Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2017-04-26

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

All times shown according to UTC.

Time Nick Message
00:01 alexlist joined #salt
00:05 woodtablet left #salt
00:10 hemebond joined #salt
00:12 Dpu84 joined #salt
00:14 Dpu84 Hi Guys, Anyone know how to pass pillar values in pepper runner
00:14 Dpu84 pepper --client runner state.orchestrate orch.install pillar="{tgt: 'mylabmachine'}"   is not working
00:16 rich87007 joined #salt
00:22 KevinAn2757 joined #salt
00:29 mosen joined #salt
00:29 KevinAn2757 joined #salt
00:31 KevinAn2757 joined #salt
00:34 tercenya joined #salt
00:38 scoates joined #salt
00:39 pcn If I have a cronjob configured, and I use 'cron.rm' on that job, why might I still see that job in the user's crontab?
00:39 hemebond pcn: Was the cron job created by Salt?
00:40 pcn Yeah, it has the identifier in the crontabl
00:40 hemebond Then I don't think Salt can remove it. Salt uses comments in the cron file to identify different cron definitions.
00:45 afics joined #salt
00:45 ntropy Dpu84: not used peper, but that looks like it should work, have you already looked at the output of trace or debug level logs?
00:45 onlyanegg Can the orchestrate runner save state? Like if it gets cut off or something, can it restart where it left off?
00:45 pcn homebond: I'm not sure I follow your logic.  Yes, the jobs have the line prefixed with "# SALT_CRON_IDENTIFIER:" and followed by the IDs of the jobs.  How does that lead to "salt can't remove it"
00:45 gmoro joined #salt
00:45 armin joined #salt
00:45 pcn Dpu84: I'm not familiar with the orchestrate runner.  Does it normally have access to pillars?  The runners normally run in the context of the saltmaster don't they?  The pillar data shouldn't be there.
00:45 aarontc joined #salt
00:45 hemebond pcn: Well, it matches the `name` in your state for removing the cron line to the identifier in the cron file.
00:46 whytewolf pcn: you can pass "pillar data" to orchestration that works on the master level. basicly it is used for passing arguments into orchestration
00:46 arif-ali joined #salt
00:47 mariusv joined #salt
00:47 mariusv joined #salt
00:48 whytewolf Dpu84: sorry, never used pepper so not sure how it works
00:48 Dpu84 @whytewolf salt-run state.orchestrate orch.install pillar="{tgt: 'mylabmachine'}"
00:48 pcn whytewolf: so you're saying that it should pass in the pillar as kwargs and the lowstate should include that in the kwargs and that should get passed to the master were the orchestrate runner knows to pull "pillar" from the kwargs
00:49 Dpu84 this is the salt-run command i used, pepper works fine for the normal salt commands
00:49 pcn Dpu84: you should quote tgt too IIRC, and have the single quotes on the outside and double on the inside so json is happy
00:49 Dpu84 but not sure how to make the orchestate to work :(
00:49 pcn Let's take a look at libpepper
00:51 whytewolf Dpu84: iirc it is --pillar='{"tgt":"mylabmachine"}' when used with salt-run
00:51 pcn Looks like the cli should be passing the pillar on as a kwarg from https://github.com/saltstack/pepper/blob/develop/pepper/cli.py#L332-L339
00:53 Dpu84 test:~$ pepper --client runner state.orchestrate orch_install.sls --pillar "{tgt: 'mylabmachine'}" {     "return": [         {             "data": {                 "mymaster_master": [                     "No matching salt environment for environment '{tgt: 'mylabmachine'}' found",                     "No matching sls found for 'orch_install.sls' in env '{tgt: 'mylabmachine'}'"                 ],                 "retcode": 1
00:54 Dpu84 --pillar doesnt help
00:54 whytewolf Dpu84: which is it. salt-run that works or pepper?
00:54 whytewolf you keep switching back and forth between them
00:54 Dpu84 salt-run works
00:54 Dpu84 pepper command doesnt work
00:55 whytewolf okay. because that is not what you just said
00:55 gmoro_ joined #salt
00:56 whytewolf but yes, pcn is right. pillar should be passed as a kwarg
00:57 whytewolf no matching env? that is an odd error
00:58 watersoul_ joined #salt
00:59 brucewang joined #salt
01:00 Reverend joined #salt
01:00 descrepes joined #salt
01:00 whytewolf Dpu84: what is the error you started with?
01:01 coldbrewedbrew joined #salt
01:01 coldbrewedbrew joined #salt
01:01 coldbrewedbrew_ joined #salt
01:02 jerrcs joined #salt
01:02 Hipikat joined #salt
01:02 benner joined #salt
01:02 Corey joined #salt
01:02 avalarion joined #salt
01:05 afics joined #salt
01:06 antpa joined #salt
01:07 tom29739 joined #salt
01:09 ntropy Dpu84: totally try quoting tgt in your pillar as pcn suggested
01:09 ntropy pepper --client runner state.orchestrate orch_install.sls --pillar '{"tgt": "mylabmachine"}'
01:13 Dpu84 i tried that command but still getting the same error
01:13 Dpu84 test:~$ pepper --client runner state.orchestrate orch_install.sls --pillar "{tgt: 'mylabmachine'}" {     "return": [         {             "data": {                 "mymaster_master": [                     "No matching salt environment for environment '{tgt: 'mylabmachine'}' found",                     "No matching sls found for 'orch_install.sls' in env '{tgt: 'mylabmachine'}'"                 ],                 "retcode": 1
01:13 scoates joined #salt
01:13 whytewolf pepper --client runner state.orchestrate orch_install.sls pillar='{"tgt": "mylabmachine"}'
01:13 _KaszpiR__ joined #salt
01:14 avalarion joined #salt
01:14 Dpu84 No matching sls found for 'orch_install.sls' in env 'base'"
01:15 whytewolf ...
