feat(etcd-manager-ctl): use backupname to delete backup instead of timestamp

[sudoforge]

This commit refactors the `runDeleteCommand` function of
`etcd-manager-ctl` to use the "backup name" instead of the "timestamp"
attribute. More often than not, users are more likely to become familiar
with the syntax of `etcd-manager-ctl restore-backup`, which uses the
`backupname` attribute to identify the target backup to restore.

By bringing the `delete-backup` subcommand in alignment with
`restore-backup`, we improve the user experience of an operator in the
process of restoring a cluster.

[sudoforge]

I ran into an issue the other day at $WORK, where two invalid commands were at the beginning of the command queue. I found the requirement to use the timestamp a little weird, when most other commands use the backup name.

Let me know if there are any changes you want here, folks. Thanks for all of your hard work!

[justinsb]

Thanks - this is a good idea. It's technically breaking, but as I agree it's so much more consistent and intuitive it seems reasonable.

If anyone objects, we could add a flag for deletion by timestamp, or just support matching either.

/approve
/lgtm

[sudoforge]

Yeah, if we think adding a flag for deleting by timestamp is a good idea, feel free to comment here and I'll take care of it.

[sudoforge]

Thanks for merging, @justinsb!

[justinsb]

I'm happy to see whether people ask for the flag; if you actually want to add it we can certainly merge it, but otherwise it's fine to wait and see I think!

BTW, have you signed the kubernetes CLA? We're trying to merge this project into the kubernetes-sigs etcdadm ( kubernetes-retired/etcdadm#180 (comment) ), and you can see we're hitting some CLA issues. I don't know if it's your account or someone else's though... Have you signed the CLA (or are you able to sign the CLA?)

[sudoforge]

Done: kubernetes-retired/etcdadm#180 (comment)

RE: the flag, I'm happy to wait and see if people ask for it.