Skip to content

Implement Saneitizer

Ebernhardson requested to merge work/ebernhardson/saneitizer-source into main

Implements the same concept as the CirrusSearch saneitizer, but in flink land instead of using the job queue. This is implemented as a custom event source that emits SanityCheck events, and an async http fetch that processes the SanityCheck events using the cirrus sanity check api. Errors found by the sanity check api are converted into UpdateEvent instances and progress through the pipeline just like the events that come from mediawiki via kafka.

Bug: T358599

Edited by Ebernhardson

Merge request reports