IBM Books

DB2 Replication Guide and Reference

Glossary

A

after-image
The updated content of a source table element that is recorded in a change data table or in a database log or journal. Contrast with before-image.

Apply program
A replication program that is used to refresh or update a target table, depending on the applicable source-to-target rules. Contrast with Capture program.

Apply qualifier
A character string that identifies subscription definitions that are unique to each instance of the Apply program.

Apply trail table
A replication source table at the control server that records a history of the full refreshes and differential refreshes performed against target tables.

archive log
The set of log files that are closed and are no longer needed for normal processing. These files are retained for use in roll-forward recovery. Contrast with active log.

asynchronous batched update
A process in which all changes to the source are recorded and applied to existing target data at specified intervals. Contrast with asynchronous continuous update.

asynchronous continuous update
A process in which all changes to the source are recorded and applied to existing target data after being committed in the base table. Contrast with asynchronous batched update.

audit trail
Data, in the form of a logical path linking a sequence of events, used for tracing the transactions that affected the contents of a record. (T)

auto-registration
A process in which replica and consistent change data tables are automatically defined as replication sources at the target server.

B

base aggregate table
A target table type that contains data aggregated form a source table or a point-in-time table at intervals.

base table
A table created with the CREATE TABLE statement. Such a table has both its description and data physically stored in the database. Contrast with view.

before-image
The content of a source table element prior to a full refresh or differential refresh, as recorded in a change data table, or in a database log or journal. Contrast with after-image.

binary large object (BLOB)
A sequence of bytes, where the size of the sequence ranges from 0 to 2 gigabytes. This string does not have an associated code page and character set. Image, audio, and video objects are stored in BLOBs.

bind
In SQL, the process by which the output from the SQL precompiler is converted to a usable structure called an access plan. During this process, access paths to the data are selected and some authorization checking is performed.

bind file
A file produced by the precompiler when the bind command or API is used with the BINDFILE option. This file includes information on all SQL statements in the application program.

BLOB
Binary large object.

C

Capture program
A replication program that reads database log or journal records to capture data about changes made to source tables. Contrast with Apply program.

cascade rejection
The process of rejecting a replication transaction because it is associated with a transaction that had a conflict detected and was itself rejected.

change aggregate table
A type of target table that contains data aggregations based on changes recorded for a source table.

change data (CD) table
A replication control table at the source server that contains changed data for a replication source table. The Capture program populates the CD table by copying the changes from the database log or journal. The contents of the CD table are then copied by the Apply program to the target table.

client
Any program (or workstation that it is running on) that communicates with and accesses a database server.

CLOB
Character large object.

cold start
A system start, using an initial program load procedure. Contrast with warm start.

common critical section table
A replication control table at the source server that is used to establish concurrency control between the Capture and Apply programs and to prevent an update replication cycle.

common pruning control table
A replication control table at the source server that coordinates the pruning of the change data and unit-of-work control tables. The values in this table indicate how much data has been replicated by the Apply program and can be safely pruned by the Capture program.

common registrations table
A replication control table at the source server that relates each source table or view to an associated change data table and consistent change data table, if applicable.

common subscription columns table
A replication control table that contains column details of target tables.

common subscription events table
A replication control table that defines the events that trigger replication, including the event name and time.

common subscription set table
A replication control table that defines the members of a subscription set including the set name, Apply qualifier, source server, target server, and status.

common subscription statements table
A replication control table used to store the optional SQL statements that can be run at the beginning or end of the set subscription cycle.

common subscription targets member table
A replication control table that maps the source and target table relationships within a subscription set.

complete
A table attribute indicating that the table contains a row for every primary key value of interest. As a result, a complete source table can be used to perform a full refresh of a target table.

condensed
A table attribute indicating that the table contains current data rather than a history of changes to the data. A condensed table includes no more than one row for each primary key value in the table. As a result, a condensed table can be used to supply current information for a full refresh.

conflict detection
The process of detecting an out-of-date row in a replica that was updated by a user application. When a conflict is detected, the transaction that caused the conflict is rejected. See also enhanced conflict detection,standard conflict detection and row-replica conflict detection.

consistent change data (CCD) table
A replication table that is used for staging data, with four replication control columns. It can be one of the following types:

consolidation replication
A replication model in which the data from multiple source tables is replicated to a single target table. Contrast with fan-out replication.

