DBT core vs DBT fusion| Workshop

Surfalytics community workshop on dbt Fusion: the next-generation execution engine. What it is, how it differs from dbt core, and how to use it locally (CLI, VS Code) and in dbt Cloud with state-aware orchestration. 📚 What you'll learn Theory • What is dbt Fusion (Rust engine, SDF acquisition) and what changed vs dbt core • Benefits: faster parse/compile, SQL error detection without warehouse, CTE preview, column lineage • Limitations: preview, license (free internally, no product-on-top without dbt Labs) • dbt architecture: authoring layer vs engine • Three tools: dbt Fusion CLI (free), VS Code extension (free up to 15 seats), dbt Cloud • State-aware orchestration: model-level state, skip unchanged, save cost (dbt Cloud) • Just-in-time vs ahead-of-time compilation; introspective models Hands-on • Compare dbt parse: core vs Fusion (e.g. 5s vs 1s); run model with dbtf • VS Code extension: column lineage, hover, CTE preview, SQL error detection • Static analysis on/off and effect on errors and helpers • dbt Cloud: enable Fusion; create standard vs state-aware daily jobs • Compare job runs: state-aware skips all models, much shorter run time • Freshness config (e.g. run model every 4 hours) • Recap and Q&A (license, three tools) ⏱️ Timestamps 0:00 – Intro and what we'll cover 0:46 – What is dbt Fusion 2:03 – Benefits (parse, compile, SQL errors) 4:57 – Limitations and license 6:55 – Architecture (authoring vs engine) 9:02 – Three tools (CLI, extension, Cloud) 12:13 – State-aware orchestration 13:03 – JIT vs AOT; introspective models 18:04 – State-aware in detail 20:47 – Hands-on: parse comparison, dbtf run 24:27 – VS Code extension features 29:34 – Static analysis on/off 34:37 – dbt Cloud and static analysis 36:23 – Standard vs state-aware jobs 39:11 – Compare run times (state-aware skips all) 42:44 – Freshness config 44:28 – Final recap 47:45 – Q&A: license and tools 📌 Resources mentioned • dbt Fusion (preview) • dbt Fusion CLI, VS Code / Cursor extension • dbt Cloud (state-aware orchestration, efficient testing) • Surfalytics community ✅ Key takeaways • dbt Fusion = Rust engine; same SQL/YAML, much faster and SQL-aware. • Use CLI for speed, extension for dev experience, Cloud for state-aware and efficient testing. • State-aware orchestration skips unchanged models; use freshness and configs to tune. • License: free for internal use; product-on-top needs dbt Labs agreement. • Still in preview—try in dev, wait for GA for production. --- Surfalytics – Community workshop. dbt Fusion: Next-Generation Execution Engine. #Surfalytics #dbt #dbtFusion #dbtLabs #DataEngineering #AnalyticsEngineering #SQL #DataTransformation #Workshop