Add documentation to describe how to backfill pre-VDF images into post-VDF subdir path

[listx]

Problem: After the vanity domain flip happened, there are now N subprojects with their own subdirs underneath k8s.gcr.io/<SUBDIR>. This is fine. But now we would like to backfill pre-VDF images that currently live in k8s.gcr.io/<OLD_IMAGE> into k8s.gcr.io/<SUBDIR>/<OLD_IMAGE>. This way, any new post-VDF references that use the SUBDIR prefix can also refer to old images that were built before the VDF happened, using the same SUBDIR prefix.

In other words, all images in k8s.gcr.io that don't live under a staging subproject's SUBDIR should be given a SUBDIR and copied there.

What we want to do is just list the old images in images.yaml to make sure they also exist under the subproject's SUBDIR. However, this won't work on its own because the staging repo won't have these images. So there are at least 2 steps involved:

1. Copy pertinent old images from k8s.gcr.io root into gcr.io/k8s-staging-SUBPROJECT.
2. Add images from Step 1 into the images.yaml for gcr.io/k8s-staging-SUBPROJECT, to promote into k8s.gcr.io/SUBPROJECT-SUBDIR/<old_image>.

This involves some trivial shell logic, probably. But we could just start with some docs.

[listx]

Which sample projects need backfilling first? @thockin @justaugustus

[listx]

/assign @listx

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[spiffxp]

/kind documentation
/wg k8s-infra
/area artifacts
/sig release
/area release-eng
/assign @justaugustus @hasheddan
FYI @kubernetes/release-engineering
Is this still relevant?

[justaugustus]

Given we've done the VDF in kubernetes/release#270 and all images were backfilled, I'd say let's close this.

For RelEng, I copied the last two or three versions of our images and promoted those ahead of VDF.
I'd suggest subprojects that need similar to do the same in a post-VDF world, instead of having us sign up for the discovery work, which I think ultimately will be low-value in comparison.

/close

[k8s-ci-robot]

@justaugustus: Closing this issue.

In response to this:

Given we've done the VDF in kubernetes/release#270 and all images were backfilled, I'd say let's close this.

For RelEng, I copied the last two or three versions of our images and promoted those ahead of VDF.
I'd suggest subprojects that need similar to do the same in a post-VDF world, instead of having us sign up for the discovery work, which I think ultimately will be low-value in comparison.

/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/test-infra repository.