Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-02-04

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

All times shown according to UTC.

Time Nick Message
00:01 frosty996 indeed!
00:05 devric joined #salt
00:05 honestly joined #salt
00:06 kermit joined #salt
00:06 chitown have a stupid question....
00:06 chitown when developing with the virtualenv
00:07 chitown do i have to keep using the "
00:07 chitown "-c" option every time?
00:07 chitown is there an easy way to change the default config path?
00:07 chitown ftr: i just followed HACKING.rst
00:14 Hipikat left #salt
00:14 mgw joined #salt
00:19 yomilk joined #salt
00:25 whyzgeek joined #salt
00:32 rager hi, guys
00:32 rager when I tell my minions to do the state.highstate thing, I don't get digests of what they've done any more
00:33 rager I'm not sure what's changed, but I used to get a list of all the actual things salt checks and runs on the minions
00:34 rager it seems to have started when I added other things to my top.sls file aside from just the base
00:35 rager http://hastebin.com/fedemuniya.vbs
00:36 rager my top.sls: http://hastebin.com/fedemuniya.vbs
00:37 honestly /ww/ww
00:37 honestly augh
00:37 austin987 joined #salt
00:37 rager when I run "salt '*' state.highstate" I still get the digest from the one minion that doesn't match things from the production/staging entries
00:47 bhosmer joined #salt
00:48 pdayton joined #salt
00:48 rager ok, found the issue - I was referring to statefiles in other directories incorrectly
00:48 rager foo.bar if it's foo/bar.sls
00:56 ajw0100 joined #salt
00:59 Gifflen joined #salt
01:11 babilen joined #salt
01:12 cedwards joined #salt
01:21 mgw cedwards: ping
01:25 cedwards mgw: hello
01:26 mgw cedwards: https://github.com/saltstack/salt/issues/9956 — I'm trying to track down the cause of that.
01:26 xzarth joined #salt
01:28 cedwards mgw: I don't have access to that machine at the moment
01:31 mgw cedwards: ok
01:31 cedwards terminalmage: ping
01:31 mgw was it only happening on one machine? Or did you not try it on others?
01:32 cedwards mgw: i didn't try it on any others.
01:32 mgw cedwards: ok, thanks
01:34 terminalmage cedwards: pong
01:34 terminalmage chitown: I use -c
01:34 terminalmage every time
01:34 terminalmage if you are using a shell with a history, it's not really that cumbersome
01:35 Hipikat joined #salt
01:36 cedwards terminalmage: I recently had this bug submitted to the FreeBSD port (http://www.freebsd.org/cgi/query-pr.cgi?pr=ports/186397)
01:36 terminalmage yeah for freebsdpkg?
01:36 cedwards terminalmage: wondered if you had any input on whether or not the patch was even still needed
01:36 terminalmage it is, for 0.17.x
01:36 terminalmage but not for after
01:36 mgw So… I created some never ending jobs with cmd.run (not intentionally). Any suggestions on how to quickly terminate them all? I don't have psutil installed on all the minions.
01:36 terminalmage since this is against 0.17.5, it's probably necessary
01:36 mgw Makes me wonder how hard it would be to have a kill function for the jobs runner
01:37 terminalmage cedwards: however, we're going to cut 2014.1.0 this week
01:37 terminalmage so, it's really a toss-up
01:38 jakubek joined #salt
01:38 cedwards hmm
01:41 jimallman joined #salt
01:41 Ryan_Lane how do I check the error status of git.fetch calls?
01:41 Ryan_Lane from another module
01:42 Ryan_Lane ugh. it throws an exception?
01:52 mbarrien joined #salt
02:02 n8n joined #salt
02:07 m87carlson oh hey cedwards, I submitted that PR for the FreeBSD salt report
02:08 m87carlson So currently, salt 0.17.4 on FreeBSD 10.0-RELEASE is broken, because by default pkgng on 10.0-RELEASE does not have /usr/local/etc/pkg.conf
02:09 m87carlson that was not the default behavior in 0.17.4 anyway, it was originally written to look at /var/db/pkg/local.sqlite, however, someone else had requested that salt should instead look for pkg.conf since they relocated their pkgng sqlitedb to somewhere else
02:13 m87carlson Part of me things that a port should not be broken out of the box, thats just bad, but I also agree that 2014.1 is so close that the port could just wait for that. However, the simple fix is to remove the sysutils/py-salt patch file for freebsdpkg.py to get 0.17.4 working on all platforms.
02:16 cedwards m87carlson: yeah. i've just been trying to recreate the issue on my end.
02:16 cedwards looks like it's broken either way on 9.2
02:16 taion809 joined #salt
02:17 m87carlson let me know if you want additional info, I've got a few 10.0-RELEASE systems now, one of them is pretty "clean" aside from running the salt bootstrap script
02:23 cedwards that's right. I already pulled that patch from the 2014.1.0rc3 port. should be ready to go when the final release arrives.
02:25 cedwards i really hope we can get the salt-call issue fixed in 2014.1 before the final
02:26 cedwards https://github.com/saltstack/salt/issues/10087
02:27 tedski joined #salt
02:28 tedski is there a way to reference nodegroups when templating a file?
02:33 m87carlson heh yeah that zfs output is a little annoying, but not critical :)
02:34 m87carlson oh, that one is different, it outputs the entire help. I might be able to dig into that
02:35 cedwards yeah. historically I've seen the "zfs help | :" one-liner, but it got worse in 2014.1
02:35 bhosmer joined #salt
02:36 tvle joined #salt
02:37 m87carlson why is the zfs help command ran? Just to determine if there is zfs support?
02:37 cedwards I'm not really sure tbh
02:38 m87carlson I mean testing the output of kldstat -m zfs could be just as good
02:39 m87carlson but /sbin/zfs help will forcibly load the zfs.ko as well
02:39 Gordonz joined #salt
02:39 ravibhure joined #salt
02:40 n8n joined #salt
02:41 nn0101_ joined #salt
02:44 Gordonz joined #salt
02:57 jalbretsen joined #salt
02:59 Dinde joined #salt
03:02 rager ok... more confusion
03:02 rager problems with
03:02 rager http://hastebin.com/matupenama.sm
03:02 malinoff joined #salt
03:03 rager The state "nginx" in sls rails-server is not formed as a list
03:03 rager am I better off just putting all those files in their own "state" blocks?
03:05 sgviking joined #salt
03:19 jkleckner joined #salt
03:20 rager also, is there a way to run a command iff a file is created for the first time?
03:20 rager well
03:20 rager iff a file is created when there wasn't one
03:21 dstanek joined #salt
03:22 jkleckner joined #salt
03:25 jasonrm joined #salt
03:28 cedwards rager: yeah. you should probably break that state up a bit
03:30 rager did that, works great, now
03:32 elfixit joined #salt
03:37 wendall911 left #salt
03:41 dstanek i have two states name 'app code'; one is a git.latest and the other is cmd.run; is there a way to require either?
03:46 rager require:
03:46 rager - cmd: app code
03:46 rager or however you deal with spaces
03:50 dstanek rager: but i want to be able to do something like ' - *: app code'
03:51 rager you want to require either/or?
03:52 rager not sure how to do that - still figuring this stuff out
03:55 zain joined #salt
03:55 dstanek rager: yeah, i have dev/stage servers pull code with git, but prod server pull code using a shell script
03:57 rager ooh
03:57 rager sounds like a job for...
03:57 rager require_in:
03:57 rager it reverses the dependency
03:57 rager just so long as only one of those states gets run per environment, it should be good
03:57 yomilk joined #salt
03:58 mgw joined #salt
03:58 rager either that or two different copies of the state you want to run on (A || B)
04:02 anitak joined #salt
04:03 dstanek rager: that won't work for me either because the thing that depends on the app code has the same problem - different implementations based on the environment
04:03 zain when I specify multiple gitfs_remotes, which of them does gitfs_root apply to?
04:05 rager in light of that, I'd just brute-froce it by having multiple copies of the state
04:05 rager one for each possible requirement
04:05 rager and wait for someone who isn't a noob to say something
04:05 dave_den dstanek: use jinja
04:07 rager it's ugly, but it works, dstanek
04:07 dstanek dave_den: that's an interesting idea
04:07 dstanek rager: that's what i'm doing now - trying to push the limits a bit :-)
04:14 EWDurbin dstanek: any room for jinja if/else in the mix?
