Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2017-11-02

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

All times shown according to UTC.

Time Nick Message
00:01 alexlist joined #salt
00:04 pipps joined #salt
00:12 Hybrid joined #salt
00:19 johnj_ joined #salt
00:20 Hybrid joined #salt
00:42 dxiri joined #salt
00:49 kellyp joined #salt
01:09 heaje joined #salt
01:15 dxiri joined #salt
01:20 johnj_ joined #salt
01:28 aanriot joined #salt
01:32 kellyp joined #salt
01:32 NV joined #salt
01:59 dxiri joined #salt
01:59 dxiri joined #salt
02:01 kukacz joined #salt
02:07 nomeed joined #salt
02:22 johnj_ joined #salt
02:24 inetpro joined #salt
02:27 tiwula joined #salt
02:31 bstevenson joined #salt
02:32 bryan joined #salt
02:39 evle2 joined #salt
02:56 ilbot3 joined #salt
02:56 Topic for #salt is now Welcome to #salt! <+> Latest Versions: 2016.11.8, 2017.7.2 <+> Support: https://www.saltstack.com/support/ <+> Logs: http://irclog.perlgeek.de/salt/ <+> Paste: https://gist.github.com/ <+> See also: #salt-devel, #salt-offtopic <+> We are volunteers and may not have immediate answers
02:57 Lenz joined #salt
03:22 bryan joined #salt
03:23 johnj_ joined #salt
03:23 icebal joined #salt
03:28 aviau joined #salt
03:32 bryan joined #salt
03:33 bstevenson joined #salt
03:44 thekevjames joined #salt
03:50 Praematura joined #salt
04:01 user-and-abuser joined #salt
04:03 pipps joined #salt
04:06 onlyanegg joined #salt
04:13 justan0theruser joined #salt
04:13 justan0theruser joined #salt
04:13 zerocoolback joined #salt
04:16 SkyRocknRoll joined #salt
04:20 thekevjames joined #salt
04:24 johnj_ joined #salt
04:29 pipps joined #salt
04:35 thekevjames joined #salt
04:39 ahrs joined #salt
04:47 omie888777 joined #salt
04:53 onlyanegg joined #salt
05:16 dxiri joined #salt
05:22 icebal joined #salt
05:24 LocaMocha joined #salt
05:25 johnj_ joined #salt
05:31 maestropandy joined #salt
05:33 kellyp joined #salt
05:38 wongster80 joined #salt
05:44 kellyp joined #salt
06:03 maestropandy joined #salt
06:11 impi joined #salt
06:26 johnj_ joined #salt
07:07 netcho joined #salt
07:20 hoonetorg joined #salt
07:21 jas02 joined #salt
07:23 maestropandy joined #salt
07:25 Ricardo1000 joined #salt
07:26 johnj joined #salt
07:27 maestropandy1 joined #salt
07:28 do3meli joined #salt
07:30 do3meli left #salt
07:32 omie888777 joined #salt
07:35 netcho joined #salt
07:35 netcho joined #salt
07:49 fracklen joined #salt
07:50 pualj joined #salt
07:51 kellyp joined #salt
07:57 fracklen_ joined #salt
07:58 maestropandy1 left #salt
08:07 omie888777 joined #salt
08:07 uptime joined #salt
08:08 felskrone joined #salt
08:10 obitech joined #salt
08:11 rathier joined #salt
08:11 gmoro joined #salt
08:12 aldevar joined #salt
08:14 oyvindmo joined #salt
08:15 uptime joined #salt
08:16 Hybrid joined #salt
08:22 fracklen joined #salt
08:28 johnj joined #salt
08:29 onlyanegg joined #salt
08:34 pbandark joined #salt
08:40 jrenner joined #salt
08:41 mbologna joined #salt
08:46 mikecmpbll joined #salt
08:46 obitech good morning my dudes
08:51 aldevar joined #salt
08:56 mikecmpbll joined #salt
09:02 Praematura joined #salt
09:02 jhauser joined #salt
09:10 fracklen_ joined #salt
09:16 impi joined #salt
09:29 johnj joined #salt
09:38 maestropandy joined #salt
09:44 Ricardo1000 joined #salt
09:54 thekevjames joined #salt
09:56 Naresh joined #salt
09:56 SkyRocknRoll joined #salt
10:00 maestropandy joined #salt
10:01 astephanh joined #salt
10:01 _KaszpiR_ joined #salt
10:08 fnords joined #salt
10:16 BarBQ joined #salt
10:19 araifr joined #salt
10:22 schasi joined #salt
10:22 fracklen joined #salt
10:30 johnj joined #salt
10:33 fracklen joined #salt
10:36 N-Mi joined #salt
10:36 N-Mi joined #salt
10:45 dRiN joined #salt
10:50 yidhra joined #salt
10:54 stooj joined #salt
10:56 fracklen joined #salt
11:01 obitech Anyone had this issue where the state which comes last in your top file gets executed at the very first? There is no depency associated with that state. It makes no sense to me
11:08 fracklen_ joined #salt
11:31 johnj joined #salt
11:44 Praematura joined #salt
11:44 aldevar joined #salt
11:51 kellyp joined #salt
12:01 absolutejam Should I have to restart master when I add a new reactor trigger?
