Analysis of OSPF external routes E1 and E2

Analysis of OSPF external routes E1 and E2 1. when ospf publishes an external route to a single asbr, 1.1 the equivalent path in the OSPF domain of a single ASBR re-publishes the rip route on the ospf of r2. When the release type is e2 or e1, there will be two routes on r5, respectively through r3 and r4 to 10.1.1.0 (because they are equal-cost routes, this part is relatively simple, not to mention) regarding the differences between ospf external routes e1 and e2 routes, the network is infinite. 1.2 when non-equivalent paths in the OSPF domain of a single ASBR are re-distributed on r2, the e2 route is released, the published metric is 20. On r5, only one route is displayed, and the route of the address reached through r4. After the lsdb library converges, there will eventually be a five-class lsa, And the metric of the received route is 20. r5 uses the spf algorithm to calculate the closest distance to the asbr. By computing, r5 selects r4 as the next hop route for the 10.1.1.0 network segment. Because the cost value in the ospf domain does not need to be added when the e2 route is placed in the route table. Therefore, the metric value of this route is 20 in the routing table. Originally, when I understood the e2 route, I always thought that the metric whose r5 reaches the 10.1.1.0 network segment does not consider the internal route, so it doesn't matter what the internal network is like, even if the path is not equivalent, it is also used as the two routes in the route table. It is actually an error that is unfamiliar with how ospf builds routes through the lsdb database. When the e1 route is re-distributed on r2, the published metric is 20. In the lsdb database of r5, there is only one lsa of Class 5, and the metric of the received route is 20, r5 uses the spf algorithm to calculate the closest distance to asbr. By computing, r5 selects r4 as the next hop route for the 10.1.1.0 network segment. Because e1 routes need to overwrite the cost value in the ospf domain when they are placed in the route table. Therefore, the metric value of this route is also 22 in the routing table. Conclusion: When the paths in the ospf domain are not equivalent, the routes of e1 and e2 are the same, the principle of selecting routes for ospf external routes e1 and e2 is essentially the same. Although e2 does not calculate the internal cost value, however, the router will still compare the size of the path value to the ASBR and select the optimal path. 2. when multiple asbrs publish external routes, ospf re-releases the rip route on the ospf of r2 and r5 with the Management Distance value of the 2.1 re-released routing protocol higher than that of the ospf route, the release type is e2, e1, and metric. At this time, there is only one route that reaches 10.1.1.0 in the entire ospf domain. Because when r2 or r5 re-releases the rip route to the ospf domain, it depends on the r2 and r5 who first receive the re-released route from the other party. If r2 receives the route first, in this case, the re-release of the 10.1.1.0 network segment fails on the r2 router. Because the re-release must follow two principles. First, the re-release route must exist in the route table (in the form of a route protocol to be re-released), that is, the ospf re-release rip route, in the routing table, the routes to be re-published must exist in the form of rip. 2. The re-release Router does not change its route table. In this case, r2 learns 10.1.1.0. There are two routing protocols, one is the r5 re-release route, and the other is the rip route, at this time, the route table deletes the Routes learned through rip. In this way, the re-release of R2. Because the re-release of r2 fails, ospf in the domain will converge to the re-release route sent by an asbr. There are infinite network differences between ospf external routes e1 and e2 routes. In fact, this is effective for re-release of any route, that is, when re-publishing, we need to consider the distance between different routing protocols. You only need to understand the re-release process. The preceding situation can be attributed to the fact that the distance value of the re-published routing protocol is higher than the distance value of the route for re-publishing other routing protocols. It's easy to understand. 2.2 The Management Distance value of the re-published routing protocol is lower than the Management Distance value of the ospf route. In this example, the routing table in the ospf domain is selected, l first, the router calculates and compares the metric of the target network based on the lsa received by the router, and selects a small metric lsa as the basis for the calculation route (E1 and E2 will not compare each other ). L second, when metric is the same, the distance calculated to the asbr is used as the basis for putting the data into the route table. Select the optimal route to the route table. In this case, the cost value in the domain is the same. All of the values are 1. If the release is an e2 route, the r3 and r4 routers will receive two 5-class LSAs from two different asbrs. In addition, the five types of lsa of e2 are released for both asbrs. r3 calculates that the metric value reaching the 10.1.1.0 network segment is 20 (excluding the cost value in the ospf domain ). Then r3 compares the cost values of two asbrs. After comparison, the cost values of the two asbrs are the same, therefore, there are two routes in the r3 route table to go to 10.1.1.0. There is no limit on the selection of ospf external routes e1 and e2 routes. Similarly, r4. R2 only has one static route in the routing table, and r5 only has one static route. If the release is based on the e1 route, the r3 and r4 routers will receive two 5-class LSAs from two different asbrs. However, the metric calculated by r3 is different from the preceding one. R3 uses the spf algorithm to calculate that the metric value of the external route of the two asbrs is 21. Then, r3 calculates the cost value of the internal path and finds that the path value of the two asbrs is the same. Therefore, the route table of r3 still has two routes, both of which learned the 10.1.1.0 network segment through r2 and r5. Similarly, r4. Both r2 and r5 have only one route. The cost value in the domain is different. The cost value between r2 and r3 is 64. If the release is an e2 route, r3 calculates two paths to reach the 10.1.1.0 Network Segment Based on the lsdb database, and metric is the same, both are 20 and are directed to two asbrs. At this time, r3 will calculate the Optimal Path to the two asbrs by performing the spf algorithm, and reach r2 to 3 and r5 to 1. Therefore, r3 will select r5 as the next hop route for the 10.1.1.0 network segment. The route table of R3 has only one route. The calculation result of r4 shows that the metric value reaching the 10.1.1.0 network segment is 20 (excluding the cost value in the ospf domain ). Then, r4 compares the cost values of the two asbrs. After comparison, the cost values of the two asbrs are the same, and the r4 route table has two routes to the 10.1.1.0 route. R2 only has one static route in the routing table, and r5 only has one static route. If the release is based on the e1 route, R3 uses the spf algorithm to calculate that there are two metric segments reaching 10.1.1.0. One is 23 and the other is 21. Then, it calculates that the ADV with the metric value 21 is r5, r3 selects r5 as the egress asbr of the 10.1.1.0 network segment. Since r3 performs the spf algorithm when calculating the metric value, the path that r3 reaches r5 is also the optimal path in the domain. The router table of R3 has only one route. Similarly, you can calculate r4. There are still two routes in the R4 route table. Both r2 and r5 have only one route. If r5 is not re-released and r2 is re-released, there is only one route table for r3, and r5 is used as the next hop. You can also think about the situation if the metric values on both sides are different, and what if the release type is different. In fact, everything is clear as long as you master the ospf principle for selecting external routes. Note: When both E1 and E2 exist, the E1 route is selected first, and the metric values of E1 and E2 are not compared.