Fully compatible with Apache Cassandra at 10x the throughput and jaw dropping low latency

Get Started

1,000,000 transactions per second per server.

Apache® Cassandra® compatible column store, with the low latency of Redis.

No pauses under compaction or garbage collection.

Fast, predictable scaling. Add or remove nodes under normal production load.

Comparing Scylla to Cassandra

Throughput of Scylla and Cassandra in a 3-node cluster with RF=3 and CL=QUORUMLearn more

Cassandra 2.1.9
Average throughput. Results were rounded with accuracy of 10K.

A Cassandra compatible NoSQL column store, at 1MM transactions/sec per server.

Scylla is the world’s fastest column-store database: the functionality of Apache Cassandra with the speed of a light key/value store.

Raw speed isn’t just a nice benchmark or a low cloud bill. Scylla’s order of magnitude improvement in performance opens a wide range of possibilities.  Instead of designing a complex data model to achieve adequate performance, use a straightforward data model, eliminate complexity, and finish your NoSQL project in less time with fewer bugs.

Performance improvements enable not just reduction in hardware resources, but also a reduction in architecture, debugging, and devops labor. Scylla combines the  simple and familiar Cassandra architecture with new power to build a data model that fits the application, not tweak the model to fit the server.

Drop in replacement solution

From the team behind KVM, OSv, and Seastar

The team behind Scylla is the one that designed and developed the KVM hypervisor—Avi Kivity, Dor Laor, Benny Schnaider and additional OSS veterans. KVM is now the default hypervisor in many cloud computing environments including Google Compute Engine and OpenStack. At ScyllaDB, we carefully apply low-level knowledge to our Big Data technology. Data structures are measured not to cross CPU cachelines, poll-mode drivers are used instead of interrupts, disk accesses are measured for write amplification and caches need to be scan-resistant. We’re proud of our very own task scheduler and TCP/IP kernel bypass. We hog the latest C++14 and gcc5.1 for latest and greatest and dream about commodity non volatile random access memory.

Learn More