IBM Books

DB2 Replication Guide and Reference


The Control Center

The Control Center is the database administration tool for the DB2 Universal Database and is the tool that you use for replication administration. The Control Center automates many initialization functions, such as creating target tables and control tables when you specify target copy information.

This section briefly explains the main replication tasks that you perform from the Control Center, the objects that you use in the Control Center, the types of target table copies that you can create, auto-registration, and the control tables that are created when you use the Control Center. See Part 3. Administering Your Replication System for more detailed information about Control Center replication tasks. See Administration Getting Started for your platform for more information about the Control Center. See the DB2 DataJoiner V2 documentation for details about heterogeneous replication administration.

Tasks

You must be a replication administrator to perform the replication administration tasks from the Control Center.

The main Control Center administration tasks for replication are:

Defining tables as sources, called replication sources
Replication sources are DB2 tables, or views, used as sources for copying data to a target table. When you define a table as a replication source, you specify how that table will be used in your replication scenario. For example, you specify which columns in the table are available for replication, whether the data existing in a column before a change should be captured in addition to the change to the column. When you define a replication source, you set the information that the Capture program will use. The Capture program uses the information when it looks at the DB2 log to find changes for the table. The Capture program then copies the changes it finds in the log to a change data table. A change data table is created for every replication source that you define.

Defining replication targets, called replication subscriptions
When you define a replication subscription, you are specifying target tables, their structure, the sources for the target tables, and the schedule for the updates to the target tables. You can create new columns in the target tables with the exact information from existing columns in the source table. You can also create new columns in the target table based on:

You can also specify SQL statements that you want to execute before or after the subscription is executed.

Other tasks include removing replication sources that are no longer sources for copies, cloning replication sources and subscriptions to other servers, and removing replication subscriptions.

You do not use the Control Center to operate the Capture and Apply programs.

Objects

You can access your replication source and target definitions through the Control Center. There are three containers that organize the objects that you use to set up and maintain your replication environment:

Tables folder
Contains user-defined DB2 relational tables, as well as system catalog tables. You select the tables in the folder to define replication sources.

Replication Sources folder
Contains tables that have been defined as replication sources. These can be DB2 tables, system tables, views, or target tables redefined as sources for replication.

Replication Subscription folder
Contains replication subscription definitions for copying updated source data to target tables.

You work with three GUI objects: Tables, Replication Sources, and Replication Subscriptions, to manage replication sources and targets.

Figure 3 shows the Control Center window.

Figure 3. The Control Center and the Replication Folders. The Control Center and the replication folders provide administration tools for replication.
The Control Center and the replication folders

Types of Copies

IBM Replication offers you many choices for target table structures. Depending on the kind of data you want in your application system, you can choose how changes are copied to the target table. Target tables function as historical or trend information, sources for update-anywhere replication, or simply an identical copy of the source table. The different types of target tables are:

User copy
A complete, condensed copy of the replication source table that must have a primary key.

Point-in-time
A complete, condensed copy of the replication source table at a certain point in time that must have a primary key. This table contains timestamp columns to indicate when a transaction occurred.

Base aggregate
A history table in which new rows are appended for each subscription cycle using the result of an SQL column function calculation against the replication source table data.

Change aggregate
A history table in which a new row is appended for each changed row in the replication source table using the result of an SQL column function calculation against only recently changed data.

Condensed, noncomplete consistent change data (CCD)
A small staging table that, when defined locally to the source, is useful for netting out "hot spot" updates before replication to other sites.

Condensed, complete CCD
A full-sized staging table useful for efficient remote staging, which allows for remote copies to be both initialized and maintained without needing to access the original source each time it is updated.

Noncondensed, noncomplete CCD
A table that is initially empty and is appended by each insert, update, and delete action; useful for auditing purposes.

Noncondensed, complete CCD
A table that starts out as a copy of the full-sized source table and is appended by each insert, update, and delete action. It retains all information about the source table and supports "as of" historical queries. For example, it can return an answer to a query as if you had run the query against the replication source table last Tuesday, or a month ago, or yesterday.

Replica
A target table that can be updated. Changes to this table are replicated back to the replication source table. This table is used in update-anywhere scenarios.

Auto-Registration

Some target table types are designed to become sources for additional replication. The CCD table and the replica table are almost always used in multiple tier or circular replication scenarios and are automatically defined as replication sources when you define them as part of a replication subscription.

The Replication Control Tables

IBM Replication tools use control tables to communicate with each other and to manage replication requests such as defining and managing sources and targets, capturing changes, and replicating changes. These control tables are located at the source, control, and target servers, which are defined in Terminology.

When you use the Control Center to perform replication administration tasks, it creates and maintains the replication control tables used by the Capture and Apply programs. These tables are created and stored at the source, control, and target servers. The workstation where the Control Center is located must be connected to all the databases where source tables exist and the Capture program runs (the source server), all the databases where the target tables will be created and the Apply program usually runs (the target server), and all the databases where the subscription control tables are stored (the control server). The exception is the "occasionally connected," or mobile environment, in which you can create and save as SQL a replication subscription definition, transfer it to another machine, and run it from there.

Most administrators need to customize the control tables for their site requirements or DB2 platforms. To customize the control tables, you must perform some steps before you perform replication requests (that is, before you define sources or subscriptions with the Control Center). For more information on customizing control tables, see Working with Customized Replication Control Tables.


[ Top of Page | Previous Page | Next Page | Table of Contents | Index ]

[ DB2 List of Books | Search the DB2 Books ]