Recover a partially completed portfolio migration
Pause the batch, read the per-domain report and retry only what failed, without restarting the whole portfolio.
Pause the batch and open its per-domain report. Separate completed items, blocked items, confirmed failures and uncertain provider results. Save the batch reference and the approved plan. Do not restart the entire inventory just because some items failed.
For blocked ownership or source-content review, resolve that requirement first. For a DNS or certificate delay, check the current external state before retrying. An unknown result may mean the provider accepted the operation but the response was lost. Reconciliation should determine what happened before another write.
Retry only eligible failed items using the supported retry control. If domains, content, DNS changes or cost have changed since approval, generate and approve a new plan. Keep the agreed cost cap and inspect progress until each item has an explained outcome.
For rollback, compare the saved before-state with the current provider records. Restore only changes covered by the approved recovery plan; never overwrite a later legitimate change blindly. If mail or an existing site was affected, stop further cutovers and escalate with the exact records and timestamps. Confirm recovery through real page and mail checks.