Database Management Systems

Oracle 11g offers ZLIB and BZIP2 compression for RMAN backups. BZIP2 provides better compression ratio but consumes more CPU. ZLIB is faster but compresses less. ZLIB is included in base licensing, while BZIP2 requires additional licensing cost for the Advanced Compression Option.

True parallel backup in RMAN occurs when multiple channels work on different sections of the same datafile simultaneously. Section size parameter divides datafiles into chunks that channels can backup in parallel. Options with parallel=2 or parallel backups=2 refer to channel parallelism, not section-based backup parallelism.

Parallel backups in RMAN divide datafiles into sections backed up by multiple channels simultaneously. Option A is correct because single-disk systems see no performance benefit from parallelism. Option C is correct because RMAN creates equal-sized sections except the last one. Option E is correct because each channel handles different sections. Options B and D are incorrect because sections aren't always same size and aren't backed up by same channel.

RMAN's backup behavior for undo data follows specific rules. COMMITTED UNDO data is never backed up because it's no longer needed for recovery (option A is correct). UNCOMMITTED UNDO data is always backed up to enable transaction rollback after restore (option F is correct). Options B, C, D, and E are incorrect because they misstate when undo data is included in backups. This distinction ensures proper recovery capabilities.

The Virtual Private Catalog feature allows a single RMAN catalog database to serve multiple database owners while maintaining security boundaries. Each administrator gets a virtual view of only their databases (option C is correct). This avoids the overhead of separate catalog databases. Options A, B, and D are incorrect because VPC doesn't create separate catalogs, restrict SYSDBA access, or relate to compliance standards.

RMAN's IMPORT CATALOG command consolidates recovery catalogs. By default, it unregisters the database from the source catalog to prevent duplicate registration, unless the NO UNREGISTER clause is specified. The destination catalog is where metadata is imported, so it is never unregistered. Thus, options supporting source unregistration and the NO UNREGISTER behavior are correct.

Database duplication using RMAN was introduced in Oracle9i Database, allowing direct duplication from backups (option B is correct). This feature was not available in Oracle8i (option A is incorrect). Oracle10g and Oracle11g came later with enhanced features, but the core functionality existed since 9i. This is an important feature for creating test databases or disaster recovery.

The DUPLICATE DATABASE command with FROM ACTIVE DATABASE clause tells RMAN to duplicate the target database over the network instead of using pre-existing backups. This is used for active database duplication where RMAN connects to the target database and copies data files directly. The other options (from target database, from live database, from source database) are not valid RMAN clause syntax for this purpose.

The Design Issue phase of data modeling includes reviewing security requirements, access controls, and data privacy policies. This is where security considerations are integrated into the design before implementation. Structure phase focuses on schema design, Data Source on identifying data sources, Storage Issue on physical storage, and Operational Process on maintenance procedures.

Metadata is 'data about data' that describes the structure, attributes, and relationships of data elements. It is self-describing and populates data dictionaries. However, metadata does NOT include the actual user data - it only describes the data's characteristics. For example, metadata might specify that a field is 'integer, max 10 digits' but not contain the actual values stored.

Linked lists use dynamic memory allocation where each node contains data and a pointer to the next node, making them ideal for collections with frequent size changes. Adding or removing elements only requires updating pointers, making these operations O(1) at known positions.

KSDS (Key Sequenced Data Set) organizes data using an INDEXED structure. Records are stored and accessed based on their key values, which are maintained in a separate index component. This indexed organization enables efficient direct access by key. Option A is correct.

NOERASE specifies that data is not physically overwritten with binary zeros when the VSAM dataset is deleted, leaving raw data on disk. Heat-sensitive or confidential data requires the ERASE parameter to ensure complete destruction. Thus, NOERASE is not desirable for sensitive data, which makes statement a not true.

Indecomposable items in a record that cannot be broken down further are called elementary items, atoms, or scalars - all three terms refer to the same concept of primitive data types that contain a single value. Therefore 'all of the above' is correct.

Arrays store homogeneous elements (same type) with natural ordering by index. Records can store heterogeneous elements (different types) without requiring a natural ordering among fields. Records can form hierarchical structures through nested records, unlike simple linear arrays. All three statements correctly distinguish arrays from records.