Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-06-02

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

All times shown according to UTC.

Time Nick Message
00:00 murrdoc joined #salt
00:02 markm joined #salt
00:11 phpdave11 joined #salt
00:17 murrdoc joined #salt
00:22 sgargan joined #salt
00:27 sunkist joined #salt
00:28 timoguin joined #salt
00:29 patrek joined #salt
00:30 ipmb joined #salt
00:32 matthew-parlette joined #salt
00:36 matthew-parlette joined #salt
00:41 baweaver joined #salt
00:50 iggy murrdoc: I like your idea of splitting th ng formulas out (php-overlycomplex-formula maybe?)
00:51 murrdoc the guy who wants to keep it in place is in the chan
00:51 murrdoc i dont want to argue
00:51 iggy nah, I don't think he wants that, he just thinks it'll be more confusing to split
00:51 iggy we just need to think of a non-confusing way to split
00:52 iggy saltstack-advanced-formulas/php-formula ?
00:52 murrdoc or make php-simple
00:52 murrdoc and nginx-simple
00:52 murrdoc and keep the 'smarter' ones in nginx
00:52 murrdoc and php
00:52 iggy possibly
00:53 iggy if you hit saltstack-formulas and then search for nginx, that's clear enough
00:53 murrdoc yeah
00:53 murrdoc nginx shows up as the main one
00:53 iggy if you hit google and search for salt nginx formula and the "ng" version comes up first
00:53 murrdoc which is fine
00:53 iggy a not at the top of the ng version that says"
00:54 murrdoc simple
00:54 cberndt joined #salt
00:54 iggy "This is for advanced uses, please see non-ng version"
00:54 murrdoc sure
00:58 timoguin joined #salt
00:59 mrbigglesworth joined #salt
00:59 notnotpeter joined #salt
01:05 hal58th joined #salt
01:05 hal58th_ joined #salt
01:05 hal58th__ joined #salt
01:30 david_an111 joined #salt
01:33 ajw0100 joined #salt
01:35 mrbigglesworth joined #salt
01:35 schuckles joined #salt
01:36 amcorreia joined #salt
01:39 antonw joined #salt
01:39 davedash joined #salt
01:39 akafred joined #salt
01:39 LinuxHorn joined #salt
01:39 joeyparsons joined #salt
01:39 m0nky joined #salt
01:39 akoumjian joined #salt
01:39 munhitsu joined #salt
01:39 natewalck joined #salt
01:40 Striki joined #salt
01:40 frankS2 joined #salt
01:40 copelco joined #salt
01:40 grepory joined #salt
01:41 neilfse joined #salt
01:41 hillna joined #salt
01:41 basepi joined #salt
01:41 lkannan joined #salt
01:42 juanito hey guys, if a file exist on a minion and we want to copy to an other destination can we use file.managed for that
01:45 juanito or we have to use file.copy ?
01:45 gazarsgo joined #salt
01:45 OliverMT joined #salt
01:45 simonmcc joined #salt
01:45 SteamWells joined #salt
01:45 collinanderson joined #salt
01:45 otter768 joined #salt
01:45 CaptTofu joined #salt
01:45 Tritlo joined #salt
01:47 mikepea joined #salt
01:47 goki joined #salt
01:47 abele joined #salt
01:48 akitada joined #salt
01:50 iggy pretty sure file.managed only takes non-local uri's
01:50 iggy juanito: but you could try file:///path/to/local.file
01:51 doriftoshoes joined #salt
01:52 dendazen joined #salt
01:55 imanc joined #salt
02:00 hasues joined #salt
02:00 hasues left #salt
02:14 murphyslawbbs joined #salt
02:15 mrbigglesworth joined #salt
02:19 murphyslawbbs joined #salt
02:23 subsignal joined #salt
02:27 murrdoc joined #salt
02:36 al joined #salt
02:40 cberndt joined #salt
02:41 ageorgop joined #salt
02:53 writtenoff joined #salt
03:03 amcorreia joined #salt
03:07 mrbigglesworth joined #salt
03:09 murrdoc joined #salt
03:11 Brew joined #salt
03:12 Aidin joined #salt
03:12 bin_005_u_j joined #salt
03:14 ageorgop joined #salt
03:15 mrbigglesworth joined #salt
03:28 ageorgop1 joined #salt
03:29 juanito ta iggy , will give it a shot
03:31 syfqmzli joined #salt
03:35 murrdoc so basically nested pillars suck
03:35 murrdoc for -C matching
03:35 CryptoMer ^^
03:35 murrdoc salt -C "I@'a:b':osprvn" test.ping
03:35 murrdoc dont work
03:35 murrdoc thanks obama
03:36 CryptoMer take out the single quotes
03:36 CryptoMer assuming osprvn is your value and a:b are the keys
03:36 CryptoMer you shouldn't need them.
03:37 CryptoMer quick example, seeing as I'm testing beacons in pillar, it's easily tested
03:37 CryptoMer salt -C 'I@beacons:service:apache2:* and FS*' test.ping
03:37 CryptoMer and it returns. ;)
03:39 hybridpollo left #salt
03:39 zer0def joined #salt
03:40 ageorgop joined #salt
03:43 favadi joined #salt
03:45 otter768 joined #salt
03:47 timoguin joined #salt
03:55 p66kumar joined #salt
04:00 mosen joined #salt
04:01 cztanu joined #salt
04:08 mrbigglesworth joined #salt
04:13 spookah joined #salt
04:14 ajw0100 joined #salt
04:20 rocket joined #salt
04:22 rocket anyone else see this error for a salt-call invokation?  NameError: global name '__salt__' is not defined
04:22 rocket I am on freebsd 9  .. with 15.1.0 installed on the client ..
