Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-12-01

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

All times shown according to UTC.

Time Nick Message
00:11 amontalban joined #salt
00:15 Rubin anyone have a snipit of jinja that can give me the name of the interface that the fqdn_ip4 address is on?
00:16 Rubin like would always used to be eth0 but now system is changing things to like enp0s31f6
00:16 Rubin need a way to figure out what that is
00:17 krymzon joined #salt
00:17 Edgan Rubin: AWS, datacenter, or other?
00:17 Rubin other
00:17 Rubin mint workstation in this specific case
00:17 Rubin but my understanding is that this is a systemd related change coming eventually everwhere by default
00:18 Edgan Rubin: I am not a fan of the new naming scheme, and the people that implemented it are even changing it. Because they realized the v1 version sucks. I just go back to the old way.
00:18 Edgan Rubin: It is not hard to disable
00:18 Rubin the problem they are solving is real though
00:19 Rubin (hardware not always indexing stuff in the same order)
00:19 Edgan Rubin: only in a subset of cases, and I would say it is the majority
00:19 Edgan Rubin: I mean not the majority
00:19 Rubin rareish, yes
00:19 Edgan Rubin: It was originally intended for physical servers with 4+ NICs
00:20 Rubin either way though, theres an fqdn_ip4 for similar reasons, right.. a system might have multiple ips but you want to know the like 'one' main real IP
00:20 Rubin so the same goes for interfaces
00:20 Edgan Rubin: The bigger common issue is each wireless driver liked to use it's own naming scheme.
00:20 Rubin yeah that too
00:20 Edgan Rubin: but for wired, 99% of the time the ethX works
00:21 Edgan Rubin: I say 99%, because even in servers most people have switched to VMs
00:21 Rubin do i just have to interate over ip4_interfaces till I find the one with fqdn_ip4 in it ?
00:21 jas02 joined #salt
00:21 whytewolf personally I know what subnet i want the ip to be from and just look up by cidr
00:22 Rubin what I'm doing is adding a [keyfile] unmanaged-devices entry to /etc/NetworkManager/NetworkManager.conf to disable management of the main wired interface in the gui
00:22 Rubin so I need to know the interface name to do that in a reliable protable way
00:22 scoates joined #salt
00:22 Edgan Rubin: Why would you want to do that on a workstation?
00:23 Rubin Edgan, long story, but i manage the IPs in /etc/network/interfaces and network manager breaks nfs
00:23 Edgan Rubin: NetworkManager handling networking on a desktop/laptop is a good thing, because then it can control DNS for things like VPNs
00:23 Rubin these are managed workstations..
00:23 Rubin the users shouldnt be tweaking the network
00:23 Edgan Rubin: ah, work, not personal
00:23 Rubin yes at work
00:24 Rubin etiher way, it seems when X starts and the user logs in, network manager 'resets' the interface that causes NFS to crash and burn
00:24 Edgan Rubin: how does networkmanager break nfs?
00:24 Rubin but only with some nfs servers .. or something
00:24 Edgan Rubin: sounds like a bug to fix. Haven't had to deal with NFS home directories since before NetworkManager became common
00:25 Rubin this is a new thing in mint18.. 17 never gave me trouble
00:25 Rubin either way, i dont want the user to have the ability to interfere with the network anyway
00:25 Klas joined #salt
00:25 Rubin its a multi-user system accessable on the network etc
00:26 Rubin they dont own it, so disabling that interface in networkmanager seems the right thing to do
00:27 Rubin anyway my question is about jinja not about network interfaces :)
00:27 Rubin is there some kind of syntactic fancy way to pick out a value that itself contains a value?
00:28 Rubin such as ip4_interfaces["the one that contains this specific ip"]
00:29 Rubin im tempted to grab the 'first' one but im not sure how the order is worked out
00:31 Brew joined #salt
00:32 whytewolf Rubin: https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.network.html#salt.modules.network.ifacestartswith
00:32 nickabbey joined #salt
00:33 Rubin whytewolf, how do i run that from jinja?
00:34 Rubin salt['network.interfacestartswith']('10') ?
00:35 whytewolf {% set iface = salt.network.ifacestartswith('10.0') %} which will get you an iface
00:35 whytewolf that way also works
00:35 Rubin sweet, thanks
00:35 whytewolf I don't know if it returns a list, dict or string so might need to play with it a bit
00:35 Rubin hmm i have 2 subnets with different class A blocks
00:36 Rubin but maybe i can put the fqdn_ip4 in there
00:36 psy0rz joined #salt
00:40 amontalban joined #salt
00:40 amontalban joined #salt
00:42 jas02 joined #salt
00:43 f4 joined #salt
00:51 sarcasticadmin joined #salt
00:53 MTecknology I don't suppose anyone's seen anything like this with 2016.11.0? http://dpaste.com/0XMWEAQ   I figure probably not, but gotta ask.
00:55 BattleChicken left #salt
00:57 cyborg-one joined #salt
01:07 whytewolf i have not
01:14 Edgan MTecknology: Sounds somewhat like, https://github.com/saltstack/salt/issues/24521
01:14 saltstackbot [#24521][MERGED] State network.managed fails on Debian (Jessie) | The following state:...
01:15 jas02 joined #salt
01:16 abednarik joined #salt
01:22 jas02_ joined #salt
01:22 pcn I've got minions that I've deleted, but I seem to have mine data from them.
01:22 pcn How should I clear that out?
01:23 sarcasticadmin joined #salt
01:25 hemebond https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.mine.html?highlight=mine.update#salt.modules.mine.update clear=True?
01:27 pcn That works even when the minions have been deleted?
01:27 hemebond According to a comment on https://github.com/saltstack/salt/issues/7226, yes
01:27 saltstackbot [#7226][OPEN] Salt Mine needs a function to flush contents | Once entered into the mine via mine.send, there is no function to remove data. We need some new functions, perhaps invoked along the lines of:...
01:28 hemebond Though that issue is still open (from 2013) so I guess there are still issues with Salt caching mine data.
01:39 MTecknology whytewolf: It turns out it was a bug in the config file that the module understandably puked on
01:40 whytewolf ahhh, humm. still not the friendlest error.
01:41 Bryson joined #salt
01:42 MTecknology I wrote a two line patch that silently stops it from breaking in that situation.
01:48 MTecknology I don't know if it's a good idea to submit, probably needs to raise an exception...
01:50 netcho joined #salt
01:51 tkharju joined #salt
01:53 MTecknology whytewolf: if you're curious, in modules/debian_ip, right next to "adapters = .." I added "iface_dict = None" and then  "# Parse the detail clauses."  I changed "elif line[0].isspace():" --> "elif line[0].isspace() and iface_dict is not None:"
01:53 MTecknology It's not a clean solution and I don't want to send it upstream, but at least sharing
02:06 krymzon_ joined #salt
02:07 _KaszpiR_ joined #salt
02:20 catpigger joined #salt
02:23 jas02_ joined #salt
02:25 orionx joined #salt
02:26 evle joined #salt
02:29 orionx_ joined #salt
02:29 JPT joined #salt
02:48 ilbot3 joined #salt
02:48 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
02:52 amontalban joined #salt
02:55 impi joined #salt
03:02 _Cyclone_ joined #salt
03:06 bocaneri joined #salt
03:22 bltmiller joined #salt
03:23 sarcasticadmin joined #salt
03:24 jas02_ joined #salt
03:44 sarcasticadm joined #salt
03:45 bastiand1 joined #salt
03:47 sebastian-w joined #salt
03:49 dxiri joined #salt
03:51 cyteen_ joined #salt
03:54 rem5 joined #salt
03:58 cyteen_ joined #salt
04:01 cyteen__ joined #salt
04:02 spuder joined #salt
04:04 Aleks3Y joined #salt
04:07 cyteen__ joined #salt
04:09 cyteen_ joined #salt
04:10 DEger joined #salt
04:14 PsionTheory joined #salt
04:20 morissette joined #salt
04:24 cyteen__ joined #salt
04:25 jas02 joined #salt
04:25 dxiri joined #salt
04:27 DEger joined #salt
04:27 rdas joined #salt
04:34 cyteen__ joined #salt
04:41 scoates joined #salt
04:57 jas02 joined #salt
05:00 onlyanegg joined #salt
05:16 amontalban joined #salt
05:16 amontalban joined #salt
05:24 tedski joined #salt
05:25 DEger joined #salt
05:25 jas02 joined #salt
05:26 nickabbey joined #salt
05:29 cyteen_ joined #salt
05:42 fracklen joined #salt
05:56 Valfor joined #salt
05:56 Valfor joined #salt
05:58 jas02 joined #salt
06:00 krymzon joined #salt
06:01 VR-Jack2-H joined #salt
06:02 netcho joined #salt
06:03 ivanjaros joined #salt
06:05 impi joined #salt
06:06 netcho_ joined #salt
06:09 fracklen joined #salt
06:14 ashokrajar joined #salt
06:15 haam3r1 left #salt
06:21 nickabbey joined #salt
06:25 nidr0x joined #salt
06:26 jas02 joined #salt
06:36 felskrone joined #salt
06:36 xenoxaos joined #salt
06:42 J0hnSteel joined #salt
07:05 colttt joined #salt
07:07 DEger joined #salt
07:17 amontalban joined #salt
07:17 amontalban joined #salt
07:19 DEger joined #salt
07:20 ivanjaros joined #salt
07:27 ivanjaros3916 joined #salt
07:27 jas02 joined #salt
07:35 fracklen joined #salt
07:45 impi joined #salt
07:47 funabashi joined #salt
07:58 ReV013 joined #salt
07:59 jas02 joined #salt
08:01 tobiasBora joined #salt
08:08 jas02 joined #salt
08:09 hojgaard joined #salt
08:09 nickabbey joined #salt
08:09 dariusjs joined #salt
08:10 o1e9 joined #salt
08:10 hojgaard Hi. Is it possible to use the facter grain without puppet? I dont want to mess with salt code, but facter is run by default with -p it seems and then it will return an error: [CRITICAL] Failed to load json facter data
08:10 hojgaard is there a way to fix this properly?
08:11 keimlink joined #salt
08:12 o1e9 joined #salt
08:16 theblazehen joined #salt
08:16 fracklen joined #salt
08:17 fracklen joined #salt
08:19 DEger joined #salt
08:20 mortis left #salt
08:25 _KaszpiR_ joined #salt
08:26 k_sze[work] joined #salt
08:26 eseyman joined #salt
08:28 jas02_ joined #salt
08:28 impi joined #salt
08:29 carmony joined #salt
08:30 ws2k3 joined #salt
08:30 jerrcs joined #salt
08:31 jab416171 joined #salt
08:32 NV joined #salt
08:33 samodid joined #salt
08:35 garthk joined #salt
08:36 Hipikat joined #salt
08:36 v0rtex joined #salt
08:36 jab416171 joined #salt
08:37 degorenko joined #salt
08:37 duckfez joined #salt
08:45 ronnix joined #salt
08:47 rdas joined #salt
08:49 raspado joined #salt
08:56 Rubin_ joined #salt
08:56 krymzon joined #salt
08:57 rdas joined #salt
08:57 ernescz joined #salt
08:58 hemebond hojgaard: Submit a PR that does... soemthing.
