Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-10-02

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

All times shown according to UTC.

Time Nick Message
00:00 honestly and then see if it is the default service provider for your distro
00:00 piffio joined #salt
00:00 honestly if not, need to set it as provider in config
00:02 atealtha_ honestly: thanks
00:12 mianos joined #salt
00:18 cshuman Hi everyone. Is there a way to cache package downloads through the master so packages are only downloaded from the web once and the minions grab packages through the master over a local (faster) connection?
00:18 msheiny joined #salt
00:19 cachedout_home cshuman: For certain distributions (ala Ubuntu) you could use something like apt-cacher. Granted, not a salty solution per-se but it's one option.
00:21 honestly could set up a caching proxy on a random server
00:21 UtahDave joined #salt
00:24 packeteer cshuman: squid or apt mirror
00:24 packeteer that issue was solved many years ago :)
00:25 mianos joined #salt
00:25 cshuman ok. thanks. I was just wondering if there was something built into salt.
00:27 packeteer you could prob make it work, but that specific wheel already works well :)
00:28 ifnull_ anyone else having trouble with gitfs in 0.17.0. I'm getting "No Top file or external nodes data matches found"
00:29 ifnull_ master: http://pastebin.com/kBcL67k9
00:31 UtahDave ifnull_: have you looked at the issue tracker? I think there's an open issue about something related
00:31 ifnull_ I'm looking at this one: https://github.com/saltstack/salt/issues/7493
00:32 ifnull_ not sure if this is it. still looking
00:33 xuru joined #salt
00:33 UtahDave k
00:36 ifnull_ doesn't appear related
00:37 teebes joined #salt
00:38 ifnull_ in my case i don't even see the attempt to connect to git
00:38 ifnull_ http://pastebin.com/raw.php?i=erpAjaq8
00:43 jaequery joined #salt
00:45 oz_akan_ joined #salt
00:46 oz_akan_ joined #salt
00:51 intchanter joined #salt
00:53 mannyt joined #salt
01:04 alekibango joined #salt
01:14 mianos joined #salt
01:20 Lue_4911 joined #salt
01:20 mannyt joined #salt
01:20 ef joined #salt
01:23 isomorphic joined #salt
01:26 redondos joined #salt
01:26 redondos joined #salt
01:32 mianos joined #salt
01:39 shinylasers joined #salt
01:40 deepakmd_oc joined #salt
01:41 shinylasers joined #salt
01:44 imaginarysteve joined #salt
01:44 Gifflen joined #salt
01:44 Lue_4911 joined #salt
01:48 Thiggy joined #salt
01:49 jefftriplett left #salt
01:49 mwillhite joined #salt
01:49 dthom91 joined #salt
01:50 cro joined #salt
01:55 forrest joined #salt
01:57 ipmb joined #salt
02:05 sixninetynine joined #salt
02:06 josephholsten joined #salt
02:20 Thiggy joined #salt
02:22 kenbolton joined #salt
02:22 dthom91 joined #salt
02:29 josephholsten joined #salt
02:29 Thiggy joined #salt
02:34 redbeard2 joined #salt
02:36 jaequery joined #salt
02:37 redondos joined #salt
02:40 cro joined #salt
02:43 cro joined #salt
02:49 juicer2 joined #salt
02:50 juicer2 left #salt
02:51 dthom911 joined #salt
02:52 juicer2 joined #salt
03:02 juicer2 joined #salt
03:03 josephholsten joined #salt
03:06 serapath joined #salt
03:39 rgarcia_ joined #salt
03:44 Ryan_Lane joined #salt
03:46 higgs001 joined #salt
03:52 d10n joined #salt
03:56 linjan_ joined #salt
03:57 cro left #salt
04:02 zach Just saw there's a SaltConf this coming January. Anyone know the cost?
04:02 UtahDave joined #salt
04:03 * zach bets UtahDave would know
04:03 zach UtahDave: How much is the SaltConf in January? If you know
04:14 Lue_4911 joined #salt
04:16 mianos joined #salt
04:17 UtahDave zach: $799 early-bird registration, $899 standard registration
04:17 forrest Are those the final numbers UtahDave??
04:17 forrest I better get a proposal in then, damn
04:17 zach fantastic, I'll have to register
04:17 * zach just got budget approval for 2014 :D
04:18 forrest Hah, that's nice zach
04:18 UtahDave forrest: I believe so.
04:18 forrest Out of pocket for me.
04:18 zach forrest: we use salt extensively at work now, and I want to utilize more of it. I know I can sell my boss on it, he's already amzed that I was able to automate our rickity environment
04:19 Ryan_Lane joined #salt
04:19 forrest zach, ahh ok, yea I'm fighting that battle (losing it to puppet)
04:19 zach I hate puppet and chef with a passion
04:19 forrest As do I
04:19 zach "we're going to use saltstack, and it's already deployed" - me
04:19 forrest but it's tough to sell someone who hasn't used config management before
04:20 zach "Okay then, guess we're going to use Saltstack..." - boss
04:20 forrest heh
04:20 UtahDave cool, zach. It's going to be awesome
04:20 zach UtahDave: can't wait, I'll put the other ones off on my list to go to saltconf after how much time it has saved me
04:20 forrest UtahDave, how many attendees does the venue handle?
04:21 UtahDave forrest: Yeah, get your talk proposal in!  :)
04:21 zach Where do you even register? I found where you can register to be a speaker but that's it
04:21 forrest I'll work on during my free time in the evenings at devops day atlanta, then try to get it in this weekend.
04:21 forrest *work on it
04:21 UtahDave forrest: I'm not sure yet
04:21 UtahDave cool
04:22 UtahDave zach: let me see if I can track it down.
04:22 zach UtahDave: thanks
04:23 Katafalkas joined #salt
04:23 zach http://lanyrd.com/2014/saltconf/ that is all I could find
04:26 UtahDave zach: Hm. That might be where we're going to have people sign up, but I'm not sure. I'm going to have to check with the office tomorrow morning.
04:27 zach UtahDave: thanks, I appreciate it
04:27 UtahDave zach: want me to email with what I find out?
04:27 zach UtahDave: I'm going to beg and plead to attend :)
04:27 zach Sure, I'll PM it to you
04:27 UtahDave great
04:28 zach done
04:30 cachedout_home Once the details are sorted, adding details to the IRC motd for this channel might generate interest as well. :]
04:31 jalbretsen joined #salt
04:31 mianos joined #salt
04:33 UtahDave cachedout_home: great idea!  I'll make sure to update that.
04:35 redondos joined #salt
04:44 Niichan joined #salt
04:49 Ryan_Lane joined #salt
04:49 krissaxton joined #salt
04:54 oz_akan_ joined #salt
04:58 CansaSCity joined #salt
04:58 krissaxton joined #salt
05:04 * rmt yawwwns.
05:05 woebtz joined #salt
05:09 packeteer pity the conference is half a world away :(
05:11 jcockhren how fitting for the conference to be in Salt Lake City
05:13 joonas jcockhren: that's also where the salt stack guys are from ;)
05:13 joonas or the company is, rather
05:14 Furao salt lake city in january is like orlando
05:14 Furao beach, sun, you can buy beer 7 days a week
05:14 jcockhren lol
05:14 cachedout_home Yes, it's exactly like that.
05:14 jcockhren I plan to be there
05:14 Furao the conference should be in basement of CEO
05:15 Furao they can save on rent
05:17 Furao I hope this will be recorded and available to watch online
05:18 cachedout_home SLC in the winter is actually pretty sweet. (No kidding.)