04:23 iggy paste more of the error message
04:25 mrbigglesworth joined #salt
04:26 rocket iggy I am going to have to recreate the issue... I downgraded to 2014.7 as I am in a time crunch
04:28 soren joined #salt
04:29 mosen hiya iggy
04:33 evle1 joined #salt
04:35 mrbigglesworth joined #salt
04:36 iggy yo
04:36 soren joined #salt
04:48 ramteid joined #salt
04:49 Singularo joined #salt
04:55 mrbigglesworth joined #salt
05:05 c10 joined #salt
05:06 c10_ joined #salt
05:07 favadi joined #salt
05:08 catpig joined #salt
05:08 joeto joined #salt
05:16 Aidin joined #salt
05:19 evle joined #salt
05:22 saifi joined #salt
05:33 oravirt joined #salt
05:35 pjs joined #salt
05:45 rdas joined #salt
05:45 otter768 joined #salt
05:53 AndreasLutro joined #salt
05:54 t7243Gues joined #salt
05:59 colttt joined #salt
06:02 stephanbuys joined #salt
06:07 rdas joined #salt
06:14 flyboy joined #salt
06:19 ndrei joined #salt
06:43 thalleralexander joined #salt
06:48 impi joined #salt
06:53 Auroch joined #salt
06:56 KermitTheFragger joined #salt
06:58 arount joined #salt
07:01 j-saturne joined #salt
07:04 soren joined #salt
07:06 tmclaugh[work] joined #salt
07:07 stephanbuys1 joined #salt
07:07 dariusjs joined #salt
07:14 eseyman joined #salt
07:16 elektrix joined #salt
07:16 favadi joined #salt
07:20 arount Hi there, not sure lot of people are connected as this moment but:
07:20 arount There is a method to access to Pillar data from Reactor ?
07:23 subsignal joined #salt
07:25 evle joined #salt
07:26 david_an11 joined #salt
07:26 bash124512 joined #salt
07:34 markm joined #salt
07:46 otter768 joined #salt
07:51 babilen arount: It is in the execution modules you run on minions
07:52 yuhl_work___ joined #salt
07:56 linjan joined #salt
07:57 arount babilen: ok, so I have to develop my own module, I can't have a /srv/reactor/<name>.sls within something like "{{ pillar['foo'] }}" ?
07:58 slav0nic joined #salt
08:01 evle1 joined #salt
08:06 s_kunk joined #salt
08:06 s_kunk joined #salt
08:09 babilen arount: What do you mean by "within" ?
08:10 arount babilen: *with in this file* something like
08:10 julez joined #salt
08:11 linjan joined #salt
08:12 babilen arount: No, pillars are specific to one minion. Why would you want that?
08:12 supersheep joined #salt
08:13 babilen In the context of reactors you typically pass information around in form of the "data" dictionary
08:13 c10 joined #salt
08:14 arount babilen: I think I have to read doc one more time, I surely in a wrong way, I will do that before bored you with inaccurate questions
08:15 arount babilen: thx anyway
08:22 bluenemo joined #salt
08:25 cberndt joined #salt
08:27 babilen I don't mind, it is just that you ask "XY problem" questions in that you don't ask how to solve what you are actually trying to solve, but about one specific step in what you think is the solution to that problem
08:28 babilen http://mywiki.wooledge.org/XyProblem
08:29 c10 joined #salt
08:29 Pulp joined #salt
08:32 julez joined #salt
08:34 Grokzen joined #salt
08:42 johngrasty joined #salt
08:43 ndrei joined #salt
08:44 arount babilen: would like to configure a Nagios server (as a minion) when a new minion (application server) is mounted
08:44 linjan joined #salt
08:45 arount babilen: So I have X app servers ups, I create a new one (salt highstate) and I would like to add one config file (representing the new server) in the nagios minion
08:46 arount babilen: obviously the config file for nagios have to be computed depending several data from the minion
08:47 babilen sure
08:47 babilen You will want to use the salt mine for maintaining the information about checks that are applicable for each minion
08:48 arount Salt can be puzzling (I don't know if this word really exists, google translate said yes) sometimes, they are many many ways to do the same things and there are not real best practice to follow
08:48 babilen That word does exist and it is fitting :)
08:48 arount babilen: :)
08:48 favadi joined #salt
08:49 babilen I am aware of people who do have such a Nagios infrastructure in place, but the name escapes me right now. I'd write to the mailing list and ask about it as they are surely willing to sketch their approach.
08:49 TheHelmsMan joined #salt
08:50 pchero_work joined #salt
08:50 viq joined #salt
08:50 viq joined #salt
08:50 arount babilen: Yes I will be interested by that
08:51 N-Mi joined #salt
08:53 kbyrne joined #salt
09:02 impi joined #salt
09:03 soren joined #salt
09:04 sgargan joined #salt
09:08 ProT-0-TypE joined #salt
09:20 supersheep joined #salt
09:25 bin_005 joined #salt
09:29 c10 joined #salt
09:32 N-Mi joined #salt
09:34 cberndt joined #salt
09:40 stephanbuys joined #salt
09:41 saifi joined #salt
09:42 Twiglet Anyone know if you're able to use an external source for minion keys (in addition to filesystem based ones)
09:44 TheHelmsMan joined #salt
09:47 otter768 joined #salt
09:51 bin_005 joined #salt
09:56 a7p joined #salt
09:56 glyf joined #salt
09:56 linjan joined #salt
10:01 fredvd joined #salt
10:06 julez joined #salt
10:07 julez joined #salt
10:10 sgargan joined #salt
10:18 babilen Twiglet: You can put them in any external pillar listed on: http://docs.saltstack.com/en/latest/ref/pillar/all/
10:27 Twiglet babilen: excellent, was poking about but didn't think to chekc in ext pillar, cheers
10:32 amcorreia joined #salt
10:40 sgargan joined #salt
10:41 Auroch joined #salt
10:43 quist``` joined #salt
10:45 permalac joined #salt
10:46 matthew-parlette joined #salt
10:47 giantlock joined #salt
10:47 Auroch joined #salt
11:00 scottpgallagher joined #salt
11:00 mage_ is it possible to get salt messages in real time and not 1000+ lines at once ?
11:05 permalac joined #salt
11:05 ndrei joined #salt
11:05 ksj I'm trying to tinker with the python in one of the modules. I know almost no python, but I just want to try this quick fix. I've changed the .py file on both the master and minion and tried restarting both, but there's no change. Do I have to do something to compile them to .pyc files? any help appreciated
11:08 babilen mage_: Look into eventlisten
11:08 babilen But not really
11:13 N-Mi joined #salt
11:15 mr_daniel joined #salt
11:17 mejiamariano joined #salt
11:17 CeBe joined #salt
11:19 sunkist joined #salt
11:25 MohShami joined #salt
11:28 MohShami hi guys, I'm running salt on freebsd 9.2, 10.0 and 10.1, which is working like a charm, but sometimes when salt auto restarts a service the command just stalls, like dovecot for example, the debug shows Executing command '/usr/sbin/service dovecot onerestart' in directory '/root', and nothing after
11:28 MohShami if I log in to the server and restart manually the execution continues without issues, any idea what this could be?
