Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-01-21

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

All times shown according to UTC.

Time Nick Message
00:01 markm_ joined #salt
00:05 scott_w joined #salt
00:07 strgcloud joined #salt
00:07 bhosmer joined #salt
00:13 mugsie joined #salt
00:15 diegows joined #salt
00:18 falican joined #salt
00:24 Gifflen joined #salt
00:26 Gifflen joined #salt
00:27 tomtomtomtom joined #salt
00:32 hunter joined #salt
00:37 yomilk joined #salt
00:37 gnu-man joined #salt
00:37 gnu-man_ joined #salt
00:39 kalloc joined #salt
00:43 timoguin_ joined #salt
00:56 hunter joined #salt
00:56 Ryan_Lane2 joined #salt
01:05 strgcloud joined #salt
01:05 cachedout joined #salt
01:06 bhosmer joined #salt
01:17 gnu-man joined #salt
01:19 kalloc joined #salt
01:20 srage joined #salt
01:21 ConceitedCode joined #salt
01:24 gnu-man_ joined #salt
01:26 xzarth joined #salt
01:30 neilf_ Can anyone weigh into the salt timeout issue https://groups.google.com/forum/#!msg/salt-users/w5xAm686w8M/Cmn5exWpDSkJ
01:39 kalloc joined #salt
01:42 test__ joined #salt
01:47 toguin__ joined #salt
01:50 yomilk joined #salt
01:53 mgw joined #salt
01:54 taion809 joined #salt
02:00 bhosmer joined #salt
02:07 Gifflen joined #salt
02:11 kmrhb joined #salt
02:21 vlcn can anyone point me in the right direction with working with a jinja dict object?
02:25 n8n_ joined #salt
02:31 n8n joined #salt
02:33 Dinde joined #salt
02:33 higgs001 joined #salt
02:36 higgs001 joined #salt
02:38 n8n joined #salt
02:39 kalloc joined #salt
02:40 Dinde joined #salt
02:40 higgs001 joined #salt
02:45 Linz joined #salt
02:52 ajw0100 joined #salt
02:52 scuwolf joined #salt
02:52 DaveQB joined #salt
03:00 n8n joined #salt
03:04 yaru22 joined #salt
03:04 srage joined #salt
03:10 Linz joined #salt
03:21 ajw0100 joined #salt
03:39 kalloc joined #salt
03:43 hunter joined #salt
03:45 possibilities joined #salt
03:46 possibilities joined #salt
03:52 Furao joined #salt
04:04 jalbretsen joined #salt
04:14 SpeeR joined #salt
04:16 n8n joined #salt
04:20 hao joined #salt
04:21 hao hello anybody here?
04:35 middleman_ joined #salt
04:36 d10n joined #salt
04:39 kalloc joined #salt
04:39 Ryan_Lane joined #salt
04:41 forresta what's up hao
04:45 yomilk joined #salt
04:45 hao I have question about salt-cp
04:45 hao can you help ?
04:49 forresta maybe, what's the question?
04:51 hao I have some AIX box can not installed minion
04:51 hao so I use the salt-ssh to management it
04:52 hao I found salt-cp seems can not work if no minion installed
04:52 hao does any other easy way I can copy file from central point (master ) to endpoint?
04:54 forresta if salt-cp isn't working, your best bet might be to use cmd.run to do an rsync/wget/scp
04:55 hao ok
04:55 hao could you give some example to use "scp" ?
04:55 hao run it in raw shell ?
04:56 forresta yea you'd use cmd.run in your state: http://docs.saltstack.com/ref/states/all/salt.states.cmd.html
04:57 forresta You'd have to look at authenticating back through though via ssh keys or something so you don't have to enter a password
04:57 forresta so it would have to be like scp user@myhost:/origin/file/path/file.asdf user@myminion:/location/to/put/file.
04:58 hao ok. thanks a lot.
04:58 forresta since it's calling it from the minion, you could always do scp user@myhost:/origin/file/path/file /location/to/put/file
04:58 forresta just depends.
04:58 forresta yea np
04:58 hao sorry , I have no minion
04:58 hunter joined #salt
04:58 hao does salt-ssh can support it  ?
04:59 forresta cmd.run?
04:59 hao yes
04:59 hao my target is AIX, I don't find the way to install it
04:59 forresta Unless something changed recently, yes.
04:59 forresta you're writing states
04:59 forresta right?
05:00 forresta You could always try to use http://docs.saltstack.com/ref/modules/all/salt.modules.state.html to call the cp module, but I have no idea if that would work.
05:00 hao Yes, I try to write some state
05:00 jimallman joined #salt
05:00 forresta ok
05:01 hao ok. I will read the doc first to get better understand  and come back here
05:01 malinoff joined #salt
05:01 hao thanks a lot
05:01 forresta Ok, your best bet is to just give it a shot!
05:01 forresta np
05:02 hao BTW, do you have suggestion for AIX management ?
05:02 hao from offical support list, I did not find AIX
05:03 forresta Yea there isn't any support as far as I know, your best bet is salt-ssh, but a bunch of stuff still might no work over there. I didn't even know AIX was used an longer.
05:04 hao ok
05:05 ajw0100 joined #salt
05:10 ndrei joined #salt
05:18 cachedout joined #salt
05:33 higgs001 joined #salt
05:34 kermit joined #salt
05:36 Linz joined #salt
05:36 sroegner joined #salt
05:39 kalloc joined #salt
05:52 strgcloud joined #salt
05:54 hunter joined #salt
06:07 taion809_ joined #salt
06:09 Psyke joined #salt
06:36 bhosmer joined #salt
06:38 strgcloud joined #salt
06:42 anuvrat joined #salt
06:43 Hipikat joined #salt
06:52 porri joined #salt
07:03 luff joined #salt
07:03 luff left #salt
07:07 leonardinius joined #salt
07:08 [vaelen] joined #salt
07:08 taion809 joined #salt
07:09 taion809 joined #salt
07:10 esogas_ joined #salt
07:24 higgs001 joined #salt
07:25 packeteer joined #salt
07:25 sroegner joined #salt
07:27 ckao joined #salt
07:28 linjan joined #salt
07:31 slav0nic joined #salt
07:35 matanya joined #salt
07:37 gammalget joined #salt
07:43 porri joined #salt
07:46 flebel joined #salt
07:53 leonardinius_ joined #salt
07:55 leonardinius_ joined #salt
07:56 leonardinius_ joined #salt
07:59 Psyke Morning, leonardinius. Did it work with the bootstrap?
08:03 balboah joined #salt
08:05 Furao joined #salt
08:09 harobed_ joined #salt
08:29 giantlock_ joined #salt
08:34 pershyn joined #salt
08:49 yomilk joined #salt
08:51 Hipikat hi… if i've got a state that ensures config files exist in a conf.d/ directory and i want to 'watch' those files and run something from an execution module (e.g. salt.modules.supervisord.reread) when the files change -- how would i do that?
08:54 MaikDerstappen joined #salt
08:54 Furao Hipikat: use state module.wait
08:55 Hipikat Furao: aah, thank you very much
08:57 luff joined #salt
08:58 luff left #salt
09:01 helminthe joined #salt
09:09 gurubert1 joined #salt
09:09 gurubert1 left #salt
09:12 zooz joined #salt
09:14 sroegner joined #salt
09:21 Psyke Is it possible to check if a grain exists? E.g. In a state file, check if the minion has got an interface "bond0" if it does use the IP in the template. If it does not exist fall back to a default interface "eth0" IP
09:21 malinoff Psyke, {% if grains.get('interface:bond0') %} #iface exist {% endif %}
09:22 gmtech_ Hey - can anyone point me to documentation about hiera ext_pillar . All seems very sketchy
09:22 malinoff Psyke, {% if 'bond' in grains.get('interface', []) %} #iface exist {% endif %} - something like that
09:23 Psyke Thanks malinoff, I will give that a try quickly.
09:27 n8n joined #salt
09:28 gmtech_ Is there a grain which represents which "environment" you're in?
09:29 leonardinius_ Psyke, morning. What do you mena by bootstrap - salt-ssh?
09:29 leonardinius_ *mean
09:29 Psyke Malinoff, thanks it works like a charm. My Jinja skills suck :P
09:30 Psyke leonardinius, the issue you were having with yesterday with debconf. Is it the git tag that is broken?
09:30 malinoff Psyke, http://jinja.pocoo.org/docs/ & http://docs.saltstack.com/ref/states/vars.html
09:31 Psyke Malinoff, thanks I will give that a read through.
