Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-07-07

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

All times shown according to UTC.

Time Nick Message
00:00 iggy joehh: we were talking about it the other day, I suggested they stop having the same info in 40 places that all had to be updated... he said he liked all the extra work
00:01 yomilk joined #salt
00:10 kal0pr joined #salt
00:16 tmclaugh[work] joined #salt
00:18 brianfeister joined #salt
00:31 furrowedbrow joined #salt
00:38 Viper233 Is it possible to parse a pillar value through a regex in a state file? I was to reference a version number, 1.9.3 as 19 and 1.9
00:40 iggy Viper233: you can use normal python string functions (if it's a string)
00:43 Viper233 iggy: okay. would I need to create a module?
00:44 iggy you _really_ don't want to use the re module in states
00:45 joehh iggy: thanks!
00:48 sunkist joined #salt
00:50 Viper233 iggy: How would I use the python string function in my salt states?
00:57 evilrob joined #salt
00:58 dopesong joined #salt
00:58 sectionme joined #salt
01:04 pcn So I've just proposed this: https://github.com/saltstack/salt-contrib/pull/127.  I'd love to know if anyone else has tried to use the ec2_pillar and if this is of interest.
01:09 kal0pr joined #salt
01:10 sunkist joined #salt
01:21 baweaver joined #salt
01:30 beauby joined #salt
01:33 hybridpollo joined #salt
01:41 clintberry joined #salt
01:46 madpenguin joined #salt
01:51 beauby joined #salt
01:54 kal0pr joined #salt
01:58 blakev__ joined #salt
01:59 jerematic joined #salt
02:07 madpenguin joined #salt
02:09 beauby joined #salt
02:16 subsignal joined #salt
02:22 jondonas joined #salt
02:23 evle joined #salt
02:27 beauby joined #salt
02:30 favadi joined #salt
02:43 napsterX joined #salt
02:46 Bryson joined #salt
02:47 sectionme joined #salt
02:47 cberndt joined #salt
02:49 beauby joined #salt
02:51 c10 joined #salt
02:56 clintberry joined #salt
02:57 quasiben1 joined #salt
02:58 dopesong joined #salt
02:59 hasues joined #salt
03:00 hasues left #salt
03:09 beauby joined #salt
03:13 ITChap joined #salt
03:15 bones050_ joined #salt
03:17 venu0336 joined #salt
03:22 ITChap joined #salt
03:28 kal0pr joined #salt
03:55 Furao joined #salt
04:01 joehh basepi: thanks for updating the docs for the windows downloads!
04:09 mosen joined #salt
04:10 zer0def joined #salt
04:18 kal0pr joined #salt
04:21 ajw0100 joined #salt
04:27 baweaver_ joined #salt
04:28 hasues joined #salt
04:30 clintberry joined #salt
04:31 jaybocc2 joined #salt
04:31 otter768 joined #salt
04:32 hasues left #salt
04:32 zer0def joined #salt
04:32 joeto joined #salt
04:33 clintberry joined #salt
04:33 ITChap joined #salt
04:36 sectionme joined #salt
04:37 kal0pr joined #salt
04:43 stanchan joined #salt
04:47 kal0pr joined #salt
04:47 ramteid joined #salt
04:50 c10 joined #salt
04:57 kal0pr joined #salt
04:57 clintberry joined #salt
05:01 jaybocc2 joined #salt
05:02 Sokel joined #salt
05:02 catpigger joined #salt
05:04 Sokel Is it at all possible to set multiple grains at the same time? Instead of doing it one at a time, like grains.setval role value business_group value environment value
05:04 napsterX joined #salt
05:06 aw110f joined #salt
05:06 ITChap joined #salt
05:07 kal0pr joined #salt
05:09 aw110f_ joined #salt
05:17 ITChap joined #salt
05:20 dopesong joined #salt
05:25 rdas joined #salt
05:27 kal0pr joined #salt
05:28 ecdhe joined #salt
05:39 cilkay left #salt
05:44 khebbie joined #salt
05:44 cliluw joined #salt
05:46 Nazca__ joined #salt
05:46 kal0pr joined #salt
05:53 jaybocc2 joined #salt
05:54 keimlink joined #salt
06:01 pipeep joined #salt
06:02 colttt joined #salt
06:03 saltsa joined #salt
06:08 desposo joined #salt
06:08 harkx joined #salt
06:09 jaybocc2 joined #salt
06:09 subsignal joined #salt
06:16 kal0pr joined #salt
06:17 danielcb joined #salt
06:19 ITChap joined #salt
06:20 AndreasLutro joined #salt
06:20 harkx joined #salt
06:24 madpenguin joined #salt
06:24 ITChap joined #salt
06:25 sectionme joined #salt
06:25 markm joined #salt
06:29 malinoff joined #salt
06:30 mick3y joined #salt
06:30 mick3y good morning
06:30 ITChap joined #salt
06:31 trikke joined #salt
06:32 mick3y has anyone encountered "minions_rejected: Data failed to compile:" after upgrading to 2015.5.2 from 2014.7.2 ?
06:32 otter768 joined #salt
06:32 dvestal joined #salt
06:33 mick3y i get that while running 'salt-key -L'
06:34 madpenguin joined #salt
06:39 stephanbuys joined #salt
06:39 kal0pr joined #salt
06:45 lb1a joined #salt
06:47 trikke joined #salt
06:49 sirex joined #salt
06:49 c10 joined #salt
06:51 sectionme joined #salt
06:53 TyrfingMjolnir joined #salt
06:59 sectionme joined #salt
06:59 jaybocc2 joined #salt
07:00 soren joined #salt
07:01 Aidin joined #salt
07:02 TyrfingMjolnir joined #salt
07:06 trikke joined #salt
07:07 mafrosis joined #salt
07:10 thalleralexander joined #salt
07:15 jaybocc2 joined #salt
07:21 kawa2014 joined #salt
07:28 Aidin joined #salt
07:28 gcfhvjbkn joined #salt
07:29 desposo1 joined #salt
07:34 NV joined #salt
07:36 madpenguin joined #salt
07:38 sectionme joined #salt
07:38 linjan joined #salt
07:44 kal0pr joined #salt
07:45 ndrei joined #salt
07:47 sectionme joined #salt
07:48 TyrfingMjolnir joined #salt
07:49 yomilk joined #salt
07:50 chiui_ joined #salt
07:50 chiui joined #salt
07:51 jaybocc2 joined #salt
07:54 kal0pr joined #salt
07:55 dkrae joined #salt
07:59 sectionme joined #salt
08:00 c10 joined #salt
08:01 ITChap joined #salt
08:02 lothiraldan joined #salt
08:05 sectionme joined #salt
08:06 jaybocc2 joined #salt
08:08 stoogenmeyer joined #salt
08:10 ITChap joined #salt
08:12 s_kunk joined #salt
08:13 Xevian joined #salt
08:16 khebbie joined #salt
08:18 N-Mi joined #salt
08:19 ITChap joined #salt
08:20 totte joined #salt
08:20 ITChap joined #salt
08:23 ITChap joined #salt
08:25 gcfhvjbkn joined #salt
08:29 ITChap joined #salt
08:33 otter768 joined #salt
08:35 mick3y any idea how to troubleshoot that?
08:35 ITChap joined #salt
08:35 dvestal joined #salt
08:41 pviktori_ joined #salt
08:43 ITChap joined #salt
08:43 pviktori joined #salt
08:51 dvestal joined #salt
08:54 keimlink joined #salt
08:57 CeBe joined #salt
08:57 julez_ joined #salt
09:07 Norrland uhm. Isn't the salt-minion services supposed to start again after upgrade?
