Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-03-05

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

All times shown according to UTC.

Time Nick Message
00:02 NV https://github.com/saltstack/salt/issues/10956 gitfs jigger
00:03 NV seems to only error out when accessing the top file
00:04 helderco joined #salt
00:06 dansalmo I have installed salt-minion and set the master IP address in the minion file.
00:07 dansalmo when I run salt-minion it says [ERROR ] The salt master has cached the public key for this node
00:07 dansalmo then it re-tries
00:08 faldridge joined #salt
00:09 forrest dansalmo, because you haven't accepted the key on the salt master
00:09 smcquay joined #salt
00:09 forrest The walkthrough talks about accepting keys
00:11 dansalmo Just read that and did it on the next step :)
00:11 dansalmo Thanks
00:13 dansalmo forrest:  I was able to ping the minion and got a True response
00:13 dansalmo Thanks for you help
00:13 ckao joined #salt
00:14 dansalmo I did get a deprication warning on the minion though
00:15 fauxtoniche joined #salt
00:15 dansalmo related to master pub message signing
00:17 dansalmo I think I was able to fix it
00:22 NV terminalmage: re gitfs bug - seems to be in tag handling - figured out how to work around the issue and in the process worked out how to reproduce it too - have updated the github issue :)
00:22 terminalmage NV: woot
00:23 terminalmage thanks
00:23 NV basically include epel-formula as a gitfs remote and you should trigger the issue :P
00:26 ajprog_laptop1 joined #salt
00:26 kintel joined #salt
00:27 pydanny joined #salt
00:29 elfixit joined #salt
00:29 dn joined #salt
00:34 dansalmo how do I stop salt-master completely?
00:34 jared_ joined #salt
00:36 brian left #salt
00:39 schimmy1 left #salt
00:41 NV dansalmo: service salt-master stop?
00:42 dyim joined #salt
00:45 taion809 joined #salt
00:51 saurabhs left #salt
00:54 Ryan_Lane joined #salt
00:56 hunter http://pastebin.com/ZLBnf6h9
00:56 hunter I see the problem - its obvious now. but I've been at this all day - how SHOULD that be structured? Whats the best salt idiom to follow here?
00:58 bhosmer joined #salt
00:58 dyim joined #salt
01:02 hunter joined #salt
01:04 wedgie joined #salt
01:04 ndrei joined #salt
01:06 fllr joined #salt
01:12 zain_ joined #salt
01:17 faldridge joined #salt
01:17 jslatts joined #salt
01:20 malinoff joined #salt
01:21 lzhang joined #salt
01:25 fllr joined #salt
01:25 pydanny joined #salt
01:26 dyim joined #salt
01:31 jeffro joined #salt
01:32 joehh ubuntu packages for newer than precise building on launchpad now
01:37 yomilk_ joined #salt
01:39 krow joined #salt
01:44 dyim joined #salt
01:48 bhosmer joined #salt
01:55 Linz joined #salt
01:57 pdayton joined #salt
01:58 quickdry21 joined #salt
01:59 n8n_ joined #salt
02:00 AdamSewell joined #salt
02:11 krow joined #salt
02:12 jeremyfelt joined #salt
02:12 conan_the_destro joined #salt
02:13 rojem joined #salt
02:13 baniir joined #salt
02:14 fllr joined #salt
02:15 baniir is it possible to use a file as input to file.blockreplace
02:16 zain_ joined #salt
02:17 krow joined #salt
02:19 timoguin joined #salt
02:22 krow joined #salt
02:28 pythopticon joined #salt
02:28 krow joined #salt
02:31 pythopticon joined #salt
02:32 redondos_ joined #salt
02:33 krow joined #salt
02:34 zain_ joined #salt
02:36 krow1 joined #salt
02:38 maxleonca joined #salt
02:38 joehoyle joined #salt
02:38 maxleonca Hello, anyone that can lend me some experience with salt-cloud and AWS?
02:38 krow joined #salt
02:39 maxleonca I have it almost ready but for some reaso the AMI always gets created with just 1GB on the root volume
02:42 akoumjian joined #salt
02:45 krow1 joined #salt
02:46 monokrome joined #salt
02:47 n8n joined #salt
02:49 GradysGhost joined #salt
02:51 krow joined #salt
02:54 zain_ joined #salt
02:55 shadowsun basepi: you around?
02:56 krow joined #salt
02:57 xl1 joined #salt
02:58 maxleonca Any salt-cloud and ec2 savy people arround?
02:59 l0x3py joined #salt
03:00 krow1 joined #salt
03:01 akoumjian joined #salt
03:04 justlooks_ joined #salt
03:04 redondos joined #salt
03:04 redondos joined #salt
03:06 malinoff joined #salt
03:06 xmltok joined #salt
03:13 raizyr joined #salt
03:14 schimmy joined #salt
03:15 justlooks_ is salt api has GA version?
03:16 malinoff justlooks_, what is GA?
03:17 justlooks_ malinoff:  general available
03:18 ajw0100 joined #salt
03:18 malinoff justlooks_, is it what you're asking for? "Generally Available (GA), also know as Production releases, are stable, having successfully passed through all earlier release stages and believed to be reliable, free of serious bugs, and suitable for use in production systems"
03:19 faldridge joined #salt
03:19 malinoff If yes, no, salt has no GA version
03:19 justlooks_ malinoff:  salt api in epel  is 0.8.2 ,it has some bug, i can not start it
03:20 justlooks_ malinoff: sorry i say salt-api project not salt
03:21 dyim joined #salt
03:23 rostam joined #salt
03:24 malinoff justlooks_, whatever :)
03:26 andrej Hmmm ... the new release 2014.1 has a non-critical flaw that irks me :)
03:26 andrej the startup scripts leave a zombie dangling - this didn't use to happen
03:26 andrej Hmmm ... the new release 2014.1 has a non-critical flaw that irks me :)
03:26 andrej the startup scripts leave a zombie dangling - this didn't use to happen
03:28 GradysGhost joined #salt
03:29 HeadAIX joined #salt
03:33 maxleonca hi again, anyone with experience with salt-cloud and ec2?
03:38 jeffro joined #salt
03:39 joehh andrej: which os?
03:41 jsm joined #salt
03:41 jsm left #salt
03:42 justlooks_ when i restart salt master it get error Unable to bind socket, error: [Errno 98] Address already in use
03:42 tr_h joined #salt
03:44 CeBe1 joined #salt
03:47 Linz joined #salt
03:48 justlooks_ here is the detail anyone know what's happened?  https://github.com/saltstack/salt/issues/10957
03:49 bdf is something else listening on 4505/4506?
03:50 bdf maybe you could include the OS and such you're running?
03:50 jmccree joined #salt
03:50 bdf but netstat -ln / lsof can probably tell you what.
03:51 joehh probably a salt-master that is hanging around
03:51 joehh kill them all off and you will be able to restart (start) a new one
03:52 twobitsprite joined #salt
03:52 [diecast] joined #salt
03:52 fllr joined #salt
03:52 justlooks_ bdf: no ,nothing except salt master, i find the reason is service salt-master  stop sometime can not work
03:53 _fllr_ joined #salt
03:54 bdf I've seen that happen beforer
03:54 bdf -r
03:54 Ryan_Lane joined #salt
03:54 bdf did anything change in the init script for the debian packages?
03:54 bdf for 2014.1.0
03:54 justlooks_ bdf: i use redhat6.2
03:57 joehh bdf no
03:59 dyim joined #salt
03:59 justlooks_ justlooks_:  here  https://gist.github.com/justlooks/9360875
03:59 schimmy joined #salt
04:00 bdf nice. :(
04:01 NV justlooks_: make your terminal wider so it doesn't cut off
04:01 joehh but there is a long standing "comprommise"  here
04:02 joehh compomise
04:02 joehh argggh laggy connection
04:02 joehh please excuse my typing
04:02 fllr joined #salt
04:03 joehh bdf: old version used to kill off all minions/masters
04:03 joehh this was too aggressive and would cause updates/upgrades of salt itself to fail badly
04:03 bdf right, I remember that being great when updating a minion with salt
04:04 joehh now the init scripts only kill the original master/minion process
04:06 AdamSewell joined #salt
04:06 joehh changed around 0.16.3/4
04:08 bdf yeah, I ran straight into it in a bad way -- unknowingly pushed out a bad sudoers to a bunch of ec2 boxes without root passwords, upgraded salt with salt
04:08 joehh I thikn salt needs to be modified to somehow deal with restarting/upgrading better
04:08 zain_ joined #salt
04:08 bdf *face palm*
04:08 bdf yeah, I was almost thinking it would be neat if it tried to spawn an upgrade with 'at' or something a minute into the future
04:08 joehh but I think it is a chicken/egg problem and not easy to solve
04:09 joehh painful
04:09 joehh that was how we had to upgrade from 0.16.3 with the debian packages
04:11 bdf but it feels like some sort of external watchdog could achieve the same result
04:11 bdf and that dog's name is... systemd
04:11 NV eh minion is easy
04:11 bdf j/k
04:11 NV when receiving a signal to die, just stop handling new requests from zmq
04:11 joehh :)
04:11 NV and terminate  once all currently running jobs are finished
04:12 NV aka, almost identical to how apache et al's graceful restart works
04:12 sroegner joined #salt
04:13 dyim joined #salt
04:15 jeffro joined #salt
04:15 jeffro joined #salt
04:15 joehh I suspect that is how the minion works
04:17 joehh it seems to behave quite nicely as long as you are happy to accept processes
04:17 Nexpro1 joined #salt
04:18 joehh finishing their work (potently some time) after the minion is "stopped"
04:22 taion809 joined #salt
04:22 naiyte joined #salt
04:24 mgw joined #salt
04:31 joehoyle joined #salt
04:37 zain_ joined #salt
04:38 dyim joined #salt
04:41 kermit joined #salt
04:45 jalbretsen joined #salt
04:49 tzero joined #salt
04:56 krow joined #salt
04:56 schimmy joined #salt
04:57 Furao joined #salt
04:58 echos joined #salt
04:58 jdenning joined #salt
04:59 Furao joined #salt
04:59 zain_ joined #salt
04:59 pdayton joined #salt
05:04 yomilk joined #salt
05:10 zain_ joined #salt
05:13 dyim joined #salt
05:13 krow joined #salt
05:15 jeremyfelt joined #salt
05:19 krow joined #salt
05:21 kintel joined #salt
05:24 krow joined #salt
05:26 schimmy joined #salt
05:30 krow joined #salt
05:34 fllr Does anyone know how to instal postgis using saltstack?
