This section explains how to make HiRDB and Datareplicator configuration changes.

If you make changes to a system configuration as described in the table below, you must initialize the Datareplicator. If the Datareplicator is not initialized, the update information and configuration information retained in the Datareplicator might no longer match the actual information, and import processing might not be performed as you intended.

The following table describes the configuration changes that require initialization of a Datareplicator.

Table 6-14 Configuration changes that require initialization of a Datareplicator

Item		Description	Subsection
Changing the configuration of the source system	Changing the configuration of the source HiRDB servers	Changing the information specified in the pdstart and pdunit operands in the source HiRDB's system common definition	6.8.2
	Changing the source HiRDB's system log	Changing the number of system log files Initializing system log files Executing HiRDB's pdlogchg -R command to forcibly change the system log status
	Changing the definition of tables subject to linkage	Adding tables subject to linkage (CREATE TABLE) Deleting tables subject to linkage (DROP TABLE) Changing the tables subject to linkage (ALTER TABLE)
	Changing the source Datareplicator's definitions#1	Changing an operand whose definition does not take effect during normal start#2 Changing the contents of the duplexing definition file
Changing the configuration of the target system	Changing the definition of tables subject to linkage	Adding tables subject to linkage (CREATE TABLE) Deleting tables subject to linkage (DROP TABLE) Changing the tables subject to linkage (ALTER TABLE)	6.8.3
	Changing the source Datareplicator's definitions	Changing an operand whose definition does not take effect during normal start#2 Changing the contents of the duplexing definition file
	Changing the configuration of the target HiRDB servers#3	Changing the information specified in the pdstart and pdunit operands in the target HiRDB's system common definition

Note:

For details about changing the configuration when the import transaction synchronization facility is used, see 6.8.4 Changing the configuration when the import transaction synchronization facility is used.

#1

If you use the hdemodq command to add or delete extraction information queue files, there is no need to initialize the source or target Datareplicator.

#2

For details about operands, see the Modifying defined information section for each definition in Chapter 5. Definitions.

#3

Initialization is required if the key range-based partitioning method or the hash partitioning method is defined in the import group definition in the import definition. Otherwise, initialization is not required.

If you need to perform initialization or re-create the target database before all data linkage is completed, see 6.5.6(1) Handling procedure for re-creating the target database.

Organization of this section

6.8.1 Changing the definition of tables subject to linkage

6.8.2 Changing the configuration of the source system

6.8.3 Changing the configuration of the target system

6.8.4 Changing the configuration when the import transaction synchronization facility is used