Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-08-29

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

All times shown according to UTC.

Time Nick Message
00:04 tim5464 joined #salt
00:06 tim5464 can a 'system.reboot' be triggered from within an .sls file ??  If so, how?  I've tried a few ways without any success.  Thx
00:07 cedwards tim5464: i don't believe there is support for the system module in the states yet
00:09 whit joined #salt
00:10 mgw tim5464: i think you have access to the salt dict within an sls
00:11 mgw salt['system.reboot']()
00:11 mgw or something like that; i've not done it
00:14 tim5464 yeah .. was not sure about support for it via states .. just throwing it out there .. thanks, I'll give your suggestion a go
00:14 Ryan_Lane joined #salt
00:15 Ryan_Lane joined #salt
00:16 mgw tim5464: you can also just use cmd.run
00:16 mgw but either way, it's going to interrupt the highstate run
00:18 teebes joined #salt
00:22 ProT-0-TypE joined #salt
00:24 StDiluted joined #salt
00:35 robertkeizer joined #salt
00:53 shadowsun grrr
00:53 shadowsun whereis forrest
00:57 ronc joined #salt
01:01 littleidea joined #salt
01:06 liwen joined #salt
01:08 teebes joined #salt
01:10 liwen_ joined #salt
01:19 whit joined #salt
01:20 liwen_ joined #salt
01:23 mgw joined #salt
01:27 racooper joined #salt
01:33 Eugene Among the trees
01:37 kenbolton joined #salt
01:41 Gifflen joined #salt
01:45 oz_akan_ joined #salt
01:46 oz_akan_ joined #salt
01:49 Gifflen joined #salt
01:52 jbunting joined #salt
01:53 isomorphic joined #salt
02:01 Gifflen joined #salt
02:06 liuyq joined #salt
02:08 liuyq joined #salt
02:09 druonysus joined #salt
02:10 mgw joined #salt
02:13 robertkeizer left #salt
02:18 liwen joined #salt
02:23 tim5464 mgw:  thx ..  cmd.run did the trick with 'shutdown /r'
02:25 Asphyxiated|Dk joined #salt
02:28 xl1 joined #salt
02:30 oz_akan_ joined #salt
02:42 StDiluted joined #salt
02:54 clintonb joined #salt
03:03 liwen_ joined #salt
03:12 [diecast] joined #salt
03:19 littleidea joined #salt
03:22 bhosmer joined #salt
03:34 malinoff joined #salt
03:36 pdayton joined #salt
03:36 dthom91 joined #salt
03:40 Ryan_Lane joined #salt
03:41 oz_akan_ joined #salt
03:43 malinoff Hi all
03:45 gareth_kun joined #salt
03:51 liwen joined #salt
03:52 jalbretsen joined #salt
04:00 qba73_ joined #salt
04:02 mwillhite joined #salt
04:06 waverider joined #salt
04:08 mmarch joined #salt
04:12 mnemonikk joined #salt
04:19 mmilano_ joined #salt
04:42 littleidea joined #salt
04:43 pdayton joined #salt
04:55 StDiluted anyone around here tonight?
04:55 malinoff StDiluted: I am
04:56 malinoff Since it's 11:56 AM here :)
04:57 berto- joined #salt
04:57 StDiluted ah nice
04:58 StDiluted have you played with the mysql_database states?
05:08 syngin malinoff: where are you? time sounds like it should be AU west, or somewhere in SE Asia...
05:08 malinoff StDiluted: Nope, but ask anyway, may be i could help
05:08 malinoff syngin: Mother Russia
05:08 syngin Ahh
05:08 syngin How many timezones in .ru, out of curiosity?
05:09 malinoff syngis: Omsk, if you want so see the exact place
05:09 syngin that's much further west than i was guessing...
05:09 malinoff syngis: Hmm... Before 2011 we had 11, now we have 9
05:09 syngin wowza
05:09 malinoff :D
05:10 syngin such a massive conitent you have, malinoff :]
05:10 syngin continent even
05:10 malinoff Sixth of the land
05:10 malinoff Well, almost
05:11 syngin wow. omsk gets some really intense heat variation over the year... your wardrobe must look like a shop...
05:11 malinoff :D
05:11 malinoff It is
05:11 malinoff From -40 in winter to +35 in summer
05:12 syngin man...
05:13 syngin we get from 5 to 40 and that kills, but -40 all the way up is some serious shit
05:14 malinoff Well, it all depends on the humidity
05:14 syngin we're very dry in melbourne
05:15 syngin and we cop a mean wind from antarctic in the winter.
05:15 StDiluted Well, I'm trying to get the mysql_database.present state to work. In order for that to work, there needs to be a mysql.config file with the username and password, so that the python mysqldb module can access the db and make the changes. Seems like things are going in the wrong order because the database is not getting created because it's trying root@localhost with no password
05:15 syngin (which is just finishing up)
05:16 malinoff syngin: Uh, Australia. I want to go there sometime
05:16 StDiluted I am trying a reload of the salt-minion once the mysql config file is in place
05:16 StDiluted but the db creation still fails
05:18 jeffmendoza joined #salt
05:18 malinoff StDiluted: Could you provide please your state file?
05:19 StDiluted hrm, it's a few inter-related state files, but let me see if i can paste them
05:21 MTecknology boy oh boy... getting salt running perfect and error free every single time is a bit of a challenge. Once you do it, though... boy oh boy it sure is swell.
05:21 littleidea joined #salt
05:22 Eugene It vomits unicorns!
05:22 StDiluted if i run the state twice, it works fine
05:23 Eugene Sounds like a requires: problem
05:23 StDiluted yep
05:23 StDiluted I'm guessing it is
05:24 StDiluted I think the salt-minion needs restarted at the right point
05:24 StDiluted but for some reason I have it watching the file but it doesnt restart the minion
05:25 StDiluted at the right point
05:25 syngin malinoff: we can swap for a few weeks. i've always wanted to see russia.
05:27 StDiluted I'm creating a mysql_config file, and I have a watch: on that with the salt-minion service restart, but it doesnt restart it as soon as the mysql_config file is created
05:27 StDiluted and it's pissing me off
05:29 StDiluted https://gist.github.com/dginther/6374542
05:29 pdayton joined #salt
05:31 malinoff syngin: Heh, i have a lot of work here, so... Come in winter, Russia is very beautiful in winter. And you will see a real snow! :)
05:31 StDiluted by looking at the highstate output, I can see that the salt-minion does not get restarted until WAY later in the highstate run
05:32 linjan__ joined #salt
05:34 StDiluted any ideas, malinoff?
05:34 malinoff StDiluted: You simply shouldn't restart minion by the salt. If it will be restarted this way, minion will not send a response to the master's response port - so, you will get no response on the master
05:35 StDiluted i dont care about the response on the master, i can get the jid info later
05:35 StDiluted but how else am i supposed to do it?
05:36 StDiluted i have to put the mysql_config there
05:36 StDiluted the minion needs to be reloaded in order to use it
05:36 StDiluted so....
05:36 MTecknology Eugene: When I run state.show_highstate on some boxes at work, it's a few thousand lines long. I ran into all sorts of dependency issues. I had to run state.highstate twice before the box would work..... not fun to figure out
05:36 Furao joined #salt
05:37 MTecknology fuuuuuun stuff
05:37 malinoff StDiluted: Can you put your mysql config in /etc/salt/minion.d directory?
