04 mohae_ joined #salt
06:06 echtish joined #salt
06:07 catpiggest joined #salt
06:23 gaker_ joined #salt
06:39 bhosmer_ joined #salt
06:40 beneggett joined #salt
06:48 hvn joined #salt
06:48 hvn joined #salt
06:49 Nazzy joined #salt
06:49 yomilk_ joined #salt
06:52 Nazca__ joined #salt
06:58 otter768 joined #salt
06:59 Nazzy joined #salt
07:00 calvinh joined #salt
07:00 colttt joined #salt
07:04 teohhanhui __number5__: the exact same error, same traceback.. i don't see any difference
07:04 beneggett joined #salt
07:04 teohhanhui anyway, is it a python-software-properties bug?
07:09 calvinh joined #salt
07:10 toanju joined #salt
07:11 bluenemo joined #salt
07:11 bluenemo joined #salt
07:13 Auroch joined #salt
07:17 calvinh_ joined #salt
07:17 teohhanhui uhh, so Salt is not yet python 3 compatible huh? but it runs fine in most cases lol...
07:19 nahamu joined #salt
07:21 teohhanhui i wonder why it can't detect the system locale properly though? it's set to en_US.UTF-8
07:23 jonatas_oliveira joined #salt
07:23 flyboy joined #salt
07:27 xv joined #salt
07:30 AndreasLutro joined #salt
07:31 jhauser joined #salt
07:31 alfemy joined #salt
07:33 alfemy Hello
07:34 alfemy Could someone help me?
07:34 alfemy I cannot save iptables using salt.
07:34 tomspur joined #salt
07:35 alfemy via CLI it works "salt node*cluster iptables.save"
07:35 alfemy but in sls file I cannot get it worked.
07:35 alfemy I got an error 'State 'iptables.save' found in SLS 'cluster/testing' is unavailable'
07:37 Furao alfemy: https://github.com/saltstack/salt/blob/2014.1/salt/modules/iptables.py#L22
07:40 alfemy @Furao do you mean that iptables is not avaibled on the host?
07:40 Furao yes
07:42 alfemy @Furao it is not true. I can run "salt node*cluster iptables.save" via cli, and it works as expected
07:42 alfemy Also iptables.flush works form sls file.
07:43 Furao http://docs.saltstack.com/en/latest/ref/states/all/salt.states.iptables.html#module-salt.states.iptables
07:43 Furao there is no iptables.save state
07:45 monkey66 left #salt
07:48 alfemy hmmm
07:48 alfemy so, and how I can flush and after that save iptable rules???
07:49 Furao module.run: - name: iptables.save
07:51 alfemy hoho
07:51 alfemy thanks, will try right now:)
07:52 dkrae joined #salt
07:54 KermitTheFragger joined #salt
07:57 trikke joined #salt
08:00 asdasd joined #salt
08:01 hebz0rl joined #salt
08:04 nettus_ joined #salt
08:06 terinjokes joined #salt
08:09 JlRd joined #salt
08:15 stanchan joined #salt
08:16 kawa2014 joined #salt
08:38 I3olle joined #salt
08:38 felskrone joined #salt
08:39 jtang joined #salt
08:45 calvinh joined #salt
08:45 lb1a joined #salt
08:48 greentea joined #salt
08:52 JlRd joined #salt
08:54 calvinh_ joined #salt
08:59 otter768 joined #salt
08:59 bud joined #salt
08:59 bud Hello everyone.
09:00 bud Is anybody running salt 2014.7 on Mac OS X 10.10?
09:00 bud I have a problem with salt 'localhost' pkg.list_pkgs.
09:00 bud The minion tells me it cannot find brew.
09:01 bud brew is at /usr/local/bin/brew and I think Salt doesn't have /usr/local/bin in $PATH.
09:01 bud I start Salt with launchd.
09:02 bud I have tried to setup PATH environment with launchctl setenv PATH /usr/local/bin:/usr/bin:/bin:/sbin:/usr/sbin.
09:02 bud Restarted master and minion and I get the same response, cannot find brew.
09:11 Hybrid joined #salt
09:12 jonatas_oliveira joined #salt
09:14 chiui joined #salt
09:24 bluenemo joined #salt
09:24 CeBe joined #salt
09:38 arno joined #salt
09:53 N-Mi joined #salt
09:53 N-Mi joined #salt
09:57 slafs joined #salt
09:57 slafs left #salt
10:06 rawkode joined #salt
10:06 rawkode joined #salt
10:17 bud joined #salt
10:20 moio joined #salt
10:25 linjan joined #salt
10:26 modafinil_ joined #salt
10:27 Guest58858 joined #salt
10:30 zadock joined #salt
10:31 Tahm joined #salt
10:39 malinoff joined #salt
10:41 bhosmer joined #salt
10:43 fredvd joined #salt
10:49 bud joined #salt
10:49 carnedepassaro joined #salt
10:52 ttrumm joined #salt
10:53 ttrumm hi is there a way to use  file.managed to cpy file from /etc/somedir/file using source: /etc/somedir/file. Want to copy file what is not in formula dir
10:57 Ligthert ttrumm: reading through http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html#salt.states.file.managed And I am not sure....
10:58 ughloru joined #salt
10:58 __number5__ ttrumm: you need file.cp
10:58 ughloru Hello, is there any way to make a "require" conditional dependent on the completion of a macro?
11:00 otter768 joined #salt
11:01 jonatas_oliveira joined #salt
11:01 calvinh joined #salt
11:02 dabb joined #salt
11:08 cruatta joined #salt
11:20 cruatta joined #salt
11:25 miner joined #salt
11:25 miner left #salt
11:29 viq ughloru: what do you mean by "macro"?
11:30 ughloru We use jinja template macros for some configuration that is similar between multiple packages.
11:31 ughloru e.g. setting up a virtualenv, pip installing some packages, running a service
11:31 viq How about wrapping the require line in {% if ... %} ?
11:31 ughloru I don't know if that's what I mean.
11:32 viq I don't either ;)
11:32 ughloru If,say, I have three formulae in my macro, and I want condition some other formula, external to the macro, on the successful completion of all three macro formulae.
11:33 ughloru I wanted to know if there was any way of doing `require: macro: <macroname>`
11:33 calvinh_ joined #salt
11:34 ughloru I've got a workaround now, where I've inserted a formula at the end of the macro that requires the completion of all other the other macro formula: end-macro: cmd.run: echo "Macro complete"  require: ...
11:35 ughloru And I require that cmd in external formulae
11:35 ughloru Which seems to work
11:35 thunderbolt_ joined #salt
11:36 ughloru I think that our whole setup of using jinja macros to abstract repeated patterns is a mistake. It would probably be better to make a custom _states module to do it.
11:38 calvinh joined #salt
11:44 octarine joined #salt
11:45 giantlock joined #salt
11:47 jrluis joined #salt
11:49 calvinh_ joined #salt
11:54 ughloru Completely different question: I'm setting up a new salt-master, but I'm having trouble with hgfs. It works on my existing master, a local Ubuntu box, but not on my new master, running on AWS EC2
11:54 ughloru I've confirmed that `ssh -Tv hg@bitbucket.org` works, so it's not an access issue.
11:55 ughloru But any salt command I run returns "Passed invalid arguments: string indices must be integers, not str"
11:55 ughloru Nothing in the logs looks relevant
11:57 ughloru Both masters are running 2014.1.11 (Hydrogen)
11:59 felskrone1 joined #salt
12:04 jeffspeff joined #salt
12:07 ughloru Actually, these two lines in the log look relevant: "[DEBUG   ] Updating fileserver cache " "[ERROR   ] Exception (1, '', '') occurred in file server update"
12:09 tkharju joined #salt
12:09 bhosmer joined #salt
12:13 istram joined #salt
12:13 Tyrm joined #salt
12:14 Grokzen joined #salt
12:14 mortis_ can i match on nodegroup in the pillar top.sls?
12:15 ughloru Actually, more info. running `salt-run fileserver.update` results in an error related to hglib
12:18 arno joined #salt
12:18 ughloru hglib.error.CommandError: (1, '', '')
12:18 ughloru Stacktrace: https://gist.github.com/tensorproduct/3b96375bcd71c18ddef7
12:25 viq ughloru: as for previous, you could put the macros in a separate file and include them in relevant files
12:26 viq ughloru: as for hgfs, are you sure dependencies are installed? No extra space in URL somewhere?
