Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-12-15

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

All times shown according to UTC.

Time Nick Message
00:01 pipps99 joined #salt
00:02 pipps99 joined #salt
00:06 pipps joined #salt
00:11 cilkay joined #salt
00:16 prg3 joined #salt
00:18 Hatifnatt joined #salt
00:21 raspado joined #salt
00:23 jas02 joined #salt
00:37 bhl joined #salt
00:39 jmickle anyone around?
00:45 Hatifnatt left #salt
00:53 whytewolf maybe
00:54 hatifnatt joined #salt
00:54 nicksloan joined #salt
00:58 jmickle :-D you any good with jinja templates?
00:59 whytewolf I have been known to dabble in the mystic arts from time to time
00:59 jmickle hehehe
01:00 jmickle i need to basically set a variable that that is multi level
01:00 jmickle and figure out how to find some values from that map
01:00 jmickle any good documentation you can point me at
01:01 whytewolf multi level as in a dict or list?
01:02 jmickle hmmm thinking
01:02 jmickle thinking of the best way to solve this
01:03 jmickle i think it would be like “appname”: { “roles”: [a,b,c]}
01:03 jmickle so if role is b then i know its in appname
01:03 whytewolf ahh okay that would be a dict of dicts
01:03 jmickle and i could have appname2
01:03 jmickle appname 3
01:04 jmickle basically i need to map to a specific configuration based on a service class of role names
01:04 jmickle so i have a list of role names, and those map to a specific service class
01:04 jmickle and i am trying to dynamically get that service class
01:04 whytewolf okay, so sounds like what you want is a pillar
01:05 jmickle but how do i traverse a pillar to see which one my role falls into?
01:07 hatifnatt joined #salt
01:11 whytewolf humm. I'm thinking for something like this you might want to explore map.jinja
01:12 whytewolf https://docs.saltstack.com/en/latest/topics/development/conventions/formulas.html#abstracting-static-defaults-into-a-lookup-table
01:13 jmickle awesome
01:13 jmickle il read through that page thank you so much for the push in the right direction
01:16 whytewolf you can also look at https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.grains.html#salt.modules.grains.filter_by
01:17 jmickle ty
01:18 KevinAn27 joined #salt
01:27 nethershaw joined #salt
01:36 debian112 joined #salt
01:36 promorph_ joined #salt
01:38 codeape joined #salt
01:41 JPT joined #salt
02:01 bltmiller joined #salt
02:04 catpiggest joined #salt
02:23 k_sze[work] joined #salt
02:25 k_sze[work] joined #salt
02:30 anotherzero joined #salt
02:36 jas02 joined #salt
02:40 sebastian-w joined #salt
02:46 cilkay Hi. I have a state that contains: {% set remote_version = salt['postgres.psql_query'](user='tcm_utility_user', password='seekrit',  port='5432', host='myhost', maintenance_db='tcm_utility', query='select max(meta.version) from meta') %}. That runs fine with 2015.5.3 but with 2016.11.0 in a masterless minion, I get: Jinja variable 'salt.utils.templates.AliasedLoader object' has no attribute 'postgres.psql_query'
02:46 cilkay Any idea how I can deal with this?
02:47 ilbot3 joined #salt
02:47 Topic for #salt is now Welcome to #salt! <+> Latest Versions: 2016.3.4, 2016.11.0 <+> 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
02:48 mattp_ cilkay: if you salt-call that from the minion with the same log level what happens (not the state, but the postgres.psql_query
02:48 mattp_ but ya, one of a,b or c :)
02:49 mattp_ its annoying that failures like that invoked from master just get swallowed/tossed. makes debugging annoying
02:49 cilkay I'm installing the Debian package in both cases.
02:49 hemebond Using v2016.11? Postgres is broken.
02:49 cilkay Yes, 2016.11.
02:49 cilkay Nice.
02:50 cilkay And we only use masterless minions so all salt-call operations are on the minion.
02:50 hemebond 2016.11 is broken. You can work around it sometimes but I found it easier to downgrade.
02:50 cilkay How far back should I go?
02:50 mattp_ hemebond: larger problems with the release you mean?
02:50 Nahual joined #salt
02:51 hemebond Well, problems that matter to me; like broken EC2.
02:51 hemebond There was nothing in 2016.11 that I really needed.
02:52 cilkay I'm only upgrading because I'm building new images using Packer so I figured I'd use the latest supported version, not because I needed a specific feature not available in 2015.5.3.
02:52 mattp_ I feel like the salt core team would do well to code freeze and just work on the issue queue for like 12 months..
02:52 hemebond That would be nice.
02:54 evle joined #salt
02:54 UtahDave joined #salt
02:58 cilkay The version in the Debian stable repos is 2014.1.13. Man Debian moves glacially.
02:59 mattp_ cilkay: lol. you can add the salt apt repo
02:59 UtahDave yep
02:59 buu debian backports has 2016.3
03:00 mattp_ hey UtahDave, youre on the core team right
03:00 UtahDave I'm on a lot of teams, lol
03:00 cilkay mattp_: I was using http://repo.saltstack.com/apt/debian/8/amd64/latest. That pulled in 2016.11.0. I'll try the backports repo from the docs.
03:01 buu cilkay: are you on stable?
03:01 cilkay Debian stable? Yes.
03:02 buu ah
03:05 cilkay Installing 2016.3.4 from the Stretch backports worked. Thanks all.
03:05 hemebond According to the Salt repo instructions you can do http://repo.saltstack.com/apt/debian/8/amd64/2016.3
03:05 hemebond But it didn't seem to work for me so I just put a hold on the packages.
03:06 cilkay It's probably safer to use from the Debian Jessie backports since that would have been packaged by the Debian packaging team.
03:06 hemebond Mmm, not sure about that.
03:08 cilkay Yay! My first VMware ESXi VM created via Packer and configured via Salt is working now.
03:08 raspado by default does salt-master auto allow minions to authenticate to it if the salt master/minion pki are the same?
03:09 hemebond raspado: I don't believe so.
03:09 hemebond Actually I have no idea.
03:09 hemebond At all.
03:10 raspado hehe
03:10 raspado let me try
03:11 skeezix-hf joined #salt
03:14 raspado ok so i changed the ip of the salt master on /etc/salt/minion but im getting an error "The Salt Master has cached the public key for this node, this salt minion will wait for 10 seconds before attempting to re-authenticate "
03:14 raspado do i need to do something else to get the new salt-master to recognize this mininion?
03:14 hemebond Delete the minion pki from the master?
03:14 bluenemo joined #salt
03:15 raspado on the new master?
03:15 raspado i did copy the pki keys across all salt-masters
03:16 hemebond The minion pki seems to be different on the master and affected minion.
