New feature
Beta
RabbitMQ connector
RabbitMQ connector is a Databricks data engineering capability, introduced June 2026.
A managed Lakeflow Connect streaming connector that continuously consumes messages from RabbitMQ classic queues into Unity Catalog streaming tables on serverless compute.
- Exactly-once is off the table - classic queues have no broker-assigned unique message ID, so it promises at-least-once and leaves deduplication to you.
- Unlike Kafka, a classic queue forgets a message once it is acknowledged, so anything another consumer already took before the pipeline started is gone for good.
- It does not even peek inside - the message body lands as raw binary, and decoding JSON, Avro or Protobuf is your job downstream.
Limitations: Classic queues only (no RabbitMQ Streams), username and password authentication only, append-only destination tables with no SCD type 1 or 2, one broker per pipeline, serverless only, continuous runs only, duplicates possible after redelivery and per-queue order is best-effort.
Open in REbricked →- Category
- Data engineering
- Also known as
- Managed RabbitMQ connector, Lakeflow Connect RabbitMQ
- Verified
- 2026-10-20