shatejas commented on PR #13985: URL: https://github.com/apache/lucene/pull/13985#issuecomment-3177238460
> it was not very large, and somewhat offset by the mysterious search slowdowns We can switch to an approach where the madvise is not changed at all. The approach leverages prefetch, read advise still changes but the reads are not as aggressive as sequential. A prefetch call after each read (inside merge) will make sure that merges don't slow down. At the same time this minimizes the search perf impact considering prefetch is not aggressive. https://github.com/apache/lucene/pull/14076#issuecomment-2563503703 -- 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: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
