Environment: Enhance 12.25.x, multi-server cluster — control-only master (Application role for the panel only, no DB role) + a dedicated worker node running Application (OpenLiteSpeed) + Database (MariaDB 10 LTS). Source site is WordPress.
Issue 1 — clone writes a doubled DB prefix into wp-config.php.
When I clone a WordPress site, the clone job succeeds and correctly creates the destination databases and users with the expected single prefix — e.g. source DB acct_wpdb → destination stg1_acct_wpdb, and the same for the user (the clone log confirms these single-prefixed names).
But the destination site’s wp-config.php gets the prefix applied twice:
define( 'DB_NAME', 'stg1_stg1_acct_wpdb' ); // doubled
define( 'DB_USER', 'stg1_stg1_acct_wpdb' ); // doubled
WordPress then points at stg1_stg1_acct_wpdb, which was never created → “Error establishing a database connection” on every page. Expected: wp-config should reference the single-prefixed names the clone actually created (stg1_acct_wpdb). Workaround: manually edit wp-config to remove the extra prefix (password/DB_HOST were already correct).
Is anyone else seeing the doubled prefix on clones, or is there a setting that changes how the clone rewrites wp-config?
Issue 2 — supported way to clone a customer-owned site into staging?
The “Clone a website” source card only lists sites owned by the org I’m currently acting as. Once a site is owned by a customer org (dedicated-VPS setup), as admin the clone card is greyed (admin owns no sites), and impersonating the customer gives a reduced Add-website flow with no source cards. The only path I’ve found is transferring the site’s ownership back to the admin org, cloning, then transferring back — which also drops the customer’s server dedication, so I have to manually pin server placement for the clone.
Is there a supported way to clone a customer-owned site into a staging site on that customer’s dedicated server, without the transfer-ownership-back-and-forth?
Thanks!
