Did you know that Continuent Clustering supports having clusters at multiple sites world-wide with either active-active or active-passive replication meshing them together?
Not only that, but we support a flexible hybrid model that allows for a blended architecture using any combination of node types. So mix-and-match your highly available database layer on bare metal, Amazon Web Services (AWS), Azure, Google Cloud, VMware, etc.
In this article we will discuss using the Active/Passive model to scale reads worldwide.
The model is simple: select one site as the Primary where all writes will happen. The rest of the sites will pull events as quickly as possible over the WAN and make the data available to all local clients. This means your application gets the best of both worlds:
- Simple deployment with no application changes needed. All writes are sent the the master node in the Primary site cluster. Multimaster topologies can be more difficult to deploy due to conflicting writes.
- Application clients are able to read data locally, so response time is much better
- Ideal for Read-heavy/Write-light applications
The possibilities are endless, as is the business value. This distributed topology allows you to have all the benefits of high availability with centralized writes and local reads for all regions. Latency is limited only by the WAN link and the speed of the target node.
This aligns perfectly with the distributed Software-as-a-Service (SaaS) model where customers and data span the globe. Applications have access to ALL the data in ALL regions while having the ability to scale reads across all available slave nodes, giving you the confidence that operations will continue in the face of disruption.
Continuent Clustering incorporates the asynchronous Tungsten Replicator to distribute events from the write master to all read slaves. The loosely-coupled nature of this method allows for resilience in the face of uncertain global network communications and speeds. The Replicator intelligently picks up where it left off in the event of a network outage. Not only that, performance is enhanced by the asynchronous nature of the replication because the master does not need to wait for any slave to acknowledge the write.
Overall, Continuent Clustering is the most flexible, performant global database layer available today – use it underlying your SaaS offering as a strong base upon which to grow your worldwide business!
Want to learn more or run a POC? Contact us.