Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-04-22

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

All times shown according to UTC.

Time Nick Message
00:00 bhosmer joined #salt
00:00 CeBe okay, will take this one then. Thanks Ryan_Lane!
00:00 Ryan_Lane from within an execution module, is it recommended to call functions in the same module directly, or via __salt__?
00:01 calvinhp_mac joined #salt
00:01 gw joined #salt
00:01 gw joined #salt
00:02 doanerock Ryan_Lane, aha the ppa:saltstack/salt  version of salt-call appear to work much better
00:02 Ryan_Lane great
00:03 joehh1 really good to hear :) some packaging updates over the weekend and was worried I had broken something
00:05 jtang1 joined #salt
00:06 sealibora joined #salt
00:10 diogobaeder left #salt
00:13 druonysus joined #salt
00:13 druonysus joined #salt
00:13 srage joined #salt
00:15 druonysus joined #salt
00:15 druonysus joined #salt
00:16 toastedpenguin joined #salt
00:16 srage_ joined #salt
00:18 ajprog_laptop joined #salt
00:20 taion809 joined #salt
00:21 Luke_ joined #salt
00:21 joehoyle joined #salt
00:21 srage joined #salt
00:24 srage_ joined #salt
00:24 amjith joined #salt
00:25 shoma joined #salt
00:26 FL1SK joined #salt
00:27 NullWagon joined #salt
00:29 srage joined #salt
00:30 ninkotech joined #salt
00:32 joehoyle joined #salt
00:33 dccc joined #salt
00:38 NullWagon joined #salt
00:38 bhosmer joined #salt
00:39 ewong_ joined #salt
00:40 NullWagon joined #salt
00:49 l0x3py joined #salt
00:56 druonysus joined #salt
00:56 druonysus joined #salt
00:59 jtang1 joined #salt
01:03 joehoyle joined #salt
01:03 mgw joined #salt
01:04 jeffg_ joined #salt
01:08 jeffg_ so... quick question; maybe
01:08 jeffg_ can salt set or update ldap attributes
01:09 jeffg_ or, is there a better way to manager user passwords?
01:09 cruatta joined #salt
01:10 conan_the_destro joined #salt
01:11 srage_ joined #salt
01:12 HeadAIX joined #salt
01:20 druonysus joined #salt
01:20 druonysus joined #salt
01:20 NullWago_ joined #salt
01:22 jtang1 joined #salt
01:22 linuxlewis joined #salt
01:28 beardo joined #salt
01:29 mgw joined #salt
01:40 gw Hi is it possible to make salt-run jobs.print_job produce colorised output with in a shell?
01:41 doanerock joined #salt
01:45 srage joined #salt
01:46 schimmy joined #salt
01:47 cowyn joined #salt
01:50 mgw joined #salt
01:50 ifnull joined #salt
01:58 garthk winrepo: how do I deal with packages which don’t report their version?
01:59 garthk pkg.list_pkgs reports stunnel: Not Found
02:00 srage_ joined #salt
02:00 meteorfox joined #salt
02:03 kickerdog1 left #salt
02:07 Daemonik joined #salt
02:08 mateoconfeugo joined #salt
02:09 jeremyfelt joined #salt
02:09 terminalmage garthk: what platform? pkg.list_pkgs shouldn't be saying "Not Found"
02:10 garthk WS2003. Not sure if it has the WMI component installed.
02:10 terminalmage I think all the required modules are included in a frozen build
02:10 NullWagon joined #salt
02:11 garthk On top of that, pkg.refresh_db is giving me “unable to cache file ‘salt://win/repo/winrepo.p'” and I can’t install an updated salt-master because it depends on an older salt-common than the one installed for salt-minion.
02:11 terminalmage garthk: have you generated your repo file?
02:12 garthk ok, uninstalling and then reinstalling both worked
02:12 garthk terminalmage: yes
02:12 terminalmage ok, might want to email the mailing list then, I'm not as familiar with the winrepo stuff
02:12 garthk terminalmage: is there some configuration of the master required that I missed in the docs somehow?
02:13 terminalmage no, I think that when you generate the repo file, it places it somewhere in the cachedir
02:13 terminalmage and it pulls from there when you need it
02:14 terminalmage winrepo.p is just a msgpack-serialized dictionary which tells win_pkg.py where to download the exe and how to install it unattended
02:14 mgw joined #salt
02:14 garthk terminalmage: naah, it puts the .p file in /srv/salt/win/repo/ which is not file-served
02:14 terminalmage ahh, ok
02:15 srage joined #salt
02:15 garthk garthk: which I’m sure I could hack around, but I want to know why the behaviour and docs are apparently out of sync first; see also https://github.com/saltstack/salt/issues/10894
02:15 terminalmage that might be something to fix in the winrepo.genrepo runner
02:15 terminalmage adding an argument for the destination of the repo file
02:16 Luke_ joined #salt
02:16 garthk … except this is Salt Minion 2014.1.3
02:17 jtang1 joined #salt
02:17 garthk bah
02:17 garthk it can wait until after lunch
02:18 cowyn_ joined #salt
02:19 terminalmage yeah, I'd email the mailing list, it's easier to get a response say, 2-3 hours from now, than it would be if you left this question in the IRC only
02:19 terminalmage and there are people that don't use the IRC but do check the mailing list
02:22 TyrfingMjolnir joined #salt
02:28 HeadAIX joined #salt
02:29 Ryan_Lane joined #salt
02:29 Luke_ joined #salt
02:30 srage_ joined #salt
02:35 TyrfingMjolnir joined #salt
02:36 anuvrat joined #salt
02:54 xl1 joined #salt
03:03 anuvrat joined #salt
03:04 TyrfingMjolnir joined #salt
03:06 jalbretsen joined #salt
03:08 ajw0100 joined #salt
03:11 jtang1 joined #salt
03:12 catpiggest joined #salt
03:13 jtang2 joined #salt
03:23 thayne joined #salt
03:24 andrej I have a wildcard SSL key/cert pair that I need to push to 5 servers.  Given the nature of the files I was wondering whether they need to go into pillar, and whether that's even possible?
03:24 andrej Or, the other way round: if I put the files into a state directory, how would I stop arbitrary other minions from accessing the keys?
03:25 Daemonik andrej, Look in to Pillar and ACLs
03:26 zain joined #salt
03:32 andrej Daemonik - I *am* looking at pillar, and can't see a way to have a file being populated via pillar
03:35 schimmy joined #salt
03:36 andrej another question ... what is housekeeping best practice?  /var/cache/salt/master/jobs is now close to 3GB in size, iss there a config option to e.g. get rid of entries older than a month? :)
03:40 andrej hmmm ... it would appear that it phases out files older than 2 weeks
03:41 mateoconfeugo joined #salt
03:41 NullWagon joined #salt
03:44 jeffg__ joined #salt
03:46 jeffG___ joined #salt
03:48 meteorfox joined #salt
03:48 |o||o| joined #salt
03:50 Gareth joined #salt
03:54 pydanny joined #salt
03:56 ravibhure joined #salt
04:00 faldridge joined #salt
04:04 Gareth joined #salt
04:04 ravibhure joined #salt
04:07 jtang1 joined #salt
04:09 amjith joined #salt
04:10 stevendgonzales joined #salt
04:14 joehoyle joined #salt
04:19 smcquay joined #salt
04:20 Gareth [I/w 33
04:24 garthk Back to my error of “unable to cache file ‘salt://win/repo/winrepo.p' from saltenv ‘base’.”
04:30 chrismoos joined #salt
04:30 chrismoos Has there been any discussion about getting something in Salt for managing formula/state dependencies, similar to Berkshelf with Chef?
04:38 abe_music joined #salt
04:42 stevendgonzales joined #salt
04:45 srage joined #salt
04:47 srage joined #salt
04:47 ravibhure1 joined #salt
04:48 ndrei joined #salt
04:52 ravibhure joined #salt
04:53 cruatta joined #salt
05:01 jtang1 joined #salt
05:17 anuvrat joined #salt
05:35 ndrei joined #salt
05:37 otsarev_ Any comments? https://groups.google.com/forum/#!topic/salt-users/8XmvQHGDBT8
05:41 srage joined #salt
05:41 Daemonik I am attempting to run salt-master 2014.1.1 in a SmartOS Zone (base64 13.4.0) and my seemingly very correct /etc/salt/master file is unparseable? http://pastebin.com/raw.php?i=RAG3T8dL
05:52 crazysim joined #salt
05:52 ramteid joined #salt
05:55 CeBe joined #salt
05:55 jtang1 joined #salt
05:55 crazysim joined #salt
05:57 __number5__ Daemonik: you need to paste your /etc/salt/master too
05:57 jasonrm joined #salt
05:59 individuwill joined #salt
06:01 sifusam joined #salt
06:04 logandg joined #salt
06:06 Daemonik __number5__, first ten lines of /etc/salt/master http://pastebin.com/raw.php?i=YGjLTNDY
06:11 fsniper is it possible to refer a state from another environment? like: prod: host1: state*dev*
06:16 djinni` joined #salt
06:20 icarus joined #salt
06:20 mekstrem joined #salt
06:20 __number5__ Daemonik: if you can't paste the whole file, check the line 262, column 2:
06:21 __number5__ jinja_lstrip_blocks: False part
06:21 __number5__ maybe that's an idention you don't want
06:24 homelinen joined #salt
06:24 homelinen joined #salt
06:25 epcim joined #salt
06:28 amjith left #salt
06:35 srage joined #salt
06:41 schimmy joined #salt
06:42 epcim_ joined #salt
06:49 jtang1 joined #salt
06:51 jtang2 joined #salt
06:52 gammalget joined #salt
06:54 xl1 joined #salt
06:59 thayne joined #salt
07:03 harobed_ joined #salt
07:04 harobed_ joined #salt
07:06 ravibhure joined #salt
07:11 slav0nic joined #salt
07:11 slav0nic joined #salt
07:15 Kenzor joined #salt
07:17 it_dude joined #salt
07:31 giannello joined #salt
07:36 bmcorser still  playing with salty-vagrant, cannot get preseeding of master to work
07:36 bmcorser https://gist.github.com/bmcorser/11168722
07:40 shoma joined #salt
07:41 SkyRocknRoll joined #salt
07:41 SkyRocknRoll joined #salt
07:44 middleman_ joined #salt
07:47 bmonty bmcorser:
07:47 bmonty bmcorser: what does “salt-key -L” show?
07:48 bmcorser before the minion runs salt-call saltutil.sync_all, salt-key is empty
07:49 bmcorser ie nothing accepted, rejected etc
07:49 bmonty so maybe the keys aren’t getting sent at all?
07:49 bmcorser after the minin runs salt-call saltutil.sync_all the minion key shows up in unaccepted
07:50 bmcorser i don't think they are
07:57 swissknife joined #salt
07:57 swissknife left #salt
08:01 toastedpenguin joined #salt
08:05 anuvrat joined #salt
08:10 gw joined #salt
08:10 ndrei joined #salt
08:14 Daemonik I just ran "./bootstrap.sh git v2014.1.3" on SmartOS base64 13.4.0 and Salt configs were placed in /etc/salt as expected but the SMF created by bootstrap.sh lists configs to be in /opt/local/etc/salt where they are not http://pastebin.com/raw.php?i=CLCTe47E
08:17 scott_w joined #salt
08:19 oc joined #salt
08:24 srage joined #salt
08:31 jY is there anyway in service to do a pattern to search a process tree to see if the service is running?
08:36 anuvrat joined #salt
08:44 scott_walton joined #salt
08:45 giantlock joined #salt
08:50 malinoff joined #salt
08:52 che-arne joined #salt
09:00 babilen jY: pgrep $FOO? "service $FOO status" ? "systemctl status $FOO" (if you already use systemd) ?
09:03 ajw0100 joined #salt
09:08 JesseC-Work joined #salt
09:08 workingcats joined #salt
09:09 Ryan_Lane joined #salt
09:10 ndrei joined #salt
09:10 babilen jY: Sorry, that answer would have been appropriate in a different channel than #salt :)
09:11 jY babilen: heh thanks.. i actually figured out there is a -sig: searchstring
09:11 anuvrat joined #salt
09:12 babilen jY: Great :)
09:15 mgarfias joined #salt
09:16 babilen I wonder if there is a good way to upgrade all my minions to a new salt-minion version (on Debian) -- Doing this through salt is probably suboptimal
09:16 babilen (or performing upgrades in a general way)
09:18 srage joined #salt
09:22 steadyonabix joined #salt
09:27 feiming joined #salt
09:28 feiming Hi guys, i had created a salt-formula for tinc
09:28 feiming https://github.com/feiming/tinc-formula
09:29 ronc joined #salt
09:30 ronc hey everyone. We suddently get: "Failed to change group to admin" in a file.directory state
09:30 ronc any idea?
09:30 kiorky joined #salt
09:31 sgviking joined #salt
09:37 bhosmer joined #salt
09:37 bezaban does the group exist?
09:38 ronc yes. I removed "recurse: group" and it now works
09:39 ronc according to the docs it should work, and it used to work flawlessly
09:39 ronc I have "group: admin"
09:43 bezaban yeah, check on the actual host if the group is there.  If not it might seem that the states are being executed in the wrong order
09:43 bezaban but if it used to work that's a little odd :)
09:44 paabu joined #salt
09:46 harobed__ joined #salt
09:46 ronc it started working by just removing the recourse clause (group: admin) is still there, so i assume the group is there
09:53 ggoZ joined #salt
09:58 bezaban right. yeah, good point.
09:59 ronc when topic says "2014.1.1 is the latest" and https://pypi.python.org/pypi/salt says "2014.1.3" what does that mean?
10:00 rofl____ that 2014.1.3 was tagged 5 days ago
10:00 rofl____ i think that carries out as the "latest"
10:00 sealibor_ joined #salt
10:01 jcsp joined #salt
10:07 babilen If I set "state_verbose: False" in the master config I get the desired output. One problem with that is, however, that successfully executed states are no longer reported in the summary
10:07 it_dude_ joined #salt
10:07 babilen Still no changelog for 2013.1.3 apparently
10:10 xl1 joined #salt
10:10 Ryan_Lane joined #salt
10:11 keekz joined #salt
10:11 workingcats babilen, that's irritated me too (the non-counting in the summary) but i havent looked into it yet
10:12 srage joined #salt
10:16 babilen workingcats: I guess that the summary is generated for the filtered list rather than all states. I would have expected the summary to report on everything even if the output is filtered
10:16 workingcats babilen, yep agreed. if you happen to find a solution let us know ;)
10:24 babilen workingcats: It's fairly easy, but would require a bit of refactoring of salt/output/highstate.py -- I would essentially generate the summary string either before _strip_clean() is run or run it on a non-stripped version (i.e. work with the original "data") in _format_host
10:27 babilen workingcats: Probably would make sense to refactor that massive function into a "_format_summary()" that returns the summary string and a _format_results() that returns the result overview of all states. The easiest "fix" is probably to simply move everything related to the summary to the beginning of the function to generate it on the non-stripped version and to append it later.
10:28 workingcats babilen, that actually sounds like something i could do even given my time constraints.. will try to find some time for it this week
10:29 coucou joined #salt
10:29 coucou hello !
10:29 babilen workingcats: I could give it a try, but will probably have to spend quite some time to get a salt test environment running. Is the whole "this is how you test salt while developing it" documented somewhere?
10:30 coucou is it possible to have a regex on a file's path with file module ?
10:30 babilen coucou: What are you trying to achieve?
10:31 coucou on windows i need to access to firefox's profile folder
10:32 coucou and le the name of this folder is a concat of a HASH and ".default"
10:32 coucou example : sd42sf45.default
10:33 coucou so i need some regex like "c:\......\*.default\....
10:34 coucou is it possible ?
10:35 babilen What will you do with it?
10:36 coucou edit a file with file.replace
10:37 babilen Okay, not sure if you can (certainly haven't seen it, but I'm fairly new still)
10:38 coucou babilen: thx guy ;)
10:44 hhenkel Hi all, is there a good way to visualize the success of highstate runs? Maybe with carbon/graphite?
10:45 hhenkel Also, is there a way to see what data is available within the salt mine?
