Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-11-03

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

All times shown according to UTC.

Time Nick Message
00:00 pipps99 joined #salt
00:10 hasues joined #salt
00:12 hasues left #salt
00:15 netcho joined #salt
00:19 fracklen joined #salt
00:24 fannet joined #salt
00:25 Klas joined #salt
00:40 bluenemo joined #salt
00:40 flowstate joined #salt
00:45 asoc_ joined #salt
01:02 swills joined #salt
01:03 infrmnt joined #salt
01:05 subsignal joined #salt
01:09 lordcirth joined #salt
01:09 lordcirth Why do the docs for 2016.3.3 have options marked 'new in 2016.3.4' ?
01:09 edrocks joined #salt
01:09 hemebond Link?
01:10 hemebond Might be because the docs are actually linked to 2016.3 and applies to all minor versions. Not sure.
01:10 systo joined #salt
01:10 lordcirth hemebond, https://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html#salt.states.file.replace
01:10 lordcirth ignore_if_missing (which I'd rather like to use)
01:11 hemebond If you click on 2015.8.12 you'll see it goes to 2015.8
01:11 hemebond So I think the docs are only grouped by the major version.
01:12 systo joined #salt
01:14 lordcirth ah
01:15 bltmiller joined #salt
01:16 lordcirth Anyone know a convenient way to sort a state.apply return by runtime?
01:16 swills joined #salt
01:16 lordcirth ie, tasks that took the longest sorted first
01:17 hemebond Oh. Can you use json return format and parse it?
01:18 lordcirth that'd probably be a good start
01:19 lordcirth Don't really need it yet, but if things keep growing
01:19 abednarik joined #salt
01:19 lordcirth rather like 'systemd-analyze blame'
01:20 swills joined #salt
01:22 hemebond Oh, that blame looks neat.
01:23 gableroux joined #salt
01:23 lordcirth It's quite handy
01:23 lordcirth A similar --blame option or so on a salt state.apply would be pretty awesome
01:24 hemebond I feel like you could just pipe the output into something.
01:24 hemebond There's some json parser than someone mentioned here before.
01:25 hemebond jq I think
01:29 akhter joined #salt
01:33 q1x joined #salt
01:33 lordcirth hemebond, looks handy
01:38 flowstate joined #salt
01:56 akhter joined #salt
01:57 jas02_ joined #salt
01:58 hemebond cat highstate.json | jq '[.db9cep01 | .[]]|sort_by(.duration)'
01:58 hemebond I think I'll probably use this in the future. Neat tool that jq.
01:59 DEger joined #salt
02:04 barajasfab joined #salt
02:06 cro joined #salt
02:08 swills joined #salt
02:09 lordcirth hemebond, does that work? Cool!
02:09 hemebond Sure does.
02:10 hemebond That's with only one minion highstate though. Will need changing to work with more than one minion.
02:10 lordcirth sure.  1 is fine for my use, though
02:16 netcho joined #salt
02:20 fracklen joined #salt
02:22 k_sze[work] joined #salt
02:22 gableroux joined #salt
02:27 gableroux joined #salt
02:29 evle joined #salt
02:30 nicksloan joined #salt
02:37 netcho joined #salt
02:39 theblazehen joined #salt
02:39 catpigger joined #salt
02:40 flowstate joined #salt
02:40 mavhq joined #salt
02:40 sebastian-w joined #salt
02:48 ilbot3 joined #salt
02:48 Topic for #salt is now Welcome to #salt! | Latest Versions: 2015.8.12, 2016.3.3 | Support: https://www.saltstack.com/support/ | Logs: http://irclog.perlgeek.de/salt/ | Paste: https://gist.github.com/ (please don't multiline paste into channel) | See also: #salt-devel, #salt-offtopic | Ask with patience as we are volunteers and may not have immediate answers
02:51 onlyanegg joined #salt
02:59 jas02_ joined #salt
03:03 jhauser joined #salt
03:10 k_sze[work] Just me or salt 2016.3.3 can't find SysV init style services on Debian 8?
03:11 k_sze[work] e.g. service.running with a SysV style service wouldn't work, salt would say the named service is not available.
03:12 edrocks joined #salt
03:14 systo joined #salt
03:16 adongy joined #salt
03:17 bastiandg joined #salt
03:17 sp0097 joined #salt
03:25 subsignal joined #salt
03:35 dyasny joined #salt
03:37 flowstate joined #salt
03:43 lompik joined #salt
03:44 blu__ joined #salt
03:57 systo joined #salt
04:03 nawwmz salt-cloud users, where do you guys usually store your map files?
04:03 jhauser joined #salt
04:04 hemebond /srv/salt/cloud/...
04:08 whytewolf I store my map files in a fileing cabnet in my cartography room
04:16 fannet joined #salt
04:18 bantone I should probably ak this during the day, but I just got laid off from Dell/EMC due to the merging
04:18 bantone so I'm looking for job opportunities to utilize my saltstack skills
04:20 fracklen joined #salt
04:24 whytewolf unforchantly not a lot of job hunting happens here day or night
04:25 whytewolf recruters don't spend a lot of time in IRC
04:32 whytewolf coarse i shouldn't say never. i have had one interview based out of this channel almost a year ago.
04:32 bltmille_ joined #salt
04:33 __number5__ maillist should have a slightly better chance than IRC
04:35 whytewolf there are jobs posted on there now and then. mostly contract work. best case would be to friend up salt people on linkedin
04:35 fracklen joined #salt
04:37 systo joined #salt
04:37 flowstate joined #salt
04:38 netcho joined #salt
04:53 netcho joined #salt
04:53 bantone yeah just word of mouth...I know a few of the salt guys in here personally
04:53 bantone just throwing it out there to see what hits :)
04:55 swills joined #salt
04:55 swills joined #salt
04:55 systo joined #salt
04:56 whytewolf always a good plan, good luck :) hope you find something soon
04:56 sp0097 left #salt
05:01 jas02_ joined #salt
05:04 jhauser joined #salt
05:04 bltmiller joined #salt
05:13 bltmiller joined #salt
05:14 ivanjaros joined #salt
05:16 nawwmz hemebond: what is the purpose of /etc/salt/cloud.maps.d
05:16 hemebond uh, probably maps.
05:17 nawwmz just wondering whats the difference between storing them there vs /srv/salt/...
05:17 hemebond Oh, I just put stuff under /srv/salt/ for my convenience.
05:18 nawwmz ah kk
05:35 samodid joined #salt
05:46 cyteen joined #salt
05:46 Ni3mm4nd joined #salt
06:05 jhauser joined #salt
06:12 impi joined #salt
06:17 nawwmz joined #salt
06:17 stooj joined #salt
06:21 west575 joined #salt
06:22 zulutango joined #salt
06:28 stooj joined #salt
06:35 sh123124213 joined #salt
06:36 swills_ joined #salt
06:38 flowstate joined #salt
06:40 prince joined #salt
06:40 prince hi
06:41 Guest18898 I am working on influxdb_returner. it has few issues. But I am stuck now.. I am not able to retrieve _options from the config file..
06:42 Guest18898 I looked at the __init__.py file in salt.returners.. tried debugging why am I not able to fetch the configs..
06:42 Guest18898 no luck..
06:42 Guest18898 @here
06:42 Guest18898 @all
06:43 sh123124213 joined #salt
06:46 Edgan joined #salt
06:47 ivanjaros joined #salt
06:48 Guest18898 @armadillo @__alex:
06:48 Guest18898 @alekti
06:48 ivanjaros3916 joined #salt
06:49 DEger joined #salt
06:50 stooj joined #salt
06:50 felskrone joined #salt
06:51 zer0def joined #salt
06:51 jhauser joined #salt
06:56 stooj joined #salt
06:59 m4rx joined #salt
06:59 Guest18898 nevermind.. got it fixed.. Thanks..
