Replace Ubuntu -> Debian SID

[cevich]

What type of PR is this?

/kind other

What this PR does / why we need it:

Replace use of Ubuntu with Debian SID CI VM Images.

How to verify it

CI will pass

Which issue(s) this PR fixes:

None

Special notes for your reviewer:

Ref: containers/automation_images#250

Does this PR introduce a user-facing change?

None

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: cevich

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [cevich]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[edsantiago]

[+1668s] not ok 467 ignore-socket
[+1668s] # (from function `die' in file ./helpers.bash, line 332,
[+1668s] #  in test file ./copy.bats, line 274)
[+1668s] #   `die "Timed out waiting for ${TEST_SCRATCH_DIR}/test.socket (is nc installed?)"' failed
[+1668s] # /var/tmp/go/src/github.com/containers/buildah/tests /var/tmp/go/src/github.com/containers/buildah/tests
[+1668s] # nc: invalid option -- 'U'

Wrong nc. Solution:

# apt-get install ncat <----- yes, "ncat", exactly like that
# update-alternatives --set nc /usr/bin/ncat

(scriptifying that is left as an exercise for the reader)

[cevich]

Thanks @edsantiago I just finished figuring that all out on my own...THEN found your comment lol.

[cevich]

Ref: containers/automation_images#256

[cevich]

@nalind could you help me with the Conformance test failures? They appear legitimate to me but I'm nowhere near an expert. This is the output from installing docker/containerd:

[+0001s] Installing previously downloaded/cached Docker packages
[+0002s] Selecting previously unselected package containerd.io.
[+0002s] dpkg: considering removing runc in favour of containerd.io ...
[+0002s] dpkg: yes, will remove runc in favour of containerd.io
[+0005s] (Reading database ... 101253 files and directories currently installed.)
[+0005s] Preparing to unpack .../containerd.io_1.6.18-1_amd64.deb ...
[+0005s] Unpacking containerd.io (1.6.18-1) ...
[+0007s] Removing runc (1.1.4+ds1-1+b2), to allow configuration of containerd.io (1.6.18-1) ...
[+0007s] Selecting previously unselected package docker-ce-cli.
[+0007s] Preparing to unpack .../docker-ce-cli_5%3a23.0.1-1~debian.11~bullseye_amd64.deb ...
[+0007s] Unpacking docker-ce-cli (5:23.0.1-1~debian.11~bullseye) ...
[+0009s] Selecting previously unselected package docker-ce-rootless-extras.
[+0009s] Preparing to unpack .../docker-ce-rootless-extras_5%3a23.0.1-1~debian.11~bullseye_amd64.deb ...
[+0009s] Unpacking docker-ce-rootless-extras (5:23.0.1-1~debian.11~bullseye) ...
[+0010s] Selecting previously unselected package docker-ce.
[+0010s] Preparing to unpack .../docker-ce_5%3a23.0.1-1~debian.11~bullseye_amd64.deb ...
[+0010s] Unpacking docker-ce (5:23.0.1-1~debian.11~bullseye) ...
[+0012s] Setting up containerd.io (1.6.18-1) ...
[+0012s] Created symlink /etc/systemd/system/multi-user.target.wants/containerd.service → /lib/systemd/system/containerd.service.
[+0013s] Setting up docker-ce-cli (5:23.0.1-1~debian.11~bullseye) ...
[+0013s] Setting up docker-ce-rootless-extras (5:23.0.1-1~debian.11~bullseye) ...
[+0013s] Setting up docker-ce (5:23.0.1-1~debian.11~bullseye) ...
[+0014s] Created symlink /etc/systemd/system/multi-user.target.wants/docker.service → /lib/systemd/system/docker.service.
[+0014s] Created symlink /etc/systemd/system/sockets.target.wants/docker.socket → /lib/systemd/system/docker.socket.
[+0016s] Processing triggers for runit (2.1.2-54) ...
[+0016s] Processing triggers for man-db (2.11.2-1) ...
[+0028s] Configuring /etc/containers/registries.conf

