How to import data over a network connection in Oracle 10g

Historically, Oracle's import and export capabilities have used disk files as storage intermediaries for unloading and reloading databases. The way to "swap files" for larger databases is a big problem. Because a large file will exceed the operating system's limit on file size, the export cannot be implemented.

Some innovative data administrators use file compression, such as the compression feature in Unix, to obtain the maximum size of a swap file. The import and export features in later versions allow multiple interchange files to be used to reduce restrictions.

In Oracle 10g, the import feature of the Data pump version can be imported directly from other database instances to completely remove the swap file.

The first step is to define a database-linked object to identify the source database and provide logon credentials. For example, a source database in Chicago can be identified through the Oracle Network service name CHI. The following CREATE DATABASE Link command can be used to define source databases in the instance, ADMIN1, logon with password windy and access to the data to be imported:

CREATE DATABASE LINK Chicago
CONNECT to admin1 identified by windy
USING ' CHI '; The Data Pump Import command--IMPDP now enables direct access to remote data using this database link. Command-line arguments Network_link point to the source database through a database link. In the Seattle local database instance, the user ADMIN2 will execute the following command (whole line):

IMPDP Admin2/market tables=customers,sales DIRECTORY=DPUMP1
Network_link=chicago Note The first Chicago instance does not have an export operation, and the swap file is not created during the operation.

If you want to query for more questions about how to import over a network link, refer to Oracle Database 10g Utilities in the Oracle file settings.