12:27 calvinh joined #salt
12:31 mortis_ i dont get complaints when i do this: dev:
12:31 mortis_ 'N@development':
12:31 mortis_ - match: nodegroup
12:31 mortis_ missing \n after this:
12:31 mortis_ but it doesnt seem to match anything
12:32 mortis_ if i try 'development' or without '
12:32 mortis_ it says bad pillar
12:32 giantlock joined #salt
12:36 calvinh_ joined #salt
12:37 bhosmer joined #salt
12:41 Grokzen joined #salt
12:42 viq mortis_: I don't see N@ notation listed on http://docs.saltstack.com/en/latest/topics/targeting/compound.html  - and with that notiation it should be match: compound, not nodegroup. Unless the nodegroup indeed is called 'N@development'
12:44 rawkode| joined #salt
12:45 jtang joined #salt
12:49 calvinh joined #salt
12:49 jonatas_oliveira joined #salt
12:56 ughloru viq: from looking at the hgfs.py source, I think that the only dependencies are mercurial and python-hglib. Both of which I have installed.
12:59 Tyrm joined #salt
12:59 hobakill joined #salt
13:01 otter768 joined #salt
13:02 ben_ko joined #salt
13:03 natewalck joined #salt
13:04 tkharju joined #salt
13:04 ben_ko Hi guys, just out of curiosity: On OS X, when I run a salt-minion, should I do it as LaunchDaemon (root) or as LaunchAgent (user)? Or is there a even better way to ensure the start of the salt-minion?
13:04 babilen ughloru: Did you restart the minion after installing them?
13:06 ughloru babilen: yes, I just tried restarting both master and minion. No change.
13:07 ughloru Also, I see the error when calling `salt-run fileserver.update` from the master, so I don't think that the minion would make any difference?
13:07 mapu joined #salt
13:07 Grokzen joined #salt
13:07 babilen No, it wouldn't.
13:08 ughloru Just tried upgrading the master to 2014.7.1, no difference.
13:08 babilen But as you restarted the master too (and have the dependencies installed there)
13:08 Grokzen joined #salt
13:08 babilen You are, btw, the first person I am aware of that uses hgfs
13:09 ughloru I've found it mostly fairly functional.
13:10 ughloru Is gitfs more actively used and maintained?
13:10 numkem joined #salt
13:11 babilen used, definitely
13:11 intellix joined #salt
13:12 ughloru If I can't resolve this problem, I might move the our salt repo over to git.
13:13 SpX joined #salt
13:13 numkem joined #salt
13:15 wnkz__ joined #salt
13:18 qybl joined #salt
13:20 vbabiy joined #salt
13:23 mndo joined #salt
13:25 GabLeRoux joined #salt
13:27 redzaku joined #salt
13:28 krelo joined #salt
13:29 cetex joined #salt
13:29 hvn joined #salt
13:29 hvn joined #salt
13:30 cetex i've got an issue. i have an upstart job that i start through salt, but it doesn't return. how do i run the equivalent of "service --no-wait xxxx start"? or "start --no-wait xxxx"?
13:31 antonw joined #salt
13:32 dexter91x joined #salt
13:32 dexter91x joined #salt
13:33 jsm joined #salt
13:35 cetex the problem is that this job is waiting for another job to start, but isn't started at this moment since salt hasn't finished running yet.
13:38 ughloru Okay, I've made some progress. A local cache of the remote hg repo is created under /var/cache/salt/master/hgfs/<hash>/, in which hglib tries to run `hg showconfig paths`, to check that the cached repo has a record of the remote repo. For some reason, my new salt master is not creating this cache correctly, and it doesn't have any default path.
13:41 mawbid joined #salt
13:41 rawkode joined #salt
13:41 rawkode joined #salt
13:41 ughloru In mercurial, default paths are located in .hg/hgrc. If I manually populate this file, then the fileserver.update works. And the salt-master behaves normally
13:41 ughloru So that's definitely where the problem is.
13:41 mawbid joined #salt
13:42 ughloru Now I just need to figure out why the file isn't being created.
13:42 kitp joined #salt
13:42 gngsk joined #salt
13:42 Whissi joined #salt
13:44 joeyparsons joined #salt
13:47 gngsk_ joined #salt
13:49 ughloru Okay, it seems that hgfs checks that the path is in the repo config using hglib, and if not, it should create it. But hglib throws an exception if the command fails, leading to the error I've been seeing.
13:50 jk_ joined #salt
13:50 jonatas_oliveira joined #salt
13:51 jk_ Hi there! the module "state.sls_id" does not include watches... Is this a bug or a feature?
13:52 eliasp joined #salt
13:53 Grokzen joined #salt
13:57 steier joined #salt
13:58 JDiPierro joined #salt
14:00 jeremyr joined #salt
14:01 kitp newb ?, how do i mine the assigned elastic IP with salt-cloud?  i can get the private_IP but not sure how to get the public one.
14:01 oldchacal joined #salt
14:10 nkuttler kitp: can't you get it with the network module?
14:10 nkuttler oh, with salt cloud, not sure..
14:11 malinoff joined #salt
14:11 kitp yeh, i need to get the elastic ip so i can set up a route53 record with it, during the provisioning process.
14:16 phx how do i refer to the ip4_interfaces / eth1 grain in a state file? salt $host grains.get ip4_interfaces lists eth0/eth1/lo0, but for ip4.interface.ipv4 it returns nothing
14:18 nitti joined #salt
14:18 phx ahh ip4_interface:eth1
14:21 bhosmer joined #salt
14:26 abe_music joined #salt
14:26 mdasilva joined #salt
14:26 glyf joined #salt
14:26 mdasilva morning all
14:27 dexter91x hello all
14:28 dexter91x i would like to know what is the importance of "with context"
14:30 oldchacal joined #salt
14:31 danemacmillan joined #salt
14:31 munhitsu___ joined #salt
14:33 mdasilva dexter91x: the jinja context of the calling template is passed through to the called jinja template when using 'with context'
14:33 mdasilva http://jinja.pocoo.org/docs/dev/templates/
14:33 mdasilva http://jinja.pocoo.org/docs/dev/templates/#import-context-behavior
14:40 signull joined #salt
14:40 cpowell joined #salt
14:41 TheoSLC joined #salt
14:42 dexter91x mdasilva: well i tried to remove "with context" and then make it "without context" but there is no difference , everything still passing
14:43 murrdoc joined #salt
14:47 rawkode joined #salt
14:49 Rawkode_ joined #salt
14:50 PovAddict joined #salt
14:51 andrew_v joined #salt
14:51 dude051 joined #salt
14:56 booly-yam-7859 joined #salt
15:00 Tyrm joined #salt
15:00 PovAddict why does the example apache state file create a user and a group? doesn't the distro package for apache do that already? http://salt.readthedocs.org/en/v2014.7.0/topics/tutorials/starting_states.html#adding-configs-and-users
15:00 che-arne joined #salt
15:00 signull joined #salt
15:02 otter768 joined #salt
15:02 elfixit joined #salt
15:03 JDiPierro joined #salt
15:07 dexter91x PovAddict: as they said "a user and group may need to be set up." so its created by default as you said but you may want to modify this default behavior
15:07 babilen PovAddict: Sure, but that would be an assumption (and you naturally want to adapt every example you see to your needs)
15:08 PovAddict is there any way to handle software that needs to be compiled because it's not packaged?
15:08 PovAddict (apart from 'package it' :P)
15:09 babilen "compile it" ?
15:09 kaptk2 joined #salt
15:10 dexter91x PovAddict: do you mean installing from source ?
15:12 TheoSLC joined #salt
15:12 housl joined #salt
15:12 twellspring joined #salt
15:15 dexter91x babilen: i don't know when should i use "with context" and when i shouldn't , in all formula examples they use "with context" but i don't know why , i checked the documentation of jinja and saltstack but i don't get it , can you help me with that
15:16 johtso joined #salt
15:17 cambridgev joined #salt
15:20 PovAddict I think I'll leave this software out of salt control for now
15:20 Ligthert I am trying to use salt-cloud with the digital ocean provider, but it seems I've entered the twilight zone. One one hand the docs specify using "personal_access_token". On the other hand salt-cloud is asking for an api_key :s
15:20 masterkorp Hmm, does salt-bootstrap covers macosx ?
15:21 babilen dexter91x: Isn't that covered in the jinja docs?
15:22 dexter91x babilen: i couldn't understand it clearly , and i tried to remove "with context" from my formula and nothing happened everything still passing
15:23 mdasilva dexter91x: according to http://jinja.pocoo.org/docs/dev/templates/#import-context-behavior jinja 2.1 passes contexts
15:23 babilen dexter91x: I'm clearly missing context then ;)
15:23 favadi left #salt
15:23 mdasilva so i think the with context is assumed/default since jinja 2.1
15:24 babilen mdasilva: Ah, good find
15:24 ponpanderer joined #salt
15:24 dexter91x mdasilva: what about when changing it to "without context" my formula still passing that why i don't understand the need of it
15:25 ponpanderer Hey, I'm noticing a salt-master CPU spike on 2014.7. CPU on a particular salt-master pid has randomly spiked (and stayed pinned) at 100%. happened on this particular salt-master on 3 occassions, but never seen it elsewhere.
