'; zhtm += '

' + pPage + ''; zhtm += ''; figDoc.write(zhtm); figDoc.close(); popUpWin.focus(); } //--> Using Oracle8 -- Glossary

Using Oracle8

Glossary

Advanced Queuing: A messaging feature built into the database that allows message data to be accessed and updated in a manner similar to traditional relational data.
Advanced Replication: A database replication technique in which the entire contents of the database or part of the database can be replicated between several sites, all of which can update data, if desired.
application transparency: A property of a networked system wherein applications access objects in remote databases as though they were on the local database.
attribute: A term from database modeling; attributes describe modeled entities and hold specific pieces of information about the entity. They become columns in relational tables.
AUD$: The name of the table Oracle uses to store audit records. It's owned by the user SYS.
audit records: The records created as a result of auditing.
auditing: The process of recording activity as it occurs. The auditing system in Oracle allows you to record different types of database activity and access to database objects.
b*tree: An index structure that uses levels of branch blocks, each level containing pointers to the next lower level, with a set of leaf blocks at the lowest level. Oracle's b*tree indexes always have the same number of levels between the top of the index and a leaf block, regardless of the value of the index entry.
basic replication: A form of database replication in which a read-only snapshot of the database is replicated between sites.
bitmap index: An index that uses a string of bits that corresponds to rows in a table to indicate whether the indexed value is stored in the row.
bootstrap segment: A block of code used to start the database; it's stored in the SYSTEM tablespace.
branch block: An index block that contains ordered pointers to other blocks.
buffer cache: The section of the system global area in which memory buffers contain images of database blocks.
cache segment: See bootstrap segment.
cardinality: The number of distinct values a column contains.
Cartesian product: A mathematical term that, when applied to relational databases, refers to the result obtained by joining all the rows of one table with all the rows of another table in every possible combination.
check constraint: A constraint based on a user-defined condition that has to evaluate to true for a record to be valid.
checkpoint: A synchronization point between data files and redo log files. Every changed block is written back to disk from memory, and the redo log records a checkpoint sequence number to indicate that all changes up to this point are now on disk.
cluster: A database structure that attempts to store rows with a common value on the same blocks with each other. Rows from one or more tables can belong to a cluster and will be grouped based on value rather than table membership.
cluster key: The column or columns used to group rows in a cluster. The term is sometimes used for a particular value around which rows are clustered.
column: The basic storage unit of a table; holds one type of data and occurs once per record (row) in the table.
column constraint: A constraint definition that's included as part of a column definition's syntax.
complex query: A complex statement in which the parent statement is a query.
complex statement: An INSERT, UPDATE, DELETE, or SELECT statement containing a subquery.
component query: A query that's part of a compound query.
composite index: An index defined on more than one column of a table.
composite limit: A measure of computer resource use that combines CPU time, memory allocation, Oracle block access, and session connect time.
composite primary key constraint: A primary key constraint defined on multiple columns.
composite unique constraint: A unique key constraint defined on multiple columns.
compound query: A query containing set operators (UNION, UNION ALL, INTERSECT, or MINUS) to combine two or more component queries.
concatenated index: See composite index.
concentrating: See multiplexing (2).
constraint: A restriction on a column or columns that defines allowable values, thus preventing unwanted values from being stored.
convert file: A file used to complete the physical migration of an Oracle7 database to Oracle8.
Data Definition Language (DDL): SQL commands that define the different structures in a database, such as CREATE TABLE and CREATE OR ALTER VIEW.
data dictionary: A set of tables that contains descriptive information about the database's components, such as the data files, tablespaces, tables, and users. This information is automatically maintained by Oracle and is available for querying a series of views documented in the Oracle8 Server Reference manual.
data file: A file containing Oracle segments and comprising all or part of a tablespace.
Data Manipulation Language (DML): SQL commands that allow for the manipulation of data in the database, such as SELECT, INSERT, DELETE, and UPDATE.
data segment: A database segment containing rows from a table or from a clustered set of tables.
data warehouse: A database used for storing historical data, which is used for data analysis.
database: A collection of information that, in Oracle's case, is organized in relational tables or objects and stored in data files.
database block: The smallest unit of storage moved between a data file and the database buffer cache in the system global area. Each block contains only one type of information, and the block header identifies the type of block.
database link: A database object that identifies a pathway through Net8 to another database and an actual or im-plied userid to connect to that database.
database management system: A set of computer programs that provides a formal user interface to a database while managing the internal database structures on behalf of the users.
datatype: In an Oracle table, associates a specific set of properties with a column, such as the types of characters allowed and the default length.
default tablespace: The tablespace in which a user's segments will be built, unless overridden by the CREATE command.
default value: A value included in a column for which no value has been supplied with the INSERT statement.
deferred constraint: A constraint that's not checked until its encapsulating transaction completes.
deferred writes: The process of writing changed blocks back to disk when a sufficient number of changes have been made, instead of writing as a result of a transaction commit or other user-driven event.
dimension table: A table, typically in a data warehouse, that contains further information about an attribute in a fact table.
driven table: The table that has its rows accessed based on the result set derived from querying the driving table's rows in a table join.
driving table: In a join statement, the table that has its rows retrieved first.
dynamic performance table: A term used inaccurately as a synonym for dynamic performance view.
dynamic performance view: Oracle provides these views to let users see real-time information derived from the memory structures used by an instance, the contents of the control file, and header information from other database files. Although these aren't built on relational tables like a normal view, they can be queried by using standard SQL SELECT statements.
equijoin: A table join that involves selecting rows from the tables based on the equality of values in columns from each table.
Export: An Oracle utility that unloads part or all of a database's objects (including definitions of users, views, and so forth) into a formatted dump file that can be read by the Import utility. Generally used to restructure free space or to move a database between releases or between different platforms.
extent: A set of contiguous blocks that belong to a single segment.
external authentication: The process of using the operating system to authenticate users to Oracle. External authentication eliminates the need for users to log in to both the operating system and Oracle.
execute: The step in processing an SQL statement in which the data is modified.
fact table: A table, typically in a data warehouse, that contains the primary data.
fast full scan: A data-retrieval method that scans all the entries in an index rather than reads the blocks from the underlying table. It can be used only if all the required columns are part of the index.
fetch: The step in processing an SQL statement in which rows are returned to the application.
foreign key constraint: A constraint that validates a record by ensuring that the value(s) in the constraining column(s) matches values already existing in some other column(s), either in the same or in a different table.
fragmentation: An undesirable collection of disjoint free extents that are too small to be useful.
free list: A linked list of blocks in a table, cluster, or index that will be examined when space is required for new or relocated data.
function: A piece of code that returns a value and may accept arguments. In Oracle8, functions can be used in SQL statements, PL/SQL commands, and TYPE definitions. A number of SQL functions are provided by Oracle; additional functions, known as user functions, can be written in PL/SQL.
hash cluster: A cluster in which rows with the same hash key values are stored together.
hash join: A method of joining two tables that relies on a hashing algorithm to group related rows in a memory work area.
heap-organized table: A table with rows stored in no particular order. This is a standard Oracle table; the term "heap" is used to differentiate it from an index-organized table.
histogram: A histogram is added to an Oracle table to store information about the distribution of values throughout the table, generally used when the different values aren't uniformly distributed.
Import: An Oracle utility that restores part or all of a database's objects (including definitions of users, views, and so forth) from a formatted dump file created by the Export utility. Generally used to restructure free space or to move a database between releases or between different platforms.
index: A structure that maintains an ordered set of entries, providing fast access to specific values.
index cluster: A cluster in which the rows from the participating tables are accessed via an index.
index-organized table: A table whose rows are stored in an index structure, sorted by the table's primary key.
index segment: A database segment that contains an index.
indicator array: An array of indicator variables. Indicator variables are used to handle NULL values when programming in compiled languages because there's no native support for NULL values.
instance: (1) A collection of background processes and the memory acquired for the system global area (SGA) to which they're attached. (2) An alternative name for a row in a relational table.
instance recovery: The process of applying the contents of the online redo log to the data files in order to reconstruct any changes made subsequent to the most recent checkpoint. This procedure is performed automatically by Oracle whenever a database is restarted. If the database were closed normally, the recovery process would have nothing to do because the most recent checkpoint will be the last entry in the redo. The recovery process has two steps: During the first, a roll forward applies all changes from the redo to the blocks, and the database remains closed until this step completes. During the second step, any uncommitted changes are rolled back out, using the rollback segment entries built as part of the roll forward step. This occurs after the database is reopened and can be done by Oracle background processes or by a user process that encounters a block needing to be rolled back.
intelligent agent: A process that runs on the machine that has the database you're trying to monitor. The processes can run autonomously, perform administrative tasks reactively, and provide information to OEM.
interim violation: A violation of a database constraint that occurs during the processing of a statement or a transaction that may or may not be resolved by the time the statement or transaction completes.
itl: See transaction entry slot.
join: The process or the result of merging data from two or more tables, or multiple views of the same table.
leaf block: A block in an index that contains ordered entries and pointers to their respective rows.
