Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-07-23

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

All times shown according to UTC.

Time Nick Message
00:00 jujube joined #salt
00:12 whitepaws joined #salt
00:16 rushmore1 joined #salt
00:17 nlb joined #salt
00:28 aquinas joined #salt
00:49 beneggett joined #salt
00:54 Theo-SLC joined #salt
01:02 schimmy joined #salt
01:03 beneggett joined #salt
01:07 jbub joined #salt
01:10 yomilk_ joined #salt
01:11 Katafalkas joined #salt
01:20 beneggett joined #salt
01:26 TheThing joined #salt
01:29 beneggett joined #salt
01:34 beardo joined #salt
01:34 Nazzy joined #salt
01:34 Nazzy joined #salt
01:35 copelco joined #salt
01:36 APLU joined #salt
01:36 mapet joined #salt
01:37 MZAWeb joined #salt
01:38 sverrest joined #salt
01:39 tyson_ joined #salt
01:43 yomilk joined #salt
01:45 yomilk joined #salt
01:45 mateoconfeugo joined #salt
01:45 mateoconfeugo left #salt
01:47 mafrosis joined #salt
01:47 mafrosis left #salt
01:50 TheThing joined #salt
01:53 jalbretsen joined #salt
01:54 ckao joined #salt
01:59 gzcwnk when you run salt 'target' state.highstate is there a way to keep it in the foreground so you can see what is happening?
02:08 bhosmer joined #salt
02:09 dimeshake run it with -v and -t360 or somesuch long timeout
02:09 dimeshake it'll stay in the foreground.
02:09 dimeshake if you want debug output, run it from the minion itself
02:09 dimeshake salt-run state.highstate
02:12 bhosmer_ joined #salt
02:12 smcquay joined #salt
02:13 jghazally joined #salt
02:14 gzcwnk is there a way to detrmine if its still running or has died? i thnk its died on me
02:17 whytewolf gzcwnk: get the job id with salt-run jobs.list_jobs then look at the output with salt-run jobs.lookup_jid <jobid>
02:18 jaimed joined #salt
02:19 acabrera joined #salt
02:22 dangra joined #salt
02:25 tkharju3 joined #salt
02:27 joehillen joined #salt
02:30 tinytub joined #salt
02:36 rallytime joined #salt
02:39 kermit joined #salt
02:41 gzcwnk all i get is {}
02:43 gzcwnk ah got an output...seems very slow
02:44 micko joined #salt
02:48 napper joined #salt
02:55 micko joined #salt
02:58 beneggett joined #salt
03:06 dangra joined #salt
03:07 andrej I'm trying a to match an optional subdomain ... {% if salt['match.pcre']('oob-[^.]+\.(internal\.)?sample.ext') %}
03:07 andrej If the minion has internal in its FQDN the match doesn't take
03:08 andrej Where am I going wrong?
03:10 Hipikat_ what is that ‘[^.]+’ trying to do, wouldn’t that match ‘one or more of not any character’?
03:14 dude051 joined #salt
03:15 m1crofarmer joined #salt
03:18 arthabaska joined #salt
03:18 BinLee joined #salt
03:21 mgw joined #salt
03:25 BinLee I meet a problem, when I execute salt command, salt 'sn*' state.sls flume.startup -t 10, it didn't stop until I stop the shell in the minion, is there any problem that the shell in the minion has "nohup" function that output message all the time, and master can't recognize it
03:28 dude051 joined #salt
03:37 gzcwnk I have a file on a server that I want to update while rinngin high.state, file.managed doesnt seem to replace an existing file with a new one? or am i doing something wrong?
03:38 gzcwnk running
03:41 picker joined #salt
03:41 joehillen joined #salt
03:42 markour joined #salt
03:43 jamesbdev joined #salt
03:54 yomilk joined #salt
04:01 bhosmer joined #salt
04:01 n8n joined #salt
04:02 n8n_ joined #salt
04:05 matthiaswahl joined #salt
04:07 ghartz joined #salt
04:09 elithrar joined #salt
04:31 TyrfingMjolnir joined #salt
04:31 schimmy joined #salt
04:36 schimmy joined #salt
04:41 elithrar Migrating some older state files to the new "no dot" format. Having an issue where I have a file.directory where require: - directory spits out "requisites were not found"
04:43 elithrar Output + relevant state file here https://gist.github.com/elithrar/6447fc02b12723a4bebd
04:43 Lomithrani joined #salt
04:44 felskrone joined #salt
04:51 jalaziz joined #salt
04:58 saravanans joined #salt
04:59 ramteid joined #salt
05:04 yomilk joined #salt
05:06 yomilk joined #salt
05:06 beneggett joined #salt
05:10 Sauvin joined #salt
05:13 joehillen joined #salt
05:15 ndrei joined #salt
05:19 saravanans joined #salt
05:35 stephanbuys joined #salt
05:37 justlooks joined #salt
05:37 evidence joined #salt
05:38 n8n joined #salt
05:40 justlooks hi ,i want do following deploy procedure , host A told to do function a ,when a done ,host B start do function b ,after b done,host A begin run function c ,how to do use salt?
05:49 justlooks anyone?
05:50 bhosmer joined #salt
05:53 ajolo joined #salt
05:55 dimeshake joined #salt
05:57 jY justlooks: reactors
05:58 dimeshake joined #salt
06:00 stephanbuys justlooks: the orchestration runners might also work
06:04 justlooks stephanbuys: what's means orchestration runners any doc link?
06:06 stephanbuys justlooks: http://docs.saltstack.com/en/latest/topics/tutorials/states_pt5.html
06:14 Katafalkas joined #salt
06:21 Katafalkas joined #salt
06:38 badon joined #salt
06:45 Valdo joined #salt
06:46 q1x joined #salt
06:50 ajolo joined #salt
06:53 vandemar joined #salt
06:57 ndrei joined #salt
06:59 saravanans joined #salt
07:00 dinamik joined #salt
07:04 xsteadfastx joined #salt
07:07 n8n joined #salt
07:11 n8n joined #salt
07:15 Katafalk_ joined #salt
07:23 ndrei joined #salt
07:23 beando__ joined #salt
07:26 alanpearce joined #salt
07:27 darkelda joined #salt
07:32 stephanbuys joined #salt
07:33 ml_1 joined #salt
07:35 giannello joined #salt
07:37 slav0nic joined #salt
07:38 bhosmer joined #salt
07:49 linjan joined #salt
07:51 luette joined #salt
07:55 eliasp does anyone know of a runner which bundles different updates (e.g. 'salt update.all' instead of fileserver.update, git_pillar.update, saltutil.refresh_pillar, saltutil.sync_all…)
07:56 eliasp to add another one: winrepo.genrepo
07:56 eliasp all the things one can easily forget to run when developing/testing
08:01 babilen eliasp: I am not and you cannot really bundle fileserver.update and sync_all as the former is master specific and the latter targeted to minions. fileserver.update should, IIRC, refresh (git) pillars though
08:02 eliasp fileserver.update won't refresh git pillars… at least here
08:02 topochan joined #salt
08:02 eliasp and by "bundling" I meant something like a "meta"-runner which calls a list of defined runners after another
08:02 eliasp but maybe I'll just quickly hack something together to do this
08:05 matthiaswahl joined #salt
08:10 ramishra joined #salt
08:11 babilen I have my pillars in git and they are being refreshed IIRC, but let me double check that
08:13 eliasp babilen: according to "ack -i pillar salt/fileserver' I'd say "no"…
08:14 babilen Yeah, I've just been reading the same source and can confirm your initial assumption.