05:18 Furao I left my home country (Canada) 2 years  ago to never ever have winter again
05:18 cachedout_home In your case, that might be a particularly tough sell then. ;]
05:19 Furao I'm in Malaysia now and it's 29c/85f
05:19 Furao and last time I transited trough US they almost rape me at custom
05:20 Ryan_Lane joined #salt
05:20 packeteer yeah, I hate transiting through anything US related
05:20 deepakmd_oc joined #salt
05:20 packeteer had to go through Guam last year, was so painful
05:21 faldridge joined #salt
05:22 Furao it was more pleasant to be interrogated for terrorism in Bulgaria than going trough US (Obama's tax farm) custom
05:23 mannyt joined #salt
05:24 krissaxton joined #salt
05:24 CansaSCity joined #salt
05:25 exanimo joined #salt
05:26 krissaxton joined #salt
05:28 CansaSCity joined #salt
05:34 packeteer ok, time to reaize my linode :/
05:41 cachedout_home left #salt
05:42 redondos joined #salt
05:50 Ryan_Lane joined #salt
05:53 krissaxton joined #salt
06:04 middleman_ joined #salt
06:06 joehh Furao: have you run 0.17.0 through your tests yet?
06:09 cshuman joined #salt
06:12 redondos joined #salt
06:13 mianos joined #salt
06:19 cshuman joined #salt
06:20 shinylasers joined #salt
06:20 Ryan_Lane joined #salt
06:25 TheCodeAssassin joined #salt
06:32 clone1018_ joined #salt
06:33 ml_1 joined #salt
06:35 MeanderingCode joined #salt
06:39 druonysuse joined #salt
06:47 CansaSCity joined #salt
06:48 matanya joined #salt
06:53 matanya joined #salt
06:54 Furao joehh: no because http://ppa.launchpad.net/saltstack/salt/ubuntu/pool/main/s/salt/ no 0.17 there yet
06:54 Furao but all my stuff had been tested with 0.16.4
06:55 Furao cp.push is fixed now :)
06:55 Ryan_Lane joined #salt
06:57 TheCodeAssassin joined #salt
07:00 f47h3r Good evening, is anyone else working on the salt-ssh passwd bug tonight?
07:01 f47h3r I can confirm that I'm getting "Permission Denied", but it's almost working!!!
07:01 f47h3r the user/pass is correct
07:03 f47h3r joined #salt
07:04 ronc joined #salt
07:05 s0undt3ch joined #salt
07:07 joehh Furao: that is understandable - normally you uncover issues pretty quickly.
07:07 joehh good to hear cp.push is working
07:13 joehh f47h3r: which os? Ive had it working with debian
07:13 Furao as soon as 0.17.0 pkgs are in ppa, I'll mirror it, branch and start a job in jenkins to see what is broken. it will take ~ 10 hours
07:14 Furao 0.15.3 test suite need 13 hours average, 10 hours with 0.16.4
07:16 joehh Furao: what do you use to mirror the ppa?
07:16 Furao wget :P
07:17 joehh :)
07:17 f47h3r joehh: RHEL... I'm Looking at the commit ajithhub just submitted like half an hour ago
07:17 joehh simple but effective!
07:17 Furao http://archive.robotinfra.com/mirror/salt/0.16.4/
07:17 f47h3r basically passwd auth is failing w/ a "Permission Denied"
07:18 Furao https://gist.github.com/bclermont/61c960cdb3b66b0e904c
07:18 joehh nice - thanks for that
07:19 joehh I've had a few requests for mirroring of the debian.saltstack.com archive, but haven't known the best way to point them
07:19 Furao I used to do it with reprepro, but I was loosing salt apt key
07:19 jgelens joined #salt
07:19 Furao almost all my files used to deploy my states are mirrored, except ruby gems and .deb
07:19 f47h3r has anyone pulled these to see if they work?: https://github.com/saltstack/salt/pull/7557
07:20 eX4n1m0 joined #salt
07:20 Furao mirror gems look harder than pip
07:20 krissaxton joined #salt
07:20 Furao and .deb, there is so much mirrors around already
07:20 Furao and it can be quite huge in term of disk space
07:20 joehh that was the direction (reprepro) I was thinking of pointing, but the key issue wasn't clear
07:22 ronc joined #salt
07:22 pdayton joined #salt
07:26 Ryan_Lane joined #salt
07:26 bejer Hi, have anyone come across a state/formula for gerrit and/or gitlab? (preferably for Debian Wheezy)
07:27 balboah joined #salt
07:29 joehh Furao: thanks very much for the gist, it will be very useful
07:37 az87c joined #salt
07:37 faldridge joined #salt
07:41 TheCodeAssassin joined #salt
07:44 krissaxton joined #salt
07:45 boite joined #salt
07:52 matanya joined #salt
07:56 Ryan_Lane joined #salt
07:57 druonysuse joined #salt
07:57 druonysuse joined #salt
08:00 ksalman joined #salt
08:02 cetex_ I've got an issue.
08:03 cetex just reinstalled a server (via debian preseed/pxe) and set it up so it pulls in salt-minion, but the minion won't use fqdn when trying to register to the salt-master
08:03 cnelsonsic joined #salt
08:05 cetex any ideas where to start looking?
08:06 felixhummel joined #salt
08:07 packeteer joined #salt
08:08 TheCodeAssassin joined #salt
08:09 cetex python -c "import socket; print(socket.getfqdn())"
08:09 cetex gives the fqdn.
08:09 cetex but the minion reports in to the master with hostname only. :/
08:09 eX4n1m0 joined #salt
08:11 backjlack joined #salt
08:17 TheCodeAssassin joined #salt
08:18 dthom91 joined #salt
08:22 TheCodeAssassin joined #salt
08:24 cetex salt-minion 0.17.0
08:27 Ryan_Lane joined #salt
08:28 cetex did: apt-get remove --purge salt-minion salt-common; rm -r /etc/salt; apt-get install salt-minion
08:28 cetex still uses hostname instead of fqdn.
08:32 ggoZ joined #salt
08:35 TheCodeAssassin joined #salt
08:36 zooz joined #salt
08:38 TheCodeAssassin joined #salt
08:38 TheCodeAssassin joined #salt
08:45 Jason-AVST joined #salt
08:47 qba73 joined #salt
08:49 TheCodeAssassin joined #salt
08:50 TheCodeAssassin joined #salt
08:52 TheCodeAssassin joined #salt
08:55 krissaxton joined #salt
08:57 Ryan_Lane joined #salt
08:58 TheCodeAssassin joined #salt
08:59 krissaxton joined #salt
09:05 shine joined #salt
09:05 shine hello :)
09:08 TheCodeAssassin joined #salt
09:18 honestly joined #salt
09:27 Ryan_Lane joined #salt
09:38 jinnko joined #salt
09:39 bhosmer joined #salt
09:45 joehh cetex: http://docs.saltstack.com/faq.html#how-does-salt-determine-the-minion-s-id
09:46 joehh I think this has changed recently
09:51 krissaxton joined #salt
09:53 p3rror joined #salt
09:53 cetex it seems it did.
09:54 cetex when i downgraded to 16.3 it works as it should.
09:57 Ryan_Lane joined #salt
09:58 krissaxton joined #salt
10:00 krissaxton joined #salt
10:04 jinnko joined #salt
10:11 krissaxton joined #salt
10:12 tempspace_ joined #salt
10:14 joehh cetex: can you set it in the minion config?
10:28 Ryan_Lane joined #salt
10:32 paluh joined #salt
10:37 cetex joehh: that should work, but it won't allow me to do a fully automatic deploy.
10:38 paluh joined #salt
10:38 xerxas joined #salt
10:44 minaguib joined #salt
10:48 jeblair joined #salt
10:48 paluh joined #salt
10:52 SpX joined #salt
10:54 joehh yeah its a tricky one - I wonder if there is a way to change it once the minion has connected and the key has been added
10:55 joehh ie update both the minion config and the name the master is using
10:58 Ryan_Lane joined #salt
10:59 joehh cetex: I've been thinking about the pros and cons of using debconf for setting the master address and minion name
10:59 joehh in your preseed environment, can you set debconf values?
