Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-10-03

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

All times shown according to UTC.

Time Nick Message
00:01 n8n joined #salt
00:07 BrendanGilmore joined #salt
00:13 TyrfingMjolnir joined #salt
00:15 jalaziz joined #salt
00:19 jensnockert joined #salt
00:19 oz_akan joined #salt
00:20 kickerdog joined #salt
00:20 jalaziz joined #salt
00:28 bhosmer joined #salt
00:36 kickerdog joined #salt
00:37 desposo joined #salt
00:38 n8n joined #salt
00:45 FeatherKing joined #salt
00:45 CF903506 joined #salt
00:47 kingel joined #salt
00:49 Gareth joined #salt
01:03 renoirb joined #salt
01:04 viq joined #salt
01:05 TyrfingMjolnir joined #salt
01:05 ndrei joined #salt
01:06 glyf joined #salt
01:13 otter768 joined #salt
01:37 possibilities joined #salt
01:41 malinoff joined #salt
01:45 jalaziz joined #salt
01:45 BrendanGilmore joined #salt
01:49 TyrfingMjolnir joined #salt
02:02 toastedpenguin joined #salt
02:04 possibilities joined #salt
02:05 jalaziz_ joined #salt
02:07 jalaziz joined #salt
02:08 jensnockert joined #salt
02:17 bhosmer joined #salt
02:17 __number5__ anyone knows why salt pip module prefer pip2 over pip?
02:18 __number5__ latest pip package (1.5.5/1.5.6) not updating pip2 script at all
02:19 possibilities joined #salt
02:27 dccc_ joined #salt
02:29 BrendanGilmore joined #salt
02:32 pssblts joined #salt
02:36 sroegner_ joined #salt
02:36 kingel joined #salt
02:41 sag47 joined #salt
02:41 sag47 What's the RHEL7 support like for salt at this time?
02:42 SheetiS1 I think that it's in epel7
02:43 SheetiS1 https://github.com/saltstack/salt/issues/10214 shows the issue as closed and that packages should be there.
02:59 quanta_ joined #salt
03:00 mechanicalduck joined #salt
03:02 possibilities joined #salt
03:05 sag47 Thanks SheetiS1, I'll check it out.
03:06 bhosmer joined #salt
03:27 semarie joined #salt
03:32 baconbeckons_ joined #salt
03:36 possibilities joined #salt
03:42 dccc_ joined #salt
03:53 N3X15 joined #salt
03:53 dlam joined #salt
03:54 N3X15 Exception must be string or buffer, not dict occurred in file server update <-- Any ideas why this would occur (from salt-master -l debug)?  Debian Jessie, gitfs using GitPython, salt-master 2014.1.10
03:57 jensnockert joined #salt
03:57 quanta_ left #salt
03:57 N3X15 http://pastebin.com/Bg2vQCv4 Master config looks like this, by the way.
04:03 baconbeckons joined #salt
04:04 __number5__ N3X15: I think per remote config is only in 2014.7.0+ http://docs.saltstack.com/en/latest/topics/tutorials/gitfs.html#per-remote-configuration-parameters
04:05 N3X15 Crap, so it has to be a public repo?
04:06 bhosmer joined #salt
04:09 heewa joined #salt
04:09 __number5__ don't have to be a public repo, just that you can't put privkey/pubkey params under specific remote, it can only be global
04:09 baconbeckons_ joined #salt
04:11 __number5__ N3X15: check this http://salt.readthedocs.org/en/v2014.1.4/topics/tutorials/gitfs.html which should work for 2014.1.10
04:11 yomilk joined #salt
04:14 mafro joined #salt
04:14 mafro lo does anyone know how I can add to the default salt-minion config as deployed by salt-cloud?
04:16 mafro nm
04:16 mafro was being a thicky thickerson
04:18 possibilities joined #salt
04:21 mafro joined #salt
04:22 Katafalkas joined #salt
04:25 kingel joined #salt
04:26 blackhelmet joined #salt
04:29 ramishra joined #salt
04:40 NotreDev joined #salt
04:41 ramishra joined #salt
04:42 possibilities joined #salt
04:43 N3X15 Yeeeah, now it's not picking up the SSH key.
04:44 N3X15 Even though it is in ~/.ssh
04:45 N3X15 Or it might be because I'm an idiot and didn't add it to authorized_keys.
04:48 SheetiS File and directory permissions are important for ~/.ssh as well
04:53 N3X15 No, pretty sure it was because I didn't add it to authorized_keys, SheetiS. :V
04:53 sag47 joined #salt
04:54 semarie joined #salt
04:54 N3X15 Because, for one, it's working now.
05:00 ramishra joined #salt
05:07 SheetiS N3X15: that can do it as well :D
05:09 N3X15 Although I'm concerned that state.highstate is hanging now.
05:19 war2 joined #salt
05:22 mafro joined #salt
05:25 N3X15 There, was some weird PKI corruption.
05:26 masm joined #salt
05:37 catpiggest joined #salt
05:40 CaptinHokk joined #salt
05:40 kingel joined #salt
05:45 yomilk joined #salt
05:45 jensnockert joined #salt
05:46 war2 joined #salt
05:46 flebel joined #salt
05:47 grove_ joined #salt
05:47 n8n joined #salt
05:52 fejese joined #salt
05:54 bhosmer joined #salt
05:56 mafro left #salt
06:03 kingel joined #salt
06:11 tinuva joined #salt
06:16 joehh back again!
06:16 joehh anyone seen problems with windows minions being unable to deserialise msgpack messages?