09:08 VSpike Is there a better place to post a question than the Salt-users group? Is Stackexchange better?
09:09 chutzpah joined #salt
09:10 sectionme joined #salt
09:11 laidback joined #salt
09:13 laidback hello, i have a question regarding the top file and formula development using gitfs. Is there a best practice, how to develop your top file without pushing to git to get it available in the master?
09:16 keimlink_ joined #salt
09:21 mrbigglesworth joined #salt
09:24 mrbigglesworth joined #salt
09:25 mrbigglesworth joined #salt
09:27 jaybocc2 joined #salt
09:27 Ztyx joined #salt
09:27 c10 joined #salt
09:33 CeBe1 joined #salt
09:35 kal0pr joined #salt
09:36 trikke joined #salt
09:39 mdupont joined #salt
09:42 jaybocc2 joined #salt
09:47 ndrei joined #salt
09:53 markm joined #salt
10:02 ndrei joined #salt
10:03 sectionme joined #salt
10:03 jaybocc2 joined #salt
10:03 markm joined #salt
10:07 pviktori_ joined #salt
10:09 julez_ joined #salt
10:19 jaybocc2 joined #salt
10:24 CeBe joined #salt
10:34 otter768 joined #salt
10:36 khebbie joined #salt
10:43 yomilk joined #salt
10:49 markm joined #salt
10:52 sgargan joined #salt
10:54 kal0pr joined #salt
11:05 aqua^c joined #salt
11:09 kal0pr joined #salt
11:17 kal0pr joined #salt
11:17 smcquay joined #salt
11:17 giantlock joined #salt
11:18 sgargan joined #salt
11:22 ndrei joined #salt
11:25 markm joined #salt
11:27 kal0pr joined #salt
11:27 impi joined #salt
11:36 TyrfingMjolnir joined #salt
11:36 kal0pr joined #salt
11:37 Gues_____ joined #salt
11:37 Gues_____ hi everyone
11:37 _sfxandy hi everyone
11:38 _sfxandy i must be going mad, but i'm getting a Jinja rendering error when i try this:
11:38 _sfxandy {% set group_name = {{ name }} %}
11:38 _sfxandy what am i doing wrong?
11:39 _sfxandy the precise error is expected colon ':', got '}'
11:39 sunkist joined #salt
11:39 Mate jinja is a simple templating language
11:39 _sfxandy agreed
11:39 Mate and not a template metatemplating language
11:39 Mate so try {% set a=b %}
11:39 jaybocc2 joined #salt
11:40 _sfxandy trying that Mate...
11:43 markm joined #salt
11:43 yomilk joined #salt
11:46 furrowedbrow joined #salt
11:47 agronholm left #salt
11:47 quinox joined #salt
11:48 KermitTheFragger joined #salt
11:48 ingslovak joined #salt
11:55 jaybocc2 joined #salt
11:56 kal0pr joined #salt
11:59 khebbie joined #salt
12:06 kal0pr joined #salt
12:08 yomilk joined #salt
12:10 dopesong joined #salt
12:11 evle1 joined #salt
12:11 sunkist joined #salt
12:16 kal0pr joined #salt
12:17 amcorreia joined #salt
12:21 canci joined #salt
12:24 denys joined #salt
12:26 kal0pr joined #salt
12:29 sgargan_ joined #salt
12:29 tmclaugh[work] joined #salt
12:30 NV joined #salt
12:32 jerematic joined #salt
12:32 sgargan_ joined #salt
12:35 otter768 joined #salt
12:36 venu0336 joined #salt
12:36 kal0pr joined #salt
12:39 CeBe joined #salt
12:42 ingslovak joined #salt
12:46 kal0pr joined #salt
12:46 jaybocc2 joined #salt
12:48 iwishiwerearobot joined #salt
12:49 anotherZero joined #salt
12:50 furrowedbrow joined #salt
12:50 subsignal joined #salt
12:54 aqua^c joined #salt
12:57 Furao joined #salt
12:57 kawa2014 joined #salt
13:00 Tecnico1931 joined #salt
13:01 julez joined #salt
13:01 julez joined #salt
13:02 jaybocc2 joined #salt
13:03 julez joined #salt
13:05 c10 joined #salt
13:05 favadi joined #salt
13:05 kal0pr joined #salt
13:10 VSpike Is it a holiday somewhere? It's crazily quiet here
13:14 vstoniest joined #salt
13:15 kal0pr joined #salt
13:17 numkem I'd say it might just be because its the summer
13:20 DammitJim joined #salt
13:22 gmoro joined #salt
13:23 SheetiS joined #salt
13:25 kal0pr joined #salt
13:28 dendazen joined #salt
13:29 Ahlee Gareth: only that i've ruled out my initial thought of it not coming up from modules/config.py:127 (option method). That function emits what it's supposed to, so I'm on to tracking down where the config file is parsed/passed in. I had a brief detour into ensuring pillars named the same as the returner config didn't interfere, and I don't believe they do
13:29 cpowell joined #salt
13:30 jdesilet joined #salt
13:30 kawa2014 joined #salt
13:30 peters-tx joined #salt
13:30 DammitJim what is the normal way to run a script on a minion?
13:31 DammitJim I have a series of things I need to do on a minion like copy folders, set permissions and stuff and I put all that into a shell script
13:31 AndreasLutro DammitJim: cmd.script
13:31 tkharju joined #salt
13:31 DammitJim AndreasLutro, does that mean I have to copy the script to a location on the minion?
13:31 DammitJim or is it truly "remote execution?"
13:32 DammitJim from the master
13:33 Deevolution DammitJim:  both are possible.
13:35 impi joined #salt
13:35 DammitJim great! any docs I should look at to do the remote execution without copying the file?
13:35 HappySlappy joined #salt
13:37 jaybocc2 joined #salt
13:38 AndreasLutro DammitJim: the cmd.script documentation? :p
13:38 pullphinger15 joined #salt
13:38 DammitJim I have seen the cmd.script documentation
13:38 DammitJim I"m confused as to how one uses it if the script isn't copied to the minion
13:39 mpanetta joined #salt
13:39 AndreasLutro technically it's copied into /tmp then removed immediately, are you using it for sensitive information?
13:40 icflournoy joined #salt
13:40 pullphinger joined #salt
13:40 DammitJim not really
13:41 DammitJim I guess I was hoping I didn't have to copy the script (I honestly don't know where I would need to put it on the minion)
13:41 AndreasLutro you don't need to think about it... salt handles it for you
13:41 ekristen joined #salt
13:41 AndreasLutro are you sure you've read the cmd.script documentation? you're not confusing it with cmd.run?
13:42 pullphinger joined #salt
13:42 DammitJim oh, I might be confusing it with cmd.run!!!!
13:43 DammitJim sorry, I need more coffee and I need to learn how to read
13:45 kal0pr joined #salt
13:48 keimlink joined #salt
13:48 twork lesson i learned yesterday: an error in one part of the configuration will (can?) cause the whole thing to fail.
13:49 twork i spent a lot of time looking in the wrong place for my mistake.
13:50 JDiPierro joined #salt
13:50 kaptk2 joined #salt
13:50 twork the error messages i get don't always give me good clues to where my mistake is.
13:51 ericof joined #salt
13:51 twork so, are there other ways to verify my setup, other than running salt itself?
13:51 twork integrity checkers, i think is the term i'm groping for
13:52 timoguin joined #salt
13:52 AndreasLutro run single .sls files before a highstate, run with test=True, be prepared to ssh in and debug that way when you hit problems
13:52 AndreasLutro make sure all your states have the proper requirements set up so they don't try to run when a requirement fails
13:53 jaybocc2 joined #salt
13:53 pullphinger joined #salt
13:54 twork AndreasLutro: that is -- i think -- more or less what i was doing. so i think the answer is, "practice, twork".
