Calculate the time required to replicate data across storage nodes or sites.
Replication time is the effective data volume (after compression reduces what actually needs to cross the wire) divided by the available link bandwidth converted from bits to bytes. WAN links are quoted in Mbps (megabits/sec) while data is measured in GB (gigabytes), so bandwidth must be divided by 8 to get MB/s before the comparison is meaningful — a common source of 8x errors when sizing replication or DR links.
Replication time
time = data_GB / (bandwidth_Mbps / 8) / compression_ratio
Network bandwidth is conventionally quoted in bits per second (Mbps) but data volumes are measured in bytes (GB), and there are 8 bits per byte — so a 100 Mbps link provides roughly 12.5 MB/s of usable throughput, not 100 MB/s.
No — this models ideal sustained throughput at the stated compression ratio. Real replication also has protocol overhead, retransmissions on lossy links, and often runs below the link's rated bandwidth due to latency (especially over long-distance WAN links where TCP window size can bottleneck throughput before bandwidth is the limit).
It depends heavily on data type — already-compressed data (video, encrypted volumes) may see a ratio near 1, while text, logs, or database dumps often compress 2-4x, directly cutting replication time by the same factor since less data crosses the link.