01:15 whytewolf pepper --client runner state.orchestrate orch.install pillar='{"tgt": "mylabmachine"}'
01:16 whytewolf not sure where the orch_install.sls came from
01:17 theblazehen joined #salt
01:28 skullone joined #salt
01:30 scoates joined #salt
01:32 pipps joined #salt
01:33 Rumbles joined #salt
01:41 toabi joined #salt
01:47 canci joined #salt
01:47 swills joined #salt
01:48 ilbot3 joined #salt
01:48 Topic for #salt is now Welcome to #salt! <+> Latest Versions: 2016.3.6, 2016.11.4 <+> 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
02:09 rem5 joined #salt
02:13 fracklen joined #salt
02:18 mpanetta joined #salt
02:27 asyncsec joined #salt
02:27 renoirb joined #salt
02:29 scoates joined #salt
02:33 onlyanegg joined #salt
02:34 zerocoolback joined #salt
02:37 mpanetta joined #salt
02:45 evle joined #salt
02:55 penguinpowernz joined #salt
02:58 antpa joined #salt
02:59 Praematura joined #salt
03:19 cyteen joined #salt
03:20 antpa joined #salt
03:21 justanotheruser joined #salt
03:22 Brew joined #salt
03:30 miruoy_ joined #salt
03:32 raspado joined #salt
03:37 onlyanegg joined #salt
03:43 mosen joined #salt
03:47 teratoma joined #salt
03:53 gmoro joined #salt
03:54 scoates joined #salt
03:55 Praematura joined #salt
03:58 onlyanegg joined #salt
04:08 swa_work joined #salt
04:25 hemebond Even Saltstack can't make Docker more pleasant to use :-(
04:26 ashmckenzie joined #salt
04:38 Bock joined #salt
04:42 it_dude joined #salt
04:43 Xenophon1 joined #salt
04:52 rem5 joined #salt
04:52 AvengerMoJo joined #salt
04:59 onlyanegg joined #salt
05:03 cyborg-one joined #salt
05:03 antpa joined #salt
05:13 felskrone joined #salt
05:19 scoates joined #salt
05:24 armyriad joined #salt
05:24 antpa joined #salt
05:25 ntropy hemebond: i think the only option might be kubernetes?
05:26 it_dude joined #salt
05:26 hemebond I hope not. I've managed to get it working. Turns out you don't need to write a systemd file just to get the container started automatically.
05:27 hemebond In fact it can prevent Salt from managing the container properly
05:32 rdas joined #salt
05:43 lasseknudsen joined #salt
05:52 rick_ joined #salt
06:03 scoates joined #salt
06:07 yuhl______ joined #salt
06:09 golodhrim|work joined #salt
06:12 zenchiken joined #salt
06:13 nikdatrix joined #salt
06:15 scoates joined #salt
06:19 mpanetta_ joined #salt
06:20 Tanta joined #salt
06:23 aldevar joined #salt
06:36 fracklen joined #salt
06:45 onlyanegg joined #salt
06:58 sh123124213 joined #salt
07:04 sh123124213 joined #salt
07:05 zulutango joined #salt
07:06 sh123124213 joined #salt
07:16 fracklen joined #salt
07:28 jas02 joined #salt
07:33 jas02 joined #salt
07:34 do3meli left #salt
07:41 it_dude joined #salt
07:42 fracklen joined #salt
07:42 _KaszpiR_ joined #salt
07:43 it_dude joined #salt
07:45 JohnnyRun joined #salt
07:48 zer0def joined #salt
07:50 antpa joined #salt
07:52 ronnix joined #salt
07:53 nikdatrix joined #salt
07:54 muxdaemon joined #salt
07:55 Rumbles joined #salt
08:00 preludedrew joined #salt
08:07 mikecmpbll joined #salt
08:09 irated joined #salt
08:13 BernhardG joined #salt
08:14 geomacy joined #salt
08:17 BernhardG Hello together, I just upgraded one minion from 2016.11.3 to 2016.11.4 (master is already running 2016.11.4). Now I get the critical error message that the "master servers public key did not authenticate". The minion was running correctly with 2016.11.3. Are there any things to look after to correct that behaviour?
08:18 hemebond BernhardG: I haven't heard of any problems like that. Did you leave the config and keys intact?
08:19 muxdaemon joined #salt
08:20 BernhardG hemebond, yes I just did a "yum upgrade" on a CentOS 7.3 installation. If I downgrade to 2016.11.3 I can run commands like 'salt-call test.ping' again. My Master is a Ubuntu 14.04 LTS system if that matters.
08:20 hemebond Master is also running 2016.11.4?
08:21 BernhardG hemebond, yes salt-master 2016.11.4 (Carbon)
08:23 jas02 joined #salt
08:25 BernhardG hemebond, there is an issue ( https://github.com/saltstack/salt/issues/40889 ) in the bug tracker of salt, which has to do with RSA key format and CentOS 7. Could that relate somehow (even though this issue is for a master which runs on CentOS)?
08:25 saltstackbot [#40889][OPEN] salt-master 2016.11.4 crashes on CentOS 7 with error "RSA key format is not supported" | Description of Issue/Question...
08:25 hemebond Ah. There are always problems with CentOS for some reason :-)
08:26 hemebond I might make that my first step when helping. "Are you using CentOS? Don't." ☺
08:26 BernhardG Well, I could just test it with a RHEL 7 (Workstation).
08:27 BernhardG hemebond, what else would you use if you need long support but not the full RedHat Support?
08:27 hemebond Long support? Like paid support?
08:28 BernhardG hemebond, update support (until 2024 in case of 7) mainly.
08:29 pbandark joined #salt
08:31 Mattch joined #salt
08:35 BernhardG hemebond, the problem also exists on my (original) RHEL 7 workstation instance unfortunately. Any way to debug this?
08:36 hemebond Run salt-minion in the foreground in debug mode?
08:38 s_kunk joined #salt
08:38 BernhardG Unfortunately no helpful information other than "Decrypting the current master AES key" just before the "The Salt Master server's public key did not authenticate!" message.
08:39 hemebond I wonder if it affects any other OSs
08:39 Kelsar joined #salt
08:40 lilvim joined #salt
08:41 BernhardG I could try it with another ubuntu 14.04 LTS instance.
08:43 BernhardG Ubuntu 14.04 LTS Minion to Ubuntu 14.04 LTS Master works. I will provide a versions report of those three machines maybe this information is helpful.
08:44 J0hnSteel Morning all, I'm trying to use my custom module inside a state file and fails with "Jinja variable has no atribute".The same module I can execute with salt and salt-call --local commands. have also set extension_modules and module_dirs and it's still failing salt 2016.11.4
08:44 jas02 joined #salt
08:45 onlyanegg joined #salt
08:45 BernhardG hemebond, this is the versions report of those systems: https://pastebin.com/WhMAhjSy
08:49 bbradley joined #salt
08:50 bdrung_work joined #salt
08:53 rem5 joined #salt
08:58 BernhardG hemebond, I found a workaround.
08:58 hemebond Oh?
08:58 BernhardG hemebond, rpm -e --nodeps python2-pycryptodomex.x86_64
08:59 hemebond A missing lib?
08:59 BernhardG hemebond, this package is not used on Ubuntu but installed as dependency on RPM based systems. If I deinstall this package I can connect.
08:59 hemebond Oh interesting.
08:59 BernhardG hemebond, no it is a lib which seems to be not needed but is a rpm dependency.
08:59 hemebond I wonder if that forces Salt to fallback on some other implementation.
09:00 asyncsec joined #salt
09:00 BernhardG seems so: pycrypto: 2.6.1 <-> pycryptodome: 3.4.3 (the last one is only there on my CentOS system)
09:02 BernhardG It seems a new rpm package is needed to solve that problem globally, right? But I could also have some othere downsides I don't know yet. What should I do now?