04:15 EWDurbin seems preferable to the bizarre dependency setup
04:16 EWDurbin and since *something* has to trigger the alternate dependency why not make it explicit in the app code state?
04:16 dstanek EWDurbin: i think i'm going to add some pillar data like 'release_from: git' and instead of having release/git.sls and release/script.sls just having release.sls with as if/else
04:16 EWDurbin and just define the state differently
04:16 EWDurbin Yeah, that :)
04:17 dstanek then require would be: ' - {{ release_from}}: app code'
04:17 EWDurbin Hmmmmm, why not just require app code
04:18 EWDurbin And wrap up the gross in that state?
04:18 EWDurbin I do some similar hijinks for prod vs dev
04:19 dstanek EWDurbin: it's somewhat like a chinese menu based on roles there are certain things that need to be installed or run
04:19 EWDurbin And just have a state named {{ app }}-code that wraps an if around git vs the package
04:19 dstanek they all depend on having the code
04:20 dstanek EWDurbin: that's not realy any different though right? just a different require line
04:20 EWDurbin Yeah, so if your dependency alway had the same name, easier to keep track of
04:21 EWDurbin Also your always just requiring the code
04:22 EWDurbin Sad my open source salt repo doesn't have a good example yet :(
04:22 dstanek the dependency is always named 'app code' right now, but when I require it i still need to know git vs. cmd
04:23 bhosmer joined #salt
04:24 EWDurbin Seems like a weak abstraction, or some workflow you can't fix
04:26 EWDurbin can you gist more info?
04:26 EWDurbin I'm about to drop for night, but you can PM if you don't hate me yet
04:27 EWDurbin G'night dstanek
04:27 dstanek EWDurbin: :-) talk to you tomorrow
04:28 dstanek EWDurbin: have a good night
04:32 Voziv joined #salt
04:37 donlightsey joined #salt
04:52 sroegner joined #salt
04:56 salt_conf joined #salt
04:59 salt_conf this link consists of mypackage.sls http://pastebin.com/WJPdhtzF
04:59 salt_conf this is  my result http://pastebin.com/ZGsHmfGP
05:00 anuvrat joined #salt
05:00 salt_conf please let me know the error needs to corrected.
05:06 jeremyfelt joined #salt
05:08 tedski salt_conf: can you pastebin an ls -l /srv/salt ?
05:08 salt_conf yes
05:10 salt_conf http://pastebin.com/fe6kn0EP contains ls -l /srv/salt
05:11 tedski wait, nevermind
05:11 tedski found your error
05:11 tedski salt_conf: '*':  <-- you're missing the colon there
05:12 salt_conf can you let me know where?
05:12 tedski line 6 of http://pastebin.com/WJPdhtzF
05:12 tedski also, convention is 2-spaces for indentation with yaml in salt
05:12 tedski you only have 1 space per level on lines 6 and 7 of the same pastebin
05:13 kedo39 joined #salt
05:13 salt_conf ok i will correct and paste in the pastebin. please do verify.
05:14 salt_conf :tedski- please do verify this link http://pastebin.com/FAVvq9d5
05:16 tedski salt_conf: why did you add newlines?
05:16 tedski salt_conf: you'd probably be best served reading up on this: http://docs.saltstack.com/ref/states/top.html
05:17 tedski salt_conf: and this: http://docs.saltstack.com/topics/troubleshooting/yaml_idiosyncrasies.html
05:17 salt_conf thanks but please give me the corrected form of mypackage,sls. which would help me in man\y ways
05:18 tedski i already did
05:20 salt_conf but i am getting the same error
05:20 tedski because you didn't take my advice, you broke your top.sls further
05:21 tedski once you familiarize yourself with the documentation, you'll see your error quite easily
05:21 salt_conf i am new to saltstack
05:21 tedski that's okay, the learning curve is nice and shallow
05:22 tedski read back a few lines and look at your original paste and then see what i told you to fix
05:22 salt_conf <tedski> you only have 1 space per level on lines 6 and 7 of the same pastebin. i am confused with it. if possible can you let me know more clearly
05:23 tedski your indentation is only 1 space
05:23 tedski you need 2 spaces per level of indentation
05:23 tedski base:
05:23 tedski '*':
05:23 tedski etc.
05:24 salt_conf for all below the base ?
05:24 Ryan_Lane joined #salt
05:24 tedski in the topfile, yes
05:27 rager hi... I'm getting more and more lost the more I edit my statefiles
05:27 rager http://hastebin.com/komelatebo.sm
05:27 rager if you have a watch that lists a bunch of different files
05:27 rager does it have to have *each* one change for it to fire off something like a service restart?
05:28 salt_conf :fedski- i have pasted once more please verify this link http://pastebin.com/WtqiRTBd
05:29 salt_conf my error is this link http://pastebin.com/ZtjaT040
05:30 tedski rager: nope, only one of the watched states needs to return changes
05:31 rager also, I've noticed that pkg.latest doesn't fulfill require... pkg: [pkgname]
05:32 rager or am I going insane?
05:33 tedski salt_conf: you're still using invalid yaml and i suspect that you're trolling because every time you pastebin something, you change everything instead of just what you say you're going to change.
05:33 tedski salt_conf: i'm going to kindly opt out of feeding the trolls.
05:34 rager my sls, for reference
05:34 rager http://hastebin.com/soretefuxi.sm
05:34 rager I keep getting stuff that says that things watching nginx don't have all the requisites
05:35 elfixit joined #salt
05:35 rager dpkg -l | grep nginx
05:35 rager shows that it's installed
05:35 tedski rager: you're mixing up ids vs pkg names
05:35 rager ids vs pkgnames?
05:36 tedski watch pkg: rsyslog says watch the pkg state with id 'rsyslog'
05:36 tedski your rsyslog id only has service and file states
05:36 rager ok
05:36 tedski the rsyslog package is installed by the pkg state of id debian-packages
05:36 rager so I'll break out the pkgs that need to be watched into the states
05:37 tedski yeah
05:38 ravibhure joined #salt
05:40 rager similarly, if I want to watch a directory that's recursed, do I just name its id?
05:40 tedski yes
05:40 tedski requisites are by id declaration
05:40 tedski well
05:41 tedski they're also on the name parameter
05:41 fllr joined #salt
05:42 tedski but your file recurse states don't have a name parameter since it's assumed as the id
05:45 rager is there a way to specify that certain files should be passed over in a file.recurse?
05:45 rager or is that asking too much?
05:45 rager or that certain files are left behind after cleaning up after a file.recurse
05:46 rager wait... exclude_pat
05:46 tedski :)
05:47 rager does that also paint the blood on the doorways for files that would be wiped out by the "clean: True" directive?
05:47 rager (trying, now...)
05:48 rager oh, silly me
05:48 rager dox says that yes, they're also cleared for passover when clean: True
05:52 pydanny joined #salt
05:56 ajw0100 joined #salt
06:12 bhosmer joined #salt
06:12 goodwill joined #salt
06:14 Psyke joined #salt
06:18 joehoyle joined #salt
06:19 joehoyle Hey - is there a way I can add another master host to all my minions? From master 1, I want to do: salt \* add-host "master 2"
06:23 MTecknology joehoyle: you mean spin up a new master and make all hosts use that master as well?
06:24 linjan joined #salt
06:24 joehoyle MTecknology: right. I have the new master already
06:24 MTecknology /etc/salt/minion.d/master.conf: \n\t file.managed: \n\t\t - source://files/minion_masters.conf
06:25 MTecknology bye :(
06:30 Ryan_Lane1 joined #salt
06:31 nebuchad` joined #salt
06:31 fatbox joined #salt
06:32 Voziv_ joined #salt
06:32 Gifflen_ joined #salt
06:32 jY joined #salt
06:32 ferai joined #salt
06:33 jimallman_ joined #salt
06:33 Xiao joined #salt
06:33 ampex_ joined #salt
06:33 rstone_ joined #salt
06:33 tru_tru_ joined #salt
06:34 emostar joined #salt
06:34 namtab_ joined #salt
06:34 xzarth_ joined #salt
06:34 Dinde851 joined #salt
06:34 cetex joined #salt
06:34 luminous_ joined #salt
06:37 EWDurbin_ joined #salt
06:37 fab0109 joined #salt
06:38 brucewang joined #salt
06:38 z3uS| joined #salt
06:40 sroegner joined #salt
06:40 oeuftete joined #salt
06:45 Nazca joined #salt
06:47 yota joined #salt
06:48 tzero joined #salt
06:54 mordred joined #salt
07:07 kolaman joined #salt
07:16 Corey Good morning.