08:59 PhilA__ joined #salt
09:04 Guest54209 joined #salt
09:04 mikecmpbll joined #salt
09:05 aarontc joined #salt
09:06 geomacy joined #salt
09:12 dariusjs joined #salt
09:13 bdrung_work joined #salt
09:15 Rumbles joined #salt
09:19 amontalban joined #salt
09:19 amontalban joined #salt
09:21 www-bukolay-net joined #salt
09:23 yidhra joined #salt
09:25 s_kunk joined #salt
09:26 dunz0r I'm using file.recurse to copy an entire directory-structure, but it doesn't seem to copy in new content that I add to the directory, am I missing something?
09:26 samodid joined #salt
09:28 dunz0r The state looks like this: http://sprunge.us/PfQC?yaml . packagerepo/packages originally contained just one file in one subdirectory, but now it contains two directories and one file in each of those
09:28 dunz0r Directory layout like this:
09:28 dunz0r http://sprunge.us/HYES
09:28 dunz0r They're of course both added to git
09:30 sjorge joined #salt
09:30 sjorge joined #salt
09:32 orionx joined #salt
09:32 keimlink joined #salt
09:36 hemebond Your states and files are in a Git repo?
09:36 hemebond So you're using GitFS?
09:36 dunz0r hemebond: Yup.
09:36 N-Mi_ joined #salt
09:39 orionx joined #salt
09:39 sjorge joined #salt
09:39 sjorge joined #salt
09:42 smcquay joined #salt
09:55 sjorge joined #salt
09:55 sjorge joined #salt
09:58 Antiarc joined #salt
10:04 brokensyntax_ joined #salt
10:04 djinni` joined #salt
10:04 rofl____ joined #salt
10:13 fredvd joined #salt
10:13 onlyanegg joined #salt
10:13 sjorge joined #salt
10:13 sjorge joined #salt
10:16 dariusjs joined #salt
10:17 N-Mi_ joined #salt
10:18 hojgaard Hi. Is it possible to use the facter grain without puppet? I dont want to mess with salt code, but facter is run by default with -p it seems and then it will return an error: [CRITICAL] Failed to load json facter data. How can this be fixed?
10:23 xet7 joined #salt
10:29 losh joined #salt
10:31 jas02_ joined #salt
10:31 __alex_ joined #salt
10:32 teclator_ joined #salt
10:33 ThomasJ joined #salt
10:34 Dr_Jazz joined #salt
10:35 dtsar joined #salt
10:36 iggy_ joined #salt
10:36 dober- joined #salt
10:38 majuscul1 joined #salt
10:38 babilen_ joined #salt
10:38 mirko hey, i guess i found an interesting bug: salt-ssh -l debug '$$HOST$$' -r "gpg --homedir /tmp/foobar"
10:39 mirko it should throw an error about /tmp/foobar not existing
10:39 mirko but i think the command interferes with the ssh connection
10:40 orionx joined #salt
10:40 mirko a very different result produces e.g.: salt-ssh -l debug '$$HOST$$' -r "gpg --homedir /root/.gnupg"
10:40 mirko which gives me "Permission denied for host jitsi.nanl.de, do you want to deploy the salt-ssh key? (password required):
10:40 mirko [Y/n]" - which is only caused by the different shell command
10:43 djinni` joined #salt
10:43 7GHAAUHJH joined #salt
10:43 rofl____ joined #salt
10:43 NeoXiD joined #salt
10:43 georgemarshall joined #salt
10:43 SWA joined #salt
10:43 jeddi joined #salt
10:43 d3c4f joined #salt
10:43 nahkiss joined #salt
10:43 pcn joined #salt
10:43 TOoSmOotH joined #salt
10:43 stotch_ joined #salt
10:43 ropes joined #salt
10:43 evidence joined #salt
10:44 djinni` joined #salt
10:45 ReV013 joined #salt
10:47 ernescz Hello! Does anyone know is it possible to pass grains information when calling runner modules (cloud)? https://docs.saltstack.com/en/latest/topics/reactor/#calling-runner-modules-and-wheel-modules
10:47 ernescz like so - https://pastebin.ubuntu.com/23562541/
10:47 cyteen_ joined #salt
10:48 ernescz can't get it to work, sadly
10:51 hemebond Wouldn't you put grains in the profiles?
10:52 hemebond I can't see that function accepting a full dict of stuff for each instance. I believe it just takes a list of names.
10:53 ernescz hemebond: I could put it in profiles but that would make the profiles even longer..
10:53 ernescz for now putting general grains in profiles is fine, and extending them per-instance specific is done in maps file and it works fine.
10:56 Dr_Jazz thats what I am doing
10:58 ernescz And it works great if called from cli with 'salt-cloud -m /my/map.cnf'. How to pass those grains to runner is another story :)
11:07 Dr_Jazz yeah, i've wresttled with it a couple of times as well. and ended by setting it in maps
11:09 ernescz well, I think I'll just set two similar profiles and then based on those grains have a state to set all the rest of the required grains.
11:09 dRiN joined #salt
11:10 babilen_ grain abuse!
11:20 cyteen_ joined #salt
11:21 fracklen joined #salt
11:21 amontalban joined #salt
11:21 abednarik joined #salt
11:24 HRH_H_Crab joined #salt
11:31 jas02_ joined #salt
11:37 pipps joined #salt
11:39 dariusjs joined #salt
11:40 orionx joined #salt
11:47 oida joined #salt
11:53 ernescz babilen: care to explain, please?
11:59 fracklen joined #salt
12:06 DEger joined #salt
12:11 we999 joined #salt
12:11 PhilA__ joined #salt
12:12 cyteen_ joined #salt
12:20 cyteen_ joined #salt
12:22 dariusjs joined #salt
12:22 amcorreia joined #salt
12:27 amontalban joined #salt
12:29 aidin joined #salt
12:32 ronnix joined #salt
12:32 jas02_ joined #salt
12:33 dariusjs joined #salt
12:41 orionx joined #salt
12:42 irctc077 joined #salt
12:43 irctc077 Hi i have 2000 minions and a lot of them are considered down and habe in their logfiles the info that they cannot return the job informations to the master. But the mater is running, has enough free cpu power and has already 64 workers. I also use -b 50 in the salt call.