11:02 peter joined #salt
11:07 SpX joined #salt
11:07 balboah joined #salt
11:13 MrTango joined #salt
11:14 krissaxton joined #salt
11:17 canci joined #salt
11:18 mwillhite joined #salt
11:20 oz_akan_ joined #salt
11:20 adepasquale joined #salt
11:21 blee joined #salt
11:21 jslatts joined #salt
11:23 krissaxton joined #salt
11:27 TheCodeAssassin joined #salt
11:28 movedx_ joined #salt
11:29 Ryan_Lane joined #salt
11:30 diegows joined #salt
11:35 sgviking joined #salt
11:41 backjlack joined #salt
11:44 eX4n1m0 joined #salt
11:48 paluh_ joined #salt
11:49 linjan_ joined #salt
11:56 giantlock joined #salt
11:59 Ryan_Lane joined #salt
12:01 canci joined #salt
12:01 cetex joehh: sure.
12:01 cetex joehh: but better yet would be to fix the regression so the minion uses fqdn ;)
12:04 joehh I feel your pain, but I'm not so sure saltstack see it as a regression. My understanding is there was a bunch of
12:05 joehh differing behaviours and they picked one and standardised it
12:06 joehh did you just send an email to the mailing list on this issue or was that someone else?
12:07 Boohbah joined #salt
12:07 shomodj joined #salt
12:08 joehh For my own purposes, I prefer it without the fqdn, but I see your side as equally valid...
12:08 joehh I don't know how the two can be reconcilled though
12:08 cetex debconf option? :)
12:08 cetex that sets an option in the minion config file
12:08 piffio joined #salt
12:08 cetex would also like a debconf option to set the master on install
12:08 joehh that is kind of where I am heading (and where I can control too :) )
12:08 cetex :)
12:09 Kholloway joined #salt
12:09 joehh do you think the debconf option be hostname v fqdn v custom or just presented with a choice for setting?
12:10 joehh given you are preseeding, I think you would be happy with just a choice to set directly
12:12 gamingrobot joined #salt
12:15 akitada joined #salt
12:16 goki joined #salt
12:16 ipmb joined #salt
12:17 deedubs joined #salt
12:18 gildegoma joined #salt
12:19 Odd_Bloke joined #salt
12:23 krissaxton joined #salt
12:30 Ryan_Lane joined #salt
12:31 patrek joined #salt
12:33 felixhummel joined #salt
12:33 krissaxton joined #salt
12:34 mannyt joined #salt
12:36 jslatts joined #salt
12:38 curtisz joined #salt
12:40 felixhummel joined #salt
12:42 tyler-baker joined #salt
12:42 tyler-baker joined #salt
12:43 honestly joined #salt
12:43 jefftriplett joined #salt
12:47 pkimber joined #salt
12:48 TheCodeAssassin joined #salt
12:50 linuxnewbie joined #salt
12:54 kenbolton joined #salt
12:54 wkf_ joined #salt
13:00 Ryan_Lane joined #salt
13:01 dthom91 joined #salt
13:05 juicer2 joined #salt
13:06 cetex joehh: Two options perhaps. One is hostname vs fqdn vs custom, then if you select custom you get asked the custom name.
13:07 cetex then i could simply select fqdn and be done with it. :)
13:07 cetex and others could select hostname..
13:08 elfixit joined #salt
13:11 toastedpenguin joined #salt
13:14 jeffasinger I'm having a hard time getting keys accepted in a 2 machine vagrant setup, I'm using seed_master with two hosts. However, the minion ids for the pre seeded keys seem to be incorrect, how do I set those?
13:14 krissaxton joined #salt
13:15 racooper joined #salt
13:15 joehh cetex: That was roughly the conclusion I came to (either that or just a single text entry defaulting to hostname/fqdn)
13:15 joehh I'm currently playing with debconf to see what works well
13:16 cetex :
13:16 cetex :)
13:16 joehh jeffasinger: you can set the id in /etc/salt/minion.conf or in /etc/salt/minion.conf.d/local.conf or similar
13:16 cetex joehh: the issue though is that hostname / fqdn might not be correct in the debian installer (which is where i'm installing the minion)
13:16 cetex since it's in a chroot and all.
13:17 lacrymology joined #salt
13:17 joehh true - so a straight text box would probably suit the most in that situation
13:17 joehh or the choice then text box
13:17 brianhicks joined #salt
13:19 cetex i'd vote for choice, then textbox.
13:20 cetex otherwise i would probably have to set the textbox value to the hostname through debconf.
13:20 cetex when preseeding.
13:23 joehh probably the biggest issue with the choice is I suspect that I may not be able to show what will be selected at the time you choose
13:24 MrTango joined #salt
13:24 joehh so you wouldn't be able to see what the script thinks the fqdn is
13:24 joehh anyway - I'll learn more and see where I end up
13:25 joehh email me at joehealy@gmail.com if you'd like to test a pre-release
13:25 joehh It's definitely one I'd like to get feedback on before releasing to the wild
13:25 joehh one = a change
13:26 logix812 joined #salt
13:29 mahul joined #salt
13:30 jeffasinger joehh: Thanks, that worked. Ideally I would have liked to change the id of the pre seeded key that vagrant is uploading, but I'll settle for this for now, and read through the salty-vagrant source at some point to get it working how I want
13:31 mgw joined #salt
13:31 Ryan_Lane joined #salt
13:31 mahul I am looking for some SaltStack patterns to use with EC2 in a auto-scale mode. I can have all the infrastructure software built into the image (Apache, Tomcat, Ruby, SaltStack etc) but I am not if there are best practises for deploying application code - any help/pointers will be appreciated.
13:32 mannyt joined #salt
13:33 mahul Do I use cmd and exec a git pull or do I have use gitfs and have the Master pack the code into a tar and then have the minions copy this over?
13:34 kenbolton joined #salt
13:34 joehh mahul: other option is git.latest state
13:36 mahul yes, I tried that and it works but it keeps asking me for username/password when going over https and I havent figured out best practises (security concerns) for have a ssh key if I want to use the ssh transport.
13:37 joehh mahul: I'm only directly familiar with the hg state and set usernames/passwords  in ~/.hgrc
13:37 joehh not too sure how to do it for git - maybe someone else will know
13:37 jeffasinger mahul: what I do for the ssh keys is a get a deploy only key setup, and just use the same key for all the machines, seeded using salt. It's not perfect, but it means in a breach people would only get read only access to your repository
13:38 mahul jeffasinger: Thanks, I could go with that for the short term.
13:38 mahul joehh: Will check out .hgrc
13:39 mahul I will test the approach of having the master keep the tarball and let minions work off of that tar ball.
13:39 jeffasinger Anyone have ideas on what might be the easiest way to get a state that handles DB migrations for me?
13:40 lemao joined #salt
13:43 redbeard2 joined #salt
13:46 aleszoulek joined #salt
13:48 scalability-junk mahul: usually with git + ssh you have ssh read only access or so called deploy keys for that
13:48 scalability-junk you can distribute these keys to your minions via file modules.
13:51 kaptk2 joined #salt
13:52 Brew joined #salt
13:53 ml_11 joined #salt
13:55 micah_chatt joined #salt
14:01 Ryan_Lane joined #salt
14:03 mahul scalability-junk: thanks - am exploring that option.
14:06 elfixit joined #salt
14:07 ebodine joined #salt
14:14 juicer21 joined #salt
14:15 ldlework Is there a no-op state?
14:18 nahamu what do you want to accomplish with a no-op state?
14:19 nahamu there's test.ping which you can issue from the master to check on the minions...