06:19 troyready joined #salt
06:19 schimmy joined #salt
06:21 astol joined #salt
06:22 schimmy1 joined #salt
06:24 mechanicalduck_ joined #salt
06:25 flyboy joined #salt
06:27 aquinas joined #salt
06:28 kingel joined #salt
06:35 kickerdog joined #salt
06:41 lcavassa joined #salt
06:47 nicolasbadia__ joined #salt
06:48 swa_work joined #salt
06:51 astol1 joined #salt
06:58 kickerdog joined #salt
06:59 CycloHex joined #salt
07:01 CycloHex Is it possible to give custom grains dynamic values so that If I deploy two minions one has the grain roles:web, the other roles:database for example
07:01 kingel joined #salt
07:03 Sweetshark joined #salt
07:04 kickerdog joined #salt
07:07 glyf joined #salt
07:07 war2 joined #salt
07:08 jensnockert joined #salt
07:14 jhauser joined #salt
07:17 yomilk joined #salt
07:17 vnico left #salt
07:17 yomilk joined #salt
07:18 Ryan_Lane joined #salt
07:33 wnkz joined #salt
07:39 wnkz__ joined #salt
07:41 N-Mi joined #salt
07:41 wnkz joined #salt
07:43 bhosmer joined #salt
07:44 war2 joined #salt
07:44 ndrei joined #salt
07:44 wnkz__ joined #salt
07:46 zerthimon joined #salt
07:47 wnkz joined #salt
07:48 vukcrni joined #salt
07:48 wnkz___ joined #salt
07:49 flyboy82 CycloHex, grains are supposed to be static info about a minion. If you want to use arbitrary data regarding minions, you're better off utilizing the mine
07:49 CycloHex ok, the closest I found to deploy 2 minions that differ, in one command
07:49 CycloHex the mine?
07:50 flyboy82 http://docs.saltstack.com/en/latest/topics/mine/
07:50 CycloHex flyboy82, ok, thanks.. I'll look in to 'the mine'
07:50 CycloHex ye, jsut found it..
07:51 CycloHex tried googling the mine salt :D haha no succes.. the mine salstack on th eother hand got me the link you specified ^^
07:51 CycloHex thanks, flyboy82
07:51 flyboy82 yea, same thing happend to me the first time... 'salt mine' search string returns actual salt mines in different parts of the world :S
07:52 wnkz joined #salt
07:55 __number5__ salt pillar doesn't help either XD
07:56 darkelda joined #salt
07:56 darkelda joined #salt
07:57 wnkz__ joined #salt
08:00 CycloHex xD
08:01 CycloHex salt pillar works for me though __number5__  but google knows me.. and lately 80% of my searches involve saltstack :d
08:02 wnkz joined #salt
08:05 fredvd joined #salt
08:07 wnkz__ joined #salt
08:10 Gnouc joined #salt
08:12 Gnouc joined #salt
08:12 Gnouc joined #salt
08:16 martoss joined #salt
08:17 tld_wrk joined #salt
08:18 agend joined #salt
08:23 jhauser joined #salt
08:24 che-arne joined #salt
08:25 tmmt a
08:27 ramishra joined #salt
08:30 intellix joined #salt
08:36 superted666 joined #salt
08:39 agend joined #salt
08:51 ramishra joined #salt
09:01 CycloHex Anyone here with experience with salt mine? I tried using it to add specific attributes to my already deployed minion... Should I better use grains for it, or try the mine with it's functions
09:04 darkelda joined #salt
09:04 istram joined #salt
09:05 viq CycloHex: mine is just a cache, you'd need to add stuff to eg grains first
09:05 CycloHex ah, ok, thanks!
09:05 adsisco joined #salt
09:06 adsisco http://pastebin.com/8rFNkYD1 , how should i rephrase the file.managed to prevent a state conflict?
09:06 viq adsisco: you need to have two (or more) top level IDs for that, you can't have multiple file statements in a single ID
09:07 adsisco so something like wrapfile: file.managed: - blah blah blah works?
09:08 adsisco sorry, i'm kinda new to salt heh
09:08 viq adsisco: http://pastebin.com/vBcSEp1z
09:08 adsisco i see ... thanks bro!!
09:08 viq though I'd probably change modes to 555 or something like that, 777 is usually _not_ a good idea
09:08 antonw joined #salt
09:09 adsisco lol yea, i agree
09:09 viq also, it will run those scripts every time you run this state
09:10 viq If you want to install redis on multiple machines, it could be better to play with something like fpm, and create your own package for it
09:11 adsisco fpm?
09:11 viq https://github.com/jordansissel/fpm/wiki
09:11 adsisco nice, i'll look into it, thanks again
09:12 grove_ joined #salt
09:17 * babilen never used 555 ever
09:23 viq 755 is more popular, indeed. But, shrug, if it's managed by salt anyway...
09:28 istram 750 not a favorite anymore?
09:32 bhosmer joined #salt
09:34 linjan joined #salt
09:34 ndrei joined #salt
09:37 xmj for what, actually?
09:38 darkelda joined #salt
09:38 darkelda joined #salt
09:52 babilen xmj: file permissions
09:53 xmj well, usually file perms serve a *purpose* :-)
09:55 kingel joined #salt
09:59 viq xmj: no, they serve a Master, us!
10:00 * xmj retreats to rabbit hole
10:00 viq Yes, follow the white rabbit
10:00 darkelda joined #salt
10:01 kingel joined #salt
10:13 fejese joined #salt
10:19 ndrei joined #salt
10:19 TyrfingMjolnir joined #salt
10:20 giantlock joined #salt
10:30 sarkis joined #salt
10:30 fredvd joined #salt
10:40 fredvd_ joined #salt
10:41 bhosmer joined #salt
10:42 TheThing joined #salt
10:48 elfixit joined #salt
10:48 jensnockert joined #salt
10:48 glyf joined #salt
10:49 saltnoob joined #salt
10:50 saltnoob Hi Guys, I have setup isolated environments and can use 'include' to include states from say a base environment but am having trouble doing the same from an environment pillar top file, just wondering if this is something that's not possible?