09:33 leonardinius_ Psyke, I've retested it with latest develop branch. Actually seems like both v0.17.4 and altest develop behaves the same https://github.com/saltstack/salt/issues/9849#issuecomment-32790198
09:34 leonardinius_ Psyke, it seeems more likely salt-ssh somehow is broken in this regard
09:34 ravibhure joined #salt
09:35 leonardinius_ Psyke, I might we wrong in my consclusions though (complete rookie, might misinterpet the evidence)
09:36 Psyke That is a strange issue. I am also just starting with salt, so don't trust me either :P
09:36 sroegner joined #salt
09:46 kalloc joined #salt
09:47 Furao joined #salt
09:55 leonardinius_ Hi All, If I want to do conditional salt state execution (e.g. ensure my xx.yy.sls is included/executed ONLY once from xx.init.sls) - how could I achieve that?
09:58 leonardinius_ some sort of {% if salt['cmd.run']('ls /tmp/mymarker') %}?
09:59 higgs001 joined #salt
10:00 leonardinius_ or is it some kind of test flag?
10:00 leonardinius_ e.g. file.managed .. test: ls ..
10:00 leonardinius_ e.g. file.managed .. test: ls ..
10:02 giantlock_ leonardinius, take a look about replace: false
10:06 leonardinius_ giantolock_, does it work only for file.manages or sls as well?
10:06 giantlock_ leonardinius_, only for file.manage
10:07 leonardinius_ giantlock_ the use case i'm trying to accomplish - is to have patches states, e.g. patches/init.sls to include named patches and to execute them only once via some kind of pillar/grain/marker whatever
10:08 leonardinius_ giantlock_, do you know something what might work in this scenario?
10:09 malinoff leonardinius_, Patch is something atomic - you should not come to a bad state via 2 or more repetitions. Otherwise, this is not a patch
10:10 leonardinius_ malinoff, so what is your recomendation? e.g. I need to drop databas and re-import it from backup as a sort of patch/initial import
10:12 giantlock_ leonardinius_, and what about condition for drop-restore?
10:12 malinoff leonardinius_, I'd create 'initialize.sls' with all these steps and never run it via highstate or whatever. This action, AFAIK, should be done only once - and I can do that manually via state.sls
10:14 leonardinius_ malinoff, ok. I'm just thinking about the states as a single point of truth of what is has been configured .. Meaning - highstate should always put instance is predefined initial state
10:15 leonardinius_ malinoff, that way `initialise.sls` is always preset during first run at least
10:15 malinoff leonardinius_, I ended up with no use of highstate ever. Too unpredictable, too difficult to customize.
10:16 leonardinius_ malinoff, what do you recommend than?
10:16 malinoff I have Jenkins ecosystem, and I can vary my builds with the help of it
10:16 scott_w joined #salt
10:16 malinoff So for each project I have prepare.sls, deploy.sls, configure.sls, backup.sls and rollback.sls state files
10:16 leonardinius_ so, what is the idiomatic / recomended pattern of salt usage then?
10:16 malinoff I don't know :)
10:16 malinoff That's how I use it
10:17 leonardinius_ how do you get `single point of truth` then?
10:17 leonardinius_ by running show_highstate?
10:18 malinoff leonardinius_, Do you mean, how do I understand that the next run will be successful?
10:18 malinoff Or what?
10:18 leonardinius_ what's the point of not having state system/highstate other than not having to write ssh scripts?
10:18 malinoff I use state system, but without highstate
10:18 malinoff Only with the help of salt \* state.sls
10:19 malinoff For each project I build rpm/deb packages, so 'deploy.sls' state looks like 'update this package from that repo'
10:20 leonardinius_ malinoff, I see. What I was trying to ask - how do you get the overview/ bird's eye view on the whole inf-re universe?
10:20 leonardinius_ meaning, as far i can understand you have partial-state state1.sls state2.sls state3.sls files with different puprose
10:21 leonardinius_ how do you get the `whole picture`
10:21 leonardinius_ how do you get the `whole picture`?
10:21 malinoff I have 'prepare.sls' state file which installs & configures everything the package do not. So I have two components: contents of 'prepare.sls' and contents of deb/rpm package of the project
10:22 Jahkeup joined #salt
10:22 malinoff I'm not saying highstate is useless. But it can't fulfill my needs
10:22 malinoff Unfortunately :(
10:23 leonardinius_ malinoff, I'm not saying highstate is a must either ;)
10:23 leonardinius_ i'm just trying to get a grasp of salt usage patterns
10:24 malinoff leonardinius_, Just try several approaches and chose the one that feels good for you. 4 months ago I was using salt environments, but now I dropped them all.
10:24 malinoff And stuck with default 'base'
10:25 leonardinius_ me myself - i'm in salt `proof of concept` phase and the use case i'm validating salt for - is creating and maintainig `single point of truth` image, so i'am being able to answer what state the system is in at current moment, reivew componets etc ...
10:25 leonardinius_ malinoff, i'm just wondering if the pattern you are describing fits my needs or not
10:26 leonardinius_ whether you are address those questions yourself or you are using salt more like bash/deployement scipts
10:27 leonardinius_ malinoff, hmm. regarding salt environments. what was the cause for dropping them?
10:28 leonardinius_ malinoff, too cumbersome or too `magic` / non -transparent?
10:28 malinoff leonardinius_, They are useless :)
10:28 leonardinius_ malinoff, how you address the need for multiple environments then? minnor specific pillars?
10:29 leonardinius_ *minion
10:30 david_a joined #salt
10:31 malinoff This is my /srv/salt tree: http://pastebin.com/grnWPvtY
10:31 malinoff I have 'project' and 'environment' grains on each minion
10:31 malinoff And I wrote a simple salt module
10:31 Psyke It is pretty interesting to see how people use salt. I personally use highstate. I have a multitude of base services which I include and extend in more complex services.
10:33 leonardinius_ Psyke, do you have a need to rollout some particular service in initial (read database import from backup) state?
10:33 leonardinius_ Psyke, if yes - how do you address that?
10:34 gmtech_ Do external pillars need configuring in master or minion configuration? (or both)
10:34 leonardinius_ malinoff, do you really mean grains or pillars? I see pilars in your pastebin
10:34 leonardinius_ malinoff, or is it combination of both?
10:35 malinoff gmtech_, pillars is a master concept
10:35 Psyke Leonardinius, not at the moment. But I am currently looking at it. Since my OpenSIPs state requires that I import a DB initialy and then not again until I specify it.
10:35 malinoff leonardinius_, I use both.
10:35 Psyke *not import again
10:36 gmtech_ Thanks @malinoff - I configured master with external pillars but pillar data not available on minion. Would I have to issue a refresh_pillar too ?
10:36 malinoff gmtech_, each time you change a pillar, you should run refresh_pillar
10:36 leonardinius_ Psyke, i'm highly interested on how it works out for you. it's something critical for me as well.
10:37 Psyke Leonardinius, If I find something I will let you know. I am currently googling the problem :P
10:38 leonardinius_ malinoff, so you use grain matchers to split boxes into environments (e.g. either by host name or ip range) and then you provide environemnt specific configuration via pillars, right?
10:38 malinoff leonardinius_, yes :)
10:39 malinoff In my setup, the only difference between environments is 'configure.sls' step - which uses environment-specific pillars
10:39 leonardinius_ malinoff, one more stupid and yet annoying question. the custom module you mentioned - what id does exactly?
10:40 Jahkeup joined #salt
10:40 leonardinius_ so many typos (facepalm)
10:42 malinoff leonardinius_, "salt -C 'G@project:project-id and G@environment:dev' project.deploy" simply makes "salt -C 'G@project:project-id and G@environment:dev' state.sls project-id.deploy" call
10:43 malinoff I parse target string and use the data provided from it to make all needed paths
10:44 leonardinius_ malinoff, (y) thanx a lot for sharing your wisdom with us
10:45 malinoff leonardinius_, my pleasure :)
10:52 Jahkeup joined #salt
10:56 scott_w joined #salt
10:59 bhosmer joined #salt
11:01 sls joined #salt
11:03 sls how to consolidate the different set of commands to run into single package to run using salt
11:03 sls how to consolidate the different set of commands into single package to run on salt minion using salt
11:05 MohShami joined #salt
11:06 MohShami Hey guys, trying to recursively set the permissions of a folder, but if the folder grows too big it fails, is there something I'm missing? http://pastebin.ca/2567675
11:07 MohShami I'm using the same thing for a smaller folder and it works without an issue, it was a copy and paste
11:07 malinoff MohShami, probably, salt reaches python's recursion limit
11:07 ravibhure joined #salt
11:08 MohShami I didn't create any more folders
11:08 MohShami only more files in the folders
11:08 aleszoulek joined #salt
11:09 scott_w joined #salt
11:12 MohShami malinoff: this is the error I'm getting Failed to change mode to 440; Failed to change mode to 440
11:12 MohShami running chmod/chown using cmd.wait works
11:12 Iwirada joined #salt
11:13 malinoff MohShami, use salt-call locally with -l debug option
11:13 MohShami already have, this is the only error I'm getting "[ERROR   ] {'mode': '440'}"
11:14 malinoff -l garbage
11:14 MohShami 1 min
11:16 MohShami malinoff: same thing
11:16 MohShami http://pastebin.ca/2567740
11:17 malinoff MohShami, Don't know what is this
11:17 malinoff Open an issue
11:17 SkyRocknRoll joined #salt
11:17 SkyRocknRoll joined #salt
11:18 MohShami malinoff: will do, thanks mate
11:24 scott_w joined #salt
11:24 diegows joined #salt
11:39 jcsp1 joined #salt
11:48 scott_w joined #salt
11:49 Whissi joined #salt
11:59 yomilk joined #salt
12:00 MohShami I figured out what was wrong, the folder had 2 symbolic links pointing to nonexisting folders, removing those fixed the issue
12:00 MohShami how can I open an issue?
12:01 bhosmer_ joined #salt
12:05 ravibhure joined #salt
12:14 bhosmer joined #salt
12:14 malinoff https://github.com/saltstack/salt/issues?state=open "New issue" button
12:15 scott_w joined #salt
12:17 JasonSwindle joined #salt
12:21 bhosmer joined #salt
12:26 bhosmer joined #salt
12:26 ndrei joined #salt
12:31 scristian joined #salt
12:31 gmtech_ is there a salt command for looking up the value of a single pillar item ?
12:32 malinoff gmtech_, http://docs.saltstack.com/search.html?q=pillar.get&check_keywords=yes&area=default
12:36 jrdx joined #salt
12:41 MohShami thanks malinoff
12:42 MohShami is there a way for salt to pull data from mysql? I remember reading about it somewhere but I can't find any examples
12:42 bhosmer_ joined #salt
12:43 malinoff http://docs.saltstack.com/search.html?q=mysql
12:43 bob2 joined #salt
12:49 taion809 joined #salt
12:50 taion809 joined #salt
12:51 taion809 joined #salt
12:52 taion809 joined #salt
12:53 taion809 joined #salt
12:54 elithrar joined #salt
12:54 taion809 joined #salt
12:58 qkaiser joined #salt
12:59 qkaiser Hi all, I have a simple question. I get this error 'Module function yumpkg.group_install is not available' when using module.run in my sls. Do I have to actually install yumpkg ? If so, how ?
13:00 malinoff qkaiser, Are you sure you run your state on yum-based OSes?
13:00 qkaiser malinoff: absolutely, Scientific Linux 6.5
13:01 malinoff Uhm. I'm sure you should use 'pkg.group_install'
13:02 malinoff And this function is present only in Hydrogen
13:08 vejdmn joined #salt
13:09 ravibhure1 joined #salt
13:10 MohShami malinoff, thanks mate, I'll check that
13:10 jslatts joined #salt
13:11 bbinet hi all
13:11 bbinet here is a question about writing custom modules
13:12 bbinet my custom module needs some specific python dependencies that can be installed in a virtualenv on every minion
13:12 bbinet but how can I extend salt sys.path so that salt can find these module dependencies?
13:13 bbinet (I prefer not to install those globally with pip)
13:17 qkaiser malinoff: changed to pkg.group_install and it works. thanks
13:25 scott_w joined #salt
13:30 Linz joined #salt
13:30 harobed joined #salt
13:34 gmtech_ Anyone know what I'm missing with ext_pillars here?   http://pastebin.com/E4aY0XsL
13:44 gmtech_ mm i just can't get it to work
13:46 jcsp joined #salt
13:47 bhosmer joined #salt
13:52 gmtech_ Weirdly, hiera.py doesn't even have a pyc file meaning it doesn't even get loaded/compiled by the python interpreter.
13:52 gmtech_ Are external pillars broken in salt or something?
13:53 juicer2 joined #salt
14:02 anuvrat joined #salt
14:02 brianhicks joined #salt
14:02 bhosmer joined #salt
14:05 renoirb Hi salters!
14:06 sroegner_ joined #salt
14:11 scott_w joined #salt
14:17 bhosmer joined #salt
14:17 honestly_ joined #salt
14:17 yaru22 joined #salt
14:18 gadams999 joined #salt
14:20 cachedout_home joined #salt
14:21 racooper joined #salt
14:23 honestly joined #salt
14:24 honestly joined #salt
14:24 jslatts joined #salt
14:24 Gifflen joined #salt
14:24 mpanetta joined #salt
14:25 honestly joined #salt
14:30 MohShami Hi renoirb
14:30 renoirb Hi MohShami
14:30 renoirb What you guys are using to monitor systems?
14:31 MohShami I'm currently using nagios
14:31 renoirb I just heard of Sensu+Graphite instead of using nagios
14:31 renoirb and sensu accepts nagios plugins
14:31 renoirb Are you using halite MohShami ?
14:31 MohShami renoirb: nope, sorry :)
14:32 honestly joined #salt
14:33 timoguin joined #salt
14:37 djaykay nagios here
14:42 racooper Zabbix, and quite happy to have replaced nagios with it.
14:43 linjan nagios on omd
14:44 quickdry21 joined #salt
14:44 gadams9991 joined #salt
14:44 psyl0n joined #salt
14:46 JasonSwindle joined #salt
14:48 balboah if state A requires state B which have a watch_in for state C, would that mean that B AND C is ran before A at all times? or would it just mean that C will run at some point after B
14:50 MohShami racooper, I'm interested in checking zabbix as well, but I'd like to do it in a way that salt does all the configuration, have you ever tried such a thing?
14:50 gmtech_ For anyone interested, hiera ext_pillar is a complete misfit. To make it work, hiera would have to return yaml for all its vars, but it returns specific lookup values. This is never going to be fixed for saltstack because puppet bought it. Therefore really the hiera ext pillar should be retired from salt because it leads you down completely the wrong path.
14:51 racooper no, but Zabbix does have an API (caveat: it is only accessible when using Zabbix's own internal auth, not external auth sources)
14:51 MohShami I see
14:51 MohShami what about nagios?
14:52 racooper nagios config is all text files, so I'm sure that can be managed via salt as well
14:53 MohShami yep, I've done some nagios configuration before, but I was wondering if the salt master can get the data and generate the files, or maybe do it on a minion
14:53 MohShami like get all servers that have a web-server role and monitor http
14:54 racooper I'm sure anything you want to script can be done :)
14:54 MohShami can you point me to where to start?
14:54 MohShami I'd also like to use salt to manage my authoritative DNS as well
14:55 MohShami we're researching apache cloud stack, so I want to use everything with salt :)
14:55 MohShami love it that much :D
14:58 jeremyBass1 joined #salt
14:59 abe_music joined #salt
14:59 rlarkin joined #salt
15:00 oz_akan_ joined #salt
15:01 vejdmn joined #salt
15:02 gadams999 joined #salt
15:03 kaptk2 joined #salt
15:05 tyler-baker joined #salt
15:05 tyler-baker joined #salt
15:08 Gordonz joined #salt
15:08 Gordonz joined #salt
15:09 Mua_ joined #salt
15:10 jimallman joined #salt
15:15 kalloc joined #salt
15:16 elfixit joined #salt
15:16 Mua joined #salt
15:16 bhosmer joined #salt
15:17 bhosmer_ joined #salt
15:17 krak3n` joined #salt
15:18 bhosmer_ joined #salt
15:19 GradysGhost joined #salt
15:20 nocturn Hi, how do I best debug that a minion does not register with the master?
15:20 nocturn I already ruled out firewalls
15:21 GradysGhost nocturn: Have you tried running your minion and master in debug mode?
15:21 GradysGhost salt-minion -l debug
15:21 GradysGhost salt-master -l debug
15:22 nocturn GradysGhost, haven't yet, will try that now
15:22 GradysGhost Okay, let me know the results if you can't decipher the output
15:25 nocturn GradysGhost, unbelievable, in debug mode it works
15:26 GradysGhost hah. ok
15:26 GradysGhost so
15:26 nocturn I restarted numerous times, tried reinstalling the minion
15:26 nocturn and in debug, it works immediately
15:26 GradysGhost Sometimes I've seen cases where the salt minion gets loads of processes running, I suspect due to hanging tasks, and a few good restarts gets things back online.
15:27 GradysGhost Try: ps aux | grep salt-minion
15:27 GradysGhost To see if you have a bazillion processes
15:27 GradysGhost I sometimes have to kill them manually
15:27 harobed joined #salt
15:27 nocturn GradysGhost, I had that on CentOS with Salt 0.16, checked now and it's ok
15:28 harobed joined #salt
15:29 GradysGhost Does it still work if you restart the minion normally again?
15:29 nocturn GradysGhost, It's still not ok though.  I did salt-key -A to accept the keys
15:29 nocturn but the minion stops
15:30 nocturn DEBUG   ] Attempting to authenticate with the Salt Master at 10.73.1.95
15:30 nocturn [DEBUG   ] Loaded minion key: /etc/salt/pki/minion/minion.pem
15:30 nocturn [DEBUG   ] Loaded minion key: /etc/salt/pki/minion/minion.pem
15:30 nocturn [INFO    ] Waiting for minion key to be accepted by the master.
