Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-11-19

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

All times shown according to UTC.

Time Nick Message
00:02 RandyT agreed
00:06 keimlink_ joined #salt
00:08 sh123124213 joined #salt
00:15 buglessdr_ joined #salt
00:21 mrueg joined #salt
00:30 sarcasticadmin joined #salt
00:34 AbyssOne joined #salt
00:38 Klas joined #salt
00:48 mikecmpbll joined #salt
01:00 jas02 joined #salt
01:02 akhter joined #salt
01:02 edrocks joined #salt
01:10 mohae_ joined #salt
01:10 fannet joined #salt
01:10 DEger joined #salt
01:20 sdelic joined #salt
01:32 setkeh Hi guys i have a quick question if i have master.example.com as my salt master and its configuring host.example.com successfully can i also make it apply states to itself 'master.example.com"
01:37 whytewolf setkeh: if you put a minion on the master and link it to it'sself yes.
01:37 whytewolf setkeh: it is pretty common thing to do and highly recommended
01:37 setkeh great that will help making master config easier :D
01:38 setkeh thanks whytewolf , next question is: is there anything special i have to do to link it or just point the master in config to it self ??
01:40 sdelic joined #salt
01:43 setkeh ok i got it figured thanks for the info :)
01:50 setkeh This brings up a new question, if i use a gitfs to update my states on the master and have that happen before the minion classes run in the top file will the sates apply as they were before i initiated the state.apply or will it use the updated states from git ??
01:53 |aaron what is the oldest supported release right now? 2016.3.4? 2015.8.12? as in, if im submitting a pull request to fix a bug, which version should it be against
02:01 jas02 joined #salt
02:03 whytewolf |aaron: Latest Versions: 2015.8.12, 2016.3.4 it is right in the title. looking at that i would say you should hit the 2015.8 develop branch
02:04 |aaron whytewolf, thank you
02:05 whytewolf setkeh: if you are using gitfs. when you run a highstate it typically updates it's cache right then. this happens outside of anything the minion does
02:07 setkeh ok sweet that will save me a whole command :P thanks
02:07 setkeh really enjoying salt atm has been really cool thanks for the help
02:08 whytewolf no problem, I've been a big fan of it for a couple of years now.
02:09 whytewolf i still find new ways of doing things.... which tends to upset my personal openstack cluster cause I end up rebuilding it
02:10 catpiggest joined #salt
02:10 setkeh its so powerful to design orchestration around i mean im still only scratching the surface but its been so simple even for things that are typicly more complicated in puppet and chef
02:11 setkeh whytewolf: lol nice i should really do openstack again currently i have just stock xenservers and doing things with the xapi is nice but means more code
02:13 whytewolf i used to just have stock xenservers, even had full enterprise vmware license for a bit. but i still ended up going with openstack because i just like to tinker
02:13 setkeh whytewolf: do you happen to keep your salt states in a public repo for perusing by chance ??
02:13 whytewolf not at this time. although i am toying with the idea for the next setup.
02:13 Nahual joined #salt
02:14 setkeh yeah tinkering is the greatest part of technology atm there is always somthing to play with :D
02:15 whytewolf I also work with openstack professioanlly so it helps if i have a tinker toy i can play with :P
02:15 setkeh ahh no worries i was interested in how you setup openstack with salt ill google a guide for now :) not something ill do this week it will take time to backup/salt state all my VM's in their current state lol
02:15 setkeh oh yeah for sure :D
02:16 setkeh or maybe its just a really great excuse to rebuild my whole stack lol
02:16 jesusaur joined #salt
02:16 rml joined #salt
02:16 whytewolf lol, always a really good excuse.
02:16 whytewolf here is my home lab http://imgur.com/gallery/HgSk1
02:18 whytewolf the 2950 on the bottom is my salt server, the 3 asus nodes are my openstack controllers. and the 2 lenovo RD450's are my compute nodes
02:18 setkeh Yours is way better than mine lol mine is mainly some older machines i have aquired over the years and a couple of old laptops with broken screens that have i5/i7's and decent ram lol
02:19 setkeh though im hoping i get a new job here shortly so i can build a whole new environment
02:19 whytewolf i started out there. my current lab is just a build up of time and money.