13:54 twork thanks.
13:54 aarseniev joined #salt
13:55 timoguin joined #salt
13:57 trikke joined #salt
13:58 gcfhvjbkn what is the preferred/habitual way of deploying large files that you don't put in the repo?
13:58 gcfhvjbkn like database dumps
13:58 gcfhvjbkn i cannot tell top.sls to look for files on my computer drive naturally
13:59 julienlavergne joined #salt
13:59 julienlavergne hello guys
13:59 julienlavergne some question
13:59 aarseniev hi! Can anyone help with salt-masterless and ext_pillar option for it? here http://docs.saltstack.com/en/latest/topics/tutorials/standalone_minion.html#external-pillars says than it works. But with http://pastie.org/10277628 such config - i ran salt-call pillar.items and can't see my pillars into output
13:59 furrowedbrow joined #salt
14:00 julienlavergne how to get data from my minion when ext_pillar s3fs is well configured on the master?
14:00 andrew_v joined #salt
14:00 hasues joined #salt
14:00 hasues left #salt
14:01 julez joined #salt
14:01 dyasny joined #salt
14:01 geekatcmu gcfhvjbkn: that's all site-specific.  2 jobs ago, we dropped it into AFS, because that's where we put *everything*.  1 job ago, it was the NetApp.  This job, it's generally Agile storage.
14:03 twork gcfhvjbkn: i'm totally new here but my guess for this is to shell out to cmd.run and handle it that way
14:03 twork i'm sure there's a better way
14:03 gcfhvjbkn geekatcmu: AFS stands for Amazon File Storage or smth? what's an Agile storage?
14:04 geekatcmu AFS == OpenAFS, which is a distributed file storage system.  I worked at Carnegie Mellon at the time, and they kind of invented it.
14:04 gcfhvjbkn twork: yeah i don't have qualms about using cmd, but i don't want to deploy salt-minion on my own drive/run salt masterless, and even then, no one else would be able to deploy but me
14:04 gcfhvjbkn oh, i see
14:04 kal0pr joined #salt
14:05 geekatcmu Agile storage == http://www.limelight.com/services/storage.html
14:05 geekatcmu Agile is the internal name for it.  Full disclosure: that's the company I work for (so, yes, we use our own product)
14:05 clintberry joined #salt
14:08 twork gcfhvjbkn: i don't follow you... what i have in mind is, 'cmd.run' tells the minion, 'hey, go get [by whatever means] [whatever source] to [whatever path]", and when salt runs, it reacts to the output of that command. (right? again, totally guessing.)
14:08 gcfhvjbkn geekatcmu: thanks for suggestions, although it's not exactly my use case i think
14:08 twork cmd.run only has to be installed on the master. or are you saying that you do run salt masterless?
14:09 JDiPierro joined #salt
14:09 spark_ joined #salt
14:09 gcfhvjbkn ie i've got  one 5 GB file to deploy and i don't want to shill out $10 to dropbox, not quite carnegie mellon scope you see
14:10 gcfhvjbkn twork: yeah i think i see what you mean
14:10 Ahlee gcfhvjbkn: We drop binary files on a web server, and serve them with source: http://webserver/file
14:11 izibi joined #salt
14:11 Ahlee alternatively, salt does have a built in file server you could expose via adding it to file_roots. I also wrote a custom _module that sync's files from an daemonized rsync instance
14:11 Ahlee so, the choice is really up to you.
14:12 Ahlee twork: a linter has been long requested, and is a very hard problem
14:12 Ahlee twork: our test flow is to run a second instance locally on the system, and develop the state against it. Once it runs locally, PR to the branch currently checked out on the master for the environment and then test against one canary minion, then 10 carnaries, then the rest
14:13 geekatcmu yeah, salt is perfectly capable of serving files.
14:13 gcfhvjbkn ahlee: thanks for input
14:14 Ahlee i found salt's fileserver to be wanting on 'large' file systems with lots and lots of files
14:14 kal0pr joined #salt
14:14 Ahlee like our NFS instance is only 4tb with a couple billion files, it didn't get along with being a part of file_roots hence the rsync daemon, but that pushed the logic of knowing how to build the path into our hands
14:14 Ahlee *shrug*
14:16 agend joined #salt
14:18 Ahlee though, given it's an amazon storage device it likely hated when it walked it it had to wait for the local sync, and dat cache churn
14:19 yudao joined #salt
14:19 Ahlee julienlavergne: ext_pillar's appear as regular pillars. So your value would be available via pillar.items
14:19 Ahlee julienlavergne: so, {{ salt['pillar.get']('my_ext_pillar_name') }}
14:19 yudao left #salt
14:21 Ahlee the workflow is the master reaches out to s3, grabs the data and merges it into the existing pillar dictionary, then ships that dictionary down to the minion(s)
14:21 Ahlee k, now work beckons. bbl
14:22 aarseniev oh.. found https://github.com/saltstack/salt/issues/18299 . Seems thats the reason why ext_pillar not works in masterless mode.. :(
14:23 julienlavergne @ahlee: i can not display the pillar data on the master or minions while my s3 ext_pillar config is good
14:23 julienlavergne 2015-07-07 16:22:35,420 [urllib3.connectionpool][INFO    ][16048] Starting new HTTPS connection (1): myfox-provisioning-salt.s3.amazonaws.com
14:23 julienlavergne 2015-07-07 16:22:35,508 [urllib3.connectionpool][DEBUG   ][16048] "GET /pillar/hcs-staging/top.sls HTTP/1.1" 200 83
14:23 julienlavergne 2015-07-07 16:22:35,509 [salt.utils.s3    ][DEBUG   ][16048] S3 Response Status Code: 200
14:23 julienlavergne 2015-07-07 16:22:35,509 [salt.utils.s3    ][DEBUG   ][16048] Saving to local file: /var/cache/salt/master/pillar_s3fs/pillar/myfox-provisioning-salt/pillar/hcs-staging/top.sls
14:24 julienlavergne 2015-07-07 16:24:01,195 [salt.loaded.int.pillar.s3][INFO    ][16048] Sync local pillar cache from S3 completed.
14:24 julienlavergne @ahlee: ubuntu@pmaster:~$ sudo salt '*' pillar.data gives me nothing
14:24 TyrfingMjolnir joined #salt
14:25 Furao joined #salt
14:26 Ahlee in the future, pastebin or gist instad of pasting to the channel.  looks like you're grabbing top.sls?
14:26 Ahlee i've never treid to do that (nor have i used s3fs as a pillar)
14:26 pullphinger joined #salt
14:27 Ahlee is this the s3fs fileserver, or an actual ext_pillar defined?
14:27 Ahlee julienlavergne: https://github.com/saltstack/salt/issues/23016
14:28 Ahlee that might be what you're running into.
14:28 quasiben joined #salt
14:28 fivmo joined #salt
14:28 fivmo joined #salt
14:29 toddnni_ joined #salt
14:30 quique joined #salt
14:30 quasiben joined #salt
14:31 julienlavergne @Ahlee: I use s3  as ext_pillar source not for my environment states (gitfs in this case)
14:32 venu0336 joined #salt
14:32 buxy joined #salt
14:33 Ahlee i don't get what you're doing with saving top.sls, then. ext_pillars don't write out files, they return dictionaries populated with hte new pillar values
14:34 Ahlee at least, not the one's i've written.
14:34 buxy What can go wrong when state.file.recurse doesn't install new files installed in the source directory on the master ? I tried to clean the cache on the minion with saltutil.clean_cache but it doesn't work.
