Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-11-20

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

All times shown according to UTC.

Time Nick Message
00:00 jslatts joined #salt
00:01 xmltok joined #salt
00:05 JulianGindi joined #salt
00:10 dpippenger joined #salt
00:14 oz_akan_ joined #salt
00:15 bemehow joined #salt
00:18 wilywonka joined #salt
00:18 oz_akan_ joined #salt
00:20 bhosmer joined #salt
00:23 Brew joined #salt
00:25 dpippenger1 joined #salt
00:25 dpippenger1 joined #salt
00:29 cocoy1 joined #salt
00:30 rawzone joined #salt
00:34 bemehow_ joined #salt
00:36 mrevans joined #salt
00:38 cdcalef joined #salt
00:39 mrevans nick snave
00:40 mrevans joined #salt
00:43 pipps__ joined #salt
00:51 jimallman joined #salt
00:51 druonysus joined #salt
00:54 NotreDev joined #salt
00:55 zandy joined #salt
00:56 pipps_ joined #salt
00:59 Ryan_Lane joined #salt
01:02 zandy joined #salt
01:13 JulianGindi joined #salt
01:20 mgw I'm getting this error trying to "download" a file with test.managed:
01:20 mgw Unable to determine upstream hash of source file file:///home/matthew/test.iso
01:21 jesusaurus are you specifying a source_hash ?
01:21 mgw just realized i was using 'hash' not source_hash
01:21 jesusaurus also, i just noticed you said test.managed not file.managed, but it probably still needs it
01:22 mgw s/test/file/
01:25 sroegner_ joined #salt
01:25 zandy joined #salt
01:26 pipps joined #salt
01:27 Nexpro joined #salt
01:29 oz_akan_ joined #salt
01:30 juicer2 joined #salt
01:35 bhosmer joined #salt
01:35 deepakmd_oc joined #salt
01:38 ajw0100 joined #salt
01:38 bhosmer_ joined #salt
01:42 cachedout joined #salt
01:58 sfz joined #salt
02:01 redondos joined #salt
02:09 cron0 joined #salt
02:11 forrest joined #salt
02:15 oz_akan_ joined #salt
02:21 cron0 joined #salt
02:22 redondos joined #salt
02:23 dwyerj man, i had the fucking repo state set totally wrong, wtf was i thinking!
02:23 xl1 joined #salt
02:24 franky__ joined #salt
02:25 dpippenger joined #salt
02:29 zsoftich1 joined #salt
02:31 dvogt joined #salt
02:37 pipps joined #salt
02:48 ajw0100 joined #salt
02:56 xmltok joined #salt
02:59 redondos joined #salt
03:05 bemehow joined #salt
03:06 bemehow joined #salt
03:07 NotreDev joined #salt
03:17 oz_akan__ joined #salt
03:20 thelorax123 joined #salt
03:22 ravibhure joined #salt
03:24 lvan joined #salt
03:26 bhosmer joined #salt
03:34 Ryan_Lane joined #salt
03:37 NotreDev joined #salt
03:38 premera_t joined #salt
03:39 dccc joined #salt
03:39 xl1 I ran "salt '*' cmd.run 'sleep 10; echo ok'" and only 1 minion returned a result, any ideas?
03:40 xl1 salt version is 0.17.1. The result looks random, most of the time only a few minions returns a result.
03:40 dwyerj output from salt-key -L
03:41 xl1 all minions are listed in Accepted Keys
03:41 dwyerj are all the minions actually connected ?
03:41 dwyerj and what about test.ping ?
03:42 xl1 I think so. test.ping result is always correct, but state.sls or cmd.run isn't.
03:45 swa_work joined #salt
03:52 xl1 Everything looks fine from "sleep 1" to "sleep 4"; "sleep 5" occasionally doesn't get all results, "sleep 6" often doesn't get all results. The cloud servers are in the same LAN.
03:56 bt joined #salt
04:00 hvn joined #salt
04:01 oz_akan_ joined #salt
04:02 hvn hi, can you guys tell me how salt release work? Since I see some commits (in the middle) already in develop for a long time but not in 0.17.2, but other later commits already there. For example https://github.com/saltstack/salt/commits/develop/salt/states/pkgrepo.py vs https://github.com/saltstack/salt/commits/v0.17.2/salt/states/pkgrepo.py.
04:09 ckao joined #salt
04:16 xl1 I wrote a little test module, the minion logs tell that all minion did get the command, but not all minions' results were returned to the master.
04:18 xl1 The master gets something like:
04:18 xl1 10-6-1-3:
04:18 xl1 ----------
04:18 xl1 10-6-2-12:
04:18 xl1 ok
04:18 xl1 10-6-5-33:
04:18 xl1 ----------
04:18 xl1 10-6-6-6:
04:18 xl1 ----------
04:19 dwyerj pastebin the module please
04:21 Corey hvn: Cherry picking.
04:21 alunduil joined #salt
04:22 xl1 http://pastebin.ca/2477410
04:24 joehh hvn: I think that is a fix terminalmage made an hour or so ago that was picked up in 0.17.2 for debian/ubuntu
04:24 junedm joined #salt
04:25 joehh no - doesn't look like he has pushed it yet
04:25 joehh co-incidence that it is the same file then... :)
04:25 hvn @joehh: I mean, the older commit was throw away since it has been merged into develop
04:26 hvn @joehh: Because I've seen a bug that should not happend (fixed one) https://github.com/saltstack/salt/blob/v0.17.2/salt/states/pkgrepo.py#L235  - failed when repo not in kwargs (develop version will not have this bug)
04:27 terminalmage joehh: it's a pending pull request against 0.17
04:28 joehh ahhh makes sense
04:28 terminalmage yeah it works in develop, but there was some change to the logic
04:28 terminalmage which is why I suspect it was a merge conflict for which the wrong fix was applied
04:29 joehh hvn: as terminalmage has hinted at, there is a fix for that and I'll be pushing 0.17.2-2 for debian tonight
04:29 dvogt joined #salt
04:29 terminalmage I'll also be patching/rebuilding for fedora/epel
04:30 hvn okay, thanks
04:31 anuvrat joined #salt
04:35 hooligan joined #salt
04:35 hooligan left #salt
04:40 redondos joined #salt
04:47 thelorax123 joined #salt
04:48 redondos joined #salt
04:56 joehoyle joined #salt
04:56 jimallman joined #salt
05:08 Furao joined #salt
05:15 bhosmer joined #salt
05:20 jalbretsen joined #salt
05:26 terminalmage EPEL5 and EPEL6 are patched, rebuild, and resubmitted for epel-testing
05:26 terminalmage I'll hit the fedora branches tomorrow
05:26 terminalmage going to get to sleep before midnight for a change, need to catch up on sleep
05:26 terminalmage later folks
05:28 forrest later
05:29 middleman_ joined #salt
05:42 noob joined #salt
05:44 Guest40297 Can anyone tell me why this doesn't bootstrap a minion?  It just invokes the cmd on my master instead: sudo salt-ssh '*' -r 'wget -O - http://bootstrap.saltstack.org | sh -s -- -P git v0.17.1'
05:51 Destro joined #salt
05:56 vipul joined #salt
06:02 zsoftich1 I'm having a bit of trouble with pillar and enironments...when I set a minions environment I lose pillar data
06:04 dpippenger joined #salt
06:07 redondos joined #salt
06:09 bemehow joined #salt
06:11 cocoy1 left #salt
06:13 Ryan_Lane joined #salt
06:13 Ryan_Lane joined #salt
06:21 Katafalkas joined #salt
06:23 zsoftich1 not sure what the problem is...but the docs seem to say control it throught the top state...guess thats what I'll do
06:27 sroegner_ joined #salt
06:36 zandy joined #salt
06:37 rawzone joined #salt
06:38 matanya joined #salt
06:38 dvogt joined #salt
06:46 bhosmer joined #salt
06:50 xl1 Still don't know the reason but "salt --verbose --timeout=60 ..." is a workaround, in case someone runs into the same problem. Took me a few weeks to realize results have been missing though.
06:55 jY does salt support restarting a service listed in another sls file?
07:03 bhosmer joined #salt
07:13 ravibhure Hi all
07:14 ravibhure Is there is salt state module or inherit functions available for jira/confluence, something like to create a new jira ticket or validate the existing while deployment using salt
07:21 rmt joined #salt
07:27 jcockhren ravibhure: nah. you should write one ;)
07:28 ravibhure ok
07:30 juasiepo joined #salt
07:41 Destro joined #salt
07:44 redondos joined #salt
07:44 Bram joined #salt
07:56 balboah joined #salt
07:59 harobed joined #salt
08:03 shinylasers joined #salt
08:04 balboah 17.1 didn't even parse the minion config for me. 17.2 works tho but has not been released officially yet :(
08:05 cocoy2 joined #salt
08:06 slav0nic joined #salt
08:06 slav0nic joined #salt
08:08 giantlock_ joined #salt
08:20 linjan joined #salt
08:24 elsmorian joined #salt
08:27 xcbt joined #salt
08:29 strind joined #salt
08:33 strind I just upgraded to 0.17.2-1precise and now can not even ping my servers. Some one knows there is an issue with this version?
08:33 _ikke_ strind: yes, there are issues
08:34 strind _ikke_: Thanks, so I gues I'd better get back to previous version for a while :P
08:44 redondos joined #salt
08:51 bhosmer joined #salt
08:52 elsmorian1 joined #salt
08:53 tinuva joined #salt
09:01 harobed_ joined #salt
09:06 shinylasers joined #salt
09:09 scott_w joined #salt
09:10 mike251 joined #salt
09:11 carlos joined #salt
09:15 scott_w joined #salt
09:20 giantlock_ joined #salt
09:20 graingert joined #salt
09:20 graingert joined #salt
09:22 giantlock_ joined #salt
09:23 scott_w joined #salt
09:23 linjan_ joined #salt
09:26 scott_w does the file state have functionality that allows you to manage a whole directory?
09:26 scott_w by that i mean: say i have a sites-enabled directory with 5 different configurations in
09:27 scott_w could i do a file.managed state that references salt://webserver/sites-enabled
09:27 scott_w which contains those five files
09:27 scott_w (or templates)
09:31 thelorax123 joined #salt
09:37 shinylasers joined #salt
09:38 xl1 scott_w: file.directory
09:40 scott_w xl1: the documentation doesn't suggest i could pass a source argument
09:41 junedm joined #salt
09:41 scott_w it might be a decent feature to add, if it's not already there
09:41 xl1 scott_w: file.recurse
09:41 xl1 scott_w: my mistake :)
09:43 mike251 i have a stupid Q: what are the diffs between salt modules and states? can the modules be run only from CLI?
09:43 mike251 can i run modules inside states? is that reccommended?
09:43 scott_w mike251: states run modules
09:43 scott_w usually
09:44 scott_w a module is a "command" that can be run
09:44 xl1 mike251: sls calls states, states and command line calls modules
09:44 mike251 xl1: scott_w thanks guys
09:44 scott_w xl1: awesome, that looks right
09:44 scott_w i'm just looking at adding s3 support, so i'm working through all the use-cases that i'd like to see it solve
09:45 scott_w (this specific one is actually sending a ton of static files to s3 to serve)
09:45 redondos joined #salt
09:49 hvn left #salt
09:49 scott_w ok, i'm off out
09:50 scott_w left a comment on the S3 stuff https://github.com/saltstack/salt/issues/6443#issuecomment-28874693 :)
09:50 tamentis joined #salt
09:51 scott_w joined #salt
09:59 mike251 left #salt
09:59 mike251 joined #salt
10:04 Ryan_Lane joined #salt
10:07 Jahkeup joined #salt
10:08 NotreDev joined #salt
10:08 chutzpah joined #salt
10:09 bhosmer joined #salt
10:10 srage joined #salt
10:15 thelorax123 joined #salt
10:16 mortis im running nagios and salt on the same server at home and wrote a quick and dirty wrapper for running nagios-plugins via salt.. if anyone want it / wants to improve it, or whatever, its avail here : https://github.com/mortis1337/nagios-plugins/blob/master/check_by_salt.py
10:17 reuss evaluating salt-ssh for testing salt -- what is required for a salt minion for that type of execution? can salt not reuse existing ssh config? i can easily log in to the server by just doing ssh $hostname -- but salt doesn't seem to like that and needs to have a user defined? and a key installed, for that user? .. That seems rather roundabout?
10:18 gasbakid joined #salt
10:18 reuss i would've thought salt-ssh made it easy to use salt for your existing ssh infrastructure?
10:19 mike251 reuss: why does salt need an username? salt communicates with the minions via its own keys
10:19 mike251 salt-ssh lets you manage your ssh access on the minions.
10:19 mike251 so if you do not have ssh access to the minion you can use a state with salt ssh and create users+ add keys
10:20 mike251 maybe i am wrong :)
10:20 mortis btw topic is wrong, latest is 0.17.2 :)
10:21 jotterbot1234 joined #salt
10:22 TomasNunez joined #salt
10:23 reuss mike251: no no - i already have ssh set up for all my servers - i just want salt to execute ssh commands as the user i'm currently running as
10:24 mike251 ah reuss got it now..
10:24 reuss and setting user: in the roster is not enough, since it *still* wants to deploy it's own keys
10:24 mike251 reuss: never tried that.
10:25 reuss and i find it odd that sshpass is apparently a dependency/requirement
10:31 srage joined #salt
10:35 zandy joined #salt
10:36 zandy joined #salt
10:36 gasbakid_ joined #salt
10:39 bhosmer_ joined #salt
10:43 baffle So, is 0.17.2 sane?
10:43 balboah trying to figure out. works better than .1 in my test so far
10:44 balboah not sure if you can see ongoing issues specific to that release somewhere?
10:45 baffle balboah: I had problems with 0.17.1, with Jinja. And someone reported that there were problems in the previously packaged versions in 0.17.2, especifally if you used repos...
10:45 balboah aha
10:45 balboah I couldn't even start the minion because of something in the minion config file
10:46 baffle What's the error?
10:46 redondos joined #salt
10:46 balboah Error parsing configuration file: /etc/salt/minion - 'str' object has no attribute 'copy'
10:46 balboah .2 works tho
10:47 baffle balboah: Have you installed 0.17.2 on the master? I had lots of strange errors when I tried to go from 0.17.1 -> 0.17.2 (homebuilt) on a minion before I restarted the master..
10:47 baffle Fuck.
10:47 baffle (Sorry)
10:47 baffle The repo bug is in the 0.17.2 packages.
10:47 balboah well I'm only testing in my dev environment atm
10:48 balboah didn't experience the jinja bug yet
10:48 balboah I just want my patch to make auto highstates and returner (aka ext job cache) option to work. and I'm too lazy to learn how to make my own ppa packages
10:49 baffle I had Jija bugs with 0.17.1. They disappeared with 0.17.2 packages that I built from develop. I did a state.highstate with my own 0.17.2 packages, and that worked. Upgraded to "proper" 0.17.2 packages, and the repo bug hit, and state.highstate fails.
