Glossary

These terms are commonly used in information about Redrock Postgres.

The basic properties of a database transaction that all Postgres Database transactions must obey. ACID is an acronym for atomicity, consistency, isolation, and durability.

A function such as COUNT that operates on a group of rows to return a single row as a result.

A member of the online redo log that has been archived by Postgres Database. The archived redo log files can be applied to a database backup in media recovery.

A mode of the database that enables the archiving of the online redo log.

The background process that archives online redo log files.

The operation of generating an archived redo log file.

An index in which data is stored in ascending order. By default, character data is ordered by the binary values contained in each byte of the value, numeric data from smallest to largest number, and date from earliest to latest value.

An index organized like an upside-down tree. A B-tree index has two types of blocks: branch blocks for searching and leaf blocks that store values. The leaf blocks contain every indexed data value and a corresponding ctid used to locate the actual row. The “B” stands for “balanced” because all leaf blocks automatically stay at the same depth.

A process that consolidates functions that would otherwise be handled by multiple Postgres programs running for each client process. The background processes asynchronously perform I/O and monitor other Postgres processes.

See also database instance; Postgres process.

A copy of data. A backup can include crucial parts of the database such as data files, the server parameter file, and control file.

A placeholder in a SQL statement that must be replaced with a valid value or value address for the statement to execute successfully. By using bind variables, you can write a SQL statement that accepts inputs or parameters at run time. The following example shows a query that uses $1 as a bind variable:

SELECT * FROM employees WHERE employee_id = $1;

A data block that is not in a recognized Postgres format, or whose contents are not internally consistent.

A part of a data block that includes information about the type of block, the address of the block, and sometimes transaction information.

Space in a data block that stores metadata required for managing the block. The overhead includes the block header, transaction items, and row directory.

In a B-tree index, a block that the database uses for searching. The leaf blocks store the index entries. The upper-level branch blocks of a B-tree index contain index data that points to lower-level index blocks.

A main memory address in the shared buffer area. A buffer caches currently and recently used data blocks read from disk. When a new block is needed, the database can replace an old data block with a new one.

A code that pairs each character from a given repertoire with a code unit to facilitate data storage.

Treatment of strings as a sequence of characters. Offsets into strings and string lengths are expressed in characters (character codes).

An encoding scheme used to display characters on your computer screen.

A constraint on a column or set of columns that requires a specified condition to be true or unknown for every row.

1. A data structure that marks the checkpoint position, which is the LSN in the redo thread where instance recovery must begin. Checkpoints are recorded in the control file and each data file header, and are a crucial element of recovery.
2. The writing of dirty data blocks in the shared buffer area to disk.

The background process that updates the control file and data file headers with checkpoint information and write blocks to disk.

Acronym for command identifier. You might see this acronym in error messages or keywords.

This is the data type of the system column rowcid. Command identifiers are 32-bit quantities.

Vertical space in a table that represents a domain of data. A table definition includes a table name and set of columns. Each column has a name and data type.

Action that ends a database transaction and makes permanent all changes performed in the transaction.

The automatic removal of lock-related transaction information (transaction item) from the blocks after a commit. The database removes the transaction item only if modified blocks containing data from the committed transaction are still in the shared memory area, and if no other session is modifying them.

Smallest logical unit of data storage in Postgres Database. Other names for data blocks include Postgres blocks or pages. One data block corresponds to a specific number of bytes of physical space on disk.

The combination of the shared memory area and background processes. An instance manages one or more databases. Every database instance is either a read/write database instance or a read-only database instance.

An event trigger is used to capture events related to DDL statements.

The Postgres feature for finding words, phrases, Boolean combinations of words, and so on within table data, in a faster, more convenient, and more flexible way than using the SQL LIKE operator or writing your own application-level search algorithm.

Logical Timestamp. A database ordering primitive. The value of an logicaltime is the logical point in time at which changes are made to a database. It is the data type of the system column rowtime.

Acronym for object identifier. You might see this acronym in error messages or keywords.

See Also Object Identifier.

Acronym for process identifier. You might see this acronym in error messages or keywords.

See Also Postgres Process.

A unit of execution that runs the Postgres database code. The process execution architecture depends on the operating system. Postgres processes include server processes and background processes.

A set of tuples.

A set of column information corresponding to a single record in a table. The database stores rows in data blocks.

A named collection of database objects, including logical structures such as tables and indexes.

A group of shared memory structures that contain data and control information for one Postgres database instance.

See also the shared memory area in the Memory Architecture chapter.

The characteristic of a SQL statement as an atomic unit of work that either completely succeeds or completely fails.

The guarantee that data returned by a single query is committed and consistent for a single point in time.

A database operation in which the effects of an unsuccessful SQL statement are rolled back because the statement caused an error during execution.

Basic unit of data storage in Postgres Database. Data in tables is stored in rows and columns.

Acronym for tuple identifier (row identifier). You might see this acronym in error messages or keywords.

This is the data type of the system column ctid. A tuple ID is a pair (block number, tuple index within block) that identifies the physical location of the row within its table.

Logical unit of work that contains one or more SQL statements. All statements in a transaction commit or roll back together. The use of transactions is one of the most important ways that a database management system differs from a file system.

A tuple is a row in a table. It has a number of attributes, each of which has a name, a value, and a data type.

See also row.

Records of the actions of transactions, primarily before they are committed. The database can use undo data to logically reverse the effect of SQL statements. Undo data is stored in undo segments.

Redrock Postgres maintains undo data in undo segment.

Acronym for transaction identifier. You might see this acronym in error messages or keywords.

This is the data type of the system column rowxid. In Redrock Postgres, the transaction ID type xid consists of the undo segment number, transaction slot number, and sequence number. Its length is 8 bytes.