05:34 joehh fllr: which os?
05:34 fllr ubuntu precise
05:34 fllr install*
05:35 jcockhren fllr: I have a state
05:35 omar joined #salt
05:35 fllr Awesome! Gah, I've been trying for hours now...
05:35 dave_den Anyone in Denver: https://twitter.com/d__a__v__e__/statuses/441084037814161408
05:35 krow joined #salt
05:35 jcockhren postgres 9.1 though
05:36 jcockhren for^
05:37 fllr jcockhren: that's actually what I'm running... :)
05:39 fllr jcockhren: Can you paste it?
05:39 favadi joined #salt
05:39 jcockhren fllr: yeah. gotta sanitze right fast. (it's part of a vagrant setup)
05:40 joehoyle joined #salt
05:40 fllr jcockhren: Ah, ok... I was just wondering. Lol.
05:40 krow joined #salt
05:44 dave_den hah, who's with scriptrock in here?
05:46 krow joined #salt
05:48 forrest joined #salt
05:48 jdenning joined #salt
05:49 fllr jcockhren: I'm gonna walk the dog in the mean time, let me know when you have it...
05:51 forrest Ugh, more companies need to use Salt. I'm tired of looking at job postings for puppet :(
05:51 krow joined #salt
05:55 malinoff forrest, it's easy - apply a puppet job posting and convince management dept. to switch from puppet to salt :)
05:55 forrest malinoff, lol, it's not always that easy unless you're at a smaller shop
05:55 malinoff that's what I'm doing right now :D
05:56 krow1 joined #salt
05:56 forrest kudos to you then
05:56 justlooks_ after i update to 2014.1.0 i find i can not send any commmand to client https://gist.github.com/justlooks/9360875
05:57 forrest justlooks_, are both the master and minions 2014.1?
05:57 justlooks_ and in master ,many zombie process
05:58 justlooks_ forrest: yes
05:58 vxitch joined #salt
06:00 forrest justlooks_, and you restarted all the services? What OS?
06:01 krow joined #salt
06:02 justlooks_ forrest: i use redhat 6.2
06:03 justlooks_ forrest: and restarted salt-master and salt-minion
06:03 justlooks_ forrest: the minion is connect other master befor
06:03 forrest weird, this is the first time so far I'm hearing of any issues like that.
06:03 justlooks_ forrest: before
06:06 forrest justlooks_, I'm not sure. I've gotta head out for the evening, best of luck!
06:07 krow joined #salt
06:09 ChaosPsyke joined #salt
06:10 thayne joined #salt
06:13 krow joined #salt
06:13 fllr jcockhren: Nothing?
06:13 jdenning joined #salt
06:16 jcockhren fllr: https://gist.github.com/jcockhren/a8b5bf473b4bb87aa291
06:16 fllr jcockhren: Thanks! :)
06:17 snuffeluffegus joined #salt
06:17 jcockhren in my vagrant it created the users as well but it craps out in 2014.1 (worked fine in 0.17.5)
06:18 jcockhren fllr: also, it assumes there's some pg_hba.conf in your salt filesystem
06:18 krow joined #salt
06:18 dyim joined #salt
06:18 fllr jcockhren: that's fine. I do have a pg_hba, and I've already got something that works for creating the users for me...
06:19 jcockhren cool
06:23 faldridge joined #salt
06:24 krow joined #salt
06:26 fllr joined #salt
06:29 krow joined #salt
06:32 krow1 joined #salt
06:37 jnials joined #salt
06:48 krow joined #salt
06:48 ckao joined #salt
06:55 ckao joined #salt
06:56 zain__ joined #salt
06:57 zain___ joined #salt
07:06 borgstrom joined #salt
07:13 zain_ joined #salt
07:17 nn0101 joined #salt
07:19 Furao joined #salt
07:20 justlooks_ if i reinstall salt master OS ,how can i register all my minion again?
07:22 snuffeluffegus joined #salt
07:23 nn0101 joined #salt
07:23 dyim joined #salt
07:24 faldridge joined #salt
07:24 justlooks_ should i need delete something in minion ,which belongs to old salt master?
07:26 ravibhure joined #salt
07:27 Furao_ joined #salt
07:27 nn0101 joined #salt
07:29 krow joined #salt
07:38 jdenning joined #salt
07:39 ndrei joined #salt
07:41 gammalget joined #salt
07:43 borgstrom joined #salt
07:45 ml_1 joined #salt
07:45 jcockhren justlooks_: no. I think you'll just have to just accept the minions again
07:46 GradysGhost joined #salt
07:46 jcockhren well... probably the minion_master.pub in the pki directory
07:46 jcockhren I haven't tried it
07:47 vxitch joined #salt
07:47 justlooks_ jcockhren: yes somtime the old master pub file cause problem,i removed it ,now it's ok
07:54 joehoyle joined #salt
07:57 lessismore joined #salt
08:01 pentabular joined #salt
08:03 TheRealBill_he-1 joined #salt
08:04 scott_walton joined #salt
08:05 aaroneous1 joined #salt
08:06 cornmand1r joined #salt
08:10 Ztyx joined #salt
08:11 ajw0100 joined #salt
08:11 gammalget_ joined #salt
08:11 Kenzor joined #salt
08:11 xmltok joined #salt
08:12 slav0nic joined #salt
08:12 slav0nic joined #salt
08:15 goki joined #salt
08:17 Ztyx joined #salt
08:18 favadi joined #salt
08:18 gmoro joined #salt
08:20 justlooks_ i install master and minion on three kvm guest which locate on some physical box,it seems very slow when i issue cmd from master (each kvm guest with 512m RAM and 1 VCPU)
08:23 Ztyx joined #salt
08:25 harobed joined #salt
08:26 nebuchadnezzar joined #salt
08:28 dyim joined #salt
08:29 linjan_ joined #salt
08:34 giantlock_ joined #salt
08:35 carlos_ joined #salt
08:36 balboah joined #salt
08:36 yomilk joined #salt
08:43 ndrei joined #salt
08:44 kadel joined #salt
08:44 zain_ joined #salt
08:44 joehoyle joined #salt
08:46 yomilk_ joined #salt
08:50 johtso joined #salt
08:52 n8n joined #salt
08:52 ravibhure joined #salt
09:00 joehoyle joined #salt
09:01 17SAAOD7B joined #salt
09:01 MTecknol1gy joined #salt
09:01 jhauser joined #salt
09:01 foresj joined #salt
09:01 kossy joined #salt
09:01 vmdsch1 joined #salt
09:01 Sypher joined #salt
09:01 babilen joined #salt
09:01 jbub joined #salt
09:01 sinh joined #salt
09:01 analogbyte joined #salt
09:01 svx joined #salt
09:01 SaveTheRbtz joined #salt
09:01 msciciel joined #salt
09:01 magenbrot joined #salt
09:01 scarcry joined #salt
09:01 andrej joined #salt
09:01 vbabiy joined #salt
09:01 sashka_ua joined #salt
09:01 balltongu joined #salt
09:01 tmwsiy_ joined #salt
09:01 openxmark joined #salt
09:01 Tinuva joined #salt
09:01 jeddi joined #salt
09:01 kaictl joined #salt
09:01 jeffro joined #salt
09:01 jnials joined #salt
09:01 funzo_ joined #salt
09:01 alienresidents joined #salt
09:01 Andrevan joined #salt
09:01 monokrome joined #salt
09:01 repl1cant joined #salt
09:01 rhand joined #salt
09:01 twiedenbein joined #salt
09:07 octarine joined #salt
09:07 jrdx joined #salt
09:07 snuffeluffegus joined #salt
09:12 MTecknology joined #salt
09:12 jforest joined #salt
09:20 jrdx left #salt
09:20 jrdx joined #salt
09:22 justlooks_ left #salt
09:23 justlooks_ joined #salt
09:23 zizzler joined #salt
09:23 smcquay joined #salt
09:24 justlooks_ left #salt
09:25 MrTango joined #salt
09:26 jhd3 joined #salt
09:26 zizzler i updated salt on ubuntu box and highstate stopped working. any advice? http://pastebin.com/ThwG9k7b
09:28 jhd3 left #salt
09:29 joehoyle joined #salt
09:34 dyim joined #salt
09:36 BogdanR_ Hello. I have an error when I try to initiate a highstate for some machines: http://hastebin.com/raw/carayutola
09:36 BogdanR_ Can you please tell me why this is not detecting the topfile?
09:37 jeffro joined #salt
09:37 zizzler BogdanR_, trying to figure this myself. i updated salt and no started getting some backtraces http://pastebin.com/ThwG9k7b
09:38 snuffeluffegus joined #salt
09:38 zizzler no == now
09:40 Furao joined #salt
09:40 zizzler i do use custom apt_repository.py plugin
09:40 zizzler maybe its that
09:43 BogdanR_ I am not using anything custom
09:45 jrdx joined #salt
09:45 zizzler i got the same error myself when salt-minion was newer then salt-server but now both are updated to 2014.1.0-1saucy1 and started getting those backtraces now
09:51 NV zizzler: without knowing more - it soundsl ike that is probably caused by your custom apt_repository plugin
09:52 NV BogdanR_: state.show_tops may assist
09:53 NV also cp.list_states
09:53 NV im guessing its an issue with your master configuration as to where your roots are located
09:53 NV which may or may not be due to environment settings
09:56 holys joined #salt
09:56 Furao custom apt_repository.py ?
09:57 zizzler i think this https://github.com/saltstack/salt-contrib/blob/master/states/apt_repository.py
09:58 NV zizzler: any reason you can't use the stock one? I note that was last updated 8 months ago...
09:58 Furao I'm the original author of that
09:58 NV mhmm, apt_repository isn't a pkg module?
09:58 NV perhaps your trace isn't from that then...
