Skip to main content

Dragonfly Cloud Enterprise

Overview​

Dragonfly Cloud Enterprise is a fully managed service designed for organizations that require extremely high-performance in-memory data stores at scale. It is suited for use cases involving:

  • Large-Scale Data: Datasets from over 500GB to multiple terabytes.
  • Extreme Throughput: Millions of requests per second (RPS) with low latency.
  • High Availability: 99.99% uptime with robust failover and multi-region backups.
  • Automatic Scalability: Intelligent autoscaling capabilities adapt to dynamic traffic patterns.
  • Security and Compliance: Advanced security controls and deployments on your existing cloud servers (BYOC).

Dragonfly Cloud Enterprise offers everything in Dragonfly Cloud plus the following enhancements and guarantees. If you are interested in Dragonfly Cloud Enterprise, please contact us to start a proof-of-concept.

Scalability and Performance​

Dragonfly Swarm​

At the heart of Dragonfly Cloud Enterprise is Dragonfly Swarm, which enables horizontal scalability on top of Dragonfly's advanced multi-threaded architecture. It first maximizes the resources of a single instance, which can utilize all available CPU cores and scale vertically to support up to 600GB of memory and millions of requests per second. When further scale is required, the system can seamlessly add nodes to form a distributed multi-shard cluster capable of handling hundreds of terabytes of memory and hundreds of millions of requests per second (RPS).

Read Replicas​

For read-intensive applications, the service provides read replicas. These replicas distribute read traffic across multiple instances to reduce latency and increase the overall throughput even further.

Flexibility and Cost Efficiency​

Autoscaling​

Dragonfly Cloud Enterprise includes an autoscaling feature that dynamically adjusts resources in response to traffic fluctuations, without any manual intervention. This helps maintain performance during demand spikes and optimize efficiency during quieter periods. Scaling policies can be customized to align with specific workload patterns. Key features and configurations of autoscaling include:

  • Supported Data Stores: At the moment, autoscaling supports any Dragonfly Swarm (distributed multi-shard) data store. It doesn't support single-shard data stores.
  • Monitoring & Autoscaling: It constantly monitors your data store memory and CPU utilization and triggers scale-up or scale-down based on your preconfigured scaling policies.
  • Resource Utilization Scaling Policies: You can specify maximum and minimum memory utilization and CPU targets. For example, "max 60% CPU utilization" means once CPU utilization surpasses 60%, a scale-up will be performed to get CPU utilization back under 60%. Similarly, "min 30% memory utilization" means once memory goes below 30% utilization, a scale-down will be performed.
  • Schedule-Based Scaling Policies: You can specify different policies for weekdays and weekends or daytime and nighttime. This suits workloads with recurring load patterns.
  • Safety Limits: You must specify maximum and minimum capacity limits to avoid scaling too large or too small. 

Bring-Your-Own-Cloud (BYOC) [Beta]​

The BYOC option of Dragonfly Cloud Enterprise can deploy fully managed Dragonfly data stores directly within your cloud account (AWS, GCP, or Azure). This model enhances security and compliance by keeping all data within your controlled environment. It can also provide cost benefits by utilizing your existing cloud servers.

Guaranteed 30% Lower Costs​

Dragonfly Cloud Enterprise is designed to offer significant cost efficiency. It provides a pricing structure that typically results in much lower costs compared to other cloud in-memory data services. For Dragonfly Cloud Enterprise, we guarantee at least 30% cost savings compared to your current managed in-memory data store solution.

Reliability and Support​

Multi-Region Backups​

Data durability is provided through multi-region backups. Critical data is replicated and stored across geographically diverse locations to ensure resilience and speedy recovery in the event of a regional outage.

Enterprise Level Support​

Enterprise includes direct support from the engineers who built Dragonfly. This provides assistance for onboarding, configuration, and ongoing operations. Enterprise support is available 24/7.

Getting Started​

Dragonfly Cloud Enterprise is designed to meet the demands of modern enterprise workloads, providing extremely high scale, performance, and cost efficiency.

To evaluate the service for your use case, contact us to begin a proof-of-concept.