Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2018-05-15

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

All times shown according to UTC.

Time Nick Message
00:00 onslack joined #salt
00:12 nidr0x joined #salt
00:17 nidr0x joined #salt
00:25 nidr0x joined #salt
00:26 Eugene Speaking of Formulas, I'm looking for an example of how one would use the nginx + php + wordpress stuff together, without repeating myself a bunch of reinventing the wheel. I'm intimidated by the ng stuff that these formula repos seem to have grown since the last time I dove into it
00:27 Eugene (or to put it another way, I want to copy-paste a thing into my top file / Pillar and have this stupid site just WOrk)
00:28 Eugene Its not clear to me that the provided pillar.example files can work out-of-the-box with each other
00:45 noobiedubie joined #salt
00:50 zerocoolback joined #salt
00:58 onovy joined #salt
01:08 noobiedubie joined #salt
01:54 stooj joined #salt
01:54 onovy joined #salt
01:56 ilbot3 joined #salt
01:56 Topic for #salt is now Welcome to #salt! <+> Latest Versions: 2017.7.5, 2018.3.0 <+> Support: https://www.saltstack.com/support/ <+> Logs: http://irclog.perlgeek.de/salt/ <+> Paste: https://gist.github.com/ <+> See also: #salt-devel, #salt-offtopic, and https://saltstackcommunity.herokuapp.com (for slack) <+> We are volunteers and may not have immediate answers
02:00 shiranaihito joined #salt
02:22 zerocoolback joined #salt
02:32 zerocoolback joined #salt
03:08 avgtechie joined #salt
03:22 Psi-Jack So. Regarding GitFS methods of Salt, what kind of basic workflow(s) should be utilized when using it, when you're ONLY using GitFS and nothing else? I know in the past, I'd had issues with it back 4~6 years ago, but this is 2018. I've set, in a specific minion's config, environment: development, created in my salt repo a develop branch, similarly done with the pillar repo.
03:31 Psi-Jack Like for example, if I make changes to the devel git branch of my pillar repo, it will not get reflected in the master until merged to master, and because of ext_pillar for the devel branch being set to env: development, it will not cross into the production environment?
04:52 tiwula joined #salt
05:17 sauvin joined #salt
05:41 briner joined #salt
05:51 masber joined #salt
05:56 sjorge joined #salt
05:56 stooj joined #salt
06:15 bluenemo joined #salt
06:19 bluenemo joined #salt
06:35 Ricardo1000 joined #salt
06:37 cbosdonnat joined #salt
06:37 cbosdonnat joined #salt
06:40 DanyC joined #salt
06:41 DanyC_ joined #salt
06:49 hrumph joined #salt
07:00 briner joined #salt
07:13 tom[] joined #salt
07:13 Hybrid joined #salt
07:16 Pjusur joined #salt
07:22 armyriad joined #salt
07:27 tom[] joined #salt
07:29 orichards joined #salt
07:33 _xor joined #salt
07:34 Hybrid1 joined #salt
07:35 marwel @Psi-Jack, yes, branches are saltenvironments
07:35 marwel and it works fine for me
07:36 rollniak joined #salt
07:37 Tucky joined #salt
07:39 jrenner joined #salt
07:40 babilen marwel: How do you deal with top files?
07:40 DanyC joined #salt
07:40 GrisKo joined #salt
07:40 marwel merge strategy and normally i just have only one in base
07:41 zerocoolback joined #salt
07:48 armyriad joined #salt
07:49 babilen marwel: Last time I had them in the same repo, they were merged between branches, is that not a problem for you?
07:51 awerner_ joined #salt
07:53 DanyC joined #salt
07:55 mikecmpbll joined #salt
08:00 toanju joined #salt
08:02 marwel in most cases they don't differ on my branches, so merge is not a problem
08:20 GrisKo joined #salt
08:21 bdrung_work joined #salt
08:33 mikecmpbll joined #salt
08:36 Mattch joined #salt
08:39 tys101010 joined #salt
08:45 defsdoor joined #salt
08:49 zerocoolback joined #salt
08:50 zerocoolback joined #salt
09:04 xist joined #salt
09:09 mikecmpbll joined #salt
09:12 toanju joined #salt
09:13 mage_ any idea how could I put a key "*" in my pillar config ?