09:03 hemebond Add a comment to that Github issue you find earlier with what you just found.
09:04 hemebond Someone from Saltstack should have a look at it soon.
09:04 jas02 joined #salt
09:07 Ricardo1000 joined #salt
09:09 antpa joined #salt
09:10 BernhardG Please, dear developers, have a look at https://github.com/saltstack/salt/issues/40889 and my additional comment. Thank you much!
09:10 saltstackbot [#40889][OPEN] salt-master 2016.11.4 crashes on CentOS 7 with error "RSA key format is not supported" | Description of Issue/Question...
09:10 antpa joined #salt
09:13 evle joined #salt
09:14 fracklen_ joined #salt
09:15 ProT-0-TypE joined #salt
09:15 antpa joined #salt
09:18 Kelsar joined #salt
09:23 BernhardG Possibly the same problem occurs there RHEL 6.9 system: https://github.com/saltstack/salt/issues/40890 - problem with cryptodome
09:23 saltstackbot [#40890][OPEN] Cryptodome breaks FIPS Compliance in 2016.11.4-1 | Description of Issue/Question...
09:24 J0hnSteel OK to rephrase my earlier question is it mandatory to put "# _modules/storage.py\n#!python" headers inside Execution Module when calling if from inside jinja ?
09:24 jas02 Hello, sorry for newbie question, but how can I send/or save output of one cmd.run to the (runtime) variable? I need to parse output of one state and use it in another one.
09:29 _KaszpiR_ joined #salt
09:33 sh123124213 joined #salt
09:33 N-Mi__ joined #salt
09:42 uglylun joined #salt
09:43 uglylun left #salt
09:50 citaret joined #salt
09:50 Neighbour jas02: If you're using jinja, try something like {%- set my_output = salt['cmd.run_all'](arguments for cmd.run_all) %}
09:50 Neighbour https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.cmdmod.html
09:50 Neighbour Not thate you can only run execution modules like this, not states
09:55 zerocoolback joined #salt
09:56 zerocoolback joined #salt
10:04 nebuchadnezzar joined #salt
10:11 o1e9 joined #salt
10:12 sh123124213 joined #salt
10:13 om2 joined #salt
10:29 jas02 Neighbour: Problem is, that jinja is evaluated at start, and the one command send output to the file, and set <variable_name> is evaluated before output file actually exist.
10:29 jas02 Neighbour: Please see code example: https://gist.github.com/jas02/150e144d975077f7453a3a077c55a3dc
10:36 scoates joined #salt
10:39 jas02 Neighbour: And I don't know how evaluated if set <variable_name> should be run or not in same manner as cmd.run/unless
10:39 Ricardo1000 How salt compare files with remote fs ?
10:42 zerocoolback joined #salt
10:46 onlyanegg joined #salt
10:48 Hybrid joined #salt
10:54 rem5 joined #salt
11:05 antpa joined #salt
11:06 sh123124213 joined #salt
11:06 Miouge joined #salt
11:09 antpa joined #salt
11:21 tobiasBora joined #salt
11:28 uglylun joined #salt
11:30 zerocoolback joined #salt
11:31 Rumbles joined #salt
11:38 jas02_ joined #salt
11:44 jas02 joined #salt
11:48 amcorreia joined #salt
11:51 sh123124213 joined #salt
11:51 Rumbles joined #salt
12:04 ronnix joined #salt
12:05 Rumbles joined #salt
12:05 Neighbour jas02: Hmm, then I suppose you could create 2 .sls-files and orchestrate them to be run sequentially
12:05 Neighbour there is no capture-output support (that I know of) like ansible's `register`
12:06 jas02 Neighbour: Thanks for suggestion, I have to figure it out. Definitely try your approach.
12:11 antpa joined #salt
12:14 kbaikov joined #salt
12:16 tobiasBora joined #salt
12:17 ProT-0-TypE joined #salt
12:17 zer0def jas02: you could try calling the first `cmd.run` in jinja, then pass them to the latter `cmd.run`
12:18 zer0def oh, right, apparently that's an issue
12:18 _JZ_ joined #salt
12:19 evle joined #salt
12:25 Ricardo1000 joined #salt
12:31 Hybrid Hi
12:31 Hybrid Is there a way to disable ipv6 support completely in salt ?
12:32 Hybrid I tried ipv6: False on master and minion but I keep getting a warning about a 10s delay due to ipv6 not properly set for hostname in dns or /etc/hosts
12:32 Hybrid but I don't use ipv6 and I don't want salt to bother with this
12:40 BernhardG joined #salt
12:45 numkem joined #salt
12:47 onlyanegg joined #salt
12:50 asyncsec joined #salt
12:55 rem5 joined #salt
12:56 antpa joined #salt
12:58 kbaikov joined #salt
13:01 gmoro joined #salt
13:10 nicksloan joined #salt
13:15 racooper joined #salt
13:15 nicksloan I'm getting "Master address could not be resolved", but I can ping (not test.ping) the hostname from that minion. Other minions that are in the same security group and VPC subnet are doing just fine. Fairly certain it's not a firewall issue.
13:19 tercenya joined #salt
13:21 tobiasBora joined #salt
13:24 pbandark Hi.. can we reuse the pillar data mentioned in external pillar in the same file. for ex. if I define external pillar data "app_location: /opt/app/" then  can I use  value of "app_location" in the same external pillar file.  ? https://paste.fedoraproject.org/paste/rfZmx5Q9M80OFWxaSQMkX15M1UNdIGYhyRLivL9gydE=
13:25 pbandark i tried but its not working. is there any other way to ahieve it ?
13:27 gmoro_ joined #salt
13:34 Brew joined #salt
13:35 pcn pbandark: I am not aware of a way to declare pillar data and reference it from other pillar data reliably, especially not in the same file.
13:35 megamaced joined #salt
13:36 Neighbour pbandark: use jinja for this {{ salt['pillar.get'](key, default) }}
13:36 pcn pbandark: What I am aware of is that you can declare data in jinja variables {% set foo = "this" %} and then dereference it with {{ foo }}/blah
13:37 Neighbour or, set a jinja variable with the pillar data, like pcn just suggested :)
13:37 pbandark ok.. let me check
13:45 rem5 joined #salt
13:45 rem5_ joined #salt
13:48 onlyanegg joined #salt
13:52 pbandark Neighbour: did you meant "{{ salt['pillar.get]('config:user:app_location') }}tomcat" https://paste.fedoraproject.org/paste/9Q8hWNb~uUd9SeGN3Q0XtV5M1UNdIGYhyRLivL9gydE=
13:53 pbandark ?
13:54 jvelasquez joined #salt
13:54 tercenya joined #salt
13:54 Neighbour pbandark: yes
13:54 pbandark its failing :(
13:55 Neighbour oh, i see..my mistake..this is a pillar file you're editing? You can't use pillar.get when compiling a pillar file
13:55 Neighbour the only data available during building a pillar file are the grains
13:55 pbandark ahh. so its not possible?
13:55 Neighbour however, you can do it differently
13:56 Neighbour set a jinja variable, and use that when defining the app_location, as well as the home-dir
13:57 pbandark ok. let me give a try
13:57 Neighbour like so: https://paste.fedoraproject.org/paste/xyIMGw4UiCp72nLl67RMR15M1UNdIGYhyRLivL9gydE=
13:58 pbandark checking
13:58 pbandark ok
13:58 dyasny_ joined #salt
14:05 nikdatrix joined #salt
14:09 nicksloan Here's an update: the minion can access the master by IP address with no trouble. ping can resolve the hostname of my master to that same IP address, but for some reason salt can't.
14:10 gord0 joined #salt
14:10 sh123124213 joined #salt
14:11 tobiasBora joined #salt
14:14 Miouge joined #salt
14:15 cro joined #salt
14:18 jdipierro joined #salt
14:19 s_kunk joined #salt
14:19 s_kunk joined #salt
14:20 pbandark Neighbour: in jinja, is it possible to set multiple variables(key value pair) in same {% set  <variable1>, <variable2> %} ? or I will need to create separate {} block for each variable ?
14:22 jdipierro joined #salt
14:22 felskrone Can i refresh grains defined in /etc/salt/grains within a module during execution? i have module that updates /etc/salt/grains, but the minion keeps running with the values it has read on startup. i tried saltutil.sync_grains, grains.items, but no success
14:24 dyasny joined #salt
14:28 zenchiken joined #salt
14:35 Neighbour pbandark: AFAIK, you will have to use separate {% %}-blocks for that
14:35 tobiasBora joined #salt
14:35 gmocanu joined #salt
14:35 pbandark ok
14:35 mpanetta joined #salt
14:37 sh123124213 joined #salt
14:42 gmoro joined #salt
14:45 aphor joined #salt
14:46 zmalone joined #salt
14:46 zmalone left #salt
14:47 sarcasticadmin joined #salt
14:49 Tanta joined #salt
14:50 ronnix joined #salt
14:55 __aphor__ joined #salt
15:07 PatrolDoom joined #salt
15:07 Ricardo1000 joined #salt
15:09 sh123124213 joined #salt
15:11 willprice joined #salt
15:12 sud0x3 joined #salt
15:15 sud0x3 Im having a bit ouf trouble enabling a service on arch linux via a salt state file
15:15 sp0097 joined #salt
15:16 jas02 joined #salt
15:16 sud0x3 i get this error "The named service dnscrypt-proxy is not available" i have tried changing the name to name.service and a few other things
15:17 sud0x3 verified that salt does find the service by calling the module.service.get_enabled
15:18 sud0x3 anyone else come across this with arch linux systems
15:24 GreatSnoopy joined #salt
15:24 GreatSnoopy Hello
15:25 GreatSnoopy I have a question: is custom tops is or not a supported feature ? https://docs.saltstack.com/en/latest/topics/master_tops/  [5:10]  doc says “imply place them into salt://_tops in the Salt fileserver and use the saltutil.sync_tops runner to sync them. If this runner function is not available, they can manually be placed into extmods/tops, relative to the master cachedir (in most cases the full path will be /var/cache/salt/master/
15:25 GreatSnoopy in 2016.11.3-1 we have salt-run saltutil.sync_tops ‘saltutil.sync_tops’ is not available.  [5:12]  indeed, if I put my code in the cache directory it works, but I do not plan to insert code undercover in the cache
15:27 iamtew joined #salt
15:27 jas02 joined #salt
15:27 whytewolf GreatSnoopy: it is a runner function salt-run saltutil.sync_tops
15:29 whytewolf on the master of coarse
15:29 GreatSnoopy #  salt-run saltutil.sync_tops 'saltutil.sync_tops' is not available.
15:29 GreatSnoopy even the doc says "if the function is not available"
15:29 GreatSnoopy why would the function be unavailable anyway ?
15:30 whytewolf could be a number of reasons but normally it has to do with __virtual__ but with saltutil it should be there.
15:30 GreatSnoopy well it's not :)
15:30 whytewolf that function says it was introduced in 2016.3
15:30 GreatSnoopy well isn't 2016.11 NEWER ?
15:31 whytewolf calm down
15:31 whytewolf yes
15:31 whytewolf one moment checking the code
15:31 whytewolf ahhh
15:32 whytewolf it isn't in 2016.11.3 it is put in in 2016.11.4
15:32 whytewolf what was just released a couple of days ago
15:33 GreatSnoopy hm
15:33 Kelsar Is there someone actively working on the portage pkg module?
15:33 GreatSnoopy whytewolf: can I also ask a question about that feature ?
15:33 whytewolf i see no signs of it in 2016.3 most likely it will be in the next release of that which will most likely be the last 2016.3 release
15:34 whytewolf I see no reason why you couldn't. but just for the record i don't work for salt. no need to be formal with me
15:34 GreatSnoopy oh ok :)
15:35 nicksloan joined #salt
15:35 GreatSnoopy its not clear to me if a custom master tops is supposed to receive parameters and how is it called : once, and the data is obtained once and eventually cached or is called for each minion
15:36 GreatSnoopy and it should return only data for a certain minion
15:36 jas02 joined #salt
15:38 whytewolf ahh that i don't know. all the modules i see just list kwargs as the input.
15:38 gmoro joined #salt
15:39 sud0x3 anyone able to give some input on my question about services on arch linux? ?
15:39 GreatSnoopy ok, i'll fiddle with it to see what it receives
15:39 whytewolf sud0x3: sorry i don't use arch linux so unable to help with that
15:40 whytewolf GreatSnoopy: might help looking at the master_tops that come with salt also. see how they handle it. https://github.com/saltstack/salt/tree/v2016.11.4/salt/tops
15:41 tiwula joined #salt
15:42 sud0x3 whytewolf: No problem, Yeah when it comes to systems provisioning for arch support is few and far between
15:43 sjorge joined #salt
15:44 ProT-0-TypE joined #salt
15:45 whytewolf Kelsar: https://packages.gentoo.org/packages/app-admin/salt
15:46 Kelsar whytewolf: that is not an answer to my question ;)
15:47 whytewolf oh doh misread the question
15:50 whytewolf Kelsar: looks like it gets patches now and then but they are few and far between
15:50 whytewolf https://github.com/saltstack/salt/commits/develop/salt/modules/ebuild.py
15:50 Kelsar whytewolf: yeah, fixed that a while agon and then somebody rewrote it... it is on my list
15:51 Inveracity joined #salt
15:51 Kelsar oh wow that was 2015... time flys
15:55 aldevar left #salt
15:58 tobiasBora joined #salt
16:00 londo joined #salt
16:00 sud0x3 left #salt
16:02 dyasny joined #salt
16:04 Kelsar is there an upcoming release for salt, looks like my issue is fixed in develop since november...
