The task of implementing a Guided Search application requires two separate environments, one for staging and one for production:
In a staging environment, you create, preview, and test a Guided Search application.
A production environment supports a live Guided Search application that serves end-user search and navigation queries.
The following diagram illustrates the workflow in staging and production environments:
This diagram introduces the differences between staging and production environments and explains how they relate to each other:
The left half of the diagram describes the data and configuration workflow within a staging environment. This workflow takes place when you provision and initialize the application to the EAC, either with the Deployment Template or by using Oracle Commerce Workbench, and then run the baseline or partial update scripts. Next, the data is prepared for indexing and is processed by the MDEX Engine. The MDEX Engine receives queries from the front-end application and returns results to the front-end application on the Application server.
The right half of the diagram describes the data and configuration workflow within a production environment. This workflow takes place when you run updates on production servers and also enable the front-end application to send user queries to the MDEX Engine server for processing. Note that the production environment does not include the Workbench, because the configuration files and operational settings in a production environment MDEX Engine are replicated from the staging environment.
Each environment has a dedicated Assembler that gathers and prepares content for rendering by your front-end application.
The arrows in this diagram that connect the staging and production environments indicate where staging data is promoted to. For information about how to promote your staging data, see Promoting application content to a production environment.
Most production environments use two or more servers and one load balancer. As system demand increases, the number of servers necessary in the implementation increases. Demand may take the form of time to crawl source data, frequent source data updates, faster query throughput, faster response time under increasing load, and so on. Several of the most common implementation architectures are described in the following sections.