12:06 coredumb absolutejam: probably yes
12:08 BarBQ joined #salt
12:09 absolutejam alright, next one
12:10 absolutejam I'm trying ot acc
12:10 absolutejam to access data from the event bus, specifically from the docker_events engine
12:10 absolutejam I know I can use the `data` dict
12:11 absolutejam but I'm trying to access the `Actor` key but it's complaining at me that there's no attribuet 'Actor'
12:12 Nahual joined #salt
12:12 Antiarc joined #salt
12:15 absolutejam https://hastebin.com/ikoduhotes.scala
12:15 absolutejam Is an example of the data going through the even bus
12:16 aldevar joined #salt
12:20 om2 joined #salt
12:20 XenophonF obitech: "no dependency associated with that state" is perhaps why it gets executed first
12:21 XenophonF you can use one of the state.show_* functions to see what things compile to
12:21 obitech no actually I found the mistake, I defined a state was defined twice in my top file
12:21 obitech -was
12:21 XenophonF oh ok cool
12:21 obitech thanks tho :)
12:28 kellyp joined #salt
12:31 Miouge joined #salt
12:32 johnj joined #salt
12:33 fracklen joined #salt
12:35 XenophonF anyone have a working example of an Office 2016 deployment
12:36 XenophonF or Cygwin
12:36 XenophonF or Visual Studio?
12:36 maestropandy joined #salt
12:38 hammer065 joined #salt
12:39 Mogget joined #salt
12:41 fracklen joined #salt
12:43 mchlumsky joined #salt
12:45 ecdhe joined #salt
12:48 netcho joined #salt
12:48 netcho joined #salt
12:48 netcho joined #salt
13:09 gh34 joined #salt
13:11 fracklen joined #salt
13:11 fracklen joined #salt
13:26 edrocks joined #salt
13:32 onlyanegg joined #salt
13:33 johnj joined #salt
13:45 racooper joined #salt
13:45 numkem joined #salt
13:50 ThomasJ|d joined #salt
13:55 thekevjames joined #salt
13:59 fracklen joined #salt
14:02 SkyRocknRoll joined #salt
14:13 user-and-abuser joined #salt
14:14 fracklen_ joined #salt
14:17 ouemt joined #salt
14:18 niluje joined #salt
14:19 niluje From a pillar using the python renderer, is it possible to dynamically load another pillar file?
14:19 cgiroua joined #salt
14:19 XenophonF you probably want pillarstack
14:20 XenophonF i don't use it and can't help you set it up, because i don't reference pillar data when rendering pillar data
14:20 numkem joined #salt
14:21 niluje I don't think I need pillarstack
14:22 niluje I'm trying to create a pillar contaning a dict "users", where each user in it is actually contained in a separate file
14:22 aldevar joined #salt
14:22 niluje this way I can reference either the "group" file or a single user from top.sls
14:24 XenophonF if i were going to set something like that up without using pillarstack, i would put the data into Jinja and import that into the desired SLS files, using e.g., the |yaml filter
14:25 XenophonF alternatively, Salt will merge Pillar SLS data for you
14:25 XenophonF so you could for example include users.xenophonf and users.niluje, and it will DTRT assuming the users dict has xenophonf and niluje keys
14:26 XenophonF include == list both Pillar SLSes in top.sls
14:27 XenophonF but maybe the include keyword works?  I dunno.  would have to RTFM.
14:27 niluje https://pastebin.com/5z75YF56
14:27 niluje [CRITICAL] Rendering SLS 'users/groups/mygruop' failed, render error:
14:27 niluje expected '<document start>', but found '{'
14:27 niluje in "<string>", line 8, column 5:
14:27 niluje {'users': {'xxx': {'fulln ...
14:27 niluje I tried to, something is wrong, I can't figure out what
14:28 XenophonF that seems far more complicated than what i have in mind
14:28 niluje oh?
14:29 XenophonF I mean literally, users/xenophonf.sls has a dictionary like `{ 'users': { 'xenophonf': { ...PUT USER INFO HERE... } } }`
14:29 XenophonF and users/niluje.sls has something similar
14:29 XenophonF and in top.sls you apply one or the other or both of users.xenophonf and users.niluje
14:30 XenophonF e.g., here's my default local account list that gets pushed to all minions - https://github.com/irtnog/salt-pillar-example/blob/master/defaults/accounts.sls
14:30 XenophonF (I'm using users-formula BTW.)
14:31 XenophonF and here's my Salt Master's local account list - https://github.com/irtnog/salt-pillar-example/blob/master/salt/example/com/init.sls#L18
14:31 XenophonF the merging happens in top.sls
14:31 XenophonF oh and you can use the `include` directive in Pillar SLS files
14:31 XenophonF https://github.com/irtnog/salt-pillar-example/blob/master/top.sls#L34
14:31 XenophonF that pulls in the defaults
14:32 XenophonF https://github.com/irtnog/salt-pillar-example/blob/master/top.sls#L91
14:32 XenophonF that assigns the Salt Master-specific configs, including user accounts
14:32 XenophonF stuff gets merged/overridden per defaults (or at least I don't remember changing the pillar merge strategy)
14:34 johnj joined #salt
14:34 niluje hm
14:34 niluje thanks
14:34 niluje I wanted to avoid duplication in the top.sls though
14:35 XenophonF if you're using Salt as a directory service, I recommend against it.
14:35 XenophonF Set up LDAP or something.
14:35 niluje I do have an LDAP
14:35 niluje I'm migrating some existing services
14:35 XenophonF gotcha
14:35 XenophonF the above is just how I do things
14:36 XenophonF it isn't necessarily Gospel Truth, you dig?
14:36 fracklen joined #salt
14:41 mikecmpbll is there a way to have a requisite on a state _file_?
14:41 mikecmpbll if i have a bunch of dynamically generated states in a .sls file, and i want another state to depend on those
14:53 fracklen joined #salt
14:53 niluje actually I can just include: - userA - userB
14:53 niluje I thought it only worked in states oO
14:53 niluje thanks for your time XenophonF
14:55 jas02 joined #salt
14:58 XenophonF mikecmpbll: I strongly recommend against it, but you can do something like this in a requisite: `sls: something.something.something.something.dark-side`
14:58 XenophonF https://docs.saltstack.com/en/latest/ref/states/requisites.html#require-an-entire-sls-file
14:59 tiwula joined #salt
14:59 Praematura joined #salt
15:00 mikecmpbll derp, how did i miss that. thanks XenophonF . for what it's worth, why recommended against? :)
15:01 bstevenson joined #salt
15:01 scbunn joined #salt
15:02 user-and-abuser joined #salt
15:04 XenophonF this is just my opinion, but i generally avoid creating requisites that reference stuff outside of the SLS file in which they appear
15:04 XenophonF I basically treat every top-level SLS like it's an independent formula
15:04 XenophonF so for example, let's say I'm deploying Shibboleth IdP
15:05 XenophonF I could add requisites on `service: apache` and `service: tomcat` states found in other SLS files, but that feels like spaghetti code to me
15:05 XenophonF plus it makes it difficult to just run the Shibboleth IdP SLS
15:05 XenophonF or to port it to a different container, e.g., Jetty
15:05 XenophonF &c
15:06 XenophonF so what I do in order to make composability more straightforward is to write the top-level SLSes independent of one another
15:06 XenophonF so e.g., there's the various tomcat SLSes from tomcat-formula, and there's the shibboleth.idp SLS from my shibboleth-formula
15:07 XenophonF and mine don't reference any of the tomcat SLSes or state IDs
15:07 XenophonF so you can call shibboleth.idp independently
15:07 XenophonF now
15:07 XenophonF one still wants to glue those things together
15:07 * mikecmpbll is following
15:07 XenophonF e.g., so a IdP config update restarts the container
15:08 XenophonF j2ee == just shoot me now
15:08 XenophonF i digress
15:08 mikecmpbll :D
15:08 XenophonF so I isolate the glue in a separate SLS
15:08 XenophonF that one SLS---let's call it `tomcat.shibboleth-idp`---includes tomcat, tomcat.conf, and shibboleth.idp
15:09 XenophonF and it mutates the appropriate states (via extends) to add the necessary dependencies and requsites and, in one case, to install an additional thing needed to make Shibboleth IdP work with Tomcat
15:09 XenophonF so to put concrete examples to this thing, here are the formulas in question:
15:09 XenophonF https://github.com/saltstack-formulas/tomcat-formula
15:09 XenophonF https://github.com/irtnog/shibboleth-formula
15:10 XenophonF and here's the glue SLS:
15:10 XenophonF https://github.com/irtnog/salt-states/tree/production/tomcat/shibboleth-idp.sls
15:10 XenophonF now in top.sls, I'll just target the `tomcat.shibboleth-idp` SLS at the appropriate minions
15:11 XenophonF but at the command line, this setup gives me a lot of flexibility
15:11 XenophonF I can run `tomcat,tomcat.conf` separate from `shibboleth.idp` if I want/need to
15:11 XenophonF I can rip tomcat out of the equation completely and write a formula for Jetty
15:11 XenophonF and then integrate Shibboleth IdP with Jetty instead
15:12 XenophonF for me, this approach makes it easier for me to write and compose (combine) independent Salt State formulas
15:13 XenophonF shibboleth-formula works, out of the box and on its own
15:13 XenophonF it doesn't quite get you to a working IdP---you still need a container for that
15:13 Heartsbane joined #salt
15:14 ParsectiX joined #salt
15:15 mikecmpbll v/interesting, thanks! always interesting to learn how people structure complicated stuff—that's the hardest thing with salt.