05:37 malinoff This should help
05:37 StDiluted sure, I can, but that still requires a restart of the minion process to re-read that, right?
05:39 MTecknology I ran into that issue a bit ago with thinking I needed to restart the minion. I wound up accepting that if I ever need to restart the minion because of a config change, I just won't ever see the return.
05:39 StDiluted yeah, i dont care about the return, honestly
05:39 StDiluted I can run the highstate with -v, and then lookup the jid to get the output
05:40 MTecknology most of my minions take over an hour to run state.highstate so they time out before I see the response. :P
05:40 StDiluted but I can't figure out why the minion reload doesnt happen until the end of the highstate
05:41 MTecknology probably the ordering
05:41 StDiluted mine takes about 30 min. Compiling ruby and a couple other libraries is time consuming
05:41 StDiluted yes but i am watching the config file with a minion restart state
05:41 StDiluted so in theory, the minion should restart as soon as the file is modified
05:41 MTecknology ruby is hell
05:41 StDiluted I thought that was how the watch: command worked
05:42 VertigoRay joined #salt
05:42 MTecknology it'll restart with the process state runs, and the process state depends on the file state; it'll restart after the file change, but that doesn't mean immediately after
05:42 StDiluted meh, I'm using rbenv to make that work. I still have an issue where it wont install the gems until the second run as well
05:43 StDiluted for now it's no big deal to highstate it twice
05:43 StDiluted but...
05:43 MTecknology I'd rather take a pineapple suppository than ever work with ruby again
05:43 StDiluted hahaha
05:43 StDiluted the app I am supporting is Ruby
05:43 StDiluted we deploy the app and it's container using capistrano
05:44 MTecknology I'd quit.
05:44 StDiluted hahaha
05:44 StDiluted meh. I like my job
05:44 StDiluted we have nerf battles and get to drink beer whenever we want during the day
05:44 MTecknology I like my rectum.
05:44 StDiluted well, that's good, at least you do
05:45 MTecknology The hot peppers I eat might disagree, but I like it. :)
05:45 StDiluted because probably you're the only one that does?
05:45 StDiluted heheh
05:45 StDiluted <- growing ghost peppers in the garden
05:45 MTecknology run state.show_highstate and state.show_lowstate to figure out what is going on and why the order isn't exactly right
05:46 MTecknology ghost peppers are tasty and hot, but not overly hot. I like them a lot
05:46 berto- joined #salt
05:46 StDiluted I'll try that. I think I'm going to go to bed tonight though
05:46 StDiluted not overly hot?
05:46 MTecknology I just bought a house. Next year I'll be growing butch t peppers
05:46 zach Gents, does this look correct: https://p.6core.net/p/0klHGHjeE3UMSzR6tQFLBQJN -- I don't really see anything in the documentation that we can do it that way
05:46 StDiluted ghost peppers are the hottest pepper in the world according to everything I've read
05:47 MTecknology no, they're rather far from it
05:47 StDiluted hm
05:47 MTecknology zach: looks fine
05:47 StDiluted In 2007, Guinness World Records certified that the Ghost Pepper (Bhut Jolokia) was the world's hottest chili pepper, 400 times hotter than Tabasco sauce; however, as of 2012 it was superseded by the Trinidad Moruga Scorpion.[7]
05:48 MTecknology yuppers
05:48 StDiluted not that i believe everything guiness says
05:48 Eugene Good beer; crappy book.
05:48 MTecknology that one is mostly true, moruga is a joke
05:48 MTecknology Eugene: the beer is iffy..
05:48 zach MTecknology: thanks
05:48 Eugene I said good, not excellent
05:49 MTecknology I have the Butch T strain of the Trinidad Scorpion pepper. I'm excited as all shit to be growing it next year.
05:49 MTecknology I think I have 5 seeds...
05:49 StDiluted ah, yeah i see that's in the 1.5 million scoville range
05:49 MTecknology Someone one THP mailed me a pure strain of them.
05:50 bemehow joined #salt
05:50 StDiluted bhut jolokia is a little over a million
05:50 StDiluted jesus
05:50 StDiluted moruga is over 2 mil
05:50 MTecknology no, it's not
05:51 MTecknology moruga is a joke and purely marketing bullshit
05:51 StDiluted hm
05:51 MTecknology it's not once been tested, the people that claim to grow it that hot do their own private testing and refuse to let others actually test it
05:52 MTecknology long rant - anything labeled moruga is a joke
05:54 MTecknology StDiluted: the record for a scorpion is 1.46 mil SHU
05:54 MTecknology I don't remember what SHU got bhut jolokia into the books
06:01 MTecknology I wonder if I have any opinions on the subject... :S
06:01 MTecknology StDiluted: Ping me next year and I'll mail you some if they grow. :)
06:02 ProT-0-TypE joined #salt
06:02 ProT-0-TypE joined #salt
06:09 StDiluted lol, ok
06:09 StDiluted SHU for bhut jolokia is around a million or so
06:09 StDiluted which is plenty hot for me
06:13 ProT-0-TypE joined #salt
06:19 middleman_ joined #salt
06:27 sgviking joined #salt
06:38 lemao joined #salt
06:47 balboah_ joined #salt
06:48 Xeago joined #salt
06:49 gildegoma joined #salt
06:51 dthom91 joined #salt
06:53 lemao joined #salt
06:55 BbT0n joined #salt
06:59 lemao joined #salt
07:02 Furao joined #salt
07:08 vaxholm_ joined #salt
07:09 az87c joined #salt
07:10 ml_1 joined #salt
07:16 black31 joined #salt
07:17 black31 left #salt
07:21 carlos joined #salt
07:24 hvn joined #salt
07:29 TheCodeAssassin joined #salt
07:32 ProT-0-TypE joined #salt
07:33 qba73_ joined #salt
07:34 Khazix joined #salt
07:37 scott_w joined #salt
07:47 quantumsummers|c joined #salt
07:48 abele joined #salt
07:52 dbrighi joined #salt
08:01 Xeago joined #salt
08:01 mmilano left #salt
08:05 felixhummel joined #salt
08:08 Furao joined #salt
08:09 lemao joined #salt
08:11 zooz joined #salt
08:27 Katafalkas joined #salt
08:36 mortis_ why does this http://pastie.org/8279471 tell me that "State file.missing found in sls gitlab is unavailable" ?
08:37 mortis_ hmm maybe absent is the thing
08:37 mortis_ ah no
08:38 mortis_ absent deletes
08:59 TheCodeAssassin joined #salt
09:24 hvn_ joined #salt
09:27 carmony joined #salt
09:28 hvn_ left #salt
09:35 MrTango joined #salt
09:38 ronc joined #salt
09:44 Furao joined #salt
09:50 waverider joined #salt
09:53 TheCodeAssassin joined #salt
09:54 ninkotech__ joined #salt
10:02 liuyq joined #salt
10:16 matanya joined #salt
10:23 Katafalkas joined #salt
10:24 felixhummel hi! Has anybody here ever encountered "Failed to change user to foo; Failed to change group to foo;"?
10:32 felixhummel file.directory cannot handle 10900 files it seems. When I checkout a prior version of the directory with 9500 files it works.
10:32 gareth_kun left #salt
10:35 giantlock joined #salt
10:51 ggoZ joined #salt
10:59 blee joined #salt
10:59 unicoletti_ joined #salt
11:04 Psy-Q if i have something that needs to be done in a certain order, is it okay if i just chain the steps with "require"?
11:04 Psy-Q i've chained two scripts, one invocation of a2ensite and two package installations that way, it seems it's the only way it can be done
11:05 Jason-AVST joined #salt
11:08 xt خ"
11:08 Katafalk_ joined #salt
11:09 xt سمَـَّوُوُحخ ̷̴̐خ ̷̴̐خ ̷̴̐خ امارتيخ ̷̴̐خ
11:09 kbar what
11:10 Katafalkas joined #salt
11:11 Katafalkas joined #salt
11:12 Katafalkas joined #salt
11:13 TheCodeAssassin joined #salt
11:16 williamthekid joined #salt
11:16 Hemebond joined #salt
11:17 Hemebond Is require: virtualenv: venv-id suppose to activate the virtualenv for that ... part of the state file?
11:18 Hemebond If not, how do I make sure a state is applied within the virtualenv, e.g., a cmd.run
11:19 santagada joined #salt
11:19 tomeff joined #salt
11:24 Furao joined #salt
11:44 jbunting joined #salt
11:45 ml_11 joined #salt
11:58 Psy-Q i have a chain of things that have to happen one after the other, but it seems i can't do that with 'requires', as even if a previously required step exits with an error, the following steps will be executed anyhow
11:58 Psy-Q how could i get stuff into a sequence?
11:58 kbar Psy-Q: notify perhaps?
11:59 kbar oh wait, wrong project, thinking of ansible. sorry :(
12:00 Psy-Q aah, was wondering where the heck that stuff is documented
12:00 kbar :( sorry
12:00 Psy-Q i'm trying again with a fresh box
12:00 Psy-Q maybe i did some requires wrong
12:00 Psy-Q but i don't see how i can get this running any other way, since one step requires the other to finish :\
12:01 brianhicks joined #salt
12:03 chubrub joined #salt
12:04 Hemebond Psy-Q: Can you use onlyif? Are these cmd.run?
12:04 ghchris joined #salt
12:04 Psy-Q Hemebond: not only cmd.run, also gem.installed
12:04 Hemebond What if you separate them out and give them their own require:
12:04 ghchris How do you use pkg.mod_repo from a .sls file?
12:05 chubrub hey all, I've got strange problem after upgrading from 0.16.2. -> 0.16.3. I'm setup one variable for my config in following way
12:05 ghchris I'm trying to add a percona.list Debian/Ubuntu in /etc/apt/sources.list.d from a .sls file
12:05 Psy-Q Hemebond: i thought that was what i was doing, but it seems to me that they get called in sequence even if an earlier one in the chain has failed
12:05 Hemebond Psy-Q: Are they all in the same require: ?
12:05 chubrub {{ salt['cmd.run'] (' [[ some test ]] && set_var || set other var ')
12:06 chubrub and started to getting :w
12:06 Hemebond Psy-Q: My states fail if their require: isn't satisfied.
12:06 Psy-Q Hemebond: no, one require per step, let me see if i can show you
12:06 Psy-Q https://github.com/zhdk/leihs/blob/rca_vagrant/salt/roots/passenger.sls  <-- this way
12:06 chubrub bin/sh: 1: [[: not found
12:07 Psy-Q i want to install the passenger gem using rvm, then compile passenger, install the module, create the loader snippet (later even reload apache)
12:07 chubrub and as far as I know default shell for cmd.run is /bin/bash
12:09 Psy-Q ah, maybe my unless is wrong
12:09 Psy-Q let's see
12:09 chubrub any idea what can be wrong here?
12:11 Hemebond Psy-Q: Is it suppose to be requires?
12:11 Hemebond Mine has require:
12:13 Psy-Q Hemebond: i'll try that
12:14 Hemebond I also just use e.g., pkg, not pkg.installed in my require sections
12:16 Psy-Q Hemebond: some from require: pkg  it automatically knows that i require the pkg to be installed?
12:16 Psy-Q and not one of the other states it could be?
12:16 tyler-baker joined #salt
12:16 Hemebond Psy-Q: Not sure. I thought it just told it to require that state id.
12:16 Hemebond e.g.,
12:17 Hemebond require: pkg: my-dependencies
12:18 Hemebond Psy-Q: e.g., https://gist.github.com/jenius/5732706
12:20 kamal_ is it possible to output only the changed bits when running highstate?
12:20 yidhra joined #salt
12:21 drawsmcgraw joined #salt
12:23 elements joined #salt
12:25 bhosmer joined #salt
12:26 abe_music joined #salt
12:27 chubrub ok, so maybe another way. How can I pass parameter shell=/bin/bash for following execution {{ salt['cmd.run'] (' bash_command ') }}?
12:32 Psy-Q Hemebond: thanks! i'll try that way
12:32 Psy-Q didn't know how the id figures into it so far
12:35 td_ joined #salt
12:36 JesseC joined #salt
12:37 teebes joined #salt
12:38 JesseC joined #salt
12:40 felixhummel kamal_: set state_verbose to false. see http://docs.saltstack.com/ref/configuration/master.html#state-verbose
12:40 felixhummel s/false/False/
12:43 kamal_ Hmm neat never knew that exists
12:45 kamal_ Thanks felixhummel
12:45 drawsmcgraw chubrub: You can't just specify shell=bash ?
12:46 felixhummel You're welcome. I asked the same question here not long ago and was very thankful for a quick response. This should really be better documented, i.e. in the tutorial...
12:46 kamal_ Mhm
12:47 felixhummel chubrub: What are you trying to accomplish? Maybe it could be modelled differently with salt...
12:52 anteaya joined #salt
12:54 Katafalkas joined #salt
12:56 Katafalkas Hey, for some reason salt-cloud fails to ssh execute command "date". I am assuming its because of the long time it takes to make a connections with aws, as if I run the command manually, it works, altho it takes some time to connect and execute it,
12:56 Katafalkas " ssh  -oStrictHostKeyChecking=no -oUserKnownHostsFile=/dev/null -oPasswordAuthentication=no -i /etc/salt/salt.pem ubuntu@10.35.161.74 "date" "
12:57 Katafalkas is there a way to increase the waiting time ?
12:57 Katafalkas or it is some other problem
12:57 Katafalkas ?
12:58 oz_akan_ joined #salt
12:59 joehh felixhummel, kamal_: thanks, that is a great thing to know!
13:00 oz_akan_ joined #salt
13:03 ProT-0-TypE joined #salt
13:05 juicer2 joined #salt
13:09 racooper joined #salt
13:14 imaginarysteve joined #salt
13:21 jeffasinger joined #salt
13:21 mwillhite joined #salt
13:21 Kilsharion joined #salt
13:21 oc joined #salt
13:24 carmony joined #salt
13:26 oc joined #salt
13:27 backjlack joined #salt
13:29 wimbo joined #salt
13:29 wimbo guys
13:30 wimbo can somebody help with states?
13:30 wimbo how i can told other sls files wait then one sls file complete? only require for everyone command?
13:31 aat joined #salt
13:34 honestly wimbo: I have no idea what you mean
13:35 alunduil joined #salt
13:36 toastedpenguin left #salt
13:36 terminalmage joined #salt
13:36 drawsmcgraw wimbo: I imagine you've read it already but just in case -> docs on Requisites --> http://docs.saltstack.com/ref/states/requisites.html
13:37 drawsmcgraw I've also hit the challenge of having an entire .sls wait for another .sls to complete
13:37 drawsmcgraw I think I wound up just re-tooling my individual states...
