Types of Replication

This page describes ClustrixDB replication, which uses a single stream to write and read binlogs. ClustrixDB also supports parallel replication, which is more performant and can be used to replicate to/from ClustrixDB only. 

ClustrixDB supports the MySQL's replication protocol as both a Slave and a Master. 

• ClustrixDB supports replication to/from MariaDB 10.3 with no special configuration.
• ClustrixDB supports replication to/from MySQL 5.7 with GTID mode disabled. 

Supported Topologies

Here is a sampling of the topologies for replication supported by ClustrixDB:

ClustrixDB can serve as a Master or a Slave to MariaDB 10.4, MySQL 5.7, or ClustrixDB:

ClustrixDB also supports slaving from multiple distinct binlogs (consolidation) and Master-Master replication:

ClustrixDB also supports the creation of multiple binary logs (binlogs), each of which can correspond to specific databases and can be accessed by distinct replication Slaves. When running multiple-Slave configurations, ClustrixDB takes advantage of each node by load-balancing replication connections in a round-robin fashion. For assistance in determining the best topology for your application or whether a specific topology is supported, contact Clustrix Support. 

Caveats for Replication:

• ClustrixDB does not support GTIDs
• ClustrixDB does not support ring replication topologies

See also section on Replication concerns when performing Online Schema Changes.

The following topics explain how to configure ClustrixDB replication:

• Using ClustrixDB as a Replication Slave
• Using ClustrixDB as a Replication Master
• Configuring Replication Failover
• Parallel Replication