LOB index segment: A database segment that contains index information for a large object (LOB).
LOB segment: A database segment that contains a large object (LOB).
local database: In a distributed database environment, a term used to distinguish between the database to which users are connected through a direct connection and any other remote database to which connections are concurrently made on behalf of users.
location transparency: In a networked system, the system's capability to make remote database objects appear local and independent of their physical location.
log buffer: See redo log buffer.
log buffer flush: The action of copying the unwritten contents of the redo log buffer to a redo log, performed by the LGWR process.
log group: A set of one or more identical redo log files written to concurrently by Oracle that can be used interchangeably for archiving and recovery.
method: An operation, defined in a function or procedure, that is applied to a user-defined datatype.
migrate: The process of upgrading an Oracle database from one major release to another.
Migration utility: A tool provided by Oracle to assist in upgrading an Oracle database from Oracle7 to Oracle8.
multiplexing: (1) An Oracle term used to refer to multiple copies of redo logs maintained synchronously by LGWR. (2) The process by which Net8 combines multiple client connection requests over a single transport connection through the Connection Manager.
National Language Support: A feature of the Oracle database that allows users and applications to interact with the database in their native languages.
network transparency: A network system's capability to behave and function similarly across different network protocols.
node: An independent system within a group of connected systems that share some common resources.
non-equijoin: A table join based on matching rows on a criterion other than equality.
NOT NULL constraint: A constraint that requires every record to contain a value in the constrained column.
object privilege: A privilege that allows specific SQL commands to be executed against a named object.
Open Database Connectivity (ODBC): A standard developed by Microsoft Corporation to provide connectivity to Windows-based tools to different ODBC-compliant databases.
optimizer: The part of a relational database management kernel that's responsible for determining how the records required to execute a statement will be located and retrieved (the execution plan).
Oracle block: See database block.
outer join: A table join in which rows from one table are included in the result set, even if they don't contain values that match those in the other table.
package: A PL/SQL structure that's stored in an Oracle database. It can contain one or more procedure, function, global variable, or cursor in any combination.
parallel server process: An Oracle background process used with other such processes to complete an execution step of a statement in parallel, with each parallel server process being responsible for concurrently manipulating only a defined subset of the data.
parent statement: The outer portion, or main portion, of a complex statement.
parse: The step in processing an SQL statement in which the syntax, object validation, and user authorization are confirmed, and during which the execution is built or retrieved from memory.
partition: (1) A subset of a data or index segment, based on a value or range of values in a key field or composite set of key fields. (2) A physical section of a disk storage system.
ping: The transfer of a block of data from one instance to another in an Oracle Parallel Server database.
pipe: An operating system feature that redirects a stream of data-typically the output from a command-into another command that can process or massage it before directing to its own output interface.
primary key: A column or set of columns that comprises a primary key constraint.
primary key constraint: A constraint that guarantees that every record is distinguishable from every other record in a table by requiring that a uniquely defined value be inserted in each row, either in a single column or in a composite set of columns.
privilege: A right to execute a particular statement or type of statement, or to access a particular object in the database.
privilege domain: The set of system and object privileges associated with a database user.
procedure: A PL/SQL program with an assigned name that's stored in the database in compiled form.
profile: A named set of resource and password limits.
pseudocolumn: Pseudocolumns can be treated like table columns in an SQL statement, although they don't exist as part of the table definition. Oracle supports the following pseudocolumns: CURRVAL, NEXTVAL, LEVEL, ROWID, and ROWNUM.
quota: The amount of space allocated to a user in a tablespace.