03:16 raspado hemebond: oh i see the host when doing salt-key -L
03:16 raspado as an unnaccepted key
03:18 raspado woo hoo
03:18 raspado accepted
03:18 amontalban joined #salt
03:20 cswang joined #salt
03:23 raspado fortunately i copied all the master keys to the sm's before hand
03:23 raspado that would have been a pita
03:30 ALLmightySPIFF joined #salt
03:32 bastiand1 joined #salt
03:42 Sammichmaker joined #salt
03:49 jas02 joined #salt
03:49 trent joined #salt
03:52 pipps joined #salt
04:18 trent joined #salt
04:19 chowmeined joined #salt
04:41 bemehow joined #salt
04:51 jas02 joined #salt
04:54 bemehow joined #salt
05:02 _KaszpiR_ joined #salt
05:10 dwfreed joined #salt
05:12 seanz joined #salt
05:12 markizano joined #salt
05:13 markizano soo... I'm getting a new error in my states: Failed to configure repo 'ppa:webupd8team/java': cannot parse "ppa:" style repo definitions: ppa:webupd8team/java
05:13 markizano ^ Never had a problem with ppa-style repos before... why all of a sudden this is an issue ?
05:13 hemebond markizano: Did you upgrade salt-minion?
05:14 markizano Salt: 2016.3.4
05:14 markizano >.<
05:14 markizano wow ...
05:14 markizano so apparently there's more fail to this server than that...
05:14 markizano Thanks, hemebond
05:14 hemebond 👍  Good luck
05:19 amontalban joined #salt
05:29 pipps joined #salt
05:30 debian112 joined #salt
05:35 rdas joined #salt
05:40 notCalle joined #salt
05:42 markizano aww...
05:42 markizano if __grains__['os'] in ('Ubuntu', 'Mint'):
05:42 markizano ^ why not Debian supported ?
05:43 bocaneri joined #salt
05:44 markizano https://github.com/saltstack/salt/blob/2016.11/salt/modules/aptpkg.py#L1610 <- just ubuntu ??
05:51 bayindirh markizano: PPA's may not be officially supported in debian yet.
05:52 markizano hrmm.. okies.
05:52 hemebond Yeah, PPA's are an Ubuntu thing.
05:52 markizano xplains why I'm just now seeing this issue.
05:53 bayindirh markizano: Which ppa you're trying to install. Maybe there's an alternative debian repo for these packages.
05:53 markizano ppa:webupd8team/java
05:54 bayindirh you can get official java packages, and build the debs with java-package.
05:55 bayindirh Then you can distribute them via salt:// or simply build a repository with aptly.
05:55 markizano even if I add 'Debian' and 'Devuan' to just this hosts' ./salt/modules/aptpkg.py list of those grains it checks, I get an error from the apt-add-repository command - which is what I'd expect,since PPA could be handled on debian, even if they aren't officially supported - perhaps that's just me opinion. it's current behaviour is probably technically the correct one.
05:56 bayindirh markizano: PPA support was on the way for Deb 9, but IIRC that was developer oriented. I also discourage using Ubuntu oriented PPAs in Debian. Some small library and/or dependency problems can cause big headaches.
05:57 markizano ^yar, that.
05:58 bayindirh java-package is a very practical way to get the debs in Debian. Then distribute it with aptly or salt://. I worked on Debian a lot tust me ;)
05:59 bayindirh You can build a proper repository in 15 minutes flat. If you add signing, you can do it in an hour.
06:01 bayindirh If you actually want webupd8's packages, aptly can mirror them for you too. Also update them automactically if you wish.
06:02 markizano hrmm
06:02 markizano Specifically, we'd need that version of oracle-java-8
06:03 hemebond Just deploy the deb via salt://
06:03 * markizano subscribes to being the updater of java now
06:04 bayindirh markizano: http://aptly.info
06:04 jmickle joined #salt
06:05 bayindirh I worked in Debian derivative development and also managed some installations which are isolated from internet, hence used aptly a lot. It's the most practical way. Java has quite number of packages, which will be a headache with salt://
06:06 bayindirh Clone the repository with aptly, sign the set if you want and publish. You can update the repository with a single command.
06:08 jas02 joined #salt
06:13 rdas joined #salt
06:14 samodid joined #salt
06:19 gladia2r joined #salt
06:20 bayindirh markizano: if you want to do that, I can help.
06:21 markizano thanks bayindirh - this is what I was looking for as far as managing apt repos were concerned.
06:23 bayindirh I'm glad that helped. Aptly can manage any number of repos with a single installation. You can also create mix-match and create new repositories. Aptly even takes care of the dependencies.
06:23 markizano sounds handy
06:26 bayindirh It's a very useful tool. It really changes a lot if you are managing big local repositories.
06:35 tobiasBora joined #salt
06:43 samodid joined #salt
06:46 debian112 joined #salt
06:47 keimlink joined #salt
06:47 pipps joined #salt
06:55 madboxs_ joined #salt
07:03 jas02 joined #salt
07:05 debian112 joined #salt
07:20 fracklen joined #salt
07:23 debian112 joined #salt
07:27 jas02 joined #salt
07:34 jhauser joined #salt
07:37 infrmnt joined #salt
07:49 netcho_ joined #salt
07:51 felskrone joined #salt
07:52 bayindirh Quick question: is there a way to get result of a last state apply's result, if the minion timed out? I have some long SLS files which takes some time to complete.
07:55 iggy you should see a jid somewhere, you can look that up
07:57 jhauser joined #salt
07:57 impi joined #salt
07:58 o1e9 joined #salt
08:00 jhauser_ joined #salt
08:04 bayindirh iggy: Like where?
08:10 ws2k3 joined #salt
08:18 pipps joined #salt
08:19 Reverend joined #salt
08:19 Reverend morning everyone
08:22 amontalban joined #salt
08:22 amontalban joined #salt
08:23 samodid joined #salt
08:27 coredumb morning
08:27 coredumb is there some script lying around that use salt to only do state rendering ?
08:28 cyteen joined #salt
08:30 coredumb rendering just like a minion running
08:33 jhauser joined #salt
08:47 fracklen joined #salt
08:48 buu you mean.. template substitutions?
08:49 s_kunk joined #salt
08:51 sarlalian joined #salt
08:54 netcho_ joined #salt
08:57 inad922 joined #salt
09:01 rai_ joined #salt
09:04 madboxs joined #salt
09:08 RandyT joined #salt
09:08 irated joined #salt
09:16 fracklen joined #salt
09:16 teclator joined #salt
09:17 fracklen joined #salt
09:25 N-Mi_ joined #salt
09:26 ivanjaros3916 joined #salt
09:41 madboxs joined #salt
09:44 bayindirh Is there any chance that using refresh: True in centos systems accidentally upgrades whole system instead of package db?
