Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-08-18

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

All times shown according to UTC.

Time Nick Message
00:08 snuffeluffegus joined #salt
00:09 bhosmer joined #salt
00:18 capitalfellow joined #salt
00:25 jalbretsen joined #salt
00:40 bhosmer joined #salt
00:50 bhosmer joined #salt
00:59 IOMonster joined #salt
00:59 IOMonster joined #salt
00:59 sectionme joined #salt
01:01 schimmy joined #salt
01:03 schimmy1 joined #salt
01:05 thayne joined #salt
01:08 rojem joined #salt
01:08 icebourg joined #salt
01:09 knot`` joined #salt
01:09 icebourg joined #salt
01:10 icebourg joined #salt
01:11 icebourg joined #salt
01:12 icebourg joined #salt
01:13 icebourg joined #salt
01:19 knot`` would anyone be able to point me in the right direction to resolve this issue?
01:19 knot`` >>> import salt
01:19 knot`` [CRITICAL] Unable to import msgpack or msgpack_pure python modules
01:19 knot`` im just trying to get salt going locally on osx10.9.3 mavericks
01:19 mosen an easier way (which might not be updated for 2014.7) is to install via homebrew
01:19 mosen but if you want, you can install msgpack via pip i think
01:19 mosen I'm not really sure about the other dependencies
01:21 knot`` install salt via homebrew? or msgpack?
01:21 knot`` when i install msgpack via homebrew, and try to import salt, I get the same error
01:22 mosen sorry i mean install salt via homebrew
01:22 knot`` np, that's actually what i did: "brew install saltstack"
01:23 mosen ahh right
01:23 knot`` is this import part unncessary then?
01:23 knot`` https://github.com/saltstack/salt/issues/7913
01:23 knot`` might be an open issue still
01:24 mosen why import salt as a python module? is that in salt-ssh?
01:24 mosen I see the output from that issue matches yours
01:25 mosen I wrongly assumed you were using salt-minion on osx
01:25 knot`` here's what im trying to do
01:26 knot`` just setup a master salt server locally on my mac for some local vagrant boxes i'll be booting up
01:26 mosen ok
01:27 knot`` here is the guide i was trying to follow, but failed at this part: http://docs.saltstack.com/en/latest/topics/tutorials/walkthrough_macosx.html#install-salt
01:27 knot`` the difference might be I used brew to install salt instead of pip. (pip install salt failed for me)
01:28 mosen naw, brew should work fine to resolve all the dependencies
01:28 knot`` ok
01:28 mosen its funny though because the guide is using python from homebrew
01:29 knot`` that is odd, why didnt he just tell people to install salt through brew as well?
01:31 mosen I'm not sure, maybe the saltstack recipe from homebrew installs into the system pythons site-packages
01:32 mosen in any case, what failed when you 'pip install salt'
01:36 knot`` I PM'ed you a gist
01:37 bhosmer joined #salt
01:41 n8n joined #salt
01:46 ml_1 joined #salt
01:53 bhosmer joined #salt
01:54 homelinen joined #salt
02:00 sectionme joined #salt
02:21 ramishra joined #salt
02:25 ramishra joined #salt
02:26 ramishra joined #salt
02:30 schimmy joined #salt
02:33 schimmy1 joined #salt
02:35 ramishra joined #salt
02:39 bhosmer joined #salt
02:39 snuffeluffegus joined #salt
02:40 tkharju1 joined #salt
02:49 srage joined #salt
02:59 ramishra joined #salt
03:01 sectionme joined #salt
03:09 TyrfingMjolnir joined #salt
03:10 vmdsch1 joined #salt
03:16 pastacino joined #salt
03:16 ramishra joined #salt
03:24 scoates joined #salt
03:27 snuffeluffegus joined #salt
03:31 schimmy joined #salt
03:31 srage_ joined #salt
03:34 srage__ joined #salt
03:41 schimmy joined #salt
03:46 schimmy joined #salt
03:49 pastacino joined #salt
03:54 rightas joined #salt
03:56 TyrfingMjolnir joined #salt
04:02 sectionme joined #salt
04:13 icebourg joined #salt
04:13 srage joined #salt
04:26 thayne joined #salt
04:27 bhosmer joined #salt
04:29 TyrfingMjolnir joined #salt
04:32 icebourg joined #salt
04:33 tkharju1 joined #salt
04:39 jalbretsen joined #salt
04:42 nnion joined #salt
04:43 tkharju1 joined #salt
04:44 IOMonster joined #salt
04:50 ramteid joined #salt
04:54 knot`` joined #salt
05:02 sectionme joined #salt
05:04 srage joined #salt
05:04 dccc__ joined #salt
05:11 kingel joined #salt
05:12 rightas salt vs chef ?
05:13 Corey rightas: Oh yes. Because this is an objective forum.
05:26 robawt aw they left
05:26 robawt i was going to say "why not both"
05:27 Kakwa and why not add puppet and ansible?
05:28 Kakwa by the way:
05:28 Kakwa channel: #ansible (4458)  channel: #puppet (4909)  channel: #salt (3792)  channel: #chef (2867)
05:28 Kakwa (statistic on freenode over last week)
05:28 Kakwa +s
05:30 mosen I think #puppet is cheating by having so many spam bots hehe
05:30 knot joined #salt
05:32 Kakwa another thing interesting is that on #salt, the 10 most active users weight 50% of the posts
05:33 Kakwa it's more like a third on the other channels
05:35 mosen hmm, why do you think that is?
05:39 mosen from anecdote, puppet is mostly one off troubleshooting questions. There are people here who are long time users or developers tend to talk a lot, and it's not always in the context of troubleshooting something.
05:40 Kakwa one interpretation could be that salt commity is smaller but really active
05:41 Kakwa but, it's only one week, it seems a bit light to be statisticaly significant ^^
05:44 mosen yep
05:45 Kakwa especialy in the middle of the summer
05:59 orion_ joined #salt
06:01 colttt joined #salt
06:01 cym3try joined #salt
06:02 martoss joined #salt
06:03 sectionme joined #salt
06:05 TyrfingMjolnir joined #salt
06:16 bhosmer joined #salt
06:25 MrTango joined #salt
06:27 kingel joined #salt
06:41 CeBe joined #salt
06:44 kermit joined #salt
06:44 pmarton joined #salt
06:44 pmarton hi
06:49 Nazca__ joined #salt
06:51 kingel joined #salt
06:53 kingel joined #salt
07:06 slav0nic_ joined #salt
07:11 matthiaswahl joined #salt
07:13 CeBe joined #salt
07:15 ml_1 joined #salt
07:15 schimmy joined #salt
07:18 alanpearce joined #salt
07:20 Flusher joined #salt
07:28 linjan joined #salt
07:34 ggoZ joined #salt
07:37 martoss joined #salt
07:49 sectionme joined #salt
07:53 CeBe joined #salt
07:54 jgelens joined #salt
08:01 N-Mi joined #salt
08:05 bhosmer joined #salt
08:31 ghartz joined #salt
08:33 n8n joined #salt
08:41 davromaniak hello
08:41 davromaniak is it possible to use file.blockreplace with a jinja template file ?
08:42 srage_ joined #salt
08:42 davromaniak I'm trying to automate the adding of bridges in my virtualization infrastructure, and I don't want to have all /etc/network/interfaces to be managed by salt, but only the bridges part
08:47 pastacino joined #salt
08:48 TheThing joined #salt
08:50 ghaering joined #salt
08:58 TheThing joined #salt
09:02 TheThing joined #salt
09:06 darkelda joined #salt
09:10 darkelda joined #salt
09:10 darkelda joined #salt
09:13 TheThing joined #salt
09:18 jhauser joined #salt
09:28 MrTango joined #salt
09:38 ggoZ joined #salt
09:43 martoss What is the suggested way to self reference pillar data? I have a service with bind ip configured via  pillar: "service.ip". Each machine has a public and a private ip and I wanna use the public_ip there. What is the recommended way to make public and private ip accessible during pillar rendering?
09:45 babilen martoss: Do you have *any* invariants for the "public" and "private" ip?
09:46 babilen Are they, for example, all within a certain subnet (e.g. 10.0.0.0/8) or is the private one always defined on eth0 whereas the public one on eth1?
09:46 martoss Yes
09:46 babilen martoss: You will, however, want to take a look at the salt mine to maintain such information: http://docs.saltstack.com/en/latest/topics/mine/
09:47 martoss well, almost - there's one exception which I would handle manually.
09:48 martoss Ah ok, thx for the hint. Looks exactly what I was looking for.
09:48 babilen martoss: Okay, then exploit that invariant and define suitable mine functions. You can probably also call those functions directly in the pillar and ip_addrs is in salt.utils.network.ip_addr
09:49 babilen Your problem will be that, as of now in 2014.1, you can apparently only call a function *once* as the "Mine Function Aliases" are a new feature in 2014.7 which, despite its name, has not yet been released AFAIK
09:49 martoss you mean without mining.
09:50 babilen (hence my reference to salt.utils.network.ip_addr() which you can import and call in pillars written in Python)
09:51 babilen Or at least I would hope that you can .. not sure actually, Not sure at all.
09:51 martoss ah ok
09:51 martoss well, I'll try out and come back. The salt mine hint was great. Thx
09:52 babilen I currently mine the private ip addresses, but having used network.ip_addrs in the mine once I cannot use it again (or at least that was my understanding and I don't see a way to do so without aliases)
09:52 babilen The mine is great, but unfortunatly useless for data that is being mined by the same function with different arguments
09:53 babilen (in 2014.1 that is)
09:53 sectionme joined #salt
09:54 bhosmer joined #salt
09:56 workingcats joined #salt
10:01 CatPlusPlus joined #salt
10:05 TomTom joined #salt
10:09 Guest31653 Hi, I've started to use salt and I'm currently experiencing strange behaviour. If I do "sudo salt '*' test.ping the command returns with true two times but if I execute this command the third time it responds with "Failed to authenticate, is this user permitted to execute commands?". Any ideas what I could have done wrong?