14:36 otter768 joined #salt
14:36 Ahlee buxy: if you salt minion cp.list_master, do you see the new files?
14:36 streetmapp joined #salt
14:36 kal0pr joined #salt
14:36 malinoff joined #salt
14:36 julienlavergne @ahlee: I use this http://docs.saltstack.com/en/latest/ref/pillar/all/salt.pillar.s3.html
14:36 buxy Ahlee: no
14:37 Ahlee buxy: doesn't sound like the files you expect it to recurse exist, then
14:37 streetmapp hey folks presently having an issue with regards to salt-ssh and sudo where requiretty is enabled. here's the issue i opened on this: https://github.com/saltstack/salt/issues/25134
14:37 stephanbuys joined #salt
14:38 Ahlee julienlavergne: sorry, don't know what to tell you. This isn't what i was expecting and I don't have s3 access to test
14:39 Ahlee julienlavergne: i guess the only other thing i can suggest is to start simple, ensure you have some pillar value in s3 that matches everything
14:40 julienlavergne @Ahlee: good advice that I already practice
14:40 Ahlee osrry i couldn't be of more help
14:41 Gareth Ahlee: have you tried other returners?  seeing the same results with the exception?
14:41 buxy Ahlee: http://paste.debian.net/279863/ the files do exist but they are not "seen", how is that possible ?
14:42 Ahlee Gareth: memcache, redis, mongo all throw exceptions, but all different ones. Haven't dove too dep due to work and procrastincation
14:42 buxy (note the missing "build-all-images" and "test-foo")
14:42 Ahlee Gareth: they all appear similar, yet different
14:43 aqua^c joined #salt
14:44 otter768 joined #salt
14:45 quasiben1 joined #salt
14:46 Ahlee buxy: i got nothing.
14:46 Ahlee bbl for real this time.
14:48 LeProvokateur joined #salt
14:53 giantlock joined #salt
14:54 jondonas joined #salt
14:55 napsterX joined #salt
14:56 kal0pr joined #salt
14:56 jalbretsen joined #salt
14:57 Gareth Ahlee: going to give the redis returner a try here, see if I can duplicate the issue.
14:57 CeBe joined #salt
14:57 Ahlee Gareth: deubg level on client is where exception is logged
14:57 Gareth Cool.
14:57 Gareth will let you know what I find.
14:58 Ahlee it looks like modules/config.py's config method is being passed in to returners/__init__.py's _fetch_option, but virtualname is undefined
14:59 Ahlee so c_cfg = cfg('{0}'.format(virtualname), {}) returns the empty dict
14:59 buxy Ahlee: found the issue in the end, it was due to a change of time backwards (after a reboot), the master was not updating its file list considering it was uptodate/newer..
14:59 Gareth Ahlee: Nod.  That's what I was thinking too before I had to run off yesterday, that was the thought behind trying the different config structure.
15:00 Ahlee buxy: ah!
15:00 Ahlee buxy: well, that's annoying.
15:00 julienlavergne anybody with some knowledges about ext_pillar with s3 ?
15:01 Ahlee sprint planning, back this afternoon
15:01 pcn can I confirm what i believe I'm seeing?  It looks like _grains needs to be available via salt://_grains to work properly.
15:02 c10 joined #salt
15:03 icflournoy joined #salt
15:05 julez joined #salt
15:06 clintberry joined #salt
15:06 venu0336 joined #salt
15:08 sdm24 joined #salt
15:11 aw110f joined #salt
15:13 fyb3r joined #salt
15:14 julez joined #salt
15:15 kal0pr joined #salt
15:15 aanriot joined #salt
15:16 aanriot left #salt
15:18 colttt joined #salt
15:19 adelcast joined #salt
15:21 SheetiS joined #salt
15:22 dfinn joined #salt
15:22 JoshuaX joined #salt
15:30 aw110f_ joined #salt
15:37 Viper233 joined #salt
15:50 murrdoc joined #salt
15:50 tmclaugh[work] joined #salt
15:50 murrdoc whats an SPM formula
15:50 murrdoc why is techchat adding that shit to everything
15:52 elfixit joined #salt
15:52 msheiny joined #salt
15:53 theologian joined #salt
15:55 iggy pretty weak release notes
15:59 aparsons joined #salt
15:59 stoogenmeyer joined #salt
15:59 sdm24 Does anyone here have a "best way" to upgrade Salt minion on Windows machines? I've tried using the winrepo, and also manually copying the installer and cmd.run, but both options require me to RDP to the Windows minion to manually restart Salt
16:00 sdm24 I'm currently working with 2 test machines, but it will be harder to RDP to each and every production Windows minion
16:00 iggy the new installer doesn't require that I don't think
16:03 pullphinger joined #salt
16:06 iggy I like how this SPM thing was talked about in the general community before people just started adding it to formulas
16:06 VSpike sdm24: I have this which is hacky but seems to work
16:06 VSpike https://bpaste.net/show/14a252f52421
16:07 VSpike So I copy the minion installer and a small script to the target (i just put them on C:\ but you could do better
16:07 VSpike )
16:07 sdm24 Thats similiar to what I was doing
16:07 sdm24 https://gist.github.com/sdm24/6113a3b9d74da259fdaa
16:08 sdm24 but then I need to manually log on to the Windows minion to restart salt
16:08 andygabby joined #salt
16:08 stanchan joined #salt
16:08 VSpike I could only get it to work if I put the batch script on and ran that
16:09 VSpike Note the batch file is called with "start" which should separate the process from the salt minion
16:09 sdm24 oh I see yeah
16:09 spookah joined #salt
16:09 sdm24 you run that script on the minion
16:09 VSpike The second one, yes
16:10 VSpike I run upgrade-windows-minion on the master, passing some kind of glob or other selection
16:10 sdm24 do you call it with salt 'minion' cmd.script 'C:/upgrade-minion.cmd'?
16:11 VSpike no just ./upgrade-windows-minion -C 'G@saltversion:2014.7.5 and G@os:Windows and G@cpuarch:AMD64'
16:11 VSpike Or something like that
16:11 VSpike THe first script runs on the master and copies the installer and the cmd file to the minions, then cmd.run's it
16:12 sdm24 oh ok
16:12 VSpike You always seem to get errors from the script .. I think from the cmd.run and the test.version lines. But, it does seem to work even so
16:13 VSpike I can understand why the cmd.run call never returns a response .. that's why i put the test.version in there. Dunno why that doesn't seem to work.
16:14 VSpike I usually just let it run and then call salt-run manage.versions
16:15 conan_the_destro joined #salt
16:17 FRANK_T joined #salt
16:17 FRANK_T If I denied a a salt-minion key how can I take that key out of the denied list/
16:17 FRANK_T "?
16:18 murrdoc "We decided against using the .yml extension."
16:18 murrdoc who is this we , iggy / forrest
16:18 iggy 12:06 < iggy> I like how this SPM thing was talked about in the general community before people just started adding it to formulas
16:18 iggy forgot the </sarcasm> tag
16:18 sdm24 FRANK_T: you can use 'salt-key -d <minion ID>'
16:19 sdm24 that will delete the key from any/all lists
16:19 sdm24 and then when the minion tries to reconnect (defaults every 10 seconds), the key will reappear in pending
16:19 darkhuy joined #salt
16:19 iggy murrdoc: I'm not opposed to the idea, but I can already see problems with it that could have been brought up if they hadn't come up with this behind closed doors
16:20 FRANK_T Thank you that worked
16:20 twork joined #salt
16:20 soren joined #salt
16:20 murrdoc iggy:  word
16:21 darkhuy hey murrdoc :-)
16:22 murrdoc sup darkhuy
16:22 FRANK_T Do you guys know if salt creates a log file very time that a run a state?
