Data Factory

A cited reference architecture for unifying data that lives across GCP, AWS, Snowflake, Databricks, and Azure into Microsoft Fabric — choosing per source between OneLake shortcuts (zero-copy), Mirroring (near-real-time CDC), and Copy Job / Copy activity (batch), all orchestrated by Fabric Data Factory pipelines and landed in OneLake for downstream serving. Grounded only in verified tier-1 claims; inferences are labelled and no limits or pricing are invented.

A cited reference pattern for replacing polling and fixed-schedule ELT with event-driven orchestration on Microsoft Fabric: event sources feed an Eventstream, Activator rules watch conditions and fire Copy Job or pipeline/Spark actions on demand, and monitoring closes the loop. Covers the fixed-schedule vs event-driven cost/freshness tradeoff, check-then-act maintenance automation, and Outbound Access Protection (OAP) governance for cross-workspace event flows. Grounded only in verified claims; GA vs preview marked; inferences labelled.

A practitioner's guide to Microsoft Fabric Data Factory — its three data-movement models, event-driven and multi-cloud orchestration, migration tooling from Azure Data Factory and Synapse, Copy activity performance tuning, and the internal plumbing behind triggers and throughput, with the operational pitfalls that catch teams by surprise.