A tool for using an interconnected network of systems to create an index for a database table. An index advisor on a primary server recommends one or more indexes to improve efficiency. While resources of the primary server are being used by various queries and processes, the primary server sends the recommendations to a secondary server (with available resources) so that the recommended indexes may be built in parallel with the processes executing on the primary server. The secondary server builds the recommended indexes based on its own copies of the database tables. The secondary server sends the built indexes to the primary server, where the primary server must reconcile the indexes with any changes that took place to the database tables subsequent to the replication of the tables on the secondary server. The primary server makes the associations between the new indexes and the tables they were built for.