The Tungsten Replicator provides open source database-neutral master/slave replication. Master/slave replication is a highly flexible technology that can solve a wide variety of problems including the following:
- Availability – Failing over to a slave database if your master database dies
- Performance Scaling – Spreading reads across many copies of data
- Data Protection - Preventing data loss by maintaining live copies with guaranteed consistency
- Cross DBMS Integration – Replicating data in real time between different database types
- Cross-Site ClusteringÂ - Maintaining database replicas across WANs
- Change Data Capture – Extracting changes to load data warehouses or update other systems
- Zero Downtime Upgrade – Performing upgrades on a slave server which then becomes the master
The Tungsten Replicator architecture is flexible and designed for rapid extensibility. It includes pluggable extractors, event filters, and appliers, as shown in the following diagram.
The Replicator has a number of specialized features designed to make it easy to install and reliable to operate.
- Simple configuration file format and startup
- Global transaction IDs for all replicated events to enable seamless promotion of any slave to master andÂ fungible backups
- Generic structure for replicated events to allow cross-platform, cross-DBMS, and cross-version data transfer
- State machine execution model to ensure deterministic behavior free from race conditions
- Smooth procedures for planned and unplanned failover.
- Command line administrative utility as well as JMX interfaces
- Native support for MySQL and replication to JDBC data sources