Indexing in Ozon's search platform
summary
Covers the design and implementation of Ozon’s document indexing subsystem for its search engine, explaining architecture choices that provided scalability and fault tolerance and practical Java-level optimizations using profilers. Describes building indexes with Apache Lucene and best practices for using Lucene efficiently in high-load, scalable Java applications.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 7:30Custom search engine
- 13:34Indexing architecture in Ozon
- 20:03Vertical scaling problems
- 30:47Memory optimization and GC
- 36:57Vertical scaling results
- 37:41Horizontal scaling
- 41:23Fault tolerance and redundancy
- 43:53Summary
- 45:17Q&A
related talks
41mArchitecture Around Search
30mCaching for Speed: Practical Performance Optimization in OpenSearch & Lucene
04mOpenSearch Performance and Monitoring
41mScaling Search and Powering Vector-Based Retrieval With OpenSea...
41mE-Commerce search at scale on Apache Lucene
51mHibernate Search talk
39mAccelerating Aggregations with Lucene-Aware Execution in OpenSearch
47mHow We Increased Image Search Throughput