A newly deployed Microwave networking consists of RTN900 series. After the service configuration finished, and found that some of the Microwave sites have service jitter.
1. Firstly we chek the NMS, and we found that the NEs can be normally managed by NMS, in another word, the DCN is normal. and we check the microwave performance, there isn't any unstablity, the Microwave link is normal.
2. the affected service is distributed on different NEs, it is not so possible that all the hardward fail at the same time.
3. Check the configuration. first according to the network topology, we check the affected service LSP, and we found that all the affected service gets through the Pop site, a RTN900 NE we call NE A, where all the service will be finally convergenced on NE A. And then we can focus our attention on NE A.
4. Accoring to the reported alarm MPLS_TUNNEL_LOCV by NMS, it is a problem occurs on the tunnels. so we check the tunnel configuration on NE A. Becasue NE A is a convergecy NE, and we found that the tunnels here have the same outgoing interface but different different NEXT HOP IP. After careful check on the configuration, one incorrect configuration is found, we found that the opposite interface is incorrect, in another word, the NEXT HOP IP i s incorrect, this makes the MAC self-learning jitters, and the ARP records are frequency delelted and learnt, which cause the service jitter.
5. Delete the incorrect configuration and re-configure the tunnel correctly and everything is normal.
1. Microwave link problem. Microwave link is unstable, cause too much bit errors which make the service blinks.
2. Hardware fault, such as the packet switch unit or Ethernet card failure which cause the bit errors.
3. Configuration mistakes.
The service forward need the destination MAC address learnt by ARP, when the ARP items a repeatedly delete and learnt, which causes the service jitter, usually we suggest the end to end configuration instead of the single station configuration, which can avoid the NEXT HOP IP mistakes.