06:59 Guest18898 :)
07:00 netcho joined #salt
07:02 stooj joined #salt
07:02 jas02_ joined #salt
07:11 bocaneri joined #salt
07:16 edrocks joined #salt
07:19 Elsmorian joined #salt
07:20 sgo_ joined #salt
07:27 nidr0x joined #salt
07:28 yuhlw____ joined #salt
07:37 flowstate joined #salt
07:40 keimlink joined #salt
07:48 amontalban joined #salt
07:48 amontalban joined #salt
07:50 CEH joined #salt
07:53 sgo_ joined #salt
07:54 toanju joined #salt
07:54 stooj joined #salt
07:58 fracklen joined #salt
07:58 fracklen_ joined #salt
07:59 fracklen joined #salt
08:00 sebastian-w joined #salt
08:03 jas02_ joined #salt
08:09 samodid joined #salt
08:18 fannet joined #salt
08:25 zer0def joined #salt
08:27 ronnix joined #salt
08:41 dariusjs joined #salt
08:41 sh123124213 joined #salt
08:55 fannet joined #salt
09:00 mikecmpbll joined #salt
09:03 geomacy joined #salt
09:13 Rumbles_ joined #salt
09:16 ecdhe joined #salt
09:19 dariusjs joined #salt
09:24 kbaikov joined #salt
09:30 jas02_ joined #salt
09:31 mikecmpbll joined #salt
09:32 jfelchner joined #salt
09:33 jfelchner Hey all, I'd like to access a pillar key so that I can use `file.manage` `contents_pillar`, however I need the key that I lookup to be dynamic.  Doesn't look like that's possible?
09:35 SaltyVagrant_ joined #salt
09:35 jfelchner nm super simple
09:36 jfelchner Just needed to rubber duck it :)
09:43 nebuchadnezzar joined #salt
09:53 N-Mi joined #salt
09:53 N-Mi joined #salt
09:55 dariusjs joined #salt
09:59 Cadmus joined #salt
10:03 akhter joined #salt
10:05 jas02 joined #salt
10:09 jas02__ joined #salt
10:13 keimlink joined #salt
10:13 ecdhe joined #salt
10:14 cdubmkvi joined #salt
10:16 amcorreia joined #salt
10:17 edrocks joined #salt
10:21 sh123124213 joined #salt
10:37 jas02 joined #salt
10:43 jas02 joined #salt
10:54 geomacy joined #salt
11:01 abednarik joined #salt
11:09 netcho joined #salt
11:10 mbersoult joined #salt
11:10 mbersoult Hello
11:14 mbersoult I have an issue, I have installed the ubuntu package of salt using repo.saltstack.com. When I list loaded modules, I does not found the module composer, anyone can help me ?
11:17 scoates joined #salt
11:27 abednarik joined #salt
11:44 keimlink joined #salt
11:52 amontalban joined #salt
12:01 akhter joined #salt
12:03 netcho joined #salt
12:07 jas02_ joined #salt
12:12 sgo_ joined #salt
12:15 dariusjs joined #salt
12:15 ronnix joined #salt
12:21 scoates joined #salt
12:26 edrocks joined #salt
12:38 DammitJim joined #salt
12:44 DammitJim joined #salt
12:45 west575 joined #salt
12:55 N-Mi joined #salt
12:55 N-Mi joined #salt
12:56 Ni3mm4nd joined #salt
12:57 pirulo joined #salt
12:58 abednarik joined #salt
12:59 Tanta joined #salt
13:03 pirulo I'm trying to see the event job on the minion using the eventlisten.py and I'm only getting back a warning but not see it the job at all.
13:04 pirulo python eventlisten.py -n centos6-host1.net WARNING: Events will not be reported (not able to access /var/run/salt/minion) /var/run/salt/centos6-host1.net/minion_event_d41d8cd98f_pub.ipc
13:05 Tanta do ls -la /var/run/salt and look at the perms
13:06 pirulo $ ls -la /var/run/salt total 12 drwxr-xr-x.  3 root root 4096 May 29  2015 . drwxr-xr-x. 40 root root 4096 Oct 29 15:18 .. drwxr-xr-x.  2 root root 4096 Oct 29 15:18 minion
13:07 pirulo it shows a minion folder
13:07 pirulo under the folder minion two files 1- minion_event_fcee64f88f_pub.ipc 2- minion_event_fcee64f88f_pull.ipc
13:08 jas02_ joined #salt
13:10 Tanta what user does salt run as, and what user owns those files? what are the perms on those files?
13:10 flowstate joined #salt
13:11 Tanta the error you posted is a read permission error
13:11 Tanta when the salt process on the master goes to access the event socket
13:11 Tanta unless the error is a red herring and this is all a waste of time, maybe someone else knows this one
13:14 ravenx joined #salt
13:14 ravenx i have salt-master running as the user:  'salt-user' on my server.  i also have jenkins on the same server
13:14 ravenx i am using the publisher_acl: https://docs.saltstack.com/en/latest/topics/eauth/access_control.html#examples with the name jenkins instead of fred
13:14 ravenx but i still cannot run any salt commands like test.ping as jenkins.  i get:  "Failed to authenticate! This is most likely because this user is not permitted to execute commands, but there is a small possibility that a disk error occurred (check disk/inode usage)."
13:15 ravenx i have even restarted salt-master a few times.
13:15 flowstate joined #salt
13:15 ravenx my salt is:  salt 2016.3.3 (Boron)
13:20 pirulo salt is running as a root
13:20 ravenx o_O?
13:21 ravenx my salt-master is running as "salt-user"
13:22 pirulo well on the master its: saltusr, but in the minion no user salt*
13:22 sgo_ joined #salt
13:23 pirulo does any one has used eventlisten.py?
13:23 ravenx i dont quite follow
13:25 CEH joined #salt
13:29 Tanta pirulo, then do chown -R saltusr:saltusr /var/run/salt
13:29 Tanta and try it again
13:32 Tanta I believe Salt assumes it is running as root
13:33 pirulo my minions don't have a user: saltusr ONLY the master. when I execute the command invalid user. I can ping my minions. Now you telling me that each minion needs to have a salt user on the minion to able to run the eventlisten.py?
13:33 subsignal joined #salt
13:34 Tanta on the salt master dude
13:34 Tanta your master can't access the IPC sockets it needs because the /var/run folder is owned by root
13:35 Tanta and the salt master process runs as saltusr, according to you
13:38 jas02 joined #salt
13:38 pirulo Yes, that is true
13:40 ravenx can someone help me with my publisher_acl problem?
13:41 cscf ravenx, can you run commands as 'jenkins' manually?
13:42 ravenx cscf: with salt?  i cannot
13:42 cscf ravenx, no, with bash.  What's the default shell for the jenkins user?
13:42 ravenx i have tried  salt 'server' test.ping
13:42 ravenx yup, i can
13:42 ravenx i can scp, cp, mv, ls, etc.
13:42 ravenx echo $SHELL says that jenkins has a bin/bash
13:46 numkem joined #salt
13:47 ravenx so my question is, does my saltmaster need to run as root, or...?
13:47 nicksloan joined #salt
13:49 racooper joined #salt
13:52 mavhq joined #salt
13:53 cscf ravenx, salt-master should be able to run as non-root, though I don't see why you'd bother
13:54 jas02_ joined #salt
13:54 nkuttler i've been using it as non-root for years
13:55 nkuttler s#using#running#
13:56 KingOfFools joined #salt
13:56 ravenx for safety reasons i suppose
13:56 dkrae joined #salt
13:56 mapu joined #salt
13:56 ravenx so i just tried something, which was tos tart salt-master as root.
13:57 ravenx and lo and behold, publisher_acl rules start applying all of a sudden.
13:57 ronnix joined #salt
13:58 ronnix_ joined #salt
13:59 gableroux joined #salt
13:59 toanju joined #salt
14:01 cscf Normally I would agree that you don't run services as root, but salt in practice has root to your whole infrastructure, so priv escalation is trivial
14:02 ravenx true.
14:02 ravenx lesson learned...;_; the hard way
14:02 ravenx i dont know if this is by design or a bug
14:02 ravenx now i dont know whether or not to write up a GH issue
14:02 cscf ravenx, what?