08:14 babilen That's suboptimal. :(
08:15 eliasp it is… besides that, the git_pillar runner is rather ugly anyways… needs a branch + repo to be defined to work, while it should simply default to what's in the master config
08:15 babilen yeah
08:15 eliasp maybe I'll get around to rewrite it after the holidays, so in early Sept.
08:16 jhauser joined #salt
08:22 TheThing joined #salt
08:24 derito joined #salt
08:25 ika2810 joined #salt
08:26 che-arne joined #salt
08:30 Outlander joined #salt
08:30 Kraln joined #salt
08:30 giantlock joined #salt
08:33 TyrfingMjolnir joined #salt
08:34 matthiaswahl joined #salt
08:40 justlooks hi ,i test reactor system on salt 2014.1.5  but when i fire event on minion some iptables help info will input to terminal why?  https://gist.github.com/justlooks/eeb9c1c29beb46daa267
08:41 eliasp justlooks: that was a mistake in this release… is fixed in 2014.1.6
08:42 justlooks eliasp: but in epel ,the newest version still is 2014.1.5
08:43 eliasp justlooks: no idea regarding the EPEL repositories, sorry…
08:43 jayfk joined #salt
08:44 justlooks eliasp: where i can get 2014.1.6? from github? no rpm package?
08:45 eliasp justlooks: you'd have to build your own RPM based on the contents of https://github.com/saltstack/salt/tree/develop/pkg/rpm
08:46 eliasp justlooks: but it's maybe easier to just wait for 2014.1.6 to make it to EPEL and update then…
08:46 eliasp the iptables output is AFAIK completely uncriticial… it's just a bit more verbose
08:46 eliasp so you could simply ignore it until you get the update via EPEL
08:48 TheThing joined #salt
08:49 justlooks eliasp: ok ,thanks
08:53 jpihl_ joined #salt
08:54 jpihl joined #salt
08:58 badon joined #salt
08:59 xzarth_ joined #salt
09:00 yomilk joined #salt
09:04 badon joined #salt
09:04 thehaven_ joined #salt
09:05 mechanicalduck_ joined #salt
09:06 bluehawk joined #salt
09:06 esogas` joined #salt
09:06 yetAnotherZero joined #salt
09:07 yomilk joined #salt
09:07 justlooks how to write - arg: when i want reactor do more then one command?
09:08 justlooks than
09:09 eightyeight joined #salt
09:12 TheThing joined #salt
09:12 babilen justlooks: List multiple commands (each with its own tgt and arg)
09:22 ghartz joined #salt
09:23 agronholm left #salt
09:23 greyhatpython joined #salt
09:27 bhosmer joined #salt
09:41 saravanans joined #salt
09:46 justlooks i try to write "salt-call event.fire_master" command in my reactor sls file ,so it can fire second event when first event hit it ,but failed ,why ?  https://gist.github.com/justlooks/eeb9c1c29beb46daa267
09:58 che-arne joined #salt
10:00 babilen justlooks: Run the master in debug mode and check for errors when it gathers the reactions, might very well be some jinja escaping fuckup. Also: How does it fail?
10:02 justlooks babilen: i can not see /tmp/demo and /tmp/second file on saltslave ,so i think the second event did not fire successfully
10:03 babilen Yes, I guess that it is an running into a problem rendering that reactor which you should see if you ran the master in debug mode
10:04 saravanans joined #salt
10:05 babilen fwiw, you can do something like https://www.refheap.com/88496 to get around those jinja issues ... I frequently battle it and simply use Python whenever I have to handle data that I don't want jinja to tinker with
10:17 saravanans joined #salt
10:18 saravana_ joined #salt
10:18 Outlander joined #salt
10:20 CeBe joined #salt
10:22 saravanans joined #salt
10:25 saravana_ joined #salt
10:27 saravanans joined #salt
10:37 saravanans joined #salt
10:39 pviktori joined #salt
10:43 biffle joined #salt
10:45 biffle hello. Is there any way to have the scheduler target only 'a' node from a group? i.e have a schedule to target the database*.example.com boxes to run a backup script but have only one of the currently available boxes pick the messages from the queue and run the script? rather than them all trying to do it at once?
10:50 biffle and on a similar note, I'm having trouble getting minions to pick up new schedules configured in their pillar. salt '*' pillar.items refreshes their pillar data correctly (so it seems) with the updated schedule but the minions never execute. I have to restart the minion entirely before the schedule starts working. Am I missing a step?
11:01 deepz88 joined #salt
11:03 diegows joined #salt
11:04 deepz88 joined #salt
11:07 beardo joined #salt
11:07 poogles joined #salt
11:09 bhosmer joined #salt
11:16 bhosmer_ joined #salt
11:17 mechanicalduck joined #salt
11:18 higgs001 joined #salt
11:18 ika2810 left #salt
11:19 yomilk joined #salt
11:23 MZAWeb joined #salt
11:28 ggoZ joined #salt
11:33 yomilk joined #salt
11:36 m1crofarmer joined #salt
11:39 saravanans joined #salt
11:41 jhauser joined #salt
11:41 masterkorp In the gem state, how do i specify to install a specific version ?
11:43 viq masterkorp: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.gem.html
11:58 zandy joined #salt
12:00 yomilk joined #salt
12:02 jas- joined #salt
12:03 intellix joined #salt
12:07 scoates joined #salt
12:09 saravanans joined #salt
12:10 Theo-SLC joined #salt
12:11 tinytub joined #salt
12:13 masterkorp thank you
12:14 blarghmatey joined #salt
12:18 saravanans joined #salt
12:19 ndrei joined #salt
12:24 pclermont joined #salt
12:26 Nazzy file.mkdir and its friends are really eww
12:32 Nazzy they seem to lack decent returns and error handling
12:36 saravanans joined #salt
12:42 jrb28 joined #salt
12:48 stephanbuys anyone aware of a method to add grains to a salt-ssh roster endpoint (normally one would just add it to the minion conf file)
12:50 eliasp Nazzy: file.mkdir? you mean file.directory?