02:20 setkeh for sure i was actually thinking of getting a slightly older blade enclosure my biggest problem will be power consumption no matter what i do lol
02:22 whytewolf keep an eye out on amazon for sales of older systems. they sometimes end up selling some pretty decent rack mount machines. thats were i got the 2950, paid like $120 for it and it is full of SAS disks.
02:22 whytewolf and yeah, power is a pain. but i actually have more heat problems then power. ac bill in the summer kills me
02:23 whytewolf coarse being in a desert doesn't help
02:23 setkeh Nice i do have 2 older Dell 1950's here but they drink power like its going out of fashion not only that i only have 5 GB of ram for them and its expencive to buy more here
02:24 whytewolf ahh yeah that 2950 came with 32 so i kind of lucked out there.
02:26 setkeh What i might do though because my current salt master is in a VM i might configure the one 1950 i have as a salt master to spin up the openstack env then setup a secondary master in a VM after the fact so when it starts getting really hot here i can shut down the 1950 without affecting the salt setup since summer is upon us here in aussieland and hear is also a huge issue
02:27 setkeh there will likely be 3-4 weeks where i only have the server running at all really late at night
02:27 setkeh servers*
02:27 whytewolf i need to get better about turning mine off.
02:27 fracklen joined #salt
02:28 setkeh WOL makes turning them off a much nicer senario
02:28 setkeh thats how i justify it atleast lol i can still VPN in and wake machines up
02:29 whytewolf lol. well unforchantly for me, the 2 RD450's that are the power hogs, also don't have the best power manegment in their ilom. so still pull a couple of kwh
02:30 whytewolf even when off
02:30 setkeh oh jeeze
02:30 whytewolf but they don't generate heat when powered off. so there is that
02:31 setkeh iirc the 1950 uses about 0.3 KWH in poweroff
02:31 setkeh that is true almost a "worth it" trade off lol
02:33 whytewolf alright, it has been fun. but i must be off for the weekend. take care. and happy salting
02:34 setkeh whytewolf: great to meet you mate have a great weekend thanks for the tips
02:38 mosen joined #salt
02:46 sebastian-w joined #salt
02:47 ilbot3 joined #salt
02:47 Topic for #salt is now Welcome to #salt! | Latest Versions: 2015.8.12, 2016.3.4 | 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
02:57 akhter joined #salt
03:02 jas02 joined #salt
03:04 edrocks joined #salt
03:05 aarontc joined #salt
03:11 pipps joined #salt
03:23 greyeax joined #salt
03:24 sh123124213 joined #salt
03:25 darthzen_ joined #salt
03:33 anotherzero joined #salt
03:44 Guest73384 joined #salt
03:47 anotherzero joined #salt
03:52 DEger joined #salt
03:57 bastiand1 joined #salt
03:58 rem5 joined #salt
04:00 onlyanegg joined #salt
04:03 jas02 joined #salt
04:22 informant1 joined #salt
04:31 lompik joined #salt
04:34 fannet joined #salt
04:43 DEger joined #salt
04:50 anotherzero joined #salt
05:03 jas02 joined #salt
05:28 fannet joined #salt
05:41 systo joined #salt
05:42 mk-fg joined #salt
05:54 mpanetta joined #salt
05:56 systo left #salt
06:04 jas02 joined #salt
06:06 edrocks joined #salt
06:28 fracklen joined #salt
06:55 hemebond joined #salt
07:00 tedski joined #salt
07:13 bocaneri joined #salt
07:17 fannet joined #salt
07:17 Miouge joined #salt
07:20 ivanjaros joined #salt
07:36 setkeh Hey guys is there a way i can list directorys in salt:// so i can verify the name of a gitfs ??
07:52 babilen setkeh: cp.list_master
07:53 babilen (or cp.list_states and other functions in cp)
07:53 setkeh babilen: perfect thanks :D
07:53 babilen Ah, you want directories .. that's be cp.list_master_dirs
07:55 mpanetta joined #salt
07:59 setkeh really having some issues updating states from within a state lol i seem to be missing somthing thanks babilen
08:00 babilen That sounds .. wrong
08:01 setkeh ??
