Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-09-09

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

All times shown according to UTC.

Time Nick Message
00:00 zz_Cidan joined #salt
00:00 forrest debian112, nothing particular, just takes some time to figure out
00:00 murrdoc do u edit reclass by hand ?
00:01 murrdoc like, if needed
00:01 forrest each server has it's own file
00:01 forrest which imports other files and such
00:01 forrest so you can modify as needed.
00:01 forrest it's cumbersome
00:02 debian112 I will check it out. Heading to gym now to destress
00:02 forrest good plan, later
00:02 hardwire joined #salt
00:05 TTimo joined #salt
00:06 murrdoc left #salt
00:06 hcr joined #salt
00:06 murrdoc joined #salt
00:08 hcr Hi folks, first time here. I need some help.
00:08 TyrfingMjolnir joined #salt
00:08 hcr I am new to salt and i am trying to learn by setting salt up to run minion/master on the same machine
00:09 hcr I am stuck with this [DEBUG   ] Attempting to authenticate with the Salt Master at 127.0.0.1
00:10 hcr I put in a debug line to confirm that i was connecting on the right ip:port
00:10 brain5ide joined #salt
00:10 hcr [DEBUG   ] Attempting to authenticate with the Salt Master at 127.0.0.1 [DEBUG   ] Loaded minion key: /etc/salt/pki/minion/minion.pem [INFO    ] Salt master_uri = tcp://127.0.0.1:4506 [INFO    ] SaltReqTimeoutError: after 60 seconds. (Try 1 of 7)
00:11 hcr I confirmed that i can telnet to this port on 4506. Master logs do not indicated that a connection attempt was made.
00:11 TTimo hcr: you can run locally, you don't need to loopback to the demon
00:11 TTimo there's a piece in the doc that explains that
00:12 hcr oh sorry i missed that
00:13 TTimo http://docs.saltstack.com/en/latest/topics/tutorials/quickstart.html
00:13 TTimo there
00:14 halfss joined #salt
00:14 snuffeluffegus joined #salt
00:15 hcr thanks TTimo!
00:15 hardwire joined #salt
00:15 hcr Let me try this
00:16 hcr TTimo: does this mean that i cannot run master/minion as separate daemons on the same machine?
00:16 tmh1999 joined #salt
00:16 Gareth hcr: you can.
00:16 Gareth they do run as seperate daemons.
00:16 aparsons joined #salt
00:19 mr_chris joined #salt
00:20 hcr Oh ok. Is the way I tried before not correct?
00:20 hardwire joined #salt
00:21 Gareth hcr: how did you try it before?
00:22 hcr Sorry. It was a simple setup. I had salt-minion bind to 127.0.0.1 and listen to 4506. The minion tries to connect on this port
00:22 hcr But the minion is stuck retrying here
00:22 hcr [DEBUG   ] Attempting to authenticate with the Salt Master at 127.0.0.1
00:22 hcr [DEBUG   ] Loaded minion key: /etc/salt/pki/minion/minion.pem
00:22 hcr [INFO    ] Salt master_uri = tcp://127.0.0.1:4506
00:22 hcr [INFO    ] SaltReqTimeoutError: after 60 seconds. (Try 1 of 7)
00:23 Gareth hcr: ahh. you have to sign the minion's key from the master....look at salt-key
00:23 hcr Tried that
00:23 hcr it is empty
00:23 hcr Accepted Keys: Unaccepted Keys: Rejected Keys:
00:24 Gareth the master is running right?
00:24 Ryan_Lane anyone have experience with salt and coreos+docker?
00:24 hcr yes
00:24 Ryan_Lane I'd really like to control coreos through salt
00:25 Gareth hcr: can you pastebine your minion and master conf. files?
00:25 hcr I can telnet localhost 4506
00:25 hcr yes will do
00:25 hardwire joined #salt
00:26 TTimo Ryan_Lane: there's not really much reasons you couldn't do that
00:27 TTimo CoreOS is a bit new, I don't know if the bootstrap/packaging system supports install
00:27 Ryan_Lane yeah, I just need to figure out how
00:27 delinquentme Ive seen some posts w people talking about having all of the minions generate keys... and then accepting the keys .... Is there a really good reason to not just generate a single keypair and then use that for all the minions??
00:27 Ryan_Lane I'm thinking of a container that runs salt and talks to coreos
00:27 jubei__ Ryan_Lane: might have to install salt from source
00:27 Ryan_Lane but I don't see how I can communicate with coreos
00:27 TTimo delinquentme: it's one key per connected minion
00:27 TTimo that's how they are identified
00:27 Ryan_Lane except through ssh
00:28 TTimo e.g. the master will say 'in use'
00:28 Ryan_Lane jubei__: oh, I'm way past that ;)
00:28 Ryan_Lane I have like 20 containers that run services that are configured via salt
00:28 Ryan_Lane but coreos itself is handled with cloud-init
00:28 halfss joined #salt
00:28 Ryan_Lane which is also fine, but in vagrant we don't automatically start all containers
00:29 Ryan_Lane so we have annoying process of starting the containers via vagrant ssh commands
00:29 Ryan_Lane what I really want is a single container that runs a web service. that web service would show all the available containers and their current status
00:30 Ryan_Lane and I'd like to be able to start/stop the containers and such via that
00:30 Ryan_Lane so, I'd like the web service to call salt, which would cause coreos to stop/stop the containers
00:30 Ryan_Lane *start/stop
00:30 hcr Gareth: http://pastebin.com/b0yZP5Q8
00:30 hardwire joined #salt
00:30 hcr got rid of the comments.
00:31 jnials joined #salt
00:31 jubei__ Ryan_Lane: have you looked at the salt rest api
00:31 Ryan_Lane it looks like I could use fleet for this, but I was really hoping to avoid using fleet
00:31 Ryan_Lane jubei__: I know how I can talk to salt, I'm trying to see if anyone had any idea of how I could make salt talk to coreos
00:36 hardwire joined #salt
00:40 halfss joined #salt
00:41 hardwire joined #salt
00:43 Gareth hcr: looks right.  You're running the master in debug mode too?
00:46 hardwire joined #salt
00:47 hardwire joined #salt
00:47 delinquentme TTimo, so you generate a key on each minion ... and then have each of those keys added to the master
00:48 digin4 joined #salt
00:48 otter768 joined #salt
00:48 delinquentme I mean that seems to be the convention ... but isn't it going to be simpler to operate w a single key ?
00:48 hcr Gareth: yes
00:49 hcr Gareth: Master does not show any signs of a connection attempt
00:50 TTimo delinquentme: you can't though
00:51 TTimo I tried .. the master identifies the minion by the key it connects as
00:51 TTimo there is documentation showing how to generate and pre-accept keys
00:51 delinquentme TTimo, Im actually doing this right now ?? and I don't have any issues identifying individual minions?
00:51 Gareth hcr: no firewall or anything on the machine?
00:51 TTimo I do this in a docker context, but it would work in a more general situation
00:51 TTimo oh really? cool .. I must have missed something then
00:52 murrdoc Ryan_Lane:  did u eever try salt.cloud.ec2 ?
00:53 hardwire joined #salt
00:54 hcr wait
00:54 hcr Gareth: let me check
00:55 hcr But that would stop me from telnetting too right?
00:55 Ryan_Lane murrdoc: nope
00:56 Ryan_Lane murrdoc: it was missing so much and didn't use boto
00:56 bhosmer joined #salt
00:57 murrdoc word, thanks
00:57 hcr Gareth: No firewall
00:57 mr_chris joined #salt
00:57 delinquentme TTimo, at current im doing this through instancing up 4 minions ... calling the 01 the "controller minion" ( im fucking w slurm ) ... generating the key on that "control-minion"  using a minion> master push ... placing it in /srv/salt ... and running $salt \* state.highstate
00:57 thayne joined #salt
00:59 hcr Gareth: Just to be sure i changed the interface to 0.0.0.0 and I am able to telnet to this port from outside
01:01 Gareth hcr: telnet to localhost 4506 works?
01:01 delinquentme actually haha im confused.  Im just copying over the local key pair and serving it up through the salt host
01:01 delinquentme ( yes I know this is security issue )
01:05 hcr yes
01:05 hcr Gareth: yes
01:05 Gareth hcr: stop both....start the master, wait a about 10 seconds, then start the minion.
01:06 hcr okay let me try that
01:07 hcr Gareth: did not work
01:08 Gareth hcr: stop both, delete the /etc/salt/pki directory, start the master...10 seconds later, start the minion.
01:08 hcr okay
01:08 Gareth then check salt-key
01:10 hcr Gareth: I use "/etc/salt-master-config/"  and "/etc/salt-minion-config/" config directories to run the master and minion. Do I need /etc/salt/ directory?
01:10 halfss joined #salt
01:11 murrdoc if u configured the salt master/minion in the config file
01:11 Gareth hcr: unless you've changed the config files they'll expect a /etc/salt directory, both the minion and master can both share that directory.
01:11 hcr okay
01:12 hardwire joined #salt
01:12 hcr Gareth: Sorry that did not work too.  salt-key -L Accepted Keys: Unaccepted Keys: Rejected Keys:
01:12 hcr and the minion is stuck trying to get in touch with master
01:14 Gareth very strange.
01:15 hcr mgw had similar problem (http://irclog.perlgeek.de/salt/2014-06-16/text) I did not understand the solution (also doubt whether that applies to my case too)
01:17 AdamSewell joined #salt
01:17 hardwire joined #salt
01:20 drags is there a way to specify which provider salt-cloud sends a destroy command at? whenever I use --destroy or --action=destroy salt-cloud trys to destroy the instance at softlayer
01:22 hardwire joined #salt
01:25 jnials joined #salt
01:26 CryptoMer joined #salt
01:26 CryptoMer is there an immutable flag with file.managed?
01:27 hardwire joined #salt
01:28 hardwire joined #salt
01:33 Hell_Fire_ joined #salt
01:36 halfss joined #salt
01:38 hardwire joined #salt
01:40 robawt hey salt folk, is there a nice way to fail a state?
01:40 robawt like if i wanted to check a grain for a value, and if i didn't like the value fail the state?
01:41 hcr Gareth: could this be zeromq not working for me for some reason. Is there some way i can debug that?
