The GitHub Actions job "Build" on jackrabbit-oak.git/oak-search-mongot-query-latency has failed. Run started by GitHub user meticulous-dft (triggered by reschke).
Head commit for run: 9186d83f5d100befbbe43f250058b2ab7da7a022 / Ming He <[email protected]> OAK-search-mongot: cache the index statistics Oak reads while planning Oak's planner reads numDocs() and getDocCountFor(property) for every query. Each call ran countDocuments against the index collection, so every query paid two or three extra round trips to MongoDB before it started, and the property counts scanned the collection because nothing indexes those fields. Serve the counts from a cache, as the Elasticsearch index does: entries are refreshed in the background after 60 seconds and expire after 10 minutes (oak.mongot.statsRefreshSeconds, oak.mongot.statsExpireSeconds). The cache belongs to the index node, which the tracker replaces whenever the index status or definition changes, so each indexing cycle that writes to the index also starts from fresh counts. Report URL: https://github.com/apache/jackrabbit-oak/actions/runs/36466348919 With regards, GitHub Actions via GitBox
