Covering Index
An index containing the columns needed to filter and return a query result, potentially avoiding additional lookups to the base table.
Finding qualifying keys in an index may still require base-table access for columns not stored in that index. When all values needed by a query are available from the index structure, those extra lookups can be reduced, lowering logical I/O and cache pressure for frequently executed narrow reads.
The cost is a wider index and more maintenance on writes. A covering index is therefore a workload trade-off rather than a universal rule: read-path savings must justify storage and write amplification. The engineering meaning of indexes in relational databases discusses the same access-economics boundary.