Introduction to ScyllaDB

Welcome to ScyllaDB University. If you’re building database-intensive applications, you’ve likely run into the database bottleneck: hard-to-maintain clusters, unpredictable latency spikes, and high costs. Today, we’re introducing you to ScyllaDB, a database that’s built from the ground up to solve these exact problems.

ScyllaDB is a hardware-optimized database designed for predictable low latency and massive throughput. ScyllaDB helps you hit your SLAs by using autotuning, data rebalancing, elastic scaling, and workload prioritization. These are all topics that we will cover in later lessons. The database runs one shard, or thread, per CPU core with isolated memory and asynchronous I/O. This shared-nothing architecture eliminates locking, providing linear scalability as you add cores or nodes.

Instead of letting the operating system guess how to manage resources, ScyllaDB pins threads directly to individual CPU cores, bypassing the Linux page cache. It manages its own disk I/O and CPU scheduling. The result? It extracts every single ounce of performance from modern, multi-core hardware. This translates to single-digit millisecond P99 latencies with a high throughput of millions of operations per second, as you will see in the next part.

Because it’s so efficient, organizations use ScyllaDB to avoid having to babysit their database. They often significantly reduce their node count while avoiding cloud vendor lock-in. ScyllaDB is the engine behind real-time applications in gaming, ad tech, streaming, cybersecurity, and IoT — anywhere that single-digit millisecond latency is strictly required at a massive scale.

You don’t have to take my word for it. In the next lesson, we aren’t just going to tell you it’s fast — we’re going to let you prove it yourself. We’ll jump into our interactive lab environment and easily spin up a benchmark to push 1 million operations per second. See you in the next lesson!