Hi Yone Louis,
This behavior is completely normal and expected. It is not a configuration error, but rather the strict, intended design of how network constraints function within Storage Replica.
When you apply Set-SRNetworkConstraint, you are explicitly commanding the Storage Replica engine to never use any other network path for replication under any circumstances. When that specific network goes offline, the underlying SMB connection drops. Because SR is aggressively programmed to maintain data sync, it continuously attempts to reconnect. However, since it is hard-locked to the dead interface and forbidden from failing over to healthy networks, it gets stuck in an infinite retry loop. This loop generates the constant "connected/disconnected" flapping Event IDs (5015 and 5014) you are experiencing in your logs.
To resolve this and achieve true high availability, you must remove the SR network constraint. If your goal is to prioritize replication traffic over a specific high-speed NIC while still allowing failover to other networks, do not use Set-SRNetworkConstraint. Instead, simply allow the cluster to use all available networks and let SMB Multichannel handle the traffic routing naturally, or use New-SmbMultichannelConstraint to guide the traffic without creating a hard-locked single point of failure.
I hope this clarifies the architecture and helps you stabilize your Hyper-V cluster. If it did, please click "Accept Answer". Should you have any questions about configuring SMB constraints instead, feel free to leave a comment!
TL.