09:58 Furao my version differ from that one
09:59 Furao but it works in 0.17.5
09:59 Furao apt_repository does some things better than pkg_repo, I wish I can stop using it
10:00 xmltok joined #salt
10:01 zizzler NV, probably i couldn't find docs and get it to work
10:03 devon_ joined #salt
10:03 NV http://docs.saltstack.com/ref/modules/all/salt.modules.aptpkg.html#module-salt.modules.aptpkg is the vanilla module ftr, usable from the pkgrepo state http://docs.saltstack.com/ref/states/all/salt.states.pkgrepo.html
10:04 Furao NV: apt_repository don't call pkg.list_pkgs so that traceback is caused by something else
10:04 bhosmer joined #salt
10:04 tmwsiy__ joined #salt
10:05 NV Furao: yeah - I realised that earlier (I initially thought it was a pkg module, but after a quick look at the source I noticed it wasn't hence my comment of 'perhaps your trace isn't from that then')
10:05 NV but still probably worth looking at using the vanilla repo module if there's no good reason to use something else
10:05 NV anyway, look at your states that use the pkg state module
10:06 NV grep for 'fromrepo'
10:06 NV my suspicion is it's probably related to that
10:07 zizzler fromrepo doesnt exist in my recipes
10:08 johtso joined #salt
10:08 NV hrm, well as far as I can tell, you have a sls that relies on pkg.installed that is breaking it
10:08 NV and something is referencing fromrepo
10:09 jrdx joined #salt
10:10 NV hrm, interesting
10:10 NV line 113 on my version of pkg.py (2014.1.0) has cur_pkgs = __salt__['pkg.list_pkgs'](versions_as_list=True)
10:10 NV from your trace you have cur_pkgs = __salt__['pkg.list_pkgs'](versions_as_list=True, **kwargs)
10:10 NV zizzler: are you on 2014.1.0?
10:11 zizzler yeah
10:12 NV kwargs is passed from the kwargs in the _find_install_targets function call, and since _find_install_targets is called with fromrepo but doesn't have it defined as a parameter, that sounds like that's the grief
10:12 NV wait hang on, im on the wrong box, that was from 0.17.5
10:12 NV disadvantage of running multiple sites with salt :P
10:12 austin987 joined #salt
10:13 NV hrm ok my 2014.1.0 site has the same code as you
10:14 NV ok i see the problem
10:15 NV i use rhel on that site - the salt/modules/yumpkg.py file has list_pkgs defined as def list_pkgs(versions_as_list=False, **kwargs):
10:15 NV meanwhile salt/modules/dpkg.py has def list_pkgs(*packages):
10:16 NV in 0.17.5 it didn't pass fromrepo, so it would have worked, but in 2014.1.0 it does, and without kwargs will error out
10:16 diegows joined #salt
10:16 NV as a temp fix you can copy dpkg.py to your states/_modules folder and edit the definition to add **kwargs at the end
10:17 NV and throw in a regression bug report on github for a long term fix
10:17 zizzler ok, i'll try
10:17 zizzler is that backtrace good enough for bug report
10:18 NV I'd include the backtrace (ignore the first line about the __virtualname__ stuff, it's irrelevant) and note what I mentioned
10:18 NV or rather, note the edit you made to dpkg.py and your success
10:19 NV if you're feeling adventerous you could probably even open a pull request if that fixes it for you :P
10:20 ash__ joined #salt
10:23 zizzler NV, i have in master. file_roots base set at /home/zizzler/salt-master/files/base. is the correct place for the file /home/zizzler/salt-master/files/base/_modules/dpkg.py
10:24 NV yes
10:24 NV be sure to run saltutil.sync_all
10:24 NV once it's in place
10:24 NV maybe even restart the salt minion to be sure if you really want
10:28 helderco joined #salt
10:32 Kenzor joined #salt
10:33 zizzler NV, i cant seem to sync the files http://pastebin.com/83RuiJ1w the command returns empty list
10:35 elfixit joined #salt
10:38 dyim joined #salt
10:45 jrdx joined #salt
10:48 BogdanR_ NV: The earlier problem that the top state was not detected was a bug in salt that has been fixed with an upgrade
10:49 BogdanR_ In 2014.1.0 it works as expected
10:49 zizzler lucky you :)
10:49 rdunlop joined #salt
10:53 jrdx joined #salt
10:54 jrdx joined #salt
10:56 helderco joined #salt
10:59 zooz joined #salt
10:59 helderco joined #salt
11:02 Furao joined #salt
11:05 yomilk joined #salt
11:06 sroegner joined #salt
11:08 ndrei joined #salt
11:10 Kenzor joined #salt
11:15 markm joined #salt
11:20 jrdx joined #salt
11:23 krow1 joined #salt
11:29 taion809 joined #salt
11:31 ndrei joined #salt
11:34 krow joined #salt
11:42 joehoyle joined #salt
11:44 dyim joined #salt
11:44 Teknix joined #salt
11:47 backjlack_ joined #salt
11:48 runnyspot joined #salt
11:49 btorch_ joined #salt
11:49 tcotav joined #salt
11:49 krow joined #salt
11:49 nebuchad` joined #salt
11:50 krow1 joined #salt
11:50 rjc_ joined #salt
11:50 dirjax_ joined #salt
11:50 abele_ joined #salt
11:50 simonmcc_ joined #salt
11:50 BogdanR joined #salt
11:51 kedo40 joined #salt
11:52 z3uS| joined #salt
11:52 honestly_ joined #salt
11:52 xzarth_ joined #salt
11:52 shadowsu1 joined #salt
11:52 pnl_ joined #salt
11:53 dpacz joined #salt
11:53 bezaban_ joined #salt
11:53 mackstic1 joined #salt
11:53 dcmorton_ joined #salt
11:53 mrchrisadams_ joined #salt
11:53 ndrei_ joined #salt
11:53 gamingrobot_ joined #salt
11:54 _Flusher joined #salt
11:54 jero_ joined #salt
11:54 krow joined #salt
11:54 modafinil__ joined #salt
11:54 crazysim joined #salt
11:54 rcsheets_ joined #salt
11:54 rogst_ joined #salt
11:54 herzi_ joined #salt
11:54 jgelens_ joined #salt
11:55 delkins_ joined #salt
11:55 mortis__ joined #salt
11:55 [vaelen] joined #salt
11:55 tempspace joined #salt
11:56 kamal_ joined #salt
11:56 bretep` joined #salt
11:56 gammalget joined #salt
11:56 renoirb_ joined #salt
11:56 tr_h joined #salt
11:56 mjulian joined #salt
11:56 cyrusdavid joined #salt
11:56 copelco____ joined #salt
11:56 scalability-junk joined #salt
11:56 mjulian joined #salt
11:56 codekoala joined #salt
11:56 juice joined #salt
11:56 scalability-junk joined #salt
11:56 copelco____ joined #salt
11:56 codekoala joined #salt
11:56 dangra_ joined #salt
11:56 luminous joined #salt
11:57 pexio joined #salt
11:57 echos joined #salt
11:57 seanz joined #salt
11:57 bejer_ joined #salt
11:57 nlb joined #salt
11:58 mgw joined #salt
12:00 Flusher joined #salt
12:00 sroegner joined #salt
12:01 abele_ joined #salt
12:02 backjlack joined #salt
12:02 Ryan_Lane joined #salt
12:02 abele joined #salt
12:04 [diecast] joined #salt
12:06 krow joined #salt
12:08 gtmtech joined #salt
12:08 gtmtech Hey all
12:09 gtmtech Got a question - in a jinja template, I've got an array of IPs - I want to output a line if and only if that IP is configured on one of the interfaces on the machine - how do I do it?
12:11 gtmtech Or to make it simpler - I've got one IP in a variable e.g. ip="10.1.0.2" - I want to write a line if and only if that IP is bound to eth0 or eth1 or eth2 etc. on a machine
12:12 blast_hardcheese joined #salt
12:19 yomilk joined #salt
12:22 xt joined #salt
12:27 bhosmer joined #salt
12:30 zz_Cidan joined #salt
12:30 devoncoder joined #salt
12:33 devoncoder Hi guys, since upgrading to salt 2014.1.0 my highstate is failing with the following error... invalid requisite type 'file.directory' in state.. any ideas?
12:35 devoncoder Requisite types must not contain dots, did you mean 'file'? is at the end of the error message
12:37 generj joined #salt
12:38 generj so I configured gitfs on my master and restarted it
12:38 generj but it doesn't pull in the repo
12:44 NV devoncoder: pretty much what the error message says, the correct syntax is '- require:\n  file: statename'
12:44 NV not 'file.directory: statename'
12:45 NV a named state can't have more than one function applied from a state module anyway
12:45 NV ie, it is invalid to have '/etc/resolv.conf:\n  file.managed:\n...\n  file.symlink:\n...'
12:45 generj joined #salt
12:46 NV so there is no need for the additional specificity that want file.directory, rather than file.symlink say, because it can only have one of any type of 'file' state applied anyway
12:46 krow joined #salt
12:47 devoncoder OK Many thanks NV for your help
12:48 NV the plus side of course is if you change how you manage that file from say managed to symlink to append, etc - you don't have to update your watch and require statements :D
12:48 yomilk joined #salt
12:48 baniir joined #salt
12:49 dyim joined #salt
12:49 luminous joined #salt
12:49 generj joined #salt
12:53 jrdx joined #salt
12:55 clemenko joined #salt
12:56 generj joined #salt
13:03 generj joined #salt
13:09 dyim joined #salt
13:11 viod joined #salt
13:12 thayne joined #salt
13:12 bryanl joined #salt
13:16 jhauser_ joined #salt
13:18 generj joined #salt
13:23 Gordonz_ joined #salt
13:24 blee joined #salt
13:26 raizyr joined #salt
13:26 rojem joined #salt
13:31 Vivi-1 left #salt
13:37 Gordonz joined #salt
13:38 generj joined #salt
13:43 jslatts joined #salt
13:44 giantlock__ joined #salt
13:44 favadi joined #salt
13:45 vuvu joined #salt
13:45 vu_ joined #salt
13:46 generj joined #salt
13:48 MrTango joined #salt
13:49 diegows joined #salt
13:50 dyim joined #salt
13:52 generj1 joined #salt
13:53 torrancew joined #salt
13:54 Gordonz_ joined #salt
13:55 Gordonz_ joined #salt
13:56 vuvu joined #salt
13:56 vu_ joined #salt
13:56 dave_den joined #salt
13:57 johtso joined #salt
13:57 helderco joined #salt
14:01 baniir joined #salt
14:01 Networkn3rd joined #salt
14:04 vu_ joined #salt
14:06 generj joined #salt
14:07 vuvu joined #salt
14:09 vu_ joined #salt
14:09 redblackbit joined #salt
14:10 rojem joined #salt
14:10 racooper joined #salt
14:11 redblackbit Is it possible to install easily user-defined dmgs on OSX minions instead of using brew as package manager?
