Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-10-28

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

All times shown according to UTC.

Time Nick Message
00:00 crocket joined #salt
00:00 crocket Yay
00:00 crocket Saltstack employees agree with me on https://github.com/saltstack/salt/issues/16917
00:01 crocket While DNS is not a bad approach, salt should be able to work without DNS.
00:06 crocket salt is not going to get away without fixing the issue.
00:09 ioio joined #salt
00:12 aquinas joined #salt
00:13 bhosmer_ joined #salt
00:13 TyrfingMjolnir joined #salt
00:16 jhulten joined #salt
00:30 patarr joined #salt
00:30 patarr joined #salt
00:33 alex-mesos joined #salt
00:35 MTecknology crocket: are you getting worked up over something that is already listed as something that needs to be fixed?
00:35 mdasilva joined #salt
00:36 alex-mesos joined #salt
00:37 micah_chatt joined #salt
00:38 aparsons joined #salt
00:38 kickerdog joined #salt
00:41 jhulten joined #salt
00:42 fragamus joined #salt
00:44 crocket MTecknology, I'm celebrating my own victory.
00:44 MTecknology oh, okay
00:50 vukcrni joined #salt
00:51 ingwaem` joined #salt
00:52 ingwaem joined #salt
00:56 aparsons joined #salt
01:01 aqua^mac joined #salt
01:15 mgw joined #salt
01:25 thayne joined #salt
01:26 harkx joined #salt
01:27 ingwaem joined #salt
01:29 dooshtuRabbit joined #salt
01:34 dooshtuRabbit joined #salt
01:35 ingwaem joined #salt
01:44 ingwaem joined #salt
01:46 crocket When I execute https://bpaste.net/show/03418f82c506 , content is an empty string.
01:51 baniir joined #salt
01:52 TyrfingMjolnir joined #salt
01:53 debian112 joined #salt
01:53 ingwaem crocket: just checking some references for you...
01:54 crocket ingwaem, Don't worry
01:55 crocket ingwaem, I found the error
01:55 ingwaem oh ok
01:55 crocket I had to specify "content" instead of "text".
01:55 ingwaem ahh
02:01 vbabiy joined #salt
02:04 ioio joined #salt
02:08 gmcwhistler joined #salt
02:09 ramishra joined #salt
02:11 thayne joined #salt
02:15 bezeee joined #salt
02:23 ndrei joined #salt
02:23 ndrei_ joined #salt
02:36 tmh1999 joined #salt
02:39 tmh1999 joined #salt
02:40 tf198_ joined #salt
02:41 tmh1999 joined #salt
02:41 n8n joined #salt
02:45 mosen joined #salt
02:54 jhauser_ joined #salt
02:58 borat joined #salt
03:03 ninkotech joined #salt
03:03 malinoff joined #salt
03:09 SheetiS joined #salt
03:11 ninkotech joined #salt
03:13 elfixit joined #salt
03:13 Mso150_f joined #salt
03:16 ioio_ joined #salt
03:40 cads joined #salt
03:51 thayne joined #salt
03:51 StDiluted joined #salt
04:01 dude051 joined #salt
04:04 dooshtuRabbit1 joined #salt
04:07 crocket salt-run is the answer when I need to bootstrap a master.
04:07 troyready joined #salt
04:08 crocket left #salt
04:15 Krrl joined #salt
04:28 TyrfingMjolnir joined #salt
04:35 fragamus joined #salt
04:37 TyrfingMjolnir joined #salt
04:37 meylor joined #salt
04:46 wt joined #salt
04:52 ramishra joined #salt
05:03 shookees joined #salt
05:08 jalaziz joined #salt
05:23 bkrram joined #salt
05:23 bkrram Quick question on the salt scheduler
05:24 bkrram I am trying to schedule the running of a python script every 60 seconds on the master
05:24 bkrram I have this at the end of my master config file :
05:24 bkrram schedule:
05:25 bkrram job1:
05:25 bkrram function: test.ping
05:25 bkrram args:
05:25 bkrram - '/home/bkrram/fractal/integral_view/monitoring/python/generate_status.py'
05:25 bkrram seconds: 60
05:25 bkrram But I keep getting a message "2014-10-28 10:49:57,895 [salt.utils.schedule][INFO    ] Invalid function: job1 in job cmd.run. Ignoring." in the logs
05:25 bkrram Is there anything else that I need to do?
05:26 bkrram Sorrt, that was  "    function: cmd.run" in the config file
05:36 ramishra joined #salt
05:43 jY bkrram: did you verify its valid yaml?
05:52 felskrone joined #salt
05:53 ramteid joined #salt
05:53 jonbrefe joined #salt
06:03 fragamus joined #salt
06:15 bhosmer_ joined #salt
06:15 SheetiS joined #salt
06:17 n8n joined #salt
06:18 ingwaem joined #salt
06:22 ek6 joined #salt
06:23 shookees joined #salt
06:23 shookees joined #salt
06:32 linjan joined #salt
06:36 bkrram ?? Any help with the scheduling would be appreciated!
06:42 CeBe joined #salt
06:42 felskrone joined #salt
06:48 scristian joined #salt
06:53 cruatta joined #salt
06:54 hintss joined #salt
06:59 ramishra joined #salt
07:00 mdasilva joined #salt
07:01 cberndt joined #salt
07:02 AirOnSkin joined #salt
07:04 Pixionus joined #salt
07:07 zlhgo joined #salt
07:13 the_lalelu joined #salt
07:14 n8n joined #salt
07:16 canci joined #salt
07:16 ramishra_ joined #salt
07:18 crocket joined #salt
07:19 crocket Why can't I use /srv/salt/_modules/minion.py as in salt['minion.blah'] in sls files used by salt runner?
07:21 crocket Huh....
07:21 crocket It's not synced?
07:21 crocket damn
07:22 babilen Yes, it helps to do things if you want them done.
07:23 crocket babilen, How can I make salt-run recognize custom modules in /srv/salt/_modules?
07:23 crocket salt-minion recognizes custom modules.
07:24 crocket It seems I have to modify runner_dirs in /etc/salt/master
07:29 ollins joined #salt
07:31 SheetiS crocket: might need to use execution modules folder on your master
07:32 crocket SheetiS, Why?
07:32 crocket SheetiS, How?
07:32 SheetiS One moment I can share
07:34 crocket ok
07:35 crocket It seems if I return a string from a runner, salt-run shows nothing.
07:35 SheetiS extension_modules: /srv/modules  <== this command in your /etc/salt/master then put your module in /srv/modules/modules since salt-run is run from the master context.  See http://docs.saltstack.com/en/latest/ref/configuration/master.html#extension-modules for more info
07:37 crocket SheetiS, I think salt-run can't run execution modules.
07:37 crocket It can't run execution modules.
07:37 SheetiS if it is a runner
07:37 SheetiS then /srv/modules/runners
07:37 crocket ok
07:37 SheetiS or whatever
07:37 crocket SheetiS, How did you find it?
07:38 crocket SheetiS, In my case, it is set to /var/cache/salt/master/extmods
07:38 crocket It's actually a cache.
07:38 SheetiS hmm
07:40 crocket SheetiS, What does 'salt' refer to in runner sls files?
07:40 crocket A bunch of execution modules?
07:40 crocket or runner modules?
07:41 crocket ok
07:41 crocket salt is not defined in runner modules.
07:42 crocket it can be imported...
07:46 SheetiS Maybe taking a look at how the orchestrate runner does it might give some insight say this: https://github.com/saltstack/salt/blob/develop/salt/runners/state.py#L96-L105
07:47 SheetiS if you need to run a function of some type
07:47 SheetiS not sure what you are needing to do
07:47 crocket SheetiS, I want to read the content of /etc/salt/minion_id
07:47 SheetiS from a targeted minion?
07:48 crocket SheetiS, no from the master
07:48 crocket Since salt-run runs state.sls on the master, it's from the master.
07:48 crocket I want to orchestrate the master.
07:50 lcavassa joined #salt
07:50 SheetiS an orchestration usually implies multiple systems.  Is that your intent?
07:50 SheetiS I'm thinking maybe you want to run your master as a minion as well
07:50 SheetiS and then handle it that way
07:51 crocket SheetiS, Actually, I want to modify /etc/salt/minion_id and /srv/pillar/top.sls
07:51 crocket I am not sure if salt-minion is suitable for that.
07:52 crocket After they are modififed, salt-minion needs to be restarted, which can't be done in the middle of state executions.
07:52 crocket Oh, I could abandon the concept of modifying them.
07:53 crocket They could be set in stone.
07:56 crocket SheetiS, I just want to modify files on the master.
07:56 crocket SheetiS, I just want to modify files on the master, which would require salt-minion to be restarted.
07:58 SheetiS well to 'modify files on the master', the minion is the right way to do it.  you could do it as an orcestration from the master (which would still require the minion) as a 2 parter... 1.) make your changes statefully 2.) restart the minion
07:59 SheetiS but it would still require the minion running and utilize the minion on the master to handle the changes
07:59 crocket SheetiS, state.orchestrate requires only salt-master.
07:59 crocket runner modules don't require minion.
07:59 trikke joined #salt
07:59 crocket SheetiS, The listen requisite would make it possible..., but I don't know.
08:00 crocket listen requisite is not in 2014.1
08:00 crocket Plus, if salt-minion is restarted, salt '*' blah.blah would fail.
08:09 bagackiz joined #salt
08:09 chiui joined #salt
08:13 duncanmv joined #salt
08:13 akafred joined #salt
08:13 crocket It feel bad.
08:14 crocket SheetiS, Do you know how to make salt wait until minion registers a key to the master?
08:15 dunz0r joined #salt
08:16 MohShami joined #salt
08:16 zsoftich joined #salt
08:17 MohShami hey guys, can I get a minion to call cmd.run on all other minions?
08:17 n8n joined #salt
08:18 fbraad joined #salt
08:20 superted666 MohShami: By default no, you can however do this through peering. http://docs.saltstack.com/en/latest/ref/peer.html
08:20 MohShami superted666: Thanks mate, checking
08:20 superted666 MohShami: Just needs enabling, it's called in a slightly different way but means you can execute commands without being on the master.
