Glossary

Replication Terminology

In Btrieve 7.x, 8.x, and 9.5, this refers to a file open mode that provides improved response time over Normal mode when updating data files. However, Accelerated mode disables the MicroKernel logging capability. Therefore, the MicroKernel cannot guarantee transaction durability or atomicity on files opened in Accelerated mode.

If you are using Btrieve 6.x, Accelerated mode is equivalent to Normal mode, except that opening a data file in Accelerated mode cancels the effect of flagging a file as transactional.

Activated Database

A database on a specific site that is capable of replicating data to or from another site's database, provided the other site is a member of the same replication network. Activated databases contain a copy of the replication template with information about the sites that constitute the replication network.

The process of creating a new database for replication – or enabling an existing one – and establishing the database as belonging to a site.

Change Capture / Change Apply

The change capture/change apply mechanism is a major subsystem of any replication implementation and is used to track and propagate modifications to replicated databases. Change capture occurs concurrent with database application execution, whereas change apply occurs during database synchronization. The DataExchange change capture mechanism is not log-based. It is based on control tables that store only the most recent changes. It reflects the current state of the database and not the history of prior states as in log-based mechanisms.

Each replicated table in the database has a control table associated with it. Control tables store a record for each record in its corresponding user table, and contain associated information such as the creation date of the record, last change date for the record, and so forth. By creating a control table for each replicated table, PSQL ensures that you do not have to make any changes to your existing tables to enable replication.

DataExchange Manager

A graphical user interface for managing tasks, users, and sites in a replication network. DataExchange Manager can:

•

Manage replication schedules

•

Manage site information

•

Manage local or remote sites

Database Network Analysis, the replication system database.

Encryption is the process of disguising a message to hide its contents. An encrypted message is referred to as ciphertext; the original, or decrypted message is referred to as plaintext.

Every replication network must have a First Site, which is the first site to be activated for replication. The First Site contains information about the whole replication network, such as, sites available for replication, etc. It also contains a copy of the whole database. Any other site in the replication network is a Partner Site.

Globally Unique IDs

A record ID that is guaranteed to be unique throughout your entire network, even if some sites create many new records in the same table between replications.

A network configuration that consists of a centralized server that maintains a complete copy of the database (the hub) and several connected servers or workstations with subsets of the replication data (spokes). The replication process typically consists of all spokes replicating with the hub at a predetermined time. Then, upon this successful initial replication, the updates are distributed to all spokes via a second replication cycle from the hub to each spoke site.

In Place Activation

Activation of an existing database by either using a replication template or by using a design exported from a template.

A nonreplicated table. Each site can have different local tables.

A network configuration that is similar to hub and spoke, but with additional levels (or tiers) in the network. An n-tiered configuration can be thought of as a hub and spoke network, where the spoke sites are also hub sites for the next tier in the configuration hierarchy. Thus, a centralized server can replicate with three second tier server sites, each of which, in turn, communicates with several desktop and notebook machines.

N-tiered configurations allow some scalability that is not available in hub and spoke networks. Configuring these network topologies and balancing the load as the network grows requires some planning.

A network is a collection of sites that can replicate among themselves. You can create separate environments, called networks, in which to run different versions of your database. For example, Version 1.0 could be running on a Production network, while Version 2.0 could be running on a Test network with its own separate and distinct data.

Open Data Base Connectivity, a standard programming interface used to connect to a variety of data sources. DataExchange connects to all of its supported databases using ODBC.

Any site that is not the First Site in a replication network.

A replication network configuration in which every node, regardless of its size, amount of data, or location, behaves as a true peer to other sites in the network. Each node with a replication engine can schedule and manage replication and provides it share of the replication processing burden. Peer-to-peer is the default replication network configuration.

A project represents a single replicated database. Each time you enable a different database, you must specify a new project. A project can contain a number of networks and releases.

A term used for a mode of synchronization that involves unidirectional flow of data, as opposed to the default bidirectional flow of data commonly referred to as replication. A refresh is used to manually override the conflict resolution rules by specifying a prevailing site. For example, a “refresh local” operation overwrites all local data with data from a specified site.

A release represents a change in your database's schema. Therefore, each time you modify your database schema you must create a new release. A replication project can contain several networks and releases.

The default database synchronization mode that involves bidirectional flow of data between two or more sites based on conflict resolution rules.

Replication Engine

The component that manages the replication process. All replication is performed in the background and can be fully automated according to schedules that you define.

PSQL Control Center contains the user interfaces to the replication engine from which you can:

•

Remotely manage replication sites

•

View current state of replication nodes

•

Query history of replication activity

•

Provide multiple views of replication statistics and information

•

Filter replication messages/logs

Replication Network

A collection of sites that can replicate among themselves. A replication network can contain several releases.

Replication Template

A copy of a database that includes the user’s database tables, replication control tables, system tables, and optionally, the user data.

A replication-enabled database that can replicate and synchronize database information with other sites across TCP/IP networks over an ODBC interface. Each local replication-enabled database is one site regardless of the number of users accessing it or the amount of data stored in it. A site can be:

•

One computer with its own local database

•

A local database shared by a local network

•

A large client-server database used by hundreds or thousands of users

A site’s authorization to replicate with other sites on the replication network.

See Replication Template.