15:15 ParsectiX hey guys is there a way to test mine functions locally? without creating a saltmaster
15:15 XenophonF it's a hard thing in general
15:15 XenophonF in my mind I take this OO/component-like approach to Salt States
15:16 XenophonF so things crossing boundaries (like requisites on stuff outside the .sls file) make me twitch
15:16 Brew joined #salt
15:23 threwahway joined #salt
15:25 sp0097 joined #salt
15:30 choke joined #salt
15:31 fracklen_ joined #salt
15:33 relidy joined #salt
15:35 johnj joined #salt
15:42 edrocks joined #salt
15:55 thekevjames joined #salt
16:01 dRiN joined #salt
16:06 kellyp joined #salt
16:17 whytewolf So quiet. Everyone at saltconf
16:22 XenophonF must be pretty good for the usual suspects not to be goofing off on irc
16:22 XenophonF well, that or the hotel has crappy internet ;)
16:23 babilen RedHat hired assassins to take out Ansible's competition?
16:23 jas02 joined #salt
16:36 johnj joined #salt
16:37 xMopxShell joined #salt
16:37 _JZ_ joined #salt
16:50 whytewolf Well I haven't seen anyone die yet. So no assasdins
16:51 coredumb Yeah that's good assassins then :P
16:52 user-and-abuser joined #salt
16:53 astronouth7303 quiet day. I guess today's a good day to run system updates
17:01 aubs joined #salt
17:05 edrocks joined #salt
17:08 charims1 joined #salt
17:08 threwahway joined #salt
17:11 pipps joined #salt
17:11 pipps joined #salt
17:16 jas02 joined #salt
17:20 jas02 joined #salt
17:23 aldevar joined #salt
17:28 Praematura joined #salt
17:28 toofer joined #salt
17:31 Edgan probably everyone is at SaltConf
17:36 thekevjames joined #salt
17:36 astronouth7303 all the users, too?
17:36 lordcirth_work Hope it's worth the ticket price
17:37 johnj_ joined #salt
17:37 toofer joined #salt
17:40 toofer joined #salt
17:42 philp joined #salt
17:43 ipmb joined #salt
17:47 Lionel_Debroux joined #salt
17:56 schasi joined #salt
17:57 ipmb trying to figure out a strange error with salt-cloud on Ubuntu 16.04 https://dpaste.de/W1ti
17:57 ipmb It dumps a traceback about requests not being able to import something from chardet. Both are installed via apt system packages.
17:58 wongster80 joined #salt
17:58 astronouth7303 looks like chardet didn't install correctly.
18:01 astronouth7303 it's trying to import one of it's own internal modules. Reinstall and if that doesn't clear it up, talk to your distro.
18:01 astronouth7303 (https://github.com/chardet/chardet/blob/2.3.0/chardet/__init__.py#L27 is the line in question)
18:01 ipmb I've uninstalled (verified it no longer exists in `/usr/lib/python2.7/dist-packages/chardet/`) and reinstalled and the problem still exists
18:02 ipmb this is Ubuntu 16.04
18:05 Edgan astronouth7303: users go the most. In my experience most of the people at a conference are newbies trying to learn.
18:05 ipmb this is interesting... just using the system python it imports from a different path https://dpaste.de/fDVS
18:06 astronouth7303 did you install with pip as well?