16:05 GreatSnoopy whytewolf: thanks
16:05 Praematura joined #salt
16:08 whytewolf Kelsar: nitrogen should be getting it's first RC soon.
16:08 whytewolf hopefully
16:09 adelcast joined #salt
16:10 onlyanegg joined #salt
16:24 woodtablet joined #salt
16:33 willprice94 joined #salt
16:34 candyman88 joined #salt
16:35 cscf Salt has some inconsistent naming with regard to "ip4" vs "ipv4"
16:35 tercenya joined #salt
16:36 it_dude joined #salt
16:39 mpanetta joined #salt
16:42 wendall911 joined #salt
16:43 Rumbles joined #salt
16:46 tercenya joined #salt
16:51 djgerm joined #salt
16:57 nixjdm joined #salt
16:59 pbandark hi. I am configuring tomcat service with salt. I am taking "password" input from the user and storing it in pillar data so that, I can use it to configure the user with provided password. But, saving plaintext password in pillar is not secure. I am trying to find out a way to encrypt provided password and then copy to tomcat service. Is it possible with salt or I need to manage it outside of salt ?
16:59 cmarzullo you can gpg encrypt pillar so only your master can decode it and provide it to the minions
17:00 TheoSLC joined #salt
17:01 pbandark cmarzullo: gpg encrypt pillar ?
17:02 cmarzullo yeah you can encrypt certain fields or whole blocks of yaml.
17:03 pbandark cmarzullo: how to do that ?
17:03 cmarzullo It's a process. Check the saltstack site they have instructions there.
17:03 pbandark ok
17:06 pipps joined #salt
17:07 pipps joined #salt
17:10 nicksloan so I've just run dns_check myself against the master hostname, and ports 4505 and 4506, with and without safe=True and it always works. The only time it fails is if ipv6 is True, but it's explicitly set to false in my minion config.
17:10 sarcasticadmin joined #salt
17:11 nicksloan Yet my minion still can't connect to the master
17:11 sarcasticadmin joined #salt
17:15 cryptolukas joined #salt
17:15 mikecmpbll joined #salt
17:17 cryptolukas After update on 2016.11.4+ds-1 on Ubuntu 16.04 I get these erros in the master log. I don't get these errors before. https://gist.github.com/anonymous/ca73795dd792e031049f66108194d32b
17:21 whytewolf lol, just looked it up. the only thing that changed with that is that warning was added. so if there was a problem before. you didn't notice it
17:21 deuscapturus joined #salt
17:21 whytewolf https://github.com/saltstack/salt/commit/cff027ddc615bde672ec6342172ea760b9b0ad25
17:22 frew if I have a managed file is there a simple way to, without changing the actual source (which is a jinja template) to prefix the content?
17:24 kojiro joined #salt
17:24 cryptolukas okay.
17:25 Deliant joined #salt
17:25 whytewolf frew: file.* are interesting. you can combine them together with require. so say a file.managed can require a file.prepend and it will take into account both.
17:26 whytewolf in thoery
17:26 frew oh huh
17:26 frew that's a cute idea
17:27 whytewolf it mostly is used for things like file.managed being used with file.recurse with clean:true
17:27 whytewolf so that the recuse doesn't delete the file.managed file
17:33 Sarphram joined #salt
17:34 Trauma joined #salt
17:36 it_dude joined #salt
17:36 GMAzrael I am trying to deploy VBox Extensions to my windows clients, how would i go about doing a cmd.run to download the extpack from the salt server and running a command to execute it, then deleting the file?
17:37 Sarphram joined #salt
17:37 rubenb joined #salt
17:37 TCShain joined #salt
17:38 TCShain Greetings helpful salt community!
17:38 TCShain Our senior dev ops engineer, that architected our salt stack, left the company yesterday.
17:39 TCShain Guess who's responsible for it now?
17:39 TCShain <-- This guy.
17:39 XenophonF joined #salt
17:39 cscf TCShain, congrats
17:39 toastedpenguin anyone doing any sort of event driven salt reactor or other actions based on alerts from tools like datadog or new relic?
17:39 TCShain I'm getting a funky problem I can't figure out, would anyone please be of assistance?
17:40 TCShain Thank you! cscf !
17:40 TCShain Several of my minions are getting: "The Salt Master has rejected this minion's public key!"
17:40 TCShain I've tried deleting the key from the master, then restarting the minion, but the error persists.
17:40 TCShain Pretty sure I'm missing something simple, but my google-fu is failing me...
17:41 TCShain Any thoughts?
17:42 TCShain I've been working with salt for over a year now, but mostly writing states.
17:42 TCShain Haven't been exposed to much of the inner workings quite yet.
17:43 sqwishy joined #salt
17:43 sqwishy Can execution models be a source of data for pillars?
17:44 whytewolf sqwishy: pillars are generated on the master. so anything you use would be in the master context
17:44 cmarzullo sqwishy: yes. it's a pain in the ass and unreliable. since the master has to call out to the minion everytime.
17:45 cyborg-one joined #salt
17:45 whytewolf TCShain: maybe actually regen the certs on the minions.
17:45 sqwishy I guess I have a state on the master to apply to minions matching on their pillar. But I want that pillar value set by an execution module on the master.
17:45 cmarzullo TCShain: is the minion pointing to the right master?
17:48 sqwishy Can I use external pillars to do what I'm trying to do?
17:48 TCShain cmarzullo: yes the minion points to the right master, and in fact, the minion pops up in a salt-key -L in the 'accepted keys' section
17:48 TCShain But then the minion stops with the rejected key message
17:48 TCShain and eventually, the minion falls off the salt-key -L list
17:49 cmarzullo Is it just one minion?
17:49 TCShain whytewolf: the instances i'm running these minions on are pretty ephemeral, I've terminated several of these instances and let the minion reinstall from bootstrap
17:49 TCShain I assume that's as effective as regening the keys manually
17:49 whytewolf it should...
17:49 TCShain cmarzullo: it is several minions
17:50 cmarzullo are you running multimaster setup?
17:50 TCShain I'm afraid to disconnect any that are currently communicating with the master, in fear they will not be able to reconnect
17:50 TCShain cmarzullo: nope -- just the one master
17:51 whytewolf what version is running?
17:51 TCShain just a minute
17:51 cryptolukas Get anyone such error sometimes. I have no production impact but i receive this erros. https://gist.github.com/anonymous/34fd21841e5e56ee6578b8df412f2dd5
17:52 whytewolf cryptolukas: that looks like gitlab is having issues now and then with authentication
17:52 TCShain whytewolf: salt-master 2015.5.8 (Lithium)
17:52 TCShain salt-minion 2016.11.4 (Carbon)
17:52 whytewolf ...
