Sinks
Sinks define destinations for your data in Cloudflare Pipelines. They support writing to R2 Data Catalog as Apache Iceberg tables or to R2 as raw JSON or Parquet files.
Sinks provide exactly-once delivery guarantees, ensuring events are never duplicated or dropped. They can be configured to write files frequently for low-latency ingestion or to write larger, less frequent files for better query performance.
Manage sinks Create, configure, and delete sinks using Wrangler or the API.
Available sinks Learn about supported sink destinations and their configuration options.