[j-nsp] Centrally-Routed Bridging

[Cristian Cardoso]

Hello
Did anyone on the list have the EVPN-VXLAN scenario active with
Centrally-Routed Bridging?
Here in my scenario, I have two spines and 2 leafs and I am testing
this form of configuration, where the IRB interfaces work with the
virtual gateway function, for access redundancy, but I realized that
when restarting spine1, the active source of the virtual gateway ,
does not change from spine1 to spine2 generating network timeout.
When I restart spine1, the source of the virtual gateway on leaf does
not change, causing network timeout. The same timeout happens, on
VNI888, if I restart spine2.

Below is the default.switch table

Routing instance : default-switch
   Vlan                MAC                 MAC      Logical
    Active
   name                address             flags    interface
    source
   VNI4018             00:00:5e:00:02:01   DR       esi.1750
    05:00:00:fd:e9:00:00:0f:b2:00
   VNI4018             40:de:ad:7b:34:40   D        vtep.32769
    172.16.11.201
   VNI4018             40:de:ad:7b:66:40   D        vtep.32770
    172.16.11.202
   VNI4018             b4:96:91:70:db:84   D        vtep.32771
    172.16.11.203
   VNI4018             b4:96:91:71:25:c0   D        xe-0/0/10.0
   VNI888              00:00:5e:00:01:02   DR       esi.1748
    05:00:00:fd:e9:00:00:03:78:00
   VNI888              40:de:ad:7b:34:40   D        vtep.32769
    172.16.11.201
   VNI888              40:de:ad:7b:66:40   D        vtep.32770
    172.16.11.202
   VNI888              4a:7b:af:da:34:69   D        vtep.32771
    172.16.11.203
   VNI888              ae:cb:39:6b:a7:ee   D        xe-0/0/10.0

The source 05: 00: 00: fd: e9: 00: 00: 0f: b2: 00 is spine1 and the
source 05: 00: 00: fd: e9: 00: 00: 03: 78: 00 is spine2

Did anyone go through this?

Best regards