07:18 MrTango joined #salt
07:19 Katafalkas joined #salt
07:26 fllr joined #salt
07:27 wendall911 joined #salt
07:27 gammalget joined #salt
07:28 _fllr_ joined #salt
07:30 tomspur joined #salt
07:31 unicolet joined #salt
07:35 NV joined #salt
07:37 Franz__ joined #salt
07:46 Iwirada joined #salt
07:46 Shenril joined #salt
07:53 unicolet left #salt
08:00 bhosmer joined #salt
08:04 goodwill o_O
08:04 goodwill morning
08:04 goodwill ?
08:05 balboah joined #salt
08:06 linjan ye, morning
08:06 * linjan looks at clock: 12:06
08:06 linjan at watch*
08:10 kedo39 http://www.total-knowledge.com/~ilya/mips/ugt.html
08:10 kedo39 it's always morning~
08:14 harobed_ joined #salt
08:15 martoss joined #salt
08:18 n8n joined #salt
08:18 Corey linjan: 9:18AM here. :-)
08:18 Corey Today has "long day" splattered all over it.
08:18 linjan kedo39: best article, bookmarked
08:18 Corey I've been up since 4.
08:18 Corey Will get to my hotel around 8PM today.
08:18 Corey ...while gaining 9 hours to my day.
08:25 zooz joined #salt
08:26 helminthe joined #salt
08:28 mike25ro joined #salt
08:28 fllr joined #salt
08:29 sroegner joined #salt
08:29 mike25ro hi guys ... is there a way of tracking what SALT does? for example all commands that are run on the master... we are 3 admins that share the salt-master and i want to make sure.. some tracking is in place..
08:36 JohnHicks joined #salt
08:43 Corey Check the logging subsystem.
08:43 kalloc joined #salt
08:44 yomilk joined #salt
08:46 giantlock_ joined #salt
08:48 kalloc joined #salt
08:49 kalloc_ joined #salt
08:54 jakubek joined #salt
09:00 zimo joined #salt
09:00 zimo hi , it's interesting to have salt able de manage configuration on network devices like cisco switch !
09:01 zimo if they are some work in this direction ?
09:07 miguitas joined #salt
09:18 miguitas joined #salt
09:20 xmj why would they?
09:20 xmj they can't sell it.
09:24 krak3n` joined #salt
09:28 fllr joined #salt
09:29 yomilk joined #salt
09:31 ninkotech joined #salt
09:32 zimo xmj: to be the best ! ;)
09:32 scott_w joined #salt
09:32 joehoyle joined #salt
09:35 simonmcc zimo: yes, salt are working on a salt proxy objects, which can be used to manage devices like switches, afaik, Juniper support is in the works
09:35 zimo puppet for example has module for that, it's very usefull for Admin sys !
09:36 simonmcc apparently cro is working on the proxy support
09:36 zimo simonmcc: good ! :) github ?
09:38 kalloc joined #salt
09:38 simonmcc zimo dunno, sorry, can't see anything at first glance
09:39 zimo simonmcc: okay, thx man !
09:40 kalloc_ joined #salt
09:40 zooz joined #salt
09:48 bhosmer joined #salt
09:55 kalloc joined #salt
10:16 ckao joined #salt
10:16 mbarrien joined #salt
10:17 sroegner joined #salt
10:18 devric joined #salt
10:27 Shenril joined #salt
10:27 n8n joined #salt
10:28 fllr joined #salt
10:30 n8n joined #salt
10:32 n8n joined #salt
10:34 robthoms joined #salt
10:34 Katafalkas joined #salt
10:37 n8n joined #salt
10:39 n8n joined #salt
10:42 n8n joined #salt
10:45 n8n joined #salt
10:48 n8n joined #salt
10:49 jeddi joined #salt
10:49 n8n joined #salt
10:52 che-arne joined #salt
10:52 n8n joined #salt
10:54 giantlock_ joined #salt
10:56 n8n joined #salt
10:56 linjan_ joined #salt
10:58 zzzirk joined #salt
10:59 n8n joined #salt
11:02 n8n joined #salt
11:04 fab0109 left #salt
11:05 n8n joined #salt
11:07 n8n joined #salt
11:08 n8n joined #salt
11:09 n8n joined #salt
11:09 shiin joined #salt
11:12 n8n joined #salt
11:14 devric joined #salt
11:17 n8n joined #salt
11:18 n8n joined #salt
11:19 shiin welp. debian squeeze doesnt use bash, but dash. that broke the /etc/init/salt-minion.sh
11:20 n8n joined #salt
11:23 n8n joined #salt
11:24 n8n joined #salt
11:25 n8n joined #salt
11:26 n8n joined #salt
11:27 n8n joined #salt
11:28 fllr joined #salt
11:31 n8n joined #salt
11:32 n8n joined #salt
11:34 n8n joined #salt
11:35 Flusher joined #salt
11:37 bhosmer joined #salt
11:38 n8n joined #salt
11:39 n8n joined #salt
11:42 n8n joined #salt
11:43 n8n joined #salt
11:45 n8n joined #salt
11:46 77CABBGX4 joined #salt
11:46 djaykay joined #salt
11:49 n8n joined #salt
11:51 n8n joined #salt
11:52 sroegner joined #salt
11:56 n8n joined #salt
11:57 joehh shiin: did you install via packages or otherwise?
11:58 shiin via packages? I added the repo and did apt-get install
11:58 n8n joined #salt
12:04 n8n joined #salt
12:06 n8n joined #salt
12:07 justBob joined #salt
12:08 n8n joined #salt
12:09 n8n joined #salt
12:12 n8n joined #salt
12:15 n8n joined #salt
12:16 n8n joined #salt
12:17 n8n joined #salt
12:18 n8n joined #salt
12:20 n8n joined #salt
12:24 n8n joined #salt
12:28 fllr joined #salt
12:29 n8n joined #salt
12:31 joehh shiin: are you referring to /etc/init.d/salt-minion or a different file?
12:31 shiin /etc/init/salt-minion.sh
12:31 shiin pardon me
12:31 shiin its /etc/init/salt-minion.conf
12:31 shiin but the contents are some form of shell script
12:32 joehh ahhh -makes more sense :)
12:32 n8n joined #salt
12:32 joehh that is for systems running upstart
12:33 joehh not actually a shell script directly
12:33 joehh are you trying to run it directly or via
12:33 joehh service salt-minion start or similar
12:33 joehh ?
12:33 shiin service, yes. salt itself does that too
12:34 shiin but if service salt-minion status fails due to this /etc/init/salt-minion.conf it tries to start itself and fails.
12:34 n8n joined #salt
12:34 joehh ok
12:34 shiin which is kinda redundant anyway, but no one else can use service salt anymore either
12:35 joehh This file is ringing some bells - I'm just looking back over my notes
12:35 joehh I have a feeling that this file shouldn't be on squeeze systems
12:35 joehh but has snuck back in
12:37 ravibhure1 joined #salt
12:39 shiin it seems inconsistent on all my squeeze machines
12:39 shiin but whenever the problem occurs I just moved it away
12:39 shiin to fix the problem
12:40 shiin I could now make a state like file.absent. heh.
12:41 n8n joined #salt
12:45 joehh nice - I'll check a bit deeper, but will probably remove the file in the next upload
12:45 n8n joined #salt
12:45 joehh thanks for the report
12:47 n8n joined #salt
12:47 shiin welcome~ and thank you for hearing me out.