10:10 sectionme joined #salt
10:14 peters-tx joined #salt
10:23 pastacino joined #salt
10:23 felskrone joined #salt
10:24 funzo joined #salt
10:25 intellix joined #salt
10:27 mechanicalduck_ joined #salt
10:29 bhosmer joined #salt
10:30 davromaniak is it possible to use regex in if conditions in a salt state ?
10:31 Damoun joined #salt
10:42 nnion_ joined #salt
10:50 babilen davromaniak: Could you elaborate on that?
10:53 Guest31653 Sorry guys, already found the reason for my problem. I was running the ubuntu service (minion, master) and hand started instances. After stopping the services everything is working fine again.
10:58 srage joined #salt
11:00 JPaul joined #salt
11:28 CeBe joined #salt
11:31 capitalfellow joined #salt
11:42 giannello joined #salt
11:43 bhosmer joined #salt
11:53 jhauser joined #salt
11:55 matthiaswahl joined #salt
11:56 mpanetta joined #salt
12:01 vbabiy joined #salt
12:06 SheetiS joined #salt
12:07 mpanetta joined #salt
12:07 davromaniak babilen: actually, I just found that I can use ".startswith" like '{% if grains['fqdn'].startswith('foo') %}'
12:07 davromaniak so my issue is now resolved
12:09 babilen davromaniak: You probably want to read: http://jinja.pocoo.org/docs/templates/ -- fwiw, I more or less gave up on jinja for anything that is a bit more involved and write most pillars in Python these days
12:10 babilen Also note that .startswith('foo-web1') would also match foo-web10 (which might or might not be what you are looking for)
12:11 davromaniak it's more like ".startswith('foo-web')", because in my arch, I have servers called "bnode*" and "vnode
12:11 davromaniak it's more like ".startswith('foo-web')", because in my arch, I have servers called "bnode*" and "vnode*", they are quite similar
12:11 davromaniak just the "bnode*" doesn't have some points of configuration
12:14 mechanicalduck joined #salt
12:15 intellix joined #salt
12:15 dhwty joined #salt
12:17 babilen Sure, just wanted to explicitly mention that (it caught out a colleague of mine)
12:21 davromaniak no problem
12:26 n8n joined #salt
12:27 j-saturne joined #salt
12:38 rawkode joined #salt
12:40 hobakill joined #salt
12:42 che-arne joined #salt
12:45 cpowell joined #salt
12:48 SheetiS joined #salt
12:53 bhosmer joined #salt
13:00 christian joined #salt
13:01 Guest51780 who
13:02 rawkode Indeed
13:02 foo__ joined #salt
13:02 racooper joined #salt
13:03 Guest93143 left #salt
13:08 toastedpenguin joined #salt
13:12 martoss joined #salt
13:14 ericof joined #salt
13:16 beardo joined #salt
13:18 jslatts joined #salt
13:21 che-arne|2 joined #salt
13:22 MrTango joined #salt
13:23 oz_akan joined #salt
13:25 Deevolution joined #salt
13:25 colttt joined #salt
13:27 acabrera joined #salt
13:29 rojem joined #salt
13:33 veselin joined #salt
13:33 jhauser joined #salt
13:35 cpowell joined #salt
13:43 ramishra joined #salt
13:45 miqui joined #salt
13:46 darrend joined #salt
13:47 ramishra joined #salt
13:47 intr1nsic joined #salt
13:49 dude051 joined #salt
13:50 jalbretsen joined #salt
13:50 badon joined #salt
13:53 mapu joined #salt
13:53 pkimber joined #salt
13:53 pkimber joined #salt
13:55 tru_tru joined #salt
13:56 che-arne joined #salt
13:58 quickdry21 joined #salt
13:59 smcquay joined #salt
14:01 orion__ joined #salt
14:02 CatPlusPlus_ joined #salt
14:02 knot` joined #salt
14:02 jut_ joined #salt
14:02 srage_ joined #salt
14:03 mpanetta_ joined #salt
14:03 msil hey, i'm having an issue building the php5-extensions package on poudriere
14:03 LordOfLA|Weechat joined #salt
14:03 dude^2 joined #salt
14:03 kaptk2 joined #salt
14:03 msil all the packages build fine, but i get the error: php5-extensions has a missing dependency: php5-odbc
14:04 msil even though the php5-odbc package has been built and i can see the .txz file with the other packages
14:04 msil is it not being indexted or something?
14:04 miqui_ joined #salt
14:05 clone1018_ joined #salt
14:05 stolitablrrr_ joined #salt
14:05 * Guest51780
14:05 mpoole_ joined #salt
14:06 twobitsprite joined #salt
14:06 lietu_ joined #salt
14:06 * Guest51780
14:06 babilen_ joined #salt
14:07 rlarkin joined #salt
14:07 jason` joined #salt
14:07 Ymage_ joined #salt
14:07 jeddi joined #salt
14:08 sectionme joined #salt
14:09 thayne joined #salt
14:10 SaveTheR0btz joined #salt
14:11 ipmb joined #salt
14:11 ajprog_laptop joined #salt
14:11 TheThing joined #salt
14:11 goodwill_ joined #salt
14:11 goodwill_ joined #salt
14:11 JPau1 joined #salt
14:12 teebes joined #salt
14:13 elfixit joined #salt
14:14 `ph8 joined #salt
14:14 * csa
14:14 gmeno joined #salt
14:15 rawkode joined #salt
14:15 tmh1999_ joined #salt
14:17 darkelda joined #salt
14:17 CSa joined #salt
14:17 lionel joined #salt
14:17 * csa
14:17 harkx joined #salt
14:17 * csa
14:18 smcquay_ joined #salt
14:18 racooper joined #salt
14:19 rap424 joined #salt
14:19 akoumjian joined #salt
14:20 tk75 joined #salt
14:22 vejdmn joined #salt
14:24 csa eval exec - expr `date +%s` - \\$F | awk '{print \"Irssi uptime: \"int(\\\\\\$1/3600/24)\"d \"int(\\\\\\$1/3600%24)\"h \"int(\\\\\\$1/60%60)\"m \"int(\\\\\\$1%60)\"s\" }'
14:24 csa oops
14:25 ecdhe joined #salt
14:25 babilen_ ETOOMANYESCAPES
14:25 kt76 joined #salt
14:26 babilen csa: Please also try to test your irssi scripts in a different channel.
14:28 catpig joined #salt
14:32 ramishra joined #salt
14:34 ramishra joined #salt
14:34 * csa
14:36 mechanicalduck_ joined #salt
14:36 jchen is this 2001? who cares about your uptime
14:37 scoates joined #salt
14:37 ramishra joined #salt
14:37 csa @babilen ok. got wrong window :-/
14:38 ramishra joined #salt
14:39 mechanicalduck joined #salt
14:39 danielbachhuber joined #salt
14:39 rockey i really love the \\\\\\\ irssi produces in that alias :)
14:40 rockey old school stuff
14:41 conan_the_destro joined #salt
14:45 ericof joined #salt
14:46 wendall911 joined #salt
14:46 pass_by_value joined #salt
14:49 scbunn joined #salt
14:50 icebourg joined #salt
14:50 vejdmn joined #salt
14:53 rallytime joined #salt
14:54 rojem joined #salt
14:55 djstorm joined #salt
14:57 dccc joined #salt
14:57 kballou joined #salt
14:58 nyx joined #salt
14:59 aparsons joined #salt
15:00 sijis joined #salt
15:00 aparsons_ joined #salt
15:00 Guest47361 I aim to install certificates in a couple different cert. stores on a Windows machine.  could anyone point me to a module or state to do this?
15:01 Guest47361 in a pinch I would be OK with invoking the power shell cmdlet 'Import-Certificate' via salt cmd.run
15:01 aquinas joined #salt
15:03 Gareth morning
15:12 jhauser joined #salt
15:16 dccc joined #salt
15:18 econnell joined #salt
15:21 linjan joined #salt
15:22 rojem joined #salt
15:23 badon joined #salt
15:24 vejdmn joined #salt
15:26 BrendanGilmore joined #salt
15:26 to_json joined #salt
15:28 sijis is there a salt syndic rpm?
15:29 screamingbanshee joined #salt
15:31 sijis ahh. its part of the salt-master package ..
15:31 conan_the_destro joined #salt
15:39 n8n joined #salt
15:47 pass_by_value joined #salt
15:50 BogdanR joined #salt
15:52 TheRealBill joined #salt
15:55 bhosmer_ joined #salt
15:56 MatthewsFace joined #salt
15:58 kelseelynn joined #salt
16:04 vmdsch1 Good day folks!  Anyone know the latest status/progress on RAET (to augment/replace 0mq at very large scale)?
16:04 MrTango joined #salt
16:05 hisPeople joined #salt
16:05 masterkorp Hello
16:05 manfred vmdsch1: it works pretty well in my testing
16:06 manfred vmdsch1: it is still considered beta though, and will for the 2014.7 release
16:06 masterkorp can i print a jinja dictonary to the salt output to check some stuff
16:06 manfred vmdsch1: but you can certainly use it in the 2014.7 release
16:07 manfred masterkorp: in develop, and in 2014.7 there is a function called show_full_context https://github.com/saltstack/salt/issues/14136
16:07 manfred masterkorp: you can use that to dump all the contents available through jinja to a file so that you can search through it
16:08 vejdmn joined #salt
16:09 masterkorp not using that version of salt :(
16:09 manfred then no
16:11 mikjaer joined #salt
16:12 mikjaer I can't find a module to manage apache modules, is there such a thing?
16:13 manfred mikjaer: i believe a2enmod and a2dismod are available, but only in 2014.7 and develop
16:13 masterkorp ok i found the problem
16:13 tmh1999_ joined #salt
16:13 masterkorp what another parameters can the grains.filter_by is able to do ?
16:13 masterkorp i want os_famility the and the osarch
16:14 manfred mikjaer: yup https://github.com/saltstack/salt/blob/develop/salt/states/apache_module.py#L30
16:14 manfred only in 2014.7 and develop
16:14 mikjaer manfred: Im using the debian packages, so no cigar for me?
16:14 manfred no apache_module state in 2014.1 https://github.com/saltstack/salt/blob/2014.1/salt/states/apache_module.py
16:14 manfred mikjaer: not until 2014.7 is done going through the release candidate phase
16:14 manfred which we just tagged as rc1
16:15 martoss joined #salt
16:16 mikjaer Are there any reason why i can't import the module as a custom module in 2014.1.7?