14:13 pdayton joined #salt
14:18 quickdry21 joined #salt
14:18 Iwirada joined #salt
14:19 vu_ joined #salt
14:19 vuvu joined #salt
14:20 vuvu joined #salt
14:20 vu_ joined #salt
14:20 jeremyfelt joined #salt
14:21 seapasulli joined #salt
14:22 mpanetta joined #salt
14:23 toastedpenguin joined #salt
14:23 vejdmn joined #salt
14:23 generj joined #salt
14:26 mpanetta_ joined #salt
14:26 sgviking joined #salt
14:27 foxx joined #salt
14:28 juicer2 joined #salt
14:30 generj joined #salt
14:30 jaimed joined #salt
14:33 redblackbit Is it possible to install easily user-defined dmgs on OSX minions instead of using brew as package manager?
14:33 jforest Has anyone tried installing pkgrepos in centos6 and then trying to install packages from them?  It doesn't work for me at all
14:34 jforest I can run salt again, and then the second time it works, because the repo is installed and the packages are there
14:34 vuvu joined #salt
14:34 vu_ joined #salt
14:34 jforest Comment: Package 'python27' not found (possible matches: python27)
14:34 jforest that is the error I get
14:35 jrdx joined #salt
14:36 krow joined #salt
14:37 ndrei joined #salt
14:39 njpatel joined #salt
14:39 CeBe joined #salt
14:43 akoumjian joined #salt
14:44 pydanny joined #salt
14:46 Marion joined #salt
14:49 yomilk joined #salt
14:50 pydanny joined #salt
14:50 tessellare joined #salt
14:52 joehoyle joined #salt
14:53 tessellare left #salt
14:56 helderco joined #salt
15:01 krow joined #salt
15:02 ndrei joined #salt
15:04 JasonSwindle joined #salt
15:05 gnugnu8 joined #salt
15:08 rdunlop joined #salt
15:08 ajprog_laptop joined #salt
15:14 kintel joined #salt
15:19 cron0 joined #salt
15:21 quantumriff joined #salt
15:23 rostam joined #salt
15:25 mgw joined #salt
15:27 forrest joined #salt
15:28 FL1SK joined #salt
15:31 Gordonz joined #salt
15:31 Gordonz joined #salt
15:32 Marion left #salt
15:33 bitmand joined #salt
15:35 bitmand Hi, I have a problem with my minions, they keep dying/shutting down randomly when/after running jobs. I wan't to debug this a bit, but unsure what log level is appropriate? ( I have tried with debug, but that may be a bit too much info )
15:36 bitmand By the way, I have 2 different salt setups, the setup running on Ubuntu doesn't have any problem - but the minions running on FreeBSD does :(
15:36 kermit joined #salt
15:40 xmj t'd be interesting to know why the freebsd minions die
15:40 forrest bitmand, what release of salt are you running, and I would suggest starting the minion with debug, then you'll get 'all the things'
15:40 xmj 'cause that sure as hell ain't the normal state of affairs
15:40 bitmand The only time i succeeded in reproducing the problem, with a non-default log level, the last entry in the minion log was: Returning information for job: 20140224... - so I suspect the job ran succesfully, but why the minion died afterwards, I don't know.
15:41 bitmand xmj: 0.17.4 og .5 ( it varies ) but happened a couple of time today on a .5
15:41 bitmand xmj: and thanks, I will set to debug on my minion to track this down
15:42 forrest bitmand, is your master 0.17.5, and all minions are 0.17.4 or 0.17.5?
15:42 bitmand forrest: sorry, got you and xmj mixed together :)
15:42 forrest bitmand, I AM SUPER INSULTED :P
15:42 bitmand forrest: hah! :)
15:43 bitmand forrest: master is 0.17.5 - double checking all minions
15:43 bitmand forrest: all minions is either 0.17.4 or .5
15:43 forrest bitmand, ok cool, the master would be the biggest concern. I'd say start the minion in debug, run a job on that single minion, and do a few things, run the command with -l debug, AND if the command is long running, use the -t option to set the timeout to something long enough that it won't drop you back to the console during the job run
15:44 ipmb joined #salt
15:45 dcmorton joined #salt
15:45 akitada joined #salt
15:45 bitmand forrest: Gotit - I am setting log_level_logfile: debug on all minions now.
15:46 kintel joined #salt
15:46 forrest bitmand, sounds good, like I said you might want to run a single test on one minion with both state.highstate, as well as state.sls to see if only specific things cause it.
15:47 bitmand forrest: The bit about the job executing take a bit more effort ( salt is not run manually ), but I will look into that as well. Salt is run with -t 60 pr. default ( and shouldn't take longer than that )
15:47 bitmand hmm
15:47 forrest bitmand, great that should be plenty of time, it's usually just the 5 second timeout that bites people
15:47 vxitch joined #salt
15:48 bitmand Yeah - confused me in the beginning too :)
15:48 markm joined #salt
15:48 jhd3 joined #salt
15:50 bitmand forrest: I'll try - thanks :) My statetable is extremely simple at the moment, only consists of keeping 2 directories sync'ed. And using cmd.run afterwards to run a script on the minion.
15:50 bitmand forrest: But know you say it, I don't really know if it's the state.highstate og run.cmd that "causes" this .. I'll investigate, thanks the for great pointers so far ( AND quick response )
15:50 bitmand but now*
15:51 forrest bitmand, Yea gotta start small and investigate from there most of the time. Not a problem, it's pretty early still here at work which helps!
15:52 jeremyBass1 joined #salt
15:53 zain_ joined #salt
15:54 faldridge joined #salt
15:56 Kenzor joined #salt
15:59 teskew joined #salt
16:00 faldridge joined #salt
16:01 kaptk2 joined #salt
16:02 vu_ joined #salt
16:02 vuvu joined #salt
16:02 krow joined #salt
16:02 jrdx joined #salt
16:03 jeremyfelt joined #salt
16:03 pdayton joined #salt
16:03 sroegner joined #salt
16:04 jergerber joined #salt
16:04 jY anyone have a minute to answer some questions on making a external pillar that is a command to read data from an external source? i'm a bit unsure how to call the pillar from the ext_pillar: in master config
16:06 vxitch joined #salt
16:07 jalbretsen joined #salt
16:08 vu_ joined #salt
16:08 vuvu joined #salt
16:08 vuvu joined #salt
16:08 vu_ joined #salt
16:08 KennethWilke joined #salt
16:11 CaptTofu joined #salt
16:13 rallytime joined #salt
16:13 crazysim joined #salt
16:14 vu_ joined #salt
16:14 vuvu joined #salt
16:15 helderco joined #salt
16:17 diegows joined #salt
16:21 vuvu joined #salt
16:21 vu_ joined #salt
16:21 pdayton joined #salt
16:23 baniir joined #salt
16:26 jared_ joined #salt
16:26 rgbkrk joined #salt
16:27 pdayton1 joined #salt
16:27 HeadAIX joined #salt
16:29 mgw joined #salt
16:34 markm joined #salt
16:37 Networkn3rd joined #salt
16:37 Iwirada left #salt
16:40 ndrei joined #salt
16:41 vxitch joined #salt
16:41 oz_akan_ joined #salt
16:42 hunter joined #salt
16:43 pdayton joined #salt
16:46 naiyte joined #salt
16:48 frasergraham joined #salt
16:50 yomilk joined #salt
16:58 eeffff joined #salt
17:01 hunter I posted this last night but naturally after everyone was gone.
17:01 hunter http://pastebin.com/BCeL6KnV
17:01 hunter I see and understand the problem - my question for the group is whats the best salt idiom to follow to accomplish it?
17:01 mgw1 joined #salt
17:02 oc what's a good strategy for targeting different OSes? if-blocks with grains['os']? how about file-structure differences (i.e. /etc/apache2/sites-enabled vs /etc/httpd/conf.d), also ifs?
17:02 redblackbit Is it possible to install easily user-defined dmgs on OSX minions instead of using brew as package manager?
17:02 krow joined #salt
17:03 hunter oc: if you look in the saltstack formulas on github, the common idiom for that is a "map.jinja" file that contains the os/platform differences
17:03 oc or do you do a {% if grains['os'] == 'redhat' %} include: redhat {% elif ... %}
17:03 oc hunter: thanks
17:04 frasergraham joined #salt
17:04 GradysGhost joined #salt
17:04 hunter oc:https://github.com/saltstack-formulas/bind-formula/blob/master/bind/map.jinja
17:04 AdamSewell joined #salt
17:04 AdamSewell joined #salt
17:05 hunter Thats the one from bind, which I had open in a tab. Most of them follow that pattern.
17:05 hunter There's doc on what the boilerplate means - let me see if I can find it
17:05 oc no problem, i've already parsed the concept :)
17:05 ndrei joined #salt
17:06 hunter oc: http://docs.saltstack.com/topics/conventions/formulas.html#map-jinja
17:06 hunter The docs go into why and how - I found them interesting. That whole page is useful actually.
17:07 hunter Anyone with a suggestion for my question?
17:08 bemehow joined #salt
17:09 fnbar joined #salt
17:09 fnbar Hi Guys
17:09 rgbkrk joined #salt
17:10 bhosmer joined #salt
17:10 rgbkrk joined #salt
17:11 fnbar Does anyone also facing this problem: (Debian Wheezy, Salt 2014.1). The salt-master as well as the minion need a long time until they're started. Both are stuck at "Reading config from $CONF_DIR". After 1-2 Minutes, they work... I don't have a central DNS server, so I'll added the IPs to the configurations...(yes, they work...)
17:11 [diecast] joined #salt
17:13 bitmand fnbar: pure guess, reverse dns of some sort?