RAID: A disk system that provides various options for striping, mirroring, and managing error correcting codes (often called parity checks) for balanced disk access and fault tolerance. RAID is an acronym for redundant array of independent (or inexpensive) disks.
raw partition: A disk partition used in its native, raw state without having a file system imposed over it.
recovery catalog: A database schema used by Oracle Recovery Manager to maintain information about the structure and status of the databases it backs up.
recursive SQL: One or more SQL statements executed by the system to complete the processing of a user's SQL statement.
redo log buffer: An area of the system global area where redo entries are buffered before being written to a redo log file.
redo log file: An Oracle database file that contains information from which transactions and their rollback information can be created. The records contain information about each change to any block and are stored in the order in which these changes were made. This allows a recovery operation to replay the changes in the exact order they occurred in the database originally.
redundancy: A number used by Recovery Manager to determine whether a sufficient number of backups are available for a data file. It's a keyword used when generating Recovery Manager reports about the status of available backups.
referential integrity constraint: See foreign key constraint.
relational database: A database that stores its information in two-dimensional tables (relations) that can be joined by using common data elements in one or more columns (attributes).
relink: The step in the process of building an executable program that involves the inclusion of one or more external libraries of common routines.
remote database: A database to which a connection is made, using a database link, while connected to a local database.
replication: The process by which copies of data are consistently maintained in the different databases of a distributed system.
reverse-key index: An index in which indexed values are sorted by reversing the order of the bits.
role: A named collection of privileges that can be granted and revoked like a single privilege.
rollback segment: A database segment that contains images of data recorded before changes being made to it. These "before" images are used to replace the changes should a transaction be rolled back, or should a query need to see data consistent with all other rows in the query.
row: A set of related fields in a table, one field per column, containing values or NULLs.
row piece: A row, or portion of a row, that's stored in a database block. A row that's too large to fit into a single block will be split into multiple row pieces.
rowid: A value that points to the specific physical location of a row in a table by identifying its file number, block number, and relative row number in the block. For tables in tablespaces with more than 1,200 files, the file number is a relative number, and the rowid also includes an object ID number. ROWID is a pseudocolumn that can be referenced in SQL statements. Rowids are also stored in b*tree indexes.
schema: A collection of objects associated with a particular user. In Oracle, the schema name always matches its username, so the two terms-schema and user-can generally be interchanged without confusion.
segment: An Oracle database object that requires its own storage to hold rows or entries, as opposed to an object that's stored simply as a data dictionary entry.
self-join: The process or the result of joining a table to itself.
self-referencing constraint: A foreign key constraint that references a column, or columns, in the same table as the constrained column(s).
sequence generator: An Oracle object that provides numbers guaranteed to be unique. The numbers can be sequential but don't need to be.
snapshot: A copy of a master table (or subset) replicated to other child sites. Snapshots can be updated at specified intervals if required.
sort run: A subset of a database sort that's copied from memory to disk due to lack of space in memory to hold the complete set of data being sorted at one time.
space utilization parameter: A parameter defined on a segment that identifies how the available space in each block will be used.
spatial data: Data that refers to and describes relative locations of objects in multidimensional space.
SQL ("sequel"): Structured English Query Language, a universal language introduced originally by IBM to allow access to data in a database.
storage parameter: A value associated with a database segment that determines a characteristic of extent allocation to that object.
striping: The process or the result of placing related data on multiple physical storage devices.
subquery: A query embedded inside another (parent) SQL statement, returning rows used by the parent.
system global area: A section of memory used to store the structures required by an Oracle instance, including the database buffer cache, the redo log buffer, and the shared pool. The system global area is frequently referred to as the SGA.
system privilege: A database privilege that allows some type of general database access or use.
table: A relational object consisting of one or more fields, where each field is the intersection of a column and a row (record) and has a value based on this specific, logical position. Although columns are identified by column names, which are defined in a specific order, rows have no inherent identifier or order. Unless at least one field value is different between a pair of rows, these rows are indistinguishable from one another.
table constraint: A constraint definition that's not syntactically included as part of a column definition.
tablespace: A named set of one or more data files that provides storage for database segments.
temporary segment: A database segment that temporarily stores records that are being processed in memory but overflow the available space.
transaction: A set of one or more commands issued against the database between bracketing COMMIT or ROLLBACK commands. Consequently, all or none of the database changes made by a single transaction will become permanent.
transaction entry slot: Part of a block that holds information about a transaction using that block, including pointer information to the transaction's rollback information.
transaction slot: See transaction entry slot.
trigger: A PL/SQL program associated with a table's DML activity. A trigger executes when the user performs the DML activity with which it's associated.
tuple: See row.
unique constraint: A constraint that enforces all non-NULL values in a column to be different from each other.
view: A logical table, storing no data itself, that allows access to one or more base tables on which the view is based, using the same SQL commands as if it were a table. Previously, views were considered pre-stored queries, but in current Oracle releases they can be used for data manipulation by DML statements.
VRML (Virtual Reality Modeling Language): A programming language used for modeling and retrieval of virtual reality environments.