The Debate: Which Search Engine?
summary
Three engines compared by concrete claims and limits: Vespa advertises native tensor support, ANN with HNSW and hybrid term+vector ranking, true real‑time partial updates and a cloud GA that backed Yahoo’s 25 billion queries/day, but it isn’t cost‑efficient for massive cold log stores. Elasticsearch touts the biggest ecosystem, HTTP/JSON developer UX, frequent releases, snapshots and data tiers, plus upcoming native model support and Lucene 9 ANN. Solr emphasizes maturity, plugin ecosystem, HA/DR work and the new Solr Operator for Kubernetes; avoid Solr for deeply nested, very large documents.
AI-written summary · about
chapters (11)
- 0:00Introduction and panel
- 3:50Elasticsearch pitch
- 10:52Apache Solr pitch
- 19:15Vespa pitch
- 30:03Query DSL debate
- 39:14When not to use specific engines
- 47:06Scalability and high availability
- 57:50Specific technical questions
- 1:08:21Query language comparison
- 1:15:59New players and innovation
- 1:26:14Conclusion and closing remarks
related talks
12mKeynote: OpenSearch, Innovation and Community
36mWhat’s New in the OpenSearch Project and Ecosystem
34mNrtSearch: Yelp’s fast, scalable, and cost-effective open source search engine
05mKeynote: OpenSearch: From Code to Community
01h 03mThe OpenSearch Community, Building the Future Together
06mSponsored Keynote: Foundation-Led Innovation: OpenSearch's Impact on Modern Data I...
40mOpen source software could save libraries
10mFrom custom Solr to out-of-the-box OpenSearch