Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-09-25

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

All times shown according to UTC.

Time Nick Message
00:00 mpanetta_ Yeah :(
00:00 frankS2 joined #salt
00:01 whytewolf mpanetta_: try it with the mine runner https://github.com/saltstack/salt/issues/11509#issuecomment-138980156
00:01 evidence mpanetta_: i believe that's on purpose.  grains.get is a silent failure to allow you to specify a default of it's not populated
00:01 voxxit joined #salt
00:01 evidence salt['grains.get']() vs grains.get()
00:02 mpanetta_ Those both do the same thing I thought?
00:02 mpanetta_ Anyway, thats not the issue.  It was me using mine in pillar heh
00:02 k00l hey whytewolf
00:02 geekatcmu joined #salt
00:02 whytewolf hi k00l
00:02 mpanetta_ whytewolf: Trying that now, thanks :)
00:02 evidence negative.. salt['grains.get']() will fail, grains.get() returns null which allows you to set a default
00:03 akafred joined #salt
00:03 evidence mpanetta_: ah good deal.. well the distinction is useful
00:03 mattl joined #salt
00:03 mpanetta_ Ahh
00:03 k00l what can an orchestrate do that a state cant ?
00:03 evidence grains.get('blah', mydefault)
00:03 twodayslate joined #salt
00:06 whytewolf k00l: orch is used for if you have an order of operations that crosses servers. such as complex clusters, or software that relys on a chain of command
00:06 mpanetta_ odd, whytewolf now I get this: TypeError: runner() got multiple values for keyword argument 'fun'
00:06 whytewolf mpanetta_: what version of salt. that might not work on older versions
00:06 mpanetta_ Bitching about this for some reason: {% set minions = salt['saltutil.runner']('mine.get', tgt=product~'-redis-'~cluster~'-'~dc~'-'~env~'-*', fun='network.interfaces') %}    <======================
00:07 k00l whytewolf: thanks, so for localized stuff states are fine ?
00:07 mpanetta_ whytewolf: 2015.5.3 I think, let me check
00:07 mpanetta_ yeah 2015.5.3
00:07 whytewolf mpanetta_: ahhh that might be the problem. it was slated for 2015.5.6 [which IIRC isn't out yet] and 2015.8
00:07 mpanetta_ Ah crap.
00:08 mpanetta_ so I can do a command.run salt-runner blah?
00:08 mpanetta_ cmd.run even
00:08 whytewolf k00l: yeah states should be fine for local states.
00:09 * whytewolf shrugs at mpanetta_, I don't try to access mine data in pillar so you are out of my depth right now
00:09 k00l whytewolf: thanks for the input
00:09 mpanetta_ whytewolf: Well I will let you know :)
00:09 whytewolf mpanetta_: thank you
00:16 zmalone joined #salt
00:17 jeffspeff joined #salt
00:20 chips joined #salt
00:20 chips Hi all
00:20 johnkeates Hi chips
00:21 johnkeates I'm eating you right now
00:21 johnkeates hope you don't mind
00:21 chips Goes well with salt :)
00:21 johnkeates exactly! :D
00:22 chips Perhaps a noob question, but if I'm using salt-cloud to deploy nodes, is there a way to set the profile I use as a grain to be accessed later?
00:22 rdxc joined #salt
00:22 johnkeates yes, use a state that sets the profile and checks to only do so when it doesn't already exist
00:23 ahammond I tried upgrading my salt master using the repos.saltstack.com stuff. The load level on the server went to 11 (normally hovers around 1). I've tried downgrading and the load is still very high. Any suggestions for finding the culprit here?
00:23 chips How would the state know which profile to set? Can it get that info from salt-cloud?
00:23 ahammond the load is high enough that it's breaking a bunch of states in the process. :(
00:24 iggy ahammond: running salt-api?
00:26 ahammond iggy not using it, I don't think. I'll check.
00:26 ahammond iggy ps axww | grep api finds nothing.
00:31 mpanetta_ whytewolf: Yeah, not working well...
00:33 chips I think I found the answer to my own question (I missed that I can have specific minion config for each profile)  Thanks all
00:35 evidence iggy: https://github.com/saltstack/salt/issues/27391
00:39 Ryan_Lane the constant refactoring of the utils directory is killing me
00:39 Ryan_Lane this is the third release in a row that the json serializer moved
00:40 druonysuse joined #salt
00:40 * robawt highfives druonysuse
00:40 Ryan_Lane I guess I should just upstream our json execution module so that it'll get fixed along with core
00:42 Ryan_Lane oh. it's a native feature...
00:42 Ryan_Lane I could just drop this module
00:45 rdxc joined #salt
00:47 Akhter joined #salt
00:49 breakingmatter joined #salt
00:49 seatan joined #salt
00:52 knite_ joined #salt
00:58 Akhter joined #salt
00:58 hal58th ahammond If you downgraded, it might not be salt... check out http://serverfault.com/questions/171494/top-shows-load-of-2-5-how-can-i-find-out-which-processes-are-waiting-for-the-cp
01:03 ahammond hal58th at this point, I've rebooted, wiped caches, deleted .pyc files and yeah, I checked top (as well as vmstat, etc) and it's salt-master grinding the CPUs
01:03 ahammond I'm also seein a bunch of these in /var/log/salt/master
01:03 ahammond 2015-09-24 17:51:41,396 [salt.transport.zeromq                    ][ERROR   ][25961] Bad load from minion
01:04 catpiggest joined #salt
01:12 Akhter joined #salt
01:15 quasiben1 joined #salt
01:16 jmreicha joined #salt
01:17 ahammond dammit. I had some old processes floating around.
01:24 Akhter joined #salt
01:24 otter768 joined #salt
01:25 quasiben joined #salt
01:26 zmalone ahammond: Something like https://github.com/saltstack/salt/issues/26161 would probably solve that problem, as well as a few others (like having multiple minion processes running by accident and consuming from the master's queue)
01:29 ahammond zmalone agreed that it would help
01:31 zmalone You could probably wrap all your init scripts in something like http://www.unixwiz.net/tools/lockrun.html , but that would be pretty klugey.
01:32 zmalone (or whatever service manager you use, if it isn't init.d)
01:34 pentabular joined #salt
01:35 teebes joined #salt
01:38 colorfuzzy joined #salt
01:38 iggy that fix is in 2015.5.5.
01:39 sunkist joined #salt
01:50 seatan joined #salt
01:54 berserk joined #salt
01:55 saltnoob joined #salt
01:55 rdxc joined #salt
01:55 ahammond manage.up and manage.down agree that my shss servers should be working. However test.ping tells me Minion did not return. [Not connected] for all of them.
01:56 ahammond master is 2015.8.0, minions are 2015.5.5
01:56 quasiben1 joined #salt
01:57 alexlist_ joined #salt
01:59 ahammond but that's only when I use a nodegroup to query the servers. WTF?!?!
01:59 ahammond ok, I'm going home because clearly computers are done working with me today.
02:03 hasues joined #salt
02:03 hasues left #salt
02:11 zmalone joined #salt
02:14 zmalone iggy: That only touches a lock file on RHEL, and it doesn't seem to prevent multiple runs if a lock file already exists.
02:15 zmalone there may be good reasons for not locking so that only one minion is running, I don't know.