14:02 ravenx the fact that publisher_acl doesn';t get realized if i run it as a non-root user
14:02 ravenx with "it" being salt-master
14:03 cscf Most likely you're missing a permission somewhere
14:03 ravenx i see.
14:03 cscf Maybe you could use strace or just max debug verbosity to trace what it's failing to access
14:04 ALLmightySPIFF joined #salt
14:04 ravenx good idea
14:04 ravenx thanks cscf
14:05 cscf np.  Salt's error messages are sometimes lacking
14:09 Tanta the error message had exactly what you needed
14:09 Tanta permissions error
14:09 Tanta like I said an hour ago
14:09 cscf Tanta, it was never clear which person you were talking to
14:09 Tanta oh, that was a different guy, my bad
14:10 Tanta I don't know anything about ravenx's issue
14:16 tapoxi joined #salt
14:18 Brew joined #salt
14:19 toanju joined #salt
14:19 fracklen joined #salt
14:20 hasues joined #salt
14:20 hasues left #salt
14:20 pirulo thanks Tanta, I will try to figure it out.  because running the bootstrap and run the minion and the minion is ready to do his job. My minions of course has the master IP nothing else.
14:21 Tanta did you try changing the ownership of the /var/run/salt folder on the master? or running the salt-master process as root?
14:23 nawwmz joined #salt
14:26 patarr joined #salt
14:31 keltim joined #salt
14:31 netcho joined #salt
14:34 DammitJim joined #salt
14:36 flowstate joined #salt
14:38 flowstate joined #salt
14:38 mpanetta joined #salt
14:39 vegasq joined #salt
14:40 toanju joined #salt
14:40 dariusjs joined #salt
14:41 tiwula joined #salt
14:44 vegasq joined #salt
14:45 abednarik joined #salt
14:46 JamieH joined #salt
14:46 DammitJim is there a way to  just list VMs through salt-cloud?
14:48 nicksloan joined #salt
14:48 ProT-0-TypE joined #salt
14:52 ravenx left #salt
14:53 DEger joined #salt
14:58 DammitJim do you guys ever just apply multiple modules to all your minions?
14:58 DammitJim or do you do 1 at a time to be able to see what is going on?
15:03 a_ghost_irl joined #salt
15:06 Tanta I have a set of states called "common" that get applied to every machine
15:06 Tanta stuff like basic security, user accounts, basic packages, hardeninig
15:07 Cidan joined #salt
15:07 Tanta another place I worked had "minimal", "server", "hardened-server" style profiles for different classes of machine
15:08 Tanta so you could select from several pre-defined base state sets
15:09 jas02 joined #salt
15:11 DEger joined #salt
15:14 vegasq joined #salt
15:16 netcho_ joined #salt
15:24 KennethWilke joined #salt
15:24 pipps joined #salt
15:24 DEger joined #salt
15:26 fas3r joined #salt
15:26 fxdgear joined #salt
15:26 fas3r Hello
15:26 danlsgiga joined #salt
15:26 nawwmz where do people typically place their salt map files? I see a /etc/salt/cloud.maps.d which may sound like an appropriate place but do they need to be there?
15:27 danlsgiga hey there folks... I have couple runners in my Salt master that I invoke through salt-api using the /run context. That is nicely working but I have a requirement to have specific http return codes for specific situations and I don't see how I can return a 50x error from the runner where the salt-api will return to the user
15:27 danlsgiga does anyone have any idea how I can do that?
15:27 komputes joined #salt
15:28 danlsgiga nawwmz: Nope... it is just a convenient place to be... salt-cloud map files are targetted using the filesystem path... salt-cloud -m /etc/salt/cloud.maps.d/mymap.conf
15:28 onlyanegg joined #salt
15:28 nawwmz yeah thats what i kinda thought to myself too, you have to define the path anyway
15:31 kojiro joined #salt
15:31 nawwmz does anyone use salt-cloud with openstack here?
15:33 rherna joined #salt
15:33 mavhq joined #salt
15:36 fas3r I try to setup Salt and openstack, using the driver nova. In the cloud.providers.d folder, I created a conf file for my openstack. In it, I set the identity_url, compute_region, user, password, tenant and driver. When I try to query openstack with "salt-cloud -Q -l debug" I get : Failed to execute 'nova.list_nodes()' while querying for running nodes: __init__() got an unexpected keyword argument 'auth_token', any idea ?
15:38 gtmanfred fas3r: can you provide the full stack trace and your sanatized configs?
15:39 fas3r gtmanfred: you mean salt-cloud --version and the conf file for the provider ?
15:39 morissette joined #salt
15:40 fas3r I will pastebin the debug.
15:40 gtmanfred no, the stack trace from salt-cloud -Q -l debug
15:40 gtmanfred the salt-cloud --versions-report would be good too
15:42 gtmanfred fas3r: please put it in gist
15:42 ProT-0-TypE I'm trying to make the gitfs example to work but:
15:42 ProT-0-TypE [ERROR   ] Error occured fetching gitfs remote 'https://github.com/saltstack-formulas/salt-formula.git': Failed to resolve address for https: Name or service not known
15:42 ProT-0-TypE on ubuntu 16.04
15:42 ProT-0-TypE anyway having the same issue?
15:42 ProT-0-TypE I'm following: https://docs.saltstack.com/en/latest/topics/tutorials/gitfs.html
15:43 ProT-0-TypE (masterless minion)
15:43 gtmanfred add git+ in front of it? or use git://github.com/saltstack-formula/salt-formula?
15:43 ProT-0-TypE I'm following the example in the tutorial
15:43 wangofett joined #salt
15:43 ProT-0-TypE basically it says:
15:43 ProT-0-TypE gitfs_remotes:
15:43 ProT-0-TypE - https://github.com/saltstack-formulas/salt-formula.git
15:44 fas3r gtmanfred: https://gist.github.com/anonymous/49f874b9543c313597df0d80bfd6d2dc
15:44 gtmanfred fas3r: and can you share your provider config?
15:44 fas3r gtmanfred: let me know if you need something else. Just in a case, I'm able to communicate with openstack with nova command ( just in a case ).
15:44 fas3r oki
15:46 fas3r gtmanfred: https://gist.github.com/anonymous/09784b3ec12149c6a16364120cb7f161
15:47 yuhlw____ left #salt
15:47 gtmanfred hrm, i have never seen that happen, can you open a /issue on salt about that?
15:51 fracklen joined #salt
15:54 nawwmz is there a way for me to run a salt command which will allow me to test basic functionality of salt-cloud -> openstack provider without having to run salt-cloud -m map_file ?
15:54 nawwmz I see some options like "salt.cloud.clouds.openstack.list_nodes_full(conn=None, call=None)" but I dont know how to translate that into a command line
15:56 gtmanfred salt-cloud -f list_nodes_full <provider>
15:56 nawwmz nice thx gtmanfred!
15:58 haam3r joined #salt
15:58 XenophonF hey all - i'm having problems with a hanging salt-minion on Windows
15:58 XenophonF it's currently stalled in the middle of a "state.apply" job, running salt.minion from salt-formula
15:59 XenophonF python is at 100% utilization of one core (I guess it's single-threaded on Windows)
15:59 XenophonF anyone have any ideas on what i could try to debug this?
16:00 XenophonF I think I have a recent version of process explorer installed.
16:01 fas3r gtmanfred: you mean on salt github ?
16:01 gtmanfred yeah
16:02 gtmanfred so we can take a look at the bug there, unfortunatley I do not have time today to look deeply into that bug
16:11 jas02 joined #salt
16:15 nawwmz ugga, getting this error "he cloud driver, 'openstack', configured under the 'fr1' cloud provider alias, could not be loaded."
16:15 XenophonF according to the stack trace, Python's stuck somewhere in PyUnicodeUCS2_Replace
16:15 bltmiller joined #salt
16:15 XenophonF it's in a file.recurse state atm, I think
16:15 nawwmz in my cloud.providers.d/openstack.conf, my driver is driver: openstack
16:15 nawwmz anyone seen this problem before?