08:21 MohShami I'm thinking of building a monitoring system with salt, so this will be helpful
08:21 MohShami thanks mate :)
08:22 superted666 No probs, be interested in what you come up with. I've read online around a few people replacing NRPE in nagios with salt. Interesting idea
08:22 alex-mesos joined #salt
08:23 MohShami I've been using nagios for years, but lately I started to hate it, so started looking around and found sensu
08:23 MohShami the problem with sensu is that it was written in ruby
08:23 MohShami and needs an agent
08:24 superted666 You building from scratch then or just replacing the agent -> server communication layer?
08:24 MohShami haven't thought about reusing parts of sensu
08:24 MohShami interesting idea
08:24 MohShami to me, salt uses a much more optimized backend
08:24 MohShami why not use that?
08:25 MohShami ZMQ vs rabbitMQ
08:25 superted666 True, only mentioned as i've come across this before.. http://russell.ballestrini.net/replace-the-nagios-scheduler-and-nrpe-with-salt-stack/
08:25 crocket Oh, there is reactor system.
08:25 crocket How can I register an event function for minion key registration?
08:26 superted666 If you end up using salt and the system is very large, might be worth looking at trying this instead of ZMQ http://docs.saltstack.com/en/latest/topics/development/raet/index.html
08:26 badon joined #salt
08:27 MohShami checking
08:27 crocket How can I specify a runner state for accepting a specific minion key when it becomes available?
08:27 alex-mesos joined #salt
08:27 slav0nic joined #salt
08:27 crocket salt-minion just registered, and the master needs to wait until it is registered and accept it.
08:28 crocket Should I preseed?
08:29 Rucknar crocket: you can auto-accept keys based on whitelist (supports regex)
08:29 crocket Rucknar, how?
08:30 irctc413 joined #salt
08:30 crocket Rucknar, I think http://docs.saltstack.com/en/latest/topics/event/master_events.html#authentication-events is more useful.
08:31 crocket Rucknar, Can you refer me to the part of documentations about auto-accept keys?
08:31 MohShami superted666: 10s of thousands of servers? My biggest environment was less than 100 servers :D
08:31 irctc413 how to confifure .~/dockercfg to pull from private repos?
08:31 Rucknar Yep - gimmie 2 ticks, just digging it out.
08:32 irctc413 when i pull  from private repo its showing me 404
08:32 irctc413 when i start salt-minion in debug mode it works
08:33 Rucknar crocket: Check out here : http://docs.saltstack.com/en/latest/ref/configuration/master.html it covers auto_accept / autosign_file and autorehect_file
08:33 Rucknar *reject
08:33 irctc413 but how to provide docker authentication?
08:34 crocket Rucknar, It seems to require changes to /etc/salt/master
08:34 irctc413 https://github.com/saltstack/salt/issues/13305
08:34 Rucknar crocket: Changing settings on the master so i'd expect that.
08:35 irctc413 in the end, all it says need to provide auth info in ~/dockercfg
08:36 Mso150_f joined #salt
08:38 irctc413 #docker
08:38 Rucknar sorry irctc413, not familiar with that stuff. There should be more people on IRC later today when the US wakes up :)
08:38 bagackiz joined #salt
08:39 cliff-hm joined #salt
08:40 intellix joined #salt
08:41 intellix joined #salt
08:43 lothiraldan joined #salt
08:43 MohShami superted666: raet looks very interesting. Can it be used for other applications as well?
08:44 Rucknar MohShami: Not too sure to be honest, it's classified as 'beta' for the 2014.7.0 release, there is some documentation floating around but i don't know a great deal beyond it's purpose.
08:44 MohShami I see, I'm designing an app based on ZMQ and thought this might be a better option
08:46 intellix joined #salt
08:49 xsteadfastx joined #salt
08:54 crocket I'm writing a state module that accepts a given id automatically within 10 seconds.
08:56 hvn joined #salt
08:56 Rucknar crocket: The salt master can do that for you with a whitelist
08:58 crocket Rucknar, by modifying /etc/salt/master?
09:00 ingwaem joined #salt
09:01 Rucknar crocket: Yep
09:01 crocket Rucknar, I'm writing states to uncomment autosign_file and put autosign_file.
09:02 Rucknar crocket: Are you dynamically updating the file or something?
09:03 crocket Rucknar, I'm writing orchestrate states for bootstrapping the master.
09:05 PI-Lloyd joined #salt
09:11 alex-mesos joined #salt
09:12 alex-mesos joined #salt
09:19 N-Mi joined #salt
09:19 N-Mi joined #salt
09:24 mortis_ when installing multiple packages at the same time in an sls, can i do like - somepkg1 and on next line set - somepkg2: latest
09:24 mortis_ or so?
09:25 mortis_ or does it have to be specified by pkg.latest
09:26 LeProvokateur joined #salt
09:30 alex-mes_ joined #salt
09:31 viq joined #salt
09:35 VSpike I guess there would be potential issues with putting install binaries in https://github.com/saltstack/salt-winrepo
09:35 VSpike Legal type issues, I mean
09:42 thayne joined #salt
09:50 ramishra joined #salt
09:51 Mso150_f joined #salt
10:00 alex-mesos joined #salt
10:04 glyf joined #salt
10:05 yomilk joined #salt
10:15 glyf joined #salt
10:20 mage_ joined #salt
10:20 mage_ hello
10:20 peters-tx joined #salt
10:21 mage_ I'm running FreeBSD (and all his ecosystem: jails, pkgng, ...) and I wondered if it's well supported in Salt ?
10:22 mage_ also: is it possible to run Salt in an Ansible-like mode (through SSH) without the need to install it on every server ?
10:22 Rucknar mage_: Seems so, http://docs.saltstack.com/en/latest/topics/installation/freebsd.html
10:22 Rucknar mage_: And yes, you can run it agentless through SSH
10:22 Rucknar mage_: http://docs.saltstack.com/en/latest/topics/ssh/
10:23 mage_ cool..!
10:23 mage_ I'm hesitating between Ansible and Salt actually
10:25 crocket left #salt
10:27 lothiraldan joined #salt
10:29 viq mage_: there is salt-ssh, but it's not working as well as the agent.
10:30 nbari joined #salt
10:31 viq mage_: also, an anecdote: we have a bunch of quite limited OpenVZ containers, some of them are assigned 30-60MB of memory. Ansible made them fail over, as it left 30+ ssh connections open, they run happily with salt-minion on them.
10:31 nbari hi all, I would like to have a file with all my users/sshkeys/loginname but I am hinking what could be the best aproach, to use a simple yalm or jinja sintax basically is a dictionary of dictionaries
10:31 ingwaem joined #salt
10:32 mage_ viq: ok :)
10:32 nbari the idea is that later I juet could import that file and select users from it, something like add only users a b c
10:32 nbari any advice ?
10:32 mage_ what do you mean by "not working as well" ?
10:32 mage_ in my case speed is not an issue (I have 10 machines not 10000)
10:33 viq mage_: also on freebsd one thing I noticed what doesn't work as well are some pkg commands, I don't remember right now whether it was pkg.list_upgrades or pkg.upgrade test=True
10:33 viq mage_: it's not as well tested, some people have issues with it I believe.
10:34 Rucknar mage_: Whilst the minion is another package to manage i personally prefer that method over ssh. As someone who has to log SSH connections under strict audit requirements, ssh based config soloutions just spam the logs as well as using a lot of memory but with less features.
10:35 viq nbari: you're talking something along of puppet's virtual resources, AFAIK salt doesn't have that. I'm waiting for 2014.7 to have pillar merging to be able to do something similiar. Though there's also https://github.com/saltstack-formulas/reverse-users-formula which does something similiar
10:36 viq mage_: also a sexy feature (that I haven't yet made use of) that requires minions is the reactor system
10:36 nbari i see, currently my idea was to create a dictionary on a pillar and later just import it
10:37 viq nbari: or it could be that I'm stupid and programming-illiterate and thorougly mistaken ;)
10:42 minaguib joined #salt
10:43 synical joined #salt
10:44 mage_ thansk for those enlightenments :)
10:51 the_drow joined #salt
10:56 eliasp did anyone of you already start to work on realmd modules/states? otherwise, I'd start this adventure now…
10:56 eliasp I'd like to make system.join_domain also available on Linux systems…
11:03 giantlock joined #salt
11:05 packeteer is Helium 2014.1.10 ? or 2014.7 ?
11:07 Rucknar Helium is 2014.7
11:07 Rucknar 2014.1 is Hydrogen
11:07 packeteer thx
11:09 DolourousEdd is helium in epel yet?
11:13 packeteer no
11:13 packeteer least not for centos 7
11:13 Rucknar No, once it's classified as Stable, it will go into EPEL-Testing then EPEL 2 weeks later
11:13 viq Helium isn't released yet
11:14 Rucknar Realisticley be about a months time from now i'd bet.
11:14 brayn joined #salt
11:15 brayn Morning. I was working on some states and all of the sudden all salt commands just hang. There is nothing useful in the salt-master log
11:17 jagardaniel1 joined #salt
11:18 viq https://tumblr.github.io/collins/index.html looks mighty interesting
11:19 msciciel_ viq: i've reconfigured minion recon parameters and it's better but i see that salt-master is constantly trying this: ssh git@git.icore git-upload-pack '/salt/pillar.git
11:19 msciciel_ viq: i have all states , pillars in git
11:19 fbraad1 joined #salt
11:20 viq msciciel_: yes, by default git gets refreshed every minute, configurable
11:21 scottpgallagher joined #salt
11:21 msciciel_ i have changed it to 3 minutes i think
11:21 viq you can also update it manually with 'salt-run fileserver.update'
11:21 yomilk_ joined #salt
11:21 msciciel_ viq: is this loop_interval parametr or sth else for git ?