17:52 whytewolf that could be the problem
17:53 whytewolf minion should be same or lower then master
17:53 TCShain Hmm
17:54 TCShain let me check my 'working' minions
17:55 TCShain whytewolf: salt-minion 2016.11.3 (Carbon)
17:55 TCShain ^ good minion
17:55 TCShain other minion is just slightly higher
17:55 TCShain the 'bad' one that is
17:56 whytewolf did these minions already exist and just update the version on them at some point?
17:57 whytewolf [new minions would regen a key, old minions would keep old keys]
17:58 TCShain We shoot our instances in the head quite regularly, so they are rejoining the master on-demand
17:58 TCShain I think, in the context you're talking about, these are effectively new minions
17:58 TCShain new linux instances with new dynamic ip address -- fresh minion installations
17:59 v12aml joined #salt
17:59 whytewolf okay. try setting up a new minion but tag the bootstrap to install an older version through git. say v2015.8.8
17:59 whytewolf err v2015.5.8
17:59 TCShain Okay, I'll give that a try.  Thank you!
18:00 TCShain I'll let you know the results either way.
18:00 TCShain Strange that I have minion on 2016.11.3 that is working
18:00 KyleG joined #salt
18:00 KyleG joined #salt
18:00 TCShain maybe I'll baby step down to that first, see what happens.
18:01 whytewolf yes it is strange. but newer minion in an older master is NOT something that is predictable. all sorts of trouble can pop up in unpredictable ways
18:01 GMAzrael_ joined #salt
18:02 tapoxi joined #salt
18:05 sh123124213 joined #salt
18:08 tercenya joined #salt
18:12 sh123124213 joined #salt
18:17 LeProvokateur joined #salt
18:22 Praematura joined #salt
18:22 onlyanegg joined #salt
18:23 nicksloan joined #salt
18:23 cmarzullo "Unable to find IPv6 record for .. causing a 10 second timeout when rendering grains." <-- is this a new thing?
18:24 whytewolf yeah it was a warning added to the grains. no other change was made to that grain.
18:25 woodtablet anyone here by any chance know how i can get in touch with _Pyro ? i didnt get a chance to /whois him 2 days ago =(
18:25 cmarzullo stupid warnings. what are they good for anyway! /s
18:25 cmarzullo guess I need to do host entries for ipv6.
18:32 fracklen joined #salt
18:33 jas02 joined #salt
18:34 pipps joined #salt
18:35 cscf cmarzullo, I think it only shows up when the machine has an IPv6 address
18:36 cmarzullo stupid ipv6. it'll never catch on /s
18:37 whytewolf it has only been 20 years. give it time
18:37 cmarzullo :)
18:37 cmarzullo I'll have to figure out what it's doing. I don't want to dns my ipv6.
18:39 aneeshusa joined #salt
18:39 cscf cmarzullo, /etc/salt/minion ipv6: False
18:39 cmarzullo cscf: my hero!
18:40 cmarzullo thanks.
18:40 whytewolf turning that off has no effect on the grain getting ipv6
18:40 cscf oh, you're right, that's a grain
18:40 cscf Didn't read the message very well
18:40 cmarzullo cscf: is no longer my hero
18:40 cscf why would a DNS lookup timeout?  Shouldn't it NXDOMAIN?
18:40 promorphus joined #salt
18:40 cscf cmarzullo, :(
18:41 TCShain ha
18:41 cscf That was brief
18:41 whytewolf you would think so
18:41 Edgan cryptolukas: gitpython or pygit2?
18:41 nixjdm joined #salt
18:41 Sketch cscf: if there's no dns server running, or it's firewalled
18:42 cscf Yeah I guess, but then shouldn't IPv4 timeout too?
18:42 Sketch eventually
18:43 pipps joined #salt
18:46 candyman88 joined #salt
18:48 pipps joined #salt
18:50 pipps joined #salt
18:51 GMAzrael_ joined #salt
18:51 GMAzrael_ I am trying to deploy VBox Extensions to my windows clients, how would i go about doing a cmd.run to download the extpack from the salt server and running a command to execute it, then deleting the file?
18:52 onlyanegg joined #salt
18:53 jas02 joined #salt
18:54 cscf GMAzrael_, file.manage it down, then cmd.run to execute && delete
18:54 GMAzrael_ interesting...
18:54 cscf Or just don't delete it, that would prevent file.manage from putting it there again
18:56 Sketch unless you know where it puts the deployed thingy, then you could only file.manage it if that doesn't exist
18:57 GMAzrael_ @sketch the idea is that I only need the file long enought for VBox to run the execution
18:57 GMAzrael_ because the ExtPack is tied to the version that is installed
18:57 GMAzrael_ through the win-repo
18:57 Sketch or if there's some command you can run like vboxmanage are_extensions_installed that will tell you whether it's installed or not
18:58 GMAzrael_ don't believe that command is in the module
18:59 Sketch you're going to have to have some way for salt to detect of it's installed or not, if you want to only put it there when it needs to be installed
18:59 Sketch unless this is just a run-one-time state
18:59 Sketch then i guess it doesn't matter
18:59 GMAzrael_ One time run
19:00 nicksloan joined #salt
19:00 ChubYann joined #salt
19:00 Sketch then just file.manage, cmd.run the install, then file.absent
19:00 Sketch much like cscf said :)
19:04 filippos joined #salt
19:04 jas02 joined #salt
19:06 geomacy joined #salt
19:13 jas02 joined #salt
19:15 _JZ_ joined #salt
19:16 Shirkdog joined #salt
19:16 Shirkdog joined #salt
19:19 pbandark joined #salt
19:20 geomacy joined #salt
19:21 jas02 joined #salt
19:22 klaas joined #salt
19:22 jas02 joined #salt
19:23 ronnix joined #salt
19:25 barkingfoodog joined #salt
19:27 GMAzrael_ @sketch How I would force file.abset to occur after the cmd.run? Is there a require I could do?
19:29 cryptolukas left #salt
19:30 aldevar joined #salt
19:31 cscf GMAzrael_, require will do it, yes
19:32 cscf Also you don't use @ on IRC
19:32 GMAzrael_ apologies
19:32 barkingfoodog It looks like the MD5 for the x64 Windows installer is wrong for 2016.11.4. https://repo.saltstack.com/windows/Salt-Minion-2016.11.4-x86-Setup.exe.md5 gives EB04FDB21EC6A51F927524C86792D1DB but I get ad22fb4cea0b3673ecf13f6393c03433 when I calculate it. Does anyone know who is responsible for maintaining the download site so I can report this?
