Where are path selections made for an SD-WAN router?

Question

Options

• Aphysical interface card
• Blocal packet forwarding engine
• Clocal routing engine
• Dcentralized controller

• A
  5% (1)
• B
  5% (1)
• C
  73% (16)
• D
  18% (4)

Explanation

In SD-WAN, path selection occurs on the local routing engine of the edge router because each device continuously monitors path quality metrics (latency, jitter, packet loss) and applies locally-held policies to select the best WAN link in real time - no round-trip to a controller is needed for each forwarding decision.

Why the distractors are wrong:

• A (physical interface card) - NICs handle physical layer transmission; they move bits, they don't make routing decisions.
• B (local packet forwarding engine) - the forwarding engine executes decisions (looks up and forwards packets), but does not make path selection choices; that logic sits in the routing engine above it.
• D (centralized controller) - this is the classic trap. The controller distributes policies and route information (e.g., via OMP in Cisco SD-WAN), but the actual real-time path selection is delegated to the local edge device so failover can happen in milliseconds without a controller round-trip.

Memory tip: Split it as "controller = policy pusher, routing engine = policy applier." The controller sets the rules once; the local routing engine makes path decisions every moment traffic flows.

No community discussion yet for this question.