11:28 TyrfingMjolnir joined #salt
11:32 michelangelo joined #salt
11:43 dariusjs MohShami: are you using salt *  service.dovecot restart ?
11:43 ProT-0-TypE joined #salt
11:43 babilen * service.restart dovecot
11:45 julez joined #salt
11:47 schuckles joined #salt
11:48 otter768 joined #salt
11:49 MohShami babilen, dariusjs, thanks guys, I'm using salt-call state.highstate, and am restarting the service on configuration file change
11:51 dariusjs ok so youve got a watch config file then if changes detected restart it? must say i dont have freebsd boxes atm to work on so dont know
11:51 dariusjs MohShami: if you do a salt * cmd.run 'service dovecot restart' is it fine
11:51 N-Mi joined #salt
11:52 MohShami dariusjs, let me check
11:52 babilen onerestart ?
11:52 MohShami onerestart is the same as restart, but doesn't require the service to be enabled in /etc/rc.conf
11:53 babilen okay, does "restart manually" mean that you run "/usr/sbin/service dovecot onerestart" manually?
11:54 rdas joined #salt
11:54 MohShami babilen, sadly it's the same thing
11:54 MohShami babilen, yep
11:54 babilen What is the same thing?
11:54 MohShami and if I do that, the state.highstate continues execution normally
11:54 MohShami salt * cmd.run 'service dovecot restart'
11:55 babilen Ah, dariusjs asked about that
11:55 ingwaem joined #salt
11:55 MohShami oh, sorry for the confusion
11:55 ingwaem left #salt
11:56 babilen So you cannot run "salt 'theminion' service.restart dovecot" (which blocks), but executing "/usr/sbin/service dovecot onerestart" directly on the minion works. What about "salt 'theminion' cmd.run "/usr/sbin/service dovecot onerestart"" ?
11:58 MohShami blocks as well
11:58 MohShami with -l debug, I get a loop with 'Checking whether jid 20150602145744360235 is still running'
11:58 MohShami now if I go to the minion and restart the service manually the master breaks out the loop and finishes execution
12:00 j-saturne joined #salt
12:02 stephanbuys joined #salt
12:03 VSpike Does anyone know if the caveat about Deb/Ubuntu systems at http://docs.saltstack.com/en/latest/ref/states/all/salt.states.network.html#salt.states.network.system still holds true?
12:03 sgargan joined #salt
12:06 slav0nic if i not set sudo:True in salt-ssh conf pkg.installed runned as non'root user, as fix this?
12:06 MohShami babilen, dariusjs, the same thing happens with amavisd-milter
12:07 jrdnr joined #salt
12:08 schuckles1 joined #salt
12:09 ponpanderer joined #salt
12:11 jrdnr joined #salt
12:11 glyf joined #salt
12:12 MohShami server irc.irchighway.net
12:12 babilen MohShami: I've seen behaviour like that when the service didn't daemonise properly (in particular: "didn't close all file descriptors and didn't double fork")
12:12 MohShami I see
12:13 MohShami so the init script needs to be rewritten, right?
12:13 babilen It might be some unfortunate interplay between the way salt determines if the service was restarted (i.e. all FDs are closed) and the init system / scripts running
12:14 lempa joined #salt
12:14 babilen If you check the logs of the service in question: Is it actually restarted?
12:15 MohShami let me check
12:19 MohShami babilen, dovecot shows it was restarted
12:19 MohShami I see a stop and start log entries
12:21 Auroch joined #salt
12:22 babilen MohShami: Okay, so it must be something that is specific to the init script. I take it that there are services you can actually restart, aren't there?
12:23 MohShami babilen, sorry I don't understand the question
12:24 babilen Can you restart other services apart from dovecot and amavisd-milter ?
12:26 MohShami yep everything else is working well
12:26 babilen Okay, could you paste the dovecot init script and an example of another, working, one to http://refheap.com ?
12:26 MohShami sure, one moment
12:27 lloesche joined #salt
12:27 babilen I'm not sure if I'll be able to spot the problem, but I guess it won't hurt to take a look. Actually ... it might be good if you could paste both the dovecot and the amavisd-milter one
12:30 MohShami sure
12:30 MohShami here is dovecot https://www.refheap.com/101997
12:30 MohShami here is clamav, which is working https://www.refheap.com/101998
12:30 MohShami let me get you the one for amavisd-milter
12:30 DammitJim joined #salt
12:31 spiksius joined #salt
12:31 MohShami this is the one for amavisd-milter
12:31 MohShami https://www.refheap.com/101999
12:31 MohShami thanks a milion babilen :)
12:33 julez joined #salt
12:33 N-Mi joined #salt
12:37 babilen MohShami: Could you, just for testing purposes, remove the "start_postcmd="amavisd_milter_postcmd"" from the amavisd-milter init script and try restarting it again?
12:37 MohShami sure
12:38 MohShami babilen, nope, didn't work
12:40 babilen pity. Just noticed that both failing ones used a postcmd and had hoped that they are the ones that are blocking
12:41 MohShami weird
12:44 dkrae joined #salt
12:44 babilen MohShami: Could you try "salt 'theminion' cmd.retcode "/usr/sbin/service amavisd-milter restart" python_shell=True" please ?
12:44 MohShami sure
12:45 snaggleb joined #salt
12:45 babilen (after having reverted the init script to its original form)
12:45 babilen Which salt version is that?
12:45 dendazen joined #salt
12:46 MohShami same thing, "Checking whether jid 20150602154609147458 is still running"
12:47 babilen Okay, that is the actual salt function that is being called.
12:48 babilen So that command doesn't return?
12:48 MohShami Salt: 2014.7.0
12:48 MohShami seems so
12:48 babilen If you were to run the minion in debug mode (stop it and then run "salt-minion -ldebug") ... do you get anything interesting in its output?
12:49 babilen Can you reproduce that issue with 2015.5.0 ?
12:49 morsik joined #salt
12:49 morsik hi
12:50 morsik it's documented behaviour that using `salt-call --return myreturner` is the same like putting 'return: myreturner' inside /etc/salt/minion with running `salt-call`?