09:14 mage_ I tried '*': auth_user = pgbouncer but then I get a     - Rendering SLS 'base:pgbouncer.config' failed: expected alphabetic or numeric character, but found '#'; line 29
09:14 mage_ with:                 &#39;*&#39;: auth_user = pgbouncer    <======================
09:15 mage_ with \* it works but then I have \* instead of * in my config file
09:17 zerocoolback joined #salt
09:23 mage_ this is the erro: https://gist.github.com/silenius/14e5154362d366e3f7d861ab014a3239
09:24 rollniak joined #salt
09:26 fredvd joined #salt
09:26 rollniak joined #salt
09:28 zulutango joined #salt
09:30 zerocoolback joined #salt
09:31 xet7 joined #salt
09:52 GrisKo joined #salt
10:08 mikecmpbll joined #salt
10:10 yidhra joined #salt
10:43 Deliant joined #salt
10:51 toanju joined #salt
11:07 crux-capacitor mage: what are you trying to accomplish using that asterisk?
11:21 xist joined #salt
11:34 zerocoolback joined #salt
11:42 mage_ crux-capacitor: setting a * = auth_user = pgbouncer in a [databases] section of an ini file
11:44 dendazen joined #salt
11:47 crux-capacitor mage: i see. have you tried wrapping that whole line in quotes in the pillar file?
11:47 crux-capacitor single quotes
11:59 tyx joined #salt
12:05 simondodsley_ joined #salt
12:19 eMBee i am stuck with this error: Failed to load ext_pillar container_state: 'minion-id'
12:19 eMBee i read that refreshing pillar data might help, but that didn't do anything
12:24 Nahual joined #salt
12:28 toanju joined #salt
12:32 briner joined #salt
12:50 exarkun joined #salt
12:51 chowmeined joined #salt
12:54 tom[] joined #salt
12:55 chowmeined joined #salt
12:57 tom[] joined #salt
13:05 jesusaur joined #salt
13:06 edrocks joined #salt
13:11 Reverend joined #salt
13:14 Hybrid joined #salt
13:16 briner joined #salt
13:19 briner joined #salt
13:19 Psi-Jack Hmm, yeah, that's what I was wondering about, base.
13:23 Psi-Jack When working with adding new stuff for example to the used states, I branched master into devel, I /added/ a development: section to the top and started adding the new states I intend to put into production after development is finished.
13:24 racooper joined #salt
13:25 DammitJim joined #salt
13:25 Psi-Jack gtmanfred: Welcome back BTW. I'm guessing you're back home by now? ;)
13:26 Pjusur joined #salt
13:27 briner joined #salt
13:34 tyx joined #salt
13:37 mchlumsky joined #salt
13:42 Hybrid joined #salt
13:45 avgtechie joined #salt
13:47 nixjdm joined #salt
13:49 Brew joined #salt
13:52 slav0nic joined #salt
13:56 cgiroua joined #salt
14:04 franksma joined #salt
14:05 AvengerMoJo joined #salt
14:07 mchlumsky joined #salt
14:08 onslack joined #salt
14:09 onslack <gtmanfred> should be fixed
14:09 onslack <psi-jack> Seems to be.
14:10 gtmanfred perfect
14:10 onslack <psi-jack> I wonder if it works with threads. ;)
14:10 onslack <psi-jack> Ohh, it does.
14:10 onslack <gtmanfred> :+1:
14:15 noobiedubie joined #salt
14:17 onslack <psi-jack> Last night I managed to convert two standard salt states I had made myself, and converted to using them as saltstack-formula based ones. Did a LOT of re-writing to the saltstack/vault-formula, because it was completely bass-ackwards from what saltstack-formula/consul was.
14:20 onslack <psi-jack> Now, I just need to figure out how to extend it, so I can add in my custom parts specifically that I wouldn't want to put into the formula itself. slightly_smiling_face
14:22 mchlumsky joined #salt
14:23 onslack <gtmanfred> psi-jack: <https://docs.saltstack.com/en/2017.7/ref/states/extend.html>
14:24 onslack <psi-jack> Yep. Already looking into it. I've never extended before. Knew it could be done, just never had a need to. slightly_smiling_face
14:27 slav0nic_ joined #salt
14:27 tiwula joined #salt
14:29 exarkun joined #salt
14:35 dxiri joined #salt
14:40 onslack <tmacey> @psi-jack I'd be interested in taking a look at your Consul and Vault formulas, as I have my own that I maintain <https://github.com/mitodl/consul-formula> <https://github.com/mitodl/vault-formula>
14:41 onslack <psi-jack> Hey @tmacey. My consul formula is the same as saltstack-formula/consul-formula, currently. But my vault-formula is modified heavily from that.
