mikemccand commented on issue #15079: URL: https://github.com/apache/lucene/issues/15079#issuecomment-3196408197
Oooh the vmstat charts show the horrific "IOWAIT". [Healthy run (08/05)](https://benchmarks.mikemccandless.com/2025.08.05.18.04.16/fastIndexMediumDocsWithVectors/) and then [unhealthy run (08/06)](https://benchmarks.mikemccandless.com/2025.08.06.18.04.40/fastIndexMediumDocsWithVectors/), and then next [healthy but still slow run (08/07) ](https://benchmarks.mikemccandless.com/2025.08.07.18.04.39/fastIndexMediumDocsWithVectors/), and then [after we went back to `NORMAL` from `RANDOM` io hint (08/08)](https://benchmarks.mikemccandless.com/2025.08.08.18.04.12/fastIndexMediumDocsWithVectors/) -- wow look at the difference in the "Memory Usage" chart. That's wild -- this sure sounds like [the MGLRU issue](https://github.com/apache/lucene/issues/14408) might be affecting the nightly benchy box? -- under `MADV_RANDOM` the `active` RAM stayed flat as if the costly random swapped in page by page didn't count in the kernel's view as "hot", sheesh. -- 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