11:23 viq I think so, judging from the description at http://docs.saltstack.com/en/latest/ref/configuration/master.html#loop-interval
11:23 tmh1999 joined #salt
11:27 tomspur joined #salt
11:28 msciciel_ viq: i think that problem is with git backend and loop interval if salt-master doesn't finish fetching from git within loop interval then next loop start fetching from git again and everything is going to fail. Maybe git fetching should be locked not to start fetching again until previous is still running
11:29 viq That sounds like bug report time to me ;)
11:29 viq Though you must have hell of a git tree to not be done in a minute
11:31 wnkz joined #salt
11:31 drybjed joined #salt
11:36 brayn This is really weird. Salt was working fine but now I can't run any command. If I do salt-master -l debug I get NameError: global name 'AES' is not defined
11:37 brayn Any job returns both an timeout and a authentication message like this http://pastebin.com/HABrERsV
11:38 brayn I've googled around all these problems and found separate github issues for them, but besides some sort of weird DNS issue I can't find a common cause. Especially as it was working fine until now
11:41 nkuttler brayn: you seem to have different problems. can you paste the traceback from when you run the master?
11:41 nkuttler it sounds like the crypto module is not installed
11:42 brayn nkuttler: I've read that it could have been one of the issues but I had it installed, since it worked fine for the last week
11:42 glyf joined #salt
11:42 brayn nkuttler: by traceback you mean the debug output?
11:43 nkuttler brayn: all the output you can get
11:45 imanc is there a way to debug why a jinja variable is not being added to an sls file?
11:49 karimb joined #salt
11:49 eliasp does anyone know whether there has been any progress on the "automated salt-minion builds for windows" front? IIRC ~3-4 months ago SaltStack started some efforts in this area…
11:56 zedr joined #salt
11:57 zedr hello, how can I apply states to a minion as another user on that minion?
12:00 eliasp zedr: masterless setup or with a master?
12:01 zedr eliasp: with a master. My laptop is the master, and I am applying states as a different user.
12:01 eliasp zedr: http://docs.saltstack.com/en/latest/ref/clientacl.html
12:01 baniir joined #salt
12:02 zedr eliasp: excellent, thanks!
12:04 gmcwhistler joined #salt
12:04 StDiluted joined #salt
12:06 imanc is it possible to get the current time in jinja?  {{ now }} or soemthing to that effect?
12:07 eliasp not without a custom filter
12:07 babilen Which oracle java formula is the one that one would want to use? https://github.com/saltstack-formulas/java-formula and https://github.com/saltstack-formulas/sun-java-formula can be found
12:08 imanc eliasp: OK - is there an easy way to append a time to a file name?  Or should I just write a custom filter to do it?]
12:09 eliasp imanc: oh, wait… http://jinja.pocoo.org/docs/dev/templates/#format this could _possibly_ allow to inject date/time values…
12:10 joehh eliasp: ask UtahDave...
12:10 joehh his repo has https://github.com/UtahDave/salt-windows-installer
12:11 joehh there are a number of sls file that called via state.sls with the right arguments will build you windows binaries
12:11 eliasp babilen: no idea, sorry… until now I was lucky enough not to deal with it :)
12:11 babilen yeah, same here.
12:11 eliasp joehh: awesome! let's see
12:12 joehh he put them together for the saltstack qa team to build test releases as required
12:13 eliasp joehh: yeah, and this could also reduce the time from a release to having Win installers available drastically…
12:13 joehh used like: salt '<minion>' state.sls installer.build  pillar='{"git_rev": "v2014.7.0rc6"}'
12:14 joehh these days, windows installers are likely to have been built and tested for a week or so before a release...
12:16 TheThing joined #salt
12:16 oeuftete joined #salt
12:17 yomilk joined #salt
12:18 bhosmer joined #salt
12:19 eliasp great!
12:23 oeuftete joined #salt
12:24 yes456 joined #salt
12:24 bezeee joined #salt
12:30 intellix joined #salt
12:32 aqua^mac joined #salt
12:34 aqua^mac_ joined #salt
12:34 fbraad joined #salt
12:35 lothiraldan joined #salt
12:36 drybjed joined #salt
12:36 XenophonF hey all - i'm getting errors running win_servermanager.installed with a "names" argument
12:36 miqui joined #salt
12:36 XenophonF here's the SLS file - https://bpaste.net/show/e5b8590d62d7
12:37 XenophonF here's the MAP file - https://bpaste.net/show/ab7ab994de24
12:37 XenophonF and here's the error - https://bpaste.net/show/38de1b77582a
12:38 TheThing_ joined #salt
12:40 ggoZ joined #salt
12:56 cpowell joined #salt
12:57 bezeee joined #salt
12:58 micah_chatt joined #salt
13:00 DolourousEdd Helium is always "a few weeks away" :)
13:01 XenophonF k
13:02 rypeck joined #salt
13:02 wrws joined #salt
13:06 XenophonF belay that - i broke the feature installs into separate states, and i'm still getting IndexErrors
13:07 XenophonF win_servermanager.installed or one of its callees is broken
13:07 mdasilva joined #salt
13:07 gngsk joined #salt
13:07 jdmf joined #salt
13:11 jdmf joined #salt
13:19 ilbot3 joined #salt
13:19 Topic for #salt is now Welcome to #salt | 2014.1.10 is the latest | Help us test the 2014.7 RC! http://bit.ly/salt-rc | Please be patient when asking questions as we are volunteers and may not have immediate answers | Channel logs are available at http://irclog.perlgeek.de/salt/
13:19 jdmf joined #salt
13:20 jdmf joined #salt
13:21 ramishra joined #salt
13:22 jdmf joined #salt
13:23 jdmf joined #salt
13:24 jdmf joined #salt
13:26 jdmf joined #salt
13:27 mpanetta joined #salt
13:27 jdmf joined #salt
13:29 yes456 joined #salt
13:29 jdmf joined #salt
13:29 mpanetta joined #salt
13:30 jdmf joined #salt
13:30 bhosmer joined #salt
13:31 bhosmer_ joined #salt
13:31 jblack joined #salt
13:32 jdmf joined #salt
13:32 thedodd joined #salt
13:33 matrix joined #salt
13:33 jdmf joined #salt
13:33 matrix
13:34 jaimed joined #salt
13:34 jdmf joined #salt
13:36 jdmf joined #salt
13:37 jblack Hello! My google skills are failing me. I need to populate a template with a list of minions that match a particular pattern.
13:37 jdmf joined #salt
13:37 ckao joined #salt
13:38 kermit joined #salt
13:38 jdmf joined #salt
13:38 jblack I found some stuff about salt mine, but was hoping there was something built into salt since I just need the minion names
13:39 mpanetta joined #salt
13:39 nitti joined #salt
13:39 jdmf joined #salt
13:40 perfectsine joined #salt
13:41 jdmf joined #salt
13:42 jdmf joined #salt
13:43 jdmf joined #salt
13:44 jdmf joined #salt
13:45 Dig joined #salt
13:46 tk75 joined #salt
13:46 jdmf joined #salt
13:47 PI-Lloyd jblack: salt mine is pretty much the only way to do it afaik
13:47 jdmf joined #salt
13:48 jdmf joined #salt
13:49 iggy everyone needs to collectively /ignore that crocket guy
13:49 kt766 joined #salt
13:50 jdmf joined #salt
13:50 yes456 joined #salt
13:50 nkuttler jblack: yeah, use a mine
13:50 PI-Lloyd iggy: I would, if I didn't find his stupidity amusing
13:51 ramishra joined #salt
13:51 nkuttler fwiw, i recognize the name as a troll from various other channels
13:51 jdmf joined #salt
13:51 iggy jblack: {% set dbhost = salt['mine.get']('G@tags:db and G@tags:primary', 'network.interfaces', 'compound') %}  OR  {% set mailhost = salt['mine.get']('tags:mail', 'network.interfaces', 'grain') %}{% set mailhost = salt['mine.get']('tags:mail', 'network.interfaces', 'grain') %}
13:52 iggy you can put network.get_hostname (or any other function you setup in your mine_functions) in for network.interfaces
13:52 jdmf joined #salt
13:53 jblack ok.
13:54 jblack So in my case, on the template side, I'm looking at something similiar to :   {% for x in salt['mine.get']('harabbit-*','get_hostname') %}
13:55 iggy network.get_hostname
13:55 dude051 joined #salt
13:56 jblack I'm looking for the list of mine functions now
13:56 iggy there are a few ways to specify them
13:56 iggy we do them in pillars
13:56 jblack oh?
13:57 iggy https://gist.github.com/iggy/035b3f2e5109e808a274
13:57 iggy like that
13:58 iggy there are other ways as well
13:58 PI-Lloyd jblack: http://docs.saltstack.com/en/latest/topics/mine/ top of the page explains how to add mine to pillar
13:58 housl joined #salt
13:58 jblack You just made my whole week, if that does what I think it does.
13:58 PI-Lloyd probably more too :p
13:58 jblack You made someone's life better, iggy.  =)
14:02 lcavassa joined #salt
14:03 whiteinge iggy ftw  :-)
14:05 wendall911 joined #salt
14:09 KennethWilke joined #salt
14:11 mdasilva err
14:11 mdasilva whats the best way to troubleshoot gitfs remotes?
14:12 jblack I should be able to do a :    sale 'harabbit*' mine.get '*' netowrk.get_hostname
14:12 jblack right? I added  the mine functiosn to my pillar/all.sls
14:13 jdmf joined #salt
14:14 jdmf joined #salt
14:15 vejdmn joined #salt
14:16 jdmf joined #salt
14:17 jdmf joined #salt
14:17 iggy mdasilva: one way is salt-run -l debug fileserver.update
14:18 iggy jblack: update mine
14:18 jdmf joined #salt
14:19 jblack Wooot!
14:20 jdmf joined #salt
14:20 jblack Does this mean I have to do a mine update each time I create a new harabbit minion, or wait an hour for mine to automatically update?
14:20 iggy fwiw, if you expect that data to change often, you can change the frequency it updates (it's 1h by default)
14:21 jblack I expect the data for individual minions to be static,  but I expect new minions
14:21 jdmf joined #salt
14:21 iggy /etc/salt/minion.d/mine.conf:  file.managed:    - contents: "mine_interval: 5"    - contents_newline: True
14:22 iggy or for that, you can use a reactor
14:22 jdmf joined #salt
14:23 murrdoc joined #salt
14:23 mdasilva thanks iggy ill try that
14:23 CeBe joined #salt
14:24 jdmf joined #salt
14:25 jdmf joined #salt
14:26 Sypher joined #salt
14:26 jdmf joined #salt
14:27 vejdmn1 joined #salt
14:28 fbraad joined #salt
14:28 jdmf joined #salt
14:31 jblack I'm not sure that pillar/all.sls is being honored.
