https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=24438 --- Comment #6 from Kyle M Hall (khall) <kyle@bywatersolutions.com> --- Created attachment 201096 --> https://bugs.koha-community.org/bugzilla3/attachment.cgi?id=201096&action=edit Bug 24438: Reindex script ( rebuild_elasticsearch.pl ) reports success even when commits fail When a commit fails ( for example a timeout ), rebuild_elasticsearch.pl logs the error, drops the batch, and still exits 0, so a reindex that lost records looks successful. This patch indexes each batch through bulk_index_records and exits non-zero if any records could not be indexed, including failures in forked children. When Elasticsearch is briefly overwhelmed it retries the failed records, waiting longer each attempt ( --retries, --retry-delay ), before giving up. Test Plan: 1) Apply the patches 2) prove t/db_dependent/Koha/SearchEngine/Elasticsearch/Indexer.t 3) Run a reindex, note the output is unchanged and it exits 0: misc/search_tools/rebuild_elasticsearch.pl -v ; echo $? -- You are receiving this mail because: You are watching all bug changes.