12:50 morsik it works for me for now… but I want to be sure i'll work later after updates too
12:50 morsik i'm unable to find any docs about 'return: ', or about that I can put command-line args inside minion config to force them
12:51 MohShami babilen, blocks on "[INFO    ] Executing command '/usr/sbin/service amavisd-milter restart' in directory '/root'"
12:51 MohShami babilen, let me upgrade and check, 1 moment
12:52 babilen ta
12:54 rocket joined #salt
12:54 giantlock joined #salt
12:57 alpiv joined #salt
12:57 alpiv good morning all...
12:58 alpiv I'm doing research on Hadoop and Azure and Salt and i figured I should ask the experts...
12:58 alpiv Are they cluster sizing (number of nodes) limits in Azure? or as any one run into that before
12:59 emaninpa joined #salt
12:59 wplatnick joined #salt
12:59 MohShami babilen, same thing
12:59 MohShami or wait, that ws 2014.7.0
12:59 Tecnico1931 joined #salt
13:01 pressureman joined #salt
13:01 jeremyr joined #salt
13:01 N-Mi joined #salt
13:03 alpiv joined #salt
13:03 alpiv sorry lost connection
13:05 rocket wow 2015.7.1 is really broken on freebsd .. :/
13:06 rocket err 2015.5.1
13:07 MohShami babilen, just checked, 2015.5.1, same thing
13:08 fusionx86 joined #salt
13:09 alpiv subject
13:09 MohShami rocket, what's wrong with it? I just upgraded :(
13:10 rocket MohShami: try doing salt-call state.highstate ...
13:10 rocket it blows up spectacularly
13:11 rocket at least on my freebsd 9 system
13:11 pythopticon joined #salt
13:11 MohShami just ran it on 9.2, works well
13:11 toastedpenguin joined #salt
13:12 rocket MohShami: really?  how did you install it?
13:12 MohShami I maintain my own poudrier repo, and installed it from there with pkg
13:12 LotR isn't the idea that you upgrade your test environment first, to catch things blowing up? :)
13:12 rocket LotR: I am in my test env
13:12 MohShami I don't always test my code, but when I do, I test it on my production server :D
13:12 rocket LotR: I never said I was in production
13:12 pfallenop joined #salt
13:13 babilen MohShami: Not sure, I';d recommend failing a butg at this point
13:13 babilen *filing a bug
13:13 MohShami babilen, thanks a million mate :)
13:13 rocket MohShami: are you using gitfs or anything like that?
13:14 dyasny joined #salt
13:14 MohShami rocket, nope, I used to, it's just one more thing that would break
13:14 rocket yeah I am finding that as well ... :/
13:14 MohShami when I push updates to my git repo I have a hook that just extracts everything to /usr/local/etc/salt
13:14 MohShami if you want the hook and script I'd be happy to share
13:15 rocket sure :)
13:16 toastedpenguin joined #salt
13:16 MohShami ok 1 moment :)
13:19 rocket at least you have given me a clue.  I bet the gitfs code broke it somehow .. since I am turning that on :/
13:20 MohShami it could be
13:20 MohShami this is the post-receive hook https://www.refheap.com/102001
13:20 MohShami rocket, and this is the script used to copy changes https://www.refheap.com/102002
13:20 TooLmaN joined #salt
13:21 rocket thanks MohShami
13:21 MohShami any time :)
13:22 rocket this project is taking longer than I realized it would... I have a client that has an ancient 8.x server install from a previous sysadmin  and very little is documented
13:23 MohShami I know the feeling, I have the exact same situation albiet with linux :'(
13:23 toastedpenguin joined #salt
13:23 rocket everything was compiled from ports, but none of the options were recorded and the config files were edited in place
13:24 MohShami that seriously sucks
13:24 rocket so I have gone through and figured out exactly what knobs to turn and install that differ from the defaults and documented that all, and/or setup clean config files that just reference the differences via a nice include structure etc
13:24 MohShami good luck man
13:24 MohShami nice
13:24 rocket but it all blew up last night when I tried to refresh my vm and salt 2015 blew up hard .. :/
13:24 MohShami ouch
13:25 rocket havent had time to get a poudriere env to get a solid pkg base :/
13:25 rocket thats next on the list
13:27 MohShami it's simple once you get the hang of it
13:28 giantlock joined #salt
13:28 rocket MohShami: it looks that way, but its just a matter of time at this point
13:28 MohShami of course
13:29 MohShami good luck with that my friend :)
13:31 N-Mi joined #salt
13:34 subsignal joined #salt
13:34 yaryarrr joined #salt
13:37 JoeHazzers joined #salt
13:37 hasues joined #salt
13:37 mpanetta joined #salt
13:37 hasues left #salt
13:42 MohShami rocket, the good thing about using a setup like mine is that you can test small changes without having to put them in git
13:45 julienlavergne1 joined #salt
13:45 julienlavergne1 Hi guys
13:46 bhosmer joined #salt
13:46 julienlavergne1 tell me how works autosign.conf for salt-master ? we can't provide minion pub key into the master ?
13:47 cpowell joined #salt
13:47 julienlavergne1 autosign seems to accept minions when we just precise the id ??
13:47 riftman joined #salt
13:48 timoguin joined #salt
13:49 otter768 joined #salt
13:53 N-Mi joined #salt
13:55 jdesilet joined #salt
13:59 marcinkuzminski joined #salt
14:00 _JZ_ joined #salt
14:01 TyrfingMjolnir joined #salt
14:01 peters-tx joined #salt
14:04 linjan joined #salt
14:08 debian112 joined #salt
14:11 murrdoc joined #salt
14:14 soren joined #salt
14:17 pfallenop joined #salt
14:20 soren joined #salt
14:21 julez_ joined #salt
14:30 rm_jorge joined #salt
14:31 dharper_ii does anyone know how to get all minions that returned false/not connect when you do a local.cmd from the API. I have tried something like: local.cmd('*', ''test.ping", kwargs=('-v') )
14:31 dharper_ii but that is not returning the failed clients like one would get from the CLI
14:33 ksj is it possible to force one "unit" (i.e. id and state) to run before everything else, without using a requiste statement in every one of the other "units". (I wish there was a better term for the "blocks", but it's not an id, it's not a state, it's not a declaration)
14:33 riftman joined #salt
14:34 ksj reason being, one of my "blocks" seems to be executing out of order....or perhaps the service just isn't starting up in time. either way, I have to run the highstate twice. on the first run it fails. second run it works
14:36 catpig joined #salt
14:41 AirOnSkin joined #salt
14:44 sporkd2 joined #salt
14:47 dimeshake joined #salt
14:49 supersheep joined #salt
14:51 fredvd joined #salt
14:55 clintberry joined #salt
14:59 SheetiS joined #salt
15:02 supersheep joined #salt
15:04 huddy joined #salt
15:07 rocket how does __salt__ get imported?
