Deployment models in Timestream for InfluxDB 3

Single-node deployment

In a single-node deployment (available for both Core and Enterprise versions), a single instance handles all database operations. This configuration features:

• One node performing multiple critical roles:

  • Writer role: Handles all write operations to the database.

  • Reader role: Processes read queries and data retrieval operations.

  • Compactor role: Optimizes storage by compacting data files (Enterprise version only).

• Core version: Single node operates with writer and reader roles only.

• Enterprise version: Single node operates with writer, reader, and compactor roles for complete functionality.

• Simplified management with all operations consolidated on a single instance.

• Lower cost compared to multi-node deployments.

• Suitable for development, testing, or smaller production workloads.

Single-node deployments are ideal for getting started with Timestream for InfluxDB 3 or for workloads with moderate performance requirements.

Multi-node deployment

Multi-node deployments (available only for Enterprise version) distribute workloads across multiple instances for improved performance, scalability, and availability. This configuration offers:

• Flexible node allocation based on workload requirements:

  • 2 nodes that can handle both write and read operations.

  • 1 dedicated compactor node for optimizing storage.

• Horizontal scaling capabilities to accommodate growing workloads.

• Enhanced fault tolerance with workload distribution across multiple nodes in different Availability Zones.

• Improved query performance.

• Better handling of concurrent operations.

Multi-node deployments are recommended for production environments with high throughput requirements, large data volumes, or the need for enhanced availability.