jpountz commented on PR #13381: URL: https://github.com/apache/lucene/pull/13381#issuecomment-2119368241
> such an application can learn whether the pages it is prefetching into the cache are still there by the time it gets around to using them This is an interesting idea! I was discussing this potential problem with @tveasey the other day. With terms and postings, we're currently only looking into loading a few pages in parallel per search thread and we then use them immediately. With GBs of capacity for the page cache, it would be extremely unlikely for these pages to get evicted in the meantime. But if/when we start looking into using `prefetch()` for bigger regions (e.g. stored fields) and/or possibly longer before needing the data (e.g. starting prefetching data for the next segment while we're scoring the current segment), then this could become a problem indeed. It would be nice if we could learn to disable prefetching when it's not working as intended. This would make this API safer to use. -- 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