17:17 oc hunter: thanks, that was a great piece of doc :)
17:17 rainmanjam joined #salt
17:17 borgstrom joined #salt
17:19 oc hunter: what's wrong with the way you do it in the pastebin?
17:19 fnbar bitmand: Hmm..why should it, dns_check is disabled in the minion config...
17:19 thayne joined #salt
17:20 fnbar Also, when running the master in foreground and putting out the debug-logs, "Found minion id from getfqdn(): $myhost"..
17:21 baniir joined #salt
17:22 TheRealBill_here joined #salt
17:25 davet joined #salt
17:26 oc fnbar: how have you set hostname? (i.e. 1) what does hostname -s, hostname -d and hostname -f report) 2) have you tampered with the getent lookup order (Name Service Switch - /etc/nsswithc.conf) ?
17:26 Gordonz joined #salt
17:27 oc sounds like the primary name service resolution method times out
17:27 jhd3 joined #salt
17:27 fnbar oc: hostname repots the correct host- and domainname. I also didnt tampered the lookup order (hosts: files, dns then networks: files)
17:28 oc test by adding your nodes to your /etc/hosts file, i'm assuming files is the first entry of your nsswitch
17:28 quickdry21 joined #salt
17:29 oc fnbar: have only seen those kind of host-lookup errors with ldap servers timing out before :>
17:29 fnbar oc: Yes, but I turned
17:30 fnbar oc: Sorry, accidently hit <ENTER>. Thats what I did first, didn't help either... :l
17:31 fnbar oc: retry_dns turned to 0, didn't help..
17:32 fnbar tcp_{pull,pub}_port is only for local communication on the minion
17:33 oc im
17:34 oc fnbar: possibly related (race condition): https://github.com/saltstack/salt/issues/5670
17:34 mgw joined #salt
17:36 oc fnbar: what does python -c "import socket; print socket.getfqdn()"
17:37 mgw joined #salt
17:37 fnbar oc: The time.sleep is already there. I increased it, but no effect..
17:38 fnbar oc: Already tried that, returns the correct fqdn...
17:38 smcquay joined #salt
17:38 oc I guess you have to report an issue then, with --versions, shell output, etc :)
17:39 oc https://github.com/saltstack/salt/blob/develop/Contributing.rst
17:39 jY is there a trick to get external pillars working.. i have a pillar file in /usr/share/pyshared/salt/pillar cmd_json works but mine says it can't load "[CRITICAL] Specified ext_pillar interface etcd is unavailable"
17:40 fnbar Thanks oc, I will report this issue until end of the week. Still trying to ensure that it isn't because my somehow-special setup. Many thanks for your help!
17:42 sporkd2 joined #salt
17:43 oc :~ sorry for not being able to sort it out :)
17:45 fnbar No problem, sooner or later it'll be sorted out :)
17:45 jY here's more info I just posted about in the mailing list
17:45 jY https://groups.google.com/forum/#!topic/salt-users/j-h30kWBR5w
17:46 danielbachhuber joined #salt
17:46 millz0r joined #salt
17:51 schimmy joined #salt
17:52 thayne joined #salt
17:54 icarus joined #salt
17:56 Heartbane robawt: seen hpfx?
17:57 jaimed joined #salt
17:57 kermit joined #salt
17:57 schimmy joined #salt
17:58 themadcanudist joined #salt
17:59 themadcanudist hey guys, what's the preferred method for debugging / pdb / walking through mako/jinja templates to fix a problem?
18:02 krow joined #salt
18:02 Ryan_Lane1 joined #salt
18:03 themadcanudist how about getting salt to spit out the resultant YAML on a run? For better troubleshooting?
18:03 diegows joined #salt
18:03 dyim is there a way to "watch" multiple services? I know for a pkg, you can have pkgs, for service, can you have services?
18:04 Thiggy joined #salt
18:06 zain_ joined #salt
18:08 Heartbane So hpfx showed me this way to sync the entire contents of a directory in salt but I can 't see it
18:08 dyim i can't seem to have multiple service under pkg and there isn't a way to list multiple services under service
18:09 Gareth morning
18:09 zain__ joined #salt
18:10 lzhang joined #salt
18:11 Heartbane Nevermind found it recurse
18:13 freerod joined #salt
18:15 socket1 joined #salt
18:16 tmwsiy_ joined #salt
18:17 schimmy1 joined #salt
18:19 Ryan_Lane1 joined #salt
18:20 vxitch hey guys, what python version does salt need? can't find it online
18:21 Vye Does salt-api 8.3 work with 2014.1 or 17.5?
18:22 Thiggy joined #salt
18:22 robawt Heartbane: fxhp
18:22 robawt :)
18:22 robawt effects hitpoints
18:23 Ryan_Lane1 is 2014 going to be built for lucid?
18:23 freerod Is it reasonable to reuse a minion key by creating a vm image that already has salt minion installed and configured?  If so, could I then trigger a minion rekey while remaining registered with master?
18:23 Ryan_Lane1 I can't upgrade without that :(
18:24 Ryan_Lane joined #salt
18:24 Ryan_Lane oh, it's already done
18:24 Ryan_Lane1 joined #salt
18:28 freerod vxitch: based on the requirements files in the project it looks like it is tested with 2.6 and 2.7.
18:28 robawt freerod: you may run into problems if your networking info stays the same.  you may want to just follow the suggestions on this page: http://docs.saltstack.com/topics/tutorials/preseed_key.html
18:29 vxitch freerod: thanks. i'm still fighting that strange python error. if its good on 2.6 then i guess i'm ok. i'm scared of trying to install 2.7 for fear of breaking yum
18:29 scoates is there an intelligent way for one state to "inform" another state? e.g. I would like one of my application states to inform my syslog state that it creates certain logs that need shipping.
18:29 scoates sort of like `watch_in` but more like `context_in`…
18:30 Heartbane robawt: fxhp (effects hitpoints) got it
18:34 Ryan_Lane1 joined #salt
18:37 phyde joined #salt
18:37 krow joined #salt
18:37 freerod robawt: Thanks. I had seen that but got hung up on key distribution. Now that I look more again I see a way and understand the potential.
18:38 freerod scoates: maybe an event?
18:38 vxitch is anyone around who would be able to help me debug a python error salt is throwing? specifically states/iptables.py . here is the error http://hastebin.com/watolanasa.rb
18:38 robawt freerod: yeah it's a grey area because there's no clear way for all setups, but there's a few options in there
18:38 vxitch any devs around?
18:39 sporkd2 vxitch: can you paste you sls?
18:39 forrest scoates, did you try prereq_in? http://docs.saltstack.com/ref/states/requisites.html#prereq-in
18:40 scoates forrest: is there a way to carry actual data along with prereq_in? I'd like fooapp's state to tell the syslog's state "my log file is called /var/log/fooapp.log"
18:41 forrest scoates, hmm, the only thing I can think of would be to run a command, then deal with output there assigned to variables...
18:42 forrest scoates, that would be slick though.
18:42 scoates forrest: was thinking that I could build a really simple registry module… but was hoping for a more… first-class… way to do it.
18:42 forrest yea of course
18:42 forrest I'm trying to think
18:43 JasonSwindle joined #salt
18:43 UtahDave joined #salt
18:43 vxitch sporkd2: http://hastebin.com/nohimaheli.sm
18:44 forrest scoates, I feel like there might be a way to do it with returners, but I'm not familar enough.
18:44 forrest *familiar
18:45 zain_ joined #salt
18:45 frederic joined #salt
18:45 bhosmer joined #salt
18:46 KyleG joined #salt
18:46 KyleG joined #salt
18:46 ajw0100 joined #salt
18:47 freerod vxitch: what version of salt are you using?
18:47 vxitch 2014.1
18:48 xmltok joined #salt
18:49 vejdmn joined #salt
18:50 hunter forrest: you've given good advice in the paste: any opinion on the _correct_ way to structure http://pastebin.com/BCeL6KnV
18:52 forrest hunter, regarding what portion of that? looks ok to me.
18:52 hunter I have two file.* things in a single stanza/state
18:52 hunter salt doesn't like that.
18:53 forrest what is the error
18:53 joehoyle joined #salt
18:54 hunter Name "local_scratch" in sls "paths.scratch" contains multiple state decs of the same type
18:54 forrest yea because they are both using file, just change the symlink to have it's own ID of local_scratch_symlink
18:54 forrest if it were file.managed, and pkg.installed it wouldn't be a problem.
18:54 hunter I wanted them under a single ID since they are both parts of a single .... config action.
18:55 hunter But yes, I can add another ID.
18:55 forrest then you'd have to get rid of file.symlink, and do something like cmd.run
18:55 hunter Pity you can't have another layer on top
18:55 forrest hunter, well, think of it like this. You require file: local_scratch
18:55 blabla_ joined #salt
18:55 forrest does that mean the symlink? Or the directory?
18:55 forrest salt doesn't know, because it doesn't do require file.symlink: local_scratch
18:56 hunter and the _file_ is the outer namespace. yeah - I can live with that.
18:56 forrest yea, there are ways around it (like the cmd.run), but it's easier to just use unique IDs.
18:56 blabla_ just saw that the latest stable is available for ubuntu. is the latest version of the master compatible minions with 0.17.5-1 ?
18:56 budrose joined #salt
18:56 forrest blabla_, as long as the master is 2014.1, your 0.17.5 minions should be ok
18:56 blabla_ forrest: tx
18:57 forrest blabla_, you aren't on bsd are you?
18:57 forrest and you don't use a multi-master setup
18:57 scoates do we know exactly what's wrong with multi-master?
18:58 scoates my minions have 2 masters, but almost all of them can only contact one of the 2.
18:58 scoates (the other is a development master that only certain minions can see, for testing)
18:58 forrest scoates: https://github.com/saltstack/salt/issues/10732
18:58 forrest scoates, cachedout already has a solution and a pull request for it.
18:58 giantlock joined #salt
18:58 scoates ok, cool. will probably await 2014.1.1 then
18:58 scoates thanks for the link
18:59 forrest scoates, a good plan
18:59 forrest yea of course.
19:00 jim__ joined #salt
19:00 n8n joined #salt
19:00 jim__ is anyone else having trouble with cron.file on the new 2014.1.0?
