Moving Forward with the On Demand Real-time Enterprise

Chapter 5. More real-time enterprise enablers 225

committed source changes are staged in relational tables before being replicated

to target systems. In Q replication, committed source changes are written in

messages that are transported through WebSphere MQ message queues to the

target systems.

There is also a solution called event publishing for converting committed source

changes into messages in an XML format and publishing those messages to

applications such as message brokers.

5.1.1 SQL replication

SQL replication captures changes to sources and uses staging tables to store

the committed transactional data. The changes are then read from the staging

tables and replicated to corresponding target tables. With staging tables, data

can be captured and staged once for delivery to multiple targets, in different

formats, and at different delivery intervals.

Replication can be continuous, at intervals, or for one time only. Replicating

continuously can be useful if your enterprise applications need data in near

real-time. Replicating at intervals can be useful for replicating large batches of

data during off-peak hours. You can also trigger replication through database

events.

There is considerable flexibility with the data that you want to replicate. For

example, you can choose to have all rows and columns replicated or you can

choose just a subset. By subsetting rows and columns, you transport across the

network only the data that you want. You can also clean, aggregate, or otherwise

manipulate the data. If you manipulate data, you can do so at the source and

then distribute the manipulated data. Or you can manipulate the data when you

replicate it so that some targets get manipulated data and others do not.

The SQL replication process is depicted in Figure 5-1. The capture program

reads the committed transaction data from the source DB2 log tables. That data

is then available to the apply program, that writes the data updates to the target

tables. Control tables are used to synchronize the activities of both the capture

and apply programs to ensure data integrity.