15:07 conan_the_destro joined #salt
15:09 riftman joined #salt
15:12 Eureka_ joined #salt
15:13 bhosmer joined #salt
15:13 Gareth basepi recently gave a talk somewhere on Salt and the dunder variables.
15:16 giantlock joined #salt
15:18 teebes joined #salt
15:19 simmel joined #salt
15:21 andrew_v joined #salt
15:22 j-saturne joined #salt
15:23 Brew joined #salt
15:23 Twiglet ksj: - order: first?
15:23 rocket the reason I ask, in freebsd 9.3 salt 2015.2.1 I am getting a traceback because __salt__ isnt found
15:23 Twiglet i think there is an order thing anyway
15:24 dfinn joined #salt
15:26 schuckles joined #salt
15:28 iggy rocket: the module loader loads it
15:28 iggy it's usually not available in __virtual__ calls (which is where most errors occur)
15:28 seev joined #salt
15:29 rocket iggy: http://pastebin.com/5SNMVTUz
15:30 goal joined #salt
15:31 iggy I'd  guess you have some egregious error in whatever config you have
15:32 goal the docs say autosign_timeout can be used to limit the period during which minions can be auto-accepted, but when I try this the default value is still applied. Looking through the code it actually has "expire_minutes = self.opts.get('autosign_expire_minutes'" which suggests the option should be actually autosign_expore_minutes. However, this is not a valid option. Anyone have a clue?
15:32 j-saturne joined #salt
15:32 rocket it worked just fine in 2014.7.5 .. :/
15:33 msheiny joined #salt
15:33 iggy rocket: well, if you file a bug include the config
15:33 iggy there's not a general problem with 2015.5 and config dirs or --local, so it's something specific to your setup
15:34 froztbyte mm
15:34 froztbyte grains[], in an SLS, are direct for that minion, right?
15:35 iggy yes
15:35 froztbyte I have a thing where I define a kind of role grain across the board for a given nodetype, and I want to "collect" all of that in an SLS
15:35 froztbyte is that a thing I get to do?
15:35 iggy mine
15:36 msheiny is anyone here using salt-cloud with vmware? i'm trying to understand some high-level operations basics like how the bootstrapping process will work
15:36 froztbyte ah. was reading that, was unclear about the module references
15:36 froztbyte ta iggy
15:36 epcim_ joined #salt
15:37 ksj Twiglet: thanks - hadn't seen that
15:37 impi joined #salt
15:39 SEJeff joined #salt
15:41 iggy froztbyte: you can put in your mine_functions: - network.get_hostname
15:41 litwol hmm. is there a documented "rule of thumb" i can read which discusses good practices when it comes to choosing between implementing minimalist solution for personal infra/system, and between designing big and all-inclusive in "formula" format?
15:41 gladiatr joined #salt
15:41 froztbyte iggy: yeah I think I have it figured out, thanks
15:42 iggy froztbyte: then use something like: {% set repohost = salt['mine.get']('tags:repo', 'network.get_hostname', 'grain').values() %}
15:42 litwol So far i've found myself very annoyed with the approach of defaults+map for very small state formulas
15:42 iggy litwol: personal preference
15:42 froztbyte litwol: I personally change things at the third SLS growth/iteration
15:42 iggy you'll notice not all formulas use that... (aptly-formula for instance is quite simple)
15:43 SEJeff left #salt
15:43 iggy some people don't do anything like that at all... they open code everything, even when it causes some duplication
15:44 iggy (see Ryan Lane's saltconf talks for some info on how/why they do that)
15:44 litwol i am conflicted because i want my states to be reusable.. so i try to *-formula everything.
15:44 litwol On the other hand i am using source distro (gentoo), and the system/pkg/service management /workflow|pattern/ doesn't translate to other binary distros at all
15:45 litwol various salt formulas insist adding 'gentoo' stuff into their map.jinja.. but all they are doing is trying to emulate binary-like behavior inside source distro
15:45 rocket iggy I found the issue
15:45 rocket if I specify providers  pkg: pkgng it blows up
15:47 litwol sounds like i shouldn't feel guilgy if i don't defaults+map abstract every state/formula i write
15:48 epcim_ joined #salt
15:48 litwol This is a fascinating example of ill-fitting pattern for source distro: https://github.com/saltstack-formulas/php-formula/tree/master/php
15:50 otter768 joined #salt
15:51 druonysus joined #salt
15:51 froztbyte that was easy
15:51 froztbyte I'm really liking this
15:51 litwol minimalist example of installing all those php5-* plugins for source: http://dpaste.com/248NRKG
15:52 sandah joined #salt
15:52 danjj left #salt
15:53 ek6 joined #salt
15:55 ndrei joined #salt
16:00 schuckles joined #salt
16:03 Auroch joined #salt
16:03 dimeshake joined #salt
16:08 riftman joined #salt
16:11 martoss joined #salt
16:12 jalbretsen joined #salt
16:15 julienlavergne1 how to prevent a master to generate its keypair ?
16:17 agend joined #salt
16:24 dimeshake joined #salt
16:24 aparsons joined #salt
16:26 L2SHO joined #salt
16:27 L2SHO can someone give me an example of how I would use salt.modules.pkgng.update_package_site to install all my packages from my own repo?  I'm not really grasping it.
16:28 julienlavergne1 master_pubkey_signature is a new feature from 2015.5.1 only ?
16:28 supersheep joined #salt
16:29 baoboa joined #salt
16:30 amcorreia joined #salt
16:31 cruatta joined #salt
16:32 pchero_work left #salt
16:33 iggy litwol: the php-formula is crap, don't follow it
16:33 bhosmer joined #salt
16:35 _mjones joined #salt
16:38 litwol i'm not
16:44 writtenoff joined #salt
16:44 pdayton joined #salt
16:44 nate_c joined #salt
16:45 KyleG joined #salt
16:45 KyleG joined #salt
16:45 j-saturne joined #salt
16:50 rocket ok iggy I have tracked this down more .. I need to be able to forcibly choose the pkg provider but its broken
16:51 kiorky hi
16:51 codehotter I can call execution modules with salt tgt module args, but can I also call state modules like that?