01:46 mr_chris joined #salt
01:47 drags robawt: https://github.com/saltstack/salt/issues/10723#issuecomment-36682086
01:48 hardwire joined #salt
01:49 hardwire joined #salt
01:50 elfixit joined #salt
01:50 cb joined #salt
01:50 halfss_ joined #salt
01:50 blarghmatey joined #salt
01:51 Jarus joined #salt
01:51 mschiff joined #salt
01:51 mschiff joined #salt
01:52 Emantor joined #salt
01:54 blarghmatey joined #salt
01:58 DaveQB joined #salt
02:01 thayne joined #salt
02:01 blarghmatey joined #salt
02:04 hardwire joined #salt
02:14 hardwire joined #salt
02:16 blarghmatey joined #salt
02:16 yomilk joined #salt
02:19 hardwire joined #salt
02:21 digin4 joined #salt
02:22 geekmush joined #salt
02:24 TyrfingMjolnir joined #salt
02:28 hcr joined #salt
02:29 nitay joined #salt
02:29 nitay is there a way to run something to the effect of => grains['id'].gsub('.','-')
02:29 nitay inside an sls?
02:30 TTimo joined #salt
02:31 jalbretsen joined #salt
02:46 hcr Gareth: I moved to an older version of salt (0.17.4) and i do not see this problem
02:46 hcr I was working off trunk
02:46 bhosmer joined #salt
02:52 thayne joined #salt
02:55 murrdoc joined #salt
02:55 ramishra joined #salt
02:58 Gareth hcr: odd.  must be a recent bug, I'm running the develop branch and don't see that issue.
02:59 Gareth hcr: 0.17.4 is a bit old...you might want to install 2014.1.10.
03:07 picker joined #salt
03:09 ramishra joined #salt
03:14 yomilk joined #salt
03:14 higgs001 joined #salt
03:25 murrdoc joined #salt
03:25 hcr Gareth: On my system it works till v2014.7.0rc1. I redownloaded the develop branch and that still hangs trying to connect.
03:32 xmly joined #salt
03:32 xmly hey
03:32 halfss joined #salt
03:32 delinquentme joined #salt
03:33 aparsons joined #salt
03:33 malinoff joined #salt
03:33 murrdoc joined #salt
03:33 Guest10835 joined #salt
03:33 wiqd__ joined #salt
03:33 copelco_ joined #salt
03:33 akoumjian_ joined #salt
03:33 JordanTesting_ joined #salt
03:33 goki___ joined #salt
03:33 grepory_ joined #salt
03:33 XenophonF829 joined #salt
03:34 TTimo joined #salt
03:34 octarine_ joined #salt
03:34 gamingrobot_ joined #salt
03:34 XenophonF829 hello
03:35 XenophonF829 hang on brb
03:35 modafinil__ joined #salt
03:35 basepi_ joined #salt
03:35 arapaho_ joined #salt
03:35 Damianz_ joined #salt
03:35 eofs_ joined #salt
03:36 aparsons_ joined #salt
03:36 DenkBret1l joined #salt
03:36 txmoose joined #salt
03:36 QuinnyPi1 joined #salt
03:36 sijis_ joined #salt
03:36 rnts_ joined #salt
03:36 gamingrobot_ joined #salt
03:37 schristensen_ joined #salt
03:37 ajw0100 joined #salt
03:38 vividloop_ joined #salt
03:38 xcbt_ joined #salt
03:39 estherbester_ joined #salt
03:39 aurynn_ joined #salt
03:39 horus_plex joined #salt
03:39 TOoSmOotH_ joined #salt
03:39 whitepaws_ joined #salt
03:40 TomD joined #salt
03:40 bonezed joined #salt
03:40 knot joined #salt
03:42 thayne joined #salt
03:42 Nazca joined #salt
03:42 neilf___ joined #salt
03:42 ajw0100 joined #salt
03:42 `ph8 joined #salt
03:42 ajprog_laptop joined #salt
03:42 davet1 joined #salt
03:42 dstokes joined #salt
03:42 Dinde joined #salt
03:42 ecdhe joined #salt
03:42 jayne joined #salt
03:42 juice joined #salt
03:42 roo9- joined #salt
03:42 gfa joined #salt
03:44 aparsons joined #salt
03:45 snuffeluffegus joined #salt
03:45 briner joined #salt
03:45 zenwryly joined #salt
03:47 snuffeluffegus joined #salt
03:49 bhosmer joined #salt
03:50 Outlander joined #salt
03:52 XenophonF joined #salt
03:54 ramishra joined #salt
04:05 aparsons joined #salt
04:07 notbmatt joined #salt
04:07 Ryan_Lane joined #salt
04:08 berto- joined #salt
04:09 yomilk joined #salt
04:12 murrdoc joined #salt
04:14 spookah joined #salt
04:18 dccc joined #salt
04:20 Jarus joined #salt
04:22 mr_chris joined #salt
04:25 arapaho joined #salt
04:29 XenophonF left #salt
04:29 troyready joined #salt
04:31 snuffeluffegus joined #salt
04:32 kermit joined #salt
04:33 notpeter_ joined #salt
04:34 delinquentme joined #salt
04:34 bhosmer joined #salt
04:36 Jarus joined #salt
04:37 mschiff joined #salt
04:37 mschiff joined #salt
04:37 Emantor joined #salt
04:42 felskrone joined #salt
04:44 yomilk joined #salt
04:44 grove_ joined #salt
04:45 mpoole joined #salt
04:46 ramteid joined #salt
04:47 dccc joined #salt
04:48 pdayton joined #salt
04:56 FL1SK joined #salt
04:56 mpoole joined #salt
04:58 juissi joined #salt
05:04 FL1SK joined #salt
05:06 jchen joined #salt
05:06 knot joined #salt
05:08 manfred hcr: just to make sure since I don't see confirmation in  my backlog... may have missed it, but you can hit 4505 too right?
05:08 schimmy joined #salt
05:08 lopez joined #salt
05:13 schimmy joined #salt
05:17 mpoole joined #salt
05:26 knot joined #salt
05:29 SheetiS joined #salt
05:29 calvinh joined #salt
05:34 viq joined #salt
05:36 calvinh joined #salt
05:37 TTimo joined #salt
05:39 grove_ joined #salt
05:42 knot joined #salt
05:42 calvinh_ joined #salt
05:42 mr_chris joined #salt
05:57 ilbot3 joined #salt
05:57 Topic for #salt is now Welcome to #salt | 2014.1.10 is the latest | Help us test the 2014.7 RC! http://bit.ly/salt-rc | SaltStack is hiring! http://www.saltstack.com/careers | Please be patient when asking questions as we are volunteers and may not have immediate answers | Channel logs are available at http://irclog.perlgeek.de/salt/
05:58 troyready joined #salt
06:09 murrdoc joined #salt
06:12 digin4 joined #salt
06:17 masm joined #salt
06:18 ndrei joined #salt
06:18 halfss joined #salt
06:19 aparsons joined #salt
06:20 aparsons_ joined #salt
06:20 jnials joined #salt
06:22 digin4 joined #salt
06:23 bhosmer joined #salt
06:23 halfss_ joined #salt
06:27 lcavassa joined #salt
06:30 tmh1999 joined #salt
06:30 alanpearce joined #salt
06:30 aparsons joined #salt
06:35 digin4 joined #salt
06:45 ecerulm joined #salt
06:46 Sweetsha1k joined #salt
06:49 tomspur joined #salt
06:49 tomspur joined #salt
07:01 kingel joined #salt
07:02 sctsang joined #salt
07:08 digin4 joined #salt
07:11 Libermentix joined #salt
07:11 calvinh joined #salt
07:12 nmadhok joined #salt
07:13 nmadhok anyone up?
07:13 Libermentix i have a couchdb formula (source install tested on Debian) that could be added to the salt-formula project, is this the right channel?
07:13 nmadhok @techhat ping
07:14 nmadhok @libermentix post the link to repo
07:15 nmadhok i'll take a look at it
07:15 Libermentix https://github.com/AISLER-NET/couchdb-formula
07:16 chiui joined #salt
07:16 Libermentix @nmadhok
07:17 Libermentix there you go... i hope it abides to the best practice, feedback is very appreciated ;-)
07:17 nmadhok alright let me take a look
07:17 TyrfingMjolnir joined #salt
07:21 nmadhok looks good to me
07:22 nmadhok it would be great if you could add support for other distributions as well
07:22 nmadhok i'm forking it and putting it under salt-formulas
07:23 nmadhok @Libermentix here https://github.com/saltstack-formulas/couchdb-formula
07:23 Libermentix yeah, i will try my best... I've set up map.jinja already, i just need to check whether the packages remain the same, and then add it
07:24 Hell_Fire_ joined #salt
07:24 nmadhok also update the read me if you can with all available states
07:24 calvinh_ joined #salt
07:26 Libermentix will do so
07:26 bhosmer joined #salt
07:27 nmadhok Thanks @Libermentix
07:30 Libermentix i'll send you a note when i have the updates readyt
07:33 felskrone joined #salt
07:36 bmcorser joined #salt
07:38 PI-Lloyd joined #salt
07:40 TTimo joined #salt
07:41 halfss joined #salt
07:42 jhauser joined #salt
07:42 martoss joined #salt
07:44 Libermentix @nmadhok original repository updated README.rst (done)
07:44 Libermentix testing other servers will take a while as we mainly use debian, i'll get back to you next week
07:46 digin4 joined #salt
07:49 martoss1 joined #salt
07:49 darkelda joined #salt
07:49 darkelda joined #salt
07:54 jdmf joined #salt
07:54 digin4 joined #salt
07:58 nmadhok sure
07:58 nmadhok i may not be available or irc
07:58 nmadhok on
08:00 Libermentix no problem ... just for my info ... is this the right channel to ask for contributions to the salt-formula project (we may have some more and are willing to contribute)
08:02 nmadhok you can ping me directly
08:03 nmadhok or send an email
08:03 Libermentix alright great
08:03 nmadhok although someone should be available on IRC to help you as well
08:05 halfss_ joined #salt
08:10 jeremmm joined #salt
08:11 bhosmer joined #salt
08:12 sectionme joined #salt
08:16 mr_chris joined #salt
08:17 aquinas joined #salt
08:18 aquinas_ joined #salt
08:18 digin4 joined #salt
08:22 CeBe joined #salt
08:23 mr_chris joined #salt
08:27 nmadhok joined #salt
08:27 _mel_ joined #salt
08:31 nmadhok left #salt
08:31 giantlock joined #salt
08:31 nmadhok joined #salt
08:31 digin4 joined #salt
08:37 yomilk joined #salt
08:43 mackstick joined #salt
08:44 jensnockert joined #salt
08:44 giannello joined #salt
08:49 ggoZ joined #salt
08:50 Twiglet joined #salt
08:53 sectionme joined #salt
08:53 davidone joined #salt
08:54 mr_chris joined #salt
08:58 davidone joined #salt
09:03 ramishra joined #salt
09:04 a1 joined #salt
09:05 Sweetshark joined #salt
09:13 kevinbrolly_ left #salt
09:15 ramishra joined #salt
09:26 che-arne joined #salt
09:29 jensnockert joined #salt
09:33 ramishra joined #salt
09:38 blackhelmethacke joined #salt
09:39 xcbt joined #salt
09:41 TTimo joined #salt
09:49 picker joined #salt
09:51 ndrei joined #salt
10:00 bhosmer joined #salt
10:03 calvinh joined #salt
10:05 johtso joined #salt
10:11 Valdo joined #salt
10:12 ggoZ joined #salt
10:13 felskrone joined #salt
10:14 alanpearce joined #salt
10:22 Outlander joined #salt
10:28 bhosmer joined #salt
10:30 bmcorser can i use the output from a cmd in my state?