15:30 GradysGhost Weird. How about salt-key -L?
15:30 nocturn [DEBUG   ] Loaded minion key: /etc/salt/pki/minion/minion.pem
15:30 nocturn shows it as accepted
15:30 GradysGhost salt '*' cmd.run hostname
15:30 GradysGhost Does that produce the expected output?
15:30 eikonomega joined #salt
15:30 JasonSwindle nocturn:  are you on RC1?
15:31 harobed joined #salt
15:31 nocturn JasonSwindle, I have salt-minion 0.17.4 from EPEL
15:31 eikonomega Hi all, could someone help me work past installation issues on RHEL5 for salt-minion?
15:31 JasonSwindle Ah, ok. :)
15:31 pdayton joined #salt
15:32 GradysGhost eightyeight: I don't run any RHEL stuff, but I'll do what I can. What have you done so far, and what's the problem?
15:32 GradysGhost ugh
15:32 GradysGhost that was meant for eikonomega
15:32 Linz joined #salt
15:32 GradysGhost That's what I get for trusting my tab key
15:33 eikonomega Thanks GradyGhost.  Here's the issue...
15:33 Ahlee eikonomega: you'll want to either build or find an RPM for python26-zmq compatible with zeromq3
15:34 eikonomega sudo yum install salt-minion spits out
15:34 Ahlee i ended up mirroring a good dozen or so RPMs from epel to get the python26 environment to a level salt requires
15:34 eikonomega python26-zmq-2.1.9-3.el5.x86_64 from epel has depsolving problems   --> Missing Dependency: libzmq.so.1()(64bit) is needed by package python26-zmq-2.1.9-3.el5.x86_64 (epel)
15:35 Ahlee eikonomega: you need an RPM that satisfies libzmq.so.1
15:36 eikonomega Ahlee: Do you know where I could find one?
15:36 GradysGhost You can try 'yum provides libzmg.so.1'
15:37 GradysGhost Except, you know, type the filename right, unlike me
15:37 Ahlee eikonomega: zeromq ships with a .spec file
15:37 eikonomega Sorry if that is a stupid question... I'm a developer... new to sysadmin stuff.
15:37 Ahlee eikonomega: if memory serves it was relatively straight forward to build from the .spec
15:37 UtahDave joined #salt
15:38 harobed joined #salt
15:39 eikonomega Yum thinks that it is provided by zeromq-2.1.9-1.el5.i386 which is already installed...
15:39 harobed joined #salt
15:39 Ahlee eikonomega: .i386, but python26-zmq is looking for .x86_64
15:39 Ahlee so, either try to install the 32 bit version of python26-zmq
15:39 Ahlee welcome to the wonderful world of architecture dependencies
15:40 harobed joined #salt
15:40 forresta joined #salt
15:41 harobed joined #salt
15:42 harobed joined #salt
15:42 eikonomega I don't like this world ;)
15:42 harobed joined #salt
15:43 Psyke joined #salt
15:43 harobed joined #salt
15:43 eikonomega Perhaps a different approach... can I set up a virtualenv, install dependencies into in an run Salt Stack from there?
15:44 patrek joined #salt
15:46 Ahlee yes
15:46 Ahlee You can
15:46 Ahlee I moved away from that to RPM packages
15:47 Ahlee as I prefer to stick to system packages for system processes
15:47 Ahlee but, whatever's easier for ya
15:47 eikonomega Sure.  I can see that.
15:47 Ahlee eikonomega: originally i launched salt ina  python2.7 venv
15:47 eikonomega Theoretically, they would work better!
15:47 Ahlee it had it's own issues, the key one that made me roll back was it couldn't handle yum
15:48 Ahlee i never troubleshot it
15:48 Ahlee as for me shipping an RPM is easy
15:48 scoates recurring jinja whack-a-mole. )-:
15:48 yota joined #salt
15:49 forresta scoates, pssh that never happens.
15:49 dave_den joined #salt
15:49 scoates never every day (-:
15:49 hunter joined #salt
15:49 salt_whitebelt joined #salt
15:50 fllr joined #salt
15:50 salt_whitebelt has anybody ever gotten cmd.script to run powershell on windows minion?
15:52 scoates "line 353 in template" is code words for "you're going to have a fun afternoon"
15:52 eikonomega Ahlee: Ok.  So tell me if I have this right.
15:52 forresta scoates, there aren't actually 353 lines in your template though right?
15:53 forresta usually that seems to be a red herring when it reports like that...
15:53 scoates forresta: almost certainly that means the salt python code, file: template, line 353
15:53 esogas left #salt
15:53 forresta yea
15:53 scoates template is 93 lines long
15:54 eikonomega yum install salt-minion attempts to update my current version of python26.x86_64.
15:54 lessismore joined #salt
15:54 opapo joined #salt
15:54 eikonomega But the attempt fails because libzmq.so.1()(64bit)
15:54 eikonomega is needed.
15:54 forresta eikonomega, is a version of zeromq available from the repo?
15:54 forresta and what OS is this?
15:55 eikonomega Yes... I've got zeromq installed... just a second for the info
15:55 forresta what version of zeromq though?
15:55 eikonomega zeromq.i386                                                 3.2.2-13.1                                               installed zeromq.x86_64                                               3.2.2-13.1                                               installed
15:56 forresta two copies of zmq?
15:56 eikonomega RHEL 5.9
15:56 eikonomega Yes... it appears that both 32-bit and 64-bit are installed on the system.
15:57 forresta do you actually have python2.6 installed? Or is it 2.4
15:58 gadams999 left #salt
15:58 eikonomega Looks like both to me
15:58 eikonomega python.x86_64                                                 2.4.3-56.el5                                           installed python-devel.i386                                             2.4.3-56.el5                                           installed python-devel.x86_64                                           2.4.3-56.el5                                           installed python-dmidecode.x86_64
15:58 eikonomega Sorry... bad output.
15:58 forresta that's just python 2.4
15:58 eikonomega python26.x86_64                                               2.6.8-2.el5                                            installed
16:00 forresta so if you do a yum update python26, what happens?
16:00 eikonomega No Packages marked for Update
16:01 forresta when you try to install the salt-minion, what does the update list look like?
16:01 eikonomega It's pretty long... I'll try to copy/paste it.
16:01 forresta just gist/pastebin it
16:01 forresta don't paste it in here
16:01 forresta you'll get disconnected for spamming
16:03 eikonomega I was just thinking that... ;)
16:03 JasonSwindle UtahDave:  Howdy!
16:04 eikonomega https://gist.github.com/eikonomega/8542862
16:04 Veticus joined #salt
16:05 lessismore Hello, I am just wandering if my salt stack is properly configured. It seems a little bit slow.  A I have only one state with one managed file with a template with 2 vars coming from pillars.  The command salt 'server' state.highstate run in around 10 seconds. A ssh connection to the system takes less than 1 sec.  I searched but i did not found any performance test hint in the doc.
16:06 scoates lessismore: offhand guess that one of your states installs a package, and it's doing `apt-get update` or similar
16:07 lessismore scoates: How can I discover that? -v?
16:07 scoates forresta: s/defaults/default/ for a typoed template var. /-:
16:07 cachedout joined #salt
16:07 scoates lessismore: from the minion: salt-call -l debug state.highstate
16:07 druonysus joined #salt
16:08 forresta eikonomega, looks like the issue is that you're trying to install a non 3+ version of zmq at the very least (might have to pick it up somewhere for RHEL5). There was a topic on this a while back: https://groups.google.com/forum/#!topic/salt-users/c9p49ZUczJo
16:08 forresta scoates, hah
16:09 davet joined #salt
16:10 eikonomega forresta: That is what it appears to be doing, which I thought was strange given that it a deprecated version.
16:10 Ahlee eikonomega: i resolved that issue by rebuilding python26-zmq and zeromq3
16:10 Gifflen Anyone know why salts docstrings use ''' instead of """ ?
16:10 eikonomega Also, I already have 3.2.2 installed remember?
16:10 Ahlee eikonomega: issue is likely in python26-zmq
16:11 eikonomega Ok... how do I rebuild it?
16:11 eikonomega using the yum tool...
16:11 Ahlee eikonomega: building an rpm development environment is a bit above the scope of what can easily be accomplished over IRC
16:12 Ahlee let me see if i can can release mine
16:12 UtahDave Morning, JasonSwindle!
16:13 Ahlee boss says I can, but not during work hours.  So it'll be around 9 hours before they're available
16:13 JasonSwindle Well, pretty much lunch for me now.  Pho for me! :)
16:14 forresta isn't it 10 am in texas?
16:14 vejdmn joined #salt
16:14 Ahlee yes
16:14 Ahlee well, it's 10:14 in CST
16:15 fllr joined #salt
16:16 JasonSwindle Yep, I lunch at 11
16:16 teskew joined #salt
16:17 Ahlee high five!