14:42 onslack <psi-jack> <https://git.linux-help.org/psi-jack/vault-formula>
14:42 onslack <tmacey> Cool, I'll take a look
14:45 onslack <psi-jack> Whom did the salt-formulas/consul-formula did a fantastic job of organization and clean stuff. Whomever did the vault-formula, looks totally different, and did not get a a lot of the vault stuff, not even enough to run it.
14:45 onslack <psi-jack> hehe. Was interesting to pretty much essentially rebuild it. ;)
14:45 onslack <tmacey> Yeah, there are some pieces of the Consul formula that I've been meaning to adopt, but in general I've been happy with the ones that I'm using
14:46 onslack <psi-jack> Yeah. The consul one is very sharp, simple, clean.
14:46 onslack <psi-jack> I'm going to write more formulas based on that style I think.
14:47 onslack <psi-jack> Mmmm, you have a dns_proxy state as well, which is handy. ;)
14:47 onslack <tmacey> Well, all of the ones that I've written are BSD-3 and available at github.com/mitodl in case there's anything there that would be useful for you
14:47 onslack <psi-jack> Except that I just configured my core DNS server to directly query the appropriate port on all servers.
14:47 onslack <psi-jack> But, still not a bad idea to include it.
14:48 onslack <psi-jack> Heh, git.linux-help.org is my own Gitea server, most of it is public. slightly_smiling_face
15:01 XenophonF psi-jack: you should totally file a PR with your changes
15:01 onslack <psi-jack> slightly_smiling_face
15:02 onslack <psi-jack> I was planning to. Hehe
15:10 onslack <psi-jack> Heh, will in fact, be trying to figure out if it's reasonably possible to do a cross-site PR from gitea->;github
15:14 tyx joined #salt
15:24 FL1SK joined #salt
15:25 immune2iocane joined #salt
15:26 noah_h joined #salt
15:33 noah_h slack talk colby
15:33 noah_h s///
15:35 nielsk joined #salt
15:49 mindo joined #salt
15:52 AvengerMoJo joined #salt
15:56 noah_h joined #salt
16:08 ecdhe joined #salt
16:10 noah_h Writing an execution module and try to run it, getting the error: "Passed invalid arguments to boto_secretsmanager.get: get_secret_value() only accepts keyword arguments.", tried executing it with named args and positional args but no luck.
16:11 onslack <gtmanfred> sounds like it doesn’t accept any positional arguments
16:11 noah_h Defined as: def get(arn, key, last_changed_date):
16:12 noah_h Do I need to define with **kwargs or something instead?
16:12 zerocool_ joined #salt
16:15 dlloyd noah_h: that looks like its complaining about get_secret_value, not get
16:15 orichards joined #salt
16:15 noah_h yea I just saw that, I think I'm blind
16:15 noah_h Do execution modules run via python2 or 3?
16:16 dlloyd afaik everything is 2 unless you explicitly force it to 3? that might be out of date though
16:16 noah_h How do you force? Via Shabang?
16:17 onslack <msmith> they run in whatever version salt is running
16:18 DanyC joined #salt
16:19 DanyC_ joined #salt
16:19 noah_h Ok thanks for rubber duckying, that was silly of me.
16:20 noah_h left #salt
16:22 druonysus joined #salt
16:24 zerocoolback joined #salt
16:38 zmalone joined #salt
17:01 sh123124213 joined #salt
17:13 tyx joined #salt
17:14 slav0nic__ joined #salt
17:15 edrocks joined #salt
17:27 tiwula joined #salt
18:04 tacoboy joined #salt
18:22 woodtablet joined #salt
18:22 woodtablet hello
18:24 woodtablet is this a bug from upgrading to 2018-03-01 ? my salt master is giving this error: Authentication failure of type "user" occurred. and i can issue commands to my minions
18:30 woodtablet i set the master to debug mode.. and i am getting very little
18:31 ymasson joined #salt
18:32 PatrolDoom joined #salt
18:37 woodtablet nm, that wasnt my issue. debug mode showed a python module that was missing that i already reinstalled, but i had to restart all the minions. i am good now =D
18:43 MyGitIsDown joined #salt
18:47 bluenemo joined #salt
18:47 MyGitIsDown is there a service like a salt master cron scheduler that I can use to delete archived salt events store in mysql?  The salt.returners.mysql only allows for archiving or deleting events older then x hours but not for the purging of archived events after x days/hours
19:00 rcvu does pkg.repo want the repo id or repo name for the name?
