High availability and disaster recovery:

replication, clustering and failover

When it comes to databases, high availability (HA) and disaster recovery (DR) should not require expensive options, separate software or complex application code. HA/DR should come out of the box. It should be built-in, configurable and easy to use.

Upcoming webinar

High Availability: replication, clustering and failover

This webinar will explore high availability strategies and their impact on performance and consistency, explain how replication and clustering work in MariaDB TX and introduce the new automatic failover/rejoin and switchover capabilities in MariaDB MaxScale 2.2.

Attendees will learn:

  • Strategies for achieving high availability
  • Pros and cons of different approaches
  • Use cases and different topologies
  • What to expect from automatic failover
  • Handling topology changes with a proxy
High Availability

March 27, 2018

Register now

Why choose MariaDB TX for HA

Advanced replication and clustering

MariaDB TX replication supports advanced configuration options: enhanced semi-synchronous replication for consistency, optimistic parallel replication for performance, delayed replication for disaster recovery, read throttling for reduced load, binlog compression for reduced IO and more. In addition, multi-master clustering is supported, and can be combined with replication to create tiered topologies.

Simplified automatic failover

MariaDB TX now supports automatic failover via its database proxy, MariaDB MaxScale. If automatic failover is enabled and the master fails, it will automatically promote the most up-to-date slave. If the previous master is recovered and automatic rejoin is enabled, it will automatically configure it as a slave. In addition, DBAs can now issue a command to perform a switchover on demand.

Transparent topology changes

MariaDB MaxScale, as a database proxy, enables DBAs to add or remove database nodes on demand – applications are not impacted because topology changes are transparent to them. MariaDB MaxScale will automatically route queries to the right database nodes when the topology changes, regardless of whether there was an automatic failover or a new database node was added.

Denial of service protection

MariaDB MaxScale includes a built-in database firewall and result set limiting to prevent malicious users from reducing database availability via denial of service attacks – excessive repetitive queries or queries returning excessive results (e.g., millions of rows).

Disaster recovery

MariaDB TX includes backup/restore tools with encryption and compression support as well as point-in-time rollback. MariaDB Flashback enables DBAs to recover data faster by rolling back transactions to a previous point in time rather than performing a restore from a backup.

Example: HA topology with intelligent routing

The example topology below uses multi-master clustering with synchronous replication and splits reads and writes – routing all writes to a single node (chosen by the database proxy) and routing all reads to the other nodes (with load balancing).

MariaDB Master Synchronous Replication

HA strategies for MariaDB TX

MariaDB TX can use master/slave replication, multi-source replication, multi-master clustering or any combination thereof – enabling DBAs to create topologies based on complex requirements, and with multiple configuration options, optimize for performance and/or consistency.

White Paper