Description
I'd like to add a Cassandra sink connector.
It's on the connector roadmap (#2753) as Sink, P2, with scylla as the suggested crate.
There's no wide-column sink today. The 13 existing sinks cover analytical and lakehouse.
The roadmap also lists ScyllaDB as its own row (P3), annotated as covered by the Cassandra connector and pointing at the same crate. scylla is pure Rust and Cassandra-compatible, so one implementation should close both rows.
The workloads the roadmap cites for Cassandra - high-write wide-column, time-series and IoT at scale are stream-fed by nature. Devices at that volume don't write to Cassandra directly, they publish to a broker and something drains the stream into the table, with buffering and replay in between. That's the gap this fills.
Affected area / component
Connectors
Proposed solution
No response
Alternatives considered
No response
Contribution
Good first issue
Description
I'd like to add a Cassandra sink connector.
It's on the connector roadmap (#2753) as Sink, P2, with
scyllaas the suggested crate.There's no wide-column sink today. The 13 existing sinks cover analytical and lakehouse.
The roadmap also lists ScyllaDB as its own row (P3), annotated as covered by the Cassandra connector and pointing at the same crate.
scyllais pure Rust and Cassandra-compatible, so one implementation should close both rows.The workloads the roadmap cites for Cassandra - high-write wide-column, time-series and IoT at scale are stream-fed by nature. Devices at that volume don't write to Cassandra directly, they publish to a broker and something drains the stream into the table, with buffering and replay in between. That's the gap this fills.
Affected area / component
Connectors
Proposed solution
No response
Alternatives considered
No response
Contribution
Good first issue