12:50 eliasp Nazzy: or do you mean the Python functions, not Salt states?
12:54 Debianmipfr joined #salt
12:54 Debianmipfr left #salt
12:55 cDR__ joined #salt
12:56 ndrei joined #salt
13:01 jslatts joined #salt
13:04 totte joined #salt
13:05 bhosmer joined #salt
13:08 bhosmer joined #salt
13:09 wendall911 joined #salt
13:12 Katafalkas joined #salt
13:12 FeatherKing joined #salt
13:16 racooper joined #salt
13:20 KGrange hey all, asked this yesterday, but no takers: what is your opinion on functions / execution modules that need to depend on a certain state?
13:21 babilen Please elaborate
13:23 anotherZero joined #salt
13:25 KGrange I'm making some execution modules meant for versions / deployment methods for a certain codebase. For example, I'd like to be able to do salt '*' codebasename.packaged to install the latest packaged version, salt '*' codebasename.git development to checkout the latest git commit on branch development, etc. I can do that fairly easily, but switching between the two requires some filesystem manipulation that
13:25 KGrange would be perfectly expressed using state modules
13:26 KGrange as of right now I just have some very long sls's with a lot of conditional jinja code... not exactly a clean way of doing it, and painful to allow variances between different minions
13:27 KGrange so, similar to how you can cross-call execution modules with __salt__["module.function"], is using __salt__["state.low"](pkg, installed, vim) (For example) considered good or bad practice?
13:28 XenophonF joined #salt
13:30 Nazzy eliasp, file.mkdir in the modules, not in the states
13:31 Nazzy https://github.com/saltstack/salt/blob/develop/salt/modules/file.py#L3226 <--- these 3 functions, specifically
13:32 logix812 joined #salt
13:34 dude051 joined #salt
13:34 hobakill joined #salt
13:37 ghartz joined #salt
13:39 KGrange so babilen what do you think?
13:40 katana_b1ade joined #salt
13:41 test_ joined #salt
13:41 babilen KGrange: I don't understand why you need an execution module for that
13:42 KGrange hmm, I should be making state modules, then?
13:42 babilen You can probably write the states in pure Python without having to turn that into an execution module.
13:42 babilen (or in pydsl)
13:43 KGrange Yeah, I should be doing that. D'oh!
13:43 babilen I think of execution modules as things that I want to call on minions from my states and states as those blocks that describe a certain state I want my minion to be in. The point of the former is essentially to be stateless (it is just called) and the point of the latter to be stateful (it is only called if necessary)
13:44 babilen jinja2 is horrible and complex things just get way too tricky with it, but luckily you can write states directly in Python if you want to do slightly more involved things.
13:44 KGrange Is there a good way of making states implicitly depend on other states? Or does every user of this module really have to make it require git: pkg.installed
13:45 babilen "implicitly" ?
13:45 KGrange like, before doing anything, it uses pre-existing salt states to make sure it can. I know that, for example, the built-in git states just fail on you, but I disagree with that
13:45 babilen If your execution module requires git then you should document that the user has to install it.
13:46 babilen But didn't we just agree that you don't want to use an execution module, but states for what you are trying to do?
13:46 KGrange yeah, I meant state module
13:46 mpanetta joined #salt
13:47 faeroe joined #salt
13:51 tligda joined #salt
13:54 hobakill Have people had much luck with chocolatey + salt?
13:55 hobakill because this seems ...y'know....wrong: http://paste.debian.net/111215/
13:56 smcquay joined #salt
13:58 oz_akan joined #salt
14:00 vejdmn joined #salt
14:01 quickdry21 joined #salt
14:04 zandy joined #salt
14:08 babilen KGrange: I consider it to be good practice for states that depend on other states to state that explicitly. You can get around it right now, but that would probably break state.id (which will be in Helium).
14:10 dude051 joined #salt
14:10 jalbretsen joined #salt
14:10 msil joined #salt
14:11 ipmb joined #salt
14:11 KGrange I'll stick to that, then. Thanks for the help!
14:11 KGrange Where is state.id, though? I can't find anything about it in http://docs.saltstack.com/en/latest/topics/releases/2014.7.0.html
14:12 ajprog_laptop1 joined #salt
14:13 ndrei joined #salt
14:15 hobakill re: chocolately, the bootstrap thing never seems to actually do what the syntax says it's going to do. the chocolatey dir never moves out of C:\Windows\TEMP
14:19 katana_b1ade joined #salt
14:21 hobakill hell, there isn't even a C:\ProgramData dir on Win2k12r2
14:24 babilen KGrange: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.state.html#salt.modules.state.sls_id
14:24 babilen state.sls_id not state.id (sorry)
14:34 Supermathie joined #salt
14:39 bmcorser joined #salt
14:43 cnelsonsic joined #salt
14:46 bombastic joined #salt
14:46 KGrange I don't quite understand how it would break that, but okay
14:48 poogles joined #salt
14:48 bombastic Does anybody have a favorite blog/video/example of implementing roles in salt ie role:develop role:webserver
14:48 thedodd joined #salt
14:49 Theo-SLC joined #salt
14:50 masterkorp Does salt-cloud has an API ?
14:54 vejdmn1 joined #salt
14:55 tkharju3 joined #salt
14:57 xerxas joined #salt
14:58 elfixit joined #salt
15:00 dude051 joined #salt
15:01 conan_the_destro joined #salt
15:04 kballou joined #salt
15:04 vejdmn joined #salt
15:08 rallytime joined #salt
15:09 smcquay joined #salt
15:10 eliasp masterkorp: http://docs.saltstack.com/en/latest/ref/clients/index.html#cloudclient
15:12 manfred masterkorp: have you seen salt-api yet? almost everything in salt can be used through salt-api
15:17 housl joined #salt
15:17 rypeck joined #salt
15:18 masterkorp manfred: http://docs.saltstack.com/en/latest/ref/netapi/all/index.html
15:18 masterkorp this ?
15:19 masterkorp i've been looking into it
15:19 manfred yes, that
15:19 manfred you can use the cloud.* modules and cloud.* states through the api
15:19 meteorfox joined #salt
15:20 masterkorp sweet thanks
15:23 seanz joined #salt
15:23 ecdhe joined #salt
15:26 dude^2 joined #salt
15:27 seanz joined #salt
15:30 derito joined #salt
15:33 higgs001 joined #salt
15:33 mpanetta joined #salt
15:34 mpanetta joined #salt
15:35 Katafalkas joined #salt
15:35 flupke is there a way to reload states without restarting the master?
