Source integration

Cassandra

Connect an Apache Cassandra cluster as a one-time migration source. NoSQLSync reads the real CQL schema directly instead of sampling rows.

Start free →

Connection requirements

  • One or more contact-point hosts, plus the cluster's local datacenter name (required by the driver)
  • A username and password if authentication is enabled on the cluster
  • The keyspace and table to migrate — both are listed directly from system_schema

One-time migration only — no CDC

Cassandra's CDC feature writes raw commit-log segments to disk in a binary format an external system must parse itself; the official Node driver exposes zero API for consuming those events. Rather than build an unverified low-level parser, this connector supports bulk (one-time) migration only, paginated through Cassandra's native paging state.

Setup

1
Add the connection
Provide contact points, local datacenter, and credentials (if auth is enabled) under Connections → Add.
2
Pick a keyspace and table
NoSQLSync lists your real keyspaces and tables from system_schema, not a placeholder.
3
Introspect and migrate
Schema comes from system_schema.columns directly — composite partition/clustering keys are combined into a single primary key column automatically.

FAQ

Why no live sync (CDC)?

Cassandra's CDC writes raw commit-log segments that need a custom binary parser to consume — a materially different and riskier engineering effort than reading a documented change-stream API. It's left out rather than shipped half-verified.

How are composite primary keys handled?

Very common in Cassandra. The partition key and clustering columns are combined into one composite key column for the SQL primary key, with every original column kept alongside it.

See every Cassandra migration path →