10:49 balboah :/
10:49 balboah would be nice if there was automated tests to catch these things before a "stable" package was deployed
10:49 baffle Strange that they pushed it out with that bug, it was known before the release..
10:50 balboah yeah I keep my own "stable" ppa and copy stuff in there once I've tested it
10:51 balboah the official has broken my states many times :)
10:53 diegows joined #salt
10:56 joehh baffle: which os are you?
10:57 joehh I'm about to do ubuntu and if there is one worth doing first, I'll start there
10:58 joehh balboah, or others: you too - if there is one to start with, let me know
11:00 srage joined #salt
11:00 srage_ joined #salt
11:07 baffle joehh: Ubuntu. :)
11:07 pniederw joined #salt
11:08 joehh any particular release?
11:08 gasbakid__ joined #salt
11:08 joehh precise?
11:08 pniederw is salt 0.17.2 safe to install? there was some talk yesterday that some packages are broken or so
11:09 joehh pniederw: there is a bug with the pkgrepo state that will affect many people
11:09 joehh packages have been updated to fix this for debian, redhat/fedora and I'm about to do ubuntu
11:09 joehh so if you don't use pkgrepo, you'll be fine now
11:09 baffle joehh: I run Precise, yeah.
11:10 joehh if you use pkgrepo and ubuntu, wait a little (could be an hour could be 6 depending on launchpad build delays- find out soon)...
11:10 joehh I'll do it first
11:10 baffle joehh: Awesome. :)
11:10 Ryan_Lane joined #salt
11:10 pniederw I do use pkgrepo (and ubuntu) for bootstrapping salt
11:11 joehh then wait a bit...
11:11 pniederw ok, thanks
11:11 joehh I'll let you know when I'm done
11:21 jeddi joined #salt
11:38 zandy joined #salt
11:40 MrTango joined #salt
11:41 sroegner_ joined #salt
11:42 gasbakid_ joined #salt
11:42 joehh precise build starts in 54 minutes... assuming no issues (not expecting any) should be available shortly after that
11:45 joehh lucid next
11:47 redondos joined #salt
11:51 balboah joehh: also in precise
11:51 balboah joehh: glad to hear you're on it
11:55 dangerousbeans joined #salt
12:05 gasbakid__ joined #salt
12:11 Ryan_Lane joined #salt
12:12 aleszoulek joined #salt
12:13 lemao joined #salt
12:16 robinsmidsrod joined #salt
12:22 gasbakid__ joined #salt
12:27 joehh balboah, baffle, pniederw: looks like the updated precise is up - I'd appreciate someone updating/installing and doing a quick check
12:28 BenBen joined #salt
12:29 balboah I will try it in a while
12:29 BenBen hi there, quick question on salt-ssh : my sls files are in a custom repository, which is indicated in the /etc/salt/master configuration file. Will salt-ssh automatically use this master configuration file to point to the correct /srv/top.sls file ?
12:31 joehh thanks for that
12:32 pniederw joeh: my machines are on raring/saucy
12:35 joehh saucy build starts in 3 minutes. Then a little bit of time for it to be published and copied to right place.
12:35 joehh raring also close
12:36 gasbakid joined #salt
12:38 BenBen Well, got my answer. it does.
12:39 pniederw how comes I get this in my minion log with latest ubuntu ppa? "[WARNING ] You have a version of ZMQ less than ZMQ 3.2!"
12:40 joehh pniederw: good question
12:40 joehh can you you run dpkg -l |grep python-zmq
12:41 pniederw joehh: 2.2.0.1-1
12:43 pniederw could this be connected to running into the pkgrepo problem? I accidentally installed "latest" instead of 0.17.1, which gave me 0.17.2
12:43 joehh no - this is different
12:43 joehh is this saucy?
12:43 joehh or raring?
12:43 pniederw raring
12:44 joehh interesting - I'll do some checking
12:45 viq That jinja bug that I saw mentioned here recently - is it fixed in 0.17.2, or was it past that?
12:46 joehh viq: I believe it was fixed in 0.17.2 - haven't seen it myself though
12:46 viq Oh, wait, I have now a different error, I'll go investigate that first
12:47 pniederw on the raring machine where I just installed salt-master 0.17.2, I get: python-zmq 13.0.0-2raring
12:47 redondos joined #salt
12:48 pniederw on the raring machine where I installed salt-minion 0.17.2, I get: python-zmq 2.2.0.1-1
12:48 joehh pniederw: I'm a little confused - is that two raring machines, each with a different python-zmq
12:48 joehh got htat now
12:48 pniederw yes, two machines
12:48 BenBen Is there anything special to know about the salt-ssh command regarding targetting ? i can't seem to have any state.highstate command do anything unless i use the '*' rule... ?
12:49 pniederw do you have a roster file?
12:49 BenBen yeap
12:49 BenBen i can issue test.ping command
12:49 alekibango joined #salt
12:49 BenBen but the top.sls file rules aren't met
12:50 pniederw I haven't had much success with highstate and salt-ssh, but `salt-shh node1 state.sls some.state` works for me
12:50 joehh pniederw: very odd - don't understand it, but it seems to have picked up the default python-zmq
12:50 BenBen what should "state.sls' be ? can it be top.sls ?
12:51 joehh you may have some connection problems until you upgrade the minion to python-zmq 13 or greater (zmq 3.2)
12:51 dzen left #salt
12:51 honestly if you want to execute top.sls, just run state.highstate
12:51 pniederw joehh: yes I do :-)
12:51 BenBen honestly: i tried that, but the rule isn't met..
12:51 joehh The ubuntu packages do not rely on particular versions of packages - something I think I will need to fix shortly
12:52 joehh what happens if you do apt-get install python-zmq
12:52 joehh I'm not familiar enough with ubuntu to force it to get a package from a particular ppa
12:52 srage joined #salt
12:52 pniederw that installed 13.0.0-2raring
12:53 joehh good to hear
12:53 joehh that should improve your connection problems
12:53 pniederw is it possible that I ended up with more outdated dependencies?
12:54 pniederw I wonder how I should adapt my salt-minion bootstrapping, to rule out this problem
12:54 joehh possible, but that is really the key one
12:54 noob joined #salt
12:55 joehh how have you been doing the bootstrapping?
12:55 pniederw via salt-ssh
12:55 pniederw I have a salt-minion sls that I push via salt-ssh
12:55 joehh can you paste that somewhere?
12:57 gasbakid joined #salt
12:57 pniederw joehh: https://gist.github.com/pniederw/ecb31aba3b994315ca54
12:58 BenBen does pillar works with salt-ssh ?
12:59 joehh pniederw: does it ever install version 0.12.1-1 rather than something newer?
13:00 pniederw after I ran that sls, I had 2.2.0.1-1
13:00 pniederw I can uninstall and try again
13:00 joehh that was salt 0.12.1-1
13:01 mgw joined #salt
13:01 pniederw `salt-minion -version` showed 0.17.2
13:01 joehh odd, my guess was that it was running the salt install before the ppa
13:01 joehh but maybe not
13:02 joehh you would definitely get pyzmq 2ish if you installed the "ubuntu (old debian) packaged" salt
13:03 joehh what happens if you separate out your pkgrepo ppa and pkg.installed bits and have the pkg.installed require the pkgrepo state
13:03 joehh ?
13:04 joehh either that or put a pkg.latest on python-zmq (and maybe ensure salt-minion is restarted afgter an upgrade)
13:09 viq What's the difference in jinja between {% something -%} and {%- something %} ?
13:10 bhosmer joined #salt
13:10 blee joined #salt
13:11 jlorence joined #salt
13:11 BenBen hmm, seems like pillars doesn't work fine with salt-ssh https://github.com/saltstack/salt/issues/8340
13:11 jrdx joined #salt
13:12 Ryan_Lane joined #salt
13:12 jsm joined #salt
13:15 balboah viq: strip whitespace on top vs bottom maybe?
13:15 viq And which would I use for "shut up about empty pillars" ?
13:16 * viq wonders if jesusaurus is around, as it was him that pointed out this stuff
13:19 balboah viq: depends on what you need as output. I'm unfamiliar with the pillar problem. But if your pillar problem goes away by removing whitespaces you should take a look at http://jinja.pocoo.org/docs/templates/#whitespace-control
13:20 viq balboah: specifically, I have a list of pillars, say group1, group2, group3, each with a list of people, then I have states, again group1 group2 group3 that fetch data from relevant pillars and create users.
13:20 viq Right now I'm getting an error, and my suspicion is that it's because some of the pillars are empty
13:21 viq Actually I guess that's easy to verify...
13:21 mgw joined #salt
13:21 sgviking joined #salt
13:25 pniederw joehh: after uninstalling salt-minion and python-zmq and running the boostrap-minion sls once more, I now have the correct python-zmq version. not sure what was going on.
13:26 xt joined #salt
13:27 Cansas joined #salt
13:27 Cansas hi, is there a way to increase by a state the ulimit of the salt slaves for a specific user?
13:28 Cansas didn't found anything in the docs
13:28 joehh pniederw: yeah - odd - should be much more reliable once you are on pyzmq 13 (zmq 3+)
13:28 Cansas except of increasing the ulimit for the saltmaster itself
13:28 joehh I'll see if we can get minimum versions into the next ubuntu packages
13:29 pniederw joehh: same problem on next machine. only works after manually uninstalling python-zmq
13:38 dangerousbeans joined #salt
13:38 pniederw how do I debug minion connection problems? minions are running, don't log anything, don't connect to master
13:39 balboah maybe increase log level and do tcpdump
13:39 balboah I'm currently having issue with my virtual machine as well, probably not salt related
13:40 joehh Are these probs on pyzmq 2.2 machines or 13+ machines?
13:40 favadi joined #salt
13:41 joehh the first step always seems to be get them running with zmq 3+
13:42 ravibhure joined #salt
13:45 pniederw is python-zmq 13.0.0-2raring ok? that's what it gave me after reinstalling the minions
13:45 joehh should be
13:45 pniederw anything else I have to do?
13:46 joehh most people report no problems after that - so I haven't dealt with anyone debugging after they have upgraded both ends
13:48 redondos joined #salt
13:51 jY does salt support restarting a service listed in another sls file?
13:53 wilywonka joined #salt
13:54 vejdmn joined #salt
13:55 blee_ joined #salt
13:55 wkf joined #salt
14:02 hazzadous joined #salt
14:03 srage joined #salt
14:03 balboah joehh: just a small test with one master and minion seems to work with the new precise ppa. not that I verified jinja bugs or anything
14:03 joehh great - thanks for that - that was the main thing I wanted to check
14:04 brianhicks joined #salt
14:05 racooper joined #salt
14:06 ipmb joined #salt
14:08 wkf joined #salt
14:08 sroegner_ joined #salt
14:09 aleszoulek joined #salt
14:09 mpanetta joined #salt
14:12 juicer2 joined #salt
14:14 balboah hmm
14:15 balboah actually one salt master command just returned and did nothing it seemed
14:17 gasbakid joined #salt
14:18 jergerber joined #salt
14:19 whiskybar joined #salt
14:22 mannyt joined #salt
14:23 jimallman joined #salt
14:23 pniederw anyone got the gitfs backend with git+ssh to work?
14:24 jeff__ joined #salt
14:25 pniederw if I use the url format advertised in the salt docs, I get "repo not found". if I use the one I use for "git clone", I get "hostname not found". git clone works fine.
14:26 Ahlee pniederw: I use Atlassian Stash, and under gitfs_remotes: - git+ssh://git@gitserver:7999/salt/saltmaster.git
14:26 pniederw the two differ in that the former has "git@github.com/", whereas the latter has "git@github.com:"
14:27 pniederw hmm, that looks the same as mine (and as the salt docs)
14:29 pniederw it's a private repo, so "repo not found" might indicate an authentication problem, but I can't spot a mistake in the setup. salt-master runs under the same user for which a manual `git clone` works fine, so ssh keys must be good.
14:30 mgw joined #salt
14:31 APLU joined #salt
14:31 th3reverend joined #salt
14:33 Ahlee pniederw: you don't have key forwarding enabled or something by chance?
14:34 dangerousbeans Hey, how do I lay out this sort of thing? http://pastebin.com/2g0w3Zk9
14:35 dangerousbeans I want to install something but it requires a series of cmd.run things
14:35 dangerousbeans but it's all WARNING NOT A LIST BLAH BLAH
14:36 Corey dangerousbeans: A few things.
14:36 Corey a2enmod just copies symlinks; I'd manage those rather than calling it via cmd.run.
14:36 elfixit joined #salt
14:36 pniederw now I'm seeing a different gitfs error in the salt-master log. it's tripping over the message "Warning: Permanently added the RSA host key for IP address ...", which it apparently didn't expected
14:36 Corey dangerousbeans: Additionally, there's a restart command you can use for apache, and have it watch the symlinks in turn.
14:37 dangerousbeans Corey: okay but those two things are working, what I'm struggling with is the syntax of the .sls file
14:37 dangerousbeans but will look into doing this a less sketchy way :]
14:37 Corey dangerousbeans: Then what's the issue? What you pasted seems valid. :-)
14:37 Corey (Sorry, I harp on best practices...)
14:38 Corey The exact error message would be useful. And if BLAH BLAH appears in it I'll eat my hat.
14:38 pniederw or maybe it's the other way around, it didn't expect that message
14:38 dangerousbeans Corey: I attempted to devide it into blocks, "passenger", "ssl", "rewrite"
14:38 pniederw sorry: did expect
14:38 dangerousbeans before it was just a series of cmd.runs
14:38 dangerousbeans which is gave a warning about
14:39 Corey Oh, I see what you're doing.
14:39 dangerousbeans old warning I was trying to fix: [WARNING ] Duplicate Key: "cmd.run" found in salt://nowa.nowa_web environment=base
14:39 dangerousbeans new warning: Stacktrace python oh god blah blah blah
14:39 dangerousbeans :P
14:40 Corey Yeah, enabble_modules: cmd.run: -name cmd1 -name cmd2
14:40 dangerousbeans [CRITICAL] Rendering SLS rvm_passenger_apache_module failed, render error: while parsing a block mapping
14:40 dangerousbeans ah so you can past a list of -name things?
14:40 dangerousbeans sweet.
14:40 dangerousbeans THNX
14:41 dangerousbeans so
14:41 dangerousbeans # enable modules
14:41 dangerousbeans enable_module:
14:41 dangerousbeans cmd.run:
14:41 dangerousbeans - name: "a2enmod passenger"
14:41 dangerousbeans - name: "a2enmod ssl"
14:41 dangerousbeans - name: "a2enmod rewrite"
14:41 dangerousbeans - name: "service apache2 restart"
14:41 dangerousbeans ?
