PostgreSQL Replicator Features

General:

“Store and forward” asynchronous data replication.
Large Object replication is now supported
No Postgresql patch required
Based on PostgreSQL catalog and inheritance
Works on 6.5.x , 7.0.x and 7.1.x versions. Large Objects replication requires 7.1.x versions.
Easy replica administration by replicator-system-tables management

Data Ownership Models:

Table level replicated database set
Table level data ownership model
Master/Slave, Update Anywhere , Workload Partitioning (and more) data ownership models are supported.
Additional replica customization parameters available at table level.

Conflicts Resolution Algorithms:

Table level conflict resolution algorithm
Automatic forwarding and rolling-back transactions during database synchronization, according to the table conflicts resolution algorithm.
Several Conflict Resolution Algorithms are provided.
Custom Conflict Resolution Algorithms can be added.

Replica Process

Socket based client server approach.
Replica process can be either scheduled or triggered by a maximum-operations threshold. Operation counters can be disabled maximizing performances
Operations that were rolled back by the conflict resolution algorithms are logged and can be submitted again.
Users whose operations were rolled back by the conflict resolution algorithms can be alerted via e-mail, and get transaction details (table level warning option) in “human readable” format.
Status of the replica process can be monitored.

Tests:

The system has been tested successfully on 6.5.x , 7.0.x and 7.1.x versions.
The system is currently used to manage databases having a master-detail relations with over 50,000 records in both master and detail table.

Auxiliary tools provided:

We are working on:

Large Objects caching on I and II level storage systems (with removable devices support)