16:16 jessegonzalez joined #salt
16:17 manfred mikjaer: you can try that, just drop it in /srv/salt/_states, you will probably also need to grab the apache.py module and drop it in /srv/salt/_modules because of stuff like this http://www.reddit.com/r/linux/comments/2dw6j4/systemd_drama/
16:17 manfred bah
16:17 manfred not that
16:17 manfred https://github.com/saltstack/salt/blob/develop/salt/states/apache_module.py#L39
16:17 manfred that
16:17 manfred where they have added stuff to the apache module
16:17 bhosmer_ joined #salt
16:17 vejdmn joined #salt
16:17 manfred it might work, but there is no gaurantee
16:17 mikjaer great, thanks :-)
16:18 kelseelynn left #salt
16:19 KFDM joined #salt
16:21 jhulten joined #salt
16:23 jessegonzalez left #salt
16:23 vmdsch1 Thanks
16:24 kelseelynn joined #salt
16:30 penguin_dan joined #salt
16:30 KFDM left #salt
16:37 schimmy joined #salt
16:37 j-saturne joined #salt
16:37 pass_by_value left #salt
16:38 pass_by_value joined #salt
16:40 alanpearce joined #salt
16:40 kelseelynn joined #salt
16:40 che-arne|2 joined #salt
16:41 aparsons joined #salt
16:42 r3vxx joined #salt
16:42 bray joined #salt
16:45 matthiaswahl joined #salt
16:47 forrest joined #salt
16:47 hisPeople left #salt
16:47 rap424 joined #salt
16:48 hisPeople joined #salt
16:50 scbunn joined #salt
16:51 KyleG joined #salt
16:51 KyleG joined #salt
16:57 thayne joined #salt
16:57 r3vxx hi all, i'm testing a simple setup with one master and one minion and i'm hitting an issue with timeouts (the minion hits two successive SaltReqTimeoutError's after loading the minion key and then continues on normally) i'm using salt version 2014.1.10 (i've also tried 2014.7rc0) for both the master and minion, i've double checked that ports 4505 and 4506 are open and that the minion key was accepted, is there something obvious that mig
16:58 r3vxx (i see the timeouts when attempting to run salt-call state.highstate on the minion)
17:02 forrest r3vxx, hmm, well the first thing I'd probably do is start by rerunning the highstate with -l debug -t 30
17:02 forrest to set up increased logging, and a higher timeout value. What actions does your highstate take?
17:02 aw110f joined #salt
17:03 jergerber joined #salt
17:04 TheThing joined #salt
17:10 r3vxx i was running with -l trace already, to see if i could get anything useful out of the log, but nothing was obvious to me, -t does not seem to be a valid argument for salt-call on the minion, only the salt command from the master, no? my high state is just doing a file copy, a file.recurse copy, and creating a symlink
17:11 RichardM_ joined #salt
17:11 kermit joined #salt
17:11 j-saturne joined #salt
17:12 RichardM_ Hello... I am having a devil of a time transferring a 3 MB file from master to minion... I would deeply appreciate someone's help.
17:12 possibilities joined #salt
17:14 davet joined #salt
17:14 forrest r3vxx, Hmm, so are you able to highstate from the master to the minion? Does test.ping even work from the master to the minion (not using salt-call on the minion)
17:14 forrest RichardM_, What issue are you encountering?
17:14 dave_den joined #salt
17:15 RichardM_ forrest, I started off trying to use states.  now I've backed off to a more straight-forward approach ...
17:16 martoss joined #salt
17:16 forrest RichardM_, what happened when using the state? And what state did you use?
17:16 RichardM_ <forrest>, salt '*' --timeout=480 salt://file /tmp/file
17:17 forrest Well, you could use salt-cp: http://docs.saltstack.com/en/latest/ref/cli/salt-cp.html
17:17 RichardM_ forrest, it's like the minion can't see / the file_root
17:17 forrest RichardM_, why would you set the timeout that high for a 3 meg file??
17:17 forrest And again, why did the state not work? Even if salt returns to the command line, the job would still be running
17:17 RichardM_ forrest, was trying anything...
17:18 jmccree Is there any easy way to run a certain state only on instances where that state would be applied?
17:18 jmccree without running the full top state
17:18 forrest jmccree, without using top file logic? You could either bake the logic into the states, or create node groups: http://docs.saltstack.com/en/latest/topics/targeting/nodegroups.html
17:19 jmccree forrest, using the top file logic is fine, I just don't want to run any state but one specific sls file
17:19 RichardM_ forrest I can't recall now why states wouldn't work... it was about a month ago... I just picked it up again and thought I'd try a more basic cp.get_file
17:20 KyleG1 joined #salt
17:20 forrest jmccree, ahh, hmm yeah I'd probably make a nodegroup, or you can target with state.sls using matchers.
17:20 RichardM_ forrest, sorry, left the command out of the example call.   salt '*' cp.get_file salt://file /tmp/file
17:20 aparsons joined #salt
17:20 forrest RichardM_, gotcha, I'd say try salt-cp, but preferably use states since then you can apply it more easily
17:21 aparsons joined #salt
17:21 RichardM_ forrest, here is the log from the minion...     2014-08-18 12:42:20,885 [salt.fileclient                             ][WARNING ] Specified file foo is not present to generate hash
17:22 vejdmn joined #salt
17:22 RichardM_ forrest, I've tried salt.cp too... it transfers smaller files, but just exits quietly on the 3 MB file
17:23 r3vxx forrest, test.ping works fine from master, state.highstate works from the master as well but it's pretty slow due to the two SaltReqTimeoutError's it hits everytime (this adds at least 120 seconds to the call) -- these timeouts happen whether it's triggered from the master or from salt-call on the minion
17:24 unprivileged joined #salt
17:24 forrest RichardM_, well, if salt times out it should still be finishing the job. It just drops you back to the command line after a while
17:25 forrest r3vxx, weird, I'm really not sure then. I'd suggest to get as much out of the logs as possible and post on the mailing list. Maybe someone will have a better idea
17:25 RichardM_ forrest, thanks for your time.  I appreciate it!
17:26 forrest RichardM_, yea np
17:30 ml_1 joined #salt
17:32 scbunn_ joined #salt
17:34 KyleG joined #salt
17:34 KyleG joined #salt
17:37 TheThing joined #salt
17:40 dalexander joined #salt
17:42 martoss joined #salt
17:42 hophophop joined #salt
17:42 unprivileged Hi folks. I'm trying to run the minion as an unprivileged user. I have changed the "id" parameter in the /etc/salt/minion file and restarted /etc/init.d/salt-minion, but the process still starts as root.
17:44 forrest unprivileged, did you already do the stuff ehre: http://docs.saltstack.com/en/latest/ref/configuration/nonroot.html ?
17:44 unprivileged Yes, except changing /var/run/salt. I'll try that
17:45 hophophop left #salt
17:45 dalexander joined #salt
17:47 cron0 joined #salt
17:47 unprivileged I changed the ownership of all the directories and restarted the minion process, but it still starts up and runs as root ...
17:48 r3vxx unprivileged, change the user flag in the config, not the id, id is the minion's name
17:48 v0rtex Hey guys, I'm getting a failure saying "Failed to confirm config of repo 'deb http://downloads-distro.mongodb.org/repo/ubuntu-upstart dist 10gen': repo 'deb http://downloads-distro.mongodb.org/repo/ubuntu-upstart dist 10gen' was not found" with this pillar/state data: http://hastebin.com/ajafuteboj.sm ...am I missing something?
17:49 possibilities joined #salt
17:49 MoarSalt joined #salt
17:49 manfred unprivileged: the id variable is just waht the server id is, you want the user: variable http://docs.saltstack.com/en/latest/ref/configuration/minion.html#user
17:50 SheetiS joined #salt
17:50 unprivileged sorry. i said that i changed the "id" parameter, but i have actually set the "user" parameter, so that is correct as well
17:51 unprivileged the only parameters i have uncommented and changed is the id parameter and the user parameter. everything else is untouched.
17:52 jergerber joined #salt
17:53 aw110f Hi, if I want to use file.recurse to distribute a bunch of subdirectories and files, and I the file modes are not uniform (some files are 755 some other are 644).  how would i do that?
17:53 forrest aw110f, you currently cannot
17:54 bmatt forrest: afaik file modes in git will be persisted
17:54 bmatt but that might be a side-effect
17:54 forrest bmatt, oh really? I thought there was an open issue about permissions not being respected.. hmm
17:54 bmatt forrest: not sure
17:55 aw110f so if I don't specify file_mode or dir_mode in file.recurse, it should respect the files that are checked in git?
17:55 bmatt aw110f: it's an open question to me but it's worth testing
17:56 bmatt is there a convenient way to make the salt command return structured data?
17:56 aparsons joined #salt
17:57 ericof joined #salt
17:59 aw110f bmatt: no it doesnt respect the file modes of the source files
18:00 ramteid joined #salt
18:00 aw110f I'm on 2014.1.3
18:02 Ryan_Lane joined #salt
18:02 jmccree A followup. Are the whitelist/exclude options of state.highstate documented anywhere. I'm having a hard time finding what they do even in the source code.
18:03 ckao joined #salt
18:04 manfred jmccree: i don't see anything specific on it
18:04 APLU joined #salt
18:04 jmccree What I want to accomplish doesn't seem like it should be this hard
18:04 manfred one second
18:04 jmccree state.highstate onlyrun='users.user1'
18:05 panchisco joined #salt
18:05 jmccree is basically what I want to do. Any highstate that involves users.users1 I want to run that only.
18:05 manfred jmccree: you just want to run that one state?
18:05 manfred ahhhh
18:05 jmccree manfred, yep.
18:05 rawkode joined #salt
18:05 manfred you only want to run a highstate on the server if it includes that state? or you only want to run that one state?
18:05 jmccree but I have no way of targeting the minions who need that state other than highstate/top
18:05 manfred yeah
18:05 manfred ok, gimme a second
18:05 jmccree manfred, only run that one state if highstate/top matches
18:05 jmccree thanks!
18:05 vejdmn joined #salt
18:06 manfred jmccree: so, the fastest way would be to take your match from top.sls and use it in salt ... salt <match> state.sls user.user1
18:06 manfred but short of doing that, i don't know of a way off the top of my head to apply the top.sls matching directly
18:06 manfred one second
18:06 chrisjones joined #salt
18:06 bmatt oh hey nice. salt 'foo' state.func --out json
18:07 jmccree manfred, Ok, It's likely going to be 20+ servers or so that return.
18:07 forrest manfred, I think nodegroups are the best option in this situation
18:07 manfred jmccree: yeah, if you had that state seperated accross several matches in your top.sls it wouldn't work
18:08 manfred jmccree: forrest might be right that the best option right now is node groups, i am still looking to see if we have a way to do top.sls matching and only run highstates if a certain state is going to be run for that minion...
18:08 jmccree This may be something not possible and something I should be using nodegroups/pillars instead of manually defining in states.
18:08 manfred i don't think we have one right now
18:08 forrest manfred, yeah I was trying to think of something as well, and couldn't
18:10 rawkode I've sync'd a custom grain from the salt-master and I can see the grain inside /var/cache/....
18:10 rawkode but it's not available with grains.ls - any advice?
18:10 jmccree manfred, forrest : what's the preferred way to open a request for that to be added. I'll take a look myself in my spare time to see if I can add a patch.
18:10 forrest jmccree, an issue on github
18:11 manfred ^^
18:11 manfred just put a feature requet
18:11 manfred request
18:11 forrest jmccree, yep
18:11 jmccree k, will do. Thanks everyone!
18:11 manfred feel free to tag @gtmanfred
18:11 manfred cause I think that is an interesting idea
18:11 forrest rawkode, I assume you already did: http://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.saltutil.html#salt.modules.saltutil.sync_grains ?
18:11 forrest if it's on the minion
18:11 rawkode Yep
18:11 manfred jmccree: http://docs.saltstack.com/en/latest/topics/development/contributing.html
18:12 rawkode It told me the grain sync'd and I can see it within /var/cache/salt/minion/extmods/grains
18:12 rawkode ^^ on the minion
18:13 forrest rawkode, and you already restarted the minion?
18:13 forrest just to be sure
18:13 rawkode Yeah
18:13 aw110f joined #salt
18:13 forrest can you try grains.get and call out the grain explicitly?
18:13 jmccree manfred, Yeah, our use case we have a ton of legacy machines we've been attempting to "saltify", but we can't necessarily run highstate randomly on them. One of our devs got a new machine and ssh key, so trying to figure the best way to deploy that out.
18:13 APLU joined #salt
18:13 unprivileged I got the salt-minion process to run as an unprivileged user by starting it up with sudo -u <the unprivileged user> /usr/bin/salt-minion. I had no success with getting the startup scripts to respect the user config parameter so far
18:13 forrest jmccree, you could look at reclass
18:14 forrest jmccree, that lets you be more specific.
18:14 forrest but it's quite a bit of work
18:14 rawkode [root@ip-172-16-224-114 _grains]# salt \* grains.get ec2_tags
18:14 rawkode i-3be1837b:
18:14 rawkode [root@ip-172-16-224-114 _grains]#
18:14 manfred unprivileged: what distro is this on?
18:14 unprivileged ubuntu 12.4
18:14 forrest rawkode, the grain does work right? Can you get the grain data from on the minion?
18:15 rawkode forrest: I've ran "python ec2_tags.py" on the minion and the python worked perfectly. How can I call the grain locally though?
18:15 forrest salt-call --local maybe?
18:15 forrest I don't usually check grains locally
18:15 cpowell joined #salt
18:16 manfred unprivileged: one second, looking through the code to see where it switches to the user
18:16 jmccree forrest, manfred : For the record, I think it would be better to have user files determined by node groups or pillars etc and avoid this whole issue, but I'll be submitting a feature request to only run one state in a high state. :) Thanks again.
18:17 rawzone joined #salt
18:17 v0rtex what would be a good approach to my state requisites for a package that depends on two other packages? In my case I am installing graylog2 and I want it to require that both my mongodb and elasticsearch packages are installed first (I have some generic states set up for both of those)
18:17 manfred jmccree: np
18:17 chrisjones joined #salt
18:18 rawkode Thanks, forrest. That has put me in the right direction. Strange that the python works on it's own, but has an error when run through salt :|
18:18 linjan joined #salt
18:19 forrest rawkode, oh it threw an error back? Nice.
18:19 unprivileged manfred, the salt-master process runs as the user when i change the user parameter in the master config file. it is only the salt-minion that is the problem
18:19 rawkode It's one of the salt-contrib grains - I assume these aren't kept up to date by salt themselves?
18:19 v0rtex It seems I can only specify a single package when using 'require'
18:19 forrest rawkode, you've looked at some of the examples in salt-contrib right? Someone has an ec2_tags in there
18:20 forrest v0rtex, did you already put the pkg call on two different lines?
18:20 manfred unprivileged: yeah, i am still looking for how the minion switches
18:20 rawkode That's the one I'm attempting to use, forrest. It fails when using IAM roles. I'm a PHP dev, so trying to stumble through the python code just now
18:20 manfred v0rtex:
18:20 manfred require:
18:20 manfred - pkg: mongodb
18:20 manfred - pkg: elasticsearch
18:20 forrest rawkode, gotcha, if you can figure out PHP, you should be able to figure out python :P
18:20 forrest sorry manfred
18:21 manfred huh?
18:21 v0rtex manfred: I'll try that - I thought that didn't work in the past when I've done it
18:21 forrest I cut your example in the middle
18:21 manfred looks fine to me :P
18:21 nicksloan joined #salt
18:21 manfred ¯\(°_o)/¯
18:21 forrest manfred, oh good, your message must have beat mine to the server then
18:21 Nazca joined #salt
18:22 forrest v0rtex, you could always do a require chain, where package 2 requires package 1, then item a requires package 2
18:22 forrest but that's messy. You should just be able to do the two requires, and salt executes in order anyways, so it should be set if you have them ordered.
18:22 APLU joined #salt
18:23 rawkode I've fixed the bug. Is it worth sending a pull request for the contrib modules? or are they mostly dead?
18:23 forrest unless you want it to bomb out if the package isn't there.
18:23 forrest rawkode, no way man, send the PR!
18:23 rawkode Will do! Thanks again for the help, forrest.
18:24 v0rtex forrest: yeah I didn't want to do the require chain idea because it feels too dirty but it appears that using 2 'pkg' requires works so thanks
18:24 forrest rawkode, yeah of course, glad you got it figured out!
18:24 forrest v0rtex, cool
18:24 forrest v0rtex, mostly manfred with his disgusting in-IRC example :P
18:24 manfred heh
18:26 nicksloan some of my minions have started getting "salt.exceptions.AuthenticationError: message authentication failed" for no reason
18:26 v0rtex heh, yeah...thanks for the suggestion manfred. I don't recall why I had trouble doing that in the past but I know I used to have issues - perhaps a bug with an older version of sal
18:26 v0rtex *salt
18:29 manfred unprivileged: can enable debug loging on the salt minion, and check if there are any errors?  It all looks to be correct in the code using os.setuid and os.setgid, but there are several errors that could happen apparently
18:30 manfred unprivileged: it does the switch here https://github.com/saltstack/salt/blob/develop/salt/utils/verify.py#L282
18:30 forrest manfred, what's with working on the IRC this weekend? I was number 1 on Friday! :D
18:31 manfred was I in here on the weekend?
18:31 manfred oh
18:31 manfred i remember
18:31 unprivileged manfred, i'll do that
18:31 forrest nicksloan, https://github.com/saltstack/salt/issues/14645
18:31 forrest nicksloan, not identical
18:31 forrest nicksloan, but I think there was another issue opened for what you are encountering
18:31 forrest I thought they were related, maybe I am wrong
18:31 Ryan_Lane joined #salt
18:32 manfred unprivileged: and it runs it over here https://github.com/saltstack/salt/blob/develop/salt/__init__.py#L250
18:32 forrest manfred, do you remember that issue for the auth errors?
18:32 manfred i do, but i don't remember how to fix it
18:32 manfred which salt version?
18:33 davet joined #salt
18:33 forrest manfred, there wasn't a fix for it last time I remember. They were still trying to dupe it in house.
18:33 manfred https://github.com/saltstack/salt/issues/14299 this one?
18:33 forrest manfred, no not that one... There's a MUCH bigger one
18:33 forrest that had a ton of replies
18:33 manfred https://github.com/saltstack/salt/issues/12248
18:33 forrest manfred, Yeah that's the one, ok different error, good
18:34 bmatt this seems like an odd choice: https://github.com/saltstack/salt/blob/2014.1/salt/grains/core.py#L1411
18:34 bmatt modulo 2^32?
18:34 rawkode Pretty happy. I have a salt-master running on a central VPC with all my application stacks running over CloudFormation in their own VPC and peering in to the central.
18:34 rawkode Definitely getting myself a beer :D
18:35 forrest rawkode, haha, wait till Ryan_Lane publishes the article he's writing about doing all sorts of loadbalancing and other work in ec2
18:35 davet joined #salt
18:35 unprivileged manfred, i just tried this: i deleted the /var/run/salt/minion directory and startet the salt-minion process again. The process actually created the /var/run/salt/minion directory again as the unprivileged user. What puzzles me is that the salt-minion process is still running as root
18:35 Ryan_Lane I'm about 1/4 finished. it's going to take a while ;)
18:35 manfred bmatt: isn't that just an unsigned long?
18:35 rawkode Interesting - I'll keep an eye out for that. Is there a blog or twitter to follow?
18:36 Ryan_Lane rawkode: ryandlane.com
18:36 rawkode Cheers, Ryan_Lane
18:36 unprivileged manfred, while the salt-master process runs as the unrpvileged user
18:36 bmatt manfred: well, yeah, but I can't grok the logic of it
18:36 forrest Ryan_Lane, it's like you care about work life balance or something, jeez...
18:36 bmatt why not take the hash digest, or take its raw value or something?