16:16 XenophonF I killed every other thread attached to the Salt process but that didn't seem to do anything
16:16 gtmanfred nawwmz: can you provide your sanatized configs and the full stack trace from -l debug in a gist?
16:16 nawwmz gtmanfred: will do
16:17 rburkholder joined #salt
16:17 fas3r gtmanfred: ok done. https://github.com/saltstack/salt/issues/37443 thanks
16:17 saltstackbot [#37443][OPEN] salt-cloud and openstack :  "got an unexpected keyword argument 'auth_token' " | Description of Issue/Question...
16:18 gtmanfred fas3r: can you also put the version of nvoaclient you are using in that gist
16:18 gtmanfred github issue*
16:19 fas3r gtmanfred: I just notice that I have 2 novaclient, the one from pip and the one from yum.
16:20 gtmanfred uninstall the one from pip
16:20 gtmanfred and see if it works
16:20 gtmanfred my guess is something changed in the new novaclient
16:20 gtmanfred then do a yum reinstall python-novaclient
16:20 gtmanfred my goal is to move away from using novaclient directly in Nitrogen and instead use shade for everything
16:21 fas3r yes !
16:21 fas3r working ! :)
16:21 gtmanfred cool
16:22 fas3r gtmanfred: I also have salt-cloud installed from pip and the one from yum
16:22 gtmanfred don't use salt-cloud from pip, that is an old old old package
16:23 fas3r I wanted to try a sdk made in python for this openstack provider and I had to install requirement via pip. Maybe I should remove it too ?
16:23 gtmanfred salt-cloud was included in the salt code back in 2014.2 or whatever that release was
16:23 fas3r oki.
16:23 fas3r Thanks for the help gtmanfred :)
16:23 gtmanfred openstacksdk should be available in the openstack rdo release packages
16:23 gtmanfred no problem
16:24 gtmanfred yeah, we have a lot of cruft in the nova stuff, mostly because novaclient and the python clients for openstack have changed so much in the last year
16:24 gtmanfred it also doesn't currently work with keystone v3 authentication
16:25 fas3r oki.
16:26 nawwmz gtmanfred: http://pastebin.com/Kpk7JWHN
16:26 krymzon joined #salt
16:27 ronnix joined #salt
16:27 gareth_ joined #salt
16:27 gareth_ windows join_domain with restart: True joins domain but reboots forever. Any ideas? Thanks
16:28 renoirb I have a question about _modules/foo.sls when we import them using GitFS and pygit2?
16:28 renoirb If I have formula loaded through GitFS, how can I load them?
16:29 shaggy_surfer joined #salt
16:30 renoirb e.g. I have a state module helper in https://github.com/renoirb/renoirb-salt-states/blob/xenial/_modules/slsutils.sls  which get's loaded using pygit2 and GitFS. But states can't work when using   slsutils.some_method  because it's not copied locally.   But I have a dynamic grain in https://github.com/renoirb/renoirb-salt-states/blob/xenial/_grains/role.py which  does copied to /var/cache/salt/minion/extmods/grains/role.py.
16:30 renoirb Is it because _modules/slsutils.sls don't get synced, and .py does?
16:31 abednarik joined #salt
16:31 shaggy_surfer how do I force a service to restart when the files are installed post install of the package?  I have nginx being installed as a package and after I have salt remove the default vhost files and I add some templated vhost files.  I have a "watch" on those new files.  However on the very first run it doesn't restart nginx after the vhosts are created.  If I modify a vhost on the server and then run again then it does.  So the watch works on the second run,
16:31 nawwmz gtmanfred: ok i got closer, I installed the python-netaddr which salt error mentioned :D
16:31 nawwmz thx though!
16:35 sh123124213 joined #salt
16:41 gtmanfred cool
16:42 gtmanfred renoirb: run the saltutil.sync_all module
16:42 Ni3mm4nd joined #salt
16:42 gtmanfred and it will sync it to the minions
16:44 Ni3mm4nd joined #salt
16:44 mpanetta_ joined #salt
16:47 pipps joined #salt
16:47 swills joined #salt
16:47 edrocks joined #salt
16:48 sh123124213 gtmanfred : I was hoping to see 2006.3.4 yesterday :(
16:49 gtmanfred it is coming
16:49 iggy that's why I always tell everyone 2 weeks
16:49 pipps joined #salt
16:50 iggy when you say today, you get people coming back in first thing next morning with sad faces
16:50 sh123124213 well, I heard 2 weeks two weeks b4 :P
16:50 gtmanfred fair point
16:51 gtmanfred i will only ever say `when it is ready` from now on
16:51 gtmanfred learned my lesson
16:51 sh123124213 oh comeon, its not that I'm blaming gtmanfred about its not being released
16:51 renoirb gtmanfred, I did that. And the file on the remote in _modules/slsutils.sls doesn't get written like _grains/roles.py does to /var/cache/salt/minion/extmods/grains/role.py
16:51 gtmanfred :P
16:52 renoirb My question is then, does _modules/foo.sls <- the extension make a difference?
16:52 sh123124213 just got excited when I heard it would be yesterday
16:52 gtmanfred renoirb: because it is slsutils.sls, make it slsutils.py
16:52 renoirb what I was wondering.
16:52 renoirb thanks I'll try that right away
16:52 gtmanfred ahh, ok
16:52 gtmanfred yeah i missed the extension in the first post
16:55 renoirb it worked.
16:55 MTecknology Good morning friends and people and others! :D
16:56 renoirb So, yeah, the file extension makes a difference.
16:56 gtmanfred yar, it is a python module, they are pretty persnickity about their extensions
16:57 renoirb Yeah, I've copy-pasted from another project and made the mistake for the file extension.  Reason for bugs: copypasta
16:59 gtmanfred :)
17:00 tapoxi joined #salt
17:00 nicksloan joined #salt
17:04 lompik joined #salt
17:07 akhter joined #salt
17:08 akhter joined #salt
17:08 mapu joined #salt
17:10 wangofett joined #salt
17:11 nawwmz hi all, do i need to source the openstack file on the salt-master or do simply adding the credentials in the cloud.providers work?
17:11 MTecknology gtmanfred: :)
17:11 jas02 joined #salt
17:14 gtmanfred you add the credentials to the cloud.providers
17:15 nawwmz thx gtmanfred
17:16 Edgan joined #salt
17:16 nawwmz gtmanfred: think if i toss a pastebin, you can help me figure out a connection problem? i got a little closer with adding the dependency but still getting an error where changing my user/pass doesnt quite help
17:17 gtmanfred you have to add /tokens to the end of your identity_url
17:17 gtmanfred looking at your previous pastebin
17:17 nawwmz oh no i dont
17:17 nawwmz let me try that
17:18 gtmanfred if you use the openstack you have to add that, if you use the nova driver you don't
17:18 gtmanfred it is dumb, but that is how those libraries works
17:18 nawwmz ah interesting ok let me try
17:19 nawwmz okay much closer, getting an error about "Something seems inconsistent
17:19 nawwmz but this is good, at least im getting an error from the provider
17:19 nawwmz thx for the tip gtmanfred!
17:20 degorenko joined #salt
17:22 foundatron joined #salt
17:24 pipps joined #salt
17:25 nicksloan joined #salt
17:26 cscf salt-ssh gives permission denied, even though I can ssh to the machine
17:27 keimlink joined #salt
17:29 DammitJim joined #salt
17:29 zillah joined #salt
17:29 cscf Oh, it's using it's own key instead of the root keyring
17:31 gtmanfred you have to specify which key to use, with either --priv or priv: in the roster
17:33 subsignal joined #salt
17:34 cscf Ah.  I ended up doing  ssh-copy-id -i /etc/salt/pki/master/ssh/salt-ssh.rsa.pub <host>
17:35 additionalgeneri joined #salt
17:35 additionalgeneri hello!
