Which Protocol Advertises WAN Edge Routes on the Service Side?

Why the Answer Is Correct

In a Cisco Catalyst SD-WAN deployment the WAN Edge holds the overlay routing table in OMP, and on the service side it peers with the LAN core, a CE router or a pair of distribution switches. Cisco's design and configuration material uses a BGP service-side peering session to push those OMP-derived prefixes into the local routing domain, because BGP carries the path attributes and filtering controls needed when many sites and VRFs are involved. OMP itself is modeled on BGP — as Faridtnx put it, "OMP itself is functionally very similar to BGP" — so the translation of OMP routes into a BGP advertisement is the natural, well-documented path. In service-provider and multi-tenant designs the service side is almost always eBGP, which is why BGP is the single option the exam expects when the stem says the WAN Edge advertises routes on the service side. That makes B the defensible verdict even though the source key was left blank.

Why the Other Options Are Wrong

OSPF is the strongest distractor and the source of nearly all the confusion here: hopatrickhk88 correctly quotes Cisco's rule that OSPF intra-area and interarea routes are redistributed into OMP automatically, but that rule governs routes entering the overlay, not routes being advertised out of it. msstanick draws the same line, noting that an OSPF network statement merely enables OSPF on an interface rather than advertising the WAN Edge's overlay prefixes. EIGRP can run on the service side in some releases, but it requires explicit redistribution both ways and is not the protocol Cisco highlights for WAN Edge route advertisement. IS-IS is not a supported service-side routing protocol in SD-WAN Manager feature templates, so it is eliminated outright.

Community Comment Notes

Several learners flag the wording rather than the technology — timtgh writes, "Bad question. Several protocols advertise routes on the service side." and Aldebeer notes "the question is not clear. it could be both OSPF and BGP." That ambiguity is real, but the exam still has one best answer, and both Faridtnx and msstanick land on BGP after reasoning through the direction of redistribution. Knowledge33 points to the Cisco unicast routing chapter, which is the right reference for the OMP redistribution behavior even though it supports the OSPF-instead-of-BGP argument only for the reverse direction. mikidvd51's frustration about question quality is understandable; when the stem is vague, answer the direction the vendor documents most explicitly — overlay routes going out to the service side.