SDWAN-failover breaks RDP sessions.
Hello everybody. New to Fortigate NGFW solutions and need some help. I'm running Fortigate 120G HA Active-Passive cluster as a hub and a 70G HA Active-Passive cluster as a spoke. There are two WAN links connecting hub and spoke which are in the SDWAN Global zone + two IPSec tunnels over these underlays which are in the SDWAN Local zone. BGP on loopbacks is the routing protocol I'm using. Point is, I expect RDP between hub and spoke to failover seamlessly when the primary link fails, but each time I try I see session drop, as if conn-reset was sent, although we have quite moderate reconnect timings. As far as I read from fortinet community, when there's a link fail, existing sessions first are being flagged as dirty, then re-evaluated and then considered may-dirty again, no asymmetric traffic issue. How to even approach this problem and where to start debugging? Any suggestions?
