ScyllaDB

Managed ScyllaDB: a Cassandra-compatible wide-column database with single-node or multi-node high-availability clusters and automated backups.

Creating a ScyllaDB instance

From the StackBlaze dashboard, click New service, then Database, then ScyllaDB. Pick a plan and storage size, and StackBlaze provisions the instance for you.

Connecting a ScyllaDB instance to a service injects the SCYLLA_CONTACT_POINTS environment variable into that service. Your application reads the contact points from this variable, so there is nothing to hardcode.

ScyllaDB is Cassandra-compatible: it is a C++ rewrite of Apache Cassandra. Because of this, existing Cassandra drivers and your existing CQL (Cassandra Query Language) work against it without changes.

Connecting

ScyllaDB uses the CQL binary protocol on port 9042, the same wire protocol as Cassandra. Any Cassandra-compatible driver can talk to it directly. Services in the same project reach the database over the cluster’s private network using its internal contact point, which takes the form below:

[service-name]:9042

Standard vs high availability

You choose between Standard and high availability when creating the instance. Standard runs a single node, which is a good fit for development and lower-stakes workloads. High availability runs a multi-node cluster that replicates data across nodes and racks, the same way a Cassandra cluster does, so the database keeps serving traffic if a node is lost.

	Standard	High availability
Topology	Single node	Multi-node cluster
Replication	None (single copy)	Across nodes and racks
Survives a node loss	No	Yes
Best for	Development and testing	Production and resilient workloads

Connecting from Python

The standard cassandra-driver package works with ScyllaDB. Read the contact points from SCYLLA_CONTACT_POINTS, connect, and run a CQL query:

import os
from cassandra.cluster import Cluster

contact_points = os.environ["SCYLLA_CONTACT_POINTS"].split(",")
cluster = Cluster(contact_points, port=9042)
session = cluster.connect()

rows = session.execute("SELECT release_version FROM system.local")
for row in rows:
    print(row.release_version)

cluster.shutdown()

Backups

StackBlaze takes automated, encrypted backups that follow the platform backup policy applied to all managed databases. For details on schedules, restores, and how the policy works, see Backups & recovery.