17:36 cscf hi
17:37 additionalgeneri maybe you can point me in the right direction
17:37 XenophonF I've narrowed down the high CPU utilization to this file.recurse state, https://github.com/saltstack-formulas/salt-formula/blob/master/salt/minion.sls#L8
17:37 additionalgeneri I need to generate a script using pillar data and actually lay down a file on disc. What's the best way to do that
17:37 XenophonF I suspect that it's tickling a bug in python.exe's string library.
17:38 XenophonF I have no idea how to debug this further.
17:38 jas02 joined #salt
17:38 cscf additionalgeneri, what sort of script?
17:38 jas02_ joined #salt
17:39 additionalgeneri one liner. /path/to/executable --option {{ grain }} --more-options {{ more_grains}}
17:39 additionalgeneri and so on
17:39 additionalgeneri I need the state to generate a file with that one line and chmod it. I'm wondering the best way to get that info into that one line
17:40 cscf additionalgeneri, file.managed with lots of jinja ?
17:41 Sketch for a one liner, you can just do file.managed and instead of source:, do contents: <yourscript>
17:41 bltmiller does the survey runner support node group targetting? https://docs.saltstack.com/en/latest/ref/runners/all/salt.runners.survey.html#salt.runners.survey.diff
17:41 bltmiller seems like it can only use globs
17:41 cscf true
17:42 ryandamour joined #salt
17:42 additionalgeneri thanks, yall!
17:42 ryandamour exit
17:42 subsignal joined #salt
17:43 icehac joined #salt
17:43 samodid joined #salt
17:44 nawwmz gtmanfred: for openstack, do you prefer the nova driver or openstack driver?
17:44 gtmanfred nawwmz: the openstack driver is getting deprecated in Nitrogen
17:44 gtmanfred bltmiller: looking at the code, it only does glob
17:45 bltmiller gtmanfred: hm, how hard would it be to add support for that in?
17:45 radhac joined #salt
17:46 win_salt does anyone know why salt makes the _schedule.conf file in the minion.d directory?
17:46 gtmanfred bltmiller: not too difficult, just need to start accepting the expr_form like the manage runner functions do in the carbon release, and then pass it through and match on it
17:46 gtmanfred win_salt: to make sure the mine gets updated
17:48 akhter joined #salt
17:48 win_salt its kind of annoying that it can act like residue if you configure a salt schedule in the config directory
17:50 fas3r joined #salt
17:53 bltmiller gtmanfred: cool thanks, I might take a stab at that. a) are PRs accepted? b) can you point me to that code?
17:53 iggy there's an old issue about that (likely closed at this point)
17:53 fas3r Hello, how to specify the network ID when using cloud.create ? where can I get the list of all the argument that cloud.create accept ? or that I can use if the profile ( when using salt-cloud). Thank you.
17:54 beowuff joined #salt
17:55 gtmanfred bltmiller: https://github.com/saltstack/salt/tree/develop/runners/survey.py
17:55 gtmanfred bltmiller: and yes, we accept lots of pull requests
17:56 gtmanfred fas3r: cloud.create takes all the arguments that would normally go into the profile file
17:56 gtmanfred so, it depends on the provider
17:58 fas3r nova :)
17:59 gtmanfred so anything that would go in a profile, https://docs.saltstack.com/en/carbon/ref/clouds/all/salt.cloud.clouds.nova.html cloud.create takes
17:59 gtmanfred https://docs.saltstack.com/en/latest/topics/cloud/misc.html
18:02 Tanta {%- set access_key = salt['cmd.run']('awk \'/keyid:/ { print $2 }\' /etc/salt/minion') %}
18:02 Tanta this sort of thing should be banned
18:09 edrocks joined #salt
18:10 fas3r gtmanfred: ok, is it the same to get the list of networks ? does it depend on the provider/driver ?
18:11 gtmanfred yeah, that is what I said... depending on which driver you are using... anything that can go in the profile can go into the cloud.create
18:11 gtmanfred so like... vmware can have different stuff in the profile than nova or ec2 ... etc
18:11 fas3r I meant if I want to list the network available from salt-cloud
18:12 gtmanfred still probably depends on the provider
18:12 gtmanfred for nova the sub command for listing networks is salt-cloud -f network_list <provider>
18:13 kojiro tfw you spend 40 minutes debugging a problem and it turns out it was a silly typo in a pillar.
18:13 gtmanfred but the ec2 driver doesn't have a function to list networks
18:14 fas3r gtmanfred: that's what I've done. I get "There was an error running the function: Not found (HTTP 404) (Request-ID: req-1e7b4f1b-5e7f-4f65-b411-8d4f24fa3133)".
18:15 gtmanfred yeah, there are some problems with the network stuff in the nova driver, because it uses novaclient to list it, when a lot of cloud providers use neutron... it was modeled after rackspace... we need to update that
18:16 kiorky joined #salt
18:19 pipps99 joined #salt
18:21 jas02_ joined #salt
18:23 fracklen joined #salt
18:31 infrmnt joined #salt
18:43 fas3r gtmanfred: last question, if I want to apply a state when the created VM, I need to also add it in the profile ?
18:43 gtmanfred what? apply a state after it has been created?
18:44 gtmanfred you could do an orchestrate, where the minion is created in the first step, then you apply a state afterwards, or you can use startup_states
18:45 Trauma joined #salt
18:48 fas3r I was thinking about using userData in openstack and execute a script using curl and salt-api to execute the sates only once when the server boot for the first time.
18:48 gtmanfred you could do a post in the cloud-init file, which posts to the salt-api webhook, that gets dumped into the event bus and kick off a reactor that runs a state on the minion
18:49 _JZ_ joined #salt
18:49 pipps joined #salt
18:50 flowstate joined #salt
18:52 wangofett joined #salt
18:57 mpanetta joined #salt
18:57 shaggy_surfer joined #salt
19:02 mikea- joined #salt
19:03 fxdgear joined #salt
19:07 mikecmpbll joined #salt
19:10 flowstate joined #salt
19:13 nawwmz can cloud.profiles and cloud.providers be put somewhere else other than /etc/salt/ ?
19:20 devtea joined #salt
19:20 infrmnt joined #salt
19:21 sjorge joined #salt
19:21 sjorge joined #salt
19:22 jas02 joined #salt
19:24 jfindlay joined #salt
19:24 GreatSnoopy joined #salt
19:27 onlyanegg joined #salt
19:31 MTecknology gtmanfred: anything exciting to share?! :D)
19:36 sh123124213 joined #salt
19:39 jas02_ joined #salt
19:41 ruxu joined #salt
19:41 ahammond I'm seeing this: [ERROR   ] Data passed to highstate outputter is not a valid highstate return: {'devops-storage-sfo-stage-0': ['Pillar failed to render with the following messages:', "Failed to load ext_pillar git: 'git init' returned with exit code 128\nstderr: 'error: could not lock config file /var/cache/salt/master/pillar_gitfs/02e6bbd37cc95b6fde742bf25728194572fe2c51e34fa1336846f159a7d50f70/.git/config: File exists\nfatal: Could not set 'core.repos
19:41 ahammond itoryformatversion' to '0''"]}
19:42 ahammond I've tried stopping the salt master, clearing /var/cache/salt/master/pillar_gitfs
19:42 ahammond I'm kind of at a loss what this means and how to fix it.
19:43 nickg I have a weird issue where salt feels pip isnt properly installed.  I know it is properly installed, and in an strace I see salt finds all of the proper pip files.  What's weird is every server that has this issue shows arguments in ps for salt.. like " KeepAlive MinionProcessManager
19:44 nickg But the server that doesn't have this problem, doesn't show that.  They are both ubuntu 14.04.4, both running salt-minion 2016.3.3.  Both were salted by the same salt scripts, both launched from the same AWS Ubuntu 14.04 community ami.
19:44 sgo_ joined #salt
19:44 bocaneri joined #salt
19:45 nickg what's even more weird, if I start salt via  salt-minion -l debug  it doesn't have the pip issue
19:47 nawwmz can cloud.profiles and cloud.providers be put somewhere else other than /etc/salt/ ?
