Database Management Systems

Oracle Warehouse Builder is certified against multiple Transparent Gateways for heterogeneous connectivity: Microsoft SQL Server, Sybase, Informix, and DB2 on OS/390 (MVS). Each individual option A-D represents a supported gateway, so E (All of the above) is the correct comprehensive answer.

OWB leverages Oracle's parallel execution capabilities for DML operations. Generated code can use parallel INSERT, UPDATE, and DELETE statements, significantly improving performance for large-scale data loads on multi-CPU systems. The answer correctly states this as True.

OWB (Oracle Warehouse Builder) supports multiple operating modes for data movement and transformation. Row-based mode processes data one row at a time, while set-based mode processes data in batches. The target-only variant applies the row-based strategy specifically to the target. The fallback from set-based to row-based (target-only) provides a mechanism to switch strategies if set-based operations encounter issues.

Informatica PowerCenter supports heterogeneous sources and targets, meaning it can extract data from one type of system or format and load it into a completely different type. This includes different databases, file formats, and applications, making it a powerful ETL tool for data integration across diverse environments.

The Informatica Repository contains metadata for transformations (A), which includes information about transformation logic, ports, properties, and relationships. While source and target schemas are stored in the repository, the option listing source schema alone (B) or target schema alone (C) is incomplete. Option D 'Transformations' is redundant because transformation metadata (A) encompasses the transformation objects themselves.

Informatica supports heterogeneous sources AND targets - meaning it can extract data from different types of source systems (Oracle, SQL Server, flat files, XML, etc.) and load into different types of targets in the same integration. This heterogeneity support is a core feature of Informatica PowerCenter. Options A and B are incorrect because they claim 'only' one side is heterogeneous. Option D is incorrect because it states homogeneous only.

Navigational systems required programmers to explicitly navigate through the database by knowing the physical structure of tables, indexes, and relationships. This meant you had to understand HOW data was stored and accessed, not just WHAT data you wanted. SQL's key advantage is being declarative - you specify the desired result without specifying the access path.

SQL statement tuning centers on optimization quality - whether the query optimizer finds an efficient execution plan. The number of CPUs (A) and degree of parallelism (B) are system resources, not tuning considerations. Bitmap indexes (C) are specific access methods, not general tuning principles. Optimization quality directly affects performance.

Counterintuitively, introducing controlled redundancy (denormalization) often improves SQL performance by reducing joins and table scans. Removal of redundancy (A) is normalization, which can hurt performance. Non-first normal form (C) relates to complex values, not performance. SQL*Plus (D) is an Oracle tool, irrelevant to performance.

ADABAS is a hierarchical/inverted-list database system, not a relational database. It uses a different data model than RDBMS systems like Oracle, SQL Server, or DB2. It pre-dates and differs from the relational model.

ADABAS stores data hierarchically: fields group into records, records into files, and files into a database. This is FIELD -> RECORD -> FILE -> DATABASE, which is the inverted-list structure ADABAS uses.

ADABAS uses natural language commands for data access: Find (search), Read (retrieve), Histogram (data distribution analysis), Get, Store, etc. 'Select' is SQL terminology, not part of ADABAS's native command set. ADABAS predates SQL and has its own access language.

ADABAS's unique and distinguishing feature is multivalued fields - the ability to store multiple values within a single field occurrence without child tables. Different views (A), data storage (B), and security (D) are common to most database systems and not unique to ADABAS. Multivalued fields were revolutionary in pre-relational systems.