Why sizing matters
If a replication link is too small, the secondary site falls behind and your real RPO grows silently. If it is too large, you pay for capacity you never use.
Step 1: Measure change, hourly
Daily averages hide the truth. Look at how much data changes in your busiest hour — database writes, file changes, VM snapshots. That is your baseline requirement.
Step 2: Identify bursts
Month-end processing, nightly batch jobs and backup windows can multiply traffic for a few hours. Decide whether replication must keep up in real time during those bursts or can catch up afterwards.
Step 3: Add overhead
Replication protocols, encryption and network headers add to the payload. A margin of 10–20% is a common starting assumption; validate it with your own measurements.
Step 4: Plan for catch-up
After a link outage or maintenance, replication must catch up on the backlog while new changes keep coming. Capacity above the steady-state need determines how quickly you recover.
Step 5: Watch the trend
Track utilisation weekly. When peaks regularly approach 70–80% of capacity, plan the upgrade before it becomes urgent.
Flexible capacity
Peer1IX DC-to-DC services can be upgraded as your needs grow, so you can start with what you need today and scale as data volumes rise.
