Which statement best describes a B-tree index in databases?

Which statement best describes a B-tree index in databases?

• A. A fixed in-memory index that must fit entirely in RAM.
• B. A self-balancing tree structure with keys and pointers sized to fit one disk page, supporting both exact-match and range queries.
• C. A write-optimized index that uses log-structured merges.
• D. A spatial index designed for 3D data.

Answer: (B)

This item tests how a B-tree index is organized to work efficiently with storage on disk. A B-tree is a self-balancing tree where each node holds multiple keys and child pointers, and the size of a node is chosen so that it fits nicely in a disk page. This design minimizes disk I/O because reading or writing an entire node can be done in a single I/O operation, and the tree’s height stays logarithmic in the number of keys, so searches, inserts, and deletes are fast even with large datasets. Because keys are arranged in sorted order within each node and the tree is shallow, you can navigate to an exact key quickly (exact-match) and you can also scan a range of keys efficiently by following the linked set of leaf nodes (range queries). In practice, many databases implement B+ trees, where internal nodes guide navigation and leaves store the actual pointers to records with their leaves linked for fast range scans; the principle remains the same: disk-page-sized nodes, balanced height, and support for both exact and range lookups. The other options point to different ideas: a fixed in-memory index isn’t the typical disk-based B-tree design; a write-optimized log-structured merge approach emphasizes append-heavy writes rather than a disk-page based balanced tree; and a spatial index is designed for multidimensional data, not the one-dimensional ordered keys and range queries of B-trees.

This item tests how a B-tree index is organized to work efficiently with storage on disk. A B-tree is a self-balancing tree where each node holds multiple keys and child pointers, and the size of a node is chosen so that it fits nicely in a disk page. This design minimizes disk I/O because reading or writing an entire node can be done in a single I/O operation, and the tree’s height stays logarithmic in the number of keys, so searches, inserts, and deletes are fast even with large datasets.

Because keys are arranged in sorted order within each node and the tree is shallow, you can navigate to an exact key quickly (exact-match) and you can also scan a range of keys efficiently by following the linked set of leaf nodes (range queries). In practice, many databases implement B+ trees, where internal nodes guide navigation and leaves store the actual pointers to records with their leaves linked for fast range scans; the principle remains the same: disk-page-sized nodes, balanced height, and support for both exact and range lookups.

The other options point to different ideas: a fixed in-memory index isn’t the typical disk-based B-tree design; a write-optimized log-structured merge approach emphasizes append-heavy writes rather than a disk-page based balanced tree; and a spatial index is designed for multidimensional data, not the one-dimensional ordered keys and range queries of B-trees.