Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-01-13

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

All times shown according to UTC.

Time Nick Message
00:01 abednarik joined #salt
00:04 tuxx [A[A
00:05 digitalhero joined #salt
00:07 onlyaneg1 joined #salt
00:19 flowstate joined #salt
00:19 hightekvagabond I have a script that is being placed on my minion and run from a salt state file, is there a way to make it so that script is only ever run one time via the state, or do I need to put a test in the script itself?
00:19 zmalone joined #salt
00:26 wangofett hightekvagabond: how do you know your state ran?
00:27 quasiben joined #salt
00:30 baweaver joined #salt
00:31 dendazen joined #salt
00:34 teryx5101 joined #salt
00:34 bhosmer joined #salt
00:35 otter768 joined #salt
00:39 dancat joined #salt
00:40 ecrane joined #salt
00:41 ecrane I want my salt state to download a file from a webaddress, but the webaddress will change, depending on the hostname of the salt master. Should I make a pillar variable for the URL? What salt variables can I tap for this? (Or do I need to write custom python)
00:45 abednarik joined #salt
00:48 dyasny joined #salt
00:54 shaggy_surfer joined #salt
00:56 gabino joined #salt
00:56 burp joined #salt
01:06 lompik joined #salt
01:08 izabera joined #salt
01:09 mapu joined #salt
01:18 flowstate joined #salt
01:19 dancat joined #salt
01:21 shpoont joined #salt
01:22 wangofett ecrane: I'd stick it in a pillar... though... if you're serving from the salt master, why not just add a symlink and use `salt://`?
01:29 CheKoLyN joined #salt
01:29 mosen joined #salt
01:36 amcorreia joined #salt
01:37 izabera left #salt
01:42 ecrane wangofett: I am not serving from the salt master, maybe I should... thanks.  I gtg go now, I'll chat about it more tomorrow. Thanks for the suggestion.
01:45 XenophonF joined #salt
01:45 johnkeates joined #salt
01:55 hasues joined #salt
01:55 hasues left #salt
01:56 hasues joined #salt
01:56 hasues left #salt
01:58 baweaver joined #salt
01:59 flowstate joined #salt
02:03 rem5 joined #salt
02:07 malinoff joined #salt
02:08 aron_kexp_ joined #salt
02:09 Grokzen joined #salt
02:10 larsfronius joined #salt
02:14 colegatron joined #salt
02:16 zmalone joined #salt
02:22 flowstate joined #salt
02:29 AdamSewell joined #salt
02:37 otter768 joined #salt
02:47 ilbot3 joined #salt
02:47 Topic for #salt is now Welcome to #salt! | Latest Version: 2015.8.3 | Paid support available for open source Salt! www.saltstack.com/support | Logs: http://irclog.perlgeek.de/salt/ | Paste: https://gist.github.com/ (please don't multiline paste into channel) | Ask with patience as we are volunteers and may not have immediate answers
02:53 catpigger joined #salt
02:59 penguinpowernz joined #salt
03:12 aron_kexp joined #salt
03:17 otter768 joined #salt
03:18 flowstate joined #salt
03:31 zmalone joined #salt
03:34 learner joined #salt
03:34 learner Hello All
03:34 antpa joined #salt
03:35 learner I am new to salt, and would like to understand the possibility of running an already developed c# exe file using salt automation
03:36 racooper joined #salt
03:38 _JZ_ joined #salt
03:39 favadi joined #salt
03:41 zmalone learner: https://docs.saltstack.com/en/latest/ref/states/all/salt.states.cmd.html
03:44 learner Thanks. Need your view on implementing Salt framework for my requirements.
03:45 learner I need to develop a framework where any number of exe files can be configured and should be triggered by picking the parameters from database
03:46 jhujhiti is there any way i can store custom grains, states, modules, etc in gitfs?
03:46 jhujhiti err, -states, that's obvious :D
03:46 zmalone learner: I have not used Saltstack on windows, but that doesn't sound too difficult, and pulling data from a database for your run would lend itself to Salt
03:55 dyasny joined #salt
04:00 quasiben joined #salt
04:03 vlcn left #salt
04:05 learner Thanks much.
04:05 writtenoff joined #salt
04:05 dancat joined #salt
04:06 neogenix joined #salt
04:07 ramteid joined #salt
04:17 cyborg-one joined #salt
04:18 flowstate joined #salt
04:24 lompik joined #salt
04:36 antpa joined #salt
04:36 bhosmer_ joined #salt
04:37 zmalone joined #salt
04:38 malinoff joined #salt
04:39 krishnak joined #salt
05:06 malinoff joined #salt
05:08 rdas joined #salt
05:10 malinoff_ joined #salt
05:11 brianfeister joined #salt
05:13 n8n joined #salt
05:18 flowstate joined #salt
05:19 dendazen joined #salt
05:19 iggy jhujhiti: _grains/_states/_modules/etc
05:22 tristianc joined #salt
05:30 antpa joined #salt
05:36 Micromus joined #salt
05:37 tristianc joined #salt
05:37 anmol joined #salt
05:50 evle joined #salt
05:55 Bryson joined #salt
05:58 impi joined #salt
06:02 dendazen_ joined #salt
06:02 dendazen_ left #salt
06:04 favadi joined #salt
06:10 andrew_v joined #salt
06:15 tzero joined #salt
06:18 flowstate joined #salt
06:23 krishnak joined #salt
06:24 calvinh joined #salt
06:25 linjan joined #salt
06:34 GnuLxUsr joined #salt
06:37 bhosmer_ joined #salt
06:53 Grokzen joined #salt
07:00 brianfeister joined #salt
07:10 TyrfingMjolnir joined #salt
07:12 burp joined #salt
07:19 flowstate joined #salt
07:21 antpa joined #salt
07:23 brianfeister joined #salt
07:23 rdas joined #salt
07:25 av_ joined #salt
07:27 duncanmv joined #salt
07:32 dgutu joined #salt
07:35 impi joined #salt
07:36 AlberTUX joined #salt
07:42 antpa joined #salt
07:44 dgutu joined #salt
07:50 auzty joined #salt
08:00 otter768 joined #salt
08:01 asco1aro joined #salt
08:02 otter768 joined #salt
08:03 linjan joined #salt
08:04 otter768_ joined #salt
08:04 opdude joined #salt
08:06 otter768 joined #salt
08:07 antpa joined #salt
08:08 otter768_ joined #salt
08:10 otter768 joined #salt
08:12 otter768_ joined #salt
08:13 otter768_ joined #salt
08:15 otter768 joined #salt
08:16 elsmo joined #salt
08:17 otter768_ joined #salt
08:17 fredvd joined #salt
08:18 flowstate joined #salt
08:19 colegatron joined #salt
08:19 otter768 joined #salt
08:21 otter768_ joined #salt
08:23 otter768 joined #salt
08:25 otter768 joined #salt
08:27 otter768_ joined #salt
08:28 cybacolt if you are running a salt-master, and have a minion which is also a syndic and master, can you register the minion it with its own salt-master?