14:41 Ahlee pniederw: yeah, rsa key warnings will throw it for a loop
14:42 pniederw can I fix this by configuring ssh?
14:43 Corey dangerousbeans: Pastebin or gist plz.
14:43 Ahlee yes. Accept the key as the user running your saltmaster
14:43 Ahlee i loged in directly as root on my master to ensure no keys/etc
14:43 Ahlee overkill of course, but proved effective
14:43 pniederw hmm, I guess I already did that when I manually cloned the repo to see if the url is right. I did use the same user then.
14:44 Ahlee sounds like you have a user mismatch somewhere if it's still prompting for rsa key trust
14:45 elsmorian joined #salt
14:46 juicer2 joined #salt
14:47 wkf joined #salt
14:48 pniederw don't think so. salt-master is definitely running under the correct user, and otherwise there doesn't seem to be anything to configure. docs say that gitfs uses same user that salt-master is running under.
14:49 redondos joined #salt
14:49 smccarthy joined #salt
14:49 anuvrat joined #salt
14:49 cachedout joined #salt
14:50 Ahlee correct.
14:50 foxx joined #salt
14:51 Ahlee You could always just set .ssh/config to ignore keys :)  Up to you to determine the security implications and if you're comfortable with them
14:53 pniederw here is the log error. I don't know which side is expected vs. actual. https://gist.github.com/pniederw/b285a5a47dde1a83f634
14:54 AdamSewell joined #salt
14:54 Voziv left #salt
14:55 bhosmer joined #salt
14:56 MTecknology *grumble* Out of 104 states, 2 are failing. Why? http://dpaste.com/1474800/  ... what?
14:57 Brew joined #salt
14:57 MTecknology wait.... regression!
14:58 xt implausible!!!
14:58 Ahlee pniederw: that's funky man. Probably a gitpython issue, which i know the guys don't like
14:59 kaptk2 joined #salt
15:01 dccc joined #salt
15:02 jslatts joined #salt
15:02 dangerousbeans I'm getting this:
15:02 dangerousbeans [ERROR   ] Got a bad pillar from master, type bool, expecting dict: False
15:02 dangerousbeans what does it mean? :C
15:03 moos3 joined #salt
15:03 dangerousbeans my other minion is happy
15:03 dangerousbeans this new one is all sadface though
15:03 pniederw starting to despair. whatever I try with salt, I'm running into problems.
15:04 joehh MTecknology: which version are you running, comment doesn't seem to be in 0.17.x
15:04 dangerousbeans pniederw: dev ops is always a pain, just different amounts of pain
15:05 dangerousbeans I agree that somehow the learning curve with salt is more intense than with chef/puppet
15:05 dangerousbeans even though it's simpler
15:05 elfixit joined #salt
15:05 MTecknology joehh: 0.17.2, I don't have any -comment: and assume it's mad because I don't
15:06 pniederw I don't have experience with other tools, and I'm a dev by training in need of cm, but my impression is that salt has a huge quality problem.
15:07 joehh line 288 of salt/modules/cron.py in 0.17 looks different https://github.com/saltstack/salt/blob/0.17/salt/modules/cron.py
15:07 joehh could you be getting versions crossed somehow?
15:07 ipmb_ joined #salt
15:07 bhosmer joined #salt
15:09 joehh pniederw: scale is a matter of opinion :) but for whatever reasons (good or bad) the early releases of a new "feature" branch/version
15:09 joehh tend to be interesting
15:10 pniederw in what ways is 0.17.2 an early release?
15:10 joehh For my purposes, I find that I stick to the old release but make sure I test and get any bugs fixed on the new branch as the releases come out
15:10 pniederw old release would be what? 0.17.1?
15:11 joehh old would be 0.16.4
15:11 gildegoma joined #salt
15:11 joehh typically I've found by x.3 things have stabilised pretty nicely
15:12 joehh I was (and am still) hoping that 0.17.2 will be a nice release too
15:12 sroegner_ joined #salt
15:12 joehh hasn't had a great start, but we'll see where it goes
15:13 pniederw that's what I'm using right now. I can try and go back to 0.16.4, but I don't have large expectations that it will make my gitfs problem go away. by biggest hope right now is configuring ssh to auto-accept keys.
15:13 joehh fwiw, I've recommended that $dayjob start moving to 0.17.2
15:13 joehh but I know what we use and am pretty comfortable it is working
15:13 quickdry21 joined #salt
15:14 joehh I'd try pinging base pi or utah dave once he shows up
15:14 balboah salt '*' pkg.install salt-minion refresh=True
15:14 balboah lets do it!
15:15 joehh they're saltstack guys and tend to have a broader and deeper knowledge
15:15 alunduil joined #salt
15:15 joehh balboah: good luck :)
15:15 balboah joehh: I'm cheating a bit moving only staging so far ;)
15:16 joehh good idea :)
15:16 micah_chatt joined #salt
15:18 sashka_ua joined #salt
15:21 dangerousbeans Okay I'm stuck
15:21 dangerousbeans help
15:21 dangerousbeans i got two minions connected
15:21 dangerousbeans http://pastebin.com/3SmZTtWy
15:21 dangerousbeans all fine
15:21 dangerousbeans web-prod-alpha is happy and runs state.highstate okay
15:21 dangerousbeans database however FREAKS OUT
15:22 dangerousbeans and says No Top file or external nodes data matches found
15:22 dangerousbeans this is my top.sls
15:22 dangerousbeans http://pastebin.com/MjQRVHCx
15:23 timoguin joined #salt
15:25 dangerousbeans tried some salt '*' saltutil.refresh_modules
15:25 dangerousbeans nothing :C
15:26 forrest joined #salt
15:26 dangerousbeans forrest: help
15:27 dangerousbeans I got some weirdness
15:27 joehh dangerousbeans: I reckon try a blank line between database and iptables?
15:27 mgw is there an issue with mount.mounted? Every state run adds another entry to fstab
15:27 dangerousbeans joehh: kay trying that
15:27 MrTango joined #salt
15:27 joehh don't thikn it will help...
15:27 forrest dangerousbeans, I'll try, what's up?
15:27 dangerousbeans i got two minions connected
15:27 dangerousbeans http://pastebin.com/3SmZTtWy
15:28 dangerousbeans all fine
15:28 dangerousbeans web-prod-alpha is happy and runs state.highstate okay
15:28 dangerousbeans database however FREAKS OUT
15:28 forrest lol
15:28 forrest so what you're saying is you broke your database states? :P
15:28 NotreDev joined #salt
15:28 dangerousbeans yes :C
15:29 dangerousbeans okay I'm trying without database
15:29 forrest cool
15:29 dangerousbeans same error
15:29 dangerousbeans it doesn't even like the * stuff
15:29 dangerousbeans Comment:   No Top file or external nodes data matches found
15:29 Elb__ joined #salt
15:29 dangerousbeans inside the minions /srv/ dir there is nothing though
15:29 forrest for your web server states?
15:30 dangerousbeans my web server guy is fine
15:30 dangerousbeans and working
15:30 dangerousbeans but my database minion is all ?????
15:31 gasbakid joined #salt
15:31 Elb__ Hello world, when i am trying to generate my self-signed TLS certificate with the the command salt '*' tls.create_self_signed_cert test , it always returns me  "tls.create_self_signed_cert" is not available. What i am doing wrong?
15:31 forrest and you can run highstate no problem on the web server, but the db freaks out? Let's start by running the highstate with -l debug if you didn't already do that
15:31 dangerousbeans kk on  it
15:32 balboah almost, http://bpaste.net/show/kSUmx4qklg0S4gyKkinV/
15:32 forrest dangerousbeans, once you do that, if the error doesn't easily present itself, let's start working through running each state on your database server, going for the ones with the least requires/includes first.
15:32 dangerousbeans forrest: http://pastebin.com/44tDqx8F
15:32 dangerousbeans it's a new thing, so there's nothing in there yet D:
15:32 dangerousbeans also if I take out any mention of it from the top file it still has the same error
15:32 teskew joined #salt
15:32 dangerousbeans I'll try changing the minion id
15:32 dangerousbeans to like
15:32 forrest any mention of the db server?
15:32 forrest you mean
15:32 forrest sorry I think I came in half way through the previous troubleshooting you guys did
15:33 dangerousbeans it's fine, thanks for helping
15:33 dangerousbeans so my top file has selectors for my 2 types of minion
15:33 dangerousbeans web guys and database guys
15:33 dangerousbeans if I delete the database part
15:33 dangerousbeans it's the same
15:33 forrest the same error?
15:33 redondos joined #salt
15:33 redondos joined #salt
15:34 dangerousbeans yes
15:34 gasbakid_ joined #salt
15:34 forrest ok, so first thing I see is [ERROR   ] Got a bad pillar from master, type bool, expecting dict: False
15:34 dangerousbeans ok I changed the minion_id to "FACE"
15:34 dangerousbeans and it's the same
15:35 forrest yea I don't think it's a minion ID problem. Let's first confirm there isn't anything wrong with Salt itself. Can you write a very simple state that's simple a cmd.run or something? Then run that with salt 'web-production-alpha' state.sls newState
15:35 forrest or whatever you name it
15:35 forrest *simply
15:36 dangerousbeans k
15:36 pniederw with `stricthostkeychecking no` I no longer see a gitfs error message in the log, although I don't see a "successful" message either
15:36 zach Has anyone tied salt output into a bot of some sort (jabber/irc/etc) ?
15:37 jsm joined #salt
15:37 zach I wonder how dangerous it could potentially be
15:37 forrest zach, I haven't seen someone try that yet, I know there was a guy in here about a week ago who was working on sanitizing/cutting down on the output so it simply returned whether the run was successful or not
15:38 zach That would be perfect. I gave the user the bot (jabber in this case) to very specific commands, so far so good
15:38 zach that the bot runs under*
15:38 forrest zach, that sounds pretty cool, you should write up a tutorial on it
15:38 forrest at the very least, things like test.ping would be nice to have.
15:39 zach indeed, I'll do a writeup on it
15:39 forrest sweet
15:39 forrest we should probably start a page on the docs for 'bots'
15:39 forrest would be nice to have them all collected in a central location, as well as links out to the repo where the configs and such live
15:40 forrest are you zfouts on github zach?
15:40 dangerousbeans okay tried a hello world test thing
15:40 dangerousbeans http://pastebin.com/myRnZFxd
15:40 zach https://pastee.org/6drw8 - so far
15:40 dangerousbeans it errors for the database guy
15:40 zach forrest: yep
15:40 joehh Furao: have you kicked 0.17.2-2 off in your tests yet?
15:40 forrest zach, ok cool, I will create an issue and ping you on it.
15:40 zach neat
15:41 zach I'll write this up and commit my code up there too when I get complete
15:41 forrest dangerousbeans, what does your top.sls look like?
15:41 BenBen joined #salt
15:41 quickdry21 Hey, I'm getting SaltReqTimeoutError: Waited 60 seconds any time I try to do a salt command - state.highsate, grains.items, etc
15:41 dangerousbeans forrest: http://pastebin.com/RevCL92M
15:42 forrest ok, so that
15:42 forrest are you using file_roots
15:42 dangerousbeans ?
15:42 forrest dangerousbeans, to keep web/database content apart
15:42 dangerousbeans I don't think so? D:
15:43 dangerousbeans I have a pillar
15:43 zach my python is pretty rusty - but here's the part that does the work: https://pastee.org/dhtz6
15:43 dangerousbeans but it's just *
15:43 forrest cool zach
15:43 zach dangerousbeans: I setup a file_root one time that had too many files, I ran into the same issue
15:43 cnelsonsic joined #salt
15:44 Furao joehh: waiting for your fixes in 0.17.2
15:44 forrest ok dangerousbeans, let's strip this back to the basics. Move your pillar file out of /srv/salt/pillar, and restart the service, then try to run the state again.
15:44 dangerousbeans not 100% sure what file_root is or means, but I've not turned it on
15:44 amahon joined #salt
15:44 dangerousbeans kk
15:44 forrest dangerousbeans, the file_roots option makes it so you can say 'ok all dev content comes from THIS location, and all prod content comes from THIS OTHER location', it can keep things organized if you have a lot of content.
15:45 dangerousbeans ah nope, not god that I don't thin
15:45 dangerousbeans k
15:45 Furao joehh: oh -2 is out
15:45 Furao I'll do that tomorrow!
15:45 dangerousbeans moved pillar folder, same issue
15:45 forrest after restarting the service?
15:46 dangerousbeans I'm just calling salt-call state.highstate
15:46 dangerousbeans is bad? :C
15:46 forrest ok don't call highstate
15:46 forrest you moved the pillar, and restarted the service, we want to specifically test that one really simple state you wrote
15:46 joehh Furao: sounds good - I'm keen to hear how they go
15:47 dangerousbeans by "The Service" what do you mean?
15:47 dangerousbeans salt master?
15:47 dangerousbeans http://bit.ly/YgHDE9
15:47 forrest yes
15:47 dangerousbeans k
15:47 forrest It's all good man, let's just slow down and work through it
15:47 forrest something is obviously broken, we just want to work through things slowly and get both machines returning first.
15:48 zach forrest: is there a way to get ping.test to return false if they are down?
15:48 dangerousbeans ran salt '*' state.sls test_hello_world
15:48 dangerousbeans got the same No matching sls found for 'test_hello_world' in env 'base'
15:48 forrest dangerousbeans, even after a salt-master service restart?
15:48 dangerousbeans it's like it doesn't have the slsss's from the master
15:48 dangerousbeans ja
15:48 zach ls /srv/salt ?
15:48 forrest zach, I don't think so zach, at least not that I've seen. Maybe the return code?
15:48 zach is it even in there?
15:48 dangerousbeans it is on the master
15:48 forrest yea that was my next thing, what's in /srv/salt
15:48 dangerousbeans the minions /srv is empty
15:48 forrest right
15:48 zach that's fine
15:48 zach they can be empty
15:49 dangerousbeans but that's the same on the working minion
15:49 dangerousbeans k good
15:49 dangerousbeans root@master:/srv/salt/test_hello_world# ls
15:49 dangerousbeans init.sls
15:49 kamal_ joined #salt
15:49 dangerousbeans it's thur.
15:49 clone1018 joined #salt
15:50 zach try renaming it to /srv/salt/testhelloworld.sls and salt '*' state.sls testhelloworld
15:50 zach just curious if that will work
15:50 dangerousbeans zach: the state runs on another minion though
15:50 dangerousbeans but will try
15:50 redondos joined #salt
15:50 zach It's only one that it does not work on? Is there a firewall between that is not permitting the port?
15:50 forrest nah zach, test.ping works fine
15:51 dangerousbeans er that broke them both
15:51 zach cool
15:51 forrest that broke both of them? So now they both can't find it?
