You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Originally posted by ardenisov January 21, 2025
Hello!
When enable debug bgp updates I see a lot of messages like below for every bgp peer:
2025/01/17 12:11:02 BGP: [TPJB2-AQ2TG] Processing EVPN MAC interface change on peer 10.37.96.8
2025/01/17 12:11:02 BGP: [TPJB2-AQ2TG] Processing EVPN MAC interface change on peer 10.37.96.104
2025/01/17 12:11:02 BGP: [TPJB2-AQ2TG] Processing EVPN MAC interface change on peer 10.37.128.62
2025/01/17 12:11:02 BGP: [TPJB2-AQ2TG] Processing EVPN MAC interface change on peer 10.37.128.126
2025/01/17 12:11:02 BGP: [TPJB2-AQ2TG] Processing EVPN MAC interface change on peer 10.37.160.62
2025/01/17 12:11:02 BGP: [TPJB2-AQ2TG] Processing EVPN MAC interface change on peer 10.37.160.126
The strangest thing is that EVPN AF is not enabled on host, all peers direct and IPV4 only.
Also the state of host interfaces and each peer is stable for weeks.
Sometimes on host flaps tap interfaces, when VMs redeployed.
But taps don't participate in bgp and not mentioned in FRR config.
Could some one help me to investigate root cause of this messages?
frr version 9.1.2
UPD:
I found, that message of "Processing EVPN MAC interface change on peer" triggers every time tap interface created/deleted and for all BGP peers.
As I understand in the code, the message of "Processing EVPN MAC interface change on peer" should be triggered only if peer has soft reconfiguration inbound configured for EVPN address family, but in my case all peers IPv4 only:
Seems that CHECK_FLAG should return false but obviously it return true for non EVPN peers which have soft-reconfiguration-inbound configured.
BTW In my opinion the structure of message "Processing EVPN MAC interface change on peer" incorrect, because bgp peer has nothing to do with interface which MAC changed.
The text was updated successfully, but these errors were encountered:
Discussed in #17890
Originally posted by ardenisov January 21, 2025
Hello!
When enable debug bgp updates I see a lot of messages like below for every bgp peer:
The strangest thing is that EVPN AF is not enabled on host, all peers direct and IPV4 only.
Also the state of host interfaces and each peer is stable for weeks.
Sometimes on host flaps tap interfaces, when VMs redeployed.
But taps don't participate in bgp and not mentioned in FRR config.
Could some one help me to investigate root cause of this messages?
frr version 9.1.2
UPD:
I found, that message of "Processing EVPN MAC interface change on peer" triggers every time tap interface created/deleted and for all BGP peers.
As I understand in the code, the message of "Processing EVPN MAC interface change on peer" should be triggered only if peer has soft reconfiguration inbound configured for EVPN address family, but in my case all peers IPv4 only:
Example of bgp peer config:
Example of bgp peer state:
Seems that CHECK_FLAG should return false but obviously it return true for non EVPN peers which have soft-reconfiguration-inbound configured.
BTW In my opinion the structure of message "Processing EVPN MAC interface change on peer" incorrect, because bgp peer has nothing to do with interface which MAC changed.
The text was updated successfully, but these errors were encountered: