Generates data access Web services for Oracle databases. Refer to “SKM SQL” in section “Generic SQL” for more details.
This SKM is optimized for the Oracle database.
Some of the Knowledge Modules for Oracle use specific features of this database. The following restrictions apply when using such Knowledge Modules. Refer to the Oracle documentation for additional information on these topics.
Using the SQL*LOADER utility
The Oracle Client and the SQL*LOADER utility must be installed on the machine running the Oracle Data Integrator Agent.
The server names defined in the Topology must match the Oracle TNS name used to access the Oracle instances.
A specific log file is created by SQL*LOADER. We recommend looking at this file in case of error. Control Files (CTL), Log files (LOG), Discard Files (DSC) and Bad files (BAD) are placed in the work directory defined in the physical schema of the source files.
Using the DIRECT mode requires that Oracle Data integrator Agent run on the target Oracle server machine. The source file must also be on that machine.
Using External Tables
The file to be loaded by the External Table command needs to be accessible from the Oracle instance machine. It must be located on the file system of the server machine or reachable from a UNC path (Unique Naming Convention) or mounted locally.
For performance reasons, it is often recommended to install Oracle Data Integrator Agent on the target server machine.
Using Oracle Log Miner
The AUTO_CONFIGURATION option provides automatic configuration of the Oracle database and ensures that all prerequisites are met. As this option automatically changes database initialization parameters, it is not recommended to use it in a production environment. You should check the Create Journal step in the Oracle Data Integrator execution log to detect configurations tasks not performed correctly (Warning status).
Asynchronous mode gives the best performance on the journalized system, but this requires extra Oracle Database initialization configuration and additional privileges for configuration.
Asynchronous mode requires the journalized database to be in ARCHIVELOG. Please review Oracle documentation carefully before turning this option on. This will help you correctly manage the archives and avoid common issues such as hanging the Oracle instance if the archive files are not removed regularly from the archive repository.
If ASYNCHRONOUS_MODE is set to "No", journalized data is available immediately after the commit. Otherwise, there is a gap between the commit and the journalized data availability. This time may vary from 1 second to a few minutes.
Always stop the journal before changing the ASYNCHRONOUS MODE option.
Knowledge Modules Reference Guide