15:51 arapaho joined #salt
15:51 zach Very interesting
15:51 dangerousbeans I think now it's just invalid or something
15:51 forrest what's the error look like?
15:51 dangerousbeans No matching sls found for 'testhelloworld' in env 'base'
15:52 micah_chatt_ joined #salt
15:52 dangerousbeans oh wait
15:52 dangerousbeans derp
15:52 zach ;-) Something simple wasnt it?
15:52 dangerousbeans I'd renamed the folder wrong
15:52 dangerousbeans now it's as before
15:52 jalbretsen joined #salt
15:53 dangerousbeans zach: no sorry still the same, I'd just renamed the folder testhelloworld.sls
15:53 dangerousbeans by accident
15:53 forrest so now you just have the sls file in /srv/salt right?
15:53 dangerousbeans forrest: nope, but I can strip everything out?
15:53 forrest nah, just move the actual sls file so it's under /srv/salt.
15:53 dangerousbeans the test one?
15:54 dangerousbeans on the master?
15:54 dangerousbeans [panic]
15:54 forrest Is this a production issue?
15:54 dangerousbeans nope it's all fine
15:54 forrest ok, so relax :P
15:54 dangerousbeans I'm rebuilding the entire stack with salt instead of chef
15:54 forrest we're in no rush.
15:55 forrest all the work we're talking about here we're doing on the master, so move the the sls from within /srv/salt/test_hello_world/ to /srv/salt/
15:55 dangerousbeans my only vague theory about this is that this minion was an 'old' one
15:55 forrest so the actual sls file lives in /srv/salt
15:55 dangerousbeans like I started it early on
15:55 dangerousbeans messed with stuff on master
15:55 dangerousbeans left it alone for ages
15:55 forrest well, they're all running the same version of salt right?
15:55 dangerousbeans then just now started to do stuff with it
15:56 dangerousbeans I ran salt-call update.everything or whatever to make it update the salt install
15:56 dangerousbeans moving files, sec
15:57 forrest can you do salt --version on the master, and then salt-call --version on the minion?
15:57 dangerousbeans mv testhelloworld/init.sls /srv/salt/testhelloworld.sls
15:57 forrest right
15:57 dangerousbeans k
15:57 forrest now run salt '*' state.sls testhelloworld
15:57 dangerousbeans master: salt 0.17.0-4338-g44c0079
15:57 dangerousbeans broken minion: salt 0.17.0-4793-g265d51c
15:57 dangerousbeans omg different O.O
15:58 dangerousbeans k upgrading master salt
15:58 forrest I'm not super concerned about that, it would have been an issue if the versions were 16 and 17
15:59 zach dangerousbeans: good choice for using salt over chef ;-)
15:59 mgw basepi: are you aware of any issues with mount.mounted added a new line to fstab on every state run?
15:59 dangerousbeans :]
15:59 dangerousbeans ok upgraded mah saltz
15:59 dangerousbeans same thing happening
15:59 dangerousbeans :C
15:59 zach I've had more issues with chef/puppet than any of the issues with salt I've ran into
15:59 forrest when you run that command?
15:59 zach dangerousbeans: ls /srv/salt - is the test_hello_world.sls in that directory?
16:00 forrest you still get http://pastebin.com/myRnZFxd
16:00 dangerousbeans yes
16:00 dangerousbeans but it's testhelloworld.sls
16:00 forrest interesting, so still no matching sls found?
16:00 dangerousbeans minion /srv still empty
16:00 zach minion /srv will be empty
16:00 dangerousbeans still not on the one minion, working minion still working
16:00 forrest ok
16:01 forrest can you go into /etc/salt/master and find file_roots
16:01 forrest and confirm that there are no entries under that?
16:01 dangerousbeans k
16:01 JasonSwindle joined #salt
16:01 dangerousbeans all commented out
16:01 zach by default it's commented out
16:01 dangerousbeans no rootz
16:02 forrest ok then, let's move your /srv/salt/top.sls to another location (or rename it to have a non.sls extension), and then restart the salt master, and try the command again
16:02 dangerousbeans k
16:02 cym3try joined #salt
16:03 dangerousbeans same :C
16:03 forrest ok, can you trash /var/cache/salt?
16:03 forrest on the master
16:03 forrest then restart the master again
16:03 dangerousbeans k
16:04 zach Can you post your sls file too?
16:04 dangerousbeans same :C
16:04 zach is it just cmd.run:\n\s\s\s\secho 'hello world' ?
16:04 dangerousbeans the test one?
16:04 dangerousbeans test_hello_world:
16:04 dangerousbeans cmd.run:
16:05 dangerousbeans - name: 'echo "Hello world"'
16:05 zach looks fine to me
16:05 dangerousbeans it also runs on the working minion
16:05 danielbachhuber joined #salt
16:05 UtahDave joined #salt
16:05 dangerousbeans web_alpha:
16:05 dangerousbeans stdout:
16:05 dangerousbeans Hello world
16:05 forrest yep the state is good, it's just failing on that one minion, what happened after you removed the cache directory and restarted the master? Did the state return the same error?
16:05 dangerousbeans yeah no change
16:05 forrest hm
16:06 dangerousbeans is there some kind of 'salt refresh plz' command?
16:06 dangerousbeans I tried some salt-call systools.refreshmodules or something
16:06 dangerousbeans but they didn't fix it
16:06 dangerousbeans but also <no_idea_what_i'm_doing_dog.jpg>
16:06 UtahDave dangerousbeans: salt '*' saltutil.sync_all
16:07 dangerousbeans k
16:07 forrest hey UtahDave
16:07 UtahDave hey, forrest
16:07 forrest You out on assignment again this week?
16:07 dangerousbeans same :C
16:08 forrest hmm, I'm trying to think what else could cause that...
16:08 dangerousbeans http://irclog.perlgeek.de/salt/2013-01-11
16:08 dangerousbeans so this IRC log
16:08 dangerousbeans shows some other guy had this problem
16:08 dangerousbeans with no solution :C
16:08 dangerousbeans omg Utah Dave is there.
16:08 dangerousbeans in history
16:08 dangerousbeans forever
16:08 cnelsonsic joined #salt
16:08 forrest did you upgrade the version of salt running on the minion as well?
16:09 forrest the salt master is on 0.17.1 now right?
16:09 forrest zach, as a heads up I tagged you on that issue if you don't get github emails
16:09 dangerousbeans sec
16:09 forrest it's 8682
16:09 zach Cool, thanks
16:09 favadi joined #salt
16:09 zach Hopefully I updated it from my @rackspace.com email ;)
16:09 forrest heh
16:09 * zach looks
16:10 zach perfect, got it :)
16:10 zach Thanks
16:10 dangerousbeans master: salt 0.17.0-4338-g44c0079
16:10 dangerousbeans minion: salt 0.17.0-4793-g265d51c
16:10 forrest I thought you said you updated the master?
16:10 zach also thanks for pushing out 0.17.1-1.el6 to epel-testing guys :)
16:10 dangerousbeans I thought I did :C
16:10 forrest or did you just mean you restarted it?
16:10 forrest zach, 0.17.2 should be there almost
16:11 forrest terminalmage pushed it in the night before last :P
16:11 zach Oh great! That's what I thought it was, oops
16:11 dangerousbeans okay I apt-get update/upgraded both machines
16:11 zach Need to get that change rolled out for the bug we encountered last week
16:11 dangerousbeans but the minion has a newer version for some reason
16:11 patrek joined #salt
16:11 zach Thankfully I've been keeping a closer eye on things
16:12 forrest zach, yea I know a few people were having issues with 0.17.2 that cropped up
16:12 forrest so be careful
16:12 terminalmage forrest: I pushed 0.17.2-2 last night to address a pkgrepo regression
16:12 forrest some guy was saying that apt repo support wasn't working
16:12 terminalmage forrest: ^^^
16:12 zach ah who cares about apt support ;)
16:12 forrest oh that fixes the issue I just mentioned then, awesome terminalmage
16:12 bhosmer joined #salt
16:12 zach at least on the environment I use it, not a big deal heh
16:12 terminalmage joehh patched for debian/ubuntu
16:13 zach My personal environment though...that would be problematic
16:13 forrest awesome
16:13 forrest dangerousbeans, so is there a new version available to you from the ppa or no?
16:13 zach we're 100% RHEL 5/6 in this particular environment that I hit that wall
16:13 dangerousbeans nope, they both think they're upgraded
16:14 forrest dangerousbeans, just to double check, can you restart both minions, and md5sum their /etc/salt/minion files to confirm the config is the same?
16:14 forrest zach, are you guys using satellite?
16:14 terminalmage zach: 0.17.1 should be in stable now
16:14 dangerousbeans forrest: kay
16:14 zach dangerousbeans: try this: http://pastebin.com/iE7AUby7
16:14 zach terminalmage: perfect - thanks
16:14 zach salt-master                                noarch                                0.17.1-1.el6                                epel                                 31 k
16:15 zach Absolutely correct
16:15 forrest zach, what makes you want to do that as an option over the echo from before?
16:15 zach I'm just curious if the 'uptime' command will work
16:15 forrest yea, I don't think the state is even getting applied to the minion, it's acting as though the machine doesn't have access to the environment the state lives in or something
16:16 zach If that works, then something else is goofy
16:16 zach Can you go into the minion and run: 'salt-call state.sls blah' ?
16:16 dangerousbeans md5s are the same
16:16 dangerousbeans trying salt-call
16:17 dangerousbeans root@web-production-alpha:~# salt-call state.sls testhelloworld
16:17 dangerousbeans success
16:17 zach Interesting
16:17 forrest what about on the db minion>?
16:17 forrest that is having the issue
16:18 dangerousbeans root@web-production-alpha:~# salt-call state.sls testhelloworld
16:18 zach no result?
16:18 dangerousbeans local:
16:18 dangerousbeans Data failed to compile:
16:18 dangerousbeans ----------
16:18 dangerousbeans No matching sls found for 'testhelloworld' in env 'base'
16:18 srage joined #salt
16:18 dangerousbeans sorry that last was from the db
16:18 dangerousbeans same output as from the master
16:18 zach Remove the key for it on both minion and master, restart the minion, salt-key -A on the master and try again
16:19 zach I DOUBT that is the issue, but process of elimination
16:19 terminalmage zach: np
16:19 dangerousbeans zach: kay
16:19 forrest yea zach, we are eliminating a lot here, the pillar and top file are already gone, I mean we're down to running single comand line states :\
16:19 dangerousbeans how do you remove the keys? :C
16:19 zach rm /etc/salt/pki/master/minions/<name>
16:19 dangerousbeans master is set to auto accept everything (yeah I know)
16:19 [diecast] joined #salt
16:20 zach rm -rf /etc/salt/pki
16:20 forrest lol
16:20 * zach hacks dangerousbeans
16:20 forrest just make sure to restart all the services after you do that
16:20 dangerousbeans oh noes mah desktop is now your server
16:20 zach ;)
16:20 zach I have a salt-master wide open to the internet somewhere
16:20 TheRealBill joined #salt
16:20 blee joined #salt
16:20 zach one of my many sandbox servers
16:21 dangerousbeans er so I remove the key from the minion too?
16:21 nahamu zach: so anyone in the world is free to make their computer one of your minions? :)
16:21 NotreDev joined #salt
16:21 zach dangerousbeans: yeah lets eliminate both
16:21 zach nahamu: :D
16:21 zach nahamu: People can become my minion and allow me to have root access to their machines, for sure!
16:21 pipps_ joined #salt
16:22 dangerousbeans rm /etc/salt/pki/minion/minion.pem ?
16:22 dangerousbeans on minion
16:22 zach just kill the whole pki dir
16:22 zach and restart salt-minion
16:22 dangerousbeans k
16:22 zach it will re-generate
16:23 aleszoulek joined #salt
16:23 dangerousbeans okay nuked it, restarted the minion and master
16:23 zach With what forrest had you do, I'm not sure what's going on with your setup
16:23 dangerousbeans running test sls
16:23 forrest zach, yea, it's almost like that one minion recognizes the master, but not the location of the states.
16:23 zach indeed
16:23 forrest since module commands work no problem
16:24 zach Do you have a custom minion config?
16:24 dangerousbeans database minion wont check in now
16:24 dangerousbeans sec
16:25 zach firewall? ;-)
16:25 zach anything in /var/log/salt/minion ?
16:26 dangerousbeans [CRITICAL] The Salt Master has rejected this minion's public key!
16:26 dangerousbeans D:
16:26 dangerousbeans oh noes
16:26 zach perfect, salt-key -A on your master
16:26 dangerousbeans The key glob '*' does not match any unaccepted keys.
16:26 dangerousbeans D:
16:26 zach heh
16:26 dangerousbeans also it's set to autoaccept mode
16:26 dangerousbeans Minion: ping salt
16:26 zach ls /etc/salt/pki/master/minions_pre ?
16:26 dangerousbeans works
16:27 dangerousbeans No such file or directory
16:27 zach maybe my memory is wrong
16:27 zach Your minion is just goofy acting
16:27 dangerousbeans ls /etc/salt/pki/minion/
16:27 dangerousbeans minion.pem  minion.pub
16:28 dangerousbeans I hate my minion :C
16:28 dangerousbeans I'm gunna nuke it and pretend it never was
16:28 dangerousbeans trying another machine
16:29 dave_den morning. just jumping in on this, and i haven't read my backlog… have you checked file_roots from the 'master_opts' when doing 'salt <inion> pillar.data'? also, 'salt <minion> cp.list_master_dirs'?
16:29 dangerousbeans hey dave_den
16:29 dangerousbeans I checked for file roots in the master config: nothing
16:29 dangerousbeans not tried those commands though
16:29 dangerousbeans but I'm just double checking another minion can connect and run
16:29 forrest He wants to see if something is cached for some reason dangerousbeans, give it a shot
16:29 dangerousbeans ok
16:29 forrest once you're done
16:30 WatfordJohn joined #salt
16:30 dangerousbeans so run
16:30 dangerousbeans salt 'database' pillar.data
16:30 dangerousbeans ?
16:30 dangerousbeans from the master?
16:30 dave_den yep
16:30 zach Is saltconf still open? I'm still trying to convince them to send me
16:30 dangerousbeans okay my minion can't connect right now though because of another issue
16:30 dave_den and look for the 'file_roots' setting in the 'master_opts' dict
16:31 nicksloan what does the editable option for pip.installed take as an argument?