19:32 wendall911 joined #salt
19:32 cscf GMAzrael_, you'd want the file.absent to require: - cmd: install_stuff , which requires the file.managed
19:33 snc joined #salt
19:33 censorshipwreck joined #salt
19:33 seanz joined #salt
19:33 Heartsbane joined #salt
19:33 Heartsbane joined #salt
19:34 pipps joined #salt
19:34 GMAzrael_ apologies
19:35 aneeshusa joined #salt
19:35 GMAzrael_ vboxext_install is getting comment, One or more requisite failed: vbox.vboxext_finish
19:37 antpa joined #salt
19:37 GMAzrael_ cscf, https://pastebin.com/h3uPvAYa
19:39 cscf GMAzrael_, file: 'c:\temp\Oracle_VM_VirtualBox_Extension_Pack-5.1.20-114628.vbox-extpack' is wrong
19:39 whytewolf GMAzrael_: unforchantly require can use name or id. i would force the use of id by instead of using the name for your first state using an id
19:39 cscf ^
19:40 cscf I give readable names to all my states anyway
19:40 cscf er, id's I guess
19:40 GMAzrael_ so file.managed would have a name directive for the path and filename?
19:41 whytewolf yes
19:41 cscf GMAzrael_, yeah, call it vboxext_file and use - name: for the path
19:42 cscf and then require the name, which is also more readable
19:43 deuscapturus joined #salt
19:49 Sketch you could just use order: for a simple one-shot state, too
19:49 Sketch https://docs.saltstack.com/en/latest/ref/states/ordering.html
19:50 promorphus joined #salt
19:50 Sketch personally, i prefer require: and states that are part of highstate
19:50 cscf Sketch, oh, that's interesting
19:50 Sketch but for one-offs, i think just using order: is fine
19:51 pipps joined #salt
19:51 cscf Sketch, with order: last, does it run before or after states without an order: ?
19:53 Sketch pretty sure last will always run last (and first will run first).
19:53 Sketch i'm not sure where unordered states will fall related to numbered states
19:54 cscf Sketch, the doc says: "Any state declared without an order option will be executed after all states with order options are executed."
19:55 cscf Which is why I'm wondering
19:55 Sketch ah, i am not sure then
19:55 Sketch i have never actually used last...
19:55 cscf It seems contradictory
19:56 btorch joined #salt
19:56 btorch have anyone experienced this https://pastebin.ca/3803393 ?
19:56 edrocks joined #salt
19:57 bantone gtmanfred: ?
19:58 pcn btorch: what version of salt are you using?
19:58 btorch on that particular paste, the compound target does not really return anything but it could have, so not sure why salt is throwing that error on " An un-handled exception" , I don't recall having that before
19:58 btorch 2016.11.3+ds-1 on 14.04/12.04
20:01 gtmanfred bantone: whats up?
20:01 squishypebble joined #salt
20:02 bantone oh btorch was asking something
20:03 pcn btorch: is that roles grain a list or a string?
20:04 btorch pcn: it's a list
20:05 btorch pcn: so like https://pastebin.ca/3803397
20:06 btorch the same worked fine and still does on 2015.5.3+ds-1trusty1 setups
20:07 cscf Don't you need something like "is in list" rather than " == proxy"?
20:07 cscf Never used grains list matching
20:07 gtmanfred not for grains
20:07 gtmanfred grains will just check if anything in that list matches what you put to the right of :
20:08 gtmanfred for targeting at least
20:08 pcn I just did a quick experiment, and trying to match an element in a list seems to work with wildcards, but I'm not sure about that match at the beginning
20:08 btorch cscf: when I use within state files than I do that but on compound targetting I never have
20:08 gtmanfred if you were trying to do if whatever in grains.get('roles') in jinja, then you would
20:08 cscf gtmanfred, I see.
20:09 nicksloan joined #salt
20:09 pcn Oh, haha, joke's on me, we renamed the devices on the tier I decided to test on.
20:09 aldevar left #salt
20:09 cscf LOL
20:11 btorch hmm I lied too
20:11 btorch stupid copy and paste :(
20:11 btorch I get the same on a "2015.5.3+ds-1trusty1" setup
20:12 btorch a bit different output but I think its all the same in regards to the batch Iteration
20:13 btorch yeah https://pastebin.ca/3803398
20:26 keldwud joined #salt
20:26 keldwud joined #salt
20:26 pipps joined #salt
20:27 pcn btorch I wonder... how are you testing this
20:27 pcn I mean these errors remind me of when I was upgrading to 2016.3, there were errors that I think were from stale .pyc files
20:27 btorch pcn: like it's on the paste on live development systems
20:28 nicksloan so, it appears that at some point salt-cloud was creating minion configs with the port as a string, and at some point recently either salt-minion stopped casting it to an integer or the python socket module started being picky about that being an integer. Lost a good bit of time on that today because the error message was pretty vague.
20:29 btorch pcn: hmmm , so you mean stale .pyc files leftover from previous versions ? cause then the same would apply for the 2014 to 2015 version upgrade I guess
20:29 cscf I thought .pyc files with old atimes were deleted and recached?
20:29 cscf Or are you running noatime?
20:29 scoates joined #salt
20:30 cscf er, mod time
20:30 whytewolf stale pyc files are rare. and normally only happen if you have time issues.
20:30 pcn @btorch: try blowing away your .pyc files on the saltmaster and see what happens.
20:41 btorch pcn: same still happens
20:42 pcn OK, and you don't have any other custom changes in there that could be affected by arity changes, etc (I also had issues with a fix to the redis returner)
20:43 ronnix joined #salt
20:44 geomacy joined #salt
20:44 Tanta joined #salt
20:48 fenlee joined #salt
20:54 btorch pcn: nope
20:55 stomith joined #salt
20:56 rem5 joined #salt
20:57 stomith So in my top.sls file, I've got base: '*':, but I've still got clients which report back with: 'No Top file or external nodes data matches found.'
20:58 stomith shouldn't I not have to declare them specifically if they at least match the wildcard?
20:58 whytewolf is there any states in the '*'?
20:58 stomith There are, yes.
20:59 stomith and I do get that error message when I explictly define the host with states.
20:59 stomith states, meaning sls files, right?
21:00 whytewolf yes. such as base: '*': - dosomething
21:01 stomith yeah. still that error. I even upgraded, making sure that it's the correct version.
21:01 whytewolf are you using enviroments or gitfs?
21:01 hemebond Pasted the top file somewhere?
21:02 stomith neither.
21:02 whytewolf checked for missing colons?
21:02 whytewolf [sounds like something a doctor whould say]
21:03 jas02 joined #salt
21:04 stomith https://www.pastiebin.com/59010b489b5cf
21:04 Vasya666 joined #salt
21:04 stomith It's pretty simple.
21:04 hemebond You have a directory called "sls"?
21:04 hemebond Paste your file_roots from the master config
21:05 stomith I do have a directory called sls.
21:05 whytewolf hemebond: wrong error for that line of thinking. that is something like sls not found
21:06 hemebond file_roots?
21:06 stomith https://www.pastiebin.com/59010bb3cdb36
21:06 whytewolf stomith: do ALL minions have that error or only some?
21:06 stomith Some.
21:06 whytewolf is state_top set on them?
21:06 stomith I am not familiar with state_top?
21:07 whytewolf it is a config option on the minion. but if you don't know about it probley didn't get set
21:07 whytewolf which would be good
21:08 hemebond So /srv/salt/top.sls exists?
21:08 stomith it's set to top.sls on this?
21:09 whytewolf stomith: default is top.sls which = salt://top.sls to the minion
21:09 whytewolf which is normal
21:09 stomith I could comment it out?
21:09 whytewolf if you want. restart the minion after words
21:10 whytewolf then answer hemebond about location of top file.
21:10 hemebond (why would you not remove comments when pasting your config into a Github issue comment????)
21:10 stomith no change.
21:10 hemebond Restarting is always good.
21:10 stomith sorry, hemebond. /srv/salt/top.sls does exist.
21:10 hemebond Your minion isn't configured or customised?
21:10 stomith root:root 644
21:11 whytewolf hemebond: because you want to make sure everyone knows what your comments are :P
21:11 stomith no, it is not customized.
21:11 stomith actually,
21:11 hemebond whytewolf: The config file comments! You know what I mean :-D
21:11 stomith it might be somehow. I did upgrade from a previous install.
21:11 stomith from another sysadmin.
21:11 stomith apt remove salt-minion, then bootstrapped
21:12 whytewolf oh fun. lots of that going around.
21:12 hemebond Oooh. Check that. Maybe paste it. Without the file comments.
21:12 hemebond Without the comments!
21:12 whytewolf hemebond: really hates comments :P
21:12 stomith the minion file?
21:13 hemebond If I have to use auto-scroll in my browser to get past your comment I'm gonna be upset :-D
21:13 hemebond /etc/salt/minion and /etc/salt/minion.d/*
21:13 stomith great. please give me a second.
21:14 whytewolf only a second? normally takes me at least 5 to get a copy without comments
21:14 stomith I have my girlfriend's daughter trying to facetime me for some reason. :/
21:16 stomith https://www.pastiebin.com/59010bb3cdb36
21:17 hemebond Wait... it actually says "environment: None" in the config file?
21:18 stomith indeed it does. :/
21:19 stomith That's kind of standard, I'm seeing
21:19 hemebond Is it set on the minions that do work? Do any minions work?
21:19 stomith but I'm seeing it usually commented out.
21:20 stomith Let me try that.
21:20 stomith some minions do work.
21:20 stomith and commenting it out worked. awesome.
21:21 stomith thank you very much!
21:21 citaret joined #salt
21:21 hemebond That config is 8.5 times the size of my config.
21:22 hemebond Makes me wonder if I should be setting more settings.
21:22 stomith I get to take over from some other failed admin.
21:22 hemebond Now whytewolf is going to hate environments even more :-D
21:22 antpa joined #salt
21:22 whytewolf hehe
21:22 whytewolf :P
21:23 whytewolf salt enviroments need to die
21:23 hemebond But they're useful!
21:24 hemebond They just have a name that has an existing meaning.
21:26 hemebond Or implies a certain behaviour.
21:28 whytewolf this is true. they can be useful. they are just a pain to use
21:28 Rumbles joined #salt
21:28 hemebond Yeah, need to have a proper understanding of top.sls and stuff before trying to use them.
21:31 whytewolf even having a full proper understanding of top you can end up messing up on accident.
21:31 hemebond That's true.
21:33 asyncsec joined #salt
21:34 antpa joined #salt
21:34 swills joined #salt
21:34 aneeshusa joined #salt
21:37 pipps joined #salt
21:44 Vasya666 joined #salt
21:49 swa_work joined #salt
21:51 lorengordon joined #salt
21:51 Rumbles joined #salt
21:57 stomith now salt-minion starts three copies. argh.
21:58 hemebond Three copies?
21:58 stomith Sorry, three processes.
21:58 hemebond That's okay.
21:59 hemebond The minion I just checked has four.
22:00 stomith oh, okay.
22:00 stomith except I'm getting that same dreaded error. :/
22:00 stomith I've commented out everything from the minion file except for the master: line.
22:00 hemebond ????
22:00 hemebond Oh
22:00 stomith this is a different comment.
22:00 stomith er
22:01 stomith client
22:01 * stomith can't type
22:01 hemebond Wasn't there settings in there for TCP protocol and ports?
22:01 hemebond Oh, you're not using TCP transport anyway.
22:01 stomith I'm not using salt-ssh?
22:02 hemebond Ah, TCP transport is an alternative to the ZeroMQ transport.
22:02 hemebond It's the protocol the minion uses to connect to the master.
22:02 stomith Oh, thank you for explaining that.
22:03 hemebond I don't actually know what tcp_pub_port does.
22:03 hemebond Doesn't look like it has anything to do with the transport.
22:03 hemebond Plus the values in your config were the default anyway :-P
22:04 hemebond So I guess you could just remove everything except the master line.
22:04 hemebond Have you restarted the new client having a problem?
22:04 stomith I did, yes.
22:05 hemebond And it's the same error?
22:05 stomith Yes. I just bootstrapped a new version.
22:05 hemebond Over the top of an existing install?
22:06 stomith apt remove salt-minion && bootstrap
22:06 hemebond apt-get
22:06 stomith I haven't figured out how to remove a bootstrapped salt yet.
22:07 hemebond Depends how it's installed.
22:07 hemebond By default it just uses packages.
22:07 edrocks joined #salt
22:07 hemebond So you just apt-get remove it.
22:07 stomith fair enough
22:10 pipps joined #salt
22:12 stomith So I can remove /etc/salt from the minion. Any other directories I should remove after a apt-get remove?
22:13 hemebond Uninstalling the package should remove other dirs like the /var/cache/salt/
22:13 stomith it doesn't seem to.
22:13 hemebond Well that's no good. I wonder why they left that in.
22:14 hemebond Even if you purge the packages?
22:14 stomith apt-get remove salt-minion
22:14 Kelsar purge would
22:14 stomith okay, let me try that.
22:16 stomith hmm. I guess I didn't need to purge. Weird. okay, it works.
22:17 stomith I need a drink.
22:19 geomacy joined #salt
22:34 pipps joined #salt
22:38 pppingme joined #salt
22:44 scoates joined #salt
22:44 pipps joined #salt
22:46 XenophonF stomith: you and me both
22:46 KyleG1 joined #salt
22:46 whytewolf someone should buy a roand all around
22:49 whytewolf round even
22:50 whytewolf salt 'whytewolf*' pkg.install aspell
22:50 whytewolf i really wish it was that easy somedays
23:02 pipps joined #salt
23:03 frew joined #salt
23:07 pipps joined #salt
23:10 deuscapturus joined #salt
23:16 paant joined #salt
23:43 fracklen joined #salt
23:49 pipps joined #salt
23:52 alem0lars joined #salt
23:55 antpa joined #salt

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