08:29 otter76__ joined #salt
08:29 cybacolt i've got an issue where some cluster nodes can get mine data about the syndic node, as its registered with the master master
08:30 cybacolt short of rsync'ing its mine data back to the syndic, i cant think of any other way to solve this neatly
08:31 otter768 joined #salt
08:32 larsfronius joined #salt
08:33 otter768 joined #salt
08:34 krishnak joined #salt
08:34 otter768_ joined #salt
08:36 larsfron_ joined #salt
08:36 otter76__ joined #salt
08:38 otter768 joined #salt
08:40 dariusjs joined #salt
08:40 otter768_ joined #salt
08:42 otter768 joined #salt
08:42 kawa2014 joined #salt
08:44 Grokzen joined #salt
08:44 otter768_ joined #salt
08:46 otter768 joined #salt
08:48 otter768_ joined #salt
08:49 Rumbles joined #salt
08:50 otter768 joined #salt
08:51 babilen cybacolt: They can't share mine data IIRC
08:52 otter768_ joined #salt
08:53 cyborglone joined #salt
08:54 geekatcmu mine data is broken if you have syndics or multi-master.
08:54 otter768 joined #salt
08:54 geekatcmu Or at least, there isn't "one mine".  It's one mine per syndic/master.
08:55 geekatcmu And the canonical way to work around that is either external pillars or something like consul
08:55 cybacolt joined #salt
08:55 cybacolt yeah, i'm aware of the salt-mine multi-master problem
08:55 Nazzy joined #salt
08:55 otter76__ joined #salt
08:56 geekatcmu as far as mines are concerned, syndics are masters
08:56 cybacolt i've (mostly) solved this with a non-destructive rsync from syndic to master master
08:56 geekatcmu We wrote runners that talk to our asset DB
08:56 geekatcmu (effectively an external pillar, but ... not usable via pillar()
08:56 cybacolt the last issue i'm seeing now is syndic mine.data is not available to its minions
08:57 otter768 joined #salt
08:58 cybacolt its a shame
08:58 cybacolt i'm so close to being having fully dynamic pillars
08:59 keimlink joined #salt
08:59 cybacolt i guess it'll be addressed in future versions tho. i've seen related tickets about it on github
08:59 otter768_ joined #salt
09:01 cybacolt geekatcmu: is that approach relatively efficient?
09:03 otter768 joined #salt
09:04 s_kunk joined #salt
09:04 s_kunk joined #salt
09:05 otter768_ joined #salt
09:07 otter768 joined #salt
09:08 denys joined #salt
09:09 otter768 joined #salt
09:11 otter768_ joined #salt
09:13 otter768 joined #salt
09:15 otter768_ joined #salt
09:17 otter768 joined #salt
09:17 KermitTheFragger joined #salt
09:17 fas3r joined #salt
09:17 fas3r Hello
09:18 simooo joined #salt
09:18 fas3r if I use salt with GIT, does it mean that each master can manage his own minions indepedantly of the others ?
09:18 flowstate joined #salt
09:19 otter76__ joined #salt
09:19 simooo is it a known bug that salt sometimes truncates the output when salt.apply is used? I'm facing the problem that I cannot see the final summary of the state.apply command. it's truncated in almost the same place and the final successful/failed section is not presented at all
09:20 otter768 joined #salt
09:22 babilen simooo: Does the same happen if you use state.highstate/state.sls/state.sls_id respectively?
09:22 otter768 joined #salt
09:24 babilen fas3r: Could you rephrase that question? How does using git affect a setup in which you use multiple, distinct, masters?
09:24 otter768_ joined #salt
09:26 fas3r_ joined #salt
09:26 slav0nic_ joined #salt
09:26 otter768 joined #salt
09:27 s_kunk joined #salt
09:28 otter768 joined #salt
09:30 otter768_ joined #salt
09:31 aurynn joined #salt
09:32 fas3r_ sorr to ask again but Ive been disconected. If I configure salt to use GIT. Does it mean that each master can manage his own set of scrpipts/pillar independantly of the other master ?
09:32 otter768 joined #salt
09:33 simoo joined #salt
09:34 otter768_ joined #salt
09:36 otter768 joined #salt
09:36 babilen fas3r_: You can configure each master separately, yes. You can also use different sets of git repositories for states and pillars on them. Not sure if that answers your question though. Separate masters are, by definition, separate and that fact has little to do with git.
09:38 otter768_ joined #salt
09:38 krishnak joined #salt
09:38 GreatSnoopy joined #salt
09:38 losh joined #salt
09:39 simoo babilen: I've not observed it with a single SLS, but I in my case the output there is significantly smaller. I thought there might be some kind of a log output module that truncates the log if it's too long
09:39 simoo babilen: surprisingly with state.highstate it works fine
09:39 otter76__ joined #salt
09:40 simoo so only state.apply seems to truncate the output
09:40 antpa joined #salt
09:41 otter768 joined #salt
09:42 babilen simoo: That could very well be a bug in state.apply. I don't really use it frequently so I haven't noticed this behaviour so far. (too set in the 'old' ways)
09:43 babilen You might want to check the bug tracker on https://github.com/saltstack/salt/issues
09:43 colttt joined #salt
09:44 otter768_ joined #salt
09:44 chiui joined #salt
09:44 fas3r_ babilen: OK, so basically only the key of the minions need to be shared between the master ? If those are using different git repo for pillar/states  ?
09:44 simoo babilen: it's not the case always. I mean it only happens when the output from the state is really long. In all other cases it works fine, as expected. i'll create a ticket then
09:45 babilen fas3r_: Share keys? Do you want to use multiple masters for the *same* minion?
09:45 fas3r_ yes but for different purpose.
09:45 babilen That isn't possible
09:45 otter768 joined #salt
09:46 babilen A minion can, as of now, only talk to a single master
09:46 antpa joined #salt
09:46 fas3r_ what about master parameter in he minion cof file ?
09:46 babilen That is for failover
09:46 fas3r_ or standard
09:47 otter768 joined #salt
09:47 babilen The idea behind that is to have multiple master in case one of them becomes unavailable. That setup is being described in https://docs.saltstack.com/en/latest/topics/tutorials/multimaster_pki.html
09:48 fas3r_ babilen: I talk about https://docs.saltstack.com/en/latest/topics/tutorials/multimaster.html
09:49 eliasp simoo: related?: https://github.com/saltstack/salt/issues/11516
09:49 saltstackbot [#11516]title: IOError in display_output function when using cmd.run | Hi all,...
09:49 babilen https://docs.saltstack.com/en/latest/topics/tutorials/multimaster.html discusses a setup in which all masters are "running hot", but the idea there is that those masters are simply redundant copies
09:49 otter768_ joined #salt
09:50 fas3r_ ok but if I use this and use gitfs, it's not possible for te master to user different pillar/states on the same minions ?
