Skip to main content
Aiven for Dragonfly is an advanced, high-scale, Redis-compatible in-memory database service designed to overcome Redis OSS limitations under high-load conditions. Handle workloads exceeding 1 TB with more than 10x the throughput performance of Redis OSS.

Overview

Dragonfly is specifically built to address the scalability and resource utilization challenges of Redis OSS. While Redis is renowned for speed and adaptability, it encounters limitations with large-scale data management and high throughput requirements. Dragonfly solves these challenges through vertical scaling, optimized hardware resource usage, and support for larger memory footprints.

Why Choose Aiven for Dragonfly

10x Performance

Achieve over 10x the throughput of Redis OSS with reduced latency

Redis Compatible

Drop-in replacement for Redis with no code changes required

Vertical Scaling

Scale up to handle 1 TB+ workloads on single instances

Efficient Architecture

Shared-nothing threading model for optimal resource utilization

Key Features

Seamless drop-in replacement for Redis:
  • Compatible with Redis API and protocols
  • Works with existing Redis clients
  • No application code changes needed
  • Support for Redis data structures
  • Redis commands and pipelines
Unique architecture for performance:
  • Shared-nothing architecture
  • Multi-threaded processing
  • Efficient CPU utilization
  • Lock-free data structures
  • Automatic thread management
Built for enterprise workloads:
  • Handle 1 TB+ datasets
  • Millions of operations per second
  • Efficient memory management
  • Optimized for modern hardware
  • Reduced latency at scale
Production-ready reliability:
  • Active-passive replication
  • Automatic failover
  • Persistence (RDB and AOF)
  • Snapshot capabilities
  • Backup and restore
Improved memory management during backups:
  • Reduced memory overhead
  • Faster snapshot creation
  • Minimal performance impact
  • Background save operations

Getting Started

1

Create Dragonfly Service

Deploy a Dragonfly service:
avn service create my-dragonfly \
  --service-type dragonfly \
  --cloud aws-us-east-1 \
  --plan business-8
2

Connect with Redis Client

Use any Redis client - Dragonfly is fully compatible:
redis-cli -h dragonfly-service.aivencloud.com \
  -p 12345 \
  -a your-password \
  --tls
3

Start Using Redis Commands

# Set and get operations
SET user:1000 "John Doe"
GET user:1000

# Hash operations
HSET product:100 name "Widget" price 19.99 stock 50
HGETALL product:100

# List operations
LPUSH queue:jobs job1 job2 job3
RPOP queue:jobs

Connection Examples

import redis

# Connect to Dragonfly (same as Redis)
client = redis.Redis(
    host='dragonfly-service.aivencloud.com',
    port=12345,
    password='your-password',
    ssl=True,
    ssl_cert_reqs='required'
)

# All Redis operations work identically
client.set('counter', 0)
client.incr('counter')

# Hash operations
client.hset('user:1', mapping={
    'name': 'John',
    'email': '[email protected]'
})

# Pipeline for performance
pipe = client.pipeline()
for i in range(10000):
    pipe.set(f'key:{i}', f'value:{i}')
pipe.execute()

# Pub/Sub
pubsub = client.pubsub()
pubsub.subscribe('notifications')
for message in pubsub.listen():
    print(message)

Performance Advantages

10x Higher Throughput:
  • Multi-threaded architecture
  • Parallel request processing
  • Efficient memory access patterns
  • Optimized for modern CPUs
  • Millions of ops/sec capability
Benchmarks:
  • SET operations: 10M+ ops/sec
  • GET operations: 15M+ ops/sec
  • Mixed workloads: 8M+ ops/sec

Use Cases

Perfect for large-scale caching needs:
  • Multi-terabyte cache layers
  • High-traffic web applications
  • Content delivery networks
  • API response caching
  • Database query caching

Best Practices

  • Test with production-like workloads
  • Monitor performance metrics
  • Use same client libraries
  • No application changes needed
  • Verify command compatibility
  • Use pipelining for bulk operations
  • Implement connection pooling
  • Set appropriate key expiration
  • Monitor memory usage
  • Use efficient data structures
  • Use business or premium plans
  • Enable automatic backups
  • Test failover procedures
  • Monitor replication lag
  • Plan for disaster recovery

Monitoring

Key Metrics

Performance

  • Operations per second
  • Command latency
  • Hit rate
  • Network throughput

Resources

  • Memory usage
  • CPU utilization
  • Connection count
  • Network I/O

Integration with Grafana

avn service integration-create \
  --integration-type metrics \
  --source-service my-dragonfly \
  --dest-service my-grafana

When to Choose Dragonfly

Choose Dragonfly when:
  • Dataset exceeds 100 GB
  • Need > 1M ops/sec throughput
  • Vertical scaling preferred
  • Memory footprint > 256 GB
Ideal for:
  • High-traffic applications
  • Low-latency requirements
  • Real-time processing
  • Heavy concurrent access
Benefits:
  • Fewer instances needed
  • Better resource utilization
  • Reduced operational overhead
  • Vertical vs horizontal scaling

Valkey

Alternative Redis-compatible in-memory store

Apache Kafka

Use Dragonfly for Kafka offset management

PostgreSQL

Cache PostgreSQL queries in Dragonfly

Grafana

Visualize Dragonfly metrics

Resources

SLA Note: Check with Aiven for specific SLA details for Dragonfly services. Performance and availability may vary based on plan and configuration.
Dragonfly is optimized for vertical scaling. Choose instance types with sufficient memory for your workload rather than scaling horizontally.

Build docs developers (and LLMs) love