09:44 AndreasLutro doubt it
09:45 bayindirh I've seen many "yum --quiet -y upgrade" while running a job.
09:45 bayindirh There's only one system upgrade command, only at the end.
09:46 chowmein__ joined #salt
09:46 bayindirh Taking a look now, will look to the code next, and will report if I see it :D
09:47 ivanjaros joined #salt
09:50 jhauser joined #salt
09:52 bayindirh Hah nice! My system update package was updating a salt minion update, which was published recently,
09:53 bayindirh and that was killing my minion in the process.
09:53 bayindirh Wow. just wow.
09:53 aw110f joined #salt
09:56 aw110f_ joined #salt
09:59 stooj joined #salt
10:00 pipps joined #salt
10:05 jhauser joined #salt
10:06 babilen bayindirh: Upgrade from which version?
10:09 bayindirh babilen: 2016.11.0-1.el6 -> 2016.11.1-1.el6
10:09 bayindirh I have no systemd on that system, so that's not unexpected.
10:10 babilen So you exptected the behaviour you saw?
10:11 bayindirh Yes. Salt forks itself a process via python popen and runs my task. That task updates salt minion and restarts itself.
10:11 bayindirh So if you kill a process' parent process, their forks (children) also get terminated.
10:12 babilen Yes, I am well aware of that issue and it has been addressed in the systemd unit file
10:12 bayindirh That's standard unix behaviour. I think systemd has something to prevent this, however Redhat 6.x family has no systemd :D
10:12 babilen (which doesn't help you as you don't use it)
10:12 babilen I am not entirely sure what you require help with now
10:12 bayindirh this is a matter of "can't" rather than "don't" :)
10:14 babilen Are you asking for procedures to upgrade the salt minion using salt on non-systemd systems?
10:14 sarlalian joined #salt
10:15 bayindirh babilen: Oh, sorry. I was talking with AndreasLutro, asked a small question about a SLS problem (about pkg.install); and that was the outcome.
10:15 bayindirh That was not a question, however, being able to exclude repositories while updating system would be nice to prevent these problems.
10:15 bayindirh Sorry for excess noise & misunderstanding again
10:16 babilen Can you exclude repositories in yum?
10:16 babilen (as I'm sure we can add that to salt if upstream has that functionality)
10:17 bayindirh from yum manual:
10:17 bayindirh --disablerepo=repoidglob => Disables specific repositories by id or glob.
10:18 babilen Should be trivial to integrate into salt then
10:18 babilen Just add a parameter that takes a list of repositories to disable and pass it to yum
10:18 babilen (to the pkg.upgrade execution module function)
10:19 florianb joined #salt
10:19 salato joined #salt
10:19 bayindirh Thanks. It also has --enablerepo which selectivelu enables repositories.
10:19 bayindirh I've just verified that "yum --disablerepo=saltstack upgrade" upgrades everything except saltstack.
10:20 salato hello, is that possible to divide pillar env like salt env does? I have pillar/prod/top.sls bub salt '*' pillar.items pillarenv=prod returns the ones of the base env only
10:20 babilen bayindirh: In fact I see disablerepo support in salt already: https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.yumpkg.html#salt.modules.yumpkg.upgrade
10:21 babilen (and enablerepo)
10:22 bayindirh I've looked to 'https://docs.saltstack.com/en/latest/ref/states/all/salt.states.pkg.html#salt.states.pkg.uptodate' and assumed that it contains the superset, my bad.
10:22 bayindirh I'm just getting familiar with the docs, sorry.
10:24 babilen bayindirh: States and modules are quite a different beast. I wouldn't really recommend to run pkg.uptodate outside of initial startup states as it is (IMHO) advisable for highstates to be idempotent
10:24 babilen Did I elaborate on the differences or are you aware of them already?
10:25 bayindirh babilen: I'm developing an initial provisioning state. This is not a high-state by any means. We use a middleware, which has a strange installation procedure. I have to apply the document as-is otherwise it causes some nasty headaches.
10:26 bayindirh After initial provisioning this state won't be used. This will not be present in top.sls.
10:26 dariusjs joined #salt
10:26 bayindirh thanks for the warning though. I'll take note.
10:27 florianb left #salt
10:28 florianb joined #salt
10:30 stooj joined #salt
10:30 babilen bayindirh: It isn't really something that is enforced, but many people found that it is a very desirable property of your highstates to have. It's just nice to be able to run them anytime and knowing that it won't break things.
10:31 babilen "highstate fear" is nothing you want to develop :)
10:32 bayindirh babilen: Actually, since it's desirable for me, I'm enforcing myself to make my top.sls idempotent. Managing hundeds of machines is a bit too brave and exciting without idempotent highstate. :)
10:33 salato also what is the correct syntax to target multiple minion ids in the same "row" ? 'dev*,prod*' seems to not work
10:34 bayindirh salato, is using a regex like (dev*|prod*) acceptable for you?
10:36 babilen salato: There are various ways to achieve that.. Sounds as if you want a compound matcher with either globbing or regex (or both)
10:36 salato sure
10:37 babilen "dev* or prod*" would do nicely though
10:37 salato uh ok
10:37 salato I'm still unable to target different pillar environments though
10:37 babilen Make sure to specify that it is a compound matcher
10:37 babilen Sorry, I am not touching environments
10:38 salato :)
10:38 salato I'm still not sure why I'm using them, but ok...
10:38 bayindirh does jinja conditionals with grain values make sense?
10:39 bayindirh salato: maybe you can use these.
10:40 salato probably, I still don't have grains set to distinguish those minions
10:41 salato so grains are evaluated on the master, right? the catch here is to distribute the correct password key set in a pillar value to the correct minion
10:41 jhauser joined #salt
10:41 bayindirh Yes, that's how I do: I have different pillars for different environments. So distribute them via top.sls
10:42 bayindirh So just sending the same state will do the correct thing on every minion.
10:42 netcho_ joined #salt
10:42 bayindirh So I'm dividing pillar envionments with file names (os1_dev.sls, os1_prod.sls, os2_dev.sls, etc.).
10:42 babilen bayindirh: And how do you set those "environments" ?
10:43 babilen (or rather, how do you target your pillars?)
10:43 bayindirh via top.sls. It can be grain, hostnames, or any other data.
10:43 bayindirh osfinger is very useful for me.