12:49 n8n joined #salt
12:51 skullone joined #salt
12:51 n8n joined #salt
12:54 n8n joined #salt
12:57 n8n joined #salt
12:57 krak3n` joined #salt
12:58 n8n joined #salt
13:00 n8n joined #salt
13:01 luminous joined #salt
13:01 jrdx joined #salt
13:02 n8n joined #salt
13:04 n8n joined #salt
13:05 n8n joined #salt
13:06 n8n joined #salt
13:08 n8n joined #salt
13:12 n8n joined #salt
13:13 n8n joined #salt
13:16 n8n joined #salt
13:17 n8n joined #salt
13:18 n8n joined #salt
13:19 n8n joined #salt
13:21 n8n joined #salt
13:21 fllr joined #salt
13:21 Flusher joined #salt
13:22 n8n joined #salt
13:23 n8n joined #salt
13:24 n8n joined #salt
13:25 bhosmer joined #salt
13:26 n8n joined #salt
13:27 dcmorton joined #salt
13:28 fllr joined #salt
13:28 n8n joined #salt
13:29 Ahlee mornin'
13:29 n8n joined #salt
13:30 n8n joined #salt
13:31 n8n joined #salt
13:32 n8n joined #salt
13:34 n8n joined #salt
13:37 n8n joined #salt
13:38 n8n joined #salt
13:39 sroegner joined #salt
13:39 n8n joined #salt
13:40 n8n joined #salt
13:41 n8n joined #salt
13:42 mgw joined #salt
13:43 rmnuvg joined #salt
13:44 n8n joined #salt
13:48 n8n joined #salt
13:49 ipmb joined #salt
13:50 n8n joined #salt
13:50 Poooogles joined #salt
13:57 elfixit joined #salt
13:57 n8n joined #salt
14:00 n8n joined #salt
14:04 n8n joined #salt
14:05 AdamSewell joined #salt
14:05 AdamSewell joined #salt
14:06 jeremyBass1 joined #salt
14:06 Gifflen joined #salt
14:08 n8n joined #salt
14:08 sroegner joined #salt
14:09 zzzirk joined #salt
14:10 n8n joined #salt
14:11 jmccree Salt now creates my monitoring rules for me.. It's so... beautiful.
14:12 n8n joined #salt
14:13 samkottler joined #salt
14:14 scalability-junk jmccree: share the code ;)
14:15 juicer2 joined #salt
14:16 jmccree scalability-junk, would if I could. :)
14:18 scalability-junk :(
14:18 danielbachhuber joined #salt
14:19 n8n joined #salt
14:20 n8n joined #salt
14:20 mpanetta joined #salt
14:22 n8n joined #salt
14:24 n8n joined #salt
14:27 kalloc joined #salt
14:28 n8n joined #salt
14:28 fllr joined #salt
14:29 pydanny joined #salt
14:29 n8n joined #salt
14:29 toastedpenguin joined #salt
14:29 toastedpenguin2 joined #salt
14:30 racooper joined #salt
14:30 rojem joined #salt
14:31 faldridge joined #salt
14:31 toastedpenguin joined #salt
14:31 n8n joined #salt
14:32 n8n joined #salt
14:34 n8n joined #salt
14:35 n8n joined #salt
14:36 Furao joined #salt
14:38 gadams999 joined #salt
14:39 n8n joined #salt
14:42 n8n joined #salt
14:46 yomilk joined #salt
14:47 tessellare joined #salt
14:47 iri- joined #salt
14:48 iri- How do I template a configuration file according to all of the nodes salt is aware of?
14:48 iri- I'd just like to get a list of strings of the FQDNs
14:48 iri- I sort of imagine I might be able to do it with the salt mine but the documentation is really not clear at all what you might do to achieve that.
14:48 iri- Like, how to automatically update the salt mine for instance.
14:48 n8n joined #salt
14:48 kermit joined #salt
14:49 iri- I'm reading http://salt.readthedocs.org/en/latest/topics/mine/index.html and http://salt.readthedocs.org/en/latest/ref/modules/all/salt.modules.mine.html#module-salt.modules.mine
14:49 krak3n` try this iri- http://stackoverflow.com/questions/17158665/how-to-get-a-list-of-all-salt-minions-in-a-template
14:50 tyler-baker joined #salt
14:50 tyler-baker joined #salt
14:50 blee joined #salt
14:50 n8n joined #salt
14:51 dccc_ joined #salt
14:52 brianhicks joined #salt
14:54 n8n joined #salt
14:54 iri- krak3n`: Strange, when I try to use network.get_hostname via publish.publish, nothing happens.
14:55 n8n joined #salt
14:55 iri- krak3n`: if I run it with `salt '*' network.get_hostname` it says "network.get_hostname" is not available.
14:55 jimallman joined #salt
14:57 n8n joined #salt
14:58 iri- krak3n`: I also don't understand how `for host in salt['publish.publish'](...)` works. If I run it at the command line, it just shows the IP address of the machine. How does `host.fdqn` get set?
14:58 fishpen0 joined #salt
15:02 blee_ joined #salt
15:02 n8n joined #salt
15:03 kalloc joined #salt
15:03 mike25ro stupid Q guys: i want to install 2 packages: glibc.i686 and libgcc.i686 ... and from SALT i get the error: The following package(s) were not found, and no possible matches were found in the package db: glibc.i686, libgcc.i686   ... i THINK it is because of the DOT in the package name... is this a known bug? i am using salt 0.17.1
15:04 n8n joined #salt
15:04 oz_akan_ joined #salt
15:04 krak3n` joined #salt
15:05 krak3n` @iri- start small, add test.ping to the mine_functions in your salt minion configs, restart the minions and then salt '*' mine.get '*' test.ping should work
15:05 krak3n` i think thats the right syntax
15:05 krak3n` build from there to get the result you wany
15:05 krak3n` want*
15:06 iri- krak3n`: what does mine_functions have to do with publish.publish?
15:07 n8n joined #salt
15:07 yota where I put my own states module?
15:07 iri- krak3n`: also, I'm not asking about the salt mine configuration, I'm just curious how the host.fqdn can possibly be available given the output I see when I run the network.ip_addrs module from the command line
15:07 yota in _modules?
15:08 yota _states
15:08 yota sorry :(
15:08 GradysGhost joined #salt
15:11 JasonSwindle joined #salt
15:12 MTecknology mike25ro: Are you using a state or the module from a salt command?
15:13 bhosmer joined #salt
15:13 warmwaffles joined #salt
15:15 jslatts joined #salt
15:18 Katafalkas joined #salt
15:19 dwiden joined #salt
15:19 dwiden I'm trying to install salt with "brew install salt", but I'm getting an error.  Can anyone help me?
15:20 anitak joined #salt
15:20 warmwaffles I'm having issues getting Salt bootstrapped on my Archlinux vagrant instance https://gist.github.com/warmwaffles/8805645 any pointers?
15:21 EWDurbin warmwaffles: Vagrant file?
15:22 warmwaffles updated the gist EWDurbin
15:23 warmwaffles The link to the vagrant box was just added
15:24 EWDurbin warmwaffles: try hopping into the vagrant box and https://github.com/saltstack/salt-bootstrap#i-found-a-bug
15:24 mike25ro MTecknology: i am using a STATE
15:25 warmwaffles I shall attempt that
15:25 mike25ro MTecknology: the package is in the repo... i can install it manually ... but from a state it fails
15:25 nn0101 joined #salt
15:25 MTecknology why all of the capitalization?
15:25 MTecknology mike25ro: put your state on dpaste.com
15:26 mike25ro ok
15:27 mike25ro MTecknology:  http://dpaste.com/1585265/
15:27 mike25ro MTecknology: thanks for taking the time to look into it.
15:28 mike25ro MTecknology:  the server i am deploying is a centos 6.4 x64 and for some reason some developers need both x86 and x64 versions of glibc
15:29 MTecknology mike25ro: on the minion, run  salt-call state.sls <that.state>  (what you put in your top.sls file
15:30 EWDurbin MTecknology: what was the salt version again?
15:30 MTecknology with centos, doesn't yum install <pkg> install <pkg>.{i686,amd64} by default?
