We are using a NGFW cluster in front of our WVD. When we test a failover from Node A to Node B, the failover is successful, however, there is a significant delay of between 4-7 minutes before the Host Session Agent established connectivity again with the RD Broker service. What seems to happen is as follows
- The UDR for the WVD subnet is updated with the Next Hop being correctly updated to the primary FW node ~ 1 minute
- The RD Client then cannot connect to the WVD resource for at least 5-7 minutes.
Does anyone know if there is a polling interval for the agent and what this setting is? I assume it is set to poll rdbroker.wvdselfhost.microsoft.com:443