Skip to main content
A Webhook sink allows you to send data to an external service via HTTP. This provides considerable flexibility for forwarding pipeline results to your application server, a third-party API, or a bot. Webhook sinks ensure at least once delivery and manage back-pressure, meaning data delivery adapts based on the responsiveness of your endpoints. The pipeline sends a POST request with a JSON payload to a specified URL, and the receiver only needs to return a 200 status code to confirm successful delivery.
Turbo’s webhook sink supports the same delivery semantics and adds a primary_key field for sink-side checkpointing, one_row_per_request, and payload_version envelope formats.
Here is a snippet of YAML that specifies a Webhook sink:

Pipeline configuration

Key considerations

  • Failure Handling: In case of failures, the pipeline retries requests indefinitely with exponential backoff.
  • Networking & Performance: For optimal performance, deploy your webhook server in a region close to where the pipelines are deployed (we use aws us-west-2). Aim to keep p95 latency under 100 milliseconds for best results.
  • Latency vs Throughput: Use lower batch_size/batch_flush_interval to achive low latency and higher values to achieve high throughput (useful when backfilling/bootstraping).
  • Connection & Response times: The maximum allowed response time is 5 minutes and the maximum allowed time to establish a connection is 1 minute.

Secret creation

Create a httpauth secret with the following CLI command:
Select httpauth as the secret type and then follow the prompts to finish creating your httpauth secret.

Example Webhook sink configuration