Multiple servers deployed at a single location

You can achieve high availability and fault tolerance (HA/FT) by deploying multiple peers and by utilizing the built-in features of the BMC Atrium Orchestrator components.

• To achieve HA, a grid must survive the failure of one or more of its components and continue to execute workflows. The minimum components required for a grid are a repository, BMC Atrium Single Sign-On, and CDP.
• To achieve HA/FT, you must configure a topology that guarantees the survival of the minimum components required for the grid to execute workflows. You can configure this setup by deploying redundant components, using built-in failover features, and by installing an external database.

In this configuration, the repository, BMC Atrium Single Sign-On, and the CDP with any optional components reside on separate servers as shown in the following figure. In this setup, the CDP handles all of the grid administration and workflow execution for the grid. The BMC Atrium Orchestrator Development Studio component is installed on a separate computer.

Because the CDP contains the largest number of deployable components and requires a dedicated server, BMC recommends this basic configuration for a production environment.

• You must install the CDP in a Tomcat instance that is separate from BMC Atrium Single Sign-On and the repository applications.
• As an option, you can install the graphing server and BMC Atrium Orchestrator Operator Control Panel components with the CDP or on separate servers.

In the following figure, each server contains a separate Tomcat instance.

The following topics describe other deployment use cases for multiple servers deployed at a single location:

• Basic production environment
• High availability for the CDP at a single location
• High availability and fault tolerance for multiple servers at a single location