14:33 jdmf joined #salt
14:33 jblack I changed that as per instructed.  I reran state.highstate on my other rmq node, then reran  salt '*' mine.update on the salt server
14:34 iggy did you assign it in the pillar top file?
14:34 jblack I have the first node (in which I did lots of stuff like temporarily edit the config file to add mine functions), but not the untouched one
14:34 bersace joined #salt
14:35 bersace Hello :-)
14:35 jblack um. so, um... whoops
14:35 snuffeluffegus joined #salt
14:37 bersace I wrote a pillar, but pillars.items doesn't show it. Even if it is rendered. See http://pastebin.com/jb8cmBzV
14:37 miqui joined #salt
14:37 bersace Do you have a clue why isn't it available ?
14:38 ajolo__ joined #salt
14:38 viq bersace: which salt version?
14:38 aqua^mac joined #salt
14:39 UtahDave joined #salt
14:39 * iggy puts money on it
14:40 viq yeah, I'm also guessing 2014.1.10 but let's wait for it ;)
14:40 bersace viq: salt-call 2014.1.11 (Hydrogen)
14:40 iggy 11 is the broken one
14:40 viq ah, sorry
14:40 bersace X(
14:40 viq bersace: yeah, .11 has a bug with pillars in local mode
14:40 iggy .13 should be out for most distros now
14:40 bersace ok :-)
14:41 jblack So, now I have all in my  top.sls for *, and I reran  salt on the not-showing mode, then ran salt 'harabbit*' mine.update; salt 'harabbit*' mine.get 'harabbit*' network.get_hostname
14:42 jblack I got 'true' for both harabbits,  but one of them is still missing.  I  tried restarting salt on the missing one
14:42 mapu joined #salt
14:44 Topic for #salt is now Welcome to #salt | 2014.1.13 is the latest | Help us test the 2014.7 RC! http://bit.ly/salt-rc | Please be patient when asking questions as we are volunteers and may not have immediate answers | Channel logs are available at http://irclog.perlgeek.de/salt/
14:45 jblack am I retarded? I'm sorry
14:46 jblack Look at that! It works great when you change the server and not your local git repo!
14:47 bersace viq, iggy upgrade fixed it. Thanks !!
14:48 fragamus joined #salt
14:49 * iggy ^5's viq
14:49 kaptk2 joined #salt
14:49 jblack thank you very much, iggy
14:50 mdasilva iggy: GitError: SSL is not supported by this copy of libgit2
14:50 mdasilva salt-run -l debug fileserver.update did the trick
14:50 iggy yeah, that's usually the first thing I check
14:50 mdasilva is pygit2 the recommended provider?
14:51 mdasilva it was kinda crappy getting that in debian wheezy
14:51 viq iggy: so now I'm viq*viq*viq*viq*viq ? ;)
14:51 murrdoc i <3 salt
14:51 murrdoc ok back to work
14:51 murrdoc morning all
14:52 iggy I'm on 2014.1, gitpython is the only option
14:52 jalbretsen joined #salt
14:52 vejdmn joined #salt
14:53 mdasilva what was i thinking using pygit2
14:54 timoguin pygit2 is supposed to be newer/better
14:54 timoguin but libgit has worked fine for me
14:55 mdasilva oh pygit2 supports added options
14:55 mdasilva thats why..
14:56 jblack time to go do my other job
14:56 jblack Thanks all!
14:57 ioio joined #salt
14:57 Ozack1 joined #salt
15:02 glyf joined #salt
15:03 withasmile joined #salt
15:04 XenophonF can anyone show me an example of calling a module from a state?
15:05 rwcs joined #salt
15:05 viq XenophonF: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.saltmod.html  ?
15:05 XenophonF I have the following, but it doesn't work - https://bpaste.net/show/6eb54245d5b5
15:06 XenophonF hm, i thought that the module.run state was what i needed
15:07 XenophonF i basically want to do a pkg.refresh_db
15:07 * viq checks
15:07 timoguin XenophonF: I have an old state that uses the same thing essentially: https://github.com/timoguin/flexnet-salt-states/blob/master/java.sls#L8
15:07 viq oh, hm
15:07 timoguin the idea was for it to refresh the apt cache after adding the repo, so the subsequent install doesn't fail on the first run
15:07 alex-mes_ joined #salt
15:07 cpowell anyone know on a Centos system when adding a repo, how to make it do a yum update prior to installing a pkg?
15:08 cpowell it can't find the pkg I want to install from the repo initally
15:08 viq cpowell: you don't need to, yum does that by itself
15:08 viq should anyway...
15:08 irctc302 joined #salt
15:08 XenophonF viq, timoguin: the state works on Amazon Linux and Windows Server 2012 R2, but not on FreeBSD
15:08 cpowell its not though...first run of the state it adds the repo and pkg.install fails, second run it installs
15:08 XenophonF I'm getting the following error: https://bpaste.net/show/a9dd1decb018
15:09 viq cpowell: do you have the pkg state require the repo state
15:09 viq ?
15:09 cpowell yes
15:09 viq cpowell: refresh: true in your package state
15:09 cpowell already done
15:10 cpowell https://gist.github.com/anonymous/07e8db3998e8c8fa555b
15:10 QiQe joined #salt
15:10 timoguin you made need to run pkg.refresh_db with module.run after adding the repo.
15:10 viq cpowell: require, not required
15:10 cpowell doh...
15:10 cpowell its always freakin typos
15:11 n8n joined #salt
15:12 rwcs guys have some whay to have different enviroments like /srv/salt/customer1 and /srv/salt/customer2 without STOP salt-master? because this server will be run for many customers...... so I cannot STOP.
15:12 conan_the_destro joined #salt
15:12 timoguin XenophonF: maybe a bug in the freebsdpkg module. maybe try running pkg.refresh_db manually to see if it gives you errors
15:13 viq rwcs: gitfs automatically maps branches to environments
15:13 hasues joined #salt
15:15 withasmile joined #salt
15:15 QiQe guys I really need some help
15:15 QiQe I have this config on my testing environment on salt master
15:15 QiQe http://pastebin.com/iF1nkdRC
15:15 QiQe the same on master environment
15:16 QiQe suddenly salt stop to update files on cache
15:16 QiQe I need to restart the salt-daemon to fix, and after a while, again
15:16 irctc302 how to provide authentication to docker registry using salt?
15:16 viq 1) anything in logs? 2) in such situations I just rm -rf /var/cache/salt/master and restart master
15:16 Rucknar Can you still contact the salt minions through something like a test.ping>?
15:17 QiQe viq, I have done that many times
15:17 QiQe and after a while the issue come back again
15:18 iggy you are using the same git repo for your states and pillars?
15:18 QiQe there isnt any issue with minions
15:18 kickerdog joined #salt
15:18 QiQe iggy: yes, Im using the smae
15:18 QiQe same
15:18 iggy that's a bad idea
15:18 XenophonF timoguin: pkg.refresh_db works fine
15:19 XenophonF I tried running salt-call state.single module.run name=pkg.refresh_db, but I got the same AttributeError 'bool' object has not attribute 'get'
15:19 QiQe so, do you think the issue is related to that?
15:19 johtso joined #salt
15:19 iggy and I never got the ext_pillar git branch stuff to work the way it was supposed to
15:19 kickerdog left #salt
15:19 ndrei__ joined #salt
15:20 QiQe and what is the config you recommend?
15:20 hintss joined #salt
15:21 schristensen joined #salt
15:24 ek6 joined #salt
15:26 racooper joined #salt
15:27 iggy we use separate git repos for states and pillars
15:28 SheetiS joined #salt
15:31 johtso What would be the correct way to write a custom module that basically just runs another module command but pulls the parameters from the pillar data?
15:31 johtso do you even have access to pillar data in modules?
15:34 ek6 joined #salt
15:35 StDiluted joined #salt
15:35 ek6_ joined #salt
15:35 timoguin johtso: yea you have access to the full __salt__ dictionary and can call other modules with it
15:35 timoguin http://docs.saltstack.com/en/latest/topics/development/dunder_dictionaries.html
15:36 timoguin you can call any of the available modules using __salt__['module.function'](args)
15:37 johtso ah right, so I'll just need to call the pillar module
15:38 ek6 joined #salt
15:42 ek6 joined #salt
15:43 druonysus joined #salt
15:46 bezeee joined #salt
15:46 bhosmer joined #salt
15:46 modafinil_ is anyone familiar with anything off the shelf to make completion with zsh work for states/tasks?
15:48 tligda joined #salt
15:51 asmaps joined #salt
15:51 bernieke fyi, the problem I was having with multimaster dividing up the minions instead of both having access to all, was due to the following tidbit in my master configs:
15:51 bernieke -reactor:
15:51 bernieke -  - 'salt/minion/*/start':
15:51 bernieke -    - /srv/reactor/sync_all.sls
15:51 bernieke -timeout: 15
15:51 bernieke (ignore the minuses in front of the lines, I c/p the diff where I removed them)
15:52 bernieke I'm not sure if this is expected behaviour or not, if not I can make a ticket for it?
15:52 bernieke this is with rc6 btw
15:55 mdasilva progress!!!!!
15:55 mdasilva we all love
15:58 keithhub joined #salt
15:58 nitti joined #salt
15:59 miqui_ joined #salt
15:59 scoates is it not possible to use a custom module in top.sls (file root/state) ? seems the dict is not yet populated at that point?
16:01 ramishra joined #salt
16:01 SheetiS scoates: I've used salt['grains.get'] in my top.sls, so I'd imagine that other modules would be available.
16:01 * scoates nods
16:02 scoates grains.get is not custom
16:02 SheetiS understood
16:02 SheetiS is your custom module working elsewhere already other than top?
16:03 scoates yep
16:03 scoates I guess that must be it.
16:04 SheetiS hmm where is top file rendered?
