Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-07-14

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

All times shown according to UTC.

Time Nick Message
00:00 flowstate joined #salt
00:05 colegatron_origi joined #salt
00:14 subsignal joined #salt
00:25 rem5 joined #salt
00:27 teryx510 joined #salt
00:38 Vye joined #salt
00:41 rem5 joined #salt
00:53 hyegeek joined #salt
00:57 scoates joined #salt
00:58 nyx joined #salt
01:00 flowstate joined #salt
01:03 Nahual joined #salt
01:06 JPT joined #salt
01:11 rem5 joined #salt
01:13 hyegeek joined #salt
01:17 JPT joined #salt
01:18 iceyao joined #salt
01:19 iceyao_ joined #salt
01:19 sagerdearia joined #salt
01:22 catpigger joined #salt
01:23 thejrose1984 joined #salt
01:27 DEger joined #salt
01:29 jesusaur joined #salt
01:29 metabsd joined #salt
01:31 colegatron_origi joined #salt
01:37 DEger joined #salt
01:39 DEger joined #salt
01:40 DEger joined #salt
01:45 ivanjaros joined #salt
01:47 ilbot3 joined #salt
01:47 Topic for #salt is now Welcome to #salt! | Latest Versions: 2015.5.10, 2015.8.10, 2016.3.1 | 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
01:50 ivanjaros joined #salt
01:59 flowstate joined #salt
02:03 rem5 joined #salt
02:07 DEger joined #salt
02:11 DEger joined #salt
02:15 iceyao joined #salt
02:18 evle joined #salt
02:21 my10c joined #salt
02:21 abolte joined #salt
02:22 my10c quick one: i like to report a bug: salt '*' cp.push_dir /tmp glob='*log' upload_path='/tmp/data'
02:22 badon joined #salt
02:22 my10c the error is :     ERROR executing 'cp.push_dir': The following keyword arguments are not valid: upload_path=/tmp/data
02:23 hemebond my10c: What version?
02:23 my10c instead id creates a dir under /var/cache/salt/master/minions/xxx/tmp
02:23 my10c version salt 2016.3.1 (Boron)
02:24 my10c both minion and master
02:27 squishypebble joined #salt
02:29 hemebond Works for me.
02:30 hemebond salt '*' cp.push_dir /etc/nginx/ glob="*.conf" upload_path='/antoher/test/'
02:32 my10c does the data show up under /antoher/test/ ?
02:32 hemebond yip
02:33 my10c or under /var/cache/salt/master/minions/xxx//antoher/test/
02:33 my10c odd
02:33 my10c mayb ein my master config  setting ?
02:33 hemebond /files/antoher/test/
02:33 my10c something i need to ser ?
02:33 hemebond /var/cache/salt/master/minions/xxx/files/antoher/test
02:33 hemebond You have to have "file_recv: True" in the master config.
02:34 my10c HA under /var/cache/salt/master/minions/xxx/files/antoher/test and not /antoher/test
02:34 hemebond Oh, not it's not under root.
02:34 my10c i have True is that i wanted to be under /tmp since it only temp ...
02:34 hemebond It always goes into the cache directory.
02:35 hemebond You can't get it to push files to the root of the master, it will always go into the cache.
02:35 hemebond You could, if you want it in the master /tmp directory, use a symlink on the master.
02:35 aberdine left #salt
02:35 DEger joined #salt
02:35 mohae joined #salt
02:36 my10c ok! i was under inpression from the doc i could do it uder any dir :)
02:36 my10c dank!
02:36 hemebond "Provide a different path and directory name inside the master's minion files cachedir"
02:37 hemebond You can change the master cache dir.
02:37 hemebond But that's a different thing.
02:37 hemebond And would affect all caching.
02:37 hemebond *all minion caching
02:37 my10c yup that would be an ugly solution : but for what i need that would work and later undo it...
02:37 my10c yeah :-(
02:38 hemebond If you create a symlink ../minions/xxx/tmp -> /tmp
02:38 hemebond Maybe that would work.
02:38 my10c yeah but i have TONS of minion :)
02:38 my10c i build a script to get oit from the cache dir
02:38 my10c thanks!
02:39 hemebond Good luck :-)
02:39 my10c :-)
02:42 debian112 joined #salt
02:49 amcorreia joined #salt
02:59 flowstate joined #salt
03:05 brent_ joined #salt
03:11 aberdine joined #salt
03:21 racooper joined #salt
03:28 treaki_ joined #salt
03:31 amcorreia joined #salt
03:33 zer0def joined #salt
03:42 manji joined #salt
03:51 k_sze[work] joined #salt
03:53 jhauser joined #salt
04:01 flowstate joined #salt
04:05 DEger joined #salt
04:27 subsignal joined #salt
04:34 DEger joined #salt
04:40 keimlink joined #salt
05:00 flowstate joined #salt
05:06 onlyanegg joined #salt
05:13 RobertLaptop joined #salt
05:13 irctc898 joined #salt
05:15 kshlm joined #salt
05:15 rdas joined #salt
05:38 squishypebble joined #salt
05:42 rem5 joined #salt
05:44 citaret joined #salt
05:45 whitenoise joined #salt
05:46 jalaziz joined #salt
05:47 onlyanegg joined #salt
05:58 flowstate joined #salt
06:01 _JZ_ joined #salt
06:14 orion left #salt
06:18 impi joined #salt
06:29 POJO joined #salt
06:31 goldielox joined #salt
06:34 jalaziz joined #salt
06:37 Rumbles joined #salt
06:42 DEger joined #salt
06:44 onlyanegg joined #salt
06:51 kawa2014 joined #salt
06:53 CeBe joined #salt
06:56 krymzon joined #salt
06:57 Sylvain31 joined #salt
06:57 ventris joined #salt
06:59 flowstate joined #salt
07:05 fracklen joined #salt
07:06 Qlawy Any ideas why salt-run manage.alived shows different list than salt-run manage.up?
07:08 fracklen joined #salt
07:08 Inver joined #salt
07:08 hemebond Qlawy: One queries the minions and one does not.
07:09 Electron^- joined #salt
07:09 Sylvain31 Qlawy: did you look at the source code? "locate locate runners/manage"
07:09 Sylvain31 only one locate of course, in order to find the source code
07:11 hemebond Looks like alived is for raet.
07:12 hemebond up runs a ping against the minions.
07:12 ravenx joined #salt
07:12 Sylvain31 "raet" ? was does it means?
07:12 hemebond "was does it means?" what?
07:12 Qlawy what raet means ;)
07:13 hemebond Oh, what does "raet" mean? It's the replacement for zeromq.
07:13 hemebond The new way for minions to connect to the master.
07:13 hemebond Not out yet.
07:13 Qlawy hmm
07:13 iggy which is already replaced by the tornado transport
07:13 hemebond In beta I think.
07:13 Sylvain31 hemebond: thanks, raet is a software name, so.
07:13 hemebond Sylvain31: Yes, name of the software.
07:14 hemebond iggy: Oh really? No zeromq now?
07:14 iggy zeromq isn't going anywhere
07:14 Sylvain31 lol, lot of changes! ;)
07:14 iggy I was saying that raet was already replaced by tornado
07:14 iggy before it ever really got off the ground
07:15 hemebond Ooooooh. What happened? I didn't know it had been dropped already.
07:15 hemebond Is that tornado the existing Tornado python library?
07:15 iggy yes
07:15 iggy the one that's already a dependency for other reasons
07:15 hemebond Wow okay. That's pretty good I suppose, re-using an existing tech.
07:16 Sylvain31 no list comprehension in jinja? {% set customers_account = [ s['CustomerName'] for d,s in sites.items() ] %}
07:17 iggy no
07:17 ravenx joined #salt
07:20 felskrone joined #salt
07:27 Qlawy Hmm... strange thing
07:27 Qlawy I have server in version 2016.3.1
07:28 tho joined #salt
07:28 Qlawy and it looks like that most of my 2016.3.1 clients are not visible in manage.alived
07:28 Qlawy but all of 2015.8.8 are
07:31 jhauser joined #salt
07:35 hemebond Qlawy: in the alived method it just returns true for all 0mq clients :-)
07:35 hemebond # Always return 'present' for 0MQ for now
07:35 hemebond # TODO: implement other states spport for 0MQ
07:36 nagri I have desktop boxes. I want to salt them, install salt-minion on them, make them minion, How should I go about it ?  I just started reading about salt, cant find documentation on this.
07:36 hemebond nagri: saltify
07:37 babilen hemebond, Qlawy: AFAIUI raet is stillborn and has already been replaced by the "tcp transport" based on tornado: https://docs.saltstack.com/en/latest/topics/transports/tcp.html
07:37 nagri not enough documentation on saltify.
07:38 hemebond nagri: What do you mean?
07:38 hemebond My understand is saltify just SSH's onto the minions you list and installs salt-minion.
07:38 babilen Isn't saltify for salt-cloud?
07:39 hemebond It's used as a salt cloud provider.
07:39 nagri this is all I can find about saltify https://docs.saltstack.com/en/2015.8/topics/cloud/saltify.html#getting-started-with-saltify
07:39 hemebond But it doesn't manage or create instances.
07:39 hemebond nagri: That's what I'm reading too.
07:39 nagri cant find how to manage versioning of salt minions
07:39 hemebond What do you mean?
07:40 hemebond How to install a specific version of the minion?
07:40 nagri how do you specify which version of salt minion you install and install only salt-minion on the target machine ?
07:40 nagri yes
07:40 hemebond By default it will only install salt-minion.
07:41 nagri ok
07:41 nagri what about version ?
07:41 hemebond I would assume you could use script_args like other profiles.
07:41 hemebond I'm also assuming it just uses bootstrap-salt like it does with other providers.
07:43 nagri http://paste.debian.net/781151/
07:43 nagri I tried this
07:43 nagri dosent work
07:47 babilen What happens instead?
07:48 hemebond The script_args goes into the profile, no?
07:48 hemebond Yours is in your provider.
07:48 nagri it runs without any error, but nothing gets installed on the target machine.
07:49 nagri Let me try with that.
07:50 subsignal joined #salt
07:52 DEger joined #salt
07:53 k_sze[work] joined #salt
07:56 ivanjaros joined #salt
08:00 mikecmpbll joined #salt
08:00 keimlink joined #salt
08:00 subsignal joined #salt
08:02 SpX joined #salt
08:03 subsignal joined #salt
08:05 krymzon joined #salt
08:05 subsignal joined #salt
08:06 Rumbles joined #salt
08:07 subsignal joined #salt
08:09 Sylvain31 how to place a file on the system once, and not overwrite it as a file.mananged? I want to have an empty block list file with a comment explaining it is not managed. I need to create it once.
08:09 Sylvain31 file.prepend dont create it…
08:09 hemebond Sylvain31: Doesn't file.managed have an option for that?
08:10 hemebond replace : True If set to False and the file already exists, the file will not be modified even if changes would otherwise be made. Permissions and ownership will still be enforced, however.
08:10 hemebond https://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html#salt.states.file.managed
08:11 Sylvain31 hemebond: oh, yes! ;)
08:12 KingJ joined #salt
08:14 lero joined #salt
08:15 deniszh joined #salt
08:18 drkrt joined #salt
08:19 GreatSnoopy joined #salt
08:21 mikecmpb_ joined #salt
08:23 fredvd joined #salt
08:28 RuriRyan joined #salt
08:31 kshlm joined #salt
08:31 keys joined #salt
08:35 traph joined #salt
08:35 traph joined #salt
08:41 sud0x3 joined #salt
08:43 emaninpa joined #salt
08:48 nagri hemebond: After the whole process test.ping still dosent work.
08:48 nagri I can see /usr/bin/python /usr/bin/salt-minion running on the target machine
08:49 hemebond does salt-key --list-all show the minion id?
08:49 nagri but salt-minion is not installed as a service
08:49 subsignal joined #salt
08:49 nagri yes it does.
08:49 nagri as Accepted
08:49 hemebond Not installed as a service.... hmm.
08:50 hemebond What OS is the VM?
08:50 nagri Ubuntu 12.04 LTS precise
08:53 fracklen joined #salt
08:53 hemebond Can you telnet from the minion to the master on 4505 and 4506?
08:53 hemebond What version is your master and what version minion are you installing (v2015.8.3?)?
08:54 hemebond *v2015.5.3
08:57 nagri salt-master 2015.5.3 (Lithium) and trying to install 2015.5.3 minion
08:58 hemebond can you do salt-call test.version on the minion?
08:59 ravenx how can i pass two pillars in the command line?
08:59 ravenx i have;  pillar='{"branch":{"git_branch": "hotfix-v5.57.4"}, "asdf":{"test_test": "values"}} but it doesn't seem to work
08:59 harkx hi, besides having your sls files, how does one document his "salt environment" ? just text files/wiki or is there a better way?
09:02 hemebond nagri: Also, when you say not installed as a service, what do you mean? There's no file in /etc/init.d/ or /etc/init/?
09:02 nagri so salt-minion command is not present on the minion.
09:03 hemebond How can that be if it's running as a process?
09:03 nagri i can see two process `/usr/bin/python /usr/bin/salt-minion` when I do ps aux
09:03 hemebond Two? That's a problem.
09:03 nagri but cant use any salt command
09:04 hemebond So there's no /usr/bin/salt-call ?
09:04 nagri nope
09:05 hemebond Can you paste your final provider and profile configs?
09:05 nagri ok
09:05 hemebond I just checked the source for saltify and it does just bootstrap the minion like the other providers.
09:05 hemebond I'll startup a VM and test it myself.
09:07 nagri http://paste.debian.net/781181/
09:07 kshlm joined #salt
09:07 nagri here you go http://paste.debian.net/781181/
09:13 nagri So I re ran the whole thing after destroying the box.
09:13 nagri now I have salt-api     salt-cloud   salt-key     salt-minion  salt-ssh     salt-unity salt-call    salt-cp      salt-master  salt-run     salt-syndic on the target machine
09:13 nagri but still test.ping not working
09:13 hemebond Ah. Was the VM one you had already tried testing with?
09:13 hemebond Key accepted?
09:13 nagri yes
09:13 nagri Accepted keys
09:14 hemebond does it show up in the list for salt-run manage.up?
09:14 hemebond Did you delete the key when you destroyed the VM?
09:14 nagri No
09:14 nagri yes Idid
09:14 nagri i did delete the keys
09:14 hemebond Okay, some strange connection issue maybe.
09:15 hemebond Can you telnet from the VM to the master on 4505 and 4506?
09:15 nagri and it does not show up in manage.up
09:15 nagri can telnet on both the ports from minion to master
09:19 hemebond What does salt-call test.version show?
09:20 hemebond (on the VM)
09:24 hemebond I just used saltify to install salt on a VM; works fine.
09:24 hemebond I didn't specify a git version though.
09:24 hemebond I have had an issue with a git version before.
09:30 Qlawy strange thing, exact same result of salt-minion --versions-report but one minion presence change is visible while another one is not
09:35 Sylvain31 can I change state_verbose:True, when I call from command line?
09:36 hemebond Sylvain31: https://docs.saltstack.com/en/latest/ref/cli/salt.html
09:36 hemebond --state-output
09:36 hemebond --verbose
09:37 hemebond nagri: What happens if you run the bootstrap script manually on a new VM with your arguments?
09:51 wolfpackmars2 joined #salt
09:51 VSpike I'm having one of those days where you start looking at one weird problem, and stumble across another. So you start looking at that one, and then something else strange starts happening
09:52 VSpike Either they are all related, or it's just really bad luck :/
09:53 VSpike Anyway, why would it be that doing `salt 'foo' pillar.get users` from the master returns different (stale) data to doing `salt-call pillar.get users` on the minion "foo"
09:55 VSpike No, me either :/
09:56 Sylvain31 hemebond: something like that, but it didn't works… I would like to have: state_verbose: False and state_output: full by default, and to be able to specify on the command line as if I chanced state_verbose: True (all state in green unchanged): when I do: salt --someoption-to-force-state_verbose: True 'web*' state.highstate
09:56 hemebond Sylvain31: The arguments go at the end of the line.
09:58 pfallenop joined #salt
10:02 hemebond or maybe not, according to that page I linked to.
10:03 Sylvain31 no, and test reveal no neighter, can be at start, after targeting, or at the end
10:04 hemebond Yeah, doesn't seem to be working for me either.
10:04 Sylvain31 but, I cant achieve the expected result :\
10:04 krymzon joined #salt
10:04 Sylvain31 modifying /etc/salt/master + restart do it… a bit long.
10:05 Sylvain31 --verbose seems not to be involved.
10:06 Sylvain31 --state-output=full dont change the behavior. only state_verbose: False do it for me.
10:06 Sylvain31 salt 2015.8.10 (Beryllium)
10:06 hemebond Yeah, state-output seems to do nothing.
10:07 hemebond At least on the command line.
10:07 hemebond In the master config it works.
10:07 Sylvain31 OK, thanks for your hint, I will check the code later do diagnoze.
10:09 hemebond Oh...
10:09 hemebond state_verbose: True in the master conf
10:10 hemebond and then --state-output=changes on the CLI call works as expected.
10:10 Sylvain31 is not the reverse? full all the time and changes with the command switch?
10:11 hemebond Some logic problem or problem with my expectations.
10:11 Sylvain31 I wanted the contrary
10:11 hemebond Well it depends what I out as state_output in the master config.
10:11 hemebond That will be the default unless I override on the CLI.
10:11 hemebond But it doesn't work at all unless I have state_verbose: True
10:12 hemebond Wait. --state-output might not be working at all.
10:12 VSpike On the master, salt-run cache.pillar aws-dev-p-db-1 produces the expected output for that minion
10:12 hemebond No, it seems to work.
10:13 VSpike Also salt 'aws-dev-p-db-1' pillar.items gives the right output
10:13 VSpike But salt 'aws-dev-p-db-1' pillar.get users gives the wrong output
10:14 hemebond VSpike: saltutil.refresh_pillar ?
10:16 VSpike hemebond: hmm, that worked. Wonder why, when salt-run cache.pillar aws-dev-p-db-1 didn't?
10:16 hemebond Oh. Hmm, very strange.
10:26 totte joined #salt
10:27 nagri on minion
10:27 nagri salt-call test.version  Failed to create directory path "/etc/salt/pki/minion" - [Errno 13] Permission denied: '/etc/salt/pki/minion'
10:27 hemebond Are you logged in as root?