18:36 manfred bmatt: /shrug
18:36 Ryan_Lane forrest: it just takes me ages to write them. it's a lot of effort :(
18:36 forrest Ryan_Lane, I know
18:37 bmatt I'm also not sure where that id opt comes from (which is what I'm actually trying to figure out - how server_id is generated)
18:37 forrest Ryan_Lane, articles/blog posts in general take a long time to write. Why do you think I haven't written one in a few months :P
18:37 Ryan_Lane heh
18:37 manfred bmatt: i would wonder if it is a limitation on 32 bit systems for certain things that might run in C?  no idea
18:37 pass_by_value left #salt
18:37 manfred unprivileged: it should switch to the unprivileged user on this line https://github.com/saltstack/salt/blob/develop/salt/__init__.py#L250 when it finds out that it isn't running as the user: you specified, it should run os.setuid("user")
18:37 forrest Ryan_Lane, 'hmm, should I write a blog post, or go do play my bass/sleep/eat/read/everything else...' :P
18:37 Ryan_Lane yep
18:37 bmatt manfred: I might think a bit more deeply about it and come up with a proposal. limiting the possible hash output space like that is just a very odd choice
18:38 pass_by_value joined #salt
18:39 manfred forrest: i really need to finish my blog post about all the stuff I have done with salt and salt-cloud since february
18:39 manfred i haven't even started
18:39 forrest manfred, Yeah that would be really cool and helpful to people
18:39 manfred cause there is a lot of it
18:39 manfred and then I have another one about how I want to integrate pyrax.py into the rackspace cloud driver
18:40 to_json joined #salt
18:41 alanpearce joined #salt
18:41 chrisjones joined #salt
18:42 ericof joined #salt
18:43 snuffeluffegus joined #salt
18:44 mechanicalduck joined #salt
18:45 scbunn joined #salt
18:46 neilofc joined #salt
18:47 bmatt okay, where the heck is __opts__ populated?
18:47 VictorLin joined #salt
18:47 forrest in the spaghetti
18:47 bmatt =[
18:47 bmatt where does the 'id' value come from? *shakes fist at sky*
18:48 forrest bmatt, for the system?
18:48 forrest I know it's using getfqdn()
18:48 bmatt yeah; I'm trying to hunt down how the system_id grain is calculated
18:48 manfred bmatt: it is injected in the loader
18:48 bmatt D=
18:49 manfred bmatt: https://github.com/saltstack/salt/blob/develop/salt/loader.py#L158
18:49 manfred somewhere around there like that
18:50 rawkode I've noticed that running salt commands sometimes does nothing. Running again works. Is this normal?
18:51 smcquay_ joined #salt
18:51 forrest rawkode, increase the timeout with -t
18:52 forrest rawkode, sometimes if a long running command takes a minute, it drops you back to the terminal, by default it's only about 5 seconds
18:52 rawkode Noted. Cheers
18:52 forrest so basically salt asks for a status update from the minion every 5 seconds, then there's an extra 2 seconds where the minion is looking to see what the status is, if it doesn't return, then it says 'nah I have nothing', and salt will drop
18:52 forrest just so you don't get locked in for long running jobs
18:53 possibilities joined #salt
18:53 pass_by_value joined #salt
18:53 aw110f joined #salt
18:54 kermit joined #salt
18:54 rawkode So I can always assume that regardless of terminal output, the job is working away?
18:55 forrest rawkode, I'd still check the job queue for active jobs.
18:55 forrest rawkode, better safe than sorry
18:56 rawkode forrest: yeah :)
18:57 unprivileged manfred, i found that the salt-minion process also creates the /var/run/salt-minion.pid file as the unprivileged user. however, the minion process cannot connect to the master for some reason. maybe it simply fails before it switches to the unprivileged user. i'll solve that problem first. thx
18:57 pssblts joined #salt
19:00 intellix joined #salt
19:01 rawkode One last question ... I've noticed that when using a jinja template to loop pillar variables, the list comes back in a random order and with extra blank lines. Is that just as expected?
19:01 manfred rawkode: yes
19:01 manfred rawkode: well, the random order not so much, but the extra lines are
19:01 rawkode Dare I ask why?
19:01 linjan joined #salt
19:01 manfred rawkode: because if you have a line in jinja
19:02 kuffs rawkode: http://jinja.pocoo.org/docs/templates/#whitespace-control
19:02 manfred {% for something in __grains__['networks'] %} that is an empty line in there
19:02 manfred be sure to add a - with the % to clear out that whitespace
19:02 kuffs also the sequence is randomized because you're iterating over a dictionary, which is unordered
19:02 kuffs you could make a list of key, value pairs if you prefer the data to be ordered
19:02 rawkode It's a completely different order from the order inside the pillar though?
19:03 manfred this is fixed in newer version of salt once we started using OrderedDictionaries for pillars
19:03 manfred not sure if this change was in 2014.1
19:03 rawkode I'm running 2014.7
19:03 manfred it should be an ordered dict for pillars in that
19:04 rawkode I have in my pillar: a: 1\nb:2\nc:3 .. or such. then the parameters.yml on the minion is bca
19:04 rawkode Not a problem really, I was merely curious
19:04 smcquay joined #salt
19:06 manfred it should be ordered, that should be an ordered dictionary for pillar data
19:07 dvestal joined #salt
19:09 smcquay joined #salt
19:10 vejdmn joined #salt
19:11 pass_by_value left #salt
19:12 ollins joined #salt
19:13 mechanicalduck_ joined #salt
19:14 j-saturne joined #salt
19:15 berserk joined #salt
19:15 davet joined #salt
19:17 mago_ joined #salt
19:18 mago_ hello I need help using the s3 module for minions
19:18 manfred mago_: s3:// prefix stuff or something else?
19:18 mago_ I set the keys in pillars
19:18 mago_ but I get an error
19:19 mago_ salt 'node1' s3.get documents
19:19 manfred what is the error?
19:21 mago_ http://paste.debian.net/116232/
19:21 smcquay joined #salt
19:22 mago_ I checked the pillar of the minion
19:22 mago_ and I can see the s3 id, key and service_url
19:23 manfred mago_: on 2014.1.5?
19:26 schimmy joined #salt
19:26 manfred mago_: looks like there is an open issue for this https://github.com/saltstack/salt/issues/7150
19:28 smcquay_ joined #salt
19:33 mago_ ah
19:33 mago_ thank manfred
19:33 mago_ and yes that is my version
19:33 aparsons joined #salt
19:35 manfred 2014.1.10 should be available, try that one, this bug was cleaned up for other modules in 2014.1.7, might have fixed it in s3 too
19:35 manfred ¯\(°_o)/¯
19:38 to_json joined #salt
19:38 CeBe joined #salt
19:38 kelseelynn joined #salt
19:39 linjan joined #salt
19:42 alanpearce joined #salt
19:45 smcquay joined #salt
19:47 badon joined #salt
19:47 smcquay joined #salt
19:50 possibilities joined #salt
19:51 wackycaptain joined #salt
19:52 SheetiS ls
19:53 manfred drwxrwxr-x 5 SheetiS SheetiS 51 2006-04-10 09:12 pr0n/
19:53 SheetiS nice
19:54 martoss joined #salt
19:54 jslatts joined #salt
19:55 kt766 joined #salt
19:56 Ryan_Lane joined #salt
19:56 KyleG1 joined #salt
19:56 bmatt oh, damn, server_id is indeed just a hash of the fqdn
19:57 fivethreeo joined #salt
19:57 Hipikat joined #salt
19:57 arochett1 joined #salt
19:57 simonmcc__ joined #salt
19:57 lude1 joined #salt
19:57 estherbester_ joined #salt
19:57 ghartz_ joined #salt
19:57 lietu joined #salt
19:57 tru_tru_ joined #salt
19:57 sirtaj_ joined #salt
19:57 HuleB joined #salt
19:57 rglen__ joined #salt
19:58 UForgotten_ joined #salt
19:58 kballou joined #salt
19:58 MatthewsFace joined #salt
19:58 ronc joined #salt
19:59 chitown joined #salt
19:59 scoates joined #salt
19:59 renoirb joined #salt
19:59 jslatts joined #salt
19:59 mpoole joined #salt
20:00 robinsmidsrod joined #salt
20:00 dalexander joined #salt
20:01 badon joined #salt
20:03 g3cko joined #salt
20:03 ahammond joined #salt
20:05 mefrackindeiros joined #salt
20:06 schimmy joined #salt
20:08 estherbester_ joined #salt
20:08 schimmy1 joined #salt
20:12 vejdmn joined #salt
20:12 schimmy joined #salt
20:28 n8n joined #salt
20:30 mefrackindeiros joined #salt
20:32 rogst joined #salt
20:32 CeBe joined #salt
20:43 alanpearce joined #salt
20:47 dvestal joined #salt
20:50 v0rtex is there an easy way I can generate a sha256 sum of a value from pillar in a jinja templated file? My use case is a config file that needs a sha256 sum of a password (which I have stored in pillar)
20:51 v0rtex my searches turned up mostly empty
20:52 forrest v0rtex, hmm, the only thing I can think is to use mako as the templating language instead of Jinja, then just do some straight up python
20:52 bhosmer_ joined #salt
20:54 v0rtex maybe that's what I'll have to do - I only have one other jinja variable in this particular file so it shouldn't be a big deal
20:55 j-saturne joined #salt
20:55 jhauser joined #salt
20:57 tk75 joined #salt
20:59 kt76 joined #salt
21:01 gothix joined #salt
21:04 sectionme joined #salt
21:04 vejdmn joined #salt
21:05 davet joined #salt
21:06 wm-bot4 joined #salt
21:07 neilofc joined #salt
21:08 beneggett joined #salt
21:09 aparsons joined #salt
21:09 beneggett joined #salt
21:10 beneggett joined #salt
21:10 sivakalyanc joined #salt
21:11 aparsons joined #salt
21:13 dalexand_ joined #salt
21:14 bhosmer joined #salt
21:16 miles32 joined #salt
21:17 BrendanGilmore joined #salt
21:17 miles32 hey, file.recursed and file.managed seem to be leaving some cruft on my windows machines of the form -> <file_name>.<file-extension><hash>
21:17 miles32 is that a common/known thing?
