Symptom
for VPWS, when remote attachment circuit is brought down, if the local attachment circuit for the vpws has propagate remote-status CLI configured, the remote failure may not be propagated to local CE device.
Conditions
Xconnect between NCS5500 over EVPN
BGP RR on three routers.
BGP RRs can cause multiple remote EAD update/withdraws and can cause this inconsistent issue.
Versions affected:
6.0.1,6.0.2,6.1.1,6.1.12,6.1.2,6.1.21,6.1.22,6.1.3,6.1.31,6.1.32,6.1.33,6.1.4,6.1.42,6.1.45,6.2.1,6.2.2,6.2.25,6.2.3,6.3.1,6.3.15,6.3.2,6.3.3,6.4.0,6.4.1,6.4.2,6.4.3,6.5.1,6.5.15,6.5.2,6.5.25,6.5.26,6.5.28,6.5.29,6.5.3,6.5.90,6.5.92,6.5.93,6.6.1,6.6.11,6.6.12,6.6.2,6.6.25,6.6.3,6.6.4,6.7.1,6.7.2,6.7.3,6.7.4,7.0.0,7.0.1,7.0.11,7.0.12,7.0.14,7.0.2,7.0.90,7.1.1,7.1.15,7.1.2,7.1.25,7.1.3,7.2.0,7.2.1,7.2.12,7.2.2
Workaround
To prevent the remote EAD churn:
or the p2p evpn-vpws configuration, the p2p should use the same target and source number on both devices.
This will allow the evi/ead routes to collapse into 1 route in bgp and should help to address the flap as well.
However, it is still possible to hit the propagate remote-status not being forwarded to the CE if the local link went down prior to the remote failure.