12:48 babilen irctc077: You might want to call that asynchronously
12:51 irctc077 babilen: how todo that?
12:52 babilen You'd pass "--async"
12:52 fracklen joined #salt
12:53 cyteen joined #salt
12:53 irctc077 babilen: how does this help? or why do the minions not time out then?
12:54 Miouge joined #salt
12:54 babilen irctc077: It relieves the master
12:54 Miouge I can’t find info about this in the doc’ so here it is: can I define a nodegroup by  targeting another nodegroup?
12:55 babilen Miouge: https://docs.saltstack.com/en/latest/topics/targeting/nodegroups.html (first example, group3)
12:55 gtmanfred only after 2015.8.0
12:55 gtmanfred Note
12:55 gtmanfred Nodegroups can reference other nodegroups as seen in group3. Ensure that you do not have circular references. Circular references will be detected and cause partial expansion with a logged error message.
12:56 Miouge Oh, I need new glasses :)
12:57 irctc077 babilen: but how do i know which one failed in that case?
12:58 gtmanfred irctc077: you can lookup the job once it is finished in the job cache
12:58 gtmanfred salt-run jobs.lookup_jid <jid>
12:58 babilen irctc077: You take a look at the job result -- https://docs.saltstack.com/en/latest/topics/jobs/
13:00 irctc077 babilen: thanks makes sense. But -b 50 deviced my 2000 machines into 40 jobs. So i have to query them all?
13:01 babilen You'd have to query per job id
13:02 babilen If you get multiple IDs you have to perform multiple lookups, if not then not
13:05 gtmanfred i don't think you can batch and do async at the same time
13:06 babilen I've never done it (just call --async)
13:06 gtmanfred yeah, doesn't look like you can
13:10 aidin joined #salt
13:10 numkem joined #salt
13:19 mirko bug report for the issue i described before: https://github.com/saltstack/salt/issues/38016
13:19 saltstackbot [#38016][OPEN] salt-ssh -r '$COMMAND' - $COMMAND seems to interfere with SSH connection | Passed shell command to `salt-ssh -r` seems to interfere with the SSH connection....
13:23 gtmanfred thanks mirko we will get to testing the bug and commenting later today hopefully
13:26 mirko gtmanfred: oh, thanks! i'm curious about the root cause (especially whether it's me.. :))
13:27 gtmanfred ch3ll should get to it when she gets on today, she is working from the newest to oldest, and I am doing the opposite this morning
13:27 gtmanfred trying to catch up from the thanksgiving weekend
13:29 mirko ah, right, i remember her commenting on most of my tickets very promptly in the past
13:32 ProT-0-TypE joined #salt
13:32 irctc077 babilen: if i call salt -t 60 -b 150 --static --state-output=mixed --state-verbose=False --async '*' state.highstate i get no job ids at all...
13:32 babilen irctc077: I was under the impression that you cannot use async in batch mode
13:33 gtmanfred when i tested it, it just ignores --async if you have -b
13:33 jas02_ joined #salt
13:34 irctc077 babilen: ah ok but i need the batch mode i do not want to update all of them to keep all services up and running all the time.
13:34 babilen Target them differently?
13:41 HRH_H_Crab im trying to get my head around proxy minions (specifically for junos devices) and having a fair bit of difficulty working out where things are going wrong.
13:41 HRH_H_Crab when i start my proxy and accept the key on the master things seem to go well
13:41 HRH_H_Crab but as soon as i go to test it the proxy crashes out.
13:42 orionx joined #salt
13:42 HRH_H_Crab if anyone can provide me with some advice on how to troubleshoot this id be extremely grateful!
13:49 lilvim joined #salt
13:49 Satyajit joined #salt
13:51 fracklen joined #salt
13:54 fracklen joined #salt
14:01 ssplatt joined #salt
14:03 fracklen joined #salt
14:09 nickabbey joined #salt
14:10 nickabbey joined #salt
14:12 _Cyclone_ joined #salt
14:17 racooper joined #salt
14:20 rem5 joined #salt
14:20 fracklen joined #salt
14:32 ronnix joined #salt
14:34 yuhlw joined #salt
14:34 _Cyclone_ joined #salt
14:37 Tanta joined #salt
14:47 dariusjs joined #salt
14:49 XenophonF Does anyone use Salt to deploy software to Macs?
14:50 spuder joined #salt
14:58 dxiri joined #salt
15:02 Miouge Not possible to do N@group1 in a compound match?
15:04 braqoon joined #salt
15:05 Rumbles having some real issues reaching the new salt repo from China, the ppa's still work fine, but reaching the new repo is very unreliable... I'm guessing that the ppa's are served from local regions better than the saltstack official repo...