16:22 Ahlee spm? where are you guys seeing this?
16:22 iggy Ahlee: formulas commits
16:22 Ahlee ah
16:22 iggy you're not one of the cool kids
16:22 Ahlee it's true.
16:23 Ahlee I don't use them
16:23 Ahlee probably should
16:23 FRANK_T I have a long list of states but if one fails I cannot see it and the only what that I can check is if a run the command and a >> to a file
16:23 iggy or I guess... you have a real job and don't have time for that shit
16:23 iggy FRANK_T: that's what the job cache is for
16:23 sectionme joined #salt
16:24 FRANK_T iggy where is that?
16:24 FRANK_T .
16:24 iggy FRANK_T: there are runners and modules to access it
16:24 FRANK_T ../var/log/salt
16:24 FRANK_T could you please point me to the doc if you have time.
16:24 iggy FRANK_T: another helpful item... state_verbose: False
16:24 iggy and maybe failhard: True
16:24 iggy if you're into that kind of thing
16:25 sdm24 or --state-output=mixed at the end of the salt call
16:25 iggy FRANK_T: http://docs.saltstack.com/en/latest/ref/runners/all/salt.runners.jobs.html#module-salt.runners.jobs
16:25 darkhuy any idea why cmd_async() is telling me "kwarg" is an unexpected keyword argument? The python api docs show kwarg as a valid param :-/
16:25 ipmb joined #salt
16:27 dfinn joined #salt
16:29 amcorreia joined #salt
16:31 tets joined #salt
16:31 tets hey gals
16:32 aqua^c joined #salt
16:32 jaybocc2 joined #salt
16:33 twork joined #salt
16:39 murrdoc sup girlie
16:42 nodens hi there
16:42 spookah joined #salt
16:42 nodens any syndic users ? I'm a bit lost
16:43 nodens I'm searching the best way to sync *pillars* on my sub-masters
16:43 spark_ wouldn't it make more sense to have different file extension for pillar files instead of using .sls for everything??
16:43 nodens (I use file.recurse for states, but it won't work with pillars)
16:43 iggy spark_: it's all yaml+jinja
16:45 spark_ iggy: I see.. it's somewhat confusing as a noob :)
16:46 nodens since I want a way to centrally manage minions, whether there are behind a syndic or not
16:46 nodens and I don't want many git repo (or submodules)
16:47 nodens but I don't want minions behind a syndic to be able to get pillars for other, either
16:47 nodens so I'm out of idea... pushing the pillar data with rsync seems messy
16:47 nodens I'd gladly take any idea
16:49 twork i'm making my first dip into formulas. the various 'gitfs_remotes' settings described at http://docs.saltstack.com/en/latest/topics/tutorials/gitfs.html#multiple-remotes all seem to suggest grabbing formulas from upstream...
16:49 twork ...but right at the top of http://docs.saltstack.com/en/latest/topics/development/conventions/formulas.html it says not to do that, instead clone them to the filesystem (or wherever, locally) and use them from there. what am i missing?
16:50 twork if "use them locally" is the recommended way, why describe them remote way first?
16:51 julienlavergne why my minion don't get my ext_pillar ?
16:51 iggy twork: clone'ing doesn't necessarily mean locally... in github terms, "fork"
16:52 twork iggy: i'm both a git n00b and a salt n00b, so bear with me... doesn't "fork" mean, "put a copy of the named archive into [wherever]", usually the filesystem?
16:53 iggy twork: no, it means put a copy under your own user account
16:53 Ahlee in github parlance, no. It means more of "create a copy of this repo at this point in time"
16:53 KyleG joined #salt
16:53 KyleG joined #salt
16:53 twork ah, ok, so we presume that twork has an account on github as a prereq for all this?
16:53 iggy gitfs_remotes: - https://github.com/iggy/salt-formula.git
16:54 Ahlee since git is a distributed version control system, the only thing that makes github the 'centralized repo' is user agreement. From git's standpoint, there is no 'main' repo, just different repos at different points in time
16:55 iggy and often different sets of branches (but not necessarily)
16:56 twork i've had all that explained to me a bajillion times. but for some reason my brain is stuck in the filesystem world.
16:56 hal58th joined #salt
16:56 Ahlee probably because git resembles a file system :)
16:56 twork where do the bits go?
16:56 iggy if that's the case, you can indeed skip the whole gitfs_Remotes route and just throw them all in your filesystem
16:56 Ahlee under the hood there's a .git directory that contians the metadata
16:56 Ahlee there 'thar be dragons'
16:57 c10 joined #salt
16:57 sdm24 VSpike: Thanks for your upgrade script. It works perfectly for me. The old .exe's need to be removed, but thats simple
16:57 iggy then you just end up with more entries under file_roots
16:58 JDiPierro joined #salt
16:58 Ahlee VSpike: indeed, thanks for linking that. totally going to steal that for when i finally upgrade
17:00 keimlink_ joined #salt
17:00 twork okay so... step me through this guys, treat me like the idiot i am:
17:00 twork i find a formula i want to use. at http://wherever
17:01 twork i make a directory, say, /salt/formulas
17:01 twork normally what i'd expect is, say, rsync from http://wherever into /salt/formulas/wherever
17:02 iggy most (all?) formulas are going to be on github, so you're going to have to use git at some level
17:03 twork yeah i get that much, but what i mean to ask is:
17:03 fxhp joined #salt
17:03 iggy but you can just git clone https://github.com/iggy/salt-formula.git into /srv/formulas and then add /srv/fromulas/salt-formula to file_roots in the master config file
17:04 twork aha, so i needn't tell salt to do the git'ing?
17:04 iggy you don't _have_ to
17:05 iggy but you can
17:05 impi joined #salt
17:05 iggy it is kind of confusing (and I'd suggest once you get it all figure out, that you suggest documentation changes to the pages in question)
17:05 twork or... i git into /wherever/snapshot; i ln -s /serever/snaphot into /got/formuas/foo ...then i point salt at that point? (for instance, i know i don't have to but i'm groping for a working case example)
17:05 darkhuy so cmd_async isn't implemented yet for salt-ssh client :-( no wonder it gave errors. derp
17:07 ajw0100 joined #salt
17:08 twork part of my trouble (obviousl) is that i don't quite grok the difference between "git [whatever]" and "copy from /there to /here"
17:08 twork i know what a snapsho is
17:09 twork and i get it that a clone takes a snapshot of a repo at a point in time
17:09 twork (right so far?)
17:10 Ahlee yes
17:10 Ahlee git clone will create a copy of the repo at the location specified
17:11 twork okay, i think the clouds are starting to part... it's a separate step, 'get, go grab stuff' and 'filesystem, copy stuff over for salt to use'
17:11 theologian joined #salt
17:11 iggy mkdir /srv/formulas ; cd /srv/formulas ; git clone https://github.com/saltstack-formulas/salt-formula.git ; nano /etc/salt/master (and add /srv/fromulas/salt-formula to file_roots:base)
17:11 iggy that's the workflow
17:13 twork garn. that's what i thought to begin with, now i'm back to confused again. doesn't that workflow expose me to trouble as the formula gets updated upstream? isn't it better to 'git clone /some/staging/spot ; cp staging/spot /srv/salt/formulas'?
17:14 Ahlee git only updates when you tell it to
17:14 Ahlee your local clone will not change until you tell it to
17:14 iggy ^
17:14 twork i know that.