10:47 stevendgonzales joined #salt
10:56 scott_w joined #salt
10:57 dmorrow joined #salt
10:58 yota joined #salt
10:59 sm1ly joined #salt
11:00 ndrei joined #salt
11:03 scott_w joined #salt
11:05 mr_chris joined #salt
11:06 srage joined #salt
11:07 mgw joined #salt
11:08 \ask joined #salt
11:11 Ryan_Lane joined #salt
11:11 ndrei joined #salt
11:16 fsniper left #salt
11:20 ninkotech joined #salt
11:20 \ask joined #salt
11:24 jrdx joined #salt
11:24 babilen basepi: It would be great if you could take care of the release notes for 2014.1.3 again :)
11:36 stevendgonzales joined #salt
11:39 stevendgonzales joined #salt
11:39 ndrei joined #salt
11:52 kadel joined #salt
11:57 diegows joined #salt
12:00 srage joined #salt
12:01 che-arne joined #salt
12:02 elfixit joined #salt
12:05 stevendgonzales joined #salt
12:06 foo_ joined #salt
12:07 vejdmn joined #salt
12:08 ekristen joined #salt
12:10 DenkBrettl joined #salt
12:11 Ryan_Lane joined #salt
12:12 jforest joined #salt
12:14 toastedpenguin joined #salt
12:19 Leech joined #salt
12:26 ckao joined #salt
12:28 gadams999 joined #salt
12:33 jslatts joined #salt
12:34 FarrisG joined #salt
12:48 HeadAIX joined #salt
12:51 tessellare joined #salt
12:52 Munk joined #salt
12:53 * Munk is a Bunk
12:54 Munk ##
12:54 jeremyBass joined #salt
12:54 jeremyBass left #salt
12:58 jraby joined #salt
13:02 tvs_ joined #salt
13:04 tvs_ I'm trying to change recursively the owner/group in a directory, but it returns a failed status if the directory already has the owner I specified, how can avoid getting this error? I need this state to return true
13:05 sroegner joined #salt
13:05 vejdmn1 joined #salt
13:05 faldridge joined #salt
13:06 ipmb joined #salt
13:07 calvinhp_mac joined #salt
13:12 Ryan_Lane joined #salt
13:13 Leech joined #salt
13:14 Ryan_Lane1 joined #salt
13:18 vejdmn joined #salt
13:21 Bunk2 http://www.youtube.com/watch?v=7k4Xk1mEwmI
13:21 Bunk2 gg
13:24 Networkn3rd joined #salt
13:24 racooper joined #salt
13:25 kadel_ joined #salt
13:29 mpanetta joined #salt
13:34 Bunk2 left #salt
13:39 bastion1704 joined #salt
13:44 it_dude joined #salt
13:44 bastion1704 good morning is it possible with the module grains.item to only get on of the ip of ipv4 ?
13:44 viq bastion1704: how about http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.network.html ?
13:45 mapu joined #salt
13:45 timoguin bastion1704: in a state, this should work: grains.get('ipv4')[0]
13:46 bastion1704 tx @viq and @timoguin
13:46 babilen bastion1704: Does "on of the up of ipv4" mean that you only want a single IPv4 address?
13:46 rojem joined #salt
13:47 babilen couldn't quite parse that
13:47 bastion1704 babilen I want the primary ip
13:47 babilen bastion1704: What is the primary IP if you have multiple interfaces?
13:48 bastion1704 I actually only have 127.0.0.1 and my primary IP:)
13:49 babilen Well, all of those should be listed by grains.get('ipv4')
13:49 srage joined #salt
13:49 mattmtl joined #salt
13:52 Jahkeup joined #salt
13:53 babilen Is there a good document that details how to develop salt itself? (e.g. coding guidelines, test suite documentation, configuration of test environments, ...)
13:55 [MT] I'm trying to make salt trigger a highstate on a minion with an updated git repo when that git repo is updated. I have a post-update hook on the git server that looks like this (http://dpaste.com/1791298/) I made sure that the git user is able to execute those commands
13:56 [MT] the events make it to the master, but, as far as I can tell, nothing is actually executing - My reactor file http://dpaste.com/1791313/O
13:58 timoguin babilen: docs for running the tests: http://docs.saltstack.com/en/latest/topics/development/tests/index.html
13:58 timoguin and writing tests: http://docs.saltstack.com/en/latest/topics/development/tests/unit.html
13:59 timoguin coding style: http://docs.saltstack.com/en/latest/topics/development/conventions/style.html
13:59 timoguin and docker images for the test suite: https://github.com/saltstack/docker-containers
13:59 vejdmn joined #salt
14:00 FarrisG joined #salt
14:00 [MT] I lied... it looks like maybe the reactor is working perfect and it actually is the hook not doing what I want it to
14:03 babilen timoguin: Thank you!
14:03 vejdmn joined #salt
14:05 mike25ro joined #salt
14:05 mike25ro hi guys
14:05 mike25ro sometimes salt command times out
14:05 mike25ro how can i see.. the status of the job that run?
14:06 babilen timoguin: I am quite surprised by the placement of the closing parantheses in, for example, https://www.refheap.com/78517 -- Is this the desired format? (salt/output/highstate.py)
14:06 babilen mike25ro: Read http://docs.saltstack.com/en/latest/topics/jobs/index.html
14:06 mike25ro thanks babilen
14:07 babilen mike25ro: You are, in particular, looking for lookup_jid and jobs.active / jobs.list_jobs
14:09 vejdmn joined #salt
14:09 meteorfox joined #salt
14:09 mike25ro babilen: i am reading through that... but where can i find the jid of salt command i just run?
14:09 meteorfox joined #salt
14:09 faust joined #salt
14:10 babilen If it is still running use "salt-run jobs.active" and you get the JID or look in the output of jobs.list_jobs for historical data
14:10 mike25ro ah that
14:10 mike25ro is the one... list_jobs
14:10 mike25ro thanks a lot babilen
14:10 faldridge joined #salt
14:11 babilen mike25ro: You are welcome -- I personally also found it to be preferable to adjust state_output or state_verbose settings in your master's config (but that is not really related to your question)
14:12 timoguin babilen: not sure it that's desired or not really
14:12 timoguin looks pretty common throughout that module though
14:12 timoguin i've never ran the linters against the code
14:13 babilen timoguin: Yes it is, but it is a completely different style than what I am used to and I personally find it ugly. I will not change it in this module (mixing styles would be even worse), but I find this to be a very extreme interpretation of the "vertical is better" guideline
14:14 timoguin those dangling parentheses really do draw the eye....
14:14 babilen yeah
14:14 kaptk2 joined #salt
14:15 Ryan_Lane joined #salt
14:17 thedodd joined #salt
14:19 calvinhp_mac joined #salt
14:20 icarus joined #salt
14:21 alunduil joined #salt
14:23 feiming joined #salt
14:23 StDiluted joined #salt
14:24 analogbyte joined #salt
14:25 ajprog_laptop joined #salt
14:27 feiming Hi guys, i had created a salt-formula for tinc
14:28 feiming https://github.com/feiming/tinc-formula
14:29 timoguin feiming: nice!
14:29 ange hey
14:29 mike25ro what is tinc?
14:30 mike25ro tinc == vpn?
14:30 timoguin yea it's a peer-to-peer vpn
14:31 timoguin feiming: you need to move that init.sls into a folder called tinc/
14:31 mike25ro is it good? :) i know it's not the place to ask...  :D
14:31 timoguin and probably the files and template folders under there too
14:31 cheus_ joined #salt
14:31 timoguin mike25ro: i used it years ago for a darknet with some friends and it was pretty fun. worked well for what we were doing
14:32 mike25ro timoguin: has a point .. it would be eaasier if all files and configs would be under a tinc folder... so you can add it easily to your own repo
14:32 timoguin feiming: take a look at how this one is structured and replicate it: https://github.com/saltstack-formulas/redis-formula
14:32 timoguin yea, that structure makes it easy to just add a formula to gitfs_remotes
14:32 HeadAIX joined #salt
14:34 feiming timoguin, ok
14:34 [diecast] joined #salt
14:34 timoguin feiming: once you've done that shoot an email to the mailing list and they'll get you added to the saltstack-formulas team
14:35 Psi-Jack_ joined #salt
14:36 feiming timoguin: fix
14:36 repl1cant joined #salt
14:36 feiming timoguin: what's the mailing list address?
14:37 mariusv joined #salt
14:38 basepi babilen: they're done, look at the pending pull requests. =) we should be announcing 2014.1.3 today, and I'll merge the release notes at the same time.
14:38 feiming timoguin: got it
14:40 UtahDave joined #salt
14:41 babilen basepi: Thank you! Am already running it :)
14:42 babilen I thought that I simply missed the announcement and didn't quite realise that the packages simply hit the mirrors before that version has been announced
14:42 Networkn3rd joined #salt
14:42 basepi babilen: awesome! Should be a good release. We'll be doing a .4 soonish because of a bug in the setup.py, but that only affects new installs (it doesn't properly dep pyzmq because of a typo)
14:42 StDiluted joined #salt
14:43 basepi So pip installs need a manual pyzmq install. But that's the only big bug I've heard about so far.
14:43 babilen basepi: I haven't encountered any problem so far, but was simply curious as to what changed
14:44 basepi Ya, check the pull req, they will be live later today.
14:44 babilen Now I only need to figure out if it is save to upgrade salt-minion on my Debian minions with "pkg.upgrade"
14:46 eriko joined #salt
14:47 basepi It's safe, but sometimes your salt minion doesn't properly restart. Best current solution is a cmd.run with an 'at' command to schedule the restart for after the highstate run.
14:48 cheus_ Anyone know of a way to grab a yaml file into a python dict in an sls?
14:48 mjulian joined #salt
14:48 basepi echo 'service restart salt-minion | at now + 5 minutes
14:48 basepi Er, missed a quote on that
14:48 basepi That's what I get for typing on my phone.
14:49 jergerber joined #salt
14:49 babilen basepi: Okay, thank you. Am slightly weary of performing this on our production boxes ...
14:50 babilen err, cautious rather
14:50 basepi babilen: I've heard nothing but good results from the 'at' method. But I would definitely target only some of your minions at a time.
14:50 basepi Ya, I assume you meant wary
14:51 basepi If you hit me up in a few hours I can get you a more complete example that's quite robust
14:51 babilen basepi: Okay, I will give it a try, but not ten minutes before I head home :) -- I don't really want a customer to scream at me now
14:51 basepi Hahaha
14:52 babilen basepi: I've written to the mailing list, would appreciate an answer there. I'll play in vagrant for the time being.
14:52 basepi babilen: I'm very behind on the mailing list but I'll look for your message later today.
14:52 tyler-baker joined #salt
14:53 babilen ta!
14:53 * basepi is afk for a few.
14:55 mike25ro is there a way ... that when i run a salt state instead of returning data to return only the jid ?
14:55 faldridge joined #salt
14:55 mike25ro something like ... salt '*' state.sls mystate out=jid ... ?
14:56 faldridge joined #salt
14:56 anuvrat joined #salt
14:56 babilen mike25ro: I haven't seen that
14:57 ipmb joined #salt
14:58 jY i can query the salt mine for interface addresses like salt-call mine.get '*' 'network.ip_addrs'  but how can i query for eth0's ip?
14:59 jY salt '*' network.ip_addrs eth0
14:59 jY that works on master.. but i need it for the salt mine also
15:00 quadmasta joined #salt
15:01 quadmasta I installed a release candidate via the bootstrap script but now the version is available in a yum repository; can anyone help me with a link or search terms on how best to undo what the bootstrap script did?
15:02 quadmasta or is it recommended to prefer the bootstrap script over yum?
15:02 mike25ro babilen: i haven't found ... i was just wondering maybe someone knows about it. thanks anyway
15:04 che-arne joined #salt
15:06 babilen mike25ro: The best I can think of is a highstate run with a short timeout followed by an immediate jobs.active and some sed/awk/python/perl magic.
15:06 Networkn3rd joined #salt
15:07 dc_ronik joined #salt
15:07 timoguin a return straight back to the command line with jid output would be nice
15:07 timoguin i.e., don't wait for any minion responses
15:07 babilen exactly
15:08 JasonSwindle joined #salt
15:08 babilen mike25ro: I would suggest to file a "wishlist" bug on https://github.com/saltstack/salt/
15:08 babilen (not that they don't have enough already, but that way it will at least not get lost)
15:08 JasonSwindle joined #salt
15:09 gadams999 is 2014.1.0 still the latest "release" code?
15:09 babilen no
15:09 babilen 2014.1.2 is the latest and .3 is just around the corner
15:10 JasonSwindle joined #salt
15:10 aleszoulek joined #salt
15:11 quadmasta so no help on the bootstrap uninstall?
15:11 gadams999 weird. 2014.1.0 is the latest showing in EPEL. Ah, I see the tags for other v2014.x.x release
15:12 it_dude joined #salt
15:12 meteorfox joined #salt
15:12 xerxas joined #salt
15:13 hhenkel gadams999: http://dl.fedoraproject.org/pub/epel/testing/6/x86_64/ => 2014.1.3 is in testing
15:14 gadams999 hhenkel: thanks! Realized I had testing turned off. will enable that bad boy
15:14 it_dude_ joined #salt
15:15 Ryan_Lane joined #salt
15:16 danielbachhuber joined #salt
15:17 timoguin mike25ro: you can add --verbose to the CLI and it will print the jid first
15:17 timoguin actually
15:17 timoguin salt '*' test.ping --verbose --out=quiet
15:17 timoguin will get you what you want
15:18 tligda joined #salt
15:21 srage joined #salt
15:21 Leech joined #salt
15:21 abe_music joined #salt
15:22 StDiluted lol, be verbose but also quiet!
15:23 Psi-Jack_ joined #salt
15:25 kvbik joined #salt
15:26 NV joined #salt
15:28 KennethWilke joined #salt
15:28 schimmy joined #salt
15:29 TheGerm joined #salt
15:29 JasonSwi_ joined #salt
15:30 kermit joined #salt
15:30 calvinhp_mac joined #salt
15:31 doddstack joined #salt
15:33 pydanny joined #salt
15:33 timoguin StDiluted: thus why the changes in develop make sense. the quiet outputter has been removed and is now called with --output_loglevel=quiet
15:34 gwmngilfen joined #salt
15:34 StDiluted makes sense
15:38 jeremyBass1 joined #salt
15:41 ajw0100 joined #salt
15:43 |o||o| joined #salt
15:43 Gareth basepi: ping
15:43 _TheDodd_ joined #salt
15:44 viq Are there any plans to put newer halite in epel?
15:45 basepi Gareth: pong
15:45 calvinhp_mac what triggers a git pull on gitfs backed salt roots and pillars
15:45 Gareth basepi: hey :) so is 2014.1.3 locked down in terms on what is being merged in?
15:45 basepi Yes, it was tagged Friday.
15:46 meteorfox joined #salt
15:46 basepi But .4 should go out later this week with a few different fixes.
15:46 viq calvinhp_mac: timer or explicit command. By default it polls every 60 seconds, unless you use salt-run fileserver.update
15:46 calvinhp_mac ok, I've been using fileserver.update
15:46 calvinhp_mac just wanted to know if there was some other event
15:47 calvinhp_mac thansk
15:47 Gareth basepi: ah okay.  someone was asking about two fixes, one got in but https://github.com/saltstack/salt/pull/11401 did not.
15:48 viq basepi: any idea whether newer halite will be put in epel?
15:48 viq since you're right now being asked about packaging and such ;)
15:48 basepi Gareth: can you mention me on that issue saying it needs to get in?
15:49 Gareth for 2014.1.4?
15:49 basepi viq:  unsure. terminalmage, did you package that originally?
15:49 basepi Gareth: yep
15:49 Gareth basepi: can do.
15:49 terminalmage basepi: package what?
15:49 terminalmage halite?
15:49 basepi terminalmage: yes
15:50 terminalmage yes, but it hasn't been updated in a while. I can build a new one
15:50 viq terminalmage: epel has 0.13, there was 0.16 released recently
15:50 basepi terminalmage: cool. Not critical, but i think Adi has released a couple of new versions since then.
15:51 * basepi is afk for a minute.
15:51 viq erm, 0.1.13 and 0.1.16
15:51 ldlework joined #salt
15:51 sroegner joined #salt
15:52 calvinhp_mac joined #salt
15:56 FL1SK joined #salt
15:56 GradysGhost joined #salt
15:57 viq WTF, I have halite configured, yet the port does not come up, and nothing in master log
16:02 JasonSwindle joined #salt
16:03 joehoyle joined #salt
16:03 NullWagon joined #salt
16:04 rgarcia_ joined #salt
16:04 teskew joined #salt
16:06 cruatta joined #salt
16:07 [MT] First I couldn't understand why salt-call * state.highstate was calling weird functions, then I realized '*', then I couldn't figure out why function * wasn't available, then I realized s/-call//
16:07 [MT] wow...
