Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-07-02

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

All times shown according to UTC.

Time Nick Message
00:00 jeddi tho' .. that link didn't take me where i thought .. hmm .. sorry.
00:00 jeddi but yeah - in the 'quick search' box i often type the state name i'm interested in and the word 'module' and then feel a bit more enlightened.
00:02 ruimarinho yeah.. I do something similar too. Good way to quickly scan the article.
00:09 adotbrown joined #salt
00:10 ruimarinho Created deb package {:path=>"nodejs_0.10.12_i386.deb"} :)
00:11 chrisgilmerproj left #salt
00:11 berto- joined #salt
00:12 jeddi ruimarinho: that's about the right timeframe :)
00:12 Gifflen joined #salt
00:13 ruimarinho yay, I can migrate my current state to 5 lines :)
00:15 ruimarinho any recommendation on directory structure when it includes binary files?
00:16 ruimarinho <state>/dist ?
00:16 jeddi soz - not from me.   you mean package or .. oh, salt.
00:17 jeddi some would suggest you deploy by pulling from git or better yet set up your own repo that you then manage.
00:17 jeddi i'd probably advise the latter, though it's some non-trivial work to just get a file into play.   far more robust and versatile though.
00:17 ruimarinho it sounds good based on my current setup.. I definitely wouldn't like to ship it with the current salt repository I'm working on
00:23 madduck joined #salt
00:23 madduck joined #salt
00:24 nrub joined #salt
00:24 ruimarinho Comment:   The following packages were installed/updated: nodejs. Changes:   nodejs: { new : 0.10.12
00:25 ruimarinho thanks jeddi and TheRealBill_
00:26 jeddi ruimarinho: np - glad my ramblings could help :)
00:26 ahammond joined #salt
00:32 carmony joined #salt
00:36 ckeene joined #salt
00:47 pweo joined #salt
00:48 Gwayne joined #salt
00:48 Gwayne joined #salt
00:53 kenbolton joined #salt
00:54 nrub_ joined #salt
00:58 dthom91 joined #salt
01:10 Nexpro1 joined #salt
01:15 baniir joined #salt
01:23 Awnic joined #salt
01:24 Awnic Hello everyone.
01:24 tpe11etier joined #salt
01:35 scooby2 joined #salt
01:41 dthom91 joined #salt
01:45 liuyq joined #salt
01:46 N-Mi joined #salt
01:50 renoirb joined #salt
01:53 kmwhite joined #salt
01:58 mgw joined #salt
01:58 oz_akan_ joined #salt
02:00 jaequery joined #salt
02:01 KyleG joined #salt
02:03 shiznit joined #salt
02:16 mchype joined #salt
02:24 dthom91 joined #salt
02:25 alazylearner joined #salt
02:27 fxhp is it possisbly to prevent ssh output from salt-cloud?
02:27 fxhp nm
02:27 fxhp display_ssh_output: False
02:28 logix812 joined #salt
02:34 berto- joined #salt
02:36 alazylearner joined #salt
02:36 StDiluted joined #salt
02:39 kleinishere joined #salt
02:43 jessep joined #salt
02:47 dthom91 joined #salt
02:58 mgw joined #salt
03:00 lemao joined #salt
03:02 milind joined #salt
03:03 jdenning joined #salt
03:13 bluemoon joined #salt
03:15 Psi-Jack I'm trying to use a list defined in a pillar to install a list of packages as a basis for every salt-minion. But, I can't grasp the logic of how to use a list in the actual salt state from the pillar data.
03:16 Corey Psi-Jack: We meet again.
03:16 Psi-Jack I have packages.sls with packages: names: - package1 - package2 - package3, etc.. Then the salt state: packages: pkg.installe: - names: .... And yeah, I'm stuck there.
03:16 Corey Psi-Jack: pillar acts as a key:value store.
03:16 Psi-Jack Corey: heh, yeah. I started looking to Salt again and it's looking awesome now. :)
03:17 Corey Psi-Jack: The state you're missing is {{ pillar['key'] }}
03:17 Corey So whatever that key maps to for the given minion will be installed.
03:18 Psi-Jack Rendering SLS base-install failed, render error: while scanning a simple key
03:18 Psi-Jack in "<unicode string>", line 4, column 7:
03:18 Psi-Jack ['htop', 'atop', 'iftop', 'iotop ...
03:18 Psi-Jack ^
03:19 johntron joined #salt
03:19 Psi-Jack That's from pillar['base-install']['names'] being used as pkg.installed: - names: {{ .mentioned-at-beginning. }}
03:19 Psi-Jack Ahhhhhh
03:19 Psi-Jack It needed to be on the same line, not the next line, indented.. It seems. :)
03:20 renoirb Hey people, I am wondering about Salt's gifts feature
03:20 renoirb Is it possible to use it on a masterless minion, for let's say, a vagrant box ?
03:20 Furao joined #salt
03:20 renoirb My attempt, it doesn't work.
03:21 renoirb AS if it only reads local and not git
03:22 Psi-Jack heh, drats!
03:22 Psi-Jack I can't seen to use a pillar from within a pillar? ;)
03:23 Furao Psi-Jack: yes but it's tricky
03:23 Furao many modules aren't available in salt
03:23 Psi-Jack Hmm, how tricky? I'm mostly just trying to get the name-mapped package from ... lemme pastebin this pillar state.
03:24 Psi-Jack http://pastebin.ca/2414176
03:24 Furao renoirb: gitfs marche juste dans le master
03:24 Psi-Jack That's what I'm trying to do, get the vim package name mapped for the appropriate distro.
03:25 Furao Psi-Jack: don't mix string and list for nslcd
03:26 Furao Psi-Jack: line 14:15 don't make sense
03:26 Psi-Jack So, do them both as a list, for that one? Okay.
03:26 Furao ah it's a pillar
03:26 Psi-Jack Yes, this is a pillar. :)
03:26 Furao that won't work
03:26 Psi-Jack So, I'll have to do the if-elif approach again?
03:27 Furao in that case, I keep the list of packages in my "base" state
03:27 Furao to avoid pillars over pillars
03:29 Psi-Jack Eh, well. This is just kinda a quick hack to insure at least the basic packages I always install get installed, until I get templating VM support in Proxmox VE 3.0 when I can upgrade. :)
03:29 Furao do the if/elif trick in your base/init.sls
03:29 Furao and not in pillar
03:29 Psi-Jack I suppose that could work too.
03:30 Psi-Jack And just pull from pillar's where package name mapping may apply, like with vim.
03:30 Furao pillars are great, usually newbies ignore pillars for a long time, but it might be better to not put too much in pillars. but that's my IMHO after 1 year of salt.
03:30 Furao you can have in pillar an list of extra packages you want to get installed
03:32 Psi-Jack Eh, pillars are awesome for defining functionality, especially in a mixed environment of CentOS, Debian, and Ubuntu.
03:32 Furao salt work with centos? all packages that come with it are 8 years old
03:32 Furao like python 2.5
03:32 Psi-Jack Not quite as good as hiera, but close enough to still be acceptable in alternative ways.
03:32 Psi-Jack Furao: You're thinking CentOS 5.x
03:32 Psi-Jack :p
03:32 Furao so python 2.6
03:33 Furao which was released 6 years ago :P
03:33 Psi-Jack CentOS 6.4, in EPEL, has salt 0.15.3
03:33 Corey Psi-Jack: So does CentOS 5.x
03:33 Psi-Jack Yep. Again, in EPEL. :)
03:33 Corey Psi-Jack: As does every current supported version of Ubuntu. :-)
03:33 Psi-Jack yep, but not in EPEL for those. :)
03:34 backjlack joined #salt
03:39 Psi-Jack Hmmm. i see an error in one of my states.. trying to create /etc/zabbix/zabbix_agentd.conf.d directory before installing zabbix20-agent.
03:41 Psi-Jack heh, simple missing require . :)
03:44 backjlack joined #salt
03:52 dthom91 joined #salt
03:53 Psi-Jack Hmmm
03:53 Psi-Jack Trying to figure out how to use salt to set selinux to permissive.
03:54 zonk1024 joined #salt
03:57 jaequery joined #salt
03:59 pweo joined #salt
03:59 cxz how do i get salt to replace variables with pillar data in a managed file (file.managed), without using 'source:'
03:59 cxz ?
04:00 Furao cxz: not sure I understand
04:00 Furao cxz: examples?
04:00 cxz so say you have a file that has pillar data referenced in it: {{ pillar['mydata']['data'] }}
04:01 cxz now you want to git clone using git.latest and a file in that repository is like the above
04:01 cxz ^
04:01 cxz so you want salt to replace that data.
04:01 Furao first git clone
04:01 Furao then use source: file:///path/to/checkouted/file
04:01 isomorphic joined #salt
04:02 Furao and render jinja using pillar data
04:03 Psi-Jack Ugh.
04:03 cxz ah ok, cool
04:03 Psi-Jack There's got to be a way to get salt to set the SELinux mode to permissive. :
04:04 Psi-Jack I mean, it's got the setenforce function..
04:04 cxz Furao: will my name also be /path/to/checkouted/file
04:04 cxz ?
04:04 cxz /path/to/checkouted/file: file.managed: source: file:///path/to/checkouted/file
04:04 cxz like that?
04:04 Furao no, the name is the output
04:04 Furao you can put whatever you want
04:04 cxz ok
04:05 shiznit joined #salt
04:07 StDiluted joined #salt
04:13 cxz so it can't be anything i want
04:13 cxz it needs to have a proper filepath
04:14 davisamo joined #salt
04:14 jbunting joined #salt
04:15 Furao cxx paste your state
04:19 StDiluted evening
04:20 linuxdingding3 joined #salt
04:22 JordanRinke logout
04:24 linuxdingding3 left #salt
04:24 jdenning joined #salt
04:26 cxz Furao: this is such a stupid solution to my problem anyway, i'm doing it wrong
04:32 bluemoon joined #salt
04:33 linuxdingding3 joined #salt
04:36 rglauser joined #salt
04:40 Gifflen joined #salt
04:55 bluemoon joined #salt
04:59 dthom91 joined #salt
05:05 shiznit joined #salt
05:10 alazylearner joined #salt
05:20 linjan joined #salt
05:21 oz_akan_ joined #salt
05:31 joehh joined #salt
05:31 abraithwaite joined #salt
05:32 abraithwaite Question: how do you set the initial root password for mysql with salt?
05:35 Furao mysql isn't a database http://www.youtube.com/watch?v=1PoFIohBSM4
05:36 abraithwaite well it's what we use :-(
05:36 abraithwaite actually mariadb
05:36 abraithwaite but still
05:40 Furao these are just data storage that might work sometimes, if you don't care about data integrity
05:42 jefimenko joined #salt
05:43 jaequery joined #salt
05:44 knightsamar joined #salt
05:45 dthom91 joined #salt
05:54 wahahajun joined #salt
05:56 wahahajun hi, now when i run the sls file i get such errors:  UnicodeDecodeError: 'utf8' codec can't decode byte 0x88 in position 10: invalid start byte
05:57 EugeneKay Not sure if troll or just stupid
05:58 wahahajun sls funcion is use file:managed to  transfer the file to windows file
05:59 EugeneKay abraithwaite - the short answer is that it is hard to definitively set a password for the "root" mysql user
05:59 Furao yes, everything is hard with mysql
06:00 jkleckner joined #salt
06:00 wahahajun i don't use anything about mysql
06:00 EugeneKay The tl;dr is you need to invoke `mysqld_safe --skip-grant-tables`, then log in to the mysqld and execute a GRANT statement.
06:01 EugeneKay The good news is that most packaging of mysql makes this a lot easier; if not more secure.
06:01 EugeneKay The bad news is that no RDBMS which guarantees sanity is any easier, despite the ramblings of $USER on IRC ;-)
06:01 Newt[cz] joined #salt
06:02 EugeneKay And on the gripping hand, doing any of this in an idempotent manner while maintaining uptime is just plain impossible. So stop worrying!
06:03 Furao EugeneKay: check that http://www.youtube.com/watch?v=1PoFIohBSM4
06:03 Furao to see how mysql is a joke
06:03 EugeneKay Fuck off, you stupid asshole.
06:03 waverider joined #salt
06:03 EugeneKay Nobody gives two shits about your youlube links
06:03 Furao so i'm an asshole because there is clear good reasons to stay away from pseuo-database?
06:03 EugeneKay abraithwaite - My advice is to set the root/admin mysql user's password once in your "vanilla" image and then set/modify users as needed with states.
06:04 EugeneKay YOu're an asshole because you don't know how to go away
06:04 EugeneKay And a moron because you don't understand the implications of doing such
06:04 Furao yay no argument, just stays sad with your shitty disk storage
06:04 EugeneKay Cool story br0
06:04 EugeneKay Go jerk yourself off over in-memory cache masturbatio
06:05 Furao I never talk of any other database
06:05 EugeneKay Good; Go talk to a goat abotu ti
06:06 shiznit joined #salt
06:06 twinshadow joined #salt
06:15 dthom91 joined #salt
06:18 adotbrown joined #salt
06:24 kleinishere joined #salt
06:39 zonk1024 joined #salt
06:39 zonk1024 left #salt
06:40 liuyq joined #salt
06:40 jaequery joined #salt
06:44 cxz can we change the topic of the channel
06:44 cxz to reflect 0.16.0
06:44 cxz and how do i get operator privileges
06:47 karakedi joined #salt
06:55 x_or joined #salt
06:56 jesusaurus UtahDave is the only one i ever see with ops
06:57 efixit joined #salt
06:58 vaxholm joined #salt
06:59 carlos joined #salt
07:06 shiznit joined #salt
07:16 rberger joined #salt
07:23 scott_w joined #salt
07:25 __gotcha joined #salt
07:25 balboah joined #salt
07:26 rglauser joined #salt
07:28 __gotcha_ joined #salt
07:28 __gotcha joined #salt
07:30 __gotcha joined #salt
07:30 p3rror joined #salt
07:33 felixhummel joined #salt
07:52 waverider left #salt
07:52 adotbrown joined #salt
08:00 jefimenko joined #salt
08:06 ioni joined #salt
08:07 shiznit joined #salt
08:19 vaxholm joined #salt
08:26 krak3n` joined #salt
08:30 kallek joined #salt
08:30 bluemoon joined #salt
08:38 backjlack joined #salt
08:40 Furao_ joined #salt
08:51 Jason-AVST joined #salt
09:00 zooz joined #salt
09:02 ProT-0-TypE joined #salt
09:08 shiznit joined #salt
09:17 adotbrown joined #salt
09:25 drdran joined #salt
09:33 bluemoon joined #salt
09:33 Mouzz joined #salt
09:34 azbarcea joined #salt
09:50 f4cl3y joined #salt
10:02 efixit joined #salt
10:07 lbotti joined #salt
10:08 lbotti hi all
10:08 lbotti just looking for some help related to iptables state
10:08 shiznit joined #salt
10:16 adotbrown joined #salt
10:17 henk left #salt
10:27 zloidemon ?
10:33 p3rror joined #salt
10:47 ruimarinho joined #salt
10:54 markm joined #salt
11:00 whiskybar joined #salt
11:09 shiznit joined #salt
11:11 faust joined #salt
11:12 Jarus I want to add a ssh_key to a user which is also created by salt. I defined following in my sls file: https://gist.github.com/jarus/fa84a75c882254673802 but salt didn't respect the require statement.
11:16 blee joined #salt
11:18 Furao Jarus: there is no user definition in pasted state
11:18 Furao and indentation is all wrong anyway
11:19 Furao line 5 must be - require: - user: fuuba
11:19 zooz joined #salt
11:19 Furao indented like line 4
11:20 Jarus Furao, the require as inline?
11:21 Furao require is an argument to ssh_auth
11:21 Furao it's not a state name
11:21 adotbrown joined #salt
11:21 Furao check github.com/bclermont/states for plenty of require usage
11:21 markm joined #salt
11:21 Jarus Furao, aarrr I'm so stupid sorry
11:23 isomorphic joined #salt
11:25 swa i'm currently running in production on 0.14.1 and created a test VM running 0.15.3 with the same states.. when I try state.highstate I get "No Top file or external nodes data matches found"
11:26 markm_ joined #salt
11:30 jrgifford_ joined #salt
11:32 techdragon joined #salt
11:52 logix812 joined #salt
11:54 lemao joined #salt
11:58 Teknix joined #salt
12:00 oz_akan_ joined #salt
12:04 kenbolton joined #salt
12:05 derelm joined #salt
12:09 shiznit joined #salt
12:10 Koma joined #salt
12:10 jbunting joined #salt
12:16 copelco joined #salt
12:20 aleszoulek joined #salt
12:39 jessep joined #salt
12:51 cron0 joined #salt
12:52 drdran joined #salt
12:55 baniir joined #salt
12:58 basti joined #salt
12:59 basti Hello, I have a question about the django module of salt
12:59 basti Can anyone guide me how to use it? I do not understand the documentation about it. And did not found any examples yet.
13:00 basti I want to do a simple collectstatic after the entire server is up and running
13:00 Newt[cz]1 joined #salt
13:03 oz_akan_ joined #salt
13:06 zooz joined #salt
13:06 anteaya joined #salt
13:06 scott_w joined #salt
13:07 oz_akan_ joined #salt
13:07 krak3n` joined #salt
13:08 shiznit joined #salt
13:08 krak3n` joined #salt
13:08 derelm joined #salt
13:10 johntron joined #salt
13:11 Jahkeup_ joined #salt
13:11 krak3n` joined #salt
13:15 __gotcha joined #salt
13:16 Kholloway joined #salt
13:21 jkleckner joined #salt
13:21 jkleckner1 joined #salt
13:22 jkleckner joined #salt
13:25 georgj05 joined #salt
13:25 juicer2 joined #salt
13:26 jkleckner1 joined #salt
13:27 jkleckner joined #salt
13:28 jkleckner joined #salt
13:29 mikedawson joined #salt
13:31 jkleckner joined #salt
13:31 toastedpenguin joined #salt
13:32 fredvd joined #salt
13:35 racooper joined #salt
13:37 __gotcha joined #salt
13:41 mgw joined #salt
13:48 MasterNayru joined #salt
14:00 bhosmer joined #salt
14:00 UtahDave joined #salt
14:02 bhosmer joined #salt
14:04 jalbretsen joined #salt
14:04 shiznit joined #salt
14:06 conan_the_destro joined #salt
14:06 Jason-AVST joined #salt
14:06 krissaxton joined #salt
14:09 swa oh god.. do not use accents in comments in your top.sls,.. I'm getting "No Top file or external nodes data matches found" on 0.15.3 while everything runs OK on 0.14.1
14:11 UtahDave swa: s0undt3ch just made a commit last night that should fix that.
14:11 Furao UtahDave: really?
14:11 ss_ joined #salt
14:11 Furao I broke my state when I used the π instead of n
14:12 krak3n` joined #salt
14:12 swa UtahDave: that was a regression ?
14:13 swa I rarely use accents in any config files, but i didn't expect it to break in a comment
14:13 Furao the power of ascii
14:13 rglauser joined #salt
14:19 ipmb joined #salt
14:19 carmony joined #salt
14:20 ipmb Hi, I just updated `salt-minion` from the PPA and running `salt-minion --version` gives me this:
14:20 ipmb salt-minion 2.0-218-gbfc8607
14:20 ipmb I was expecting something more like 0.15.x
14:21 Furao http://ppa.launchpad.net/saltstack/salt/ubuntu/pool/main/s/salt/
14:22 pweo joined #salt
14:22 kho joined #salt
14:23 ipmb yes, dpkg tells me 0.15.3-1precise is installed
14:24 ipmb why would --version be telling me it is 2.0-218-gbfc8607
14:25 ipmb hmm, looks like somebody pip installed it on this server at one point
14:28 cedwards UtahDave: have there been any dependency changes in the latest release?
14:28 hazzadous joined #salt
14:29 isomorphic joined #salt
14:29 fragamus_ joined #salt
14:34 jbunting joined #salt
14:38 shiznit joined #salt
14:46 Gifflen joined #salt
14:46 sashka_ua Furao: do you have some issues, when salt-master caches some grains or pillar values
14:46 sashka_ua Furao: suing salt-call I got output I expect
14:47 sashka_ua running "salt model pillar.data" I got pillar I expect
14:47 sashka_ua but then "salt model state.highstate" give me previous grains or pillar values
14:48 Gifflen joined #salt
14:55 brianhicks joined #salt
14:55 alazylearner joined #salt
14:57 timoguin joined #salt
14:57 timoguin http://lincolnloop.com/blog/2013/jun/14/introducing-salmon/
14:58 timoguin ^^ have you guys seen that? got posted to the saltstack subreddit recently
15:00 rglauser joined #salt
15:00 UtahDave Furao and swa, Oh, I guess we haven't merged it yet. Wanted to have Tom do some testing: https://github.com/saltstack/salt/pull/5873
15:01 UtahDave sashka_ua: have your run salt \* saltutil.refresh_pillar     ?
15:01 UtahDave timoguin: Yeah, that's pretty cool, huh?
15:01 bensix joined #salt
15:01 timoguin indeed.
15:01 jessep joined #salt
15:01 timoguin trying to find time to play with it.
15:02 teskew joined #salt
15:02 sashka_ua UtahDave: I checked everything. Basically its grains
15:03 sashka_ua UtahDave: I checked all caches, all files
15:03 UtahDave if you delete all your caches it should fix that.  Then I wonder if it will occur again.
15:03 sashka_ua it occurs all the time
15:04 sashka_ua but not that often
15:05 s0undt3ch ipmb: have you set any tags locally?
15:05 ipmb s0undt3ch: I ran `pip uninstall salt` and everything is good again :)
15:05 sashka_ua UtahDave: you know, refresh_pillar helped
15:06 sashka_ua UtahDave: but the issue was related to grains
15:06 s0undt3ch ipmb: nice
15:06 ipmb PPA wasn't running because it was installed via pip too
15:06 Jahkeup_ joined #salt
15:07 sashka_ua UtahDave: I did in /srv/pillar/top.sls : http://pastebin.com/7ztJKRiu
15:07 __gotcha joined #salt
15:07 davisamo joined #salt
15:08 sashka_ua and it gave me : http://pastebin.com/2FVx9HEf
15:08 hebz0rl joined #salt
15:08 brianhicks is it possible yet to write a state that depends on the changes in another state - for example to run a command IFF a certain file in git.latest is modified?
15:08 sashka_ua all the time despite of any changes and syncs
15:08 Sailor_ joined #salt
15:09 sashka_ua brianhicks: ordering, require, require_in, watch, watch_in
15:09 racooper Is there a place to submit enhancement requests for the EPEL packages?
15:09 brianhicks sashka_ua: right, last time I checked it didn't actually pass the output of the previous state to the current one - did that change?
15:09 sashka_ua brianhicks: more specificly you need: watch, watch_in
15:10 ange UtahDave: hey
15:10 sashka_ua brianhicks: output of what?
15:10 efixit joined #salt
15:10 UtahDave brianhicks: you'll want to use a  'require' or a 'watch'
15:11 brianhicks sashka_ua: the problem in my case was that the state didn't get the old and new refs of git. require and watch worked fine otherwise.
15:11 UtahDave sashka_ua: does /srv/salt/levels.level4.sls  exist?
15:11 sashka_ua UtahDave: previously not, but it still showed that error even when I created it
15:12 UtahDave racooper: Just open an issue on the github repo for your EPEL enhancement. herlo is our packager for that.
15:12 racooper it's actually an existing issue from a year ago, #1012 to add logrotate scripts to the -master and -minion packages
15:13 UtahDave racooper: cool. Do you have a pull req for it?
15:16 racooper looks like...thatch45 added the necessary files under the 0.11 tree and closed the ticket, and in 0.15 they've been removed/changed, but never added to the .spec file
15:18 LyndsySimon joined #salt
15:19 berto- joined #salt
15:24 ruimarinho is there a roadmap for a built-in augeas state?
15:24 faust joined #salt
15:24 johntron joined #salt
15:27 swa root@salt-master-015:/srv# salt-key -d salt*
15:27 swa No keys to delete.
15:27 swa root@salt-master-015:/srv# salt-key -d salt-*
15:27 swa No keys to delete.
15:27 swa root@salt-master-015:/srv# salt-key -d salt-c*
15:27 swa No keys to delete.
15:27 swa root@salt-master-015:/srv# salt-key -d salt-cl*
15:27 swa The following keys are going to be deleted:
15:27 swa Accepted Keys:
15:27 swa salt-client01
15:27 swa Proceed? [N/y] ^C
15:27 jeddi swa .. pastebin is a good thing, huh?
15:27 jeddi are you assuming that your shell doesn't glob these wildcards?
15:28 teskew joined #salt
15:28 Sacro salt-cl\*
15:29 swa yeah, sorry about that. Why does it work at some point with salt-cl* ?
15:30 UtahDave swa: try this:   salt-key -d 'salt-*'
15:30 jschadlick joined #salt
15:31 yidhra joined #salt
15:34 whit joined #salt
15:37 diegows joined #salt
15:39 Furao sashka_ua: I run salt-call saltutil.sync_all every time my pillar git repo get updated
15:40 chrisgilmerproj joined #salt
15:41 m_georgez joined #salt
15:43 oz_akan_ joined #salt
15:43 m_georgez left #salt
15:43 sashka_ua Furao: that didn't help as well
15:44 Gifflen joined #salt
15:50 lvicks joined #salt
15:51 jacksontj joined #salt
15:52 spmurray joined #salt
16:07 dahunter3 joined #salt
16:07 alazylearner joined #salt
16:08 milind joined #salt
16:09 dahunter3 The name reported with I run a salt '*' test.ping. Is that from what the minion_server learned about the minion via reverse DNS at the time I was updating the key? If I've since updated reverse DNS, how do I get the master to update how it refers to them?
16:10 Jahkeup__ joined #salt
16:11 swa dahunter3: as long as I know (not much) it's based on the hostname of the machine, not DNS lookup
16:12 KyleG joined #salt
16:12 dahunter3 swa: I don't think so. Because the machines in question had their hostname set correctly but the master named them their reverse DNS entry.
16:13 UtahDave dahunter3: When the minion starts up for the first time if you haven't set its id in the minion config, then it will use its dns hostname. I think it uses pythons   get_socket()  or something like that. I can go find the code if you're interested
16:13 UtahDave dahunter3: you can set the minion's ID in its config file, then you'll have to re-accept the minion's key on the master
16:13 cron0 joined #salt
16:13 gmoro joined #salt
16:14 Bastion2202 joined #salt
16:14 dahunter3 UtahDave: Gotcha, thanks.
16:14 f4cl3y joined #salt
16:14 shiznit joined #salt
16:14 Bastion2202 hey guysm
16:15 UtahDave dahunter3: you're welcome.
16:15 UtahDave hey, Bastion2202
16:15 Bastion2202 salt does not seem to work with iptables     v1.4.16.3. iptables.get_rules output nothing and on the minion log : salt-minion-python2.7: error: unrecognized arguments: ! SYN
16:16 UtahDave Bastion2202: I think iptables support is still new and in flux
16:17 Bastion2202 UtahDave: ok tx for the info
16:18 UtahDave Bastion2202: you're welcome! You might find info on its progress on the mailing list. I think there's an open issue on github as well.
16:19 kermit joined #salt
16:20 ruimarinho UtahDave, is there any way I can take advantage of the augeas module that salt ships with to manage states?
16:22 jaequery joined #salt
16:22 UtahDave ruimarinho: I don't think we have an augeas state yet.  So you're wanting to manage your sls files with augeas?
16:23 ruimarinho some of them yes.. I really like the power that templating provides, but on huge configuration files where only one or two settings required tweaking, it is unnecessary imho
16:25 KennethWilke joined #salt
16:25 UtahDave ruimarinho: So do you find using augeas easier for that situation?
16:26 oz_akan_ joined #salt
16:26 ruimarinho yes.. not as simple as templates, since you need to figure out the right key syntax, but with a couple of lines I can easily list 3 or 4 managed changes for a hundred-line config file
16:29 dthom91 joined #salt
16:29 UtahDave ruimarinho: I think we just need to write the augeas state to complement the augueas execution module.
16:30 ruimarinho that would be fantastic
16:31 carmony joined #salt
16:32 jpadilla joined #salt
16:35 techdragon joined #salt
16:37 jefimenko joined #salt
16:37 jaequery joined #salt
16:38 Gifflen joined #salt
16:40 sashka left #salt
16:45 ipmb joined #salt
16:46 jbunting joined #salt
16:49 isomorphic joined #salt
16:50 bluemoon joined #salt
16:53 alazylearner joined #salt
16:54 Timitos joined #salt
16:55 Timitos left #salt
16:57 jacksontj joined #salt
16:58 krak3n` joined #salt
16:59 UtahDave cedwards: I don't know of any new dependencies.  Sorry it took so long to respond.  :)
17:00 shiznit joined #salt
17:01 chrism Is there a way to show all minions that didn't respond to a specific cmd, like a test.ping?
17:01 UtahDave chrism: add    -v   to your command
17:01 chrism ooh fancy =)
17:01 chrism tyvm
17:01 UtahDave you're welcome!
17:02 jkleckner joined #salt
17:02 chrisgilmerproj1 joined #salt
17:03 UtahDave Hey, ange! How's it going?
17:05 diegows joined #salt
17:08 sashka joined #salt
17:09 oliv_mc joined #salt
17:10 Ryan_Lane joined #salt
17:14 Xeago joined #salt
17:16 jbunting1 joined #salt
17:18 lbotti joined #salt
17:18 Nazca joined #salt
17:18 Nazca joined #salt
17:19 bejer joined #salt
17:26 oz_akan_ joined #salt
17:28 steamraven joined #salt
17:29 shiznit joined #salt
17:33 LyndsySimon joined #salt
17:35 whit joined #salt
17:36 mollerstrand joined #salt
17:36 majoh joined #salt
17:37 aat joined #salt
17:41 ruimarinho does pillar.get() have a finite level nesting limit?
17:41 UtahDave ruimarinho: I'm not aware of it having a nesting limit.  Are you running into one?
17:42 ruimarinho not yet, but since I'm defining the pillar keys while I'm configuring the template, sometimes I run 3 or 4 levels deeps and just wanted to make sure I wouldn't hit any known limitation
17:43 UtahDave Yeah, I think you're fine
17:43 ruimarinho thanks
17:44 jacksontj ruimarinho: I just checked the code-- no limit :)
17:44 teskew akoumjian: you should make it apparent that people should add information to their README about the states they have in their repos for saltstarters.org so that it would be so much easier to search and find specific states
17:44 ruimarinho nice :) thanks a lot
17:44 akoumjian teskew: yes, I actually have been neglecting saltstarters for a while. that's in the bucket list.
17:44 akoumjian maybe this week
17:46 UtahDave UtahDave.eat_lunch()
17:46 returntosender joined #salt
17:47 ruimarinho it would be very nice to have a better stacktrace when a jinja template fails to compile due to a missing variable
17:47 bluemoon joined #salt
17:48 jlund joined #salt
17:49 Xeago joined #salt
17:50 f4cl3y joined #salt
17:56 Gifflen joined #salt
17:57 auser joined #salt
18:05 bluemoon joined #salt
18:06 jaequery joined #salt
18:07 auser joined #salt
18:07 milind joined #salt
18:08 whit joined #salt
18:10 oliv_mc left #salt
18:10 oliv_mc joined #salt
18:11 oliv_mc left #salt
18:16 spmurray ruimarinho: belated response to your note, but my answer to that issue has been to write tests to assert structure, existence of keys, etc, of my custom grains. Im working with a pretty uniform data model though, so that solution fit well
18:19 jacksontj joined #salt
18:20 Jahkeup_ joined #salt
18:21 koolhead17 joined #salt
18:22 chrism so i've got a bit of an odd problem that I'm not sure how to solve here
18:22 chrism I have a VM image that has salt-minion installed, and set to come up for the first time when a new VM is provisioned.  it automatically hits my master salt server
18:23 chrism now if I reinstall that image, I lose connectivity between master and minion due to the key change
18:23 chrism is there any sane way to get around that?
18:23 dthom91 joined #salt
18:23 spmurray is this for dev? or more of a production environment?
18:24 chrism salt serves both environments
18:24 spmurray well one option would be to turn on auto-accept in the master config
18:24 spmurray but clearly less secure
18:25 chrism that actually doesn't work because of the matching hostnames
18:25 chrism since the VM is reprovisioned w/ the same hostname, just different keys
18:25 spmurray oh I see, its a key mismatch
18:25 chrism aye
18:26 spmurray theres always open_mode, which turns off authentication, but unless the environment is secure thats probably a bad idea
18:26 chrism aye, i wouldn't wanna go down that road
18:26 chrism hehe
18:26 spmurray is the VM sandboxed?
18:26 spmurray e.g. does it come up in a fresh state each time?
18:27 chrism yep, it's a xen vm and / on it is essentially a tarball that gets extracted
18:27 __gotcha joined #salt
18:27 chrism like i know i could snapshot each environment w/ the key in place, but that get really hairy w/ all the VMs i've got
18:28 spmurray you could add a user + ssh key that has ssh access to your master box, then automate the 'salt-master -d <host-id>' each time it wakes up for the first time
18:28 blee joined #salt
18:28 chrism hrm
18:28 spmurray (im thinking of hacks now, because Im not sure of a Salty way to do it)
18:28 chrism yeah
18:28 chrism i getcha
18:29 chrism there might be a way to do that in a secure fashion
18:29 chrism actually hrm.
18:32 anteaya joined #salt
18:34 EWDurbin joined #salt
18:35 jessep joined #salt
18:37 __gotcha_ joined #salt
18:38 redondos how can write this in an sls? if /etc/foo does not exist, execute `bar'
18:38 rglauser joined #salt
18:40 __gotcha_ joined #salt
18:42 zooz joined #salt
18:42 chrism redondos: one sec, I use a cmd.run I believe to do that
18:43 chrism oh actually I do it with an unless
18:43 dthom91 joined #salt
18:43 chrism redondos: https://gist.github.com/cmeisinger/6ddd3a7b55571762b123
18:43 redondos thanks, chris!
18:43 chrism np
18:44 chrism obv. strip out the conditionals there, but yeah that should getcha started
18:45 krissaxton joined #salt
18:48 mikedawson_ joined #salt
18:51 Koma joined #salt
18:52 pentabular joined #salt
18:52 jhujhiti joined #salt
18:53 jaequery joined #salt
18:54 __gotcha_ joined #salt
18:54 __gotcha_ joined #salt
18:55 dthom91 joined #salt
19:00 jacksontj joined #salt
19:08 milind joined #salt
19:09 basti joined #salt
19:11 chadhs joined #salt
19:13 redondos what's a good way of executing a command only once? creating a file?
19:13 LyndsySimon joined #salt
19:13 UtahDave redondos: from the cli on your master?
19:14 kula or do you need to do a "if this command has never been ran on a particular minion, run it once and only once"?
19:14 kula (so that it catches new minions in the future)
19:14 clusterfudge joined #salt
19:15 clusterfudge Greetings, all!
19:15 clusterfudge Anyone know if there's a way to make the pkg.installed module to download packages only?
19:15 clusterfudge as opposed to download and install
19:17 UtahDave clusterfudge: I'm not sure if Salt has that built in yet.
19:18 UtahDave let me see
19:18 clusterfudge @UtahDave: not according to the public docs, i thought i'd come here to see if there was deeper knowledge
19:19 UtahDave terminalmage: do you know if there's a way to do what clusterfudge is attempting?
19:19 mgw clusterfudge: what distro are you on?
19:20 clusterfudge ubuntu 12.04.2
19:21 clusterfudge apt-get will support it, but neither the pkg.installed nor the apt.install modules seem to support it (per the documentation)
19:21 ange UtahDave: hey hey
19:21 UtahDave hey, ange. How's it going?  Sorry I missed you earlier!
19:23 ange UtahDave: fine, back to square one for the job search :/, so a bit more time to look into salt and the rbenv thing
19:23 ange UtahDave: and you ?
19:23 UtahDave good!
19:24 DeepThoughts left #salt
19:25 ange UtahDave: oh I just saw that the salt rbenv doc has been updated
19:25 steamraven Hey guys, is there any way in a top file to say: if state.sls exists then apply, else skip?
19:26 oliv_mc joined #salt
19:26 cnelsonsic joined #salt
19:26 UtahDave ange: yeah, i think terminalmage updated the docs a bit to make it obvious that it's new
19:26 ange UtahDave: and apparently the 0.16 is getting ready https://github.com/saltstack/salt/issues/5703
19:27 cnelsonsic left #salt
19:27 Lucas_- joined #salt
19:27 UtahDave steamraven: what version of Salt are you on?  We changed the behavior so that if the sls file doesn't exist, Salt will log that that state file didn't exist and then continue with the highstte
19:27 mgw UtahDave, clusterfudge: maybe there should be a pkg.downloaded
19:27 UtahDave ange: yeah, we'll probably have an official release announcement this evening
19:28 ange congrats
19:28 UtahDave thanks
19:28 UtahDave mgw: I can see that being useful
19:29 mgw clusterfudge: Are you just wanted it not to run?
19:29 mgw Or do you want it downloaded but not installed?
19:29 baniir joined #salt
19:29 steamraven utah dave: git develop, and its actually the pillar top
19:29 clusterfudge @mgw: downloaded but not installed
19:29 ange UtahDave: I am going to upgrade the .py locally to check if it works, and I'll wait for the update
19:29 clusterfudge i'm building some deployment tools on top of salt, and I want to transfer all bits to the box (from git or apt or whereever) without disrupting the running service
19:29 clusterfudge as separate steps
19:29 mgw clusterfudge so apt-get —download-only
19:30 clusterfudge @mgw: correct
19:30 terminalmage UtahDave: catching up on scrollback now
19:30 clusterfudge I'm just going to push it into a script and invoke that from salt
19:30 clusterfudge it's somewhat less elegant
19:30 clusterfudge but meh
19:30 UtahDave clusterfudge: if it's just a matter of adding a --download-only,  you might be able to just pass that into the state
19:31 terminalmage UtahDave: states pass through kwargs, so we can do apt-specific stuff in the pkg state
19:31 terminalmage clusterfudge: please open an issue, should not be difficult
19:31 mgw UtahDave: What's the policy on features across distros? E.g., if I added this and/or the don't start functionality for apt but not rpm...
19:32 mgw terminalimage: you answered my question, thanks
19:32 mgw well, partially
19:32 terminalmage UtahDave: basically, the install function in apt.py would just check for the value in the kwargs
19:32 terminalmage it's what we do for specifying a repo
19:33 UtahDave mgw: It's fine to add distro specific stuff.  If someone wants it, they can implement it for rpm, if you don't get to it
19:33 terminalmage mgw: yeah, look at yum, it uses enablerepo/disablerepo/etc.
19:34 mgw UtahDave, terminalimage: Thanks
19:34 terminalmage that means you can put "- enablerepo: epel-testing" in the state and it gets passed through to the install function in the kwargs dict
19:34 terminalmage but if you put it there and it is not supported it is just ignored
19:34 clusterfudge @mgw, UtahDave, thanks for the help, guys
19:35 backjlack joined #salt
19:37 auser hey all
19:38 StDiluted joined #salt
19:38 StDiluted quick question, I want to use a different user on a salt minion to check out a git repo. Is it possible to remote exec as a different user?
19:39 UtahDave hola, auser
19:39 mgw terminalimage: earlier I was talking to UtahDave about preventing apt from starting/enabling services automatically. Would you suggest just adding enable: True/False as a pkg.installed param, similar to service?
19:39 UtahDave StDiluted: are you using cmd.run?
19:39 StDiluted UtahDave, yes, but not married to that if there is a better way
19:39 cedwards joined #salt
19:39 UtahDave StDiluted: there's a runas option for cmd.run
19:41 akio Where can I find information on the salt virt-runner module?
19:42 akio just found readthedocs, nvm
19:42 StDiluted UtahDave, what would be the CLI syntax for that?
19:45 UtahDave http://docs.saltstack.com/ref/modules/all/salt.modules.cmdmod.html#salt.modules.cmdmod.run
19:45 StDiluted cool, thanks
19:45 UtahDave StDiluted: salt '<the minion name>' cmd.run ls runas=cooluser
19:46 krissaxton joined #salt
19:50 krissaxton1 joined #salt
19:50 StDiluted That's exactly what I needed
19:50 StDiluted I have trouble navigating the docs
19:50 StDiluted not sure if that's me or the docs
19:56 toastedpenguin is it possible to run interactive commands from the master?
19:57 toastedpenguin I'd like to be able to install RHEL FreeIPA from the master interactively if possible
19:59 UtahDave toastedpenguin: do you mean so that you can see the different user prompts that pop up when installing FreeIPA?
20:00 toastedpenguin yes
20:00 toastedpenguin I mean I could always resort to --unattended
20:00 toastedpenguin but was curious if that sort of interaction was possible
20:01 chrisgilmerproj joined #salt
20:02 jacksontj joined #salt
20:02 UtahDave toastedpenguin: No, it's not possible right now.  Commands are sent out through the pubsub interface and then the result is returned to the master
20:02 UtahDave so you don't get to see anything until the command has already finished executing
20:02 toastedpenguin ok, is that something that is planned?
20:02 UtahDave There are some tentative plans to build something like that in the future, but nothing has been implemented yet
20:03 toastedpenguin or even possible
20:03 mgw toestedpenguin, at least with apt, you can preseed the selections
20:04 toastedpenguin ok, I'll keep an eye open for it, we have started investigating setting up hosts that don't have logins, so while I used the FreeIPA example in reality we would like it so no one logs into a server for security purposes
20:04 zooz joined #salt
20:04 toastedpenguin so we kickstart everything and use salt to manage the host and if we need to upgrade we just rebuild the host
20:04 mgw http://docs.saltstack.com/ref/states/all/salt.states.file.html#salt.states.file.accumulated  <— docs say string or list, can it be a dict or list of dicts as well?
20:06 toastedpenguin mgw:  ya FreeIPA can be completely unattended install so there are options, a lot of services can be installed or enrolled unattended
20:06 toastedpenguin but having the ability to not no allow ssh access to a host would allow our production servers to be more secure
20:07 toastedpenguin we currently use ramdisks for storing data so its "recordless"
20:07 toastedpenguin just trying to find ways to make our production instances uber secure
20:09 UtahDave toastedpenguin: we have several users who disable ssh on certain production servers and completely manage them through Salt
20:09 UtahDave toastedpenguin: I haven't set up FreeIPA myself yet, but I'm sure you can completely do it with Salt.
20:09 bluemoon joined #salt
20:10 toastedpenguin UtahDave: actually if we eliminated ssh we probably wouldnt need FreeIPA on the hosts
20:10 UtahDave mgw: I'm not sure on that. You'd probably just have to test it out.
20:11 UtahDave toastedpenguin: true.
20:11 toastedpenguin we run either tomcat or cassandra
20:11 toastedpenguin so we only need  to stop/start/restart services
20:11 toastedpenguin and change files etc.
20:11 kermit joined #salt
20:12 redondos kula: exactly what you said: only run this command on new minions once
20:12 toastedpenguin like to eliminate having to have ssl on the tomcat hosts because the keystore pw is in clear text and put that on a LB or nginx proxy
20:12 redondos UtahDave: on SLS, not the command line
20:12 toastedpenguin will have to give this some testing, with the help I have gotten so far I am much further with salt in our env than I ever was with puppet.....
20:13 toastedpenguin and being able to run remote commands is HUGE
20:13 toastedpenguin well without SSH and on multiple hosts at the same time
20:16 KennethWilke toastedpenguin: that is awesome to hear :P
20:16 KennethWilke never going back, are you? :)
20:17 DerekRBN never ever
20:17 ipmb joined #salt
20:17 toastedpenguin no, I actually removed  what limited amount of puppet configs I had and did it all in salt 2 days after finding salt
20:18 toastedpenguin and I couldn't get the 3 data center different config files need to be served based on location thing working in puppet without a @#$% ton of issues
20:18 KennethWilke oh yeah i remember discussing that with you
20:18 toastedpenguin one file a few lines and it works well
20:18 KennethWilke did you go with pillar in the end for that?
20:18 toastedpenguin no, but I might eventually
20:19 kula redondos: then yeah, i'd create a file that indicates the command has been ran, and run the command as a pre-req to that file being created
20:19 toastedpenguin I think part of my issue is not fully grasping the power of pillar
20:19 faulkner joined #salt
20:19 ydavid joined #salt
20:19 redondos thank you, kula
20:19 KennethWilke fortunately it's not a necessary tool, it just fills a space to add convenience
20:20 toastedpenguin I used the packages example to get a feel for it, but not quite sure how to properly use pillar
20:20 KennethWilke often a great deal of convenience
20:20 toastedpenguin beyond that
20:20 KennethWilke yeah, i use it for endpoint configuration, similar to how it shows it for different package names in different os's
20:20 toastedpenguin so the template I used worked for the resolv.conf need I had
20:20 jaequery joined #salt
20:20 KennethWilke so theres a pillar sls for dev and prod
20:21 toastedpenguin I only have CentOS but it is nice to use for converting a common name to the distro equvilant
20:21 toastedpenguin see that is where I need to figure out to best use pillar
20:21 KennethWilke like in my case i have a redis box for dev and a redis box for prov, i think the best use is always just based on your need
20:21 toastedpenguin because I have dev, staging, prod and netops environments, plus 3 differen data centers
20:22 KennethWilke yeah that makes it a bit trickier
20:22 blee_ joined #salt
20:22 jaequery joined #salt
20:22 KennethWilke oh, i also use git to manage my pillar stuff
20:23 KennethWilke which is kinda quirky, but it works for me
20:23 pentabular joined #salt
20:23 KennethWilke more than anything i think it's just there to provide separation if you need it at that level
20:24 toastedpenguin think I need to attend one of those online Salt training sessions or see if anyone here in Chicago does any meetups etc.
20:24 KennethWilke and then external pillar really makes a lot of fun things possible
20:24 chrisgilmerproj joined #salt
20:24 toastedpenguin I am a see an example learning type of person
20:24 KennethWilke yeah that makes sense
20:24 toastedpenguin once I see an example that works I start groing it
20:24 KennethWilke i usually have to do it myself to fully understand it
20:24 toastedpenguin *growing from it
20:25 toastedpenguin ya I need that too but having a working example lets me grasp the concepts being shown, like ah ok thats why that is done that way
20:25 Jahkeup_ joined #salt
20:26 KennethWilke yeah i don't think there's a lot of good sample work out there for it yet, it is something i'd like to contribute towards
20:26 toastedpenguin plus seeing/hearing how others use Salt will get my mind churning with ideas I hadn't thought of
20:28 toastedpenguin WTF no Saltstack meetup group in Chicago...
20:28 toastedpenguin lol
20:29 jaequery joined #salt
20:30 faulkner None in SF either.
20:30 redondos how can i capture the output of cmd.run and store it in a file on the minion?
20:31 KennethWilke redondos: i think you can use bash output redirection for that
20:31 faulkner Oh, nm, there is one in SF.
20:31 redondos oh, right
20:31 KennethWilke :p
20:32 redondos yeah big duh :)
20:33 mgw I think status.pid needs some love:
20:33 mgw $ sudo salt-call status.pid asdfasdfas3ew
20:33 mgw local:
20:33 mgw 5287
20:33 mgw 5288
20:34 mgw it's broken when used with salt-call, and I know why
20:40 opapo joined #salt
20:40 Newt[cz] joined #salt
20:42 carlos__ joined #salt
20:44 toastedpenguin if I want to have config files that have host specific attributes appended to it, e.g. hostname/fqdn is that possible using salt?
20:44 toastedpenguin for instance I have a motd and an ssh banner where I specify the hostname of the host being accessed
20:45 toastedpenguin so I'd like to have the file maintained by salt but with the hostname of the host in the file on the minion
20:45 racooper toastedpenguin,  you'd need a template
20:46 toastedpenguin racooper: would you by any chance have an example of such a template?
20:47 UtahDave toastedpenguin: you're in Chicago?
20:47 toastedpenguin yep
20:47 toastedpenguin work downtown live in the burbs
20:48 toastedpenguin work in one of the worlds tallest and most secure building....lol
20:48 UtahDave cool. We're probably going to have a Sprint location there. One of our employees lives outside Chicago.
20:48 jschadlick is 0.16.0 still set to come out tonight?
20:48 StDiluted UtahDave, we have three people attending on July 27 from my office
20:48 racooper toastedpenguin, look at https://github.com/daviddyball/salt/tree/master/salt/denyhosts
20:49 UtahDave racooper: thanks!
20:49 UtahDave StDiluted: where's your office?
20:49 StDiluted Albuquerque, NM
20:49 mgw Does it make sense that if service.running with enable: True results in a failed start, but the service is already enabled…. that it responds "no changes" rather than failure?
20:50 toastedpenguin racooper: ah so you use the grain in the conf file/template and salt master appends this
20:50 mgw I'd fix it, but I'm not sure how a state should respond if it does one thing but fails to do another...
20:50 bluemoon joined #salt
20:50 LarsN UtahDave: for the record I love this project & it's community of volunteers, helpers, etc.
20:50 racooper apparently so. the denyhosts is one I copied for my own use :)
20:51 Xeago joined #salt
20:51 toastedpenguin UtahDave:  Sprint location?
20:51 UtahDave LarsN: We love you, too!!   :)
20:51 LarsN UtahDave: I just cause more work for y'all.
20:51 UtahDave he he.
20:51 UtahDave toastedpenguin: let me find out if we've finalized the location yet
20:51 LarsN filing bugs without patches.
20:52 toastedpenguin cool
20:52 UtahDave LarsN: :)
20:54 jschadlick1 joined #salt
20:58 UtahDave toastedpenguin: ok, so I was right. We're going to have a Sprint venue in Chicago, we're just nailing down the location still.
20:59 toastedpenguin UtahDave: ok cool, I will be there when it occurs!
21:00 UtahDave awesome! that will be awesome to have you there, toastedpenguin
21:00 baoboa joined #salt
21:02 raulp joined #salt
21:03 toastedpenguin racooper: thanks for the example, works like a charm
21:04 toastedpenguin add template: jinja and all the magic happens
21:05 hazzadous joined #salt
21:06 bluemoon joined #salt
21:07 torandu joined #salt
21:10 lunarben joined #salt
21:12 RickW I'm looking into adding Google Compute Engine support to Salt Cloud.  Should I be developing against the code in 'master' or the code in 'develop'?  (Also any other comments that you have about that effort would be appreciated)
21:14 alazylearner joined #salt
21:15 UtahDave RickW: You'll want to code against develop
21:16 UtahDave RickW: redbeard2 has already done some work towards that. Google gave us an early account
21:16 UtahDave RickW: I'm sure he could use some help because it hasn't been released yet.  Have you done anything toward that yet?
21:16 RickW OK, thanks.  That's what I thought, just wanted to double-check.
21:17 RickW I've done some work, but primarily, I just finished coding up support for GCE for libcloud, so I'm looking at salt cloud now
21:17 Gifflen joined #salt
21:17 RickW The libcloud change hasn't been merged in yet, but I expect that to happen in the relatively near future
21:17 UtahDave RickW: that's awesome.
21:18 UtahDave RickW: let me get redbeard2. I think he'd like to discuss that with you.   Just a minute
21:19 RickW Here is the libcloud pull request that is currently under review: https://github.com/apache/libcloud/pull/115
21:20 redbeard2 joined #salt
21:22 UtahDave cool, RickW
21:24 toastedpenguin can yo uhave more than one pkg require under a managed file?
21:24 UtahDave toastedpenguin: you can have as many as you want or need
21:25 toastedpenguin ok cool
21:25 toastedpenguin thx
21:28 ipmb left #salt
21:28 up_the_irons joined #salt
21:29 mgw anybody have experience with accumulators?
21:30 mgw I was under the impression from the docs that there would be an 'accumulator' dict available to the templated file.
21:32 jacksontj joined #salt
21:37 jaequery joined #salt
21:37 rberger joined #salt
21:43 bluemoon joined #salt
21:45 baniir joined #salt
21:45 teskew UtahDave: still going to have the new salt release in launchpad ppa this evening?
21:45 UtahDave teskew: I'm pretty sure that's the plan!
21:49 * akoumjian needs to start deploying from his own debs because autoscaled minions sometimes break when they upgrade automatically and old versions disappear from launchpad
21:50 UtahDave akoumjian: yeah, that's a common solution. We're actually building a saltstack repo where we can keep all the versions.
21:50 akoumjian UtahDave +1
21:51 davea joined #salt
21:55 baniir joined #salt
21:59 kermit joined #salt
22:02 StDiluted does file.symlink work on directories as well?
22:05 EugeneKay symlinks(in UNIX) don't care if the target is a file or a folder, so yes, they should work just fine.
22:06 StDiluted true enough
22:07 jacksontj joined #salt
22:15 x_or joined #salt
22:17 StDiluted popular opinion: better to use rbenv via salt, or better to compile the version I want and package it using fpm?
22:18 UtahDave StDiluted: the ruby-heads seem to like the rbenv option. Sysadmin types will strangle you if you don't package it properly.  :)
22:18 StDiluted <- sysadmin of the servers
22:18 StDiluted soooo
22:19 StDiluted what I say goes
22:19 StDiluted I don't so much care about the package being pretty
22:20 blee joined #salt
22:20 jeddi StDiluted: my vote would be package it and pop it on a repo that you manage, or at least ship the file out and dpkg --install it locally.   guessing rbenv is the pip of the ruby world?  i guess it's pretty pervasive.  still, dev tools on production machines is a bit ugly.
22:20 StDiluted jeddi: that's my approach, and I'd rather not have build tools on the hosts.
22:21 StDiluted i don't want to manage a repo
22:22 jaequery joined #salt
22:27 dave_den rbenv doesn't require build tools
22:28 dave_den only if you want to actually install the ruby version from rbenv do you need them.
22:31 jeddi StDiluted: yeah - i'm not keen on repos where i am now either - too much overhead.  i'd just deploya nd install via salt.
22:31 jeddi dave_den: okay, taa.   i've probably cut-n-paste some rbenv stuff in the past, but ruby's not my thing.
22:32 * jalbretsen strangles UtahDave for not packaging things properly
22:33 dave_den i hear ya. rbenv and rvm are super handy when you need to have multiple ruby versions installed in one server, but otherwise there's no reason not to compile the version you need and package it. or find one in someone else's repo...
22:34 jalbretsen I'm still push my devs to stop using old old old versions of MySQL.
22:34 jalbretsen *pushing
22:43 jacksontj Having some issues with prereq requisite :/ looks like i have it doing my oor stuff, and isntalling properly and stopping the service, but not re-starting it properly
22:44 chrisgilmerproj left #salt
22:49 whit joined #salt
22:56 kenbolton joined #salt
23:00 jetblack joined #salt
23:03 conan_the_destro joined #salt
23:04 conan_the_destro joined #salt
23:05 ahammond joined #salt
23:07 adotbrown joined #salt
23:08 dthom91 joined #salt
23:16 bhosmer joined #salt
23:17 pygmael joined #salt
23:22 bluemoon joined #salt
23:28 aranhoide joined #salt
23:29 whit joined #salt
23:29 cxz joined #salt
23:34 rfgarcia joined #salt
23:36 johntron joined #salt
23:37 davisamo joined #salt
23:43 baniir joined #salt
23:52 aat joined #salt
23:53 ajw0100 joined #salt
23:54 jschadlick1 left #salt
23:59 adotbrown joined #salt

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