Daffodil Replicator is a data replication software that synchronizes data residing on corporate backend databases and desktop databases operating in the network. Daffodil Replicator performs data replication based on the 'Publish and Subscribe' model (in client and server architecture). A publication is a collection of one or more tables required for data replication. A subscription is a copy of tables involved in publication on client side.

Replicator supports bi-directional data replication by either capturing a data source snapshot or synchronizing the changes. It monitors data changes in the tables involved and synchronizes all data changes made by the subscriber and the publisher on periodic basis or on-demand by the subscriber. While synchronizing with one or more target data source, Replicator uses pre-defined conflict resolution algorithms to resolve conflicts between publisher and subscriber. The publications and subscriptions are defined using GUI or APIs on existing database servers.

Daffodil Replicator has been developed for easy replication between homogeneous and heterogeneous data sources. It supports data replication between Oracle, SQL Server, DB2, Daffodil DB, PostgreSql and Derby, enabling high-performance distributed connectivity with corporate databases residing on Windows, Linux, Unix or Solaris. Replicator's publish-and-subscribe model allows a client application to have subscriptions to publisher defined on the server side, which can consist of single or multiple data tables. Replicator filters enable application developers to specify conditions for individual tables, through which the client can subscribe to a subset of rows.

Daffodil Replicator supports the following methods of Data Replication:

Snapshot Replication

Snapshot replication simply takes a "snapshot" of the database on new server and moves that data to another database

Merge Replication

Merge replication compares data from Publisher and Subscriber and updates the changes to both the databases. It uses initial synchronization by taking the snapshot of data on the Publisher and moving it to subscribers. Merge Replication is also called as Change-based Replication or Synchronization.

Data Replication Model

Conformation to Publication and Subscription for Data Replication

The Publication (collection of one or more tables defined by developers) and Subscription (a subset of a particular publication on a Publisher) allows a Publisher to publish data to one or more Subscribers where Replication server enacts the role of a listener on a pre-defined communication port and the Subscriber makes periodic requests for Data Replication operations.

Publisher

Publishers are the basic building blocks of a Replicator's synchronization network. Each publisher forms part of a server-client network and is assigned a unique name. The name as well as the other concerned data of each publisher are stored in the publisher database i.e. all relevant information about the publishers, is stored in a database.

Subscriber

Subscriber subscribes to Publisher's publications.

Replicator Publisher and Replicator Subscriber are responsible for managing data replication across the network. Both the Publisher and Subscriber are objects instantiated by user applications and perform the core functions associated with the synchronization of data. The Replication Server API gives method calls to Publisher and Subscriber, which enable Subscriber to perform synchronization operations.

Each Publisher and Subscriber defines one or more JDBC-enabled databases as its data source. In the publish and subscribe mode, Subscribers pull complete copies or partial (change-only) copies of data to each other to maintain Data Replication.

Configuration:

Configuration involves creation of Publishers and Subscribers, selection of data sources, and also defining or creating of publications and subscriptions. While creating a publication, one or more tables must be specified those are to be published. Each table to be specified in publication must have a primary key. During configuration, you can specify Row Filtering and Conflict Resolution as well.

Execution:

Execution consists of the actual operations that are performed for data replication between Publishers and Subscribers. During runtime, Snapshot () and Synchronize () methods are called where snapshot () invokes an operation to retrieve a complete copy of refreshed data and synchronize () ensures the integrity and consistency of data propagated amid various data sources.

Row Filtering:

A publication represents data from selected Publisher tables, which is filtered according to a specified criterion. Filters specified by Publishers on the tables of a publication determine what data will be made available to a Subscriber. Additional filters specified by Subscribers on the tables of a publication further refine the data according to the specific needs of the Subscriber. Regardless of where the filter is specified, the filter applies only to data in publication tables accessed by the Publisher. Filters may be specified for tables on publications or subscriptions. Filters, however, are always applied to publication only.

Conflict Detection and Resolution

Replicator is capable of detecting and resolving conflicts that tend to occur while data replication operations. Conflict resolver comes into play only when operation has been performed on same records on both Server side as well as Client side.