09:50 fas3r_ the masters*
09:50 babilen fas3r_: You mean to explicitly violate https://docs.saltstack.com/en/latest/topics/tutorials/multimaster.html#file-roots ?
09:51 fas3r_ babilen: more like this : https://docs.saltstack.com/en/latest/ref/pillar/all/salt.pillar.git_pillar.html#git-pillar-2015-8-0-and-later
09:51 otter76__ joined #salt
09:52 babilen Well, that's the way in which you plan to violate it. The advice is still to keep those synchronised (pillars and states).
09:52 babilen You could try it, but what do you want to happen if the masters contain conflicting configuration?
09:52 simoo eliasp: not really because there is no exception in my case. it just silently truncates the end of the output even though there are some Failed states
09:52 fas3r_ babilen: what do you mean by conflicting configuration ?
09:53 otter768 joined #salt
09:53 fas3r_ each branch would be independant and general cfiguration should not change.
09:53 babilen fas3r_: One configuration says "delete /etc/foo/bar.conf" and the other says "manage /etc/foo/bar.conf" for example
09:54 |Trend| joined #salt
09:54 fas3r_ babilen: if they are in different branch does it matter
09:54 fas3r_ ?
09:54 fas3r_ because at the end this would be configured in the pillar/state
09:55 otter768_ joined #salt
09:56 babilen fas3r_: Well, a configuration like that would never be able to converge as the work done by one master will be undone by the other. That is basically a question of strict separation of concerns though
09:56 babilen What's your usecase for multiple masters btw? (just curious)
09:57 otter76__ joined #salt
09:58 fas3r_ I would like to have dedicated master for each team/project that we manage. The idea would be they could manage their server by themself from salt without intefiring or impacting the other teams which would have also their own master.
09:59 otter768 joined #salt
09:59 babilen fas3r_: Okay, "their server" refers to a single minion or will multiple people work on the same minion from different masters?
09:59 fas3r_ multiple people work on the same minion from different masters?
09:59 fas3r_ sorry without the ?
09:59 babilen ah!
09:59 babilen (that makes more sense!)
10:00 fas3r_ that's why I thouht that gitfs was the best approach...
10:00 babilen fas3r_: How will you ensure that the states don't conflict?
10:01 fas3r_ using different git based on the master
10:01 otter768_ joined #salt
10:01 fas3r_ depends on*
10:01 babilen I mean, you could still just connect the minion to a single master and just include multiple separate GitFS repositories. A highstate run would then be able to use states from all of them.
10:02 fas3r_ babilen: we would like to avoid to provide them access to the 'main' master if I ca say it like his.
10:02 fas3r_ this*
10:02 babilen fas3r_: Yeak, but if team A says "absolutely make sure that /etc/foo/hosts" is managed by us and *only* contains "bar" and another team has a similar state it would flip back and forth forever.
10:02 fas3r_ I see what you maen.
10:02 otter768_ joined #salt
10:03 fas3r_ thats sad hahah :D
10:03 babilen I mean that is something to keep in mind. I understand where you are coming from and, in the end, you cannot ensure that you don't have such conflicting states when using a single master either, but the likelihood simply becomes higher.
10:04 babilen What I would do is to give the teams access to a "generic" master that configures everything that is not absolutely specific to one team. You can control who does what with https://docs.saltstack.com/en/latest/topics/eauth/access_control.html
10:05 otter768 joined #salt
10:06 babilen You would then allow different teams/users/... to run different states and give them separate repositories. If you really want to you can still use multiple masters with different configuration, but I'm not sure what that buys you.
10:06 otter768_ joined #salt
10:06 babilen People don't even have to login to the masters as you could just fire off a highstate every hour (or so)
10:06 malinoff fas3r_: or you can install jenkins and use its matrix-based security to launch salt cli, leaving cli itself unsecured
10:08 otter768 joined #salt
10:10 otter76__ joined #salt
10:10 CeBe joined #salt
10:11 anmol joined #salt
10:12 babilen fas3r_: I'd rather think about appropriate git workflows and repository layouts. Salt should, at the end of the day, simply "do the thing" and you might rather want to concern yourself with defining workflows that allow for control as to which areas of the configuration are touched by which team
10:12 otter768 joined #salt
10:12 babilen That might be as easy as "You guys only touch pillars in foo/*"
10:14 otter768_ joined #salt
10:16 otter76__ joined #salt
10:17 GnuLxUsr joined #salt
10:18 otter768 joined #salt
10:18 flowstate joined #salt
10:20 otter768 joined #salt
10:22 otter768_ joined #salt
10:23 tristianc joined #salt
10:23 otter768 joined #salt
10:25 otter768 joined #salt
10:27 otter768_ joined #salt
10:29 otter768 joined #salt
10:30 chiui joined #salt
10:37 traph joined #salt
10:38 bhosmer joined #salt
10:40 amcorreia joined #salt
10:41 giantlock joined #salt
10:41 Kage` joined #salt
10:42 Kage` I feel like I'm missing it somewhere, but is there a Salt method for executing HTTP requests?  Or would we just need to make like, a Runner and do it manually?
10:43 babilen https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.http.html
10:44 smARMie joined #salt
10:46 smARMie hi. i know this has been previously discussed, but i'm struggling with including pillar files only when that file exists. i found a solution on the web, using an if['salt.file_exists'](<pillar_file>), but this doesn't seem to work with gitfs. does anyone have any pointers for this?
10:46 permalac joined #salt
10:46 Kage` Ah, perfect, thanks babilen
11:01 zerthimon joined #salt
11:01 colttt joined #salt
11:02 jamesp9 joined #salt
11:07 ericof joined #salt
11:08 k00mi joined #salt
11:10 invalidexception joined #salt
11:10 AirOnSkin joined #salt
11:11 aurynn joined #salt
11:12 tehsu joined #salt
11:15 digitalhero joined #salt
11:17 oida joined #salt
11:19 flowstate joined #salt
11:26 GnuLxUsr joined #salt
11:31 LondonAppDev joined #salt
11:37 antpa joined #salt
11:43 Grokzen joined #salt
11:43 favadi joined #salt
11:43 Rumbles joined #salt
11:45 dariusjs joined #salt
11:48 asco1aro joined #salt
11:51 antpa_ joined #salt
11:53 antpa joined #salt
11:58 asco1aro joined #salt
12:12 impi joined #salt
12:16 denys joined #salt
12:24 harkx joined #salt
12:24 SVQTQ joined #salt
12:30 otter768 joined #salt
12:35 SVQTQ Hey guys. Is it possible to use the v8 search method in controllerS?
12:36 SVQTQ ok, using env instead of registry works. Thanks documentation :D
12:37 quasiben joined #salt
12:37 SVQTQ :< Wrong channel.
