[WIP] Disable retries on the CI

[apostasie]

There is a lot of context about this in #3303.

TL;DR:
Retrying tests was a good idea to make the CI less painful to work with, but unfortunately, the side-effect is that it is just making things worse by basically allowing tests (and code) that can fail half of the time to still "pass".

Merging this PR will make things very painful for us for some time as we'll have to "pay back" and fix all the things that have probably been failing for months, but it will make things much better in the long run.

LMK your thoughts...

[apostasie]

Note: I can't make this to go green right now obviously...
I would still avocate we need to merge this so that people can engage and fix these bugs.

[apostasie]

@ktock maybe you could help on these

TestRunStargz is failing twice on this run:

• https://github.com/containerd/nerdctl/actions/runs/10686581365/job/29622148138?pr=3400
• https://github.com/containerd/nerdctl/actions/runs/10686581365/job/29622151095?pr=3400

=== RUN   TestRunStargz
    container_run_stargz_linux_test.go:36: assertion failed: res.ExitCode is not exitCode: 
        Command:  /usr/local/bin/nerdctl --namespace=nerdctl-test --snapshotter=stargz run --rm ghcr.io/stargz-containers/fedora:30-esgz ls /.stargz-snapshotter
        ExitCode: 1
        Error:    exit status 1
        Stdout:   
        Stderr:   ghcr.io/stargz-containers/fedora:30-esgz: resolving      |--------------------------------------| 
        elapsed: 0.1 s                            total:   0.0 B (0.0 B/s)                                         
        ghcr.io/stargz-containers/fedora:30-esgz:                                      resolved       |++++++++++++++++++++++++++++++++++++++| 
        index-sha256:5286767fa09878e16acd75ab13bfa5b985473a7cf5442599c837153050b4122c: downloading    |--------------------------------------|    0.0 B/235.0 B 
        elapsed: 0.2 s                                                                 total:   0.0 B (0.0 B/s)                                         
        time="2024-09-03T16:21:58Z" level=error msg="server \"ghcr.io\" does not seem to support HTTPS" error="failed to prepare extraction snapshot \"extract-426242182-kaxG sha256:0e9db48c579d098b58a078ee45fc6490b2c297d677ee63a3da3ae032c59eb4d6\": connection error: desc = \"transport: Error while dialing: dial unix /run/user/1001/containerd-stargz-grpc/containerd-stargz-grpc.sock: connect: connection refused\": unavailable"
        time="2024-09-03T16:21:58Z" level=info msg="Hint: you may want to try --insecure-registry to allow plain HTTP (if you are in a trusted network)"
        time="2024-09-03T16:21:58Z" level=fatal msg="failed to prepare extraction snapshot \"extract-426242182-kaxG sha256:0e9db48c579d098b58a078ee45fc6490b2c297d677ee63a3da3ae032c59eb4d6\": connection error: desc = \"transport: Error while dialing: dial unix /run/user/1001/containerd-stargz-grpc/containerd-stargz-grpc.sock: connect: connection refused\": unavailable"
        
        
--- FAIL: TestRunStargz (0.39s)

Thanks in advance.

[apostasie]

Closing.
This is getting subsumed into the next "testing" PR with the introduction of the Flaky marker.