15:25 mdasilva dexter91x: lol odd
15:26 ponpanderer strace on the pid that is pinned yields only: restart_syscall(<... resuming interrupted call ...> <unfinished ...>
15:26 mattl joined #salt
15:26 beneggett joined #salt
15:27 peters-tx joined #salt
15:27 scoates joined #salt
15:28 lothiraldan joined #salt
15:28 oldchacal left #salt
15:30 mdasilva whats a good way of getting notified of a job completion or getting the full job report from a reactor event?
15:34 rudi_s joined #salt
15:35 randomuser joined #salt
15:35 jab416171 joined #salt
15:39 iggy Ligthert: personal_access_token is old (afaik)
15:39 iggy Ligthert: so those docs probably need to be upgraded
15:40 iggy masterkorp: I think there's a bug open about it (i.e.e not yet
15:41 ProT-0-TypE joined #salt
15:42 iggy mdasilva: event.fire? (yes, fire another reactor at the end of your reactor) also returners are the canonical way to get output from jobs (aside from looking through master logs)
15:42 mdasilva thanks iggy i'll look into that
15:44 rojem joined #salt
15:46 msheiny joined #salt
15:46 johtso joined #salt
15:46 teogop joined #salt
15:50 racooper joined #salt
15:51 jhauser joined #salt
15:52 hasues joined #salt
15:52 hasues left #salt
15:53 gladiatr joined #salt
15:56 twellspring joined #salt
15:56 wrench joined #salt
15:57 twellspring joined #salt
15:59 Furao joined #salt
16:00 nitti_ joined #salt
16:01 Ligthert iggy: The docs are up to date, but the code in helium haven't been update since release... while the v2 API at DO have.
16:01 Ligthert iggy: There was an issue on github on this.. now closed..
16:02 murrdoc yo iggy, i was looking at  https://github.com/saltstack-formulas/collectd-formula/blob/master/collectd/map.jinja
16:02 murrdoc any reason why we dont mention explicitly what the grain is that we are looking up stuff by
16:05 booly-yam-7566 joined #salt
16:06 Ozack-work joined #salt
16:07 booly-yam-7361 joined #salt
16:09 jngd joined #salt
16:11 ralala joined #salt
16:12 intellix joined #salt
16:14 ralala joined #salt
16:14 wnkz__ when you use salt-api + reactor + custom returners ; is there a way to pass arguments to the returner ?
16:14 redzaku joined #salt
16:14 twellspring joined #salt
16:15 copelco joined #salt
16:16 davet joined #salt
16:22 davet joined #salt
16:23 oldmantaiter joined #salt
16:24 ProT-O-TypE joined #salt
16:24 pjota joined #salt
16:24 pjota hi everyone
16:25 whiteinge wnkz__: do you mean to specify a returner? I don't think you can pass arguments to a returner
16:25 schristensen joined #salt
16:26 phpdave11 hi pjota
16:26 pjota What are dashboard you recommend for salt?
16:26 thedodd joined #salt
16:26 pjota =-/
16:27 ProT-O-TypE joined #salt
16:27 malinoff pjota, cli, it's the only one that stable enough :)
16:27 wnkz__ whiteinge: no, i already have a small custom returner ; and i have data incoming on a webhook ; problem is ; i want to my returner to post to a callback url that is inside the json I get from the webhook
16:28 _JZ_ joined #salt
16:28 phpdave11 i only use command line
16:28 Morbus joined #salt
16:28 wnkz__ I could do it directly in the reactor but I need to know my jobs' states before sending anything
16:29 ProT-0-TypE joined #salt
16:30 pjota phpdave11, malinoff i know, but in my environment is necessary
16:30 wnkz__ pjota: I'm not aware of a lot of dashboards except Halite but it is in very early stage
16:30 wnkz__ pjota: https://github.com/saltstack/halite
16:32 iggy murrdoc: because it's the default (os_release or whatever)
16:32 whiteinge wnkz__: Returners only have access to the data from the job. So the url would have to live in that
16:32 pjota someone uses foreman with salt?
16:32 jonatas_oliveira joined #salt
16:33 iggy pjota: unfortunately, halite is (effectively) dead... there's salt-pad
16:33 mpanetta joined #salt
16:34 pjota iggy, hmm
16:34 whiteinge Foreman recently got salt support. Rundeck isn't quite a dashboard but also has good salt support. salt-pad is a new project that looks nice
16:34 mpanetta joined #salt
16:35 iggy os_family
16:35 murrdoc salt-pad ?
16:35 Vye Is there an equivalent to test.ping to verify the minion process isn't hung _without_ pinging the master? I don't care if it is connected for this purpose.
16:35 iggy https://github.com/tinyclues/saltpad
16:36 iggy Vye: salt-run manage.up ?
16:36 Vye One thought was to use a schedule that logged a heartbeat using logger, but not sure if that exercises the same workflow as functions being received.
16:37 murrdoc nice flask app
16:37 whiteinge Vye: there's the presence system that watches active tcp connections. It's master-side only
16:37 murrdoc uses mongo!
16:37 murrdoc dude why come
16:37 Vye iggy, that requires master -> minion communication, no?
16:38 tligda joined #salt
16:38 whiteinge Manage.up wraps test.ping
16:39 flavio2 joined #salt
16:39 bhosmer_ joined #salt
16:40 cruatta joined #salt
16:40 flavio2 is it possible to have a disconnected minion keep running the latests version of the states he obtained from the salt server until the connection with the master is restored?
16:41 iggy I don't know the docs are pretty vague on what it actually does and I've never read the source
16:41 murrdoc ah i see why mongo
16:41 Vye whiteinge, If there ever was an issue with a minion I want that to be detected and signal supervisor to restart it. Any thoughts on how I could do that without pinging the master?
16:42 Vye whiteinge, I'm not against pinging the master, but I don't want it to restart the minion just because it lost its connection with the master.
16:42 Brick___ joined #salt
16:43 wnkz and for now I'm using 'cmd.state.sls' in my reactors ; should I consider changing that to a runner and 'state.orchestrate' ? What's the real difference ?
16:44 f3d hello, my master is suffering (100% CPU) when the minions are returning (~350)
16:44 iggy wnkz: orchestrate can do things across different hosts
16:44 f3d it looks like every worker is looping on every minion key
16:45 iggy wnkz: i.e. we use it to do backups... db hosts run a prepare command -> master snapshots volumes -> db hosts end prepare
16:45 tligda1 joined #salt
16:45 timoguin joined #salt
16:46 twellspring joined #salt
16:46 wnkz iggy: oh .. seems cool ; can you chain states and have dependencies ?
16:47 whiteinge Vye: oh all from the minion, I see. Yeah that would be valuable. Hm...
16:47 f3d could someone explain how the work load is spread among the minions ?
16:48 f3d the workers I meant (not the minions)
16:49 iggy wnkz: well, they (like everything in Salt) are run in order... I don't know about dependencies, none of my reactors are that complex
16:50 iggy wnkz: fwiw, your reactors are formatted like normal states, they are more like "run this state file (using the normal process) on these hosts" "then run this other state on these other hosts" etc
16:50 wnkz iggy: well they run in order but asynchronously, aren't they ?
16:51 saffe joined #salt
16:51 iggy async on hosts not between stanzas
16:52 Rawkode joined #salt
16:53 giantlock joined #salt
16:54 wnkz iggy: so if I have multiple states in a reactor ; they will run in order only when the previous state finished ?
16:54 iggy it wouldn't be very useful if it didn't wait for each step to be done before moving on
16:54 iggy (that said, I haven't actually read the source to see if that's how it works)
16:55 iggy that does seem to be the way it works on our orchestrate jobs
16:55 wnkz iggy: well, I'm not currently using orchestrate but I guess it should work the same way
16:56 nickdew joined #salt
16:56 aparsons joined #salt
16:56 jalbretsen joined #salt
16:57 wnkz iggy: on the opposite if you specify multiple reactors to an event (in different files) those states would be run async don't they ?
16:57 iggy I believe
16:58 wnkz iggy: ok, thanks
17:00 aparsons joined #salt
17:01 wnkz iggy: any chance you know a way, with the reactor system, to chain states and get their return value (mostly success/faillure) to take an action, for example a notification or a callback ?
17:01 beneggett joined #salt
17:01 murrdoc is there a variable for the current state dir ?
17:02 murrdoc so that     - source: salt://path/to/conf/ping.conf becomes less tedious
17:02 iggy wnkz: no, you'll find that most salt functions don't return appropriately anyway... there are several bugs open on the topic
17:02 murrdoc slspath only works in init.sls
17:03 otter768 joined #salt
17:03 xliiv joined #salt
17:04 murrdoc https://github.com/saltstack/salt/blob/9446633e9d5e4148ec1c44a8309dbd1c4491042d/salt/utils/templates.py#L63-L72
17:04 murrdoc uses init.list in line 67
17:04 murrdoc any chance we can change that to __file__
17:04 murrdoc or the equivalent
17:04 murrdoc wait nvm
17:07 wnkz iggy: yeah .. to be honest ; I have a few states that manage docker containers and I started working on a small returner to have notifications (or other callbacks) when changes happens ; but the data I had to parse was so inconsistent I almost abandoned :-/
17:08 hobakill any one had much luck setting up rest_cherrypy in 2014.7.0 ?