19:47 bltmiller joined #salt
19:47 nickg If I manually start salt-minion via: salt-minion -d  it works, but if I use the service handler, it doesn't work. I don't see any environmental details with ps
19:48 sp0097 joined #salt
19:49 ruxu joined #salt
19:50 amontalban joined #salt
19:51 RandyT trying to deploy a freebsd image on AWS. Finding that the process is ignoring ssh_username:
19:51 RandyT something I am forgetting? ssh process hammers on every possible linux image login, but does not use 'root' which is what I have assigned for the profile.
19:51 ruxu joined #salt
19:52 RandyT nawwmz: the answer is yes
19:52 RandyT just need to symlink it though
19:52 nawwmz RandyT: ahhh ok
19:52 RandyT I'm not aware of a way to set that root.
19:53 nawwmz makes sense, is that what people typically do?
19:53 RandyT cannot say... I am usually not in the "typical" camp... :-)
19:53 nawwmz heh
19:54 RandyT works well for me. I keep the configs under git scm
19:54 subsigna_ joined #salt
19:54 nawwmz RandyT: do you use git as a file_system?
19:55 joe__ joined #salt
20:01 subsignal joined #salt
20:01 RandyT nawwmz: I use git for some filesystems as well as s3
20:02 RandyT I don't use git for my states filesystem with one minor exception of a remote git repo
20:03 nawwmz is there any nuances with haviing states managed by git fs? were thinking of doing that
20:03 nickg I've completely unset my PATH and still running salt-minion manually works.  what would be different between the service handler and running salt-minion manually?
20:04 DammitJim joined #salt
20:04 RandyT nawwmz: no "nuances" from my perspective, but again, I'm not using a remote repo for that stuff and probably wouldn't.
20:04 DammitJim /join #python
20:04 RandyT works well for me to clone the repo to my master and use it that way.
20:04 RandyT when working on fixing state issues, makes it easy to generate diffs on the changes to capture back in the repo
20:08 ahammond nawwmz my experience with salt is that it is best to avoid anything you don't absolutely need. Right now we're having an issue with the pillar_gitfs stuff that makes absolutely no sense to me.
20:09 RandyT ^^
20:11 netcho_ joined #salt
20:11 sjorge joined #salt
20:11 sjorge joined #salt
20:16 LostSoul joined #salt
20:18 klausfiend MajObviousman / iggy / hemebond: my problem turned out to be moot, at least for now, because of a fix for the underlying package that was causing the problem
20:18 klausfiend (nss-softokn, specifically)
20:18 flowstate joined #salt
20:18 hemebond klausfiend: Was that the YUM environment variable thing?
20:18 klausfiend hemebond: yep
20:18 hemebond ????
20:19 klausfiend it affected other things that used the NSS libraries as well, and it turned out to be a known bug so i didn't have to get dirty things to happen
20:19 klausfiend still, it would be pretty sweet to be able to pass environment variables to the package install operations
20:19 klausfiend e.g., 'http_proxy'
20:21 hemebond http_proxy can be set via minion config I think.
20:21 hemebond There was/is a bug open for it.
20:21 hemebond Not sure if they implemented a generic system or if they specifically only supported http_proxy
20:22 MajObviousman I concur that would be mega useful
20:22 MajObviousman but my quickie fix is to make an ephemeral state file and run that instead of using an execution module
20:22 klausfiend it's not possible to chain salt functions on a command line either, is it?
20:23 klausfiend (we do some internal dev around our salt implementation, so i'm still not 100% what's "vanilla" vs. what's ours)
20:23 MajObviousman you mean, apart from cmd.run `command1 && command2 && command3` right?
20:23 klausfiend MajObviousman: indeed
20:23 klausfiend it didn't seem to me that that was possible so i didn't spend too much time in the weeds on it
20:24 MajObviousman I don't know of a way. Something requiring more than one change probably belongs in a state file
20:24 klausfiend yeah that was my conclusion
20:24 MajObviousman though I've been getting mega upset at the library of state files I inherited that there was no precondition checking or targetting checking
20:24 ivanjaros joined #salt
20:25 sjorge joined #salt
20:25 sjorge joined #salt
20:25 MajObviousman an admin ran a state to update /etc/ntpd.conf on everything ... including the two servers attached to stratum 1s and serving as time sources for everyone else
20:25 klausfiend ugh, gross
20:25 MajObviousman hello clock drift. Kerberos let us know it was unhappy in, oh, 2-3 days?
20:26 klausfiend you mean as critical services like LDAP start failing randomly all over the place
20:26 klausfiend heh
20:26 MajObviousman well, one box was drifting like mad and crossed +- 5min threshold in that amount of time. The rest were cruising along OK and had weeks or months before it'd become a problem
20:26 MajObviousman one of our remaining physical boxen that's very old
20:27 klausfiend i recall having an issue with clock drift years ago on IBM hardware that eventually required me to restart NTP on an hourly basis
20:27 MajObviousman I'm kinda thankful we had that box around to catch the misapplied state quickly
20:27 klausfiend like, in a day, it would be off by enough time that ntpd couldn't recover gradually
20:27 MajObviousman yeowch
20:27 klausfiend yeah gnarly
20:28 MajObviousman and, it being IBM hardware, I'm sure they said "That's within tolerances, deal with it" and billed you $8,300 for the service call
20:28 klausfiend hahaha
20:28 klausfiend so, if i wanted to make a request for enhancement
20:28 klausfiend e.g., allow all of the pkg state methods to pass in an environment dictionary
20:28 klausfiend what's the process for that?
20:29 MajObviousman simple
20:29 MajObviousman you clone the repo and go look for the file that might control the thing you want to change
20:30 MajObviousman then suggest changes by changing the code to see if it works :)
20:30 dyasny joined #salt
20:30 klausfiend oh, ok
20:30 MajObviousman there's feature requests on github too
20:30 klausfiend so clone the master branch, make a change branch, and submit a PR
20:31 MajObviousman well, if you open an issue and start the title with Feature Request that'd be a good place to start
20:31 MajObviousman I would search existing issues to see if it's already been suggested, and if so what was the project response
20:31 MajObviousman if not, I think you'd get the most traction by opening an issue with the feature request and ask for help on where you should look to make this enhancement
20:31 MajObviousman do you know python at all?
20:32 MajObviousman if you've done any non-trivial statefile work with jinja, you probably have
20:32 WKNiGHT joined #salt
20:32 tkharju joined #salt
20:33 klausfiend i do, i wouldn't consider myself an expert, though
20:34 MajObviousman thankfully, making changes or adding small functionality doesn't require you to be a python expert so much as a python decent-ert and good at pattern recognition
20:34 nickg ok i needed requests==2.2.1.  2.11.1 was too old and didnt work
20:35 nidr0x joined #salt
20:35 mgresser joined #salt
20:36 tkharju joined #salt
20:36 klausfiend MajObviousman: okay, great, thanks for the help
20:36 MajObviousman you're welcome
20:40 nawwmz getting this error when trying to run my map file, "No provider for the mapped 'centos7_v2-t1tiny-test' profile was found. Skipped VMS: blah-postgresql"
20:41 nawwmz but i am defining a profile in cloud.profiles.d with centos7_v2-t1tiny-test
20:41 nawwmz what does no provider for the mapped... mean?
20:41 tkharju joined #salt
20:42 hemebond It means you haven't specified a cloud provider in the map file. No?
20:42 hemebond Oh
20:42 hemebond Looks like you haven't specified it in the profile.
20:43 nicksloan joined #salt
20:43 gtmanfred it means that there hasn't been a provider: key specified in that cloud profile
20:43 nawwmz ahhh crap ya that was it hemebond, i had the provider named as openstack which wasnt the same name in my config within cloud.providers.d :) thx!
20:44 hemebond ????