15:05 braqoon hi Guys, does salt have some mirrors repo in Asia ? i got lot's of time outs on current ones
15:09 dxiri joined #salt
15:11 ReV013 left #salt
15:11 mpanetta joined #salt
15:12 babilen Miouge: It's the other way round: You need a compound matcher if you want to use N@group1
15:13 Heartsbane joined #salt
15:18 Miouge “salt -N group1 test.ping” works, but not “salt -C N@group1 test.ping” moreover the N letter is not in the coumpound matchers page (https://docs.saltstack.com/en/latest/topics/targeting/compound.html#targeting-compound)
15:30 spuder joined #salt
15:33 LeProvokateur joined #salt
15:34 _Cyclone_ joined #salt
15:34 debian112 joined #salt
15:35 jas02_ joined #salt
15:37 mpanetta joined #salt
15:40 Rumbles I guess no one uses the salt repo in China then? :)
15:42 XenophonF i'm not sure the ppa repos are up to date
15:42 Rumbles they're not
15:42 Rumbles the ppa's haven't been updated since march
15:42 Rumbles but the repo just isn't reachable during most of the day in China
15:43 Rumbles seemed okay yesterday evening GMT, but half n hour ago everything timed out
15:43 cscf Rumbles, do you have multiple servers?  Perhaps you should mirror locally yourself
15:43 Rumbles we have 2 servers, which will be run as hypervisors
15:44 Rumbles but storing the whole salt repo isn't ideal :)
15:44 braqoon cscf, this is a bit overkill just for one repo with two packages
15:44 Sketch just mirror the version you need
15:44 Sketch it's not very big
15:44 cscf there's also apt-cacher-ng
15:44 Sketch 26M     salt-2016.3-el7-x86_64/
15:44 XenophonF Rumbles: what operating systems are you supporting?
15:44 Rumbles something I've never done...
15:44 Sketch for  https://repo.saltstack.com/yum/redhat/7/x86_64/2016.3/
15:45 cscf Though I forget if it still reaches out of metadata
15:45 cscf Rumbles, really, if you're running a bunch of VM's, they should all be using a package cache of some sort.
15:45 braqoon XenophonF, Ubuntu
15:45 cscf To run that many machines off the real mirrors is wasteful and impolite
15:45 Rumbles yeah sorry, me and braqoon are working together :)
15:46 XenophonF ah well if RHEL/CentOS/FreeBSD/Windows, I was going to suggest some combo of spacewalk/poudriere/etc.
15:46 braqoon cscf, i think that's your point of view, but point taken. Probably need to look into local repo for this.
15:47 Rumbles do you expect everyone to have their own mirrors?
15:47 Rumbles I've never heard that before :/
15:47 cscf If you're running like 50 VMs, syncing them all from the mostly-volunteer distro repos is rude
15:47 cscf And it takes forever anyway
15:48 Sketch yep, local repos are much quicker
15:48 Rumbles well, currently we're mainly using the ppa
15:48 cscf apt-cacher-ng or squid-deb-proxy are much easier to set up than a real repo, though.
15:48 braqoon cscf, will check it
15:48 cscf Most of my servers use our local repos, at gigabit.
15:49 Sketch some of my servers don't have internet access, so i have to have a local repo anyway
15:49 cscf And at home I'm using squid-deb-proxy
15:49 cscf Sketch, yeah, that's true of some here too
15:50 cscf squid-deb-proxy is nice because it doesn't break if the cache goes away, so that's why I use it at home, for laptops, etc
15:50 Brew joined #salt
15:51 nZac joined #salt
15:58 pcdummy I use apt-cacher-ng
15:58 pcdummy Need a formula for it?
15:59 Pa1adin joined #salt
16:00 Pa1adin hello world
16:01 pcdummy hi Pa1adin
16:01 sarcasticadmin joined #salt
16:08 fracklen joined #salt
16:09 _Cyclone_ joined #salt
16:10 cscf pcdummy, there's an official formula, irrc
16:10 cscf iirc*
16:13 bltmiller joined #salt
16:15 _JZ_ joined #salt
16:15 amontalb1n joined #salt
16:16 Miouge Ok, I answer my own question “The N@ classifier cannot be used in compound matches within the CLI or top file, it is only recognized in the nodegroups master config file parameter.”
16:18 jas02 joined #salt
16:18 orionx joined #salt
16:21 onlyanegg joined #salt
16:22 dxiri joined #salt
16:22 ronnix joined #salt
16:23 dxiri joined #salt
16:27 pipps joined #salt
16:29 nicksloan joined #salt
16:34 cyborg-one joined #salt
16:36 jas02_ joined #salt
16:36 fracklen_ joined #salt
16:39 hasues joined #salt
16:39 hasues left #salt
16:45 Miouge joined #salt
16:46 Reverend joined #salt
16:46 Reverend oh god.
16:46 Reverend oh sweet mother of god.
16:46 Reverend hello chaps
16:46 Reverend i've been in #saltstack for the last few days wondering why everyone went dark
16:47 fracklen joined #salt
16:47 cscf Reverend, lol welcome back
16:48 Reverend thanks cscf - I've been raging at my desk like "nobody iun the salt channel is fucking here!
16:48 Reverend turns out I was jsut derping
16:48 * Reverend lolz
16:49 ramblinpeck left #salt
16:49 ericprice joined #salt
16:55 gtmanfred interesting, i had no idea that channel existed
16:58 samodid joined #salt
17:00 DEger joined #salt
17:00 anotherzero joined #salt
17:01 orionx_ joined #salt
17:02 Reverend gtmanfred: me either, until just now
17:02 Reverend gtmanfred: i got a little lost on my way back after a windows restart
17:02 Reverend ooo - centos got the carbon update.
17:02 Reverend win
17:04 orionx__ joined #salt
17:07 jas02 joined #salt
17:10 gtmanfred yup, all of them should have it now
17:11 Reverend noiiice
17:11 Reverend not sure on patch notes
17:12 Reverend but it didn't break ote
17:12 Reverend so... im pleased
17:12 gtmanfred yeah, we fucked up on the release notes this time...
17:12 gtmanfred didn't get all the things in there at all
17:13 sagerdearia joined #salt
17:14 ericprice joined #salt
17:23 Reverend gtmanfred: well, no problem. I've not read them... so no harm done :) just don't fuck my infrastructure yeah? :P
17:24 gtmanfred :)
17:28 MTecknology so many modules/debian_ip regressions in 2016.11.0!!! :'(
17:29 MTecknology it looks like it may actually be stripping out the iface lines that it then later gets upset about not being there because ... I ranted that enough last night.