15:36 eliasp flupke: what do you mean? states are available on the master from the moment on they're placed in the file roots or pulled from GitFS via fileserver.update…
15:39 Katafalkas joined #salt
15:39 dude051 joined #salt
15:53 beneggett joined #salt
15:55 bhosmer joined #salt
15:59 oz_akan joined #salt
16:01 seanz joined #salt
16:02 ndrei joined #salt
16:04 arthabaska joined #salt
16:06 chiui joined #salt
16:07 teskew joined #salt
16:09 jaimed joined #salt
16:10 KyleG joined #salt
16:10 KyleG joined #salt
16:10 poogles joined #salt
16:11 dude051 joined #salt
16:11 druonysus joined #salt
16:11 patarr how do you pros handle file locations like salt://x in your state files in case they move? Do you have some sort of external file or do you just hardcode them and not alter your state structure much?
16:12 tligda joined #salt
16:13 bdf_ patarr: those sorts of things are more appropriate for pillar variables
16:13 linjan joined #salt
16:13 bdf_ but what you can also do is return a default value in case pillar isn't defined.
16:14 patarr Ah thank you. Also, why does salt always cache state files and not purge them even after resending a modified state?
16:14 patarr I'm constantly having this problem and doing a state.clear_cache does not help
16:15 bdf_ {{ salt['pillar.get']('pkgs:apache', 'httpd') }}
16:15 bdf_ are you using git or svn or something?
16:16 patarr as a matter of fact I have the states in svn..
16:17 Gareth morning morning
16:17 patarr I committed it and I still have these caching issues.
16:18 schimmy joined #salt
16:20 bdf_ I think the idea is to not hit the repository too hard -- but I have similar problems with git
16:20 patarr I'm not getting any of the data from svn.
16:20 bdf_ what I ended up doing was just checking out a working copy on the salt master and leaving salt out of the interface with the SCM
16:21 patarr I'm jus thave my /srv/salt directory version controlled.
16:21 bdf_ oh
16:21 patarr And I just deleted my /var/cache/master directory and ruined my salt :D
16:22 TyrfingMjolnir joined #salt
16:22 patarr "Failed to authenticate, is this user permitted to execute commands?"
16:22 ajprog_laptop1 joined #salt
16:22 robawt wuddup Gareth
16:23 Theo-SLC joined #salt
16:23 Comradephate joined #salt
16:24 patarr Might I add, I am root.
16:25 schimmy joined #salt
16:26 patarr I've just reinstalled salt-master and I get the same error. Anyone have any ideas?
16:27 patarr And this is all that's in the log Authentication failure of type "user" occurred.
16:28 rypeck Try running as sudo
16:28 rypeck *with sudo
16:28 rypeck patarr:
16:28 patarr I'm root. I've always had no problems with running this stuff as root.
16:29 n8n joined #salt
16:31 patarr I just removed all /etc/ and /var/cache and reinstalled. But it's pretty scary that this could happen in production.
16:31 patarr I still don't know why salt has such persistent caching that doesn't get purged on state changes.
16:32 Theo-SLC joined #salt
16:36 rallytim_ joined #salt
16:36 rallytim_ left #salt
16:37 viq patarr: I guess it's mostly relevant for getting states from revision control, but still may be worth a try: 'salt-run fileserver.update'
16:38 rallytime joined #salt
16:40 MatthewsFace joined #salt
16:41 stephanbuys joined #salt
16:44 stephanbuys1 joined #salt
16:45 forrest joined #salt
16:45 oz_akan joined #salt
16:48 MatthewsFace joined #salt
16:49 joehillen joined #salt
16:51 Ryan_Lane joined #salt
16:51 hobakill has anyone played around with halite?
16:52 oz_akan joined #salt
16:52 rap424 joined #salt
16:53 patarr hobakill: still has lots of work to be done..
16:53 robawt halite == the web frontend right?
16:53 hobakill patarr: yeah i just have a question on the initial setup. POC stuff
16:53 oz_akan_ joined #salt
16:54 TyrfingMjolnir joined #salt
16:54 hobakill patarr: the instructions say to do a salt-call on the master for tls.create_self_signed_cert but there is no salt-call binary
16:54 patarr I believe i've found a bug in state.archive It says if you omit tar_options it uses the python tarfile module that supports gz and bz2. Salt just says tar archive need argument tar_options. Is this a versioning issue?
16:54 patarr hobakill: there is. But I think it might only be on the minion.
16:54 ipmb joined #salt
16:55 hobakill that's what i meant patarr - no salt-call binary on the master
16:55 patarr The guide may be assuming you're salt controlling the server that runs halite or something. Idk. Sounds weird to me.
16:55 troyready joined #salt
16:56 hobakill patarr: the reason i'm going down this path is that i don't really know how to do "reporting" in salt. just exploring options. thanks.
16:56 MatthewsFace joined #salt
16:58 patarr How do I access previous salt stack docs? There is a sidebar links that says "Docs for previous releases are available on salt.rtfd.org." but it just redirects to salt.readthedocs.org/en/latest
16:59 patarr aah. There is a formatting error. Perhaps my browser. All the way on the bottom of the page the versions appear in broken CSS formatting
17:00 Ahlee salt-call is a minion provided script
17:01 Ahlee salt-call runs on minions and reaches back to the master (unless configured to be standalone and provided the necessary base config for state resolution)
17:02 hobakill Ahlee: yep. however in the documentation for halite it says to run 'salt-call blahhh' on the master.
17:02 seanz retrospek: ping
17:02 hobakill running that same string using sall-run doesn't work since tls.create_self_signed_cert tls doesn't actually exist
17:03 Ahlee hobakill: I have not touched halite in a long time due to them tracking develop so closely
17:03 hobakill ahale: it would be nice if i could just get it started to take peak at. :)
17:03 Ahlee I haven't looked at it since January at salt-conf
17:03 Ahlee it's pretty, and impressive
17:03 hobakill 'netstat' doesn't even show anything listening on 8080 tho.... so.... i don't know. annoying.
17:04 Ryan_Lane howdy. we'd like to add tests for all the boto_* states and modules, but at minimum for integration tests we'll need to make calls to amazon, which will cost money
17:04 Ahlee so in this case, tls is an exectution state
17:04 Ryan_Lane who at SaltStack Inc should we talk to about this?
17:04 Ryan_Lane we'll probably need to make api calls for some of the unit tests too
17:05 Ahlee hobakill: what version of salt are you running?
17:05 hobakill 2014.1.5
17:05 Ahlee ok, lemme spin up a 2014.1.5 env
17:07 Ahlee yay not working from the office today and being able to reach github
17:08 carmony hrm, does anyone have any idea as why salt-call could take 11 seconds to just run salt-call test.ping ?
