Observability Pipeline Query Languages: Present and the Future
Talkat KubeCon + CloudNativeCon Europe 2025 · on Thursday, 2025-04-03· at 16:45· Level 1 | Hall Entrance N10 | Room G·
summary
Proposes collapsing many vendor languages into a SQL-centered approach (semantic standard first) and demos a proof-of-concept: an open source Grafana plugin backed by ClickHouse that parses logs, extracts fields, performs enrichment (threat intel joins) and compiles to large SQL for execution. Claims a month-long prototype works end-to-end and can be transpiled to vendor dialects, aims to add extensibility for ops primitives like a rate operator and concise pipe SQL composition, solving awkward patterns such as top N plus others aggregation.
AI-written summary · about
chapters (12)
- 0:00Observability pipeline query languages
- 3:49From Unix pipes to modern languages
- 7:01Observability market growth and fragmentation
- 10:13Problems with proprietary languages
- 12:41SQL as unified solution
- 16:34Why SQL wasn't adopted
- 19:24Challenges of SQL for observability
- 21:09Improving SQL with extensions
- 22:56Demo of SQL extensions
- 26:49Path to standardization
- 28:59Future of AI and query languages
- 31:32Q&A
related talks
34mThe State of SQL-Based Observability
29mA Practical Blueprint for Evolving Observability Schemas
37mOTeL/Piped Processing Language: Simplifying Observability Queries With Open...
24mWhat Observability Can Learn From BI
19mDo We Still Need to “Observe”? the Future of AI & Observability
52mObservability: it's not just an ops thing
18mThe Future of Database Observability is Open: SQL Commenter Merges W...
24mBuilding Queryable O11y with the Model Context Protocol