16:08 [MT] viq: I still need to get gitlab-ci working... :(
16:09 sealibora joined #salt
16:10 cruatta joined #salt
16:10 KyleG joined #salt
16:10 KyleG joined #salt
16:11 mateoconfeugo joined #salt
16:12 jeremyfelt joined #salt
16:16 Ryan_Lane joined #salt
16:18 srage joined #salt
16:19 cruatta joined #salt
16:20 smcquay joined #salt
16:20 scoates joined #salt
16:21 anuvrat joined #salt
16:23 jaimed joined #salt
16:24 obimod joined #salt
16:24 obimod hey all
16:24 ajolo joined #salt
16:25 obimod i'm trying to grab the ip address of the minion for use in a yaml configuration context, like a local_address variable inside a configuration... i know a bash command to grab it but can't figure out how i'd stick it in the yaml config file
16:26 bastion1704 joined #salt
16:27 obimod as soon as i get this configuration right i'll try to replicate #11784
16:27 obimod i think my cloud configs will be ready for testing
16:27 cruatta joined #salt
16:27 obimod and clean that issue up
16:27 obimod issue/pull request w/ a strace
16:27 UtahDave obimod: I might be misunderstanding your question, but can you use jinja to template that config file with the ip address?
16:28 obimod well the ip address changes
16:28 obimod it would be nice to have it as a dynamic read-in
16:28 viq obimod: yeah, {{ salt['network.ipaddrs']('eth0') }}
16:28 obimod @viq sweet!
16:29 obimod i'll check out the salt network docs, thanks @viq
16:30 chrisjones joined #salt
16:30 Topic for #salt is now Welcome to #salt | 2014.1.1 is the latest | SaltStack / CoreOS meetup tonight in SLC.  See http://goo.gl/DDXUkS | 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/
16:31 viq UtahDave: "2014.1.1 is the latest" - I don't think that's accurate ;)
16:32 schimmy joined #salt
16:33 mpanetta joined #salt
16:33 Topic for #salt is now Welcome to #salt | 2014.1.3 is the latest | SaltStack / CoreOS meetup tonight in SLC.  See http://goo.gl/DDXUkS | 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/
16:34 UtahDave thanks, viq!
16:35 viq ;)
16:36 viq OK, WTF? Logs say "Halite: Starting up ..." yet nothing listens on port 8080 and and no further mention of halite in logs
16:36 viq log_level: info
16:39 veselin joined #salt
16:39 [diecast] joined #salt
16:40 linuxlewis joined #salt
16:41 veselin Hello all, I'm trying to match a string against regex in a Jinja template. E.g. {% if grains['id'] == "webserver*" %} My Text {% endif %}
16:41 veselin is this the correct way to populate 'My Text' on that line in the jinja template?
16:42 veselin if the ID form the minion's grains matches that is
16:42 timoguin you can't do filters like that in jinja for conditionals.
16:42 timoguin you can use the match module though: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.match.html
16:44 googolhash joined #salt
16:45 HeadAIX joined #salt
16:49 veselin timoguin: Thank you. Could I use conditionals with match module?
16:49 smcquay joined #salt
16:52 mgw joined #salt
16:52 jtang1 joined #salt
16:53 ajw0100 joined #salt
16:53 srage joined #salt
16:54 timoguin veselin: yes. {% if salt['match.glob']('webserverfoo*') %}
16:54 timoguin i think that's how you'd do it
16:55 ajolo_ joined #salt
16:55 veselin timoguin: perfect, thank you
16:55 srage joined #salt
16:56 conan_the_destro joined #salt
16:57 pydanny joined #salt
16:57 cheus_ Anyone know if there's a good way to include environments (eg, the setting the 'env' variable) without also imposing the environment directory structure or branching scheme?
16:58 ajw0100 joined #salt
16:58 ajolo__ joined #salt
17:00 cruatta_ joined #salt
17:00 joehoyle joined #salt
17:03 ajw0100 joined #salt
17:04 ajolo__ joined #salt
17:04 bastion1704 joined #salt
17:05 schimmy1 joined #salt
17:06 AdamSewell joined #salt
17:06 AdamSewell joined #salt
17:07 ajw0100 joined #salt
17:07 jmpf joined #salt
17:09 [MT] I wonder if a 2.5GB pile of files makes sense in a git repo...
17:09 viq [MT]: I heard git dislikes big repos/binary files
17:10 viq On the other hand I heard of a company (a big portal, think ebay) that uses git to keep and distribute to machines their ~3GB of PHP code
17:10 [MT] that's what I've heard too.. 1.8GB could be it's own repo which is just a giant pile of ms office templates
17:10 druonysus joined #salt
17:10 druonysus joined #salt
17:11 viq [MT]: I heard somewhere there's a recent presentation by facebook about scaling mercurial, I think they mentioned it's much better suited for that
17:11 viq So you may want to look that up. If mercurial is an option that is ;)
17:12 TheRhino04 joined #salt
17:12 [MT] gitlab is git only, isn't it?
17:12 viq yeah...
17:12 timoguin we have some 7+GB mercurial repos
17:13 timoguin but plz don't
17:13 viq oh?
17:13 jrdxusa joined #salt
17:13 timoguin it can be done, but it's not fun.
17:14 Eugene In #git we usually point at git-annex
17:14 UtahDave New varnish salt formula!  https://github.com/saltstack-formulas/varnish-formula
17:14 Eugene git itself has minimal problems with large quantities of files, even binary ones, but your system will run out of RAM on repack operations
17:14 [MT] The problem is that I want the benefit that I did with these other servers... someone pushes to a git repo and as soon as that happens, highstate runs on a group of servers
17:15 Eugene I don't see why you couldn't make annex work in a salted environment, though
17:15 [MT] I'll take a look at it
17:15 whiteinge yay varnish!
17:15 timoguin git-annex looks interesting
17:16 [MT] the alternative option is that I give people access to the salt master so they can drop in the files and run state.highstate themselves (via a wrapper script)
17:16 viq [MT]: have a look at rundeck
17:17 Eugene I often turn the question around: why do you need that many jiggabytes of files tracked? What are they?
17:17 timoguin [MT]: maybe you could open up a share and run a filesystem watcher on it
17:17 jrdx joined #salt
17:18 [MT] Eugene: in this case, I care much less about tracking and more about how people will work with it; with git, I could set up the repo for them and done
17:19 [MT] rundeck looks interesting
17:19 logix812 joined #salt
17:19 [MT] it looks possibly overkill for this specifically... the share idea might be best
17:20 timoguin [MT]: https://github.com/gorakhargosh/watchdog#shell-utilities
17:20 Eugene SFTP-restricted user ;-)
17:20 timoguin check out watchdog and the CLI scripts that come with it
17:21 timoguin should be pretty easy to trigger whatever you need on filesystem update
17:21 [diecast] joined #salt
17:21 viq timoguin: question is whether a filesystem update won't be a partial write of the file
17:22 [MT] I wonder if I could have a script launch on fs update that kills other instances of the same script and then sits and waits 30min before running state.highstate
17:22 possibilities joined #salt
17:22 timoguin viq: that's true. i know the create event fires as soon as the file is touched
17:22 timoguin i used a sleep function in the application i did
17:23 timoguin not sure about modify etc.
17:24 [MT] I imagine there are a lot of ways to skin this kitty. The git repo option would be easiest up front, but probably the most broken in the long run.
17:24 druonysus joined #salt
17:24 druonysus joined #salt
17:24 viq [MT]: fileshare somewhere, in cron rsync and fire off highstate if anything changed?
17:24 [MT] ya.. probably best
17:24 [MT] I'm gonna go with that one
17:24 viq And you can get out of rsync info whether something did/is going to change
17:25 viq Just make sure to use a lock file or something, to not run two rsyncs at once
17:25 will joined #salt
17:25 ajolo__ joined #salt
17:25 viq Also while on the subject of writing scripts, I'll plug http://rerun.github.io/rerun/
17:26 jeremyfelt joined #salt
17:26 [MT] I need to build the file server anyway. That's going to be a pain. I want multiple samba file servers that use the same data source (disk) that are fully redundant and load balanced and are capable of being brought up instantly in our other DC if this DC were to blow up.
17:27 [MT] It's going to replace our netware server
17:28 [MT] viq: nice, that looks neat
17:28 dfinn joined #salt
17:28 [MT] on that subject... I should consider moving my personal website back to github.io
17:29 stanchan joined #salt
17:29 Ryan_Lane joined #salt
17:32 happytux joined #salt
17:33 mapu joined #salt
17:33 epcim joined #salt
17:36 UtahDave viq: did you write rerun?