12:40 bhosmer joined #salt
12:42 flowstate joined #salt
12:50 akhter joined #salt
12:57 flowstate joined #salt
13:05 mapu joined #salt
13:07 N-Mi joined #salt
13:13 londo joined #salt
13:17 antpa joined #salt
13:18 chesty joined #salt
13:28 teryx510 joined #salt
13:36 favadi joined #salt
13:37 ericof joined #salt
13:37 flowstate joined #salt
13:40 dendazen joined #salt
13:46 wangofett joined #salt
13:50 dyasny joined #salt
13:52 ericof joined #salt
13:56 AdamSewell joined #salt
13:57 gh34 joined #salt
13:58 lompik joined #salt
13:58 krishnak joined #salt
14:02 evle1 joined #salt
14:02 andrew_v joined #salt
14:10 giantlock joined #salt
14:13 toastedpenguin joined #salt
14:14 tristianc_ joined #salt
14:14 banderas07 joined #salt
14:15 geekatcmu cybacolt: Not terribly effecient.  One of the down sides of runners is that they get called on every include, so if, for example, your hadoop/map.jinja fills in any data items via runner, on a single minion you'll see a roundtrip to the master for hadoop/init.sls, hadoop/namenode.sls, hadoop/resourcemanager.sls, and hadoop/zookeeper.sls.
14:15 banderas07 hi, can a minion connect to two distinct salt-masters in the same time ?
14:18 toastedpenguin joined #salt
14:19 antpa joined #salt
14:19 quasiben joined #salt
14:21 colegatron joined #salt
14:22 hasues joined #salt
14:25 tehsu joined #salt
14:27 racooper joined #salt
14:27 LondonAppDev joined #salt
14:29 MikeyYeahYeah joined #salt
14:31 otter768 joined #salt
14:31 morissette joined #salt
14:32 AndreasLutro am I understanding it right that /etc/salt/pki/master/master.{pub,pem} are regenerated every 24 hours?
14:33 cpowell joined #salt
14:34 antpa joined #salt
14:34 scoates joined #salt
14:35 babilen AndreasLutro: yeah
14:35 babilen (hence the ping_on_rotate option)
14:36 jhujhiti iggy: re custom grains/etc in gitfs, so i can just create a _grains directory in the root of the branch and it will do the right thing? i didn't see documentation for it anywhere
14:36 babilen jhujhiti: Is "do the right thing" means: "make it available for syncing" then, yes
14:38 babilen jhujhiti: _grains (and _modules, _states, ...) work more or less the same and _grains is documented/mentioned on https://docs.saltstack.com/en/latest/topics/targeting/grains.html#writing-grains for example
14:39 babilen (excuse me for answering if you wanted to hear that from iggy)
14:40 quasiben joined #salt
14:40 bhosmer joined #salt
14:42 winsalt joined #salt
14:42 BogdanR joined #salt
14:43 BogdanR Is it possible to specify the cloud provider when running a salt map?
14:44 BogdanR Currently I have several providers configured and salt-cloud interrogates all of them when it's attempting to start the VMs
14:47 Tanta joined #salt
14:48 AndreasLutro babilen: when a new key gets generated, if ping_on_rotate is false, how do the salt minions know the new key is valid?
14:49 babilen They don't ... they try, fail and ask again
14:50 AndreasLutro that seems like a potential security volnurability...
14:51 AndreasLutro if an attacker could figure out the rotation time, they could spoof the master
14:52 rim-k joined #salt
14:52 rim-k hello everyone
14:52 Priyanka joined #salt
14:53 rim-k does anyone know how to make salt renders gpg pillar in a file using contents.pillar?
14:53 babilen AndreasLutro: No, it fails and they try to re-authenticate
14:54 Priyanka Hi.. I want to use Salt for database testing on windows/linux... Is this possible?>
14:54 Priyanka to connect to database server, extract the data out of that, then         do database testing mainly
14:56 Priyanka And run exe file also...
14:56 quasiben joined #salt
14:57 amcorreia joined #salt
14:57 DammitJim joined #salt
14:57 Priyanka Has nobody got any answer on that ..?
14:59 zmalone joined #salt
14:59 rim-k does anyone know how to make salt renders gpg pillar in a file using contents.pillar?
15:00 DammitJim is there anything that does laptop configuration management?
15:00 DammitJim knowing a laptop might be offline at the time that the request is fired?
15:01 flowstate joined #salt
15:02 federicob joined #salt
15:02 tpaul joined #salt
15:03 federicob Hi guys, anybody has experience pulling docker images from private docker registries?
15:05 quasiben joined #salt
15:07 brianfeister joined #salt
15:10 CheKoLyN joined #salt
15:12 vieira_ joined #salt
15:13 vieira_ Hello, when using {% set var = salt['grains.get']('something') -%} and #!yaml|gpg I get an error
15:13 spiette joined #salt
15:13 vieira_ wach of these alone work
15:14 vieira_ *each
15:14 vieira_ but not together. is this a known issue? am I doing something wrong?
15:14 rem5 joined #salt
15:18 krishnak joined #salt
15:22 kaptk2 joined #salt
15:23 rm_jorge joined #salt
15:24 digitalhero joined #salt
15:26 bhosmer joined #salt
15:29 vieira_ found character '%' that cannot start any token; line 3
15:29 vieira_ :\
15:37 perfectsine joined #salt
15:40 dgutu joined #salt
15:41 thetoolsmith joined #salt
15:42 dijit anyone got an example of an idempotent script with salt? I dont' know how to formulate the json output if nothing changes.
15:43 winsalt dijit, are you talking about something with cmd.script? or a custom salt module?
15:45 dijit cmd.script
15:45 dijit or actually, cmd.run: stateful: yes
15:45 Brew joined #salt
15:45 chadhs joined #salt
15:47 winsalt I use that too, and I have the scripts return nothing if there are no changes
15:48 dijit oh, maybe I should do that.
15:48 dijit because the 'changed=no' line now returns as if it failed.
15:48 flowstate joined #salt
15:48 dijit and changed=yes comment="something changed' has been failing for a while.
15:48 dijit :(
15:48 rim-k does anyone know how to make salt renders gpg pillar in a file using contents.pillar?
15:48 dijit I dont' mind changing things, but I'd like to know what to change to.
15:49 dijit rim-k: yeah
15:49 dijit I have an example somewhere.
15:49 winsalt yeah, they are tiny scripts, basically just an if statement to check old vs new, and if they are the same return nothing at all
15:49 rim-k dijit: thx :D
15:49 dijit https://github.com/darkscience/ds-salt/blob/master/states/inspircd/init.sls
15:49 dijit rim-k: ^
15:49 dijit you can see me doing it for my certificates.
15:50 spuder joined #salt
15:50 impi joined #salt
15:51 rim-k [ERROR   ] Could not decrypt cipher
15:51 rim-k :/
15:51 dijit sounds like you don't have your gpghomedir set up properly.
