Partitions and Data Management

Many databases today are designed to manage terabytes of data, and provide a formidable support and maintenance challenge for the DBA.
Partitions are a way to efficiently manage these very large databases, by allowing the DBA to manage procedures including

1. index creation,
2. data loading, and
3. data purge at the partition level, rather than on the entire table.

The result is a drastically reduced maintenance and support time.

In the past when business computing was called "Data Processing", Master Data Management was known as "Data Administration", the job of defining and managing the companies metadata. There was even a job role, Data Administrator, which was a job where the IT professional cataloged all common data attributes for the organization. Formally, the definition for Master Data Management is identical to those from the 1980's Data Administrator (also known as Metadata administrator):

Master data is the consistent and uniform set of identifiers and extended attributes that describe the core entities of the enterprise.

The predecessor to the MDM specialist was the metadata specialist, commonly used in data warehouse and business intelligence systems where data from shared sources was integrated. This was usually a senior Oracle data administrator (DA) who had an intimate knowledge of the data in the operational systems, whether the systems are legacy databases such as SQL Server or IMS systems.

MDM is a process that may be

1. workflow-driven or
2. transactional

in nature, in which business units and IT departments collaborate, publish and protect common information assets that must be shared across the enterprise
In plain English, Oracle master data management is the idea that tracking and managing shared data resources allows companies faster and more seamless access to shared data, usually things like master lookup tables. Back in the days of pre-relational databases and COBOL, master data was shared across the enterprise much as it is today, with Data Administrators defining and managing the shared data components across the enterprise.
For Oracle databases, MDM allows for several important data management functions:

1. Data Uniformity: One major role of MDM is calling attributes by common names, and a large job of an MDM administrator is collecting and referencing all names for shared data, usually from mining the Oracle data dictionary. In some shops, this goes as far as collecting PL/SQL variable name, and these variables are stored in the dictionary in extension tables with the dba_synonyms view.
2. Data Synonyms: Most corporations have context-sensitive entries. For example, a person can be a "customer" to the sales department, while a person can be a "employee" to the HR department. MDM allows for the management of common entities and the documentation of department-specific information. In our example, a person will need an employee_type flag in HR and a customer_status flag in sales.
3. Data keys: The usage of shared data components varies between department, as does the keys for the data. For example, a person of type customer may be identified with a customer_id, while a person of type employee may have an emp_id column. Both serve to identify a person.
4. Cross platform data usage: In addition to understanding the data structures and data relationships within the corporate schema, the master data management specialist must also be well-versed in SQL and understand the ramifications of not null data characteristics. The MDM specialist will often work closely with the legacy systems to develop a "common data format" for the extraction and to devise methods for cleansing, unifying and validating shared data.