Stop using podman2 for loading and building images for cri-o

[afbjorklund]

We need to upgrade from podman2 (2.0/2.1/2.2) to one of the podman3 (3.0/3.1/3.2)

Currently using 2.2.1, and upgrading to 3.1.2. The minor versions are not compatible.

Eventually it should be possible to upgrade to 3.2.1, but it is still in "testing" (not "stable")

• https://build.opensuse.org/package/show/devel:kubic:libcontainers:stable/podman

• https://build.opensuse.org/package/show/devel:kubic:libcontainers:testing/podman

See https://podman.io/getting-started/installation#linux-distributions for details

RHEL versions (of Podman):

• v1.6.4
• v2.0.5
• v2.1.1
• v2.2.1
• v3.0.1
• v3.2.3
• v3.3.1
• v3.4.2

It is possible to use a newer version of podman for running with the podman driver.
But the version of podman-remote needs to match what is running on the machine.

Development is now happening on 4.0.0-dev, which might even will become Podman 4.0.
(Podman 3.0 was the final release for older distributions*, like CentOS 7, Ubuntu 18.04)

* see #10937

[afbjorklund]

#11704 was reverted in #11711

[afbjorklund]

#11727 is re-opened

[afbjorklund]

Podman 3.3.0 is out now, but it is impossible to know which of the podman3 to choose before it will all be podman4.

• v3.0.1
• v3.1.2
• v3.2.3

https://github.com/containers/podman/releases/tag/v3.3.0

[k8s-triage-robot]

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

This bot triages issues and PRs 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 or PR as fresh with /remove-lifecycle stale
• Mark this issue or PR as rotten with /lifecycle rotten
• Close this issue or PR 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 and PRs.

This bot triages issues and PRs 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 or PR as fresh with /remove-lifecycle rotten
• Close this issue or PR with /close
• Offer to help out with Issue Triage

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

/lifecycle rotten

[afbjorklund]

I think we should update podman to version 3.4.2 (or 3.4.4)

I don't plan on adding podman4 like ever, since it is not needed...

It is mostly about IPv6 support and about breaking the remote API ?

But I think we can drop support for the older distros, that only had 3.0

[afbjorklund]

Finally settled on podman 3.4.2, same as final version for Ubuntu 20.04 (Kubic)

• ISO: Podman 3.4.2 #13126

Using minikube for running podman is not recommended, it was mostly for cri-o
It is possible to run containers using sudo podman run, but it is not supported.

There are no plans to add podman 4.0.2, use Podman Machine for that instead ?

[afbjorklund]

Note: it is not possible to use minikube podman-env with a podman4 client.

Error: unable to connect to Podman socket: server API version is too old. Client "4.0.0" server "3.4.2"

You can download compatible podman clients from here:

https://github.com/containers/podman/releases/tag/v3.4.2