The delay introduced by data replication adds risk to some business decisions. For example, a banking application that approves cash withdrawals uses the most current account information available to verify that a customer’s balance is sufficient to cover the withdrawal. If withdrawals processed in a primary database have not reached the replicate database, an application using the replicate database risks approving a withdrawal that exceeds the funds available in the customer’s account.
To limit risk, the banking application can distinguish between high-value transactions and low-value transactions. For example, it might approve a $100 withdrawal based on the account balance in the local replicate database, but it would log in to the primary database to check the account balance before approving a $1000 withdrawal.
