Re: Proposing a 10.3.1 bug fix release
Ignacio Vera <[email protected]>
| Newsgroups | gmane.comp.jakarta.lucene.devel |
|---|---|
| Message-ID | <CAKryLwuwqJmYDckN8neZb-_0Tg+418Qpq2RtjEROcqAd+5AZKQ@mail.gmail.com> |
The thing is even stranger as we had the same bug in the 10.2 release and it was fixed in the 10.2.1 release. The PR that fixes is here: https://github.com/apache/lucene/pull/14511 What it seems is when backporting to the lucene 10_x branch, the branch did not contain the lucene 10.3 posting format: https://github.com/apache/lucene/pull/14557 <https://github.com/apache/lucene/pull/14557> When the format was backported it still contained the bug: https://github.com/apache/lucene/pull/14563 On Wed, Oct 1, 2025 at 1:25 PM Benjamin Trent <[email protected]> wrote: > Michael, > > Lucene Benchmarks actually caught this. Then other optimizations made the > regression go away (with the original cause not being directly fixed). Then > later it was directly fixed, and never actually backported. > > Ignacio has indeed added a new test on main for this particular bug (see: > https://github.com/apache/lucene/pull/15266) > > Here is the benchmark in Lucene that caught the issue, it was partially > addressed, but regressed again on 'IR', and "fixed" accidentally by other > optimizations. > > https://benchmarks.mikemccandless.com/FilteredAndHighHigh.html > > [image: image.png] > > On Wed, Oct 1, 2025 at 6:43 AM Michael Sokolov <[email protected]> wrote: > >> It sounds like this is needed. Do we also have any ideas about how we >> could have avoided this? Should we have some added test coverage, maybe a >> luceneutil enhancement? Did we already have a test that would have caught >> but we did not notice? >> >> On Wed, Oct 1, 2025, 4:03 AM Ignacio Vera <[email protected]> wrote: >> >>> Hi, >>> >>> In Elasticsearch, we have detected some of our benchmarks to degrade >>> badly after upgrading to Lucene 10.3.0. We have nailed down the issue to a >>> bug <https://github.com/apache/lucene/pull/15263> in the posting >>> formats where the impacts returned when there are no frequencies is wrong, >>> leading to a miscalculation of scores. >>> >>> In order to fix this regression, I would like to propose a bug fix >>> release for 10.3 line. I will be pushing the bug fix today so we can start >>> the release process as soon as this is done as far as there are no >>> concerns. I will try to find a release manager or if I cannot I can do it >>> myself. >>> >>> Thanks, >>> >>> Ignacio >>> >>> >>> [image: image.png] >>> >>
image.png
(image/png, 165.1 KB) - not displayed
image.png
(image/png, 43.6 KB) - not displayed