Before data replication can start, the target database must have a copy of the database objects that are going to be replicated. It is also necessary that the data in the objects to be replicated are also completely in sync. If the data is not in sync before replication starts, there is a strong possibility that data conflicts can occur. For instance if a record is updated or deleted on the source database that does not exist on the target database.
It is important that the replication starts at the same point as when the source and target data are in sync so that no changes are lost when replication starts. This is done through capturing the Oracle SCN (System Change Number) at the point the table is replicated or "prepared" in Dbvisit Replicate.
Dbvisit Replicate can assist with creating and setting the starting points to ensure no changes are lost when the replication starts, even when it is not possible for an outage on the source database.
The following options are available:
- Oracle Data Pump
- Standby Database
- Rman Duplicate
- Rman Restore and Recover
- Transportable Tablespace ( Enterprise Edition Only)
- The Data Pump option (option 1 above) is best used when the whole database does not need to be replicated, but only a select number of schemas and tables. Also use this option when the source and target databases are different Oracle versions and on different Operating Systems.
- Options 2, 3 and 4 above can be used when the entire database or many schemas need to be replicated.
- The Transportable Tablespace option can be used for heterogeneous environment when the whole database needs to be replicated or there is a very high data volume.
The options are available through the setup wizard. In the following pages, each option will be described.
Only when the target database is Oracle, Dbvisit Replicate assists in creating the starting point. If the target database is not Oracle, then this will have to be done manually. Please see Data instantiation for non Oracle for more information.
Creating the starting point can also be referred to as:
- Data instantiation
- Seeding the target database
- Creating the baseline or baselining
- Loading the historical data