10:27 nagri on master
10:27 nagri salt-call test.version [ERROR   ] The Salt Master has cached the public key for this node, this salt minion will wait for 10 seconds before attempting to re-authenticate Minion failed to authenticate with the master, has the minion key been accepted?
10:27 nagri yes
10:28 cyborg-one joined #salt
10:28 nagri no
10:28 nagri salt-call test.version  local:     2015.5.3
10:28 nagri on minion
10:28 nagri when as root
10:28 nagri but all this gets installed on minion
10:28 nagri salt-api     salt-cloud   salt-key     salt-minion  salt-ssh     salt-unity salt-call    salt-cp      salt-master  salt-run     salt-syndic
10:29 nagri and test.ping still not working
10:29 nagri it says
10:29 nagri minion-01:     Minion did not return. [No response]
10:29 hemebond You always have to run commands as root.
10:29 hemebond Okay. Stop the minion.
10:29 hemebond Delete the key on the master (let me know if you need the command for that)
10:29 hemebond Start the minion
10:29 hemebond Accept the key
10:30 hemebond Run salt 'minion' test.ping on master
10:35 jakshi joined #salt
10:36 netcho joined #salt
10:36 netcho joined #salt
10:36 netcho joined #salt
10:38 jakshi Hello All. What is the best way in saltstack to reload service on config files changing only if it's already running and do nothing if service is not running?
10:39 hemebond jakshi: Probably a cmd.run with an onlyif test.
10:39 jakshi @hemebond thank you. I'll try that.
10:39 hemebond Or a module.run + service.restart with an onlyif
10:39 hemebond ^ probably better than a cmd.run
10:41 jakshi @hemebond thank you, I'll try second one then :)
10:41 colegatron_origi joined #salt
10:45 * jakshi after rtfm: can be even better with salt.states.module.wait + service.restart with an onlyif
10:46 babilen jakshi: Would it normally reload services that are not running ? What about those that have been disabled ?
10:46 nagri did exactly what you said
10:46 nagri still minion-01:     Minion did not return. [No response]
10:47 hemebond nagri: You'll need to put the minion into debug mode and restart it to read the log file then.
10:47 babilen nagri: How long did you wait after accepting the key?
10:47 hemebond I can't tell what's happening on it.
10:47 nagri 2-3 min
10:47 babilen That should be long enough
10:50 jakshi @babilen if I add onlyif service is running - it shouldn't reload service if it's not running
10:50 jakshi I think so at least
10:50 nagri minion in debugmode
10:50 nagri stuck here
10:50 nagri http://paste.debian.net/781212/
10:51 hemebond nagri: Is that a clean log?
10:51 teryx510 joined #salt
10:53 Ard_ joined #salt
10:53 Ard_ Hi!
10:54 Ard_ Greetings,   I'm trying to automatize a bareos deploy on several minions, I'll need 4 states due to 4 backup types. Wondering about how can i make both configurations (Client and Director)... Here are my ideas:  1. Make a trigger State, as soon i launch bareos-fd.conf it 'll include Bareos-Dir.conf with client job.  2. Make a Mine State,  launching bareos-fd.conf on client and using son custom grain to make  a pillar for each backup. The
10:54 KermitTheFragger joined #salt
10:54 Ard_ I'm seeking across in internet but don't have any idea to make possible the first one.  Could it be possible? Or maybe isn't possible to push different host with one state?.  Any help 'll be appreciated.  P.S: i'm new in saltstack, it's evident? xD
10:57 teryx5101 joined #salt
10:58 flowstate joined #salt
11:00 hemebond nagri: That minion actually looks fine.
11:01 nagri yes its clean
11:02 nagri I just pressed enter a lot before accepting the key on the master
11:02 hemebond Do you see anything in the minion log when you issue a command from the master?
11:02 nagri nothing
11:02 nagri nothing happens on the minion
11:02 hemebond Okay, time to put the master into debug and watch its log :-)
11:07 nagri master logs
11:07 nagri http://paste.debian.net/781215/
11:08 nagri when I accept the key and then issue a test.ping from master
11:08 nagri minion logs stay the same
11:09 hemebond "sudo_vagrant" < what's that?
11:10 hemebond Your master a vagrant VM too?
11:11 nagri http://paste.debian.net/781216/
11:11 nagri yes
11:11 nagri is that a problem ?
11:11 hemebond Not if you're sudo'd and salt is running as root.
11:12 hemebond *salt-master is running as root.
11:12 nagri yes salt is running as root
11:14 Harris joined #salt
11:14 hemebond if you do a test.ping right now, still fails?
11:14 nagri if I do test.ping on master the log gets appended. but on minion logs dont change
11:14 nagri and on master I get minion-01:     Minion did not return. [No response]
11:15 nagri so the test.ping command is not reaching minion
11:15 nagri i think
11:15 hemebond Well, the minion picks it up from the master.
11:15 hemebond The fact that the minion is authenticated and the key accepted means connectivity is fine.
11:15 hemebond Have you changed the master or minion configs at all?
11:16 Harris Think you can change minoin config not to accept ping
11:16 nagri i havent
11:16 hemebond No other VMs running, correct?
11:16 nagri yes
11:16 nagri none
11:17 hemebond test.version times out too?
11:18 Inver nagri, I'm wondering if there's a hanging salt-master process in the background, tried doing a 'vagrant reload' for your master and minion?
11:18 nagri test.version, grain.items both timed out
11:18 nagri minion I did.
11:19 hemebond What about running on the minion "netstat -anp | grep '192.168.33.10'" ?
11:19 nagri let me do that for master too
11:19 hemebond That netstat is just a bash command, btw, not run via Salt.
11:19 nagri root@minion-01:/home/vagrant# netstat -anp | grep '192.168.33.10' tcp        0      0 192.168.33.11:39195     192.168.33.10:4506      ESTABLISHED 7278/python      tcp        0      0 192.168.33.11:55219     192.168.33.10:4505      ESTABLISHED 7278/python
11:20 babilen nagri: Did you try restarting the master and mnion?
11:20 hemebond Like inver suggested, stop the master, check that all processes are gone.
11:20 babilen And there are often left over processes
11:20 nagri doing that now
11:21 XenophonF joined #salt
11:21 amcorreia joined #salt
11:22 nagri reloaded both master and minion
11:22 nagri still all the same
11:22 nagri :(
11:22 nagri same timeout
11:22 babilen Okay, stop both and check if there are still salt-master or salt-minion processes
11:22 babilen Which init system do you use?
