Press Releases

Redis Labs Removes the Performance Barrier of Running Redis on Clouds to Achieve Sub-Millisecond Latency

Customizable, regular expression-based (RegEx) sharding technology enables sub-millisecond latencies at any throughput while maintaining full compliance with existing client libraries and Open Source Redis Cluster APIs

SANTA CLARA, Calif., October 29, 2014

Redis Labs (, the leading commercial Redis provider, today announced the launch of a new sharding feature that will enable developers to easily scale both the size and throughput of their Redis datasets, while maintaining sub-millisecond latency. This new addition to the Redis Cloud database-as-a service offering supports all Redis commands and data types.

Redis is an open source in-memory NoSQL database that has taken the developer community by storm. The fastest data store available today, Redis is used for high performance use cases such as high speed transactions, real time analytics, counting and ranking, job management, and social capabilities. Redis was created by Salvatore Sanfilippo.

"Enabling more than 100,000 developers to consume 9,000 APIs, Mashape is a leading API marketplace for developers," said Marco Palladino, founder and chief technology officer of Mashape. "With Redis Cloud’s new sharding technology, we were able to reduce response time from 4 to 0.70 milliseconds, during our 80,000 ops/second daily peaks."

Redis Labs’ new sharding technology enables:

  • Over 10 times more requests per second straight out of the box. Traditionally, developers have been limited to 80-90,000 requests per second when using a standard single-threaded Redis on the strongest cloud instance. This limit is caused mainly by virtual machine overhead and packets per second restrictions. By distributing the load across multiple cores and multiple cloud instances, Redis Cloud’s new sharding technology enables developers to supercharge their stack and easily reach over one million requests per second, while maintaining sub-millisecond latency.
  • Custom sharding by regular expression (RegEx) rules. Users can easily distribute a dataset across shards without changing the existing application design and while supporting all Redis commands and data types.
  • Fully managed clustering and sharding. This new sharding technology lets developers manage their Redis dataset from a single endpoint and scale to near infinite size and throughput without the overhead of managing a cluster, nodes, shards configuration, shards migration, replication or data-persistence settings. Redis Cloud guarantees highly available Redis deployment within and across data-centers.
  • Full compliance with existing clients libraries and open source Redis Cluster APIs. The new technology supports all existing Redis clients without requiring changes to the clients’ code. It fully complies with new open source Redis Cluster APIs.

"We love Redis because of its speed. With the Redis Cloud cluster, we were able to drop our application's response time from 700 down to 50 milliseconds with very little engineering effort," said Tung Nguyen, senior director of engineering at Bleacher Report. "Redis Labs has been extremely crucial in helping us meet the high expectations of Bleacher Report users."

"Developers building high-speed applications on the cloud are looking for high throughput, low latency and an easily configurable Redis solution," said Yiftach Shoolman, chief technology officer and co-founder of Redis Labs. "Our new customizable, regular expression-based sharding technology meets these needs exactly."