Advanced search for your legacy application
summary
Shows how to add Elasticsearch to a Java/Postgres legacy app to replace slow LIKE joins with real search, using the Elasticsearch bulk API to index batches (demo reaches ~10k/sec then scales to 1M documents), wiring index calls into the save transaction or a message queue for reliability, fixing relevance by supplying custom analyzer and mappings (autocomplete, boosted name field), and visualizing results with Kibana aggregations; code is live-coded, repository and working dashboard provided.
AI-written summary · about
chapters (21)
- 0:00Legacy app search intro
- 1:32What is Elasticsearch
- 4:29Legacy data model overview
- 7:03Current search limits
- 9:38Why Elasticsearch fits
- 11:28Real-time DB sync
- 15:14Java client implementation
- 21:45Setting up Elasticsearch and Kibana locally
- 24:01Indexing data into ES
- 28:54Switching to Elasticsearch
- 31:38Bulk API for performance
- 33:36Analysis & tokenization
- 37:11Custom analyzers & mappings
- 41:00Documents with custom settings
- 42:55Boosting field relevance
- 44:24Measuring performance
- 45:26Aggregating with Elasticsearch
- 47:54Nested aggregations
- 49:20Fuzzy search for typos
- 50:47Kibana dashboards
- 52:20Summary
related talks
54mTesting in Legacy: From Rags to Riches
55mModernizing Legacy Code - While Not Bringing Down Prod
49mWhat Does a Legacy System Teach Us?
18mPlatform-Oriented Legacy Modernization
21m50 Shades of Legacy
01h 35mRedesigning Legacy Systems - Keys to Success / Lessons Learned
43mHow to Innovate in a Legacy
01h 27mRedesigning Legacy Systems : Keys to Success