Standard OSPF behaviour
According to the OSPF definition, for a router to behave as an OSPF ABR router, it must have at least two configured areas: one would be the backbone area 0, and the other a specific area used for a group of routers which, due to route summarization or administrative clarity, are placed into a non-backbone area, for example area 10.
If we further say that area 10 is configured as a stub area, by default the ABR router will generate a default route 0.0.0.0/0 for area 10.
This is typical behavior in networks where traffic routing is controlled by the OSPF protocol. How does this look like in Cisco SD-WAN?
At first glance, the answer would be that there is no and should be no difference. OSPF as such is not one of the Cisco SD-WAN protocols like OMP, and in general it is independent of SD-WAN and is not even aware of the protocols that are part of the SD-WAN setup.
However, OSPF behavior in a Cisco SD-WAN environment is somewhat different from what I expected to be.
SD-WAN use case:
The typical customer scenario involves modernizing a legacy WAN environment using SD-WAN technology, where traditional routers are replaced with Cisco cEdge devices connected to one or more WAN links. This enables the later implementation of AAR policies and application-aware traffic steering across multiple WAN transports.
On the LAN side in a branch office, the customer may have a pure Layer2 design, where the cEdge acts as the default gateway for local users, or a Layer3 design, where a Layer3 switch or router provides LAN connectivity. In the Layer3 case, a dynamic routing protocol, typically OSPF is established between the cEdge and the Layer3 device in the LAN.
In some deployments, the OSPF area within the LAN is an extension of backbone area 0, while in others it is a non-backbone area, such as area 10, and it might be configured as a stub area. This is often done because LAN Layer3 devices do not need to maintain external routes in their routing table, and it also simplifies route summarization.
In the HQ, where the data center is also located, the legacy Layer3 router in the LAN has one interface in area 0 towards the data center resources and another towards HQ users in area 10, taking on the ABR role in OSPF, which also includes generating a default route for area 10 by default. There are some more legacy routers in area10.
These remaining routers in the LAN do not participate in area 0, and their connectivity to the rest of the network is provided through the ABR router.
The new cEdge device on the LAN side (referred to as a VPN service in Cisco SD-WAN terminology) in the HQ is connected only to area 10 and does not have any interfaces in another OSPF area for VPN services, the same as remaining legacy routers. Connections towards SD-WAN controllers are established over WAN interface, called VPN0, and it belongs to the underlay network.
At first glance, the design does not appear to introduce anything unusual or any limitations compared to the legacy setup. However, in practice, this is not entirely the case.
Result:
Once the cEdge router is configured via vManage as part of OSPF area 10, it starts to behave like ABR router by generating a default route for area 10. In a traditional OSPF design, this behavior would typically imply connectivity to area 0 through at least one interface; however, this is not the case in this scenario.
The cEdge ignores the default route received from the legacy ABR, which is expected behavior when a default route is received from another ABR via a non-backbone area. If the OSPF database is checked on other legacy routers in area 10, cEdge will be listed as an border area router, show ip ospf border-routers.
This behavior indicates that the cEdge introduces an abstraction where it effectively assumes a backbone-like context internally. When an additional OSPF interface is configured in a non-backbone area, it act as an ABR and originate a default route accordingly. As a result, it can lead to suboptimal routing in area 10, where remaining routers in area 10 receive multiple default route sources from both the cEdge and the legacy ABR.
Workaround:
Therefore, it is recommended to avoid configuring multiple OSPF areas in the traditional sense and instead design the network as an extension of area 0. This approach will also require adjustments in the legacy network, particularly for service VPNs and LAN segments.
What does Cisco say:
For such OSPF behavior in SD-WAN environments, I am not aware of any Cisco document that clearly describes this specific behavior (if you have one, please share it), except for the general requirement that OSPF must be configured in Area 0, without further explanation of the underlying reasons.
This raises an interesting question: why does the vManage GUI even allow the configuration of OSPF areas other than Area 0, and why does it also allow stub area configuration?
Cisco Press about OSPF setup:
https://www.cisco.com/c/en/us/td/docs/routers/sdwan/configuration/routing/vEdge-20-x/routing-book/m-unicast-routing.html