10:31 bmcorser my use case is to loop over the installed python packages and see if there are any packages installed that have corresponding states
10:32 bmcorser basicically something like "for package_name in $(pip freeze): - include package_name"
10:40 ndrei joined #salt
10:44 jensnockert joined #salt
10:46 TheThing joined #salt
10:46 vbabiy joined #salt
10:52 TheThing joined #salt
11:00 davet joined #salt
11:02 tmh1999 joined #salt
11:09 Guest98699 joined #salt
11:10 calvinh_ joined #salt
11:13 dccc_ joined #salt
11:22 diegows joined #salt
11:22 TyrfingMjolnir joined #salt
11:27 davidone joined #salt
11:29 TheThing joined #salt
11:40 AdamSewell joined #salt
11:43 TTimo joined #salt
11:49 bhosmer_ joined #salt
11:54 giannello joined #salt
11:55 hobakill joined #salt
11:55 nkuttler puzzling.. why would a highstate care about the installed gobject..
11:57 juice joined #salt
11:57 Dinde joined #salt
11:58 juissi left #salt
12:01 bhosmer joined #salt
12:02 Voziv left #salt
12:04 jensnockert joined #salt
12:05 juice joined #salt
12:05 alanpearce joined #salt
12:05 ajprog_laptop joined #salt
12:05 dstokes joined #salt
12:05 ecdhe joined #salt
12:05 jayne joined #salt
12:05 roo9- joined #salt
12:08 Nexpro joined #salt
12:09 TyrfingMjolnir joined #salt
12:11 TTimo joined #salt
12:12 nkuttler joehh: i see http://dpaste.com/0SKTWE6 when attemting a highstate of a sid minion. not sure if/where to report?
12:17 AdamSewell joined #salt
12:18 nyx joined #salt
12:20 jchen joined #salt
12:21 jchen joined #salt
12:23 alanpearce joined #salt
12:25 nebuchad` joined #salt
12:27 knot` joined #salt
12:27 calvinh joined #salt
12:28 colttt joined #salt
12:33 jslatts joined #salt
12:33 marnom joined #salt
12:34 mechanicalduck_ joined #salt
12:35 vejdmn joined #salt
12:35 felskrone joined #salt
12:38 kingel joined #salt
12:38 dvestal joined #salt
12:39 winmutt joined #salt
12:40 winmutt puppet has this concept of exported resources, what does salt do that is similar? I want to generate ssh keys on servers and then collate them into authorized keys for a mercurial server
12:40 winmutt sorry to compare but puppet is what i use at the job
12:41 nkuttler winmutt: i think you want to use a salt mine
12:42 nkuttler http://docs.saltstack.com/en/latest/topics/mine/
12:47 mr_chris joined #salt
12:48 hobakill any word if salt will support domain user accounts any time soon? this says it currently only supports local accounts: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.win_useradd.html
12:51 satish_ joined #salt
12:51 giantlock joined #salt
12:52 che-arne|2 joined #salt
12:56 Setsuna666 joined #salt
13:02 thayne joined #salt
13:05 micah_chatt joined #salt
13:05 wiqd joined #salt
13:05 wiqd joined #salt
13:05 XenophonF joined #salt
13:07 jaimed joined #salt
13:08 micah_chatt_ joined #salt
13:09 racooper joined #salt
13:09 racooper joined #salt
13:14 AdamSewell joined #salt
13:16 cast joined #salt
13:16 XenophonF the part of me that loves scope creep wants to use salt for patch management :)
13:16 mike25de hi all ... has anyone confrunted with a need for a rollback situation?    salt deploys files/configs... but if someone does a mistake,,, i would need a rollback situation. Any ideas ? ...  maybe someone can give me a hint on how to start .... THANK you all!
13:17 XenophonF mike25de: if you want to roll back changes, you need to write states that do exactly that
13:17 hobakill mike25de: if you use gitfs you could revert to an older commit.
13:17 mike25de XenophonF:  do you have maybe an example?
13:18 mike25de hobakill: gitfs isn't just another way of storing my configs on the salt master?
13:19 XenophonF mike25de: no because i haven't needed to do it
13:19 mike25de my configs are generated from jinjas... so the only place i have the real/final config is on the minion.
13:19 mike25de XenophonF: thank you for the tip anyway!
13:19 XenophonF mike25de: note that file states can create backup files
13:20 mike25de XenophonF: i missed that !!!! awesome!
13:20 cast XenophonF: seems better than using the likes of spacewalk...
13:21 cast i roll forward rather than back
13:22 mechanicalduck joined #salt
13:24 nyx joined #salt
13:24 XenophonF mike25de: other issue is that the backups will get overwritten anytime the file state needs to make changes
13:24 XenophonF so you'd need to capture them somehow, maybe at the end of a state run or something
13:24 XenophonF i dunno
13:26 XenophonF all of my stuff is virtual
13:26 XenophonF so if i need to roll back, i just spin up a new pristine VM
13:26 mike25de minion cachedir  ... where is that ? what is the path for that cache folder
13:27 mike25de XenophonF: ... that's what i also do... but an AUDIT ... finding is that i need to have a rollback mechanism
13:28 gngsk joined #salt
13:34 englishm joined #salt
13:36 ndrei joined #salt
13:38 nitti joined #salt
13:38 nitti joined #salt
13:39 XenophonF mike25de: what about throwing the state tree into a version control system?
13:40 XenophonF or the on-disk configs, prior to having salt make any changes?
13:40 acabrera joined #salt
13:41 XenophonF not 100% sure how i'd do that, maybe by adding a "vcs" SLS with a vcs_init state that _all_ other states require
13:42 XenophonF and that runs a script that basically checks the contents of /etc, /usr/local/etc, etc. (hah) into sccs/rcs/cvs/svn/git/hg (if it isn't already)
13:43 XenophonF then then adding a vcs_update state that _all_ other states require_in
13:43 XenophonF and which commits any deltas to the VC repo
13:45 XenophonF i dunno
13:45 XenophonF iirc there was a way to programmatically alter salt states
13:45 XenophonF if there is, you wouldn't have to spam requires: or requires_in: lines all across your state tree
13:45 mike25de right
13:45 mike25de thanks man
13:46 XenophonF alternatively, before you do something that you know will touch a file, back it up (again, this could be to a VCS or something)
13:46 XenophonF or try to convince the auditor that your backup/DR system is generally how you'd revert those changes
13:46 XenophonF i dunno what would fly
13:46 mike25de i want to see if file.list_backups  works... at this moment.. it doesn't work... on my salt 0.17.1
13:46 XenophonF oh
13:47 XenophonF that's really old
13:47 mike25de XenophonF: yeah... i tried to update in Janyuary -- but all my states were fucked :) ... all 90 states that i had.. so i kept using 0.17.1
13:48 XenophonF yikes
13:48 mike25de at some point... i will set up a new salt master... and redo all states... to match the new version.
13:48 mike25de ironically the most basic stuff was broken... like pkg and file. manage
13:48 mike25de i am not sure why ... but i have freaked out and reverted to my working version... which is fine... -ish
13:49 mike25de the option with the backup seems to work... but the master does not read the backups from the minion.
13:50 mike25de i have the files backed up on the minion but salt 'my-vm file.list_backups /tmp/remove.txt   doesn't return anything
13:50 mike25de could be  a bug in my old version...
13:50 XenophonF man I dunno
13:50 XenophonF that sucks
13:51 cpowell joined #salt
13:51 dvestal joined #salt
13:52 mike25de XenophonF:  thanks anyway :)  i was thinking that ... before the state runs on the minion ... a backup will be created and those backups added to a git ... seems the easiest for me now
13:53 XenophonF yup yup
13:54 mike25de i actually think that is even easier ... the backups are done automatically by salt using   - backup: minion for each file.manage     and then i will add to git only /var/cache/salt/minion/file_backup which contains all the files that are being backed up
13:55 XenophonF I didn't know about the file state backups mechanism detailed here at http://docs.saltstack.com/en/latest/ref/states/backup_mode.html
13:55 XenophonF that sounds pretty good to me
13:55 XenophonF hopefully, it passes muster with your auditor
13:55 mike25de i was reading the same thing
13:55 mike25de it should be :)
13:55 XenophonF those guys can be a little funny about things sometimes
13:56 XenophonF like some things required by FIPS or PCI DSS are downright insecure if you ask me
13:57 mike25de everyone has to do their job.... is just a matter of perception and interpretation in the end...
13:57 XenophonF very true
13:58 quickdry21 joined #salt
14:00 jayfk joined #salt
14:01 quickdry21 joined #salt
14:04 aquinas joined #salt
14:07 troyready joined #salt
14:11 oz_akan joined #salt
14:12 dccc_ joined #salt
14:13 rnr9 joined #salt
14:14 rnr9 The jinja template replaces "on" to "True" when applied. is there an option to prevent it?
14:17 englishm joined #salt
14:17 englishm joined #salt
14:17 xintron Is there any way to not have the salt-master timeout (I want it to wait for the result from all minions)?
14:17 xintron salt -t 0 '*'?
14:20 rypeck joined #salt
14:21 felskrone joined #salt
14:21 schristensen joined #salt
14:21 Micromus joined #salt
14:22 toastedpenguin joined #salt
14:24 ajolo joined #salt
14:26 peters-tx joined #salt
14:26 jensnockert joined #salt
14:27 mr_chris joined #salt
14:28 rnr9 answer to my own question: parameter (not sure what's it called correctly) has to have single quotes: 'on'
14:28 rnr9 Otherwise jinja will replace it with True (or False for off)
14:28 SheetiS rnr9: correct.  It's a yaml thing
14:29 rnr9 Is there a way to prevent it?
14:29 SheetiS do as you said (place it in single quotes).
14:29 rnr9 because it's much artificial intelligence
14:30 rnr9 pillar file becomes clumsy then, somewhere quotes are must, somewhere not
14:30 pdayton joined #salt
14:31 ajprog_laptop joined #salt
14:31 SheetiS I tend to always use quotes in those cases.  There are some other things that will require it as well.  For example setting a 4-digit permission on a file (e.g '0644') that has a leading zero so it doesn't get parsed as a number and lose the 0.