17:29 jhauser joined #salt
17:36 Salander27 joined #salt
17:37 jas02_ joined #salt
17:41 nicksloan joined #salt
17:42 Edgan joined #salt
17:43 orionx joined #salt
17:46 pcdummy MTecknology: i only have one bug
17:46 MTecknology what's that?
17:47 pcdummy MTecknology: https://github.com/saltstack/salt/issues/37942
17:47 saltstackbot [#37942][OPEN] Regression with debian network management and the loopback interface. | Description of Issue/Question...
17:51 orionx joined #salt
17:51 armin joined #salt
17:55 jas02 joined #salt
17:57 wwalker yesterday, I ran a state with pkg.latest: name librados2   and it (apt-get) UNINSTALLED an application that depended on the older library version that was already installed.  I can't find anything in pkg.installed, or pkg.latest to prevent this (assinine) behavior of apt-get.  Any recommendations?
17:58 scsinutz joined #salt
17:59 s_kunk joined #salt
18:00 Reverend well how are you going to keep on 'latest' without updating your packages?
18:00 Reverend I'd use package installed and specify the version
18:00 iggy wwalker: fix your packages
18:00 UtahDave joined #salt
18:00 Reverend also - what iggy said.,
18:02 spuder_ joined #salt
18:05 KajiMaster joined #salt
18:05 UtahDave pcdummy: Hey, man.  I haven't had time to thoroughly review your formula, but I think it's definitely good enough for a PR and discussion/review there.
18:06 pcdummy UtahDave: will do so, thanks.
18:06 UtahDave anytime!  Thanks for putting the time and effort into that code. It will be a great addition
18:06 demize joined #salt
18:07 mikecmpbll joined #salt
18:09 Bryson joined #salt
18:10 fuhnoo joined #salt
18:11 gtmanfred !stats
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot Total Words: 27883
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot Total Words: 27885
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot Total Words: 27887
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot ...
18:11 saltstackbot Total Words: 27888
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot Total Words: 27890
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot Total Words: 27891
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot was kicked by gtmanfred: gt...fo. ♥, gtmf
18:11 saltstackbot joined #salt
18:11 saltstackbot ...
18:11 saltstackbot Total Words: 27891
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot Total Words: 27891
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 pcdummy :)
18:11 saltstackbot ...
18:11 saltstackbot Total Words: 27889
18:11 saltstackbot Stats for gtmanfred in #salt
18:11 gtmanfred lol
18:12 gtmanfred so that happened
18:12 wwalker iggy: the packages aren't broken.  apt-get is.  it should have refused to uninstall the app (and therefore refused to upgrade the library).
18:12 MTecknology oopsies :P
18:12 gtmanfred !restart
18:12 gtmanfred !quit
18:12 gtmanfred hrm...
18:14 gtmanfred !salt runner.pillar.show_pillars
18:14 gtmanfred !salt states.cmd.run
18:14 saltstackbot https://docs.saltstack.com/en/latest/ref/states/all/salt.states.cmd.html#salt.states.cmd.run
18:14 MTecknology bots suck :(  (to write, maintain, and fix bugs in)
18:15 gtmanfred !salt runners.pillar.show_pillars
18:19 infrmnt joined #salt
18:22 impi joined #salt
18:23 iggy it only understands states and modules
18:23 iggy I can add runners or whatever
18:24 iggy and if people think it isn't useful, I can get rid of it
18:24 gtmanfred i like it :)
18:24 jas02 joined #salt
18:25 gtmanfred but that !stats one seems like a bit of a bug
18:25 iggy yeah, I'll look into it (it's sopel, so if anybody else wants to do it...)
18:26 gtmanfred cool
18:33 jas02 joined #salt
18:34 fuhnoo hi all... question about using a signed public key from the master on a minion
18:35 fuhnoo on the master, i added the following lines in /etc/salt/master:
18:35 fuhnoo master_sign_pubkey: True
18:35 fuhnoo master_use_pubkey_signature: True
18:35 fuhnoo master_pubkey_signature: master_name.pub
18:36 fuhnoo then i ran `salt-key --gen-signature --auto-create` which generated all the keys i needed. i copied master_name.pub to my minion in /etc/salt/pki/minion and added the following lines to /etc/salt/minion:
18:36 fuhnoo verify_master_pubkey_sign: True
18:36 fuhnoo master_sign_key_name: master_name
18:36 fracklen joined #salt
18:36 fuhnoo when i start the minion in debug mode, i get "Exception: RSA key format is not supported"
18:36 fuhnoo any ideas? sorry for the wall of text
18:37 writteno1 joined #salt
18:37 jas02_ joined #salt
18:37 o1e9 joined #salt
18:42 cscf fuhnoo, master and minion versions?
18:43 fuhnoo both are 2016.3.3 (Boron)
18:48 Trauma joined #salt
18:48 jbellone joined #salt
18:48 jbellone Hi, anyone familiar with Salt Enterprise topologies mind answering few questions? We're trying to figure out if we are unable to use Salt Open with the eAPI.
18:49 Miouge joined #salt
18:49 gtmanfred jbellone: The salt enterprise api talks to salt open
18:50 pipps joined #salt
18:50 jbellone That means that we can use syndics as well, right?
18:50 gtmanfred yes
18:50 gtmanfred and multimaster
18:51 gtmanfred it should just be a fileserver plugin and a couple other things that get plugged into the master to make it push upwards to the enterprise api
18:51 jbellone So any reason why I would ever use the enterprise master?
18:52 gtmanfred For more granular rbac control
18:52 jbellone And the enterprise master cannot talk to syndics?
