Hello everyone,
Today, I'm going to introduce the discussion about six router networking in the new HCIE lab.
Topic
1. How to speed up the convergence time when OSPF is running between aggregation and access and it is an Ethernet link?
2. The downstream of R5 and R6 are connected to some PCs through Layer 2 switches, and these PCs continue to receive useless OSPF protocol packets. Please give an example of a solution.
3. If OSPF is running, OSPF area 0 is between the core and the aggregation, and OSPF area 1 is between the aggregation and access. In which area should the link between the aggregation routers be placed? Please analyze.
4. In this network, under normal circumstances, the access layer routers can shunt each other from left to right. Now it is found that the traffic of all access devices at the access layer is upstream from the left. Please list the possible reasons.
Question 1
How to speed up the convergence time when OSPF is running between aggregation and access and it is an Ethernet link?
1. Modify the network type to P2P to save the time of DR/BDR election.
2. OSPF and BFD linkage. BFD is associated with routing protocols, once the link fails, the fastness of BFD can speed up the convergence speed of routing protocols.
3. Modify the time of each timer
Use the ospf timer hello command to modify the hello packet sending interval to speed up the perception of neighbor status.
Use the spf-schedule-interval command to modify the SPF calculation interval and shorten the running time of the two SPF algorithms to speed up the convergence.
Use the lsa-originate-interval command to set the LSA update interval. You can set the interval to 0 to cancel the LSA update interval, and it can be published to the network immediately after the topology changes.
Use the lsa-arrival-interval command to set the LSA receiving time interval so that it can quickly detect changes in the network.
Use the ospf trans-delay interval command to set the delay time of LSA transmission to speed up the transmission speed of LSA.
In the MA network, without modifying the network type, use the smart-discover command to speed up the DR/BDR election (send the hello message when the neighbor status changes, and there is no need to wait for the timer).
Question 2
The downstream of R5 and R6 are connected to some PCs through Layer 2 switches, and these PCs continue to receive useless OSPF protocol packets. Please give an example of a solution.
To prevent devices on other networks from obtaining OSPF routing information and prevent the local device from receiving routing updates advertised by other devices on the network, run the silent-interface command to disable the interface from receiving and sending OSPF packets. By this solves the problem that a downstream device continuously receives useless OSPF protocol packets. After an interface is disabled from receiving and sending OSPF packets, the direct routes of the interface can still be advertised. However, Hello packets of the interface are blocked and neighbor relationships cannot be established on the interface. This enhances OSPF networking adaptability and reduces system resource consumption.
The interface connected to the Layer 2 switch does not need to be added to the OSPF process, and the route is advertised by importing direct routes. At this time, the PC will not receive OSPF protocol packets.
Question 3
If OSPF is running, OSPF area 0 is between the core and the aggregation, and OSPF area 1 is between the aggregation and access. In which area should the link between the aggregation routers be placed? Please analyze.
It is more appropriate to place the link between R3 and R4 in area 0 to provide a backup path for the backbone area. Assume that the same link between R3 and R4 is in area 1 (non-backbone area) if the link between R1 and R2 is disconnected, In this case, the backbone area is divided, and R1-R3 is in area 0 and R2-R4 is in area 0. In this case, normal communication is affected. To ensure the continuity of the backbone area, it is recommended that the link between aggregation routers be placed in area 0.
The link between R3 and R4 can also be placed in the non-backbone area. However, a virtual link (Vlink) needs to be established between R3 and R4 because the backbone area may be split and traffic cannot be normally accessed. Logically add a backup path to the backbone area. If the link between R1 and R2 is down, you can access the destination address through the path between R1 and R3 and R4 and R2.
R3 and R4 use two sub-interfaces. One sub-interface is deployed in area 0, and the other sub-interface is deployed in the non-backbone area for access and aggregation. This prevents discontinuity of the backbone area caused by a link fault between R1 and R2, ensuring service access continuity.
Question 4
In this network, under normal circumstances, the access layer routers can shunt each other from left to right. Now it is found that the traffic of all access devices at the access layer is upstream from the left. Please list the possible reasons.
If the link between R4 and R6 is congested or faulty, traffic cannot be load-balanced between the two links and can only be accessed through the left uplink. As a result, all traffic is sent out from the left.
The stub-router on-startup command is run on R4. When a network fault occurs or the device restarts, the maximum cost of 65535 is advertised within the recovery time (configured by the administrator), affecting route selection and preventing the downstream device from forwarding traffic upstream through R4, which breaks load balancing.
That is all I want to share with you!
