Redis

The Redis data service in PDS offers flexible deployment configurations to suit a variety of use cases, ranging from standalone setups to highly available and scalable deployments. This section provides detailed guidance on the available deployment options and the use of Redis Sentinel for ensuring high availability.

Deployment configurations

The Redis data service supports two deployment configurations:

• Standalone deployment (1 node):

  • When you deploy the Redis data service with a single node, it functions as a standalone Redis instance.
  • This configuration is best suited for non-critical use cases, testing environments, or applications that do not require fault tolerance or high availability.

• High availability deployment (3 to 10 nodes):

  • For deployments with 3 to 10 nodes, Redis instances are automatically configured with Redis Sentinel for high availability.
  • Recommended for production workloads where high availability and automatic failover are essential.
  • Ideal for applications that require consistent uptime and fault tolerance.

High availability with Redis Sentinel

When deploying more than one Redis instance (3 to 10 nodes), Redis Sentinel is integrated into the architecture to provide the following benefits:

• Automatic failover: If the master node becomes unavailable, Redis Sentinel promotes one of the replicas to the master role automatically, ensuring minimal downtime.
• Monitoring: Redis Sentinel continuously monitors the health of the master and replica nodes, ensuring early detection of issues.
• Scalable configuration: Users can deploy Redis clusters with up to 10 nodes, leveraging Redis Sentinel for consistent monitoring and failover management. This setup is ideal for production-grade applications where reliability and availability are critical.

For more information, refer to the Redis documentation.

Benefits of Redis Sentinel-enhanced deployments

• Enhanced reliability: Redis Sentinel ensures that your data service remains operational even in the event of node failures, making it highly reliable for mission-critical applications.
• Greater flexibility: By supporting node counts from 3 to 10, users can tailor their deployments to their specific requirements, whether for smaller setups or large-scale applications.
• Scalability: With support for up to 10 nodes, the Redis data service can grow alongside your application's demands.

How to deploy a Redis or Redis Sentinel data service in PDS

Refer to the Deploy a data service page for more information:

1. In the PDS UI, make sure you select Redis in the Data Service Overview page -> Deploy Data Service dropdown.
2. When choosing the node count in the Deploy Redis page -> Configuration section -> No. of instances dropdown:
   • Select 1 for a standalone deployment.
   • Select a number between 3 and 10 for a clustered deployment with Redis Sentinel.

Use the PDS dashboard to monitor the connection and health of your Redis nodes. Redis Sentinel will handle failovers and notify you of any issues that require attention.

Limitations

You cannot switch between the existing Redis Sentinel and Redis Cluster deployment patterns and vice-versa. For example, scaling from a 3-node Redis Sentinel deployment to a 6-node Redis Cluster deployment is not supported. However, scaling:

• within Redis Cluster deployments is supported, allowing a 6-node Redis Cluster to be scaled to an 8-node or 10-node deployment
• from a single-node Redis deployment to a 3-node or larger Redis Sentinel deployment is supported