10:52 saltnoob tried this - include:   - base: users and get error render error: sequence item 1: expected string, OrderedDict found
10:52 CycloHex joined #salt
10:53 saltnoob if I remove the - from "- base" I don't get any error but also the data is not loaded
10:57 viq saltnoob: include is not for top.sls
10:58 saltnoob it works for local sls files
10:58 saltnoob but want to reference files from another environment
10:59 viq saltnoob: top.sls is not a normal sls file
11:00 saltnoob I have isolated environments
11:00 saltnoob I need a top file for each file and pillar root
11:00 saltnoob for states / file roots I can 'include'
11:01 saltnoob appears you can do the same for pillar roots - http://docs.saltstack.com/en/latest/topics/pillar/#including-other-pillars
11:01 saltnoob but not sure if we can pull from other environments
11:02 viq I think you can't, but I don't feel fully qualified to answer that
11:03 fe92 joined #salt
11:03 CycloHex pulling from other environments, don't you have to put a ?env=dev
11:04 CycloHex behind the file that hsould be pulled fomr the environment
11:05 CycloHex - salt://foo.conf?saltenv=dev
11:05 saltnoob not sure what you mean
11:05 saltnoob oh
11:05 CycloHex not sure if it works with sls though.. this is for the fil state
11:05 CycloHex file*
11:05 saltnoob I'm not copying the file to the minion
11:05 saltnoob purely data in pillar for use in states and formula's
11:06 CycloHex Why have you split up into environments?? I had it setup as well, but went back to jus tone environment, because I dnd't really get the point of the environments
11:06 johtso joined #salt
11:06 ramishra joined #salt
11:07 jdmf joined #salt
11:07 saltnoob want to ensure minions on one environment can't read sensitive data in other environments should it get compromised
11:07 saltnoob e.g. passwords and such in pillar
11:07 viq saltnoob: environments make no difference here
11:07 viq Either you can access pillars, or you can't
11:08 saltnoob when they are isolated they can only access it's predefined pillar root
11:08 viq I'm not sure what you mean by "isolated"
11:08 saltnoob each environment has their own roots (file and pillar)
11:08 saltnoob and their own top.sls files
11:09 saltnoob environment is hardcoded on the minion
11:09 saltnoob e.g. environment: dev
11:09 scottpgallagher joined #salt
11:09 viq In case of a compromise it doesn't stop someone from changing that
11:09 jensnockert joined #salt
11:10 saltnoob good point
11:10 saltnoob what if resolve that by setting it from the master instead
11:11 viq Then you have control on the master what pillars the minion gets, and it doesn't matter what environment it's in
11:11 viq (or does, for convenience, etc)
11:11 saltnoob reason I went down this road of isolated environments..
11:12 saltnoob is because when trying to re-use formula's
11:12 saltnoob e.g. say a user formula
11:13 saltnoob it wasn't obvious in other environment states that it will run (without explicitly calling the formula) if you had the same pillar data in another environment
11:13 saltnoob e.g. for the users-formula
11:13 viq I don't follow
11:13 saltnoob if I add users.sls pillar data in the base, and call the state then new users get applied
11:14 saltnoob now in another environment, without explicitly calling the state, if I just drop another users.sls in the pillar root it would add those users too
11:14 saltnoob so for clarity I wanted to split into environments
11:15 saltnoob cos salt combines all the pillar data
11:15 viq Huh? How will get users added without explicitly calling state?
11:15 saltnoob but doesn't in isolated environments
11:15 saltnoob exactly
11:15 saltnoob as a test you can try the 'users-formula'
11:15 TheThing joined #salt
11:15 saltnoob so you lets say you add users.sls with a bunch of users in the base pillar
11:15 saltnoob you also call the formula from the base
11:15 saltnoob now
11:16 saltnoob in another environment, e.g. prod
11:16 gmcwhistler joined #salt
11:16 saltnoob add another users.sls in the pillar root
11:16 saltnoob and it will add new users
11:16 saltnoob without explicitely calling the state from the prod environment
11:16 viq You can assign the state (formula) only to some nodes. You can give various values of that pillar to different groups of nodes. Environments is one way to do it, but there are others
11:16 saltnoob salt seems to compile all the 'users.sls' pillar data
11:17 viq (please don't use enter instead of comma)
11:17 jensnockert joined #salt
11:17 viq I still don't get how a state gets executed without you explicitly calling it, unless one of us is misunderstanding something
11:18 saltnoob you call from base, e.g. '*' for all nodes
11:18 saltnoob but for say prod environment you might want to add some extra users
11:18 saltnoob but simply dropping a users.sls in the prod pillar root would add the users
11:18 viq Then you did explicitly assign that state.
11:18 saltnoob which is confusing when you don't explicitly call it
11:18 diegows joined #salt
11:19 viq And as mentioned, you can define different contents of pillars for different machines
11:19 viq saltnoob: you explicitly add it to highstate and then get confused when it gets executed when you run highstate?
11:20 saltnoob let's say for base pillar users I have admin1 and admin2
11:20 saltnoob so I want this applied to every server
11:20 saltnoob now lets say for qa servers I want to add qa1 and qa2
11:21 saltnoob so I create a users.sls (pillar data file) with those users
11:21 bhosmer joined #salt
11:21 saltnoob for clarity it should not add those users unless I explicitly call the formula from qa state
11:21 saltnoob but it gets added because it's called from base ('*')
11:21 viq saltnoob: that's not how it works
11:22 saltnoob please test
11:22 saltnoob it is
11:22 saltnoob salt combines all the pillar roots
11:22 saltnoob which adds to confusion when using formula's
11:22 viq saltnoob: and by adding it to base you explicitly told it to run this state when you execute highstate. You don't want it to execute, don't add it to '*' in base
11:22 saltnoob yes for admin1 and admin1 data
11:23 saltnoob but I didn't tell it to add the qa users
11:23 saltnoob the combining is what I'm trying to work around
11:23 saltnoob by using isolated environments and includes
11:23 saltnoob which is why I initially came here
11:24 VSpike can I get salt to show me pillar data rendered in a python syntax?
11:24 viq saltnoob: I see two ways around the users thing. Either create two pillars, say prod-users.sls containing admin1 and admin2, and qa-users.sls containing admin1 admin2 qa1 qa2, and then assign those pillars to appropriate machines, or in pillar file do an if that will add the qa users to list when some conditions are met
11:25 saltnoob that would mean creating new formula's per environment which I wanted to get away from
11:25 saltnoob e.g. by reusing formula's from the base
11:26 saltnoob guess you're right
11:26 saltnoob will need to look into structuring things a bit different
11:26 saltnoob and not reuse formula's
11:26 saltnoob just stick to simple states and pillar data
11:26 VSpike ahh .. --out
11:26 VSpike I knew I'd seen a mention of that somewhere
11:27 ramishra joined #salt
11:30 vbabiy joined #salt
11:30 sarkis joined #salt
11:31 VSpike A list of lists is quite hard to do in YAML, isn't it?