14:32 SheetiS These can be attributed to 'quirks' in the yaml specification and would be outside the scope of a fix inside saltstack.
14:36 jalbretsen joined #salt
14:36 rnr9 just curious, is it documented somewhere in salt?
14:38 rnr9 http://docs.saltstack.com/en/latest/topics/development/conventions/formulas.html - is this a good place for such sorts of single quote quirks single quote?
14:39 yomilk joined #salt
14:40 ndrei joined #salt
14:41 babilen rnr9: I would rather expect to find it in http://jinja.pocoo.org/docs/dev/templates/
14:44 Ozack1 joined #salt
14:45 rnr9 yeah, I think you're right. I always forget jinja is not native to salt
14:46 calvinh joined #salt
14:46 Guest2369 joined #salt
14:48 vejdmn1 joined #salt
14:51 hardwire joined #salt
14:52 calvinh_ joined #salt
14:53 esogas_ joined #salt
14:55 esogas left #salt
14:56 XenophonF not just jinja
14:56 XenophonF it's YAML too
14:57 XenophonF http://www.yaml.org/spec/1.2/spec.html
14:58 XenophonF The YAML you write in SLS files is a textual representation of the data structures Salt uses.
14:58 XenophonF or in some cases, the code that Salt executes
15:00 XenophonF like, when you do some_state: file.managed: etc., that's effectively a function call, only it isn't written like a funcall in your average ALGOL-like programming language
15:01 econnell joined #salt
15:02 XenophonF so don't think of the SLS files as purely textual config files
15:02 XenophonF they aren't
15:02 XenophonF when you write an SLS, you're programming
15:03 workingcats joined #salt
15:03 XenophonF that's why 0644 turns into a number, but '0644' is a string
15:05 aquinas joined #salt
15:06 logix812 joined #salt
15:08 esogas_ joined #salt
15:09 dfinn joined #salt
15:09 XenophonF speaking as a Lisper, I think the saltstack devs should just take this to its logical conclusion and use S-expressions for SLS files ;)
15:09 dfinn has 2014.1.10 been pretty stable for people?
15:10 hobakill dfinn: i like it more than any other release i've used
15:11 hobakill dfinn: but i stated with 2014.1.4 which was ...not great....
15:11 dfinn ah, interesting that you say that
15:11 dfinn that's what we are on now and it has been good for us
15:11 dfinn I was thinking it's probably time to think about updating but I was hesitant because 1.4 has been good to us
15:12 alanpearce joined #salt
15:12 hobakill dfinn: man...you're a rarity i think. i think the general consensus in was 2014.1.4 was a buggy release. but i'm certainly no salt release expert.
15:13 hobakill dfinn: i can't type for shit today. :(
15:13 dfinn appreciate the feedback, gonna upgrade our test env today I think
15:14 hobakill dfinn: for us - things like test.ping and test.version take like 1/2 a second now where as with 2014.1.4 it was a solid 5seconds of lame
15:14 rodo joined #salt
15:14 rodo hi
15:14 ajprog_laptop joined #salt
15:14 alanpear_ joined #salt
15:14 dfinn yeah, we definitely aren't seeing that here
15:15 rallytime joined #salt
15:15 XenophonF LOL @ "5 seconds of lame"
15:16 hobakill XenophonF: didn't demo well for CTO and SVP-level folks. :)
15:17 rodo FYI i'm the author of last 2 issues #15620 and #15621 come here to discuss with whom who wants
15:18 tinuva joined #salt
15:20 XenophonF I bet it didn't, hobakill.
15:20 babilen XenophonF: +1 on s-expressions -- https://github.com/edn-format/edn
15:23 babilen rodo: I don't think there is any need for a "pre-PR discussion". If there are issues that might have to be addresses they can very well be discussed within the context of your PR.
15:23 babilen *addresses
15:23 babilen argh
15:23 babilen hobakill: You infected me with your typing!
15:23 conan_the_destro joined #salt
15:23 rodo babilen: Ok, I just try to follow the Contribution guide and inform other before doing the PR, so I won't do next time
15:24 hobakill babilen: yeah sorry. i've been working since 5 CDT... brain is starting to slip already
15:24 dvestal joined #salt
15:25 babilen rodo: Oh, is that being asked for in the contribution guidelines? I thought the procedure was "fork, branch, push, PR, discuss, merge" (simplified)
15:28 rodo babilen: https://github.com/saltstack/salt/blob/develop/Contributing.rst =>  Before starting on a new feature an issue should be filed for it. , but may be I misunderstood (I'm french), but it's ok I'll do my PR now, thanks for your explanation
15:30 rodo babilen: my PR https://github.com/saltstack/salt/pull/15623 hope the Team found it usefull
15:30 SheetiS joined #salt
15:32 babilen rodo: http://docs.saltstack.com/en/latest/topics/development/contributing.html is probably what you want to read and base your workflow on
15:33 higgs001 joined #salt
15:34 z0rc joined #salt
15:34 rodo babilen: thanks for that, I'll read
15:36 CeBe joined #salt
15:38 babilen rodo: Is the s/create/update switch really necessary for the connlimit functionality? (If you simply renamed it to be in line with "def user_update" → "postgres.user_update" (and not → "postgres.user_create") I'd rather see that in a different PR.
15:39 UtahDave joined #salt
15:39 oz_akan joined #salt
15:40 rodo babilen: yes, I just fix the doc by the way, may be It's more clear to do a second PR for that
15:41 TheThing joined #salt
15:41 babilen rodo: I think that would be better, yeah. (already commented on GH)
15:44 XenophonF babilen: saw the EDN page, just not the same without the parens :-/
15:46 babilen rodo: Did you rebase your branch on develop before (or what exactly is/was a5d7e18 (Merge to upstream))
15:47 rodo babilen: I made a mistake on rebase, is the reason why I prefer to close this PR and made a new one with rebase
15:47 babilen XenophonF: What do you mean? Clojure is valid EDN (for example)
15:48 babilen rodo: Wonderful, thank you! :)
15:48 rodo ur welcome
15:48 estherbester joined #salt
15:49 babilen XenophonF: I mean (...) are just lists and it is mere convention to treat the first element in a list somewhat special (i.e. function))
15:49 estherbester joined #salt
15:49 babilen But you would know that.
15:49 * babilen ♥ EDN
15:50 babilen But this is getting rather off-topic. /p
15:52 XenophonF heh
15:52 TheThing hehe
15:54 XenophonF babilen: I am enough of a Lisp nerd that I used to sign my name #\XenophonF (because I'm a character).
15:54 XenophonF anyway...getting back on topic...
15:57 ndrei joined #salt
15:57 tligda joined #salt
15:59 catpig joined #salt
16:00 darkelda joined #salt
16:03 kingel joined #salt
16:03 thayne joined #salt
16:03 racooper joined #salt
16:05 spookah joined #salt
16:06 iggy how do you guys normally go about figuring out who's the point of contact for a formula? or do you just submit a pull request and hope someone pulls it?
16:08 vejdmn joined #salt
16:09 ericof joined #salt
16:09 KyleG joined #salt
16:09 KyleG joined #salt
16:10 wendall911 joined #salt
16:11 kaptk2 joined #salt
16:14 Tahm joined #salt
16:17 aparsons joined #salt
16:22 apergos Ahlee: just so you know, that fix wasn't it, sadly. Guess I'll beat on the problem some more
16:22 Jahkeup joined #salt
16:23 UtahDave iggy: Just go ahead and open a pull request.  That's a great place to open a discussion on a potential change.
16:24 UtahDave iggy: if you don't hear anything in a day or so, feel free to ping us in here.  :)
16:25 ggoZ joined #salt
16:25 brandon__ joined #salt
16:27 metaphore joined #salt
16:30 mechanicalduck_ joined #salt
16:31 aparsons joined #salt
16:32 mapu joined #salt
16:36 higgs001_ joined #salt
16:38 swilcox joined #salt
16:39 higgs001 joined #salt
16:42 mechanicalduck joined #salt
16:43 iggy UtahDave: okay, I had a couple of questions about direction, but I guess GH is as good a place to have that discussion as any
16:44 UtahDave Yeah, plus then the people who are following that repo will be notified and can chime in if they have an opinion
16:47 schimmy joined #salt
16:48 gngsk joined #salt
16:49 schimmy1 joined #salt
16:49 * robawt highfive UtahDave
16:50 blackhelmethacke joined #salt
16:50 giantlock joined #salt
16:54 UtahDave hey, robawt
16:54 robawt hey hey
16:54 robawt you guys keep adding goodies to 2014.07
16:55 robawt i am digging the check command requisite and the gpg renderers
16:55 UtahDave :)
16:56 martoss joined #salt
16:57 chrisjones joined #salt
16:57 giannello is there a ZooKeeper module/state planned?
16:57 martoss1 joined #salt
17:00 forrest joined #salt
17:02 Daviey joined #salt
17:02 mattmtl joined #salt
17:02 penguin_dan joined #salt
17:03 ndrei joined #salt
17:03 kingel joined #salt
17:06 Gareth morning morning
17:06 forrest hey Gareth
17:06 murrdoc joined #salt
17:07 Gareth hey forrest, hows it going?
17:07 forrest Gareth, pretty good, you?
17:08 Gareth not bad.  just arrived at work, some big announcement today...which should be interesting.
17:08 KyleG do you work for…apple? lol
17:09 Gareth KyleG: I do not.  Won't be that interesting :)
17:13 forrest New announcement Gareth is never working on-call again!
17:14 murrdoc \o/
17:15 Gareth forrest: haha highly unlikely.  I think it's more likely about the vaccant VP position and whose going to fill it.
17:15 debian112 joined #salt
17:16 eriko joined #salt
17:18 Ryan_Lane joined #salt
17:19 Ahlee apergos: :( sorry to hear that
17:19 chrisjones joined #salt
17:19 kermit joined #salt
17:19 rap424 joined #salt
17:19 kermit joined #salt
17:20 shaggy_surfer joined #salt
17:20 apergos well 0.17.1, it's an old version, who knows what things might have been fixed in the meantime
17:20 eunuchsocket joined #salt
17:20 apergos but I can break it on a single client so I'm just doing my testing there, which is a relief
17:20 eunuchsocket joined #salt
17:23 sijis_ left #salt
17:29 pdayton joined #salt
17:29 otter768 joined #salt
17:30 Setsuna666 joined #salt
17:32 midacts `salt-key -L` lists the salt-minion, and i added the salt-minion on the salt-master by running `salt-key -a hostname`. but when i go to run `salt '*' test.ping` that host doesnt respond. i have tried restarting the salt-minion service and rebooting that host. Does anyone else have any ideas?