08:07 jas02 joined #salt
08:09 edrocks joined #salt
08:11 fannet joined #salt
08:18 ernescz joined #salt
08:19 Miouge joined #salt
08:25 preludedrew joined #salt
08:26 Trauma joined #salt
08:27 breza joined #salt
08:31 breza #anonops
08:31 breza left #salt
08:35 sh123124213 joined #salt
08:41 sh123124213 joined #salt
08:45 Miouge joined #salt
08:56 DEger joined #salt
09:03 Miouge joined #salt
09:05 fannet joined #salt
09:06 alinuxninja joined #salt
09:08 keimlink joined #salt
09:10 ivanjaros3916 joined #salt
09:17 ilbot3 joined #salt
09:17 Topic for #salt is now Welcome to #salt! | Latest Versions: 2015.8.12, 2016.3.4 | 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
09:17 Trauma joined #salt
09:22 ivanjaros joined #salt
09:26 fannet joined #salt
09:27 IdoKaplan joined #salt
09:35 IdoKaplan Hi, I have a problem with indentation when using for. Can you please help? http://pastebin.com/2VYBdsek
09:43 sh123124213 joined #salt
09:51 ernescz IdoKaplan: Hi, maybe you could find useful Jinja's indent (http://jinja.pocoo.org/docs/dev/templates/#indent)
09:55 mpanetta joined #salt
10:00 aidin joined #salt
10:01 IdoKaplan ernescz: Thank you for the follow up! Is there any chance that are you helping me with this specific scenario?
10:05 fannet joined #salt
10:09 jas02 joined #salt
10:11 edrocks joined #salt
10:27 jas02 joined #salt
10:35 aidin joined #salt
10:37 fannet joined #salt
10:48 keimlink joined #salt
10:56 mikecmpbll joined #salt
10:58 fannet joined #salt
11:06 theblazehen__ joined #salt
11:29 sjorge joined #salt
11:29 sjorge joined #salt
11:30 joehh joined #salt
11:33 jas02 joined #salt
11:38 fannet joined #salt
11:39 fannet joined #salt
11:47 babilen setkeh: I simply meant that "updating states from withing a state" sounds, well, prone to errors and I am not sure if what you want to do can be done that way
11:47 babilen Hard to say though without actual information and errors
11:48 aidin joined #salt
11:54 fannet joined #salt
11:57 DEger joined #salt
12:33 Trauma joined #salt
12:34 fannet joined #salt
12:37 jas02 joined #salt
13:05 aarontc joined #salt
13:05 sh123124213 joined #salt
13:16 greyeax_ joined #salt
13:19 samodid joined #salt
13:24 eichiro joined #salt
13:27 stooj joined #salt
13:28 fannet joined #salt
13:33 concernedcitizen joined #salt
13:33 concernedcitizen joined #salt
13:41 jeddi joined #salt
13:57 mpanetta joined #salt
14:14 edrocks joined #salt
14:22 samodid joined #salt
14:23 fannet joined #salt
14:36 raspado joined #salt
14:41 jas02 joined #salt
14:53 stooj joined #salt
14:57 armguy joined #salt
14:59 cyborg-one joined #salt
15:03 amcorreia joined #salt
15:18 aidin joined #salt
15:32 anotherzero joined #salt
15:33 _weiwae__ joined #salt
15:42 jas02 joined #salt
15:50 pppingme joined #salt
15:57 o1e9 joined #salt
16:02 mohae joined #salt
16:17 Norrland joined #salt
16:32 jake_d joined #salt
16:33 keimlink joined #salt
16:36 Nightcinder joined #salt
16:43 jas02 joined #salt
16:47 Aleks3Y joined #salt
16:48 justanotheruser joined #salt
16:52 _weiwae__ joined #salt
17:13 akhter joined #salt
17:16 edrocks joined #salt
17:17 DEger joined #salt
17:17 DEger joined #salt
17:38 DEger joined #salt
17:44 jas02 joined #salt
17:59 mpanetta joined #salt
18:00 NeoXiD I use LocalClient() for additional execution logic. By using client.cmd('*', 'state.sls', kwarg={'mods': ['state1', 'state2', ...]}) I can just run usual SaltStack states on all minions
18:00 NeoXiD Afterwards, I'm using salt.output.display_output(results, 'highstate', __opts__) to display the output
18:01 NeoXiD Is there any better way to do that, any less complicated one?
18:03 NeoXiD It works, but seems overcomplicated, it is blocking and I have to do manual error checking
18:07 orionx joined #salt
18:12 iggy !xyproblem
18:12 saltstackbot What are you actually trying to achieve? Do you have an xy problem? http://mywiki.wooledge.org/XyProblem
18:13 NeoXiD iggy: Was that related to me? If so, the runner executes several states (the easy part), then executes various commands on various minions, collects them into dictionaries, uses them in other commands on different minions, ...