21:18 n8n joined #salt
21:19 miles32 (for instance global.asax.cs becomes global.asax.cshjpz3f)
21:20 kermit joined #salt
21:22 jergerber joined #salt
21:23 gzcwnk :)
21:28 jacksontj joined #salt
21:29 orion_ joined #salt
21:29 miles32 I think it's what happens when a file is in use
21:29 miles32 so it can't be replaced
21:30 miles32 :(
21:30 drawks is syndic usage pretty common or is it more typical to have a fully flat architecture
21:30 drawks ?
21:31 forrest drawks, depends on the complexity of the architecture I'd imagine
21:33 drawks maybe I should restate my question a bit :) Is syndic usage common enough to be well supported and generally devoid of suprises? At what minion count does it make sense to split up the architecture to have this intermediary syndic layer? Is subordniating minions meant to optimize out round trip latency for distant installations? Is there some other bottle neck in the salt stack architecture that may be addressed by syndic delegation?
21:33 drawks the docs are a little slim on the "why" part of exomplaining syndics
21:34 forrest drawks, very true
21:34 forrest drawks, my only answer to your other questions is 'I do not know'
21:34 forrest I don't use the syndic extensivel
21:34 forrest *y
21:34 forrest I've messed with it a bit at home, but nothing else
21:35 forrest drawks, there's an open issue to improve them: https://github.com/saltstack/salt/issues/8383
21:36 drawks thanks :) maybe someone else will chime in eventually. IRC works well for the patient
21:36 forrest drawks, yeah for sure.
21:37 tyson_ joined #salt
21:37 miles32 is good for the trolls too
21:39 miles32 alright, the minion is running as local system, everything and there mother is turned off
21:39 miles32 this should work
21:40 DaveQB joined #salt
21:44 alanpearce joined #salt
21:47 robawt does anyone get a crazy infinite loop running ext_pillar in 2014.1?
21:47 robawt running salt-call on a masterless minion with debug flag
21:47 forrest 2014.1?
21:48 forrest which sub item
21:48 forrest 2014.1.?
21:48 forrest robawt, if you're just using 2014.1, please upgrade.
21:48 robawt forrest: 2014.1.x?  set git to pull down the latest
21:48 forrest there's a bunch of bugs in that release.
21:48 forrest robawt, oh lame then
21:48 forrest should have pulled 2014.1.10
21:48 robawt i didn't see the branch, should've looked harder at the tags
21:48 robawt ok i'll reconfig
21:50 robawt and this is why I run it in VMs
21:50 robawt ty forrest
21:50 robawt 2 beers now?
21:50 robawt :)
21:50 forrest robawt, ALL THE BEERS!
21:50 robawt all the [natty ice] beers!
21:50 forrest that picture where the guy is hold stuff and it's falling out of his arms is how many beers I've acquired at this point
21:51 forrest robawt, bleh, I don't even like good beer :P
21:51 forrest natty ice I'd just pour out
21:51 robawt ha
21:51 p3llin0r3 joined #salt
21:51 p3llin0r3 Hey guys
21:52 p3llin0r3 what's the difference between SaltStack free / Enterprise?
21:52 p3llin0r3 ( Also CI options for Saltstack free )
21:52 forrest p3llin0r3, I'm not sure what you mean by CI options for saltstack, but the biggest difference between free and enterprise is you get support with enterprise from saltstack.
21:53 forrest with free, obviously you just get some help in IRC, and the mailing list.
21:53 p3llin0r3 http://www.saltstack.com/enterprise/
21:53 jason` joined #salt
21:53 p3llin0r3 the bullet says:
21:53 Gareth p3llin0r3: code wise they are identical.
21:53 p3llin0r3 Application provisioning and continuous deployment
21:54 p3llin0r3 Is this a feature of free, and their Enterprise page just isn't clear about what I get with free saltstack?
21:54 miles32 pretty sure enterprise is just support
21:54 miles32 and everything else you get outta normal salt
21:54 Ryan_Lane yep. enterprise is just support
21:54 kt76 p3llin0r3, they also will fix bugs and implement features with enterprise
21:54 Ryan_Lane salt is 100% open source. their business is based on services and support model
21:54 Gareth p3llin0r3: what are you hoping to do with Salt Stack?
21:54 forrest p3llin0r3, yeah the fatures there you get with free salt. I imagine it's just there as it's a bit quicker to explain
21:55 forrest p3llin0r3, agreed, tell us about your goals
21:55 p3llin0r3 I want to write a saltstate that will deploy my django apps
21:55 p3llin0r3 based on their git tag
21:55 p3llin0r3 and have a Jenkins instance executing the saltstates when I push changes to their individual branches
21:56 p3llin0r3 ( or whatever is better than Jenkins, if you guys can recommend something )
21:56 forrest p3llin0r3, terminalmage wrote something to handle his django stuff: https://github.com/terminalmage/djangocon2013-sls
21:56 miles32 I got a guy named bob
21:56 forrest but it doesn't obviously handle the jenkins and basing on specific git tags, you could figure out how to do that though
21:56 forrest miles32, I'll take Jenkins over bob
21:57 p3llin0r3 Do you guys feel like Saltstack is mature enough for a small startup?
21:57 TheThing joined #salt
21:57 forrest p3llin0r3, absolutely. I work at a smaller shop, myself and one other guy run the whole devops group.
21:57 godber joined #salt
21:57 tcotav shell scripts are mature enough for a small startup and salt is leaps beyond that
21:57 forrest tcotav, heh
21:57 p3llin0r3 Yeah, that's me in my startup. I REALLY like their Docker stuff as well
21:58 miles32 I'm not sold on Docker
21:58 p3llin0r3 I feel like Fig was a mistake. Salt is basically the same thing with way more features
21:58 manfred p3llin0r3: have yall used shipyard yet?
21:58 p3llin0r3 I've yet to get it to work, and I lost interest
21:58 forrest p3llin0r3, Yeah, as with any 'newer' config management, some things aren't quite as mature as they could be, but most features are pretty well implemented, and adding to them is easy if you (or your devs) write python
21:58 p3llin0r3 Right now IMO, it's a pain
21:58 manfred https://github.com/shipyard/shipyard
21:58 manfred shipyard looks awesome
21:58 manfred shipyard and drone
21:58 manfred replace jenkins with drone, and run all jobs in docker containers
21:59 p3llin0r3 Ambassador containers are a friggin hack job, and nobody has a well documented service discovery implementation
21:59 kelseelynn left #salt
21:59 forrest manfred, I hadn't heard of drone, this looks cool
21:59 p3llin0r3 forrest, thanks, that's good advice.
22:00 jhauser joined #salt
22:00 forrest p3llin0r3, yeah of course. Have you started using salt at all yet
22:00 forrest *?
22:00 manfred forrest: my buddy is using it for continues integration for his dockershell bastoin, that overrides a users shell using nss override in ldap, and drops them into a docker container on login
22:00 forrest manfred, Nice. We still can't user docker here because of some missing features
22:00 kelseelynn joined #salt
22:00 forrest but we are using lxcs
22:00 kelseelynn left #salt
22:00 forrest so I want to head towards docker for our devs
22:00 manfred i want to use it for an automated build system, so I can just hook up a docker cluster to it, and not have to maintain a jenkins slaves, and the jenkins docker plugin
22:00 forrest they're using it, but I want to make it awesome
22:00 forrest manfred, right
22:01 p3llin0r3 forrest, yes, I've run a few small scripts, and I've done a bunch of reading. It's way way way cleaner than Puppet and other similar techs, and I'm in love with the fact that it's rendering with Jinja2 templates
22:01 p3llin0r3 forrest, since I'm a python nerd
22:01 forrest p3llin0r3, Yeah for sure, that's good. I was going to give you a sample project I had written that's just a simple blog, but you seem to be past that already
22:01 jhulten_ joined #salt
22:02 p3llin0r3 forrest, :) yeah, I've got a basic grasp. Their getting started docs are AMAZINGLY bad though, IMO, XD
22:03 pssblts joined #salt
22:03 p3llin0r3 forrest, They should just start you out explaining masters and minions and such, instead of the confusing scaffolding stuff. I ended up with a Moe, Curly, etc on my dev environment! XD
22:03 forrest p3llin0r3, Yeah we've been trying to figure out how to make them better
22:03 panchisco_ joined #salt
22:04 forrest p3llin0r3, right, I've been trying to go that way with some stuff I'm writing on my own
22:04 vbabiy_ joined #salt
22:04 xmj_ joined #salt
22:04 rglen joined #salt
22:04 miles32 last time I read the getting started, you did all sorts of minioning before you got to the master
22:04 miles32 it was silly
22:04 p3llin0r3 forrest, it's a common problem. Take a look at the Celery docs! :X Or heck, the Docker docks are terrible as well! They cut off the commands your'e supposed to run!
22:05 claytron joined #salt
22:05 srage__ joined #salt
22:05 forrest miles32, Yeah, even in the stuff I'm writing you set up a masterless minion first, but you then join it to your master, so that's not as big a deal
22:05 p3llin0r3 I was gonna swoop em'
22:05 forrest p3llin0r3, Yeah I agree, I was pissed at the docker documentation, so I felt Salt was at least a bit better than that :P
22:05 forrest There's only a few of us that seriously work on the documentation though, so that makes it rough
22:06 miles32 welcome to open source
22:06 forrest yep
22:06 Ryan_Lane well, people use salt differently. it should really have links to getting started for different use cases
22:06 p3llin0r3 Question: What's the best way to set up my dev environment? I don't like editing stuff in /srv/salt
22:06 Heartsbane_ joined #salt
22:06 Rojematic joined #salt
22:06 jeblair_ joined #salt
22:06 yidhra_ joined #salt
22:06 andabata- joined #salt
22:07 miles32 I have git automatically check things out to /srv/salt/
22:07 miles32 I never actually visit my salt-master
22:07 Ryan_Lane Getting started with masterless mode. Getting started with remote execution using master/minion. Getting started with state management using master/minion.
22:07 j-saturne1 joined #salt
22:07 forrest miles32, why don't you just use gitfs?
