Collect Events

Accept events via HTTP POST. Buffer on disk. Flush to DuckLake during pipeline runs.

1. Define the Schema

-- models/raw/events.sql
-- @kind: events

event_name VARCHAR NOT NULL,
page_url VARCHAR,
user_id VARCHAR,
received_at TIMESTAMPTZ

2. Start the Collector

ondatrasql events 8080

Public endpoint on port 8080. Admin endpoint on 8081 (localhost-only).

3. Send Events

curl -X POST localhost:8080/collect/raw/events \
  -d '{"event_name":"pageview","page_url":"/home","user_id":"u42"}'

Batch endpoint:

curl -X POST localhost:8080/collect/raw/events/batch \
  -d '[{"event_name":"pageview"},{"event_name":"click"}]'

4. Flush to DuckLake

ondatrasql run

Endpoints

Public (port from CLI):

Admin (port + 1, localhost only):

Validation

NOT NULL columns must be present. Unknown fields are ignored during flush. received_at is auto-populated if not provided.