How does Tungsten Cluster compare to Galera Cluster for MySQL high availability?

Tungsten Cluster supports a broader range of topologies than Galera Cluster, including disaster recovery, active-passive, active-active, and geo-scale deployments. Galera Cluster is limited to single-site high availability and multi-primary within one region, making it unsuitable for multi-region or geographically distributed deployments.

What are the performance limitations of Galera Cluster compared to Tungsten?

Galera Cluster uses synchronous replication, which introduces performance overhead that grows significantly as workload and transaction sizes increase. Tungsten Cluster uses asynchronous replication, which adds almost no performance penalty and scales predictably under high write loads.

Does Galera Cluster include a built-in proxy like Tungsten?

No. Galera Cluster does not include a built-in proxy. Read/write splitting and load balancing require separate third-party tools. Tungsten Cluster includes an integrated proxy (Tungsten Connector) that automatically handles read/write splitting without any application changes.

Can Galera Cluster support multi-region or geo-distributed deployments?

No. Galera Cluster is limited to single-site deployments. It does not support multi-region active-passive, active-active, or geo-scale topologies. Tungsten Cluster supports all of these through its CAP, CAA, and DAA topology options.

How does Tungsten Cluster support compare to Galera Cluster support?

Tungsten Cluster includes 24/7/365 expert support with an average urgent response time under 3 minutes, provided by engineers with 20+ years of DBA and site reliability experience. Galera Cluster's standard support offers a 4-hour response time, which may be inadequate for mission-critical production environments.

Tungsten Cluster vs. Galera Cluster

MySQL and MariaDB Clustering Solutions Comparison

Learn why Tungsten Clustering is a better (complete, proven) and more cost-effective (software, 24/7 support and consulting all included) MySQL High Availability (HA), Disaster Recovery (DR) and Geographically-Distributed (Geo-Scale) solution compared to Amazon AWS Aurora for MySQL.

For any MySQL clustering to be a complete high-availability and disaster recovery solution, it requires three core components working in sync: connectivity (proxy), management (orchestration) and replication (local/global).

Tungsten Cluster

Tungsten Clustering includes three highly integrated and fully tested components Tungsten Proxy, Tungsten Manager and Tungsten Replicator working together to provide both local (individual clusters) and global (multiple clusters working together) data service for continuous 24/7 MySQL operations.

Connectivity Complete and Integrated (Tungsten Proxy)

Orchestration Integrated Business Rules (Tungsten Manager)

Replication Advanced Local and Global (Tungsten Replicator)

 Connectivity
 Orchestration
 Replication

Galera Cluster

Galera Cluster is a decent solution for local MySQL HA needs, especially when dealing with smaller sized transactions. As workloads increase, transactions become larger, or if geo distribution is needed, Galera Cluster becomes less attractive due to significant performance overhead. Also a 3rd-party proxy or load balancer would be needed for maximum performance and availability benefits.

Connectivity Direct (benefits from 3rd party proxy or load balancer)

Orchestration Limited

Replication Sync replication for local needs, N/A for multi-site/global needs

Galera Cluster Comparison

Use Cases Supported

	Tungsten Cluster	Galera, MariaDB or XtraDB Cluster
High Availability	Supported	Supported
Disaster Recovery	Supported	Not Supported
Zero Downtime Maintenance	Supported	Supported
Continuous Operations	Supported	Supported
MySQL Multi-Primary	Supported	Supported
Active/Passive	Supported	Not Supported
Active/Active	Supported	Not Supported
Geo-Scale	Supported	Not Supported