14:19 ldlework nahamu: I have a database state that either includes .mysql or .postgres
14:19 kermit joined #salt
14:19 ldlework the states are different in each file
14:19 ldlework but in a different state, I want to require that the database is setup
14:19 ldlework you for some reason can't require whole states
14:20 ldlework and so I need a generic no-op state at the end of the 'database' state, so I can require it generically from other states
14:20 ldlework I don't want to have a conditional everywhere I need to require that a database be created
14:20 nahamu hopefully someone smarter than I am will chime in now that you've more clearly described what you're trying to do. :)
14:21 ldlework heh
14:21 m_george|away joined #salt
14:21 ldlework right now I'm simply doing a cmd.run with echo "database done"
14:21 m_george left #salt
14:21 ldlework it'd be nice if I could require a whole state..
14:22 honestly ldlework: should be able to do that with jinja
14:22 redondos joined #salt
14:22 ldlework honestly: I don't want to have an if conditional anywhere I require the database
14:24 honestly you have something like db_mysql.sls and db_postgres.sls?
14:24 honestly they could both have a state called "db_setup"
14:24 ldlework honestly: but then you need to couple the state name with a state function
14:24 honestly and you could require that state from anywhere
14:24 ldlework like you have to do something like
14:24 ldlework require:
14:24 ldlework - mysql_grant: db_setup
14:24 ldlework in the postgres version, its not called mysql_grant:
14:25 ldlework there's no way to just require the entirety of "db_setup"
14:25 ldlework at least as far as I can tell wrt to the docs
14:25 honestly hmm
14:25 honestly try making an empty state that only has requires in it
14:25 ldlework huh?
14:25 ldlework oh
14:25 ldlework nice
14:25 ldlework genius
14:26 honestly does it work?
14:28 ldlework honestly: well...
14:28 ldlework what function do you couple with that state?
14:28 ldlework how do you require that state that only has requires?
14:28 mgw joined #salt
14:28 honestly isn't there a require for a pure state?
14:28 ldlework no!
14:29 ldlework at least it isn't mentioned anywhere I could find :)
14:29 honestly damn.
14:29 honestly let's check!
14:29 ldlework you can do sls: state-filename
14:29 ldlework but that doesn't work when that file only contains includes
14:29 honestly hrm
14:29 honestly sounds like issue-file-time to me
14:29 ldlework (in this case my database/init.sls just contains a conditional include)
14:30 ldlework I guess I'll use the cmd.run for now
14:31 Ryan_Lane joined #salt
14:33 JaredR joined #salt
14:34 jalbretsen joined #salt
14:35 honestly please file an issue
14:35 honestly this is an important usecase imo
14:36 mwillhite joined #salt
14:36 LGSilva joined #salt
14:37 LGSilva hey guys. Is it possible for the salt-master generate a pillar with all the IPs from all iterfaces from all minions?
14:38 bhosmer joined #salt
14:38 mgw joined #salt
14:39 oz_akan_ joined #salt
14:40 cnelsonsic joined #salt
14:40 lemao joined #salt
14:41 cron0 joined #salt
14:43 ldlework How can I deal with host verification issues for example like when cloning with git?
14:43 dthom91 joined #salt
14:44 cnelsonsic left #salt
14:44 anteaya joined #salt
14:44 ldlework nm
14:45 StDiluted joined #salt
14:47 never2far joined #salt
14:48 m00dy joined #salt
14:52 diegows hi
14:52 Gifflen joined #salt
14:53 diegows short question, if I use "service: ID" in a watch statement of a service. The service will be restared/reload when the other service is restarted?
14:53 cro1 joined #salt
14:55 ldlework diegows: logic seems sound
14:55 diegows yes, but not sure :)
14:56 diegows I'll try it
14:57 diegows yes, works
14:57 diegows :)
15:02 Ryan_Lane joined #salt
15:06 cshuman joined #salt
15:06 SunSparc joined #salt
15:07 redondos joined #salt
15:08 cshuman joined #salt
15:09 Ryan_Lane joined #salt
15:11 cwright_ is there a way to have salt exit if a sls file does not exist?
15:12 cwright for example, I encrypt my pillar secrets.sls file as secrets.sls.aes, and secrets.sls is not stored in the repository.  i want the state.highstate call to fail if secrets.sls.aes has not been decrypted to secrets.sls
15:12 cwright currently it just continues and fills in None for the values that are missing
15:17 rmt forrest, I created an issue to cover my use-case, which hopefully sums it up succinctly:  https://github.com/saltstack/salt/issues/7556 ('ENC'/plugin enforced environments/custom SLS paths (multi-tenant support))
15:18 rmt oh, he's not here. :)
15:18 timoguin joined #salt
15:19 pipps joined #salt
15:21 terminalmage joined #salt
15:25 jaequery joined #salt
15:27 lemao_ joined #salt
15:35 LGSilva hey guys. Is it possible for the salt-master generate a pillar with all the IPs from all iterfaces from all minions?
15:36 ldlework LGSilva: use jinja
15:36 opapo joined #salt
15:37 cwright LGSilva: i don't think so, but you could use salt mine for that
15:39 saysjonathan joined #salt
15:40 ldlework I think with some jinja loops you could get that grain information into a pillar
15:40 pmrowla joined #salt
15:41 LGSilva ldlework: how could I target different minions inside the jinja?
15:42 ldlework LGSilva: I don't know how that's related to your original question so I dunno
15:42 LGSilva nm, cwright suggestion will work better. thanks :)
15:42 * ldlework shrugs
15:44 shinylasers joined #salt
15:47 ldlework btw, to "target minions in the jinja" you just target the minions normally with the pillar targeting, and then use jinja in those pillar files.
15:49 lemao joined #salt
15:50 Ryan_Lane joined #salt
15:53 Gifflen_ joined #salt
15:53 mahul joined #salt
15:55 mahul Is there a way to have a new salt minion connect to the master without the manual step of "accepting the key"? This is for an auto-scale setup where a host needs to come up and configure itself without any manual intervention
15:55 blee_ joined #salt
15:55 MTecknology mahul: ya, you can tell the salt master to auto-accept new keys
15:55 mahul I would assume this use-case has been explored before but I am not able to find anything in the documentation or by googling.
15:56 MTecknology I can't remember what it is, but the default commented config has it in there near the top
15:57 MTecknology Personally, I use the reactor system to verify the host is who it says it as before purging any old matched keys or accepting new ones
15:57 scalability-junk seems like my 20% time for salt will perhaps grow into a 60-80% time next month...
15:57 MTecknology scalability-junk: hm?
15:57 mahul MTecknology: Thanks let me chase that down... :) so simple in the end - found it in the documentation
15:57 scalability-junk mahul: salt-cloud can preseed if you use aws etc.
15:58 lemao_ joined #salt
15:58 scalability-junk MTecknology: one of my bigger projects just got axed down it seems :( but hey new stuff then
15:58 MTecknology a cloud of salt sounds horribly unpleasant
15:59 scalability-junk MTecknology: how to you use the event system for verifying hosts?
15:59 MTecknology the event system?
15:59 scalability-junk reactor system ;)
16:00 pipps joined #salt
16:00 mahul scalability-junk: will check out salt-cloud as well. I downloaded saltstack just a couple of days ago and I am in the discovery phase :)
16:01 scalability-junk mahul: don't worry. the only things you need to know is. don't count of one solution as in salt you can use more than 5
16:01 scalability-junk it depends on your preference and your setup etc.
16:01 MTecknology If system failed to auth and system matches this, then remove the key, ssh to the box, and tell the salt service to restart (since failed means the client won't try again), then if it comes on and auth is pending, then run a script and if the script succeeds, it will accept the key
16:01 scalability-junk the docs are mostly helpful.