17:09 carmony here is a gist, and I've marked where it runs a long time: https://gist.github.com/JustinCarmony/b972d4df06c81d9ed7a2
17:11 acidrainfall joined #salt
17:11 manfred hobakill: 2014.1.5 has a lot of bugs
17:11 manfred just fyi
17:11 hobakill manfred: teriffic (
17:11 manfred 2014.1.7 is in epel-testing
17:12 Ahlee carmony: do you have the verify environment option set?
17:12 hobakill is chocolatey sucking balls one of the bugs?
17:12 Ahlee like, could it be just terribly slow disk IO getting to te cert?
17:12 eliasp hmm, file.directory is really slow applying permissions recursively … waiting since ~2h for a 750GB folder
17:12 carmony Ahlee: I'm not sure about the verify environment option, let me check
17:12 acidrainfall Anybody use Salt *with* puppet, rathehr than as a replacement?
17:13 Ryan_Lane acidrainfall: I have
17:13 Ryan_Lane then I replaced puppet
17:13 Ahlee carmony: I've seen this before (0.16 days I think?) I don't think it ever went away, I just never bothered looking for the solution
17:13 eliasp acidrainfall: you know this one?: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.puppet.html
17:13 Ryan_Lane I may be able to answer questions you have, though
17:13 vbabiy joined #salt
17:14 endersavage joined #salt
17:14 acidrainfall Ryan_Lane: Yeah largely I'm wondering from an experience standpoint how it worked out
17:14 Ahlee oh neat, a facter wrapper
17:15 Ryan_Lane it works ok depending on what you're going to do
17:15 Ryan_Lane if you're going to use salt for remote execution and events it should be fine
17:15 acidrainfall We have two separate infrastructures - one works just fine with the "check in whenever you want" methodology - the other is a clustered services ... cluster, and it's managed by an in-house (10 year old) SALT-like method.
17:15 jalaziz joined #salt
17:15 acidrainfall Yeah
17:15 acidrainfall That's what I want
17:15 Ryan_Lane yeah, use the facts custom grain
17:15 carmony Ahlee: let me try turning off verify env, I can cat the minion.pem without issue
17:15 Ryan_Lane then you can target by facts
17:16 Ryan_Lane don't try to integrate hiera
17:16 Ryan_Lane hiera's model is simply different
17:16 Ryan_Lane it won't work
17:16 Ahlee carmony: it's a long shot, but it's the only thing I can think of that salt would be doing at that point
17:16 n8n_ joined #salt
17:17 acidrainfall We don't currently use Hiera
17:17 acidrainfall A lot of what Hiera can do, we do with Foreman.
17:20 acidrainfall Ryan_Lane: Thanks
17:20 carmony Ahlee: looks like its an issue with our VM's networking being borked
17:22 beneggett joined #salt
17:24 n8n joined #salt
17:25 aw110f joined #salt
17:27 Ryan_Lane acidrainfall: yw
17:35 logix812 joined #salt
17:36 Ahlee carmony: ah, word
17:37 arthabaska joined #salt
17:42 zandy joined #salt
17:45 Eliz Is there a way to do file.recurse while preserving permissions?
17:45 ndrei joined #salt
17:45 mapu joined #salt
17:47 bradbeam_ joined #salt
17:48 patarr Eliz: looks like you can. Check out the docs http://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html
17:48 patarr Search for "recurse"
17:49 Eliz patarr: While it looks like you can tell it what permissions to set, I'd rather it looked at the files in the source and copied them appropriately.
17:49 Eliz Because not every file has the same permission set
17:50 kermit joined #salt
17:51 patarr Eliz: https://github.com/saltstack/salt/issues/2707
17:52 Eliz Well that's a shame.
17:52 ghartz joined #salt
17:53 ramteid joined #salt
17:57 Eliz I guess I'll just split up my stuff into several bits then.
17:57 layer3switch1 joined #salt
18:01 LordMike01 joined #salt
18:02 LordMike01 Hi there
18:02 LordMike01 Is it possible for me to generate files on one salt minion, based on all other salt minions in my setup?
18:02 LordMike01 Specifically, I want to generate a series of files in my monitoring server (munin), one for each other node in my network
18:04 Ahlee LordMike01: http://docs.saltstack.com/en/latest/topics/mine/
18:05 LordMike01 Cool
18:05 LordMike01 Mines seem to be targeted actual data fetched from minions
18:05 LordMike01 But is it possible to iterate just the hostnames?
18:09 saravanans joined #salt
18:09 acidrainfall okay another use case question
18:11 cnelsonsic joined #salt
18:12 acidrainfall Eh.  Let me look more before I ask it.  Nevermind.
18:12 zandy joined #salt
18:14 saurabhs joined #salt
18:15 Ahlee LordMike01: http://stackoverflow.com/questions/17158665/how-to-get-a-list-of-all-salt-minions-in-a-template
18:15 azylman joined #salt
18:16 LordMike01 Damn.
18:17 LordMike01 Nice, Ahlee
18:18 laubosslink joined #salt
18:19 druonysus joined #salt
18:19 druonysus joined #salt
18:20 bhosmer joined #salt
18:20 ajolo joined #salt
18:22 manfred laubosslink: it runs on the master, because pillars are rendered on the master
18:23 layer3switch joined #salt
18:23 roolo joined #salt
18:27 MZAWeb joined #salt
18:27 Ahlee If you need to execute a command on the minion, you're looking at a custom state (placed into _states in your environment root)
18:27 Ahlee laubosslink: http://docs.saltstack.com/en/latest/ref/states/writing.html#full-state-module-example
18:31 scoates in roots/top.sls, is there a way for me to say "* except [list]" ? e.g. I don't want my `apps` state on one specific node id.
18:33 bmatt scoates: see the docs for "compound matchers"
18:34 scoates bmatt: nice. thanks.
18:37 bhosmer_ joined #salt
18:38 vbabiy joined #salt
18:38 matthiaswahl joined #salt
18:39 SpeeR can I use grains in files that I'm pushing to servers? in this example syslog-ng.conf... I have different configs needs for different servers
18:44 lude SpeeR: use jinja templating for that
18:45 cpowell joined #salt
18:48 ecdhe joined #salt
18:48 bhosmer joined #salt
18:49 druonysus joined #salt
18:49 druonysus joined #salt
18:50 felskrone joined #salt
18:54 Corey What's the latest "Salt for Morons" tutorial / install guide?
18:55 hhenkel joined #salt
18:57 rypeck forrest: is your salt-book online anywhere?
