Managed fault tolerance in TIBCO ActiveMatrix BusinessWorks establishes an active-standby configuration. It requires the group persistence mode, which uses a database for both process state persistence and for coordinating among AppNodes to elect a single active leader. This coordination inherently makes the AppNodes "aware" of each other's presence and status.

Non-managed fault tolerance creates an active-active configuration where all AppNodes run concurrently and independently. An external load balancer is required for failover. This model requires the datastore persistence mode for a shared state, but the AppNodes themselves do not coordinate and are "unaware" of other instances in the AppSpace.

A. The persistence modes are incorrect. Memory persistence is not for fault tolerance, and group persistence is for managed, not non-managed, fault tolerance.

C. Both fault tolerance models support checkpointing. It is the fundamental mechanism for saving process state to enable recovery after a failure in either model.

D. This describes the outcome (single vs. multiple active engines) but not the core configuration requirement or architectural principle as accurately as options B and E.

1. TIBCO BusinessWorks™ Container Edition Application Development, Release 2.8, "Fault Tolerance" chapter.

Section: Managed Fault Tolerance: "The persistence mode must be set to group." and "The AppNodes are aware of each other and one of the AppNodes is elected as the primary (active) node." This directly supports options B and E.

Section: Non-managed Fault Tolerance: "The persistence mode must be set to datastore." and "The AppNodes are unaware of each other." This also directly supports options B and E.

2. TIBCO BusinessWorks™ Container Edition Application Monitoring and Troubleshooting, Release 2.8, "Engine Persistence Modes" section.

Subsection: Group: "The group engine persistence mode is used for managed fault tolerance." This confirms the requirement for option B.

Subsection: Datastore: "The datastore engine persistence mode is used for non-managed fault tolerance." This also confirms the requirement for option B.