15:51 dijit at least it's trying to render.
15:51 dijit do you have a password on your key?
15:51 rim-k no
15:52 dijit ok, and where is the gpg home directory?
15:52 rim-k the salt-master is the one that will do the decript right?
15:52 dijit /etc/salt/?
15:52 rim-k yes
15:52 dijit yeah
15:52 dijit strange.
15:52 dijit what's the name of the directory?
15:52 dijit mine is /etc/salt/gpgkeys
15:53 fas3r joined #salt
15:53 fas3r Hello
15:53 rim-k same
15:53 vieira_ do you have python-gpg
15:53 vieira_ ?
15:54 vieira_ the public key is trusted?
15:54 rim-k you mean python-gnupg
15:54 rim-k ?
15:54 fas3r I deleted an salt key and restart the deamon minions on my target host. Unfortunately my master does not list /receive the key. Do I need to do someting particular
15:54 vieira_ rim-k: yup
15:55 v0rtex does anyone know of a simple/clever way to know if a Ubuntu system is using systemd or upstart within state files? Right now I have a hacky bit of jinja that is just looking at the oscodename grain to see if it's a 'vivid' system but I'd like something more flexible.
15:55 rim-k vieira_: yes
15:56 vieira_ rim-k: when you do echo -n "secret" | gpg --armor --encrypt -r your_key
15:56 vieira_ from the cmd line
15:56 vieira_ does it work? is the key trusted ?
16:00 rim-k yes
16:00 Kage` left #salt
16:01 vieira_ no warnings?
16:02 vieira_ I did that yesterday and it is working :\
16:02 rory joined #salt
16:02 vieira_ 2015.8.3
16:03 vieira_ 2015.5.3 I mean
16:05 rim-k 2015.8.3
16:08 rim-k mmm
16:08 babilen v0rtex: You could write a custom grain that checks for /run/systemd/system and returns True if it is being set
16:08 rim-k is there something else to do with the master configuration file ?
16:08 vieira_ If you want to compare any settings, let me know
16:08 babilen s/being set/present/
16:08 rim-k in my pillar I have #!yaml|gpg
16:08 vieira_ nope, I changed nothing there
16:08 vieira_ yes, that is needed
16:09 rim-k then I'd like to have more verbosity
16:09 rim-k so I can troubleshoot and not bothering you with this :/
16:10 fas3r how to generate salt minion key please ?
16:10 babilen v0rtex: Take a look at /var/lib/dpkg/info/libc6:amd64.postinst -- it contains various "init system detection" methods. You could even set the grain to the appropriate value that way.
16:11 rim-k ohhhh
16:11 rim-k I found something
16:11 rim-k vieira_:
16:12 rim-k gpg: invalid armor header: hQEMA5pN+0gaidemAQf/UeXHOTOA0mM40fdaSaPZCZzSTPftOGqb+tt31+U/MW/l\n
16:13 rim-k gpg: decryption failed: secret key not available
16:13 babilen :(
16:16 v0rtex babilen: I'll take a look, thanks
16:16 vieira_ are you putting it with gpg header?
16:16 vieira_ in the pillar?
16:17 ekristen joined #salt
16:17 vieira_ -----BEGIN PGP MESSAGE----- ...
16:17 vieira_ ?
16:19 vieira_ rim-k: like in https://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.gpg.html#salt.renderers.gpg.render ?
16:20 _JZ_ joined #salt
16:21 edrocks joined #salt
16:22 FreeSpencer How would I go about getting grain data from other minions in a template?
16:22 bbradley can anyone see why this conditional is true when the value is None?
16:23 bbradley {% if settings.config != None %}
16:23 rim-k vieira_: thx
16:23 bbradley it is in a formula, trying to think about if it needs to be changed.
16:23 rim-k I think I know what I did wrong
16:23 rim-k vieira_: you've been very helpful
16:23 rim-k :D
16:24 vieira_ hehe glad you got it
16:24 FreeSpencer Looks like I need to use the salt mine
16:25 FreeSpencer I don't NEED to use the mine though since I dont care about caching it
16:25 bhosmer joined #salt
16:27 N-Mi joined #salt
16:29 lothiraldan joined #salt
16:32 giantlock joined #salt
16:32 otter768 joined #salt
16:32 hasues left #salt
16:36 vieira_ ahh when using jinja and gpg the bang should be #!jinja|yaml|gpg instead of #!yaml|gpg
16:37 DammitJim joined #salt
16:43 mapu joined #salt
16:44 jnilsson joined #salt
16:44 MichaelRomig joined #salt
16:45 MichaelRomig_ joined #salt
16:46 MichaelRomig joined #salt
16:51 dfinn joined #salt
16:54 thetoolsmith joined #salt
17:00 digitalhero joined #salt
17:01 Bryson joined #salt
17:01 oida joined #salt
17:02 krishnak joined #salt
17:03 impi joined #salt
17:03 ecrane joined #salt
17:04 thetoolsmith hello all. I have about 3 weeks with salt in poc mode, but seem to be stuck on some pillar rendering.
17:04 thetoolsmith I’m using git backend only for state and pillar.
17:05 thetoolsmith I get this   _errors:  - Specified SLS 'ne_engine' in environment 'base' is not available on the salt master
17:05 thetoolsmith i have verified that the sls file does exist in my pillar base as well as state base tree
17:05 danlsgiga joined #salt
17:06 danlsgiga hey guys... I'm writing a custom runner that needs to run a cmd_batch in a number of minions
17:06 danlsgiga but I also need to pass the --failhard
17:06 FreeSpencer How do I have multiple mines? Seems only the first one is taking
17:07 danlsgiga I can't find any references in the docs or on the internet if the LocalClient cmd_batch accepts the failhard arg
17:08 teryx510 joined #salt
17:08 digitalh_ joined #salt
17:14 racooper joined #salt
17:21 CheKoLyN joined #salt
17:21 wnkz joined #salt
17:21 viq thetoolsmith: I'm not sure how (stuff in the middle) likes underscores
17:25 winsalt thetoolsmith, is the sls matched to a minion in your top file?
17:26 larsfronius joined #salt
17:27 thetoolsmith viq: no. the sls files match '*'
17:27 thetoolsmith fwiw, i cloned the git pillar repo to the local file system on the master, enabled pillar_roots in master config and commented all the gitfs_xx stuff, and it renders fine.
17:28 thetoolsmith viq: so the issues when pillar is external git repo
17:29 viq thetoolsmith: I use pillars from git
17:30 thetoolsmith viq: I do have another pillar that I am successful when using gitfs remote, but for some reason this ne_engine pillar is not rendering
17:32 bbradley about to try to add the t2.nano instance to salt-cloud.