18:57 forrest rypeck, no, I haven't finished it yet
18:58 rypeck ah k.
18:58 rypeck by the way - thanks a bunch for helping me yesterday.
18:58 forrest rypeck, you can look at what I have in the repo right now, but that's really only chapter 1, and part of chapter 2
18:58 forrest rypeck, yea np
18:58 rypeck I was able to make the leap and start getting productive.
18:58 forrest rypeck, awesome, always glad to hear that!
18:58 dude^2 joined #salt
18:59 forrest Corey, you can use chapter 1 and 2 from here if you want: https://github.com/gravyboat/salt-book/tree/master/en
18:59 forrest Corey, other than that it's just the main tutorial
19:01 banana_ joined #salt
19:02 SpeeR ls
19:05 hobakill forrest: nice. a book! do a kickstarter. i'll contribute for sure :)
19:05 forrest hobakill, nah I'd rather just work on it for free, just need to stop being a slacker. I have all my tasks ready to go in my kanban board, just need to work on it
19:06 forrest hobakill, it would be a rip off for you anyways since I wasn't planning on doing a Windows section :P
19:06 dude051 joined #salt
19:06 hobakill forrest: i wish i didn't have to have a windows environment to begin with - trust me!
19:06 forrest hah
19:06 smcquay joined #salt
19:07 hobakill forrest: literally nothing has worked right apart from the minion installation. i'm having a call with saltstack inc this week to discuss troubles.
19:07 forrest hobakill, Nice, are you talking to Dave?
19:07 forrest I wish he was in the IRC this week |
19:07 forrest *:\
19:07 hobakill but someone said 2014.1.5 was buggy. waiting for 2014.1.7 to hit epel stable
19:08 forrest he's probably at training or some other activity
19:08 forrest hobakill, yea at least use 2014.1.6
19:08 forrest I wish you weren't having so many issues, it looks bad and is frustrating :(
19:09 hobakill forrest: linux side has been flawless once i wrapped my brain around the basics. sadly we have 500+ win vm  to go with our 1k+ linux boxes. so - y''know - sadly windows is still alive and well
19:10 bhosmer_ joined #salt
19:10 bradbeam_ joined #salt
19:11 biffle is there a list of bugs in 104.1.5 - or a release log for later releases/
19:11 KyleG bugs are tracked in Github
19:11 KyleG https://github.com/saltstack/salt/issues?direction=desc&amp;sort=created&amp;state=open
19:11 biffle I'm having scheduling troubles - wonder if it's related, I'm also running 2014.1.5
19:11 forrest hobakill, yea but it should work, and it has worked in the past
19:11 forrest so the fact it isn't is quite annoying
19:11 babilen biffle: There are release notes at http://docs.saltstack.com/en/latest/topics/releases/
19:12 biffle thanks!
19:12 zandy joined #salt
19:12 hobakill forrest: i created an issues for my chocolatey issue but now i'm thinking they'll just have me update to > 2014.1.5 as a start.
19:12 babilen .5 has well known problems communicating (not sure what the issue # is though and I didn't see it in the release notes, an upgrade to .7 would be recommended by me (but then you are probably using pesky rhel or other rpm based distro))
19:13 peters-tx joined #salt
19:13 hobakill 2014.1.7 has been in EPEL6-testing for nearly the required 2 weeks. i'd expect to see it soon in stable. maybe before end of week.... /me fingers crossed
19:14 babilen It should have never made it into stable, how did that happen?
19:14 hobakill babilen: 2014.1.5? i have no clue
19:15 thehaven joined #salt
19:17 babilen When will Helium be released? Given that it has been decided that it will be 2014.7.0 there aren't really that many days left. (Or might it still become 2014.8 / 2014.9 ?)
19:18 rypeck forrest: sorry to be making puppet comparisons - but in puppet I can manage settings for an environment/set of nodes. What's the equivalent in salt?
19:18 rypeck or anyone else!
19:18 jcockhren babilen: the date isn't so much of a indication of release month. it's a indication of what month the new release became RC
19:19 hobakill rypeck: you can do that using salt state environments. http://docs.saltstack.com/en/latest/ref/states/top.html
19:20 rypeck thanks hobakill
19:20 babilen jcockhren: Oh, I thought it would be the month of the release. Thanks for clarifying that!
19:20 rypeck ah I see. I had kinda glazed over that whole file thus far...
19:20 rypeck thanks!
19:20 rypeck hobakill: can I change pillar facts with that file?
19:20 rypeck or would I have to separate that out in a pillar?
19:22 hobakill pillar_roots != file_roots so i think you'd have to do something else. i haven't done much with pillars myself.
19:24 hobakill but if i'm honest rypeck what i don't know about pillars could fill an ocean. mostly working with states.
19:26 forrest rypeck, there are a few ways to handle it, take a look at http://docs.saltstack.com/en/latest/topics/pillar/#declaring-the-master-pillar
19:26 cnelsonsic joined #salt
19:34 kermit joined #salt
19:37 SpeeR can someone tell me why the changes I've added to the source file aren't updating? http://pastebin.com/raw.php?i=vx5yh6xQ
19:37 SpeeR I have the roles specified in the minion's config
19:39 TheThing joined #salt
19:40 smcquay_ joined #salt
19:40 saravanans joined #salt
19:43 renderer24 joined #salt
20:01 nlb joined #salt
20:03 dude051 joined #salt
20:03 oz_akan joined #salt
20:12 to_json joined #salt
20:12 zandy joined #salt
20:13 saurabhs left #salt
20:13 m0nky joined #salt
20:14 whytewolf joined #salt
20:14 zandy joined #salt
20:15 layer3switch1 joined #salt
20:16 smcquay_ joined #salt
20:18 bradbeam_ joined #salt
20:18 eightyeight i'm managing ssh keys with salt, and there are extra line breaks in the ~/.ssh/authorized_keys file. how can i eliminate these extra line breaks?
20:18 acidrainfall left #salt
20:18 bradbeam_ left #salt
20:18 seanz joined #salt
20:20 renderer24 joined #salt
20:20 smcquay joined #salt
20:25 eightyeight here is a pastebin of my salt config, and the result: http://ae7.st/p/9fp
20:26 eightyeight the shuffling is strange, but meh. but the extra line breaks are just annoying
20:34 szicari joined #salt
20:37 jaimed joined #salt
20:41 ajolo joined #salt
20:43 stevednd eightyeight: That's how it has always been. Same thing in mine. I agree it's a bit annoying
20:44 smcquay_ joined #salt
20:45 eightyeight ugh
20:45 eightyeight it works, so meh. but where are the extra newlines coming from?
