When using VMware HCX to extend networks between an on-premises data center and Azure VMware Solution (AVS), users observe significant packet loss or intermittent connectivity. This behavior typically manifests as failed pings from on-premises to AVS (or vice versa) unless traffic is initiated in both directions simultaneously.
Symptoms:
The issue is caused by a misconfiguration in the Teaming and Failover policy of the Distributed Port Group associated with the extended VLAN. Specifically, the Active and Standby uplinks are reversed or do not match the physical network uplink mapping. This causes asymmetric routing or MAC learning failures on the physical switches unless traffic is active on both paths.
To resolve this issue, you must align the Distributed Port Group uplink configuration with the verified working physical network configuration.
Procedure: