The ScyllaDB Story

We’d like to share with you how our company and our technology came to be. Our founders, Dor and Avi, met 15 years ago at a company called Qumranet, a start-up that pivoted not once but twice. Qumranet’s penultimate idea was around the Xen hypervisor. The complexity of Xen drove Avi to invent a new hypervisor model, one that uses hardware acceleration and aligns with Linux’s everything-is-a-process design. Avi gave it the name KVM, or Kernel Virtual Machine. Initially the name wasn’t a great choice as everybody attributed KVM to Keyboard-Video-Mouse, but as you know things changed. Since KVM was a latecomer to the market, a battle began around type-1 (Xen OSS) vs type-2 hypervisor (KVM).

However, over time the KVM technology became industry standard. Google adopted it for their cloud, Red Hat acquired Qumranet, and AWS, which had been an early customer of Red Hat’s Xen, eventually moved to KVM.

After finishing their work at Red Hat, Dor and Avi decided to start a new company in 2012. Having identified that everyone runs a single application in their virtualized OS, they came up with a novel idea — to create a new operating system from scratch, one that runs a single application with better performance and manageability. This was Cloudius Systems and its product OSv. OSv’s main targets were JVM-based applications that are weakly bound to the OS due to the managed runtime. This is important to ScyllaDB’s history because it was from OSv that the team learned about Apache Cassandra. Through much effort, they found it almost impossible to accelerate Cassandra with a new kernel since the bottleneck lies within Cassandra. (By comparison, Redis runs 70% faster on OSv than on Linux).

After much research into the market and technology, in mid-2014 the team decided to pivot away from OSv (still a live project). Instead they decided to embrace the Cassandra architecture but rewrite it from scratch in native code with all of the know-how of years of kernel/hypervisor coding. This would become the ScyllaDB database.

Fast forward six years and you’ll find that ScyllaDB not only has a large performance advantage over Cassandra, it also provides better functionality and better stability.

You can think of ScyllaDB as Cassandra the way it should have been. ScyllaDB is an open-source-first project. We have a small but growing community but it is very active — by github history, ScyllaDB has a larger number of active committers than Cassandra. And thanks to our Project Alternator, ScyllaDB now also has an Amazon DynamoDB-compatible API. We invite you to join our journey. You’ll be in good company, since Samsung, IBM, GE, Comcast, Grab, Starbucks as well as smaller businesses like Reddit, Medium, Opera and many others run multiple ScyllaDB clusters.

Start scaling with the world's best high performance NoSQL database.

Get Started
Github

Product

Resources

Company

Social

Legal

2024 ©ScyllaDB | ScyllaDB, and ScyllaDB Cloud, are registered trademarks of ScyllaDB, Inc.

Apache® and Apache Cassandra® are either registered trademarks or trademarks of the Apache Software Foundation in the United States and/or other countries. Amazon DynamoDB® and Dynamo Accelerator® are trademarks of Amazon.com, Inc. No endorsements by The Apache Software Foundation or Amazon.com, Inc. are implied by the use of these marks.