16:51 kiorky does someone have recently problems with cmdmod.run (on develop, essentially)
16:51 garphy`aw joined #salt
16:51 kiorky See https://github.com/lxc/lxc/issues/548#issuecomment-108012672 for details
16:52 mpanetta joined #salt
16:52 desposo joined #salt
16:53 arount joined #salt
16:53 sunkist joined #salt
16:53 pdayton Can I install minions on Ubuntu 10.04?
16:55 smcquay joined #salt
16:58 bhosmer joined #salt
17:00 codehotter Can I trigger a state without sls file, from python api or command line?
17:00 codehotter or is that what execution modules are for
17:00 codehotter like I want a certain package to be installed
17:00 bmac2 joined #salt
17:00 codehotter but if it's already installed, no error
17:02 codehotter I can probably just use the execution module
17:02 cruatta_ joined #salt
17:02 schuckles joined #salt
17:05 Dan_ joined #salt
17:06 Guest24163 Hello, Is anybody experienceing a problem with 2015.5.0 where targeting 100+ minions results in the salt command hanging?
17:06 Guest24163 I searched the bugs and didn't see anything similar to what I'm seeing.
17:07 Guest24163 Windows minions are a mix of 2014.7.1, 2014.7.2, 2014.7.5, and 2015.5.0.
17:09 cruatta joined #salt
17:11 ramaseshan joined #salt
17:13 ramaseshan joined #salt
17:17 32NAB3CSF joined #salt
17:17 forrest joined #salt
17:18 ajw0100 joined #salt
17:19 hal58th__ codehotter, you can write your state out manually with "state.single". It looks like you can also call one id in a state with "state.sls_id".
17:19 hal58th__ codehotter. Look up the docs with sudo salt-call sys.doc state
17:20 hal58th__ Guest24163: Unfortunately you won't find a lot of help since you have Windows minions.
17:21 jespada joined #salt
17:22 codehotter I'm finally starting to feel like I understand salt a little bit, what's the difference between states, modules, etc.
17:22 murrdoc joined #salt
17:23 hal58th__ states use the state system (state system is itself a module). modules are python programs.
17:23 hal58th__ So when I do state.highstate, state is the python program (module), highstate is a python function
17:25 32NAB3CSF left #salt
17:25 aphor joined #salt
17:28 p66kumar joined #salt
17:28 hal58th__ joined #salt
17:31 ryuhei joined #salt
17:32 tiadobatima joined #salt
17:33 ramaseshan joined #salt
17:34 theologian joined #salt
17:35 baweaver_ joined #salt
17:36 racooper joined #salt
17:37 baweaver_ joined #salt
17:39 impi joined #salt
17:41 ramaseshan joined #salt
17:42 cruatta joined #salt
17:43 cruatta_ joined #salt
17:43 cruatta joined #salt
17:46 murrdoc yes
17:47 pdayton joined #salt
17:47 linjan joined #salt
17:47 ndrei_ joined #salt
17:48 baweaver_ joined #salt
17:50 bhosmer joined #salt
17:51 ryuhei Is there any doc for updating from 2014.7.* to 2015.5.* ?
17:51 otter768 joined #salt
17:55 jrdnr joined #salt
17:56 forrest ryuhei: In what way past just doing a package update?
17:56 ageorgop joined #salt
17:56 forrest Usually we recommend people update a test environment first to make sure everything is fine with the upgrade.
17:57 rvankleeck joined #salt
18:00 rvankleeck I'm trying to use the docker formula, but getting 'Unable to manage file: Jinja variable 'dict object' has no attribute 'aws_bucket''
18:01 ryuhei just curious if there's any documentation like that before I update.. on stage environment.
18:01 rvankleeck i have my pillar set up like this example: https://github.com/saltstack-formulas/docker-formula/blob/master/pillar.example
18:01 rvankleeck however, it seems that 'merge' isn't actually merging
18:01 drawsmcgraw joined #salt
18:02 rvankleeck salt --versions-report:  http://dpaste.com/
18:02 wendall911 joined #salt
18:02 rvankleeck woops...http://dpaste.com/02ST04P
18:04 rvankleeck any idea why merge would only work if the setting is also included in the 'default' statement as opposed to adding in additional settings from the pillar?
18:04 julienlavergne joined #salt
18:04 jalbretsen AGH!!!
18:04 * jalbretsen beats Jinja whitespace with a stick
18:05 Heartsbane jalbretsen: find its inner and beat its little ass
18:05 Heartsbane s/inner/inner child/
18:07 pdayton joined #salt
18:10 * rvankleeck slaps forehead
18:10 rvankleeck you know what's a good idea when using the pillar? adding pillar_roots to the master.conf >_<
18:11 baweaver_ joined #salt
18:16 bin_005 joined #salt
18:16 baweaver_ joined #salt
18:22 pdayton joined #salt
18:22 murrdoc joined #salt
18:25 MTecknology I need to start testing out RAET
18:26 MTecknology I'd love to make the switch over to it. It might alleviate a butt ton of problems I've been seeing
18:28 tiadobatima joined #salt
18:28 pdayton joined #salt
18:32 murrdoc1 joined #salt
18:33 mr_daniel joined #salt
18:34 murrdoc joined #salt
18:36 murrdoc joined #salt
18:37 pdayton joined #salt
18:38 scoates_ joined #salt
18:41 denys joined #salt
18:43 nate_c joined #salt
18:44 dfinn joined #salt
18:45 RabidCicada joined #salt
18:46 N-Mi joined #salt
18:47 RabidCicada joined #salt
18:50 Lyfe joined #salt
18:53 pdayton joined #salt
18:57 teebes joined #salt
18:58 iggy rocket: file a bug I guess... I imagine that's not an often tested feature... it probably should be (if it's possible to unit test something like that)
18:59 iggy ryuhei: the release notes have some things to watch out for (cmd.run changes probably being the big one)
19:00 ajw0100 joined #salt
19:01 arount joined #salt
19:01 rocket iggy I did open a bug for it .. I have a workaround that works for me but doesnt address the real issue
19:01 iggy MTecknology: afaik, you still can't run both together
19:02 ryuhei iggy: thanks!  I've been reading release notes
19:02 tiadobatima joined #salt
19:03 iggy rocket: #?
