Dragonfly is a drop-in Redis replacement that is 25x more performant and designed to effortlessly scale up heavy workloads. The Dragonfly Operator for Kubernetes provides easy deployment with automated replication and fail over for high availability.
→ ~/ helm upgrade --install dragonfly oci://ghcr.io/dragonflydb/dragonfly/helm/dragonfly --version v1.13.0
Dragonfly is a drop-in Redis replacement, meaning it uses the same APIs and is compatible with all of the same SDKs and tooling. Teams that switch from Redis to Dragonfly get huge performance gains and a simpler system to operate, all without changing code.
The power of modern cloud hardware can never be realized with legacy software. Dragonfly is optimized for modern cloud computing, delivering 25x more throughput and 12x lower snapshotting latency when compared to legacy in-memory data stores like Redis, making it easy to deliver the real-time experience your customers expect.View the benchmarks
QPS benchmark on AWS c6gn.16xlarge. Snapshot benchmark on AWS r6gd.4xlarge.Source
More QPS than Redis
Faster snapshotting than Redis
Dragonfly is architected to deliver the scalability that modern data workloads require. Say goodbye to complex multi-node environments with endless configuration options and infinite points of failure. Dragonfly can effortlessly scale up to handle over 1TB of data and 4 million QPS on a single node, with a simple primary-replica pattern for high availability.
In-memory datasets on a single instance
Less memory usage
Dragonfly is designed to fully utilize the processing power of modern hardware, reducing infrastructure costs by up to 90% when compared to legacy data stores like Redis. Dragonfly’s simple, vertical scaling model also means significantly lower total cost of ownership.
Memory usage under BGSAVE. Filling with 5GB of data using debug populate 5000000 key 1024, sending the update traffic with memtier, and snapshotting with bgsave. Source
Reliable By Design
Dragonfly makes extremely efficient use of memory (i.e., no spike in usage during snapshotting), dramatically reducing the risk of out-of-memory outages.
Dragonfly natively supports an eventually consistent primary-replica model to achieve high availability. If an outage on the primary occurs, Dragonfly will automatically failover to the replica and promote it as the primary.
Dragonfly is a simple system with a single instance for workloads up to 1TB. This eliminates the fragility inherent to distributed clusters and dramatically improves the reliability of the system.