Meeting the demands of today’s web hosting and cloud hosting applications, Oracle has announced the general availability of MySQL Fabric, providing an integrated system that makes it simpler to manage groups of MySQL databases.
MySQL Fabric is an open source framework to manage farms of MySQL servers. The new Oracle system delivers both high availability – via failure detection and failover – and scalability through automated data sharding.
Available for download at http://dev.mysql.com/downloads/fabric/, this integrated and extensible system would make it easier to scale out or add high availability to MySQL applications.
MySQL Fabric capabilities:
Automatic failure detection and failover when used with MySQL Replication for high availability, including:
- Monitoring of the master database; if the server fails, MySQL Fabric selects one of the slave databases and promotes it to be the new master.
- Automatic routing of transactions to the current master and load-balancing of queries across slave databases. Topology and state of the servers are transparent to the application.
Automated data sharding and re-sharding make it easier for development and operations, or DevOps, teams to manage the process. This enables users to:
- Allow tables to be sharded for scale-out of both reads and writes.
- Select which tables should be sharded and specify the columns to be used as sharding keys, including whether to use a HASH or RANGE based mapping.
- Move existing shards to new servers or split them into multiple shards.
Extensions to PHP, Python, and Java connectors allow transactions and queries to be routed directly to the correct MySQL Server, eliminating the extra latency caused by passing through a proxy.