Jijo K Jose

Oracle Data Pump is a new and unique feature of Oracle Database 11g Release 2. A new public interface package, DBMS_DATAPUMP, provides a server-side infrastructure for fast data and metadata movement between Oracle databases. It is ideal for large databases and data warehousing environments, where high-performance data movement offers significant time savings to database administrators.

Key Features

Fast Performance

Operations performed with the new Data Pump Export and Import utilities are typically much faster than operations performed with the original Export and Import utilities. With Data Pump Export, when the direct path method of unloading is used, a single stream of data unload is about two times faster than original Export. This is because the direct path API has been modified to be even more efficient than before. Depending on the level of parallelism, the performance improvement can be even greater.

Improved Management Restart

Every Data Pump operation has a master table that is created in the schema of the user running a Data Pump job. The master table maintains information about all aspects of the job, such as the current state of every object exported or imported and its location in the dump file set. In the event of a planned or unplanned job stoppage, Data Pump knows which objects were currently being worked on and whether or not they completed successfully.

Monitoring and Estimating Capability

In addition to the standard progress and error messages printed by the client and to the log file, the new interactive STATUS command displays cumulative status of the job, along with a description of the current operation. An estimated completion percentage for the job is also returned. The user can also specify a time period, in seconds, for an automatic status update at specific intervals.

Network Mode

Data Pump Export and Import both support a network mode in which the job's source is a remote Oracle instance. When you perform an import over the network, there are no dump files involved because the source is another database, not a dump file set.