11:23 nagri what do you mean by init system ?
11:23 babilen (and which salt versions do you have on the master and minion?)
11:23 babilen nagri: Is it SystemD or sysvinit or upstart or ...
11:23 hemebond babilen: upstart probably.
11:23 hemebond Ubuntu 12 I think.
11:23 nagri root@minion-01:/home/vagrant# salt-minion --version salt-minion 2015.5.3 (Lithium)
11:23 nagri root@master:/home/vagrant# salt-master --version salt-master 2015.5.3 (Lithium)
11:23 babilen That's ancient
11:24 kevinquinnyo1 joined #salt
11:25 nagri so should I try with newer versions?
11:26 babilen Have you stopped the processes and checked that there are no leftovers?
11:26 hemebond There was a reason for using an old version, yeah?
11:26 Inver nagri, are you using vagrants built-in salt bootstrap?
11:27 nagri yes. stoped everything
11:27 nagri no process running
11:27 babilen okay
11:27 babilen One question down, two to go :)
11:28 nagri what are the other questions ?
11:29 babilen "There was a reason for using an old version, yeah?" -- " nagri, are you using vagrants built-in salt bootstrap?"
11:29 Inver :)
11:29 nagri yes. and yes.
11:30 nagri we have a dependency on that version
11:30 babilen How come?
11:30 nagri our current master in production is that version
11:30 nagri minion installation is not automated
11:30 nagri thats why testing with this version
11:31 babilen You might want to try current versions in vagrant, just to check if they behave differently
11:31 Inver how much memory are you assigning the VMs, as I recall vagrant only assigns 512mb, I've had trouble with too few resources in the past
11:32 hemebond nagri: Question about the minion; when you said that nothing was in the minion log, did you mean nothing or nothing new when you issue the command?
11:33 nagri nothing new
11:33 nagri memory is 400 some MB
11:33 babilen That really isn't much
11:33 nagri http://paste.debian.net/781212/
11:33 nagri ^ minion log
11:34 nagri minion never makes past this.
11:34 nagri let me allocate more memory
11:34 Inver cool beans
11:36 hemebond 2016.3.1 uses zeromq still.
11:36 hemebond According to the minion log.
11:38 XenophonF zeromq4 iirc
11:39 nagri now its 1 GB
11:39 nagri but still the same
11:39 nagri times out
11:39 hemebond What if you telnet from the minion VM to the master VM and just type in garbage?
11:40 yomateo joined #salt
11:40 nagri on port 4505 ?
11:40 hemebond Yeah
11:40 hemebond and/or 4506
11:41 nagri it lets me
11:41 nagri I can
11:42 hemebond So strange.
11:42 nagri master gets the minion key so I think connection should not be the issue
11:42 hemebond This has happened to me, I think, and it was a leftover master process.
11:42 hemebond Yeah.
11:42 nagri but I reloaded the vagrant box thrice
11:43 hemebond I can't think of anything else.
11:43 nagri master I mean
11:43 nagri and minion too
11:43 nagri does this mean its time I sit in a corner and cry ?
11:43 hemebond I suspect so :-(
11:43 nagri :'(
11:44 hemebond Hmm. What if you install a minion on the master?
11:44 Inver damn, I felt pretty confident it might have been lack of memory
11:44 nagri master already has a minion. but that I installed manually
11:45 nagri I can try uninstalling and then reinstalling with the salt-cloud
11:45 nagri Inver: I would have been so happy if it was just memory issue.
11:52 hemebond How can we get the pem fingerprint on the minion?
11:53 hemebond nagri: salt-key --list-all lists only minion-01, correct?
11:53 numkem joined #salt
11:53 DEger joined #salt
11:54 nagri master too
11:55 nagri master's minion key was also accepted
11:55 hemebond Okay, that's cool.
11:55 hemebond On master:$ sha256sum /etc/salt/pki/master/minions/minion-01
11:56 hemebond On minion:$ sha256sum /etc/salt/pki/minion/minion.pub
11:56 hemebond Do they match?
11:57 nagri 11f288138e28bb7e0db5f29e7fd85cebc0a7bd3af525e071f8f8408ffddc9a92  /etc/salt/pki/master/minions/minion-01
11:57 nagri 11f288138e28bb7e0db5f29e7fd85cebc0a7bd3af525e071f8f8408ffddc9a92  /etc/salt/pki/minion/minion.pub
11:57 nagri yes they do
12:01 kpostrup joined #salt
12:01 kpostrup Hello, I have created a new saltstack formula. Who do I contact?
12:03 hemebond nagri: How did you install the master and the minion?
12:04 hemebond nagri: Stop the master and minion services, check that all processes are gone.
12:04 nagri manually
12:04 nagri ok
12:04 nagri then ?
12:04 hemebond On the master run "salt-master -l debug"
12:05 nagri ok
12:05 hemebond On the minion run "salt-minion -l debug"
12:05 nagri ok
12:05 hemebond You should see stuff on the master when the minion connects. Yeah?
12:06 Harris joined #salt
12:06 Harris In order for the orchestrator to perform tasks we need some pillars - can those be defined by 'saltenv' somehow? So far we can only use 'base' pillars for orchestrator
12:06 Harris I want the orchestrator to know which nodes to run stuff to, and that changes from environment to environment
12:06 nagri yes I see
12:07 nagri hemebond: I do see
12:07 hemebond Okay, new session on the master, issue the test.version command
12:08 nagri logs of master get updated but nothing on minion.
12:08 nagri and it times out
12:08 Inver kpostrup: Not sure, but you could possibly ask to have your formula forked into https://github.com/saltstack-formulas, by asking on https://github.com/saltstack/salt/issues
12:11 hemebond nagri: How did you install the master and minion?
12:12 babilen (and did you try a newer version by now?)
12:12 hemebond salt 'minion-01' test.ping -t 300
12:13 nagri hemebond: I followed this https://www.digitalocean.com/community/tutorials/how-to-install-and-configure-salt-master-and-minion-servers-on-ubuntu-14-04
12:13 babilen nagri: Could you try http://repo.saltstack.com/#ubuntu
12:14 babilen Pesky "third-party tutorials"
12:14 nagri alright.
12:14 hemebond nagri: Did you put master_finger into your minion config?