15:30 higgs001 joined #salt
15:30 MTecknology EWDurbin: his? 0.17.1
15:30 EWDurbin whoops yeah
15:31 EWDurbin my scrollback is missing mike25ro's error as well
15:31 MTecknology 09:03 < mike25ro> stupid Q guys: i want to install 2 packages: glibc.i686 and libgcc.i686 ... and from SALT i get the error: The following package(s) were not found, and no possible matches were found in the package db: glibc.i686, libgcc.i686   ... i THINK it is because of the DOT in the package name... is this a known bug? i am using salt 0.17.1
15:31 EWDurbin thanks
15:32 grep_awesome joined #salt
15:32 Katafalkas joined #salt
15:33 Katafalkas joined #salt
15:34 mike25ro MTecknology:  but for any package... that has a dot in the name... the state fails
15:34 mike25ro doesn't matter is glibc or whatever
15:35 EWDurbin mike25ro: can you try with a newer version of salt? current EPEL is 0.17.4
15:35 EWDurbin i'm checking to see where the bug i remember with package archs was
15:35 Katafalk_ joined #salt
15:35 EWDurbin i got hit with similar issues
15:36 mike25ro EWDurbin: i can workaround... like creating an shell script that runs a yum install x y z  ... but i wanted to be sure is a bug/feature :) and not me being stupid
15:37 EWDurbin that's why i'd reccommend trying a newer release of salt first
15:37 EWDurbin if it's still broken in 0.17.4, it's a bug that needs more attention :)
15:37 alunduil joined #salt
15:38 Katafal__ joined #salt
15:38 rgbkrk joined #salt
15:38 Bosse joined #salt
15:40 EWDurbin btw, just worked fine for me on 0.17.2
15:40 EWDurbin going to try to repro on 0.17.1
15:41 mike25ro EWDurbin: don't waste your time with this crap... i just wanted to know .. that is not only @ my end... do something else.. i can work around this
15:41 mike25ro EWDurbin: MTecknology thanks guys
15:41 EWDurbin it is reproducible on 0.17.1
15:43 EWDurbin also works fine on 0.17.4, so it is a resolved bug mike25ro
15:44 Gifflen joined #salt
15:44 esogas_ joined #salt
15:45 anuvrat joined #salt
15:46 rojem joined #salt
15:47 mike25ro EWDurbin: thanks, i will try and update
15:49 fllr joined #salt
15:50 ekristen joined #salt
15:50 cnelsonsic joined #salt
15:50 ekristen how can I make salt stack uncomment the backports repo in ubuntu?
15:51 GradysGhost joined #salt
15:51 anitak joined #salt
15:51 btorch joined #salt
15:52 Ahlee ekristen: better to take control of hte file with file.managed then try to manipulate parts of the file
15:53 Linuturk ^ +1
15:53 Linuturk http://docs.saltstack.com/ref/modules/all/salt.modules.apt.html#module-salt.modules.apt
15:54 Linuturk might want to look at this module as well
15:54 Linuturk it has some aspects of repository management
15:54 gadams999 left #salt
15:58 a__jennings joined #salt
16:02 esogas left #salt
16:04 pydanny joined #salt
16:05 warmwaffles joined #salt
16:07 dave_den ekristen: i would recommend this: http://docs.saltstack.com/ref/states/all/salt.states.pkgrepo.html
16:08 dave_den ekristen: otherwise, if you just want to go the uncomment route, you can use http://docs.saltstack.com/ref/states/all/salt.states.file.html#salt.states.file.uncomment
16:09 Marion joined #salt
16:09 dave_den joined #salt
16:10 Marion left #salt
16:14 ekristen thanks dave_den
16:17 shiin left #salt
16:17 tempspace joined #salt
16:18 JasonSwindle joined #salt
16:33 m87carlson joined #salt
16:37 jkleckner joined #salt
16:37 Psyke joined #salt
16:38 jalbretsen joined #salt
16:39 jkleckner joined #salt
16:39 fllr How do I get the hostnames of all the machines connected to my salt master? I'm already using mines, and network.interfaces, but I've only been able to get the ip addresses of the machines.
16:40 GradysGhost fllr: Try this... salt '*' cmd.run 'hostname'
16:40 GradysGhost You can also use grains to get the fqdn
16:40 GradysGhost salt '*' grains.get fqdn
16:41 fllr GradysGhost: I like the grain idea...
16:42 fllr GradysGhost: this is my hosts file that I'm trying to generate with mines. How would it change? https://gist.github.com/felipellrocha/4da41cb6de08cab6507a
16:42 agend joined #salt
16:42 fejjerai joined #salt
16:43 kermit joined #salt
16:45 GradysGhost fllr: I'm not sure. I've been using a pillar to define my nodes. It's more explicit that way, too. Changes to /etc/hosts can result in different fqdns, so having a static list of nodes ensures no changes.
16:48 fllr GradysGhost: I see... thanks, though...
16:48 MTecknology fllr: I've been wanting to do something like that, but for backup scripts
16:48 Kenzor joined #salt
16:48 Katafalkas joined #salt
16:49 fllr MTecknology: Yeah... I just think it'd be nice to just add a node to my master, and know that my hosts file would be up-to-date...
16:50 MTecknology I really only need a list of accepted keys, but I should probably be pulling it from the mine
16:51 dave_den fllr: if your minion id's already match the hostname, then you already have that info
16:51 quickdry21 joined #salt
16:52 fllr dave_den: They do. How so, though?
16:52 dave_den "for host, hostinfo"…  the host variable in your jinja loop is the hostname.
16:52 fllr Ooooh. dumb... yes...
16:53 fllr Yessss! it worked! :D thanks
16:53 dave_den no prob :)
16:55 Ryan_Lane joined #salt
16:55 jslatts joined #salt
16:55 pydanny joined #salt
16:57 troyready joined #salt
16:57 jeremyfelt joined #salt
16:57 xmltok joined #salt
16:59 fllr Oh... it looks like if I add grains.items to my mine_functions I get all that information that I need, all very nicely parsed. I'm gonna use that! :D
17:00 Whissi joined #salt
17:00 diegows joined #salt
17:01 bhosmer joined #salt
17:08 mattmtl joined #salt
17:09 mattmtl left #salt
17:10 racooper joined #salt
17:12 rojem joined #salt
17:14 miguitas Hi guys, I am finishing a nagios module for salt, my question is what is the best way to send to the project: pull request, gist, issue + pull request...?
17:15 KyleG joined #salt
17:15 KyleG joined #salt
17:19 cedwards miguitas: github pull requests are usually looked at pretty quickly
17:19 miguitas cedwards: thanks
17:20 anitakrueger hey guys…has anyone tried to use salt-bootstrap with XenServer?
17:21 Iwirada left #salt
17:22 tvle joined #salt
17:27 mgw joined #salt
17:27 alexthegraham joined #salt
17:42 rlarkin joined #salt
17:44 foxx joined #salt
17:45 jpaetzel_ joined #salt
17:45 Bosch[] joined #salt
17:45 ze-_ joined #salt
17:46 Troic_ joined #salt
17:46 Troic_ joined #salt
17:46 BenCoinanke_d joined #salt
17:46 Ahlee joined #salt
17:46 tsantero joined #salt
17:46 delkins joined #salt
17:46 gwmngilfen joined #salt
17:47 sinh__ joined #salt
17:47 Linuturk joined #salt
17:47 packeteer joined #salt
17:48 sirtaj joined #salt
17:48 canci joined #salt
17:48 rockey joined #salt
17:48 Gordonz joined #salt
17:48 svx joined #salt
17:48 sashka_ua joined #salt
17:49 lionel joined #salt
17:49 pt|Zool joined #salt
17:51 Nazzy anitakrueger, bootstrapping the Dom0 or bootstrapping a DomU?
17:51 BogdanR joined #salt
17:52 rogst joined #salt
17:52 zach Hmm, with salt-cloud it would be great if it created the cname as well
17:53 anitakrueger I'm trying to install on dom0
17:54 anitakrueger thought I'd like to manage it with my salt master :)
17:54 bemehow joined #salt
17:54 martoss joined #salt
17:56 fejjerai joined #salt
17:56 scoates is there a way to get `salt -l debug nodename state.highstate` to report in realtime, instead of a huge dump at the end of the run? I'd like to see progress on slow-provisioning nodes.
17:56 zach automating a jar installer from salt: http://i.imgur.com/RsXb7XG.jpg
17:57 rojem joined #salt
17:58 Nazzy anitakrueger, honestly I wouldn't attempt to install salt on XenServer's Dom0 ... it's reaaaally fragile ime
17:59 anitakrueger @Nazzy…well…I would have liked to have my whole standard setup on there as well…monitoring agent, logstash and such…so I thought if I could only install a minion on it...
18:00 kasanop1 joined #salt
18:01 jacksontj joined #salt
18:01 tvle joined #salt
18:09 amckinley joined #salt
18:09 kalloc_ joined #salt
18:10 jdenning joined #salt
18:11 n8n joined #salt
18:12 jdenning Is anybody using GitFS file backend to distribute custom execution modules?
18:13 amckinley1 joined #salt
18:16 Nazzy anitakrueger, it'll run out of memory probably ... xenserver's dom0 is usually stupidly light on resources
18:17 anitakrueger hmm…well, I can't even get the minion installed...