18:52 gtmanfred the enterprise master talks to the master of the syndics
18:52 freezey joined #salt
18:53 jbellone (I am asking this because documentation is pretty sparse on enterprise, apologies if I seem dense.)
18:55 lws joined #salt
18:55 gtmanfred gimme a minute
18:55 CampusD joined #salt
18:56 freezey k
18:56 freezey curious as well
18:56 freezey emasters do not connect to syndics
18:56 ivanjaros joined #salt
18:57 gtmanfred the connect to the master of the syndics, and then that master connects to the syndics is how i understand it should work
18:57 UtahDave jbellone: your masters talk directly to the enterprise api server.  It doesn't matter if syndics are talking to your masters
18:57 whiteinge joined #salt
18:57 freezey fun
18:58 freezey @UtahDave: the syndics will goto the masters and the eapi talks to the masters?
18:58 freezey the emasters can only talk to the eapi?
18:59 CampusD hi all, quick question, has anybody experienced certain files not being picked up by the master from the backend gitfs?
19:00 freezey Dave would you have a few minutes to go over some topology questions? We are looking to purchase the eapi stuff.
19:00 CampusD I can clone the repo from multiple locations and I see the files that are missing from the master
19:01 CampusD these missing files have something in common, they have no file extensions, if that matters
19:01 CampusD are they being filtered out?
19:01 UtahDave freezey: Yeah for sure.
19:01 gtmanfred CampusD: how are you trying to reference those files?
19:01 oida joined #salt
19:01 UtahDave freezey: do you want to pm me?
19:01 freezey sure
19:02 CampusD gtmanfred, I am using file.recurse to put in the the /etc/logrotate.d/ directory of systems
19:02 CampusD *to put them
19:02 gtmanfred hrm, that should be fine... i can't think of a reason that would fail because of no extension
19:03 CampusD this is a small development environment running with a few minions, the master and a git file server
19:04 CampusD is there some kind of cache for gitfs backends?
19:04 gtmanfred there is, but it should refresh every 60 seconds/on the loop_interval
19:05 aarontc joined #salt
19:05 gtmanfred you could run `salt-run fileserver.update backend=git`
19:05 gtmanfred or `salt-run fileserver.clear_cache backend=git`
19:09 austin_ joined #salt
19:10 CampusD ok, I might have found the issue, so files won't get copied to the minions if say there was a misspelling on the state file that has the path, right?
19:10 stupidnic How would I copy a file to a minion and then delete it in the same state (it's a temp, short lived file)? I tried to set "order" on my states, but they don't seem to be respected.
19:14 beowuff joined #salt
19:14 ronnix joined #salt
19:22 CampusD thanks gtmanfred, clearing the cache and fixing the spelling solved the issue
19:22 gtmanfred cool
19:27 Eugene joined #salt
19:33 nicksloan joined #salt
19:38 dendazen joined #salt
19:39 jas02_ joined #salt
19:40 cro joined #salt
19:46 armin joined #salt
19:48 armin joined #salt
19:50 cro joined #salt
19:51 gtmanfred joined #salt
19:54 dRiN joined #salt
20:02 _Cyclone_ joined #salt
20:07 jas02 joined #salt
20:08 pcdummy joined #salt
20:08 numkem joined #salt
20:13 Rumbles joined #salt
20:13 jas02_ joined #salt
20:14 jhauser joined #salt
20:15 fuhnoo joined #salt
20:18 Trauma joined #salt
20:22 jas02_ joined #salt
20:32 jas02_ joined #salt
20:33 impi joined #salt
20:34 armin joined #salt
20:34 scsinutz joined #salt
20:35 snergster joined #salt
20:37 onlyaneg1 joined #salt
20:37 ivanjaros3916 joined #salt
20:38 irctc187 joined #salt
20:38 irctc187 question.. how do you access modules from jinja templates?
20:39 irctc187 ie.. for pillars you can do salt['pillar.get']
20:39 hemebond salt['module.fun']()
20:39 irctc187 thanks
20:39 irctc187 :)
20:42 jas02_ joined #salt
20:43 irctc260 joined #salt
20:43 irctc260 hmm... so I'm trying to access network.get_hostname
20:44 irctc260 would it be salt['module.network']('get_hostname')
20:44 irctc260 ?
20:44 irctc260 I must be doing something wrong
20:44 cmarzullo where are you trying to acceess that module from?
20:44 hemebond No, it would be salt['network.get_hostname']
20:44 cmarzullo inside a state file?
20:45 irctc260 jinja template
20:45 irctc260 ie... {{ salt['network.get_hostname'] }}
20:46 irctc260 I guess I just need to pass it in
20:46 hemebond ?
20:46 Neighbour yep, that works in any jinja templated file
20:47 Neighbour be it a state, pillar or file.managed-file with template:jinja
20:48 armin joined #salt
20:49 irctc260 inside the jinja, I have {{ salt['network.get_hostname'] }}
20:49 irctc260 and here is what gets rendered:
20:49 irctc260 name = <function get_hostname at 0x7fa47c6005
20:49 hemebond No, you have to call it.
20:50 hemebond {{ salt['network.get_hostname'](args go in here) }}
20:50 sjorge joined #salt
20:50 sjorge joined #salt
20:50 irctc260 hemebond: by default ther are no args for it though
20:51 hemebond Then just ()
20:52 irctc260 there, looks good.. thanks!
20:52 hemebond ????
20:58 sh123124213 joined #salt
20:58 sh123124213 how can I forward events from syndic to all masters ?
21:01 Miouge joined #salt
21:02 jas02_ joined #salt
21:06 pcn Is anyone aware of whether this is a known issue that an upgrade can  fix?  trying to find this in github issues isn't working for me at the moment: https://gist.github.com/pcn/b10189bf7bfb4fa8cbaff5e599e62816
21:07 Miouge joined #salt
21:08 freezey joined #salt
21:16 UtahDave pcn: I haven't seen that problem before, so I'm not sure.  Can you try on a test system?
