Re-replication for re-balance utilization doesn't check that the destination chunk server for re-balancing already has the chunk stored on it. If the chunk initially had 2 replicas, one on the re-balance source chunk server and one on the re-replication destination chunk server, after the rebalancing the file will only have one replica, on the destination one.
I can provide additional logs from the environment I'm using it to prove it if needed.
Re-replication for re-balance utilization doesn't check that the destination chunk server for re-balancing already has the chunk stored on it. If the chunk initially had 2 replicas, one on the re-balance source chunk server and one on the re-replication destination chunk server, after the rebalancing the file will only have one replica, on the destination one.
I can provide additional logs from the environment I'm using it to prove it if needed.