Features & Benefits

	Tungsten Cluster	Galera, MariaDB or XtraDB Cluster
Continuous Operations	4.67	2.50
Zero Downtime Maintenance	Both database, operating system and hardware changes & upgrades	Supported
High Availability	Automatic failover of primary	Multi-master in use
DR with Active/Passive Cluster	Automated site level switch with one failover command	Must mix replication technologies, involves DIY
DR with Active/Active Cluster	Only requires repointing the apps	Active/Active not supported
DR Failback	Single command, fast	Manual rebuild needed
Backup & Data Recovery	Flexible backups using existing external tools	Standard tools; however requires additional steps compared to other technologies
Scalability/Read replicas	4.40	3.80
Scalability/Read replicas	Supported	Supported (all active)
Scalability/Physical location	Deploy anywhere, on-prem and in the cloud (also hybrid- and multi-cloud)	Deploy anywhere but single site only
Scalability/Instance size	Choose the instance size, minimal requirements	Choose the instance size, minimal requirements
DB Node Auto-Scale	No auto-scale. Requires manual steps	No auto-scale. Requires manual steps
DB Cluster Scaling	Ideal for the SaaS apps (Pod Architecture)	Pod Architecture
Geographic Scaling	5.00	0.67
Scalability/Read replicas	Across regions, availability zones, and clouds	Mix replication technologies, no management
Active/Active Geo-Clusters	Multi-site active/active supported and widely deployed	Not supported
Active/Passive Geo-Clusters	Supported and widely deployed in cross-regions and cross-clouds, also on-premises	Complicated due to mixing of replication technologies and no management layer
Performance	4.75	2.75
Smart-Scale	Tungsten Proxy with automatic read/write split, no application changes	No proxy included
Load Balancing	Highly configurable with automatic read/write splitting	All masters, but requires application logic or additional proxy (not included)
Up to Date Read Replicas	Using asynchronous replication	All masters with synchronous replication
Robust Database Performance	Tune your favorite database, asynchronous replication adds almost no performance penalty	Depends on workload due to synchronous replication
Transparency	5.00	2.67
No Application Changes	Works out of box without app changes	Some features require application changes, certain SQL constructs not supported
All MySQL Versions	Native MySQL: Community, MySQL Enterprise, MariaDB, Percona Server, all versions	Based on Galera builds of MySQL
Compatibility With All Database Engines	All MySQL engines	InnoDB only
Security	4.33	4.33
Security/Encryption at rest	Can leverage existing cloud or filesystem encryption technologies	Can leverage existing cloud or filesystem encryption technologies
Security/Encryption in flight	Supported for all traffic	Supported for all traffic
Compliance (PCI, HIPAA)	Can be used in security/privacy standards	Can be used in security/privacy standards
Management and Monitoring	4.80	3.40
Deployment	Easy to deploy	Complex, easier with 3rd party tools such as ClusterControl; or Galera Manager (on AWS only)
Cluster Management	Easy management via web-based GUI	Complex due to issuing commands to control replication; Galera manager available, but brand new (AWS only)
Monitoring - GUI	Full GUI based cluster monitoring. Performance monitoring with Prometheus	Complex due to issuing commands to control replication, Galera manager available but brand new (AWS only)
Alerts / Notifications	Integration with notifications systems	Scripted
OS/DB Patching	Assistance with scripts, no downtime	Assistance with scripts, no downtime
Real-time Data (Analytics, Data Lake, Reporting)	5.00	3.00
Replication Into Analytics	Real-time to the data analytics engines (AWS Redshift, Hadoop, Vertica, many others)	Using native replication, 3rd party or DIY tools
Filtering	Complex filtering at schema and data level	Basic filters, not useful in multi-master
Cluster Aware	Replication stream not interupted by loss of a single cluster node	Replication stream not interrupted by loss of a single cluster node

Support Services

	Tungsten Cluster	Galera, MariaDB or XtraDB Cluster
24/7 Support	5.00	3.00
24/7 Support	24/7 support with high, near 100% customer satisfaction	24/7 directly from Codership, or through Percona or MariaDB, 4 hour reponse time
Fast Support Reponse Time	The fastest in the industry, averaging less than 3 minutes for Urgent support request.	4 hours, could be catostrophic for enterprise
Support Team Experience	Each member has 20+ years experience on DBA and Site Reliability	Core developers among support staff

Intangibles

	Tungsten Cluster	Galera, MariaDB or XtraDB Cluster
MySQL Version Support	5.00	3.00
MySQL Community	Fully supported, all versions up to 8.0	8.0.21
MySQL Enterprise	Fully supported, all versions up to 8.0	Not supported
MariaDB	Fully supported, all versions up to 10.5	Most MariaDB versions with MariaDB Galera
Percona Server	Fully supported	Most Percona versions with Percona XtraDB
Infrastructure and Platform	5.00	3.00
On-premises Deployment	Fully supported	Fully supported
Cloud Deployment	All cloud environments (AWS, Google Cloud, MS Azure) are supported	All cloud environments (AWS, Google Cloud, MS Azure) are supported
Multi-Cloud	Fully supported	Except for complicated DIY deployments
Hybrid-Cloud	Fully supported	Except for complicated DIY deployments
Product Maturity	5.00	2.33
Complete	One solution for HA, DR & Geo-Scale	No proxy, no multi-site
Integrated	Comes with Manager, Proxy & Replicator	No proxy, which is critical for MySQL clustering
Fully Tested	Proven in business-critical customer environments	Proven in business-critical customer environments
Total Cost of Ownership (TCO)	4.33	3.67
Low Subscription Cost, High Value	Instance size does not matter, non-linear pricing (additional clusters are at increasingly lower cost), no cost TrueUps	Support costs unknown
24/7 Support	Included	Offered, but could be up to 4 hour response time
Few DBA Resource Needed	Large number of clusters can be managed with small number of DBAs	Large number of clusters can be managed with small number of DBAs