Control Center
A graphical interface that shows database objects (such as databases and tables) and their relationship to each other. From the Control Center you can perform the tasks provided by the DBA Utility, Visual Explain, and Performance Monitor tools.

control server
The database location of the applicable subscription definitions and Apply trail table.

control table
A table in which replication source and subscription definitions or other replication control information is stored.

copy table
See target table.

D

database log
A set of primary and secondary log files consisting of log records that record all changes to a database. The database log is used to roll back changes for units of work that are not committed and to recover a database to a consistent state. See also primary log and secondary log.

DAO
Data Access Object

database server
A functional unit that provides database services for databases.

DBA Utility
A tool that lets DB2 users configure databases and database manager instances, manage the directories necessary for accessing local and remote databases, back up and recover databases or table spaces, and manage media on a system using a graphical interface. The tasks provided by this tool can be accessed from the Control Center.

DBCLOB
Double-byte character large object.

DBMS
Database management system.

delimited identifier
A sequence of characters enclosed within quotation marks ("). The sequence must consist of a letter followed by zero or more characters, each of which is a letter, digit, or the underscore character.

Design Master
The original copy of a database. Only the Design Master supports changes to the database structure (table, query, and form design).

differential refresh
A process in which only changed data is copied to the target table, replacing existing data.

distinct type
A user-defined data type that is internally represented as an existing type (its source type), but is considered to be a separate and incompatible type for semantic purposes. See also user-defined type (UDT).

DMS table space
Database managed space table space.

E

enhanced conflict detection
Conflict detection that guarantees data integrity among all replicas and the origin table. The Apply program locks all replicas in the subscription set against further transactions, and begins detection after all changes made prior to locking have been captured. See also standard conflict detection and conflict detection.

external source table
A non-DB2 table that is manually updated to match the consistent change data table structure and defined as a replication source. See also consistent change data (CCD) table.

F

fan-out replication
A replication model in which data from one source table is copied to multiple target tables, thereby distributing the data to multiple locations. Contrast with consolidation replication.

foreign update
An update that was applied to a target table and replicated to the local table.

full refresh
A process in which all of the data of interest in a user table is copied to the target table, replacing existing data. Contrast with differential refresh.

G

gap
A situation in which the Capture program is not able to read a range of log or journal records and there is potential loss of change data.

I

internal CCD table
A consistent change data table that is a join of the change data table and the unit-of-work table at the source server.

J

join
A relational operation that allows for retrieval of data from two or more tables based on matching column values.

K

key
A column or an ordered collection of columns that are identified in the description of a table, index, or referential constraint.

L

large object (LOB)
A sequence of bytes, where the length can be up to 2 gigabytes. It can be any of three types: BLOB (binary), CLOB (single-byte character or mixed) or DBCLOB (double-byte character).

local database
A database that is physically located on the workstation in use. Contrast with remote database.

lock

  1. A means of serializing events or access to data

  2. A means of preventing uncommitted changes made by one application process from being perceived by another application process and for preventing one application process from updating data that is being accessed by another process.

locking
The mechanism used by the database manager to ensure the integrity of data. Locking prevents concurrent users from accessing inconsistent data.

long table space
A table space that can store only long string or large object (LOB) data.

