A VM is protected in an Async Protection Policy with a schedule for every 1...

In a Nutanix environment, Synchronous replication (including Metro Availability and Synchronous Protection Policies) relies on a specialized service known as Anduril to manage the real-time mirroring and state transitions of virtual machines. While standard asynchronous replication primarily uses ports 2009 (Stargate) for data transfer and 2020 (Cerebro) for metadata and replication management, synchronous replication introduces additional requirements for low-latency communication between the Anduril services on the source and destination clusters.

The error message " Failed to connect to remote Anduril " specifically indicates that the local cluster ' s management plane is unable to communicate with the remote cluster ' s synchronous replication manager. TCP port 2036 is the dedicated port for cross-cluster communication between Anduril services. This port is used to negotiate the " stretch " of the VM across sites, coordinate the initial sync, and maintain the health heartbeat required for high-availability failover decisions. If port 2036 is blocked by a physical or virtual firewall between the clusters, the " VmSyncRepEnable " request will time out or fail, preventing the VM from ever entering the " Sync " (Active-Active) state. Troubleshooting synchronous replication always begins with verifying that the 5ms latency requirement is met and that port 2036 is open alongside the standard ports 2009 and 2020 to ensure both data and control traffic can flow without interruption between availability zones.