18:06 stewgoin joined #salt
18:06 toofer joined #salt
18:07 pbandark joined #salt
18:07 ipmb not that I recall, but looking into that
18:10 ipmb ok, pip uninstall, apt uninstall, apt install and now I get the right path in a python shell, but still see the error in Salt https://dpaste.de/Th1z
18:11 astronouth7303 ok, that's weird
18:13 Edgan ipmb: how is that an error?
18:13 ipmb https://dpaste.de/W1ti
18:14 Edgan ipmb: what version of chardet?
18:14 Edgan ImportError: cannot import name universaldetector
18:14 Edgan That looks like at least one issue
18:15 ipmb 2.3.0
18:15 ipmb imports work fine when I run python manually
18:15 Edgan ipmb: what about the module above?
18:16 astronouth7303 Edgan: fyi, universaldetector is part of chardet, not an external dependency
18:16 Edgan ipmb: For some reason it isn't finding, /usr/lib/python2.7/site-packages/chardet/universaldetector.py
18:16 Edgan astronouth7303: yes
18:17 ipmb https://dpaste.de/jwzU
18:18 Edgan ipmb: dpkg -L python-chardet | grep __init__.py
18:19 ipmb /usr/lib/python2.7/dist-packages/chardet/__init__.py
18:19 Edgan ls -al /usr/lib/python2.7/dist-packages/chardet/universaldetector.py
18:20 ipmb -rw-r--r-- 1 root root 6840 Sep 17  2014 /usr/lib/python2.7/dist-packages/chardet/universaldetector.py
18:22 Edgan ipmb: python -c "import sys; print('\n'.join(sys.path))"
18:23 ipmb https://dpaste.de/73AJ
18:26 Edgan salt -V
18:26 nixjdm joined #salt
18:27 ipmb this also works fine, https://dpaste.de/qLGw
18:27 ipmb salt -V > https://dpaste.de/W3W7
18:30 Edgan ipmb: updatedb ; locate chardet
18:32 Edgan ipmb: I think I figured it out.
18:33 Edgan ipmb: You somehow have salt running with 2.7.12, but the Xenial version of Salt is 2.7.11
18:33 Edgan ipmb: You somehow have salt running with 2.7.12, but the Xenial version of Python is 2.7.11
18:36 Edgan ipmb: Though your dpastes show 2.7.12 too
18:36 ipmb https://dpaste.de/GTNe
18:36 Edgan ipmb: So how did python get overridden?
18:36 Edgan copy in /usr/local?
18:37 Edgan ipmb: Hmm, I googled it and couldn't find a 2.7.12 mentioned as an update, but ok
18:37 ipmb https://packages.ubuntu.com/xenial/python2.7 2.7.12 comes from ubuntu
18:37 Edgan google fail
18:38 Edgan ipmb: Is this happening on more than one system?
18:38 johnj_ joined #salt
18:38 ipmb results of locate -> https://dpaste.de/COch
18:38 ipmb this is the only system I have with this setup
18:39 Praematura joined #salt
18:40 Edgan ipmb: VM?
18:40 threwahway joined #salt
18:40 ipmb It's on EC2
18:41 Edgan ipmb: What does pip | grep chardet say?
18:41 ipmb pip freeze?
18:42 ipmb chardet==2.3.0
18:42 Edgan ipmb: I know the latest version of salt works on up to date xenial. So it seems like this system got corrupted somehow.
18:42 Edgan ipmb: even in EC2
18:43 Edgan ipmb: echo $PYTHONPATH
18:43 ipmb empty
18:45 Edgan ipmb: I would make a new xenial instance and install salt. See if you get the same outcome. I bet you don't.
18:46 ChubYann joined #salt
18:46 ipmb considering this is a salt master in production, that's easier said than done
18:47 mikecmpbll joined #salt
18:48 Edgan ipmb: No, I mean as a test
18:48 Edgan ipmb: Not literally replace it
18:48 Edgan ipmb: Prove my point that it is a corrupt system
18:48 ipmb corrupt just means "we dont' know what the problem is yet" :)
18:49 Edgan ipmb: yes
18:49 Edgan ipmb: If you don't fear it too much, mv /usr/local/lib/python2.7 /usr/local/lib/python2.7.bak and try salt again
18:50 phileus0 joined #salt
18:51 ipmb same error
18:51 ipmb but interesting, it spit this out in the beginning: [WARNING ] Missing dependency: 'requests'. The ec2 driver requires 'requests' to be installed.
