Request 522278: Submit cloud-netconfig

This request supersedes: request 520607 (Show diff)

Overview

Request 522278 accepted

new package cloud-netconfig (scripts to automatically manage network interfaces in public cloud frameworks)

Created by jgleissner over 6 years ago
In state accepted
Supersedes 520607

Submit cloud-netconfig

Comments for request 522278 1
Comments for request 520607 8

Staging Bot (staging-bot) - over 6 years ago

reviewer

@aosthof, @bear454, @dirkmueller, @glaubitz, @jdsn, @rjschwei, @sax2, @svollath, @thardeck: review reminder

Dirk Mueller (dirkmueller) - over 6 years ago

I'm not sure _multibuild can be submitted to Factory already. can it?

Dominique Leuenberger (dimstar) - over 6 years ago

reviewer

_multibuild can be accepted to Factory - with some limitations (mostly when we hit rings, it can be nasty) = but here we should be safe

There are file conflicts between various packages though:

found conflict of cloud-netconfig-azure-0.3-3.1.noarch with cloud-netconfig-ec2-0.3-3.1.noarch:
  - /usr/lib/udev/rules.d/75-cloud-persistent-net-generator.rules
  - /etc/netconfig.d/cloud-netconfig
  - /etc/sysconfig/network/scripts/cloud-netconfig-cleanup
  - /etc/sysconfig/network/scripts/cloud-netconfig-hotplug
  - /etc/sysconfig/network/scripts/functions.cloud-netconfig

That needs to be addressed

Joachim Gleissner (jgleissner) - over 6 years ago

author source maintainer

The whole point for me to use multibuild was so that I can have the same file in different versions. It implements the cloud framework specific bits, and the main script simply sources it. Only one of the packages (the one matching your public cloud framework) is supposed to be installed, installing both doesn't make sense.

Jimmy Berry (jberry) - over 6 years ago

Indeed, so just add conflict to other providers so users don't.

Joachim Gleissner (jgleissner) - over 6 years ago

author source maintainer

Since the packages clash, they can't really. Do the packages need to conflict with each other explicitly?

Jan Engelhardt (jengelh) - over 6 years ago

reviewer

+# persistent net generator was split off from udev in Leap

Since SLE15 is likely going to inherit udev-persistent-ifnames, the condition should probably be dropped, or rewritten to not just require it on Leap, but also on SLE15. And Tumbleweed. And uh...

Jan Engelhardt (jengelh) - over 6 years ago

reviewer

Or rather, it should exclude SLE12.

Joachim Gleissner (jgleissner) - over 6 years ago

author source maintainer

Ack, I'll change it.