ScyllaDB — NoSQL Data Store Compatible with Cassandra

What it is

ScyllaDB is a high-performance NoSQL database built on the Seastar framework in C++. It is API-compatible with Apache Cassandra (CQL) and Amazon DynamoDB, meaning you can migrate existing applications without code changes. The shard-per-core architecture dedicates each CPU core to its own data shard, eliminating contention.

ScyllaDB targets teams running latency-sensitive workloads at scale: time-series data, IoT telemetry, user profiles, and real-time analytics. It handles millions of operations per second with predictable tail latencies.

How it saves time or tokens

ScyllaDB's Cassandra compatibility means no learning curve for teams already using Cassandra. Drop it in, point your existing drivers at ScyllaDB, and get better performance with fewer nodes. The shard-per-core design reduces the number of servers needed compared to Cassandra for the same throughput.

Automatic tuning (CPU, memory, I/O) means less time spent on JVM garbage collection tuning that Cassandra requires.

How to use

1. Run ScyllaDB with Docker:

docker run -d --name scylla -p 9042:9042 scylladb/scylla:latest \
  --smp 2 --memory 2G --overprovisioned 1

1. Connect with CQL shell:

docker exec -it scylla cqlsh

1. Create a keyspace and table:

CREATE KEYSPACE myapp WITH replication = {
  'class': 'SimpleStrategy',
  'replication_factor': 1
};

USE myapp;

CREATE TABLE users (
  user_id UUID PRIMARY KEY,
  name TEXT,
  email TEXT,
  created_at TIMESTAMP
);

INSERT INTO users (user_id, name, email, created_at)
VALUES (uuid(), 'Alice', 'alice@example.com', toTimestamp(now()));

Example

# Python driver (same as Cassandra driver)
from cassandra.cluster import Cluster

cluster = Cluster(['127.0.0.1'])
session = cluster.connect('myapp')

# Insert
session.execute(
    'INSERT INTO users (user_id, name, email, created_at) VALUES (%s, %s, %s, toTimestamp(now()))',
    (uuid.uuid4(), 'Bob', 'bob@example.com')
)

# Query
rows = session.execute('SELECT * FROM users LIMIT 10')
for row in rows:
    print(f'{row.name}: {row.email}')

Related on TokRepo

• AI Tools for Database -- Database tools and storage engines
• AI Tools for DevOps -- Infrastructure and database management tools

Common pitfalls

• ScyllaDB uses CQL (Cassandra Query Language), which is not SQL. JOINs, subqueries, and aggregations across partitions are not supported. Design your data model around query patterns.
• Single-node Docker deployments are for testing. Production requires at least 3 nodes for replication and fault tolerance.
• ScyllaDB wants dedicated hardware. Running other services on the same machine causes performance degradation because Seastar expects exclusive CPU and memory access.