19:00 jim__ it can't find my source file
19:01 johtso joined #salt
19:01 jim__ I'm running everything locally, no servers
19:02 zain_ joined #salt
19:02 johtso joined #salt
19:02 scoates pull request is merged, btw, forrest
19:03 freerod vxitch: I see the problem.  If you edit the file in question and change the ] to a } on line 490, I suspect your problem will be resolved.
19:03 bitmand forrest, salt minion 0.17.5 just died twice - last thing in log both times: Returning information for job: 2014030... :(
19:03 djinni` joined #salt
19:04 wendall911 joined #salt
19:04 freerod There are also a few mixups when mapping the family in the formation of comments elsewhere in the file, but those shouldn't have any significant impact.
19:05 kermit joined #salt
19:05 amckinley joined #salt
19:06 bitmand forrest: and it seems to die/shut down after the state.highstate ( last lines is INFO about how the 3 directories is in the correct state ( file.recurse ).
19:07 bitmand ( and before it even get's to do the cmd.run it is supposed to run afterwards )
19:07 oz_akan_ joined #salt
19:09 bitmand it's just wierd :/
19:13 oz_akan_ joined #salt
19:13 ajw0100 joined #salt
19:16 pentabular joined #salt
19:18 hardwire yay for new salt in debian
19:18 scoates "yay"
19:18 vxitch freerod: I've been fighting this for 3 or 4 days now
19:18 scoates (-;
19:18 vxitch freerod: THANK YOU so much
19:18 hardwire http://debian.saltstack.com/debian/pool/main/s/salt/
19:20 ndrei joined #salt
19:20 hardwire so I've seen a few things like salt-ssh like python-execnet
19:21 hardwire where code is exported.
19:21 fxhp Heartbane - robawt : fxhp ~= foxhop (or effects hitpoints)
19:21 freerod vxitch: very happy it helped.
19:21 fxhp Heartbane - robawt - : P
19:22 vxitch freerod: can't find an issue for this bug in github, gonna open one and send a pull request
19:22 robawt :D
19:23 Gareth vxitch: was this with the iptables state?
19:23 freerod vxitch:  I was just about to do the same thing when I noticed that it was fixed just after 2014.1.0 release.
19:23 jacksontj_ joined #salt
19:24 vxitch Gareth: yep
19:24 vxitch freerod: oh okay, even easier then
19:25 vxitch so there's a new ver? or just the dev version?
19:25 krow joined #salt
19:25 Gareth 2014.1 is the new version.
19:26 vxitch ah yeah, i see in the latest dev version the } issue is fixed
19:26 generj joined #salt
19:26 generj can somebody help me with gitfs?
19:27 generj I want to pull in my repo but its simply not grabbing it
19:27 freerod There are several newer tagged versions but I don't thing any have been officially distributed. So yes basically dev.  Ah, as you've seen.
19:28 generj I get erros like: Comment: Source file salt://mydjangoappfolder/ not found
19:28 generj *errors
19:28 nextdoorwarren joined #salt
19:29 themadcanudist hey guys, any way to have a highstate or state.sls output the resultant yaml from a .sls template rendering? ie. to debug?
19:31 vxitch themadcanudist: check out the output modules available with salt, and then set which one to use with --output= when calling salt
19:31 vxitch http://docs.saltstack.com/ref/output/all/index.html
19:31 generj if I just do a "file.directory: -name: /var/www/app/ -source: salt://app/ -makedirs: True"
19:31 generj it does make the directory
19:31 generj but doesn't fill it with the repo contents
19:31 vxitch # salt --output=yaml_out '*' state.sls mystate
19:31 generj if its "file.managed" it does nothing
19:31 themadcanudist vxitch: i did, but I may be misunderstanding… it doesn't output the yaml from a rendered .sls, it outputs all the salt-call output in a particular format
19:31 vxitch try state.show_sls instead
19:32 themadcanudist k,
19:32 vxitch the state execution module has really useful debugging tools. things like show_top, show_sls, show_highstate, etc
19:33 generj vxitch: are you talking to me? :)
19:34 vxitch no, to themadcanudist. but we can talk too, i'm not opposed to a poly relationship
19:34 vxitch <3
19:35 generj vxitch: aww, I accept, heh
19:35 generj <3
19:35 generj did you get my gitfs question? I just want it to pull in my repo
19:35 vxitch i did, but i dont know anything about gitfs unfortunately
19:36 vxitch did you read all the docs on gitfs and using it in states?
19:36 generj vxitch: these docs: http://docs.saltstack.com/topics/tutorials/gitfs.html
19:36 generj but they just let you specify ONE file
19:36 generj which it will source from salt://whatever and grab from git
19:36 generj I want to pull in the whole FOLDER :)
19:38 sporkd2 i thought gifts used the entire repo unless you specified
19:38 vxitch ahhhhhh....crap well i have no idea other than specify 'folder/*' or 'folder'
19:38 AdamSewell_ joined #salt
19:39 generj sporkd2: me too!
19:39 sporkd2 what did you specify in your fileserver_backend?
19:40 allanparsons joined #salt
19:40 allanparsons i was wondering if there was an acceptable way to have a minion delete itself from a master
19:40 generj sporkd2: /etc/salt/master fileserver_backend: -roots -git
19:40 generj sporkd2: just like the docs say
19:40 allanparsons like, if an ec2 instance terminates, i remove the dns entry from Route53
19:41 allanparsons and if a new server comes online, i'll take over that available hostname (prod-myhost-001.mydomain.net)
19:41 allanparsons but i get a key error unless i delete the key from the master
19:41 vxitch Gareth, freerod: in case anyone comes in here asking why iptables.flush or iptables.set_policy throws py errors in the future. the {2] typo is present in both of those functions in salt/states/iptables.py . the first at line 490, the second around 440. the current dev release has this fixed. thanks everone for all the help :)
19:41 themadcanudist vxitch: thanks… i just noticed that rendered output of the .sls files shows up on minion debug logs
19:42 oz_akan__ joined #salt
19:42 amckinley joined #salt
19:42 scoates allanparsons: guessing here, but I bet there's a key conflict / auth fail event that you could listen for, and maybe you could have reactor (or just a python script) remove the key
19:43 allanparsons the salt-call on load will fail then
19:43 allanparsons i def need that salt-call to go throuhg the first time an instance boots (because these are part of an autoscale group)
19:43 * scoates nods
19:44 scoates not sure of an elegant solution… is this VPS or public?
19:44 scoates *VPC
19:45 edlane joined #salt
19:45 seapasulli joined #salt
19:45 allanparsons VPC
19:45 allanparsons and i start everything w/ cloudformation scripts
19:45 JasonSwindle left #salt
19:45 atealtha joined #salt
19:46 scoates this needs very careful consideration, but *maybe* open_mode?
19:46 allanparsons i use open mode
19:46 allanparsons and lock down by security groups
19:46 allanparsons but, i re-use hostnames
19:46 scoates ah. didn't realize it still stored keys the same way. sorry, then.
19:47 allanparsons so if i have prod-001, prod-002, prod-003.  if prod-002 dies/is terminated, it gets removed from route53.  the next server that comes up will see that prod-002 is an avail hostname and register itself / rename itself
19:47 allanparsons but i'll have a key mismatch on salt-call for prod-002
19:48 * scoates nods
19:49 edlane seeking feedback on this issue: https://groups.google.com/forum/?hl=en&amp;fromgroups#!topic/salt-users/fTpLOzflXy4
19:54 funzo joined #salt
19:55 Guest94499 joined #salt
19:55 Guest94499 تحذير
19:55 Guest94499 warning          you may be  watched
19:55 Guest94499 do   usa & israel   use  the internet ( youtube, facebook, twitter, chat rooms,skype,ect...)to collect informations,,,,can we call that spying ?
19:55 Guest94499 do they record&analyse everything you type????
19:55 Guest94499 why they ask for ur name,age,gender,occupation,place of work,ur mobile number.............ect, can they use these informations harming you?!!
19:55 Guest94499 warning          you may be  watched
19:55 Guest94499 do   usa & israel   use  the internet ( youtube, facebook, twitter, chat rooms,skype,ect...)to collect informations,,,,can we call that spying ?
19:55 Guest94499 do they record&analyse everything you type????
19:55 Guest94499 why they ask for ur name,age,gender,occupation,place of work,ur mobile number.............ect, can they use these informations harming you?!!
19:55 Guest94499 warning          you may be  watched
19:55 Guest94499 do   usa & israel   use  the internet ( youtube, facebook, twitter, chat rooms,skype,ect...)to collect informations,,,,can we call that spying ?
19:55 Guest94499 do they record&analyse everything you type????
19:55 Guest94499 why they ask for ur name,age,gender,occupation,place of work,ur mobile number.............ect, can they use these informations harming you?!!
19:55 Guest94499 warning          you may be  watched
19:56 meteorfox joined #salt
19:57 funzo joined #salt
19:57 naiyte Wow, I sure am glad Guest94499 warned me that I'm being watched. I would have no idea!
19:57 Guest94499 joined #salt
19:57 funzo joined #salt
19:57 ipalreadytaken joined #salt
19:57 juniorzoid joined #salt
19:58 Guest94499 joined #salt
19:59 forrest lol
20:00 Guest94499 joined #salt
20:00 Guest94499 left #salt
20:01 KyleG joined #salt
20:01 KyleG joined #salt
20:01 seanz Can someone give me a really quick overview on what states.cmd.mod_watch does exactly?
20:01 seanz Will it only execute a command when a command it's watching actually changes something?
20:02 funzo joined #salt
20:03 sporkd2 seanz yes, if the watched state changes
20:03 seanz sporkd2: Thank you.
20:03 funzo joined #salt
20:03 sporkd2 seanz: mod_watch is really just an alias for cmd.run
20:04 seanz sporkd2: I thought cmd.run would run its commands every time.
20:04 frasergraham joined #salt
20:04 andrej Hmmm ... the new release 2014.1 has a non-critical flaw that irks me :)
20:04 andrej the startup scripts leave a zombie dangling - this didn't use to happen
20:04 forrest andrej, someone brought that up earlier this morning as well
20:05 forrest lots of zombie processes.