12:16 teryx510 joined #salt
12:16 nagri shouldnt that be done by salt-cloud ?
12:16 hemebond No.
12:16 hemebond But it isn't required for testing.
12:16 hemebond So I just wanted to check.
12:17 nagri let me put it in and then test again
12:17 hemebond No, it's not necessary.
12:19 hemebond I thought you had installed using the bootstrap script.
12:19 teryx5101 joined #salt
12:19 TooLmaN joined #salt
12:19 kpostrup Inver: ok, I will do that. Thanks
12:21 pcdummy joined #salt
12:22 nagri hemebond: Did the finger thing still no change
12:22 babilen nagri: This is still with 2015.3 isn't it?
12:22 hemebond nagri: Yeah, it's not required.
12:24 nagri yes
12:24 nagri I'll follow this http://repo.saltstack.com/#ubuntu and will bother you guys again
12:25 nagri I have given up on life for today :(
12:25 babilen heh
12:25 nagri hemebond: Thanks
12:25 nagri babilen: Thanks
12:25 babilen Keep us posted
12:26 nagri I will.
12:26 hemebond Last time someone had an issue like this they were trying to run as a restricted user. But everything here seems fine :-(
12:26 hemebond Good luck :-)
12:26 babilen I wouldn't be surprised if it is due to that ancient version
12:34 mikecmpb_ joined #salt
12:34 krymzon joined #salt
12:37 yidhra joined #salt
12:53 teryx510 joined #salt
12:56 zirpu joined #salt
12:58 subsignal joined #salt
13:01 teryx5101 joined #salt
13:02 subsignal joined #salt
13:03 pid1 joined #salt
13:04 squishypebble joined #salt
13:08 subsignal joined #salt
13:08 totte left #salt
13:10 subsigna_ joined #salt
13:11 emaninpa joined #salt
13:23 mapu joined #salt
13:23 fracklen joined #salt
13:25 iceyao joined #salt
13:26 deus_ex joined #salt
13:28 racooper joined #salt
13:30 iceyao_ joined #salt
13:30 flowstate joined #salt
13:31 martoss joined #salt
13:31 rem5 joined #salt
13:31 west575 joined #salt
13:35 patrek joined #salt
13:36 flowstate joined #salt
13:39 JohnnRun joined #salt
13:42 catpig joined #salt
13:43 dyasny joined #salt
13:49 perfectsine joined #salt
13:50 nyx joined #salt
13:52 iceyao joined #salt
13:53 gngsk joined #salt
13:57 ajv joined #salt
14:06 flowstate joined #salt
14:06 iceyao_ joined #salt
14:09 lightus joined #salt
14:10 aphor whoa. I guess everyone's showing up at work 9AM CDT
14:10 fredvd joined #salt
14:11 CTaylor joined #salt
14:11 gngsk_ joined #salt
14:15 renoirb joined #salt
14:17 blw0rm joined #salt
14:24 iceyao joined #salt
14:31 POJO joined #salt
14:35 brent_ joined #salt
14:36 pcdummy joined #salt
14:37 corichar joined #salt
14:44 zmalone joined #salt
14:51 fracklen joined #salt
14:53 kstewart joined #salt
14:54 kstewart Good morning!  I am reading the documentation about using xmpp in states, but I feel like there's a piece of documentation I should also read, but I'm unsure what it is.
14:54 kstewart https://docs.saltstack.com/en/latest/ref/states/all/salt.states.xmpp.html  -- where does one define the profile that gets referenced here?
14:55 flowstate joined #salt
15:03 iceyao joined #salt
15:03 DX1 joined #salt
15:03 anotherZero joined #salt
15:09 bastiandg joined #salt
15:17 bastiandg joined #salt
15:17 aphor kstewart: I suspect it's borrowing code from the returner https://docs.saltstack.com/en/latest/ref/returners/all/salt.returners.xmpp_return.html
15:18 ALLmightySPIFF left #salt
15:22 iggy the state modules also generally are tied to the execution modules, so that's a good place to look too
15:22 iggy !salt modules.xmpp
15:22 saltstackbot https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.xmpp.html
15:22 iggy yeah, it's on that page
15:24 pseudonyma joined #salt
15:26 fracklen joined #salt
15:28 pseudonyma Dear all, I have a quick question regarding the interaction of pkgrepo.managed and pkg.installed for an apt-based system. I'm grateful if you can help me sort out the issue that pkg.installed does not install usb-modeswitch from jessie-backports despite using fromrepo. What am I doing wrong? The example SLS file is available here: http://pastebin.com/7CK0qRPW
15:29 iggy some of your indentation is off
15:30 PuppyPoker joined #salt
15:30 iggy or it might just be that pastebin.com is broken (which is why we don't use it... well that and the terrible ads)
15:30 aphor who loves some salt-cloud?
15:31 iggy and your pkgrepo.managed should have a require_in on pkg: mobile_broadband
15:31 aphor I'm about to work on something that looks like it goes waaaay back to 2014 vintage.
15:31 onlyanegg joined #salt
15:31 cyborg-one joined #salt
15:31 aphor https://github.com/saltstack/salt/issues/34669
15:31 saltstackbot [#34669][OPEN] sad error detection/logging/handling in salt.cloud when driver does not implement fun | Description of Issue/Question...
15:31 iggy that's like the 60's in salt time
15:31 pseudonyma Thanks for the suggestions. The indentation issue is due to pastebin.
15:32 aphor pseudonyma: try doing 'salt-call --loglevel=all state.apply' on the minion to get under the hood.
15:34 aphor So.. salt.cloud.stuff has a lot of 'if fun not in self.clouds: <log an error>'
15:34 pseudonyma aphor: It's --log-level, but thanks! I hadn't known about that!
15:34 aphor pseudonyma: correct, sorry about my typing.
15:35 * aphor suffering a migraine
15:35 Harris joined #salt
15:35 Harris Can you use custom execution modules on orchestrator??
15:35 pseudonyma aphor: No worries :) By the way, where would you recommend that I post the output of salt-call? Anywhere better than pastebin?
15:37 aphor pseudonyma: pastebin is OK, but try to *read* the debug/trace logging and understand what's happening.
15:37 aphor .. give a man a fish/teach a man to fish ..
15:39 pseudonyma aphor: Doing just that... The log is huge.