16:01 xuru joined #salt
16:02 scalability-junk MTecknology: alright sounds kinda legit
16:02 mahul I would assume some patterns will emerge for the most common use-cases or saltstack might provide built-in solutions for these...
16:02 MTecknology scalability-junk: it was a bitch to come up with, but once you figure it out, it's pretty sweet
16:02 scalability-junk perhaps I come back to that :)
16:03 MTecknology mahul: if salt lacks something, write a state and a module; they're easy and actually fun
16:03 scalability-junk mahul: the thing is that there are always more than one built-in solution
16:03 pipps3 joined #salt
16:03 scalability-junk so you have no real guide on how to do things
16:03 scalability-junk that's the bad thing about starting with salt I think.
16:03 mahul MTecknology: I am still a newbie .. maybe after a bit...
16:04 scalability-junk mahul: modules are quite easy as it's mostly common python
16:04 scalability-junk but anyway have fun exploring gotta eat
16:04 Lue_4911 joined #salt
16:05 MTecknology salt is flexible enough to do whatever you want (which is great and terrible); it lets you shoot yourself in the foot, but if you do things right, it's better than sex
16:05 ldlework It'd be nice if it had better debugging facilities
16:05 ldlework Like showing you full renders of pillar and state yaml files
16:05 MTecknology -l debug isn't enough?
16:05 ldlework post includes and jinja rendering
16:05 ldlework not even close
16:06 ldlework When there is a syntax/logic errors when rendering stuff, having more information on those errors would be nice. Like a filename.
16:06 MTecknology state.show_highstate, state.show_lowstate, -l debug, and the error messages from things failing when running state-call state.highstate has almost always been plenty
16:07 MTecknology salt-call state.highstate will show you render errors
16:07 ldlework The render errors are vague ime
16:07 MTecknology file name and line number are included...
16:07 MTecknology and why it failed
16:07 ldlework No. Make a yaml syntax error
16:08 ldlework I can't recall off hand if I come into ambiguous error messages again I'll mention it
16:09 lemao joined #salt
16:09 MTecknology Twice I had to run salt inside of a python debugger, but that was a bug in the core system that was incredibly rare and hard to figure out, but thatch was there and he figured out what was going on.
16:09 pdayton joined #salt
16:09 MTecknology that was an interesting ride... guy is brilliant
16:09 matanya joined #salt
16:10 redondos joined #salt
16:14 lemao joined #salt
16:19 elfixit joined #salt
16:20 esogas_ joined #salt
16:22 alunduil joined #salt
16:23 dualJok joined #salt
16:23 matanya joined #salt
16:35 jacksontj joined #salt
16:37 teebes joined #salt
16:40 Thiggy joined #salt
16:42 Katafalkas joined #salt
16:44 higgs001 joined #salt
16:44 danielbachhuber joined #salt
16:44 quist joined #salt
16:45 Ryan_Lane joined #salt
16:48 xuru Ok, I've been  beating my head against this for a while now, and my forehead is starting to hurt…   I get the following message trying to use the state mysql_database.present:  State mysql_database.present found in sls ncl.mysql is unavailable
16:48 xuru Anyone know what is wrong?
16:52 robertkeizer joined #salt
16:53 xuru Here is the gist:  https://gist.github.com/xuru/6796751
16:54 Thiggy @xuru what version are you running?
16:54 xuru 0.16.4
16:54 dthom91 joined #salt
16:58 Gifflen joined #salt
16:58 genkimind joined #salt
16:58 jefftriplett left #salt
16:59 xuru hmm… it may be due to not having python-mysqldb installed...
17:00 pipps joined #salt
17:00 Gifflen joined #salt
17:00 xuru yep, that was it...
17:02 higgs001 joined #salt
17:05 pentabular joined #salt
17:09 Gifflen joined #salt
17:10 kenbolton joined #salt
17:13 KyleG joined #salt
17:13 KyleG joined #salt
17:15 SpX joined #salt
17:17 seanz Quick question: I noticed I can use <state>.<function> in a require statement.
17:17 seanz If I do that and have two states with different functions under the same ID, can I reference them separate in a require statement in a separate ID block
17:17 seanz ?
17:18 pipps1 joined #salt
17:19 pipps3 joined #salt
17:21 seanz So if I have an ID block that has both cmd.watch and cmd.run, can I "require" just the "cmd.run" statement in a different ID block?
17:22 faldridge joined #salt
17:23 robawt hi seanz
17:23 matanya joined #salt
17:23 robawt have you checked with your local operations staff?
17:23 seanz robawt: Hi, unknown human.
17:24 seanz robawt: I have not, because they're busy discussing other important things.
17:24 seanz I'd rather have asked this to the salt community at large to keep everyone working.
17:24 robawt mmhmm
17:24 robawt likely story
17:25 seanz robawt: If you have the much-coveted answer, I welcome your input.
17:25 robawt i offer none
17:25 travisfischer joined #salt
17:25 robawt speak of the devil
17:26 travisfischer Hey guys, we are really struggling with communicating about our salt implementations because the terminology for various things isn't clear to us. we are using this page as a reference http://docs.saltstack.com/ref/states/highstate.html Is there a name we can use to refer to the entire block (or stanza) under an id declaration? What do you guys call them?
17:27 fxhp What is the block of yaml called that lives under one 'id/label' in an sls file?
17:27 fxhp travisfischer - you always beat me to the punch
17:27 fxhp : )
17:28 Thiggy joined #salt
17:29 matanya joined #salt
17:32 logix812 joined #salt
17:33 dualJok left #salt
17:34 jefimenko joined #salt
17:36 Drawsmcgraw joined #salt
17:37 matanya joined #salt
17:37 Drawsmcgraw Anyone ever have problems getting a minion to connect to a Master in AWS?
17:37 TheRealBill joined #salt
17:38 Drawsmcgraw I've tried several times now. Master can ping Minion and vice versa. The security group is wide open.
17:38 Drawsmcgraw The Minion says it's waiting for the key to be accepted by the master, but the minion's ID doesn't show up in salt-key
17:38 renoirb Drawsmcgraw: Communication with salt stack is between them using a message queue
17:39 CheKoLyN joined #salt
17:39 uta joined #salt
17:40 Drawsmcgraw renoirb: Right, the 0mq bus across ports 4505 and 4506, right?
17:40 renoirb What is the salt.renderers.stateconf, an extension of what is by default? (http://docs.saltstack.com/ref/renderers/all/salt.renderers.stateconf.html)
17:41 renoirb http://docs.saltstack.com/topics/tutorials/firewall.html
17:41 renoirb Drawsmcgraw: http://docs.saltstack.com/topics/tutorials/firewall.html   But funnily enough on my Windows8 Vagrant run VM I could run my own master on an other machine and the VMs in Windows8 had no network connectivity but was receiving states from salt mastre
17:41 brianhicks joined #salt
17:41 renoirb got to go
17:44 jpcw_ joined #salt
17:45 timoguin joined #salt
17:46 matanya joined #salt
17:46 jpcw_ joined #salt
17:46 Drawsmcgraw renoirb: sunovabitch! It was iptables. I'm using the barebones CentOS 6.4 and (wrongly) assumed that it didn't have iptables running
17:47 Gifflen joined #salt
17:51 teebes Is there a maximum time that highstate is given to return its result on the master? I'm seeing a weird behavior where the master eventually exits out of state.highstate with no output at all, but I can see on the minion logs that there is still work being done, and eventually it completes there…
17:52 micah_chatt joined #salt
17:52 robbyt joined #salt
17:52 cetex i'm experiencing the same thing..
17:52 cetex minion did not return..
