< Back to list

Galera Cluster for MySQL

Coderhip's Galera Cluster for MySQL is a database clustering solution that is using synchronous replication to build a single cluster entity, a so-called multi-master cluster, in which all servers have identical data each time. Galera Cluster is active-active cluster. The application can read and write to any server, which provides scale-out opportunity for both read and write transactions. Adding nodes to the cluster is completely automated, unlike most of the alternatives. Removing nodes (scale-in) is simply a matter of shutting down unwanted nodes. Since the application developer no longer needs to implement cumbersome read-write splitting logic, the scale-out potential can be immediately realized without the need to change the application logic, which saves time when developing applications and when migrating existing ones to Galera Cluster. There is no delay in replicating the data and therefore Galera offers better security against data loss and inconsistent databases. If there is a failure in one of cluster nodes - or the whole datacenter - the application/service does not notice anything but will continue serving the users using the other nodes - which may also be located in other data centers. The replication implementation can use multiple slave threads, without limitations on the application or database schema that is used. This ensures good performance and benchmarks have shown that Galera does not cause any performance overhead, on the contrary it only increases performance when used as a scale-out solution. Galera Cluster powers MariaDB Enterprise Cluster and Percona XtraDB Cluster.

Documentation URL: http://galeracluster.com/documentation-webpages/
Download URL: http://galeracluster.com/downloads/

Key Features and Benefits

  • Galera Cluster: significant improvements to MySQL high availability/replication

  • True Multi-master, active-active topology. Read and write to any node at any time

  • Synchronous Replication. No slave lag, no data is lost at node crash. Parallel Replication on Row Level. For better performance. For any workload.

  • Automatic Node Provisioning and Elasticity. Scale out, Scale in. Works in LAN, WAN, Hybrid Clouds between Data Centers.

  • No complex (master-slave) Failover needed.

  • Transparent to Applications. No need to do read/write splitting (MySQL).

Category

Database & Data Management

Red Hat Certifications

This product has been certified to run on the following Red Hat products and technologies:

Target Product Level
Red Hat Enterprise Linux 7.x Self-Certified