19:01 TTimo joined #salt
19:01 rcvu woops, found it
19:01 rcvu It wants repo id
19:04 zer0def joined #salt
19:04 rcvu Is there any significance to differenct colours of returned grains?
19:05 rcvu I am running salt '*' grains.get osmajorrelease and my 2 minions are both returning 6
19:05 rcvu But one 6 is green and the other yellow?
19:06 TTimo Hello! .. I would like to make sure that packages matching a particular pattern are not installed or being removed, namely anything matching "libnvidia-*" .. any recommendation on how to implement that cleanly?
19:06 TTimo rcvu: green is no change, yellow means changes made afaik ?
19:07 rcvu Hmm seems weird, osmajorrelease shouldn't change
19:14 TTimo sometimes 'change' is a bit different for saltstack than the usual human meaning
19:15 TTimo that means doing some operations .. even if that might result in no actual change on the system, for saltstack it still did something
19:22 rcvu Ah okay
19:29 bachler joined #salt
19:30 exarkun joined #salt
19:33 rcvu Hmm well it seems to be breaking the {% if %} block I have
19:33 rcvu Is there any resources on what it means specifically? I can't find anything online for it
19:34 tiwula joined #salt
19:34 rcvu And it seems to be skipping the if block I have entirely because the grain is yello
19:34 rcvu (The other minion which comes back green works fine )
19:52 rojem joined #salt
20:04 dendazen joined #salt
20:25 fgimian joined #salt
20:26 mindo joined #salt
20:32 sjorge joined #salt
20:41 gdot joined #salt
20:44 mikecmpbll joined #salt
20:50 tkharju joined #salt
20:52 dendazen joined #salt
21:34 vexati0n is there anything i can do about Windows minions being so ridiculously slow to respond? I've had to increase the default timeout to 30 seconds just because it takes 3 years for Windows to load python or something
21:37 onslack <gtmanfred> no
21:37 gtmanfred the windows minion is slow because there is no `os.fork` equivalent on windows
21:38 gtmanfred so the entirety of the loaded modules have to be reloaded each time a fork happens on windows, because what windows does is pickle all of the memory to disk, and then reload it, which you can't pickle function definitions
21:38 gtmanfred linux is faster, because it just forks the memory, so all of the modules are already loaded
21:38 gtmanfred windows has to reload all modules from disk
21:58 vexati0n why would anyone write an OS like that.
21:58 vexati0n ugh
22:00 pcn They had reasons - at the time, threads were the future (see: solaris)
22:01 vexati0n solaris is the other bane of my existence.
22:02 vexati0n our prod environment is 95% decade-old solaris 10 servers
22:02 onslack <gtmanfred> hey, think of it this way…. at least you don’t have AIX
22:02 vexati0n that is truly a blessing
22:02 bluenemo joined #salt
22:03 vexati0n when is someone going to be super nice and implement a nice, clean minion renaming function that handles the keys and everything
22:03 vexati0n asking for a friend
22:11 Edgan vexati0n: I don't think you need a function for this. You actually need to do this minion and master. I could easily implement this as a salt-ssh script. Rename files in /etc/salt/pki/master, change /etc/salt/minion_id on the minion, and restart the minion
22:12 vexati0n oh i have scripts that do it
22:12 Edgan vexati0n: Unlike puppet, it looks like Salt doesn't include the hostname as part of the key. It is just keys not certs.
22:13 vexati0n but it would be nice to have a one-step module that takes care of it. the issue in my environment is the support team keeps renaming windows minions but i don't know what name they're going to use or when they're going to do it, so i end up with minion names that no longer match computer names.
22:13 bachler joined #salt
22:14 vexati0n i have a script i can run from the master that renames the minion and then re-accepts the key after its name has changed, but that's... hacky. i would rename the key, but like i said, i don't know what name they're going to use ahead of time.
22:15 vexati0n it's more of an organization problem than a salt problem, though. :P
22:15 gdot was just going to ask if that isn't one of those challenges best overcome via a non-technical solution :P
23:11 gdot2 joined #salt
23:17 dxiri_ joined #salt
23:32 dxiri joined #salt
23:49 woodtablet left #salt
23:53 noobiedubie joined #salt
23:55 dendazen joined #salt
23:59 zerocoolback joined #salt

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