17:32 sm1ly_ joined #salt
17:33 Test joined #salt
17:33 UtahDave midacts: do you have other minions that are working?
17:34 midacts UtahDave: Yes sir, 6 of the 9 minions I have are working.
17:34 dalexander joined #salt
17:35 UtahDave what version of salt on the master and on the minion?
17:36 midacts hmm, looks like the salt service isnt running?  ( salt-master 2014.1.10+ds-1~bpo70+1 )  (  salt-minion 2014.1.10+ds-1~bpo70+1 )
17:36 alanpearce joined #salt
17:36 midacts ps -A | grep salt* doesnt return anything on those servers (that `salt '*' test.ping` is not working on)
17:38 notpete__ joined #salt
17:38 gngsk midacts: what distro is the minion running?
17:39 dalexander I'm testing out halite (0.1.17) on my salt master installed with pip on OSX.  The live doc search feature does not work.  It's showing up in the gui but is not selectable.  should it be working in this version?  is there further configuration required to enable it?
17:40 midacts gngsk: Debian 3.2.60-1+deb7u3 x86_64 GNU/Linux
17:45 rjc joined #salt
17:46 jalaziz joined #salt
17:52 swilcox joined #salt
17:52 aw110f joined #salt
17:52 obimod joined #salt
17:52 midacts Had to delete the salt-minion's key on the master and then restart the salt-minion service and then re-accept the problem salt-minion's key   (http://paste.ubuntu.com/8301814/)
17:53 elfixit joined #salt
17:56 mr_chris UtahDave, Any word on if https://github.com/llamallama/reverse-grains-formula and https://github.com/llamallama/reverse-users-formula could be included into the salt formulas repo?
17:56 eagen joined #salt
17:56 mr_chris If not, is there anything I could change or do better about them?
17:56 obimod joined #salt
17:56 forrest mr_chris, ahh sorry about not getting back to you
17:57 forrest mr_chris, I confirmed Seth was out at training this week, so we won't hear from him. Let me make some repos
17:57 UtahDave cool, thanks, forrest!
17:57 mr_chris forrest, Thanks!
17:57 forrest UtahDave, mr_chris,  yea sure, should have just done it in the first place. :P
17:57 mr_chris I'm curious, do people typically avoid the py rendered states.
17:58 mr_chris ?
17:58 forrest mr_chris, well we don't have any formulas that use them, because the formulas are mostly 'intro' stuff
17:58 forrest that people can build off of, so we try to keep them simple
17:58 dusel joined #salt
17:58 UtahDave mr_chris: py rendered states can be very powerful, but I usually recommend keeping things simple unless it's really necessary
17:58 mr_chris Makes sense.
17:59 mr_chris UtahDave, You've seen that mess of a jinja MySQL user manager state I built.
17:59 mr_chris py really cleaned that up.
17:59 mr_chris I should probably submit that too.
17:59 jhauser joined #salt
17:59 forrest mr_chris, https://github.com/saltstack-formulas/reverse-grains-formula and https://github.com/saltstack-formulas/reverse-users-formula
17:59 esogas_ joined #salt
17:59 UtahDave mr_chris: cool!
17:59 mr_chris forrest, You've made my day!
17:59 mr_chris Thanks again.
17:59 forrest mr_chris, can you please fork those two, do the upstream for yours, modify the README to be RST and normal like the other formulas, and copy over the license?
18:00 obimod joined #salt
18:00 forrest mr_chris, yeah np, I was trying to look for you when I got home a few nights ago, but couldn't remember who it was that I had talked to, lol
18:00 mr_chris forrest, lol. NP. I'm patient and persistent.
18:00 forrest mr_chris, cool, let me know when you've put in the merge request and I'll merge them
18:01 obimod joined #salt
18:01 pdayton joined #salt
18:01 mr_chris Merge request for?
18:01 mr_chris OH.
18:01 mr_chris Right.
18:02 druonysus joined #salt
18:02 forrest yeah you have to fork, push, and then do a merge request, can't push straight to it
18:02 UtahDave dalexander: are your master and minions the same version?
18:03 devx joined #salt
18:04 kingel joined #salt
18:06 dalexander UtahDAve: my minion is slightly newer actually  salt-master 2014.1.4, salt-minion 2014.1.7
18:06 UtahDave dalexander: Is your master on OSX?
18:07 dalexander UtahDave: yes it is
18:07 UtahDave OK, I was just told that there was some wonkiness on OSX with the manage.present function, which apparently the live docs feature depends on
18:08 UtahDave He thinks that using another OS for the master should fix that.  Like Ubuntu or CentOS
18:08 dalexander that's fine.  I can change to a linux master, this is just my dev setup.
18:08 UtahDave let me know if that works for you. The live docs feature is pretty cool.
18:08 dalexander UtahDave: I know I want some of that.  I'll let you know.  Thanks!
18:08 jrb28 joined #salt
18:08 jubei__ left #salt
18:09 kermit joined #salt
18:09 heise joined #salt
18:11 nkuttler i get a http://dpaste.com/0SKTWE6 from a salt minion running debian sid. i'm not sure if this a sid bug of if salt's module reload thing is to blame
18:12 mr_chris forrest, Pull requests created.
18:14 kballou joined #salt
18:15 forrest mr_chris, cool, merged
18:15 mr_chris forrest, Many thanks. I hope someone finds them helpful.
18:15 forrest I wish gitlab would show PRs in addition to stars/forks on the main page in a list of repos...
18:15 forrest mr_chris, yeah np
18:17 forrest err github
18:23 druonysuse joined #salt
18:23 druonysuse joined #salt
18:23 ksalman has anyone tried creating thousands of users? I have to create 5000+ and it takes over 60 minutes :S
18:24 ksalman i don't know if any thing could be done about it?
18:24 mr_chris ksalman, Creating or keeping current?
18:25 dfinn that time doesn't seem that unreasonable, is that something you are doing on a regular basis?
18:25 ksalman mr_chris: I have to do both. Just creating took over 60 minutes
18:25 ksalman dfinn: Yea
18:25 mr_chris ksalman, You could rule out if it's a salt issue or not by making a bash loop that creates 5000+ users.
18:25 ksalman just updating takes 35 minutes
18:25 TTimo joined #salt
18:25 mr_chris Oh wait, 60 minutes.
18:26 mr_chris I read seconds.
18:26 forrest ksalman, you're creating 5k users per box?
18:26 ksalman forrest: yes
18:26 forrest ksalman, you can't use *shudders* ldap or AD or something?
18:26 ksalman forrest: i wish =)
18:26 forrest ksalman, or better yet, pack all the users into the image you use to build systems if they aren't changing that often
18:26 forrest but man, 5k people who can log into a system, sounds like a security nightmare
18:27 ksalman forrest: it's a lot of test machines for qa folks
18:27 mr_chris ksalman, Man this sounds familiar.
18:27 mr_chris Just not on this scale.
18:27 ksalman =(
18:27 forrest ksalman, but why are they local users on the system? If you're slamming an app shouldn't they be in some sort of QA database or something that the application accesses?
18:28 forrest ksalman, honestly though if they are test users, screw using Salt to add them, add them to the image (docker, lxc, etc) and use that golden image to build your machines off of.
18:28 ksalman salt calls useradd and usermod for each user and i don't know how else it could do it to make it faster
18:28 mr_chris ksalman, Still, maybe you should try the bash loop thing just to get a base line. Still salt has to check for the user, compare the results with what the state says it should expect., and apply changes if not. That's a lot of work being done.
18:29 forrest ksalman, that's what I'm saying, I don't know if it could be faster, but a LOT of that has to be compile time, thus why I'm suggesting to have an image that contains them all, then configure with salt from there.
18:29 forrest ksalman, also, have you tried running the state locally?
18:29 forrest ksalman, minion solo is faster than the master.
18:29 ksalman forrest: they are not test users. they are accounts for all the qa folks on a bunch of test machines
18:29 forrest ksalman, oh ok
18:30 forrest ksalman, I'd give it a shot with salt-call --local state.sls 'state_name'
18:30 forrest so you bypass the master completely (unless you are already doing so)
18:30 ksalman mr_chris: i'll try the bash loop and see how fast that is
18:30 forrest to see if it improves times
18:31 ksalman forrest: i have not tried --local
18:31 forrest ksalman, give it a shot, that will force it to use local state/pillar data, and not contact the master at all
18:31 forrest so obviously you'd need the content on the minion, but it's worth trying
18:31 ksalman can i copy the stale_name via salt and call salt-call --local from that state? :)
18:32 forrest uhh what do you mean?
18:32 forrest you treat it just like normal, it just uses content from the local /srv/salt and /srv/pillar dirs
18:32 eunuchsocket Hi all.  I'm struggling to figure out why I'm getting this error: http://pastebin.com/ftLHjD6L when running this state: https://github.com/emuehlstein/snmp-formula
18:32 forrest oh you mean copy it over? Sure I don't see why not, lol
18:32 ksalman i mean, run a salt state from the master which just copies the state over, then the same state does a command.run for salt-call --local
18:32 ksalman lol
18:33 forrest eunuchsocket, should be snmp.options.source
18:33 forrest eunuchsocket, and snmp.options.snmpdoptions
18:34 eunuchsocket forrest: I'll give that a try thanks.
18:34 forrest eunuchsocket, you need to go all the way down to the variable you want, with the way you have it now it's looking for options: \n source: value
18:34 forrest eunuchsocket, nop
18:34 forrest *np
18:34 eunuchsocket forrest: I'm curious why I don't get the same error for the 'config.source' in the preceding stanza
18:34 cnelsonsic joined #salt
18:35 forrest eunuchsocket, yaml rendering is really weird, both of them do need to be fixed though
18:36 eunuchsocket forrest: that didn't work
18:36 forrest eunuchsocket, ok, maybe the state is formatted differently then, I haven't looked at the whole formula
18:36 mr_chris ksalman, I may have an alternative if you're allowed to introduce new tech into your QA setup.
18:37 eunuchsocket forrest: I just forked this yesterday... that specific SLS starts with a bunch of "{%- from "snmp/server/lib.sls" import options with context -%}" which I haven't seen before
18:37 mr_chris ksalman, FreeIPA. Dead simple LDAP+Kerberos based authentication setup.
18:37 forrest eunuchsocket, yeah I'm looking at your fork now, I see that's a new state.
18:37 mr_chris Three commands and you're authenticating.