15:42 DEger joined #salt
15:42 iggy there is an alternative to pastebin in the /topic (but also refheap.com paste.debian.net etc...)
15:42 iggy gist is the tits though... multiple file per paste is the killer feature
15:42 onlyanegg joined #salt
15:43 aphor +1 gist
15:45 s_kunk joined #salt
15:45 pseudonyma iggy: perfect, thanks! The missing require directive seems to have fixed the issue. At least now, the correct packages get installed. By the way; is there a general preference of using a single state ID for multiple state directives or one ID per directive?
15:48 iggy personal preference
15:48 pseudonyma Thanks so much for your help, aphor and iggy. Saltstack is an incredible piece of software!
15:49 gtmanfred :)
15:55 debian112 joined #salt
16:04 ivanjaros joined #salt
16:04 briansteffens joined #salt
16:04 tapoxi joined #salt
16:06 tapoxi hi, anyone know what "Exception AttributeError: "'NoneType' object has no attribute 'close'" in <bound method IPCMessageSubscriber.__del__ of <salt.transport.ipc.IPCMessageSubscriber object at 0x206a190>> ignored" means? I'm getting it in my dev environment after running commands through salt
16:06 krymzon joined #salt
16:07 POJO joined #salt
16:07 Jaremy joined #salt
16:09 iggy did you search for issues?
16:13 brent_ joined #salt
16:14 mohae joined #salt
16:18 snaggleb joined #salt
16:20 teryx510 joined #salt
16:26 wryfi i have a question about external auth
16:26 wryfi i'm setting up the chrerrypy rest api for my salt masters
16:27 wryfi so that i can integrate salt with a couple of external systems
16:27 wryfi yet the only auth options seem to be PAM and LDAP
16:27 wryfi i'd rather not use either of these
16:28 wryfi is there no way to set up a "virtual" user or "machine account" for salt to auth against?
16:32 Jaremy Not sure if this answers your question but we are using a local user on the salt master to auth, username and password in the api call without problems.
16:32 Jaremy We have since hashed the password and set it to an HTTPS connection for security.
16:33 wryfi yeah, i have a fundamental issue with creating what is essentially a system user, but giving it the ability to log at a console on with a password
16:33 wryfi not really a good security model
16:33 wryfi but it seems like that is the only way
16:33 CimmerianX joined #salt
16:33 Jaremy Nope not at all. Just depends where you calls our coming from as well. External or internal network and so on, but i have yet to find another way.
16:34 wryfi does the api user need a shell?
16:35 wryfi Jaremy: do you happen to know?
16:35 wryfi i'm guessing not?
16:35 Jaremy What we ended up doing was going in making adjustments to what commands and calls the api was allowed to make and accept in attempts to add some security. cause cmd.run_all 'sudo rm -rf /' could be a bad day across minions.
16:35 Jaremy no it does not.
16:36 wryfi thanks Jaremy
16:36 Ard_ joined #salt
16:37 CimmerianX Guys - I'm trying to use salt-cloud to spin up Windows hosts in EC2.     I'm at a point where I'm getting an error during the spin up..... "Attempting WinRM connectio to host 10.10.24.176 on port 5986"  then There was a query error:  global name 'winRMTransportError' is not defined.
16:37 tapoxi joined #salt
16:38 CimmerianX Can someone give me a push in the right direction here?    Not sure where to look.
16:39 aw110f joined #salt
16:39 woodtablet joined #salt
16:40 Ard_ Hi, got one question. Anyone know if can make a state on host A with trigger a second state on Host B ?   I just wanna manual run first state
16:42 ageorgop joined #salt
16:46 Jaremy You could write a custom state in python that would do that but I cannot think of anything using the built in states.
16:54 nyx joined #salt
16:57 manji joined #salt
16:58 iggy wryfi: shell=/bin/false ?
16:58 wryfi iggy: yeah, that's what i went with
16:58 * wryfi still doesn't like the design
16:58 iggy PR's welcome ;)
17:00 lero joined #salt
17:00 tracphil joined #salt
17:01 tracphil Hi all. Is there any way when using gitfs to map the master branch to a particular environment (in my case production)? I don't use a "base environment" for my hosts.
17:13 pcdummy Can i make salt fail hard when a single pillar file fails ?
17:13 pcdummy like on states.
17:14 pcdummy tracphil: yes
17:15 pcdummy tracphil: https://docs.saltstack.com/en/latest/topics/tutorials/gitfs.html
17:17 pcdummy s/single/included/g
17:23 coval3nce joined #salt
17:26 DammitJim joined #salt
17:27 mntnman joined #salt
17:28 west575_ joined #salt
17:29 teryx510 joined #salt
17:34 mikecmpbll joined #salt
17:35 tracphil pcdummy, all I see is "base: salt-base" to change the salt base branch. I don't see how to make master: production
17:37 flowstate joined #salt
17:37 iggy tracphil: base environment is required
17:38 tracphil Right, I don't mind requiring it. I just want to make master map to the production environment.
17:38 fracklen joined #salt
17:40 cyborg-one joined #salt
17:42 iggy no clue... gitfs+environments is black magic voodoo
17:43 onlyanegg joined #salt
17:48 tongpu joined #salt
17:48 ponyofdeath hi, is there a way for pkg.installed: - sources: to check if package is installed before downloading again?
17:49 morissette joined #salt
17:49 tawm04 joined #salt
17:50 wryfi can anybody help me track down an error i'm having with the cherrypy netapi server?
17:50 wryfi when i try to call /login
17:50 wryfi i get a 401 back
17:51 wryfi and in my master log
17:51 wryfi all it says is "[WARNING ] Authentication failure of type "eauth" occurred."
17:51 wryfi and this is in debug mode
17:51 wryfi not sure where to look next
17:53 wryfi oh wait, nevermind. my salt state didn't set the user's password somehow.
17:53 infrmnt joined #salt
17:59 onlyanegg joined #salt
18:00 ruxu joined #salt
18:01 tapoxi joined #salt
18:05 ruxu joined #salt
18:07 gimpy2938 I'm trying to schedule a state to apply but it never runs; do I need to use states.schedule.enabled in addition to states.schedule.present?  What does that look like?  https://gist.github.com/jwhite530/1610d89bf1cf349cb3c82a2c13ef08cf
18:07 writtenoff joined #salt
18:09 felskrone joined #salt
18:10 alvinstarr joined #salt
18:14 sud0x3 joined #salt
18:17 teryx5101 joined #salt
18:23 gngsk_ joined #salt
18:27 tapoxi joined #salt
18:38 gimpy2938 ... does anyone have any working example of states.schedule.present?