02:16 bhosmer joined #salt
02:26 pentabular joined #salt
02:32 favadi joined #salt
02:34 clintberry2 joined #salt
02:34 quasiben joined #salt
02:42 PeterO_ joined #salt
02:44 neogenix joined #salt
02:45 quasiben joined #salt
02:45 DammitJim joined #salt
02:50 breakingmatter joined #salt
02:51 seatan joined #salt
02:55 ITChap joined #salt
03:04 cppking left #salt
03:06 pratikmallya joined #salt
03:12 writtenoff joined #salt
03:14 hod0r joined #salt
03:16 evle joined #salt
03:19 jan_ joined #salt
03:25 otter768 joined #salt
03:42 mosen joined #salt
03:45 rhodgin joined #salt
03:52 seatan joined #salt
03:58 Akhter joined #salt
04:05 PeterO_ joined #salt
04:08 andrew_v joined #salt
04:09 iggy you were the one that pointed it out
04:09 iggy there are... there's a closed ticket about it
04:13 clintberry2 joined #salt
04:24 msciciel joined #salt
04:46 markm_ joined #salt
04:51 breakingmatter joined #salt
04:52 seatan joined #salt
05:20 armguy joined #salt
05:26 otter768 joined #salt
05:27 felskrone joined #salt
05:34 ramteid joined #salt
05:35 armguy If I can telnet without issue to 4506 on the master but the minion just hangs at Initializing new AsyncZeroMQReqChannel for ('/etc/salt/pki/minion', 'test1.lan', 'tcp://192.168.89.23:4506', 'clear') does anyone know what I might be doing wrong? The minion will not show up to accept the key
05:37 sunkist1 joined #salt
05:40 traph joined #salt
05:42 felskrone armguy: do your zerome versions match on both ends? was an issue for me a while ago
05:42 felskrone zeromq
05:42 armguy lemme double check thanks
05:43 armguy fingers crossed they were different. Would not have thought to check that thanks
05:47 armguy felskrone~ negative still same result. its puzzling no FWs on either box both on lan, telnet works just fine to both ports from minion to master
05:48 felskrone what does tcpdump say?
05:48 iggy to be clear, you mean you can telnet from minion to master on 4506/4505?
05:48 armguy correct iggy
05:51 armguy let me get the tcpdump output pasted
05:54 malinoff joined #salt
05:54 armguy http://codepad.org/In1MBwfR
05:54 JavaAtom joined #salt
05:55 seatan joined #salt
05:58 colttt joined #salt
05:58 felskrone hm, does not look that wrong. besides the first three lines of FIN/ACKs from an older connection, the lines below seem fine to me
05:59 mohae joined #salt
05:59 armguy yeah everything seems like it should work
06:02 felskrone do you see open connections on 4505/4506 with lsof -ni:<port>?
06:03 felskrone there should only be one on 4505, 4506 is actually just for returning data to the master, it does not stay open all the time
06:05 armguy hrm yeah no 4505 on minion, but see both on master
06:06 felskrone ok, so that definately broken then, but we knew that already  :-)
06:06 Fiber^ joined #salt
06:06 felskrone do you see anything in the master log when running it in debug mode?
06:06 lb joined #salt
06:07 armguy nothing about the minion in the master logs Just that its accepted 0 minion keys
06:07 |Trend| joined #salt
06:07 An_T_oine joined #salt
06:11 felskrone my final idea, what kernel are you running, we had problems with ancient 2.6.x kernels last year and had to rebuild the zeromq packages with older kernel sources
06:11 armguy 4.1.6-1-ARCH
06:14 felskrone well, thats pretty new :-)
06:14 armguy same kernel on both master and minion 2015.8.0 running both on Arch Linux. Brand new setup first minion to this master
06:15 mohae joined #salt
06:17 felskrone hm, maybe to new? no idea if there might by anything broken with that combination
06:18 armguy OK np appreciate the time :) I will play around some more
06:19 svinota joined #salt
06:20 perfectsine joined #salt
06:26 iggy armguy: does a minion on the master work?
06:27 ramteid joined #salt
06:27 armguy I just set up one on another box and its working. The master is in a VM(if that matters) the one working is hosting the VM that is running the master
06:29 armguy I am seeing this on the minion not working that does not show up on the working one. Please install 'virt-what' to improve results of the 'virtual' grain.
06:32 armguy but its not installed on the working one
06:37 armyriad joined #salt
06:39 DanyC joined #salt
06:41 BlackFX joined #salt
06:49 armguy joined #salt
06:51 oravirt joined #salt
06:53 breakingmatter joined #salt
06:55 seatan joined #salt
06:58 pratikmallya joined #salt
06:59 rubendv joined #salt
06:59 rubendv joined #salt
07:00 GreatSnoopy joined #salt
07:02 akafred joined #salt
07:02 esharpmajor joined #salt
07:03 Ztyx joined #salt
07:03 pratikmallya joined #salt
07:04 kukacz joined #salt
07:09 kukacz joined #salt
07:09 malinoff joined #salt
07:10 armguy joined #salt
07:10 twork https://gist.github.com/mjinks/31267d8e2945b3764011
07:11 twork short version: i can't figure out how to turn a single config template into one that loops over a list of them.
07:12 twork i'm sure there must be a standard (obvious) construction for what i want to do. but so far no joy.
07:12 Ztyx left #salt
07:12 Ztyx joined #salt
07:13 twork ...and having posted that my wheels are falling off for the night. think i'll re-post to the list. sorry for the dupe.
07:15 eseyman joined #salt
07:20 Grokzen joined #salt
07:20 Bryson joined #salt
07:21 ingwaem joined #salt
07:24 kawa2014 joined #salt
07:25 aqua^c joined #salt
07:27 otter768 joined #salt
07:39 katyucha joined #salt
07:39 jhauser joined #salt
07:39 babilen twork: "{%- for user, conf in pillar.get('vsftpd_user_config', {}).items() %}" looks about alright (or do you need support for absent users? (i.e. users you want to remove like in the users formula you copied this from)
07:39 linjan_ joined #salt
07:43 babilen )
07:44 traph joined #salt
07:46 KermitTheFragger joined #salt
07:47 favadi joined #salt
07:47 cppking joined #salt
07:48 cppking hello guys, what's the magic behind  cp.get_dir ? does it use rsync ??
07:48 fersur3 joined #salt
07:49 fersur3 When doing ./bootstrap git $version what version string should be used for 2015.8 ?   2015.8.0 ??
07:50 cppking just try all of them
07:51 saffe joined #salt
07:51 chiui joined #salt
07:52 cppking I found that cp.get_dir can't keep the mode of file
07:54 Antiarc joined #salt
08:00 babilen fersur3: You can either specify a branch, tag or SHA
08:00 babilen (IIRC)
08:01 fersur3 Yes.  What I want to know is what is the newest / release / stable?    v2015.8.0 and v2015.8 look different
08:04 Grokzen Is it possible to get the username of the user that run a custom salt module function?
08:04 larsfronius joined #salt
08:05 babilen fersur3: That is because they are different. v2015.8.0 is the latest release
08:05 fersur3 Thank you babilen that's exactly what I needed to know
08:05 mattiasr joined #salt
08:06 fersur3 babilen, seems a few dependencies don't get raked in on SmartOS base64 15.2.0 can you have someone check that out?   I'll submit issues on GitHub but a Salt dev should definitely QA 2015.8.0 on SmartOS base64 15.2.0
08:07 babilen I might like to point out that I am not a saltstack employee
08:07 fersur3 shut up, get back to work
08:07 babilen Even though if saltstack felt the urge to pay me I wouldn't say no ;)
08:08 fersur3 SaltStack is also welcome to employ me, my fingers are ready
08:12 elsmo joined #salt
08:15 s_kunk joined #salt
08:15 s_kunk joined #salt
08:16 Antiarc joined #salt
08:19 alexlist__ joined #salt
08:27 kawa2014 joined #salt
08:28 thalleralexander joined #salt
08:29 alexlist_ joined #salt
08:31 aristedes joined #salt
08:31 markm joined #salt
08:32 aristedes left #salt
08:32 aristedes joined #salt
08:32 seatan joined #salt
08:33 aristedes left #salt
08:34 Lionel_Debroux joined #salt
08:34 favadi joined #salt
08:40 chiui joined #salt
08:42 voxxit joined #salt
08:53 Antiarc joined #salt
08:54 breakingmatter joined #salt
08:55 stephanbuys joined #salt
09:09 DanyC joined #salt
09:11 favadi joined #salt
09:13 traph joined #salt
09:26 tampakrap using ssh_auth, is it possible to purge the unmanaged ssh authorized_keys?
09:26 Antiarc joined #salt
09:28 trph joined #salt
09:28 otter768 joined #salt
09:28 svinota joined #salt
09:34 zerthimon joined #salt
09:38 babilen tampakrap: You could use a file.absent with a prereq on the ssh_auth states (so that one of them has changes the file is blown away)
09:39 tampakrap thank you, I'll try that
09:40 babilen There might be a "consolidate" option hidden away somewhere, but that would be an alternative way
09:41 babilen (use either prereq or prereq_in depending on where you'd like to maintain that association)
09:46 DanyC hi, anyone with salt-cloud experience/ hands-on here ?
09:47 DanyC if yes, anyone can pls help me understand where/ how you can say via salt-cloud: i want x VMs of type x ? in which files is this info defined ? I know on OpenStack HEAT but not on salt-cloud
09:49 favadi joined #salt
09:53 trph joined #salt
10:00 sgargan joined #salt
10:01 Grokzen joined #salt
10:01 elsmo Morning all am having issues with  "TypeError: unhashable type: 'dict'" with an ssh_auth.present state, which was previously working..
10:02 colorfuzzy sls not found error: https://gist.github.com/ColorFuzzy/7ed9b45cd63894414eb3  thanks very much
10:03 elsmo http://pastebin.com/rDgpzMLb is my state..
10:04 elsmo colorfuzzy: is your server directory the root for your state files?
10:04 colorfuzzy @elsmo /srv/salt is the root state file
10:04 colorfuzzy server is a directory in /srv/salt
10:04 elsmo DanyC: https://docs.saltstack.com/en/develop/topics/cloud/index.html might help?
10:05 colorfuzzy Thanks, let me have a look
10:07 elsmo colorfuzzy: I can't see anything glaringly obvous, however am not quite a salt expect :)
10:08 elsmo (*expert)
10:08 quasiben joined #salt
10:08 colorfuzzy https://github.com/saltstack/salt/issues/10852
10:08 colorfuzzy maybe is a bug, I use 2015.5.3
10:09 sgargan joined #salt
10:09 Antiarc joined #salt
10:10 giantlock joined #salt
10:15 culexor joined #salt
10:15 geekatcmu joined #salt
10:15 kossy joined #salt
10:15 [vaelen] joined #salt
10:15 adendrag joined #salt
10:15 gchao joined #salt
10:17 sgargan joined #salt
10:18 blergh- joined #salt
10:18 blergh- So, what's the web-gui of choice these days?
10:18 bhosmer joined #salt
10:21 tkiel_ joined #salt
10:23 sgargan joined #salt
10:29 colorfuzzy joined #salt
10:29 DanyC elsmo: thanks
10:30 colorfuzzy joined #salt
10:30 DanyC and anyone knows if via salt-cloud i can create volumes on OpenStack provider ? (i couldn't find any on doc)
10:31 tkiel_ joined #salt
10:33 sgargan joined #salt
10:36 elsmo DanyC: I don't think you can, or couldnt when I last looked a while ago.
10:36 seatan joined #salt
10:36 DanyC elsmo: i looked on AWS and is possible according to docs but haven't seen anything on OpenStack
10:37 DanyC elsmo: i.e in case anyone is interested - http://salt-cloud.readthedocs.org/en/latest/topics/aws.html#creating-volumes although that might be outdated
10:37 elsmo I solved that by using Fabric and pyrax to create servers, sort out DNS and loadbalancing etc, and then called highstate on them.
10:37 traph joined #salt
10:37 traph joined #salt
10:37 elsmo Yeah the openstack stuff always seem to trail AWS a little which is a bit of a shame
10:38 DanyC elsmo: sure but then i can easily stay on using HEAT + salt master & salt minion rather then salt-cloud ;)
10:38 elsmo You could probably create a salt state that calls some python code that attachs the volumes through the API?
10:41 DanyC elsmo: sure that is another option
10:45 sgargan joined #salt
10:46 tkiel_ joined #salt
10:48 fredvd joined #salt
10:53 DanyC elsmo: and i just found this https://docs.saltstack.com/en/develop/topics/cloud/features.html which is pretty shocking to see how little feature set is between ec2 vs the rest of cloud providers
10:54 less joined #salt
10:54 gchao joined #salt
10:54 kossy joined #salt
10:54 less left #salt
10:54 adendrag joined #salt
10:54 culexor joined #salt
10:55 [vaelen] joined #salt
10:55 breakingmatter joined #salt
10:56 daemonkeeper Does anybody have an idea how determine in a grain or wherever on what commit the most recent highstate is based on? I'd like to write to /etc/motd something like "hello sir, you are logged in to {{ grains.fqdn }} this system was last updated by a salt highstate called at {{ highstate.time }} based on gitfs commit {{ whatever }"
10:57 daemonkeeper Minus highstate.time to avoid noise maybe.
10:59 elsmo DanyC: heh, wow, yeah you're right!
11:00 felskrone daemonkeeper: would you minding posting that to the mailing list? i would be very interested in approaches on how to do something like that
11:00 tkiel_ joined #salt
11:01 elsmo daemonkeeper: not sure through salt directly, as a part of a salt state you could cat the current date into the ssh MOTD file..
11:01 elsmo daemonkeeper: you might also be able to use a mine function to update the current gitversion, but thats only a vague theory ;)
11:01 daemonkeeper Yeah. More important is the commit though. Having the time there is probably more noisy than anything else since it would change even with void highstates.
11:02 aqua^c joined #salt
11:02 daemonkeeper How would you _know_ the git version in the first place elsmo? I don't think I'd have the cloned repository on the minion somewhere which runs the code in the end
11:03 alexlist_ joined #salt
11:03 daemonkeeper Oh. But a reactor could possibly do
11:03 daemonkeeper Hurm.
11:04 elsmo daemonkeeper: i was thinking if the master was also a minion, then some mine function that finds the commit from the cached gitfs dirs on the master might do it
11:04 elsmo feels a bit hacky though
11:04 Ztyx left #salt
11:05 daemonkeeper Not sure how that would work. It would only work by automatically creating a meta-data file at commit/push time in a repository hook.
11:05 felskrone daemonkeeper: do you generate the motd or just the motd.tail? there are different approaches on how its generated depending on the distro
11:05 daemonkeeper In my case I'd overwrite motd directly, but that's a distribution detail as you pointed out.
11:06 felskrone with a file from the master or do you generate it?
11:07 daemonkeeper Well. I'd do a jinja template which comes from the master, but then, the real question underneath was: where/how to gather the information I would like to put there.
11:08 felskrone a git-hook always updating template somehow would do, but that sounds a little wrong, there must be a better way
11:08 daemonkeeper Indeed.
11:11 felskrone do you have a link to the documentation regarding the git stuff you're using? i might take a look at the code if i find the time
11:11 daemonkeeper https://docs.saltstack.com/en/latest/topics/tutorials/gitfs.html
11:12 breakingmatter joined #salt
11:13 daemonkeeper I think, a reactor talking to the master would be the most suitable data bridge.
11:13 daemonkeeper But I have no idea how the reactor code itself could determine the information needed.
11:13 giantlock joined #salt
11:14 felskrone easy, write a setup a reactor that executes a runner
11:14 felskrone write a reactor that executes a runner that runs the git-comand you want
11:14 daemonkeeper Well, and where? :>
11:14 felskrone i think whats more interesting is how to get info back to the minion
11:15 felskrone because you are already running a highstate where the info is needed.
11:15 felskrone where what? :-)
11:17 daemonkeeper Hence a reactor, that should allow me to fire an event from the minion pointing to the reactor's master giving the information back. But then again, how could the reactor know, e.g. where the repository is located and which branch it used to highstate, since that'S a configuration detail of the master.
11:19 felskrone aah, ok…. i only run a single master branch on my masters and the development is being done in vagrant vms. so for me getting that info would not be too hard
11:19 felskrone but how to get it into the highstate / to the minion still needs a solution :-)
11:21 felskrone firing events is async, so there is nothing you could wait for
11:22 felskrone what about setting up a reactor that listens on state.highstate events and updates the minions pillars (somehow…) before executing the highstate?
11:23 daemonkeeper That's even worse than a post-receive git commit hook im
11:23 daemonkeeper *imo
11:23 daemonkeeper Well. For me. Since pillars are also in gitfs here.
11:24 daemonkeeper Also, if the event system is inherently asynchronous, updating the pillar while the highstate already started is likely already too late.
11:24 daemonkeeper Since on of the first thing highstate does is to update the minion pillar, right?
