How the estimate works
ObservKit first normalizes daily ingestion to GB, then multiplies it by the retention period. Replication is applied to the complete raw dataset. The growth buffer is applied last, to the replicated capacity.
Formula
recommended = daily × retention × replication × (1 + growth / 100)
Example
At 500 GB/day for 90 days, raw data is 45,000 GB. With 2× replication it becomes 90,000 GB; a 25% buffer adds 22,500 GB, producing 112,500 GB of recommended capacity.