As shown in the following networking diagram, the NE40E and the UMG are directly connected and a BFD session is set up between the two devices. The BFD session is associated with the interface status using the process-interface-status command. BFD Session Flapping Occurs during peak hours on the network every day.
1. The on-site Huawei technical support personnel check the BFD session flapping log "BFD/4/STACHG_TODWN" and find that DetectDown is displayed in the Diagnostic field. The BFD session goes Down on the NE40E because the NE40E fails to receive BFD packets within the specified detection time.
2. The BFD entries on the NE40E are correct. However, BFD for PIS does not respond to the Layer 3 protocol change by delivering a new entry.
3. BFD session flapping does not sustain for a long time. Its occurrence is related to the number of services. If a large number of services are transmitted on the network, BFD session flapping is likely to occur. The BFD-enabled NE40E sends packets logically irrespective of the number of services. Meanwhile, the interface is not congested.
4. The UMG processes BFD packets using software. The on-site Huawei technical support personnel replicate the problem in the laboratory and capture the packets for analysis. Based on the analysis, the fault occurs on the UMG.
One-way audio no audio (no-way audio) is configured on the UMG. The detection mechanism has a bug, which affects the BFD detection performance.
Remove the bug on the UMG that affects the BFD performance.