17:12 nitti joined #salt
17:14 stanchan_ joined #salt
17:18 spookah joined #salt
17:19 hobakill i'm following the basic setup here but can't seem to curl: http://docs.saltstack.com/en/latest/ref/netapi/all/salt.netapi.rest_cherrypy.html
17:19 Ryan_Lane joined #salt
17:20 rickh563 joined #salt
17:22 hobakill basically salt isn't doing anything on port 8000 according to netstat
17:25 theologian joined #salt
17:26 arno joined #salt
17:26 nickdew joined #salt
17:27 twellspring joined #salt
17:29 I3olle joined #salt
17:30 StDiluted joined #salt
17:31 signull_ joined #salt
17:32 Gareth morning morning
17:32 TheoSLC joined #salt
17:35 beneggett joined #salt
17:39 jerematic joined #salt
17:40 nitti joined #salt
17:41 KyleG joined #salt
17:41 KyleG joined #salt
17:43 saffe joined #salt
17:44 saffe_ joined #salt
17:46 Ahlee iggy: where'd you end up booking a room for saltconf?  I'm eyeing the doubletree just around the corner
17:46 booly-yam-7361_ joined #salt
17:47 iggy Peery
17:47 iggy it's only 3.5 stars , but I'm not planning on spending a lot of time there, so...
17:47 Ahlee yeah, exactly
17:49 brendanashworth joined #salt
17:49 iggy it was the cheapest place within walking distance when I looked
17:50 brendanashworth left #salt
17:50 vbabiy joined #salt
17:51 Brew joined #salt
17:53 RedundancyD joined #salt
17:53 stoffell joined #salt
17:54 shaggy_surfer joined #salt
17:55 harkx joined #salt
17:58 moos3 joined #salt
17:59 desposo joined #salt
17:59 vbabiy joined #salt
18:01 saffe joined #salt
18:01 saffe joined #salt
18:01 whiteinge hobakill: if you're using SSL there's a problem with later versions of CherryPy. Try downgrading to 3.2.x.
18:01 stevednd the documentation for restricting mine data eludes me at the moment. can someone  point me to the right page?
18:02 saffe joined #salt
18:02 hobakill whiteinge, yeah i tried using it with and without SSL - no love on either
18:03 basepi iggy: passive aggressive e-mails on salt-users are a really good way to make me happy. ;)
18:03 basepi Anyway, I replied.
18:04 stevednd basepi: what is the syntax, or docs location for restricting what mine data minions can access?
18:04 stevednd or am I imagining that exists
18:04 iggy well, you won't have to deal with me much longer, we're move everything to ZOMGKUBERNETES
18:04 whiteinge hobakill start it in the foreground to see what the error is:  salt-api -l debug
18:04 stevednd whiteinge: is there a way to query a minion's grains from the master? like during the orchestration process?
18:05 basepi stevednd: as far as I know mine data is available to all minions, and does not have access control.
18:05 hobakill whiteinge, maybe this is part of it... i don't have the salt-api installed.... :/
18:05 rlarkin has anyone here managed to fire a custom event with a tag using the helium release?
18:05 whiteinge hehe
18:05 basepi iggy: I was mostly joking, just had to rib you a little. ;)
18:05 basepi iggy: no more salt for you??? =(
18:06 basepi Well, I guess we're in kubernetes, but still
18:06 hobakill whiteinge, i figured it was like halite and just kind of baked in. :/
18:06 whiteinge stevednd: there's the cache runner
18:07 iggy halite isn't baked in... at all
18:07 forrest joined #salt
18:07 whiteinge hobakill: ah, halite was kind of an outlier. salt-api is a separate daemon
18:07 stevednd whiteinge: not familiar with that. Does it work in any way similar to mine.get?
18:07 hobakill whiteinge, ok... i can curl stuff now! thanks man. gosh i feel dumb.
18:07 whiteinge hobakill: the salt-api code ships in salt-core but the init scripts and bin script are in a split package
18:07 _JZ_ joined #salt
18:08 whiteinge stevednd: yes, it's similar: http://docs.saltstack.com/en/latest/ref/runners/all/salt.runners.cache.html#salt.runners.cache.grains
18:08 hobakill iggy, i swear all i did was add the config in the master file and it 'worked'.... not that i use it.
18:09 neogenix joined #salt
18:10 stevednd whiteinge: but how woudl one access that from inside of an orchestration runner, or reactor sls? runners aren't made available inside of the `salt` dict are they?
18:10 stevednd because I think I tried that
18:10 whiteinge reactors have this syntax: `runner.cache.grains`
18:10 racooper is there an option in status.diskuage to return the equivalent of a "df -h"?
18:11 saffe joined #salt
18:11 whiteinge stevednd: orchestrate has http://docs.saltstack.com/en/latest/ref/states/all/salt.states.saltmod.html#salt.states.saltmod.runner
18:14 hobakill is there a place we can make documentation change suggestions?
18:15 whiteinge hobakill: definitely! most of the docs live right in the code
18:15 beneggett joined #salt
18:15 whiteinge hobakill: if you link a doc i'll tell you where to edit them
18:16 hobakill whiteinge, http://docs.saltstack.com/en/latest/ref/netapi/all/salt.netapi.rest_cherrypy.html
18:16 whiteinge hobakill: https://github.com/saltstack/salt/blob/develop/salt/netapi/rest_cherrypy/app.py
18:16 hobakill thanks so much
18:17 hobakill whiteinge,  ^^^^
18:17 seblu joined #salt
18:18 stevednd whiteinge: maybe I'm dumb, but I don't think that will let me use the data for templating, will it? Maybe I should describe what I'm trying to do. salt-cloud creates a minion, the created event fires and I run an orch file to do basic initial setup. At the end of that orch, an event is fired saying that the minion is ready to join any clusters it belongs to. Right now I have this happening as a state call which then uses jinja t
18:18 stevednd emplating to determine the clusters(from grains), and then sends off another series of events saying which clusters to join. This way I don't have to highstate the world and update every known cluster for no reason. I'd like to eliminate that middle step, and could do so if duringn the templating of the reactor sls I could query the grains and generate the proper calls there
18:18 shaggy_surfer joined #salt
18:18 stevednd well that was long
18:18 forrest JEEZ stevednd, make a forum post would you? :P
18:19 stevednd working on a novel
18:19 PovAddict stevednd: at least your IRC client split the message correctly, some people write stuff like that and get their messages truncated by the server :P
18:22 murrdoc forrest:  do u know about the collectd formula
18:22 murrdoc i have an implentation question regarding formulas in general
18:22 murrdoc is the init.sls state supposed to include all the other states as needed
18:22 murrdoc or do i have to include the ones i need
18:22 twellspr_ joined #salt
18:23 rap424 joined #salt
18:23 cpowell joined #salt
18:24 forrest murrdoc: I know of it but probably haven't really looked at it. the init just includes the basics of what you would need to get it running it looks like. So you'd want to include the other chunks of that formula in your top.
18:24 murrdoc hmm
18:24 murrdoc ok but all the other states outside of init include: - collectd at the top level
18:24 stevednd basepi: it looks like `mine_get` can be set in the master config, but it only restricts by mine values. can't filter out specific minions unfortunately
18:25 murrdoc so now if i include a bunch of them i guess salt will handle the duplicates
18:25 forrest murrdoc: Right because for that sls to work, it needs the init and service.
18:25 forrest murrdoc: And yes salt will handle the dups, don't worry about that.
18:26 murrdoc i d rather the init.sls figure out what states it has to include based on the plugins
18:26 forrest murrdoc: just think of it like it's building a huge dict, it doesn't add on to the dict what it already has.
18:26 basepi stevednd: Ya, a feature request may be the way to go.
18:26 murrdoc yeah i get the merging part
18:26 forrest murrdoc: If you wanted to do something based on pillar or something I guess you could do that, you can always just fork and modify as you see fit as well. The formulas are just a starting point so people can get up and running.
18:27 murrdoc and there is no way to do 'if state exists' include
18:28 smcquay joined #salt
18:32 whiteinge stevednd: i would not suggest triggering on the cloud created event. the minion may not yet be available for running commands. i would suggest reacting on the minion start event instead.