16:31 dave_den oh, heh
16:31 forrest dave_den, so basically just one minion complains about the state not existing in the base env, we've removed the pillar and top files to eliminate as much as possible, cleared the caches, and restarted the services (note that the running version is 0.17.0)
16:31 forrest zach, yea as far as I know
16:31 dangerousbeans http://xkcd.com/349/
16:31 zach THe $899 price tag wasnt a problem, it was putting me in a hotel for a week they didnt like
16:31 forrest lol
16:31 nicksloan seems like that ought to be a boolean
16:32 zach Even said I'd pay for hotel out of pocket...he's hehawing around now
16:32 zach "Well we're waiting on our budget......" etc
16:32 ebodine i am using hgfs as a backend w/ version 0.17.1, it doesn't appear to be updating and I don't see any indicative message in the log.  anyone know what the update interval is?
16:32 jcockhren dangerousbeans++
16:32 forrest zach, let me know, I'm still looking for someone to split a room with since it's the same price for a double as it is for a single bed, unless you snore really loud or something.
16:33 zach I'd be open to it, I'll harass the boss again today
16:33 zach He might be against it since I'm taking a trip to vacation that wasnt approved :D
16:33 JasonSwindle If I am able to go, the SSCE will be out of packet
16:33 zach er, s/vacation/vegas/
16:33 dangerousbeans alright
16:33 dangerousbeans brought online another machine
16:33 dangerousbeans 'messenger'
16:33 dangerousbeans it now shows the same error
16:34 dangerousbeans :Q_____
16:34 mgw I found the fstab issue, but mount.mounted is still running umount/mount and then reporting changes on very state run. Is this expected behavior?
16:34 zach dangerousbeans: do you have ufw/iptables/etc running on your master/
16:34 bhosmer joined #salt
16:34 forrest dangerousbeans, can you make sure you have the latest ppa, and then upgrade to 0.17.1?
16:34 dangerousbeans ok will check
16:34 dangerousbeans the only weird thing I saw was
16:34 dangerousbeans on the messenger machine
16:34 dangerousbeans I ran salt --version to get what it was
16:35 dangerousbeans and it just errored a screenfull of python stacktrace
16:35 dangerousbeans so I did the bootstrap script
16:35 dangerousbeans (again)
16:35 dangerousbeans and then it fixed the screen
16:35 forrest zach, the registration deadline is jan 3 according to the site
16:35 dangerousbeans but this was the same sequence of events that happened with the database machine which is now a mess
16:35 forrest lol
16:35 pipps joined #salt
16:35 dangerousbeans so maybe upgrading with the bootstrap is like... upsetting it?
16:36 forrest I guess it might? But technically it shouldn't, and you used the bootstrap on the new machine right?
16:36 zach dangerousbeans: did you do a clean uninstall?
16:36 forrest and that was a clean installation
16:36 dangerousbeans nope I just ran the bootstrap
16:36 dangerousbeans copy pasta
16:36 dangerousbeans is bad? D:
16:37 dangerousbeans I have 3 more machines we can try this on
16:37 dangerousbeans S:
16:37 MrTango joined #salt
16:37 forrest lol
16:37 forrest fix the first ones, before we bring in more broken boxes.
16:37 zach I would just install them from your package manager
16:37 zach then you can blame the packager when it breaks
16:37 zach and keep your job
16:37 UtahDave zach: Yeah,  saltconf.com is where you can sign up.
16:38 jrdx joined #salt
16:38 * Gareth waves
16:38 zach UtahDave: thanks. I will try again today. he was onboard first until I told him hotel and airfare costs. I'll get him to commit to letting me go now :)
16:38 BenBen forrest: Finally i didn't use salt-ssh
16:38 forrest zach, what part of the country are you coming from?
16:38 aleszoulek joined #salt
16:38 zach Austin, TX
16:38 forrest BenBen, oh yea?
16:39 BenBen forrest: Yeah, i choose to have both a master and a minion on the same machine
16:39 forrest zach, that's not even that bad... it's like 250 round trip from Arizona for the flight, and the hotel is only 120 a night or whatever.
16:39 zach y'all just need to come down here for SXSW or something
16:39 dangerousbeans ok on master:
16:39 dangerousbeans did: sudo add-apt-repository ppa:saltstack/salt
16:39 forrest BenBen, ahh masterless minion, good idea.
16:39 dangerousbeans but it wont upgrade anymore
16:39 BenBen forrest: nope, master and minion
16:39 BenBen both on the same machine
16:39 dangerousbeans master is at: salt 0.17.0-4338-g44c0079
16:39 forrest BenBen, oh ok
16:39 BenBen I saw a tutorial on digitalocean with this setting and i found it quite nice
16:39 forrest we should try to get the master to 0.17.1 before anything else.
16:40 dangerousbeans ok is there another ppa?
16:40 BenBen and i didn't go for ec2, but choosed digitalocean instead
16:40 dangerousbeans thanks for helping gaiz :D
16:40 jY 6:20 i heard both kids get up at the same time and took off
16:41 forrest BenBen, yea I like digitalocean, a lot cheaper, and I like their interface more for small jobs.
16:41 jY opps
16:41 forrest dangerousbeans, did you already run apt-get update to pull in the new repo data?
16:41 dangerousbeans yep
16:41 forrest 0.17.1 should be in the ppa right joehh?
16:42 MTecknology joehh: I finally tracked down the problem code, now I'm trying to figure out a proper fix
16:43 BenBen forrest: I also choosed this minion/master configuration because i had too many issues with salt-ssl
16:43 BenBen salt-ssh i mean
16:43 forrest oh yea?
16:43 BenBen i couldn't get the targetting to work
16:43 forrest ahh ok
16:43 forrest ahh crap I've gotta go to a meeting
16:43 forrest let me know if you guys figure that out
16:43 BenBen and i also couldn't manage to get pillar settings injected in managed files
16:44 BenBen ok, see u
16:44 amckinley joined #salt
16:45 aleszoulek joined #salt
16:46 jY does salt support restarting a service listed in another sls file?
16:46 dangerousbeans bye forest, thanks for the help
16:47 UtahDave jY: yep!  just include the other file first
16:47 jY UtahDave: what if it's included already in the top.sls?
16:50 UtahDave Hm. I think it has to be included by the sls file that wants to restart the other one.
16:50 srage joined #salt
16:50 jY UtahDave: ok it seems if i double include it.. it doesn't run the highstate
16:50 zandy joined #salt
16:50 jY that is once in top.sls and in the other sls file
16:51 zach dangerousbeans: did you get it working?
16:51 dangerousbeans zach: nope :C
16:51 dangerousbeans sorry boss came back
16:51 dangerousbeans zach: I spun up a new machine and it started to show the same problem
16:51 MTecknology >100 states where a few sync many (a few hundred) files takes a long time...
16:51 dangerousbeans well, by new I mean
16:51 dangerousbeans different
16:51 UtahDave jY: It won't cause the included file to be executed twice
16:52 dangerousbeans it still had an older version of salt, which I upgraded with the bootstrap thing
16:52 jY UtahDave: ok i'll have to re-work things then.. thanks
16:52 dangerousbeans my gut instinct tells me that upgrading like this makes it unhappy
16:52 zach I would never upgrade that way
16:53 dangerousbeans but I gotta run
16:53 dangerousbeans thanks for the help
16:53 dangerousbeans I'll come back and let you know if I figure it out :D
16:53 zach enjoy
16:53 dangerousbeans baii
16:53 dangerousbeans xxx
16:54 bhosmer joined #salt
16:54 pipps_ joined #salt
16:55 pniederw joined #salt
16:56 srage joined #salt
16:58 heewa Is there a way of vaguely giving ordering of states being executed during a highstate run? Like maybe a priority number?
16:58 ovorp joined #salt
16:58 pniederw anyone else seeing salt-ssh calls routinely hang? sometimes they work, sometimes they just hang. I can see two salt-call processes on the target host.
16:58 MTecknology UtahDave: Should I have an issue created for a commit that fixes something in the develop branch?
16:59 redondos joined #salt
16:59 redondos joined #salt
17:00 MTecknology or should I just skip the issue part and request a merge?
17:00 UtahDave heewa: there is the    order:  option
17:00 xcbt joined #salt
17:01 UtahDave MTecknology: you can probably just send in the pull request
17:02 ovorp joined #salt
17:02 MTecknology Is it possible to tell a job to stop running?
17:03 MTecknology like... salt-call job.terminate 20131120110105193655
17:03 dave_den MTecknology: http://docs.saltstack.com/ref/modules/all/salt.modules.saltutil.html#salt.modules.saltutil.kill_job
17:04 dave_den or term_job/signal_job
17:05 Voziv joined #salt
17:05 MTecknology Signal 9 sent to job 20131120110105193655 at pid 14429  ..  but it's still running
17:05 Voziv When configuring the master, if I set the git fs to my own git repo, can I use that instead of maintaining a /srv/salt folder?
17:06 krak3n` joined #salt
17:07 srage joined #salt
17:07 KyleG joined #salt
17:07 KyleG joined #salt
17:09 srage_ joined #salt
17:10 jslatts Voziv: yes
17:11 Voziv do I still need to include -roots to have included states recognized?
17:12 MTecknology da frick? When I try to run state.highstate, it just hangs and acts like it's waiting to connect to the master
17:13 heewa UtahDave: Nice. Tried to use that just now to have the state that manages salt-minion itself be run first, to upgrade it and get a fix for a bug that a different state hits. But that bug was in rendering the sls file, so ordering didn't matter (they're all rendered, then execution starts).
17:13 MTecknology heck.. if I run salt-call state.highstate on the master, I get the same thing
17:13 heewa UtahDave: No worries, though, I guess I just have to do a 'state.sls components.salt_minion' first, then a highstate
17:13 MTecknology something is really screwed up, and I have no idea what
17:14 UtahDave heewa: have you tried putting   - order: 1    on those items you want run first?
17:15 mwmnj joined #salt
17:15 heewa UtahDave: right, but the problem I was having was in rendering the sls files themselves, so ordering didn't help me.
17:16 UtahDave heewa: ah, ok
17:17 kermit joined #salt
17:17 MTecknology Why would salt-call hang at this point?  http://dpaste.com/1475106/
17:18 jslatts Voziv: file roots? no, they will be ignored
17:19 ovorp1 joined #salt
17:19 jslatts Voziv: the different environments can map to different branches in git if that is what you need
17:19 jslatts Voziv: docs should have examples
17:19 troyready joined #salt
17:19 Voziv I'm trying to get it to work. Currently "salt '*' state.highstate" returns saying there's no top.sls file
17:22 terminalmage Voziv: roots is the default
17:22 UtahDave MTecknology: I'm not sure.
17:22 Voziv terminalmage: Yeah, I changed it to use gitfs
17:22 terminalmage ok, so if you change to use gitfs you then need to explicitly define a "- roots" line
17:23 terminalmage and keep in mind that any change to the master config file requires a restart of the salt-master service
17:26 pniederw I've changed the value for ext_pillar in the master config, but even after restarting the salt-master service, it still sees the old value. how is this possible?
17:28 pniederw my ext_pillar is pointing to a github repo, and no matter how I change the url, it's still trying to fetch from the old one
17:28 davet1 joined #salt
17:31 Katafalkas joined #salt
17:33 ovorp1 joined #salt
17:34 Voziv terminalmage: I did define - roots line. It still doesn't seem to be checking my repo out though
17:38 jacksontj joined #salt
17:38 pniederw Voziv: git-backed file server is working for me by now, but I can't get git-backed pillar to work.
17:39 terminalmage Voziv: so where is your top file?
17:39 terminalmage is it in git?
17:39 terminalmage I thought it was still in roots
17:39 pniederw yes
17:39 terminalmage /srv/salt or whatever
17:40 pniederw I have it under /salt in my git repo
17:40 terminalmage pniederw: ?
17:40 terminalmage was talking to Voziv
17:40 pniederw ah ok
17:41 terminalmage pniederw: can you pastebin your ext_pillar from the master config
17:41 terminalmage ?
17:41 zandy joined #salt
17:41 shinylasers joined #salt
17:41 pniederw I can. I don't know what the expected format is, so I tried a few. however it seems it isn't picking up my changes and still tries the very first version.
17:42 cron0 joined #salt
17:43 zandy_ joined #salt
17:44 _ikke_ pniederw: Have you tried to update the states?
17:44 _ikke_ saltutil.syncall
17:44 Topic for #salt is now Welcome to #salt - SaltConf in January!! http://saltconf.com | 0.17.2 is the latest | 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/
17:45 terminalmage pniederw: have you read the docs?
17:45 terminalmage pniederw: because there shouldn't be any confusion on what the format should be
17:45 _ikke_ is 17.2 officially released?
17:46 basepi _ikke_: announcement hasn't gone out
17:46 terminalmage pniederw: http://docs.saltstack.com/ref/pillar/all/salt.pillar.git_pillar.html
17:46 basepi but it's available on most packaging systems at this point
17:46 _ikke_ ok
17:48 pniederw terminalmage: I saw that site some time ago, but this time I only looked at "GitFS Backend Walkthrough". anyway, I think my problem is that master config change isn't picked up. even when I change url to FOOBAR, it still says it has a problem with "https://github.com".
17:50 pniederw terminalmage: http://pastebin.com/V8TMexq7
17:50 viq joined #salt
17:50 viq joined #salt
17:54 matanya joined #salt
17:57 jimallman joined #salt
17:58 Katafalkas joined #salt
17:59 wkf joined #salt
17:59 ovorp1 joined #salt
18:01 moos3 joined #salt
18:03 Linz joined #salt
18:03 terminalmage pniederw: do you have a top.sls in your git repository?
18:03 oraqol joined #salt
18:04 terminalmage (the one you're using for the git pillar)
18:04 Voziv terminalmage: Sorry grabbed lunch. It's in the root of my git repository. Contains a simple entry base: '*': -htop
18:04 oraqol Hi, I've been using salt on a few clusters and everything was working fine until this morning.  Now when I try to run anything including test.ping I get the following, any ideas?
18:04 oraqol Traceback (most recent call last):   File "/usr/bin/salt", line 10, in <module>     salt_main()   File "/usr/lib/python2.7/dist-packages/salt/scripts.py", line 103, in salt_main     client.run()   File "/usr/lib/python2.7/dist-packages/salt/cli/__init__.py", line 117, in run     for full_ret in cmd_func(**kwargs):   File "/usr/lib/python2.7/dist-packages/salt/client.py", line 357, in cmd_cli     **kwargs)   File "/usr/lib/python2.7/dist
18:04 pniederw yes I have
18:04 oraqol salt.exceptions.SaltReqTimeoutError: Waited 60 seconds
18:04 elfixit joined #salt
18:04 oraqol I tried googling but to no avail
18:05 terminalmage Voziv: there should be a space between them: "- htop"
18:05 Voziv terminalmage: There is
18:05 pniederw it's a symlink from /top.sls to /pillar/top.sls, but hopefully that won't matter
18:05 terminalmage what?
