The correct answers are A and C. In an SRX chassis cluster, redundant Ethernet interfaces are configured as reth interfaces. Juniper defines a reth interface as a pseudointerface that includes at least one physical interface from each node in the cluster. Therefore, assigning a physical interface from node0 and a physical interface from node1 to the same reth interface is mandatory for redundant Ethernet operation. Juniper also states that before configuring chassis cluster redundant Ethernet interfaces, you must set the number of redundant Ethernet interfaces.

Option C maps to the required reth-count configuration under the chassis cluster hierarchy. Without defining the number of reth interfaces, Junos does not know how many redundant Ethernet pseudointerfaces are available for the cluster configuration. Option B is wrong because setting a retry interval is not a required step for creating a reth interface or redundancy group. Option D is wrong because heartbeat behavior belongs to cluster control-link monitoring and chassis-cluster node health, not to the required configuration steps for Ethernet reth membership. The required design steps are: define reth capacity, create/configure the reth interface, assign child physical links from both nodes, and associate the reth with the proper redundancy group. Reference topics: HA Clustering, redundant Ethernet interfaces, reth-count, reth child interfaces, redundancy groups.