Which two design recommendations are most appropriate when OSPF is the data center core routing protocol? (Choose two.)

Question

Options

• ANever use passive interfaces.
• BUse NSSA areas from the core down.
• CUse totally stub areas to stop type 3 LSAs.
• DUse the lowest Ethernet interface IP address as the router ID.
• EManipulate the reference bandwidth.

• A
  5% (2)
• B
  81% (30)
• C
  3% (1)
• D
  11% (4)

Explanation

Using NSSA (Not-So-Stubby Areas) from the core down allows external routes to be redistributed into stub-like areas, reducing LSA flooding while still permitting limited external routing information - useful in data center designs where some external connectivity is needed at edge areas. Manipulating the reference bandwidth (choice E) is essential because OSPF's default reference bandwidth is 100 Mbps, which makes all interfaces of 100 Mbps or faster appear to have the same cost (1). In modern data centers with 1G, 10G, 40G, or 100G links, the reference bandwidth must be increased so OSPF can differentiate link costs correctly. Choice A is wrong - passive interfaces are recommended on interfaces with no OSPF neighbors (e.g., server-facing ports) to suppress unnecessary hellos. Choice C is wrong - totally stub areas block type 3 LSAs (inter-area summaries), which would be too restrictive for a data center core. Choice D is wrong - using the lowest Ethernet IP as the router ID is not a best practice; a stable loopback address should be used.

No community discussion yet for this question.