20:05 andrej Happens on ubuntu 12.04 and Centos 6.3
20:05 forrest andrej, https://github.com/saltstack/salt/issues/10867
20:05 andrej Thanks forrest
20:05 forrest andrej, yea np, can you add onto that (or open a new issue) talking about the startup scripts themselves leaving zombie processes?
20:07 ndrei joined #salt
20:07 ajprog_laptop joined #salt
20:12 freerod left #salt
20:12 JordanRinke joined #salt
20:15 hunter joined #salt
20:19 jeffro joined #salt
20:19 JordanRinke joined #salt
20:19 [diecast] joined #salt
20:23 Networkn3rd joined #salt
20:25 whiskybar joined #salt
20:25 Ryan_Lane joined #salt
20:25 vlcn anyone know how I would go about using event.fire_master with the python renderer?
20:25 andrej Forrest - kind of did straight after you posted the link
20:25 forrest cool
20:27 scoates vlcn: __salt__['event.fire_master']({'data': 'here'}, 'tag') ?
20:27 n0c joined #salt
20:28 joehoyle joined #salt
20:28 n0c hi, is it possible to use grains to match in the 'peer communication' configuration
20:28 alunduil joined #salt
20:28 n0c i.e., http://salt.readthedocs.org/en/latest/ref/peer.html
20:28 vlcn scoates, that looks right?  Am I missing some documentation somewhere?
20:29 scoates I don't understand what's wrong.
20:29 dave_den n0c: no. only by regex of the minion id.
20:29 aaroneous joined #salt
20:30 zain_ joined #salt
20:31 n0c that seems like something worth implementing, me thinks
20:31 bhosmer joined #salt
20:32 dave_den n0c: it should be possible to implement, since peer communication still goes through the master
20:32 whiskybar left #salt
20:32 themadcanudist guys, using the ssh_auth salt state, is there a concept in this state or in others, where it will ensure that ONLY the ssh keys that have been specified as "present" exist in the file, anything that isn't, it implicitly absents?
20:33 krow joined #salt
20:33 themadcanudist if not, is there an excepted "salty" pattern that is recommended to accomplish that?
20:34 themadcanudist ie. can I "suck in" the current ssh_auth file and have the entries available as objects that I can compare to what I want provisioned>
20:34 jhauser joined #salt
20:36 [diecast] joined #salt
20:40 frasergraham if i'm writing custom runners, where do I put them?
20:42 jab416171 joined #salt
20:42 n0c I'm also wondering that, frasergraham
20:42 frasergraham it says the "runners" folder on this page https://salt.readthedocs.org/en/latest/ref/runners/index.html?highlight=runner
20:44 frasergraham but where is that?
20:44 frasergraham can it be custom? I want it in source control
20:46 n0c frasergraham: I just found "runner_dirs" in the example master configuration file
20:46 n0c if there's no default, that could be a good place to start
20:46 frasergraham ok, cool - i'll give that a go
20:46 joehoyle joined #salt
20:49 themadcanudist let me rephrase. during a highstate, can you request all provisioned ssh_auth states in a data structure to make rendering decisions?
20:49 [diecast] joined #salt
20:50 rojem joined #salt
20:51 kainswor @frasergraham we use a vc'd runners dir. infact our file-root is a git repo. (we don't use gitfs for other integration issues, but all your modules/extensions can be controlled well this way)
20:52 n0c kainswor: out of curiousity, where did you opt to put it in relation to the state roots and such?
20:52 n0c I'm trying to decide
20:54 maxleonca joined #salt
20:55 kainswor it's within the state-roots
20:55 maxleonca Hello, someone that can give a perl of knowledge with salt-cloud and ec2?
20:55 kainswor we also dynamically manage the top file, so we don't have to worry about conflicts there
20:56 maxleonca having trouble with an AMI that starts with just 1GB of space on the root device, but I have not figure out how to make it larger
20:56 kainswor otherwise, I'd probably have them in different directories at the root of the repo
20:57 dave_den i would not put runners in your file_roots path
20:58 n0c I'm debating /srv/runners, since /srv/pillars and /srv/salt are defaults
20:59 dave_den by default salt-master looks for runners in /etc/salt/runners
21:01 n0c awesome, thanks!
21:01 n0c I dunno if that's documented anywhere?
21:01 themadcanudist hey dave_den, sorry to harass you daily ;), but you seem well versed. Any insight in to my question above?
21:01 zain_ joined #salt
21:02 themadcanudist i posted it to the mailing list, but in the meantime, if anyone has some thoughts (or not)?
21:02 dave_den themadcanudist: i would personally just manage each user's .authorized_keys file as a jinja tempalte using keys from pillar.
21:02 themadcanudist dave_den: that's what I do
21:02 krow joined #salt
21:02 themadcanudist However, I'm thinking of some more advanced functionality
21:02 themadcanudist that I've built in other cfg managers
21:02 cewood joined #salt
21:03 dave_den what would be the difference in using ssh_auth.present versus a file.managed?
21:04 korylprince joined #salt
21:05 themadcanudist dave_den: that is a good point, it's just a few well defined columns
21:05 themadcanudist i hear you
21:05 abe_music joined #salt
21:05 themadcanudist a *very* good point
21:05 themadcanudist much simpler.
21:05 themadcanudist thx again
21:05 zooz joined #salt
21:05 vu joined #salt
21:05 n0c dave_den: /etc/salt/runners isn't working for me, is that a new thing?
21:05 dave_den if you wanted to use ssh_auth.present, you can build the a temporary .authorized_keys file, loop over your ssh_auth.present into that file, then move that file into the correct place when it's built
21:05 themadcanudist dave_den: that's what i've done previously
21:05 themadcanudist but it's not as nice as a file.managed
21:06 themadcanudist as diffs would be shown properly
21:06 themadcanudist crystal clear
21:06 abe_music UtahDave, noticed that one of last pull requests from 30 days ago (https://github.com/saltstack/salt/pull/10154) was merged, but it's not in the v2014.1.0 release?
21:06 korylprince Hello all. I'm trying to figure out how to use file.directory and file.recurse in a salt state. Here's my attempt: http://pastie.org/8877871  What am I doing wrong/is there a better way to achieve this?
21:07 korylprince (With that state I get 'Too many functions declared in state "file" in sls backup')
21:07 vu joined #salt
21:07 dave_den n0c: try /etc/salt/_runners
21:08 UtahDave abe_music: Hm. If it's not in there, could you add a comment asking basepi to cherry pick it?
21:08 Ryan_Lane joined #salt
21:08 cpenner461 joined #salt
21:09 n0c dave_den: nope, they only work if I explcitly define them in /etc/salt/master runner_dirs
21:09 abe_music UtahDave: sure thing
21:09 n0c using 2014.1.0
21:09 dave_den n0c: hrm, then yeah it may depend on where you point extmods and runner_dirs config in your master
21:10 ipmb joined #salt
21:10 n0c alrighty, thanks :)
21:11 dave_den extmods defaults to cachedir + 'extmods/', which has always seemed odd to me
21:11 UtahDave thanks, abe_music!
21:11 xt joined #salt
21:12 dave_den or rather 'extension_modules' defaults to cachedir + 'extmods'
21:14 jeffro joined #salt
21:16 travis joined #salt
21:17 basepi abe_music: yep, i'll grab that when i cherry-pick next.
21:17 Guest53477 Is a 2014.1.0 minion incompatible with 0.17.4 master?
21:18 forrest Guest53477, yes
21:18 abe_music basepi: thanks :) curious what happened there?
21:18 basepi Nobody pinged me that it was a bugfix, so I missed it when I was cherry-picking.
21:18 abe_music makes sense haha, thanks again
21:18 basepi No worries.
21:19 Guest53477 So I have to update my master, then minions?
21:19 sgviking joined #salt
21:19 forrest Guest53477, usually you'd want to update your minions using your master, then update the master.
21:20 forrest Guest53477, there are multi-master issues on 2014.1, so if you use that functionality, stick with 0.17.5
21:20 Guest53477 I just spun up aserver and pulled the latest package form EPEL, key-exchanges when well but highstte failed and then I ntoiced what had happened
21:21 Guest53477 I'm probably just going to downgrade the minion for now as I can't update all servers at the moment.
21:22 forrest Guest53477, sounds good
21:22 skynet joined #salt
21:22 Guest53477 thanks for the help
21:23 forrest you're welcome
21:23 jhauser_ joined #salt
21:23 [diecast] joined #salt
21:23 TheRealBill joined #salt
21:24 Guest53477 crap EPEL only has 2014.1.0-1, they don't seem to keep around old versions...
21:25 hardwire I had the opposite problem in debian until a few days ago
21:25 hardwire thankfully the salt bootstrap can get you what you need quickly
21:25 hardwire any version
21:25 abe_music joined #salt
21:29 druonysus joined #salt
21:29 druonysus joined #salt
21:29 quickdry21 joined #salt
21:30 pentabular left #salt
21:30 Guest53477 Is breaking backwards compatibility every release going to be a normal thing or is this just growing pains?
21:31 n8n joined #salt
21:31 seapasulli joined #salt
21:32 quantumriff joined #salt
21:32 Thiggy joined #salt
21:33 EWDurbin Guest53477: http://kojipkgs.fedoraproject.org/packages/salt/ could help you
21:34 bemehow joined #salt
21:34 Guest53477 Thanks EWDurbin I didn't think to check Koji, I already started with the salt-bootstrap since this is a temporary server
21:34 harobed joined #salt
21:35 EWDurbin Guest53477: TBH I'm curious too where backwards compatibility stands as well, as salt completely glossed past the elusive 1.0.0 release
21:35 harobed joined #salt
21:36 harobed joined #salt
21:37 harobed joined #salt
21:37 vxitch how are the states applied? in what order?
21:37 johtso joined #salt
21:37 harobed joined #salt
21:37 frasergraham joined #salt
21:38 Guest53477 EWDurbin: thanks for the koji suggestion, bootstrap had some problem and I just grabbed the RPMs and install went well, I'm going to have to decided to upgrade or put these rpms into our private repos
21:38 harobed joined #salt
21:38 vxitch im not sure where to put my iptables.flush rule so it doesnt nuke everything, and where to put the iptables.set_policy drop rule so i dont lose access before the rest of the rules are applied
21:42 n0c so I'm trying to configure haproxy to direct traffic to a bunch of backend web servers based on the host header.. what would be the proper way to get minions to report back thier ip address along with all the enabled nginx vhosts they've got installed
21:42 n0c I guess part of my question, because I've looked at both of these things, is whether or not an execution module or a runner is a better place to start
21:45 [diecast] joined #salt
21:45 amckinley joined #salt
21:45 helderco joined #salt
21:45 schimmy joined #salt
21:46 jcockhren n0c: so you're trying to get out the ip addresses of the minions and have it injected in a config file?