18:05 oraqol also tried restarting the salt-master daemon, restarting all the minion would be time consuming so I'm trying to save that troubleshooting step til last
18:06 terminalmage ok, unfortunately I need to step away
18:06 terminalmage done with lunch
18:06 gary joined #salt
18:06 _ikke_ oraqol: What version of salt?
18:06 terminalmage pniederw: Voziv: if you're not getting answered in a timely manner, I would suggest the mailing list
18:06 pniederw ok, thanks
18:06 terminalmage the IRC will scroll by quickly, but on the mailing list people can see it even hours later
18:07 xmltok joined #salt
18:07 jstrunk joined #salt
18:07 pniederw so far I've had more success with irc, but I'll try
18:07 Voziv terminalmage: Thanks
18:07 Guest83671 whats the problem ?
18:07 Voziv dwyerj: I'm trying to get gitfs backend to work. I get "no top.sls found"
18:08 dwyerj ahh fair enough, ive never tried gitfs.
18:08 pniederw Voziv: do you have top.sls at the root of the git repo?
18:09 dwyerj yeah otherwise you need to set gitfs_root
18:09 pniederw that's why I was asking
18:10 pniederw `/salt/top.sls` and `gitfs_root: salt` works for me
18:10 dwyerj i think i might use gitfs once ive finished writting all my states.
18:11 reuss left #salt
18:13 Voziv pniederw: yeah
18:13 Voziv pniederw: Can you find your top.sls under /var/cache/salt/gitfs?
18:14 anitak joined #salt
18:14 shinylasers joined #salt
18:15 JasonSwindle joined #salt
18:16 pniederw I can see it in /var/cache/salt/master/gitfs/refs/master/salt/top.sls
18:16 Voziv yeah, I don't see mine there
18:16 Voziv so I think I have a checkout issue
18:17 Voziv Salt runs as root, right?
18:17 dwyerj usually, you can run as non-root if you want
18:17 pniederw it's configurable. my salt-master doesn't
18:17 Voziv by default it does?
18:17 dwyerj http://docs.saltstack.com/topics/nonroot.html
18:18 forrest Voziv, you can modify the directories that your system looks at as well to run as a non root user if you don't want to chown those directories
18:18 Voziv forrest: Thanks, I'm not too concerned with running it as non-root yet. Just trying to get this gitfs thing working
18:19 forrest Voziv, I've created an issue to get that updated over here: https://github.com/saltstack/salt/issues/8649
18:19 forrest gotcha
18:19 forrest basepi, regarding your comment on 8649, I thought the same thing
18:20 basepi forrest: yep.  though i can also see a case where i might want to just specify a different config file for testing without otherwise modifying my environment
18:20 forrest yea, maybe it justifies another option as well
18:20 basepi but it seems like the most common use case will be wanting to change the config directory and root directory
18:21 forrest yea
18:21 jesusaurus viq: did you figure out your empty pillar thing?
18:21 basepi but you're right, for backwards compatibility we should just add another arg
18:21 forrest you just expect it to pull everything from the specified dir
18:21 basepi forrest: can you create a new issue for the new arg?  don't want to clutter the documentation issue with it
18:21 forrest sure
18:21 basepi thanks
18:21 MTecknology DAMNIT!!
18:22 basepi MTecknology: ?
18:22 Voziv pniederw: What does your remote syntax look like? git://github.com/my/repo/here.git ?
18:22 MTecknology This issue I'm fighting... it's already been fixed in develop. Why didn't I look at the changelog first?
18:22 forrest lol
18:23 basepi MTecknology: Don't know whether to reply with a smiley face or a frowny face.
18:23 basepi (:(
18:23 basepi there, that's both.  or something
18:23 forrest that bothers me
18:23 forrest it's like a sad penguin face, or a dude whose ears were chopped off.
18:23 basepi hahaha
18:24 Voziv lego head!
18:24 matanya joined #salt
18:24 srage_ joined #salt
18:24 pniederw Voziv: pillar git is now working. it's yet another salt bug, will file an issue.
18:25 Voziv pniederw: But for your gitfs, whats your syntax look like? I'm trying to get mine to check out properly :P
18:25 pniederw ah, ok
18:25 MTecknology basepi: HA! nah, it's fine. I just need to repackage and upgrade and all better
18:25 basepi =D
18:25 zach for apache.signal - is there a way to implement "graceful" as a possible arg?
18:25 MTecknology back later
18:25 Voziv pniederw: I'm trying this now:  - git+ssh://git@github.com/my/repo.git
18:25 srage joined #salt
18:25 pniederw gitfs_remotes:   - git+ssh://git@github.com/xxx/yyy.git
18:26 Thiggy joined #salt
18:26 Voziv Then I should restart salt-master and do salt '*' state.highstate  ?
18:26 pniederw did you turn on debug logging and check master log?
18:26 Voziv ha! it works now
18:27 Voziv pniederw: I did run it in debug mode but it didn't spew out any git errors that I could see
18:27 zach nevermind, it's actually there -     valid_signals = ('start', 'stop', 'restart', 'graceful', 'graceful-stop')
18:27 Voziv Except that GitPython wasn't installed which I installed an hour ago,  once I did that it stopped complaining but still wasn't working
18:28 sfz joined #salt
18:28 srage_ joined #salt
18:28 forrest basepi, https://github.com/saltstack/salt/issues/8691
18:28 forrest I think that covers it, feel free to comment with ideas or whatever.
18:30 wilywonka joined #salt
18:34 berto- joined #salt
18:34 matanya joined #salt
18:39 Linz joined #salt
18:39 pniederw ok now it fetches my ext_pillar, but it still fails to substitute pillar values in states.
18:41 zandy joined #salt
18:42 pentabular joined #salt
18:43 aleszoulek joined #salt
18:43 pniederw when it says "undefined jinja variable" for "hostname {{ pillar['fqdn'] }}:", doesn't this mean that it can't resolve "pillar" (rather than "fqdn")?
18:45 druonysus joined #salt
18:45 _ikke_ No, it can mean both
18:46 rmt joined #salt
18:47 Katafalkas joined #salt
18:48 jesusaurus which is why its useful to use the pattern: {{ salt['pillar.get']('some:key', 'default value') }}
18:49 zach Do you guys know if the mysql module is compatible with http://dev.mysql.com/downloads/connector/python/ vs the version in RHEL/Epel
18:50 zach Since we do not use the RH MySQL version, the python-MySQL package is incompatible
18:51 oraqol Hi guys, sorry, afk for lunch
18:51 oraqol My version is 0.16.0
18:51 tty-1 joined #salt
18:51 oraqol and I'm getting the above error
18:51 pniederw jesesaurus: the value is always defined, unless something goes wrong with salt. how is the long form better then? looks boilerplatey to me.
18:52 oraqol This in reference to that long error ending in salt.exceptions.SaltReqTimeoutError: Waited 60 seconds
18:53 pniederw turns out the same state works when applied via state.sls, but doesn't work when applied via state.highstate.
18:53 pniederw seeing this in the minion log when trying state.highstate: "TypeError: 'NoneType' object is not iterable"
18:54 foxx[cleeming] joined #salt
18:54 foxx[cleeming] joined #salt
19:01 mgw does cp.fileclient work behind a squid proxy?
19:01 mgw (using http scheme)
19:02 Katafalkas joined #salt
19:05 zfouts joined #salt
19:06 JulianGindi joined #salt
19:06 MatthewT1 joined #salt
19:10 lineman60 joined #salt
19:12 NotreDev joined #salt
19:13 Damoun joined #salt
19:13 MTecknology Yup.. that's almost irritating. I find an issue and then fix it. Once I fix it, I go to create a pull request but after checking out the latest, it's already fixed.
19:14 pniederw anyone using a template engine other than jinja?
19:14 MTecknology are there others?
19:14 MTecknology jinja4lyfe
19:14 pniederw it's completely configurable
19:15 pniederw I've seen `marko` stated as another option
19:15 pniederw for example
19:15 MTecknology There are others, but I don't think anyone in here uses them.
19:16 pniederw that's my impression too, everyone seems to be using jinja. just curious.
19:24 lineman60 joined #salt
19:24 zfouts left #salt
19:25 logix812 joined #salt
19:25 logix812 left #salt
19:26 elsmorian joined #salt
19:28 aloki joined #salt
19:28 Voziv How can I do string concatenation when using pillar?
19:28 jesusaurus what do you mean?
19:29 mkoistinen I'm on Mac OS 10.8. I've installed Salt via mac ports, but I'd like to start using the latest dev releases. How do I go about doing that?
19:29 Voziv jesusaurus:  I'm using {{ pillar['mysql']['database'] }}  but I want the final product to be "database.*" when I assign privileges. I don't know how to add the .*
19:30 Voziv Is any python valid inside the {{ }} ?
19:30 ajw0100 joined #salt
19:30 jesusaurus oh, you can just add that after the jinja: "{{ pillar['mysql']['database'] }}.*"
19:31 Voziv all wrapped in quotes?
19:31 jesusaurus or use a second variable instead of hard-coding it
19:31 DredTiger joined #salt
19:31 jesusaurus jinja is just a templating engine. if you want quotes in the final file, put them in the template
19:33 Voziv gotcha
19:33 Voziv is that what sls is?
19:35 davet joined #salt
19:36 dvogt joined #salt
19:37 jesusaurus no, sls is just salt (or Scheme according to github), but by default the sls files are jinja templates of yaml files
19:38 Linz joined #salt
19:40 quickdry21 joined #salt
19:40 ajw0100 joined #salt
19:40 Voziv so which style convention would they follow? Scheme or yaml?
19:41 Voziv yaml seems to agree in my editor
19:42 quickdry21 joined #salt
19:44 _jslatts joined #salt
19:44 StDiluted joined #salt
19:44 StDiluted afternoon all
19:44 moos3__ joined #salt
19:47 StDiluted anyone using salt for rails/unicorn app deployment or using salt in conjunction with capistrano?
19:48 Voziv Does jinja work in configuration files when using salt:// ?
19:48 StDiluted yes
19:48 StDiluted as long as you have templet: jinja
19:48 StDiluted err template: jinja
19:48 Voziv What do you mean?
19:49 StDiluted in your state declaration
19:49 StDiluted - source: salt://wherever/whatever.file
19:50 StDiluted - template: jinja
19:50 dwyerj http://docs.saltstack.com/topics/nonroot.html
19:50 dwyerj woops, excuse that
19:51 Ryan_Lane joined #salt
19:51 dwyerj i need to use something like bluetooth with my phone to unlock/lock my machine. im always scared im going to type my password in irc without realising if i forget to lock my machine
19:52 zach Son of a bitch. https://pastee.org/hf3fq
19:52 zach Anyone seen that before?
19:52 Corey Let's see.
19:52 Corey Nope, that's new.
19:53 zach :'(
19:53 UtahDave zach: Yeah, that's the new security token that was added to the communication protocol
19:53 zach What do i do to fix it?
19:53 zach Took down all my minions
19:53 UtahDave what version is your master and your minions?
19:53 zach master: salt 0.17.1
19:54 zach salt-minion 0.17.1
19:54 micah_chatt joined #salt
19:55 zach # salt-key -L|wc -l
19:55 zach 80
19:55 zach Not too many I have eto login to manually to restart it :P
19:56 UtahDave Hm. that popped up in _ext_nodes.
19:56 zach I do like this bit now:
19:56 AviMarcus joined #salt
19:56 zach Total:     193
19:56 zach Summary bit
19:56 zach I like that
19:57 dwyerj what does the hash need to be for password in user.present ?
19:57 zach shadow hash
19:57 dwyerj thans
19:57 zach use SHA512 ;-)
19:57 _ikke_ dwyerj: you can use python's crypt library
19:57 viq jesusaurus: it wasn't empty pillar, though I have to make sure of it; immediate problem was non-ascii characters
19:57 dwyerj do i need to do hash=_the_hash ?
19:57 dwyerj like hash_type=the_hash
19:57 zach I wrote this nasty thing to generate hashes: https://pastee.org/yt5g8
19:58 dwyerj like with source_hash in files ?
19:58 zach random ones
19:58 StDiluted no one using salt to deploy a rails/unicorn app?
19:58 _ikke_ nope
19:58 UtahDave zach: are you using master tops?
19:58 zach UtahDave: yeah
19:58 dwyerj thanks for th perl, ill just rip out the random and drop in the rubbish passwords i have
19:59 UtahDave Hm. I wonder if that is a bug that we haven't caught yet.
19:59 zach Thankfully the servers salt-call every hour, I'll wait until 2:45 when it runs to see if it brings them back up
19:59 zach UtahDave: I seem to run into them all ;-)
19:59 UtahDave could you try disabling master tops?
19:59 zach # netstat -napt|grep :4505|wc -l
19:59 zach 39
19:59 zach ABout half of them are up now
20:00 zach Let me look real quick
20:00 zach interesting
20:00 zach did cmd.run uptime
20:00 zach most spit this out:
20:00 zach TypeError encountered executing cmd.run: __init__() got an unexpected keyword argument 'with_communicate'. See debug log for more info. Possibly a missing arguments issue: ArgSpec(args=['cmd', 'cwd', 'stdin', 'runas', 'shell', 'python_shell', 'env', 'clean_env', 'template', 'rstrip', 'umask', 'quiet', 'timeout', 'reset_system_locale'], varargs=None, keywords='kwargs', defaults=(None, None, None, '/bin/sh', True, (), False, None, True, None, False, None, Tr
20:02 * zach is cursed
20:04 srage joined #salt
20:08 zach UtahDave: weird thing is it seems to only be affecting a handful
20:08 UtahDave can you double check that those are running 0.17.1?
20:08 EntropyWorks slightly annoyed that it took me longer than I liked to fix a simple syntax error in my state. the dirfference between "- require: " and "- requires:"
20:08 zach UtahDave: yes, checking now
20:09 forrest EhtropyWorks, that seems to happen a lot
20:09 forrest have you by chance writte npuppet code?
20:09 forrest *written
20:10 EntropyWorks been awhile since I touched puppet. but yea I recall I had similar problems typing in something just slighlty wrong
20:11 EntropyWorks is there any syntax checker for states that anyone is aware of?
20:11 forrest EntropyWorks, there isn't one currently
20:11 zach what the hell - all sorts of versions running. 0.16.4-0.17.2
20:11 forrest jesusaurus worked a bit on this: https://github.com/jesusaurus/salt-shaker
20:11 srage_ joined #salt
20:11 forrest but it's basically just the test option :\
20:11 forrest yea zach, no 0.16.4 and 0.17 mixing :\
20:12 zach indeed - not sure how that happened...
20:13 jesusaurus yeah, salt-shaker basically just tries to run salt in masterless mode with test=true to see if the pillar and state compiles
20:13 jesusaurus it doesnt really do much yet
20:13 forrest yea I know. because you are slacking :P
20:13 jesusaurus i know...
20:14 jesusaurus if only i had time to work on it...
