The case that the BFD session between S6800 and CS equipment could not UP
- 0 Followed
- 0Collected ，1429Browsed
Network Topology
NULL
Problem Description
The 6800 of our company is connected with CS N3K, and the BFD session cannot up noemally.We can see that our company has transfered messages, but it has not received any response.
< H6800HF>dis bfd session verbose | be 1/0/9
Session State: Down Interface: Ten-GigabitEthernet1/0/9.
440
Min Tx Inter: 100ms Act Tx Inter: 1000ms
Min Rx Inter: 100ms Detect Inter: 5000ms
Rx Count: 0 Tx Count: 57518
Connect Type: Direct Running Up for: 00:00:00
Hold Time: 0ms Auth mode: None
Detect Mode: Async Slot: 1
Protocol: BGP
Version: 1
Diag Info: No Diagnostic
< H6800HF >dis bfd session verbose | be 1/0/9
Session State: Down Interface: Ten-GigabitEthernet1/0/9.
440
Min Tx Inter: 100ms Act Tx Inter: 1000ms
Min Rx Inter: 100ms Detect Inter: 5000ms
Rx Count: 0 Tx Count: 57547
Connect Type: Direct Running Up for: 00:00:00
Hold Time: 0ms Auth mode: None
Detect Mode: Async Slot: 1
Protocol: BGP
Version: 1
Diag Info: No Diagnostic
Process Analysis
The switch checks the message, and when a CS device sends a message with extra VLAN.TRAILE information((00:00:00:00:00:07: a1:20:00:07: a1:20:00:07: a1:20), our device will treat such a message as an error message and discard it.
Currently, there is no avoidance method for this version of the device, and it is suggested that the opposite end avoid generating messages with such trailer information.
This kind of strict check exists in all R26** versions, but not in R2702 versions and above.
Solution
So there are two ways to avoid it
1.Our equipment has been upgraded to R2702 and newer versions.
2.CS devices don"t carry extra data like VLAN and TRAILE