17:14 iggy and salt doesn't update it either
17:15 Ahlee so, yes. If you update your git clone (git pull), you're susceptible to any bugs/breaking changes introduced
17:15 Ahlee if you don't update, you don't run into problems
17:15 iggy so no worries (as long as the formula is in a working state when you clone it the first time ;)
17:15 twork har
17:15 iggy we break things from time to time
17:15 viq joined #salt
17:15 gcfhvjbkn joined #salt
17:15 Ahlee since i'm old and crotchity I don't run anything off repos i don't control
17:15 twork ok but it still sounds like it's a good idea to 'git pull' into a staging area rather than right into the live area?
17:16 Ahlee twork: indeed.
17:16 iggy true
17:16 Ahlee always better to work in controled environments
17:16 Gareth Ahlee: fyi, haven't been able to duplicate this returner issue.  Installed redis and it worked without exception.
17:16 iggy but unless you need new features that are added, there's likely very little reason to update later on at all
17:17 Ahlee Gareth: salt 2015.5.2, python 2.x?
17:17 Gareth Ahlee: yessir.
17:17 * Ahlee blows away the directory, starts over
17:17 Gareth tried both the packaged version and the current 2015.5 branch.
17:17 twork ok. so: "hey salt, git [formula]" is only wise if [formula] is a copy that i've staged and vetted. the docs make it look like the suggested method is 'git with trust'.
17:18 Ahlee twork: different people trust different things
17:18 twork in which case, having salt do the git'ing strikes me as bad advice to the n00b.
17:18 Ahlee twork: it's the same as some people point their systems at the distro's mirrors, instead of their own
17:18 iggy twork: that's why the docs say to fork your own copy first
17:19 SeeDickCode joined #salt
17:19 iggy don't add https://github.com/saltstack-formulas/salt-formula.git to gitfs_remotes:
17:19 Ahlee Gareth: this is really, really strange. Are you using the system's interpretor?
17:19 twork i must have missed that point in the cocs
17:19 iggy fork your own copy on github and add https://github.com/iggy/salt-formula.git
17:19 twork oh oh! the lights come on!
17:20 Gareth Ahlee: the packaged version of python?
17:20 twork 'git fork into github' != 'git whatever's up there now into use by salt now and forevermore'
17:20 twork (even as it changes up there)
17:21 Ahlee Gareth: sorry, disregard. something's clearly messed up on my system. I'm bootstraping another system and going to start form scratch with 2015.5.2
17:21 twork the meaning of a fork that stays on github wasn't sinking in
17:21 iggy twork: if you go look at https://github.com/saltstack-formulas/salt-formula
17:21 Gareth Ahlee: no worries.
17:21 iggy you should see a button that says fork
17:21 iggy (top right of the page)
17:21 twork iggy: i did, but for some reason i figured that was for use by developers. but, of course, git is distributed! a developer is me!
17:22 iggy when you click that button you'll end up with https://github.com/twork/saltstack-formula
17:22 Ahlee Gareth: thanks for looking into it. sorry to probably lead you astray
17:22 twork iggy: yes yes yes, thank you. it becomes clear.
17:22 pdayton joined #salt
17:22 SeeDickCode joined #salt
17:22 Gareth Ahlee: no worries :)
17:22 twork lordy i feel better. thank you.
17:23 SeeDickCode joined #salt
17:26 darkhuy Hey guys, is there a way to hide debug/info log messages when using the python client?
17:27 gcfhvjbkn joined #salt
17:27 Ahlee do you mean the output of the state runs?
17:28 Ahlee or is your client being invoked with salt -l debug \* test.ping and you no longer want that to happen?
17:28 julienlavergne joined #salt
17:28 darkhuy when using the python client via salt.cmd('*', 'test.ping')
17:29 JDiPierro joined #salt
17:29 Ahlee yes, there's a way to surpress that, let me find the invocation
17:29 darkhuy the logging defaults to debug and outputs loads of info
17:29 Ahlee i think it's quiet=True
17:30 darkhuy as a keyword arg for salt.cmd?
17:30 Ahlee i thought so, but that doesn't look right.
17:31 Sokel Can a grain have multiple values rather than using sub-keys? If so, how does that get set
17:31 napsterX_ joined #salt
17:32 JDiPierro joined #salt
17:32 darkhuy Sokel: I think it would have to be a grain value that contains a dict of your other values
17:32 Ahlee Sokel: grains.setval foo [bar,baz]
17:32 Ahlee would create a list under foo of bar and baz
17:33 Ahlee darkhuy: sorry, it's been a while since i used local client and have been going over the api
17:34 darkhuy Ahlee: no rush
17:38 Ahlee darkhuy: well, there's the cheap way here of salt.cmd_async
17:39 Ahlee i know there's a way to do this, i'm just blanking on it
17:39 spark_ joined #salt
17:40 darkhuy Ahlee: problem is salt-ssh client doesn't have async implemented yet
17:42 SeeDickCode joined #salt
17:42 markm joined #salt
17:44 hasues joined #salt
17:44 hasues left #salt
17:47 JoeHazzers is there any way of getting a "satisfies" behaviour, e.g. a firefox formula can satisfy "web-browser", and so can a chrome formula at the same time, and something depends on "web-browser", which can be satisfied by one or both?
17:47 Ahlee darkhuy: looks like I was doing this with a runner. status = runner.cmd('manage.status', ['output=False'])
17:47 Ahlee passing hte same option to the client doesn't seem to be implemented
17:48 darkhuy hmmm, wonder if initializing a python logger before importing salt.client will pipe it to the new logger where i can have it write file instead
17:49 dimeshake joined #salt
17:51 Ahlee Probably.
17:53 quasiben joined #salt
17:53 riftman joined #salt
17:53 quasiben joined #salt
17:54 spark_ joined #salt
17:55 jasonrm joined #salt
17:55 gcfhvjbkn joined #salt
17:58 denys joined #salt
17:59 Sokel left #salt
17:59 X67r joined #salt
18:00 jasonrm joined #salt
18:01 ajw0100 joined #salt
18:04 jasonrm joined #salt
18:05 c10 joined #salt
18:06 dimeshake joined #salt
18:08 ageorgop joined #salt
18:09 spark_ joined #salt
18:10 lictor36 joined #salt
18:13 jasonrm joined #salt
18:17 baweaver joined #salt
18:19 Slimmons joined #salt
18:19 twork JoeHazzers: i think a flat "any of x, y, or z" function would run counter to the spirit of the system. "perfectly predictable control" and all that.
18:20 aqua^c joined #salt
18:21 twork but in some cases, like under file.managed for instance, there are fallthrough functions.
18:21 twork dunno if there's the same for pkg.managed
18:22 twork [thus spake the complete n00b]
18:24 spark_ joined #salt
18:26 Slimmons I'm trying to use pip.installed using a requirements.txt file, but I'm getting this error, "Failed to install package Scrapy. Error: Downloading/unpacking Scrapy                 Cannot fetch index base URL https://pypi.python.org/simple/                 Could not find any downloads that satisfy the requirement Scrapy               Cleaning up...               No distributions at all found for Scrapy".  When I go to that site, the Scra
18:26 twork JoeHazzers: looks like not, at least from a look through salt.states.pkg
18:28 linjan joined #salt
18:29 twork ...and it looks like 'pkg.managed' was a figment of my imagination...
18:29 riftman joined #salt
18:29 julez joined #salt
18:31 quique joined #salt
18:38 spark_ joined #salt
18:45 aron_kexp joined #salt
18:45 losh joined #salt
18:45 dimeshake joined #salt
18:46 bja joined #salt
18:47 iggy Slimmons: does that system have access to the internets?