16:04 scoates haven't found that yet either (-:
16:05 SheetiS I'll make a quick check
16:06 thayne joined #salt
16:10 SheetiS well if the top file is rendered on the master, (which for states I am not sure, but all of the pillar is rendered on the master)
16:10 SheetiS Then the module will need to live in another place
16:10 * scoates nods
16:11 SheetiS http://docs.saltstack.com/en/latest/ref/configuration/master.html#extension-modules
16:11 SheetiS wouldn't hurt to make it available that way
16:11 scoates it's for a helper function. I can implement in Jinja
16:11 SheetiS as well as in _modules
16:11 ek6_ joined #salt
16:11 SheetiS to test
16:11 scoates it's actually a pain for us to make stuff available that way (-:
16:11 scoates race condition between the saltmaster starting and the directory in which our file_roots exist being mounted
16:11 SheetiS scoates: ahh okay.
16:12 scoates was just hoping to confirm my theory about modules not being available in top; no big deal.
16:12 scoates thanks for checking into it
16:13 ek6_ joined #salt
16:14 SheetiS scoates: sorry I wasn't more help.
16:14 scoates s'ok. I appreciate it anyway.
16:19 StDiluted joined #salt
16:20 KyleG joined #salt
16:20 KyleG joined #salt
16:21 glyf joined #salt
16:21 vukcrni joined #salt
16:21 aparsons joined #salt
16:22 slav0nic joined #salt
16:23 troyready joined #salt
16:23 masterkorp murrdoc: hellp
16:24 masterkorp murrdoc: can I pm you ?
16:24 jngd joined #salt
16:27 bezeee are there any known issues on the rc branch with the postgres_user state?  I'm getting tracebacks when boolean attributes are set: https://gist.github.com/brianz/a27bf7ec3cf3b15b68f0
16:27 dude051 joined #salt
16:31 Jarus joined #salt
16:33 thedodd joined #salt
16:34 rwcs1 joined #salt
16:35 rwcs1 .
16:37 iggy bezeee: someone else was reporting the same yesterday... you might see if they opened an issue for it
16:38 ekristen joined #salt
16:38 rwcs1 guys how to RELOAD salt-master??? I have to create one more environment like /srv/salt/customer/customer1, but I cannto stop/star I need to RELOAD.
16:38 bezeee iggy: thanks...i didn't see anyting in the issues but i'll poke through irc logs from yesterday
16:39 wt joined #salt
16:39 iggy rwcs1: the suggestion someone already gave you is probably your only solution (i.e. gitfs branches)
16:41 baconbeckons joined #salt
16:41 UtahDave joined #salt
16:42 baconbeckons joined #salt
16:44 yes456 joined #salt
16:46 rlima joined #salt
16:47 bernieke scoates: my experience is that you can only have custom modules in your top.sls, after you've done your first sync_all. Basically sync_all without a top.sls, and only then put your top.sls in its place. (So, in short, no :))
16:48 scoates bernieke: that makes a lot of sense
16:48 ek6 joined #salt
16:48 scoates though perhaps the sync should take place before trying to render the top unless there's a great reason not to
16:48 intr1nsic joined #salt
16:48 jngd joined #salt
16:49 Ahlee Is there any way to make salt-call report back to the master it ran, short of defining a returner and enforcing it on call?
16:49 jhulten joined #salt
16:49 bernieke yes, that would be my expectation as well, but like you said, there might be a reason it's not :)
16:49 * scoates nods
16:49 scoates bernieke: thanks
16:50 ekristen so this isn’t necessarily a salt question — anyone have experience/recommendation for ulimit settings as it pertains to open file descriptors on linux systems that handle a considerable amount of http connections?
16:50 debian112 joined #salt
16:50 Ahlee unlimited and let the server die a fiery death
16:51 ekristen Ahlee: lol, thanks for that
16:52 ek6 joined #salt
16:53 wramthun joined #salt
16:53 Ahlee *shrug* :)  It works.  Either way, the system is going to be constrained, so it's either an artificial limit that may allow you to get back in to reset the system, or no limit which offers no chance of getting in
16:53 scoates Ahlee: you could catch a successful event with reactor: http://paste.roguecoders.com/p/5929ef8fdd21ef5dd9516343efead8bc.txt
16:53 smcquay joined #salt
16:54 troyready joined #salt
16:54 felskrone joined #salt
16:54 Ahlee scoates: thanks.  I guess it's one more reason to look into processing the events directly
16:54 intr1nsic joined #salt
16:56 linjan joined #salt
16:59 wramthun I need help finding the latest Solaris minion.  The most recent one I can find is verision 0.14.0 which I obtained via OpenCSW.  This lags the linux minion version I'm using - 0.17.5.
16:59 wramthun Any help would be appreciated.
16:59 ekristen Ahlee: PM
17:00 glyf joined #salt
17:00 jngd joined #salt
17:01 cliff-hm joined #salt
17:03 ek6 joined #salt
17:03 Gareth morning morning
17:03 wramthun doctor doctor
17:05 hasues wramthun: The bootstrap script does not run on Solaris?
17:05 druonysuse joined #salt
17:06 jalaziz joined #salt
17:07 iggy 0.17.5 is old, 0.14 is older... sadly there aren't any newer packages (that I've heard of)
17:08 nitti_ joined #salt
17:09 aparsons joined #salt
17:09 hasues left #salt
17:09 wramthun hasues: We opted for the package install. So we were looking for an updated package. Is one method perferred over another?
17:10 wramthun hasues: I'll give the bootstrap a try and let you know what installs.
17:11 basepi Thanks to the Cisco WebEx team, tonight's SaltStack Silicon Valley meetup (hosted by Cisco) is now available via live stream. The meetup and video stream will begin at 6pm on Tuesday, Oct. 28 (today). Speakers include Tom Hatch (Salt Helium release highlights), Matt Meservey (SaltStack Enterprise sneak peek), and Ed Kern (how Cisco uses SaltStack). Register
17:11 basepi here for access to WebEx details: http://www.saltstack.com/saltstack-events/saltstack-silicon-valley-meetup-cisco
17:11 basepi (Just a little spam for everyone. =) )
17:11 jalaziz_ joined #salt
17:12 glyf joined #salt
17:14 iggy I'm assuming those times are local?
17:15 iggy well, yeah, Si Val and SLC are the same tz, so...
17:16 basepi Actually, they're note
17:16 basepi not*
17:17 basepi We're figuring out timezones now
17:17 zlhgo joined #salt
17:17 Gareth basepi is in the future.
17:18 * iggy goes back to geography class
17:18 basepi hehe
17:21 basepi the livestream is 6pm Pacific Time
17:22 baconbeckons joined #salt
17:25 Ryan_Lane joined #salt
17:25 vejdmn joined #salt
17:32 nitti joined #salt
17:33 jeddi joined #salt
17:36 thomasfr joined #salt
17:38 ek6 joined #salt
17:43 Gareth UtahDave: ping
17:43 UtahDave Gareth: pong!
17:43 Gareth UtahDave: got a second for a PM?
17:43 UtahDave Yeah, totally
17:45 jalaziz joined #salt
17:47 iggy the docs say scheduler settings should be able to be set in the master or minion config files
17:47 iggy but I can't get master schedule settings to work
17:47 jalaziz joined #salt
17:50 XenophonF so i looked at the output of the "module.run name=pkg.refresh_db" state
17:50 XenophonF https://bpaste.net/show/066c2027ca77
17:50 ek6 joined #salt
17:50 XenophonF this is on FreeBSD 10.0-RELEASE-p11
17:51 XenophonF it's definitely running "pkg update"
17:51 XenophonF you can see that at lines 9 and 10
17:51 XenophonF so i don't understand where the AttributeError is coming from
17:58 mdasilva when using salt.pillar.git_pillar, how does one specify credentials?
17:59 sjaredj is there something you can run on the master that will tell you the last time a minion ran a highstate?
18:00 iggy sjaredj: not by default
18:00 lothiraldan joined #salt
18:02 bhosmer joined #salt
18:03 sjaredj iggy: so lets say you have a cron that runs highstate every 15 minutes.  some other admin user comes in and needs to do something and doesn't wan to be bothered with salt so he comments out the cron that runs highstate.  when he is done he forgets to uncomment out the cron.  Is there a way to detect that easily?
18:04 viq iggy: wouldn't that be in the event cache?
18:04 * viq is guessing
18:04 viq And if it is, you could possibly wrap something around salt-eventsd
18:04 iggy I really hope events aren't cached or I'm going to be out of disk space rsn
18:05 viq iggy: well, jobs are... And the cache has a limit, IIRC
18:05 iggy I was trying to think of something with the job list
18:05 iggy but aside from just walking over every job and checking...
18:05 viq Yeah, I wonder if scheduled highstates would show up in completed jobs listing
18:05 iggy it should
18:06 sjaredj I'm very new to salt so I'm trying my best to follow what you are saying…may take some research though :)
18:06 viq Well, there are scripts and whatnot to yell if an event did _not_ happen in a scheduled time. So you would need a list of minions, and keep checking for their jobs, and alert on that
18:07 sjaredj but I could use cmd.run to kind of inventory it?
18:07 viq hm?
18:07 iggy all our stuff is in git and requires a PR, so... it would be hard for someone to do something like that without anyone else knowing
18:07 viq iggy: ssh to machine, comment out cron job
18:08 viq sjaredj: on the other hand, you're thinking cron job - salt has it's own scheduler, controlled via pillars, which ideally live in git - therefore there's no changing it without leaving a trace
18:08 iggy if you ssh directly to machines and run stuff, I get to punch you in the junk
18:08 viq http://docs.saltstack.com/en/latest/topics/jobs/schedule.html
18:08 viq There's that, yes :P
18:08 iggy it's in my contract
18:09 viq :D
18:10 viq I still get people not wanting to use salt in our infrastructure "because that's the only machine that's going to have that configured" and "each machine is different" and "it's much slower to have to edit the file on my computer, then push it to git, then fileserver.update on saltmaster, then push it out than to just edit file on machine, especially if I'm trying to figure out how to fix something"
18:10 sjaredj I will have to checkout schedules. I was under the impression I needed to setup a cronjob in order for my minion to checkin to see if anything changed
18:11 viq sjaredj: no, there's scheduling. And it's going to get a boost in 2014.7
18:11 sjaredj very good to know
18:11 iggy we don't have minions "check" for updates (it's salt after all... not chef)
18:12 iggy when we push changes to git, we then login and run a highstate
18:12 viq Or schedule it otherwise
18:12 iggy you could also use a reactor to automatically highstate on every push
18:12 iggy but we are sticking with the manual step for now
18:12 viq Can you do batches in reactor?