18:37 higgs001 joined #salt
18:37 ksalman mr_chris: thanks! i'll check it out
18:37 eunuchsocket forrest: trying to separate out the install and config per best practices
18:37 mr_chris ksalman, Also has a web interface for the whole thing.
18:38 forrest eunuchsocket, yeah for sure, this formula definitely could use some updates, I've got to run to lunch since my coworkers are waiting on me, if you figure it out can you let me know? Maybe try writing a very simple state that references that pillar data
18:38 forrest otherwise we can work on it a bit when I get back
18:38 eunuchsocket forrest: you bet.  Thank you
18:38 forrest eunuchsocket, np
18:40 mattmtl joined #salt
18:41 AdamSewell joined #salt
18:43 iggy when using merge in a map, the value from the pillar should override the map, correct?
18:43 jut joined #salt
18:43 bhosmer joined #salt
18:44 dalexander what's a good example of making pillars available at /srv/pillar when pillars are stored in git, where the pillar directory is off the root of the of the git repo?  this is what I'm starting with:
18:44 dalexander ext_pillar:
18:44 dalexander - git: master https://stash./******.com/scm/sys/salt.git root=pillar
18:47 n8n joined #salt
18:47 ajolo joined #salt
18:48 martoss joined #salt
18:48 esogas left #salt
18:48 metaphore joined #salt
18:49 martoss1 joined #salt
18:51 rojem joined #salt
18:54 esogas_ joined #salt
18:56 ipmb joined #salt
18:57 TheThing joined #salt
18:57 Me joined #salt
18:58 n8n_ joined #salt
19:00 grove_ joined #salt
19:01 rnr09 joined #salt
19:01 murrdoc joined #salt
19:02 rnr09 Is there a way to get a nicely formatted list of servers that satisfy some requirements (for example have a string in the conf file)?
19:02 rnr09 Of course it can be done with "salt -C '* ' cmd.run ' grep servera outputs.conf ' " - but output is not really pretty and I'm not sure how to filter or aggregate it.
19:03 eunuchsocket rnr09: write a custom grain?
19:04 younqcass joined #salt
19:05 kingel joined #salt
19:06 hobakill by default does PCRE ignore case in a compound match?
19:09 martoss joined #salt
19:11 martoss1 joined #salt
19:12 mr_chris hobakill, Just tested. Yes.
19:12 hobakill mr_chris: was just doing that myself. thanks.
19:13 blackhelmethacke joined #salt
19:14 smcquay joined #salt
19:16 dvestal joined #salt
19:16 grove__ joined #salt
19:17 nitti_ joined #salt
19:18 felskrone joined #salt
19:21 Pixionus joined #salt
19:21 adizam joined #salt
19:24 satya_ joined #salt
19:24 patarr how's salt support for network gear?
19:24 patarr Puppet has JunOS integration and such
19:24 murrdoc hook it to trigger
19:24 murrdoc and pew pew
19:24 picker joined #salt
19:25 goodwill salt does it exactly with that sound
19:25 goodwill pew pew
19:25 patarr coooool
19:25 ajprog_laptop2 joined #salt
19:26 nmadhok joined #salt
19:27 ajprog_laptop joined #salt
19:28 shaggy_surfer joined #salt
19:29 AdamSewell joined #salt
19:37 jalaziz joined #salt
19:37 babilen patarr: I can think of http://docs.saltstack.com/en/latest/topics/topology/proxyminion/index.html
19:38 patarr thanks babilen
19:38 dalexander UtahDave: just installed halite on linux and the livedoc search is working.  just GYI.
19:38 dalexander FYI
19:38 manfred patarr: we are working on expanding the support for thouse, but just haven't goetten to them yet
19:38 babilen patarr: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.junos.html is the referenced module in there
19:39 manfred I am trying to get a BigIP F5 so I can write a proxyminion for it
19:41 UtahDave dalexander: ah, great! thanks for the update
19:44 patarr you guys should mooch some of the kibana devs. Get some sexy angular.js action going on in halite
19:46 jhauser joined #salt
19:49 delinquentme joined #salt
19:50 ckao left #salt
19:50 Ryan_Lane +1 on angularjs
19:51 ksalman bash loop to create 5000 users take 8m
19:51 ksalman ^ forrest: mr_chris
19:52 Ryan_Lane UtahDave: https://github.com/saltstack/salt/issues/13146 <-- is there any way this can be fixed for 2014.7? It's a pretty major issue for ordered runs.
19:53 mr_chris ksalman, Quite a difference.
19:53 ksalman yea
19:54 mr_chris I'd be curious to know if the running the states locally is any faster.
19:54 ksalman can i even run the state locally if the state is a python sls?
19:54 mr_chris Should be able to.
19:55 mr_chris I'm assuming you should be able to.
19:55 mr_chris I'd consider it a bug otherwise.
19:56 delinquentme 2014-09-09 17:02:23,857 [salt.utils.event ][DEBUG   ] LocalClientEvent PULL socket URI: ipc:///var/run/salt/master/master_event_pull.ipc << Whats this ipc about??
19:56 ksalman either that or i am running it incorrectly
19:56 ksalman shouldn't it be salt-call --local state.sls testusers?
19:57 ksalman No matching sls found for 'testusers' in env 'base'
19:57 hop do i have access to pillar data from a execution module?
19:57 hop i'm looking at /ref/modules/index.html
19:58 hop is there more documentation?
19:58 delinquentme This is an event where its pulling from the salt-master file server right?
20:00 eunuchsocket Anyone else see a solution to this error
20:00 eunuchsocket :http://pastebin.com/U0gJKe8c running this formula: https://github.com/emuehlstein/snmp-formula
20:01 grove_ joined #salt
20:01 eunuchsocket I've been poking around for an hour and I'm no closer to find the root cause
20:01 forrest eunuchsocket, sorry came back and had to deal with a few things
20:01 forrest eunuchsocket, I looked at it
20:01 forrest eunuchsocket, you're trying to use pillar data and I was thinking incorrectly
20:01 eunuchsocket but I haven't actually set any pillars
20:01 aparsons joined #salt
20:01 nmadhok Pedro you here?
20:02 eunuchsocket forrest: I'm trying to validate that the defaults are working before I set an pillars
20:02 forrest nmadhok, he's not, you can message him directly at s0undt3ch.
20:02 ckao joined #salt
20:02 forrest eunuchsocket, yeah, so those values are using information from the map files
20:02 aparsons joined #salt
20:02 eunuchsocket forrest: I want them to use the values from lib.sls
20:02 nmadhok thanks forrest!
20:02 forrest eunuchsocket, is lib.sls working?
20:03 forrest nmadhok, np, he works nights, and lives in Portugal so he might not be around for a bit, but he'll hit you up
20:03 eunuchsocket forrest: it is for the rest of the references (ie. config.source)
20:04 eunuchsocket forrest: I took the file.manage stanza for the config file and tried to follow the same model to manage the options file
20:05 mr_chris ksalman, Have you properly set up the state tree locally?
20:05 ksalman mr_chris: oh, I guess I have not. Didn't know i needed to do that
20:05 forrest eunuchsocket, yeah but look at https://github.com/emuehlstein/snmp-formula/blob/master/snmp/server/lib.sls#L22
20:05 kingel joined #salt
20:06 mr_chris ksalman, I think the state itself has to reside locally on each box. http://docs.saltstack.com/en/latest/topics/tutorials/quickstart.html
20:06 mr_chris I've personally never tried it.
20:06 ksalman okay
20:07 XenophonF joined #salt
20:07 forrest eunuchsocket, are you on rhel 5?
20:07 forrest or 6?
20:07 mr_chris I think you may be pushing salt's limits ksalman.
20:07 eunuchsocket forrest: testing on 6 at the moment
20:07 forrest ok, so snmp.options drops in the name from the map.jinja...
20:08 eunuchsocket why doesn't the same thing happen with 'config'?
20:08 forrest and options.source should be working as well, hmm
20:09 jalaziz_ joined #salt
20:09 eunuchsocket forrest: admittedly things get a little bit confusing in this formula with all the name duplication
20:09 forrest I agree, it's not a formula from the repo so I can't really complain
20:10 forrest eunuchsocket, oh man I'm a moron
20:10 forrest eunuchsocket, you can't have two file.* under a single ID
20:10 eunuchsocket forrest: lol, not as much of a moron as me :-P
20:10 forrest for your configured.sls
20:10 forrest you have two managed files under snmp.server.configured
20:10 eunuchsocket forrest: oooooooooooooohhhh
20:10 eunuchsocket forrest: simplest answer :-D
20:10 forrest eunuchsocket, hopefully that fixes it
20:10 forrest we'll see.
20:14 ekristen how does salt-call talk to the minion to talk to the master to get stuff done? is it though the IPCs?
20:14 delinquentme so using salt-cloud is outdated right? As this is currently how Im throwing up instances ...
20:14 forrest delinquentme, salt-cloud is now in mainline salt
20:14 delinquentme and I've got an issue associated with that command =]
20:14 forrest ekristen, salt-call is a command available on minions where they then make requests to the master.
20:15 forrest ekristen, unless you pass --local, which then makes calls locally
20:15 ekristen forrest: I get that, but I’m curious as to how it communicates with the minion
20:15 ekristen TCP, Unix socket etc
20:15 UtahDave ekristen: salt-call does not talk to the locally running salt-minion
20:15 forrest ekristen, over the established zeromq connection
20:15 ekristen UtahDave: interesting, so does salt-minion have to be running for it to work?
20:15 UtahDave ekristen: over zeromq
20:16 UtahDave ekristen: nope. salt-minion does not need to be running for salt-call to work
20:16 bhosmer joined #salt
20:16 ekristen I’m assuming it needs access to the right config files though to work?
20:16 ekristen I
20:16 ekristen I’m working on running salt-call from inside a docker container to trigger an event on the master, but I don’t want to run the salt-minion in the container, so just trying to figure out what I need to make it work
20:17 UtahDave patarr: that would be cool!  halite does use angular
20:18 UtahDave Ryan_Lane: I'll see who's on that issue
20:18 UtahDave ekristen: yeah, it needs to access /etc/salt/minion and probably /var/cache/salt/*  and a few other things
20:19 ekristen hrm ok
20:19 dfinn left #salt
20:19 delinquentme forrest, so the salt-cloud package should be removed completely and running everything with salt.
20:19 eunuchsocket forrest: I split out the second managed file and still no love
20:19 forrest eunuchsocket, in that same state?
20:19 eunuchsocket nope, created a new state called 'optioned'
20:19 forrest delinquentme, if you're on a newer release yes, the salt-cloud command will still be available.