11:34 CycloHex what's the default format for the attr expire in states.user.present?
11:45 bhosmer joined #salt
11:51 elfixit joined #salt
11:52 ramishra joined #salt
12:01 astol joined #salt
12:01 intellix joined #salt
12:03 CycloHex anyone else using the postgres_user? always when I call a highstate test=true it says the user will be updated.. why is this? it should say use ris present and up to date like it does for my other users
12:07 linjan joined #salt
12:08 jaimed joined #salt
12:16 canci joined #salt
12:17 bhosmer joined #salt
12:20 raccoon_ joined #salt
12:21 raccoon_ hello! how can i make a minion output a source file diff between the current state and it's top state? like it does after an upgrade, in the master console
12:21 raccoon_ but without actually writing anything
12:22 sarkis joined #salt
12:24 wnkz joined #salt
12:24 bezaban http://docs.saltstack.com/en/latest/ref/states/testing.html
12:24 raccoon_ bezaban: thanks!
12:24 bezaban np :)
12:25 jdmf joined #salt
12:30 CeBe joined #salt
12:33 jemejones joined #salt
12:34 wnkz__ joined #salt
12:34 bhosmer joined #salt
12:35 to_json joined #salt
12:46 QiQe joined #salt
12:47 QiQe hey guys, our repo is full of war files (binaries) that dev guys upload each time that they need to do a new deploy using salt
12:47 QiQe any idea to manage war files in a better way
12:59 TheThing joined #salt
13:01 SkyRocknRoll__ joined #salt
13:03 cpowell joined #salt
13:04 cpowell joined #salt
13:04 scottpgallagher joined #salt
13:07 FeatherKing joined #salt
13:07 CF975553 joined #salt
13:08 nitti joined #salt
13:08 eunuchsocket joined #salt
13:11 ggoZ joined #salt
13:14 racooper joined #salt
13:18 rypeck joined #salt
13:20 kermit joined #salt
13:23 sarkis joined #salt
13:26 CaptTofu joined #salt
13:27 linjan joined #salt
13:28 vejdmn joined #salt
13:30 Katafalkas joined #salt
13:30 prosper_ joined #salt
13:33 vbabiy joined #salt
13:39 Ahrotahntee is there a way to specify the order in which commands inside a state are run? I need to deliver a keyfile before attempting a git.latest, but it happens in the reverse order
13:39 war2 joined #salt
13:40 viq Ahrotahntee: require
13:40 viq Ahrotahntee: http://docs.saltstack.com/en/latest/ref/states/requisites.html
13:43 Ahrotahntee I was using requires_in, this appears to be working thanks
13:43 Ahrotahntee which is backwards
13:43 Ahrotahntee now that I understand the implication
13:43 darkelda joined #salt
13:46 lionel joined #salt
13:49 oyvjel joined #salt
13:54 linjan joined #salt
13:55 Bateau_ joined #salt
13:55 elfixit joined #salt
13:56 djstorm joined #salt
13:57 TheThing joined #salt
13:57 scottpgallagher joined #salt
13:57 intellix joined #salt
13:59 darkelda joined #salt
13:59 darkelda joined #salt
14:00 perfectsine joined #salt
14:01 kickerdog joined #salt
14:02 scott_w joined #salt
14:06 glyf joined #salt
14:10 yannj_fr joined #salt
14:11 yannj_fr hi there, I want to know if there is a way to use os.listdir in a state
14:13 vbabiy joined #salt
14:14 romanparish joined #salt
14:14 romanparish Ahrotahntee: m00
14:14 micah_chatt joined #salt
14:16 darkelda joined #salt
14:16 kickerdog joined #salt
14:17 Ahrotahntee boop
14:19 luminous yannj_fr: in .sls, or a _state?
14:20 smcquay joined #salt
14:21 luminous QiQe: we use maven to build the WARs, repos have source and maven build/make stuff, but no WARs. When we build a WAR, we push to CDN/cloud files for storage/retrieval. States know how to retrieve WARs as well as build one and use the built one
14:21 cbulist joined #salt
14:21 luminous VSpike: how is a list of lists difficult?
14:24 sarkis joined #salt
14:24 vbabiy joined #salt
14:25 linjan joined #salt
14:26 ipmb joined #salt
14:26 yannj_fr luminous: in .sls
14:26 smcquay joined #salt
14:26 QiQe luminous: thanks that is cool
14:30 ckao joined #salt
14:31 micah_chatt joined #salt
14:33 intellix joined #salt
14:34 prosper__ joined #salt
14:34 mpanetta joined #salt
14:38 StDiluted joined #salt
14:46 istram yannj_fr: what about jinja? like {% for dir in salt['os.listdir'](' path? ') %} something {% endfor %} ?
14:47 pdayton joined #salt
14:47 istram yannj_fr: not sure about the os.listdir invocation
14:47 war2 joined #salt
14:49 istram yannj_fr: damn, os.listdir. Though they made it into salt module too :-D I'm not sure how to use the python modules. ATM I'd maybe use something like salt['cmd.run']('find . blah blah')
14:50 desposo joined #salt
14:50 yannj_fr joined #salt
14:52 yannj_fr istram: I will test
14:52 bhosmer__ joined #salt
14:53 rallytime joined #salt
14:56 linjan joined #salt
14:57 DaveQB joined #salt
14:58 yannj_fr joined #salt
14:58 grove_ joined #salt
15:03 yannj_fr joined #salt
15:03 marnom joined #salt
15:08 spiette joined #salt
15:08 dalexander joined #salt
15:09 to_json joined #salt
15:09 conan_the_destro joined #salt
15:12 yannj_fr istram : the os.listdir invocation is not ok
15:12 istram yannj_fr: that's because it
15:12 wendall911 joined #salt
15:13 istram yannj_fr: I think it's because you don't have that module available in the namespace you're in
15:13 darkelda joined #salt
15:13 yannj_fr how could I make it available
15:13 istram yannj_fr: as I wrote under it I have did the workaround with cmd.run and find.
15:13 linjan joined #salt
15:14 istram yannj_fr: that's what I don't know yet, still a salt beginner
15:14 yannj_fr ok
15:14 yannj_fr but cmd.run is distant not local right?
