[00:04] <ShellcatZero> Does anyone here do a lot of configuration management?  Do you get much benefit of something like Ansible over shell scripts?  I've personally found the translation of scripts to Ansible to non-trival and rather redundant, and I've found no such tools which seem to assist in automatically creating such scripts or playbooks from an OS environment.
[00:07] <hermande> ShellcatZero, Are you saying that transitioning to Ansible or Puppet from shell scripts are a PITA?
[00:07] <rbasak> ShellcatZero: it's handy for sharing common things between multiple instances, shipping templated files to instances, and so on. Sure, you can use shell, but then you're just reimplmenting some library that configuration management tools already provide.
[00:09] <Disaster_Area> i think typing dbeaver -vm /usr/lib/jvm/java-8-openjdk-amd64/jre/bin/java
[00:09] <Disaster_Area>  into my terminal has worked! :o
[00:10] <ShellcatZero> hermande: Just based on my experience, yes, and it seems redundant.  From my understanding, scripted provisioning at scale is just as well acomplished via ssh, rendering scalability benefits of Ansible or Puppet rather moot.
[00:16] <hermande> ShellcatZero, And for what it's worth, A well configured Scripting system can be powerful.
[00:17] <hermande> ShellcatZero, Ansible and Puppet have known interfaces that don't require special knowledge or code.
[00:18] <hermande> ShellcatZero, If I showed you my old scripting system and asked you to deploy it across a set of servers, you would insist that my coding didn't fall into your box of "Good Things"
[00:19] <hermande> With Puppet/Ansible, it's easier to describe using a common language.
[00:19] <ShellcatZero> hermande: Ah, I see what you're saying.
[00:20] <mybalzitch> did the docker snap break for anyone else? I get https://gist.github.com/thisisnotmyrealname/61f2215cb5dc23d2184d4c37474f7f92 after docker was running fine for ages
[00:20] <hermande> In otherwords, you are making yourself replaceable....
[00:20] <mybalzitch> removing the snap and readding it let me launch things again, but it seems the next day its broken
[00:42] <ShellcatZero> hermande: Based on what you describe I don't think Ansible or Puppet is applicable to my use cases, because I need to enforce domain knowledge for those using my scripts, which must also be extended as the need arises.
[00:47] <hermande> ShellcatZero, That's a fun topic.  So many tools enforce learning the API/UI over what is actually happening.
[00:47] <hermande> ShellcatZero, I've seen people pull up a full UI to make a small change in the /etc.
[06:28] <lordievader> Good morning
[09:55] <loru> Hello. I'm on ubuntu 18.04, using netplan on a server. I have addresses: [ x.x.x.50/24, '2x:x:x:x::ffff/64' ] and I need to add one more ipv4 and v6 (currently in a container, attaching them on the host instead).
[09:55] <loru> Would "addresses: [ x.x.x.50/24, y.y.y.51/24, 'x:x:x:x::ffff/64', 'y.y.y.y::ffee/64' ]" be fine?
[10:01] <lotuspsychje> idle a bit here ok loru -server channel might wakeup more at US wakeup
[10:02] <loru> lotuspsychje: sure
[10:02] <loru> https://gurdeep.ninja/ovh-soyoustart-ip-failover-ubuntu-18-04/
[10:02] <loru> I'm not using OVH but it seems to be it
[10:02] <loru> looks simple, I'll try :)
[10:02] <lotuspsychje> !netplan | loru see also
[10:40] <loru> It works anyway, thank you :D
[12:34] <paride> ahasenack, not sure why your git-ubuntu MP was being ignored by git-ubuntu-ci
[12:34] <paride> I had a look at the code and couldn't spot a limit to 10 MPs
[12:35] <paride> there are currently 11 active reviews in https://code.launchpad.net/usd-importer/+activereviews , the last one is from bryce and I can see it in the git-ubuntu-ci-trigger-job
[12:36] <paride> e.g. https://jenkins.ubuntu.com/server/view/git-ubuntu/job/git-ubuntu-ci-trigger/57812/console
[12:36] <paride> it's ce9afc69d8e6199ec96e1f28b68b23d3db9a8459
[12:41] <ahasenack> paride: it's because the MP was in the "approved" state already
[12:41] <ahasenack> paride: I switched it back to "needs review", kicked the job, and then it picked it up and voted
[13:05] <paride> ahasenack, OK, good to know
[14:12] <mason> Is there an apt(8) equivalent to "apt-get source" that I'm missing, or is there some tool that's preferred? Or is the goal for apt-get to always be there for its source-getting chops?
[14:23] <albert23> mason: you mean apt source <package> ?
[14:24] <mason> albert23: I don't know. Half a sec and I'll tell you if I mean that or not. :)
[14:25] <mason> albert23: Well, look at that. I hadn't thought to try that since it's not listed in the man page.
[14:25] <mason> albert23: I feel a docbug coming on.
[14:26] <ahasenack> rbasak: what's the name of that tool that organizes d/control? Puts each depends in its own line, orders it alphabetically, etc
[14:26] <mason> albert23: Thank you.
[14:26] <rbasak> ahasenack: wrap-and-sort
[14:27] <ahasenack> thx
[14:27] <albert23> mason: np
[14:30] <mason> albert23: Ah, and I guess my question is mildly confused, as when I snagged the apt source to whip up a patch, there was apt-get in the same package. Cough.
[14:44] <mason> albert23: Do you know if -a/--host-architecture is vestigial? I wanted to check options before snagging the apt-get man page section on source intact, and I seem to be missing some clue about how to apply -a.
[14:48] <mason> https://bpaste.net/show/25cb9c46b310 anyway - I'll defer the docbug until I know if what I'm using as a source is confused
[17:19] <JanC> mason: not only missing from the man page but also from --help apparently
[17:21] <JanC> but bash completion knows the 'source' command
[17:47] <mason> JanC: Do you know about -a there? I haven't dug into the source, but it seems not to work, and copying bad info from one man page to another might not be optimal.