11:26 felskrone i think so, yes
11:26 felskrone i'd post that topic to the mailing list, i bet it can be done nicely
11:29 otter768 joined #salt
11:30 daemonkeeper Possibly so
11:30 wm-bot4 joined #salt
11:38 elsmo I have joined the mailing list soley to see this discussion! Another thing I wonder about is changing the gitfs refresh from a time based one, to one that can be triggered by a github hook for example
11:40 daemonkeeper elsmo: https://docs.saltstack.com/en/latest/topics/tutorials/gitfs.html#refreshing-gitfs-upon-push
11:41 elsmo Aha, so I'd need to write a little github-hook-listener to auto update it, but still thats quite nice
11:42 sgargan joined #salt
11:50 mike25de joined #salt
11:52 rhodgin joined #salt
11:53 vfdfvdfv joined #salt
11:54 sgargan joined #salt
11:58 losh joined #salt
12:00 ferbla joined #salt
12:01 dendazen joined #salt
12:11 adendrag joined #salt
12:14 evle joined #salt
12:16 babilen daemonkeeper: Also keep in mind that there are potentially many GitFS backends at play and that it is entirely unclear which SHA you want
12:20 mapu joined #salt
12:25 evle1 joined #salt
12:42 numkem joined #salt
12:42 DammitJim joined #salt
12:44 teebes joined #salt
12:49 markm joined #salt
12:51 dijit hey guys.
12:51 dijit I'm havng some weirdness with my "name"s, they seem to always substitute a "_" with a "-"
12:51 dijit which causes some headaches with postgres.
12:56 subsignal joined #salt
12:57 martyni joined #salt
12:59 martyni Does anyone have a good guide for using pre-existing salt states for deploying into docker containers?
12:59 av_ joined #salt
13:00 malinoff martyni: docker containers suppose to be idempotent already - maybe you want to _build_ a container instead?
13:00 malinoff or maybe run it into a docker server?
13:01 racooper joined #salt
13:03 av_ guys I need some help. I posted on reddit so linking here, hope thats okay: https://www.reddit.com/r/saltstack/comments/3mbwqm/whats_wrong_with_my_salt_config/
13:03 Aleks3Y joined #salt
13:06 subsignal joined #salt
13:07 JDiPierro joined #salt
13:07 martyni malinoff: Yes building the containers would be fine. Basically we already have an a lot of stuff written for deployment using salt and we were looking into the best way to use that to deploy containers
13:07 quasiben joined #salt
13:08 malinoff martyni: well, you can throw all you have for now :) docker is a brand new technology (tm) which provides completely new ways of doing usual things
13:08 honestly Completely new, completely unencrypted
13:09 malinoff martyni: think of docker as an alternative to deb/rpm packaging, not as deployment platform/system/whatever else
13:09 TooLmaN joined #salt
13:09 racooper av_,  have you tried (on line 14 of your venv.sls) using sls: git_repo instead of git: git_repo ?
13:09 av_ no
13:09 av_ let me try that
13:10 racooper actually..it would be sls: git but that may not be what you're wanting
13:11 av_ failed :/
13:11 rdxc joined #salt
13:14 sgargan joined #salt
13:15 trph joined #salt
13:15 racooper in that case...someone with more experience with requisites will be able to help better. sorry!
13:16 av_ no problms, thanks!
13:17 pcn av_ how are you kicking off your runs?
13:18 pcn Because from what I see, if you're running them one at a time, there's no reason that git.sls will be loaded
13:20 pcn have you tried salt '*' state.sls base.git,base.venv ?
13:20 cpowell joined #salt
13:20 bezaban joined #salt
13:22 dendazen joined #salt
13:22 pcn Or use top.sls to construct a highstate of these two states?
13:23 jryberg joined #salt
13:23 av_ I run one after another
13:23 av_ is that not enough?
13:23 jryberg Hi, Has anyone tried to restart services in Windows? I'm trying to restart salt-minion but got the following result back: ERROR: The system cannot find the file specified.
13:24 av_ I will try running them together now
13:24 av_ failed
13:24 pcn av_: for a requires, you may need to include one in the other, or run a highstate with both - I haven't tried this, I want to keep dependencies in the same sls file and order them from a separate state
13:25 av_ I ran: sudo salt '*' state.sls base.git,base.venv -l debug -v
13:25 jryberg my sls file: http://pastebin.com/dc8ujUeV
13:25 pcn I thought that perhaps that would be the case: could be the same as two separate state runs?
13:25 av_ pcn I don't understand
13:25 av_ can you elaborate
13:26 av_ anyways, with high state also it fails. I started running one after another as specified in `init.sls` and it fails at this point
13:28 cpowell joined #salt
13:28 av_ this is output when I ran it with high state: http://dpaste.com/2KQ8AWV
13:29 protoz joined #salt
13:29 otter768 joined #salt
13:31 pcn av_ sls: git_repo?
13:31 perfectsine joined #salt
13:32 An_T_oine joined #salt
13:32 pcn It should be git: git_repo shouldn't it?
13:33 pcn And in your venv.sls, I think your dep on pkg: instal_deps needs a second "l"?
13:34 edulix joined #salt
13:34 av_ I removed pkg require and it still fails at git
13:34 av_ yes, it is `git: git_repo`
13:35 pcn Not according to that failure message?
13:35 av_ wait I posted a wrong one. actually when I was testing I removed it and ran just to make sure it has nothing to do with instal_deps. However even if I remove it, it still fails
13:36 trph joined #salt
13:38 sunkist joined #salt
13:40 pcn av_ since git_repo in your case relies on a ssh key etc. could you try constructing a different dependency, perhaps on a file or directory existing, and see if that works?  That can give you a reproducer that other people can try out.
13:41 av_ sure, I will do that
13:45 timoguin joined #salt
13:45 colorfuzzy joined #salt
13:49 murrdoc joined #salt
13:50 dyasny joined #salt
13:52 sgargan joined #salt
13:52 jdesilet joined #salt
13:52 seatan joined #salt
14:00 anmol joined #salt
14:01 kaptk2 joined #salt
14:03 zmalone joined #salt
14:06 theologian joined #salt
14:08 dyasny joined #salt
14:10 tmclaugh[work] joined #salt
14:12 cyborg-one joined #salt
14:12 danemacmillan joined #salt
14:14 rim-k joined #salt
14:14 pravka joined #salt
14:14 rim-k hello everyone!
14:14 rim-k got a quick question
14:15 rim-k I see that the last debian package for wheezy is for 2015.5
14:15 rim-k did you stop packaging for wheezy?
14:16 rim-k i.e. any chance to get 2015.8 for wheezy?
14:18 DammitJim joined #salt
14:18 Nazca__ joined #salt
14:18 Akhter joined #salt
14:20 cyborg-one joined #salt
14:20 seatan joined #salt
14:22 Akhter joined #salt
14:22 neogenix joined #salt
14:23 breakingmatter joined #salt
14:24 Akhter joined #salt
14:25 numkem joined #salt
14:30 TheoSLC joined #salt
14:30 clintberry2 joined #salt
14:30 andrew_v joined #salt
14:31 darix any ideas why this snippet leads to salt doing the cleanup on every run? https://gist.github.com/darix/3f0c135125a97e4f63c6
14:31 murrdoc - clean: True /
14:32 av_ I am not sure, but you have     - clean: True
14:33 darix murrdoc: the docs says ... "Make sure that only files that are set up by salt and required by this function are kept. If this option is set then everything in this directory will be deleted unless it is required"
14:33 murrdoc yeah
14:33 murrdoc rm -rvf <dir>/*
14:33 murrdoc except the one fle
14:33 darix so for me that sounds like "it should not do delete files generated by salt in the previous run"
14:35 bhosmer joined #salt
14:35 darix which it can track by the checksum of the files in the directory
14:36 An__T__oine joined #salt
14:36 stephanbuys joined #salt
14:37 murrdoc which file is it deleting
14:37 murrdoc the required one too ?
14:37 aristedes joined #salt
14:38 aristedes joined #salt
14:38 darix yes
14:39 darix basically each run gets the entry for the directory + file
14:39 murrdoc salt —version
14:39 bluenemo joined #salt
14:39 TheoSLC joined #salt
14:39 simmel joined #salt
14:40 aqua^c joined #salt
14:41 darix salt 2015.5.3 (Lithium)                                                                                                                                                                                                                      Ã¢â€â€šÃ‚·
14:41 _JZ_ joined #salt
14:42 darix hmm let me retest with 2015.8
14:43 darix still broken
14:43 darix ^^
14:45 TheoSLC joined #salt
14:46 murrdoc that is wierd
14:46 murrdoc wfm in 2014.7
14:46 murrdoc :)
14:46 qwerty joined #salt
14:46 murrdoc but then i took it out
14:46 murrdoc cos i hate state re ordering
14:47 Guest21682 Hi there, can someone out here help me with salt loader?