22:07 forrest Ryan_Lane, heh
22:07 miles32 the performance dropped with gitfs
22:07 aparsons joined #salt
22:07 forrest miles32, fair enough
22:08 miles32 it was a 10 minute difference
22:08 forrest what/?
22:08 forrest gitfs should be caching
22:08 forrest that's.. weird
22:08 p3llin0r3 I guess I could do rsync...
22:08 forrest p3llin0r3, :(
22:08 miles32 oh yeah, it was nuts man I was like holy hell
22:08 forrest p3llin0r3, just use gitfs to start
22:09 forrest p3llin0r3, miles32 is also doing windows
22:09 miles32 oh I'm just starting windows
22:09 forrest miles32, nope, once windows, always invalidated! :D
22:09 p3llin0r3 ahh, I'm on Ubuntu
22:09 p3llin0r3 Would sym links work?
22:09 forrest p3llin0r3, we use gitfs and have no issues with execution times
22:09 forrest p3llin0r3, I've never tried it
22:09 miles32 symlinks work
22:09 p3llin0r3 Ok, thanks. Super cool, thanks for the help! :)
22:10 \ask_ joined #salt
22:10 forrest p3llin0r3, For sure
22:10 miles32 originally I had the entire /etc/salt directory symlinked
22:10 SaveTheR0btz joined #salt
22:10 forrest miles32, you mean /srv/salt?
22:10 miles32 (and /srv/salt symlinked too)
22:10 forrest oh ok
22:10 forrest lol
22:10 fridder joined #salt
22:10 p3llin0r3 For the record: Salt and Docker are going to be great friends
22:11 toddnni joined #salt
22:11 rawzone joined #salt
22:11 forrest Yeah
22:11 miles32 salts neat, it's just rough around the edges, and not always how an admin would do it
22:11 miles32 it's developer focused
22:11 miles32 (in my humble opinion)
22:11 p2_ joined #salt
22:11 homelinen joined #salt
22:11 forrest miles32, you feel that way? I thought it was way more sysadmin focused than puppet/chef
22:11 p3llin0r3 I believe https://github.com/GoogleCloudPlatform/kubernetes uses saltstack
22:11 jab416171 joined #salt
22:11 p3llin0r3 hopefully they don't bail and switch to Fig
22:11 Blacklite joined #salt
22:11 tcotav forrest: agreed -- more admin centric than chef.  equiv to puppet (imo)
22:12 miles32 it's closer
22:12 scooby2 joined #salt
22:12 miles32 but that doesn't mean it's admin focused
22:12 Damoun joined #salt
22:12 forrest miles32, yeah, maybe I am just too jaded by puppet
22:12 tcotav it's tomorrow's admin focused
22:12 conan_the_destro joined #salt
22:12 miles32 yes, but I'm having to teach todays admin
22:12 miles32 ;)
22:12 Ryan_Lane yeah. I think it's a bit more dev focused, too
22:12 tcotav to quote caddyshack -- the world needs ditch diggers too :-p
22:13 Ryan_Lane (I like that, though)
22:13 tcotav I think the yaml makes it very accessible
22:13 forrest miles32, yeah a small python crash course can help
22:13 forrest but if your admins are still using bash and perl for all their scripts, bleh
22:14 tcotav they're two generations behind
22:14 forrest tcotav, yeah the yaml isn't perfect, but it's pretty clear
22:14 tcotav yep -- exactly
22:14 jslatts joined #salt
22:15 miles32 I'm not saying that it's not something that's needed, I'm saying that it's difficult to get someone up and running on is all
22:15 miles32 in a month, I've duplicated work that took 6 months to do by hand
22:15 forrest nice
22:16 tcotav are you working with linux or windows folk, miles32?
22:16 p3llin0r3 IMO, it's by far the cleanest syntax I've seen
22:16 miles32 windows, I'm the linux guy
22:16 tcotav yeah, my experience (previous life as chef consultant) is that the windows guys push back hardest
22:16 tcotav but then you whip out the ol' powershell and they continue to look at you funny
22:17 forrest tcotav, heh
22:17 godber Hi Salt people, glad to see you're still here and keep up the great work!
22:17 forrest I always feel bad for the windows guys when they get super excited they'll finally be able to use powershell
22:17 forrest hey godber
22:17 ajprog_laptop joined #salt
22:17 godber I had a question, but in preparing it to ask, I answered it myself I think
22:17 forrest godber, awesome
22:17 miles32 I like this question
22:18 forrest godber, your rubber duck is on the way :P
22:18 godber the docs helped :P
22:18 forrest godber, wooo
22:18 esogas_ joined #salt
22:18 godber I think, ... I should try and run it probably, otherwise I'll feel exceptionally stupid
22:18 forrest godber, can't hurt to try
22:19 miles32 alright, somethings getting screwed on permissions
22:19 miles32 if the destination directory is empty life is kosher
22:19 miles32 but running highstate the second time
22:19 miles32 says that it can no longer write to the directory
22:19 miles32 (permissions)
22:20 forrest miles32, hmm, do the perms change between runs?
22:20 miles32 not from what I can see in the advanced properties
22:20 forrest lame
22:20 miles32 I run the second as soon as the first completes
22:20 forrest miles32, yeah I was just wondering if something was setting it back. Someone brought that up last week
22:21 miles32 I might try creating a local admin and assigning that as the owning user in file.recurse
22:21 aquinas joined #salt
22:24 aquinas_ joined #salt
22:29 martoss joined #salt
22:31 p3llin0r3_ joined #salt
22:32 mefrackindeiros joined #salt
22:37 godber forrest: it worked as expected ...
22:37 forrest godber, awesome
22:38 godber an example of state.git.latest using 'name' would have eliminated my initial confusion
22:39 forrest godber, add one!
22:40 godber :)
22:40 godber https://gist.github.com/godber/20d3bc25185d79507d61
22:40 godber basically that situation ...
22:40 mefrackindeiros left #salt
22:40 godber lessee if I can add one quickly
22:41 godber are the docs with the code?
22:42 forrest godber, they should be in the comments
22:43 godber you want a patch or want me to fork the repo and send a pull request
22:43 forrest godber, please fork and send a PR
22:43 forrest then it will get pushed through jenkins (not a concern in this case, but still)
22:44 ninkotech joined #salt
22:44 godber can I put something like [noci] in the comment to prevent the ci from running?
22:45 ipmb joined #salt
22:45 alanpearce joined #salt
22:45 godber travis does that
22:51 pssblts joined #salt
22:57 orion_ joined #salt
23:01 bmatt hm. can I access a grain from with a grain module?
23:01 bmatt let's assume that the loader has already populated that particular key
23:03 bhosmer joined #salt
23:04 manfred bmatt: i do not believe you can do that.
23:05 bmatt hm
23:05 manfred you can use __opt__, but __grain__ doesn't appear to be added
23:06 bmatt durn.
23:06 manfred https://github.com/saltstack/salt/blob/develop/salt/loader.py#L374
23:06 manfred feel free to try it and see though
23:10 TheThing joined #salt
23:11 gmeno joined #salt
23:11 orion_ joined #salt
23:12 teebes joined #salt
23:14 aquinas_ joined #salt
23:14 mosen joined #salt
23:15 mordonez joined #salt
23:16 KyleG joined #salt
23:16 KyleG joined #salt
23:16 godber forrest: whats the general feeling on how concise documentation inside a functions docstring should be
23:16 joehoyle joined #salt
23:17 forrest godber, http://docs.saltstack.com/en/latest/topics/development/conventions/documentation.html
23:17 godber I feel like I am making too much of an example that should be elsewhere
23:17 forrest godber, *shrug*, I mean if it's just another code block example, seems fine to me
23:17 forrest Have you ever looked at docs and gone 'gosh, I wish there were fewer examples!'?
23:18 aparsons joined #salt
23:19 godber http://3df8a554.ngrok.com/ref/states/all/salt.states.git.html#module-salt.states.git
23:19 godber search for "Different branches"
23:19 forrest godber, yours is the global?
23:19 Nazca joined #salt
23:19 godber its the stuff at the bottom of the 'latest'
23:20 forrest did you add the 'different branches' text?
23:20 godber yes
23:20 forrest godber, can you make that a ..note::
23:20 manfred godber: +1 looks fine to me
23:20 forrest I agree, it looks good, just wrap that text in a note block
23:20 godber oh ok
23:20 godber lemme try
23:20 forrest then it's more noticeable
23:21 aparsons joined #salt
23:21 n8n joined #salt
23:21 forrest godber, .. note::
23:21 forrest there's a space, sorry
23:21 forrest between the dots and note
23:21 godber forrest: no sweat, tracking down example
23:22 godber oh its like any :: block
23:22 forrest godber, yep!
23:22 forrest the joys of RST
23:22 mosen i just started RST coming from markdown, and i have no idea what im doing :)
23:23 forrest mosen, it's pretty easy!
23:23 n8n joined #salt
23:23 manfred mosen: http://docutils.sourceforge.net/docs/user/rst/quickref.html
23:23 manfred that is the quickref I use
23:23 panchisco joined #salt
23:23 forrest manfred, yep the quick ref is good. I've used a lot of stuff over in the best practices docs as well mosen: https://raw.githubusercontent.com/saltstack/salt/develop/doc/topics/best_practices.rst
23:23 mosen thanks manfred ill keep it handy
23:24 n8n joined #salt
23:24 mosen forrest: ahh thanks for that document, i was looking at the source of other modules to document mine
23:24 forrest mosen, for sure
23:24 aparsons_ joined #salt
23:25 n8n joined #salt
23:25 forrest godber, that looks better, pull the example blog out so it's not in the note block
23:25 forrest and it will be perfect
23:25 godber is the nomenclature I use correct?
23:25 godber "attribute" for example
23:26 n8n joined #salt
23:26 kelseelynn joined #salt
23:26 n8n joined #salt
23:26 forrest godber, I can't remember
23:26 forrest basepi, are you around?
23:27 godber oh and I didn't even finish the sentance
23:27 basepi sorta.  was just heading out
23:27 forrest there's a page that has all of the nomenclature, but I can't find it
23:27 forrest basepi, do you have that link for the nomenclature page handy?
23:27 basepi the glossary?