13:38 drawsmcgraw wimbo: You can, though, have on .sls file include another .sls file. I think you can have individual states require other individual states across .sls files. That may be what you're looking for.
13:39 wimbo what do including?
13:39 drawsmcgraw wimbo: Here's the doc on Include: http://docs.saltstack.com/ref/states/include.html
13:41 Kholloway joined #salt
13:42 kenbolton joined #salt
13:43 m_george left #salt
13:43 tomviner joined #salt
13:44 honestly so I'm needing to extend the users-formula (https://github.com/saltstack-formulas/users-formula/) to support NFS homedirs
13:44 honestly wondering how to go about that
13:44 honestly (I know the obvious answer is "use ldap")
13:46 kaptk2 joined #salt
13:48 DredTiger joined #salt
13:50 cron0 joined #salt
13:54 fxhp normally service restarts happen at the very end
13:54 fxhp what if I needed to run a state/cmd after a service restarts?
14:02 Psy-Q hrm, the salt-call state.highstate only worked on the second run with my requires
14:03 fxhp Psy-Q - that is a dependency issue
14:04 fxhp Psy-Q - you have the order wrong somewhere
14:04 StDiluted joined #salt
14:07 Gifflen joined #salt
14:10 entil I got a lead on my oddball TypeError!
14:10 entil self.auth.crypticle.loads() in RemoteClient returns an empty string
14:10 Kilsharion joined #salt
14:10 entil it's this weird thing where salt/state.py expects a dict but RemoteClient.master_opts() can return ''
14:11 entil it has a try/except SaltReqTimeoutError, and the except: code returns '' - but I'm not ending up in that, this is not a timeout
14:11 entil what happens in the try: block, it does mojo with self.auth.crypticle.loads() and that returns a '' too
14:11 entil wtf's up with that, anyone?
14:12 honestly I'm getting confused with state and pillar and everything... what's the correct place to put a designation for my nodes? I want to have one node designated as "nfs-master" and the rest as "nfs-client" and switch on that in state templates
14:13 entil honestly: I use aws tags for that
14:13 mmilano joined #salt
14:13 honestly entil: well... I don't use aws at all
14:13 honestly should i just add my own grains?
14:13 drawsmcgraw honestly: Can you name your machines accordingly?
14:13 drawsmcgraw Then you could key off of their names in the Topfile and/or Pillar files.
14:14 entil encode it in real life anyway
14:14 honestly drawsmcgraw: I want to key off of their names in state template files
14:15 honestly they are named accordingly but I want a layer of indirection there
14:15 entil I have this magick state called environment-selector which looks at aws tags and works from there
14:15 honestly entil: can I take a look at that?
14:15 entil but I'm seriously bothered by why this whole setup suddenly crumbled... I know python has issues with types, like not having strong types, but all of a sudden getting bs back from the master does not sound ok
14:16 jbunting joined #salt
14:16 entil honestly: it's basically include: - environment-selector.{{ grains['ec2_tags']['environment'] }}
14:16 entil honestly: I have that in environment-selector/init.sls
14:16 honestly entil: are you managing your setup in a git repo?
14:17 felixhummel honestly: a) DNS, because minion ids default to hostname b) minion config on boot (id param) c) pillar
14:17 vaxholm_ joined #salt
14:17 entil honestly: yes
14:17 honestly entil: well then, should be easy to bisect and find out what's causing the collapse
14:18 entil honestly: but I debug the collapse by logging stuff in salt itself, I'm not so convinced this is a state issue anymore
14:18 honestly felixhummel: how do I do this in pillar
14:18 entil the exception I get stems from salt itself
14:18 honestly ideally I just want to list the node IDs in the pillar data
14:19 honestly oh
14:19 honestly I can use grains.id
14:23 felixhummel honestly: are we talking about roles? so do you already have useful IDs and want to say that host1 is "nfs-master" and host2, 3, 4 are "nfs-client" boxes?
14:23 honestly felixhummel: https://gist.github.com/duk3luk3/6378758
14:24 KennethWilke joined #salt
14:24 honestly this is basically what I want to put in my pillar
14:24 honestly the syntax is probably wrong
14:24 mannyt joined #salt
14:25 honestly and then I can do {% if grains['id'] in pillar.get('nfs_id.nfs_master') %} or something like that
14:25 honestly right?
14:25 felixhummel correct
14:26 honestly what's the right syntax?
14:26 felixhummel http://docs.saltstack.com/topics/tutorials/pillar.html
14:26 felixhummel just yaml
14:26 felixhummel your example should work
14:26 honestly ok
14:28 honestly and the template command should be... {% if grains['id'] in pillar.get('nfs_id:nfs_master',{}).items() %}
14:29 honestly correct?
14:30 jaredr joined #salt
14:30 jergerber joined #salt
14:30 entil that might work, though I think the .items() is redundant.
14:30 entil test it :)
14:31 entil this is soooo strage. I ventured farther into the code and got this binary-looking response from what probably is the salt master, and it goes to crypticle.loads() and it returns an empty string
14:32 entil this made me look at the master's log file
14:32 entil self.pid = os.fork() -> OSError: [Errno 12] Cannot allocate memory
14:32 entil so what's actually a memory issue on the master looks like a TypeError on the client
14:33 entil 22M free memory, that sounds like plenty for a system as scalable as salt, but apparently not
14:35 jams1 joined #salt
14:36 entil so it's os.fork() that fails, sending a response that's a pickled '' which gets tumbled around in the minion code, until it hits a path where a method expects a (populated) dictionary but gets '' instead
14:36 kenbolton joined #salt
14:36 entil I hope later versions of salt would include some field for "error happened" and return values of the expected type
14:37 [diecast] joined #salt
14:38 abe_music joined #salt
14:41 honestly entil: fix it!
14:41 entil I'm on 0.15.2, so in order to fix it I'd need to upgrade first
14:42 entil furao warned me the newer versions are incompatible, so I'd have to migrate everything
14:42 entil and then start hacking
14:42 honestly ah, damn
14:42 Xeago left #salt
14:42 entil as much as I'd love to contribute code, I don't have time for that
14:45 DredTiger joined #salt
14:45 honestly ah
14:47 Katafalkas does the "block_device_mappings:" actually work on salt-cloud 8.9.0 ?
14:48 mrpull joined #salt
14:51 zwevans joined #salt
14:53 pdayton joined #salt
14:55 Psy-Q fxhp: hrm, the sequence seems fine for me, let me show you
14:56 Psy-Q https://github.com/zhdk/leihs/blob/rca_vagrant/salt/roots/passenger.sls
14:56 Psy-Q it can't go through everything in one step
14:57 cnelsonsic joined #salt
14:57 teskew joined #salt
14:58 Psy-Q i thought i had it set up so that it goes through each id in sequence, but i have to run it twice for it to work
14:58 ckao joined #salt
14:59 jalbretsen joined #salt
15:02 SunSparc joined #salt
15:04 StDiluted joined #salt
15:04 samsalt joined #salt
15:04 samsalt left #salt
15:04 StDiluted morning
15:06 lineman60 joined #salt
15:07 chrisgilmerproj joined #salt
15:07 StDiluted Anyone here using the mysql_database states?