17:32 bbradley weeee
17:32 writtenoff joined #salt
17:34 digitalhero joined #salt
17:35 tim_s007 joined #salt
17:37 drawsmcgraw joined #salt
17:43 anotherZero joined #salt
17:43 onlyanegg joined #salt
17:47 impi joined #salt
17:49 deniszh joined #salt
17:50 stevednd joined #salt
17:50 stevednd anyone here using salt-cloud with linode? I currently seem to be unable to connect. Keep getting "connection reset by peer". This has previously worked hundreds of times.
17:52 rihannon joined #salt
17:57 DammitJim is there a product that will work with salt and laptops?
17:57 DammitJim to be able to do stuff with laptops that might be offline at the time that the call is made?
18:00 hacks Is anyone using the mysql external pillar? Trying to find some more information on how it works specifically.
18:00 drawsmcgraw joined #salt
18:03 andrew_v joined #salt
18:05 iggy I've never heard anyone mention using it
18:07 ageorgop joined #salt
18:10 OliverMT joined #salt
18:10 ramblinpeck joined #salt
18:10 munhitsu_ joined #salt
18:10 hacks joined #salt
18:10 Phtes joined #salt
18:14 serverascode joined #salt
18:15 supermike___ joined #salt
18:16 akitada joined #salt
18:17 hacks iggy: Yeah, looks like the other external pillars are a lot easier to understand :/
18:17 mattl joined #salt
18:17 akoumjian joined #salt
18:18 czchen joined #salt
18:18 iggy I think most people either end up using regular file pillars or they go and write their own to interface with whatever system they already have running
18:18 huddy joined #salt
18:18 bbhoss joined #salt
18:18 lkannan joined #salt
18:18 ashb joined #salt
18:18 antonw joined #salt
18:18 davedash joined #salt
18:18 copelco joined #salt
18:18 goki joined #salt
18:18 Striki joined #salt
18:18 gazarsgo joined #salt
18:18 ramishra_ joined #salt
18:19 Edgan DammitJim: Should work with laptops, but you would have to expose your salt master to the internet, or have the laptops on VPN
18:19 wiqd joined #salt
18:19 simonmcc joined #salt
18:19 johtso joined #salt
18:19 hightekvagabond joined #salt
18:19 hillna joined #salt
18:20 twodayslate joined #salt
18:20 nahkiss joined #salt
18:20 shaggy_surfer joined #salt
18:20 kutenai joined #salt
18:20 tcolvin joined #salt
18:20 jay_d joined #salt
18:21 mihait joined #salt
18:21 Edgan DammitJim: The potential issue I would worry about with salt and laptops that they would regularly disconnect from the salt master in the middle of salt runs.
18:21 Edgan DammitJim: and be left in broken states
18:22 tpaul joined #salt
18:22 m0nky joined #salt
18:22 CaptTofu joined #salt
18:22 mrbobbytables joined #salt
18:22 Edgan DammitJim: The should have the list of things to do without salt master contact, but the salt master is also the file server. So they will break there on disconnect.
18:23 gtaylor joined #salt
18:23 snarfy joined #salt
18:23 Edgan DammitJim: They also need constant access to whatever you use for package management repositories
18:25 trave joined #salt
18:25 tmkerr joined #salt
18:25 NachoDuck joined #salt
18:25 mikepea joined #salt
18:26 iggy I'd probably use masterless with laptops
18:26 imanc joined #salt
18:27 frankS2 joined #salt
18:27 JonGretar joined #salt
18:27 Ryan_Lane joined #salt
18:28 jirwin_ joined #salt
18:29 donmichelangelo joined #salt
18:29 michelangelo joined #salt
18:29 Score_Under joined #salt
18:31 iggy just package up your states, deploy package to laptops, have the laptops run the states via cron/etc
18:31 SteamWells joined #salt
18:31 shawnbutts joined #salt
18:31 MichaelRomig joined #salt
18:31 mansquid joined #salt
18:31 iggy you really don't want a bunch of keys hanging around on  your master for machines that are only sometimes connected
18:32 DammitJim iggy, Edgan thanks for your input
18:32 bmcorser joined #salt
18:32 DammitJim so, there is no mechanism that works together with salt where a state run is queued if the minion is not available?
18:32 DammitJim and it will keep re-trying?
18:33 iggy no
18:33 otter768 joined #salt
18:33 bhosmer joined #salt
18:34 bstaz joined #salt
18:34 bVector joined #salt
18:37 LinuxHorn joined #salt
18:37 Edgan DammitJim: Normally you run salt as a cron job on the salt master. So it will just fail when the laptop is unavailable and start working again when the laptop connects to the salt master again.
18:37 catpig joined #salt
18:39 Edgan DammitJim: But it isn't all or nothing. I don't think it transfers everything that is needed before starting the salt run to the minion. So if you kill network access between them in the middle, the run breaks. It would be a good thing to know for sure though.
18:39 DammitJim oh, I never run salt as a cron job on my salt master!
18:39 winsalt What about a reactor that listens for the auth event when the laptop starts running
18:40 Edgan DammitJim: It is the preferred way with -b for batch mode. Then you can make sure that the salt master never gets overloaded by a thundering herd of minions.
18:45 darvon joined #salt
18:46 deus_ex joined #salt
18:50 baweaver joined #salt
18:56 Rumbles joined #salt
18:57 flowstate joined #salt
18:58 denys joined #salt
18:59 DammitJim winsalt, I thought about that!
18:59 DammitJim so, then the minion will call home?
19:00 iggy it's really not a good idea to have minions that won't be always connected
19:03 cberndt joined #salt
19:03 flowstate joined #salt
19:04 winsalt the minions do periodic auth checks, so yes
19:05 winsalt but i agree, a minion should always be in contact.  it would be better to have an isolated salt setup on the laptop
19:09 viq joined #salt
19:10 Brew joined #salt
19:12 Dev0n joined #salt
19:13 iggy the problem with minions not being connected constantly is you'll end up getting timeouts and whatnot
19:13 iggy the annoyance alone would drive me batty... not to mention what goes on behind the scenes
19:14 wt joined #salt
19:15 baweaver joined #salt
19:15 wt Hi, I am having a problem creating a dummy interface on CentOS7 with salt. Salt is adding the HWADDR (mac address) field to the config without my having specified it.
19:16 bhosmer joined #salt
19:17 wt I am using "type: eth"
19:17 wt however, that type implies setting the addr context var.
19:18 Rumbles joined #salt
19:18 aron_kexp joined #salt
19:18 wt Looks like there is no way to prevent the HWADDR field in that case.
19:18 wt Any suggestions?
19:20 GreatSnoopy joined #salt
19:22 wt https://github.com/saltstack/salt/blob/develop/salt/modules/rh_ip.py#L630 <-- this is wrong for dummy interfaces
19:23 wt The block starting at that line.