16:17 Ahlee when your day starts at 6, you get hungry at 11 :)
16:17 JasonSwindle Yeah, really.
16:17 Ahlee up at 5, work at 6, breakfast at 6:30-7ish, lunch at 11
16:17 scoates thought JasonSwindle might be in Halifax, for a min. (-:
16:18 JasonSwindle No, in the heartland of TexMex food.
16:19 higgs001 joined #salt
16:20 eightyeight GradysGhost: bad tab-complete. :)
16:21 timoguin JasonSwindle, stop making me hungry!
16:22 lessismore scoates: there was one dependency with some package, removing it for test, and the time goes down to arond 3 sec.  Not blazing fast but resonable, tnx
16:23 strgcloud joined #salt
16:26 JasonSwindle timoguin:  :P
16:26 JasonSwindle timoguin:  It is also Food Truck day, too.
16:26 timoguin stop it!
16:27 JasonSwindle timoguin:  Are you going to SaltConf?
16:27 timoguin not this time. :(
16:28 JasonSwindle Darn
16:29 salt_whitebelt best tex mex on the face on this planet! is a little unknown in the middle of nowhere restaurant called cueva de leon in fort davis. it really is in the middle of nowhere as fort davis is a isolated small town in the middle of the desert.
16:29 harobed_ joined #salt
16:30 salt_whitebelt i live here in texas all my life and i've eaten all over but i have never ever tasted any chips an salsa as good as this place. everything is guud.
16:31 JasonSwindle salt_whitebelt:  What part?  SATX for me.
16:31 salt_whitebelt H town
16:32 scott_w joined #salt
16:33 gmtech_ Is it possible in a (jinja) sls file to include all files in a subdirectory, rather than name them specifically ?
16:34 * GradysGhost knows a few people in H town
16:35 troyready joined #salt
16:35 dave_den gmtech_: no, but as a workaround you can just have the include of all files in the directory's init.sls. then you only need to include that directory name in your other includes
16:36 gmtech_ Yeah I thought about that. Trying to clean up by trying to avoid duplication of redundant stuff, but I guess its not possible. Thanks
16:37 fllr joined #salt
16:37 eikonomega Hey ahlee:  I think that I'm going to play around with this on a new vm.
16:37 eikonomega Get around some of the dependency issues.
16:37 eikonomega Thanks for your help though.
16:38 [diecast] joined #salt
16:38 MTecknology I keep getting highlighted
16:39 SpeeR joined #salt
16:39 dave_den mtec
16:39 eikonomega Also thanks to GradysGhost
16:40 GradysGhost No problem!
16:40 GradysGhost oh
16:40 GradysGhost well, i guess i was late to that one
16:40 * GradysGhost can't seem to fill the coffee-shaped hole in his soul today
16:40 forresta GradysGhost, That's because you need to fill it with tea.
16:41 GradysGhost I guess I could do some tea this afternoon
16:41 GradysGhost I like mixing the earl gray with the mint stuff.
16:42 mgw joined #salt
16:42 MTecknology dave_den: I get highligted on [Mm][Tt]ec[kh] and [Mm][Tt]ec[kh]nology
16:44 scoates just told one of our guys "Yes, it would be faster to port everything over to Salt than to move PuppetDB to a new AWS account."
16:44 scoates and I think that's accurate.
16:45 rojem joined #salt
16:46 fllr joined #salt
16:46 JasonSwindle I have never used Puppet, but I hear a lot about it.
16:46 MTecknology I've tried to use it. It's too cumbersome
16:46 vlcn agreed.  I tried puppet before Salt.
16:47 JasonSwindle I tried Chef before Salt...
16:47 MTecknology I looked at chef
16:47 nineteeneightd joined #salt
16:48 MTecknology I looked at a lot and thought, "if this is what it takes to just install these tools and make them work with any level of usefulness, screw it... I'll do it by hand"
16:48 hunter joined #salt
16:48 MTecknology the first time I looked, salt wasn't very usable either
16:48 [diecast] joined #salt
16:51 fllr joined #salt
16:53 MTecknology salt's become incredibly usable and useful in an incredibly short amount of time
16:53 zach chef is a nightmare
16:54 MTecknology It doesn't feel like a fast growing child anymore, it feels like a massive project with many contributors that's ready for the big time; the stability and compatibility issues are less and less with every release
16:54 MTecknology chef looked scary, but I never used it
16:54 zach I don't like Chef or Puppet
16:54 zach They are a nightmare to maintain
16:54 zach salt is easy as pie
16:55 forresta Yea the ease of getting it going is a win
16:55 amckinley joined #salt
16:55 forresta I should have taken a video of a presentation I did a while back where I had Salt up and running in 15 minutes. The shock/excitement from people was pretty funny
16:57 abe_music salt ftw!!!!
16:57 zach my last day of using salt in a largescale environment is today
16:57 forresta before you implmenet at a new environment?
16:57 zach so glad to be done with the mess of the infrastructure here
16:57 zach yep!
16:58 zach Next one will be 95% AWS based
16:58 zach 5% in-house
16:59 bhosmer joined #salt
16:59 zach so it shall be entertaining!
17:01 KyleG joined #salt
17:01 KyleG joined #salt
17:01 rgbkrk joined #salt
17:02 bemehow joined #salt
17:03 vejdmn joined #salt
17:04 hunter joined #salt
17:06 hotbox_ joined #salt
17:06 dave_den1 joined #salt
17:07 BenCoinanke_k joined #salt
17:07 Shish_ joined #salt
17:07 psyl0n_ joined #salt
17:08 jeremyBass2 joined #salt
17:08 eclectic_ joined #salt
17:08 _david_a joined #salt
17:08 jefferai_gone joined #salt
17:08 maci_ joined #salt
17:08 puppet_ joined #salt
17:08 pexio_ joined #salt
17:09 gmtech joined #salt
17:09 jgiraldo_ joined #salt
17:09 Psyke joined #salt
17:09 bbinet joined #salt
17:09 xmj joined #salt
17:09 rjc joined #salt
17:09 delkins joined #salt
17:09 shadowsun joined #salt
17:09 xmj joined #salt
17:09 rjc joined #salt
17:09 beardo_ joined #salt
17:09 kaptk2 joined #salt
17:09 whyzgeek joined #salt
17:10 trevorj joined #salt
17:10 stotch joined #salt
17:10 jeblair joined #salt
17:10 djaykay joined #salt
17:10 harobed_ joined #salt
17:11 matanya joined #salt
17:11 dave_den joined #salt
17:12 MedicalJaneParis joined #salt
17:12 KyleG left #salt
17:12 dave_den what did the shoes say to the pants?
17:13 Jahkeup what's up
17:13 Jahkeup :D
17:13 dave_den what's up, britches!
17:13 Jahkeup LOLOL you's a funny one you are
17:13 esogas_ joined #salt
17:13 dave_den har har har
17:14 Iwirada left #salt
17:14 nkuttler joined #salt
17:14 nkuttler joined #salt
17:14 esogas left #salt
17:14 Jahkeup Anyone have any experience with salt-minion on windows?
17:15 gmtech I tried this in my pillar top.sls :
17:15 gmtech include:
17:15 gmtech - environments/*
17:15 gmtech didn't work
17:15 gmtech can you use wildcards like this?
17:15 bhosmer_ joined #salt
17:15 jtang joined #salt
17:15 bhosmer_ joined #salt
17:16 nahamu anyone using Salt on a SunOS flavor using it to set user passwords successfully (or also seeing that that is broken)?
17:16 jdenning joined #salt
17:17 forresta gmtech, I don't believe so.
17:25 pdayton joined #salt
17:25 mgw joined #salt
17:26 pdayton1 joined #salt
17:26 tessellare joined #salt
17:27 scott_w joined #salt
17:30 Ahlee oh yay today turned into a compliance day
17:32 rgarcia_ joined #salt
17:33 batulin-s joined #salt
17:33 rgarcia_ joined #salt
17:36 david_a joined #salt
17:37 gadams999 joined #salt
17:42 oz_akan_ joined #salt
17:42 Linz joined #salt
17:43 jalbretsen joined #salt
17:43 vejdmn joined #salt
17:44 [diecast] joined #salt
17:45 joehillen joined #salt
17:46 Linz joined #salt
17:47 pdayton joined #salt
17:48 Xiao joined #salt
17:49 Veticus joined #salt
17:51 toastedpenguin left #salt
17:52 toastedpenguin1 joined #salt
17:53 oz_akan_ joined #salt
17:54 n8n joined #salt
17:55 jeremyBass1 question, the order the the debug outputs the list of what happend is the order it ran?  ` Name: /etc/resolv.conf - Function: file.managed - Result: True /  Name: curl - Function: pkg.installed - Result: True  /  Name: dos2unix - Function: pkg.installed - Result: True` means that  /etc/resolv.conf happened then curl then dos2unix?