14:48 rim-k anyone for my question regarding the wheezy packaging?
14:49 darix murrdoc: so calling highstate twice does not do the operation twice?
14:49 darix hmm
14:49 darix good to know. to the bugtracker!
14:50 murrdoc darix:  yeah i work with 0 requires
14:50 murrdoc in my code
14:51 darix ok
14:51 darix *test*
14:51 murrdoc which means i cant use clean:True
14:51 darix murrdoc: so how would you make sure only managed files are in a directory?
14:52 darix :)
14:52 teebes joined #salt
14:53 murrdoc i believe in the lord jesus
14:53 murrdoc and hope that some day file.directory will work with requre_in
14:53 murrdoc lord base of pi will provide
14:54 murrdoc iggy:  can i get an amex
14:54 murrdoc amen*
14:54 breakingmatter joined #salt
14:54 darix murrdoc: amex might be more useful ;)
14:54 murrdoc iggy is broke
14:54 murrdoc but u are correct darix
14:55 numkem joined #salt
14:58 dthom91 joined #salt
14:59 TheoSLC joined #salt
15:00 hoonetorg joined #salt
15:01 sgargan joined #salt
15:03 Jeff_ joined #salt
15:04 colorfuzzy left #salt
15:05 Guest31149 I'm trying out salt on Ubuntu 14.04.   Various install instructions for the minion have gotten me salt 2015.5.3, which appears two releases old.  How do I install the latest stable release?
15:08 seatan is there a way to get the salt version from a grain?
15:09 pravka joined #salt
15:11 zmalone Guest31149: check the mailing list, 2015.5.3 is currently the latest version supported from the Salt PPA
15:12 zmalone A new repo was just announced on the mailing list, but I don't actually see anything on the Salt website yet.
15:12 zmalone Salt packages on Ubuntu seem to be in flux right now
15:13 bhosmer joined #salt
15:14 PeterO_ joined #salt
15:14 whytewolf the repo.saltstack.com does have ubuntu packages http://repo.saltstack.com/apt/ubuntu/ubuntu14/
15:15 whytewolf [just doens't have ubuntu install instructions in the docs]
15:16 pcn zmalone: Yss
15:16 pcn Yes, the repo.saltstack.com seems to have 2015.8.0 per whytewolf, and seems to work
15:17 tmclaugh[work] joined #salt
15:18 traph joined #salt
15:19 darix so we figured it out: 1. only file resoures are allowed to be kept. 2. you need to pass the *full* filename and not the name of the resource
15:19 neogenix joined #salt
15:21 dayid_ joined #salt
15:22 andrew_v joined #salt
15:28 Guest31149 ok. Thanks.
15:28 mpanetta joined #salt
15:29 nofxroky joined #salt
15:30 discypher joined #salt
15:30 otter768 joined #salt
15:31 nicmoski joined #salt
15:32 pratikmallya joined #salt
15:32 PeterO_ joined #salt
15:32 teryx5101 Hey guys, are the only state modules that work with Windows the ones that begin with win_ ? Or are there others that will work as well?
15:34 discypher joined #salt
15:34 pgoetz joined #salt
15:34 pratikma_ joined #salt
15:37 felskrone joined #salt
15:38 giantlock joined #salt
15:38 hasues joined #salt
15:39 hasues left #salt
15:40 nicmoski teryx5101: No, thought it's not well-documented. Here is an example of a module that works on multiple platforms: https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.status.html#salt.modules.status.procs
15:41 nicmoski For specific system commands (ps in this case) it uses grains to know which command to run based on the platform. A lot of modules are like this, but some are not. Again, it's not well documented. The best way to know is to test.
15:41 denys joined #salt
15:41 teryx5101 nicmoski: okay, thanks!
15:43 danlsgiga joined #salt
15:43 pcn Is there a way to specify multiple mailto or other environments in a cronab?
15:43 Gareth morning morning
15:43 pcn I'd like to have one per job
15:44 denys release 2015.8.0 does not seem to be available in ppa:saltstack/salt.  is someone working on it?  that would be great :-)
15:46 discypher joined #salt
15:47 pcn denys: it seems that salt is moving to repo.saltstack.com instead of a ppa
15:50 mapu joined #salt
15:51 discypher joined #salt
15:52 breakingmatter joined #salt
15:53 mpanetta Morning all.
15:54 mpanetta I'm trying to run the workaround for mine in pillar outlined in https://github.com/saltstack/salt/issues/11509 (the execution module AusIV put in the comments) and it isn't working.
15:54 mpanetta ANyone else here tried that yet?
15:55 mpanetta I've no idea what I am doing wrong
15:55 mpanetta The minions load the module fine without complaints.
15:56 mpanetta But when I do a pillar.items the pillar rendering crashes out saying it cant find my module (in the master logs)
16:00 dendazen joined #salt
16:02 Cyis anyone used git.latest to pull a repo down to a Windows server? I've done so with Linux servers with no problem, wondering how difficult it would be to duplicate for Windows
16:03 denys pcn: ah, ok, thanks!
16:07 alemeno22 joined #salt
16:09 sPollk joined #salt
16:12 mattiasr joined #salt
16:13 discypher left #salt
16:13 sunkist joined #salt
16:14 snarfy joined #salt
16:17 pdayton joined #salt
16:20 bhosmer joined #salt
16:20 jodv joined #salt
16:20 godber joined #salt
16:21 sunkist1 joined #salt
16:23 chiui joined #salt
16:29 writtenoff joined #salt
16:31 hasues joined #salt
16:31 hasues left #salt
16:41 danlsgiga hey folks... I'm starting to work with the orch runner and I'd like to create a var with data from a pillar.get 'tgt' 'mypillar:value' inside the orchestrate
16:41 danlsgiga is that possible?
16:41 danlsgiga Since the orch runs on the master, I don't want to bother passing data in the event from the minion to the master that will run the orchestrate
16:41 cruatta joined #salt
16:42 pcn I'm using pepper with 2015.8.0, and I'm trying to use jobs.lookup_jid, and it's failing:
16:42 danlsgiga to be honest I'm trying to avoid the use o mine
16:45 pcn pepper/cherrypy api failing at jobs.lookup_jobid: https://gist.github.com/pcn/9f18726562f9686d82ca
16:45 cruatta joined #salt
16:47 snarfy joined #salt
16:50 teak__ joined #salt
16:50 Akhter joined #salt
16:52 teak__ Hello! how does salt handle merging config keys that are specified multiple times across files in /etc/salt/minion.d ?    It seems the last instance wins, is there a param for telling salt to merge these config keys?
16:52 aparsons joined #salt
16:53 dthom91 joined #salt
16:54 hal58th teak__: I doubt it. I could see some things wanting to merge like schedules, but you wouldnt want multiple salt masters being defined :P
16:54 trph joined #salt
16:55 teak__ hal58th: defining multiple masters is fine, last I checked
16:55 hal58th In comma seperated fasion yes.
16:56 hal58th What I am trying to say is, I want to make sure my minion is pointed at the right salt-master. Maybe that was a bad example
16:57 timoguin joined #salt
16:57 bhosmer joined #salt
16:59 pcn Looks like my problem
17:00 breakingmatter joined #salt
17:01 chiui joined #salt
17:02 pcn No, OK, it is a bug in 2015.8.0.  Is it supported to run an older master with a 2015.8.0 client?