15:14 rlarkin joined #salt
15:14 istram yannj_fr: well, the state will run on minion
15:15 hasues joined #salt
15:15 yannj_fr ok
15:15 yannj_fr I will try another way
15:16 jergerber joined #salt
15:17 kingel joined #salt
15:17 CeBe joined #salt
15:18 viq yannj_fr: maybe a guiding question - why do you want to use it?
15:25 sarkis joined #salt
15:26 ericof joined #salt
15:26 anotherZero joined #salt
15:28 thedodd joined #salt
15:28 jroof joined #salt
15:29 vbabiy joined #salt
15:30 scottpgallagher joined #salt
15:30 glyf joined #salt
15:32 vbabiy joined #salt
15:33 prosper_ joined #salt
15:33 possibilities joined #salt
15:34 dccc_ joined #salt
15:41 Ahrotahntee Can I map existing vms to salt-cloud?
15:41 Ahrotahntee or must they be explicitly be spun up by it?
15:41 timoguin Ahrotahntee: how do you mean? install salt on existing systems?
15:42 Ahrotahntee I mean add existing vms to a cloud map
15:42 Ahrotahntee I think.
15:42 Ahrotahntee actually I'm not sure what I want here; I'll come back when I have a better understanding
15:46 darkelda joined #salt
15:47 sarkis joined #salt
15:49 fannet Does anyone know if its possible to dynamically map git branches to environments?
15:52 fannet http://docs.saltstack.com/en/latest/ref/pillar/all/salt.pillar.git_pillar.html?highlight=ext_pillar%20git  seems to indicate that you need to define every branch that salt should use
15:52 pipps joined #salt
15:52 SheetiS1 joined #salt
15:54 war3 joined #salt
15:54 stephanbuys joined #salt
15:55 viq fannet: for states or pillars?
15:55 stephanbuys hi all, I'm only executing a change on "changes" using the "onchanges" requirement. It works as expected but results in an error during highstate if there are no changes. Is there a way to suppress the error? As its not actually an "error" if it doesn't execute?
15:57 timoguin fannet: for pillar you have to directly map them
15:57 timoguin for file_roots it will automatically assume each branch is an environment
15:58 jemejones left #salt
15:59 darkelda joined #salt
16:01 fannet sooo no then :)
16:02 jroof Is there an ETA for an updated Solaris Minion install?
16:03 XenophonF joined #salt
16:06 ndrei joined #salt
16:10 prosper_ joined #salt
16:10 schimmy joined #salt
16:12 NotreDev joined #salt
16:14 schimmy1 joined #salt
16:14 KyleG joined #salt
16:14 KyleG joined #salt
16:17 bkrram_ joined #salt
16:18 bkrram_ Hi,
16:18 bkrram_ Since which version of salt was multimaster PKI supported?
16:18 bkrram_ I'm using 2014.1.10 and it does not seem to work for me
16:19 bkrram_ Even The signed key pairs are not getting generated
16:21 conan_the_destro joined #salt
16:22 viq bkrram_: you have a link to the docs to what you're trying to do?
16:23 bkrram_ viq: Yes, I was following http://docs.saltstack.com/en/latest/topics/tutorials/multimaster_pki.html down to the letter!
16:23 snuffeluffegus joined #salt
16:25 viq hm, from the docs looks like it should work with 2014.1
16:27 bkrram_ Which is why it sounded strange. I found a link at http://grokbase.com/t/gg/salt-users/148v7sgjg9/enable-master-sign-its-public-key which says it will be in the 2014.7 release. Is that an upcoming release?
16:28 viq yeah, 2014.7 is hoped to have a final release candidate in a week or so, and a release about a week after that, from what I've heard
16:29 * viq can't wait for pillar merging
16:29 bkrram_ So am I right in assuming that the PKI based failover will only work with that release?
16:31 CF314258 joined #salt
16:31 CF705724 joined #salt
16:31 possibilities joined #salt
16:35 SheetiS1 joined #salt
16:36 war2 joined #salt
16:36 viq apparently so
16:37 viq you can already play with it, as the bootstrap script is able to install from git., but I would do it only on test machines
16:38 prosper_ joined #salt
16:39 bkrram_ viq: Thanks.. will do!
16:42 aparsons joined #salt
16:43 n8n joined #salt
16:44 godber joined #salt
16:48 sarkis_ joined #salt
16:48 wnkz joined #salt
16:54 schimmy joined #salt
16:54 TheThing joined #salt
16:58 kingel joined #salt
16:58 primechuck joined #salt
16:59 stephanbuys1 joined #salt
17:00 StDiluted joined #salt
17:01 schimmy joined #salt
17:01 primechuck Short of changing cherrypy code, Is there a way to set the timeout in cherrypy for greater than 5 minutes?  We have a few things like high state that can take 10 minutes but after 5 cherrypy times out the connection.
17:02 repl1cant submit the job via local_async, grab the job id from that, and then query against it
17:03 repl1cant instead of via local which will eventually timeout
17:04 patarr joined #salt
17:04 patarr joined #salt
17:04 kingel joined #salt
17:05 schimmy1 joined #salt
17:06 derelm joined #salt
17:07 Gareth morning morning
17:07 tristianc joined #salt
17:08 cpowell joined #salt
17:08 gmcwhistler joined #salt
17:09 tristianc install_patched_bash:
17:09 tristianc pkg.installed:
17:09 tristianc - sources:
17:09 tristianc - bash: salt:///BASH/rpms/bash-3.2-33.el5_11.4.x86_64.rpm
17:09 tristianc - skip_verify: True
17:09 tristianc Ugh, aplogies
17:09 tristianc Too many windows
17:10 ndrei joined #salt
17:10 Ryan_Lane joined #salt
17:13 bhosmer joined #salt
17:17 primechuck hmmmm..we do all out calls to salt through cherrypy and we rely on sync rather than trying to do job tracking...