10:44 babilen Please keep in mind that minions can claim to have whatever grains they want .. so I wouldn't target sensitive information based on grain data
10:44 stooj joined #salt
10:44 jhauser_ joined #salt
10:44 salato backing up, using a regex like (dev*|prod*): does not work...weird, "dev*" or "prod*" gives me error on the "or" part
10:45 salato ok the or should have been inside the quotes
10:45 babilen salato: It is "dev* or prod*" not "dev*" or "prod*"
10:45 babilen It's a single compound matcher
10:45 bayindirh I know. In my environment pillars doesn't contain any usernames or passwords. That's why I'm rather lax.
10:46 bayindirh I don't forsee that I will use any custom grains for targeting. Our hostnames are also very descriptive and useful for targeting.
10:46 babilen salato: That regex would have to be (dev.*|prod.*) if you want to match anything after dev or prod
10:46 babilen bayindirh: Sure
10:47 salato clearly... thanks
10:47 bayindirh babilen: Thanks for all the warnings, these are greatly appreciated btw.
10:47 chowmeined joined #salt
10:51 irctc643 joined #salt
10:51 irctc643 Hi is it known that onchanges does not work if aggregate is on by default?
10:52 salato is it possible to list available environments? maybe the master doesn't recognizes this pillar env
10:53 fracklen_ joined #salt
10:53 Arendtse1 joined #salt
10:59 amcorreia joined #salt
10:59 jhauser joined #salt
11:00 Deliant joined #salt
11:07 jhauser_ joined #salt
11:10 impi joined #salt
11:12 cyteen joined #salt
11:18 cyborg-one joined #salt
11:19 dariusjs joined #salt
11:29 stooj joined #salt
11:37 stooj joined #salt
11:56 dariusjs joined #salt
11:57 jhauser joined #salt
11:57 zer0def joined #salt
11:58 gladia2r joined #salt
12:01 jas02 joined #salt
12:02 pipps joined #salt
12:04 jhauser joined #salt
12:06 d3c4f_ joined #salt
12:07 bayindirh Does salt automatically update pillars if they are changed?
12:08 demize- joined #salt
12:08 jhauser_ joined #salt
12:08 bayindirh e.g. When a new state.apply is submitted?
12:09 Nightcinder joined #salt
12:17 frew joined #salt
12:21 cscf joined #salt
12:21 saintaquinas[m] joined #salt
12:22 hemebond joined #salt
12:22 _Cyclone_ joined #salt
12:26 dariusjs joined #salt
12:27 stooj joined #salt
12:39 jas02 joined #salt
12:42 stooj joined #salt
12:55 amontalban joined #salt
12:55 amontalban joined #salt
13:03 pipps joined #salt
13:05 amontalban joined #salt
13:05 amontalban joined #salt
13:13 stooj joined #salt
13:16 netcho_ joined #salt
13:40 stooj joined #salt
13:50 stooj joined #salt
13:57 darioleidi joined #salt
13:57 darioleidi joined #salt
13:58 darioleidi joined #salt
13:59 tkharju joined #salt
14:01 stooj joined #salt
14:08 stooj joined #salt
14:09 mrchoc joined #salt
14:11 mrchoc Hi there! Just wondering, can I call grains on the command line? Like "salt server1 cmd.run "foo { host }""
14:11 mrchoc or rather use grains
14:12 dariusjs joined #salt
14:12 Witol joined #salt
14:15 Witol Hi - It is possible to use salt-ssh with sudo user and not using NOPASSWD in sudoers ?
14:16 numkem joined #salt
14:17 Tanta joined #salt
14:23 bayindirh mrchoc: you can use grains to target servers and other stuff. What are you trying to do?
14:26 cyteen joined #salt
14:28 gableroux joined #salt
14:28 nicksloan joined #salt
14:31 mpanetta joined #salt
14:32 mpanetta joined #salt
14:34 Witol It works when i put passwd in roster file but it not work if i prompt salt-ssh with   --askpass
14:37 drawsmcgraw joined #salt
14:44 dxiri joined #salt
14:47 netcho joined #salt
14:47 racooper joined #salt
14:50 stooj joined #salt
14:50 bowhunter joined #salt
14:52 ekristen joined #salt
14:52 mavhq joined #salt
14:56 nickabbey joined #salt
14:57 hatifnatt Has anybody used users-formula? https://github.com/saltstack-formulas/users-formula Is it any way to use it within another state, or only one option to create pillar(s) with list of users and "connect" it to some hosts?
14:58 anotherzero joined #salt
15:03 Straphka_ joined #salt
15:03 dxiri joined #salt
15:05 stooj joined #salt
15:07 impi joined #salt
15:12 babilen hatifnatt: What do you mean by "use in another state" ?
15:12 keltim joined #salt
15:13 babilen The basic assumption with that formula is that you target "- users" to all minions and that the pillar data decides what is being done. You can, naturally, include "- users" in other SLSs and define requisites and so on.
15:13 babilen You just have to make sure that  you target suitable pillar data (which can be structured in any way you like when you rely on pillar merging)
15:17 hatifnatt I'm writing state wich includes user creation, but default 'user' state isn't enought I need sudo management and authorized_keys management. Sure I can write anyting by hand but users-formula can make anyting what I need.
15:18 dxiri joined #salt
15:18 cscf hatifnatt, then you don't put it in the state, basically.  You assign that info in Pillar, assign - users in salt/top.sls, and put a comment in your state about it.
15:19 hatifnatt Am I right that I can use 'users:' in pillar only once?
15:20 cscf hatifnatt, Well, you can use it multiple times, but the last one wins
15:20 cscf The list won't be merged
15:20 jmickle joined #salt
15:21 hatifnatt So I must use one pillar and describe all users I need?
15:23 bemehow joined #salt
15:23 cscf 1 for each host, yes
15:23 cscf list merging is on the wishlist, I think
15:25 babilen hatifnatt: You can use "users:" as often as you want and use pillar merging
15:26 babilen hatifnatt: https://docs.saltstack.com/en/latest/ref/configuration/master.html#pillar-merging-options -- That way you can structure your pillar in any way you like
15:26 Mattch joined #salt
15:26 babilen That allows you to easily "mix in" additional users and/or their options
15:27 babilen You might, for example, decide to use a single separate SLS for each individual user or combine users into "groups" in a SLS. You might also want to keep, say, sudoers options in separate SLSs
15:27 babilen The way you structure this is obviously entirely up to you and your needs
15:28 babilen Please read through the pillar merging documentation and configure it to your liking
15:28 babilen smart + list merge sounds like the right way for you
15:29 nickabbey joined #salt
15:29 hatifnatt babilen: thanks I will read docs merging sound like what I need.
15:36 sp0097 joined #salt
15:39 sp0097 Hi, I am setting up the orchestration runner for deploying a long running task.  How do I setup a compound match?