18:14 NeoXiD Basically collecting data and execution commands on minion set A, then use that data to do HTTP REST API requests, then use the results of that for minion set B
18:14 NeoXiD That kind of flow seems to be impossible with the usual highstate thing, so I went for a custom runner as I'm familiar with Python.
18:16 _weiwae joined #salt
18:21 amontalban joined #salt
18:21 amontalban joined #salt
18:28 NeoXiD If there's an X instead of Y, please tell :)
18:34 iggy yeah, you mentioned none of that in your original question(Y), so that's the X
18:34 NeoXiD I asked about X before and was told to go the Y-route, that's why I skipped
18:34 stooj joined #salt
18:34 NeoXiD (few weeks ago)
18:35 alxchk left #salt
18:36 iggy have you tried a workflow (or is there a reason it wouldn't work) more like: run states on A, they push data to sdb, fire events when they are done, events kickoff reactor to run states on B
18:36 iggy where the states on B could use the data from sdb
18:37 NeoXiD I thought about that too, but it didn't seem that easy to use more complex REST APIs within Salt states
18:37 sdelic joined #salt
18:37 NeoXiD And those REST API queries -have- to be executed on the Salt master, as they require confidential credentials
18:38 iggy hmm
18:40 iggy orch job that does states on A (which push data to sdb), API stuff on master (using data from sdb/phase1 and pushing data for next phase to sdb), then run states on B ?
18:40 iggy the orch will make sure each phase is complete before moving on
18:42 iggy I guess that doesn't really solve the problem of data output, error checking/correction, but at least it doesn't block ;)
18:43 jas02 joined #salt
18:44 iggy back to your original question, the only runner I can think of that does something similar is the orch runner, and it just doesn't output anything... so not a prior solution to look at there
18:47 sh123124213 joined #salt
18:54 fracklen joined #salt
18:58 hwtt joined #salt
19:03 zer0def joined #salt
19:11 XenophonF is there a salty way to configure /etc/defaults/ conf files on debian
19:12 XenophonF other than templating the files or doing a bunch of file.replace states, i mean
19:18 edrocks joined #salt
19:23 shaneod joined #salt
19:23 shaneod left #salt
19:24 nidr0x joined #salt
19:25 wolfpackmars2 joined #salt
19:26 keimlink joined #salt
19:34 Nahual joined #salt
19:34 cDR_ joined #salt
19:35 cDR_ hi, what is the best way to set a grain after an highstate completed with succes?
19:37 cDR_ I want to set a grain with the results of the last highstate
19:39 sh123124213 joined #salt
19:50 fracklen joined #salt
20:06 jas02 joined #salt
20:10 akhter joined #salt
20:11 akhter cDR: What's an example result?
20:14 cDR_ for now I got a state file that sets the 'highstate' grains with a timestamp and a status. Since I include this statefile in my top.sls it is always run. But I would like to know if the highstate failed so that I can set a 'status: failed' grain
20:14 cDR_ might need a reactor for this
20:16 raspado joined #salt
20:17 akhter Why not use a fail_hard and order your states?
20:17 akhter So to begin with, set your grain to "failed" on your states, if any states failed, set a statement on those states to fail_hard, meaning they won't continue if they failed.
20:18 akhter Use requirement statements along the states in top.sls and make sure your grains.set success statement is last.
20:19 akhter https://docs.saltstack.com/en/latest/ref/states/failhard.html
20:19 akhter https://docs.saltstack.com/en/latest/ref/states/ordering.html
20:20 akhter cDR: ^
20:21 cDR_ Interesting approach, I like it. Failhard does seem to be a 'big' setting considering the impact
20:21 akhter cDR: It's a global setting and can be used everywhere as far as I know.
20:22 akhter I personally think failhard should be true by default.
20:22 cDR_ I will give it a try!
20:22 akhter Good luck.
20:22 cDR_ it feels safer I agree if enabled
20:24 akhter cDR: I believe failhard is new, please test it before changing all of your states.
20:24 cDR_ are states in the top file ordered as listed?
20:25 akhter That's right.
20:26 akhter You use one of the target schema's (grains, pillars, pcre, or a compound of the previous 3).