20:19 otter768 joined #salt
20:20 eunuchsocket forrest: and pulled out all the option setting stuff to that state
20:20 forrest eunuchsocket, so the other state you have with just the first item now runs correctly though
20:20 eunuchsocket yes
20:20 forrest but same error on the other one?
20:20 eunuchsocket forrest: I'm getting the same error on the new state
20:20 delinquentme Ohhhhhhh alright
20:20 forrest I don't get it, you're import the options with context, and the first item, also imported with context works fine
20:20 forrest there must be something obvious we are missing, lol
20:21 eunuchsocket forrest: are there any builtin dicts or anything like that which would conflict with the name 'options'?
20:21 forrest eunuchsocket, I don't know
20:21 forrest maybe try changing it?
20:21 forrest in that lib file, and then the import?
20:21 eunuchsocket forrest: I'm going to try giving things more unique names to make sure there aren't any conflicts.  I'll let you know how it goes
20:22 forrest eunuchsocket, that sounds like a good plan
20:25 CeBe1 joined #salt
20:25 chrisjones joined #salt
20:27 eunuchsocket forrest: holy cow.  I can't explain it but it looks like the error was actually coming from the line above the one indicated.
20:28 forrest sometimes that happens
20:28 forrest so it was an issue with the name value then?
20:28 eunuchsocket forrest: and ultimately not having the right 'name' set in the map.jinja
20:28 forrest ahh alright
20:28 forrest that's good
20:29 eunuchsocket forrest: I guess I need to read up on mapping.  I don't understand why it works when I put the options file name under the 'RedHat' grain but doesn't work under the '
20:30 eunuchsocket Red Hat Enterprsie Linux Server-6' grain.
20:30 eunuchsocket forrest: anyways, thanks for the help
20:30 forrest eunuchsocket, probably because it's not being detected as that. list the grains and see what the OS grain returns
20:30 forrest eunuchsocket, I did a talk on salt formulas and states if you want the basics: https://www.youtube.com/watch?v=OCbtD2r7bPo
20:31 eunuchsocket forrest: osfinger: Red Hat Enterprise Linux Server-6
20:31 eunuchsocket forrest: the way I read map.jinja I expected it to match on any grain not just OS
20:32 bluenemo_ joined #salt
20:32 forrest eunuchsocket, no, by default grains.filter_by uses the 'os_family' grain
20:32 forrest so that's probably why it didn't work correctly
20:32 eunuchsocket forrest: sure enough.  I made a bad assumption there.
20:32 forrest oh well, you figured it out
20:33 eunuchsocket yeah, time to git reset and fix the real problem
20:33 forrest heh
20:33 forrest yeah that original formula needs some work as well
20:33 forrest I've never talked to the author I don't think
20:33 forrest I'm surprised we don't have an official snmp formula
20:34 bluenemo_ hi guys. How can I just print a string on the master executing highstate? i dont want to cmd.run or anything.. like print in python.. is that possible?
20:34 timbermaniac joined #salt
20:34 eunuchsocket forrest: would you suggest I start from scratch using the template-formula?
20:35 forrest eunuchsocket, I probably would
20:35 forrest there's a LOT of usage of set and other stuff in there
20:35 forrest I just don't know if it's necessary.
20:36 forrest eunuchsocket, I'd be happy to make an snmp formula repo once you're done and confident it's working so you can make a PR against it and we can have an official one
20:36 wt joined #salt
20:36 wt Is it possible that the s3fs downloads a file every time it updates?
20:37 timbermaniac wt: from my experience, it does that.
20:37 wt I am running on 2014.1.10 on amazon linux
20:37 wt Really, WTF?
20:37 eunuchsocket forrest: that sounds great.  I just started with salt a couple days ago so I'm still getting my feet wet but it seems like low hanging fruit and a relatively easy contribution.
20:37 wt it doesn't look at the md5s first?
20:38 forrest eunuchsocket, Yeah there are a LOT of example formulas, as well as some great docs that whiteinge wrote: http://docs.saltstack.com/en/latest/topics/development/conventions/formulas.html#writing-formulas
20:38 forrest eunuchsocket, so between that, the presentation, and a look at other formula docs you should be set to go.
20:39 timbermaniac Can somebody help me get GitFS working? I'm always getting "No Top file or external nodes data matches found" when I highstate a minion, making me thing it's not configured correctly
20:39 dalexand_ joined #salt
20:39 dvestal joined #salt
20:40 iggy is your top file at the root of your git repo?
20:41 bhosmer joined #salt
20:41 timbermaniac It's under salt/
20:42 wt Ok, so it looks like the md5 is in the s3cache/buckets_files.cache file
20:42 timbermaniac and 'gitfs_root: salt/' is set
20:42 iggy timbermaniac: did you set gitfs_root?
20:42 wt I don't know if that is an mdt
20:42 wt md5, that is
20:42 wt but is looks like one
20:42 KyleG joined #salt
20:42 KyleG joined #salt
20:42 englishm joined #salt
20:43 timbermaniac iggy: 'gitfs_base: develop' and 'gitfs_remotes: - git+ssh://<remote>.git'
20:43 wt Salt is downloading the same 80MiB file every 30s or so from s3.
20:44 martoss joined #salt
20:44 wt that's kinda lame
20:45 iggy timbermaniac: can you clone the repo as the user running salt?
20:45 timbermaniac iggy: yup
20:46 iggy try running the master with -l debug maybe?
20:46 martoss1 joined #salt
20:46 timbermaniac iggy: unless running as root just doesn't work (I realize this is bad practice)
20:46 iggy and see if there's anything actually making it's way to /var/cache/salt/master/gitfs
20:46 iggy I think everybody pretty much runs salt as root for now
20:47 timbermaniac oh, there is stuff in that dir
20:48 mrlesmithjr joined #salt
20:48 iggy there should be some hash looking dir names
20:48 timbermaniac you know how when you clone a repo down, it creates a dir with the repo name? that name doesn't need to be included in the gitfs_root, does it? only the subdir if that's how you're doing it
20:49 iggy no, don't include the repo name in the root
20:49 timbermaniac k. so yeah, I did see one directory with a hash, but it's empty
20:49 iggy empty or just has .git dir?
20:50 timbermaniac oh, there is a .git
20:50 aparsons joined #salt
20:50 iggy maybe try taking the trailing / off the gitfs_root?
20:51 iggy I don't think it should matter, but everything else seems like it should be running
20:51 mrlesmithjr joined #salt
20:52 bluenemo_ I'm wondering why all pillar data is listed under recipe:lookup by default. What does the lookup accomplish?
20:53 timbermaniac iggy: same thing. what's the best test to verify if gitfs is working? surely running a highstate can't be the most reliable
20:54 iggy state.show_top
20:55 KyleG1 joined #salt
20:55 quist joined #salt
20:56 timbermaniac doesn't work on the master, returns nothing. the minion is returning 'local: ----------'
20:56 timbermaniac that's a good test, thanks
20:56 iggy should say "local:\n    ----------\n    base:\n"....
20:58 timbermaniac should the master config show 'file_roots:\n  base:\n  - etc'?
20:58 timbermaniac I commented that out due to introducing git for fileserver_backend
20:59 Cidan joined #salt
21:00 iggy it shouldn't matter if you changed fileserver_backend to "- git"
21:00 iggy if it's set to roots and git, it'll default to /srv/salt
21:01 dalexander joined #salt
21:01 rogst joined #salt
21:02 marco_en_voyage joined #salt
21:03 marco_en_voyage left #salt
21:04 ajprog_laptop joined #salt
21:05 kermit joined #salt
21:05 marco_en_voyage joined #salt
21:06 kingel joined #salt
21:07 metaphore joined #salt
21:10 BrendanGilmore joined #salt
21:12 Setsuna666 joined #salt
21:14 timbermaniac iggy: in the documentation (http://salt.readthedocs.org/en/latest/topics/tutorials/gitfs.html) where it says "And each repository contains some files", that doesn't mean /etc/salt/master actually has to have all state files listed, does it?
21:14 jalaziz joined #salt
21:14 aparsons joined #salt
21:15 wt I see what's wrong
21:15 wt S3fs assumes an Etag is an md5 for the file.
21:15 wt But it's not always.
21:17 ksalman forrest: running locally takes the same time when creating 5k users
21:17 wt I don't think there's any way to reliably calculate the etag locally.
21:17 wt I think salt probably needs to write the etag somewhere when downloading the file.
21:18 forrest ksalman, good to know
21:18 forrest ksalman, then I'd say just do a cmd.run
21:18 marco_en_voyage left #salt
21:18 marco_en_voyage joined #salt
21:18 forrest ksalman, and run your bash script if you don't want to create a 'golden' server image
21:18 timbermaniac wt: you are using s3fs as the filesystem for salt backend?
21:18 ksalman yea, i suppose so
21:19 wangofett anyone ever get KeyError: 'file.gid_to_group' ?
21:20 marco_en_voyage left #salt
21:20 marco_en_voyage joined #salt
21:20 shaggy_surfer joined #salt
21:20 wt timbermaniac, yes
21:22 wt actually, awscli seems to know how to sync in this case
21:22 wt let me take a look
21:23 timbermaniac wt: I was originally using s3fs for the same purpose, but found it was quite slow. looking into gitfs instead
21:24 delinquentme is this channel logged?
21:24 Ryan_Lane yes
21:24 Ryan_Lane in two places
21:24 wt I am plunking binaries like a tarball for logstash into s3
21:24 wt didn't seem appropos for git
21:25 wt I will use git if I really need to though.
21:25 Ryan_Lane delinquentme: see the topic for the primary location
21:25 wt I actually use rootfs, s3fs, and gitfs.
21:26 jalaziz joined #salt
21:27 iggy timbermaniac: no, a basic gitfs config shouldn't be more than about 8 lines of config file
21:29 martoss joined #salt
21:31 martoss1 joined #salt
21:31 wt s3fs doesn't work with multipart files
21:33 delinquentme UtahDave, were you the one who gave the random_reauth_delay suggestion?
21:33 delinquentme tried it out -- still hasing issues
21:33 TheDodd joined #salt
21:33 wangofett ah, looks like I had an old version hanging about
21:35 schimmy joined #salt
21:37 pdayton joined #salt
21:42 aparsons joined #salt
21:46 wangofett where are the actual .py states stored on a salt minion?
21:47 wangofett because apparently I"m missing some, I guess
21:48 TaiSHi Hi, I'm trying to use set up a cron state with an * in the cron line, but it's failing
21:48 77CAAEEHU joined #salt
21:48 TaiSHi Tried double and single quotes, with no success
21:49 ajprog_laptop joined #salt
21:49 SunSparc joined #salt
21:49 SunSparc Query: can salt minions listen to more than one master
21:50 TaiSHi Yes they can
21:50 murrdoc left #salt
21:50 murrdoc joined #salt
21:51 SunSparc TaiSHi: Good to know. Do you have a link to any docs on that?