18:51 Edgan ipmb: I agree, but I am more trying to prove my assumption that latest salt and latest xenial work together
18:51 Edgan Oh
18:51 ipmb I believe that
18:51 Edgan This is boto
18:51 ipmb in this specific instance it doesn't and that's what I care about :)
18:51 Edgan ipmb: How are boto/boto3 installed?
18:52 ipmb pip, probably
18:52 Edgan That is probably your problem
18:52 Edgan chardet is also used by requests
18:52 Edgan and it likes it's own internal version
18:52 edrocks joined #salt
18:52 ipmb well, I moved the whole local/lib directory, so they aren't tecnically installed anymore
18:52 Edgan But I think your boto is using it's own requests which is then using the system chardet
18:54 Edgan AWS likes to internalize python modules in their python modules
18:54 Edgan I really hate it
18:54 Edgan It is even worse than your datadog and scrambler having their own independent versions of python and modules
18:55 ipmb those are in isolated virtualenvs
18:55 Edgan ipmb: yes, that is a side comment
18:55 Edgan ipmb: try putting /usr/local back
18:56 Edgan ipmb: then uninstall boto and boto3, and then apt-get install them
18:56 Eugene joined #salt
18:56 pipps joined #salt
18:56 Edgan ipmb: Either salt needs to virtualenv itself, or people need to stop pip installing at the system level random stuff
18:57 Eugene joined #salt
18:57 Edgan ipmb: My personal solution is I use fpm to package up the versions of python modules I want. It just pip installs it and makes a deb/rpm for me. But then the deb/rpm are trivial to uninstall cleanly.
18:59 Edgan salt also doesn't help itself in that so many python modules are optional
19:00 Edgan Then there is cherrypy where the apt-get install version is the wrong version for salt
19:00 ipmb same error with local/lib back and boto removed
19:00 Edgan ipmb: boto and boto3?
19:00 ipmb botocore and boto3 (no boto)
19:00 Edgan ipmb: What about requests vs apt-get install requests?
19:01 ipmb it was pip removed and apt installed
19:02 astronouth7303 tbh, i pip install everything for salt, but i also don't have anything else using system python
19:02 Edgan ipmb: then why did we get the requests error when you moved /usr/local?
19:03 ipmb I think Salt's python paths are messed up or something
19:03 Edgan astronouth7303: using pip is a good way to get random results, you aren't version locking. You do the same pip install sequence on two different days and you are likely to get different results
19:03 ipmb everything works as expected when I run `python` from the cli
19:03 Edgan ipmb: yes, but why would it only not work on this system?
19:04 Edgan ipmb: dpkg -l | grep -i salt
19:05 Edgan ipmb: you didn't pip install salt ?
19:05 ipmb I don't
19:06 ipmb they are all listed there, 2017.7.2+ds-1
19:08 noobiedubie joined #salt
19:09 Edgan ipmb: one advantage to testing then on another instance is you are working with salt-cloud, where as most people, myself included, are using salt master/minion
19:14 pipps joined #salt
19:14 ipmb this is interesting... "File "/usr/lib/python2.7/dist-packages/salt/utils/cloud.py", line 2770, in update_bootstrap"
19:15 ipmb it's line 2820 in the file I'm looking at :/
19:17 pipps joined #salt
19:18 sjorge joined #salt
19:20 nomeed joined #salt
19:20 Edgan ipmb: to answer your path question, https://paste.fedoraproject.org/paste/Ws8AvTIjlO9wAlucUYUZ4A
19:20 Edgan ipmb: then run salt -V
19:20 astronouth7303 Edgan: is that different from any other unpinned package management?
19:21 ipmb ok, this is interesting...
19:21 Edgan astronouth7303: my way you will be version locked to what ever you put in the deb/rpm
19:21 Edgan astronouth7303: just like a linux distro
19:21 ipmb I put some code that will crash in the update_bootstrap function
19:21 Edgan astronouth7303: pip will give you 1.0.1 today and 1.0.2 tomorrow
19:21 ipmb it crashes when I run `sudo salt-cloud -u`
19:21 astronouth7303 so will apt, if they've updated in between :P
19:22 astronouth7303 but yes, building your own packages means you know when you rebuilt packages
19:22 Edgan astronouth7303: yes, but they go through QA to make sure it works
19:22 astronouth7303 sometimes....
19:22 Edgan astronouth7303: more than pip does
19:22 ipmb it is never hit when I try to create a server
19:22 astronouth7303 yeah, but there's been problems, especially in the more obscure packages, and i certainly wouldn't rely on it
19:23 Edgan ipmb: salt's python path?