18:13 viq Also I wonder if there's a decent way to push highstate only to affected machines.
18:13 krak3n` joined #salt
18:13 iggy automatically? no clue
18:13 iggy we do it manually
18:14 iggy our biggest "cloud" (we actually have 3 different gce projects with 3 different salt masters) is only about 100 intances, so we don't bother batching
18:14 cberndt joined #salt
18:15 iggy but it'd probably be a good feature request
18:15 SheetiS1 joined #salt
18:15 iggy (if it's not already supported)
18:17 sschwartz_ee joined #salt
18:18 sschwartz_ee Me again, still trying to figure out why I can't get rest_wsgi to execute state.sls successfully -- has anyone here managed to get it to work?
18:20 krak3n` joined #salt
18:21 sschwartz_ee For example: curl -sS -i -H 'Content-Type: application/json' -d '[{"eauth":"pam","username":"jenkins","password":"<pasword>":"local","tgt":"ip-10-83-98-227.ec2.internal","fun":"'state.sls'","args":[{"vararg":"test.apitest"},{"saltenv":"/srv/salt"}] }]' localhost:8999
18:22 sschwartz_ee Generates a "state.sls: sls() takes at least 1 non-keyword argument (0 given)" error, but for the life of my, I cannot figure out how to give it said non-keyword argument.
18:23 wramthun hauses: I tried installing via bootstrap script onto a sparc solaris 10 instance.  The bootstrap script includes #!/bin/sh but in linux /bin/sh is linked to bash whereas in solaris 10 /bin/sh is really bourne shell.  So first run, the "set -o nounset" failed.
18:23 wramthun hauses: so I replace #!/bin/sh with #!/bin/bash and reran.
18:24 timoguin the bootstrap script is supposed to be fully POSIX compliant. if it isn't then a bug should probably be filed
18:24 bhosmer joined #salt
18:26 wramthun hauses: and then ran into a "sed -r" options issues.  I'll take a look at that as well, but feel like I'll run into issues with all the achient-ass tools included in solaris.
18:27 wramthun timoguin: Thanks.  I'll see about opening a bug.
18:27 thayne joined #salt
18:27 jalaziz joined #salt
18:30 rwcs1 .
18:30 rwcs1 .
18:30 nebuchadnezzar joined #salt
18:32 Dig Hello guys... Is it possible to run two jobs simultaneously ? I will call salt-api to run a job with several state for one target and then, I need to run another job to install in another target others states... I see that when I do it,it returns me a message saying that "The function "state.highstate" is running as PID XXXX and was started.........." but I don't if there is conf to do enable it. Would it be possible?
18:34 aparsons_ joined #salt
18:36 jalaziz joined #salt
18:37 druonysus joined #salt
18:40 bluenemo joined #salt
18:42 bluenemo hi guys. I'm trying to use user.info on the values of a list i'm iterating over via a for loop, sadly it gives me back 'dict object has no attribute'. I dont really get why it is trying to find the attribute to the users i'm iterating over, instead of looking up the attribute from user.info output. Hope you have a hint for me. Code, pillar and output is here: http://paste.debian.net/hidden/0a33ccc2/
18:42 keithhub left #salt
18:46 nitti joined #salt
18:46 mdasilva joined #salt
18:47 dstokes there any way to run a single state (by id) in a state file? i.e. salt \* state.by_id path.to.state state_id?
18:48 dstokes i tried state.sls_id but doesn't seem to work. think it relies on module system
18:49 timoguin dstokes: sls_id should work but that's only in 2014.7+
18:49 dstokes it complains about needing three arguments, docs aren't very clear. i'm using the same method sig as above
18:49 viq joined #salt
18:49 dstokes also looking at these docs http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.sls_id
18:49 jalaziz joined #salt
18:49 dstokes and running rc2 (>2014.7)
18:49 viq meh, sorry, my connection decided to be uncooperative
18:51 timoguin dstokes: yea docs are sparse. doesn't look like it needs 3 args
18:52 dstokes "TypeError encountered executing state.sls_id: sls_id() takes at least 3 arguments (2 given). See debug log for more info.  Possibly a missing arguments issue:  ArgSpec(args=['id_', 'mods', 'saltenv', 'test', 'queue'], varargs=None, keywords='kwargs', defaults=(None, False))"
18:52 timoguin oh i see
18:52 timoguin saltenv doesn't have a default
18:52 timoguin but it should
18:52 timoguin try passing saltenv=base for now
18:53 dstokes ah, cool. was able to make it work with `.. state.sls_id sls_id path.to.sls base`
18:54 nitti joined #salt
18:55 kballou joined #salt
18:58 iggy Dig: no, use orchestrate
18:59 felskrone joined #salt
19:01 evidence w00t, the syndic_master fix made it into dev.. wonder if it'll be pulled into rc7.  are there plans to cut that soonish?
19:03 bezeee joined #salt
19:07 sschwartz_ee OK. Trying now with rest_cherrypy, instead of rest_wsgi -- I can get it to recognize a password, and give me back a 200 state -- but not actually *execute* anything.
19:07 sschwartz_ee Any advice?
19:07 schristensen joined #salt
19:08 aparsons joined #salt
19:09 rostam joined #salt
19:09 timoguin dstokes: saltenv now defaults to base for that function in the 2014.7 branch
19:09 micah_chatt joined #salt
19:11 ggoZ joined #salt
19:13 KyleG joined #salt
19:13 KyleG joined #salt
19:16 darvon joined #salt
19:17 n8n joined #salt
19:17 buhman joined #salt
19:18 spookah joined #salt
19:18 rwcs1 guys, I need some consulting if you really good on saltstack send me a email....  (cjason88888 at gmail.com)
19:18 debian112 hey guys, I posted a question about regex and includes, but I missed the response because I lost connection to the Internet - I will post again
19:19 debian112 Can I use a regex here: http://paste.debian.net/128979/
19:19 iggy we talked about it
19:19 iggy you were responding
19:20 iggy realistically, you should tell us what you're actually trying to achieve, it sounds like you could do things better than what you're trying
19:20 iggy i.e. if you have a if/elif/endif list that's pages long, you're doing something wrong
19:21 sschwartz_ee Ah-hah! I knew that asking the question in public would cause it to start working. Thank you!
19:21 debian112 iggy: I am looking to shorting this process with a regex, basiclly - match the grain ID with the regex
19:22 iggy 18:59 < iggy> debian112: regex, probably not... possibly a: {% if 'server1' in grains['id'] %}
19:22 iggy and you replied back after that, so I know you saw it
19:23 debian112 yeah I saw that, but anything after that I didn't see
19:23 iggy that was it
19:23 debian112 just checking
19:23 iggy what else do you need to know
19:24 buhman http://ix.io/eXs :(
19:24 mdasilva debian112: perhaps consider setting a grain or pillar to define a property that can be use in the state
19:24 debian112 iggy: so basically I just need to continue what I am doing: Looking for something like this: {% if 'server[0-9]' in grains['id'] %
19:24 mdasilva rahter than trying a regex match against the minionid
19:25 iggy I think you should actually tell us what you are trying to achieve, not how you are trying to achieve it
19:25 iggy as I said... if you are going to end up with a page full of if statements, you are almost certainly doing it wrong
19:25 SheetiS joined #salt
19:26 debian112 OK, I have some servers that need to have unique settings: IP address, and some pillars.
19:27 debian112 I could add every server in top.sls, but this does not seems logical
19:27 iggy shouldn't the servers already have IPs (you know... to be able to talk to the master)
19:28 iggy and be more specific about "some pillars"
19:28 thayne joined #salt
19:30 aparsons joined #salt
19:31 buhman that's the first salt-boostrapp'ed master I've done in awhile--I'd read that a bunch of relevant things were merged into 2014.07 in https://github.com/saltstack/salt/issues/12248 yet using this on the master and all minions didn't seem to accomplish anything useful.
19:34 aparsons joined #salt
19:35 iggy it sounds like your config/cache is fubar
19:35 murrdoc joined #salt
19:36 iggy I notice it says root@localhost... might it be easier to copy your states/pillars/etc and start over fresh?
19:36 tkharju joined #salt
19:38 buhman heh
19:38 buhman so I nuked all of the minion/master keys, fixed master_finger, and things magically started working.
19:39 buhman salt-cloud is pretty weird sometimes :S
19:39 buhman iggy: thanks
19:39 * iggy keeps his mouth shut
19:40 tkharju joined #salt
19:44 rostam hi newbie question, how would I change the log level on salt. thanks
19:44 KyleG joined #salt
19:44 KyleG joined #salt
19:45 buhman rostam: I use 'salt-{minion,master} -l debug' to look at what's happening in the foreground.
19:45 buhman salt-call is fun for looking at what states/modules are doing.
19:46 babilen ... or run 'salt -ldebug '*' some.function ...' and salt-call
19:46 babilen s/or/and
19:47 timoguin there's a log_level and log_level_logfile setting for the master config as well.
19:47 timoguin as well as a --log-level arg for the command line
19:47 rostam thank you all for great pointers.
19:50 rostam ok so as it was suggested I ran: salt-minion -l debug to find out what is going on, and I see this warning: [WARNING ] SaltReqTimeoutError: Waited 60 seconds
19:51 iggy there are some other commands that are useful in special circumstances too (salt-run -l debug fileserver.update being one)
19:53 mdasilva iggy: do you got another one like that to help debug ext_pillar git remotes?
19:54 iggy did you try that one?
19:54 mdasilva yeah that fixed my gitfs issue
19:54 mdasilva it identified my libgit2 not having SSL built in
19:54 mdasilva thank you
19:54 babilen rostam: That is typically indicative of a communication error between master and minion. can you ping them?