17:55 andrewclegg joined #salt
17:55 sroegner_ joined #salt
17:56 bhosmer_ joined #salt
17:57 srage joined #salt
17:58 jeremyBass1 can someone confirm that is a truth?
17:59 jimallman joined #salt
17:59 strgcloud joined #salt
18:01 UtahDave jeremyBass1: let me check on that.
18:01 Ryan_Lane2 joined #salt
18:02 jeremyBass1 the issue I'm facing is somehow what is said here.. http://docs.saltstack.com/ref/states/top.html#how-top-files-are-compiled .. `The base environment's top file is processed first` .. this doesn't seem to be true.. i have an environment's top file being done first and then the base..
18:03 abele_ joined #salt
18:04 neilf_ joined #salt
18:04 copelco__ joined #salt
18:05 viq Is there a way to do equivalent of 'yum groupinstall' from salt state?
18:05 mattmtl is it possible to avoid multiple execution of a module call if the first hasn't finished yet?
18:06 kermit joined #salt
18:07 Psyke joined #salt
18:08 JesseC joined #salt
18:09 srage joined #salt
18:11 scott_w joined #salt
18:12 thrashr888 joined #salt
18:13 fridiculous joined #salt
18:14 jslatts joined #salt
18:15 fridiculous does anyone know how to manage multiple file_roots with a standalone minion?  More generally, I'm trying to run srv/salt/dev in a standalone local minion with --local, but I can't target the minion
18:15 Ryan_Lane2 joined #salt
18:17 UtahDave fridiculous: are you setting file_client: local  in your minion config?
18:17 UtahDave mattmtl: You can do that in a custom execution module.
18:18 fridiculous yes file_Client is set to local in minon
18:18 UtahDave mattmtl: here's an example: https://github.com/saltstack/salt/blob/develop/salt/modules/state.py#L78
18:19 UtahDave fridiculous: ok, with file_client: local  you don't have to add the --local to the command.  It's implied.
18:19 UtahDave Then you just add the    file_roots:    option to your minion config
18:19 jeremyBass1 in short what seems to be happening is , https://github.com/washingtonstateuniversity/WSUMAGE-base/blob/master/provision/salt/top.sls seems to be running before https://github.com/washingtonstateuniversity/WSU-Web-Serverbase/blob/master/provision/salt/top.sls .. even though that would conflict with the doc's as i understand.. i even have the minion matching https://github.com/washingtonstateuniversity/WSU-Web-Serverbase/blob/master/prov
18:19 mattmtl UtahDave: I already have a custom module... this is pretty nice, thanks!!
18:20 fridiculous let me try it
18:20 mattmtl UtahDave: you have an idea if it is possible yield output (stdout) and send it back to master as a module is running?
18:21 UtahDave mattmtl: no, not yet.  The module can only return one time at the end of the execution.   We plan on having that eventually
18:21 UtahDave mattmtl: a work around would be to use event.fire_master inside your module to send messages up the event bus during execution
18:22 UtahDave jeremyBass1: The top.sls in your base environment should be first.
18:22 mattmtl UtahDave: thanks, I will investigate that avenue
18:22 fridiculous without the --local tag, i get : [ERROR   ] This master address: 'salt' was previously resolvable but now fails to resolve! The previously resolved ip addr will continue to be used [WARNING ] Master hostname: salt not found. Retrying in 30 seconds
18:23 fridiculous even with the implied flie_client: local
18:24 mattmtl btw, I did a custom module that return the list of devices (mounted and unmounted) from a minion... is it something that could be interesting for the project?
18:24 forresta mattmtl, yea that could be cool! You should look at making a pull request on the contrib: https://github.com/saltstack/salt-contrib
18:24 ekristen joined #salt
18:24 fridiculous nvm i think i see the problem
18:24 ekristen so if I want to grab network stuff from publish.publish network.* needs to be added to peers?
18:25 elfixit joined #salt
18:26 JasonSwindle1 joined #salt
18:26 fridiculous (thumbsup)_
18:26 rojem joined #salt
18:26 fridiculous my minion file was getting swapped the wrong way in the docker container i was running
18:27 rojem joined #salt
18:27 SpeeR joined #salt
18:28 rojem joined #salt
18:28 KyleG joined #salt
18:28 KyleG joined #salt
18:29 jeremyBass1 @UtahDave so if the debug has the non-base coming before the base is there a way to correct that?  As i see it, it seems to be ordered as required
18:29 SpeeR joined #salt
18:29 dangra joined #salt
18:30 xmltok joined #salt
18:31 ajw0100 joined #salt
18:33 Ryan_Lane21 joined #salt
18:35 ekristen JasonSwindle1: you there?
18:36 gadams9991 joined #salt
18:36 Ryan_Lane2 joined #salt
18:36 JasonSwindle1 ekristen:  I am
18:36 Ryan_Lane2 joined #salt
18:37 ekristen is mine.get and publish.publish interchangeable in terms of what comes in ( ) after it? assuming the peer stuff is setup properly?
18:37 Ryan_Lane21 joined #salt
18:38 JasonSwindle1 Never used mine.get
18:38 JasonSwindle1 Sorry……. publish was around before mine when I started Salt. :)
18:38 JasonSwindle1 or never used mine in general
18:38 ekristen JasonSwindle1: fair enough
18:38 ekristen I think its interchangeable how the searching works
18:39 JasonSwindle1 give it a try. :)
18:39 ekristen that is what I am doing ;)
18:40 jeremyBass1 hmm.. in the log even the rendering is do the non-base first then the base.. what can cause that to happen?
18:45 srage joined #salt
18:45 qkaiser left #salt
18:45 Mua joined #salt
18:46 dmwuw joined #salt
18:47 gadams999 joined #salt
18:48 ashtonian joined #salt
18:50 dangra is GitPython going to be the default gitfs backend for 2014.1 or is libgit2 bindings taking over it?
18:51 ekristen hrm I’m getting “TypeError: 'NoneType' object is not iterable” all of a sudden on highstate
18:52 dangra ta, 2014.1 sources haven't references to libgit2
18:52 ajw0100 joined #salt
18:52 baffle Is anyone using the Postgres state? Things like postgres_user.present; How is that supposed to work? I get auth problems, but here is no users in the DB yet, except the local postgres user.
18:56 forresta dangra, that would be a question for terminalmage
18:57 SpeeR joined #salt
18:57 terminalmage dangra: pygit2 is experimental
18:57 terminalmage it will not be in 2014.1
18:57 terminalmage it was just added in the last week and a half
18:58 terminalmage or 2 weeks
18:58 terminalmage so, it's too new to put into a feature release
18:58 terminalmage since it required a complete rewrite of the gitfs backend
18:59 terminalmage dangra: also, pygit2 doesn't support credentialed access
18:59 SpeeR_ joined #salt
18:59 jtang joined #salt
18:59 terminalmage so, if you have a git repo which requires http(s) authentication, or requires ssh auth, it will not work
18:59 ekristen is there anyway to debug why highstate is failing to merge top.sls files?
19:00 GradysGhost ekristen: I suggest running salt master and minion in debug mode to see if they provide any addition details.
19:00 GradysGhost salt-master -l debug
19:00 GradysGhost salt-minion -l debug
19:00 ekristen GradysGhost: already doing that
19:01 GradysGhost Failing that, what I've done in the past is to comment out parts of the top file and run the highstate to find the sls that's causing the problem.
19:01 GradysGhost It's trial-and-error, but chances are you have a bad require or a missing pillar or something.
19:01 GradysGhost What kind of output are you getting?
19:02 ekristen http://pastebin.com/dHvNUCCe
19:02 thrashr888 joined #salt
19:02 GradysGhost That is pretty vague.
19:02 [diecast] joined #salt
19:03 GradysGhost I suggest doing the trial-and-error thing. Comment out most of the top file, then uncomment them one by one, running the highstate each time to see where the error pops up.
19:03 ekristen yeah
19:03 GradysGhost Someone else here may have better advice, but I've done it before with success.
19:03 ekristen I’ve got 5 environments :/
19:03 gadams9991 joined #salt
19:04 GradysGhost ekristen: http://i.imgur.com/syUEq4e.jpg
19:04 GradysGhost I keep a separate environment for me to test salt configurations before putting them into staging/whatever
19:05 hunter joined #salt
19:05 ekristen yeah, I have separate environments, but for some reason even though the minion is pinned to an environment, it still looks at all top.sls across all environments for matches
19:05 ekristen its rather annoying actually
19:06 zach "Testing in production is the only way"
19:09 costi joined #salt
19:09 vlcn is there such a thing as exclude in a top file?
19:10 SpeeR joined #salt
19:11 alekibango joined #salt
19:13 costi is there a way to organize runners in subdirs?
