https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=31925
--- Comment #62 from Martin Renvoize (ashimema) <[email protected]> --- Created attachment 205889 --> https://bugs.koha-community.org/bugzilla3/attachment.cgi?id=205889&action=edit Bug 31925: Document that thesaurus-aware linking depends on index freshness QA (comment 58/60) suggested LinkerConsiderThesaurus only works with Elasticsearch, having observed duplicate authorities being created when testing with Zebra. Reproduced this directly: with AutoCreateAuthorities and LinkerConsiderThesaurus on, saving two bibs with the same new FAST heading back-to-back (no pause) creates a duplicate authority under Zebra, but waiting for Zebra's indexing daemon to catch up (a few seconds) between saves lets the second heading link to the first-created authority correctly, as designed. The same duplicate-on-rapid-succession behaviour also reproduces with LinkerConsiderThesaurus off and a plain LCSH heading - it's a pre-existing race between AutoCreateAuthorities and Zebra's asynchronous indexing daemon, not something specific to thesaurus-aware matching or this bug. Elasticsearch indexes near-synchronously within the same request, so the same race window is far less likely to be hit there, which is almost certainly why this only showed up in manual Zebra testing. This documents that dependency on the LinkerConsiderThesaurus system preference description, rather than gating the feature to Elasticsearch only, since the underlying logic is search-engine agnostic and works correctly on Zebra given normal indexing latency. Sponsored-by: Heythrop Library <https://heythroplibrary.co.uk/> Co-Authored-By: Claude Sonnet 5 <[email protected]> -- You are receiving this mail because: You are watching all bug changes. _______________________________________________ Koha-bugs mailing list -- [email protected] To unsubscribe send an email to [email protected] website : http://www.koha-community.org/ git : http://git.koha-community.org/ bugs : http://bugs.koha-community.org/
