kubelet-1.18.4-0 cannot be installed alongside cri-o on fedora 32

[davemac30]

What happened:

Tried to install kubelet-1.18.4-0 RPM on a Fedora 32 box which had cri-o installed. Got the following error:

Error:
 Problem: problem with installed package containernetworking-plugins-0.8.6-1.fc32.x86_64
  - package kubelet-1.18.4-0.x86_64 conflicts with kubernetes-cni provided by containernetworking-plugins-0.8.6-1.fc32.x86_64
  - package kubelet-1.18.4-0.x86_64 conflicts with kubernetes-cni provided by containernetworking-plugins-0.8.5-1.1.gitf5c3d1b.fc32.x86_64
  - conflicting requests

What you expected to happen:

As with previous versions of kubelet RPM (e.g. 1.18.3-0), I expected kubelet-1.18.4-0 to install without errors.

How to reproduce it (as minimally and precisely as possible):

On a clean, new vagrant box based on fedora/32-cloud-base, run the following:

sudo dnf -y module enable cri-o:1.18
sudo dnf -y install cri-o
sudo dnf config-manager --add-repo https://packages.cloud.google.com/yum/repos/kubernetes-el7-\$basearch
sudo rpmkeys --import https://packages.cloud.google.com/yum/doc/yum-key.gpg https://packages.cloud.google.com/yum/doc/rpm-package-key.gpg
sudo dnf -y install kubelet-1.18.4-0

Anything else we need to know?:

Environment:

Vagrant 2.2.9, VirtualBox 6.1.10, MacOS 10.15.5

Vagrantfile

# -*- mode: ruby -*-
# vi: set ft=ruby :

Vagrant.configure("2") do |config|

  config.vm.box = "fedora/32-cloud-base"

  config.vm.provider :virtualbox do |virtualbox|
    virtualbox.linked_clone = true
    virtualbox.customize ["modifyvm", :id, "--audio", "none"]
    virtualbox.memory = 1024
  end

end

[saschagrunert]

Hey @davemac30 👋, with the release of Kubernetes 1.18.4 we removed packaging kubernetes-cni, which means we probably can fix this error with #1367, too.

[saschagrunert]

Alright #1367 (comment) clarifies that the solution is not as trivial as initially thought.

Calling for @dougsland 👋 do you have a suggested way how we can use both, the Fedora module and the packages on yum.k8s.io together?

cc @lsm5 @haircommander

[dougsland]

Alright #1367 (comment) clarifies that the solution is not as trivial as initially thought.

Calling for @dougsland do you have a suggested way how we can use both, the Fedora module and the packages on yum.k8s.io together?

cc @lsm5 @haircommander

I think it's hard to keep a statement that both repos will work 100% all the time without conflicts (as independent repos). @lsm5 what you say?

[justaugustus]

Potential fix being discussed here: #1375

[justaugustus]

@davemac30 -- We've published new packages. Can you try again w/ 1.18.4-1?

[davemac30]

Seems to work fine - cri-o and kubelet installed together no problem:

[vagrant@localhost ~]$ sudo rpm -qa cri-o kubelet
cri-o-1.18.1-2.module_f32+9078+15f1cda4.x86_64
kubelet-1.18.4-1.x86_64

Thanks all for the speedy resolution!

[jlpettersson]

This still seem to be a problem.

I tried to install kubelet and cni-tools.

Error:
 Problem: conflicting requests
  - package kubelet-1.28.0-150500.1.1.aarch64 from kubernetes requires kubernetes-cni >= 1.2.0, but none of the providers can be installed

and if I let kubernetes-cni be installed when I have cni-tools I get:

Error: Transaction test error:
  file /opt/cni/bin/bandwidth from install of kubernetes-cni-1.2.0-150500.2.1.aarch64 conflicts with file from package cni-plugins-1.2.0-1.amzn2023.0.3.aarch64

The binaries are installed from https://pkgs.k8s.io/core:/stable:/v1.28/rpm/

[haircommander]

cni-plugins-1.2.0-1.amzn2023.0.3.aarch64 I think this may be something to take-up with the folks that put together your distro. is it Amazon Linux?

[jlpettersson]

@haircommander yes, it is Amazon Linux 2003.

But I will workaround by using curl instead of the yum repositories, it seems not to be a good way to install the kube binaries.

[xmudrii]

/reopen
/assign
I'll take a look into this

[k8s-ci-robot]

@xmudrii: Reopened this issue.

In response to this:

/reopen
/assign
I'll take a look into this

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle stale
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[k8s-triage-robot]

The Kubernetes project currently lacks enough active contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle rotten
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle rotten

[xmudrii]

/remove-lifecycle rotten

[xmudrii]

I'm not working on this at the moment.
/unassign

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle stale
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[xmudrii]

/close

[k8s-ci-robot]

@xmudrii: Closing this issue.

In response to this:

/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.