PowerExchange Change Data Capture (CDC) provides the ability to capture insert, update, and delete operations performed against z/OS data sources. When capturing changes, PowerExchange uses techniques that help minimize the impact on the performance and availability of the source table, database, or data set.
Sometimes, PowerExchange CDC captures changes in near real time by integrating with the transaction that performs the change. This technique is called synchronous change data capture. In other cases, PowerExchange CDC captures changes from the source database or source relational database logs. This technique is known as asynchronous or log-based change data capture.
PowerExchange can capture changes from the following z/OS data sources:
Adabas files
Datacom databases
DB2 for z/OS tables
IDMS databases
IMS databases
VSAM data sets
PowerExchange uses the following components for change data capture:
PowerExchange Agent
On a z/OS system, provides and verifies capture registration information for ECCRs.
PowerExchange Condense
Optionally creates condense files that contain a condensed version of the change data in the change stream.
On a z/OS system, captures change data from a data source and passes the captured changes to the PowerExchange Logger for recording.
PowerExchange Listener
Manages data maps for nonrelational files and DB2 tables and capture registrations and extraction maps for all data sources. It also handles extraction requests for bulk data and change data.
PowerExchange Logger
On a z/OS system, receives captured change data from the ECCRs that are connected to it and stores the change data in log data sets.
PowerExchange Navigator
The graphical user interface that you use to define and manage data maps, capture registrations, and extraction maps for the data sources from which you want to extract bulk data or capture change data.
The PowerExchange Navigator runs on Windows. All of the other components run on z/OS.
The PowerExchange architecture is flexible enough to provide for many alternative configurations. You can configure PowerExchange to handle large volumes of change data using multiple instances of PowerExchange CDC components on a single z/OS system. You can capture change data from different source types to multiple PowerExchange Loggers.
The following figure shows an example configuration on a single z/OS system:
This sample configuration contains the following components:
Multiple ECCRs writing to a single PowerExchange Logger.
Multiple instances of the PowerExchange Logger running simultaneously.
Multiple instances of PowerExchange Condense running simultaneously to extract changes from the logs of one PowerExchange Logger.
Multiple instances of the PowerExchange Listener running simultaneously and extracting changes from the logs of a PowerExchange Logger.
One PowerExchange Listener extracting changes from both the logs of a PowerExchange Logger and from condense files. To prevent data loss, the PowerExchange Logger provides dual logging for both the active and archive log data sets.
You can use PowerCenter® to propagate the change data to one or more relational or nonrelational targets in your enterprise. PowerExchange CDC works in conjunction with PowerCenter to perform the following tasks:
Capture change data for sources from which you want to propagate data
Create an inventory of captured change data that you can use for auditing, recovery, and data propagation
Provide data transformation capabilities that enable you to propagate changes that are captured from a database on one system to another type of database that is on another system