An SAP solutions architect is designing an SAP HANA scale-out architecture for SAP Business Warehouse (SAP BW) on SAP HANA on AWS. The SAP solutions architect identifies the design as a three-node sca

Question

Options

• ACreate a cluster placement group. Launch the instances into the cluster placement group.
• BCreate a spread placement group. Launch the instances into the spread placement group.
• CCreate a partition placement group. Launch the instances into the partition placement group.
• DBased on the operating system version, verify that enhanced networking is enabled on all the
• ESwitch to a different instance family that provides network throughput that is greater than 25 Gbps.

• A
  76% (34)
• B
  4% (2)
• C
  7% (3)
• E
  13% (6)

Explanation

Two steps are required for optimal inter-node network performance in an SAP HANA scale-out cluster. First (A), a cluster placement group physically co-locates instances on the same high-bisection-bandwidth networking fabric within an Availability Zone, providing the lowest possible inter-instance latency and highest throughput - critical for SAP HANA node-to-node communication during distributed queries and delta merge operations. Second (D), enhanced networking must be verified to be enabled on all nodes; it uses SR-IOV to provide higher packets-per-second, lower latency, and higher bandwidth compared to traditional virtualized networking. A spread placement group (B) maximizes fault isolation by placing instances on separate hardware, which sacrifices network proximity. A partition placement group (C) is designed for distributed/replicated workloads needing fault isolation, not tight coupling. Switching instance families (E) is unnecessary since x1e.32xlarge already supports high network throughput.

No community discussion yet for this question.