Tungsten Cluster

Tungsten Clustering is the only complete, fully-integrated, fully-tested MySQL HA, DR and geo-clustering solution running on-premises and in the cloud combined with industry-best and fastest, 24/7 support for business-critical MySQL, MariaDB, & Percona Server applications.

It allows enterprises running business-critical MySQL database applications to cost-effectively achieve continuous global operations with commercial-grade high availability (HA), geographically redundant disaster recovery (DR) and geographically distributed active/active.

Continuous Operations

Local clustering for enterprise-grade high availability (HA) and multi-site clustering for best-in-class disaster recovery (DR).

High Availability

Automated local failover, automated recovery and failback for MySQL and MariaDB clusters.

Disaster Recovery

Composite “clusters of clusters” spread across two or more sites or cloud data centers for a geographically distributed managed mesh of MySQL or MariaDB clusters.

Global Scaling

Geo-distributed and global clusters that are locally and globally aware. Serve a geographically-distributed user base with high availability and performance.

Interested to Learn More? Get in Touch!

Please use the button below to contact us. We will reply within 24 hours.

Related Resources

Read the Comparison Blog

Galera - High Noon with Tungsten Clustering for MySQL High Availability (HA), Disaster Recovery (DR) and Geographic Distribution

This is the second blog in our ‘High Noon’ comparison series in which we look at the main solutions for MySQL high availability, disaster recovery and geographic distribution. Here we focus on highly available, geo-scale, multi-region MySQL for mission-critical sites and apps with Codership’s Galera and its variants (MariaDB Galera and Percona XtraDB Cluster) as compared to MySQL clusters with Continuent Tungsten, the only complete, fully-integrated solution for highly-available MySQL clustering solution - for any version of MySQL, MariaDB, and Percona Server - on-premises, in the cloud, hybrid-cloud or multi-cloud.

Matthew Lang

Feb 16, 2021

High Noon Competitor Comparisons, Disaster Recovery, MySQL High Availability, Tungsten Clustering

Watch the Comparison Presentation

MySQL HA/DR/Geo-Scale - High Noon #2: Galera Cluster

This second installment of our High Noon series of on-demand webinars is focused on Galera Cluster (including MariaDB Cluster & Percona XtraDB Cluster). It looks at some of the key characteristics of Galera Cluster and how it fares as a MySQL HA / DR / Geo-Scale solution, especially when compared to Continuent Tungsten Clustering.

Matthew Lang

Upgrade From the Following Solutions

Do It Yourself Solution
(such as with ProxySQL and Orchestrator)

Industry-Best Continuent MySQL Support

24/7 MySQL support

Service is Always Available. Experts are on-duty 24/7/365 no matter which timezone you are in.

Highly Experienced

Our support engineers have 15 or more years of MySQL DBA and Site Reliability experience.

Quick Response

24/7 support comes with 1-hour SLA, with response times for urgent requests averaging less than 3 minutes.

Highly Available

MySQL uptime measured in months or years.

Why Continuent?

Our customers are leading SaaS, e-commerce, financial services, gaming and telco companies who rely on MySQL and Continuent to cost-effectively safeguard billions of dollars annual revenue.

Learn more

Tungsten Cluster vs. Galera Cluster

Galera Cluster Comparison

Use Cases Supported

Features & Benefits

Support Services

Intangibles

Tungsten Cluster

Continuous Operations

High Availability

Disaster Recovery

Global Scaling

Interested to Learn More? Get in Touch!

Related Resources

Read the Comparison Blog

Galera - High Noon with Tungsten Clustering for MySQL High Availability (HA), Disaster Recovery (DR) and Geographic Distribution

Watch the Comparison Presentation

MySQL HA/DR/Geo-Scale - High Noon #2: Galera Cluster

Upgrade From the Following Solutions

Industry-Best Continuent MySQL Support

24/7 MySQL support

Highly Experienced

Quick Response

Highly Available

Why Continuent?

These Companies Trust Our Solutions