Database and SQL

In relationship revision options within Matrix database systems, 'Float' allows relationships to exist without strict enforcement, 'None' indicates no revision control is applied, and 'Replicate' creates copies of related data. Clone and Copy are not standard revision options in this context - they may describe data operations but not relationship revision behaviors.

When calling JCL procedures, you have extensive modification capabilities. Option B is most complete because procedures allow overriding, nullifying, or adding parameters on all statement types (JOB, EXEC, DD), plus adding new DD and OUTPUT statements. Options A, C, and D are incomplete because they don't capture the full range of modification capabilities available when calling procedures.

DISTINCT removes duplicate rows from result sets, returning only unique rows. It does not select all rows, affect ordering, or perform filtering - it purely eliminates duplicates based on selected columns.

UNION can be used with outer joins because UNION combines result sets from multiple queries, and outer joins return matched plus unmatched rows. UNION ALL preserves duplicates while UNION eliminates them, but both are compatible with join operations.

Views serve two key purposes: they can restrict data access by showing only selected columns or rows (security/control), and they can simplify complex queries by presenting pre-joined or calculated data (usability). Both functions are standard use cases.

DB2 provides three main DML (Data Manipulation Language) statements for modifying data: INSERT adds new rows, UPDATE modifies existing rows, and DELETE removes rows. All three operations change the database content.

In SQL, a subquery (nested query) can be placed in several clauses, including the WHERE clause to filter rows and the HAVING clause to filter groups. Since both options are valid, 'Either of the above' is correct.

A foreign key is a key specified in the definition of a referential constraint to link two tables. It is not necessarily a unique key on its own, and it is part of the constraint definition rather than a basic table column definition.

When SQL retrieves data, the output is called a result table - a temporary set of rows matching the query criteria. A view is a stored query definition, a synonym is an alternate name for an object, and an alias is an alternate name for a table - none of these are the retrieved data itself.

WHERE is a SQL clause that introduces conditions, not a comparison operator itself. AND is a logical operator that combines conditions, IS NULL checks for null values, and BETWEEN tests range conditions - these are all valid operators or predicates used within WHERE clauses.

Column functions (also called aggregate functions) operate on sets of values. SUM calculates the total of numeric values, AVG computes the average. Both are standard column functions in SQL.

SQL*Loader supports INSERT (into empty table), APPEND (to existing table), and REPLACE (which can be combined with TRUNCATE to clear table first). DELETE and UPDATE are not valid SQL*Loader loading methods - SQL*Loader is designed for data loading, not DML operations.

This multiple-choice multi-select question correctly defines SQL*Loader options: INSERT requires an empty table, APPEND adds to existing data, REPLACE deletes existing rows, and TRUNCATE truncates the table.

Direct path load requirements include: non-clustered tables, no active transactions, and special handling for referential constraints (loading parent and child together). BFILE columns are also supported in direct path loads. All conditions listed are valid.

In Sybase IQ, the Fast Projection (FP) index is the default index type automatically created for every column in a table when the table is defined. Other index types like LF, HG, and HNG must be explicitly created.