18:32 felskrone joined #salt
18:33 murrdoc forrest:  like so https://gist.github.com/anonymous/7add4337835b46a7bec7
18:33 murrdoc now if only there was a include_if_exists:
18:34 whiteinge stevednd: i would then recommend having the reactor kick off an orchestrate script that sets up a minion in stages. base config -> specific config
18:34 aw110f joined #salt
18:34 neogenix joined #salt
18:36 forrest murrdoc: Yeah you could totally do something like that on your fork. I like the idea of include_if_exists. That might be a good feature request if someone didn't already think of it.
18:36 murrdoc my question is more why dont formulas do that already
18:37 Ryan_Lane joined #salt
18:39 neogenix joined #salt
18:39 UtahDave joined #salt
18:40 murrdoc k i ll move on, imho including the init.sls of a formula should setup stuff fully from the pillar
18:40 murrdoc thanks for your time
18:42 ajw0100 joined #salt
18:43 forrest murrdoc: Yeah no worries. I think we've discussed a lot of that stuff before, but it just comes down to ease of use, as well as making the formula complicated for newer users. It's pretty much why I've adopted the stance of people forking stuff, because we can't make everyone happy with the design unfortunately.
18:44 felskrone joined #salt
18:45 stevednd basepi: https://github.com/saltstack/salt/issues/20009
18:45 murrdoc which is cool
18:46 murrdoc thanks again forrest
18:46 murrdoc u aight, despite what iggy says
18:46 forrest murrdoc: haha
18:46 mpanetta joined #salt
18:46 forrest murrdoc: I really do like that include_if_exists though
18:46 forrest if you have time, please see if an issue is open for it, or create one
18:47 glyf joined #salt
18:47 stevednd basepi: thanks for looking at it. I don't know anything about how mine  data is stored. Would it be really messy to try and do something like my latter example with more complex matching?
18:47 SheetiS joined #salt
18:48 stevednd whiteinge: isn't the start event fired every time a minion starts?
18:49 basepi stevednd: Probably not, the master has enough information to stop the minion from accessing stuff it shouldn't. Will just take a substantial rewrite of the mine handling code. *shrugs*
18:50 stevednd also, what you describe with stages is kind of how I'm doing it. It's just when it gets to the part about updating the cluster(s) a minion belongs to I won't have that at the start of the orchestration, since that is inferred from the name as the process goes
18:50 stevednd that's why I fire off an event at the end of that first orchestration to say the minion is ready to join its clusters
18:51 A||SySt3msG0 joined #salt
18:51 felskrone joined #salt
18:52 stevednd I just need a way to get that grain data in the reactor sls instead of having to fire another event on top of it
18:52 Guest96923 joined #salt
18:52 murrdoc https://github.com/saltstack/salt/issues?page=3&amp;q=is%3Aopen+label%3AFeature+state+include doesnt have anything about include a state only if it exists, forrest
18:52 forrest murrdoc: sweet!
18:52 murrdoc i ll try writing it up but i have problems putting my thoughts on paper, its why i write code
18:53 forrest murrdoc: Then use code, show that example you gisted, then show how via the new method it would resolve the error you could possibly receive with invalid pillar data.
18:53 jagardaniel Hi :) Does anyone knows why a git pull through cmd.run takes ~ 38 seconds, but instant in windows CMD? I don't have any issues with git clone. How it looks: http://pastie.org/private/t5lwaxp9h8g5qfa9dsfua
18:54 forrest murrdoc: 3 code examples (misspelling in pillar or something) would provide a good amount of details, then just flesh it out a bit and you're good to go.
18:54 whiteinge stevednd: i would recommend against putting too much logic in a reactor file. that said, if you want grain data within an event so you can use it in a reactor sls file, then you have to put those grains in the event when you fire it. look at the  `event.send` execution function and read the `grains` argument
18:54 jeffspeff I'm often finding that the salt-master is unresponsive. It won't show results from commands sent to host I know are online. I can restart the service and it works again. Here's a line from running top. 3639 root      20   0 2523108 154628   1780 S 100.4  1.9 801:09.54 salt-master
18:56 jeffspeff the CPU percentage often goes above 100% which leads me to believe that it's not actually eating the entire cpu; not sure how an app would use 110% of a cpu. any suggestions on how to fix this?
18:57 saffe joined #salt
18:57 homerding joined #salt
18:57 jeffspeff i have about 300 minions and have configured 30 worker threads on the master.
18:57 stevednd whiteinge: thanks. It's not too much logic, just basically templating out a few state calls to target and run dependent on cluster memberships. I'll look at the event.send docs
18:58 chris-telus joined #salt
18:58 chris-telus good day
18:58 chris-telus $ ps gxww| grep salt 15351 pts/9    S+     0:00 grep salt 26145 ?        S      0:00 /work/infra/salt-2014.7.0/bin/python /work/infra/salt-2014.7.0/bin/salt-master --config-dir=/work/infra/salt-2014.7.0/etc/salt 26171 ?        S     13:47 /work/infra/salt-2014.7.0/bin/python /work/infra/salt-2014.7.0/bin/salt-master --config-dir=/work/infra/salt-2014.7.0/etc/salt 26172 ?        Sl     0:00 /work/infra/salt-2014.7.0/bin/python /wor
18:59 chris-telus Happens when I call /work/infra/salt-2014.7.0/bin/python /work/infra/salt-2014.7.0/bin/salt-master --config-dir=/work/infra/salt-2014.7.0/etc/salt  Used to only start one process
19:01 signull joined #salt
19:01 chris-telus anyone know why this is?
19:03 murrdoc ```saltstack works in markdown, which is cool
19:04 otter768 joined #salt
19:05 echtish joined #salt
19:06 rojem joined #salt
19:08 thedodd joined #salt
19:10 murrdoc forrest:  https://github.com/saltstack/salt/issues/20010
19:12 nickg joined #salt
19:12 nickg is there a way to do a mine.get search based on another mine.get result in a sls or jinja template?  I need to find everyone with a grain role:X based on the grain key_name's value.  So basically everyone who has the same key name as the server with the role:X
19:13 rap424 joined #salt
19:13 rvankleeck joined #salt
19:13 rvankleeck joined #salt
19:16 forrest murrdoc: looks good!
19:16 rvankleeck I have some weird issue with acl and auth going on. I have my group set in external_auth to be able to run anything ('.*') on all servers. When I do 'pkg.install', for example, I get a message that I failed to authenticate.
19:16 rvankleeck However, the package install goes through anyway (as seen in the minion log)
19:16 chupetito joined #salt
19:19 hal58th joined #salt
19:20 TaiSHi joined #salt
19:20 TaiSHi Hi everyone
19:20 TaiSHi It seem some minions are failing to respond to ping, but it wont show me which ones
19:21 TaiSHi Is there any option that might have hidden it? (output_changes i.e.)
19:21 chupetito hi everyone
19:22 chupetito can anyone take a look at this https://gist.github.com/ppchan/778be1b1ca47ba06a58d and tell me what I am doing wrong? its related to order of printed output using unless
19:22 hal58th TaisHi http://docs.saltstack.com/en/latest/faq.html#when-i-run-test-ping-why-don-t-the-minions-that-aren-t-responding-return-anything-returning-false-would-be-helpful
19:22 druonysus joined #salt
19:22 druonysus joined #salt
19:23 * TaiSHi withdraws slowly to a corner
19:23 TaiSHi thanks hal58th
19:24 glyf joined #salt
19:24 hal58th welcome TaiSHi
19:25 chris-telus any reason why the master is generating multiple processes during startup?
19:26 chris-telus ie. /work/infra/salt-2014.7.0/bin/python /work/infra/salt-2014.7.0/bin/salt-master
19:26 TaiSHi chris-telus, worker threads might be one
19:26 TaiSHi Also minions connecting to send mine data
19:27 chris-telus Hi TaiSHi thanks for the quick response :)
19:27 ipmb joined #salt
19:27 TaiSHi I had to increase my mine_interval (was using 2, quite suicidal tbh)
19:27 TaiSHi It was causing my master to be constantly at 1.6 load
19:27 hal58th chris-telus I don't know the specific reason, but master will always spawn multiple processes. Probably worker threads as noted.
19:28 chris-telus ok, noted. so not necessary a bad thing then :) thx TaiSHi
19:28 rvankleeck so i'm wondering: if the packge install goes through anyway, why do i get a 'failed to authenticate' message from the master?
19:28 TaiSHi rvankleeck, did you debug it ?
19:29 dexter91x joined #salt
19:30 dexter91x what is the different between salt.module.cmd and salt.state.cmd ?
19:30 druonysuse joined #salt
19:30 rvankleeck TaiSHi, yeah. and in the minion logs I see the command running/output, etc.
19:30 TaiSHi State is for state files I guess, the module is to run through cli
19:30 lamasnik joined #salt
19:30 twellspring joined #salt
19:30 REdOG joined #salt
19:31 TaiSHi rvankleeck, -l debug from the master, might give you some insight
19:31 TaiSHi Minion clearly is executing the job
19:31 rvankleeck TaiSHi, and then on the master i see it authenticate, and then fail authentication
19:31 dexter91x TaiSHi: well both can be used within SLS or CLI , but is there a main difference ?
