Add retry option to container_push rule

[jumbosushi]

🚀 feature request

Relevant Rules

container_push in contrib/push-all.bzl

Description

It would be great if container_push rule supports a retry option. (this is an idea briefly explored in #673). We recently ran into a transient networking issue which was resolved by retrying manually:

# note: gcr info is deducted
2021/12/19 19:16:45 Unable to publish images: unable to push image gcr.io/xyz:abc: Head "https://gcr.io/v2/xyz/blobs/sha256:abc": read tcp 10.248.250.64:53422->142.250.148.82:443: read: connection reset by peer

Describe the solution you'd like

• Add a new retry attribute to container_push rule like the following.

container_push = rule(
    attrs = {
        ...
        "retry": attr.bool(
            default = False,
            doc = "If true, retry push images on transient errors.",
        ),
    }
    ...
)

• Update container/go/cmd/pusher/pusher.go to support --retry boolean flag to enable retry logic

Describe alternatives you've considered

We could add a retry logic to our bazel run invocations as a workaround, but we run it at various places within our CI system :/

---

I would be interested in working on the implementation! Let me know if the feature would be useful

[uhthomas]

This would be great. Transient network failures when pushing dozens of images using container_bundle is really inconvenient.

[github-actions]

This issue has been automatically marked as stale because it has not had any activity for 180 days. It will be closed if no further activity occurs in 30 days.
Collaborators can add an assignee to keep this open indefinitely. Thanks for your contributions to rules_docker!

[github-actions]

This issue was automatically closed because it went 30 days without a reply since it was labeled "Can Close?"