21:52 TaiSHi Sec, let me find it
21:52 UtahDave SunSparc: http://docs.saltstack.com/en/latest/topics/tutorials/multimaster.html
21:53 TaiSHi http://docs.saltstack.com/en/latest/ref/configuration/minion.html#master
21:53 TaiSHi Thanks Dave
21:53 UtahDave also, in the latest develop (maybe 2014.7 RC as well) you don't have to share the keys among your masters
21:54 SunSparc Handy
21:54 SunSparc TaiSHi++ UtahDave++
21:54 SunSparc Thanks
21:54 KyleG joined #salt
21:54 KyleG joined #salt
21:54 TaiSHi OHHHH
21:54 TaiSHi 2014.7 has a RC ?
21:54 TaiSHi packages as well?
21:54 TaiSHi I might fap to that
21:54 murrdoc for ubuntu there isa  ppa:salt-daily
21:55 murrdoc but thats more for develop
21:55 murrdoc not for 2014.7 rc
21:55 TaiSHi Since when is that daily up? I keep cloning on my deploys :P
21:55 murrdoc but the git repo ships with the info needed to build a package
21:56 TaiSHi Ah, my idea was to actually speed up the deplosy
21:56 TaiSHi deploys*
21:56 TaiSHi salt-daily is rather outdated :P
21:58 bhosmer joined #salt
21:58 n8n joined #salt
21:59 nitti joined #salt
21:59 wt awscli uses the filesize and the date to check whether a file should be copied
22:00 delinquentme forrest, im installing salt off of my package manager -- thats probs a bad idea huh?
22:00 forrest delinquentme, huh?
22:01 delinquentme forrest, so I've attempted to run salt-cloud without installing the specific salt-cloud pip package and it says " command not found "
22:01 forrest delinquentme, what version of salt did you install?
22:01 delinquentme 2014.1.10
22:01 forrest hmm
22:02 forrest it's definitely here in 2014.1.10
22:02 forrest https://github.com/saltstack/salt/tree/v2014.1.10/salt/cloud
22:03 forrest delinquentme, is libcloud installed?
22:03 delinquentme weirrrrd
22:03 delinquentme salt-cloud -P -m cloud.map
22:04 delinquentme no go.
22:04 ndrei joined #salt
22:04 forrest Ryan_Lane, am I forgetting something about the salt-cloud command being part of main line salt now?
22:04 forrest Ryan_Lane, delinquentme isn't showing the command as available on 2014.1.10
22:06 aparsons joined #salt
22:07 delinquentme yeap just verified.
22:07 smcquay joined #salt
22:07 Ryan_Lane forrest: I don't know. I don't use salt-cloud :)
22:07 delinquentme specifically this is without the pip installation of the salt-cloud package
22:08 delinquentme Ryan_Lane, you instance servers with another command?
22:08 Ryan_Lane delinquentme: are you using aws?
22:08 Ryan_Lane or some other cloud provider?
22:08 delinquentme rackspace
22:08 Ryan_Lane yeah, you need to use salt-cloud
22:08 delinquentme noted.
22:09 Ryan_Lane it would be nice if salt-cloud eventually supported the code I wrote, too, but it doesn't yet
22:09 Ryan_Lane (I wrote the boto_* states and modules, so aws-only)
22:10 troyready joined #salt
22:10 aw110f joined #salt
22:11 UtahDave Ryan_Lane: I actually brought that topic up in a discussion today. I think a lot of the Salt Cloud functionality ought to be split out into execution modules and states and have Salt cloud use those modules, where possible
22:11 UtahDave then it would be easy to use your boto_* states and modules
22:12 Ryan_Lane yeah, doing that would mean you could manage resources in environments with no master
22:12 iggy any way to list the environments salt thinks is in the pillar?
22:12 n8n joined #salt
22:13 iggy I set the pillar_ext git branch to something other than master and all hell has broken loose
22:16 wt looks like salt would need to store the etag somewhere when fetching the file
22:16 mapu joined #salt
22:16 aparsons joined #salt
22:17 raptor22x joined #salt
22:18 raptor22x so i keep getting this weird message when running a highstate
22:18 dalexander joined #salt
22:18 raptor22x I have an sls file in a subfolder from the base:
22:19 raptor22x if I call it salt 'nameofminion' state.sls production.apps
22:19 raptor22x it goes just fine
22:19 raptor22x however when i place the
22:19 bluenemo_ I think I found kind of a bug. When I use - unless: ! true, it does not execute the state, saying unless execution succeeded. - unless: $(! true) works as expected. Using salt 2014.1.10 on both master and minion.
22:20 bluenemo_ I'm want to invert the exit status of the command running for unless.
22:20 raptor22x line in the top.sls file in the base directory, it tells me "No Matching sls found for 'apps' in env 'production'
22:21 iggy raptor22x: you have an env and a module called production?
22:21 metaphore joined #salt
22:21 raptor22x yes
22:21 heise raptor22x: whats in app.sls
22:22 iggy I would probably avoid doing that python is weird about naming conflicts
22:22 raptor22x ok, let me try that
22:23 raptor22x this is the content of my apps.sls
22:23 raptor22x include:   - users  nginx:   pkg:     - installed      service:     - running
22:23 raptor22x sorry, using a web IRC client
22:24 raptor22x there is an includes for the users.sls, then it installes the nginx package and starts the service
22:24 iggy that was better than pasting 8 lines in here
22:24 aparsons joined #salt
22:24 schimmy joined #salt
22:24 raptor22x to start off, i will change the name of either my env or my module.
22:26 englishm joined #salt
22:26 heise raptor22x: also try and change the include: users to include: production.users
22:26 heise or whatever the name of the module is
22:26 raptor22x well the users.sls file is actually in the base
22:26 raptor22x for here is the structure
22:27 schimmy1 joined #salt
22:27 raptor22x base: contains top.sls and users.sls, there there is base/production  which contains the apps.sls file
22:29 heise any reason you dont want to put users.sls in top.sls instead ?
22:31 raptor22x I think the env and module being named the same might be root of the all evil. I will do a second run but I changed the env name to prod and ran a test whcih completed succesfully. Build a brand new minion and see what happenes
22:31 raptor22x happens
22:32 raptor22x nope, that was nto it
22:32 snuffeluffegus joined #salt
22:33 raptor22x ok so let me ask, I'm trying to make sure the users.sls is checked prior to installing the nginx package
22:33 raptor22x should I do a require or includes statement
22:33 heise require
22:33 heise require makes it an explict requirement
22:34 raptor22x - require:   -sls: users
22:34 raptor22x is that correct
22:35 aparsons joined #salt
22:36 pdayton joined #salt
22:39 DaveQB joined #salt
22:41 zemm joined #salt
22:42 aparsons joined #salt
22:43 yomilk joined #salt
22:44 delinquentme is there a salt-developer IRC channel?
22:44 Gareth delinquentme: there are a few.
22:44 snuffeluffegus joined #salt
22:44 delinquentme Im still wondering what the advantages of individuall generate SSH keys are vers shared between the minions
22:44 forrest yeah they're used for development, probably won't get any questions answered there
22:45 aparsons joined #salt
22:45 forrest delinquentme, you mean each minion generating it's own key?
22:45 forrest security
22:45 delinquentme it could be implicated in the permissions Issue im currently running into  -- and wondering if the additional load on the system of individual keys would actually solve it
22:45 forrest You might want to post to the mailing list
22:46 murrdoc whats the recommended formula for managing 'mkfs.ext4'
22:47 kralla joined #salt
22:48 murrdoc i d rather not cmd.run stuff
22:49 TTimo joined #salt
22:50 rojem joined #salt
22:51 dalexander murrdoc: here's an example of how I'm using extfs.mfks
22:51 dalexander https://gist.github.com/anonymous/3a868125308613dca846
22:54 thayne joined #salt
22:55 kingel joined #salt
22:56 murrdoc thanks dalexander
23:01 mosen joined #salt
23:01 iggy having a hell of a time getting a pillar _not_ using the master branch to show up in the base environment
23:03 yomilk joined #salt
23:03 aparsons joined #salt
23:05 aparsons joined #salt
23:08 wendall911 joined #salt
23:08 aparsons joined #salt
23:08 pentabular joined #salt
23:11 aparsons joined #salt
23:13 jkeys joined #salt
23:14 Outlander joined #salt
23:14 jkeys left #salt
23:14 longdays joined #salt
23:14 oz_akan joined #salt
23:16 longdays I am setting some jinja variables by getting the ouput of a salt cmd.run like so {% set log_disk_size = salt['cmd.run']("lsblk --nodeps --noheadings -b -o SIZE "+ log_disk) %}
23:16 longdays it looks like this gets executed before the state begins processing top down. Do any of you know if this is the case?
23:16 mapu joined #salt
23:16 longdays this line is of course in an init.sls file
23:16 kralla joined #salt
23:17 longdays it is in the middle
23:17 iggy I would expect the cmd.run to occur when jinja parses the file, then salt goes back and parses it as yaml
23:17 longdays ah, ok. So it parses it first
23:18 longdays that makes sense
23:18 longdays well now it makes sense based on the behavior I am seeing
23:18 iggy you can think of jinja as a pre-processor
23:18 ajolo joined #salt
23:19 iggy so everything in {}'s get's processed first, then that output is handed off to salt
23:19 iggy which is also why error lines are frequently off from where they should be
23:20 longdays ok, do you know if those jinja commands are processed top down or in parrallel?
23:20 longdays looks like top down like other scripts
23:21 iggy I'm not that familiar with jinja's processing, but I'd guess it's top-down
23:24 longdays thanks iggy. Help clarify some things
23:27 n8n joined #salt
23:29 acabrera joined #salt
23:32 laxity joined #salt
23:32 eunuchsocket joined #salt
23:33 active8 joined #salt
23:34 jalaziz joined #salt
23:34 delinquentme will decreasing my worker_threads speed up the instancing process of my minions?
23:35 yomilk joined #salt
23:35 delinquentme this shit takes like 10 minutes for 4 servers
23:35 joehh nkuttler: I would report it as a salt bug - it could be due to a new version of something in sid, but best to track at https://github.com/saltstack/salt/issues
23:38 n8n_ joined #salt
23:40 bhosmer_ joined #salt
23:43 Shish left #salt
23:46 delinquentme New / disturbing error: Unable to get IP for 00:10:00
23:46 delinquentme This means it cant get to salt-master?
23:58 scooby2 joined #salt
23:59 warmwaffles joined #salt

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