23:27 n8n joined #salt
23:27 forrest yes
23:27 basepi http://docs.saltstack.com/en/latest/glossary.html
23:27 forrest basepi, thanks, I couldn't remember the name of the page
23:27 forrest godber ^
23:28 basepi The bottom of this could be useful as well, if talking strictly about state structure:  http://docs.saltstack.com/en/latest/ref/states/highstate.html
23:28 sherbs80 joined #salt
23:28 n8n joined #salt
23:28 * basepi is out for now
23:28 forrest bye, thanks!
23:28 n8n joined #salt
23:29 n8n joined #salt
23:30 sherbs80 hey Salt folks…possible bug found in pkgs: ?
23:30 n8n joined #salt
23:30 sherbs80 when listing,
23:30 forrest what's up sherbs80
23:30 sherbs80 mypkgs:
23:30 sherbs80 pkg.installed:
23:30 sherbs80 pkgs:
23:30 aparsons joined #salt
23:30 sherbs80 - pkg1
23:30 sherbs80 - pkg2
23:30 forrest sherbs80, could you please gist it?
23:30 sherbs80 crap, sry
23:30 forrest formatting gets all screwed up
23:30 sherbs80 yes
23:30 forrest np
23:30 p3llin0r3 joined #salt
23:30 forrest just makes it easier, also the error in the gist/pastebin as well
23:31 n8n joined #salt
23:31 sherbs80 basically, if one of the pkgs listed under ‘pkgs:’ is missing, none of the pkgs get installed
23:31 sherbs80 this is on centos
23:31 sherbs80 6
23:31 manfred sherbs80: pkgs needs to be a list
23:31 sherbs80 happens on 5 too
23:31 forrest sherbs80, that is correct. The way pkgs works is it does all the calls in a single command
23:31 sherbs80 sure, but yum will still install other pkgs if one of the pkgs is missing
23:31 n8n joined #salt
23:31 p3llin0r3 Hey guys, sorry to like, creep around and mooch info, but! I have a question:
23:31 sherbs80 it’ll fail if there’s a dpeendency issue
23:32 sherbs80 but not if there’s just one missing
23:32 sherbs80 @manfred: how would I write that?
23:32 manfred sherbs80: http://paste.gtmanfred.com/oJhbiJ/
23:32 n8n joined #salt
23:32 sherbs80 yeah, that’s exactly what I have
23:32 p3llin0r3 So, what I'm aiming to do is to have it so that when I push a fix to dev/qa/prod, it will automatically propagate to my configured servers. Any tips on what this is calledit / how to accomplish this?
23:32 sherbs80 and thanks
23:32 manfred no it isn't
23:32 p3llin0r3 ( git branches )
23:32 manfred sherbs80: you have pkgs: i have - pkgs:
23:33 forrest manfred, you're using pkgS.installed, that is inaccurate
23:33 forrest but other than that, looks good
23:33 manfred forrest: no it isn't
23:33 manfred │18:30:30         ╰ » mypkgs:
23:33 manfred │18:30:32         ╰ »    pkg.installed:
23:33 manfred │18:30:33         ╰ »       pkgs:
23:33 n8n joined #salt
23:33 manfred that needs to be - pkgs:
23:33 manfred not pkgs:
23:33 forrest manfred, I'm looking at your gist
23:33 forrest pkgs.installed:
23:33 forrest :P
23:33 manfred meh
23:33 forrest so much hate from manfred
23:33 manfred the pkgs: should still eb a list in his state and it isn't
23:34 forrest completely agree
23:34 manfred http://paste.gtmanfred.com/r0jnX/
23:34 manfred there :P
23:34 n8n joined #salt
23:34 forrest I just didn't want him to use pkgs.installed
23:34 * forrest throws manfred an exaggerated thumbs up
23:34 sherbs80 @manfred: confirmed, we have - pkgs: in our state
23:34 manfred oh, if this wasn't #salt... you would have a giant ascii thumbs up right now
23:34 sherbs80 same behavior
23:35 forrest sherbs80, so you want the state to fail if ANY of the packages fail to install?
23:35 icebourg joined #salt
23:35 forrest where as currently it's saying it completes even though some fail to install?
23:35 sherbs80 @forrest, no I expect pkg.installed to behave the same way as yum
23:35 sherbs80 so if one pkg is missing, the rest will still install
23:36 icebourg joined #salt
23:36 sherbs80 the thing is that the rest of the state keeps going
23:36 sherbs80 shouldn’t it fail?
23:36 sherbs80 I mean, I’m trying to guarantee state, so if Salt can’t guarantee that for whatever reason, I’d expect it to stop
23:36 icebourg joined #salt
23:37 manfred sherbs80: you are saying the rest of the states fail, or the rest of that state fails?
23:37 sherbs80 manfred: I’m saying nothing fails :)
23:37 manfred ok good
23:37 smcquay joined #salt
23:38 manfred one second
23:38 sherbs80 manfred: http://pastebin.com/0qnRdgsK
23:39 sherbs80 “vim” should be “vim-enhanced"
23:39 manfred yes yes, hold on one second
23:39 sherbs80 ok
23:39 manfred sherbs80: oh
23:39 manfred sherbs80: that one won't error
23:39 manfred sherbs80: a psuedo name of vim-enhanced is vim
23:40 johtso joined #salt
23:40 sherbs80 the way we have our repos configured doesn’t allow vim, but does allow vim-enhanced
23:40 sherbs80 that’s not the issue though — hte issue is that b/c it can’t find vim, NONE of the packages in that pkgs: block installs
23:40 sherbs80 …but then the state keeps rolling
23:40 * robawt highfives sherbs80
23:40 forrest sherbs80, are you using a requires?
23:41 forrest sherbs80, I assume it says that specific ID failed to finish right? It just did in my test.
23:41 manfred it doesn't matter, the state still shouldn't succeed if it fails
23:41 sherbs80 right
23:41 manfred and it currently is failing
23:41 manfred but still marks it as succeeds
23:42 manfred sherbs80: best option right now... use names: instead of pkgs: this will run 1 yum command per package in the names list.  I am still looking for why this might succeed when it should fail
23:42 forrest manfred, totally fails for me when I try to install 'shitpackage' and 'nano' and shitpackage doesn't exist
23:42 sherbs80 ok, awesome...thanks
23:42 forrest manfred, but 'vim' and 'nano' work fine
23:42 forrest this is on ubuntu though
23:42 manfred forrest: but this one does exist, he has just stripped out the alternative name
23:43 manfred sherbs80: what version of salt?
23:43 sherbs80 2014.7
23:43 sherbs80 er, latest
23:43 robawt forrest: confirmed on Centos too
23:43 forrest manfred, yeah, I wonder if it's a repo issue? sherbs80 are you adding the repo prior to this?
23:44 sherbs80 yes, the repos are added
23:44 forrest ok, can you add refresh: True to that pkg.installed state?
23:44 forrest and try it again
23:44 sherbs80 forrest: processing...
23:45 sherbs80 sry, I’m doing this for someone who’s not here atm
23:45 forrest no worries
23:45 druonysus joined #salt
23:45 alanpearce joined #salt
23:46 manfred it is failing here on my centos6 box
23:46 manfred http://paste.gtmanfred.com/vUeWV3/
23:47 manfred but it is failing because Python is capitalized
23:47 manfred ¯\(°_o)/¯
23:47 skyler Is there are way to get a summary of the results of a salt command? Specifically, I would like to be able to quickly tell if anything failed even if I target 20 minions. Right now, I have to scroll up and look for each minion's summary.
23:47 manfred when that is fixed, it failes on vim
23:48 manfred skyler: yes
23:48 manfred let me find it
23:48 manfred skyler: http://docs.saltstack.com/en/latest/ref/output/all/salt.output.highstate.html
23:49 sherbs80 manfred: does the rest of the state continue? are the other packages installed? or does it stop altogether?
23:49 manfred sherbs80: it stops
23:49 Eugene manfred - vim is indeed not a package in centos; vim-enhanced has provides: set, but Salt doesn't seem to Do The Right Thing like good ol' `yum install vim` does.
23:49 manfred Eugene: we know, read up
23:50 manfred sherbs80: http://paste.gtmanfred.com/7lt/
23:50 manfred it should be behaving as you want it to
23:50 Eugene Actually it's "read down", stupid scrollbar :-/
23:50 sherbs80 ok, so all of the packages don’t install
23:50 * Eugene crawls back under rock
23:50 manfred sherbs80: if you use names... it would install the other packages, but fail on the vim one
23:50 jsm_ joined #salt
23:51 sherbs80 manfred: perfect…ok, I”m definitely going to try that
23:51 manfred yeah
23:51 manfred the difference is, it is installing all of the packages in one yum command if you use pkgs
23:51 sherbs80 ah right, you just said that above…pardon me
23:51 manfred while names splits each entry int he names list into it's own sub state, completely seperate (that is done on the internals of salt) and then runs one yum command for each package
23:52 manfred sherbs80: http://docs.saltstack.com/en/latest/ref/states/highstate.html#names-declaration
23:52 mordonez joined #salt
23:52 manfred that overrides portion is actually new in 2014.7 that i added :)
23:52 sherbs80 ha, werd
23:53 manfred the names portion was there, just the part where you could override other stuff in the state underneath each entry in names is what I added
23:53 jsm_ with salt-cloud, is there a way to specify what location a new VM is created in?  From what I ready, you can only set the default location.  I can create, and specify alternate configurations, but that seems awkward
23:54 manfred jsm_: you can only set it in the provider file, there isn't a command line argument for that stuff yet
23:54 manfred jsm_: https://github.com/saltstack/salt/issues/10149
23:57 sherbs80 manfred: if you use names: on a bogus package, does the state still fail?
23:57 manfred only the state in the lowstate fails iirc.
23:57 manfred http://docs.saltstack.com/en/latest/ref/states/highstate.html#names-declaration check how it expands in the low state data
23:58 manfred sherbs80: yup, only the bad state fails, cause it actually returns back 3 states in the final information
23:59 manfred sherbs80: http://paste.gtmanfred.com/Fqxf/
23:59 manfred http://paste.gtmanfred.com/paFD/

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