More tests …
Neither hwmp-prep-lifetime nor ageing-time changed a thing but I saw something that could be a bug.
Aging seems to work differently between ARP/switchport MAC table and Mesh-learned MAC.
For ARP and switch port MAC, an entry is purged after aging, that is that the association ARP/IP or switchport MAC has been inactive for a period of time. The Mesh-learned MAC, this seems to be from the first occurrence of the learning.
On R1 (192.168.100.1, 18:FD:74:FA:21:FD), I have a continuous ping to R2 (192.168.100.2, 18:FD:74:FA:22:34).
At first, everything is fine and on R2, bridge host and mesh fdb show that the entries were just learnt.
> /interface bridge host print
Flags: X - disabled, I - invalid, D - dynamic, L - local, E - external
# MAC-ADDRESS VID ON-INTERFACE BRIDGE AGE
9 D 18:FD:74:FA:21:FD 10 mesh.1 bridge 0s
10 DL 18:FD:74:FA:22:34 10 bridge bridge
> /interface mesh fdb print detail
Flags: A - active, R - root
mesh=inter-ap type=larval mac-address=18:FD:74:FA:21:FD lifetime=0s age=3s metric=0 seq-number=0
A mesh=inter-ap type=local mac-address=18:FD:74:FA:22:34 age=3s metric=0 seq-number=22
After a few moments, the MAC on the bridge has not aged however the entry in the mesh FDB has. Additionally, the local entry for 18:FD:74:FA:22:34 has also aged.
> /interface bridge host print
Flags: X - disabled, I - invalid, D - dynamic, L - local, E - external
# MAC-ADDRESS VID ON-INTERFACE BRIDGE AGE
9 D 18:FD:74:FA:21:FD 10 mesh.1 bridge 0s
10 DL 18:FD:74:FA:22:34 10 bridge bridge
> /interface mesh fdb print detail
Flags: A - active, R - root
mesh=inter-ap type=larval mac-address=18:FD:74:FA:21:FD lifetime=0s age=1m58s metric=0 seq-number=0
A mesh=inter-ap type=local mac-address=18:FD:74:FA:22:34 age=1m58s metric=0 seq-number=22
When the ping on R1 times out, the entry for the local MAC on R2 shows as “unknown”, even though there were frames sent every second.
/interface mesh fdb print detail
Flags: A - active, R - root
AR mesh=inter-ap type=neighbor mac-address=18:FD:74:FA:22:00 on-interface=wds4 age=8h55m37s metric=50 seq-number=2860
mesh=inter-ap type=unknown mac-address=18:FD:74:FA:22:34 lifetime=0s age=8s metric=0 seq-number=0
Showing the FDB on R1 and R2 shows that
R1: inter-ap unknown 18:FD:74:FA:22:34 0s 5s
R2: inter-ap unknown 18:FD:74:FA:22:34 0s 7s
So the current questions:
- How is the aging defined for the FDB entries?
- Why does a local entry go into the “unknown” state?
- Am I doing something wrong in my configuration?