15:41 ecdhe I'm running masterless salt on vagrant boxes.  My /etc/salt/minion file says "file_client: local".  Everything is working as I expect, all my states are applied, but when I look in the logs it appears salt is at some point timing out trying to find a master in "multiple-master" mode.
15:41 ecdhe How can I keep salt minion from checking for masters?
15:42 sp0097 I was trying salt.state: -tgt: {{ data[‘name’] }} and G@servertype:db
15:42 sp0097 -tgt_type: compound
15:43 hatifnatt babilen: great pillar mering exactly what I need, now I can structure files in more logic way - different users in different pillars without need to include everyting in one top-level pillar.
15:43 AndreasLutro ecdhe: try stopping the salt-minion service (systemd/sysvinit/whatever), since that's probably still using your old config
15:44 mavhq joined #salt
15:44 nickabbey joined #salt
15:46 babilen hatifnatt: Exactly
15:53 FireSnake joined #salt
15:53 Witol joined #salt
15:54 sarcasticadmin joined #salt
15:56 fracklen joined #salt
16:03 sp0097 In addition, I am passing pillar information on the command line.
16:04 sp0097 salt-run state.orch orchestrate.multisid pillar='{"event_data": {"name": "db1"} }'
16:05 sp0097 A non-compound globbing works just fine.
16:06 ivanjaros joined #salt
16:06 ecdhe Thanks AndreasLutro.  Because I'm running through vagrant, this should be the first instance of salt-minion to have ever run on the host.  But I'm going to look into salted vagrant to see what command it's calling to kick salt-minion off in the first place.
16:07 babilen sp0097: Not sure if it's related, but event_data != data
16:09 babilen And you might want to use proper whitespace (no space after '-') and also replace data[‘name’] with data['name'] (single quotes are needed here not your fancy shite)
16:09 sp0097 balilen: Yeah, you are right, I omitted one important detail.  At the top of my orchestration file, I have {% set data =  salt.pillar.get(‘event_data’) %}
16:11 babilen Are you sure you can access pillar data there?
16:11 babilen grr ... have to be afk for a while
16:11 sp0097 yeah, If I simply have tgt: {{ data[‘name]] }} it works through command line and salt-cloud
16:12 sp0097 Just the compound match doesn’t work through the command line
16:16 babilen sp0097: Could you paste the actual SLS to http://paste.debian.net ?
16:16 stooj joined #salt
16:16 windowsrefund joined #salt
16:17 windowsrefund hi
16:21 Ztyx joined #salt
16:22 sp0097 sure
16:22 jweede joined #salt
16:22 jmickle joined #salt
16:23 lws joined #salt
16:23 agend joined #salt
16:24 sp0097 http://paste.debian.net/902339/
16:24 sp0097 Take a look at setup-oracle-db as an example
16:24 dyasny joined #salt
16:25 Trauma joined #salt
16:26 amontalban joined #salt
16:26 amontalban joined #salt
16:27 cyteen joined #salt
16:28 rlatimore joined #salt
16:28 jas02 joined #salt
16:29 babilen sp0097: '{{ data['name'] }}' uses single quotes twice. It shouldn't cause issues, but you might want to check that. You also seem to enclose your compound matchers in ( ... ) -- Why do you do that?
16:30 babilen Does setup-oracle-prereqs work? What does actually happens if you run that?
16:34 impi joined #salt
16:35 babilen sp0097: '( {{ data['name'] }} and G@servertype:db))' also suffers from unbalanced parantheses
16:43 sp0097 yeah, setup-oracle-prereqs works find
16:43 sp0097 *fine
16:43 teclator joined #salt
16:43 sp0097 oh, I wonder if that’s the issue, the unmatched parans.
16:45 sp0097 I had a more complex match and then tried to back it down, and made a mistake when trying to test something less complex.
16:47 sp0097 I took out all of the parans and going to test it
16:47 sp0097 hopefully that does the trick
16:47 babilen parens away!
16:47 sp0097 :)
16:47 sp0097 thank you!
16:47 babilen Did that solve it?
16:48 sp0097 I am going to run it now, but it’s going to take a little while to test…I’ll report back when the process finishes
16:48 babilen https://github.com/Fanael/rainbow-delimiters ftw
16:49 sp0097 oh that’s funny…
16:49 pipps joined #salt
16:50 babilen afk, good luck!
16:51 pipps joined #salt
16:52 sp0097 thanks
16:52 nickabbey joined #salt
16:56 dxiri joined #salt
16:56 stooj joined #salt
16:58 sp0097 babilen: It says it ran successfully, but no changes were listed.  So, I don’t think it ran the way I expected.  :(
17:02 sp0097 Here is the output to show what I mean: http://paste.debian.net/902345/
17:02 dyasny joined #salt
17:04 pysen joined #salt
17:04 mavhq joined #salt
17:05 pipps joined #salt
17:05 aarontc joined #salt
17:05 impi joined #salt
17:07 jor joined #salt
17:08 upb joined #salt
17:08 nledez joined #salt
17:08 nledez joined #salt
17:08 lubyou joined #salt
17:08 sybix joined #salt
17:08 tbrb joined #salt
17:09 fhh joined #salt
17:11 tercenya joined #salt
17:11 nickabbey joined #salt
17:13 pipps joined #salt
17:14 tiwula joined #salt
17:18 nebuchadnezzar joined #salt
17:18 raspado joined #salt
17:18 raspado what is the difference between running state.apply vs state.highstate?
17:19 FireSnake joined #salt
17:19 raspado is it considered the same thing?
17:23 nickabbey joined #salt
17:24 woodtablet joined #salt
17:26 netcho joined #salt
17:36 ajw0100 joined #salt
17:37 zer0def joined #salt
17:41 raspado is there a good way to upgrade the salt-minions ?
17:42 pysen joined #salt
17:48 Bico_Fino joined #salt
17:50 cscf raspado, you could just call pkg.upgrade on the minions first?
17:50 numkem joined #salt
17:52 numkem joined #salt
17:52 onlyanegg joined #salt
17:58 nickabbey joined #salt
17:59 nickabbey joined #salt
18:00 nidr0x joined #salt
18:02 pipps joined #salt
18:08 nixjdm joined #salt
18:08 Bico_Fino_ joined #salt
18:12 raspado when i do a salt-run manage.up
18:12 raspado i see shortnames of hosts, where is the minion getting this hostname from?
18:12 Bico_Fino joined #salt
18:19 dxiri joined #salt
18:21 fracklen joined #salt
18:21 MTecknology heh... I didn't think about the PR's not being one shot. Now we have this.  "525c746 May Cthulhu take mercy on my soul for this commit."