18:17 Nazzy *nod* they use a custom repo iirc
18:17 Nazzy you might be able to control it via salt-ssh
18:27 mgw joined #salt
18:31 Gareth morning
18:37 jslatts joined #salt
18:39 synfin joined #salt
18:41 kermit joined #salt
18:41 bitz joined #salt
18:43 Ryan_Lane1 joined #salt
18:43 felixhummel joined #salt
18:45 Gordonz joined #salt
18:48 jimallman joined #salt
18:49 bhosmer joined #salt
18:51 jacksontj joined #salt
18:52 Poooogles joined #salt
18:56 Ahlee or proxy minion
18:57 dancat joined #salt
18:58 forrest joined #salt
19:00 Gordonz joined #salt
19:01 JasonSwindle joined #salt
19:01 Gordonz joined #salt
19:03 torrancew joined #salt
19:03 frosty996 left #salt
19:05 jimallman joined #salt
19:05 amckinley joined #salt
19:13 alexthegraham At SaltConf someone mentioned that there was (or would be in 2014) a way to list minions that had been accepted but aren't currently connected. Anyone have any details on that?
19:14 alexthegraham I've been using an alias: unsalted='diff <(salt '\''*'\'' test.ping | grep -v True | cut -f1 -d: | sort) <(salt-key -L | grep -v Keys | sort)'
19:16 cedwards alexthegraham: if I understand you correctly, it sounds like you're looking for 'salt-run manage.down'
19:16 dave_den alexthegraham: you can use http://docs.saltstack.com/ref/runners/all/salt.runners.manage.html
19:16 alexthegraham Bingo.
19:16 zach that's neat, never seen it
19:16 Ryan_Lane1 is there any way to get a list of minions that should match a grain from the mine?
19:16 cedwards manage.up and manage.down will compare keys with available/unavailable minions and give you a list
19:16 Ryan_Lane1 whether they are up or down? :)
19:18 dave_den Ryan_Lane: from a runner, you can use http://docs.saltstack.com/ref/runners/all/salt.runners.cache.html to inspect the mine
19:18 mgw joined #salt
19:19 Ryan_Lane1 dave_den: cool
19:19 clearclaw If I set a grain with grains.present in a statefile (say one that isn't under top.sls for that host), what is the expected lifetime of that grain?
19:19 clearclaw I'm finding that I can set a grain with grains.present via `salt foo state.sls ...` but that any subsequent highstate will erase that grain, despite the fact that none of the states under top.sls change the grain...  Is this expected?
19:21 retran joined #salt
19:22 clearclaw Also, has anything happened on the develop branch that could affect creating EC2 instances with ephemeral storage?
19:22 clearclaw The same configs that worked for me a fortnight ago are now building images with no ephemeral storage for me now...and I don't seem to be able to get them to create ephemeral storage for blood nor money.
19:24 jacksontj joined #salt
19:31 joehoyle joined #salt
19:32 Eugene Try a sacrifice to the booze gods
19:33 kermit joined #salt
19:34 kermit joined #salt
19:34 clearclaw Did that.  The bottle of Laphroaig on my desk is noticeably lighter...
19:34 scoates there's yer problem. booze gods don't appreciate peatiness
19:34 * clearclaw is now staring adoringly at a bottle of 1921 anejo tequilla...
19:37 retran i thought only whisky has peaty attributes
19:39 Mua joined #salt
19:39 Corey retran: You would be wrong. :-)
19:45 mr_chris joined #salt
19:46 Ryan_Lane1 joined #salt
19:50 elfixit joined #salt
19:51 sroegner clearclaw: have you ever been able to use more than one ephemeral device per instance?
19:53 clearclaw sroegner: I've never tried.  I've only needed the one.
19:54 sroegner darn
19:54 sroegner https://github.com/saltstack/salt/issues/8631
19:54 sroegner nvmd then
19:54 clearclaw Ooof.  That's a pisser.
19:55 clearclaw I did find that the value of DeviceName was seemingly irrelevant.
19:55 sroegner no kidding
19:56 clearclaw (No matter what I set it to it ended up at the same device)
19:56 sroegner i still wonder if it is just some yaml magic i am missing
19:56 Ahlee going back to the manage.{status,up,down}, anybody happen to remember what the 'present' falls under? i thought it was part of the reactor, but not seeing it
19:56 JasonSwindle left #salt
19:56 clearclaw That's always the haunting suspicion.
20:00 dwiden joined #salt
20:02 zooz joined #salt
20:03 elfixit1 joined #salt
20:04 g3cko joined #salt
20:05 dwiden I'm running a salt master/minion locally on my macbook pro (10.7.5).  They can successfully communicate.  I'm trying to learn the basics of salt states.  When I run "salt '*' salt.highstate", my minion esponds with "'salt.highstate' is not available."
20:05 cro dwiden: Try 'state.highstate'
20:05 openxmark joined #salt
20:06 dwiden cro: thank you, user error as usual -_-
20:06 Chrisje joined #salt
20:06 cro dwiden: no worries
20:07 cro dwiden: Mac support isn't perfect.  It's better than Win support but not as good as Linux
20:07 cro That being said, I put the 2014.1 RC on all my Macs at home this weekend
20:08 dwiden cro: I'm just learning the basics on my Mac, but will eventually put the master on linux
20:09 kalloc joined #salt
20:10 retran what would the world be like if everyone used mac for servers
20:12 Gareth A lot of spinning beach balls. :)
20:12 cro And permissions issues.  The change to SMB protocol by default in Mavericks has made me want to jump off a building.
20:12 andrej Yesterday someone told me that I should be using ansible rather than salt; their primary argument after I refuted some of the things they said was that a salt-minion is capable to get to other minions pillar data when exploited. Is this correct?
20:13 cro andrej: No
20:15 andrej heh
20:15 andrej thanks
20:15 toastedpenguin is it possible to use custom metadata with salt, so if say I wanted salt to manage a whitelist of IPs for something like iptables etc. I could assing IPs to a customer name so I wouldn't have to remember the IPs just the customer?
20:16 toastedpenguin and that is assign...assing is something entirely different
20:16 andrej lol
20:16 zach dwiden: just run vbox and put a few centos vms in it ;-)
20:18 dwiden zach: I was thinking about using a few VMs, but my computer can barely handle one
20:18 dwiden I'll just wait until I get a real linux box
20:19 zach oh that's no fun
20:21 retran oh so you have a Pentium 3 Coppermine
20:21 retran once you upgrade from that you'll do fine with VMs
20:22 zach ;-)
20:24 Nazzy retran, I have a box somewhere at home that has dual P3 Coppermine 700Mhz :p
20:24 retran Dual that's srs
20:25 UtahDave joined #salt
20:25 Nazzy it was actually a pretty reasonable server spec when I got my hands on it ... like a decade and a half ago
20:25 retran in the 1990s i remember a customer was really disapointed he didn't notice performance increase using IE when he got a Dual p2 system
20:26 retran had to explain about threads... etc
20:26 zach Those P3's will run for ever
20:26 retran but he never understood
20:27 elfixit joined #salt
20:29 forrest Did you explain it to him as though the processor was a highway, and IE decided to close all the lanes except one?
20:29 zach haha ^^^
20:30 zach I had a P3....thing ran for 5 years straight, would still be running had I not moved states
20:30 retran that would been great to have that analogy handy :p
20:30 zach Was the most amazing CPU I've ever had
20:31 zach One of the early days gateway/dell computers - not sure which been too long
20:31 zach I also have an HDD that has been working since 1986
20:32 zach One of the first years that PATA came out
20:32 zach Could have been 1989, somewhere around there. Makes horrible noises now though
20:33 zzzirk joined #salt
20:34 rojem joined #salt
20:35 zach with salt-cloud is there a way to pull the EC2 DNS that is set?
20:36 cewood joined #salt
20:36 retran if only p3s had hypervising
20:36 bhosmer joined #salt
20:36 cewood joined #salt
20:36 Ryan_Lane1 joined #salt
20:37 martoss joined #salt
20:38 Vye toastedpenguin: have you looked at Pillar?
20:39 pydanny joined #salt
20:39 toastedpenguin Vye: some, I take it that Pillar has this capability?
20:40 Nazzy zach, I have some 5.25" scsi disks that still run fine and some 200Mhz boxes to put them in
20:41 Nazzy I have debated racking one of them in the DC for giggles
20:42 Nazzy "Hey support, my test server is running a bit slow, can one of you pop down and reboot it?" >:)
20:42 jrdx joined #salt
20:42 Vye toastedpenguin: I'm not familiar with the iptables state but you should be able to have it iterate through values (IPs) stored in Pillar.
20:43 toastedpenguin Vye: I was using iptables as an example but I am looking to have metadata with assigned values so I can
20:43 Vye toastedpenguin: How you organize the IPs in Pillar is completely up to you (nesting within customers).
20:43 toastedpenguin add/delete
20:43 toastedpenguin ah ok so it would do what i need
20:44 toastedpenguin I want to whitelist IPs based on a customer and want to store the value and the customer in one place so if I have to delete a customer and their IPs its one place and done
20:44 driks joined #salt
20:45 Nazzy the other one I want to try, purely to troll the support department, is to get a big 24 bay rack mount chassis and shove an arduino inside, wire the power button to the reset pins, then stick rgb LEDs behind the drive activity lights... maybe have a speaker as well to make beeping noises
20:46 Vye toastedpenguin: give these a read: http://docs.saltstack.com/topics/pillar/index.html http://docs.saltstack.com/topics/tutorials/pillar.html
20:47 fllr Hey guys. Is there anyway I can download a directory from s3 using salt and file.recurse?
20:47 Nazzy then I send support to investigate why the server is down ... the lights flash randomly so they "reboot" it and after making random noises it does a colour fade with the LEDs before changing to doing trollish things like larson scanning
20:47 gadams999 joined #salt
20:47 toastedpenguin Vye: thanks I will, I am supporting a combination of WIndows and Linux in a AWS VPC...well building it now and I am trying to collect ancillary services I can put in to make admin of this environment easier
20:48 retran and speaker makes defening alarm  noises and their eardrums bleed
20:48 Nazzy they go round the back to the cables and find that not only does it only have a power cable, but it doesn't even have a visible motherboard
20:48 retran and a boxing glove springs out and punches them in the balls
20:48 jacksontj joined #salt
20:48 Nazzy finally they open up the box and find this little arduino and a whole mess of wires and LEDs
20:48 retran a million mice crawl out of a cage
20:49 forrest fllr, if the salt master was in s3 sure, but I don't know if that would be the best way to do it.
20:49 Nazzy and maybe a piece of paper saying "trollolololol"
20:49 fllr forrest: How would you recommend doing it?
20:49 retran then a midget comes out of the rack too
20:49 retran naked and urinated
20:49 retran asking what counrty he's in
20:49 Ryan_Lane1 joined #salt
20:49 forrest fllr, depends on the number of files, to reduce bandwidht it might be good to tar them up first, then just pull that file down
20:50 fllr forrest: I thought about that, but that would complicate keeping everything up to date...
20:50 Nazzy oooooh, those servers have intrusion detect switches too, so it can have a backup battery ("I unplugged it but the lights are still flashing!") and when they take the lid off it starts playing a rick roll at them ^_^
20:50 retran include man traps
20:51 tru_tru joined #salt
20:51 retran so their co-workers die randomly
20:51 Nazzy I want to troll my co-workers, not get fired ¬_¬
20:52 retran yeah killing is only grounds for being fired
20:52 retran true
20:53 jimallman joined #salt
20:55 Gifflen joined #salt
20:59 fllr Agh. Why is the online documentation set to a development branch, instead of some release tag? that's confusing...
21:00 forrest The docs always build from the develop branch, usually stuff has a versionadd regarding what release it is in. You can check out the readthedocs version and it supports tags
21:00 Ryan_Lane2 joined #salt
21:03 pdayton joined #salt
21:08 oz_akan_ joined #salt
21:10 hunter joined #salt
21:11 _jslatts joined #salt
21:13 clearclaw On the off-change: Does anyone know of a command line utility that will retrieve versioned objects from S3?  s3cmd, s3cp, s4cp, etc all only handled unversioned objects.
21:15 ajw0100 joined #salt
21:17 mgw clearclaw: Have you looked at the official AWS cli tools?
21:18 clearclaw Yeah, they don't handle versions either.
21:18 clearclaw I can see them, get their
21:18 clearclaw I can see them, get their version IDs etc, but not retrieve by version.
21:19 yomilk joined #salt
21:19 * clearclaw mutter mutters blah blah yet another for of someone's code and a pull request that's never accepted...mutter mutter.
21:19 clearclaw s/for/fork/
21:20 ferai joined #salt
21:21 mgw clearclaw: does adding ?versionId=<versionid> to the object not get the correct version?
21:26 juso joined #salt
21:26 fejjerai joined #salt
21:27 hunter I should know this- how do I mark a state so that if it fails, salt just bombs?
21:28 cedwards hunter: failhard option i believe
21:28 joehillen joined #salt
21:29 mgw hunter: http://docs.saltstack.com/ref/states/failhard.html
21:30 KyleG joined #salt
21:30 KyleG joined #salt
21:31 pydanny joined #salt
21:31 fejjerai joined #salt
21:32 kedo39 joined #salt
21:39 cewood joined #salt
21:40 druonysus joined #salt
21:40 druonysus joined #salt
21:45 cstelma joined #salt
21:46 philipsd6 What's the easiest/best way to install salt from development branch when you don't have internet access on the target machine? Can I clone the repo on a machine that has access, and then point the bootstrap script to that?
21:47 Nazzy philipsd6, probably
21:47 philipsd6 duh, I should really read things before I post
21:47 philipsd6 -g  Salt repository URL. (default: git://github.com/saltstack/salt.git)
21:48 Nazzy :)
21:57 ferai joined #salt
22:00 fejjerai joined #salt
22:03 fridiculous joined #salt
22:04 zach getting this when building out a new vm with salt-cloud: https://p.6core.net/p/tLiDJVz9ocjgoDBBsh7tfJPb
22:04 hunter anyone know how well (if at all) salt-cloud works with openstack essex?
22:04 mbarrien joined #salt
22:05 fridiculous what's the best way to define a path that's minion specific, as opposed to environment specific
22:05 codekoala anyone know what the compatibility rules are for Python? like should new code be compatible with 2.5+? 2.6+?
22:06 Nazzy fridiculous, pillar?
22:06 fridiculous e.g. with grains, i can get the minion id, and then with that minion, would i call pillar for a path deifinition??
22:06 pdayton joined #salt
22:07 yomilk joined #salt
22:07 codekoala fridiculous: I'm new to salt, but that seems like it would work well enough
22:07 codekoala not sure about the "best" way...
22:07 codekoala hehe
22:08 fridiculous pillar may work, but it's a bit bulky the way i see it
22:08 fridiculous for example imagine each minion has a custom path
22:08 bdf anyone have a good example of /proc manipulation in a state?
22:08 codekoala fridiculous: are you talking about an actual path? or the environment variable?
22:09 fridiculous it could be either
22:09 fridiculous not the environmental variable
22:09 fridiculous maybe i need custom grains... hmm
22:09 codekoala that could work too
22:10 codekoala ti would probably just be a bit heavier on the maintenance side, compared to the pillar
22:10 DaveQB joined #salt
22:15 hunter is it possible to have a state that WILL NOT run in a highstate? I'm thinking about states I only want run "manually" because they are potentially dangerous and want to prevent them from accidentally being included in highstate.
22:16 Nazzy hunter, don't put the state top.sls and don't call it from any state that is in highstate, problem solved :)_
22:17 btorch I got some issues with the file.replace .. I tried just searching and gave the search_only=True  but it zerod out the whole file
22:18 codekoala ouch
22:18 Ahlee so, just had a case, 4 minions targeted with grains.setval same value
22:18 btorch this is on 0.17.4
22:18 Ahlee three minions gook the value, fourth required a restart
22:18 Ahlee 0.17.2, thoughts?
22:18 ferai joined #salt
22:18 btorch codekoala: yeah I'm glad I was testing it on a copy of the file
22:19 hunter Nazzy: not helpful
22:19 codekoala btorch: oh whew, glad it was testing
22:19 hunter Nazzy: Sadly I won't be the only one with write access to the repo.
22:20 btorch if I read correctly I thought this would have been ok "sudo salt-call file.replace  /tmp/test.log 'Reset adapter' '' search_only=True "
22:20 Ahlee hunter: ultimately salt can't stop people problems
22:20 Nazzy hunter, if you can't stop someone adding a state to highstate, how would you prevent them from removing whatever prevents it running as well?
22:21 hunter Ahlee: thanks
22:21 Ahlee make sure you have an external job cache up to capture it
22:21 Ahlee and, fire whomever fucks up
22:21 fridiculous i find myself figuring out my problems as i materialize them into text for this chat
22:21 fridiculous yay!
22:21 Ahlee i have numerous states not targeted in high state we only call via state.sls
22:21 hunter Nazzy: its a quick catch - like the --force command on lots of tools. It doesn't STOP you - but it helps prevent the quick "oops"
22:22 Ahlee fridiculous: hooray! That means you're understanding :)
22:22 Ahlee hunter: well, people need to know those states exist, so is documenting the top letting them know the dangers enough?
22:22 hunter Ahlee: Thats what I'm going for, I was just hoping for a way to say "if highstate: fail "
22:22 Nazzy hunter, right, so you don't put it in highstate so it has to be called explicitly
22:22 Ahlee or are your corworkers actually gonna run it to find out what it does?
22:22 hunter Ahlee: absolutely - I've never known a developer to not read sysadmin docs.
22:23 Nazzy Ahlee, coworkers like that should be kicked off the systems and have official complaints filed against them ¬_¬
22:23 hunter Nazzy: that takes on average about 2 years around here - when management can be bothered at all
22:24 jp_cognet joined #salt
22:24 hunter I'll just go with "salt doesn't (yet) have that feature"
22:24 Ahlee yeah, you're up shit creek man
22:24 gadams999 left #salt
22:24 Ahlee like I said, salt can't solve people problems
22:24 hunter Ahlee: I'm not in complete agreement this is solely a people problem. (--force and all)
22:25 basepi apergos: ping
22:25 jp_cognet I can't seem to find this anywhere in the docs, but lets say i only wanted 1 piece of information from loadavg, so I do "salt '*' status.loadavg" but I just want it to show me the 1min load. Is there a way to slice the output or something with the salt command? or do I just have to use my shell?
22:26 Nazzy lol, my solution to that was "deal with this now or I file it with the security committee as an ISO violation"
22:26 bhosmer joined #salt
22:30 Nazzy important life lesson: don't fight the system, use it instead
22:32 [diecast] joined #salt
22:33 cedwards can anyone tell me if pillar data is available in salt-ssh?
22:34 Mua joined #salt
22:42 philipsd6 cedwards: most definitely it is
22:43 philipsd6 hunter: I'm in agreement with you on this. It's not enough to not reference it in the top file, if it can be included in an other state.
22:44 philipsd6 I'm not sure I have a valid use case for this, but I was looking for the ability to disable a state from being called at all. I.e. I want to edit mybuggystate.sls and add disabled: True
22:45 philipsd6 and it won't be called no matter where it is referenced.
22:45 philipsd6 but this isn't exactly your use case, since you want to run it manually/directly.
22:46 macorsen left #salt
22:49 jacksontj joined #salt
23:00 bhosmer joined #salt
23:02 Ryan_Lane1 joined #salt
23:04 bretep Does anyone here use the gitfs?
23:04 kermit joined #salt
23:05 tedski yup
23:08 bretep How long is the delay for gitfs to update? Seems pretty random to me. Can I add a hook into the repo to tell the salt-master that the repo has been updated?
23:11 hunter philipsd6: I'm tempted to file a very low priority feature request for this
23:11 tedski the gaps i notice are with github enterprise updating, not with salt knowing
23:11 tedski bretep: on average, about 5-10 seconds for us
23:12 bretep I'm seeing 30-60 seconds to notice the repo has changed. I've done a tcpdump on the salt-master and it seems like it only checks every so offten not every time of running highstate
23:12 tedski interesting
23:12 tedski it does cache gitfs
23:13 Gordonz joined #salt
23:13 dave_den http://docs.saltstack.com/ref/runners/all/salt.runners.fileserver.html
23:13 Gordonz joined #salt
23:15 zain when using salt-cloud, does anyone know how I can set AssociatePublicIpAddress to be True? it's for a VPC instance in my public subnet. without it, the instance doesn't have a public IP and can't connect to the internet.
23:25 zain found my answer here https://github.com/saltstack/salt-cloud/pull/889
23:29 justBob_ joined #salt
23:30 rgbkrk_ joined #salt
23:32 scoates I'm trying to use the py renderer in pillar, and I'm getting __grains__ is undefined. any ideas? http://docs.saltstack.com/ref/renderers/all/salt.renderers.py.html#module-salt.renderers.py
23:32 Thiggy joined #salt
23:37 forrest scoates, what does your file look like?
23:38 scoates forrest: http://paste.roguecoders.com/p/b7f0a7a16f93529813d1b10cf62bb7df.txt
23:38 scoates (I also tried with __grains__)
23:38 scoates since they're both passed in the code
23:39 Ryan_Lane1 joined #salt
23:39 forrest did you try a different value for grains, such as __grains__['os'] like in the example?
23:39 scoates I'll try.
23:39 forrest ok
23:39 scoates yeah. same error for 'os': http://paste.roguecoders.com/p/0bb9079cdc1d0b31961820435b2c87eb.txt
23:40 scoates I also did have a `def run(): … ` in there, and it failed, too. This is a reduced reproduce case.
23:41 scoates this is on 0.17.5 (and 0.17.2) … this renderer *has* been around for a while, right?
23:41 lzhang joined #salt
23:42 forrest while there are changes between the py.py renderer files in 0.17.2 versus 0.17.5, it does exist
23:42 * scoates nods
23:42 forrest I'm not quite sure, can you run it with debug to see if there are more details?
23:43 alunduil joined #salt
23:43 warmwaffles joined #salt
23:43 warmwaffles anyone able to help me with Archlinux on Vagrant? https://github.com/saltstack/salt-bootstrap/issues/295
23:43 scoates that's with debug on the master
23:43 scoates let me try on the minion
23:43 forrest scoates, cool
23:44 bemehow joined #salt
23:44 bemehow hello all, quick question about state ordering, If for the particular Id I have more than one state definied.. how can I assure the dependency between the ids
23:45 forrest bemehow, you mean two files named the same thing?
23:45 bemehow {{mddev}}:
23:45 bemehow raid:
23:45 bemehow - present
23:45 bemehow - opts:
23:45 bemehow - level=0
23:45 bemehow - chunk=256
23:45 bemehow {%- for edisk in edisks %}
23:45 bemehow - {{edisk}}
23:45 forrest don't paste that here, you will get removed
23:45 bemehow {%- endfor %}
23:45 bemehow - raid-devices={{edisk_count}}
23:45 forrest just use pastebin or something
23:45 bemehow lvm:
23:45 bemehow - pv_present
23:45 bemehow - require:
23:45 bemehow - raid : {{mddev}}
23:45 bemehow blockdev:
23:45 scoates forrest: nothing obvious in the minion output. no errors. just regular stuff. pillar renders on the master side, too, AFAIK.
23:45 bemehow - tuned
23:45 bemehow - read-ahead : 65535
23:45 bemehow - filesystem-read-ahead : 65535
23:45 scoates I need to go, unfortunately.
23:45 bemehow - require:
23:45 bemehow - lvm: {{mddev}}
23:45 bemehow sorry for the spam
23:45 bemehow just clickd the wrong one
23:45 forrest scoates, weird let me know if you figure it out
23:46 fllr joined #salt
23:46 scoates I mean not THAT unfortunately. I need to go… drink beer.
23:46 forrest heh
23:46 scoates thanks for the sanity check (-:
23:46 * scoates out
23:46 bemehow CqS05igx
23:47 scoates ^ definitely not a password
23:47 bemehow http://pastebin.com/CqS05igx
23:47 scoates ah. heh
23:47 brucewang Your guys nickname align very well on my screen :P
23:47 warmwaffles ahhah
23:48 forrest bemehow, I guess I am confused, where is the conflict here?
23:48 bemehow i've written extfs and block states
23:48 bemehow the order is not preserved. The lvm.lv_create does not wait for the vg to be created and fails
23:50 forrest is there an issue going on here with the indent level?
23:51 lzhang left #salt
23:52 bemehow I've found one issue with the syntax error extent vs extents (failing lvcreate state)
23:52 forrest so why is - devices indented, when - vg_present (as an example) has no colon?
23:52 bemehow ok i think I got it.. require does not require_success
23:53 forrest ok
23:53 rojem joined #salt
23:53 bemehow sorry for the spam, mystery solved :)
23:54 forrest no worries
23:54 bemehow if anyone is interested in extfs states I will try to contribute them after some cleanup
23:54 Ryan_Lane1 joined #salt
23:54 rojem joined #salt
23:54 forrest yea that would be cool to have in the formulas repo
23:56 bemehow am new to the formulas concept. just hacked on the source of _modules/ _/states but definitely will do

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