20:14 joehh forrest: probably already realised it, but 0.17.2-2 is in the ppa for ubuntu
20:14 forrest Do you have a family?
20:14 jcockhren jesusaurus: aaaah. You took my name idea. ;P
20:14 forrest jcockhren, just start committing on that one :P
20:14 EntropyWorks it wouldn't catch the "- requires:" which seems to just get ignored and
20:14 forrest joehh, yep  saw you bring it up (even though I don't use ubuntu)
20:15 shinylasers joined #salt
20:15 forrest jesusaurus, if you have a family, just get rid of them, same with pets
20:16 forrest more time to code, duh
20:16 [diecast] joined #salt
20:16 jesusaurus ha
20:16 dwyerj yeah i got a gf, she wants to go get coffee now
20:17 jslusher joined #salt
20:17 dwyerj terrible timing
20:17 teebes joined #salt
20:17 forrest I'm really just joking
20:17 jslusher I'm having a hell of a time getting a version of salt and salt-cloud to work together
20:17 forrest work life balance goooooooooo
20:17 forrest which versions jslusher?
20:17 forrest they pushed salt-cloud into the main salt branch with the latest release to try and help alleviate that
20:18 jslusher forrest: I know that 17.2 only works with 0.8.10
20:18 jslusher forrest: and that < 17.1 won't work with > 0.8.9
20:18 jslusher forrest: I'm on Fedora
20:18 jslusher 18
20:19 jslusher yum installs 17.1
20:19 jslusher and it has 0.8.8
20:19 jslusher which don't work together
20:19 NotreDev joined #salt
20:19 jslusher forrest: pip installs 0.8.10
20:20 jslusher even if version 0.8.9 is specified
20:20 forrest :\
20:20 [diecast] joined #salt
20:20 jslusher forrest: If I try installing salt using pip, it doesn't set the binaries for me
20:21 forrest that's lame
20:21 jslusher and I'm not sure how to set them
20:21 zach since I don't know exactly which IPs were running salt-minion before, I'm running some perl automation to 480 IPs heh
20:21 ovorp joined #salt
20:21 jslusher forrest: I know that they are sorting things out right now with these latest salt/salt-cloud version, but I just need something that works
20:22 forrest yea I understand
20:22 jslusher forrest: if you know where I could simply get a working salt-cloud 0.8.9 rpm that would hold me over
20:22 jslusher I have a salt and salt-master 16.4 rpm that I know works with 0.8.9
20:23 forrest well, there's this one
20:23 forrest http://rpmfind.net/linux/RPM/opensuse/factory/aarch64/noarch/salt-cloud-0.8.9-1.2.noarch.html
20:23 forrest but I don't know how much you trust rpmfind
20:23 pdayton joined #salt
20:23 jslusher forrest: I tried that, but now that you mention it, I think there was something else out of place
20:24 forrest ahh
20:24 forrest weird, 0.8.9 isn't in epel
20:25 jslatts joined #salt
20:25 forrest jslusher
20:25 forrest did you try this? https://pypi.python.org/packages/source/s/salt-cloud/salt-cloud-0.8.9.tar.gz
20:25 zach mix of .17.1 and .17.2 - what are the implications of running both?
20:25 forrest then manually run a pip install locally?
20:25 dave_den http://koji.fedoraproject.org/koji/packageinfo?packageID=15694
20:26 jslusher forrest: I think that will probably be the way to go.
20:26 dave_den oh, those are old
20:26 forrest yea that's what I was looking at dave_den
20:26 forrest was surprised the newer ones weren't in there
20:26 jslusher forrest: the rpm gives me a lot of dependency errors, so the local pip might be easier.
20:26 forrest yea it's worth a shot at the very least.
20:30 pipps joined #salt
20:30 jslusher forrest: that got my versions in line. thanks a ton!
20:32 forrest jslusher, yea np. Hopefully it will be easier going forward!
20:33 jslusher forrest: I really like salt/salt-cloud. I hope they can get past this compatibility mess of the last few months
20:34 forrest jslusher, now that it's integrated more directly into the main salt repo I think that will be the case.
20:35 Ryan_Lane is it possible to specify a returner for a state?
20:35 Ryan_Lane basically I want to call a module from a state and want the return to go to a returner
20:36 Ryan_Lane ah, seems it's possible: http://docs.saltstack.com/ref/states/all/salt.states.module.html
20:37 wilywonka joined #salt
20:39 georgebuth joined #salt
20:39 jslusher forrest: salt-cloud launches minions with version 17.1
20:39 forrest ?
20:39 jslusher I have version 16.4 of salt on the master
20:39 forrest oh
20:39 jslusher the last good version for 0.8.9
20:39 forrest are your minions also 0.16.4?
20:39 jslusher but it installs 17.1 on the minions
20:40 forrest yea that won't work
20:40 jslusher is there a way to specify what salt version gets installed?
20:40 forrest I don't know, I haven't played with salt-cloud very much
20:40 forrest I feel like there must be
20:41 zach OK. All of the ones I got back up to 17.2 are working, the ones on 17.1 are still broken. That's good news :)
20:41 zandy joined #salt
20:42 pipps_ joined #salt
20:42 jslusher forrest: either figure that out or I have to get a proper install of 17.2 to use with 0.8.10
20:43 forrest I'd say figure out how to get salt cloud to install the same version
20:43 bhosmer_ joined #salt
20:44 forrest someone around here probably knows jslusher, or you could always ask the mailing list
20:44 aloki joined #salt
20:45 Voziv How can I force salt to refresh the git fs?
20:47 BenBen joined #salt
20:48 forrest Voziv, did you already do salt-run fileserver.update?
20:48 Voziv forrest: Nope
20:49 zach man that was a fun fix
20:51 DredTiger Hey does anyone know of anything like GitHubs Boxen ( http://boxen.github.com/ ) in the saltstack world?
20:51 forrest well, don't mix versions so poorly next time zach, problem solved :P
20:52 zach forrest: it should have never happened :) package.latest was setup for salt and salt-minion
20:52 forrest DredTiger, what exactly is it that you want out of that?
20:52 forrest zach, hah makes sense
20:52 forrest DredTiger, just the fact that it's config management for your laptop?
20:52 Voziv Hmm, I get "Specified SLS salt.modules.mysql in environment base is not available on the salt master"   Aren't these on master by default?
20:52 zach At least I got to audit our network with whats online with this huge almost 600 internal IP list I generated
20:53 forrest if that's the case no, there is a brew install now available for saltstack however, and a few of the guys in here have wanted to hack on an automated 'laptop build'
20:53 forrest maybe boxen could work with salt?
20:53 DredTiger forrest: I can of course do salt stuff myself to set up my laptop as I like
20:53 forrest DredTiger, yea I gotcha
20:53 DredTiger Boxen just seemed to be a neat project
20:53 mgw joined #salt
20:53 forrest I agree, it seems very cool
20:53 DredTiger It uses puppet
20:53 forrest other than only supporting maco f course.
20:54 forrest yea I saw that, I'm curious as to what the salt integration would look like
20:54 jcockhren DredTiger: That's something forrest and I was talking about a few nights ago
20:54 jacksontj joined #salt
20:54 jcockhren called salt-shaker
20:54 jcockhren jesusaurus: ^^
20:54 forrest no no no
20:54 forrest salt-shaker is the state checker
20:54 forrest and why would you shake your laptop?
20:54 forrest makes no sense :P
20:54 jesusaurus ha
20:54 DredTiger I guess I'll clone the repo and look more closely at what it does
20:54 forrest fork boxen, make it salt-boxen, boom
20:55 jcockhren salt-shaker is like... you're spinkling your laptop with salt
20:55 forrest *watches jcockhren stretch to make this work*
20:55 jcockhren like saltify, but "saltify" is already taken
20:55 jcockhren b/c of salt-cloud
20:55 jcockhren lol
20:56 DredTiger haha
20:56 cewood joined #salt
20:56 dustyfresh joined #salt
20:56 forrest ugh
20:56 forrest boxen is written totally in ruby
20:56 DredTiger jcockhren: so you guys were just talking about it, no project as of yet right?
20:56 forrest DredTiger, no.
20:57 dave_den DredTiger: salt-formulas are similar to boxen in that boxen are just puppet configs
20:57 forrest Now that there's the brew version of salt though, it wouldn't be too bad
20:57 jcockhren nope. DredTiger that project is most likely to be an elaborate formula
20:57 dave_den https://github.com/saltstack-formulas
20:57 forrest just needs to be a way to easily bootstrap the box and install the formulas
20:57 Voziv How can I put the default salt modules so that master can access them?
20:58 dave_den Viziv: make your salt master a minion as well
20:59 Voziv dave_den: It is
20:59 [diecast] joined #salt
20:59 dave_den Voziv: i may not understand what you need to do - can you clarify/
21:00 Voziv dave_den: " Specified SLS salt.modules.mysql in environment base is not available on the salt master"
21:01 forrest DredTiger, the only thing that sucks is that macs only come with ruby by default, so it would be hard to keep it OS agnostic in that sense. If you take a look at Boxen and you think it would be easy to integrate, we can work on converting their puppet manifests for software to salt.
21:01 forrest I just don't have time to spin that up as yet ANOTHER project I should be working on :P
21:01 dave_den Voziv: that message means that you are references /srv/salt/base/salt/modules/mysql.sls  - that is probably not what you want
21:02 KyleG joined #salt
21:02 KyleG joined #salt
21:02 jslusher forrest: for the record, I got them to work together using a local install of both 17.2 and 0.8.10 with pip
21:02 forrest jslusher, lol
21:02 Voziv dave_den: my include:  is - salt.modules.mysql
21:02 dave_den forrest: eh? OSX includes python 2.7
21:02 forrest jslusher, did the one that comes with salt not work for you?
21:02 jslusher for whatever reason, the local install of 17.2 set the binaries
21:03 jslusher I actually didn't try it.
21:03 forrest dave_den, sorry I should have been more specific, it doesn't come with pip and stuff, it comes with brew right?
21:03 forrest jslusher, ahh ok
21:03 dave_den Voziv: - include means include another state file, not a module/
21:03 jslusher from 17.2 salt-cloud comes with it?
21:03 forrest it's merged into the main repo now at least, it might be a different RPM stil
21:03 dave_den forrest: nah, you have to install homebrew
21:03 forrest *still
21:03 forrest I haven't messed with 0.17.2 at all yet
21:03 Voziv dave_den: It works in masterless mode :/ How do I include the module in that case?
21:04 forrest jslusher, but it's now part of the main repo to try and avoid issues at the very least.
21:04 DredTiger sorry, someone came to my cube asking for help on a problem that happened at 0230 this morning
21:04 DredTiger It is now 1604
21:04 forrest lol
21:04 DredTiger "=/
21:04 dave_den Voziv: default modules don't need to be included. they will be loaded when the minion starts
21:04 DredTiger :-/
21:04 forrest 'dear sir, did you add these machines to the monitoring system?'
21:05 dave_den Voziv:  can you pastebin your statefile?
21:05 DredTiger I just got a new MBA at work so I'll take a closer look at Boxen and how to saltify it
21:06 teskew jslusher: salt-cloud has been merged into salt core, but it won't make a "stable" appearance until .18.0/Hydrogen release. redbeard2 and i are trying to get salt-cloud stable with the current stable salt core. there may be a bit of backporting from dev and another release of salt-cloud
21:06 mkoistinen Anyone know when the Mac Port of Salt will also get updated to 0.17.2?
21:06 forrest teskew, thanks for the thorough answer.
21:06 Voziv dave_den: http://hastebin.com/xiyepapiha.yaml
21:06 cdcalef joined #salt
21:06 jslusher teskew: yes. thank you. I have a working pair at the moment, and I'm looking forward to the merger. :)
21:07 DredTiger I hate having to use "homebrew" on MacOS X instead of a nice package manager like apt-get and deb files
21:07 dave_den Voziv: where's the include?
21:08 Voziv I removed it as per your statement, currently waiting on the minion to respond
21:08 DredTiger I ditched OSX for Ubuntu and now Mint on my current MBP. Don't want to start an OS flamewar or anything though
21:08 dave_den gotcha
21:08 * Voziv likes Mint
21:09 forrest I actually just ordered one of the new chromebooks
21:09 forrest gonna get a flash drive and put some sort of linux flavor on there.
21:09 DredTiger The Pixel? Those look sweet
21:09 forrest nah
21:09 forrest the acer c720
21:09 DredTiger The job wasn't ready to go for those
21:10 forrest Well, chrome os still has some wonky bits
21:10 forrest but overall, I like my first gen chromebook
21:10 DredTiger I got a groupon the other day for the previous version of the acer for like $129. Would've got one but my wife would have been like "why do you need another computer?"
21:10 forrest only thing I don't like on it is the performance, which the new one will solve.
21:10 forrest lol
21:10 matanya joined #salt
21:11 ncjohnsto joined #salt
21:11 forrest we'll see how the new chromebook is
21:12 forrest Has anyone run a linux distro off of one of those 'half size' thumb drives?
21:12 Voziv DredTiger: To be honest the pixel isn't really worth it
21:12 forrest usb flash drives, whatever you want to consider them known as?
21:12 hazzadous joined #salt
21:12 forrest Voziv do you have one?
21:12 Voziv Nope
21:13 DredTiger Voziv: Really? Not even with all those pixels?!?! LOL
21:13 Voziv I was thinking about buying, but at the pricepoint it's at, you can get a better functioning laptop
21:13 forrest I just don't see the point of a touch screen laptop
21:13 forrest I hate finger marks
21:13 forrest Voziv, have you looked at the acer c720?
21:13 thelorax123 joined #salt
21:13 Voziv nope, going to look it up tonight though
21:14 Voziv can salt open firewall ports?
21:14 DredTiger man Amazon is out of the 720
21:14 KyleG left #salt
21:14 forrest Yea mine just shipped today, was on backorder. I don't get what's with the one they have there though, pretty sure mine was 4 gigs of ram
21:14 forrest Voziv http://docs.saltstack.com/ref/states/all/salt.states.iptables.html
21:15 Voziv I was afraid of that
21:16 forrest DredTiger, interesting, so when I bought mine (http://www.amazon.com/Acer-Chromebook-11-6-Inch-Haswell-micro-architecture/dp/B00FNPD1OY)
21:16 forrest it was 4 gigs of ram and shipping straight from amazon
21:16 anitak joined #salt
21:16 forrest Looks like this is now the one amazon ships: http://www.amazon.com/Acer-Chromebook-11-6-Inch-Haswell-micro-architecture/dp/B00FNPD1VW
21:16 UtahDave Voziv: Yeah, salt can open up firewall ports
21:16 Voziv I have no idea how to configure iptables, (I use ufw)
21:17 DredTiger forrest, yeah that's the one I see with only 2GB of Ram
21:17 pentabular joined #salt
21:17 forrest DredTiger, yea, that first link I provided is the one I bought
21:18 forrest but it shipped straight from amazon, apparently they are only offering the 199 version now *shrug*
21:18 forrest I ordered just in time it seems
21:19 forrest DredTiger, just don't go with the samsung series 3 over the c720, that proc is much faster
21:19 forrest I'll have to try and dig up the comparison sheet in the numbers they tested.