20:26 akhter And then list the states you want to run underneath.
20:27 akhter https://docs.saltstack.com/en/latest/ref/states/top.html
20:27 akhter cDR: *
20:28 cDR_ testing now
20:31 akhter cDR: Let me know how it goes.
20:32 cDR_ I got Unable to render top file: Conflicting ID '*'
20:33 cDR_ my top.sls; https://gist.github.com/githubcdr/d7cb6090b15011e55d7588e1b65104af
20:34 akhter One second.
20:35 akhter Line 3 and line 68.
20:36 cDR_ true, I now use 'kernel:*': match grain as a workaround
20:36 akhter That would work.
20:39 akhter You might also be able to test with the show_highstate module.
20:39 akhter or show_lowstate, that might show the order it will run the states.
20:39 akhter cDR: ^
20:39 cDR_ wow, great success ^^
20:40 akhter https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html
20:40 cDR_ this works, got a nice timestamp grain on my highstate as a bonus
20:40 akhter Nice.
20:40 cDR_ thanks akhter
20:41 akhter You're welcome.
20:41 cDR_ Might open a feature request to always run a pre and/or a post state to every highstate :)
20:43 akhter That will help, although I know people who run highstates in CI apps (jenkins, teamcity, bamboo, etc), they will want an ignore option too :)
20:44 ivanjaros joined #salt
20:54 zer0def joined #salt
21:06 LeProvokateur joined #salt
21:18 ThomasJ joined #salt
21:18 fannet joined #salt
21:51 rem5 joined #salt
21:51 DEger joined #salt
22:00 mpanetta joined #salt
22:03 onlyanegg joined #salt
22:08 jas02 joined #salt
22:12 fracklen joined #salt
22:14 scottcrooks joined #salt
22:14 sc250024 Hello, can anyone here provide some feedback on the php-formula ? I am wanting to update it based on the new ondrej/ppa repo (it needs it desperately) and just have some questions
22:15 sc250024 Specifically, the difference between "ng" and non ng states / formulas
22:20 edrocks joined #salt
22:28 iggy sc250024: what do you want to know?
22:30 robbintt joined #salt
22:37 sc250024 @iggy The `php-formula` badly needs to be updated. The `ondrej/php` repo is now the default repo, and the formula needs to be updated to account for the new repo structure
22:37 sc250024 Currently the `php-formula` has the NG (Next-Gen) and non-NG version. I guess my biggest questions are...
22:38 sc250024 1. How much / little does the non-NG version of formulas need to be updated? Can we deprecate them entirely? If not, what's the goal of maintaining a non-NG version
22:39 fracklen joined #salt
22:39 sc250024 2. Through the formula, can I force use of the PPA exclusively? Or does the formula need logic built in for people who solely want to use the native repositories of Ubuntu / Debian (the formula update I will write really only applies to Debian family OSes)
22:48 iggy I don't specifically know the difference between the non/ng versions of this formula, but I don't think they're expected to have feature parity
22:49 iggy if the formula currently uses the upstream packages, you should keep that functionality and add an optional flag to enable the ppa
22:50 sc250024 Alright
22:50 iggy i.e. don't break backward compatibility without a _really_ good reason
22:50 iggy ^ those are general rules... I don't consider myself all that up to speed on that specific formula, so somebody else may have more input
22:51 iggy but you probably won't get that feedback until you do the work and submit the PR
22:51 hemebond Are formulas tagged?
22:51 iggy :/
22:51 iggy lol
22:51 hemebond aw
22:51 sc250024 Yeah, that's true
22:51 sc250024 Hence why I'm trying to ask before to avoid an argument in the PR
22:52 iggy we barely have the manpower to keep up with the formulas as is... much less to maintain tags and branches
22:54 hemebond I was thinking more along the lines of versioning using tags.
22:54 hemebond Similar to what ng does.
22:55 hemebond And also matching the formula to a salt version.
22:56 iggy yeah, we don't have the manpower for such things
22:56 iggy (not to say that's a bad idea, just we can't do it right now)
22:57 iggy there's actually a lot of things that have been suggested over the years that are in the same boat (writing actual tests for formulas being one of the other big ones)
23:08 jas02 joined #salt
23:12 akhter joined #salt
23:13 sc250024 joined #salt
23:27 preludedrew joined #salt
23:34 abednarik joined #salt
23:54 abednarik joined #salt

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