15:07 mwillhite joined #salt
15:09 KennethWilke howdy, i have some light usage of it
15:09 conan_the_destro joined #salt
15:11 StDiluted hey KennethWilke, I'm running into an ordering problem with the config file
15:11 StDiluted I'm creating a config file as it says to in the docs
15:12 StDiluted I'm managing that file with Salt
15:12 StDiluted but that means the minion needs a restart after the file is placed
15:14 StDiluted but that means no return from the minion (i dont care about that, really) but the restart it not happening until the end of the highstate
15:14 StDiluted is*
15:14 dave_den joined #salt
15:14 StDiluted which means the database creations I'm trying to do fail
15:15 KennethWilke hmm, what config file is it you're managing?
15:15 KennethWilke the minion config?
15:16 KennethWilke if i understand correctly it sounds like you have salt-minion and mysql configs in the same state file?
15:16 KennethWilke or running in the same highstate
15:16 kenbolton joined #salt
15:16 StDiluted yep
15:16 StDiluted that is the case
15:16 StDiluted the file I'm managing is not the minion config file
15:16 StDiluted it's the mysql_config
15:17 StDiluted https://salt.readthedocs.org/en/latest/ref/modules/all/salt.modules.mysql.html#module-salt.modules.mysql
15:17 StDiluted mysql.default_file
15:17 KennethWilke yeah, but i believe that setting goes in the /etc/salt/minion config file
15:18 StDiluted it can, or you can use a defaults file that points to it
15:18 StDiluted or rather
15:18 StDiluted you can use a defaults file that the minion points to
15:18 StDiluted I am using an AMI with the minion baked in
15:18 dave_den joined #salt
15:18 StDiluted perhaps I ought to just bake in the defaults file, however, I lose some flexibility doing that
15:19 KennethWilke hmm maybe, you could bake in a good default file then manage that file with salt if you need to make any settings changes afterward
15:20 KennethWilke i haven't used that default_file setting before, is that looking for a yaml config that salt will use or for something ala .my.cnf
15:20 kenbolton joined #salt
15:21 ccase joined #salt
15:21 JoAkKiNeN joined #salt
15:22 StDiluted it's just got the connection details for the mysql root account in it
15:23 StDiluted so that the mysql_database module can do it's job
15:24 StDiluted I guess I'll just bake it into the AMI. That causes me issues when I move to using salt-cloud, but that's down the road, I guess
15:26 Eugene Look into how Debian does it with their deb-maint-user
15:27 StDiluted I'm configuring the database to begin with using a debian conf file
15:27 StDiluted that's not the issue
15:27 StDiluted the issue is that the mysql_database module for salt requires a config
15:27 devinus joined #salt
15:27 StDiluted so that the python-mysqldb module can access and edit the database
15:28 Eugene Ah, requires: ordering. Got it. Sorry, didn't read all the way up.
15:28 StDiluted yeah it's an ordering issue, but it's a weird-ish one
15:28 Eugene Yeah, I think this is (yet another) situation where two highstate runs are needed.
15:28 dave_den joined #salt
15:28 StDiluted yeah unless I bake the config into the AMI which is (I guess) what I'm going to do
15:29 StDiluted The other ordering issue I have is that rbenv installs ruby properly but the rbenv shims are not there to install gems until the second run
15:29 Eugene This sort of thing is why my Vanilla vmware image is already running a minion. ;-)
15:29 StDiluted yeah, that's why I've got an AMI with a minion baked in
15:29 StDiluted I wanted to make it as generic as possible
15:29 Eugene I have a script that nukes the keys when the hostname changes
15:29 StDiluted but I think I need to specialize it a bit
15:30 Eugene AMIs are cheap.
15:30 dave_den left #salt
15:30 dave_den joined #salt
15:30 Jahkeup_ joined #salt
15:34 scalability-junk joined #salt
15:35 m_george|away joined #salt
15:38 jams2 joined #salt
15:38 dave_den joined #salt
15:43 Jahkeup_ joined #salt
15:44 mgw joined #salt
15:45 honestly so salt will resolve requisites dynamically and it doesn't matter in which order I specify them as long as I don't create a cycle, correct?
15:45 vaxholm_ joined #salt
15:47 StDiluted no
15:47 StDiluted you need to specify order in some cases
15:48 StDiluted using require: and require_in:
15:48 StDiluted as well as include:
15:49 ipmb joined #salt
15:50 dthom91 joined #salt
15:53 mgw joined #salt
15:54 carxwol joined #salt
15:54 honestly StDiluted: yes, using require is what I meant
15:54 StDiluted as long as you are using require, it will resolve the order for you
15:54 honestly k
15:54 honestly what happens if I declare the same state more than once?
15:55 honestly i.e. a file.directory state
15:55 StDiluted it should only run the state once
15:55 jams1 joined #salt
15:55 honestly ok
15:55 StDiluted according to UtahDave when I asked the same q
15:55 honestly what happens if there is a conflict?
15:55 honestly will it just throw a compile errorß
15:55 honestly ?
15:55 StDiluted well, how would there be a conflict
15:56 honestly say I declare /etc/blah: file.directory: user: foo
15:56 honestly and then the same thing again with user: bar
15:56 StDiluted you can't declare /etc/blah: more than once
15:56 StDiluted declarations have to be unique
15:56 StDiluted declaration IDs
15:56 StDiluted it will throw an error
15:56 StDiluted even if they are in different states
15:57 StDiluted because it makes one big dict out of all your states
15:57 StDiluted so if you have two delcaration IDs that are the same but with different values, you will get a stacktrace
16:01 mgw left #salt
16:03 UtahDave joined #salt
16:06 KyleG joined #salt
16:06 KyleG joined #salt
16:06 honestly StDiluted: ah, damn
16:07 honestly well I can bypass that
16:07 tyler-baker joined #salt
16:07 mfournier joined #salt
16:08 jams2 joined #salt
16:12 lemsto left #salt
16:12 dave_den technically, you can have the same IDs
16:12 dave_den but you cannot have two of the same functions for those same IDS.
16:13 jams1 joined #salt
16:13 dave_den for example, this works fine: https://gist.github.com/dlanderson/55d050d0739674f19192
16:14 chrisgh joined #salt
16:14 chrisgh how do you use pkg.mod_repo from a .sls file?
16:14 dthom91 joined #salt
16:16 dave_den chrisgh: what are you trying to do?
16:17 chrisgh I'm trying to Add Percona MySQL apt source in Ubuntu via Salt
16:18 dave_den chrisgh: https://gist.github.com/dlanderson/bc9bb6944d74854d03f0
16:18 juanlittledevil joined #salt
16:18 chrisgh Nice dave_den
16:18 waverider joined #salt
16:18 chrisgh Thanks a lot! :)
16:18 dave_den welcome :)
16:18 chrisgh Didnt find that one when searching
16:19 dave_den i just posted it
16:19 jdenning joined #salt
16:20 honestly dave_den: hrm, I see
16:21 dave_den honestly: that's also identical to just putting cmd.run and file.managed under one single ID
16:22 chrisgilmerproj left #salt
16:22 dave_den but if you try to do a file.managed and a directory.managed under the same ID, even with different -name arguments for each, it will fail
16:22 honestly dave_den: how about this? https://gist.github.com/duk3luk3/0ce105ea7f77009a95a5
16:22 honestly i.e. same file.managed twice under different IDs
16:22 StDiluted ah, yes, he was asking about conflicts, where two declarations with the same ID did different things to the same state
16:23 bitz joined #salt
16:23 honestly well ideally I will have different IDs doing the same thing :P
16:23 dave_den honestly: there is no point in doing that if they will be run during the same state/highstate
16:23 StDiluted same file managed twice under different IDs will work, however, make sure you know what order they are in so you know what state they will end up in
16:23 StDiluted I guess I can see where you might order things and need a certain value at first but then a different final value
16:24 honestly dave_den: I am quite aware - however this will be the result of some templating I'm doing now
16:24 StDiluted but that seems sort of sketch to me
16:24 honestly ah
16:24 jaredr I want to be able to configure individual mongo servers with salt -- port numbers, peers, replication set number, etc. What's the best way to do the individualized-host configuration? Do it with "salt cmd.run" from a script? Or something else?