17:04 pcn fg
17:10 alemeno22 joined #salt
17:10 hasues joined #salt
17:10 hasues left #salt
17:13 pcn OK, I opened https://github.com/saltstack/salt/issues/27409
17:13 linjan joined #salt
17:14 vovcia left #salt
17:21 forrest joined #salt
17:21 bhosmer joined #salt
17:21 JDiPierro joined #salt
17:22 [vaelen] joined #salt
17:22 geekatcmu joined #salt
17:23 pcn Because of the issue I just posted ^^^ I need to ask if it's supported to use 2015.8.0 clients with a 2015.5.3 server?  I've been waiting to get the bug with the md state fixed for a while, and I want to get first highstates working (finally)
17:24 nocturn joined #salt
17:24 nocturn joined #salt
17:26 culexor joined #salt
17:26 gchao joined #salt
17:26 jodv pcn: generally newer masters are supported with older minions, but not necessarily the other way around
17:28 pcn OK, that makes me sad
17:28 jodv i think the issue you're reporting it probably with pepper...i just did a tested against cherrypy salt-api 2015.8.0 firing off a job and looking up the jid for results with raw http calls and it works as expected
17:29 pcn Could b
17:30 pcn Sorry, network hiccup - jodv could be, but what's changed?
17:30 JDiPierro joined #salt
17:31 otter768 joined #salt
17:34 blergh- joined #salt
17:35 jodv donno off the top of my head, but I dont think the jobs endpoint of the cherrypy api did
17:39 twork i know this isn't enough to solve my problem. what i want to know is, where should i be looking? https://gist.github.com/mjinks/b40e11cc57d6b2f823a7
17:40 protoz joined #salt
17:40 twork oops, correction, let me amend that...
17:41 rim-k joined #salt
17:41 pcn jodv: the jobs get has definitely changed between them, not sure why it would cause this effect yet.
17:43 pcn Looks like the lookup_jid argument type was changed
17:43 danlsgiga how can I use the salt['module.function'] inside an orchestrator and be able to target the system at the same time?
17:44 twork ok. there are two error messages in the "error" panel of that gist. the first one contains actual information (a string that should be filled in from {{ var }}_vsftpd_user_config is instead getting just "l") ...
17:45 pcn Not just the type, but the kwarg key
17:45 twork and, where the error points, is a hunk of code that's worked dozens of times leading up to now.
17:45 pcn jodv what did you pass to test?
17:46 twork i know the answer is "why isn't that field being filled correctly?" but i don't know where to hunt for likely answers why that could be happening.
17:46 cruatta joined #salt
17:46 danlsgiga something like salt['pillar.get']('apache:vhost:test.example.com:certs', 'P@httpd:proxy', expr_form=compound)
17:46 hal58th pcn You should always assume that salt-master needs to be equal or higher version then the salt-clients. Also the issue you opened is lacking the "salt --versions-report" for master and minion
17:47 bhosmer joined #salt
17:48 twork probably should have re-posted that link, to be clear it's been added to: https://gist.github.com/mjinks/b40e11cc57d6b2f823a7
17:49 subsignal joined #salt
17:49 svinota joined #salt
17:50 pcn hal58th: I wish I had it, but that is blocking other things I'm doing, so I reverted and kept the log.  I'll try it again later once I can record a demo to mark my work.
17:50 pcdummy joined #salt
17:50 pcn However, I see this in 2015.5.3 and in develop: https://github.com/saltstack/salt/blob/2015.8/salt/netapi/rest_cherrypy/app.py#L1146
17:50 viq joined #salt
17:51 pcn err backwards, that's 2015.8, and this is in develop and 2015.5: https://github.com/saltstack/salt/blob/develop/salt/netapi/rest_cherrypy/app.py#L1158
17:52 pcn It looks like that difference breaks the runner pepper is calling
17:52 hal58th Sounds good. I just know they will ask you for more info. An example setup would be good too
17:52 pcn hal58th: all true
17:53 bhosmer joined #salt
17:55 Ryan_Lane the pecl module is broken every single release
17:55 Ryan_Lane without fail
17:56 Ryan_Lane someone must be making changes to it, but what I don't understand is that it doesn't work at all
17:56 Ryan_Lane so how are the changes going in? are people just making the code change and never running it?
17:57 Ryan_Lane Executing command "yes '' | pecl -d preferred_state=beta install 'x h p r o f - 0 . 9 . 4'" in directory '/root'
17:57 Ryan_Lane seriously.
17:57 Ryan_Lane wtf
17:59 jodv pcn: ya, looks like it's the underlying RunnerClient that changed https://github.com/saltstack/salt/issues/25107
17:59 jodv anyway, pepper just needs to be updated
17:59 JavaAtom Dumb question -- any good docs on migrating from Chef to Salt?
18:00 baweaver joined #salt
18:02 twork JavaAtom: that question has come up here before, and iirc the answer boiled down to, short version, not really, and longer version, it'd be such a mismatch of the two approaches that such a doc would be tough to write.
18:02 twork ...or, if you come up with one, the world will think you
18:02 bVector_ JavaAtom: http://www.artofmanliness.com/2011/12/04/rip-off-the-band-aid/
18:03 dthom91 joined #salt
18:03 twork <- wrong as usual
18:04 bVector_ everyone that I've talked to that has tried a 'migration' has resorted to just starting over from scratch
18:04 JavaAtom twork: Thanks for the update.
18:04 Ryan_Lane JavaAtom: the easiest approach is to do sequentially ordered saltstack
18:04 iggy there's some stuff on ryan lane's blog too (think that might be more puppet -> salt though)
18:04 JavaAtom bVector_: I totally believe it :( Just needs a lot of work.
18:04 dthom91 joined #salt
18:04 Ryan_Lane since chef is generally sequentially ordered
18:05 pyaaron joined #salt
18:05 JavaAtom Ryan_Lane: Isn't chef sequentially ordered by default as well?
18:06 JavaAtom err
18:06 JavaAtom Salt runs things in sequence, no?
18:07 pyaaron Just a general question, for 'salt.state.ssh_known_hosts.present' does the 'fingerprint' argument only work if you are the repo owner of the Git repo that you are trying to pull from for example?
18:08 An__T__oine joined #salt
18:08 pyaaron I was able to do a Git pull with Salt using the 'key' argument, and specifying the path, but not with the 'fingerprint' argument of that same public ssh key.
18:08 pcn jodv: I'm not clear on what the fix is here.  Pepper is passing in jid.  What is the expected api here?
18:08 Akhter joined #salt
18:09 JavaAtom bVector_: I guess I'll have to start over with salt. :) Not a bad thing for a clean reimplementation
18:09 szhem joined #salt
18:10 bVector_ salt usually runs in sequence, but aggregation functions and other conditionals can mix things up
18:10 denys joined #salt
18:10 bVector_ so its not guaranteed unless you specify order
18:11 JavaAtom bVector_: Sweet. Thanks!
18:12 druonysus joined #salt
18:12 Ryan_Lane JavaAtom: it runs things in order, but it may also reorder your states if you aren't careful
18:13 Ryan_Lane I wouldn't use the order arg :)
18:13 Ryan_Lane http://ryandlane.com/blog/2015/04/02/saltconf15-sequentially-ordered-execution-in-saltstack-talk-and-slides/
18:13 Ryan_Lane (and http://ryandlane.com/blog/2014/07/14/truly-ordered-execution-using-saltstack/ and http://ryandlane.com/blog/2015/01/06/truly-ordered-execution-using-saltstack-part-2/)
18:14 Ryan_Lane the talk/slides are roughly a presentation version of those blog posts with a bit more context
18:15 aqua^c joined #salt
18:19 jodv pcn: not sure off the top of my head, would have to replicate and take a closer look at it
18:20 Ryan_Lane JavaAtom: my recommendation is to just not use any features that reorder, unless you explicitly want that :)
18:20 JavaAtom Ryan_Lane: Gotcha. Thank you.
18:21 baweaver joined #salt
18:23 Ryan_Lane yw
18:25 pcn jodv: No worries, apparently it was supposed to be reverted
18:26 Ryan_Lane basepi: https://github.com/saltstack/salt/pull/27415
18:26 Ryan_Lane blocking us on upgrade
18:26 Ryan_Lane you may get some more pings like this next week too :)
18:31 stupidnic joined #salt
18:32 zmalone joined #salt
18:38 Bryson joined #salt
18:39 PI-Lloyd joined #salt
18:41 DammitJim can I have something like this in a pillar?
18:43 DammitJim https://gist.github.com/anonymous/beeba7bd42ab6e659739
18:43 DammitJim and be able to just pick out road from a state?