17:18 whiteinge primechuck: we could certainly make the default reponse timeout configurable. it's an easy addition
17:19 jgarr joined #salt
17:20 whiteinge please do feel free to file an issue to get it added
17:20 jgarr is there a reason a module would work on the salt command line but not in a state? error is Comment: State iptables.check_chain found in sls test is unavailable
17:21 jgarr but salt-call iptables.check_chain filter INPUT works
17:22 whiteinge primechuck: polling is very likely to be a more reliable solution, as repl1cant suggested, but polling (currently) requires manual checking of which minions haven't returned yet. there's a feature request to get a "summary" display in the jobs response, e.g., (X number of minions out of Y have returned)
17:22 msh14103 joined #salt
17:22 tristianc joined #salt
17:25 druonysus joined #salt
17:25 druonysus joined #salt
17:26 troyready joined #salt
17:28 Setsuna666 joined #salt
17:29 repl1cant yup, i basically just poll and wait for a result. could also tap into the event bus and wait for the result there
17:29 SheetiS1 joined #salt
17:30 Setsuna666 basepi: Did you have any report so far of reactor .sls file not rendering with 2017.7rc2 ?
17:31 Setsuna666 2014.7rc2*
17:31 logix812 joined #salt
17:31 robawt 0/bu2
17:31 vejdmn joined #salt
17:31 robawt ug
17:32 basepi Setsuna666: there was this comment: https://github.com/saltstack/salt/issues/12827#issuecomment-57669343
17:32 basepi Though apparently whiteinge couldn't reproduce it.
17:32 mpanetta What are the repercussions of removing /var/cache salt on a salt-master?  My salt master is somehow getting the wrong pillar data, I have checked all my top files and everything looks right, so I want to clear the cache..
17:33 Setsuna666 hum, kk. I installed the Salt Master with bootstrap and since then, all reactor sls files cannot render, but using earlier release of 2014.7, I do not experience this issue
17:33 whiteinge Setsuna666: i'd be interested to see your ``file_roots`` and ``reactor`` settings
17:33 basepi mpanetta: just make sure the master is stopped when you do it, and otherwise you're good to remove the cache.
17:33 mpanetta basepi: Ok cool, thanks.
17:33 logix812 joined #salt
17:34 Setsuna666 whiteinge: Give me a couple of seconds and I'll pastebin the info
17:34 MasterJ joined #salt
17:36 Setsuna666 whiteinge: Here it is: http://pastebin.com/iSNs1sQ0
17:37 logix812 joined #salt
17:37 logix812 left #salt
17:38 whiteinge Setsuna666: thanks!
17:38 aw110f joined #salt
17:38 ndrei joined #salt
17:41 whiteinge basepi: i was under the impression the 'base' environment is always required (even if unused). do you know details on that?
17:43 yomilk joined #salt
17:43 whiteinge (obviously we don't have error-checking in place for either approach, so that will need to be addressed anyway)
17:47 astol joined #salt
17:49 primechuck Oh, looks like there is already a request.  #159
17:50 murrdoc joined #salt
17:55 bhosmer joined #salt
17:56 vbabiy joined #salt
17:58 primechuck Would tornado still have the same timeout issue?
18:01 dalexand_ joined #salt
18:01 possibilities joined #salt
18:05 basepi whiteinge: I honestly am not sure on that, I've never not had it present.
18:05 basepi I'm*
18:06 basepi oh, nevermind, I just misplaced the "am" in that sentence, I didn't typo it after all. -)
18:06 basepi =)*
18:07 ozzzo1 joined #salt
18:08 desposo joined #salt
18:10 jchen so I added gitfs_provider: pygit2, have it installed, and restarted salt-master, but still seeing "Git fileserver backend is enabled in configuration but could not be loaded, is GitPython installed?". what do?
18:11 whiteinge primechuck: https://github.com/saltstack/salt/pull/16359
18:12 whiteinge Setsuna666: will you please add a 'base' environment to your ``file_roots`` setting to double-check this? you can point it to an empty dir
18:15 war3 joined #salt
18:16 Setsuna666 whiteinge: So an empty 'base' environment should do it ?
18:17 whiteinge that's my theory :)
18:17 Setsuna666 whiteinge: Will try it, and let you know the outcome
18:22 Setsuna666 whiteinge: Seems to solve the issue :)
18:24 whiteinge woot. thanks for checking
18:24 whiteinge i believe the 'base' environment is required, but if so we should output an error/warning if one is not found at least
18:25 whiteinge we can discuss further in that issue
18:26 micah_chatt joined #salt
18:29 smcquay joined #salt
18:31 higgs001 joined #salt
18:33 kingel joined #salt
18:37 kermit joined #salt
18:37 kermit joined #salt
18:40 Chris______ joined #salt
18:42 Chris______ How do I make new sensors and connect them to interfaces on the network?
18:44 ndrei joined #salt
18:44 Chris______ Can anyone help me?
18:45 Ahrotahntee Does anyone off-hand know how I ought to specify authentication credentials for ext_pillar when using git as an external pillar source? The documentation doesn't include anything about keys for ext_pillar
18:46 possibilities joined #salt
18:47 kermit joined #salt
18:47 Gareth joined #salt
18:50 SheetiS1 Ahrotahntee: the easy answer is to put your private ssh key in the ~/.ssh/id_rsa of the user that salt runs as, but I there might be something cleaner.
18:51 SheetiS1 assuming you use ssh keys for git auth
18:51 rap424 joined #salt
18:51 Ahrotahntee I do
18:52 vbabiy joined #salt
18:54 yannj_fr joined #salt
19:03 yannj_fr joined #salt
19:04 jchen how do i refresh the gitfs roots for a salt master
19:05 jchen i want to use saltstack/users-formula, returning "no matching sls found"
19:07 N3X15 jchen, git.update afaik
19:07 jchen what
19:07 kingel joined #salt
19:07 primechuck whiteinge: (hug)
19:08 jchen found what i was trying to remember: `salt-call event.fire_master update salt/fileserver/gitfs/update` (which is strange)
19:08 jchen however, still getting "no matching sls found for 'users' in env 'base'"
19:08 jchen is there some better documentation than the super sparse document on formula usage?
