# OrderGuard existing-store cutover: stop the workers

Status: procedure template for qualification. It is not permission to use the cutover on a live store.

The cutover is permanent. Use it only after the procedure for that store's actual host has been confirmed. Stopping WP-Cron alone is not enough: a PHP request, WP-CLI process, system cron, queue runner or Action Scheduler worker may still be executing.

## Required sequence

1. Put the store into a maintenance window which prevents new checkout and administrator writes.
2. Disable the server cron entries which invoke `wp-cron.php`, WP-CLI or an OrderGuard/WooCommerce queue command.
3. Stop or restart the site's PHP workers, then verify that the old worker processes are gone.
4. In WooCommerce Scheduled Actions, verify that no `orderguard_*` or `og_*` action is `in-progress`. Do not delete actions merely to pass the check.
5. Verify that WordPress Tools → Erase Personal Data has no unfinished OrderGuard request and that OrderGuard shows no erasure protocol fault.
6. Keep checkout, server cron, WP-CLI, queue runners and every background PHP worker stopped. Restart only the web PHP service needed to serve the administrator, then open Tools → OrderGuard takeover, tick the confirmation and perform the switch once. Do not allow public traffic during this step.
7. Confirm the lifecycle reads `active`, the action is present in the cutover audit, and the native eraser is attached with its reviewed receipt key.
8. Restore PHP workers and scheduled jobs, then leave maintenance mode.
9. Run one non-customer smoke check and the reviewed test erasure. If either fails, keep customer erasures unavailable and contact support; never edit the lifecycle row.

## Host-specific qualification still required

For each live store, record the host, control panel, installed PHP version, current OrderGuard version, exclusion count, the exact PHP restart/stop control, and whether scheduled jobs use real cron or WP-Cron. The six live stores listed in the worklist are not qualified by this generic template.

- cPanel: identify the host's PHP-FPM restart control or have support stop/restart the account's PHP workers; disable the relevant Cron Jobs entries.
- Plesk: identify the domain's PHP-FPM restart control and disable the relevant Scheduled Tasks.
- Managed WordPress: obtain the host's written procedure or support confirmation for draining PHP and background jobs.
- VPS/Docker: stop the site-specific PHP/worker containers and cron/queue services without stopping the database; verify no old containers or processes remain before cutover.

Record the operator, UTC time, procedure/version used, evidence that workers were stopped, cutover audit result, and restart result. Do not record customer identifiers.