20:46 smcquay_ joined #salt
20:46 hhenkel joined #salt
20:47 smcquay joined #salt
20:47 kermit joined #salt
20:49 poogles joined #salt
20:51 stevednd eightyeight: my guess is somewhere around here https://github.com/saltstack/salt/blob/develop/salt/modules/ssh.py#L573
20:52 stevednd the other places where it actually adds the key, it does a writeline()
20:52 stevednd so maybe something where it screws up, and is doubling up the newlines at the end of the file
20:55 thehaven joined #salt
20:56 EWDurbin joined #salt
21:05 druonysus joined #salt
21:12 zandy joined #salt
21:14 n8n_ joined #salt
21:14 smcquay joined #salt
21:15 layer3switch joined #salt
21:18 druonysuse joined #salt
21:26 n8n joined #salt
21:26 ghartz joined #salt
21:28 n8n__ joined #salt
21:29 kermit joined #salt
21:36 alekibango joined #salt
21:44 pfallenop joined #salt
21:48 dimeshake_ joined #salt
21:51 whytewolf is there any way to do a file.manage_file with windows. [I'm not a windows guy but i need baby steps to get the windows guys out of using cmd.run [with out the drastic leap to states that they can't seem to grasp either]
21:54 forrest whytewolf, you already tried http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.file.html#module-salt.modules.file ?
21:55 whytewolf thats what got me file_managed
21:55 whytewolf err manage_file i mean
21:56 whytewolf which works great in linux. but is blowing up for windows
21:57 whytewolf TypeError encountered executing file.manage_file: get_user() got an unexpected keyword argument 'follow_symlinks'. See debug log for more info.
22:02 forrest whytewolf, ahh yea I am not sure on that one, sorry
22:02 whytewolf humm, I wonder if i can use cmd.run to run salt-call -l debug [I don't have access to the box directly only through salt] with out causing a rift in space
22:05 whytewolf wholey crap that worked. "Passed invalid arguments: manage_file() got multiple values for keyword argument 'name'"
22:05 forrest why are the windows guys adverse to using states?
22:05 forrest they make things so amazing
22:06 whytewolf I have no idea. my life would be so much simpler if they would.
22:06 whytewolf but they have this idea of running scripts, that download other scripts that they then run.
22:06 whytewolf and abou that time i get an eye twich
22:09 bradbeam_ joined #salt
22:12 zandy joined #salt
22:22 dstokes would it be crazy to write a salt runner to manage multi-minion code deployments and auto-rollback on error?
22:23 jeremyBass1 joined #salt
22:24 laubosslink joined #salt
22:25 forrest dstokes, sounds cool to me
22:25 druonysuse joined #salt
22:25 druonysuse joined #salt
22:27 jeremyBass1 joined #salt
22:30 forrest laubosslink, http://docs.saltstack.com/en/latest/ref/states/all/salt.states.cmd.html#salt.states.cmd.wait
22:31 oz_akan joined #salt
22:33 dstokes forrest: should ./runners/runner.py work for creating new runners, or do you need master config to specify a dir
22:33 stevednd dstokes: You mean something like orchestration?
22:33 dstokes in other words, is there a default runners dir like like _states + _grains
22:34 forrest dstokes, http://docs.saltstack.com/en/latest/ref/configuration/master.html#runner-dirs
22:34 dstokes so that's required, got it
22:35 dstokes stevednd: sort of. i need deployments to be managed at the master level. a single minion failure needs to rollback all minions that succeeded
22:36 stevednd dstokes: yeah, orchestration gets you up to the point of failure. Right now I just have a second orchestration script that reverses the process if things fail
22:36 ajprog_laptop1 joined #salt
22:36 dstokes stevednd: you talking about overstate?
22:37 jamesbdev joined #salt
22:37 forrest laubosslink, np
22:37 stevednd dstokes: the orchestration runner is the newer fancier version of overstate
22:38 stevednd http://docs.saltstack.com/en/latest/topics/tutorials/states_pt5.html#the-orchestrate-runner
22:38 dstokes stevednd: i'm hoping for automated error handling, which this almost gets me..
22:39 n8n joined #salt
22:39 dstokes stevednd: i assume you're running the error orchestration manually
22:39 stevednd dstokes: yeah, nothing built in has any automated error handling. It would be a great addition to the orchestrate runner
22:39 stevednd yeah, I just use a separate salt-run call
22:40 dstokes yeah, that's my problem. error handling for ci is kinda non-existent atm
22:40 dstokes w/ a runner i can capture minion errors and handle accordingly (afaik)
22:45 Outlander joined #salt
22:45 stevednd dstokes: I wonder if this new functionality in helium  could be used inside the runners http://docs.saltstack.com/en/latest/ref/states/requisites.html#onfail
22:46 tkharju4 joined #salt
22:48 stevednd I believe it's a global addition, so I'm thinking that it should, but I haven't tried yet
22:52 bhosmer joined #salt
22:55 alekibango joined #salt
22:56 whytewolf ...  i might have talked them into states. for a least the file management.
22:56 forrest did you show them how easy it was?
22:57 mosen joined #salt
22:57 whytewolf apperently their big gripe was they didn't understand pillar. so when i posted a pillar based example at least one of them started to at least get a glimmer
22:58 Luke joined #salt
22:58 ldlework pillar is amazing
22:59 whytewolf coarse it wasn't a large jinja heavy example that i normally use which is my logstash state formula that i am working on
23:00 whytewolf [coarse that one also sucks currently]
23:00 savvy-lizard joined #salt
23:00 ldlework I still don't understand why most people I meet like how ansible is yaml, where the yaml strings are templated
23:00 ldlework instead of, y'know, the entire damned yaml file
23:00 ldlework blows my mind
23:01 dstokes stevednd: i thought about that too. i'm building the app on each node (for now) so there's too many pts of failure for onerror to be useful
23:01 dstokes i was thinking tho, that i could use orchestration to run 'build' and then 'release' states so that apps are only reloaded if builds succeed on all noes
23:01 dstokes nodes*
23:01 whytewolf ldlework: cause people like buzz words? and the more things they can fit in the more buzz words they can use?
23:02 ldlework like looping in Anisble is just so awkward
23:02 ldlework whytewolf: I think its more momentum and Salt is patently hard to understand
23:02 whytewolf i wouldn't know. never touched anisble myself
23:03 ldlework Salt committed seppuku by naming its things grains and pillar and so on
23:03 ldlework As soon as I start trying to explain how Salt works, people's brains turn off
23:03 ldlework So I literally never mention the words Grain or Salt, ever, except behind closed doors with only actual Salt users around.
23:03 ldlework err Grain or Pillar
23:04 Blacklite "pass the salt"
23:04 Blacklite "to what?"
23:05 whytewolf I have thought about writing a Salt module library called Pepper. just to mess with people.
