Query Insights Deep Dive: From Architecture to Production Optimization
summary
OpenSearch Query Insights tracks the top N queries by latency, CPU, and memory, collects thread-level resource metrics via the resource tracking framework, and buffers requests for an async 5s drain into per-metric priority queues that reset on a configurable time window (default 5 minutes). Data can be exported to a local index or S3 remote repository (OpenSearch 3.6) for historical analysis. Features include query-shape grouping to collapse repeated templates, live query cancellation, workload-group limits integration, and a planned rule-based recommendation engine and finer latency breakdown.
AI-written summary · about
chapters (17)
- 0:00Query Insights intro
- 1:05Why Query Insights needed
- 3:12How Query Insights works
- 5:00Dashboards config
- 7:52Top queries analysis
- 9:16Query latency breakdown
- 10:19Query grouping patterns
- 11:24Workload management integration
- 12:28Live query monitoring
- 13:11Query recommendations
- 15:20Dev Tools profiling
- 16:24Backend architecture
- 19:37Export top query data
- 22:06Query grouping details
- 23:32Future features
- 25:19Q&A
- 37:19Summary
related talks
17mMastering the Query Insights Dashboard in OpenSearch
20mUnified Performance Observability in OpenSearch: Integrating Workload Management and Query Insights
04mOpenSearch Performance and Monitoring
18mEnhance Query Execution Visibility in OS with OpenTelemetry
06mWhy OpenSearch?
22mFrom Telemetry to Insight: Building AI-Powered Observability Pipelines with OpenSearch
29mThe Search for Speed
20mSeeing Is Believing: OpenTelemetry and OpenSearch in Next-Gen Observability