16:24 dave_den honestly: you should use jinja templating to dynamically compile that state instead of making an a_motd and b_motd.
16:25 dave_den IMO
16:25 honestly dave_den: here's what I'm doing - I'm extending users-formula to include the possibility of creating group homes
16:25 honestly and the group home directory is derived from the first group assigned to the user
16:26 honestly and it will create the home dir declaration for every user, regardless of group home or not
16:26 honestly so if I have three users with the same primary group, it will have three declarations for the same directory
16:27 troyready joined #salt
16:27 honestly (and I'm doing that using jinja)
16:28 StDiluted that seems fine, you're not going to have it creating a different directory on each declaration, and each declaration will have a unique ID
16:28 StDiluted the name of the dir will be the group name, not the username
16:30 honestly yeah, I'm making the declaration ID to be "homedir_{{ user }}"
16:31 kula jaredr: this is the kind of thing pillar is often used for.
16:32 StDiluted honestly: seems fine to me, it will just tell you the directory is already there on subsequent passes.
16:32 mgw joined #salt
16:34 honestly StDiluted: yeah, that's cool
16:34 TheCodeAssassin joined #salt
16:35 waverider joined #salt
16:35 honestly only problem is that it won't fail if I screw up
16:35 honestly but eh, I'll just not screw up
16:35 StDiluted haha, that's always the goal, right?
16:38 KyleG joined #salt
16:38 KyleG joined #salt
16:39 carmony joined #salt
16:40 littleidea joined #salt
16:41 madduck when is 0.17 due?
16:42 jacksontj joined #salt
16:42 TheCodeAssassin joined #salt
16:44 dthom91 joined #salt
16:45 jdenning joined #salt
16:48 jdenning_ joined #salt
16:51 KennethWilke hmm, i'm having some funky pillarness, when i do a state run i get an error saying my sls file isn't available, but when i run a pillar.data the data from that sls file shows up
16:51 UtahDave madduck: 0.17 should be released sometime next week
16:54 KennethWilke and also a bit odd... my file_roots is pointing to /srv/pillar though it's set to /srv/salt in my master config :x
16:55 jhenrypb joined #salt
17:02 aat joined #salt
17:04 devinus joined #salt
17:09 dthom91 joined #salt
17:10 madduck UtahDave: \o/
17:11 madduck that is exactly what I wanted to hear
17:11 tomviner joined #salt
17:11 littleidea joined #salt
17:11 jaequery joined #salt
17:12 UtahDave :)
17:12 Eugene KennethWilke - obvious question; did you restart the master?
17:12 KennethWilke yep, few time
17:12 KennethWilke s
17:13 bashcoder It sure feels like a release is imminent - there's been quite a flurry of Github activity over the past two weeks. UtahDave, s0undt3ch and basepi are coding monsters!
17:13 basepi heh
17:13 basepi ya, this will be one of our most commit-heavy releases
17:14 basepi lots of fixes and lots of features
17:14 bashcoder I can barely keep up with the RSS feed haha
17:14 basepi Hehe, I can barely keep up with the issue tracker, and I work on salt full time!  =P
17:16 bashcoder you ARE the issue tracker! haha
17:17 UtahDave lol
17:18 Anb_ joined #salt
17:20 mwillhite I'm having problems formatting my sls to create multiple mysql databases, if anyone could please take a look I'd appreciate it: https://gist.github.com/miwillhite/6380814
17:20 KennethWilke Eugene: i suspect this may be the cause https://github.com/saltstack/salt/blob/6b3b047980bd23718718728d78807bb129375618/salt/pillar/__init__.py#L99
17:20 mwillhite thats the complete mysql-formula.sls
17:21 Eugene That does look rather suspect
17:21 mwillhite with error at the bottom…
17:21 StDiluted mwillhite: each one of your database present states should be it's own declaration
17:21 StDiluted not under a 'database-setup:' declaration
17:22 KennethWilke yeah looks like that may be accidental whitespace in there?
17:22 eculver joined #salt
17:22 jhenrypb left #salt
17:22 StDiluted python wont like that indention
17:22 mwillhite not an accident, I'm just new to this :)
17:22 mwillhite ok, I'll try with those at the parent level
17:22 KennethWilke ah yeah, i think you just need to remove 2 spaces on lines 24-37 to fix it
17:22 StDiluted I think if you bring them back it will work
17:23 isomorphic joined #salt
17:23 mwillhite thanks, building…
17:23 KennethWilke or at least to get a different error :p
17:23 KennethWilke you're welcome
17:23 StDiluted lol Kenneth
17:23 StDiluted ain't that the truth
17:23 KennethWilke my biggest problem with salt is always my hands :(
17:23 KennethWilke and sometimes the brain
17:24 bashcoder ^^ this
17:25 mwillhite joined #salt
17:25 bashcoder even my dreams are now indented.
17:27 felixhummel joined #salt
17:32 smcquay joined #salt
17:34 dthom911 joined #salt
17:35 aat joined #salt
17:35 robawt hey folks
17:36 robawt has anyone successfully used bootstrap.saltstack.org but specified a particular version for Linux clients?
17:36 cedwards robawt: I don't know that it supports anything other than latest stable or git
17:37 robawt hmm
17:37 Eugene No, I just `yum install` and call it a day.
17:37 robawt I know from git install we can specify a tag/version
17:37 Eugene "bootstrap" scripts outside of the system's package management have always ended in tears for me
17:38 Eugene The ones which create a /etc/yum.repos.d/foo.repo and then `yum install`(or the apt equivalent) aren't /too/ bad, but when they get wedged it's hilarious.
17:38 jms_ joined #salt
17:41 d10n joined #salt
17:41 g4rlic Eugene: I agree.  We just went through bootstrapping salt into a very isolated network last week.  Kickstart, tftp, http, dhcpd, and yum set up on my work laptop ruled the day.
17:42 g4rlic (and some ugly, kludgy %post to pull the saltmaster configs from my laptop directly)
17:42 Eugene Aheheheh
17:43 NotreDev joined #salt
17:44 g4rlic I didn't say it was pretty. but it got the job done. ;)
17:48 d10n joined #salt
17:51 honestly christ, I totally ran out of steam on this  >.<
17:51 chrisgh joined #salt
17:57 Ryan_Lane joined #salt
17:58 ml_1 joined #salt
18:05 druonysus joined #salt
18:05 druonysus joined #salt
18:07 bwq joined #salt
18:08 blee joined #salt
18:11 cedwards is devel currently broken, or have i done something wrong?
18:11 mattbarto joined #salt
18:12 dthom91 joined #salt
18:12 TheCodeAssassin joined #salt
18:13 whit joined #salt
18:13 vaxholm_ joined #salt
18:14 aat joined #salt
18:18 Ryan_Lane joined #salt
18:19 waverider joined #salt
18:20 UtahDave Corey and bhosmer got mentioned in this great SaltStack article on Linux.com: http://www.linux.com/news/enterprise/cloud-computing/735742-5-things-to-know-about-salt-cloud-management
18:21 blee congrats guys!
