We have determined the problem at the level of the vPC of one of the switches managing these networks. We have deactivated the traffic on the switch which is causing this problem and are working with cisco for the troubleshooting.
According to our monitoring, the routing is OK for these networks. It is running on a single switch. The troubleshooting with cisco is ongoing.
The FO IPs towards these 2 networks were impacted by the problem as well. We have modified the routing in order to fix the problem.
Now that the situation is stable, we will try to understand what is the problem on the switch exactly. Cisco confirm the problem but did not yet identify the causes. The offending switch is still bypassed. We suspect a dysfunctioning at the level of the vPC.
Powered by Flyspray