17:36 viq UtahDave: no, just passing a useful project on
17:36 UtahDave cool
17:37 viq In my previous job I showed it and we started using it, quite nice to have auto-self-documenting scripts
17:38 viq Dumped them in svn, used jenkins to build RPMs out of them, and satellite to push onto machines ;)
17:40 viq OK, time to run, cya
17:41 CeBe joined #salt
17:44 abe_music joined #salt
17:45 ajolo__ joined #salt
17:45 shoma joined #salt
17:45 quist joined #salt
17:46 colsandurz45 joined #salt
17:46 sandbender1512 basepi: got that patch for the mysql_user state to add support for the optional REQUIRE <ssl, etc> stuff, just testing it now
17:48 colsandurz45 Hi, I have a question about returners.  I'm using the sqlite3 returner and I'm noticing that each minion only logs to it's own database.  Is there any way to get the master to log everything in it's database?  Is there some config option for the master I'm missing?
17:50 gothix_ joined #salt
17:51 haroldjones joined #salt
17:52 stevendgonzales joined #salt
17:54 MrTango joined #salt
17:55 mpanetta_ joined #salt
17:57 KennethWilke joined #salt
17:57 calvinhp_mac joined #salt
17:57 Gordonz joined #salt
17:58 TheGerm joined #salt
17:58 Gordonz joined #salt
17:58 linuxlewis joined #salt
17:59 haroldjo_ joined #salt
18:02 srage joined #salt
18:03 jmpf joined #salt
18:05 happytux joined #salt
18:07 travisfischer joined #salt
18:08 shm_get any idea what is wrong with http://pastebin.com/GpKqX19s  ?
18:09 druonysus joined #salt
18:09 druonysus joined #salt
18:09 UtahDave shm_get: looks like bugzilla is a list not a dict
18:09 quist` joined #salt
18:10 KennethWilke you're a list!
18:10 KennethWilke whoops, sorry :)
18:10 shm_get UtahDave what is the syntax to make 'bugziall' a dict rather than alist ?
18:10 cruatta joined #salt
18:10 JasonSwindle joined #salt
18:11 UtahDave shm_get: it's how it's set up in your pillar.  can you pastebin a sanitized version of your pillar definition for bugzilla
18:11 jimklo joined #salt
18:13 shm_get UtahDave no... no pillar.. my intent is to have a sls that derive a home directory out of a pillar 'user' if deined and fallback to 'bugziall' as user if not...  I want to allow to overide via pillar somehow.. but the idea is that by default I would get 'bugziall' as user and whatever the home directory of 'bugzilla' user is
18:13 KennethWilke ah yeah it looks like your bugzilla is a list of tuples
18:13 [MT] guiding people through replacing a server in a network rack that barely speak english and understand very little of it is difficult
18:13 shm_get oh.. so I should use {} istead of [] to define by variable ?
18:13 [MT] irc is helping me here...
18:14 KennethWilke yeah i think jinja does the normal python dict format
18:14 shm_get to get a dict rather than a list... /me try thanks
18:14 KennethWilke so {} instead of [] and : instead of ,
18:14 KennethWilke you'll need : for the key:value pairing
18:14 KennethWilke instead of , as the list item separator
18:14 shm_get KennethWilke Thanks... will do...
18:15 KennethWilke then your bugzilla['webhome'] should work, and in jinja it might also work as bugzilla.webhome but the choice is a matter of prefernece
18:15 KennethWilke shm_get: you're welcome
18:15 mateoconfeugo joined #salt
18:16 possibilities joined #salt
18:16 terminalmage viq: http://koji.fedoraproject.org/koji/packageinfo?packageID=17175
18:16 KennethWilke [MT]: i hear you there, those are tough calls
18:18 Ryan_Lane1 joined #salt
18:20 [diecast] joined #salt
18:21 vejdmn joined #salt
18:23 shm_get KennethWilke  \o/ it works.. Thanks again
18:23 basepi sandbender1512: awesome!
18:23 stephanbuys joined #salt
18:23 KennethWilke shm_get: woo! anytime!
18:24 vejdmn joined #salt
18:25 stephanbuys hi all, where does Salt configure the paths that it searches for modules? When running salt from my IDE is complains that Function 'state.highstate' is not available
18:25 linuxlew_ joined #salt
18:25 shm_get coming to salt without prior knowledge of python makes the experience a bit steeper... but so far, I'm quite impress with how powerful and expressive this is....
18:27 KennethWilke stephanbuys: those modules are in the salt source tree, so i think it's relative to where salt was installed
18:27 sandbender1512 Gareth J. Greenaway in the house? (msg me priv if u want)
18:28 Gareth hm?
18:28 sandbender1512 modules/iptables.py:197 calling startswith on an int
18:28 stephanbuys KennethWilke: My salt lives in a virtualenv, which I’ve configured - I’m suspecting there is a “loader” method somewhere that’s chasing the wrong paths, if I knew what it was searching for I could set it in the IDE
18:29 sandbender1512 I think you want to wrap kwargs['dport'] in str() on those lines (testing it now)
18:29 KennethWilke stephanbuys: that specific function is defined in https://github.com/saltstack/salt/blob/develop/salt/modules/state.py
18:29 tzultacah joined #salt
18:30 KennethWilke shm_get: salt might be a good gateway into the realm of python :)
18:30 KennethWilke based on 'shm_get' are you a C developer?
18:31 thayne joined #salt
18:31 UtahDave New OS install.  Went to make a git commit and it dropped me into nano.  AAGGGHHH MY EYES!!!!
18:31 UtahDave I couldn't type this fast enough:   git config --global core.editor vim
18:32 KennethWilke but but but
18:32 shm_get KennethWilke yep... an old crusty one :-)
18:32 whiteinge UtahDave: that's a known Ubuntu bug
18:33 torrancew UtahDave: save yourself further pain: update-alternatives --set /usr/bin/vim.basic (or .tiny, whichever you have)
18:33 torrancew errr
18:33 KennethWilke shm_get: nice :) i know for me i tried to not like python, then the ctype library won me over because i could work with both languages for whatever i wanted
18:33 torrancew update-alternatives --set editor /usr/bin/vim.basic
18:33 KennethWilke and my colleagues were all doing python
18:35 druonysuse joined #salt
18:35 druonysuse joined #salt
18:36 stephanbuys KennethWilke: seems the error originates in salt/cli/caller.py - busy tracking
18:36 timoguin UtahDave: you should put some config management on that new OS install process. ;-)
18:36 timoguin i think there are softwares that can help with that
18:36 UtahDave Yeah, I should start a project that would make that easy.  :)
18:36 dfinn left #salt
18:37 harobed joined #salt
18:37 marnom joined #salt
18:37 Gareth sandbender1512: Yeah.  Probably a good idea for the dport and sport code, and the sports and dports too.  I *think* the contents of kwargs is always a string though, could be wrong.
18:38 sandbender1512 yeah I would have assumed, but it seems to be getting converted to an int somewhere (didn't look) b/c that was throwing an exception for me
18:38 shm_get KennethWilke my problem is that I got complaisant and lazy with age.. and I rely way to much on the compiler finding my many typos and omissions for me... which in an interpreted world does not work that well :-)
18:39 Gareth really? hmm. could have sworn I tested that.  I may have kept the port wrapped in quotes though.
18:40 smcquay joined #salt
18:40 sandbender1512 nps :) Just thought I'd ping here in case you were around, faster than email or adding a comment on github/etc :)
18:41 possibilities joined #salt
18:41 KennethWilke shm_get: indeed, I prefer compiled languages as well! i often can get more done in these interpreted ones though :\
18:42 Gareth sandbender1512: are you going to make the change and submit a pull request?
18:43 harobed joined #salt
18:44 sandbender1512 I can - will be submitting a pull shortly anyways and I already added it to my local copy, so yeah if you want?
18:44 harobed joined #salt
18:44 Gareth sandbender1512: go for it.
18:44 sandbender1512 kk will do
18:44 possibilities joined #salt
18:45 sandbender1512 (you work for Saltstack, or just a contributor?)
18:46 Gareth sandbender1512: just a contributor.
18:46 linuxlewis joined #salt
18:46 smcquay joined #salt
18:46 sandbender1512 cool, likewise
18:47 bastion1704 joined #salt
18:51 gothix_ I am woring with api and i can auth as myself but when i set up a user other than myself and try to use that i get the following {"status": 500, "return": "An unexpected error occurred"}
18:52 gothix_ the user exists in /etc/salt/master withthe same perms as i have
18:52 fishernose joined #salt
18:54 fishernose If I have a few .deb files that I want to deploy to ubuntu minions, what's the optimal way to do that?
18:54 fishernose I could do a file.managed: source: package.deb
18:55 fishernose And then use salt to remotely execute a dpkg -i command, but I'd prefer to have a pkg.installed format.
18:57 timoguin fishernose: that way will work fine. i feel like i saw .deb support added for pkg.installed, but i can't find it atm.