19:26 Brew joined #salt
19:39 aron_kexp_ joined #salt
19:49 AdamSewell joined #salt
19:55 darvon joined #salt
19:56 flowstate joined #salt
19:57 Grokzen joined #salt
19:58 cpowell_ joined #salt
19:59 CheKoLyN joined #salt
19:59 boargod3 joined #salt
20:01 indispeq joined #salt
20:03 manfred joined #salt
20:05 stanchan joined #salt
20:12 indispeq joined #salt
20:18 colegatron joined #salt
20:18 slk_ joined #salt
20:19 giantlock joined #salt
20:19 slk_ is there a way to run a state.orchestrate via the salt api? i can't find any documentation and my googlings return nothing
20:21 beardedeagle joined #salt
20:22 baweaver joined #salt
20:22 beardedeagle Query: would be it ill advised to salt your multimasters? Ie multimasters have minion installed and can see eachother?
20:22 whytewolf slk_: orchestrate is a runner. so yes. you just need to run the orchestrate runner
20:23 indispeq joined #salt
20:23 digitalhero joined #salt
20:23 slk_ @whytewolf i've tried various things, including sending the post data client=runner, fun=state.orchestrate, but i don't know how to specify what orchestrator to run. i keep getting a 401.
20:25 slk_ including arg=$path_to_orchestrate_sls
20:26 whytewolf well 401 is unauthed. so i would work on that first.
20:27 slk_ that's what's confusing me. i'm sending a valid auth token.
20:27 Rumbles joined #salt
20:27 slk_ i can run states using the same auth token
20:28 whytewolf slk_: can you get any runners to work, such as manage.up
20:28 federicob joined #salt
20:29 indispeq joined #salt
20:31 whytewolf but the arg should be the same you pass in from the cli. so if the you normally run salt-run state.orch newserver then the json would look something like [{"client": "runner", "fun": "state.orchestrate", "mods": "newserver"}]
20:33 slk_ @whytewolf can't get the jobs runner to work, so seems like there's something else afoot. here's an example of some data i've tried to pass, which results in a 401.
20:34 slk_ data = {'client': 'runner', 'fun': 'jobs.lookup_jid', 'jid': '20160113121156320160'}
20:34 whytewolf slk_: but yeah your eauth should be setup so the user can access @runner
20:34 otter768 joined #salt
20:34 whytewolf slk_: https://docs.saltstack.com/en/latest/topics/eauth/index.html#acl-eauth
20:34 slk_ ah that may well be it
20:34 baweaver joined #salt
20:34 controversy187 joined #salt
20:34 slk_ @whytewolf thanks, i think eauth may be the culprit
20:34 whytewolf beardedeagle: I would advise that master of any kind be salted.
20:35 whytewolf slk_: no problem
20:35 beardedeagle @whytewolf: right, setting up MULTI-MASTER-PKI and figured that it would be a good method of keeping things in sync between the two.
20:37 whytewolf beardedeagle: sounds good. honestly I'm waiting for better multimaster support. right now it is just horrendous.
20:37 slk_ yeah multimaster right now is not so great :(
20:40 keimlink joined #salt
20:42 baweaver joined #salt
20:44 Bryson I need to pass an arbitrary switch and value to user.present. specifically, i need to add `-w no` to the command salt runs to create a user. how do I add this in my YAML file?
20:47 stevednd anyone here using salt-cloud with linode? I currently seem to be unable to connect. Keep getting "connection reset by peer". This has previously worked hundreds of times.
20:49 CheKoLyN joined #salt
20:49 zmalone stevednd: I recall someone else having that problem when Linode began getting ddosed
20:51 wendall911 joined #salt
20:57 forrest joined #salt
20:57 foundatron joined #salt
20:58 whytewolf Bryson: well you can drop to useing cmd.run. or find the module being used for your OS and edit it and put it in _modules. but right now I'm wondering about -w no. I find no reference to this switch in and useradd program on any system. nor in the mac and windows way of adding a user.
20:58 Bryson freebsd
20:58 Bryson https://www.freebsd.org/cgi/man.cgi?query=pw&amp;sektion=8&amp;apropos=0&amp;manpath=FreeBSD+10.2-RELEASE
20:59 whytewolf ahhh pw not useradd
21:00 Bryson yeah freebsd has useradd but it's just a wrapper for pw and I don't think its man page includes the -w arg
21:00 whytewolf Bryson: it doesn't https://www.freebsd.org/cgi/man.cgi?query=useradd&amp;apropos=0&amp;sektion=0&amp;manpath=Red+Hat+Linux%2Fi386+9&amp;format=html
21:01 stevednd zmalone: any idea what they did to fix it? I can reach the linode api through curl calls on the machine just fine, but through salt-cloud it's a no go
21:01 zmalone No idea, sorry.
21:03 bhosmer joined #salt
21:07 linjan joined #salt
21:10 whytewolf Bryson: in theory having no password set and empty_password: False has a simalar functionality.
21:10 Bryson that disables password login but won't disable login with something like an ssh key
21:10 Bryson `-w no` completely disables logging in with the account from any method
21:11 Bryson at least this is how i'm understanding the man pages so far.
21:11 Bryson and salt's state for .present doesn't pass along args or kwargs, so i can't just add arbitrary things to the yaml file and let them pass through. it just ignores them.
21:12 whytewolf Bryson: correct there is no way to pass something like that through the yaml. like i said you would need to basicly add the functionality to pw_user.py and put it in _modules.
21:12 whytewolf or use cmd.run
21:13 bbradley start t2.nano instances with salt-cloud
21:13 bbradley https://github.com/saltstack/salt/pull/30337
21:13 saltstackbot [#30337]title: salt-cloud: add t2.nano instance to Amazon EC2 | Change was very simple!
21:13 whytewolf but also I don't think it disables the account compleatly. since just a single passwd unlocks it
21:15 whytewolf but i also don't have a freebsd box to test on right now. as  i need to rebuild a new openstack setup to get one :/
21:15 zmalone Bryson/whytewolf: Those match up with special password fields in /etc/passwd
21:16 zmalone ! is the "hash" you probably want to set using the normal command
21:16 wm-bot4 You are not authorized to perform this, sorry
21:16 whytewolf ...
21:17 whytewolf !list
21:17 whytewolf aww
21:17 zmalone *lk* is another, which might match up better
21:17 zmalone but if you run the useradd -w no command on freebsd manually, you can probably figure out which "special" hash it's using
21:17 neogenix joined #salt
21:17 shaggy_surfer joined #salt
21:18 Bryson whytewolf: where is the documentation on dynamic modules? the only thing i can find in the salt docs is a reference to syncing them. also, if the state calls the module and the state doesn't accept any additional arguments, how is it going to pass that info when it calls the module?
21:20 whytewolf Bryson: good point about the state. you might need to update user.py state and put that in _states
21:20 seblu joined #salt
21:20 whytewolf anyway dynamic modules https://docs.saltstack.com/en/latest/ref/modules/
21:21 racooper has anyone done a state module for authconfig on CentOS/RHEL?
