Database+Administration

media type="custom" key="4399783" **[|Database Administrator]** is responsible for maintaining and updating the database and the DBMS software. Administrators are critical players in the success of any organization, since they are largely responsible for preventing computer downtime, or time in which the system is unavailable.

[|Data Loss or Corruption] - **Data corruption** occurs when data is unreadable, incomplete, or damaged. Stategies for backing up data are the major method for recovering lost or corrupted data. A **Backup and Recovery Plan** is a key part of any DBMS. To lessen the chance of accidental data loss, it is important that backup files are stored separately from original material. Tape backups are a commonly used backup method, as tape are cheap mass storage media that can be placed in a company safe for protection.

[|Record Locking] is an automatic protection process that occurs when users attempt to edit existing records in a multi-user system. The system automatically checks to see if anyone else is working on the same record, and only allows one user at a time to edit or delete the record.

[|Data Integrity] is used to describe the accuracy of the information provided to database users. Redundancy, or the duplication of data in several fields, is an enemy of data integrity. Having the same value in multiple places creates opportunities for error when changes are made. If an address is entered in several database locations, any changes to that address would have to be entered in all the locations. Another example of this problem might occur if an person signs up more than once for the same account. Redundancy errors are difficult to weed out. One technique used is called normalization, a process intended to eliminate redundancy among fields in relational databases.


 * Data Contamination** is the spread of incorrect information. In many situations data contamination can have serious consequences, such as in credit rating checks. Preventing data contaminations is one of the goals of data validation. **Data validation** is summed up by the phrase garbage in, garbage out (GIGO). Data validation checks are methods of restricting input so that false data cannot be as easily entered into the system. Validation methods include range, alphanumeric, consistency, and completeness checks.

[|Referential Integrity] involves a check to make sure that deleting a record in one table will not affect other tables. If you were to delete the accounting department record in your database, then all the employee records that listed a key to the accounting department would have an invalid reference to a record that no longer existed.

A **range check** is a simple error-checking system usually performed on numeric data entries. An **alphanumeric check** allows only letters of the alphabet and digits to be entered. A **consistency check** may be made against previously entered data that has already been validated //(will indicate an error if a user attempts to enter numbers that do not match previously validated information)// A **completeness check** ensures that every required field is filled out. One of the greatest threats to data integrity is the natural human tendency to tire of entering data. To prevent this, one of the first checks performed on any data entry is a check for completeness, meaning that valid entries were provided for every field marked "required".