feat: add --txt-prefix-override flag to support apex domains safely

[u-kai]

What does it do ?

This PR introduces a new flag, --txt-prefix-override, as a safe and flexible workaround
to support apex domains when using the TXT registry.

The flag allows users to override the TXT record prefix per domain by specifying
one or more mappings in the following form:

--txt-prefix-override=domain=txt-prefix

Multiple overrides can be provided, and each mapping applies only to the explicitly
specified domain.

Motivation

In #5864, we discussed limitations and edge cases around managing apex domains with the
TXT registry.

While a flag such as --apex-domains was considered, it was intentionally avoided for
the following reasons:

• A flag named --apex-domains could be misunderstood as applying special behavior to
  all apex domains automatically.
• The underlying problem is not limited to apex domains; it is more generally about
  domain-specific TXT prefix customization.
• Introducing a domain-scoped override mechanism keeps the behavior explicit and avoids
  surprising users.

By using --txt-prefix-override, users must opt in per domain, making the behavior
clear and predictable.

More

• Yes, this PR title follows Conventional Commits
• Yes, I added unit tests
• Yes, I updated end user documentation accordingly

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign mloiseleur for approval. For more information see the Code Review Process.

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

Details

Needs approval from an approver in each of these files:

• OWNERS

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

[k8s-ci-robot]

Hi @u-kai. Thanks for your PR.

I'm waiting for a github.com member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Details

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-sigs/prow repository.

[ivankatliarchuk]

This should be in it's own PR.

[ivankatliarchuk]

I may explain a bit more - some changes, like refactoring are easy to revew/approve/merge. While adding new capabilities or fixing certain behaviour has different SDLC

[u-kai]

Understood, thanks for the clarification.
I’ve split this into a separate PR.

[ivankatliarchuk]

Conceptually the change aligns with ExternalDNS’s goal of offering pragmatic knobs for tricky DNS layouts like apex records (this is just single opinion)

What I'm unsure

1. Another new flag with prefix and domain mapping
2. Missing any practical documentation - as without clear use case, no clear reason of adding features

We need to review pros-cons of adding flag complex vs annotation external-dns.alpha.kubernetes.io/txt-prefix as example.

Even then the change may still be reviewed and rejrected by project owners

[u-kai]

@ivankatliarchuk

Another new flag with prefix and domain mapping

Thanks — I’m not sure I fully understand this concern yet.
Do you mean a case where multiple domains need different TXT prefix overrides?

If so, the current design supports this by allowing the flag to be specified multiple times at runtime.
If you meant something else, could you clarify what kind of mapping or complexity you’re concerned about?

Missing any practical documentation - as without clear use case, no clear reason of adding features

I’ve already added a detailed description to the flag itself, including the intended use case for apex records and support for multiple domains.

Would you prefer:

• expanding the flag description further, or
• adding a separate piece of documentation (e.g. a short guide or example section) that explains the practical use cases in more detail?

We need to review pros-cons of adding flag complex vs annotation external-dns.alpha.kubernetes.io/txt-prefix as example.

My current view is:

Annotation approach:

• Pros: behavior is more visible and intuitive to users
• Cons: requires implementation across all sources and TXT registry, and deciding how to propagate txt-prefix data through Endpoint

Flag approach (this PR):

• Pros: change is mostly contained to TXT registry logic
• Cons: ExternalDNS restart is required when flags change

From a maintenance and scope perspective, I chose the flag-based approach for now.

[ivankatliarchuk]

Let's w8 for other parties to review and consider which approach should we choose.

• Annotation over flag, could help conditionally configure it for specific resources
• Flag of course has it's own advantages. But composite flags, when there is domain+prefix, not sure, feels quite tricky. If we have a configuration file support, it could be easier.

This PR slowly slips into area of #3757

Relates

• Creating dns records for apex / root domains #449
• Multiple design failures in new-format TXT registry ownership records #3757

[k8s-ci-robot]

PR needs rebase.

Details

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-sigs/prow repository.

[ivankatliarchuk]

I don't see --txt-prefix-override as something we should support. The problem is with TXT registry and how it handles records for apex domains. This feels like a patchwork and tech debt, not a solution.

[ivankatliarchuk]

Last attempt to change/introduce new metadata format #4624. Was not successful, as jumped straight in implementation, without cleanup older formats and concrete specification.

[ivankatliarchuk]

The problem that we trying to solve feels like has to be clearly explained in the PR body. Not sure what problem we are solving here. There is on WHAT pr is doing, not WHY.