In systems where you dont have direct access to the source system you can now take advantage of remote data mining using Oracle Archive log files only. You can use this method of data extraction against the Source Oracle Database, Standby database, Meta data database or even BCV copy of the source database. One advantage of extracting from archive log files is that the whole process can be offloaded to another system and literally have zero impact on the source system. This method has been used to offload very high volume systems in order to load reporting databases.
This document is intended to document the best practice for setting up and configuring GoldenGate Extract to read from Oracle Archive log files only. The Object ID database can be a Source database, Standby database or even a stub database where only a minimal database is configured that has the object IDs from the source database
This document is organized by the different configurations for setting up archive only processing:
Oracle GoldenGate Software install - Provides an overview of GoldenGate software install.
Object ID Database, Source: The Object ID are resolved in the source database. An sqlnet net connection to the source database can be used but the Archive files that need to be processed must be local to the extract process.
Object ID Database, Oracle Standby Database: The Object IDs are resolved in a standby database that was created from the source database. Connection to the standby database can be thru an sqlnet connection, but the archive log files must be local to the extract process. If the extract resolves the object ids at startup, resolve immediate, then the standby database can be put back into managed recovery mode while extract is still running.
Object ID Database, Object database: An Object ID database can start as a standby database or a restored backup of the primary database. The only tablespaces needed in an object ID database are the system tablespace, rollback tablespace. All other tablespaces can be offline dropped before the database is opened for read/write.
Extract Parameter settings: This section covers the different parameter that need to be set for archive log processing.
Performance tuning: This section covers the recommended changes to the extract parameter file that may help extract performance.
Limitations and restrictions: This section covers restrictions and limitations for running GoldenGate in archive log only mode.
The complete document can be found on the Oracle Support site under the document ID:1358342.1
Another document that is related is Oracle GoldenGate Best Practice - Configuring Downstream Integrated Extract v1.4 (Doc ID 1485620.1)