19:05 hybridpollo joined #salt
19:06 colonD joined #salt
19:08 drawsmcgraw joined #salt
19:08 pdayton joined #salt
19:08 N-Mi joined #salt
19:09 ProT-0-TypE joined #salt
19:09 rocket iggy: apparently I never submitted it. 1 sec..
19:10 rocket https://github.com/saltstack/salt/issues/24340
19:11 tiadobatima joined #salt
19:13 iggy rocket: put the full command that gave the traceback (and preferably the complete output)
19:16 rocket iggy: I have put the command I ran and full output... but the full output isnt much different than the single traceback
19:17 iggy yeah, I'm just looking at it from their point of view
19:17 iggy you and I have been talking about it so we have some context
19:17 iggy that's an easy way to give them context
19:18 jacksoncage joined #salt
19:18 supersheep joined #salt
19:22 perfectsine joined #salt
19:22 druonysus joined #salt
19:22 N-Mi joined #salt
19:24 MTecknology iggy: I meant deploy a few new boxes for testing
19:24 MTecknology You do much yet with RAET on Debian 8?
19:24 iggy I have not
19:25 iggy 2015.2 fixed all of our connectivity problems, so I moved RAET to the back burner
19:26 rocket iggy: I added a few more notes to the ticket and my workaround... so that I didn't need to turn on the provider setting
19:27 * MTecknology is still at 2014.1.13
19:31 sdm24 joined #salt
19:34 iggy MTecknology: poor bastard
19:35 murrdoc joined #salt
19:39 N-Mi joined #salt
19:39 MTecknology iggy: I hear that frequently... :(
19:40 fyb3r joined #salt
19:41 fyb3r Wanted to come in and mention that the documentation http://docs.saltstack.com/en/latest/ref/clients/#salt.wheel.WheelClient there is incorrect
19:41 fyb3r looks like someone just copy pasted the code for the Runner Client
19:43 iggy fyb3r: mind filing a bug on the github project?
19:45 jschroeder joined #salt
19:46 SEJeff left #salt
19:48 fyb3r Not at all. It may be a bit later though
19:51 giantlock joined #salt
19:51 Ph-x joined #salt
19:51 rdutch joined #salt
19:52 otter768 joined #salt
19:56 N-Mi joined #salt
19:57 baweaver|work joined #salt
20:00 ahammond iggy speaking of bugs, has anyone taken a look at https://github.com/saltstack/salt/issues/24198
20:01 iggy ahammond: I couldn't say (I don't have any more insight into the company than anyone else)
20:01 iggy but if nobody has added tags/commented, I think it's safe to say nobody has looked at it yet
20:02 ahammond I'm hoping it's me doing something wrong, but I can't figure out what. Seems to me that event.send should cause events to be visible on the master.
20:04 ek6 ahammond: two questions that pop into my head... a: running multimaster?  b: was salt-minion process running at the time?
20:04 supersheep joined #salt
20:05 iggy iirc, you also tried event.fire_master ?
20:06 ahammond yup, tried fire_master, too. not running mulitmaster, yes, I think salt-minion was running at the same time.
20:07 baweaver|work joined #salt
20:11 adelcast is it possible to use one set of keys for both ZeroMQ communication and ssh?
20:12 adelcast nothing wrong with the current configuration, but seems like Salt generates it's own keys always, while it could be leveraging keys already generated....
20:13 N-Mi joined #salt
20:20 druonysuse joined #salt
20:21 hal58th__ zeromq communication might use different keys
20:24 Pulp joined #salt
20:28 druonysus joined #salt
20:28 druonysus joined #salt
20:33 spookah joined #salt
20:39 fyb3r left #salt
20:41 sdm24 Can I have multiple "unless:" statements in a file? would I format them like a list?
20:44 iggy sdm24: yes, yes
20:44 sdm24 great, thanks
20:45 sdm24 do you know if it would run with "AND" logic, or "OR" logic?
20:46 arount joined #salt
20:48 rdutch sdm24: you can use jinja
20:49 _mjones left #salt
20:50 evilrob joined #salt
20:50 sdm24 ooh good call, thanks
20:55 baweaver|work joined #salt
20:59 rdutch left #salt
21:02 baweaver|work joined #salt
21:02 iggy the docs are kind of tricky on the requisites
21:02 sporkd2 joined #salt
21:02 iggy some are AND some are OR
21:02 iggy you'll have to read the docs to see which
21:03 sporkd2 hey everyone, when I launch an instance with salt-cloud it outputs a lot of aws specific grains, how do I get those onto the minion so when running grains.items I can see them?
21:03 rocket when including an sls file .. how is the ordering defined
21:03 sporkd2 joined #salt
21:03 sporkd2 rocket I believe includes run first
21:03 iggy true
21:04 supershe_ joined #salt
21:04 baweaver|work joined #salt
21:04 rocket even if they are at the end of the file?
21:04 iggy sporkd2: you'd have to setup a custom grain module to pull the ec2 metadata
21:04 iggy rocket: if you mean include: yes
21:04 sporkd2 iggy: what is salt-cloud using to output after deploying a VM?
21:05 iggy sporkd2: no idea, never used salt-cloud with ec2
21:05 iggy https://github.com/saltstack/salt-contrib/tree/master/grains
21:05 iggy there are a few different ec2 ones there
21:06 bin_005 joined #salt
21:08 timoguin anyone know how to reference a Jinja variable within a Jinja set block?
21:09 iggy {% set foo = 'something' ~ var1 ~ 'bar' %}
21:09 iggy like that?
21:09 timoguin Yea, so ~ does concat?
