Hello everyone,
Have you encountered a question about dividing OSPF areas in the HCIE exam? Today I will share with you the solution to this problem.
OSPF area division
Design OSPF areas based on the following topology to ensure that the routing table of R7 is not affected when the network changes.
OSPF areas are divided as follows:
1. A backbone area (area 0) is established between the core layer and the aggregation layer. Including the direct links between R3 and R4 to prevent the backbone area from being split.
2. A non-backbone area (area 1) is established between the aggregation layer and the access layer.
3. To prevent the routing table of R7 from flapping due to topology changes, set the area where R7 resides as a special area. This ensures that R7 uses only the default route to access inter-area routes and external routes to other areas. Therefore, the routing table of R7 is not affected when the network changes.
Special areas can be set to Totally Stub and Totally NSSA.
If external routes are not required in the area where R7 resides, set the area where R7 resides to a totally stub area. Otherwise, set the area where R7 is located to a totally NSSA area.
Note:
The area where R7 resides cannot be set to the NSSA or Stub area because type-3 LSAs can exist in the NSSA and Stub area. When routes in other areas change, the routing tables of the NSSA and Stub area are affected.
4. Route summarization can also be performed on the network. When the specific route disappears, the summarization route on the R7 device still exists, and the route entries are not affected.
Egress redundancy
There are two ways to ask this question, but the test points are basically the same.
Question 1: R1 and R2 each have an uplink connected to the public network. Redirection to the next hop is required. When the uplink of R1 is faulty, services are switched to R2.
Question 2: R1 and R2 use static or default routes to access external networks. When the network is normal, R1 functions as the primary link, and R2 functions as the backup link.
Answer
In normal cases, R1 and R2 function as egress routers on the campus network and use default routes to access public IP addresses. When configuring the default route to the ISP on R1 and R2, need to associate BFD or NQA. This ensures that the incorrect static route is immediately deleted from the routing table when the uplink fails. In addition, OSPF delivers default routes for other devices to access the public network.
Generally, the next hop of a route is the carrier's device. To preferentially access the public network through R1, run the default-route-advertise cost 10 command to set the cost to 10 when configuring a default route on R1, and run the default-route-advertise cost 20 command to sets the cost value to 20 when configuring a default route on R2. In this way, R1 is preferentially used to access the external network. When the uplink of R1 is faulty, the downlink device accesses the external network through the default route of R2.
That is all I want to share with you! I hope it helps!