[cevich]

Update: I fired up the current (from main) Ubuntu CI VMs and find the following:

containerd.io_1.6.15-1_amd64.deb
docker-ce-cli_5%3a20.10.22~3-0~ubuntu-jammy_amd64.deb
docker-ce-rootless-extras_5%3a20.10.22~3-0~ubuntu-jammy_amd64.deb
docker-ce_5%3a20.10.22~3-0~ubuntu-jammy_amd64.deb
docker-scan-plugin_0.23.0~ubuntu-jammy_amd64.deb
pigz_2.6-1_amd64.deb

So the conformance issues may be due to the newer versions.

[nalind]

Is it possible to use the same versions of docker-ce on sid and Fedora?

[cevich]

Is it possible to use the same versions of docker-ce on sid and Fedora?

Possibly, and likely by manually hard-coding then rebuilding the VM images 😱 Then they're also unlikely to represent the versions run by Debian users 😞 Are these versions known-broken in some way, or is there some other good reason to not use them?

For ref: I'm already hard-coding the Debian release name (for lack of an easy way to get SID-1) when installing Docker. In other words, "bullseye" is Debian 11 and SID is 12 ("bookworm").

[nalind]

It's... difficult to conform to the behavior of multiple different target versions at the same time.

[nalind]

When I install unstable here, the default docker.io version is 20.10.23+dfsg1-1+b1. Is the VM getting its packages from somewhere else?

[cevich]

It's... difficult to conform to the behavior of multiple different target versions at the same time.

Gotcha, yeah that's probably a PITA. At least these tests are only running on Debian and not Fedora also 😁

When I install unstable here

Clarification, the link I posted (above) is to the VM image build script that installs docker. It's pretty stimple, though I believe it's the latest stable version, not unstable. Only the Debian OS is "unstable". I could change this, but I think it might make the problem worse.

Another idea...would it be appropriate for me to just file issues and skip these tests for now with "FIXME reasons"?

[cevich]

Force-push: rebased.

[cevich]

In case it helps, I ran docker version and got:

Client: Docker Engine - Community
 Version:           23.0.1
 API version:       1.42
 Go version:        go1.19.5
 Git commit:        a5ee5b1
 Built:             Thu Feb  9 19:46:54 2023
 OS/Arch:           linux/amd64
 Context:           default

Server: Docker Engine - Community
 Engine:
  Version:          23.0.1
  API version:      1.42 (minimum version 1.12)
  Go version:       go1.19.5
  Git commit:       bc3805a
  Built:            Thu Feb  9 19:46:54 2023
  OS/Arch:          linux/amd64
  Experimental:     false
 containerd:
  Version:          1.6.18
  GitCommit:        2456e983eb9e37e47538f59ea18f2043c9a73640
 runc:
  Version:          1.1.4
  GitCommit:        v1.1.4-0-g5fd4c4d
 docker-init:
  Version:          0.19.0
  GitCommit:        de40ad0

[cevich]

@nalind I went trolling through the test-code referenced by the failure message in the logs. I see a pair of calls that seem relevant, around tests/conformance/conformance_test.go:456:

    deleteLabel(originalBuildahConfig, buildah.BuilderIdentityAnnotation)
    deleteLabel(ociBuildahConfig, buildah.BuilderIdentityAnnotation)

Those make me think, could this moby.buildkit.buildinfo.v1 label simply be deleted/ignored by the tests? (I'm assuming its presence is what's causing the tests to fail). Or is the problem more complicated/nuanced than that? All the tests I looked at appear to fail with this same error.

[cevich]

force-push: rebased

[cevich]

Opened #4639 for conformance test failures.

[TomSweeneyRedHat]

LGTM
but want a head nod from @nalind, especially concerning the conformance test changes
Happy green test buttons.

[cevich]

@nalind ping

[nalind]

#4639 solves my main issues, the rest looks fine.

[rhatdan]

/lgtm

[cevich]

force-push: Rebased.

Thanks Nalin.

[edsantiago]

/lgtm