18:47 iggy salt 'minion' http.query https://pypi.python.org/simple/
18:47 murrdoc ~# salt-call config.get master
18:47 murrdoc [INFO    ] SaltReqTimeoutError: after 60 seconds. (Try 1 of 3)
18:47 murrdoc [INFO    ] SaltReqTimeoutError: after 60 seconds. (Try 2 of 3)
18:47 murrdoc sk[INFO    ] SaltReqTimeoutError: after 60 seconds. (Try 3 of 3)
18:47 murrdoc watup with that
18:48 iggy the minion isn't running?
18:48 murrdoc the minion is running
18:48 murrdoc status salt-minion
18:48 murrdoc salt-minion start/running, process 23669
18:48 iggy the minion isn't running properly?
18:48 jdesilet joined #salt
18:49 * iggy runs
18:49 murrdoc …
18:50 iggy did you look at the logs? is that PID correct? Is there only one copy of the minion running?
18:50 Slimmons it does iggy, but I think you're on the right track.  I think this is a corporate firewall issue.
18:50 murrdoc iggy:  yeah only one
18:50 Slimmons Once I rule that out (or resolve it), I'll get back to you
18:52 iggy murrdoc: I've seen that before, trying to remember what I did
18:52 murrdoc its the returner
18:52 * murrdoc fixed it
18:52 bja is it possible to capture the result of executing a module during sls state rendering?
18:52 * iggy stops helping murrdoc
18:52 murrdoc sank u
18:53 murrdoc i mean for the help
18:53 murrdoc not for stopping :D
18:53 iggy bja: question unclear, do you want to see the rendered yaml? (salt-call -l debug will show that)
18:56 gcfhvjbkn1 joined #salt
18:58 s_kunk joined #salt
19:00 dimeshake joined #salt
19:00 [dee] joined #salt
19:04 wedgie joined #salt
19:04 bja is there a way to run debug salt modules locally via pdb?
19:08 gadams joined #salt
19:10 soren joined #salt
19:11 dimeshake joined #salt
19:11 hybridpollo joined #salt
19:14 jbub joined #salt
19:15 iggy there was some talk about that at saltconf, but I Think it was in one of the tech sessions (I don't know if htey were filmed)
19:15 iggy look for Mike Place's videos from saltconf
19:16 sgargan joined #salt
19:17 DammitJim joined #salt
19:19 fllr joined #salt
19:19 fllr Guys. The rabbitmq state is completely broken...
19:19 fllr I keep getting errors after errors, and it used to work before
19:20 ajw0100 joined #salt
19:22 dopesong joined #salt
19:24 icflournoy joined #salt
19:27 evidence joined #salt
19:29 numkem fllr: I'd suggest to open an issue with what you find
19:29 spark_ joined #salt
19:30 dimeshake joined #salt
19:32 spark__ joined #salt
19:32 Bryson joined #salt
19:34 quasiben1 joined #salt
19:41 dopesong_ joined #salt
19:43 JoeHazzers twork: thanks (for earlier)
19:46 khebbie joined #salt
19:49 conan_the_destro joined #salt
19:51 baweaver joined #salt
19:53 dimeshake joined #salt
19:58 spark_ joined #salt
19:59 chiui joined #salt
20:00 sgargan joined #salt
20:01 quasiben joined #salt
20:01 theologian joined #salt
20:01 iwishiwerearobot joined #salt
20:02 c10 joined #salt
20:03 Nazzy joined #salt
20:06 quasiben1 joined #salt
20:09 aqua^c joined #salt
20:12 mapu joined #salt
20:21 stanchan joined #salt
20:29 Aoreias joined #salt
20:29 yota joined #salt
20:30 _JZ_ joined #salt
20:35 spark_ joined #salt
20:38 tkharju joined #salt
20:44 dimeshake joined #salt
20:47 quique joined #salt
20:48 aparsons_ joined #salt
20:49 Tecnico1931 joined #salt
20:50 Ahlee Gareth: You around? Did you happen to have your returner running on a host that resolves to salt?
20:53 pullphinger15 joined #salt
20:53 Ahlee fresh install of ubuntu 14.04, 2015 from salt's ppa, only config is to point minion to localhost, and redis config'd per https://docs.saltstack.com/en/develop/ref/returners/all/salt.returners.redis_return.html to point to a redis server (that the python interpreter can access). Still wants to return to 'salt'
20:53 Ahlee if i have to set this on the minion i'm gonna lose it
20:56 CeBe joined #salt
20:56 Ahlee god. damnit.
20:56 Ahlee so masters shipping returner config to minions is broken in 2015
21:00 gcfhvjbkn joined #salt
21:00 spark_ joined #salt
21:01 Ahlee I'm seriously mad enough to punch babies over this
21:02 murrdoc ?
21:02 Ahlee i spent most of yesterday dicking around in modules/config.py and returners/__init__.py
21:03 Ahlee in 0.17.ahlee, i set my returner config in the master config, which is shipped to the minions at start
21:03 Ahlee in 2015.5.2, that appears non-functional. as soon as i dropped it in the minion config *poff*
21:03 Ahlee which is like poof, only less cool
21:04 Ahlee opening issue now, and hopefully PR to restore functionality
21:04 cmb joined #salt
21:04 cmb joined #salt
21:05 sectionm1 joined #salt
21:11 c10 joined #salt
21:13 TheoSLC joined #salt
21:13 TheoSLC Greetings!
21:13 amcorreia joined #salt
21:13 TheoSLC Is it possible to use the salt returner to only return failures or changes to the returner?
21:15 Cliph joined #salt
21:16 Ahlee TheoSLC maybe with --state-output ?
21:17 Ahlee Gareth: https://github.com/saltstack/salt/issues/25215
21:18 TheoSLC Ahlee: I'll give that a try.  thanks!
21:19 Cliph Hello all. Is AssociatePublicIpAddress known to be broken when using salt-cloud to create ec2 instances? (using 2015.5.2)
21:20 spark_ joined #salt
21:21 Cliph profile looks like this: https://gist.github.com/cliph/df64877fdbe0879aaac3
21:25 baweaver joined #salt
21:26 Ztyx joined #salt
21:35 TheoSLC Ahlee: the --output and --state-output flags seem to have no affect on the returner data, only the cli output.
21:37 hal58th joined #salt
21:37 SheetiS joined #salt
21:38 Gareth Ahlee: I had the redis config in the minion config.
21:38 Gareth Ahlee: and the host was just localhost.
21:41 Ahlee TheoSLC: to a _returner you must go, then
21:41 Ahlee if not ret.get('result'): do_the_needful()
21:42 Ahlee Gareth: yeah. I'm now tearing apart how configurations got shipped in 0.17 vs how they do now
21:42 Ahlee probably going to give up though and just move the returner config into the minion config files
21:43 Ahlee anyway, bbl
21:45 bluenemo joined #salt
21:45 bluenemo joined #salt
21:46 Gareth Seems like a major issue if it worked and then didn't though.
21:49 campee joined #salt
21:50 aparsons joined #salt
21:50 tvinson i'm trying something like http://pastebin.com/ErBw0aHH am i getting too far afield? it's not throwing errors but is also not adding the role specified in the sls. is this a reasonable way to manage host grouping in pillars? is there a way to get salt to dump the post-template top.sls?
21:50 iwishiwerearobot joined #salt
21:53 tvinson errr...i think i spotted my own problem. missed the match: list
21:54 keimlink joined #salt
21:55 tvinson ...now to figure out how to iterate over the groups
21:56 murrdoc -b batch mode
21:56 TheoSLC Ahlee: Yes, It looks like any filtering of the returner will have to be added to the returner code.
