Configuring an Oracle Endeca Workbench deployment in a preview environment

This deployment approach requires a standalone Deployment Template installation, which runs its own baseline and partial index updates.

1. Remove the step that uploads post-Forge dimensions to Oracle Endeca Workbench during the baseline update. This step should only be performed in production, to ensure that new dimension values are only generated in the production environment and that all configuration created in Oracle Endeca Workbench is based on dimension values in the production environment. The following lines should be removed from the preview environment's baseline update script.

   ConfigManager.cleanDirs();
   Forge.getPostForgeDimensions();
   ConfigManager.updateWsDimensions();
   

2. In the separate production environment, the following configuration is required:
   1. Update AppConfig.xml to specify the host and port of Oracle Endeca Workbench in the preview environment.
   2. As these deployments may end up on different sides of a production firewall, this deployment may also require opening the Oracle Endeca Workbench port for communication through the firewall. This communication goes in both directions, as the production environment needs to retrieve files from Oracle Endeca Workbench and also update dimension values in the Workbench.
3. In addition to these configuration changes, it may be necessary to keep the preview environment up to date with the latest data and configuration changes applied to production. The following approaches may be taken to fulfill this requirement.
   • Deploy the production index on the preview server. Each time a new index is built by the production baseline update process, the index can be copied to the preview environment and applied to the preview Dgraph(s). This approach may require deploying partial updates onto the preview server as they are processed in production, if it is important for Oracle Endeca Workbench to see these partial updates while previewing.
   • Deploy production data and configuration on the preview server. Each time new data is made available for processing in production or new configuration files are deployed to production, the same data and configuration should be made available to the preview environment. With this approach, the preview environment runs baseline and partial updates to process the same data as production, ensuring that the same index is built in the preview environment just after that index is built or updated in production. This approach may also require copying the Forge autogen state file from the production environment to the preview environment, so that the dimension values used in the preview index are identical to those used in production. To ensure that no new values are generated in the preview environment, the --noAutoGen may be added to the Forge processes in the preview environment.