RonDB

The fastest open-source key-value store, now optimized for the cloud, offering unmatched performance and scalability for large-scale distributed storage systems.

Real-Time Machine Learning with Low Latency and High Availability

RonDB shows higher availability and the ability to handle larger data sets in comparison with Redis, paving the way to be the fastest key-value store available.

Comparing RonDB on AWS, Azure and GCP using Sysbench

RonDB 21.04.0 has integrated benchmark scripts to execute various benchmarks towards RonDB. We bring now the results of RonDB benchmark on AWS, Azure, and GCP.

A Real-Time Database for Real-Time AI Systems

Learn more about how RonDB outperforms AWS Sagemaker and GCP Vertex in latency for real-time AI databases, based on a peer-reviewed SIGMOD 2024 benchmark.
Dashboard mockup

The first LATS database in the cloud

RonDB combines cutting-edge technology with an easy-to-use interface, enabling clusters to launch in under a minute and scale to handle hundreds of millions of operations per second — an achievement unmatched by any other managed database today.

Beyond high-performance online feature storage, RonDB excels with LATS — Low Latency, high Availability, high Throughput, and scalable Storage — making it the ideal solution for an online feature store with superior key capabilities.

Throughput Sysbench results

Sysbench testing with RonDB demonstrates impressive performance:
Peak Throughput: RonDB clusters with r5n.8xlarge data nodes reached 34,870 transactions per second, demonstrating exceptional performance.
Latency Performance: Maintained sub-20ms latency at 64 threads, highlighting RonDB's efficiency in high-load scenarios.
Filtering Efficiency: The filtered Sysbench benchmark showed RonDB’s ability to handle high throughput while reducing the amount of row data transferred, emphasizing its scalability and processing efficiency.
Dashboard mockup
Dashboard mockup

High Availability & Robustness

RonDB ensures high availability and robustness through a multi-layered architecture. It features redundant node groups with datanodes for data durability, management nodes for system operations, and API nodes for database access.

The metadata layer is managed by RonDB’s distributed, in-memory database, which avoids single points of failure with stateless applications and multiple replicas.

This design guarantees resilience and reliability across various operational contexts.

Automatic Thread Configuration

RonDB's ThreadConfig allows advanced users to fine-tune thread assignment, CPU locking, and scheduling, while the managed version provides optimized configurations based on best practices.

The system uses adaptive CPU spinning and automatic thread configuration to enhance performance and reduce latency, with sophisticated scheduling and load balancing to ensure efficient CPU usage.
Dashboard mockup

Try now free

Hopsworks - Real-time AI Lakehouse

Enhanced MLOps with Hopsworks Feature Store

Contact us and learn more about how Hopsworks can help your organization develop and deploy reliable AI Systems.