18:23 Anb joined #salt
18:23 Anb joined #salt
18:25 iambossjones joined #salt
18:26 kermit joined #salt
18:31 dthom91 joined #salt
18:32 teebes joined #salt
18:33 dave_den joined #salt
18:34 bhosmer joined #salt
18:35 bossjones left #salt
18:36 UtahDave bhosmer: you missed my recent comment. You were quoted in this linux.com article about Salt! http://www.linux.com/news/enterprise/cloud-computing/735742-5-things-to-know-about-salt-cloud-management
18:37 bhosmer I hadn't seen that. I'm going to go look at it now.
18:38 bhosmer I see Corey go in there too!
18:39 dave_den left #salt
18:39 dave_den joined #salt
18:43 cwright is there an expected timeframe for 0.17?  reading that article about the changes that are coming makes me wonder if I should wait to use the salt-ssh system
18:43 danielbachhuber joined #salt
18:43 UtahDave bhosmer: yep!
18:44 gldnspud joined #salt
18:44 UtahDave cwright: We're planning on releasing 0.17 sometime next week.
18:44 cwright UtahDave: thanks.  thats very exciting :)
18:44 cedwards basepi: what would you like me to add and where? re: #6954
18:45 drawsmcgraw Wait... would salt-ssh let me manage ssh sessions to appliances?
18:46 UtahDave drawsmcgraw: not initially, but it's built to be able to do that.
18:47 isomorphic joined #salt
18:48 drawsmcgraw UtahDave: Neat! So... if we had, say, some Netapps, we could (eventually) use Salt to connect to, then configure, them for a deployment?
18:48 drawsmcgraw a la Expect, but without the suck ?
18:48 basepi cedwards: just add a log.debug or a print before line 88 (where it stacktraced) to print whatever's in the variable `line`
18:48 UtahDave drawsmcgraw: Yes, a la Expect, but without the suck.   :)
18:49 UtahDave drawsmcgraw: but again, this first release can't do that yet, but the foundation is there and that ability will be added soon
18:49 basepi cedwards: we're obviously expecting very exact contents in `line`, and somehow those contents have changed
18:49 giantlock joined #salt
18:55 d10n joined #salt
18:55 linjan_ joined #salt
18:56 NotreDev joined #salt
18:57 cedwards basepi: I've added a print just above 88, but I don't see anything different in the output..
18:57 cedwards basepi: what am I missing?
18:57 kenbolton joined #salt
18:58 basepi cedwards: how did you start the minion?
18:58 aurigus is there a way I can run an external script in a state file, with return data, and set it as a variable (or grain), or do I have to write a module for that?
18:58 basepi any print statement you put in will manifest not in the log, but in the shell which started the minion.  if you did it via init scripts, it might be lost, you're better off starting it via `salt-minion -d` in a normal shell
18:59 basepi aurigus: so you want to put the return of that script into grains?
18:59 aurigus yes
19:00 aurigus I can set a grain with grains.present but not sure how to use the return call of a command
19:00 basepi aurigus: off the top of my head, i'm not sure of a way to repurpose the return in that way.  you're probably better off writing a custom grain which runs the script and stores the result.
19:00 cedwards basepi: salt-minion -d
19:00 KennethWilke aurigus: i think you can use jinja for that
19:01 aurigus ok, let me read up some on jinja
19:01 basepi cedwards: then the print should show up in that shell.  did you hit the traceback again?
19:01 KennethWilke grains:\n thingy: {{ __salt__['module.function']('do_stuff') }}
19:01 basepi ah, good call, KennethWilke
19:01 p3rror joined #salt
19:02 KennethWilke here's the first example that comes to mind on that: https://github.com/saltstack-formulas/mysql-formula/blob/master/mysql/package-map.jinja
19:02 KennethWilke which also makes me realize there's not __'s on that one
19:03 aurigus where does that go?
19:03 aat joined #salt
19:04 aurigus ah i see
19:05 KennethWilke i can slap a gist together if you want a more solid example
19:05 KennethWilke this kinda shows you how jinja has {% .. %} blocks and {{ ... }} blocks: http://jinja.pocoo.org/docs/templates/
19:06 cedwards basepi: ok. I see the output
19:06 KennethWilke the {{ ... }} ones will return whats ran inside as text into your template
19:06 aurigus got that part
19:07 aurigus where would i add the external command though, within the jinja file?
19:07 aurigus that example looked like static variables
19:07 cedwards basepi: it looks like if there are any other pending upgrades it's going to choke: http://dpaste.com/1360059/
19:08 basepi looks like the only thing that doesn't match the split is the '[salt]' on the end
19:08 jhenrypb1 joined #salt
19:09 basepi does it keep track of the fact that salt installed the package?  where does that '[salt]' come from?
19:09 cedwards basepi: the [salt] comes from the salt repo i have configured pointing to freebsd.saltstack.com
19:09 basepi cedwards: also, can you add that paste to the issue for terminalmage?  it's useful
19:09 basepi ah, gotcha
19:10 terminalmage cedwards: sorry, still head down working on djangocon
19:10 basepi terminalmage: hehe, no worries, i just saw you assigned yourself to it
19:10 basepi or.....someone assigned you to it
19:11 terminalmage we talking about that pkg.latest bug?
19:11 basepi ya, for FreeBSD
19:11 terminalmage ok
19:11 cedwards terminalmage: this is different than the previous one I submitted.
19:11 cedwards ..i think
19:12 basepi https://github.com/saltstack/salt/issues/6954
19:17 cedwards basepi: i discovered this issue using a git checkout. do you want me to test against develop again on a fresh machine?
19:17 basepi cedwards: might be useful.  when you say git checkout, i think `develop`.  what branch/tag did you checkout?
19:18 terminalmage ok
19:18 cedwards basepi: I didn't specify a branch/tag
19:19 cedwards i'm guessing just the latest develop
19:20 basepi i mean, i won't complain if we can narrow down the exact conditions of what will make it fail, cedwards, so it's up to you.  =)
19:21 basepi i'm going to lunch, so i'll reply to further mentions when i return
19:22 bhosmer joined #salt
19:24 mgw1 joined #salt
19:25 TheCodeA_ joined #salt
19:26 SunSparc joined #salt
19:27 chrisgh joined #salt
19:27 mgw1 joined #salt
19:29 jacksontj joined #salt
19:30 forrest joined #salt
19:32 jm-ck joined #salt
19:32 forrest shadowsun, you around?
19:32 nrub joined #salt
19:33 log0ymxm joined #salt
19:34 log0ymxm I'm running salt-minion 0.16.0 on a CentOS machine, it seems to hang at "Caching directory '_outputters/' for environment 'base'"
19:34 log0ymxm any pointers on what I can do to remedy this?
19:36 log0ymxm trying it with debug doesn't offer too much more info
19:38 whit joined #salt
19:39 Eugene Alcohol?
19:40 log0ymxm None present
19:41 log0ymxm looks like it's a potential bug in 0.16.0
19:45 UtahDave log0ymxm: any way you can test against 0.16.3?
19:45 log0ymxm UtahDave, trying that now
19:47 gldnspud joined #salt
19:49 gildegoma joined #salt
19:50 chrisgh joined #salt
19:50 carxwol joined #salt
19:51 log0ymxm UtahDave: stuck at "MinionEvent PULL socket URI: ipc:///var/run..."
