Microsoft Azure · Software · YoctoIT tech page

Fabric

The unified data platform: data engineering, warehouse, real-time and Power BI on a single lake (OneLake) and a single license.

FOCUS · UNIFIED DATAThe end of seven glued-together tools: ingestion, transformation and reporting on the same lake
Request the Free Assessment Learn more
YoctoIT material for clients and partners · Microsoft, Azure, Entra, Defender and the other products mentioned are trademarks of Microsoft Corporation.
01 · What it is

Microsoft Fabric, made clear.

Fabric gathers the data chain into a SaaS platform: Data Factory for ingestion, Spark notebooks for engineering, a SQL warehouse, real-time intelligence and Power BI — all on the same OneLake, in the open Delta format. One copy of the data, one capacity to govern, one experience.

OneLake
one lake for everything: data written once, used by every engine
Delta
open format: no data lock-in, Spark and SQL on the same file
Direct Lake
Power BI reads the lake without imports: fresh reports without nightly refreshes
Microsoft Fabric
OFFICIAL MICROSOFT BRANDING · FABRIC
CONSOLE REALE · MICROSOFT FABRIC HOME · FONTE: MICROSOFT LEARN
REAL CONSOLE · MICROSOFT FABRIC HOME · SOURCE: MICROSOFT LEARN
02 · How to use it well

The things that make the difference.

The chain in one

Reports and decisions · Power BIwhere data becomes a choice
Data Factory
Spark & Warehouse
Real-Time
ingestion · transformation · streaming
OneLake (Delta)the single copy of the data
Fabric capacity (F-SKU)a single meter to govern
From the ERP to the dashboard, without removals

Medallion architecture

Bronze, silver, gold: the lake structure keeping order between raw and certified data.

SAP and IBM i data inside

Connectors and mirrors from the ERPs: the company's data estate finally together.

Governed capacity

F-SKUs saturate: monitoring, smoothing and priorities so month-end reports don't choke.

Governance with Purview

Lineage, classification and permissions: self-service BI without anarchy.

03 · In depth

Capacity, OneLake and Direct Lake

Fabric is bought as capacity (F-SKU): a pool of Capacity Units shared by all workloads with smoothing and throttling — capacity governance IS the operations; OneLake is the single lake in Delta format with shortcuts to ADLS/S3 (zero copy); Direct Lake lets Power BI read the Delta natively without imports or DirectQuery; deployment pipelines promote across dev/test/prod; mirroring replicates entire databases (SQL, Cosmos, Snowflake) into the lake.

  • F-SKU & CU — one capacity for everything: the single metric to monitor
  • Smoothing — peaks spread over time: bursting without purchases
  • Shortcut — external data mounted without copying: S3/ADLS in OneLake
  • