Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-09-24

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

All times shown according to UTC.

Time Nick Message
00:00 nu7hatch joined #salt
00:03 rasp hello, new salt user here. I'm trying to install a simple package, redis, with salt, however, it doesn't appear to create the redis user (in the .rpm install, I assume). If I simply yum install the package the user is correctly installed. the config runs as the root user. here's a paste bin with some info. any ideas? http://pastebin.com/ypjRAZTc
00:04 g4rlic rasp: shut off selinux, remove the redis rpm, then try highstate again.
00:05 g4rlic (just a hunch)
00:05 rasp g4rlic: thanks, I'll try that
00:05 g4rlic rasp: there's a well known bug in CentOS that prevents the salt-call and salt binaries from correctly executing %post scripts in RPM's
00:07 g4rlic https://github.com/saltstack/salt/issues/6735
00:09 rasp that was it, many thanks :)
00:09 g4rlic No problem. ;)
00:15 saurabhs joined #salt
00:17 pipps1 joined #salt
00:20 cdsrv joined #salt
00:21 f47h3r UtahDave: I've followed up with the issue, from what I can tell the roster file is not an option that can be set in the __opts__ data structure
00:22 f47h3r and the roster/flat.py fails to use a configuration option other than config_dir/roster
00:23 f47h3r UtahDave: I'm not sure if this helps, but I'd be willing to help out.
00:23 f47h3r I'm brand new to the code base, so I'm not sure how all the moving parts work
00:27 eliasp I reworked the Gentoo support for salt-bootstrap quite a bit: https://github.com/saltstack/salt-bootstrap/pull/228 … feel free to review
00:28 cdsrv eliasp,
00:28 cdsrv do you think there is a need for a universal salt-bootstrap script?
00:28 jslatts joined #salt
00:29 eliasp cdsrv: what do you mean? it's already there…
00:30 cdsrv yes, of course..
00:30 cdsrv not many dependencies ahead of the bootstrap script
00:34 cdsrv what about abstraction layers for application roles? is there anything out there currently for that type of thing?
00:35 dthom91 joined #salt
00:39 Gwayne joined #salt
00:45 faldridge joined #salt
00:45 pdayton joined #salt
00:50 mmilano joined #salt
00:52 UtahDave f47h3r: look in salt/config.py for the list of defaults for the master.
00:53 robawt aloha folks
00:53 robawt can I order events in top.sls?
00:53 robawt is it just - order: 1
00:53 robawt ?
00:53 mmilano joined #salt
00:58 nu7hatch joined #salt
01:00 nu7hatch joined #salt
01:07 higgs001 joined #salt
01:12 morganfainberg joined #salt
01:24 dthom91 joined #salt
01:31 jslatts joined #salt
01:31 saurabhs joined #salt
01:36 dthom91 joined #salt
01:36 shinylasers joined #salt
01:39 NV joined #salt
01:39 NV so, perl
01:40 NV anyone discovered a non-painful way to manage perl and modules, ideally with salt?
01:40 NV perlbrew perhaps?
01:43 jbunting joined #salt
01:43 pipps1 joined #salt
01:46 cdsrv joined #salt
01:55 jalbretsen joined #salt
01:56 dthom91 joined #salt
02:02 __number5__ what's the best way to manage AWS VPC (create/destroy dynamically) with salt? salt-cloud seems not having much support for VPC
02:07 redondos joined #salt
02:08 morganfainberg joined #salt
02:10 deepakmd_oc joined #salt
02:12 rgbkrk joined #salt
02:13 Brew joined #salt
02:17 xl1 joined #salt
02:18 qba73 joined #salt
02:31 oz_akan_ joined #salt
02:32 malinoff Hi all
02:36 saurabhs joined #salt
02:37 morganfainberg joined #salt
02:39 saurabhs1 joined #salt
02:48 salt_ joined #salt
02:49 hosmelQ joined #salt
02:50 Ryan_Lane joined #salt
02:54 hosmelQ hi guys, i get this error "no top external file.  No Top file or external nodes data matches found" when i try "state.highstate", someone help me please ? :(
02:56 salt_ hello, good night... I have some doubts with the deployment of states saltstack
02:56 salt_ :(
03:00 redondos joined #salt
03:02 jpeach joined #salt
03:04 salt_ I feel ignored
03:22 mwillhite joined #salt
03:28 heckman joined #salt
03:30 Lue_4911 joined #salt
03:30 packeteer dont we all
03:31 malinoff hosmelQ: Do you have a top.sls file in /srv/salt/ ?
03:31 malinoff salt_: What's the question?
03:40 morganfainberg joined #salt
03:41 packeteer hmm, my minions have failed after a reboot: [ERROR   ] The Salt Master has cached the public key for this node, this salt minion will wait for 10 seconds before attempting to re-authenticate
03:41 packeteer any ideas?
03:42 packeteer nm, my screw up :)
03:42 packeteer i changed the minions domain
03:43 jpeach joined #salt
03:46 oz_akan_ joined #salt
03:52 faldridge joined #salt
03:57 forrest joined #salt
04:03 StDiluted joined #salt
04:08 Gifflen joined #salt
04:09 z3uS joined #salt
04:19 morganfainberg joined #salt
04:19 CheKoLyN joined #salt
04:23 sssslang joined #salt
04:29 Kraln joined #salt
04:35 morganfainberg joined #salt
04:54 cdsrv joined #salt
04:54 fragamus joined #salt
04:55 rgbkrk What does "faster AES" in "faster AES encryption for payload communication" mean from this section of the docs: http://salt.readthedocs.org/en/latest/topics/index.html#building-on-proven-technology
04:57 oz_akan_ joined #salt
04:58 forrest It's just saying that it uses public keys for authentication, then uses Advanced Encryption Standard for pushing payloads to the boxes.
04:58 forrest because it's faster than using keys
04:59 g4rlic Specifically, AES is a symmetric cipher, which is always faster than asymmetric (public key) ciphers.
04:59 g4rlic rgbkrk: https works in a similar fashion
04:59 g4rlic Public key algorithms are slow, and can't encrypt a whole lot of data.
04:59 rgbkrk That sounds like normal usage of AES
04:59 g4rlic It is.
04:59 g4rlic hence the "proven technology" bit.
05:00 g4rlic (full disclosure: I'm not a saltstack dev, I just know enough about crypto to be dangerous.)
05:00 rgbkrk Yeah, I'm mostly nitpicking about the verbage in that section.
05:00 rgbkrk Because it doesn't say *what* it's faster than.
05:01 g4rlic rgbkrk: it's a subtle dig at Asnible, Slack, and other CM systems that rely entirely on making a new SSH tunnel every time they need to do something.
05:01 Kraln joined #salt
05:01 forrest it's faster than basically every other CM system out there
05:01 Kraln- joined #salt
05:02 rgbkrk The subtle dig at Ansible I get, and that clears it up.
05:02 g4rlic forrest: to be fair, I believe you can make CFengine 3 work in the same manner.
05:02 forrest oh can you?
05:02 forrest I've never messed with chef enough to do that
05:02 g4rlic yep.
05:02 forrest that's good they have that functionality
05:02 g4rlic wait, chef?  No idea bout chef.
05:02 g4rlic I said CFEngine.
05:02 g4rlic you may have misread,.
05:02 g4rlic :D
05:03 forrest the joys of multitasking :P
05:03 g4rlic rgbkrk: Glad I could help. ;)
05:03 robawt aloha folks
05:04 rgbkrk Which path should I be going down if I'd like to eventually get to a state of auto-deploying from a git repo (master -> dev server, candidate tag -> staging, release tag -> production)
05:05 rgbkrk g4rlic - Thanks. Just seemed odd in the docs.
05:06 g4rlic rgbkrk: np. The lay-person really would be confused reading that, iffn' they didn't know that public key == slow.
05:07 rgbkrk I think for a layperson, they would probably go "yay, crypt, cool"
05:07 rgbkrk I read it and was skeptical, as I saw it as "we sped up AES"
05:07 rgbkrk Sped up AES by reducing rounds?
05:07 rgbkrk Using a default key and pre-computing?
05:07 rgbkrk :P
05:08 rgbkrk I'm used to the whole public key to encrypt your symmetric key, symmetric key encrypts your payload
05:09 rgbkrk It's the kind of thing in the docs where I'd make a pull request if I knew *exactly* what the devs were going for.
05:11 rgbkrk Any ideas about git automation?
05:12 rgbkrk Currently reading http://blog.gibbon.co/posts/2013-06-12-salting-your-django-stack.html
05:12 rgbkrk Would love to find other awesome examples or tips
05:12 xinkeT joined #salt
05:14 isomorphic joined #salt
05:15 xinkeT joined #salt
05:21 forrest check out https://github.com/terminalmage/djangocon2013-sls rgbkrk
05:21 forrest it uses git for pulling down the content.
05:22 sssslang Hi guys, if i use jinja in an sls file, is there a way to show the render result?
05:22 robawt sssslang: won't a test=True show the file change?
05:22 robawt it should show a diff of the final version
05:23 sssslang robawt: yep, but I want to ensure the rendered template is correct.
05:24 rgbkrk forrest - Thanks!
05:25 forrest yea p
05:25 forrest *np
05:34 Kraln joined #salt
05:36 dthom91 joined #salt
06:02 faldridge joined #salt
06:08 drdran joined #salt
06:30 linjan joined #salt
06:42 ml_1 joined #salt
06:49 az87c joined #salt
06:52 elsmorian joined #salt
06:55 balboah joined #salt
07:01 drdran joined #salt
07:07 qba73 joined #salt
07:13 drdran left #salt
07:18 Furao joined #salt
07:27 adepasquale joined #salt
07:30 scalability-junk hey anyone got some ideas about docker and salt?
07:30 scalability-junk seems like docker could be a great way to distribute stuff or prebuilt images... sorta and then let salt run the continues upgrades?
07:34 malinoff joined #salt
07:34 ml_1 joined #salt
07:36 agend joined #salt
07:43 kiorky scalability-junk: im writing atthe moment some modules around
07:43 kiorky i saw a bug around this 3 days ago
07:44 scalability-junk kiorky: any idea how you want it to be integrated? should it be a plain dicker linux container and salt does the setup as usual after that or do the presetup before deploying the container?
07:44 scalability-junk *docker not dicker :D
07:45 kiorky scalability-junk: as a state and module like other things
07:46 kiorky scalability-junk: in the module, it is more or less than a docker-py wrapper
07:46 kiorky uhm
07:46 kiorky let me commit my current wok :p
07:46 kiorky (its not working, i begun it 10 minutes aggo)
07:47 scalability-junk no worries I just wanted to get the general idea
07:48 kiorky scalability-junk: https://github.com/makinacorpus/makina-states/blob/master/services/docker.sls
07:48 kiorky whoops
07:48 scalability-junk thanks
07:48 kiorky wrong branch
07:48 kiorky :p
07:48 scalability-junk forshare i assume ;)
07:48 kiorky scalability-junk: http://goo.gl/s2R7uA
07:49 Ryan_Lane joined #salt
07:49 kiorky http://goo.gl/LN7Ida
07:49 kiorky scalability-junk: http://goo.gl/1Mbm1W
07:50 scalability-junk cool thanks
07:50 kiorky scalability-junk: uhm, wait i forgot one file
07:53 Furao joined #salt
07:55 kiorky scalability-junk: http://goo.gl/4qPLpF http://goo.gl/hPznkO http://goo.gl/jM0erU
07:55 kiorky scalability-junk: morning is always hard
07:55 kiorky :p
07:56 scalability-junk true
07:57 Furao_ joined #salt
08:00 kiorky scalability-junk: let me find the bug, but im doing concurrent work as we must advance here,i dont know if my stuff will enough good quality to ne proposed in salt without shame.
08:01 kiorky scalability-junk: https://github.com/saltstack/salt-cloud/issues/615
08:02 scalability-junk saltstack staff is mostly helpful in this regard.
08:05 AviMarcus joined #salt
08:12 Furao_ joined #salt
08:19 matanya joined #salt
08:24 Ryan_Lane joined #salt
08:36 zooz joined #salt
08:53 qba73 joined #salt
08:56 Gwayne joined #salt
08:57 tuxIO joined #salt
09:01 lemao joined #salt
09:05 tomi joined #salt
09:07 kiorky uhm, is someone can have a look to this pull request: https://github.com/saltstack/salt/pull/7426
09:08 kiorky s/is/if/
09:14 geak joined #salt
09:15 tomi joined #salt
09:16 ronc joined #salt
09:17 zooz joined #salt
09:21 mortis in  ssh state_auth you can give salt a source like for instance salt://blddiblah.pub, but how do you specify multiple keys here?
09:22 spacedentist left #salt
09:22 mortis do you have to define a ssh_auth pr user, or what?
09:24 mortis found out :)
09:25 bhosmer joined #salt
09:34 krissaxton joined #salt
09:37 NV mortis: what was the solution out of curiosity?
09:37 mortis using one ssh_auth pr user
09:37 mortis NV: ^
09:38 mortis but can also do like the last example here http://docs.saltstack.com/ref/states/all/salt.states.ssh_auth.html
09:42 NV ah i thought you meant for multiple ssh keys for a single user, with each key being in its own file
09:47 maveas Does a sls and directory with the same name conflict during an include? Ie. I have dir/file.sls and dir/file/somefiles
09:47 honestly yes, that conflicts
09:48 honestly don't know what happens off the top of my head, but it's in the tutorial somewhere
09:49 maveas honestly: Super, no worries then. Thanks :)
09:53 NV iirc the file.sys takes precedence over file/init.sls
09:53 NV that said
09:53 NV why not just use file/init.sls?
09:54 NV If both webserver.sls and webserver/init.sls happen to exist, webserver/init.sls will be ignored and webserver.sls will be the file referred to as webserver.
09:54 NV http://docs.saltstack.com/topics/tutorials/states_pt1.html#sls-file-namespace
10:10 giantlock joined #salt
10:10 Ixan I'm having an issue with cp.get_file. the file appears to be corrupt after transfer. cp.hash_file also shows that is has changed between master and minion. any suggestions where i should start to debug?
10:12 Kraln joined #salt
10:19 bhosmer_ joined #salt
10:26 krissaxton joined #salt
10:31 mordred joined #salt
10:41 matanya joined #salt
10:49 jesusaurus joined #salt
10:56 diegows joined #salt
11:01 halfss joined #salt
11:07 xl1 left #salt
11:09 gildegoma joined #salt
11:22 farra joined #salt
11:42 jbunting joined #salt
11:47 cyp___ joined #salt
11:49 krissaxton joined #salt
11:51 jamiem joined #salt
11:56 krissaxton joined #salt
11:56 felixhummel joined #salt
11:56 blee joined #salt
11:58 GoKage joined #salt
12:07 ede joined #salt
12:09 eliasp Ixan: have you tried to reproduce with an easily comparable file (small text-file)? to see what's actually changing?
12:09 eliasp Ixan: does the size change significantly?
12:11 jfalco joined #salt
12:15 lied joined #salt
12:16 lied hi
12:17 Damoun joined #salt
12:17 lied i'm new to configuration management systems. why should I choose salt over ansible? (I don't want to troll or start a flamewar…) for me both systems seems like copies besides the different approach on ordering the directives to the clients
12:20 ml_1 joined #salt
12:20 xl1 joined #salt
12:22 bhosmer joined #salt
12:22 jslatts joined #salt
12:25 carlos__ joined #salt
12:25 ml_11 joined #salt
12:25 Ixan eliasp: i've narrowed it down a bit. it occurs when i push, not with get_file. i think it might be caused by zmq
12:26 Ixan master is running linux and minion is aix 5.3, so there's bound to be some fun in debugging it
12:26 eliasp eeek, ok… that'll be quite a bit fun… good luck ;)
12:27 Ixan hehe, thx ;)
12:29 heckman joined #salt
12:30 balltongu joined #salt
12:34 scristian joined #salt
12:36 mwillhite joined #salt
12:38 nliadm joined #salt
12:40 olenz joined #salt
12:42 techdragon joined #salt
12:42 xl1 left #salt
12:46 svx_ joined #salt
12:46 bdf joined #salt
12:46 zfouts joined #salt
12:46 sinh joined #salt
12:46 eskp joined #salt
12:46 bui joined #salt
12:46 Heartmender joined #salt
12:46 rorski joined #salt
12:46 intchanter joined #salt
12:46 austin987 joined #salt
12:46 Tekni joined #salt
12:46 octarine joined #salt
12:46 Heartmender joined #salt
12:46 eskp joined #salt
12:47 bawnzai joined #salt
12:47 jbub joined #salt
12:47 bejer joined #salt
12:47 ldlework joined #salt
12:47 balltongu joined #salt
12:47 geak joined #salt
12:47 pjs joined #salt
12:48 akoumjian joined #salt
12:49 bhosmer joined #salt
12:50 absolute joined #salt
12:51 MK_FG joined #salt
12:53 linjan_ joined #salt
12:54 kiorky joined #salt
12:55 nu7hatch joined #salt
12:55 ipmb joined #salt
12:55 \ask joined #salt
12:56 danielbachhuber joined #salt
12:56 tseNkiN joined #salt
12:56 ggoZ joined #salt
12:57 elsmorian joined #salt
12:59 mmilano joined #salt
12:59 bhosmer_ joined #salt
13:02 oz_akan_ joined #salt
13:04 oz_akan_ joined #salt
13:04 farra joined #salt
13:04 imaginarysteve joined #salt
13:06 brianhicks joined #salt
13:07 Teknix joined #salt
13:08 juicer2 joined #salt
13:08 techdragon Anyone used the sentry returner ? Im having issues debugging it not working. No events arriving in sentry (sentry is tested working ok with the DSN details Ive configured in sentry) and I've tried running the returner with 'salt-call -l all'
13:09 Gifflen joined #salt
13:10 Teknix joined #salt
13:12 bemehow joined #salt
13:13 racooper joined #salt
13:22 halfss joined #salt
13:23 aleszoulek joined #salt
13:30 jeff__ joined #salt
13:36 Furao techdragon: I'm original author I'm not sure latest version in 0.16.x works anymore
13:36 Furao I maintain my own version I should publish
13:38 Furao https://gist.github.com/bclermont/dc857c1c4abd8dd5c723
13:42 gmoro joined #salt
13:42 qba73 joined #salt
13:44 alunduil joined #salt
13:45 techdragon @Furao I did see that earlier, wasn't sure what to make of it while searching. Is bclermont your repo ?
13:46 piffio joined #salt
13:47 tyler-baker joined #salt
13:47 balboah joined #salt
13:48 piffio anyone using overstate with gitfs here?
13:48 shinylasers joined #salt
13:49 toastedpenguin joined #salt
13:51 alunduil joined #salt
13:51 Brew joined #salt
13:52 Furao techdragon: yes
13:53 Furao but this is quite old (~ 6 months old), today I have a team of 7 persons and we work full time on these states
13:53 Furao well 2 are half-time
13:54 techdragon I'll take the opportunity to thank you, then Furao, I learned quite a bit from your repo when i got started using Salt. Was a huge help. And Ive been meaning to follow up on the "these states aren't publicly maintained anymore. but they had been improved a lot in private. if you're interested contact me." you left in your states repo here https://github.com/bclermont/states
13:57 Furao the reason I made it public in first place was because nowhere else I could find complete example, doc was incomplete (it was worst that time), all public repo of states were so simple they couldn't helped me
13:57 agend joined #salt
13:57 devinus joined #salt
13:58 adepasquale joined #salt
13:58 balltongu joined #salt
13:58 pipps1 joined #salt
13:59 adepasquale joined #salt
13:59 derelm joined #salt
13:59 adepasquale joined #salt
13:59 copelco joined #salt
14:00 Furao techdragon: today it's 3000 commits, a lot more states, a state testing framework, CI trough jenkins that all states are executed in different way (a single test run take between 10 to 15 hours on SSD drive) and it's almost 800 tests. I do have automatic discovery for monitoring, salt-cloud web ui, and fancy doc coming soon.
14:00 techdragon That was exactly why it was such a huge help. I've been meaning to make some contributions back myself. Haven't had time to polish anything up.
14:01 adepasquale joined #salt
14:01 shinylasers joined #salt
14:02 techdragon Ooooh Furao thats very impressive.
14:02 Furao the graphite and nrpe integration are now separate files
14:02 Furao it's optional
14:02 Furao and switched from gsyslog (my own syslog daemon) to rsyslog (config syntax is horrible, but easier to tame with salt)
14:02 Furao backup
14:03 Furao clustering in many state that support it
14:03 Furao find . -type f | grep -v '/.git/' | xargs wc -l | tail -n 1
14:03 Furao 34011 total
14:04 Furao same command in last version of github return 10344 lines
14:04 techdragon Thats a heck of a diff.
14:05 Furao and these states are now the recipe for bunch of clients :)
14:05 techdragon shared hosting stack or templates used by multiple top.sls files?
14:05 Furao totally separated infra, I don't do hosting
14:05 nu7hatch joined #salt
14:06 Furao it's client that all have their own requirements, some are in EC2, some private cloud, other "bare metal"
14:06 Furao public, private. all of them with custom states that depends on the evolution of what was in that git repo
14:06 hughsaunders left #salt
14:07 Furao it's not for me that I ended with PHP support in uWSGI :P
14:08 techdragon haha
14:09 techdragon So all your work stack side?
14:09 techdragon or is that just the base layer for your other work?
14:13 Furao the product I sell is access to low-level states, doc, process and consulting time. low-level states allow to build form scratch a complete infra (monitoring, stats, centralized logging, config management, backup, etc) and also pre-existing state of FOSS stuff, such as wordpress or mongodb. then, for clients, we create custom states, if they need to, or just the "roles" and the basic to deploy all that in their infra
14:15 Furao in few two cases, from scratch to deployed infra, it took 2 days, the problem was more gathering info than anything else :)
14:16 mohae joined #salt
14:19 shinylasers joined #salt
14:20 techdragon Furao thats quite a nice line of work.
14:22 nu7hatch joined #salt
14:24 pdayton joined #salt
14:24 Furao yes, I'm still unhappy with few things, but it progress fast, with such so much in the team :)
14:26 Furao my states are still all open source, but not on a public repo
14:26 Furao techdragon: what kind of changes you made?
14:29 abe_music joined #salt
14:30 techdragon nothing based off your states, I've mostly used them as reference. Most of the stuff I've been meaning to contribute back is little things, a few documentation fixes, a couple of extra utilities, returners, etc.
14:31 techdragon Looking forward to contributing formulas after 17 lands.
14:31 cnelsonsic joined #salt
14:32 Furao as soon as formulas format is complete, I'll switch to it
14:33 Furao but for now, it don't cover everything I do
14:33 Furao last time I checked formulas weren't impressive
14:34 lineman60 joined #salt
14:37 micah_chatt joined #salt
14:37 juicer2 joined #salt
14:38 opapo joined #salt
14:38 Furao and also it's problematic if you use gitfs and multiple environment
14:38 vipul joined #salt
14:39 Furao as you have to create a branch into ALL repos
14:39 Ahlee ain't that the truth.
14:39 techdragon Yeah theres still a few issues I see, but they are a nice step forward to modular app deployment.
14:40 Ahlee I'm very disenfranchised to gitfs backends
14:40 techdragon I don't use the gifts backend myself.
14:40 Ahlee finding init files in first repo, but then dropping to next gitfs repo for any corresponding state files
14:41 Ahlee that and being unable to manually manipulate expiration time are the two reasons I'm trying to drop it from the environment
14:41 techdragon My use of formula's has been by pulling the important content into the state tree as a dir, then using the example.pillar as a commented out header for my own pillar files.
14:42 piffio Ahlee: have you tried playing with salt-run fileserver.update?
14:42 piffio it will force a git pull basically
14:42 Ahlee Interesting
14:42 Ahlee Thaks piffio
14:42 Ahlee i will look into that
14:42 piffio actually, I wonder why it's not done implicitely
14:42 piffio you're welcome
14:43 piffio I'm trying to understand why overstate is not working on my configuration and just found out the fileserver runner
14:43 piffio :)
14:47 nu7hatch joined #salt
14:49 jimallman joined #salt
14:53 sgviking joined #salt
14:55 backjlack joined #salt
14:55 techdragon Furao how do you normally setup that returner you pasted into the gist? Im having trouble confirming if its activated when I'm calling it
14:57 mackstick Is there anyway to get the current function name as a variable in a state file?
14:59 elsmorian Hi all
14:59 StDiluted joined #salt
14:59 elsmorian Is there a way to make sure a service is dead, unless a file exsits?
14:59 elsmorian I've tried service.dead:
14:59 elsmorian - name: postgresql
14:59 elsmorian - unless:
14:59 elsmorian - file: postgres-file
15:00 jrb28 joined #salt
15:00 elsmorian But I don't think the service state module accepts "unless"
15:01 mapu joined #salt
15:01 geak_ joined #salt
15:02 Furao joined #salt
15:08 seanz joined #salt
15:13 raghavp80 joined #salt
15:15 forrest joined #salt
15:16 morganfainberg joined #salt
15:18 UtahDave joined #salt
15:20 pipps joined #salt
15:20 blee joined #salt
15:25 higgs001 joined #salt
15:28 redbeard2 joined #salt
15:29 whyzgeek I am trying to see if something is possible. I would like to extract the name of template directory and use it as part of template. Does jinja support that?
15:31 faldridge joined #salt
15:31 geak joined #salt
15:31 jpeach joined #salt
15:33 m_george|away joined #salt
15:37 bastion2202 joined #salt
15:39 bastion2202 Hey everybody ! Is there a way, with salt to : remove a specific key from /home/*/.ssh/authorized_keys
15:39 m_george left #salt
15:39 dthom91 joined #salt
15:39 bastion2202 the prob for me at the  momen is the *
15:40 piffio you might be able to list all those files with a pillar definition
15:40 UtahDave bastion2202: so you want to remove it from all user's authorized_keys files?
15:40 piffio and then loop through that list in your state file
15:41 bastion2202 UtahDave yes, I use salt to manage user. I disable and remove the key from his account but like to double check if the pub key is not present in other users authorized_keys
15:43 whyzgeek piffio: the same problem happens for that. I need to get list of directories. The problem is that this list is on master only. I can call a minon on the master and get it. However these files get procesed before the minion so I have to somehow find a way to get that during loading.
15:45 faldridge joined #salt
15:47 UtahDave bastion2202: You could get users.list_users to get the list of users,  then loop through that list to make sure the keys have been removed
15:49 nu7hatch joined #salt
15:49 lyddonb joined #salt
15:51 pentabular joined #salt
15:53 jbunting joined #salt
15:53 bastion2202 UtahDave: good idea ! I will also check if the file is present and then look inside the file. Tx !!
15:59 lacrymology joined #salt
15:59 felixhummel joined #salt
16:01 kenbolton joined #salt
16:01 Lue_4911 joined #salt
16:04 lacrymology joined #salt
16:07 redondos joined #salt
16:08 Ryan_Lane joined #salt
16:09 ronc joined #salt
16:09 pentabular joined #salt
16:17 [diecast] joined #salt
16:27 jpeach joined #salt
16:28 teebes joined #salt
16:33 giantlock joined #salt
16:33 KyleG joined #salt
16:33 KyleG joined #salt
16:37 morganfainberg joined #salt
16:38 lacrymology is there an _init function for execution modules? I'm setting up some module-level variables, but I need to read them from pillars, and doing __salt__ at the module-level fails
16:40 Ahlee is there a log level deeper than debug? I'm having issues tracking down host matching for _modules distribution, with the top.sls residing a gitfs branch matching the environment: on the minion
16:42 jcockhren Ahlee: when running the salt-minion or salt-master binaries you can use '-l all' instead of '-l debug'
16:43 jcockhren Ahlee: we are tied. ;)
16:43 Ahlee i'll give you a chance to pull ahead ;)
16:44 jcockhren ha
16:44 Ahlee jobs.lookup_jid isn't responding when I request output from a job I see in jobs.active
16:44 Ahlee neermind, it hasn't returned
16:46 jacksontj joined #salt
16:46 jalbretsen joined #salt
16:47 Ahlee yeah, salt-run jobs.kill_job and salt-run jobs.term_job both return funtion jobs.*_job are unavailable
16:50 jacksontj i want to use gitfs for my state tree etc, but i'd like the environments to be cumulative-- meaning if a host is in the DEV environment i'd like it to grab all modules from DEV branch then the rest from base (basically dev as overrides of base)
16:50 jacksontj anyone know of a nice way to do that?
16:50 Ahlee thanks btw, jcockhren
16:50 dthom91 joined #salt
16:50 jacksontj you can do it with regular file_roots by putting the base dir as the second on each environment
16:51 Ahlee jacksontj: My understanding is that's possible based on ordering of your gitfs remotes
16:51 Ahlee oh
16:51 Ahlee i see the issue. nevermind
16:51 jacksontj yea, i'm not sure it supports what i want :/
16:55 jcockhren jacksontj: you mean have various branches overwrite the base branch?
16:55 jacksontj basically
16:55 bastion2202 joined #salt
16:55 jcockhren nah. the states from env are combined I think
16:55 jacksontj i want to be able to comitt a module (for example) which will be applied to a subset of hosts (based on environment)
16:56 jacksontj some way to push to a subset of hosts instead of all at once
16:56 Ahlee I know UtahDave mentioned that the a giant dict is built of all top.sls's, so I assume it's the same
16:57 Ahlee If you come up with a sane solution for that, let me know. I do similar by setting arcane grains on minions to denote they're targetable
16:57 jacksontj well, its a bit weird because the file_roots is modules AND states (sls)
16:57 Ahlee Same
16:57 jacksontj but i think we can make it work-- i'll have to hack at it
16:57 Ahlee although I'm still fighting getting _modules/* to reliably match
16:58 jcockhren jacksontj: can't you set the branch specifically for you gitfs remote
16:58 jcockhren I know for a specific repo (in an env) all branches are merged (hence UtahDave suggestion to have only one top.sls)
16:59 jacksontj nope-- just subdirectories (by the look of it)
16:59 jacksontj oh, if thats the case it might do what i want already
16:59 jacksontj :D
16:59 mr_chris In the top.sls file, is it possible to match on multiple grains? For example, if roles matches "database" AND "production"...
16:59 jacksontj it was just a bit unclear with the docs
16:59 jcockhren jacksontj: yeah. the docs...
16:59 jacksontj i'll hack on it a bit-- and update ;)
16:59 jcockhren cool
16:59 Ahlee mr_chris: compound match
17:00 mr_chris Ahlee, Thanks.
17:00 cro joined #salt
17:00 jacksontj i really want to free up some time to work on refactoring some of this gitfs stuff :/
17:00 jcockhren jacksontj: you should. ;)
17:00 jacksontj well, dropping off-- be back after some hacking :D
17:00 Ahlee wow.  You weren't just whistling dixie with -l all
17:07 teebes joined #salt
17:08 shinylasers joined #salt
17:10 geak joined #salt
17:11 anteaya joined #salt
17:12 saurabhs joined #salt
17:15 berto- joined #salt
17:15 Ahlee oh. is there a way to exclude branches with gitfs?
17:15 jcockhren Ahlee: not that I know of
17:16 Ahlee well then.
17:17 forrest Will including branches like this: http://docs.saltstack.com/topics/tutorials/gitfs.html#branches-environments-and-top-sls-files not work for you?
17:17 forrest or do you just have a ton of branches you need
17:18 dthom91 joined #salt
17:18 Ahlee no, i just noticed in -l all that it was building the top.sls from even the alpha branches
17:19 Ahlee the recommended setup is to have the top.sls file only in your master branch, and use environment-specific branches for states definitions. - so master (base) gets top.sls, the other branches have no top.sls
17:20 Ahlee how does merging work in that? or do you create your dev/qa branches, then check into the master a top.sls, branch off it for any changes to base.sls, while managing in-work states off branches off dev and qa a la gitflow?
17:22 jcockhren Ahlee: yeah. I've been wondering the same
17:22 jcockhren which is why I have yet to implement that
17:22 jcockhren So far I've been doing it with grains
17:22 Ahlee yeah, i pushed against it as it just doesn't seem flushed out
17:22 Ahlee but, i was over ruled
17:23 Ahlee wondering if i'm hitting the bug on http://docs.saltstack.com/topics/tutorials/gitfs.html#branches-environments-and-top-sls-files that speaks to custom modules not syncing and talks of a version that's not available
17:24 ronc joined #salt
17:24 blee_ joined #salt
17:25 SunSparc joined #salt
17:25 nu7hatch joined #salt
17:26 countdigi joined #salt
17:29 vmdsch joined #salt
17:30 djn joined #salt
17:31 vmdsch left #salt
17:33 mgw joined #salt
17:33 krissaxton joined #salt
17:33 seanz joined #salt
17:47 xinkeT joined #salt
17:47 mwillhite joined #salt
17:47 jheise joined #salt
17:48 imaginarysteve joined #salt
17:49 vmdsch1 joined #salt
17:49 pipps joined #salt
17:50 boite joined #salt
17:51 druonysus joined #salt
17:51 druonysus joined #salt
17:53 piffio joined #salt
17:54 maveas What is the status of https://github.com/saltstack/salt/pull/3594 ?
17:54 piffio joined #salt
17:55 forrest well, iptables are in
17:55 forrest http://docs.saltstack.com/ref/states/all/salt.states.iptables.html
17:55 forrest are you referring to asomething specific in that issue?
17:56 forrest *something
17:57 vmdsch1 left #salt
17:57 KennethWilke joined #salt
17:59 KennethWilke howdy
18:00 maveas I were just wondering if every idea is fully implemented / ready for production use? And if there is a function to create a new chain..
18:01 pentabular joined #salt
18:03 jdenning joined #salt
18:03 NotreDev joined #salt
18:04 NotreDev is there a place to see states that other people have written that I can grab from?
18:04 vmdsch2 joined #salt
18:05 maveas NotreDev: http://saltstarters.org
18:06 honestly also https://github.com/saltstack-formulas/
18:09 jhenrypb joined #salt
18:10 robawt hey all
18:11 boite joined #salt
18:11 shel3over left #salt
18:13 bhosmer_ joined #salt
18:13 mesmer joined #salt
18:13 ede joined #salt
18:13 druonysus joined #salt
18:15 robawt what's the relationship between requiring other salt states and the top.sls file?
18:15 robawt If i have one state require another, must they both be identified on the top file?
18:15 SEJeff_work robawt, no
18:16 robawt SEJeff_work: when I require Y in X, it'll tell me the state wasn't found.  Is there more than the require statement needed?  Do I need to include the other state via Jinja templating?
18:17 SEJeff_work robawt, You need to include it via include :)
18:17 SEJeff_work you can't require something and have it magically find it. You have to include it
18:17 SEJeff_work think of include like prepending the contents of another state file to the top of the state file you're including it into
18:18 robawt SEJeff_work: that's a fair point
18:18 SEJeff_work you include state_fileA inside state_fileB, then you (in top.sls) add state_fileB
18:18 robawt I was hoping for magical salt stack powers i suppose, but nothing wrong with include
18:18 SEJeff_work It can't read your minds and magically guess were the IDs come from
18:18 jimallman joined #salt
18:18 SEJeff_work salt is simple, not magical
18:19 robawt I disagree SEJeff_work
18:19 robawt i think it is both magical and simple
18:19 SEJeff_work :)
18:19 SEJeff_work Having looked extensively at the code, the magic is gone for me
18:19 SEJeff_work but it works well heh
18:19 SEJeff_work It is still quite the awesome
18:20 SEJeff_work But I equate magic with doing things behind the scenes I neither want or expect
18:20 SEJeff_work so perhaps our definitions are different
18:21 pentabular joined #salt
18:22 jimallman joined #salt
18:26 cdsrv joined #salt
18:29 robawt agreed SEJeff_work
18:32 pdayton joined #salt
18:33 abe_music joined #salt
18:34 lacrymol1gy joined #salt
18:46 Ahlee anybody got an example of using cmd inside a state?
18:47 KyleG Ahlee: Sure 1 sec
18:48 KyleG Ahlee: I use this to check if splunk is installed, and if it's not then install it: https://gist.github.com/kylegato/e14963e7a840b7314e51
18:48 KyleG To keep my servers time up to date: https://gist.github.com/kylegato/87c5909116a6b66373d3
18:48 Ahlee gotchaso
18:51 Ahlee awesome, thanks KyleG
18:52 mapu joined #salt
18:52 forrest KyleG, any particular reasoning behind simply dropping the command in as opposed to using the name option?
18:53 colinbits joined #salt
18:59 KyleG forrest: No specific reason
19:00 zooz joined #salt
19:05 whyzgeek is there anyway that I can use include function in jinja templates? when I do it in sls templates, it fails.
19:07 tracphil joined #salt
19:12 EugeneKay joined #salt
19:14 pipps joined #salt
19:18 dthom91 joined #salt
19:19 whyzgeek I am specifically looking for have a variable get set in one of the sls templates and then reuse it in others, this is to avoid copy pasting the same code everywhere
19:19 whyzgeek sls jinja
19:34 wilywonka joined #salt
19:38 ccase joined #salt
19:39 danielbachhuber joined #salt
19:45 micah_chatt joined #salt
19:51 dthom91 joined #salt
20:01 Jarus joined #salt
20:01 c0bra joined #salt
20:03 UtahDave whyzgeek: I'm not sure on that. You might ask on the mailing list.
20:04 c0bra Quick newbie question: Would salt-stack be useful for managing semi-connected devices? (i.e. laptops that move around on trucks and have occasional vpn connectivity)
20:06 UtahDave c0bra: yep!
20:07 UtahDave when the devices are disconnected you obviously won't be able to run any commands from the salt master, but you can when they connect again.
20:07 c0bra Okay, great.
20:08 c0bra UtahDave: is there any way to handle devices that for whatever reason CAN'T get network activity?  I'm learning that we have to sometimes mail out CDs with batch files that perform updates. It would be great if something like salt could do some sort of static packaging, if that's the right term
20:09 UtahDave c0bra: you can run salt in standalone mode.
20:10 c0bra hmm
20:10 lacrymology joined #salt
20:11 micah_chatt would anyone be interested in a docker (http://docker.io) salt module?
20:11 c0bra OK, so from reading the docs you could just call state.highstate --local and it would check the local file_roots?
20:11 saurabhs joined #salt
20:12 UtahDave micah_chatt: yes, we're VERY interested in docker support
20:12 micah_chatt ok, I've been working on a module for it
20:12 UtahDave c0bra: yeah, you can also set file_roots to local and then it will always assume --local
20:12 micah_chatt @UtahDave
20:12 micah_chatt I have some state information written, but would
20:12 micah_chatt like to have more control
20:12 micah_chatt with remote execution
20:13 UtahDave micah_chatt: that's awesome!
20:13 micah_chatt UtahDave: like 'G@apiserver' docker.pull <container_id>
20:13 c0bra UtahDave: is there a way to "fail over" to running locally?
20:13 micah_chatt salt 'G@apiserver' docker.run <container_id>/<tag>
20:14 rmt joined #salt
20:14 UtahDave c0bra: Hm. I don't think we have that functionality yet.
20:14 c0bra ok
20:14 UtahDave micah_chatt: sweet!
20:14 jimallman joined #salt
20:15 c0bra well, I would think I could have a utility check for connectivity and run state.highstate if it has a connection, state.highstate --local if not
20:15 c0bra perhaps?
20:15 pipps joined #salt
20:15 c0bra Or, hell, have autorun.inf execute it with --local
20:15 UtahDave c0bra: sure
20:16 forrest c0bra, it would be kind of hacky, but you could always add to the init script to support something along those lines if you wanted it to check on startup
20:16 imaginarysteve joined #salt
20:16 forrest as long as the priority was after networking obviously
20:17 c0bra yea
20:17 c0bra well that's great; that at least is a viable option worth checkin gout
20:17 forrest yea the masterless minion works really well cobra
20:18 forrest *c0bra
20:18 c0bra ok
20:21 jrb28 joined #salt
20:23 Brew1 joined #salt
20:23 c0bra oh, you know what. I think there's are XP machines
20:23 c0bra it looks like the install procedure is quite a bit more hefty
20:23 forrest saltstack supports windows... sort of?
20:23 forrest I haven't messed with that at all
20:24 c0bra it's worked great for my testing on server 2008
20:24 c0bra x64
20:24 forrest cool
20:24 forrest I don't know what the development path is looking like for windows, or what features are supported
20:24 forrest that's brutal you have to deal with XP though
20:26 shinylasers joined #salt
20:26 UtahDave we have installers for Windows.
20:26 forrest Yea he's looking at those I believe
20:27 mwillhite joined #salt
20:28 abe_music joined #salt
20:29 Ryan_Lane joined #salt
20:30 forrest Hey abe_music
20:30 abe_music hey forrest
20:31 saurabhs joined #salt
20:33 mwillhite joined #salt
20:36 c0bra yes
20:36 c0bra the 32-bit XP instructions have 24 steps
20:36 c0bra which I'm assuming you could script... but... yea
20:37 mwillhite joined #salt
20:37 berto- joined #salt
20:45 Brew joined #salt
20:47 mwillhite joined #salt
20:47 tyler-baker joined #salt
20:49 pipps joined #salt
20:51 geak joined #salt
20:55 whyzgeek forrest: do you know how to include another jinja template from current directory?
20:56 shinylasers joined #salt
20:56 forrest between states?
20:56 whyzgeek forrest: yes
20:57 whyzgeek basically I have a peace of code which I don't want to repeat everywhere
20:57 whyzgeek I refactored it out but tried using include in jija seems not to be able to find the other templates
20:57 whyzgeek i used {% include="blah.sls" %}
20:57 forrest Hmm, I don't know, because there's no assurance that unless you use includes that the order will be correct, so state 2, might run before state 1 if you don't have requires and such.
20:58 forrest Hmm
20:58 lacrymol1gy joined #salt
20:58 forrest can you try to maybe just include sls1 in sls2?
20:58 forrest since that ports the actual yaml and such over so you can do requires, maybe it would pull the jinja data in as well?
20:58 forrest I don't know, I haven't tried that before...
20:58 forrest now I'm curious about it thoug
20:58 whyzgeek forrest: I did that using yaml include but it seems the actual jinja variables are not recognizable
20:58 forrest Hmm ok
20:59 forrest is there any way that your logic in the state, could instead be stored in pillar?
20:59 forrest so you could simply reference it via a lookup there?
20:59 whyzgeek forrest: it can but makes it complex
20:59 forrest or are you doing legit python code
20:59 forrest I understand
20:59 forrest try this
20:59 whyzgeek so if there is anyway I can share some peace of code between templates
20:59 whyzgeek that would be awesome
20:59 forrest {% from 'sls1' import jinjavar %}
21:00 whyzgeek forrest: I tried that already
21:00 forrest in that exact format?
21:00 whyzgeek it says sls1 not found
21:00 forrest are they in the same directory?
21:00 whyzgeek yes
21:00 forrest are they in a subdirectory?
21:00 whyzgeek yes
21:00 pentabular joined #salt
21:01 whyzgeek so when I use static variable
21:01 whyzgeek everything works
21:01 whyzgeek however as soon as I refactor
21:01 forrest how about this {% from 'path/file.sls' import jinjavar %}
21:01 whyzgeek it out it stops working
21:01 forrest I see
21:01 whyzgeek let me try this
21:02 forrest so you could import a static variable using that syntax (such as a set), but when you have logic within the state, and you try to import that variable, it fails
21:02 whyzgeek forrest: correct
21:02 forrest did you try to do it with context?
21:02 whyzgeek basically it fails to find the other adjacent template
21:03 whyzgeek I tried with both
21:03 whyzgeek with and without
21:03 forrest interesting, I'd think that using 'with context' would run through the logic..
21:03 whyzgeek however I didn't try the path/file.sls yet
21:03 forrest ok
21:03 whyzgeek let me double check
21:04 g4rlic trying to write a simple custom static grain.  On the minions, it's installed in /etc/salt/grains.  How do I get the master to recognize that this grain exists?  salt '*' grains.ls shows it on the correct node. I'm clearly doing something wrong, but the documentation is slightly confusing on what to do *after* defining a custom grain on a minion.
21:05 forrest did you already sync the grain g4rlic?
21:06 UtahDave KennethWilke: do you have a good example of using includes in jinja?
21:06 KennethWilke UtahDave!
21:06 KennethWilke i do
21:06 whyzgeek forest: the path trick works!
21:06 whyzgeek :)
21:06 forrest great
21:06 whyzgeek thanks man
21:06 forrest yea np
21:06 whyzgeek but I have to give it from rooty
21:06 forrest all the way down the whole dir tree??
21:06 KennethWilke UtahDave: https://github.com/saltstack-formulas/salt-formula/blob/master/salt/master.sls
21:06 whyzgeek which is a little bit annoying
21:06 c0bra so...
21:07 c0bra where would local /srv/salt be on a windows machine?
21:07 whyzgeek forrest: yes
21:07 forrest that seems odd whyzgeek.
21:07 UtahDave c0bra: are you using the Salt Windows installer here: http://docs.saltstack.com/topics/installation/windows.html  ?
21:07 UtahDave thanks, KennethWilke!
21:07 whyzgeek forrest: this would be unmaintable as I might move directories and rename them
21:07 c0bra UtahDave: yea, on this 2008 server I used the installer
21:07 whyzgeek and evertime I have to change this
21:07 KennethWilke no problem!
21:07 forrest agreed, you should at MAX have to reference it from the main salt dir
21:08 c0bra UtahDave:just trying to figure out where I should put a local top.sls to give it a test run
21:08 forrest can you try salt/path/to/state.sls whyzgeek?
21:08 whyzgeek as far as I know jinja should work based on relative path
21:08 lineman60 joined #salt
21:08 UtahDave c0bra: ah, ok.   c:\salt\srv   I think.
21:09 whyzgeek forrest: sorry I guess I confused you, I did it down from /srv/salt (not including)
21:09 forrest oh I see.
21:09 forrest so you have salt/path/to/state.sls?
21:09 whyzgeek yes
21:09 whyzgeek and this works
21:09 UtahDave g4rlic: you should have to do anything after setting your custom grain.
21:09 forrest ok yes that should work
21:09 g4rlic forrest: I do not believe so. The docs imply that sync will happen during high state. But when I highstate the master, it borks on that grain.
21:09 Brew joined #salt
21:09 forrest ok I wasn't sure if you had run highstate yet from what you put in earlier
21:10 UtahDave g4rlic: can you pastebin the output when your highstate fails?
21:10 forrest whyzgeek, I think you'll have to do that, because you're applying the state, and it starts to traverse from the main salt dir
21:10 forrest what about it would make it unmaintainable? The logic you're importing is always the same right?
21:11 g4rlic UtahDave: http://pastebin.centos.org/4561/
21:11 g4rlic that's what I see.
21:12 Lacrymology left #salt
21:12 shinylasers joined #salt
21:13 g4rlic I *am* running 0.16.0, since 0.16.4 isn't packaged for CentOS afaict
21:13 whyzgeek forrest: yes however, this a setup for application and part of it is application name which matches the diectory it resides. I made it as such to keep it organized. Now I have to change the whole thing. Because if I know the directory I already know the application name which makes the whole refactoring useless. Also I tend to copy folders and use them as template to setup other apps.
21:14 forrest woah woah woah
21:14 UtahDave g4rlic: your grain seems to be out of whack.    roles: testrole
21:14 forrest whyzgeek
21:14 forrest why don't you set a pillar value, that you then use as part of your import path?
21:14 whyzgeek forrest: if I can somehow tell it to use current dicetory my problem will be solved
21:15 g4rlic UtahDave: oh?
21:15 g4rlic Did I do something stupid like override a built-in grain?  Or did I format it wrong?
21:15 whyzgeek forrest: ya I have to use pillar now
21:15 UtahDave whyzgeek: have you tried  {%  include="./blah.sls" %}
21:15 forrest I mean just use pillar for the path variable
21:15 whyzgeek UtahDave: ya doesn't work
21:15 forrest that would be funny if the ./ fixed it :P
21:15 forrest oh you tried it already like that?
21:15 whyzgeek only full path from /srv/salt works
21:16 UtahDave g4rlic: I think your grain is formatted wrong.  It needs to be valid yaml
21:16 forrest what I'm saying whyzgeek, is to set a value based on the app with jinja, and then use THAT value when you reference the application and where you want to import it from
21:16 UtahDave whyzgeek: can you paste the exact import line that worked?
21:17 whyzgeek forrest: I didn't know you can do that, how you can set a value pillar from jinja?!
21:17 whyzgeek forrest: that would solve my problem if I can do it
21:17 UtahDave whyzgeek: KennethWilke's example here doesn't use the full path: https://github.com/saltstack-formulas/salt-formula/blob/master/salt/master.sls
21:17 g4rlic UtahDave: So, it could be incorrectly formatted, even if salt '*' grains.ls shows the grain correctly?
21:17 forrest no you can't as far as I'm aware (or at least I've never tried it)
21:17 UtahDave whyzgeek: {{ salt['pillar.get']('mypillarval', 'defaultval') }}
21:18 forrest he means set the pillar value from a state UtahDave
21:18 whyzgeek UtahDave: {% from 'pyapps/<app_name>/app_name.sls' import app_name with context %}
21:18 forrest right, so use jinja logic to determine <app_name>
21:20 whyzgeek UtahDave: I think that's because KennethWilke use it at root level
21:20 forrest whyzgeek, take a look at my fail2ban formula
21:20 forrest https://github.com/gravyboat/fail2ban-formula/tree/master/fail2ban
21:20 c0bra UtahDave: I've tried putting a top.sls file in c:\salt, c:\salt\srv c:\salt\srv\salt c:\salt\srv\pillar and c:\salt\srv\salt\pillar, nothing seems to get picked up when running with the --local flag
21:20 forrest Here I've got my default values set, but based on data I lookup, I change the file value (for debian versus redhat systems)
21:21 diegows does saltstack saves for information about the minion in the master?
21:21 diegows because I'm trying something, revert a snapshot and is not copying some files
21:21 whyzgeek forrest: I think this also assumes that fail2ban is under root
21:21 whyzgeek ie if you put this code in root/blah/fail2ban
21:21 forrest yes this does, but the only thing that matters from here is where I'm setting my values
21:22 whyzgeek it shouldn't work
21:22 forrest Yea I haven't tested that. But what I'm suggesting is to use jinja logic
21:22 forrest https://github.com/gravyboat/fail2ban-formula/blob/master/fail2ban/jail.sls
21:22 forrest see here where I have that data being based off a grain?
21:22 faldridge joined #salt
21:22 forrest what if you did something similar (based off some sort of value), and then populated the field for the include
21:23 whyzgeek forrest: this is getting the value from grain not set it
21:23 lyddonb joined #salt
21:23 morganfainberg joined #salt
21:23 forrest I know
21:23 g4rlic UtahDave: so I'm not entirely the best @ yaml, is there any tool I can use to ensure that I'm formatting it correctly?  A lint checker, perhaps?
21:23 Kyle joined #salt
21:23 forrest but there's nothing to say you couldn't use jinja logic to perform local look ups right?
21:23 forrest the jinja logic can be just plain python
21:24 whyzgeek forrest: ya this bit I know I already use it extensively with pillar and grains
21:24 wilywonka joined #salt
21:24 whyzgeek forrest: no it won't
21:24 whyzgeek forrest: the python is limited in jinja
21:24 whyzgeek I tried already that
21:24 forrest oh ok
21:24 forrest I know it's limited, but wasn't sure if you had tried that already, that's a bummer
21:24 whyzgeek for example I can't figure out the current absolute path in jinja
21:25 whyzgeek there is no way to find that
21:25 whyzgeek otherwise I can grab that and use it
21:25 g4rlic this is from top.sls, and what's causing the traceback: http://pastebin.centos.org/4566/
21:26 whyzgeek forrest: I think the only way as you suggested is pillar
21:26 whyzgeek which is a pitty, this design I did, is very powerful
21:26 g4rlic http://pastebin.centos.org/4571/  <-- this is the contents of /etc/salt/grains on the node.
21:26 forrest yea I understand whyzgeek, I'm sure there's a way to do it, I'm just not smart enough to figure it out :D
21:26 KennethWilke whyzgeek: when jinja overcomplicates what i'm doing i use mako
21:27 whyzgeek KennethWilke: good point, I think mako is more flexible on this. Any docs on how to activate that?
21:28 Corey Feh. g4rlic is here.
21:28 KennethWilke whyzgeek: sorry i've missed most of this conversation, but i think all of python is available in mako
21:28 * g4rlic blows Corey a kiss
21:29 whyzgeek KennethWilke: you are right I know mako is more flexible
21:29 forrest whyzgeek, just change the shebang to #!mako|yaml
21:29 KennethWilke whyzgeek: and i'm not sure i understand what you mean by activate, do you mean to use it instead of jinja?
21:29 forrest or #!yaml|mako
21:29 KennethWilke i think just #!mako would do it too
21:29 whyzgeek KennethWilke: yes forrest mentioned it
21:29 whyzgeek thanks
21:29 KennethWilke no problem
21:29 whyzgeek let me try this then
21:29 forrest KennethWilke, trying to break the state with just mako :P
21:30 whyzgeek :)
21:30 KennethWilke i break things all the time
21:30 KennethWilke but mostly when i touch things
21:30 KennethWilke or write code
21:30 forrest well, as long as you fix it, and acknowledge that you broke it, who cares
21:30 KennethWilke forrest: here here!
21:30 KennethWilke some knowledges are gained along the way
21:31 forrest yepo
21:31 forrest *yep
21:32 UtahDave g4rlic: oh, you can't put state commands in your top.sls
21:35 BrendanGilmore joined #salt
21:36 c0bra argh
21:37 g4rlic UtahDave: oh derp, did not know this.  Will correct asap.
21:39 g4rlic I'll be damned, grain parsing error was a red herring. >.<
21:39 cnelsonsic joined #salt
21:39 forrest lol
21:39 c0bra "No Top file or external nodes data matches found" no matter where I put files :\
21:40 cnelsonsic left #salt
21:49 ggoZ joined #salt
21:51 pjs joined #salt
21:53 bhosmer_ joined #salt
21:55 whyzgeek KennethWilke: mako works like a charm! thank you! However, it seems that old mako for Centos doesn't work with salt. I grabed an rpm from somewhere and it seems to be working with that!
21:56 KennethWilke whyzgeek: awesome! i'm glad that worked out for you :)
21:58 c0bra I give up; till tomorrow at least
21:59 whyzgeek also the #!mako is beautiful I like it! :)
22:00 forrest the mako usage of <% %> reminds me of puppet templates, gives me bad flashbacks
22:00 whyzgeek forrest: lol
22:01 nu7hatch joined #salt
22:03 g4rlic boo, iptables.append() no worky in 0.16.0.  I really need to get an updated package built for 0.16.4 for our CentOS environment.
22:04 ahammond hmm, UtahDave, is 16.4 out? If so, you might want to update the MOTD. :)
22:04 forrest ahammond, let's pretend 0.17 is out :P
22:04 forrest it has a release candidate
22:06 ahammond forrest, just nagging about the channel's MOTD
22:06 Ahlee So anytime I put a watch: on a service, salt hangs until that process exits
22:06 Ahlee the salt run calling that state
22:09 forrest ahammond, I know
22:09 forrest your service watches something else Ahlee?
22:09 forrest can you show us your sls?
22:15 Ahlee http://pastebin.com/1eQYEW0i
22:15 Ahlee Exhibitor will continue running if that watch: is there
22:17 Ahlee without the watch, the process won't start.  With the watch, the salt run hangs until the process exits
22:18 alunduil joined #salt
22:19 jacksontj joined #salt
22:20 oz_akan_ joined #salt
22:22 Ahlee I believe this is going to be related to how Exibitor launches, in that it calls nohup *shudder* java -jar
22:22 Ahlee i'll see if the dev behind this can change to sh -c like a sane person
22:22 UtahDave ah, yeah.  thanks, ahammond
22:22 Topic for #salt is now Welcome to #salt - http://saltstack.org | 0.16.4 is the latest | Please be patient when asking questions as we are volunteers and may not have immediate answers - Channel logs are available at http://irclog.perlgeek.de/salt/
22:24 cro joined #salt
22:34 Ryan_Lane joined #salt
22:36 nu7hatch joined #salt
22:43 berto- joined #salt
22:43 redondos joined #salt
22:45 nu7hatch joined #salt
22:46 kermit joined #salt
22:47 xuru joined #salt
22:48 RebStilly joined #salt
22:49 xuru hey everyone…  I'm having trouble getting gitfs to work.  I keep getting "Unknown include: Specified SLS base: sudoers is not available on the salt master in environment(s): base".  I'm going off of this walk through:  https://docs.saltstack.com/topics/tutorials/gitfs.html
22:49 krissaxton left #salt
22:51 Jahkeup joined #salt
22:53 UtahDave xuru: can you pastebin your top.sls
22:54 xuru sure
22:56 xuru UtahDave: not a whole lot too it: http://pastebin.com/f1TXWG4r
22:57 xuru in ncl I do a include:  - sudoers
22:58 UtahDave what's the path to sudoers.sls?
22:59 xuru I guess that's where I'm confused.  I have my master file setup like https://docs.saltstack.com/topics/tutorials/gitfs.html#simple-configuration
22:59 xuru so it should pull it down from git correct?
22:59 UtahDave is sudoers.sls in your master branch in git?
22:59 xuru git
23:00 Brew joined #salt
23:01 xuru UtahDave: I have this set in my salt/master
23:01 xuru http://pastebin.com/qh9sgNEj
23:01 UtahDave xuru: is sudoers.sls in the root directory of your master branch?