18:46 breakingmatter joined #salt
18:48 cpowell_ joined #salt
18:48 Akhter joined #salt
18:48 nate_c joined #salt
18:51 szhem joined #salt
18:53 Bryson joined #salt
18:53 ujjain- joined #salt
18:53 ujjain joined #salt
18:53 ujjain joined #salt
18:54 XenophonF left #salt
19:00 dean joined #salt
19:00 blergh- joined #salt
19:01 Akhter joined #salt
19:06 twork okay, it's becoming clear to me that my present hookup is an error somewhere in my template file. but staring and staring turns up nothing wrong. are there tools that will unravel a jinja template and point to there syntax error?
19:08 Deevolution left #salt
19:08 Deevolution joined #salt
19:09 iggy twork: salt-call -l debug will show the rendered output
19:09 iggy but no, not really
19:09 twork thanks
19:09 twork that reminder will at least help, been a while since i did anything with -l
19:09 iggy it can also be helpful to gist it and let a second (or more) set of eyes take a peek
19:10 twork yeah... i've been gisting an awful lot lately, but yeah
19:11 anotherZero joined #salt
19:14 bVector_ I went down the rabbit hole of trying to find the lines in jinja where it processes statements
19:14 bVector_ just to add some debug statements
19:15 twork i see you're still with us...
19:15 bVector_ I lost a piece of my self down there, and have nothing to show for it...
19:15 twork friend.
19:17 pcdummy joined #salt
19:18 twork okay, concocting a gist. stay tuned. (or don't, those of you who've earned your purple hearts. or gotten tied of twork already.)
19:18 cruatta joined #salt
19:18 twork looking through -l debug didn't teach me anything i didn't already know.
19:20 bluenemo joined #salt
19:21 jdesilet joined #salt
19:22 baweaver joined #salt
19:23 mpanetta Hrm
19:23 mpanetta Ok Either I found an odd bug in jinja templating or I am doing something wrong...
19:23 mpanetta ANyone can help me with an issue?
19:23 felskrone joined #salt
19:24 mpanetta This one is real weird
19:24 mpanetta It's like jinja is ignoring my set statements
19:24 DammitJim how do I update my grains on a minion?
19:24 cruatta_ joined #salt
19:24 DammitJim I just added the minion to the pillar top
19:24 DammitJim and created a new grain for it
19:25 mpanetta grains.setitem or whatever ?
19:25 jartsu joined #salt
19:25 mpanetta any salt devs hanging around?
19:28 _pravka_ joined #salt
19:29 darix is there any command to run on the minion to ask "is this file maintained by salt"?
19:34 markm joined #salt
19:35 baweaver joined #salt
19:37 pcdummy joined #salt
19:37 mpanetta Ok I am pretty sure this is a bug...
19:38 mattiasr joined #salt
19:38 basepi Ryan_Lane: thanks.
19:38 mpanetta basepi: Good afternoon :)
19:39 basepi mpanetta: heya. :)
19:39 mpanetta basepi: I have what I think is a jinja bug
19:39 mpanetta Pretty sure it is, let me gist it.
19:40 twork ...oh, i think i see a clue. posting my gist here anyway because while i see at least one symptom, i know better than to think i'll unravel the thing: https://gist.github.com/mjinks/40701f3bfa41a58dbff6
19:41 danlsgiga DammitJim: grains.setval
19:41 twork the clue i spot is in the panel named "weird: the service salts fine on its own"
19:41 twork ...which i only thought to gist as an afterthought
19:42 mpanetta basepi: https://gist.github.com/bentwire/48ed931bd40b4d961bcf
19:42 mpanetta The jinja if statement is not qorking
19:42 mpanetta working...
19:43 * twork qorks
19:43 mpanetta I've had this in the past where if I replace the set with a do and an arry it works fine, but the sets are silently ignored.
19:43 mpanetta twork: haha
19:43 mpanetta This is 2015.5.3
19:43 mpanetta BTW
19:44 mpanetta Weirdest issue I have ever seen in a piece of software... heh
19:44 mpanetta Anyone else had an issue similar to this?
19:45 mpanetta Is it possible to have a jinja template that is too complicated to run? :P
19:45 twork ..or one that isn't
19:45 basepi mpanetta: Nah, I've seen much more complicated jinja templates
19:45 basepi See some of the salt formulas....
19:46 basepi Nothing appears to be wrong in your linked template....
19:46 s_kunk joined #salt
19:46 basepi Are you certain it's the if statement that is borking it? Have you tried removing the if statement, for example?
19:47 ajw0100 joined #salt
19:48 twork oh, on my own gist: just realized i didn't post any pillar. moment.
19:48 zmalone joined #salt
19:49 pcdummy joined #salt
19:51 twork ok, there's a snippet of pillar there, if it helps at all. https://gist.github.com/mjinks/40701f3bfa41a58dbff6
19:51 snarfy joined #salt
19:52 twork ...and i'm going afk for just a minute, will scroll back if anybody has anything to say
19:52 twork and thanks
19:54 GreatSnoopy joined #salt
19:57 rhodgin joined #salt
19:58 trph joined #salt
19:59 rhodgin joined #salt
20:01 neogenix joined #salt
20:01 pratikmallya joined #salt
20:03 Akhter joined #salt
20:05 pcdummy joined #salt
20:05 pcdummy joined #salt
20:08 twork a remark on my present troubles: notice the error there about "...global name 'l_vsftpd_user_config' is not defined". earlier, the templat that (i think) it's griping about was "{{ name }}_vsftpd_user_conf". that bare "l" had me puzzled so i changed the template to "vsftpd_user_conf_{{ name }}"... but i still get the same error. so i have no idea where that's coming from. and yes i've pillar i[
20:08 twork ..ahem. and yes i'v up'ed my pillar.
20:08 twork at least three times now.
20:08 blergh- joined #salt
20:14 pcdummy joined #salt
20:16 bhosmer joined #salt
20:18 mpanetta basepi: I have not removed the if statement yet, but if I have defaults set at the top of the file it ignores those too
20:19 mpanetta It is really really weird
20:20 Ryan_Lane basepi: woot
20:20 Ryan_Lane thanks
20:20 zerthimon joined #salt
20:21 snarfy joined #salt
20:21 KyleG joined #salt
20:21 KyleG joined #salt
20:22 twork another non-surprise: find+grep through my whole tree of salt and pillar finds no instance of the string 'l_vsftp' anywhere.
20:22 mpanetta Is there any way to debug the templating engine?  Like to see if it is acually running the set statements?
20:23 mpanetta basepi: ^
20:23 twork ...likewise any lingering '}_vsftp'
20:25 mpanetta GRR this is frustrating...
20:26 mpanetta Ok, I'm a dumbass...
20:27 twork try me.
20:27 mpanetta I was looking in the wrong place in the pillar for the value...
20:28 mattiasr joined #salt
20:29 mimianndi joined #salt
20:30 Bryson joined #salt
20:30 mimianndi anyone here running high number of worker_threads?
20:30 mimianndi and saw any issues ?
20:32 riftman joined #salt
20:33 Ahlee I run 50 here, across 8 CPUs
20:34 mimianndi Ahlee: how many events are you able to send at one time?
20:34 Ahlee mimianndi: I don't know, i've never bench marked that. I have ~1200 minions on that master
20:35 mimianndi Ahlee: do you use publish ?
20:35 mimianndi or use api or just command line from the salt master?
20:35 Ahlee mimianndi: I do not. Most of my interaction with saltstack is through salt-api
20:36 DammitJim why is this a problem? https://gist.github.com/anonymous/02f69ce318dfff0c1128
20:37 mimianndi Ahlee: i see, just out of curiousity how many api calls are you ble to make ?
20:37 DammitJim Missing end of comment tag?
20:37 teebes joined #salt
20:37 mimianndi Ahlee: is salt-api able to handle more than 50 at one time ?
20:37 felskrone joined #salt
20:38 Ahlee mimianndi: I don't know how to benchmark those items, nor do I know a way to land 50+ requests at once
20:38 Ahlee mimianndi: i've never tried apache bench or similar against hte API
20:39 Ahlee that sounds like a fun little project, though
20:39 mimianndi Ahlee: ah i see.  I see that publish module performance is highly dependent on the number of mworkers
20:39 mimianndi Ahlee: just curiout is api has same issue or not
20:39 mimianndi Ahlee: thanks for the info
20:40 Ahlee sorry I couldn't provide more useful information.
