MFE-IT

MariaDB Galera Cluster Training Course - Deploy a Robust, Fault-Tolerant Architecture for Service Continuity

Reference: W/EN/MG
1,690 € excl. VAT
3 days
21 h
11 Jan. 2027
Remote
1 Feb. 2027
Remote
1 Mar. 2027
Remote
5 Apr. 2027
Remote
3 May 2027
Remote
7 Jun. 2027
Remote

Each session will take place even if only one person is registered (except in cases of force majeure).

MariaDB Galera Cluster Training

Description of the MariaDB Galera Cluster Training Course

This MariaDB Galera Cluster training course will enable you to understand the fundamentals of MariaDB, configure a functional Galera cluster, and master failover, recovery and monitoring scenarios.

In critical environments, databases must remain available at all times. MariaDB, coupled with Galera Cluster, offers real-time synchronous replication between multiple nodes, ensuring business continuity, fault tolerance, and horizontal scalability.

Also discover our MySQL Training Course.

Format

Remote (recorded sessions). 

GOOD TO KNOW

This training course includes numerous exercises (60% practical) to enhance learning. Each session will take place even if only one person is registered (except in cases of force majeure). A preliminary interview is held between the participant and/or a company representative in order to fully assess the participant’s profile (level, needs, professional context, challenges, etc.).
Assessment : during the training course, the trainer assesses the participants’ progress through multiple-choice questions, role-playing exercises and practical work. Participants receive a certificate of completion at the end of the training course. 

objectives of this MariaDB Galera Cluster Training Course

By the end, each participant will be able to :

  • Understand MariaDB architecture and how Galera Cluster works.
  • Install, configure, and administer a secure and redundant Galera cluster.
  • Ensure data synchronisation and avoid conflicts.
  • Manage failure, split-brain, and automatic recovery scenarios.
  • Optimise performance and monitor cluster health.
  • Integrate the cluster into production environments.

Prerequisites

  • Good foundation in Linux administration
  • Basic knowledge of databases (MySQL or MariaDB)
  • Experience in managing production services (CI/CD, monitoring, etc.)
  • Because each participant has a unique background and expectations, a preliminary interview with our expert allows us to precisely identify their objectives, level and professional challenges.
    This enables us to tailor the training content to ensure relevant and personalised learning.

Target Audience

Perfect training for system administrators, DevOps engineers, infrastructure managers, or IT architects.

Detailed of this MariaDB Galera Cluster Training

MariaDB architecture and replication principles

MariaDB vs MySQL, storage engines, master-slave vs multi-master architecture, synchronous replication concept.

Operation, SST vs IST, quorum, write sets, multi-node topology, system requirements.

Step-by-step deployment of MariaDB with Galera on multiple machines (Linux), firewall, SSL, communication testing.

Adding/removing nodes, conflict management, control tools (galera_new_node, wsrep_status, etc.)

Securing inter-node exchanges, audits, monitoring with Prometheus, Grafana or ClusterControl.

Behaviour in the event of failure, split-brain, recovery, automatic restart, failover testing, PRA/PCA.

This training course :

Combines advanced replication mechanisms + high availability
Includes practical multi-node exercises on simulated environments
Focuses on resilience and service continuity for critical contexts
Prepares you for real-world deployments in cloud or on-premise production environments

FAQ – MariaDB Galera Cluster Training

MariaDB Galera Cluster is a multi-master clustering solution for MariaDB, based on wsrep synchronous replication. All nodes can read and write, with near-immediate transaction propagation and fault tolerance. It delivers high availability without manual failover. MFE-IT trains DBAs on its deployment and operation.

Asynchronous master-slave replication only allows one node to accept writes, with potential lag. Galera provides synchronous multi-master replication: every node can write, and consistency is guaranteed before commit. Galera thus delivers active/active HA and transparent failover, at the cost of greater sensitivity to network latency.

You use wsrep_* variables (wsrep_cluster_size, wsrep_local_state, wsrep_flow_control_*), MaxScale or HAProxy as a load balancer, and monitoring tools (Prometheus, Zabbix). The MFE-IT training covers cluster bootstrap, split-brain resolution and performance optimization.

The training lasts 3 days (21 hours), remote or on-site, with a maximum of 3 participants per session and 30 days of post-training support included.

Would you like to know about upcoming sessions ?

Would you like to schedule this Training Couse on a specific date ? Contact us by email or by filling out the contact form.