23:03 xuru hmm…   now that I look at it, I don't see it there:  https://github.com/saltstack/salt-states
23:04 xuru So I think that's my problem
23:04 xuru I have the wrong link
23:04 UtahDave ah, ok.
23:05 xuru So, if I wanted to use any of these:  https://github.com/saltstack-formulas/
23:05 geak joined #salt
23:05 xuru I have to add each of them to my gitfs_remotes?
23:06 xuru which I guess makes sense…  the documentation is a bit misleading in that regard
23:07 forrest xuru, in what way do you feel it is misleading?
23:08 xuru https://docs.saltstack.com/topics/tutorials/gitfs.html#simple-configuration
23:08 UtahDave xuru: yep!
23:08 xuru That part uses git://github.com/saltstack/salt-states.git as the remote
23:09 xuru so I assumed it had all the formulas
23:09 forrest ahh I understand
23:09 xuru UtahDave: thanks for the help
23:09 geak_ joined #salt
23:09 forrest I'll email myself to put in a pull request to update that when I get home, thanks for pointing it out
23:09 xuru sweet, np
23:09 xuru thank you
23:09 xuru Loving salt btw
23:11 geak__ joined #salt
23:13 forrest xuru, yea I do too
23:14 xuru hmm…  putting in https://github.com/saltstack-formulas/sudoers-formula.git for the gitfs_remotes doesn't work either
23:14 pentabular joined #salt
23:14 xuru I'm assuming because there is no top.sls in that repo?
23:15 forrest it shouldn't require a top.sls
23:19 xuru ah, an error was masking the real issue…  Think I'll look at this later when I'm fresh
23:19 xuru thanks again for the help guys
23:19 xuru left #salt
23:20 forrest it's too bad salt-states couldn't be automatically populated based on saltstack-formulas
23:20 forrest that would be pretty slick
23:20 abe_music forrest: i should be able to put a bunch of *.conf files in cloud.providers.d and cloud.profiles.d in place of two gigantic confg files, right?
23:20 pentabular left #salt
23:20 forrest I don't know abe_music, I'm not familiar enough with salt-cloud yet :\
23:20 UtahDave abe_music: yep
23:21 forrest so they operate like any other .d dir then UtahDave?
23:21 forrest or do the giant configs include include cloud.profiles.d/* abe_music?
23:21 abe_music forrest, UtahDave: thought so…any particular format they need to be done? my salt-cloud is not picking them up
23:22 abe_music forrest: i think the *.d directories can be used in place of the two individual cloud.providers and cloud.profiles files
23:22 forrest gotcha
23:22 abe_music for my purposes the directory approach will be great
23:22 forrest yea I wasn't sure if it's just like apache where the main confi includes content within the .d directories or what
23:22 forrest *conf
23:24 forrest I'll be back on later, have a good one.
23:24 abe_music forrest: would make sense…i seem to recall a warning/error if you tried to use both though
23:24 UtahDave abe_music: I'm not 100% sure. I've only used the monolithic files
23:24 forrest Ahh ok abe_music
23:24 forrest let me know what you figure out
23:24 abe_music UtahDave: i'm digging, trying to find where it attempts to load the providers up
23:24 pentabular1 joined #salt
23:30 alekibango joined #salt
23:33 abe_music UtahDave: is there an environment variable to cause salt to look in other places than /etc/salt?
23:34 abe_music i believe what's happening is i'm setting things up in a virtualenv and not using the global /etc/default (shared system)
23:34 abe_music so the cloud.providers.d/*.conf glob is looking relative to /etc/salt
23:34 abe_music *i think* :)
23:34 Ryan_Lane joined #salt
23:35 UtahDave that makes sense.  I'm not sure.
23:35 UtahDave you can pass in -C at the cli to specify that config file to use.
23:36 micah_chatt joined #salt
23:39 abe_music UtahDave: yeah that can be overridden with SALT_CLOUD_CONFIG
23:39 abe_music i'm wondering if maybe it needs to take the dirname of that path if it exists and use it
23:43 redbeard2 joined #salt
23:44 kenbolton joined #salt
23:45 __number5__ abe_music: AFAIK, salt-cloud will look every thing from /etc/salt/* unless you overrided it with command-line options
23:46 abe_music __number5__: i know environment variables have been added, but i don't think they've been fully integrated throughout right now
23:46 abe_music i'm overriding the salt master and cloud config files at the moment using env vars
23:47 svx joined #salt
23:47 Micromus_ joined #salt
23:47 kvbik joined #salt
23:47 pt|Zool joined #salt
23:47 scooby2 joined #salt
23:47 scooby2 joined #salt
23:47 Vivek joined #salt
23:48 rachbelaid joined #salt
23:48 kamal_ joined #salt
23:48 s0undt3ch joined #salt
23:49 Koma joined #salt
23:49 fxdgear joined #salt
23:54 MTecknology joined #salt
23:54 [vaelen] joined #salt
23:54 [ilin] joined #salt
23:54 godog joined #salt
23:54 crashmag joined #salt
23:54 errr joined #salt
23:54 SoR joined #salt
23:54 gadams joined #salt
23:54 pmrowla joined #salt
23:54 chutzpah joined #salt
23:54 g4rlic joined #salt
23:54 honestly joined #salt
23:54 sssslang joined #salt
23:54 mordred joined #salt
23:54 jslatts joined #salt
23:54 bawnzai joined #salt
23:54 jbub joined #salt
23:54 kiorky joined #salt
23:54 tseNkiN joined #salt
23:54 jeff__ joined #salt
23:54 vipul joined #salt
23:54 tyler-baker joined #salt
23:54 Jahkeup joined #salt
23:54 oc joined #salt
23:54 quantumsummers|c joined #salt
23:54 Ryan_Lane joined #salt
23:54 Ryan_Lane1 joined #salt
23:54 NV joined #salt
23:54 kevinbrolly joined #salt
23:54 letterj joined #salt
23:54 packeteer joined #salt
23:54 twinshadow joined #salt
23:54 nocturn joined #salt
23:54 setient joined #salt
23:54 puppet joined #salt
23:54 indymike joined #salt
23:54 carxwol joined #salt
23:54 iMil joined #salt
23:54 kermit joined #salt
23:54 alunduil joined #salt
23:54 saurabhs joined #salt
23:54 rmt joined #salt
23:54 vmdsch2 joined #salt
23:54 anteaya joined #salt
23:54 Lue_4911 joined #salt
23:54 raghavp80 joined #salt
23:54 Furao joined #salt
23:54 mmilano joined #salt
23:54 \ask joined #salt
23:54 bhosmer joined #salt
23:54 akoumjian joined #salt
23:54 bejer joined #salt
23:54 heckman joined #salt
23:54 jesusaurus joined #salt
23:54 LordOfLA|Weechat joined #salt
23:54 benturner joined #salt
23:54 Daviey joined #salt
23:54 sixninetynine joined #salt
23:54 keekz joined #salt
23:54 zach joined #salt
23:54 Greboid joined #salt
23:54 gamingrobot joined #salt
23:54 zakm joined #salt
23:54 SEJeff_work joined #salt
23:54 _mzito___ joined #salt
23:54 cloq joined #salt
23:54 lahwran joined #salt
23:54 mr_chris joined #salt
23:54 andrew_seattle joined #salt
23:54 keee joined #salt
23:54 utahcon joined #salt
23:54 Odd_Bloke joined #salt
23:54 mattikus` joined #salt
23:54 clone1018 joined #salt
23:54 jcockhren joined #salt
23:54 aurigus joined #salt
23:54 dcmorton joined #salt
23:54 ggherdov joined #salt
23:54 ahammond joined #salt
23:54 scalability-junk joined #salt
23:54 Ixan joined #salt
23:54 17SADD470 joined #salt
23:54 eliasp joined #salt
23:54 tedski joined #salt
23:54 akio joined #salt
23:54 ktenney joined #salt
23:54 renoirb joined #salt
23:54 jbirdman joined #salt
23:54 lorimar joined #salt
23:54 dave_den joined #salt
23:54 basepi joined #salt
23:54 Zurullo joined #salt
23:54 EntropyWorks joined #salt
23:54 goodwill joined #salt
23:54 coolj joined #salt
23:54 talso joined #salt
23:54 holmb joined #salt
23:54 vbabiy joined #salt
23:54 Heartsbane joined #salt
23:54 nocturn joined #salt
23:54 Ryan_Lane joined #salt
23:56 NotreDev joined #salt
23:57 sald19 joined #salt
23:58 berto- joined #salt

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