19:24 Edgan astronouth7303: ubuntu is a spectrum, if it is in universe it isn't that well tested
19:26 numkem joined #salt
19:31 ipmb this is nuts, I'm editing the files to raise exceptions and they aren't getting hit
19:33 ipmb ah, salt-cloud goes through salt-minion which I wasn't restarting... may be onto something here
19:33 Edgan ipmb: Do you install the evil parallel universe connection python module?
19:33 Edgan ipmb: ah
19:33 numkem joined #salt
19:39 johnj_ joined #salt
19:41 phileus0 joined #salt
19:45 kellyp joined #salt
19:51 MajObviousman how's the conference getting along? I'm kinda bummed I couldn't attend
19:51 kellyp joined #salt
19:52 kellyp joined #salt
20:04 mechleg left #salt
20:05 threwahway joined #salt
20:09 SkyRocknRoll joined #salt
20:14 absolutejam Sorry to pester
20:15 absolutejam but I'm hoping some more eyes might drift across my issue
20:15 absolutejam docker_events reactor - I can't access data['Actor'] in my reactor states - Any ideas why/how/who/where?
20:16 absolutejam https://hastebin.com/ikoduhotes.scala is an example of what I get on the event bus
20:16 absolutejam And I'm trying to access it with {{ data['Actor'] }}
20:18 pbandark joined #salt
20:19 NeoXiD joined #salt
20:25 hemebond absolutejam: Is it possibly under data['data']['Actor']?
20:26 mechleg joined #salt
20:26 SkyRocknRoll joined #salt
20:28 absolutejam ohhh
20:28 absolutejam let me give it a go!
20:29 absolutejam You wonderful person
20:30 absolutejam not sure why I didn't think of that
20:30 absolutejam derp
20:32 hemebond 👍
20:32 threwahway joined #salt
20:34 cgiroua joined #salt
20:35 _KaszpiR_ joined #salt
20:38 ipmb Edgan interestingly, this only happens with LXC. I don't have the same issue with EC2 cloud profiles
20:39 XenophonF man, Salt on Windows makes Baby Jesus cry
20:40 XenophonF the same salt.minion state that works great across Linux and FreeBSD manages to put salt-minion into an infinite loop
20:40 XenophonF on windows
20:40 XenophonF :(
20:40 absolutejam I've had no issues really
20:40 johnj_ joined #salt
20:40 absolutejam besides bigger delay with zmq
20:41 astronouth7303 i haven't had that problem yet, but i've found module quality to be somewhat inconsistent
20:41 absolutejam I just wish PowerShell support was first class
20:41 absolutejam I tried - with my very limited python skills - to run a persistent powershell process that takes multiple commands, but it kept crapping out on me
20:42 absolutejam Because running a new powershell process per command is expensive and slow
20:45 Edgan ipmb: So sounds like a very corner case bug
20:45 ipmb this is what fails https://github.com/saltstack/salt/blob/2017.7.2/salt/modules/lxc.py#L3515-L3516
20:46 Edgan ipmb: I would reproduce it locally and use PyCharm to debug it
20:46 brd is it possible to filter results in Jinja? like I am trying to check if an IPv6 address includes "^fe80".. So I have tried this: {% if interface.igb0.inet6.0.address | regex_match('fe80:', ignorecase=True) -%}
20:46 ipmb salt-call config.gather_bootstrap_script works fine, however
20:46 ipmb so it is some difference between what happens with salt-call and the __salt__ object
20:48 toofer joined #salt
20:48 jas02 joined #salt
20:49 XenophonF brd: the only filters available to you come with whatever version of Jinja's installed + Salt's special sauce extensions
20:50 absolutejam They recently added a regex match iirc, but I added an execution module to do it because I was needing it
20:50 absolutejam which is pretty easy to do
20:50 XenophonF https://docs.saltstack.com/en/latest/topics/jinja/index.html
20:50 brd XenophonF: yeah, I am looking at: https://docs.saltstack.com/en/latest/topics/jinja/index.html#regex-search
20:50 XenophonF note the `New in version 2017.7.0` bit
20:51 XenophonF if you have the right version installed, maybe post an error message to help clue us in :)
20:51 brd yeah, I am running 'salt 2017.7.1 (Nitrogen)'
20:51 brd XenophonF: well the if just outputs anything
20:51 pipps joined #salt
20:52 brd XenophonF: er, it just never outputs anything
20:52 XenophonF where do you get the interface variable from?
20:53 thekevjames joined #salt
20:53 brd XenophonF: poking through salt '*' network.interfaces
20:54 toofer joined #salt
20:55 brd maybe I should switch to network.ip_addrs6
20:56 XenophonF ok
20:56 jas02 joined #salt
20:57 bvenable joined #salt
20:58 bryan joined #salt
20:59 user-and-abuser joined #salt
21:04 jas02 joined #salt
21:05 user-and-abuser joined #salt
21:14 toofer joined #salt
21:17 XenophonF would it kill Jinja to tell me which variable is the non-iterable NoneType please?