21:46 schimmy1 joined #salt
21:47 n0c more than that
21:47 n0c I need the IP, and a list of hostnames
21:47 jcockhren n0c: salt mine comes to mind about the ip addresses
21:47 n0c because I won't have the same sites on all my web minions
21:48 jcockhren as for the vhosts, I guess there's two ways:
21:48 jcockhren 1. the DNS way of matching A records to ip addresses
21:48 jcockhren or
21:48 teskew how do you initiate salt-cloud now that it's integrated?
21:49 timoguin teskew, salt 'configuredminion' cloud.function args
21:49 jcockhren 2. have the haproxy minion call the command module to get all the files in sites-enabled for nginx
21:49 timoguin just a regular exec module now
21:49 teskew cool
21:49 dave_den n0c: i wrote MasterPillarUtil in utils.master.py to achieve similar goals. i would use that in a runner that your haproxy minion can call, and the runner will return a data structure of the IPs of the nginx servers and their vhosts, as you have them defined in pillar.
21:49 teskew thanks timoguin
21:50 teskew so, no maps?
21:50 vxitch how is the top file processed? the docs explain how different environments are treated, but not in which order states are applied
21:51 n0c dave_den: eh, I'm assigning vhosts to minions in my top file
21:51 n0c is there a better way to do that?
21:52 zooz joined #salt
21:52 joehoyle joined #salt
21:53 n0c my file_roots top, mind you
21:53 n0c i.e., not pillar
21:53 dave_den n0c: you mean via your states top file? are they all static configs files?
21:53 darrend joined #salt
21:53 n0c lemme see if I can't pull a few examples out
21:54 dave_den you should leverage pillar
21:55 dave_den if you're only defining the vhost configs to be applied in your state tops, you are losing the dynamic ability of salt and pillar
21:56 n0c dave_den: here's my top file: http://sprunge.us/KHbH
21:56 timoguin teskew, there is a cloud state module now to, so most of that can go in actual SLS files.
21:56 n0c dave_den: and here's an example vhost sls: http://sprunge.us/PWNY
21:56 timoguin teskew, i want to say the salt-cloud command can still be used though, but i may be wrong about that
21:56 n0c (I love that that one came out as pwny)
21:57 n0c I want to use pillar as much as possible, but I dunno how I'd say 'put this vhost on this minion' in pillar
21:57 n0c that seems to be what the top file does
21:57 [diecast] joined #salt
21:59 cpenner461 joined #salt
22:02 dave_den n0c: you would just assign vhosts in the pillar top to minions, similar to waht you do in the state top, but your state files would be templated so they can iterate over the pillar data to apply configurations for each vhost
22:02 krow joined #salt
22:02 dave_den {% set hostname = sls.replace("vhosts.","").replace("_",".") %}
22:02 n0c :)
22:03 joehoyle joined #salt
22:03 dave_den instead of trying to interpret the hostname from the name of the sls file itself, assign vhost names in pillar and interate over them in the state file, using {% for hostname in pillar['vhosts'] %}
22:03 dave_den iterate
22:03 ndrei joined #salt
22:04 dave_den DRY yourself off
22:04 xmltok joined #salt
22:06 timoguin teskew, I just used salt-cloud 0.8.11 successfully with a 2014.1.0 master and minion, so map files should still work that way if you don't want to use the states
22:07 n0c so I started down that path before daveden, and for some reason I came to the conclusion that it wasn't flexible enough
22:07 n0c BUT I've picked up quite a bit in the last couple weeks, so I'll revisit that idea
22:07 n0c dave_den: *
22:09 xmltok_ joined #salt
22:09 hunter How do people manage ssh keys for hosts?
22:09 hunter If I rebuild a  host, I hate changing the ssh keys on users.
22:09 hunter Do you have salt pregen keys for new hosts, or suck them back from existing hosts or ... what?
22:10 rgarcia_ joined #salt
22:11 n0c /technically/ this makes clever use of symlinks, so it's still DRY-ish ;)
22:11 timoguin hunter, I have a users pillar defined, and I put the public keys in for each user.
22:12 joehoyle joined #salt
22:12 timoguin the users-formula will put the key in place if it's defined in pillar
22:13 hunter timoguin: that would be useful as well, but I meant host keys
22:14 joehoyle joined #salt
22:14 dave_den hunter: you could make use of minionfs for that.
22:17 hunter dave_den: I've not heard of minionfs yet. I was thinking salt mine, but I can't find any docs on it and am not sure what all is required/usable there.
22:18 dave_den minionfs allows your minions to push files up to the master, where they are accessible at salt://<minion_id>/
22:19 yomilk joined #salt
22:19 dave_den the problem would be that you would also need to push your private keys, and minionfs is readable from all minions, i believe
22:22 diegows joined #salt
22:22 hunter That would be an issue
22:23 hunter Are there docs for minionfs anywhere?
22:24 timoguin http://docs.saltstack.com/topics/tutorials/minionfs.html
22:24 hunter and is that in 0.17.5?
22:25 sroegner joined #salt
22:25 dave_den no, it's in 2014.01
22:27 hunter shucks
22:28 hunter what about the salt mine - does it have the same issue with private keys?
22:29 dave_den yes, and the mine only stores output from module functions
22:30 manicouman joined #salt
22:33 herzi_ joined #salt
22:33 andersb hunter: I used the frame of : https://github.com/brutasse/states/tree/master/certs for awhile. Decent enuff to get you started?
22:34 manicouman Hola... I'm looking for a very good beginner's tutorial for vagrant/saltstack
22:34 druonysuse joined #salt
22:34 druonysuse joined #salt
22:35 andersb hunter: Sorry, worng link.
22:35 zain_ joined #salt
22:36 helderco joined #salt
22:38 andersb hunter: Hmm: https://github.com/uggedal/states/tree/master/users
22:39 andersb hunter: Combined with some thing like: https://github.com/brutasse/states/tree/master/git should get you going. It's pillarbased.
22:39 analogbyte joined #salt
22:40 Steve_ joined #salt
22:41 Steve_ Hi All... how does one escape double curly braces in jinga templates?
22:42 chrisjones joined #salt
22:42 skynet joined #salt
22:44 Steve_ Thoughts?
22:44 Steve_ I want to output a template which supervisord can process
22:45 dave_den Steve_: http://jinja.pocoo.org/docs/templates/#escaping
22:45 Steve_ Cool.. thanks for that Dave
22:45 dave_den np
22:47 kermit joined #salt
22:48 vxitch as far as ordering goes, what's the easiest and cleanest way to ensure iptables.flush executes before any other iptables states? id like to avoid setting order: 1
22:49 vejdmn joined #salt
22:50 codysoyland so I'm having trouble with the new release. I was using salt-cloud to provision new nodes, but I can't figure a way of making salt-cloud use an older version. so I upgraded salt-master to 2014.1.0 but now it times out and gives an authentication error. "Salt request timed out. If this error persists, worker_threads may need to be increased" and "Failed to authenticate, is this user permitted to execute commands"
22:51 codysoyland I increased worker_threads and created an external_auth setting with myself (was not using that before)
22:52 Ryan_Lane joined #salt
22:52 codysoyland on that note, I'm really getting frustrated with salt-cloud failing every time a salt release comes out...
22:52 krow joined #salt
22:52 btorch so is there a way to view what the order will really be when executing highstate ? I have requisites in place and I thought things were in a quite a good order but when actually running highstate it's not working out so great
22:53 btorch I have read over http://docs.saltstack.com/ref/states/layers.html?highlight=compiled
22:55 bhosmer joined #salt
22:56 vu_ joined #salt
22:56 vu joined #salt
22:56 dave_den btorch: state.show_highstate should be what gets executed, and you will see the orders in the output
22:56 Steve_ Has anyone got  start_action: state.highstate to work on EC2  (with salt-cloud) ?
22:56 manicouman joined #salt
22:57 btorch dave_den: yeah I checked that out and the order just seems to get crazy
22:57 honestly_ /ww/28
22:57 honestly_ augh
22:58 dave_den btorch: there was this, for highstate order: https://github.com/saltstack/salt/pull/10873
23:01 Steve_ Also it seems that an old version of the minion gets installed by default (i'm using the RC1 but 17.5 gets installed)
23:03 timoguin btorch, dave_den, state.show_lowstate should have the final word for the ordering
23:03 amckinley joined #salt
23:05 dave_den timoguin: ah, yes you are correct
23:05 themadcanudist joined #salt
23:06 HeadAIX joined #salt
23:09 dancat joined #salt
23:11 zain joined #salt
23:11 manicouman joined #salt
23:15 oz_akan_ joined #salt
23:21 schimmy joined #salt
23:24 schimmy1 joined #salt
23:27 rojem joined #salt
23:27 vbabiy joined #salt
23:27 alunduil joined #salt
23:30 Ryan_Lane joined #salt
23:33 druonysus joined #salt
23:33 druonysus joined #salt
23:43 rojem joined #salt
23:43 krow joined #salt
23:45 [M7] joined #salt
23:45 oz_akan_ joined #salt
23:47 christopherl joined #salt
23:47 rgbkrk_ joined #salt
23:47 christopherl I'm getting a really strange jinja error, can someone help me out? http://pastebin.com/Z3D44Fvb
23:48 sporkd2 joined #salt
23:48 allanparsons any idea why auto-accept wouldnt be working?
23:48 allanparsons root@dev-salt:~# cat /etc/salt/master | grep auto_accept
23:48 allanparsons auto_accept: True
23:48 pydanny joined #salt
23:50 conan_the_destro joined #salt
23:51 ndrei joined #salt
23:52 zain_ joined #salt
23:55 KyleG1 joined #salt
23:58 krow joined #salt
23:58 rojem joined #salt
23:59 rojem joined #salt

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