AnishMahto commented on PR #58209: URL: https://github.com/apache/spark/pull/58209#issuecomment-5372753571
@szehon-ho for review. This is actually a fairly well contained change, the majority of the diff comes from tests + code comments. The non-test, non-comment diff is actually under 150 LOC total. We're just swapping out the mechanism for determining affected rows for reconciliation. Rather than doing the "find first row that precedes the earliest event in the microbatch per key" search per table independently, we union the auxiliary and target tables then do the search across both together. This reduces us from potentially finding two anchor rows, to [correctly] finding at most one. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