19:54 jalaziz joined #salt
19:54 mdasilva iggy: im testing ext_pillars with a git remote
19:54 KyleG joined #salt
19:54 KyleG joined #salt
19:55 mdasilva the salt.pillar.git_pillar doc doesnt say where it picks up its git creds
19:56 iggy it doesn't
19:56 iggy we use aliases in the ssh config file
19:56 rostam I keep seeing these type of messages in syslog : 2014-10-28 19:55:14.531 - 2014-10-28 12:55:14,531 [salt.loader      ][DEBUG   ] Loaded json_out as virtual json
19:57 AmrMostafa joined #salt
19:57 rostam not sure why salt.loader keeps getting run...
19:57 XenophonF it appears that the function call interface for salt.modules.pkg.refresh_db has changed, but salt/modules/pkgng.py wasn't updated
19:58 XenophonF hence why "salt-call state.single module.run name=pkg.refresh_db" throws AttributeError even though pkg.refresh_db runs properly
19:58 XenophonF https://bpaste.net/show/066c2027ca77
19:58 mdasilva iggy: how does that tie in git credentials?
19:59 iggy mdasilva: oh, we use keys to talk to the git server, so we setup an alias with specific keys
19:59 XenophonF basepi or UtahDave, can either of you point me to an interface definition for pkg.refresh_db()?
20:00 XenophonF i'll patch pkgng.py but i'm having a hard time understanding what that module is expected to return
20:00 XenophonF i'm specifically looking at https://github.com/saltstack/salt/blob/develop/salt/modules/pkgng.py#L213
20:01 AmrMostafa I'm pretty confused with environments. Where exactly do you tell the master which environment to apply to the minion? I was expecting something like an "environment = qa" in minion configuration, or a parameter to salt state.highstate (i.e. --env=qa). Anyone care to elaborate on how this works?
20:01 XenophonF as compared to https://github.com/saltstack/salt/blob/develop/salt/modules/aptpkg.py#L320 et seq
20:01 mdasilva iggy: is there documenation somewhere on configuring that?
20:01 iggy it's an ssh client thing, pretty well documented
20:02 timoguin mdasilva: when i used gitfs i just gave the salt user (user salt runs as) a read-only deploy key
20:02 iggy XenophonF: it returns True|False|None
20:02 XenophonF iggy: the module.run state appears to expect a tuple or something back
20:02 XenophonF there's also this https://github.com/saltstack/salt/blob/develop/salt/modules/yumpkg.py#L629
20:03 KyleG joined #salt
20:03 KyleG joined #salt
20:03 XenophonF and this https://github.com/saltstack/salt/blob/develop/salt/modules/win_pkg.py#L480
20:03 XenophonF so i'm confused by this stack trace - https://bpaste.net/show/066c2027ca77
20:03 murrdoc joined #salt
20:03 ek6 joined #salt
20:04 iggy yeah, it's expecting a dict back
20:04 ek6 joined #salt
20:05 iggy I suspect it's different for each one
20:06 XenophonF i can run state.single module.run name=pkg.refresh_db on Amazon Linux and on Windows
20:06 XenophonF and it works properly
20:06 XenophonF the same state run on FreeBSD fails with the given stack trace
20:06 XenophonF I don't really understand what's going on
20:07 iggy I think you are looking in the wrong place
20:07 iggy the pkg.refresh_db is completing, it's something after that that fails
20:08 thedodd joined #salt
20:08 iggy the "All repositories are up-to-date." says (to me at least) that it worked
20:09 XenophonF yes
20:10 XenophonF the "pkg update" command runs successfully
20:10 XenophonF so if i understand correctly, cmd.retcode picks that up and returns the process exit code somehow
20:11 XenophonF then refresh_db() compares that to 0, which should be true?
20:11 XenophonF and that's what should get returned to module.run()
20:11 XenophonF via the variable mret
20:11 iggy if you run pkg update, does it return 0?
20:11 XenophonF lemme double check
20:13 XenophonF yes it does
20:13 XenophonF https://bpaste.net/show/6886c9fd9d31
20:14 XenophonF and in fact, if i run just pkg.refresh_db from the CLI, it returns True as expected
20:14 iggy no idea then
20:14 XenophonF https://bpaste.net/show/cafe2f7eca2c
20:15 XenophonF is there any way to trace the execution of "salt-call state.single module.run name=pkg.refresh_db"?
20:15 XenophonF like "sh -x"?
20:16 XenophonF at least I have an easily replicated test case ^_^
20:18 nickg left #salt
20:19 UtahDave XenophonF: There isn't an actual spec for that. Your best bet would probably be to look at apt.py or yum.py and see how they handle it.
20:19 bhosmer joined #salt
20:19 UtahDave AmrMostafa: The top file determines which salt environments a minion pulls from
20:20 felskrone joined #salt
20:20 iggy there is also an environment setting in the minion file
20:21 mpanetta Are there any known issues with master cache corruption on 2014.1.x ?
20:21 iggy mpanetta: like what?
20:21 mpanetta Actually any version of salt.  It happened to me about 2 weeks ago on 0.16.4
20:22 mpanetta iggy: Like a compound match matching everything
20:22 mpanetta When it shouldn't
20:22 iggy YES!
20:22 mpanetta Both times it was fixed by clearing the master cache and restarting
20:22 iggy one sec
20:23 AmrMostafa UtahDave: thanks, and then I believe there is no special "environment" property specified anywhere, it's just how I decide which conditions (based on my top.sls targeting) match which environment?
20:23 iggy mpanetta: like this? https://github.com/saltstack/salt/issues/15673
20:24 * mpanetta looks
20:25 mpanetta iggy: That looks exactly like the issue I had in 0.16.4, and similar to the one I just had on 2014.1.10
20:26 XenophonF UtahDave: I have looked at yumpkg.py and aptpkg.py and win_pkg.py
20:27 XenophonF it looks like a True return should work
20:27 XenophonF I dunno.
20:27 XenophonF Maybe tracing that execution would clue me in.
20:27 UtahDave AmrMostafa: Yeah.  There's also an "environment" config option in the minion config that will force the minion to only pull from a certain salt environment, but you still have to use the top file.
20:27 UtahDave XenophonF: let me see if terminalmage is around. He's done a lot with the package modules
20:28 iggy mpanetta: what cloud provider (if any) are you using?
20:28 mpanetta rackspace
20:28 XenophonF UtahDave: thanks
20:28 iggy not familiar with their setup
20:28 XenophonF UtahDave: i'm more than willing to do the work to fix this
20:29 UtahDave XenophonF: yeah, that's cool.  sometimes getting someone to point you in the right direction is very helpful, though
20:29 mpanetta iggy: I don't think it is related to provider.
20:29 XenophonF I'm just not sure what's broken.
20:29 XenophonF yeah
20:30 alex-mesos joined #salt
20:31 shookees joined #salt
20:32 mpanetta iggy: Looks like patches for the issue were merged in to 2014.7
20:32 iggy mpanetta: yeah, it's definitely a bug, but it seems to be exacerbated by something
20:32 mpanetta iggy: I think master restarts
20:32 iggy because we didn't see it all the time
20:34 cpowell iggy: we noticed it with master restarts, more than likely not clean restarts
20:34 mpanetta iggy: We were fine with 0.16.4 until the master rebooted uncleanly.
20:34 mpanetta I assume the OS removes /var/cache on reboots
20:34 mpanetta well cleans it
20:35 iggy we never restart the instances (and very rarely restart the process)
20:35 mpanetta Hmm, then maybe that's not it :(
20:35 cpowell ...it wasn't on purpose lol
20:35 mpanetta haha
20:35 iggy part of my rolling out the new clusters is to roll out a fresh salt master with everything else
20:36 smcquay joined #salt
20:42 aw110f joined #salt
20:42 n8n joined #salt
20:43 UtahDave XenophonF: Hey, terminalmage should have PM'ed you.
20:44 XenophonF got it
20:44 XenophonF thanks UtahDave
20:46 debian112 joined #salt
20:46 cliff-hm joined #salt
20:47 AmrMostafa joined #salt
20:47 debian112 joined #salt
20:48 * robawt highfives UtahDave
20:49 * UtahDave highfives robawt!
20:49 faust joined #salt
20:49 clemensb joined #salt
20:54 DaveQB joined #salt
20:54 duncanmv joined #salt
20:55 jessec joined #salt
20:56 AmrMostafa UtahDave: thanks again for clearing that up
20:56 UtahDave you're welcome, AmrMostafa!
20:57 kivihtin joined #salt
20:58 KyleG joined #salt
20:58 KyleG joined #salt
20:58 t0rrant joined #salt
21:00 mephx joined #salt
21:00 alex-mesos joined #salt
21:01 ahammond I'm turning a .sls formula into pydsl. How do I handle the map.jinja import stuff? Is there a pattern?
21:03 XenophonF UtahDave: I think I'm good to go.  terminalImage found out that it's patched in 2014.7, so I'm going to deploy states/module.py via _states.
21:04 fragamus joined #salt
21:04 smcquay joined #salt
21:07 mephx joined #salt
21:08 XenophonF and just to close the loop, calling pkg.refresh_db from module.run works on FreeBSD now
21:08 XenophonF so fixed in 2014.7 :)
21:09 smcquay_ joined #salt
21:13 rap424 joined #salt
21:14 UtahDave XenophonF: ah, cool!  Thanks for helping track that down, XenophonF
21:16 murrdoc Failed to configure repo 'deb [ arch=amd64 ] http:/HI.mom/sup/upbunut/ precise main': request() got an unexpected keyword argument 'stream'
21:16 murrdoc thoughts ?
21:17 iggy http://
21:19 nethershaw joined #salt
21:19 murrdoc (headbang)
21:19 murrdoc :D
21:19 kermit joined #salt
21:19 murrdoc figured it out
21:19 murrdoc repo was broke
21:20 smcquay joined #salt
21:21 giantlock joined #salt
21:32 XenophonF UtahDave: my pleasure
21:33 tf198_ joined #salt
21:33 XenophonF thanks for your support
21:35 dooshtuRabbit1 joined #salt
21:35 JerryB joined #salt
21:36 glyf joined #salt
21:36 dooshtuRabbit joined #salt
21:38 sdmflk joined #salt
21:39 perfectsine joined #salt
21:42 JerryB Morning! I had some problems earlier today but haven't had time to put together a log until now. The issue is that salt worked fine until this morning but now all commands stall. I get AES is not defined but also timeout and failure to authenticate. Hare's a compiled log of what happens https://bpaste.net/show/41a2d6ddc139
21:43 sschwartz_ee OK. Just checking here before I go bug the cherrypy people: anyone got rest_cherry.py up and running, and what did you use for your configs to make it visible to the outside world?
21:43 whiteinge sschwartz_ee: what are you seeing?
21:44 sschwartz_ee It runs fine from the localhost, but everywhere outside gives me a connection timed out.  It's in EC2 and the security group has port 8888 open.
21:46 loz-- joined #salt
21:47 geekatcmu iptables
21:48 sschwartz_ee geekatcmu: Very turned off. ;)
21:49 geekatcmu Service bound to localhost.
21:49 sschwartz_ee geekatcmu: From salt-api.conf: port:8888 host: 0.0.0.0
21:50 geekatcmu Do you have IPv4 configured, or just IPv6?
21:50 sschwartz_ee IPv4.
21:50 geekatcmu Then either you have a gateway host that blocks it or it isn't in the security group you think it is.
21:51 geekatcmu You're running out of reasons why it doesn't work.
21:52 sschwartz_ee That's what I was figuring. Very frustrating. Will keep digging, and thank you for the help!
21:52 JerryB I've google all the errors that I'm getting but I only find separate causes, I can't find anything that makes sense for these symptoms, especially as it worked fine until now
21:53 lopezio joined #salt
21:53 lopezio left #salt
21:54 lopezio joined #salt
21:55 fragamus joined #salt
21:56 lopezio hi.. anyone here using freebsd? i'm trying to use the 'fromrepo' in the pkg.installed state but it seems to get ignored
21:57 srage joined #salt
22:08 bhosmer joined #salt
22:09 mosen joined #salt
22:11 span joined #salt
22:12 aw110f joined #salt
22:13 bhosmer joined #salt
22:17 t0rrant joined #salt
22:18 mephx joined #salt
22:18 glyf joined #salt
22:19 thedodd joined #salt
22:21 Ryan_Lane1 joined #salt
22:25 CeBe joined #salt
22:25 span joined #salt
22:26 jhauser joined #salt
22:32 fragamus joined #salt
22:34 ndrei joined #salt
22:34 ndrei_ joined #salt
22:34 ndrei__ joined #salt
22:48 glyf joined #salt
22:49 smcquay joined #salt
22:49 fxhp joined #salt
22:50 robawt fxhp in the house
22:50 UtahDave woot
22:50 mosen we're up all night to get salty
22:51 robawt haha mosen
22:51 scoates joined #salt
22:52 fxhp sup
22:52 fxhp robawt: my Internet is terrible now that AT&T sold my state to frontier
22:52 fxhp : (
22:53 robawt haha
22:53 fxhp Its really bad, I used to never go offline, now TV and Internet drops multiple times a day.
22:53 Pixionus what state?
22:54 fxhp CT
22:54 fxhp They sold all land lines to frontier (including Uverse)
22:54 fxhp There is no competition, its terrible.
22:55 fxhp </rant>
22:55 Pixionus Lame.  You're right there in New England near all the trade and between the big cities.
22:56 Pixionus Monopolies suck, and current legislation just feeds them
22:56 fxhp Oh yeah, totally.
22:58 fxhp We need to treat Internet like water and eletricity.
22:58 mordonez joined #salt
22:58 fxhp IMHO
22:58 fxhp to get out of this mess
22:58 KyleG I've been trying to get AT&T to activate my connection for about 2 weeks now lol
22:58 KyleG They sent me the modem
22:58 KyleG and were supposed to send techs
22:58 KyleG and yeah…got a call twice from the techs…but they never showed up
22:59 fxhp Yeah, AT&T isn't any better, but they just sold me to a worse company.
22:59 KyleG luckily AT&T is my backup connection @ home
22:59 KyleG I'm actually VERY happy with TWC
22:59 KyleG They lower my bill whenever I ask, they're always giving me free upgrades and shit
22:59 KyleG and they're on time, and their people are good at customer service (now-a-days)
22:59 KyleG A++
22:59 fxhp "would buy again"
23:00 * fxhp smiles
23:00 KyleG for sure lol
23:00 fxhp You all see that unmanned rocket explosion ?
23:04 aquinas joined #salt
23:08 Pixionus yup
23:08 Pixionus was just watching that actually
23:10 Pixionus electricity is still monopoly in much of the us
23:10 murrdoc joined #salt
23:10 fxhp thats true too, I'm not sure what the fix is/
23:10 Pixionus In fact, I don't think I have a choice for whom I pay for electricity or water, so Internet is the same over here
23:10 badon joined #salt
23:12 Pixionus And it's hard to do it any differently because it takes a company to build an infrastructure.  If it was one company's business to build the infra and other companies provided the water and electricity to the grid then there would be a little more competition.  Not sure there is really an analogue for the Nets though...
23:13 KyleG the country needs to be like a meet-me room
23:13 KyleG oh you want verizon? *takes fiber strand for your house/apartment and plugs it into verizon port*
23:13 KyleG oh you want <x> provider same thing
23:13 Pixionus That's a very costly infrastructure to build...
23:13 KyleG yes, yes it is
23:14 KyleG but it'd work
23:14 KyleG lol
23:14 Pixionus but yes, I see it's benefits
23:14 KyleG If money is the only obstacle, that's a good problem.
23:14 fxhp Plus the major companies purposely don't mess with each others "turf"
23:14 KyleG I thought that was due to agreements
23:14 Pixionus I could choose to upgrade my local wire to fiber and use googles fiber ports at the local substation or stick with basic and choose from whomever was available there.
23:14 fxhp Its like a bunch of back room agreements.
23:15 X86BSD joined #salt
23:15 fxhp which put the customer last and profits first
23:15 Pixionus That's business for you.
23:15 geekatcmu There are a bunch of US cities doing exactly that under a "dig once" rule.  Any time the street gets dug up, new cable conduit gets run, with the city owning it.  To be leased out to whomever wants to play.
23:15 Pixionus Automation can mean we don't have to work for basic necessities or it can mean we don't have work so we can't afford basic necessities.  What do you think companies will pay for?
23:16 geekatcmu fxhp: yes, that's pretty much how things work.
23:16 KyleG San Antonio and Austin seem to be doing well
23:16 KyleG they have Grande, TWC, and AT&T and I think some others
23:16 KyleG cool stuff, all offer 1 Gb i think
23:17 KyleG or at least 300Mbit
23:17 fxhp : /
23:18 fxhp I have 12Mbit
23:18 Pixionus Dallas and Houston similarly had choices.  Now that I am living in Phoenix, however, I have no choice.  I have SRP for Elec, and COX for the Nets
23:19 iggy Houston doesn't really have choice for internet (elec, yes)
23:20 Pixionus That said, I get something crazy like 40mbps dl (on average) and basic cable for $30/m because my apartment complex has a deal with them.
23:20 iggy there are some spots that have one (different than the rest of the city) choice, but for the most part it's Comcrap
23:21 Pixionus I think I had choice when I was living in SugarLand
23:21 KyleG I need to upgrade my wireless router…I'm not able to pull my full connection speed at home, at least on download lol
23:21 KyleG I have 100Mbit down 5 up, which wtf why still only 5 up -_-
23:21 iggy well... that's not exactly Houston is it ;)
23:22 Pixionus True iggy.  True.
23:22 Pixionus But for the uninitiated, it may as well be.
23:22 jonbrefe joined #salt
23:23 iggy I actually live up north (of Houston) now, and it's the same as downtown... only not Comcast
23:23 iggy so at least I don't have to deal with them anymore
23:24 bagackiz joined #salt
23:25 Pixionus one of my friends is rich and lived out in a private land spot south of us and his net was worse than anyone elses.  They could afford to lay down their own line but the only provider for the area was some shitty company I can't remember that was always having problems and worked off of last decades routers...
23:26 Pixionus I swear they used all the hand me down hardware from the other bigger providers, and it SUCKED.
23:26 Pixionus sorry leaving out my apostrophes everywhere.
23:26 vividloop joined #salt
23:26 Pixionus though, I suppose that's the least of my grammer problems.
23:27 Pixionus grammar
23:27 Pixionus :P
23:27 thayne joined #salt
23:28 iggy yeah, you get into "boondocks" real quick past Sugar Land
23:28 KyleG If I was rich and had a private island within 10 miles of land, I'd probably get a 10Gbe connection on land within view of the island (not view of your eyes, but in the general direction), and then throw up my own little mini internet tower
23:28 KyleG http://www.ubnt.com/airfiber/airfiber5/
23:28 KyleG 100 KM+ range
23:29 KyleG that's 62 miles
23:29 KyleG bam done
23:29 KyleG rich guy gets internet
23:29 geekatcmu Have fun licensing the spectrum
23:29 KyleG it uses the 5 Ghz unlicensed band.
23:29 KyleG says so right there on the page
23:30 Pixionus I understand there are a number of us (sysadmins, etc) that live up in Alaska and the crazy north.  I kinda want to do that some day, but I'm afraid the net will suck and of course, can't keep a remote job if you have spotty internet.
23:30 geekatcmu Yep, I didn't go down far enough
23:31 KyleG I'm actually looking to get into this business with a buddy possibly
23:31 KyleG Deals w/ apartment complexes and whatnot, providing internet
23:31 KyleG over zeh air
23:31 jonbrefe joined #salt
23:32 iggy Houston actually started building out a lot of that (my old boss did a lot of it), then the state passed something that said communities couldn't be in the internet biz
23:34 Pixionus iggy, you know tx/rx labs perchance?
23:37 Pixionus if so I was wondering how they were doing, if not, you should check out one of the better makerspaces in the US.
23:38 Pixionus Also, for a totally unrelated and yet pertinent question: is test.ping just always going to be unreliable or is there something I can do to fix it?
23:53 murrdoc so the update-alternatives state needs love
23:55 mgw joined #salt
23:59 n8n joined #salt

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