Oracle Data Block Structure Analysis

Oracle Data Block Structure Analysis oracle data block is the smallest unit of database storage, generally N times the operating system block. Its structure is: www.2cto.com block header --> empty row --> data, which is actually a vertical row structure. The standard block size is specified by the initialization parameter DB_BLOCK_SIZE. A Block with a Standard size is called a Standard Block ). The Block size is different from the standard Block size. A non-standard Block is called a non-standard Block ). In the same database, Oracle9i and later versions support the use of both standard and non-standard blocks in the same database. Oracle allows you to specify five non-standard blocks ). Oracle data blocks include :? Block header: The Block header contains segment types (such as tables or indexes), data block addresses, table directories, row directories, and transaction processing slots. The size of each slot is 23 bytes. These slots are used when the row in the block is modified. The block header is expanded from top to bottom. ? Row data: this is the actual data of the row. The row data space is expanded from bottom to top. ? Idle space: the idle space is located in the middle of the block. In this way, the header and row data space can be expanded as needed. After some new rows are inserted, or some columns of existing rows are updated using the greater limit, row data occupies free space. If the row directory requires more row entries, or the number of transaction slots required exceeds the number originally configured, the header extension will appear. At first, the free space in the block is adjacent. However, deletion and update operations may cause fragments in the free space of the block. If necessary, the Oracle server joins the free space in the block. Row Directory: If the block contains Row data, the information of these rows is recorded in the Row Directory. This information includes the address of the row.