Description: Two NE40s are connected through an Eth-Trunk interface that is formed when five 100 Mbit/s optical interfaces are bundled together. OSPF is running on the Eth-Trunk interface. Based on the test result, NE40-1 fails to ping the address of the interface on NE40-2. However, the two NE40s can ping each other's loopback address. Additionally, the OSPF neighbor relationship can be set up between the two NE40s.
Since the OSPF neighbor relationship can be set up and the two NE40s can ping each other's loopback address, the logical link between the two NE40s is not faulty. Because five physical links are aggregated, one of the physical links may go faulty.
Check the optical interfaces on the two NE40s that are bundled together and find that one of the optical interfaces on NE40-2 goes Down. However, all the optical interfaces on NE40-1 are Up.
The cause for the problem is that the optical module on the interface that is Down on NE40-1 goes faulty. As a result, NE40-2 cannot receive optical signals.
the optical module on the interface that is down goes faulty.
Shut down the interface that is Down on NE40-1, and the problem is resolved.
Resolve the unidirectional link problem (an optional solution).
Configure self-negotiation mode on the optical interface.
Use OAM to detect the unidirectional link.
Use BFD to detect the unidirectional link. If there is a Layer 2 interface, configure multicast BFD.
Note: A Huawei switch has restrictions on BFD. In a version earlier than V100R006, the interval at which BFD control packets are transmitted can only be set to 1000 ms; otherwise, BFD session flapping occurs. In V100R006, the number of BFD control packets that are transmitted at the interval of 100 ms is limited. Therefore, confirm the operation with the Huawei technical support personnel before you use BFD.
Use the static Link Aggregation Control Protocol (LACP).
Use the Device Link Detection Protocol (DLDP). Confirm whether the device support this feature or not.