21:18 toofer joined #salt
21:18 Edgan XenophonF: what is the actual full error?
21:21 Neighbour XenophonF: A variable which you use to loop over (or call `.items()` on, again to loop over), but which turns out to be `None` at that point
21:21 bbradley joined #salt
21:21 SkyRocknRoll joined #salt
21:25 XenophonF no it got thrown because i ref variables in my jinja macro that i forgot to put in the macro arglist
21:25 XenophonF like, 6 or 7 that i've found so far
21:25 XenophonF guess i got ahead of myself
21:25 XenophonF it's just tedious tracking them all down
21:26 XenophonF chow time ttyl
21:27 pipps joined #salt
21:28 pipps joined #salt
21:29 threwahway joined #salt
21:34 ipmb joined #salt
21:37 cyborg-one joined #salt
21:42 ipmb Edgan I think it was a version mismatch between the master (where salt-cloud was running) and the minion that served as the lxc host. No idea why it manifested like that though :/
21:43 johnj_ joined #salt
21:46 fracklen joined #salt
21:46 Edgan ipmb: interesting
21:46 Edgan ipmb: what was the remote version?
21:47 ipmb 2015.8 iirc
21:47 Edgan ah, very old
21:47 pipps joined #salt
21:58 Praematura joined #salt
22:02 toofer joined #salt
22:08 onlyanegg joined #salt
22:21 nethershaw joined #salt
22:23 threwahway joined #salt
22:27 aldevar left #salt
22:29 kellyp joined #salt
22:32 teratoma joined #salt
22:40 kellyp joined #salt
22:44 johnj_ joined #salt
22:47 kellyp joined #salt
22:47 kellyp joined #salt
22:53 RandyT curious if anyone here has dealt with reattaching lvm ebs volumes?
22:53 RandyT not finding a set of tools to do this and wondering if I have gotten off the sanity path
22:53 RandyT a set of "salt" tools I should say
22:56 deathstar joined #salt
22:56 deathstar left #salt
22:58 deathstar joined #salt
23:14 deathstar joined #salt
23:16 pipps joined #salt
23:28 jfelchner joined #salt
23:31 Praematura joined #salt
23:39 dxiri hello everyone, need some help, can anyone tell why I am getting this error when running this state?
23:39 dxiri state is https://seashells.io/v/9NebMRd3
23:39 dxiri and output https://seashells.io/v/X9YJetXP
23:39 edrocks joined #salt
23:40 hemebond dxiri: Where is your network state?
23:40 hemebond And your libvirt state.
23:43 dxiri libvirt state I just shared it
23:43 cyborg-one joined #salt
23:44 dxiri don't have a network state, just a bridge on br0
23:44 hemebond Your libvirt state has no libvirt module in it.
23:44 hemebond And your libvirt service.running state requires a network: br0 state.
23:45 hemebond Dependencies point to states, not stuff.
23:45 johnj_ joined #salt
23:45 tiwula joined #salt
23:45 hemebond So "- network: br0" means a state called "br0" with a "network.blah" in it.
23:46 Whissi joined #salt
23:47 hemebond (and requires that state to succeed)
23:48 pipps joined #salt
23:48 dxiri ah I see, I took that one from the example state here: https://docs.saltstack.com/en/latest/topics/tutorials/cloud_controller.html
23:48 dxiri k then I should be fine just removing that
23:49 hemebond Right, and https://docs.saltstack.com/en/latest/topics/tutorials/cloud_controller.html#hypervisor-network-setup shows the network state.
23:49 hemebond br0: network.managed
23:49 Lenz joined #salt
23:50 dxiri thanks! only one question remains :)
23:51 dxiri don’t get the require: libvirt: libvirt thing
23:52 hemebond Well there is a vert.keys in there
23:52 hemebond Perhaps it's supposed to point to that.
23:52 hemebond *virt.keys
23:53 hemebond Possibly a typo in the require
23:53 hemebond Might have to be virt: libvirt
23:53 hemebond But you also need to add the virt.keys to your state.
23:54 dxiri I removed that cause a)I don't need the migrations and b) It was causing an error (maybe because I am on Centos6?)
23:55 hemebond Then you need to also remove the reference from the requires
23:56 dxiri what reference? don't see any "virt" reference
23:58 hemebond - libvirt: libvirt

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