Data selection and filtering

In addition to column mapping, GoldenGate offers two data filtering options:

• Complex: Complex data evaluations use the FILTER clause. FILTER can select rows and columns for a given operation, whereas WHERE just selects rows. FILTER can also use GoldenGate's built-in functions.

• Noncomplex: This filtering is achieved through the WHERE clause in a TABLE (Extract) or MAP (Replicat) statement.

As a rule of thumb, it is best practice to adopt the following strategy to avoid any performance overhead associated with data selection and filtering:

• Avoid data filtering on the primary Extract group, thus avoiding slow extract of LCRs from the database redo logs.

• Use a secondary Extract group (data pump) to filter data and potentially reduce data volume for transmission to the target system.

• Use a data pump on an intermediate system to filter data, thus avoiding any performance overhead on the source and target systems.

• Use a Replicat to optimize data filtering on the target....