19:31 TaiSHi Did you try to salt-call on the minion ?
19:31 TaiSHi dexter91x, not sure, I don't think so tho
19:32 rvankleeck TaiSHi, no, i didn't. just 'salt <minion> pkg.install <package>' on the master
19:32 TaiSHi Go to the minion and execute salt-sall -l debug pkg.install <package>
19:33 shaggy_surfer joined #salt
19:33 StDiluted so has anyone written a capistrano module for salt?
19:33 hal58th chupetito still researching your problem. never used "unless" before
19:34 StDiluted seems like that could be a thing
19:34 chupetito @hal58th, thanks!!
19:36 TaiSHi Autocompletion is seriously slow on salt
19:36 murrdoc tab completion is for the lazy
19:36 rvankleeck TaiSHi, should that be run as root or my user?
19:36 murrdoc i enter my commands in binary
19:37 murrdoc /bofh
19:37 rvankleeck TaiSHi, sort of assuming my user. but i don't have access to any of the minion configs that way (e.g. /etc/salt/minion, /etc/salt/pki/minion, etc.)
19:37 mapu joined #salt
19:38 TaiSHi rvankleeck, root, given salt runs privileged
19:38 rvankleeck TaiSHi, kk
19:38 TaiSHi murrdoc, I have no shame to admit I -am- lazy
19:39 hal58th chupetito ok, well one thing odd about your states. I have never seen "names" and I am not sure you can run multiple cmds in the same state. Maybe you can show me documentation otherwise.
19:39 beneggett joined #salt
19:39 PovAddict I guess I have to make this software packageable and package it before trying to use salt :/ getting salt to build it from source and run its weird installation script would be more of a pain I think
19:39 dexter91x joined #salt
19:40 rvankleeck TaiSHi, salt call from the minion works fine
19:40 saffe joined #salt
19:40 rvankleeck as root
19:40 TaiSHi No weird stuff on the log ?
19:40 StDiluted you definitely cannot run more than one command in a cmd.run
19:41 rvankleeck TaiSHi, nothing aside from 'The `dmidecode` binary is not available on the system. GPU grains will not be available.'
19:41 dexter91x joined #salt
19:41 chupetito @ hal58th... well, i dont know if im doing it right. I just went with names form an example i found online. Do you have a better way to do what I am doing (see the link i sent)
19:41 murrdoc TaiSHi:  i was being facetious
19:41 PovAddict can I get salt to install software from source by running custom commands or something?
19:42 hal58th - name:
19:42 hal58th - echo "#Requirment 123.2" >> /etc/sysctl.conf; echo "net.ipv4.conf.default.send_redirects = 0" >> /etc/sysctl.conf; sysctl -p
19:42 StDiluted chupetito: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html#salt.states.file.append
19:42 StDiluted use the file.append module to put the data into sysctl.conf instead. cmd.run is a bad way of doing things in general
19:42 chupetito hal58th, so basically using ; ???
19:42 StDiluted you can append multiple strings that way also
19:42 hal58th yes, colons, StDiluted also works
19:43 StDiluted semicolons
19:43 StDiluted not colons
19:43 TaiSHi rvankleeck, all I can think of is run master# salt -l debug package.install <package>
19:43 TaiSHi And check both outputs in real time to see if there is anything amiss
19:43 StDiluted but file.append would be the saltier way of doing things
19:43 saltnewB joined #salt
19:43 hal58th yes, semicolons if you want to get all picky and stuff
19:43 TaiSHi It -might- be a connection issue
19:44 StDiluted hahaha, hal58th
19:44 StDiluted :)
19:44 chupetito @ hal58th and StDiluted.... i will take a look at both suggestions and report back. Thakns for the pointers.
19:44 StDiluted linux is picky, I’m not. :)
19:45 CeBe joined #salt
19:45 saltnewB can anyone help point me to why unless isn't working like I think it should in my salt state file? http://pastebin.com/nbmvP7XH My box with the sftponly group and the /data/ directory first applies .nosftp and then the .sftponly versions of the sshd_config file and I can't figure out why
19:45 Diaoul joined #salt
19:49 SheetiS joined #salt
19:49 PovAddict for compiling software from source, would it make sense to make a set of states like 'run configure, unless Makefile exists, depend ln libfoo-dev and libbar-dev pkgs' and 'run make unless the compiled binary exists'?
19:50 rypeck joined #salt
19:50 shaggy_surfer joined #salt
19:50 rvankleeck TaiSHi, logs: https://www.refheap.com/96418
19:50 iggy saltnewB: assuming you're using 2014.7, I'm pretty sure unless/onlyif take shell command to run (at least that's the only way I've ever seen them)
19:50 TaiSHi rvankleeck, sec, gotta deal with a couple minions :P
19:50 rvankleeck TaiSHi, no worries
19:51 rvankleeck i'm still chugging away over here
19:51 rvankleeck or is that 'beating my head against the wall' instead of 'chugging away'?
19:51 hal58th iggy, I think unless works on all states now but I am also not sure.
19:51 TaiSHi Can I restart salt-minion service with other than service.restart ?
19:51 hal58th iggy https://github.com/saltstack/salt/issues/17238
19:52 iggy hal58th: yes, as of 2014.7
19:52 iggy which is I why I said assuming that was in use
19:52 hal58th gotcha
19:52 saltnewB hm... Ok, then I guess I will have to use multiple source lines and use the {% grains['id'] %} to override which file to pull. thanks!
19:53 twellspring joined #salt
19:53 casey|sfe joined #salt
19:53 murrdoc or
19:53 murrdoc make the source a pillar
19:54 murrdoc and then specify it as a pillar based on grains.id
19:54 druonysus joined #salt
19:54 druonysus joined #salt
19:54 saltnewB yeah, was just hoping to have it 'autodetect' the presence of the group and the directory, so I didn't have to remember to do something on the salt server just to apply it
19:54 saltnewB also the grains syntax is cleaner than the pillar.get['blah','blah'] syntax
19:55 beneggett joined #salt
19:55 iggy technically you should probably be using grains.get too
19:55 felskrone joined #salt
19:57 smcquay joined #salt
19:58 rvankleeck TaiSHi, when you get back: I can see the actual install happening if I turn on trace logging on the minion - https://www.refheap.com/96420
19:59 aparsons joined #salt
19:59 TaiSHi Question, it seems you're executing salt command as unprivileged user
19:59 Tyrm_ joined #salt
20:00 TaiSHi Is there a reason for it ?
20:00 TaiSHi (just curious, it might or might not be related)
20:00 rvankleeck TaiSHi, yes. I'm trying to use external_auth to limit salt usage. So that i can give other people the ability to run a specific command without giving them sudo
20:01 Tyrm__ joined #salt
20:01 TyrfingMjolnir joined #salt
20:01 wendall911 joined #salt
20:02 rvankleeck TaiSHi, maybe I have to have both external_auth and client_acl configured?
20:03 TaiSHi I haven't played with external_auth
20:03 rvankleeck ah
20:03 TaiSHi But from what I can see on the logs it might be the reason
20:04 TaiSHi Although it would be pretty insecure if people could execute commands and then authfail
20:05 rvankleeck TaiSHi, yeah, i think i'm going to put in a bug report with the linked log entries
20:05 * rvankleeck smacks forehead.
20:05 TaiSHi Sorry I couldn't be of assistance
20:05 rvankleeck i copied the external_auth to client_acl (with some slight tweaks for proper formatting) and everything works now
20:06 rvankleeck TaiSHi, Thanks for all the help!
20:06 TaiSHi I might have crashed all my minions
20:06 * TaiSHi is proud of himself
20:06 rvankleeck ouch...
20:06 TaiSHi I tried to restart them to apply a change in the config
20:06 TaiSHi It seem they did stop
20:07 TaiSHi rvankleeck, wow that's great
20:07 TaiSHi I will dig on it sometime, might give me tools for the devs
20:07 TaiSHi I don't give them full access anywhere
20:07 rvankleeck smart move
20:07 TaiSHi Only lead dev, but he's awesome
20:08 rvankleeck we have an after-hours support team that I want to give access to only run 'pkg.upgrade'
20:08 rvankleeck which is what led to this. because they are not the most linux savvy group...
20:09 neogenix_ joined #salt
20:09 TaiSHi I'll dig into that after I finish with my icinga2 salt setup
20:10 TaiSHi Also I'm trying to figure out why the master is always at >1.5 load
20:10 PovAddict left #salt
20:10 ajw0100 joined #salt
20:20 rap424 joined #salt
20:20 keeth joined #salt
20:21 TaiSHi My single-core master is handling >70 minions but constantly runs at >1.5 load, any reason why this might be?
