Figaro High Availability, or HA, adds support for replicated systems to Figaro Transactional Data Store (TDS). There are two basic reasons to build a replicated highly available system: scale and reliability.
- Scale your transactional application beyond the processing constraints of a single system.
- Deliver a highly reliable system where the failure of one node does not cause system-wide failure.
Replicated databases come in some different configurations. The HA layer of the Oracle Berkeley DB database engine can be configured for any of the replicated designs that involve a single writer as a master and multiple replicas allowing for read-only access to data.
Figaro HA supports the entire range of configurations from a simple hot standby with failover setup to a geographically diverse set of tiered replica clusters. Figaro HA makes no assumptions about how the nodes communicate with one another; it will support TCP/IP communications networks.
Figaro HA supports any topology regardless of the number of replicated nodes within it. Upon notification of a failure of the master node, an election algorithm will identify a new master so that processing can continue uninterrupted. Figaro HA can even be used as a completely diskless in memory reliable distributed storage system.
Choose Figaro HA when:
- Single-system throughput isn't enough to meet your data requirements.
- System uptime of %99.999 or better is a requirement.
- The system must be able to survive node failure.
- There is a need for consistent online backup of system data.
- Your architecture calls for a hot-failover strategy.
- Your hardware device has multiple redundant systems and must be able to fail from one to the other gracefully.
See http://bdbxml.net/blog/figaro-6-1-4-changelog.html for more information.
This package has no dependencies.