The article describes the life of a Full Text Search query as it originates from a client and is processed through a distributed Couchbase system. The search service is configured to partition the index across multiple servers, allowing for efficient querying and indexing of large amounts of data. The article explains how the coordinating node receives the search request, looks up the index, obtains the plan, sets up a list of eligible servers, dispatches internal requests, waits for responses, and merges results from local and remote partitions before shipping the response back to the client. The article also discusses the contents of the search response, including the "status" field, which contains information on the total number of partitions, successful and failed responses, and errors that occurred. Additionally, it highlights the importance of investigating cluster health and resolving issues with unresponsive or failing index partitions.