18:58 timoguin fishernose: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkg.html#salt.states.pkg.installed
18:59 timoguin look at the sources example towards the bottom of that section
18:59 fishernose Will do!
18:59 timoguin you may be able to pass the .deb URI
19:00 stevendgonzales joined #salt
19:00 fishernose Thanks timoguin!
19:01 chrisjones joined #salt
19:01 tr_h joined #salt
19:03 haroldjones joined #salt
19:06 jrdx joined #salt
19:08 jcsp1 joined #salt
19:13 mateoconfeugo joined #salt
19:15 jalbretsen joined #salt
19:15 gothix_ anyone here using the api?
19:17 jcockhren gothix_: I am
19:17 jcockhren gothix_: is your user an actual unix user?
19:18 jcockhren also, are permissions granted in the acls
19:18 ajw0100 joined #salt
19:19 gothix_ jcockhren, did you notice any oddities in using it to athenticate local pam users? I am trying with my user and can authenticate no issues but when i create a system user i get an  {"status": 500, "return": "An unexpected error occurred"}
19:19 gothix_ cant seem to find anything in debug logs eigther
19:22 jcockhren gothix_: yes. I ran into similar issues
19:22 mapu Good afternoon- i was wondering if there is a good mechanism to manage elastic IP addresses with EC2 instances using salt (salt-cloud)?
19:23 whiteinge gothix_: can you auth with that user without salt-api? i.e.:   salt -a pam '*' test.ping
19:24 jcockhren gothix_: have you included your system user in the 'client_acl' setting?
19:24 jcockhren as well as the 'external_auth' setting?
19:25 gothix_ jcockhren, i tried both a low uid systems user under 500 and high regular user over 500 uid
19:27 gothix_ whiteinge, i can auth like that salt -a pam works no issue
19:27 [MT] well... that's interesting, I'm suddenly not able to run salt commands anymore. Apparently I'm out of worker threads
19:28 whiteinge gothix_: ok, good. next step is to add 'debug: True' to your rest_cherrypy config. the traceback from the 500 will then be put in the output
19:28 gothix_ jcockhren, no client acl just set in the master as external auth
19:32 [MT] apparently retrieving git repos went boom too
19:33 ggoZ joined #salt
19:33 gothix_ whiteinge, it  spits out a traceback now Error while processing request for: /login and error in hypermedia_handler and something about KeyError: 'salt_user  the usereim trying to use
19:34 whiteinge gothix_: great. pastebin, please?
19:34 gothix_ sure
19:34 whiteinge oh, and what version are you running?
19:34 fishernose timoguin, the salt minion from the ubuntu repos was complaining about the "sources" part of pkg.installed:\ - sources.
19:35 fishernose I installed the git develop version of the minion instead.
19:35 fishernose But I'm still getting the same issue.
19:35 otsarev__ joined #salt
19:35 timoguin fishernose: is your master the same version?
19:35 fishernose I'm just running locally.
19:36 timoguin fishernose: what's the error?
19:36 gothix_ whiteinge, http://www.pastebin.ca/2700962
19:38 whiteinge gothix_: ah, i see what's going on. can you also pastebin your external_auth configuration? sanitized is great, i just need to see the structure of that config
19:39 TheRhino04 joined #salt
19:39 Ahlee boo.  virtualenv_mod.py only supports pip
19:39 Ahlee guess it's back to cmd.run easy_install :(
19:41 gothix_ whiteinge, http://pastebin.com/2uBegtqn
19:42 whiteinge thx. looking...
19:49 fishernose timoguin, the error is "Data failed to compile"
19:49 jimklo joined #salt
19:50 fishernose timoguin, Unknown yaml render error; line 3
19:50 fishernose timoguin, "vagrant\n pkg.installed\n    - sources:    <======================"
19:51 timoguin fishernose: looks like you forgot a colon after pkg.installed
19:51 timoguin can you pastebin to be sure?
19:51 scalability-junk joined #salt
19:52 fishernose timoguin, I'm running it now--I did indeed forget a colon
19:53 fishernose I have about 10 years of C programming experience, you'd think I'd be over having single-character syntax errors by now.
19:53 timoguin nevar!
19:56 obimod joined #salt
19:57 Ahlee nah, it's pretty clear programming languages are terrible and humans will never be good at them
19:58 timoguin a simple syntax checker for states would be nice
19:59 obimod can you stack pkg.installed module calls under a single name?  i.e. powerdns: \n\t pkg.installed: \n\t\t - name: pdns-server \n\t pkg.installed: \n\t\t - name: pdns-backend-pgsql
19:59 calvinhp_mac Ahlee:  re: pip, shouldn't you be using pip, easy_install is the deprecated way to install packages now
19:59 Ahlee calvinhp_mac: until pip can handle binary distributions it will never be deprecated for me
20:00 whiteinge gothix_: do you have more than one "external_auth" section in your config?
20:00 calvinhp_mac Ahlee:  doesn't the new wheel format do this?
20:01 obimod ooo - powerdns: \n\t pkg.installed: \n\t\t pkgs: ['pdns-server','pdns-backend-pgsql']
20:01 Ahlee calvinhp_mac: maybe, dunno.
20:02 Ahlee I don't have compilers on my systems, nor do I want them, so I just build eggs and distribute them
20:02 Ahlee suppose i'll have to look into wheel now
20:03 gothix_ whiteinge, i have only one external_auth: section but i do have a seperate ldap subsection that occurs after it. could it only be reading the ldap section?
20:03 gothix_ that would explain it
20:03 ajolo__ joined #salt
20:04 whiteinge gothix_: yeah, worth checking. if you can pastebin (sanitiszed) that too, I can say for sure. but the structure of the "salt_user" is good
20:04 gothix_ ok
20:06 calvinhp_mac Ahlee:  I've not done much with them, but that seems to be "the way™"
20:08 gothix_ whiteinge, this is originally how i had it http://pastebin.com/Jjhr77k3
20:09 bhosmer joined #salt
20:09 gothix_ whiteinge, i going to flip flop the section to see if the ldap setion is the only one getting processed
20:10 ajolo__ joined #salt
20:18 gothix_ yeah it makes no difference to flip flop them around
20:22 jrdx joined #salt
20:23 calvinhp_mac joined #salt
20:23 tyler-baker joined #salt
20:23 kballou joined #salt
20:26 whiteinge gothix_: i don't see a problem with either version of the config. did you restart salt and salt-api after changing the config?
20:26 ldlework left #salt
20:28 gothix_ always
20:28 gothix_ on not salt-api
20:29 whiteinge that might be it. salt-api loads the master config into memory same as salt does
20:30 gothix_ whiteinge, salt-api doesent get restarted with salt-master ?
20:31 kermit joined #salt
20:31 whiteinge gothix_: no, it's a separate daemon
20:32 gothix_ i cant stop it and restart it with the service command it fails :(
20:32 gothix_ probably the issue
20:32 vejdmn joined #salt
20:33 whiteinge what is the failure?
20:33 whiteinge you can skip the service command and start it in the foreground:  salt-api
20:33 gothix_ whiteinge, works now
20:34 gothix_ thanks
20:35 joehoyle joined #salt
20:42 aleszoulek joined #salt
20:43 gothix_ whiteinge, so the api reads the same config file as the master then ?
20:43 whiteinge gothix_: yes
20:48 jslatts Ryan_Lane: you around?
20:48 vejdmn joined #salt
20:48 Ryan_Lane jslatts: yep
20:48 jslatts Ryan_Lane: I am doing some work involving s3, noticed that you were one of the recent committers on the s3 util, can I ask you a few things?
20:52 vejdmn joined #salt
20:56 KennethWilke joined #salt
20:56 rawtaz jslatts: dont ask to ask :)
20:58 haroldjones joined #salt
20:58 jslatts rawtaz: yeah, should have phrased that differently.
20:59 timoguin yes. ask to ask to ask.