19:16 pdayton1 joined #salt
19:16 faldridge joined #salt
19:18 pdayton2 joined #salt
19:19 f3d joined #salt
19:21 baffle Anyone know how to bootstrap PostgreSQL from salt? state.postgres_user doesn't seem to work with the "user: postgres" setting, wich if I understand things correctly is the new "runas:" function to do things locally..
19:22 hunter ekristen: I just read that doc - thats how salt wants to do it - IE read all the top.sls files configured in the repo.
19:23 hunter gmtech: let me know if you figure that glob question out
19:23 kermit joined #salt
19:23 ekristen hunter: according the minion docs if environment is set then it pins itself to that env
19:24 hunter Oh? Hmm.
19:24 jeremyBass1 ok I'm pounding my head here.. how would env XXX get forced to run before the base?  that is what is happening, the docs as i read them say it can't, so there has to be something that is making it right?
19:24 hunter ekristen: how are you pinning a given client to a given environment?
19:24 ekristen minion config, like the docs say
19:24 hunter ekristen: I was hoping not to do it there - I would really like to set the env for a given client on the server somehow.
19:25 hunter IE, I don't want to have to edit something ON the client to promote it from dev to prod
19:26 * apergos is looking for cro and hoping that today will be better for talking about 0.17.1 testing
19:26 druonysuse joined #salt
19:27 amckinley1 joined #salt
19:32 vlcn Is there a way I can specify an 'expected' return code?
19:39 faldridge I'm changing the nginx conf files in a state that requires my statefile that actually includes the service and watch directives.  When the files are changed in the child state (the one that includes the other other), the service is not getting reloaded.
19:39 faldridge Does the child statefile need to have watch_in?
19:41 zooz joined #salt
19:43 Ryan_Lane2 joined #salt
19:43 cedwards blist online
19:43 cedwards erg
19:47 Ryan_Lane2 joined #salt
19:48 faldridge yes, watch_in fixes it. :D
19:48 fridiculous whats' the best way to organize environments?   You can use separate config files in the file roots as is recommended by the documentation,  but if you do that, you'll end up with multiple config files per environment, when in reality all you want to do is just pass a couple variables to each config file through jinja
19:49 fridiculous e.g. i don't want redis_server.conf to live in each file root, because it increases the liklihood that he environments aren't matching
19:50 fridiculous i'd rather have 1 redis_server.conf with jinja that is shared between qa / prod / dev
19:53 fridiculous i guess, i'm saying if each environmental variable could be managed through jinja, what's the point of dividing it up with a file root ; assuming you want the environments to mirror one another as much as possible
19:53 [diecast] joined #salt
19:54 scoates second fun today "SaltRenderError: Undefined jinja variable; line 353 in template"
19:55 [diecast] joined #salt
19:56 scoates I wonder what we can actually *do* about that. Because it's worth a few hours of my time to attempt to fix it, at this point.
20:03 blee joined #salt
20:03 Kax_ joined #salt
20:03 SpeeR joined #salt
20:03 Ryan_Lane2 joined #salt
20:04 blee can someone point me at the salt changelogs
20:04 blee for the rc1?
20:04 blee i feel like im failing at the internet right now
20:05 forresta blee, http://docs.saltstack.com/topics/releases/2014.1.0.html
20:05 Gordonz joined #salt
20:05 thrashr888 joined #salt
20:05 felixhummel joined #salt
20:05 forresta those also probably aren't final yet blee, since we're still using the RCs
20:05 btpier joined #salt
20:07 btpier Hello. Does anyone here have experience using the pkg.upgrade command in the yum module and setting a batch size?
20:08 UtahDave btpier: the batch option isn't tied to the pkg.upgrade command, so it should just work.   what have you tried so far?
20:08 btpier what I see is that the batch mode doesn't actually wait for pkg.upgrade to run
20:09 monokrome joined #salt
20:09 btpier so when issuing pkg.upgrade against 65 dev VMs they all run at the same time (which heavily loads the VM Host servers)
20:10 forresta are you specifying a batch size of 65?
20:10 btpier nope, size of 5
20:10 forresta btpier, https://github.com/saltstack/salt/issues/8597
20:13 jeremyBass1 am i miss understanding the docs here http://docs.saltstack.com/ref/states/top.html#how-top-files-are-compiled ,, i read it as the base env top.sls will render and process first, then XXX env top.sls will render next and as long as there is not a base: in the XXX env top it is suppost to mantain that order of processing.. is that how it's suppose to work?
20:13 btpier thanks forresta. that's exactly what I'm encountering. I expect a batch to run and wait to finish before firing the next batch
20:14 JasonSwindle joined #salt
20:14 forresta btpier, yep we discovered that a while ago. Try increasing your timeout value for these specific kind of jobs until that slacker basepi finishes rewriting the batch system with all the free time he has :P
20:14 btpier especially with something as I/O intensive as pkg.upgrade on potentially hundreds of VMs
20:14 basepi forresta: ha!
20:14 * basepi goes back into his cave
20:14 forresta btpier, yep! The timeout increase is a hack, but it does help a bit
20:15 btpier forresta, thanks!
20:15 forresta btpier, np.
20:17 JasonSwindle forresta:  got more issues in for you :P
20:17 JasonSwindle sass away
20:17 forresta heh
20:17 forresta too busy today :\
20:17 JasonSwindle To sass, my my my....
20:19 ccase joined #salt
20:19 jslatts joined #salt
20:19 forresta JasonSwindle, yea its' sad.
20:19 forresta *it's
20:27 bemehow joined #salt
20:29 efajardol joined #salt
20:31 SpeeR joined #salt
20:32 ChoHag joined #salt
20:32 ChoHag How does a salt minion know that its initial connection prior, to generating the master/minion communication keys and certificates, is to a safe salt master?
20:33 ChoHag Specifically, assuming it uses pki, how do you tell the minion the ca of the certificate the master will present?
20:38 UtahDave ChoHag: By default, the salt-minion accepts the Salt Master's key when it first connects.
20:38 efajardol left #salt
20:38 UtahDave ChoHag: You can add the salt master's public key fingerprint to the minion's config file
20:38 ChoHag How do you get it to a) not do that, and b) only accept an appropriately signed certificate?
20:38 UtahDave ChoHag: master_finger: lksljaslkfoweroj2j32lkj32lk2j
20:39 UtahDave ChoHag: if the master doesn't present a public key matching the fingerprint in master_finger option, then it shuts down the salt-minion daemon
20:40 ChoHag master_finger is not listed in the minion configuration reference.
20:42 thrashr8_ joined #salt
20:43 UtahDave ChoHag: Here's the definition: https://github.com/saltstack/salt/blob/develop/conf/minion#L397
20:44 ChoHag That says "to double verify the master is valid"
20:44 ChoHag What is to singly verify?
20:44 UtahDave what do you mean by "singly verify"?
20:45 ChoHag If with master_finger it will double verify, without it will single verify.
20:46 Marion joined #salt
20:48 UtahDave ChoHag: If the public key that the master presents doesn't match the key fingerprint in master_finger, then the minion will reject that master.
20:48 ekristen JasonSwindle: you there?
20:49 bemehow_ joined #salt
20:49 ChoHag So is the comment wrong when it refers to a doubling of the verification steps? ie. master_fingerprint provides the only (single) verification of the salt master?
20:49 jeremyBass1 can someone help figure out why http://pastebin.com/Fh4XHT97 i see the [DEBUG   ] Jinja search path: '['/srv/store.wsu.edu/salt']' before the [DEBUG   ] Rendered data from file: /srv/salt/top.sls: ?
20:50 mchype joined #salt
20:50 JasonSwindle ekristen:  I am
20:51 jeremyBass1 the minion is http://pastebin.com/Mbjzc8fa .. and the order is base then env .. but it's going the other way
20:51 JasonSwindle just working on a PPT, which I hate doing.
20:51 ekristen do you have an example of a salt[‘publish.publish’] to grab ipaddrs or network adapters?
20:51 UtahDave ChoHag: well, I think it's referring to the fact that the minion stores the public key of the master it has been accepted to.  If the master presents a different public key, then it rejects the master
20:52 ChoHag Right. So the initial verification step it's referring to is the cached public key.
20:52 UtahDave ekristen: look at the second file here: https://gist.github.com/UtahDave/5217462
20:53 JasonSwindle ip adders, sure
20:53 ekristen UtahDave: how do I target the search at a grain?
20:53 UtahDave ChoHag: I believe so.
20:53 JasonSwindle ekristen:  one moment
20:53 ekristen JasonSwindle: thanks
20:53 JasonSwindle ekristen:  https://gist.github.com/JasonSwindle/5f5c38f1ea78a25ca2f6
20:54 JasonSwindle it grabs the IP adder of eth0
20:54 ekristen sweet, thanks!
20:54 ChoHag Thankyou.
20:54 JasonSwindle and uses grains
20:54 forresta JasonSwindle, did you fix that yet?
20:55 JasonSwindle forresta:  yes and no
20:55 forresta was it the grain?