19:09 N3X15 try salt  '*' saltutil.sync_all
19:10 aparsons joined #salt
19:10 N3X15 I use gitfs (although with gitpython) and that usually seems to get everything synced after master updates from the repo
19:10 iggy jchen: salt-run fileserver.update
19:11 iggy it syncs by default once a minute
19:13 jchen iggy: thank you
19:13 jchen that did the needful
19:16 thayne joined #salt
19:20 romanparish left #salt
19:21 thedodd joined #salt
19:23 ozzzo joined #salt
19:26 ckao joined #salt
19:29 mapu joined #salt
19:29 skullone Phibs:
19:34 aparsons joined #salt
19:36 Katafalkas joined #salt
19:37 kingel_ joined #salt
19:47 oz_akan joined #salt
19:48 aparsons joined #salt
19:54 Setsuna666 basepi: Currently playing with 2014.7.0rc2 and noticed that debconf prompt seems to fail package installation instead of choosing default options
19:56 nitti_ joined #salt
19:56 Katafalkas joined #salt
19:56 nitti__ joined #salt
20:01 basepi Setsuna666: looks like it's been fixed since RC2: https://github.com/saltstack/salt/issues/16307
20:01 aparsons joined #salt
20:02 zooz joined #salt
20:02 aparsons joined #salt
20:03 Ahlee Anybody familiar with how pillars are resolved within custom _states?  I'm trying to pass "pillar={foo:{'bar':'baz'}}, but if i __pillar__['foo'] from within my _state it's throwing an exception, which I imagine is a KeyError (logs show dict object has no attribute 'foo')
20:03 Setsuna666 kk, thanks
20:03 Ahlee does this mean I'll need to hand those off to variables in a sls state which in turn will hand them through to the _state?
20:04 aparsons joined #salt
20:05 possibilities joined #salt
20:05 SheetiS1 Ahlee: if a sane default is possible, you could use .get() inside the module like you can in jinja
20:05 SheetiS1 and get the pillar or use the default if not.
20:05 aparsons joined #salt
20:06 Ahlee SheetiS1: these pillars only exist for the length they're passed - they're generated by another application and passed to salt via the rest api
20:06 druonysus joined #salt
20:06 druonysus joined #salt
20:06 Ahlee We don't want to introduce an at rest state by storing them in a database or similar
20:07 bhosmer joined #salt
20:07 Ahlee i think i'm just going to have to resolve all the pillars in the sls, rather than trying to parse out the dunder varibales from within the _state
20:07 mpanetta Um, is environ.setenv not available in 2014.1.x?
20:08 mpanetta Because it is telling me it is unavailable
20:10 SheetiS1 https://github.com/saltstack/salt/tree/2014.1/salt/states <== environ is not there, correct
20:10 pssblts joined #salt
20:11 mpanetta Yeah I just checked on my master :(
20:11 mpanetta SheetiS1: Think I could copy it from 2014.7 and put it in a _states dir?
20:11 SheetiS1 checking
20:12 mpanetta SheetiS1: Thanks
20:12 Supermathie joined #salt
20:12 SheetiS1 It looks clean of external dependencies from what I can see
20:12 SheetiS1 I bet it would work
20:12 garphy joined #salt
20:12 mpanetta Sweet, I will let ya know in a few mins :)
20:13 SheetiS1 It just imports os and then string_types from salt._compat
20:13 Supermathie Hey all, what's the right way to listen for events? I tried doing this but it didn't work: for data in event.iter_events(tag='job/*/ret/*'):
20:14 mpanetta I'm not sure you can use wildcards in the tag...
20:14 mpanetta and you need it to start with salt
20:14 mpanetta salt/job
20:14 mpanetta Supermathie: ^
20:16 mechanicalduck joined #salt
20:16 Ahlee assuming the reactor event is tagged salt/job
20:17 mpanetta Ahlee: all salt events start with salt now
20:17 mpanetta http://docs.saltstack.com/en/latest/topics/event/index.html
20:17 Supermathie mpanetta, Ok, so you have to listen for the event prefix. Any way to retrieve the tag so you can do further matching if you can't wildcard?
20:18 Ahlee Supermathie: there's a sample consumer in the test directory, eventlisten.py
20:18 mpanetta Supermathie: IIRC add the kwarg full = True I think
20:18 fannet does jinja support a ternary operator?
20:24 Supermathie mpanetta, ah yes that does it, now returns {'tag': ..., 'data': ...}
20:24 SheetiS1 Supermathie: it looks like the tag is matched with 'startswith' inside the salt/utils/events.py, so you might be able to limit your matches a little bit that way, but it might be easiest on you if examined the  returned dict itself.  I think you get back ret['tag'] and ret['data'] when getting an event with get_event
20:25 SheetiS1 But it looks like you just saw that yourself :D
20:25 Supermathie yeah you only get the data back without full=True
20:25 mpanetta SheetiS1: Unfortunately my idea failed :(
20:25 SheetiS1 mpanetta: what happened?
20:25 mpanetta Threw an exception
20:26 mpanetta https://gist.github.com/anonymous/39d0359cb2c2ab84b1e4
20:26 mpanetta SheetiS1: ^
20:28 SheetiS1 hmm
20:29 SheetiS1 the only thing that was external was _compat.string_types and they are the same in 2014.1 and 2014.7 (basestring on python < 3)
20:30 mpanetta Hmm
20:30 mpanetta Did module imports change from 2014.1 to 2014.7?
20:31 mpanetta It seems to be failing here: https://github.com/saltstack/salt/blob/2014.7/salt/states/environ.py#L136
20:31 __TheDodd__ joined #salt
20:31 p2 joined #salt
20:31 ozzzo joined #salt
20:31 mpanetta Which implies to me that whatever is responsible for setting the values in that dunder dict isn't...
20:31 SheetiS1 you need the module to go with the state
20:31 mpanetta Oh!
20:31 mpanetta Duh...
20:31 SheetiS1 https://github.com/saltstack/salt/blob/2014.7/salt/modules/environ.py
20:32 mpanetta I should know this since Ihave already written a pthon state module before :P
20:32 SheetiS1 hehe
20:32 mpanetta I wrote something to add pub keys to github :)
20:33 mpanetta If I ever get a chance to flesh out the other bits of the github api I will submit it.
20:33 mpanetta I find it to be a useful module when setting up jenkins servers
20:33 SheetiS1 Indeed
20:34 SheetiS1 I approve :D
20:34 mpanetta That fixed it!