20:44 nawwmz been looking at this for way too long
20:44 nawwmz think i need to grab a bite to eat
20:44 gtmanfred nawwmz: also, you can put them in /etc/salt/cloud.profiles.d and /etc/salt/cloud.providers.d, or if you are using the cloud execution module (and not the salt-cloud cli command) you could put them in pillar data
20:44 gtmanfred cloud:
20:44 gtmanfred providers:
20:45 gtmanfred profiles:
20:45 gtmanfred and then put the dictionaries from teh providers.d and profiles.d files in those locations
20:45 nawwmz oh nice
20:45 toastedpenguin joined #salt
20:46 gtmanfred but that cannot be used with the cloud.create or cloud.profile commands with salt-run, and it cannot be used with /usr/bin/salt-cloud cause those two dont get the pillars
20:47 nawwmz kk, im sticking the configs in cloud.*.d
20:47 nawwmz for now
20:48 dyasny joined #salt
20:53 jas02_ joined #salt
20:55 nidr0x joined #salt
21:00 toastedpenguin is there a way to reference pillar data when executing a module from the command line?
21:03 pipps joined #salt
21:04 abednarik joined #salt
21:05 gtmanfred what are you trying to do?
21:08 sarasfox joined #salt
21:08 sarasfox help what do i do with this error winexe -U \'Administrator%XXX-REDACTED-XXX\
21:09 gtmanfred that isn't an error, that is the command that salt runs to try and login to the windows server when bootstrapping, XXX-REDACTED-XXX would be where the password would be, but it was removed so the Administrator password wouldn't be in the logs
21:10 sarasfox nice it is failing to login
21:11 nawwmz need support boost :) anyone come across this before using salt-cloud/openstack "The following exception was thrown by libcloud when trying to run the initial deployment: 400 Bad Request Multiple possible networks found, use a Network ID to be more specific"
21:11 fracklen joined #salt
21:12 sarasfox Error: There was a profile error: Failed to authenticate against remote windows host (smb)
21:13 zulutango joined #salt
21:13 gtmanfred nawwmz: did you specify any networks to attach to the vm in the profile?
21:14 nawwmz negative, i didnt define any nics in my profile
21:14 nawwmz in my profile though, i added securitygroupid, subnetid
21:14 perfectsine joined #salt
21:15 nawwmz gtmanfred: is there a way to see the list of available network ids i can attach to with nova or openstack cli?
21:16 sarasfox So i should not see that winexe 10 times
21:16 toastedpenguin gtmanfred:  trying to use the s3 module, at the moment just a simple s3.get, I have s3 creds defined in pillar
21:17 toastedpenguin so something like: salt 'minion-name' s3.get
21:18 nawwmz gtmanfred: actually, is it a network_interface that i need to define in the profile?
21:18 toastedpenguin since the AWS creds for accessing s3 are defined as pillar data I imagine I need to reference these to be used for the s3.get?
21:20 hemebond toastedpenguin: I don't think you need to explicitly reference them.
21:20 RandyT sarasfox: you may be running into issue where your winexe does not have required features.
21:20 RandyT https://github.com/beardedeagle/winexe-rpm
21:21 nawwmz gtmanfred: heres my profile http://pastebin.com/GvjEyDRC
21:21 RandyT sarasfox: or you need to open some firewall ports on the windows minions
21:22 gtmanfred nawwmz: nova network-list or salt-cloud -f network_list <provider>
21:22 gtmanfred toastedpenguin: the s3 module should already use specifically wahtever you have in pillars
21:23 debian112 joined #salt
21:23 nawwmz gtmanfred: ah ok got it, whats the var for the profile?
21:23 SaucyElf joined #salt
21:23 sarasfox randyt do you think i was using window 2016 might be a issue
21:24 CEH joined #salt
21:24 toastedpenguin gtmanfred: hmm ok, I am getting an error when I execute the s3.get: "The authorization header is malformed; a non-empty region must be provided in the credential" I added the s3.location to the pillar data and it didnt fix it
21:24 nawwmz gtmanfred: think i found it, its networks:, then you pass -net-id
21:25 nawwmz thx for the support boot! gtmanfred
21:25 gtmanfred and you see the pillar in salt \* pillar.get ? did you do a pillar_refresh after it?
21:25 gtmanfred nawwmz: yes
21:25 nawwmz boot/boost :)
21:25 toastedpenguin yes
21:25 RandyT sarasfox: I've not tried to deploy a 2016 minion yet but I suspect it is more about the winexe capability
21:25 Rebus_ joined #salt
21:25 gtmanfred toastedpenguin:that I don't know then, because if you have the s3.key and s3.keyid in the pillar data, then it should work
21:27 Bryson joined #salt
21:27 hemebond Hmm. I think I've had that issue before.
21:28 SaucyElf Good afternoon, if someone has some spare cycles I've run into an issue with salt-cloud,map files and Jinja templating that I've been beating my head against.
21:28 gtmanfred i am working on an s3 customer issue, i will let you know if i run across it :)
21:28 hemebond What of you provide the location on the command line?
21:28 toastedpenguin so it should do a recursive lookup through all pillar data?
21:28 gtmanfred toastedpenguin: no, it just does a salt['config.get']('s3.key') and then uses the data from that, which looks through a bunch of places for the config, but one of those places is the pillar data
21:28 hemebond SaucyElf: Using Jinja for the map files?
21:28 sarasfox you would think salt give erre if winexe was not installed
21:29 SaucyElf @hemebond Yes. Expressions work fine (ie random 1:3 sort of things) however I can not get any variable substitution to work
21:29 tercenya joined #salt
21:30 hemebond SaucyElf: I wasn't even aware you could use Jinja in map files.
21:30 * gtmanfred had no idea that jinja was used to render map files... the more you know
21:30 toastedpenguin gtmanfred: so I understand,  it looks for the s3.key; how is it going to know which pillar to refer to or is it based on the pillar key s3.key?
21:31 gtmanfred toastedpenguin: s3.key is the pillar
21:31 sarasfox wtf it work it you install winexe
21:31 SaucyElf So what I was trying to do was have a pillar with my site specific configuration (ips, hosts etc...) and have the map file render with that data, rather than store it in the map file and the pillar
21:31 gtmanfred toastedpenguin: the master just delivers a giant dictionary to the minion as pillar data... then one of the top level keys in that dictionary needs to be s3.key and s3.keyid
21:31 geomacy joined #salt
21:32 abednarik joined #salt
21:32 hemebond SaucyElf: That doesn't sound like it will work. Pillars are applied to minions.
21:32 gtmanfred SaucyElf: how are you getting pillar data into the map file
21:32 hemebond *Pillars are compiled per minion
21:32 SaucyElf @gtmanfred That's the issue I am running into.
21:32 gtmanfred yeah, you can't do that
21:32 hemebond It's not possible.
21:33 hemebond I'm still shocked that Jinja works at all. I've never seen any mention of it in the docs.
21:33 gtmanfred the {{ pillar }} and {{salt}} objects just aren't available in the map file
21:33 SaucyElf So how about from a state file then? I tried includes in the map file also to no luck
21:33 toastedpenguin gtmanfred: ok so it should see it, got it
21:33 gtmanfred you could do it in a state file for cloud.profile or cloud.create and not use a map file, but then you lose the parallel builds
21:34 nawwmz gtmanfred: for defining the networks, does this get plopped into the provider config or in the profile config?
21:34 gtmanfred SaucyElf: you could also use a file.managed to build the map file, and then reference it in a cloud.create with the map file
21:34 gtmanfred nawwmz: eitehr
21:34 lordcirth joined #salt
21:34 gtmanfred either*
21:34 lordcirth So I see that there's a module for nfs3 exports.  Isn't there a state, too?
21:35 Elsmorian joined #salt
21:35 lordcirth I don't see one
21:35 gtmanfred lordcirth: doesn't look like it
21:35 onlyanegg joined #salt
21:36 Elsmorian joined #salt
21:36 lordcirth hmm, how hard is it to write a state wrapper for a simple module like this?
21:36 gtmanfred not very
21:36 SaucyElf so file.managed on the master, then call the salt-cloud as needed. Hmmm liking it.
21:36 whytewolf since that is what most states are not very
21:36 lordcirth Would such a pull request be accepted?
21:36 gtmanfred yes
21:36 gtmanfred we accept all kinds of stuff
21:37 Elsmorian joined #salt
21:37 gtmanfred lordcirth: make sure you include __opts__['test'] == True check
21:37 gtmanfred lordcirth: https://docs.saltstack.com/en/latest/ref/states/writing.html
21:37 Elsmorian joined #salt
21:37 gtmanfred there is all the information you need to write a state from scratch
21:38 lordcirth Awesome, thanks
21:38 SaucyElf @hemebond Not sure where I picked up that maps were rendered with Jinja *shrug* seems to work though
21:38 gtmanfred lordcirth: and here is our contributing doc https://docs.saltstack.com/en/latest/topics/development/contributing.html
21:38 Elsmorian joined #salt
21:38 lordcirth gtmanfred, since it's python, I can develop things like this while still using the Salt from the packages, right?
21:38 gtmanfred of course
21:39 gtmanfred just drop it into /srv/salt/_states/nfs.py
21:39 gtmanfred lordcirth: https://docs.saltstack.com/en/latest/ref/file_server/dynamic-modules.html
21:39 whytewolf salt makes the whole custom module thing SUPER simple
21:40 lordcirth Excellent.  Now, for the pull request, I'd put it somewhere else, I assume?
21:40 gtmanfred you put it in salt/states/nfs.py
21:40 sarasfox after installing winexe it seems to work but can't test.ping
21:40 gtmanfred I think I am going to write a blog post tomorrow about contributing to salt and all the little things to know
21:41 lordcirth gtmanfred, that sounds like a good idea
21:42 Rebus_ The documentation for cpntribution was really complex last time I checked, so this would be really appreciated
21:42 gtmanfred https://github.com/saltstack/salt/blob/develop/HACKING.rst
21:42 gtmanfred and this is what we think about when we merge pull requests https://docs.saltstack.com/en/develop/topics/development/pull_requests.html
21:42 Rumbles_ joined #salt
21:43 gtmanfred excpet don't follow the line about Posting patches to the mailing list, just open a PR on github
21:45 gtmanfred https://github.com/saltstack/salt/blob/develop/.github/CONTRIBUTING.rst
21:47 Rebus_ Some kind of howto on setting up a dev environment for contributing might also be cool - last time I tried i had one hell of a trouble getting virtualenv etc. up and running until I could start the dev branch in pararell to my already running master
21:48 lordcirth Are the links on https://docs.saltstack.com/en/latest/topics/development/contributing.html broken?  They look strange and don't seem to do anything
21:49 gtmanfred lordcirth: the coding style ones should link here
21:50 lordcirth I just found https://docs.saltstack.com/en/latest/topics/development/conventions/style.html
21:50 lordcirth But the links need fixing, I think
21:50 gtmanfred yeah, that one
21:50 gtmanfred yar, already told someone
21:50 ziggit joined #salt
21:51 gtmanfred aight, i am going to go cook dinner
21:51 gtmanfred o/
21:52 klausfiend left #salt
21:53 netcho_ joined #salt
21:59 akhter joined #salt
21:59 jas02_ joined #salt
22:00 netcho_ joined #salt
22:01 danlsgiga anyone knows how I can return a specific http code status through the salt-api rest_cherrypi implementation?
22:01 danlsgiga I want to spit out a 50x error in case my automated deploys fail
22:06 debian112 joined #salt
22:08 abednarik joined #salt
22:09 pipps99 joined #salt
22:16 nawwmz gtmanfred: :( for some reason i still get the same error use a network id to be more specific
22:16 nawwmz maybe im doing it wrong
22:16 nawwmz let me toss a pastebin
22:18 nawwmz http://pastebin.com/JNp0W8ES heres my profile, im adding networks but still get an error trying to build a host with the map file
22:18 flowstate joined #salt
22:20 nawwmz the error is "The following exception was thrown by libcloud when trying to run the initial deployment: 400 Bad Request Multiple possible networks found, use a Network ID to be more specific."
22:26 whytewolf nawwmz: are you using the openstack driver or nova?
22:27 nawwmz whytewolf: openstack driver
22:27 nawwmz compute_name: nova
22:27 Heartsbane joined #salt
22:27 Heartsbane joined #salt
22:28 nawwmz whytewolf: ooooo
22:28 nawwmz its slightly different huh? i need to pass in fixed or floating?
22:28 whytewolf exactly. openstack driver and the nova driver are way different beasts
22:28 nawwmz fml
22:29 keimlink joined #salt
22:29 whytewolf compute_name: nova is pretty normal since the keystone server type is compute and tpyically the name of that server is nova
22:29 whytewolf s/server/service
22:29 nawwmz ok cool, so I got 3 different networks all depending on environment, do I specify all three in the config?
22:29 nawwmz or should i define the network field accordingly inside my profiles.d configs?
22:30 nawwmz actually yeah, think i might separate it by environment in the profiles.d config
22:30 nawwmz thx for the assist whytewolf
22:31 whytewolf no problem.
22:31 * whytewolf really can't wait for the carbon drop when the nova drive can use auto assign floating ip's
22:34 whytewolf also can't wait for gtmanfred's openstack updates to nitrogen.
22:34 nawwmz whytewolf: ever seen an endless loop checking for misidentified IPs ?
22:34 whytewolf I have... a lot time ago. don't remeber how i resolved it
22:35 whytewolf i used to have that problem a lot with icehouse and the openstack driver while i was at IGT
22:38 nawwmz whytewolf: Waiting for VM IP. Giving up in 00:04:50 hmmmm
22:39 whytewolf ssh enabled on the security group you assign it?
22:39 whytewolf or default
22:40 nawwmz whytewolf: is there a place where i can define an internal IP or a public IP?
22:40 netcho_ joined #salt
22:40 whytewolf nope. no defining ip's is not supported
22:41 whytewolf you do need to have a free floating ip that can be assigned it doens't create one
22:42 whytewolf [if you are using neutron with floating ips]
22:42 nawwmz ah ok
22:45 akhter joined #salt
22:51 onlyanegg joined #salt
22:53 sh123124213 joined #salt
23:03 netcho_ joined #salt
23:08 APLU joined #salt
23:10 GreatSnoopy joined #salt
23:14 rpb joined #salt
23:15 APLU joined #salt
23:18 vegasq joined #salt
23:20 cliluw joined #salt
23:25 RandyT joined #salt
23:25 Qwazerty joined #salt
23:26 xet7 joined #salt
23:28 johnkeates joined #salt
23:28 Dev0n joined #salt
23:30 pipps joined #salt
23:31 subsignal joined #salt
23:33 stooj joined #salt
23:38 flowstate joined #salt
23:39 patarr joined #salt
23:40 pipps joined #salt
23:41 stooj joined #salt
23:44 lordcirth So it turns out that not knowing any python at all is a slight difficulty when writing a state
23:44 hemebond Really?
23:44 lordcirth What a surprise.  Think I'll do a bit of reading first
23:44 hemebond What do you need to know Python for?
23:44 lordcirth Sorry, as in making a new state module
23:45 hemebond Oh. Yes, definitely :-D
23:46 __number5__ yep, as hard as writing a Java program without any Java knowledge XD
23:48 akhter joined #salt
23:49 lordcirth That worked surprisingly well the last time I tried, actually
23:50 lordcirth Actually, it looks like the nfs3 module only has list_exports and del_export
23:50 lordcirth Perhaps I should first make a change to add add_export
23:50 johnkeates i just got the 'salt stack enterprise 5.0' spam
23:51 johnkeates is this the direction salt is taking? like, salt foss and salt gui?
23:51 stooj joined #salt
23:51 johnkeates i know halite got canned because it couldn't keep up, but does this mean there will never be a salt gui for non-enterprise ?
23:51 lordcirth What's the best way to test a change to an built-in module?  If I copy it to /srv/salt/_modules will it override the existing one?
23:57 __number5__ lordcirth: that's correct
23:57 lordcirth __number5__, figured, thanks
23:59 patarr joined #salt

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