20:21 saltnewB joined #salt
20:23 chupetito @ hal58th and StDiluted....I tried both aproches you provided earlier to the issue i was experiencing with unless. both aproaches pretty cool. I think the better one for my use case is file.append. thank you both for your help.
20:23 druonysus joined #salt
20:23 keeth Setting up salt master for the first time, trying to use gitfs with pygit2 and ssh repo access, but getting ‘Unsupported URL protocol' caught while fetching gitfs remote git@github.com:org/repo.git - any ideas?
20:23 druonysus joined #salt
20:23 hal58th welcome chupetito
20:24 jY is there anyway to debug a reactor sls file.. i'm getting failed to render.. but i can run the command just fine on the target.. the yaml for the sls passes a lint test
20:26 saltnewB you can use multiple sources for a file, like using "source: salt://ssh/sshd_config.{{ grains['id'] }}" and then a "source: salt://ssh/fallback.conf", and if the grain's specific file is there it should use it, right? Is there any way to debug why this isn't happening?
20:28 Pixionus joined #salt
20:28 Ahrotahntee joined #salt
20:28 jY saltnewB: pastebin the sls file
20:28 Ahrotahntee is there a way to permanently change the ownership/group of /var/cache/salt/master ?
20:29 Ahrotahntee every time the master restarts it ends up as root:root
20:30 saltnewB jY: http://pastebin.com/nbmvP7XH
20:31 jY Ahrotahntee: do you not run salt-master as root?
20:31 saltnewB I can verify the file exists, I've copy/pasted the grain's value from one box to the other to ensure I didn't mistype it, and it shows up correctly when I 'salt '*' state.highstate' it
20:31 jY saltnewB: wrong syntax
20:32 jY http://pastebin.com/PuxPUb7v
20:33 saltnewB OOOHHH, that makes sense. I hate YAML. Ok, thanks!!
20:34 murrdoc :)
20:34 murrdoc does salt state.show_sls return error codes correctly
20:34 murrdoc if so i recommend using that as write hook in your editor
20:35 murrdoc on write if salt state.show_sls <currentfile> returns 0 .. then keep working, else exclaim
20:36 pdayton joined #salt
20:38 signull Is there a url anywhere that shows the new features/bugs fixed that is going into the 2015.02 release?
20:38 murrdoc good question
20:38 TheoSLC joined #salt
20:38 signull I have the repo checked out and im on the branch but i am unable to find a changelog type file
20:38 murrdoc i d expect to be here https://github.com/saltstack/salt/milestones
20:40 linjan joined #salt
20:40 iggy anybody using 2015.2 (or develop) seen "[ERROR   ] Invalid outputter  specified, fall back to nested"
20:40 signull murrdoc: thats nothing specific to the future stable release
20:40 iggy I don't have any outputter's setup
20:41 murrdoc signull:  nope , i expected it to be there
20:41 murrdoc like i was hoping it was
20:41 signull i see.
20:42 yetAnotherZero joined #salt
20:45 viq joined #salt
20:47 kitplumm_ joined #salt
20:48 kitplumm_ i'm trying to use boto_route53 but i keep getting an saying that it is unavailable.  what am i missing?  boto is installed on both master and minion now.  any ideas?
20:53 rbjorklin Hi, is docker.absent supposed to remove containers or just stop them?
20:53 Ryan_Lane kitplumm_: howdy
20:54 Ryan_Lane kitplumm_: it should only require boto
20:54 kitplumm_ Ryan_Lane: heyo.
20:54 rbjorklin Documentation says "destroy" which I would assume means remove but it doesn't seem to do that
20:54 Ryan_Lane kitplumm_: did you do a restart of the minion daemon?
20:54 Ryan_Lane after you installed boto?
20:54 Ryan_Lane you may be able to call reload modules
20:54 kitplumm_ ah, probably not, because i installed it with salt. :)
20:55 Ryan_Lane kitplumm_: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.saltutil.html#salt.modules.saltutil.refresh_modules
20:55 Ryan_Lane you'll probably need to call that
20:55 Ryan_Lane did you install it via an execution module, or a state?
20:55 Ryan_Lane if you install it via the state, you can pass the reload_modules: True param to the state
20:55 kitplumm_ state.  ah, just ran it with log-level=all and it's getting a DNSServerError
20:55 aparsons joined #salt
20:55 beneggett joined #salt
20:56 Ryan_Lane http://docs.saltstack.com/en/latest/ref/states/#reloading-modules
20:56 Ryan_Lane DNSServerError?
20:56 kitplumm_ in the boto python
20:56 Ryan_Lane hm. what version of boto are you using?
20:56 Ryan_Lane from pip, or from a package?
20:56 Ryan_Lane the package versions are usually ancient
20:56 kitplumm_ 2.35.2 via Pip
20:57 Ryan_Lane what line in boto_route53 is it erroring at?
20:57 kitp_ connection.py:475 - returns 400 Bad Request
20:58 Ryan_Lane are you passing in region/key/keyid? using a profile?
20:58 kitp_ it's just minion config, but yes.
20:59 Ryan_Lane can you show me a sanitized version of what you have defined?
20:59 Ryan_Lane and how you're calling it?
20:59 giantlock joined #salt
20:59 kitp_ yep, let me gist my config.
20:59 Ryan_Lane cool
21:00 Ryan_Lane if you're calling it from within AWS, you should consider using IAM roles. then you don't need to configure or pass anything in :)
21:00 shaggy_surfer1 joined #salt
21:01 kitp https://gist.github.com/kitplummer/340377c1bb995664097f
21:01 signull_ joined #salt
21:01 kitp yeh, i'll get there...just trying to get the complete process to work.
21:02 nettus_ joined #salt
21:04 Ryan_Lane kitp: hm. interesting. that looks like it should be valid
21:04 otter768 joined #salt
21:05 Ryan_Lane kitp: are you sure the value being passed in is a string?
21:06 kitp not sure.  but salt 'rabbitmq*' grains.item external_ip does return the value.
21:07 nickg is there anyway in with ubuntu/debian via ppa to force the salt-minion to 2014.1.13 and not 2014.7.0 ??
21:07 Ryan_Lane kitp: it returns it as a string, or a list?
21:08 kitp ah.  yeh, it's a list.  good catch.
21:08 Ryan_Lane yep. that's your problem :)
21:08 Ryan_Lane of course, it would be sane for that module to support a list
21:08 Ryan_Lane since you can add multiple values to an A record
21:09 kitp ok, silly list...how do i get the first one in the jinja stuff?
21:09 kitp (sorry for the ignorance)
21:09 hal58th nickg, the short answer is no. It seems that even if you had an archive ppa, it still has the latest available.
21:09 Ryan_Lane no worries
21:09 Ryan_Lane kitp: {{ grains.get('elastic_ip')[0] }}
21:09 Ryan_Lane just like python
21:09 signull joined #salt
21:10 kitp ok...i'm getting an error saying that [0] is undefined.  so i'm guess that the grains.get isn't returning anything
21:11 nickg hal58th: what a shame.  for some reason salt thinks most of my minions are down, even though they arent.  would love to upgrade minor versions but just dont have the time right now for a major upgrade
21:11 theologian joined #salt
21:12 saffe joined #salt
21:12 Ryan_Lane kitp: one way to know for sure is this: {{ grains.elastic_ip }}
21:12 rojem joined #salt
21:12 Ryan_Lane it'll throw a jinja error saying the value doesn't exist
21:13 booly-yam-5895 joined #salt
21:14 kitp ah.  i think i have two probs.  elastic_ip doesn't exist yet.  but, i'm getting the same error with external_ip  will try that.
21:14 iggy nickg: try a couple test.ping's in a row
21:15 Ryan_Lane oh. right. boto_route53 /does/ support a list. it just needs to be comma separated.
21:15 iggy that sometimes "wakes" the minions up
21:15 Ryan_Lane comma separated string
21:15 Ryan_Lane this is one of the first modules I added. it could use some love :D
21:16 kitp :) no prob, appreciate the help.
21:16 Ryan_Lane no worries
21:16 kitp so {{ grains.get('external_ip') }} works - its returning a single value.
21:16 Ryan_Lane cool
21:17 Ryan_Lane are you using autoscale groups at all?
21:17 Ryan_Lane or ELBs?
21:17 signull_ joined #salt
21:17 Ryan_Lane I should say ELBs, since it's the one that's relevant.
21:17 kitp nope...this is just a simple single machine.
21:17 Ryan_Lane gotcha
21:17 Ryan_Lane because the boto_elb module can manage cname records on your behalf
21:18 kitp that's good to know.
21:18 Ryan_Lane kitp: it would be great to have an elastic ip module
21:18 kitp so...should I use a CNAME and reference the external DNS instead of an A record pointing to an elastic IP or external IP?
21:18 Ryan_Lane I definitely wouldn't use the external IP
21:19 Ryan_Lane I'd point a cname at the node's public DNS
21:19 kitp yeh, just dev'ing the state stuff now...but will definitely use the elastic IP
21:19 kitp ok, can do that too.  i'll let you know what happens.