20:34 mpanetta Thanks :)
20:35 SheetiS1 ok good deal
20:36 jchen so i'm using salt-cloud to spin up a new Linode, and it's returning "Error: There was a profile error: Failed to upload /tmp/tmptBuA8P to /tmp/.saltcloud-e7076003-8406-455c-bd1a-4ed57450984c/minion.pem. Exit code: 0"
20:36 jchen however, the file is definitely uploaded correclty
20:36 jchen I see a kwarg for allow_failure which appears to allow us to ignore those failures, but setting it in the profile didn't do anthing
20:36 jchen how do I convince salt-cloud that it succeeded in scping the minion.pem
20:37 higgs001 joined #salt
20:38 Supermathie mpanetta, SheetiS1, for the future: https://github.com/Supermathie/salt/commit/c928b09a9d1c483d886501f01db65167e8a297b1
20:39 SheetiS1 I've not used Linode via salt-cloud before.  (My personal stuff is at Rackspace and work stuff AWS).  I won't be much help there jchen, sorry :(
20:39 pdayton joined #salt
20:39 jchen SheetiS1: don't think it's Linode-specific, pretty sure it's the general salt module something is effed
20:42 kingel joined #salt
20:42 mpanetta Supermathie: Awesome!
20:43 SheetiS1 ^^
20:46 yannj_fr joined #salt
20:47 kermit joined #salt
20:49 possibilities joined #salt
20:51 bhosmer joined #salt
20:53 Whissi joined #salt
21:00 ajolo joined #salt
21:01 nitti joined #salt
21:05 wnkz__ joined #salt
21:05 TheThing_ joined #salt
21:07 fannet can someone take a look at this state snipped and tell me how I would pass the current iterated server name to the Jinja template?   http://pastebin.com/6Z0AxARM
21:07 kballou joined #salt
21:07 jgarr left #salt
21:07 druonysus joined #salt
21:07 TheThin__ joined #salt
21:10 TheThing joined #salt
21:11 murrdoc - context:
21:11 murrdoc server_name: {{server_name}}
21:11 TheThing_ joined #salt
21:11 ajolo_ joined #salt
21:13 N3X15 joined #salt
21:13 TheThin__ joined #salt
21:13 whytewolf joined #salt
21:14 TheThing joined #salt
21:14 XenophonF joined #salt
21:15 Setsuna666_ joined #salt
21:16 pdayton joined #salt
21:17 TheThing_ joined #salt
21:18 TheThin__ joined #salt
21:20 yomilk joined #salt
21:20 aparsons joined #salt
21:20 TheThing joined #salt
21:21 TheThi___ joined #salt
21:22 TheThing_ joined #salt
21:23 desposo joined #salt
21:25 pdayton joined #salt
21:26 N3X15 joined #salt
21:30 KyleG left #salt
21:33 kickerdog joined #salt
21:37 kickerdog I need to spin up 10,000 aws vm's, does anyone know how to use salt-cloud to do that?
21:43 aparsons joined #salt
21:43 kickerdog1 joined #salt
21:47 TheThing joined #salt
21:48 TheThin__ joined #salt
21:49 TheThi___ joined #salt
21:50 peters-tx joined #salt
21:51 TheTh____ joined #salt
21:54 chitown has there been any talk about having "more than 2 levels" for states/commands/...?
21:54 Eugene 10,000 is a bit above the standard EC2 limits....
21:54 chitown eg. in 2014.7 has boto_* stats
21:54 chitown states*
21:54 Dinde left #salt
21:54 chitown it would be cool if it were boto.route53.do_something
21:55 Eugene You can just call state.sls dosomething
21:59 aparsons_ joined #salt
22:04 dave_den joined #salt
22:04 jalbretsen joined #salt
22:09 desposo joined #salt
22:15 renoirb Hi all
22:15 renoirb Is there a way to list the available states (i.e. i want to read what is available through gitfs backends
22:16 alex-solidfire joined #salt
22:16 mlanner_ joined #salt
22:18 murrdoc yeah
22:18 murrdoc there is a way
22:20 murrdoc ah man its like state.list
22:21 aparsons joined #salt
22:21 murrdoc salt '*' state.show_sls
22:21 murrdoc or salt '*' state.show_sls saltstackfomula-name
22:24 druonysus joined #salt
22:25 desposo1 joined #salt
22:32 pdayton joined #salt
22:36 mlanner joined #salt
22:36 snuffeluffegus joined #salt
22:43 ede joined #salt
22:43 TheThing joined #salt
22:43 iggy renoirb: salt-call cp.list_master can be helpful too
22:45 aparsons joined #salt
22:46 ntropy joined #salt
22:49 renoirb iggy, it makes me realize that cp.list_master also considers files in .git/
22:49 renoirb i haven’t thought of that
22:51 iggy it considers a _lot_ of stuff
22:51 mlanner #pfsense
22:56 renoirb So, salt salt cp.list_states should list all available states or only what is configured in the compounded top.sls ?
23:04 fannet murrdoc -- I want to take {{server_name}} being used in that iterator inside the jinja not inside the yaml
23:06 fannet I want to pass the current iterated server name to the Jinja template that is building the /tmp/{{server_name}}.conf   http://pastebin.com/6Z0AxARM
23:10 possibilities joined #salt
23:15 murrdoc joined #salt
23:18 spookah joined #salt
23:19 murrdoc yeah fannet thats how the -context stuff works
23:20 fannet ahhh I get it
23:20 fannet had a brain melt
23:21 fannet ls
23:23 bhosmer_ joined #salt
23:27 jaimed joined #salt
23:29 mschiff joined #salt
23:30 bhosmer joined #salt
23:33 aparsons joined #salt
23:40 jcockhren I can confirm thr behavior cited by jchen in regards to salt-cloud failing during key exchange. The issue appears for a master version 2014.1.11. Reverted to 2014.1.10 and the issue went away
23:48 micah_chatt joined #salt
23:51 druonysus joined #salt
23:51 druonysus joined #salt
23:53 mpanetta joined #salt
23:55 mpanetta joined #salt
23:56 Katafalkas joined #salt

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