M

member
See subscription set member.

migration

  1. The process of moving data from one computer system to another without converting the data.

  2. Installation of a new version or release of a program to replace an earlier version or release.

mobile client
The node, usually a laptop computer, where the mobile enabler and replication source and target tables used in a mobile environment are located. The mobile replication mode is invoked from the mobile client.

mobile replication enabler
A replication program that starts the mobile replication mode at the mobile client.

mobile replication mode
A mode of replication in which the Capture and Apply programs operate as needed rather than autonomously and continuously. This mode is invoked from the mobile client and allows data to be replicated when the mobile client is available for a connection to the source or target server.

N

nullable
The condition where a value for a column, function parameter, or result can have an absence of a value. For example, a field for a person's middle initial does not require a value.

null value
A parameter for which no value is specified.

O

object

  1. Anything that can be created or manipulated with SQL--for example, tables, view, indexes, or packages.

  2. In object-oriented design or programming, an abstraction consisting of data and operations associated with that data.

ODBC
See Open Database Connectivity.

ODBC driver
A driver that implements ODBC function calls and interacts with a data source.

Open Database Connectivity (ODBC)
An API that allows access to database management systems using callable SQL, which does not require the use of an SQL preprocessor. The ODBC architecture allows users to add modules, called database drivers, that link the application to their choice of database management systems at run time. Applications do not need to be linked directly to the modules of all the supported database management systems.

ordinary identifier
In SQL, a letter, which might be followed by zero or more characters, each of which is a letter (a-z and A-Z), a symbol, a number, or the underscore character, used to form a name.

P

package
A control structure produced during program preparation that is used to execute SQL statements.

partitioning key

  1. An ordered set of one or more columns in a table. For each row in the table, the values in the partitioning key columns are used to determine on which database partition the row belongs.

  2. In replication, an ordered set of one or more columns in a table. For each row in the source table, the values in the partitioning key columns are used to determine in which target table the row belongs.

performance monitor
A tool that lets database administrators use a graphical interface to monitor the performance of a DB2 system for tuning purposes. The tasks provided by this tool can be accessed from the Control Center.

point-in-time
A type of target table whose content matches all or part of a source table, with an added system column that identifies the approximate time when the particular row was inserted or updated at the source system.

predicate
An element of a search condition that expresses or implies a comparison operation.

primary key
A unique key that is part of the definition of a table. A primary key is the default parent key of a referential constraint definition.

primary log
A set of one or more log files used to record changes to a database. Storage for these files is allocated in advance. Contrast with secondary log.

Q

query
A request for information from the database based on specific conditions; for example, a request for a list of all customers in a customer table whose balance is greater than $1000.

R

RDBMS
Relational database management system.

referential integrity
The state of a database in which all values of all foreign keys are valid.

registration
The process of identifying a source table to DPROPR to make the table available for subscription.This is done by using the DataJoiner replication administration tool. The result is a row inserted into the ASN.IBMSNAP_REGISTER TABLE, which is located at the source server. See also replication source.

regular table space
A table space that can store any nontemporary data.

rejected transaction
A transaction containing one or more updates from replica tables that are out of date in comparison to the origin table.

replica
A type of target table that can be updated locally and receives updates from a user table through a subscription. It can be a source for updating the user table or read-only target tables.

replication
The process of taking changes that are stored in the database log or journal at the source server and applying them to the target server.

replication administrator
The user responsible for defining replication sources and subscriptions. This user can also run the Capture and Apply programs.

replication source
A database table that is defined as a source for replication. This type of table can accept copy requests and is the source table in a replication subscription set. See also subscription set and registration.

replication subscription
(1) A specification for copying changed data from replication sources to target tables at a specified time and frequency, with the option of enhancing data. It defines all of the information that is required by the Apply program to copy data.
(2) (For MS Jet) The process of requesting an automatic refresh or update of a set of one or more members. This is done by using the DataJoiner replication administration tool . The result is two rows into the ASN.IBMSNAP_SUBS_SET table, one or more rows into the ASN.IBMSNAP_SUBS_MEMBR and ASN_IBMSNAP_SUBS_COLS tables. These tables are located at the control server.

remote database
A database that is physically located on a workstation other than the one in use. Contrast with local database.

row-replica
A type of update-anywhere replica maintained by DataPropagator for Microsoft Jet without transaction semantics.

row-replica conflict detection
Conflict detection done row by row, not transaction by transaction, as is done for DB2 replicas.

S

secondary log
A set of one or more log files used to record changes to a database. Storage for these files is allocated as needed when the primary log is full.

SMS table space
System managed space table space.

source server
The database location of the replication source and the Capture program.

source table
A table that contains the data that is to be copied to a target table. The source table can be a replication source table, a change data table, or a consistent change data table. Contrast with target table.

spill file
A temporary file created by the Apply program that is used as the source for updating data to multiple target tables.

staging table
A consistent change data target table that is used as the source for updating data to multiple target tables during a full refresh.

standard conflict detection
Conflict detection in which the Apply program searches for conflicts in rows that are already captured in the replica's change data tables. See also conflict detection and enhance conflict detection.

subscription
See replication subscription.

subscription set
The specification of a group of source tables, target tables, and the control information that governs the replication of changed data. Updates are committed in the same transaction.

subscription set member
A member of a subscription set. There is one member for each source-target pair. Each member defines the structure of the target table and which rows and columns will be replicated from the source table.

T

table space
An abstraction of a collection of containers into which database objects are stored. A table space provides a level of indirection between a database and the tables stored within the database. A table space:

See also system managed space (SMS) table space, database managed space (DMS) table space, long table space, regular table space, temporary table space.

target server
The database location of the target table. Normally this is also the location of the Apply program.

target table
The table on the target server to which data is copied. It can be a user copy table, a point-in-time table, a base aggregate table, a change aggregate table, a consistent change data table, or a replica table.

temporary table
A table created during the processing of an SQL statement to hold intermediate results.

temporary table space
A table space that can store only temporary tables.

trace table
A table that contains a high-level record of the execution of the Capture program.

trigger
In DB2, an object in a database that is invoked indirectly by the database manager when a particular SQL statement is run.

tuning parameters table
A table at the source server that contains timing information used by the Capture program. The information includes:

two-phase commit
A two-step process by which recoverable resources and an external subsystem are committed. During the first step, the database manager subsystems are polled to ensure that they are ready to commit. If all subsystems respond positively, the database manager instructs them to commit.

U

uncommitted read (UR)
An isolation level that allows an application to access uncommitted changes of other transactions. The application does not lock other applications out of the row it is reading, unless the other application attempts to drop or alter the table.

unit-of-work table
A replication control table at the source server that contains commit records read from the database log or journal. The records include a unit-of-recovery ID that can be used to join the unit-of-work table and the change data table to produce transaction-consistent change data. For DB2, the unit-of-work table optionally includes the correlation ID, which can be useful for auditing purposes.

update
A process in which the changes to data in a source table are used to refresh a target table. This process is also known as differential refresh.

update replication cycle

UR
Uncommitted read.

user copy table
A target table whose content matches all or part of a source table and contains only user data columns.

user-defined type (UDT)
A data type that is not native to the database manager and was created by a user. See also distinct type.

user table
A table created for and used by an application before it is defined as a replication source. It is used as the source for updates to read-only target tables, consistent change data tables, and replicas.

V

view
A logical table that consists of data that is generated by a query. Contrast with base table.

Visual Explain
A tool that lets database administrators and application programmers use a graphical interface to display and analyze detailed information on the access plan of a given SQL statement. The tasks provided by this tool can be accessed from the Control Center.

W

warm start
A restart that allows reuse of previously initialized input and output work queues. Contrast with cold start.

warm start table
A table used by the Capture program to save position in a DBMS log for later reference during warm start.

[ Top of Page | Previous Page | Next Page | Table of Contents | Index ]

[ DB2 List of Books | Search the DB2 Books ]