17:52 bhosmer joined #salt
17:54 pdayton joined #salt
17:54 timoguin docs say default timeout is 5 seconds, looks like
17:54 timoguin http://docs.saltstack.com/ref/cli/salt.html
17:55 mwillhite joined #salt
17:56 cetex :o
17:56 cetex timeout of 5 seconds seems waay to short.
17:57 cetex we need something like 5 minutes at least.
17:57 teebes I saw that but it definitely waits for more than 5 seconds before exiting out. I think that might be timeout for the minion to acknowledge and accept the task.
17:57 ccase joined #salt
18:01 TheRealBill did a git install of a new master today, getting an unhandled exception. traceback at https://gist.github.com/TheRealBill/6797785 I didn't see anything about this one, so looking for assistance on running it down.
18:09 dthom91 joined #salt
18:10 uta joined #salt
18:10 terminalmage joined #salt
18:13 newbie joined #salt
18:13 renoirb Drawsmcgraw: glad to hear :)
18:14 renoirb Anybody has experience with module publish.publish
18:15 Corey renoirb: Some.
18:15 renoirb If I am on a minion and a run a command such as: salt-call publish.publish '*' test.ping
18:15 saras tab make sls go boom
18:16 renoirb Corey: then, I only get at the stdout
18:16 renoirb [INFO    ] Publishing 'test.ping' to tcp://10.5.240.3:4506
18:16 renoirb But no answer on the console, nothing in the master's log nor in the minion's
18:17 Ryan_Lane renoirb: you need to configure the master to allow peer calls
18:17 Ryan_Lane and you shouldn't allow * ;)
18:18 Ryan_Lane otherwise any host getting owned will own every single other host
18:18 djn I am having problems deploying halite... everytime I try to login it fails, and the console prints the following error:
18:18 djn Exception AttributeError: "'SREQ' object has no attribute 'poller'" in <bound method SREQ.__del__ of <salt.payload.SREQ object at 0x3279cd0>> ignored
18:18 djn thoughts?
18:19 diegows joined #salt
18:20 renoirb Ryan_Lane: sure!! I agree I should't have .* allowed. But I am still figuring out which part is what where :)
18:21 matanya joined #salt
18:21 Ryan_Lane renoirb: in the master config, there's a peer configuration
18:21 Ryan_Lane http://docs.saltstack.com/ref/peer.html
18:21 renoirb All I need is that the minion can call grains.item and peer: {'*':[{'grains.item'}]} should be enough
18:22 renoirb Ryan_Lane: I know, I am there :)
18:22 renoirb Ryan_Lane: thread with my discoveries https://github.com/SmartReceipt/salt_state/issues/4
18:22 Ryan_Lane I think there's a new feature for caching data on the master
18:22 Ryan_Lane then pulling that data from the master
18:22 Ryan_Lane you should likely use that, rather than publish
18:22 renoirb I also ran saltutil.sync_all :D
18:22 Ryan_Lane it'll be more reliable
18:23 Thiggy joined #salt
18:23 renoirb Agreed
18:23 quist left #salt
18:23 renoirb My current point is to figure out how publish.publish works but not necessarily use it for now as I have a deliverable by tomorrow.
18:23 renoirb But If I can figure it out, it might just make me finish what I am doing
18:24 Ryan_Lane http://docs.saltstack.com/topics/mine/index.html
18:24 renoirb No way!!!
18:24 renoirb :)
18:25 Thiggy Can you access the salt logger from a salt runner or module or is that a Bad Idea?
18:25 renoirb But, mine is similar to grains isn't it?
18:25 Ryan_Lane yes
18:25 Ryan_Lane it's similar to exported resources in puppet
18:26 renoirb Do you know if mine has same configuration convention as grains (e.g. /etc/salt/mine)
18:26 Ryan_Lane not terribly sure
18:26 renoirb Will read doc
18:26 Ryan_Lane the docs on mine are somewhat thin
18:27 Ryan_Lane oh, wow. a state runner in 0.17?
18:27 Ryan_Lane want
18:28 quist joined #salt
18:28 renoirb I see that :(
18:28 renoirb Like the capability to use salt['publish.publish'] in a state file :P
18:28 Ryan_Lane awww. no puppet key support in 0.17 :(
18:28 a1j salt master uses 800m virtual and 300-400M resident memory per process. Is that normal? How can i debug memory?
18:29 Ryan_Lane a1j: which version of salt are you using?
18:30 Ryan_Lane mine is using 800M virt and 30-60M resident per process
18:30 a1j Ryan_Lane: 0.16.4
18:30 a1j well if i restart it now it will be manageable then it will grow.
18:31 Ryan_Lane sounds like a memory leak
18:31 a1j slowly
18:31 Ryan_Lane I know there was a leak in older systems due to the key signing system, but that was fixed. I wonder what your leak is from...
18:32 a1j i wonder too, i'm going though python debugging tools and cant find anything obvious to debug memory
18:32 honestly a1j: python is managed
18:32 honestly I have no idea *how* python's MM works
18:32 a1j reference counting?
18:32 rlarkin anyone, I'm stuck looking for a repo that provides python-jinja2 for ubuntu precise.  tired and hungry and missing the obvious apparently.  "Depends: python-jinja2 but it is not installable"
18:33 honestly I wonder if salt uses any c modules
18:33 honestly rlarkin: why do you want to install that? jinja is integrated in salt
18:34 honestly oh.
18:34 honestly wait.
18:34 honestly hm
18:34 rlarkin apt-get install salt-minion pulls in salt-common which pulls in python-jinja2, but I don't have it
18:34 honestly right, yeah
18:34 rlarkin aptitude search jinja2 is no hits
18:35 renoirb rlarkin: apt-cache search jinja|grep python
18:35 rlarkin been googleing for an hour and all indications are that it should be there
18:35 honestly rlarkin: I am running Ubuntu 12.04.3
18:35 honestly and I have it
18:36 rlarkin Ubuntu precise (12.04.3 LTS
18:36 honestly so, I *am* on precise
18:36 honestly 64-bit though
18:36 rlarkin hmm
18:36 honestly https://gist.github.com/duk3luk3/6798408
18:36 a1j honestly: of course it does, openssl, sqlite, zmq...
18:37 honestly a1j: I mean salt-specific
18:37 williamthekid joined #salt
18:38 shomodj joined #salt
18:39 jacksontj joined #salt
18:40 anuvrat joined #salt
18:41 ifnull I'm picking up with a problem that I was having yesterday where I was getting "No Top file or external nodes data matches found". I have tried both 0.16.4 and 0.17.0. Another set of eye would be greatly appreciated. https://gist.github.com/ifnull/4836cfa1a08f85c0c9fc
18:41 a1j honestly: nope.
18:42 Corey Pfft, utahdave is still sleeping? :-)
18:47 renoirb I got an answer from the author Ryan_Lane https://github.com/SmartReceipt/salt_state/issues/4#issuecomment-25564871
18:47 renoirb I tried peer: I am now trying peer_run: in case it is different
18:48 xuru Q: I have a django.syncdb state that requires a git checkout, but when I run it from vagrant, I don't want to do the git checkout.  How can you conditionally require something?
18:52 Gifflen joined #salt
18:54 druonysus joined #salt
18:55 mgw Does anybody know what would happen if I put a 'schedule' option in multiple minion config files? e.g., /etc/salt/minion.d/config_a and /etc/salt/minion.d/config_b?
18:55 mgw Will it combine them?
18:59 Brew1 joined #salt
19:00 kiorky uhm salt HEAD seem terribly broken, can someone confirm : https://github.com/saltstack/salt/issues/7565
19:00 Brew joined #salt
19:00 Ryan_Lane renoirb: peer_run calls a runner, which exists on the server
19:02 ronc joined #salt
19:02 pentabular joined #salt
19:03 newbie93 joined #salt
19:04 newbie93 has anyone used salt with rebar
19:05 jpcw joined #salt
19:05 felixhummel joined #salt
19:07 saras anyone here work with erlang or elixir code
19:09 alunduil joined #salt
19:10 jpcw joined #salt
19:11 renoirb By the way redbeard2, I might be able to give stuff for https://github.com/saltstack-formulas/piwik-formula  soon...ish.