21:17 UtahDave sh123124213: what do you mean by "all masters"?
21:18 sh123124213 UtahDave: 2 masters in order_masters: True, 2 syndics. I want all commands I run from any master to return to both masters event bus
21:18 pcn UtahDave: try what on a test system?
21:18 UtahDave try upgrading
21:18 UtahDave see if you still have the proble
21:18 UtahDave problem
21:20 UtahDave sh123124213: Have you tried pointing the syndics to both top level masters?  I'm not sure if we support that use case explicitly
21:21 jhauser joined #salt
21:22 jas02_ joined #salt
21:22 edrocks joined #salt
21:29 pcn OK, thanks
21:31 spuder joined #salt
21:36 thekabal joined #salt
21:37 sh123124213 UtahDave: its already pointing to them. problem is that syndic forwards events to random master
21:38 thekabal I have multiple syndics & one master_of_masters, and after upgrading, *one* of my syndics is giving this error in the syndic log: Unable to call _return_pub_multi on (master_master_hostname) , that syndic is not connected
21:38 sh123124213 I run command in master01 and it returns to master02
21:39 thekabal And the odd part is that the error in the SYNDIC log starts with [salt.minion/ERROR], not [salt.syndic/ERROR]
21:39 thekabal Later, it also gives [salt.minion/CRITICAL] Unable to call _fire_master on any masters! , despite it not being a multi-master environment.
21:39 barry_ joined #salt
21:46 sjorge joined #salt
21:46 sjorge joined #salt
21:46 bluenemo joined #salt
21:46 fracklen joined #salt
21:49 wayne__ joined #salt
21:59 bluenemo where can i find packages for x86 debian 8? I'm looking for something more or less recent.. debian 8 ships 2014.1 by default.. 2015.3 upwords would be lovely
22:00 schemanic joined #salt
22:00 schemanic joined #salt
22:01 stupidnic blue: x86? not amd64?
22:01 sh123124213 UtahDave: which part of the code deals with sending events to masters ?
22:02 UtahDave bluenemo: looks like repo.saltstack.com only provides amd64 and armhf for Debian 8
22:02 bluenemo yes
22:03 bluenemo hm https://github.com/saltstack/salt/issues/27228#issuecomment-144551132
22:03 saltstackbot [#27228][OPEN] salt-minion not installable on debian jessie x86 | I'm getting this...
22:03 bluenemo I dont get this - whats this then http://debian.saltstack.com/debian/dists/jessie-saltstack/main/binary-i386/ (2015.3 stuff)
22:04 bluenemo hmpf. thats what happens when you let the customer install the machine. who on earth installs x86 anyway...
22:04 stupidnic 3.5GB of RAM is all you will ever need
22:04 UtahDave bluenemo: Mm.  that's old
22:04 bluenemo its newer than 2014.1, which is shipped by debian 8 by default :D
22:05 KajiMaster joined #salt
22:05 bluenemo but yes still it wont run my custom stuff.. bad luck for me today ;)
22:05 stupidnic Debian isn't known for bleeding edge packages
22:05 dxiri joined #salt
22:07 bluenemo hehe yes it isnt ;)
22:09 iggy install via pip?
22:10 fracklen joined #salt
22:10 voileux joined #salt
22:11 bluenemo good idea
22:16 bluenemo doesnt seem to be as simple as pip install salt-minion.. hm. not sure if I want to go that way- I will suffer under x86 later on anyway
22:16 bluenemo I will request a new box
22:17 bluenemo I like when all boxes run similar stuff ;)
22:18 iggy I think you install salt (which will do a complete install) and then only run the minion
22:19 iggy but yeah, making it like the rest of your cattle is a good idea
22:19 nkuttler bluenemo: it's in backports
22:20 tercenya joined #salt
22:22 babilen bluenemo: jessie-backports has 2016.3.3+ds-3~bpo8+1
22:23 babilen stupidnic: The whole point of Debian stable is that it is, well, stable. "stable" in this context means "not changing" so you do not get newer versions by definition.
22:24 babilen If you don't want it to be stable you'd install packages from backports or track testing .. but then, who would want that on production boxes?
22:26 orionx_ joined #salt
22:29 fracklen joined #salt
22:29 scsinutz joined #salt
22:32 eseyman joined #salt
22:32 Vaelatern Without any other setup than intalling salt, can I run a salt-call command and apply a statefile I have in a local directory?
22:32 Vaelatern salt-call --local seems to not be what I want
22:32 jas02_ joined #salt
22:34 nidr0x joined #salt
22:34 fracklen joined #salt
22:36 Vaelatern Hmm, --file-root=./ seems promising
22:38 jbellone joined #salt
22:39 eseyman joined #salt
22:43 J0hnSteel joined #salt
22:53 UtahDave Vaelatern: yeah, you can.  look in the docs for masterless salt
22:54 J0hnSteel joined #salt
22:54 fuhnoo joined #salt
22:55 fracklen joined #salt
22:56 sp0097 joined #salt
22:59 scsinutz joined #salt
23:02 lws joined #salt
23:16 jbellone joined #salt
23:16 dendazen joined #salt
23:17 J0hnSteel joined #salt
23:21 jbellone joined #salt
23:24 jbellone_ joined #salt
23:29 jbellone joined #salt
23:32 jbellone_ joined #salt
23:34 jbellone joined #salt
23:34 jas02_ joined #salt
23:37 jbellone1 joined #salt
23:39 jbellone joined #salt
23:41 cliluw joined #salt
23:49 cliluw joined #salt
23:50 jbellone joined #salt
23:56 fracklen joined #salt

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