21:58 aqua^c joined #salt
22:00 JDiPierro joined #salt
22:03 Guest89 joined #salt
22:05 twork i need some guidance. here's my situaton: i have a class of machines that all need to have the same package installed. they're all configured in mostly the same way, except that on each host, the group membershp of a class of files needs to differ. same path, different gid depending on the minion
22:05 twork i know this is a template situation but i haven't found a way to do quite what i want
22:05 TheoSLC Ahlee: The content of ret.get('return') should be formatted according to --output and --state-output.  Agreed?
22:07 dimeshake joined #salt
22:07 TheoSLC And then the returner can do. if not ret.get('return'): DO_NOTHING
22:07 twork i'm reading here: http://docs.saltstack.com/en/latest/topics/best_practices.html#modularity-within-states
22:08 twork ...and i get the gist but i'm not yet to the point of generalizing things to fit my situation
22:08 jasonrm joined #salt
22:10 TheoSLC twork: What needs to differ?
22:10 twork the gid
22:10 c10 joined #salt
22:10 TheoSLC twork: I mean, what does salt need to do differently on your minions depending on gid?
22:11 twork oh, sorry. a couple of paths under /var/log need to be owned by that group. does that answer your question?
22:12 twork ack, let me try this again:
22:13 twork salt needs to do the same thing, in each case. install the package, control the metadata of the paths. but what the gid is needs to depend on the minion.
22:14 twork did that make more sense?
22:14 TheoSLC twork: http://paste.ubuntu.com/11838349/
22:14 twork sorry, it's been a long day
22:15 TheoSLC twork: set a "gid" grain for each minion and do a state that looks like the one posted?
22:15 twork that looks like exactly what i need
22:15 sgargan joined #salt
22:16 twork well close enough. the paths happen to be files not directories. but this should get me going. thanks much.
22:16 TheoSLC np
22:17 lictor36 joined #salt
22:18 iggy Ahlee: #pillar_opts: True
22:20 iggy config.get reads pillar:master:<var> and that default changed in 2014.7 which broke a lot of people's expectations of what config.get would return
22:22 yomilk joined #salt
22:23 Cliph I appear to be experiencing this issue; https://github.com/saltstack/salt/issues/19288 with 2015.5.2
22:24 Cliph Have something changed or has there been a regression?
22:26 dendazen joined #salt
22:27 TheoSLC Cliph: somebody earlier was saying the same thing.
22:28 TheoSLC Cliph: you should re-open the issue
22:32 giantlock joined #salt
22:35 baweaver joined #salt
22:36 Cliph Ok TheoSLC, will do
22:37 iggy that was merged to develop after 2015.2 was forked
22:37 iggy so it should be in 2015.8
22:37 Cliph ooo
22:38 iggy err, 2015.5
22:38 iggy that was merged to develop after 2015.5 was forked
22:38 iggy I still call it 2015.2 sometimes for some reason
22:43 Cliph OK, good to know, I'll try EIPs until 2015.8 ships
22:44 Ztyx joined #salt
22:52 clintberry joined #salt
22:57 sgargan joined #salt
23:04 Bryson salt:// URLs start at /srv/salt right?
23:04 slav joined #salt
23:04 Bryson by default i mean. so salt://top.sls is /srv/salt/top.sls
23:04 murrdoc yes
23:05 murrdoc no
23:05 slav hi do you guys know how to deploy salt minion agents in an active directory domain?
23:05 slav is there any msi package for salt?
23:05 twork argh. i'm back with my file permission setting issue again. i thought TheoSLC clued me in, but i think i was mistaken. trying again to clarify my issue:
23:05 iggy twork: what's the difference between the systems
23:06 twork let's say something like:
23:07 twork system class 'foo', i want the gid to be 'x'. system class 'bar', 'y'. and so fourth. in each case though, the path '/var/log/this' should be owned by that group.
23:07 twork did that make sense?
23:07 iggy yes, but what is different between the system "classes"?
23:08 twork all those systems are the same otherwise for the purpose of this module
23:08 iggy are they different distros?
23:08 twork no
23:08 iggy are they grouped somehow?
23:08 whytewolf slav: is there an easy way to tell them apart in software
23:08 twork the difference is that we, the site, classify them differently. that's pretty much it.
23:08 whytewolf err twork
23:08 twork we file them under a category.
23:08 twork of our own devising.
23:08 iggy that's non-specific
23:08 iggy do you have that mapping in salt somehow?
23:09 twork none defined yet but they could (maybe the answer is "that's what i'm trying to figure out")
23:09 iggy do you have that mapping anywhere?
23:09 twork i'll get more concrete:
23:10 Cliph could salt determine the appropriate gid depending on a nodegroup, a tag or the hostname?
23:10 iggy or subnet
23:10 iggy or .... anything?
23:11 X67r_ joined #salt
23:11 iggy database lookup?
23:11 iggy parse a csv?
23:11 twork on one host (or set of hosts), there's a group, let's call it "foo". on another set of hosts, the group is "bar". in each case, we want that group to own "/var/log/mumble". does that help?
23:12 twork this seems like it should be easy.
23:13 iggy {% if salt['group.getent']('foo') %}{% set gid = '500' %}{% else %}{% set gid = '600' %}
23:13 twork ahhhhh
23:13 iggy no, that won't work
23:13 iggy I read the docs wrong
23:13 twork drat!
23:14 twork what do i pay you for!
23:14 whytewolf goofing off
23:14 twork it seems like it's getting me closer though
23:14 iggy There's something in here
23:14 ericof joined #salt
23:14 iggy ahh
23:15 iggy just group.info
23:15 iggy {% if salt['group.info']('foo') %}{% set gid = '500' %}{% else %}{% set gid = '600' %}
23:16 * twork gives iggy a great big wet smooch
23:17 iggy _but_
23:17 twork i never would have figured this out on my own, not at my present state
23:17 twork oh dear
23:17 iggy long term, you probably want to make that mapping accessible to salt as either a grain or via ext_pillar or something
23:17 sgargan joined #salt
23:17 iggy that way you don't end up with a bunch of if's all over the place
23:18 twork hm... that may have been the direction i was looking when i started this on my own, and got all turned around
23:19 twork but this should help me hit my deadline
23:19 twork the grain thing i mean. i've only glanced at ext_pillar so far
23:21 iggy a lot of people also don't realize you can write pillars using the python renderer (vs a full blow ext_pillar)
23:22 TheoSLC joined #salt
23:23 pdayton joined #salt
23:27 twork oh, just to answer a question further up, about my sets of hosts, is there some way to programmatically determine which category a minion is in? the answer is no, not really.
23:27 twork that's a problem i should solve.
23:27 twork but no not yet.
23:28 sgargan joined #salt
23:28 twork ideally their hostnames would fit a pattern or something, but they don't.
23:31 SheetiS I use a yaml file and external pillar to accomplish that, but there are a number of ways.  Some people use the reclass ext pillar as well: http://docs.saltstack.com/en/latest/ref/pillar/all/salt.pillar.reclass_adapter.html.  You could also just have a very large and ugly top file (I don't like that option personally).
23:32 whytewolf i typically use a role grain that gives servers roles and seperate them out like that
23:33 SheetiS that can work as well, but just be mindful of the potential to expose additional data to a minion when it can set its own roles -- especially if different pillar data is exposed to different roles.
23:33 iggy ^
23:34 whytewolf SheetiS: correct. I don't use pillars for sensative data
23:34 whytewolf or only use minion_id for anything that would be sensative
23:38 sgargan joined #salt
23:39 iwishiwerearobot joined #salt
23:45 Bryson I run `sudo salt-call state.highstate` and I get an endless (literally) stream of “[INFO    ] Failed to get mtime on /var/db/etcupdate/current/sys, dangling symlink ?” …
23:47 aqua^c joined #salt
23:51 aqua^c joined #salt
23:59 subsignal joined #salt

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