21:21 whytewolf I would start witht he existing code. things in _modules do override things built in
21:22 whytewolf racooper: doubt it
21:23 racooper yeah that fits with my lack of useful google results.
21:26 blckbit10 joined #salt
21:29 slav0nic_ joined #salt
21:30 stanchan joined #salt
21:30 PeterO joined #salt
21:31 snarfy joined #salt
21:33 aboe joined #salt
21:35 Karunamon Hi folks - I'm having a heck of a time getting salt-cloud to talk to a vcenter server. It's complaining about certificate verification failure, even though i've set the provider as verify_ssl: False
21:37 baweaver joined #salt
21:43 onlyaneg1 joined #salt
21:44 tristianc joined #salt
21:48 Rumbles joined #salt
21:57 rotbeard joined #salt
22:01 snarfy joined #salt
22:03 MindDrive joined #salt
22:08 larsfronius joined #salt
22:11 Aleks3Y joined #salt
22:18 M-MadsRC Does anyone know when SaltStack plan to make a new release? There's a ton of features I'd love that's in develop right now :P and I don't feel like patching my salt install all over my minions
22:20 jfindlay next major release probably happening in March
22:21 whytewolf should act like google and push that off to saltconf and announce it then in a big fan fare :P
22:22 M-MadsRC jfindlay (IRC): Ah okay, so if there's a feature or fix in develop right now, I can count on it being part of the next release?
22:24 jfindlay whytewolf: after the frenetic mania last year we decided not to do a release during saltconf, e're going to take things one at a time this time :-)
22:25 whytewolf hehe, good idea :)
22:25 jfindlay M-MadsRC: yes, everything currently in develop is going into Boron until Boron is forked
22:26 M-MadsRC Awesome! You can't imagine how thrilled I am to see some of my own code in a release :P The updated beacons are gonna be awesome! Can finally do proper service monitoring now
22:27 digitalhero joined #salt
22:27 jfindlay nice
22:29 RandyT are beacons working for windows platform?
22:30 M-MadsRC Atleast some of them are cross-platform in 2015.8.3
22:30 RandyT M-MadsRC: awesome
22:31 N-Mi joined #salt
22:31 N-Mi joined #salt
22:34 mosen joined #salt
22:35 otter768 joined #salt
22:36 ajw0100 joined #salt
22:38 afics joined #salt
22:42 brianfeister joined #salt
22:43 MikeyYeahYeah joined #salt
22:44 zenlot1 joined #salt
22:44 afics joined #salt
22:48 digitalhero joined #salt
22:49 digitalhero joined #salt
22:50 gimpy2938 joined #salt
22:50 gimpy2938 whe using startup states is there a way to specify a returner to be used?
22:51 gtmanfred i do not believe so for that specific thing
22:51 gtmanfred but you could specify a default returner
22:51 gtmanfred iirc*
22:52 digitalhero joined #salt
22:52 gimpy2938 dang...I have minions that call state.highstate when they start but I haven't found a good way to get Salt to tell me A: was it successful? and B: If not, what went wrong?
22:53 gtmanfred you could key off the reactor, instead of using startup_states
22:53 gimpy2938 would be really nice to just have Salt shove the highstate result into Mongo :)
22:53 gtmanfred and kick off a highstate the first time the connect to the master, and then have the reactor event pass a returner to use ( i think that thing can pass it)
22:53 gimpy2938 reactor?
22:54 gtmanfred i set the default returner to mongo, and put everything in mongo
22:54 gtmanfred then it returns to the commandline and mongo
22:54 RandyT I'm in the process of starting down the path of implementing returners, beacons and log aggregation. Was wondering if anyone has any pointers toward good examples and tech to use.
22:54 gtmanfred gimpy2938: https://docs.saltstack.com/en/develop/topics/reactor/index.html
22:54 RandyT Leaning toward a Grafana console. Undecided about Carbon, Elasticsearch, ...
22:54 gimpy2938 gtmanfred: if you set Mongo as the default returner what happens when you do something by hand via CLI?  do you still get the same console output but it also goes into Mongo?
22:54 gtmanfred RandyT: i have done it a couple times, my advice is to be patient, cause the documentation could be better
22:55 seblu joined #salt
22:55 gtmanfred gimpy2938:you get both
22:55 gtmanfred yeah
22:55 gtmanfred i don't know that it was a returner
22:55 gimpy2938 gtmanfred: oh, that might be the best approach then...just shove everything into Mongo
22:55 gtmanfred but it was the jobcache for salt master, and the master puts everything into the mongodb
22:55 gtmanfred yup
22:55 gtmanfred i have moved to postgres now, but i did mongo for a minute
22:56 gimpy2938 postgres?  ...like SQL? *shudders*
22:57 gtmanfred heh
22:57 gtmanfred i would say the same thing about mongos query interface
22:57 gtmanfred jsonb hstore, and gin indexes ftw
22:57 gtmanfred https://github.com/gtmanfred/pg_shard_example
22:59 RandyT I was also considering infuxdb for time series data. anyone with any experience there?
22:59 gtmanfred i have been playing with the postgres 9.6 beta last week while setting up openstack on the rackspace public cloud
22:59 gtmanfred not a good reason, i just hate stable things heh
23:01 ashmckenzie joined #salt
23:01 burp joined #salt
23:01 rihannon joined #salt
23:02 onlyanegg joined #salt
23:02 baweaver joined #salt
23:07 denys joined #salt
23:07 jfindlay gtmanfred: so you run arch? :-)
23:08 * jfindlay hides
23:09 gtmanfred yeah, i package salt for it :P
23:09 baweaver joined #salt
23:09 gtmanfred https://www.archlinux.org/people/trusted-users/#dwallace
23:11 jfindlay I guess I can't really say anything without hypocrisy since I run gentoo
23:11 gtmanfred heh +1
23:18 flowstate joined #salt
23:23 snarfy joined #salt
23:27 hasues joined #salt
23:29 hasues left #salt
23:34 tj_devries joined #salt
23:34 whytewolf ... anyone ever use the IFTTT module? I wanted to play around with it a little. but i can't find any documentation on where i can get an api_key from ifttt
23:44 whytewolf Nevermind, got it working
23:53 GreatSnoopy joined #salt
23:53 scoates joined #salt
23:55 aron_kexp joined #salt
23:55 MikeyYeahYeah joined #salt
23:56 stooj joined #salt
23:58 Karunamon Vmware always finds new and interesting ways to disappoint me
23:58 whytewolf sweet. nothing like 5 min of work to add a small stanza for alerting when ip changes. to a route53 home made dyndns setup.
23:58 Karunamon consider, salt.cloud.clouds.vmware can't create any disks if the template doesn't already have a scsi controller.
23:59 Karunamon vmware won't let you add a scsi controller through the UI unless either the host has a scsi device in passthrough mode, or you create a hard drive

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