23:05 ldlework I have thought about forking the documentation and rewriting it to say Attributes and Context
23:05 forrest that sounds awful
23:05 ldlework because god forbid, we speak to our audience
23:05 manfred whytewolf: libpepper is out there
23:06 manfred it is a command line salt-api client iirc
23:06 ldlework forrest: yeah and Grain and Pillar are just so intuitive
23:06 manfred https://github.com/saltstack/pepper
23:06 ldlework I can see how reusing words that actually mean the thing they aim at is 'awful'
23:06 forrest ldlework, I never said they were :P
23:06 yomilk joined #salt
23:07 whytewolf huh, thanks manfred. didn't even notice that before.
23:07 manfred ldlework: so... this is what i have been doing today https://github.com/gtmanfred/salt/blob/inventories/salt/roster/ansible.py
23:07 ldlework I'm not exactly sure what that's doing
23:07 manfred it is reading your ansible inventory and using it to generate a salt roster
23:07 manfred for salt-ssh
23:08 oz_akan joined #salt
23:08 manfred and when I am done with it, salt will have better documentation about ansible dynamic inventories than ansible has, cause ansible's is attrocious
23:09 m1crofarmer joined #salt
23:10 ajolo joined #salt
23:12 zandy joined #salt
23:17 MZAWeb joined #salt
23:17 schmutz joined #salt
23:19 rypeck Hmmm - is there a way to run a command before a file gets changed - and only if it will be changed?
23:19 rypeck Previous work around in puppet was to make the file, see if it was different form the existing one, run the command and replace the file.
23:21 forrest rypeck, I don't think there is unfortunately.
23:21 rypeck So I'll just do the same work around then.
23:21 forrest yea probably
23:21 rypeck No biggie - it seems to be a rare use case.
23:21 rypeck Unfortunately I'm managing a service that needs the configuration files to remember how many processes it started.
23:21 forrest yea most people want to trigger commands after files are updated, not before only if the file is updated
23:22 forrest rypeck, ahh that's... weird
23:22 rypeck If you change it before stopping - you have random process that gather around.
23:24 manfred rypeck: you can pass test='' to the state to see if the state is going to change the file
23:24 joehillen joined #salt
23:24 manfred actually
23:24 manfred that won't work
23:24 manfred because that is passed on the commandline
23:24 rypeck manfred: yea that's what I thought looking at this - http://docs.saltstack.com/en/latest/ref/states/testing.html
23:25 manfred http://docs.saltstack.com/en/latest/ref/states/all/salt.states.test.html
23:25 manfred nm
23:25 manfred that doesn't do it either :/
23:25 rypeck I'll have it create a file - do a comparison - and a command either to stop the service if it differs then replace it - then start again.
23:25 rypeck Or if it doesn't I'll delete the file.
23:25 manfred rypeck: the only way, may be to use the onlyif and unless statesments that are now available on every state in 2014.7
23:26 manfred yeah, that would work
23:26 rypeck Yea executing commands. No way to run a command before salt goes to change a file.
23:26 rypeck As I understand it. Still a noob.
23:26 Ryan_Lane manfred: what salt-ssh needs more than anything is cloud inventories
23:26 Ryan_Lane I'd bet very few people use file inventories
23:29 rypeck manfred: forrest thanks again! saving me hours of poring over documentation.
23:29 DaveQB joined #salt
23:29 forrest rypeck, yea np
23:31 aquinas joined #salt
23:32 m1crofarmer joined #salt
23:32 m1crofarmer joined #salt
23:33 dstokes anyone know if it's possible to include / extend in salt-run files
23:33 dstokes include seems to work but extend is failing
23:33 m1crofarmer joined #salt
23:34 m1crofarmer joined #salt
23:35 m1crofarmer joined #salt
23:36 m1crofarmer joined #salt
23:36 m1crofarmer joined #salt
23:37 m1crofarmer joined #salt
23:38 m1crofarmer joined #salt
23:39 m1crofarmer joined #salt
23:40 ajolo joined #salt
23:40 m1crofarmer joined #salt
23:41 m1crofarmer joined #salt
23:42 m1crofarmer joined #salt
23:43 m1crofarmer joined #salt
23:43 m1crofarmer joined #salt
23:44 m1crofarmer joined #salt
23:46 Luke joined #salt
23:46 m1crofarmer joined #salt
23:46 FeatherKing joined #salt
23:47 m1crofarmer joined #salt
23:47 m1crofarmer joined #salt
23:48 m1crofarmer joined #salt
23:49 che-arne|2 joined #salt
23:50 darrend joined #salt
23:50 m1crofarmer joined #salt
23:50 FeatherKing is there a way to salt bootstrap a specific minion version? i have minions pulling 2014.1.7 and my master only has 2014.1.5 available in EPEL
23:51 m1crofarmer joined #salt
23:51 manfred FeatherKing: i believe you have to pull from get to do that
23:51 m1crofarmer joined #salt
23:52 manfred curl -sL https://bootstrap.saltstack.com | sh -s -- git v2014.1.5
23:52 m1crofarmer joined #salt
23:52 FeatherKing looks like right now we are curl -L https://bootstrap.saltstack.com -o /home/lspeed/install_salt.sh
23:52 FeatherKing the non one liner
23:53 FeatherKing so
23:53 m1crofarmer joined #salt
23:53 forrest yea I have examples of how to do that on the bootstrap page: http://docs.saltstack.com/en/latest/topics/tutorials/salt_bootstrap.html
23:54 manfred ./install_salt.sh git v2014.1.5
23:54 m1crofarmer joined #salt
23:54 FeatherKing yeah looking now
23:54 FeatherKing is 1.7 submitted to EPEL?
23:54 FeatherKing for the master
23:54 forrest I don't think so
23:54 forrest oh 2014.1.7
23:54 forrest it might be
23:54 forrest probably isn't through their process yet
23:54 m1crofarmer joined #salt
23:55 FeatherKing yeah 2014.1.7, i didnt know we had an issue until i went to run a highstate
23:55 FeatherKing highstate was failing with newer minions, but state.sls worked
23:55 FeatherKing so i never ran into it
23:55 manfred it is in epel-testing
23:55 m1crofarmer joined #salt
23:56 m1crofarmer joined #salt
23:57 m1crofarmer joined #salt
23:57 FeatherKing if i re bootstrap with 1.5 will it remove 1.7?
23:58 m1crofarmer joined #salt
23:58 manfred unless you use the system package, i do not believe so
23:58 manfred i think you will just have to python eggs installed
23:58 manfred and it should only run the one you installed most recently
23:59 m1crofarmer joined #salt
23:59 m1crofarmer joined #salt

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