Database Management Systems

Data skew is an uneven distribution of data across processing units (SPUs). When some SPUs receive significantly more data than others, they become bottlenecks while other SPUs sit idle. This defeats the purpose of parallel processing and severely degrades query performance. It's typically caused by poor distribution key choice in MPP databases.

nzload is transactional - if it fails mid-operation, all partially loaded records are rolled back to maintain data integrity. This rollback is typically implemented by logically deleting the records (marking them as deleted). They remain on disk temporarily but are invisible to users and will be cleaned up by maintenance processes.

Zonemaps in Netezza are automatically created and maintained by the system as metadata structures that track min/max values per data slice. Unlike indexes, users do not need to manually create or maintain zonemaps - they are built automatically during data loads.

JDBC classes are distributed across java.sql (core JDBC API including Connection, Statement, ResultSet) and javax.sql (optional package including DataSource, ConnectionPool, RowSet). Options A and B incorrectly use 'jdbc' instead of 'sql', and option D invents non-existent 'rdb' packages.

Transactions require auto-commit to be disabled BEFORE starting transactional operations. This allows multiple statements to be grouped together and either committed (all succeed) or rolled back (all fail) as a unit. Option C is false because auto-commit must be turned off before, not after, the transaction begins.

ServerName is the correct parameter in tm1s.cfg that uniquely identifies the TM1 application server. DataBaseDirectory specifies where data is stored, AdminHost is for administration, and PortNumber defines the communication port.

Creating multiple hierarchies within a single dimension reduces the total number of dimensions needed in a cube, which consequently reduces the number of empty cells. Using separate dimensions for each hierarchy would dramatically increase sparsity.

When loading data into a cube, each dimension must be mapped to a simple (leaf-level) element to identify the specific cell where data should be written. Consolidations and attributes are not used for data loading.

A reporting cube and lookup cube can share data even with zero common dimensions. TM1's ability to reference cells across cubes doesn't require dimensional overlap - it uses cube-level references and dimension mappings.

MigrateFileStorage Utility is the specific tool for migrating from file-based to database-based domain data repository storage. The domain utility is general-purpose, TIBCO Administrator doesn't perform migrations, and MigrateDomainStorage is not the correct utility name.

JDBC core classes are in java.sql package. Extended JDBC API (for connection pooling, distributed transactions) is in javax.sql. There are no java.jdbc or java.rdb packages.

In JDBC transactions, auto-commit must be disabled BEFORE beginning transactional operations, not after. Setting setAutoCommit(false) before the transaction ensures statements are grouped. Setting it after would be meaningless as the transaction would have already executed in auto-commit mode. Options A, B, D, and E describe true transaction characteristics.

All three statements (A, B, C) accurately describe drilldown object components: the source field is a foreign key pointing to the target, the destination field is the primary key in the target BC, and the business component specifies the master view. Option D ('None') correctly indicates there is no wrong statement among them.

When sorting spans multiple tables, database engines cannot effectively use indexes because the sort operation requires combining data from different sources. Indexes work best for single-table sorts where column order matches. Option C is correctly identified as false - using both indexes simultaneously would be inefficient and is not supported.

Picklists enforce business rules by restricting data entry to predefined valid values, ensuring data consistency and preventing invalid inputs. They also speed up data entry by providing dropdown selections instead of manual typing, reducing errors and improving efficiency.