20:59 timoguin that is the polite thing. :)
21:01 kermit joined #salt
21:01 jslatts i'm wondering how far out the current develop branch is from a release. It has a bunch of s3 enhancements that would save us some extra effort
21:02 jslatts i'm not familiar with the release cycles
21:02 UtahDave jslatts: we're hoping to have the first RC out in a couple of weeks
21:02 rawtaz what's the work involving s3, what does it do? :)
21:03 haroldjo_ joined #salt
21:03 jslatts I am writing a state to pull our app builds from S3 and deploy them
21:03 joehoyle joined #salt
21:03 jslatts i want to take advantage of s3 backed sources for file.managed (not in current release) as well as IAM roles for minions
21:04 joehoyle joined #salt
21:04 jslatts i think I saw that S3 back filesystem is in there as well and that would be huge for us
21:06 Ryan_Lane jslatts: well, I didn't really do much
21:06 Ryan_Lane I just split some functions out of and into a more generic location
21:07 Ryan_Lane but yeah, the S3 backend looks awesome
21:07 joehoyle_ joined #salt
21:07 jslatts Ryan_Lane: yeah, i saw you were had the last commit, then I looked at it and realized :) so sorry
21:07 Ryan_Lane no worries
21:07 Ryan_Lane I'm working on a bunch of aws states/modules right now
21:07 ajolo joined #salt
21:07 Ryan_Lane if you're doing something similar, let me know so that we don't duplicate effort ;)
21:08 Ryan_Lane from what I can tell the S3 file backend should work with IAM roles
21:08 jslatts I think I found a bug with SSL cert verification, though with the move from urllib2
21:09 jslatts it looks like its trying to validate the whole bucket name + endpoint as the CN in the cert
21:10 jslatts but before I submit a one-liner patch for that, I wanted to make sure I wasn't violating some security principal
21:10 Ryan_Lane ah. heh. well, submit a bug (and maybe a patch ;) )
21:10 Ryan_Lane well, SSL should be verifying the hostname
21:10 Ryan_Lane nothing else
21:10 Ryan_Lane the FQDN
21:11 googolhash joined #salt
21:11 Ryan_Lane can you give me an example of what it's checking vs what it should check?
21:12 jslatts this line: https://github.com/saltstack/salt/blob/develop/salt/utils/s3.py#L143 passes in the constructed URL for the bucket + endpoint
21:12 jslatts so its looking for mybucket.s3-us-west-2.amazonaws.com or something similar
21:12 Ryan_Lane yeah, that's correct
21:12 jslatts and AWS has a wildcard cert so that would validate
21:12 jslatts BUT, it is valid to have a bucket named my.bucket.name.s3-us-west-2.amazonaws.com
21:13 jslatts which doesn't match on wildcard
21:13 twlight joined #salt
21:13 Ryan_Lane folks shouldn't name their bucket with dots, then :)
21:13 Ryan_Lane hell, amazon should be warning people about that
21:13 Ryan_Lane you'd need to turn off validation for that
21:13 jslatts http://docs.aws.amazon.com/AmazonS3/latest/dev/BucketRestrictions.html
21:13 jslatts they actually recommend it
21:14 nebuchadnezzar joined #salt
21:14 Ryan_Lane -_-
21:14 jslatts :)
21:14 Ryan_Lane assuming you're connecting over http that's fine
21:15 jslatts sure, but I guess allowing a flag to turn off cert validation would be the preferable fix? rather than turning off verification completely
21:15 twlight Hi everyone, I've started to work with the orchestration salt stuff and had a question. can you specify an environment where the state files within your orchestration.sls live? I'm seeing errors based on not matching sls files found in env base.
21:15 wendall911 joined #salt
21:16 Ryan_Lane jslatts: I guess that's the most reasonable solution
21:17 vejdmn joined #salt
21:21 fishernose I'm having some trouble getting packages installed from .deb files.  http://pastebin.com/yFVuhWty
21:21 racooper joined #salt
21:21 mateoconfeugo joined #salt
21:25 fishernose I caught that double-colon in pastebin but I'm still having the same issue.
21:31 yomilk joined #salt
21:32 TheRhino_ joined #salt
21:34 happytux joined #salt
21:34 cewood joined #salt
21:34 googolhash_e joined #salt
21:37 patrek My git repo is only available by https. Looking for the documentation to setup gitfs to send the right credentials.
21:37 logix812 joined #salt
21:54 Heartsbane Why doesn't this work: http://pastebin.com/bMdvr2mM
21:54 Heartsbane I blame whiteinge.
21:55 zdong joined #salt
21:59 Daemonik joined #salt
22:02 Gareth Heartsbane: error?
22:05 yusuket joined #salt
22:13 Heartsbane ya hold on I am running it one more time
22:14 Heartsbane I think I may have had what alcoholics refer to as a moment of clarity
22:15 jeffg1 joined #salt
22:16 Heartsbane Gareth: false alarm, it is always they way it rolls I ask a question ... then figure it out myself
22:16 Heartsbane I hate when that happens
22:17 possibilities joined #salt
22:19 alunduil joined #salt
22:22 Gareth Heartsbane: no worries :)
22:23 zdong joined #salt
22:27 yusuket joined #salt
22:28 fishernose So does anyone have working example of a pkg.installed working with a .deb file?
22:30 Whissi_ joined #salt
22:33 fishernose Perhaps it would be better to ask tomorrow?
22:34 Ryan_Lane fishernose: you have one package defined with two sources?
22:35 Ryan_Lane fishernose: try this, maybe? http://pastebin.com/9Ha6bB3m
22:36 fishernose Ryan_Lane, I'm checking it out.
22:37 JasonSwindle joined #salt
22:37 kermit joined #salt
22:37 vejdmn joined #salt
22:37 mjulian left #salt
22:38 fishernose Ryan_Lane, I'm running that example, it's got a few other states to check up on.
22:39 fishernose Ryan_Lane, wow looks like at least one of them worked.
22:39 Ryan_Lane great
22:39 fishernose That's more progress than I've had, thanks!
22:39 Ryan_Lane heh. yw.
22:43 jimklo What does "Requisite declaration install-ant in SLS ant-install is not formed as a single key dictionary" mean?  Heres my sls: http://pastebin.com/NgaSSMku
22:43 ajw0100 joined #salt
22:45 ajprog_laptop joined #salt
22:47 jslatts joined #salt
22:48 Luke_ joined #salt
22:48 haroldjones joined #salt
22:49 tyler-baker joined #salt
22:49 schimmy joined #salt
22:51 Whissi joined #salt
22:53 Whissi joined #salt
22:55 Whissi joined #salt
22:57 faldridg_ joined #salt
22:58 schimmy1 joined #salt
22:58 vejdmn joined #salt
22:59 rojem joined #salt
23:05 mgw joined #salt
23:06 kiorky joined #salt
23:09 ajw0100 joined #salt
23:12 AdamSewell joined #salt
23:13 kballou joined #salt
23:18 ajprog_laptop1 joined #salt
23:19 jgarr joined #salt
23:20 jimklo can anyone please explain why I the archive.tar module might not be available here? http://pastebin.com/index/NgaSSMku
23:20 abe_music joined #salt
23:21 Ryan_Lane I don't understand what I'm supposed to document here: http://jenkins.saltstack.com/job/salt-pr-rs-ubuntu12.04/2875//console
23:22 jaimed joined #salt
23:24 TheGerm joined #salt
23:26 Ryan_Lane jimklo: tar isn't available in the state, only the module
23:27 Ryan_Lane so, if you want to use it, you'll need to do so via the state module.run
23:27 Ryan_Lane jimklo: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.module.html#salt.states.module.run
23:27 jimklo i just figured that out... thanks... Ryan_Lane... how can I use the salt:// prefix for the file that should be untared?
23:28 jimklo or can i?
23:28 Ryan_Lane I think it should work with the module
23:29 jimklo I get the following error: tar: salt\://ant-install/apache-ant-1.9.3-bin.tar.bz2: Cannot open: Input/output error
23:29 mateoconfeugo joined #salt
23:29 Ryan_Lane ah. not sure, then
23:29 jimklo does the tar file go in the pillar or the state?
23:29 TheGerm joined #salt
23:30 patrek_ joined #salt
23:31 schristensen joined #salt
23:31 jimklo oh... i think i kind of got it... think i need to manage the file then throw it into temp on minion
23:37 darless__ joined #salt
23:38 rglen joined #salt
23:40 meteorfo_ joined #salt
23:41 meteorfo_ joined #salt
23:42 kiorky joined #salt
23:45 eliasp joined #salt
23:48 Teknix joined #salt
23:48 rojem joined #salt
23:51 andrej /msg basepi no worries, you're welcome  ... love the project! :)  And while I'm not much of a developer I'm happy to chase bugs ;}
23:51 andrej meh
23:51 andrej fat fingers
23:53 KyleG psht
23:56 * whiteinge rightly blames Heartsbane right back
23:59 Ryan_Lane whiteinge: any idea what's going on with this failure? http://jenkins.saltstack.com/job/salt-pr-rs-ubuntu12.04/2875//console

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