19:51 log0ymxm CentOS release 5.9 (Final)
19:52 UtahDave log0ymxm: can you pasetbin the output of salt '<minion name>' test.versions_report
19:52 Linuturk I'm trying to come up with a series of state files that keep a set of servers patched, including salt itself. a set of maintenance states that handle updates, ssh keys, security of services (ssh and such). are there good state files out there for this now? any tips on writing these?
19:55 log0ymxm UtahDave: http://pastebin.com/kik88tVQ
19:55 log0ymxm empty response when running from the master
19:58 log0ymxm looks similar to this thread except I'm not running local, https://groups.google.com/forum/#!msg/salt-users/OL4Y9vczZRA/u32siqCWUpsJ
19:58 carxwol joined #salt
19:58 zooz joined #salt
19:59 teebes joined #salt
20:01 pdayton joined #salt
20:01 UtahDave log0ymxm: oh, ouch.
20:02 UtahDave log0ymxm: any way you can upgrade zmq?  zmq 2.1.x is horribly buggy
20:03 log0ymxm yeah looks like 2.2.0-4 is available in epel-testing
20:04 log0ymxm UtahDave: It's still getting stuck in the same spot
20:06 pdayton left #salt
20:06 carmony joined #salt
20:08 jacksontj joined #salt
20:11 UtahDave log0ymxm: have you made any changes to the default configs?
20:11 littleidea joined #salt
20:12 log0ymxm on the minion, I've just set the master & id
20:12 log0ymxm nothing out of the norm on the master either
20:12 log0ymxm I was able to accept this boxes key, no problem
20:12 conan_the_destro joined #salt
20:13 UtahDave both master and minion are on Salt 0.16.3 now?
20:13 log0ymxm yeah
20:13 Thiggy joined #salt
20:13 UtahDave are you running as root?
20:14 log0ymxm no
20:14 Thiggy If I use gitfs to store my states, does it map the master branch to base or do I have to make a base branch?
20:14 UtahDave Thiggy: Yeah, the master branch gets matched to the base environment.
20:15 UtahDave log0ymxm: try running salt as root
20:15 Thiggy @utahdave - ok thanks. Switching over to that today. Trying to learn all the things.
20:16 UtahDave LEARN ALL THE THINGS!!!!!!!
20:16 Thiggy YES!
20:17 Thiggy I'm setting up 2 environments, staging and production. Does this mean I should probably make a "master" branch that's completely empty and/or only contains the things that I want shared?
20:17 Thiggy (was planning on having a production branch and a staging branch)
20:18 juanlittledevil joined #salt
20:18 log0ymxm UtahDave, I think it might network or firewall issues, though it's accessible the master isn't on the same vpn, I'm going to testing that out.
20:20 Anb joined #salt
20:21 racooper joined #salt
20:25 carxwol joined #salt
20:25 Anb joined #salt
20:27 ronc joined #salt
20:31 drawsmcgraw UtahDave: Looking forward to future iterations of salt-ssh ! We've been looking into Python's paramiko module but we can certainly keep ourselves busy while we wait for salt-ssh to mature :)
20:36 UtahDave drawsmcgraw: that's cool!  It will be very useful
20:36 littleidea joined #salt
20:41 santagada joined #salt
20:42 dthom91 joined #salt
20:44 mgw joined #salt
20:46 carxwol joined #salt
20:50 jdenning joined #salt
20:56 matanya joined #salt
20:57 opapo joined #salt
20:57 alunduil joined #salt
21:02 jacksontj joined #salt
21:04 druonysuse joined #salt
21:07 Thiggy Is the gitfs stuff not the best at logging what it's doing? It seems to be doing a lot of silent failing even with debug logging turned on on the master
21:08 m_george left #salt
21:08 Thiggy The logs complained about the git-python not being installed, which I did, and now… nada
21:09 opapo joined #salt
21:11 jacksontj joined #salt
21:12 dthom91 joined #salt
21:14 TheCodeAssassin joined #salt
21:14 lemao joined #salt
21:16 UForgotten left #salt
21:16 Ryan_Lane joined #salt
21:18 mattbarto joined #salt
21:23 log0ymxm joined #salt
21:23 Anb_ joined #salt
21:24 chrisgh How do you preseed the mysql root password with Salt?
21:24 lemao joined #salt
21:24 ccase joined #salt
21:24 tseNkiN joined #salt
21:32 bashcoder Thiggy: are you using the develop branch of salt?
21:33 Thiggy I'm not, I'm currently on the .16.0, but that reminds me I should upgrade to .3
21:34 cwright how should a list be represented in a pillar file?
21:34 bashcoder If you can, please try using the develop branch by either using salt-bootstrap or cloning github. From my experience, 0.16.0 is broken for gitfs, and I think 0.16.3 is too.
21:35 bashcoder But I've seen several issues get resolved relating to it recently
21:35 Thiggy @bashcoder ok great thanks for the info, I knew I was getting a little close to the bleeding edge
21:36 bashcoder Haha - yes. Let me know how it works with develop.
21:40 mwillhite joined #salt
21:47 bwq joined #salt
21:50 littleidea joined #salt
21:51 bhosmer joined #salt
21:51 littleidea joined #salt
21:56 ggoZ left #salt
21:56 ggoZ joined #salt
21:57 Linuturk http://docs.saltstack.com/topics/troubleshooting/yaml_idiosyncrasies.html#yaml-does-not-like-double-short-decs This would have saved me tons of frusteration the other day
22:01 dthom91 joined #salt
22:02 kermit joined #salt
22:03 jdenning joined #salt
22:04 whit joined #salt
22:05 xt joined #salt
22:07 aat joined #salt
22:08 forrest yea yaml has some fun stuff
22:08 forrest like no leading 0s without quotes
22:14 aat joined #salt
22:16 tyler-baker joined #salt
22:16 tyler-baker joined #salt
22:17 wickster joined #salt
22:20 isomorphic joined #salt
22:21 oz_akan_ joined #salt
22:24 g4rlic left #salt
22:26 Jahkeup joined #salt
22:35 druonysus joined #salt
22:36 aat joined #salt
22:39 Jahkeup joined #salt
22:45 Gareth joined #salt
22:47 jacksontj joined #salt
22:51 pdayton joined #salt
22:53 bhosmer joined #salt
22:57 Ryan_Lane joined #salt
23:02 TheCodeAssassin joined #salt
23:05 kiorky joined #salt
23:09 cewood joined #salt
23:19 yota joined #salt
23:23 jacksontj joined #salt
23:28 aat joined #salt
23:29 berto- joined #salt
23:29 berto- i have a pillar that makes an HTTP request for data and am trying to track down why i get the error: `Targeted pillar "server_cluster" not found`  if i run `salt '*' pillar.data` i see the server_cluster pillar data listed.
23:30 berto- the command I'm running is: salt --compound 'I@default and I@roles:router' --return=console pillar.data
23:31 jams1 left #salt
23:36 emocakes joined #salt
23:41 carmony joined #salt
23:44 cyp joined #salt
23:47 juanlittledevil joined #salt
23:47 sgviking joined #salt
23:47 kenbolton joined #salt
23:51 ccase Would anyone be interested in a coffeescript style JSON format? I was thinking that this would be a nice replacement for YAML.
23:52 oz_akan_ joined #salt
23:54 littleidea joined #salt
23:59 pdayton joined #salt

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