You can optionally customize apply settings for individual target tables, provided that the target table has the same name and schema as the source table. You can customize the types of DDL and DML operations to apply and whether to distribute change records by rows or tables.
Data Replication does not replicate DDL operations to
Apache Kafka
targets.
On the
Map Tables
tab, click
Settings
.
The following image shows the
Customize Settings for the Selected Tables
dialog box:
To indicate the types of DDL changes to apply to a table, select the
Add Column
,
Drop Column
,
Alter Column
,
Truncate Table
,
Add Index
, and
Drop Index
check boxes, as needed. By default, these check boxes are not selected.
To select a DDL operation type for a single table, select the check box in the row for the table. To select a DDL operation type for all listed tables, select the check box in the
Select All
row.
To not apply some types of DML changes, clear the
Insert
,
Update
, and
Delete
check boxes, as needed. By default, these check boxes are selected and Inserts, Updates, and Deletes are applied.
To not apply a DML operation type for single table, clear the check box in the row for the table. To not apply a DML operation type for all of the listed tables, clear the check box in the
Select All
row.
In the
Distribute by
column, select the type of algorithm that Data Replication uses for distributing Insert, Update, and Delete records to threads for parallel apply processing. Options are:
Rows
. Change records for a table are distributed across threads. This option is the default for Oracle targets.
Tables
. Change records for a table are assigned to a separate thread. This option is the default for DB2, Microsoft SQL Server, and Merge Apply targets. Informatica recommends that you select the
Tables
option when replicating a large number of small tables or when replicating data to an appliance such as Netezza or to a target on AIX, Solaris SPARC, or HP-UX systems that have certain CPU architecture such as Itanium 64-bit. With this option, Data Replication selects the thread to use for a table and dynamically distributes records as follows:
In SQL Apply mode, if the apply.subtasks_enabled runtime parameter is set to 1 and the change volume on a thread begins to cause a bottleneck, Data Replication dynamically spawns a subtask to process some of the change records. Data Replication uses the thread with the least load. If necessary, Data Replication can spawn multiple subtasks on multiple subtask threads. To prevent Data Replication from spawning too many threads, set the minimum amount of data, in MB per cycle, that is available to distribute to a subtask before spawning a new thread. To specify this minimum threshold, enter a value for apply.distribute_by_obj_size_threshold parameter on the
Runtime Settings
tab >
Advanced Settings
view. In Merge Apply and Audit Apply modes, Data Replication does not support Applier subtasks.
If the change volume on a thread is low, other tables that also have low change volumes can share the thread.
This setting is synchronized with the
Distribute by
setting in the
Map Columns
dialog box.
Click
OK
to save changes and close the dialog box, or click