Multi-region deployments

Multi-region deployments in RayDB enable your PostgreSQL clusters to be distributed across different geographical locations. This helps reduce latency, improve redundancy, and enhance disaster recovery.

Benefits of Multi-Region Deployments

Reduced Latency: Deploy clusters closer to users for faster query response times.
High Availability: Minimize downtime by distributing replicas across regions.
Disaster Recovery: Protect against regional failures by maintaining geographically redundant instances.

How Multi-Region Deployments Work

Primary Cluster in One Region: Your main PostgreSQL cluster operates in a chosen region.
Read Replicas in Other Regions: Additional replicas can be deployed in different regions to distribute read traffic.
Failover Between Regions: If the primary region fails, a read replica can be promoted to minimize downtime.

Setting Up a Multi-Region Deployment

Log in to the RayDB Dashboard.
Navigate to the Clusters section and select your primary cluster.
Open the Replication tab.
Click Create Read Replica and select a different region.
Choose the instance size and confirm the deployment.

Managing Multi-Region Failover

Automatic Failover: If high availability is enabled, a replica in another region can be promoted automatically.
Manual Failover: Administrators can manually promote a read replica if needed.

Considerations

Replication Lag: Data replication across regions may introduce minimal latency.
Cross-Region Costs: Be aware of potential networking and data transfer fees.
Application Routing: Configure your application to direct read queries to the nearest replica.

Best Practices

Distribute Read Traffic: Use a load balancer to optimize regional read requests.
Monitor Regional Performance: Regularly check latency metrics and adjust deployments as needed.
Plan for Failover: Ensure disaster recovery policies include cross-region failover scenarios.

For more details on configuring read replicas, visit Read Replicas.