21:10 iggy you can technically use + or `
21:10 iggy or ~
21:10 iggy not `
21:10 timoguin thanks
21:10 iggy ~ means convert to string and concat
21:10 pdayton joined #salt
21:10 iggy otherwise if it's an int or something, you'll get errors
21:13 clintberry joined #salt
21:15 cberndt joined #salt
21:18 linjan joined #salt
21:18 denys joined #salt
21:28 hal58th joined #salt
21:28 hal58th_ joined #salt
21:28 hal58th__ joined #salt
21:30 forrest basepi: Can you let Jacob know this 404s now? http://docs.saltstack.com/topics/tutorials/quickstart.html
21:30 arount joined #salt
21:30 basepi thanks
21:35 murrdoc joined #salt
21:35 sgargan joined #salt
21:37 c10b10 joined #salt
21:38 rdavis joined #salt
21:40 sgargan joined #salt
21:43 rdavis Anyone familiar with the AD auth stuff in the master?  I'm trying to set up group authentication and it works fine if I have my username in the external_auth section of the master, but if I remove it and put in group%, I get 'Authentication failure of type "eauth" occurred.'.
21:45 rdavis Looks like I'm hitting something in the mk_token part of the master, but that might also be because I'm running an older version, lemme update adn see if it helps.
21:46 basepi forrest: should be fixed
21:46 basepi refresh of mpt
21:46 basepi if not*
21:47 forrest Looks good to me, thanks basepi
21:49 dimeshake joined #salt
21:52 tiadobatima joined #salt
21:53 otter768 joined #salt
21:53 rdavis left #salt
21:58 spookah joined #salt
22:01 c10b10_ joined #salt
22:02 iggy forrest: I have a bug open about it already
22:02 forrest iggy: Ahh okay, well the page is back
22:03 iggy there weren't forwards setup to go to en/latest for everything
22:03 iggy so a ton of external links broke
22:06 kwork joined #salt
22:06 kwork joined #salt
22:07 murrdoc joined #salt
22:08 sunkist joined #salt
22:09 sunkist1 joined #salt
22:10 pcdummy joined #salt
22:11 sgargan joined #salt
22:17 baweaver|work joined #salt
22:19 amcorreia joined #salt
22:30 baweaver|work joined #salt
22:33 hal58th_1 joined #salt
22:33 hal58th_2 joined #salt
22:35 pdayton joined #salt
22:35 hal58th_3 joined #salt
22:35 JonGretar joined #salt
22:36 mattl joined #salt
22:39 pdayton joined #salt
22:41 cedwards joined #salt
22:42 antonw joined #salt
22:42 hal58th joined #salt
22:42 cedwards I'm not able to find any documentation on what can and cannot be set as configuration options within pillar data.
22:42 Striki joined #salt
22:43 bin_005 joined #salt
22:43 LinuxHorn joined #salt
22:44 natewalck joined #salt
22:46 joeyparsons joined #salt
22:50 grepory joined #salt
22:53 diegows joined #salt
22:55 diegows who do I configure a specific service for pam auth? I'm trying with
22:55 diegows external_auth:
22:55 diegows pam:
22:55 diegows service: salt
22:55 diegows but no luck
22:56 copelco joined #salt
22:56 davedash joined #salt
22:57 cedwards1 joined #salt
22:57 MatthewsFace joined #salt
22:57 Singularo joined #salt
22:58 cruatta_ joined #salt
22:58 hal58th cedwards what's your goal?
22:59 cruatta joined #salt
22:59 cedwards hal58th: well I've noticed I can put scheduler config in pillar, just curious what else is a good candidate
23:00 munhitsu joined #salt
23:00 cedwards and it looks like mine_functions can be defined there as well..
23:00 akoumjian joined #salt
23:01 baweaver|work joined #salt
23:02 yidhra_ joined #salt
23:02 timoguin anyone know if the file.copy state will copy a directory?
23:03 aurynn I think there's a recurse option?
23:07 dRiN joined #salt
23:07 sgargan joined #salt
23:08 Aidin joined #salt
23:09 timoguin Well there's file.recurse, but as far as I know that's for copying from a master to a minion
23:09 timoguin I want to copy from one location on a minion to another
23:10 KyleG Should be easy enough to test
23:10 litwol you cant
23:10 litwol look up file push
23:10 litwol timoguin: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.cp.html#salt.modules.cp.push
23:12 iggy timoguin: cmd.run it :/
23:14 forrest yeah just use cmd.run
23:15 forrest someone else asked this exact question earlier this week, timoguin what software are you using that you need to copy files that aren't managed by salt :\
23:15 forrest some crappy software with no installer or something?
23:15 forrest just dumps files, no choice on the directory? ;)
23:21 smcquay joined #salt
23:28 mosen joined #salt
23:34 ajw0100 joined #salt
23:35 binocvlar joined #salt
23:38 MindDrive So I just tried to set up gitfs using Dulwich (0.10.0), and have a 'gitfs_remotes' entry like the following: git@mycompany.com:salt/salt_modules.git .   I got the following error in the logs: "Exception 'unknown scheme ''' caught while fetching gitfs remote ..."  Does the dulwich support not recognize that remote entry form?
23:39 MindDrive (The docs don't seem to indicate as such.)
23:39 viderbit joined #salt
23:39 evilrob joined #salt
23:40 timoguin forrest: the Bamboo agent installer is an archive
23:40 forrest disguisting
23:40 timoguin fetching it and extracting it
23:40 clintberry joined #salt
23:40 forrest Not surprising for an atlassian product though
23:41 timoguin They recommend and only support using their own AMIs, but I'm building a custom one
23:42 forrest Yeah I'm sure the ydo
23:42 forrest *they do
23:43 timoguin Yea if I don't build a custom one I'm stuck with customizing one of theirs and taking a snapshot, which can be attached to an instance that they spin up
23:43 timoguin but that's way limited
23:44 MindDrive Never mind, https://github.com/saltstack/salt/issues/20993 seems to indicate is does not support it... bleah.
23:48 forrest timoguin: Yeah I agree, just sucks that their suggestion is to use their image because they don't provide a proper package.
23:48 forrest Seems oddly common with java apps, and I don't get why
23:50 timoguin yea. they do require a few other things to be installed, the main thing being ec2-api-tools. that's actually the zip i need to move because there's a folder inside the archive so i can't put it exactly where i want with archive.extracted
23:51 packeteer joined #salt
23:52 macinrack joined #salt
23:54 otter768 joined #salt
23:54 whytewolf forrest: cause java devs tend to want to control every last detail of the enviroment. and don't like writeing requirements.
23:54 forrest heh
23:55 hal58th_ joined #salt
23:55 hal58th__ joined #salt
23:55 aurynn reproducibility is important
23:55 hal58th_3 joined #salt
23:59 CeBe joined #salt

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