Skip to content

usagestats: redact info from URLs, send to BigQuery

Administrator requested to merge fa/redact-urls-for-bq into main

Created by: attfarhan

This PR re-adds URLs to event payloads sent to BigQuery, with redaction for URL paths and query parameters not explicitly specified for marketing usage. This is being added to ensure we can implement a marketing attribution solution and determine the effectiveness of our campaigns.

URLs were completely removed in the past to make sure we didn't leak any private info to our analytics systems from Sourcegraph Cloud. With redaction in place for everything except marketing UTMs / query params, we can send URLs without leaking any private data.

Test plan

Added unit tests for the new function. Tested locally that events are properly sent to a mirror of the BQ data warehouse.

Merge request reports

Loading