MySQL

You’re including the table prefix, but $wpdb takes care of that. Change: $query = $wpdb->prepare( “SELECT * FROM $wpdb->wp_terms INNER JOIN $wpdb->wp_term_taxonomy ON ($wpdb->wp_terms.term_id = $wpdb->wp_term_taxonomy.term_id) WHERE $wpdb->wp_terms.name LIKE %s AND $wpdb->wp_terms.count > 0”, $wpdb->esc_like($keyword) . ‘%’); To: $query = $wpdb->prepare( “SELECT * FROM $wpdb->terms INNER JOIN $wpdb->term_taxonomy ON ($wpdb->terms.term_id = $wpdb->term_taxonomy.term_id) WHERE $wpdb->terms.name LIKE … Read more

You don’t need to write a custom SQL for this. You can always leverage WP’s get_post_custom() that returns array of all custom fields for a particular page or post. https://codex.wordpress.org/Function_Reference/get_post_custom

Instead of replace url directly in mysql db Try with Velvet Blues Update URLs wordpress plugin. It is safe to use this plugin. Because mysql db has some serialize values for post meta key. In that case your solution will not work proper. I had used velvet blue update urls plugin many times while migrating … Read more

See Marko’s link for InnoDB tables and the caveats. For MyISAM, there isn’t a dead easy “this is the offending query” solution. You should always start with a processlist. But be sure to include the full keyword so that the printed queries aren’t truncated: SHOW FULL PROCESSLIST; This will show you a list of all … Read more

Use the relational database for its intended purpose – create the separate table. The rich set of time/date functionality available in your SQL queries for this example, is well worth it. Sticking searchable/reportable data into a serialized column in the database should be the last resort not the first 😉

Because $pageposts is an array of post IDs, but you need to pass a post object to setup_postdata – pull the objects instead: SELECT $wpdb->posts.* [rest of query] Note that, since you’re not paginating, you don’t need SQL_CALC_FOUND_ROWS