20:40 mimianndi no worries man
20:40 seatan joined #salt
20:42 mpanetta Ok, mine in pillar is rocking...
20:45 whytewolf mpanetta: you got it working? you should write a blog post on the steps. and share the magic
20:45 mpanetta whytewolf: I manually patched my master with the patch that is supposed to go in to 2015.5.6 heh
20:46 mpanetta Or do you mean write a blog post specifically showing the use of mine in pillar?
20:46 mpanetta As opposed to in states.
20:46 whytewolf well since technically that patch is in release i would say write the post for mine in pillar
20:47 felskrone joined #salt
20:47 mpanetta whytewolf: Oh? 2015.5.6 is out?
20:47 whytewolf 2015.8 is out
20:47 mpanetta whytewolf: Oh damn
20:47 whytewolf and that patch is in it
20:47 mpanetta Yeah I saw that
20:47 mpanetta didn't think it was out yet
20:48 whytewolf if you are on the ppa that might be why. since it looks like the ppa isn't really being supported anymore. repo.saltstack.com has the ubuntu pacakges now [2015.8 anyway]
20:50 whytewolf the docs don't seem to have been updated to reflect that ubuntu packages are on the salt repo now either :/
20:50 zmalone A coworker tried to convert from the PPA package to repo.saltstack.com and says his local apt repository went off the rails as a result, so it's probably worth playing with a little before converting anything important
20:50 zmalone (I have not tried it yet)
20:50 mpanetta whytewolf: :(
20:50 mpanetta Yeah we are using the ppa I think.
20:50 mpanetta Actually we are using the salt formula to create our master, so however it gets the package is how we do it.
20:51 whytewolf zmalone: any time you mess repos it tends to go heywire. :/
20:51 giantlock joined #salt
20:51 whytewolf humm. I don't think the salt formula has been updated with the new repo
20:51 whytewolf I can't be sure though
20:52 zmalone The bootstrap script probably hasn't either.
20:52 mpanetta whytewolf: Here is the magic: https://gist.github.com/bentwire/e6a972e1df257c16a17e :)
20:52 subsignal joined #salt
20:52 whytewolf since the packages were added earlyer this week i doubt anything really has been updated to reflect the changes
20:53 mpanetta It needs a bit of cleanup, but it works :)
20:53 whytewolf huh, so the saltutil.runner works like a charm with that patch? wonder if it will start working in orch also than
20:53 mpanetta like I used grains.name where I should have used grains.id...
20:53 mpanetta whytewolf: Donno.  Have not tried that yet :)
20:54 ts_ joined #salt
20:55 dthom91 joined #salt
20:56 danlsgiga joined #salt
20:57 nate_c joined #salt
20:57 pentabular joined #salt
20:57 aqua^c joined #salt
20:57 danlsgiga guys... I'm looking for quite some time to find a way to get results for grains.get or pillar.get inside an orchestrate state
20:58 DammitJim how does one copy a directory with subfolders and files?
20:58 danlsgiga I know that salt['pillar.get'] doesn't work because I can't target the machines
20:58 DammitJim I am extracting an archive
20:58 DammitJim then renaming it and configuring it
20:58 DammitJim I have to do this 4 times for different instances
20:59 DammitJim but the archive.extracted only runs once while going through the
20:59 DammitJim oh, hold on
21:02 felskrone joined #salt
21:07 markm joined #salt
21:10 derelm joined #salt
21:13 rdxc joined #salt
21:18 baweaver joined #salt
21:24 seatan joined #salt
21:27 larsfronius joined #salt
21:43 twork i just hit a data point, maybe, on the weird failure i'm chasing. the trouble (as above) is a minion with about 16 states that fail with the same non-helpful, non-truthful error. i've just learned that if i call for those states individually, they claim to run without error, but they lie.
21:43 hal58th DamnitJim, you want to copy contents on the salt-minion to somewhere else on the salt-mininon?
21:44 twork if i change the fieles, they don't get touched. if i remove the files, they do get touched, but no contents get filled in.
21:45 twork a repeat, but now edited with a little more detail of what i just described: https://gist.github.com/mjinks/40701f3bfa41a58dbff6
21:50 twork right now the closest thing i have to a clue is (still) that business of the service.running module failing, with a long requisites line made up of a lot of wrong path names that should point to the config files i'm trying to manage.
21:54 twork and with that... looks like i turn into a pumpkin for a while. what a week. thanks for all the help, channel.
21:56 trph joined #salt
21:58 baweaver joined #salt
22:00 murrdoc joined #salt
22:00 murrdoc [WARNING ] Returner unavailable: Error -2 connecting to salt:6379. Name or service not known.
22:04 Bryson joined #salt
22:04 mpanetta joined #salt
22:06 mpanetta_ joined #salt
22:06 pedja Is anyone else using vagrant-1.4.7/Virtualbox-5.0.4 for salt-vagrant-demo?
22:07 pedja master and minions can't see each other, so I was wondering if it is just on my system
22:09 bfoxwell joined #salt
22:10 baweaver joined #salt
22:15 dthom91 joined #salt
22:19 baweaver joined #salt
22:23 iggy murrdoc: you need to set the config for the redis returner
22:23 murrdoc i had it set
22:24 murrdoc the file was named .confi instead of conf
22:24 iggy it doesn't default to config.get(master) for some reason
22:24 iggy lol
22:24 murrdoc rookie mistake
22:29 giantlock joined #salt
22:31 Grokzen joined #salt
22:31 eitzei joined #salt
22:31 dthom91 joined #salt
22:32 Garo_ joined #salt
22:33 snarfy joined #salt
22:34 protoz joined #salt
22:36 mpanetta joined #salt
22:36 protoz joined #salt
22:39 pravka joined #salt
22:41 TheoSLC joined #salt
22:42 mpanetta_ joined #salt
22:44 kevinquinnyp joined #salt
22:47 kevinquinnyp hey all, I am wanting to pull down a git repo in a state, but i only need one file from it -- https://github.com/varnish/varnish-devicedetect/blob/master/devicedetect.vcl
22:48 mvensky joined #salt
22:48 protoz joined #salt
22:48 kevinquinnyp is there a best way to do this with salt's git module, or should i just have the state clone it to a specific directory on the minion, then copy or symlink it where i need it?
22:48 Bryson joined #salt
22:49 kevinquinnyp or should i just clone it on the master, and use it as a template, updating it myself as I need?
22:50 protoz joined #salt
22:55 Aleks3Y Hey guys, is there any cool quick start tutorials or videos for salt?
23:03 pedja Aleks3Y: https://docs.saltstack.com/en/getstarted/index.html
23:04 Rebus joined #salt
23:06 Aleks3Y pedja: yeah, went over it
23:08 eitzei joined #salt
23:08 druonysuse joined #salt
23:08 murrdoc i can sell u some
23:08 murrdoc for tree fiddy
23:09 murrdoc you got that tree fiddy ?
23:09 trave joined #salt
23:09 whytewolf damn you loch ness
23:10 murrdoc :D
23:12 SunPowered joined #salt
23:15 baweaver joined #salt
23:23 seatan joined #salt
23:23 solidsna_ joined #salt
23:24 solidsna_ left #salt
23:25 solidsna_ joined #salt
23:30 pravka joined #salt
23:32 solidsna_ joined #salt
23:34 otter768 joined #salt
23:36 khaije1 anyone aware of interoperability between Salt and Hashicorp's Vault? Looks to make an awesome pillar source.
23:37 ajw0100 joined #salt
23:38 aurynn khaije1, I'm looking at writing an integration, but I'm not aware of one
23:38 iggy a number of us have thrown the idea around... nobody has done the work
23:40 khaije1 aurynn: iggy: Cool! It's nice to know that I'm not the only one who sees potential there.
23:41 sjorge joined #salt
23:41 sjorge joined #salt
23:43 pentabular joined #salt
23:47 sjorge joined #salt
23:47 sjorge joined #salt
23:47 _JZ_ joined #salt
23:52 fivehole joined #salt
23:53 armguy joined #salt
23:55 fivehole Is there something special that needs to be done after installing python-gnupg?  I did the install through yum but salt --versions-report still shows python-gnupg: Not Installed

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