18:22 lws joined #salt
18:22 iggy I saw that
18:22 zero_shane joined #salt
18:23 iggy shook my head and then I saw it was you
18:23 iggy shook my head more vigorously
18:24 zero_shane hey all - I'm starting to integrate RKT containers in to our environment - we use Saltstack pretty heavily ... was poking around looking for States for RKT management, but am not seeing anything in the docs - does anyone have the low-down on RKT support in Saltstack ??
18:24 iggy coredumb: best I've found is -l debug mixed with test=True
18:24 iggy zero_shane: there is none
18:24 zero_shane was (sadly) gathering that to be the case - but was hoping it might be in the "roadmap" somewhere ?
18:25 cscf MTecknology, what's the commit do, anyway?
18:28 MTecknology iggy: I was "thinking" it would be merged in one shot. I knew better...
18:28 cscf MTecknology, what's the commit do, anyway?
18:28 cscf oops sorry
18:30 cyteen joined #salt
18:30 MTecknology cscf: It's the commit where I realized I'm about to go down a scary rabbit hole of brainfuckery the likes of which I've never experienced.
18:30 cscf I see
18:30 MTecknology It's where I started reworking a lot of logic and I knew it was gonna come with many dragons.
18:32 pipps joined #salt
18:33 dxiri joined #salt
18:34 MTecknology apparently we broke docs.saltstack, eh?
18:34 jmickle joined #salt
18:35 pipps99 joined #salt
18:36 MTecknology gtmanfred: ping?
18:39 eprice joined #salt
18:42 fracklen joined #salt
18:45 aw110f joined #salt
18:49 hatifnatt It seems like 'ssh_known_hosts' still broken in 2016.11.0+ds-1
18:51 mavhq joined #salt
18:53 Edgan joined #salt
18:53 johnkeates joined #salt
18:53 Praematura joined #salt
18:59 beardedeagle joined #salt
19:01 cyborg-one joined #salt
19:04 iggy zero_shane: I haven't heard anybody talk about it... best bet is to search around on github... maybe also look for a rkt formula and see if there are modules/states there
19:06 sh123124213 joined #salt
19:07 madboxs_ joined #salt
19:08 numkem joined #salt
19:09 raspado our salt minions are on salt-minion 2015.5.10 (Lithium)
19:09 raspado but our salt master is newer, is there a safe way to upgrade the minion?
19:10 Grokzen joined #salt
19:10 raspado to salt-minion 2016.11.0 (Carbon)
19:12 iggy salt '*' pkg.install salt-minion
19:12 iggy should do it
19:12 Salander27 joined #salt
19:12 raspado will that update the salt-minion to the version the master is?
19:13 edrocks joined #salt
19:13 raspado im migrating some salt-minions from an old salt master (2015) but my new salt master is carbon
19:13 Bico_Fino joined #salt
19:14 iggy no, it'll upgrade it to whatever it thinks the latest version is
19:14 iggy you would of course have to setup repos and all that beforehand
19:14 raspado oh... i need it to be the same version as the salt master
19:14 raspado ahhh ok ok i see what you mean
19:15 iggy there's not a pkg.magic_install_same_version_as_master function
19:15 raspado on the salt master i specify the version in bootstrap so new minions dont install the latest version, then we have all kinds of versions of minions out there
19:15 raspado found out the hardware after seeing errors in logs because by default the repo is set to latest -_-
19:16 raspado hardware/hard way :)
19:16 raspado anyway, thx iggy
19:16 Edgan raspado: I build my own debs of salt(with patches) and bootstrap by putting salt in the AMI in EC2.
19:17 bowhunter joined #salt
19:18 mavhq joined #salt
19:18 jhauser joined #salt
19:19 AndreasLutro we run our own apt repository named "salt-stable" which is our own cherry-picked version that's been tested in our own environment
19:20 jas02 joined #salt
19:22 Edgan AndreasLutro: I use the official salt repository for dependecies, and then use apt pinning to make sure my custom salt packages win
19:23 wendall911 joined #salt
19:24 raspado interesting
19:25 AndreasLutro sounds like there's a risk of dependency version mismatches
19:25 fracklen_ joined #salt
19:28 raspado hmmm
19:28 raspado so if i do pkg.install
19:28 raspado its doesnt upgrade the existing package
19:29 raspado oh i suppose i can use pkg.upgrade
19:29 AndreasLutro pkg.install just does whatever apt-get install does
19:29 dyasny joined #salt
19:29 AndreasLutro no, pkg.upgrade is `apt-get upgrade`
19:31 raspado oh that would be bad
19:32 jenastar joined #salt
19:32 jenastar left #salt
19:32 raspado is there a way to mimic an apt-get install --only-upgrade <packagename>
19:32 AndreasLutro probably, if you look at the pkg.install documentation
19:32 AndreasLutro !salt modules.pkg.install
19:32 saltstackbot https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.pkg.html#salt.modules.pkg.install
19:33 AndreasLutro !salt modules.aptpkg.install
19:33 saltstackbot https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.aptpkg.html#salt.modules.aptpkg.install
19:33 AndreasLutro that's it
19:34 jas02 joined #salt
19:34 jhauser joined #salt
19:34 raspado thx!
19:35 raspado only_upgrade nice
19:41 gtmanfred MTecknology: pong
19:42 raspado how do i pass the only_upgrde to the salt command
19:42 raspado is it pkg.install only_upgrade:true salt-minion ?
19:43 AndreasLutro = instead of :
19:43 Edgan AndreasLutro: Depends on if you work of a versioned apt repo, or latest. I work off latest, but I also keep on top of new versions. I am generally waiting for the new version so I can drop patches on my version.
19:43 gtmanfred and move salt-minion before the keyword argument like you would in python
19:43 gtmanfred salt \* pkg.install salt-minion only_upgrade=True
19:43 raspado ah kk thx AndreasLutro
19:43 raspado thx gtmanfred
19:44 _JZ_ joined #salt
19:44 gtmanfred also, i don't see an only_upgrade on the module
19:44 Ztyx joined #salt
19:44 gtmanfred so it might only work with the pkg.installed and pkg.latest states
19:44 gtmanfred infact, i bet it is only available from the state
19:44 sh123124213 joined #salt
19:46 raspado gtmanfred: so id have to somehow invoke this through an sls file?
19:46 gtmanfred yes
19:47 gtmanfred i bet that the pkg.installed state checks to see if it is installed first, before doing anything else if only_upgrade is set to True
19:47 raspado hmmm, could I simply do a yum --upgrade salt-minion on the remote hosts instead?