18:39 amcorreia joined #salt
18:43 PuppyPok_ joined #salt
18:49 brent_ joined #salt
18:50 fracklen joined #salt
18:50 cliluw joined #salt
18:51 PuppyPoker joined #salt
18:56 nyx joined #salt
18:59 whytewolf gimpy2938: I know i don't since I setup my schedules through piller. not useing the states
19:04 whytewolf gimpy2938: wait that doesn't look like you are using the state.schedule present anyway. thats a pillar with a name of schedule.present
19:05 whytewolf change the test: to schedule:
19:05 fracklen joined #salt
19:06 deniszh joined #salt
19:07 whytewolf gimpy2938: here is an example pillar for sceduler https://gist.github.com/whytewolf/22893a918cf6ee928810
19:08 felskrone joined #salt
19:09 whytewolf if you want to use scedule.present you need to put it in with your states. not your pillars
19:14 felskrone joined #salt
19:17 scoates joined #salt
19:17 mavhq joined #salt
19:20 cableninja_ joined #salt
19:25 ajw0100 joined #salt
19:26 GreatSnoopy joined #salt
19:35 TyrfingMjolnir joined #salt
19:37 mikecmpbll joined #salt
19:37 felskrone joined #salt
19:41 TyrfingMjolnir joined #salt
19:47 cableninja__ joined #salt
19:52 ruxu joined #salt
20:02 gimpy2938 whytewolf: I got further and now I see my state in the minion's schedule but the state isn't being ran (nothing in journald saying it ran) ... https://gist.github.com/jwhite530/1610d89bf1cf349cb3c82a2c13ef08cf
20:02 perfectsine_ joined #salt
20:04 yomateo joined #salt
20:04 gimpy2938 nevermind, looks like it is being ran but Salt logs nothing about it
20:04 whytewolf yes
20:04 ruxu joined #salt
20:08 keimlink_ joined #salt
20:10 keimlink joined #salt
20:11 jalaziz joined #salt
20:15 protoz joined #salt
20:15 ruxu joined #salt
20:15 rlarkin joined #salt
20:17 brent_ joined #salt
20:21 pcdummy joined #salt
20:22 manji joined #salt
20:22 teryx510 joined #salt
20:23 POJO joined #salt
20:30 jalaziz joined #salt
20:36 pocketprotector joined #salt
20:39 perfectsine joined #salt
20:42 v12aml joined #salt
20:47 jalaziz_ joined #salt
20:49 jalaziz joined #salt
20:50 mohae_ joined #salt
20:55 jalaziz joined #salt
20:55 ajw0100 joined #salt
21:04 ivanjaros joined #salt
21:05 bluenemo joined #salt
21:07 pcdummy Howto get my new formula: https://github.com/pcdummy/saltstack-limits-formula into https://github.com/saltstack-formulas/ ?
21:09 iggy gtmanfred: ^
21:11 jalaziz joined #salt
21:11 gtmanfred pcdummy: can you send an email to the salt-users google group, just for documentation, and I will add the formula and add you as a contributor?
21:11 iggy pcdummy: actually, you should grab tamplate-formula and base off that (i.e. different license, rst docs instead of md, etc)
21:11 johnkeates joined #salt
21:11 gtmanfred pcdummy: https://groups.google.com/forum/#!forum/salt-users
21:12 gtmanfred yeah, that too
21:16 gtmanfred pcdummy: it also needs to be relicensed as an apache license
21:17 gtmanfred I think it would also be preferred if it was in yaml instead of #py
21:19 [dee] left #salt
21:23 pcdummy joined #salt
21:23 brotatochip joined #salt
21:24 lero joined #salt
21:28 lero joined #salt
21:29 flowstate joined #salt
21:41 protoz joined #salt
21:43 protoz_ joined #salt
21:44 protoz_ joined #salt
21:56 brotatochip joined #salt
21:56 pcdummy gtmanfred, iggy: ty
21:58 pcdummy gtmanfred: not sure i want yaml instead of #py, well that should be doable with jinja https://github.com/pcdummy/saltstack-limits-formula/blob/master/limits/init.sls#L11
22:00 iggy I'm not hard on the jinja+yaml requirement
22:00 iggy I've wanted to see a #py formula for a while
22:00 iggy but that's just me
22:01 babilen No, it's good
22:11 badon left #salt
22:12 brotatochip joined #salt
22:14 daishan joined #salt
22:19 cableninja_ joined #salt
22:27 TOoSmOotH joined #salt
22:31 ajw0100 joined #salt
22:35 DEger joined #salt
22:38 armguy joined #salt
22:46 debian112 joined #salt
22:48 badon joined #salt
22:48 clevodearia joined #salt
23:00 artemz joined #salt
23:03 snc joined #salt
23:04 cableninja__ joined #salt
23:06 flowstate joined #salt
23:24 kevinquinnyo1 joined #salt
23:34 mort joined #salt
23:34 mort has anyone here run into an issue with pkg.removed not finding packages that are installed ?
23:35 pcdummy joined #salt
23:35 pcdummy joined #salt
23:37 rem5 joined #salt
23:39 mort hitting a weird issue where a 'pkg.removed' is telling me the package I'm trying to remove is already absent... but it's not. This is across several distros.
23:39 johnkeates left #salt
23:41 rem5_ joined #salt
23:45 Eugene Never seen that myself. Random guess, are you using the friendly package name(openssl) or the full package name(openssl.x86_64) ?
23:48 pmcg joined #salt
23:51 brotatochip joined #salt
23:57 mort @Eugene - using the friendly/short name
23:58 mort I figured it out for cent - the arch was appended at the end of the package name in pkg.list_pkgs for some reason
23:58 zero_shane joined #salt
23:58 mort ubuntu doesn't show the packages at all in pkg.pkg_lists
23:58 flowstate joined #salt
23:58 Eugene Yup, that's what I was guessing
23:59 Eugene That sounds like a bug in handling package names
23:59 JPT joined #salt

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