21:19 * Ryan_Lane nods
21:19 Ryan_Lane elastic IP is the best bet
21:20 Ryan_Lane mostly you want to avoid changing DNS when a node terminates
21:20 Ryan_Lane since DNS has a long tail of nearly a day
21:20 Ryan_Lane in AWS you always want to do DNS -> ELB or DNS -> elastic IP
21:21 signull joined #salt
21:21 nickg iggy: i've tried that.. i restart the minions.  The minion will authenticate, but the master is just backed up. something strange
21:22 Ryan_Lane kitp: are you using salt-cloud to create your single instances?
21:22 kitp yes.
21:22 kitp it's creating a few other instances too.
21:22 * Ryan_Lane nods
21:23 kitp all the inter-connects are using internal IP stuff which works through mine.get but i need to expose one of 'em.
21:23 * Ryan_Lane nods
21:23 Ryan_Lane I haven't actually used salt-cloud :)
21:24 signull joined #salt
21:24 kitp i'm coming from Puppet and scripts.  so there's some things i like, and some things i don't  ;)
21:24 Ryan_Lane :D
21:28 kitp so...salt thinks it failed to create the record, but it in fact did.
21:28 Ryan_Lane does the record have a * in it?
21:29 kitp no, i don't think so.
21:30 kitp the record worked too.
21:30 Ryan_Lane kitp: can you show me the error message?
21:31 kitp comment:  Failed to add rabbitmq-demo.rbtcloud.com. Route53 record.
21:31 cpowell joined #salt
21:31 Ryan_Lane route53 is a pain. the module will ensure the record was created and replicated
21:31 Ryan_Lane which takes a long time
21:31 Ryan_Lane it's possible it timed out
21:32 kitp ah, ok.  subsequent run is all green.
21:32 Ryan_Lane if you run it in debug log-level it should show you what's going on
21:32 Ryan_Lane if you get a chance to test again, tell me what the failure is
21:32 kitp wilco.  gonna rerun it now.
21:32 Ryan_Lane false failures need to be fixed :)
21:33 kitp yeh, i'll go back to the external_ip in a minute too.
21:33 kitp which i could have the records deleting when the cloud map gets deleted.  :)
21:34 nickg any reason why this would work in a template but not a state file, and how can i fix it?  {{ set seeds = salt['mine.get']('roles:cassandra.seed', 'grains.items', expr_form='grain').items() }}
21:36 Ahlee anybody successfully use service.restart (or win_service.restart)?
21:38 Ahlee sorry, on a windows host.  Important tidbit there.
21:38 Ahlee Comment:   State service.restart found in sls tnwest is unavailable
21:39 kitp Ahlee: if you run with --log-level=all it might give you a bit more info if you're not already
21:40 keeth turns out my ‘Unsupported URL protocol’ error was due to libgit2 not having libssh2 enabled.
21:40 keeth building pygit2 was a real pita
21:40 Ahlee kitp: yeah, I shoved the windows minion up to trace, and it just comes back saying unavialbale
21:41 ajw0100 joined #salt
21:42 iggy if you open up python, import salt.utils and run salt.utils.is_windows() what do you get?
21:45 UtahDave Ahlee: is service.restart actually a state function?
21:45 SheetiS joined #salt
21:47 CeBe joined #salt
21:48 iggy oops, totally missed that
21:48 zeroC left #salt
21:48 GabLeRoux joined #salt
21:49 UtahDave yeah, service.restart is an execution module function not a state module function
21:50 UtahDave Hey everyone, Tom Hatch and I are going to be in Belgium next weekend for FOSDEM in Brussels and then CfgMgmtCamp in Ghent.
21:50 UtahDave https://fosdem.org/2015/
21:50 UtahDave https://www.eventbrite.com/e/cfgmgmtcamp-2015-registration-12899912987
21:51 Ryan_Lane <3 fosdem
21:51 UtahDave If any of you are going to be there we'd love to meet up. We're going to be teaching several classes at CfgMgmtCamp, too
21:51 UtahDave are you going this year, Ryan_Lane?
21:51 Ahlee UtahDave: oh.  well that's annoying
21:51 Ryan_Lane unfortunately not
21:52 UtahDave darn.
21:52 kitp Ryan_Lane: failed during salt-cloud run, because ec2-public_hostname didn't exist.  failed, assuming timeout, during state.highstate salt run.  running it again, is all green with CNAME record working.
21:52 UtahDave Ahlee: usually you have a file.managed or something watched by the service.running
21:52 schristensen joined #salt
21:53 Ahlee UtahDave: right
21:53 Ryan_Lane kitp: weird. I use this extensively and never get failures
21:53 UtahDave Ahlee: you could use   service.running with the module.run
21:53 Ryan_Lane I guess I never add A record
21:53 Ahlee it's not usual you want to just restart a service outta the blue.
21:53 Ahlee UtahDave: good call, i always forget about module.run
21:54 kitp Ryan_Lane: i just ran it with the CNAME.  haven't gotten back to the A record yet.
21:54 Ryan_Lane kitp: can you look at your minion logs to see what the errors were?
21:55 kitp sure.
21:55 Ryan_Lane it should show you stuff like 'attempt 1' 'attempt 2', etc.
21:55 glyf joined #salt
21:56 keeth i think finally have gitfs working - is there any way to dump/print the salt-master states filesystem as a minion would see it?  want to make sure it looks right.
22:01 spookah joined #salt
22:01 iggy Ahlee: it is in states... that's what watch/listen/etc are for
22:02 iggy keeth: cp.list_master
22:02 keeth iggy: thanks
22:03 beneggett joined #salt
22:07 kitp Ryan_Lane: definitely timing out.  "Timed out waiting for Route53 status update."
22:07 Ryan_Lane yep. weird
22:07 Ryan_Lane how many attempts?
22:07 kitp 10
22:07 Ryan_Lane kitp: please open an issue in github for this and add @ryan-lane into the description
22:08 Ryan_Lane it's really not necessary for the module to wait for replication
22:08 kitp which repo?
22:08 Ryan_Lane it if gets a success from boto, it should just assume it worked
22:08 kitp i agree.
22:08 Ryan_Lane there should be an option to validate, and a timeout option too
22:08 Ryan_Lane kitp: saltstack/salt
22:09 Brew joined #salt
22:12 TheoSLC joined #salt
22:13 JDiPierro joined #salt
22:20 ProT-0-TypE joined #salt
22:25 rojem joined #salt
22:25 jonatas_oliveira joined #salt
22:27 dstokes would anyone be interested on a write up about testing pillar / topfile / highstate config with docker?
22:28 murrdoc ooh with docker ?
22:28 murrdoc i havent seen that
22:28 murrdoc kitchen-salt does vagrants
22:28 murrdoc i would totally be interested dstokes
22:28 dstokes vagrant is ok, but if you want to test in parallel..
22:29 murrdoc yeah , write it up
22:29 dstokes murrdoc: just got it working today so need to do some cleanup, but i'll start on something next week
22:30 mapu joined #salt
22:30 murrdoc ./me subscribes
22:35 kermit joined #salt
22:37 UtahDave dstokes: Yeah, I'd totally be interested in reading about that!
22:38 dstokes :D
22:41 beneggett joined #salt
22:41 bhosmer_ joined #salt
22:46 murrdoc u should present at the conf
22:47 murrdoc not sure if the call for speakers is still on
22:47 Vye dstokes, send a link out to salt-users when it is up. I also would like to read what you've got.
22:48 dstokes that a mailing list?
22:48 Vye yes: https://groups.google.com/forum/#!forum/salt-users
22:49 dstokes will do
22:50 Vye thnx!
22:52 ipmb joined #salt
22:53 theologian joined #salt
23:00 doobi-sham-46893 joined #salt
23:03 lothiraldan joined #salt
23:05 otter768 joined #salt
23:06 smcquay joined #salt
23:06 smcquay joined #salt
23:08 shaggy_surfer joined #salt
23:09 kitp joined #salt
23:09 CeBe joined #salt
23:09 nethershaw joined #salt
23:11 Nazca__ joined #salt
23:12 ckao joined #salt
23:16 glyf joined #salt
23:17 Nazzy joined #salt
23:19 KyleG joined #salt
23:19 KyleG joined #salt
23:23 dstokes looking pretty good.. https://s3.amazonaws.com/f.cl.ly/items/162u3I3A1c2H1P1M0r3Q/Screen%20Shot%202015-01-23%20at%203.22.20%20PM.png
23:40 I3olle joined #salt
23:40 linjan joined #salt
23:46 jerematic joined #salt
23:53 jY dstokes: i'd be interesting in seeing that.. i was just about to re-write kitchen-salt to make it more robust
23:55 spookah joined #salt
23:58 theologian joined #salt

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