19:47 raspado then i can just do a cmd.run "yum --upgrade salt-minion" after I set the salt repo to carbon
19:49 iggy if you are using yum, then you probably want to look at the yumpkg docs instead of the aptpkg docs
19:49 iggy !salt modules.yumpkg.install
19:49 saltstackbot https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.yumpkg.html#salt.modules.yumpkg.install
19:49 MTecknology gtmanfred: Still interested in helping me with those tests? :)
19:50 nicksloan joined #salt
19:50 promorphus joined #salt
19:51 gtmanfred MTecknology: i am, unfortunately i am dealing with customer issues right now
19:52 DammitJim joined #salt
19:54 MTecknology I didn't figured you'd be ready this instant. Any guesses when you might be free?
19:54 sh123124213 joined #salt
19:54 gtmanfred after christmas probably
19:55 sh123124213 can somebody which has access to a server with a minion connected to the master get commands which are intended for another minion ?
19:57 amontalb1n joined #salt
19:57 iggy sh123124213: yes
19:59 sh123124213 is there any way I can prevent or detect that ?
20:00 buu switch to ansible?
20:00 raspado gtmanfred: so looks like pkg.upgrade is avail in the module
20:00 buu stop letting untrusted users have root on your machines?
20:00 lws joined #salt
20:01 iggy sh123124213: not really, it's inherent to salt's design, it blasts stuff on the zmq bus
20:01 gtmanfred don't put secret data on the event bus? store it in pillars like it is designed
20:02 netcho joined #salt
20:02 iggy sh123124213: I'm not sure if that is true of the tcp or raet transports, but I'd guess it is
20:04 gtmanfred it should still be true
20:04 gtmanfred because the minion is the thing that does its matching to see if it should run the command
20:04 gtmanfred so they check all the events, and look for ones they match
20:04 buu So, do you guys use salt \* ...; or salt '*' ...;
20:04 MTecknology gtmanfred: sounds good, I can ping you then. :)
20:04 iggy thought so, just wanted to leave a small glimmer of hope for them
20:04 gtmanfred buu: salt \*
20:05 gtmanfred GTMANFRED: DREAM CRUSHER!~
20:05 iggy I try to avoid sending commands to *
20:05 buu And at one point was the idea of using non-bash globbing operators discarded
20:05 buu ....
20:05 buu er, what point
20:05 buu Jesus god how do you have word dyslexia
20:07 sh123124213 gtmanfred : I heard something about getting rid of zmq and having only tornado ? is that true ?
20:07 jas02 joined #salt
20:08 lws joined #salt
20:09 raspado is there a way to do a dry run of "salt "stage1*" pkg.upgrade name=salt-minion" ?
20:10 sh123124213 going back to my initial use-case, I want client to download large files and the command will be from the module but I don't want all client to have all files so I pass some authentication. I'm worried that somebody else might use this.
20:11 nickabbey joined #salt
20:12 raspado ah upgrade_available
20:12 raspado jesus
20:12 iggy sh123124213: salt really isn't designed for multi-tenancy
20:13 raspado salt "stage1*" pkg.upgrade_available salt-minion just returns true
20:13 raspado can i get the version that it would upgrade to
20:15 netcho joined #salt
20:16 raspado oh yes, list_upgrades :)
20:17 sh123124213 https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.yumpkg.html#salt.modules.yumpkg.upgrade_available
20:19 writtenoff joined #salt
20:20 raspado oooo wait a second, so salt will not install pkg dependencies as well?
20:20 gtmanfred sure it will
20:20 gtmanfred it just uses yum or apt or whatever package manager
20:21 raspado i got this error on the minion that is requires python-futures http://pastebin.com/QA866x9C
20:22 raspado could i pass a param to upgrade dependencies?
20:22 gtmanfred no
20:22 gtmanfred do -l debug, and get the yum command that it ran to get that error
20:22 jmickle joined #salt
20:22 gtmanfred aslo, will need to run it from the minion, salt-call
20:23 nickabbey joined #salt
20:23 jhauser joined #salt
20:23 raspado k let me try
20:24 eprice joined #salt
20:27 mavhq joined #salt
20:28 nicksloan joined #salt
20:28 jhauser_ joined #salt
20:28 jhauser joined #salt
20:30 SaucyElf joined #salt
20:30 johnkeates joined #salt
20:31 pipps joined #salt
20:31 raspado gtmanfred: should I do the upgrade from the minions?
20:31 raspado i see i did it from the salt master and it timed out :)
20:33 jas02 joined #salt
20:34 Praematura joined #salt
20:35 babilen raspado: With current versions and systemd you can just upgrade the salt packages like any other package
20:36 SaucyElf joined #salt
20:37 jhauser_ joined #salt
20:38 gtmanfred but make sure to upgrade your master first
20:39 gtmanfred also, you have to have salt 2016.3.4 or newer to have the upgrade work correctly
20:39 gtmanfred and be using a distro with systemd
20:40 debian112 joined #salt
20:40 hatifnatt Sure ssh_known_hosts.present works fine only with fingerprint, specifing enc and key broke whole thing in different manners. Broken permissions, broken known_hosts file for example.
20:42 whytewolf hey gtmanfred where is the salt offices in ut?
20:42 dxiri joined #salt
20:43 raspado yes, the master is at carbon
20:43 dxiri joined #salt
20:44 raspado i mean boron
20:44 whytewolf cause i think i just got sent a possabile contract position for Saltstack ...
20:44 gtmanfred whytewolf: lehi
20:44 whytewolf is american fork around there?
20:45 gtmanfred yeah
20:45 whytewolf lol
20:45 gtmanfred there are a bunch of startups and companies in lehi
20:45 whytewolf Role : Devops Engineer with SaltStack (UT, 6+Months) Skype Interview
20:45 gtmanfred microsoft has an office there, and so does adobe
20:45 gtmanfred oh lol
20:45 gtmanfred weird
20:45 gtmanfred did you fill out the form on our bamboohr page?
20:46 eprice joined #salt
20:46 whytewolf nope. this is from a head hunter
20:46 whytewolf IDC Technologies
20:46 gtmanfred ¯\(°_o)/¯
20:46 dyasny joined #salt
20:46 whytewolf if it wasn't a contract i would consider it
20:46 whytewolf :P
20:47 whytewolf also if i didn't have to live in utah
20:47 raspado saltstack needs to move to SF :)
20:47 jhauser joined #salt
20:47 whytewolf naw, las vegas. lots of sun and sin
20:48 raspado haha yes
20:48 whytewolf and me....
20:48 gtmanfred i don't want to live in SF,
20:48 JoeJulian who does?
