Direct-Path Insert

The Oracle database engine provides two types of INSERT statements to load data into a table: conventional inserts and direct-path inserts. Conventional inserts, as the name suggests, are the ones that are generally used. Direct-path inserts, instead, are used only when the database engine is explicitly instructed to do so. The aim of direct-path inserts is to efficiently load large amounts of data (they can have poorer performance than conventional inserts for small amounts of data). They are able to achieve this goal because their implementation is optimized for performance at the expense of functionality. For this reason, they are subject to more requirements and restrictions than conventional inserts. I'll now discuss how direct-path inserts work, when they should be used, and some common pitfalls and fallacies related to them.