20:55 JasonSwindle this one works great
20:55 JasonSwindle but the issue I am facing is more complex
20:56 ConceitedCode joined #salt
20:56 JasonSwindle forresta:  my issue is with salt adding [['IP']] on one single sls
20:56 JasonSwindle and at random
20:57 forresta at random? that seems weird.
20:57 JasonSwindle I have been able to nail to down to a "*" vs targeted highstate
20:59 amckinley joined #salt
20:59 UtahDave JasonSwindle: are all the minions on the same version of Salt?
20:59 jeremyBass1 I'm i missing something simple and why no one is pointing it out? lol..
20:59 JasonSwindle Yep, made sure of that.
20:59 JasonSwindle UtahDave:  Salt Maps version locking.
21:00 UtahDave jeremyBass1: sorry.  I'm guessing it has something to do with the kind of non-standard way you're pulling your different projects together.
21:02 JasonSwindle UtahDave:  forresta https://github.com/saltstack/salt/issues/9857
21:02 jeremyBass1 @UtahDave, oh i guess I'm at a loose to why it's none standard thou.. it's based on http://docs.saltstack.com/ref/states/top.html#states-top-file-roots .. I only think I'm doind is building the minion file is all ..
21:02 farra joined #salt
21:03 cewood joined #salt
21:03 forresta jeremyBass1, You might have to just play with it and review the output to see what order things need to be in.
21:05 ekristen JasonSwindle: I have the fix
21:05 jeremyBass1 @forresta, for sure that is what i have been doing, it's just that the base keeps coming last to the other env is the issue i need to solve.. I've been trying to see why, and ensuring that I'm matching the docs on the state top..
21:05 ekristen JasonSwindle: https://gist.github.com/JasonSwindle/5f5c38f1ea78a25ca2f6
21:06 JasonSwindle ekristen:  What did you change?
21:06 forresta he just added a comment
21:06 JasonSwindle Oh I see, dur
21:06 forresta where he specifies the items in the list
21:06 ekristen I fixed the comment
21:06 forresta which is dumb to have to do :P
21:07 JasonSwindle Funny thing is, this one works in RC1
21:07 JasonSwindle it is hit and miss.  That is why I have an issue in and Thomas looks to be poking at it
21:08 jeremyBass1 so maybe i can ensure the parts .. this is the minion, http://pastebin.com/Dv9F1MVQ which matchs http://docs.saltstack.com/ref/states/top.html#environments right?
21:08 JasonSwindle forresta:  I agree, having to do .pop().pop() or etc is silly
21:08 gadams999 joined #salt
21:09 JasonSwindle forresta:  and it was not always like this
21:09 JasonSwindle RC1 Return
21:09 JasonSwindle redis_url: redis://[['11.11.11.11']]:6379/2
21:09 JasonSwindle 0.16.x Return
21:09 JasonSwindle redis_url: redis://11.11.11.11:6379/2
21:09 JasonSwindle and some things return without [['']]
21:11 farra whenever I run a salt command from the master, there is a big delay after the final minion response before the command finishes
21:11 farra example: http://d.pr/i/HqVN
21:12 UtahDave farra: do you have any minions in the list provided  by   salt-key -L  that are down or no longer exist?
21:12 farra any idea why this might be happening? I've tried running the master in debug mode but the output doesn't provide any helpful errors
21:12 farra @UtahDave: let me check
21:14 farra @UtahDave: one minion was not responding. I've restarted the salt-minon service on that minion and still no reply. I'm going to try running salt-minion in debug mode to see what it tells me
21:14 farra thanks
21:16 farra @UtahDave: that was it! got that minion up and running and no more delays in running a test.ping against all minions.  Thanks for your help
21:17 UtahDave farra: Ah good!  That kind of delay is usually the salt master waiting for a minion that it expects a return from
21:17 farra good to know. really appreciate your help
21:18 UtahDave anytime, farra
21:19 timoguin joined #salt
21:19 zooz joined #salt
21:20 n8n joined #salt
21:21 ekristen UtahDave: your example appears to need [0][0] after .values() to work properly
21:21 UtahDave ekristen: Yeah, that makes me think that the datastructure has changed.
21:23 JasonSwindle UtahDave:  it look to have.
21:23 Ryan_Lane2 joined #salt
21:23 JasonSwindle UtahDave:  https://github.com/saltstack/salt/issues/9857
21:24 JasonSwindle but what is more funny is, if I use https://github.com/saltstack/salt/commit/d4d1f692b1c6b4a51e4e621d31373fcbac9dba08 , it seems to be fine....
21:25 ConceitedCode joined #salt
21:26 forresta so what you're saying is if you use the fix, it fixes the problem? :P
21:26 JasonSwindle somewhat
21:27 JasonSwindle some of the SLSs are still needing the [0][0]
21:27 JasonSwindle and then at random don't need it
21:28 de_bigt joined #salt
21:28 JasonSwindle forresta:  I get random IndexError: pop from empty list
21:28 JasonSwindle errors
21:29 forresta so are you passing an empty list?
21:29 JasonSwindle forresta:  You must not be too busy, back to sassing again. :P
21:29 forresta no I still am
21:29 forresta otherwise I'd actually be looking at this
21:29 JasonSwindle I am passing it the same code that works when I do a targeted highstate vs '*'
21:29 JasonSwindle This issue confuses me
21:30 mchype joined #salt
21:31 jslatts joined #salt
21:33 alunduil joined #salt
21:34 Marion joined #salt
21:38 zwevans joined #salt
21:40 jacksontj joined #salt
21:46 jtang joined #salt
21:48 n8n joined #salt
21:48 clearclaw Are there any known problems with mixed case host tags?
21:48 clearclaw (not DNS names, but the targetting tags)
21:49 * clearclaw is getting some really odd/annoying behaviours.
21:50 srage joined #salt
21:52 hunter Is there a way to tell user.present to not put the commas in /etc/passwd if the roomnumber, homephone and workphone aren't specified?  The extraneous commas annoy.
21:52 yomilk joined #salt
21:54 Whissi joined #salt
21:55 scott_w joined #salt
21:55 njpatel joined #salt
22:00 rojem joined #salt
22:01 UtahDave clearclaw: I don't think that should be a problem. What issues are you seeing?
22:01 gadams999 left #salt
22:04 leonardinius joined #salt
22:04 leonardinius_ joined #salt
22:04 sulky joined #salt
22:05 bemehow joined #salt
22:07 jslatts joined #salt
22:07 rojem joined #salt
22:12 elfixit joined #salt
22:12 leonardinius__ joined #salt
22:16 timoguin joined #salt
22:22 timoguin joined #salt
22:24 jdenning joined #salt
22:25 ninkotech_ joined #salt
22:26 vejdmn joined #salt
22:30 davet joined #salt
22:34 david_a joined #salt
22:35 kermit joined #salt
22:35 psyl0n joined #salt
22:38 ninkotech__ joined #salt
22:44 cachedout joined #salt
22:47 jcsp1 joined #salt
22:59 farra is there a fix for the salt-cloud warning messages that get generated with every command?
22:59 david_a joined #salt
22:59 farra ie.  [WARNING ] The function 'cloud_config()' defined in '/usr/lib/python2.7/dist-packages/saltcloud/config.py' is not yet using the new 'default_path' argument to `salt.config.load_config()`. As such, the 'SALT_CLOUD_CONFIG' environment variable will be ignored
23:01 forresta farra, is that in the log?
23:02 farra in the output on the console
23:02 farra command still seems to work
23:02 JasonSwindle joined #salt
23:03 forresta farra, did you try changing: http://docs.saltstack.com/ref/configuration/master.html#log-level
23:06 farra not yet, I'll give that a try thanks
23:07 Ryan_Lane2 joined #salt
23:08 fridiculous is there a simple way to pull all the pillar items for a given environment in commandline?
23:13 de_bigt joined #salt
23:24 rojem joined #salt
23:25 rojem joined #salt
23:26 rojem joined #salt
23:26 rojem joined #salt
23:27 de_bigt joined #salt
23:27 rojem joined #salt
23:28 jtang joined #salt
23:28 rojem joined #salt
23:29 rojem joined #salt
23:29 rojem joined #salt
23:30 rojem joined #salt
23:31 mgw joined #salt
23:33 rowleyaj joined #salt
23:34 JasonSwindle joined #salt
23:34 JasonSwindle left #salt
23:38 esogas_ joined #salt
23:40 BenCoinanke joined #salt
23:47 CheKoLyN joined #salt
23:48 yomilk joined #salt
23:49 Gifflen joined #salt
23:49 ninkotech__ joined #salt
23:50 rowleyaj joined #salt
23:50 Gifflen joined #salt
23:52 CheKoLyN Hello All,  I'm having a weird error.  I'm getting a TemplateNotFound  msg; but it only happens when using salt-ssh. Anyone had this issue?
23:53 ninkotech__ joined #salt
23:54 bemehow joined #salt

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