20:49 gtmanfred whytewolf: we are not actually working with any contractors
20:49 gtmanfred JoeJulian: apparently raspado
20:49 johnkeates why would anyone still want to live in the USA
20:49 raspado :)
20:49 JoeJulian Hehe
20:49 djgerm joined #salt
20:49 johnkeates maybe hawaii, since it's far enough away from the mainland
20:49 johnkeates or alaska, since canada will buffer you ;-)
20:49 whytewolf gtmanfred: strange. this is defintly a Saltstack position... this line kind of gave it away. " Provide help desk support for questions about SaltStack development and systems deployment automation."
20:50 gtmanfred whytewolf: well, that job is available for remote for the right fit
20:50 whytewolf forget hawaii, more expensive then new york
20:52 whytewolf gtmanfred: i do find it strange if you guys don't work with contractors why this position came through as a contract position. I almost turned it down flat because of that.
20:52 gtmanfred yeah, jeff has no idea where it is coming from
20:53 jas02 joined #salt
20:53 gtmanfred this is the job if you want to apply to it directly https://saltstack.bamboohr.com/jobs/view.php?id=47
20:54 whytewolf humm, wonder what the drop in pay would be like
20:55 whytewolf if it would be enough to compensate for the moving to a product i actually support
20:56 aawerner joined #salt
20:56 XenophonF i work from home so "well stocked breakroom" isn't much of a draw ;)
20:56 debian1121 joined #salt
20:57 pipps joined #salt
20:57 jmickle joined #salt
20:59 JoeJulian Unless they would stock your kitchen for you.
21:02 stooj joined #salt
21:03 jas02 joined #salt
21:04 nickabbey joined #salt
21:08 netcho joined #salt
21:18 sp0097_ joined #salt
21:19 pipps joined #salt
21:19 sh123124213 12 paid holidays is standard in US ?
21:22 whytewolf not really. currently at the back i work at there are 8 paid holidays
21:22 whytewolf s/back/bank
21:24 sh123124213 europe has at least 20 :)
21:24 whytewolf thats nothing. china i know has weeks per holiday.
21:25 jas02 joined #salt
21:25 whytewolf and they were goverment mandated.
21:25 canci joined #salt
21:29 whytewolf gah, so many places have issues with my prefered email domain
21:29 whytewolf whats wrong with a guru domain?
21:29 mavhq joined #salt
21:29 whytewolf or ninja
21:29 buu No one believes you!
21:29 whytewolf most don't believe it is a valid email domain
21:30 whytewolf i also have problems with my .us domains.
21:32 buu I don't use email at funny domains any more =[
21:32 buu I'm old and busted
21:33 whytewolf how old is old?
21:33 KevinAn27 joined #salt
21:33 voileux joined #salt
21:33 Ztyx joined #salt
21:34 stooj joined #salt
21:35 buu old enough!
21:35 whytewolf cause I'm 40 and still do silly stuff
21:35 buu Every timme I learn a new javascript framework I age 5 years
21:35 buu So at this point I'm 125
21:35 chayden joined #salt
21:36 chayden left #salt
21:37 ivanjaros joined #salt
21:39 sh123124213 joined #salt
21:42 sh123124213 joined #salt
21:42 onlyanegg joined #salt
21:42 stooj joined #salt
21:46 giany hi, if someone modifies a file on a minion is it possible to make salt master to keep the original version of that file?
21:47 jas02 joined #salt
21:48 djgerm that's a great use case for salt
21:49 djgerm files.managed should cover that for you
21:52 jas02 joined #salt
21:52 giany ok, but who triggers the update on file change?
21:52 stooj joined #salt
21:53 whytewolf giany: depends. you can have a beacon do it [inotify beacon if linux] or a schedule
21:54 mikecmpbll joined #salt
21:58 racooper giany,  take a look at https://docs.saltstack.com/en/latest/topics/beacons/ . their example sounds a lot like what you're looking for.
22:02 giany thx, will look into it.
22:02 pipps joined #salt
22:05 aarontc joined #salt
22:05 johnkeates joined #salt
22:06 PerfectDeception joined #salt
22:06 sh123124213 joined #salt
22:12 teclator joined #salt
22:13 iggy there's not really a way to get a file's contents _before_ it's saved (there is an interface in Linux that can do something similar, but afaik, nobody has made salt interface with that)
22:14 nebuchadnezzar joined #salt
22:15 debian112 joined #salt
22:17 raspado joined #salt
22:17 raspado is there a command i can run on salt master to return the minions version
22:17 iggy salt '*' test.version
22:17 nledez joined #salt
22:17 nledez joined #salt
22:17 raspado noiceee thx iggy
22:18 iggy there are also grains you can look at
22:18 devster31 joined #salt
22:18 titilambert joined #salt
22:18 raspado kk, this test.version is perfect
22:19 ReV013 joined #salt
22:19 MeltedLux joined #salt
22:19 CustosLimen joined #salt
22:20 fhh joined #salt
22:20 jas02 joined #salt
22:22 iggy generally I just have a state that enforces a salt-minion version instead of trying to do all this via execution modules
22:23 raspado iggy: we've locked it down via salt-cloud provisioning with "script_args: stable 2016.3.4"
22:24 raspado is that okay to do?
22:24 iggy except when you want to upgrade
22:24 raspado yeahh true
22:24 iggy then you have to go back and remember all these commands
22:24 raspado hmmm, thx for reminding me :)
22:24 sh123124213 joined #salt
22:24 iggy whereas if you just bump a number in a pillar file... bam!
22:25 djgerm joined #salt
22:25 raspado true
22:27 djgerm if I wanted to react on an event from a minion checking in with salt-master after say service salt-minion restart or reboot, what would that event look like? salt/minion/*/started ?
22:27 djgerm better yet: where in the world can I see where those events are defined?
22:28 iggy the code
22:28 iggy there aren't events for reboot, only for minion startup
22:28 djgerm oh
22:28 djgerm ah but that's ok, cause minion starts up after reboot
22:29 iggy and honestly, I've found the easiest place to find events is actually watching the event bus :/
22:30 Ztyx joined #salt
22:41 MTecknology djgerm: https://docs.saltstack.com/en/2015.8/topics/event/index.html
22:47 ReV013 left #salt
22:51 anotherzero joined #salt
22:53 fracklen joined #salt
22:54 djgerm thanks. i just had to google it :slaps forehead:
22:55 djgerm and now I remember I read over those like… a year ago too *sigh* getting old and tired
22:57 Salander27 joined #salt
23:01 edrocks joined #salt
23:05 jas02 joined #salt
23:13 sarcasticadm joined #salt
23:23 stooj joined #salt
23:27 MTecknology my one grain for a known environment in a relatively static state is >400 lines and almost no comments...
23:27 * MTecknology sighs
23:27 MTecknology that's worse than my commit messages
23:32 mirko joined #salt
23:44 stooj joined #salt
23:50 woodtablet left #salt
23:56 stooj joined #salt

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