19:12 redbeard2 renoirb: are you a big user of piwik?
19:13 renoirb We are using it, I am actually running a set of Piwik nodes under NGINX using Salt
19:13 cshuman joined #salt
19:13 renoirb And now that I see your guidelines in Salt Formulas… I might adopt the conventions for our own too.
19:14 renoirb Objective is to use remotes with gitfs
19:14 felixhummel_ joined #salt
19:16 druonysuse joined #salt
19:16 CansaSCity joined #salt
19:21 renoirb Yes, Salt can salt itself! L337 :D  https://github.com/saltstack-formulas/salt-formula
19:21 cjh renoirb: i like it :D
19:21 toastedpenguin joined #salt
19:24 jeffasinger Does running an overstate ever make sense in a masterless environment?
19:25 dthom91 joined #salt
19:31 shinylasers joined #salt
19:31 pipps1 joined #salt
19:32 never2far joined #salt
19:33 linuxnewbie joined #salt
19:34 jacksontj joined #salt
19:37 TheRealBill kiorky: I confirm that. I get the same issue.
19:38 ipmb joined #salt
19:38 kiorky TheRealBill: confirm also on github please !
19:38 kiorky (if not already done ^^=
19:39 TheRealBill kiorky: done did. :)
19:39 kiorky thx!
19:39 mmilano joined #salt
19:41 pipps joined #salt
19:41 pipps1 joined #salt
19:41 TheRealBill yw
19:41 mgw1 joined #salt
19:41 bastion2202 joined #salt
19:44 redondos joined #salt
19:48 EugeneKay Apparently I'll be in SLC week-after-next.
19:49 lacrymology left #salt
19:49 TheRealBill kiorky: I"m guessing the change which causes it is https://github.com/saltstack/salt/commit/77fff17f1bddd70e16e42f8d95948ce0648f5484 I"m going to comment that out and test it
19:52 Anita joined #salt
19:52 TheRealBill yes, commenting out the event call on line 1524 avoids the problem.
19:56 jeffasinger Anyone know of any good overstate examples?
19:57 kiorky TheRealBill: those introducing jid
19:58 cshuman joined #salt
19:58 kiorky TheRealBill: i mean the whole bunch of commits , thats why i indicated which worked, from where it dont work
19:58 kiorky TheRealBill: did you looked at the commits i indicated on the bug report ?
19:59 TheRealBill yeah the line that broke it was 1524 in the state.py file. I just followed the traceback backward and compared that to recent commit history on github.
20:00 kiorky TheRealBill:
20:00 kiorky http://goo.gl/ubwDgK
20:00 TheRealBill kiorky: the commit I linked above is where it appears to have broken as it added the self.event call
20:01 kiorky TheRealBill: yep, but in between, there were fixes which in fact do not fix that well :)
20:01 kiorky fixes need other fixes
20:01 kiorky :)
20:01 TheRealBill kiorky: yeah, i've commented there on what I've found. Looking into *why* that call breaks.
20:02 TheRealBill it isn't limited to using names, I don't use that function and still see it.
20:02 TheRealBill (well I did until I removed the aforementioned line)
20:02 kiorky TheRealBill: personally im cowardly waiting for the main commiter of those change to do the fix ^^ i rebased my copy to  8ce5ce3
20:04 jacksontj joined #salt
20:05 anitak Guys, I got a tiny question about the new release 0.17.0. I am running Ubuntu and have the ppa added and installed the packages from there. Looks like there are no 0.17 packages on launchpad yet. When will they be released/uploaded?
20:06 Ryan_Lane what's the difference between overstate and the new state runner in 0.17?
20:07 druonysus joined #salt
20:07 druonysus joined #salt
20:08 cshuman joined #salt
20:08 pdayton joined #salt
20:09 pdayton joined #salt
20:10 pdayton joined #salt
20:22 shinylasers joined #salt
20:26 xuru joined #salt
20:27 djn has anybody halite up and running? still gettig the error I posted above
20:27 cedwards djn: i've got it running on a couple systems. what error do you get?
20:28 cedwards djn: my setups are using packages based on 0.1.0 on Arch and FreeBSD..
20:29 redondos joined #salt
20:29 cachedout joined #salt
20:30 cshuman joined #salt
20:30 djn i am getting the following:
20:30 djn Exception AttributeError: "'SREQ' object has no attribute 'poller'" in <bound method SREQ.__del__ of <salt.payload.SREQ object at 0x2deb3d0>> ignored
20:30 djn cedwards: sound more like a issue in the python code to me...
20:30 djn *sounds
20:30 cedwards are you using a git checkout?
20:31 djn I am using the version packaged for debian wheezy
20:31 cachedout For issues related to documentation that needs to be clarified, is it best to file an issue in the Github tracker or is another method preferred?
20:31 cedwards cachedout: file an issue on github or submit a PR with a fix
20:31 cedwards djn: what does your halite config look like?
20:32 cedwards djn: i assume you have the halite: block and an external_auth: block?
20:32 djn I'll do a paste for that, in a second
20:32 djn yes
20:32 timb joined #salt
20:32 cachedout cedwards: Ah, I missed that doc directory. I'll go ahead and roll a PR. Thank you.
20:33 Ryan_Lane renoirb: did you see that salt has logstash support now?
20:33 micah_chatt joined #salt
20:33 uta joined #salt
20:33 Ryan_Lane renoirb: in 0.17. is this separate from what you were looking at?
20:34 renoirb I did not know!
20:34 Ryan_Lane renoirb: I was thinking this would be a good way to debug deployments. send tagged logs to logstash, so that we could track a deployment via logstash as it happens
20:34 renoirb Let's see later :)
20:34 renoirb yup!
20:34 Ryan_Lane well, a kibana3 search anyway
20:34 wkf_ joined #salt
20:34 djn cedwards: https://paste.selfnet.de/A2f/ here
20:34 djn cedwards: ldap auth is working on the commandline so far, but no dice via the webinterface
20:35 renoirb But, as I see it Ryan_Lane, deployment will be ran hundreds of time locally using vagrant, then in a separate, then the real deal.
20:35 Ryan_Lane renoirb: well, I mean the salt deployment system I wrote
20:35 djn cedwards: I think I did notice that only when I asked before but not this time: said error occurs when trying to log in
20:35 Ryan_Lane I guess we haven't upgraded to the git version yet
20:36 renoirb I can't wait to see https://github.com/saltstack/halite
20:36 cedwards djn: have you tried it with pam auth to test if it's an ldap issue?
20:36 Ryan_Lane you're working on pre-requisites to that, though :)
20:36 cedwards renoirb: halite is nice, although still very alpha
20:37 renoirb Sure, but at least to see what has been issued is more helpful than editing states.
20:37 renoirb or emmiting commands
20:37 giantlock joined #salt
20:37 cedwards djn: i'm also curious about line 27-31 in your paste..
20:38 djn cedwards: thats a remnant of trying various configurations for halite :) I don't need it?
20:38 cedwards djn: i don't think so. the block starting at line 20 looks good.
20:39 cedwards djn: but just to test you might change line 21 from ldap to pam and see if you get anything different.
