Readers will find a generalized discussion of one of the most important features in MySQL Cluster 7.2, namely, Multi-Site Clustering, in Mat Keep's blog. While, as Keep points out, MySQL has long offered Geographic Replication, Multi-Site Clustering enables the deployment option of splitting data nodes across data centers. This feature, he writes, allows users to synchronously replicate updates between data centers with no need to modify applications or schema for conflict handling. Moreover, he adds, automatic failover between sites in the event of a node failure is part of the new feature. He notes as well that MySQL Cluster offers high availability by maintaining a configurable number of data replicas, all of which are synchronously maintained by a built-in 2 phase commit protocol. This enables data node and communication failures to be detected and handled automatically. On recovery, data nodes automatically rejoin the cluster, synchronize with running nodes, and resume service, he writes.
Scaling Web Services with MySQL Cluster
Read More ...