Preserve original volumes during Longhorn V2 cutover #6005

🤖 Generated by the Agentic Engineer

Summary

Keep each original Longhorn V1 volume recoverable while its application moves to V2.

Motivation

Most existing application claims still have a reclaim policy that deletes their storage with the claim. A reviewed cutover needs to keep the original backend so the application can recover if its replacement fails.

Proposed Solution

Provide a separately confirmed operation for one bound V1 application claim. Bind approval to the original claim, persistent volume and Longhorn backend. Change only the reclaim policy, then read all three original identities back.

Acceptance Criteria

• The operation requires the first protected run from current main and confirms the exact original storage identities.
• Changed or unbound objects, backup scratch resources and other engines are rejected.
• An interrupted response is observed without repeating the write.
• The protected cutover verifies the original backend survives claim replacement until its recovery and retirement checks pass.

Additional Context

This is a safeguard for the V2 migration. The platform-wide reclaim policy and scheduled audit remain tracked in #4441.

Part of #5582