21:19 teskew Voziv: the nomenclature and configuring of iptables is simple. you should learn it, anyway :)
21:20 DredTiger brb
21:20 Voziv I aggree teskew, however I need to get this up and running tonight :P
21:20 aleszoulek joined #salt
21:20 forrest UtahDave, is there any good workaround for ufw for firewall stuff?
21:21 Voziv Does this look about right? http://hastebin.com/xeqifofigi.sm
21:22 forrest should the connstate only accept new? Or should it support established as well?
21:22 Voziv What's the difference?
21:22 * Voziv googles
21:22 forrest and I assume the - file: nginx-vhost-default is somewhere in that file, or included?
21:23 Voziv I only pasted the first configuration block
21:23 Voziv where I've included the ip tables
21:23 forrest ok cool
21:23 forrest http://serverfault.com/questions/371316/iptables-difference-between-new-established-and-related-packets
21:25 Voziv Looks like I should include established
21:25 forrest for web traffic yea usually
21:25 Voziv Will salt flush this change if it detects that the chain wasn't present?
21:26 bhosmer joined #salt
21:26 forrest uhhh you know I'm not actually sure
21:26 Voziv well I'll test
21:26 forrest ok
21:27 ajw0100 joined #salt
21:30 aloki joined #salt
21:32 Voziv "State iptables.chain_present found in sls nginx is unavailable"
21:32 Voziv hmm
21:33 Voziv salt '*' iptables.version works though
21:34 dave_den Voziv: https://github.com/saltstack/salt/blob/v0.17.2/salt/states/iptables.py
21:34 dave_den unfortunately chain_present is only in teh develop branch and the docs are built against develop
21:35 Voziv but append will only append if it doesnt exist?
21:35 dave_den so if you are running 0.17.x you do not have that state function
21:35 Voziv the append function that is
21:35 dave_den it should, yes.
21:37 matanya joined #salt
21:38 ajw0100 joined #salt
21:41 zandy joined #salt
21:41 Chrisje joined #salt
21:44 isomorphic joined #salt
21:45 matanya joined #salt
21:45 ovorp joined #salt
21:45 Voziv It seems like iptables is very broken at the moment
21:47 Voziv http://hastebin.com/nasomohafo.txt
21:48 forrest Voziv, you're using chain_present
21:48 Voziv forrest: this is append
21:48 forrest oh you changed it?
21:48 forrest from this: http://hastebin.com/xeqifofigi.sm
21:48 Voziv yeah, all I care about is getting port 80 open :P
21:49 fllr joined #salt
21:49 Voziv but as you can see, it's trying to pass --__env__ base and --__sls__ nginx  in the command
21:49 Voziv which of course is going to break
21:49 shinylasers joined #salt
21:49 forrest did you do - connstate: NEW,ESTABLISHED?
21:49 pipps joined #salt
21:49 Voziv forrest: http://hastebin.com/bosovuxeso.sm
21:50 Voziv thats the whole file now
21:50 forrest can you try to only do one option in the connstate?
21:50 forrest none of the examples use 2
21:51 KyleG joined #salt
21:51 KyleG joined #salt
21:51 matanya joined #salt
21:51 Voziv forrest: If you look at http://hastebin.com/bosovuxeso.sm  you'll see that it gets passed properly
21:52 forrest yea I can see that it does
21:52 forrest but it's also adding in a bunch of stuff
21:52 Voziv but those stupid --__env__
21:52 forrest yea
21:52 Voziv I'll just enable the firewall manually for now
21:53 Voziv yay my webserver says "File not found" (which is great)
21:54 forrest heh, it's up at least
21:54 Voziv exactly
21:54 Voziv now I have to put my application on it. Going for a manual process on this one for now
21:54 forrest psssssssssh
21:54 forrest salt, salt, salt
21:54 Voziv someday
21:55 pipps_ joined #salt
21:55 fllr Hey guys. I've heard in a lot of places, you can use pillars to keep all sensitive information, stuff like db passwords, and such, away from repositories. What I don't understand is, how is that possible? Certainly I also have to commit those pillar files, no?
21:56 Voziv fllr: Heh. You can do a few things
21:56 Voziv 1. Separate private repository where only people who are allowed to see credentials can access it
21:56 Voziv or 2. Maintain the files outside of git
21:57 Voziv The idea is you version your setup but keep the passwords and such away from the repo so that you can share it (be it with coworkers or anyone)
21:58 fllr Voziv: Aaaah... makes sense...
21:59 bhosmer joined #salt
22:01 harobed_ joined #salt
22:02 harobed_ joined #salt
22:02 ovorp1 joined #salt
22:03 aleszoulek joined #salt
22:07 [diecast] joined #salt
22:09 Marion_ joined #salt
22:11 NotreDev joined #salt
22:13 Katafalkas joined #salt
22:16 jimallman joined #salt
22:21 koblas joined #salt
22:22 koblas doing a salt bootstrap and ending up with two packages in /etc/apt/sources.list.d/  — salt.list and saltstack-salt-precise.list — which is causing apt-get update to barf….
22:22 wkf joined #salt
22:23 koblas oh - wait - might have found a bad state file...
22:24 Marion_ koblas: oh my, someone else having troubles with apt-get update
22:25 koblas I think I found that one of my older sls files was "making sure" that there was a correct salt.list file…
22:29 elfixit joined #salt
22:29 thelorax123 joined #salt
22:29 HeadAIX joined #salt
22:31 sroegner_ joined #salt
22:32 alunduil joined #salt
22:32 StDiluted joined #salt
22:39 NV joined #salt
22:41 zandy joined #salt
22:43 Corey joined #salt
22:49 snave joined #salt
22:51 Ryan_Lane so, I have a deployment system that's using salt: https://github.com/trebuchet-deploy/trebuchet
22:52 pentabular joined #salt
22:52 Ryan_Lane I wanted to know how to make that easily includable by people using salt
22:52 Corey Yay!
22:52 Corey Build a formula!
22:52 Ryan_Lane formula?
22:52 Corey Ryan_Lane: https://github.com/saltstack-formulas
22:53 Ryan_Lane well, this needs to install stuff mostly on the master
22:53 Ryan_Lane it adds returners, modules, runners, files, etc, etc
22:53 Ryan_Lane and then the master needs to push those out to the minions
22:54 Ryan_Lane basically I'd like to include the whole set of code as is and have the master immediately know how to reference all of it
22:54 ajw0100 joined #salt
22:55 Ryan_Lane basically something that would add the modules to the modules path, the runners to the runners path, etc, etc.
22:55 Ryan_Lane puppet has modules for this...
22:55 Ryan_Lane it seems in salt you need to manually do all of this
22:56 pentabular joined #salt
22:56 Ryan_Lane yeah, definitely seems formulas won't help me here
22:57 jsm joined #salt
22:59 Ryan_Lane I kind of need something like a master formula where you can something in and it'll register a module in the module path, a runner in the runner path, a returner in the returner path, etc
23:02 dave_den you could create a formula that does that
23:02 dave_den formula for a formula
23:02 * dave_den eyes go crossed
23:02 Ryan_Lane except how would I modify the runner path and such?
23:03 Ryan_Lane or is that info available via grains/pillars?
23:03 Ryan_Lane even then, though, if there's multiple locations in the path, you'd have no clue where to put it
23:04 jacksontj joined #salt
23:05 Ryan_Lane is there a way to extend configuration yaml?
23:06 dave_den it would be fairly straightforward to do it with a pydsl state
23:06 Ryan_Lane extending config yaml?
23:07 Ryan_Lane you can use pydsl state for the config?
23:07 dave_den i'm talking about a state file that uses pydsl to configure your config
23:08 Ryan_Lane well, basically I'm looking for a way to modify the file roots, pillar roots, state roots, etc.
23:08 dave_den right.
23:08 dave_den you cn manage your master as a normal minion
23:09 Ryan_Lane yes, but that requires everyone that uses my deployment system to also use my salt master state?
23:10 dave_den you have to make changes on your master somehow
23:10 Ryan_Lane yeah :)
23:10 fllr joined #salt
23:10 Ryan_Lane what I want is a drop-in mechanism for this
23:11 Ryan_Lane so that someone can drop my code in, include a single file and now the states, returners, runners, etc, were then available
23:11 dave_den how much more drop-in could it be than 'salt "mymaster" state.sls configure_trebuchet' ?
23:11 Ryan_Lane how would configure_trebuchet actually work, though?
23:12 dave_den it would do whatever you need it to do to install trebuchet like you already do by hand
23:12 dave_den just like writing any other state config
23:12 Ryan_Lane there's no way for me to know how to configure runner dirs, file dirs, etc
23:12 dave_den why not?
23:13 Ryan_Lane if the master has base: /srv/runners, /srv/runners2, where do I put the file?
23:13 dave_den you can even use the py rendered to have a full python env for the state: http://docs.saltstack.com/ref/renderers/all/salt.renderers.py.html
23:13 blee_ joined #salt
23:13 Ryan_Lane this is more than just a state
23:13 ovorp1 joined #salt
23:13 dave_den why don't you extend the runners dir list to include your trebuchet runners dir?
23:13 Ryan_Lane that's what I want to do :)
23:14 Ryan_Lane I can do that from within a state?
23:14 dave_den you can easily do that
23:14 dave_den yes :)
23:14 Ryan_Lane same with modules, etc?
23:15 Ryan_Lane interesting
23:16 cnelsonsic joined #salt
23:16 JasonSwindle joined #salt
23:18 jimallman joined #salt
23:18 Ryan_Lane let me start an etherpad to see if what I'm thinking is the right approach
23:19 Ryan_Lane etherpad from a plane is really rough
23:20 forrest Ryan_Lane, you still have internet in the sky
23:20 forrest except paying for internet in the air, which is dumb
23:20 Ryan_Lane oh, not a complaint, it's just funny. finally something worse than SSH on a plane
23:20 forrest lol
23:20 jacksontj joined #salt
23:20 forrest and it only cost 10 dollars!
23:20 Corey Ryan_Lane: Terrorists.
23:20 Corey And I *refuse* to support in flight wifi. That's my refuge!
23:21 Ryan_Lane heh. it was $26
23:21 Ryan_Lane for the whole flight
23:21 forrest holy crap
23:21 forrest how long is the flight?
23:21 Corey It's the glorious time when the other world can screw right off.
23:21 Ryan_Lane 5 hours
23:21 forrest ehhh I guess that isn't too bad then
23:21 Corey Ryan_Lane: Where to where?
23:21 Ryan_Lane Corey: it's such a quick way to pass the time, though
23:21 Ryan_Lane SFO -> JFK
23:21 Corey Ryan_Lane: Friday I'm PDX -> SEA -> SFO -> LAX.
23:22 Ryan_Lane ouch
23:22 Corey Ryan_Lane: Wait, is that United?
23:22 Ryan_Lane delta
23:22 Corey Bah.
23:22 forrest Corey, were there no straight throughs??
23:22 Corey Greyhound of the skies.
23:22 forrest lol
23:22 Corey forrest: I'm just a *hair* shy of my goal, so I'm taking interesting routings to hit it.
23:22 jesusaurus haha
23:22 Corey That doubles the mileage I fly for the trip.
23:22 forrest Corey for what?
23:22 forrest ahh ok
23:22 Corey Top tier United status for the year.
23:23 forrest honestly I like Delta
23:23 forrest flew to Chicago last a while back, and it was the first time my ass hasn't fallen asleep on an airplane in probably 5 years
23:23 forrest the joys of being tall on an airplane
23:23 forrest *no last, can't type apparently
23:24 Corey forrest: Yeah, they're actually all fairly similar, but I have status on United. :-
23:24 forrest lol
23:24 forrest hey man, unless you're flying first class everywhere :\
23:24 Corey forrest: Uh... pretty much, yeah. :-)
23:24 Corey Upwards of 50% of the time anyway.
23:25 forrest let's see...
23:25 Ryan_Lane dave_den: https://etherpad.wikimedia.org/p/trebuchet-formula
23:25 Ryan_Lane ?
23:25 Ryan_Lane would something like that work?
23:26 forrest ok, ~250 for normal from PHX to salt lake in january, or I could pay 500 for first class
23:27 forrest Let me just dive into my uncle scrooge money vault over here.. :P
23:28 Ryan_Lane if that etherpad is possible it would be awesome because it would mean that environments just work
23:28 Corey forrest: I don't think you're getting the concept here. :-)
23:28 Corey forrest: I pay the cheap rate. The upgrade is a perk of status.
23:28 forrest Corey, probably not, I almost always fly on my own dime
23:28 forrest how much do you pay for a first class ticket usually?
23:28 Corey That's the point, I never pay for a first class ticket.
23:28 forrest ahh
23:29 Corey I pay the ~250 or whatever it is, and they shove me into the $1300 (or whatever nutty markup the first class is) seat.
23:29 forrest how many flights are you doing a year?
23:29 Corey It's November, and this year I'm at... 84 segments.
23:29 Katafalkas joined #salt
23:30 forrest a segment = a flight in one direction?
23:30 Corey "A lot."
23:30 kermit joined #salt
23:30 Corey forrest: It's a flight, period. If I fly from LA to Chicago to NY, that's two segments.
23:30 forrest ahh ok
23:30 Corey But I fly 100K miles a year, or will by the time December 31 hits.
23:30 bemehow joined #salt
23:30 forrest heh
23:30 forrest that's a lot
23:31 Corey Ever see Up in the Air?
23:31 forrest yes
23:31 Corey I thrive on it. :-)
23:31 Corey <-- traveling consultant
23:31 Ryan_Lane I'm pretty tired of travelling
23:31 Corey Well I bank the points, and then blow them on Mrs. Quinn.
23:31 forrest lol
23:31 Corey Next week we're doing a week and a half in London.
23:32 Ryan_Lane hm, I can probably assume that the state path is also a file path, eh?
23:32 forrest and on that note, I'm going home for the day.
23:32 Ryan_Lane I guess that's a bad assumption
23:32 forrest later guys
23:32 Corey Good call. :-)
23:32 Corey Bah.
23:33 Ryan_Lane it would really be nice to have a standard way of extending the paths from states
23:34 Ryan_Lane I guess that's just asking to be abused, though
23:39 micah_chatt joined #salt
23:41 zandy joined #salt
23:41 StDiluted joined #salt
23:42 NV joined #salt
23:48 shinylasers joined #salt

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