20:40 Ryan_Lane renoirb: hm. the same deployment system could be used for salt-vagrant and production
20:40 Ryan_Lane renoirb: if we treat it like a continuous deployment
20:41 Ryan_Lane let's talk more in webplatform-site
20:41 djn cedwards: ok, thanks for the input, I removed those lines... changing ldap to pam (in the config and in the webui) does not help
20:41 RoboTeddy joined #salt
20:42 renoirb sure Ryan_Lane
20:42 cedwards djn: might be a bug related to your platform or package. maybe submit something an issue on github
20:42 renoirb But anyway it doesn't get logged on webplatform-site only on the %s/-site//
20:43 djn cedwards: thanks, I feared so much...
20:44 djn joehh: do you know if theres a known issue like this in the debian packages?
20:44 djn 'this' meaning a problem when authenticating to halite
20:44 timb can someone help a newbie with pillar?  i'm trying to use two files in the same namespace but only the last one it finds is registering.  using salt 0.16.3 on rhel6.  http://pastebin.com/8zDhjDp3.  thx!
20:45 cedwards djn: this is the first i've heard of one, and a quick search doesn't show one on github
20:46 djn cedwards: I got it
20:47 djn cedwards: I disabled IPv6 in my transports, and now it is working
20:47 cedwards djn: interesting
20:48 djn cedwards: I guess I will hunt that one down and submit a pull request... It seems like the salt guys are not testing ipv6 at all, shortly after the introduction of ipv6 it broke again too
20:51 Jahkeup joined #salt
20:51 anitak Any idea when 0.17 packages will on the Ubuntu ppa?
20:55 cedwards anitak: I don't see a mention of the status of those packages on the packagers list, but i'd imagine it'd be any day now
20:55 anitak awesome, thanks
20:55 anitak been waiting for those packages to show up :)
20:56 cedwards they are usually pretty quick with them
20:58 lineman60 joined #salt
20:58 anitak cool
20:59 anitak I got one more question actually...it's about the virt module
20:59 anitak I have a xen hypervisor that I manage with Salt
20:59 anitak master is 0.17, minion is 0.16.4 right now
21:00 mr_chris Just made my first contribution to salt and my first code contribution to an open source project.
21:00 mr_chris https://github.com/saltstack/salt/pull/7567/files
21:00 mr_chris A single 0.
21:00 anitak when I run salt 'hostname' virt.is_xen_hyper, I get this
21:00 anitak it returns True
21:00 cedwards mr_chris: congrats!
21:00 anitak when I run salt 'hostname' virt.list_vms though, nothing is returned
21:01 anitak any ideas what I am missing on that xen box?
21:01 jslatts joined #salt
21:01 anitak I have python-libvirt and python-xenapi installed
21:02 anitak master and minion are Ubuntu 12.04 64-bit
21:02 mr_chris The code that handles MySQL grants that have "WITH GRANT OPTION" was breaking.
21:03 cedwards anitak: interesting. i can run the same on my kvm host and it returns all the vms
21:03 cedwards anitak: may be a bug in the xen code
21:03 mr_chris cedwards, Thanks.
21:04 anitak I already had master and minion running with debug, but don't see anything :(
21:04 anitak it only shows the commands being run
21:04 anitak but no further output
21:05 ronc joined #salt
21:06 pgollakota joined #salt
21:07 jacksontj joined #salt
21:15 jacksontj joined #salt
21:16 pgollakota Hi guys, I deleted one of the minion keys from my master. The I deleted the minion keys in /etc/salt/pki/minion/ and restarted salt-minion which generated new keys. However the minion is not being listed in the unaccepted keys in the master.
21:17 pgollakota I checked that ports 4505 and 4506 were open on the minion.
21:17 pgollakota Any ideas on how I can get the minion keys appear on my master again?
21:19 higgs001 joined #salt
21:19 mgw1 pgollakota: anything in the minion logs?
21:19 mgw1 or master for that matter
21:19 JaredR joined #salt
21:19 cshuman joined #salt
21:22 pgollakota mgw1: You're right, I did not notice the logs! There is this error message: RSAError: oaep decoding error
21:23 pgollakota I'm using salt 0.17.0
21:23 pgollakota I think my PyCrypto is not installed correctly. Let me check.
21:23 mgw1 pgollakota: that's in the minion or master's log?
21:23 pgollakota That's the minion's log
21:24 dave_den pgollakota: did you also delete the minion's key from the master?
21:25 dave_den sounds like the master still has the old minion public key and using that to encrypt the aes handshake and the minion is trying to decrypt is using the new key
21:26 dave_den oh, you said you deleted form the master
21:26 dave_den derp
21:26 pgollakota dave_den: Yeah, I ran salt-key -d ….
21:26 pgollakota Also verified that /etc/salt/pki/master/minions does not have the corresponding hostname key
21:31 shinylasers joined #salt
21:31 kermit joined #salt
21:38 forrest joined #salt
21:39 JaredR joined #salt
21:46 dthom91 joined #salt
21:52 msuarez_ joined #salt
21:54 msuarez_ Hi
21:55 msuarez_ Anyone know when salt 0.17 will be available in Ubuntu?
21:56 jefftriplett joined #salt
21:58 jacksontj joined #salt
21:59 cedwards msuarez_: should be soon
21:59 cro joined #salt
22:07 msuarez_ thanks
22:07 lineman60__ joined #salt
22:08 jefftriplett left #salt
22:09 Thiggy joined #salt
22:10 lineman61 joined #salt
22:21 RoboTeddy hmm, I'm accessing a custom grain from pillar. when I run salt myhost state.highstate, pillar fails to render due to the key for the custom grain missing in the `grains` dict. however, if I restart salt-minion on myhost, the running state.highstate on the master works
22:22 RoboTeddy do minions cache their grains?
22:23 Thiggy joined #salt
22:23 RoboTeddy (and.. assuming they do: anyone have an idea of why the custom grain isn't being cached correctly at first?)
22:25 RoboTeddy I wonder if the minion doesn't purge its grain cache after it receives custom grains during sync_grains
22:25 pgollakota left #salt
22:28 msuarez joined #salt
22:30 cshuman_ joined #salt
22:33 msuarez joined #salt
22:36 higgs001 joined #salt
22:42 jacksontj joined #salt
22:48 cshuman joined #salt
22:49 dthom91 joined #salt
22:54 mianos joined #salt
22:54 JaredR joined #salt
22:55 cro joined #salt
22:56 jaequery joined #salt
22:59 ifnull If anyone else is having trouble with GitFS not loading gitfs_remotes and failing silently, make sure the dependency GitPython is installed. It is not currently included in salt-bootstrap. I put a pull request in but I'm not sure if it makes sense or not to have it in the bootstrap.
22:59 ifnull https://github.com/saltstack/salt-bootstrap/issues/247
23:00 anteaya joined #salt
23:01 druonysus joined #salt
23:02 jcockhren ifnull: which version are you having issues with?
23:03 xuru joined #salt
23:06 mianos joined #salt
23:11 micah_chatt joined #salt
23:11 micah_chatt Is anyone going to be at devops day Atlanta tomorrow?
23:15 pentabular joined #salt
23:15 austin987 joined #salt
23:16 eculver joined #salt
23:16 eculver joined #salt
23:17 ifnull @jcockhren, all version. GitPython is a dependency for GitFS but not included in bootstrap
23:18 ifnull https://github.com/ifnull/salt-bootstrap/commit/668a24226c43d52203234860bf0ab316bf5a45e6
23:18 packeteer joined #salt
23:25 joehh djn: Not sure exactly the issue, but I tested halite on debian with the release candidate
23:25 joehh It took a little bit of reading to docs "carefully" and playing to get it right
23:26 msuarez__ joined #salt
23:26 joehh from memory, i needed to install python-cherrypy3 rather than python-cherrypy
23:34 pipps joined #salt
23:34 jacksontj joined #salt
23:38 mianos joined #salt
23:40 arapaho joined #salt
23:42 taylorgumgum joined #salt
23:48 oz_akan_ joined #salt
23:50 austin987 joined #salt

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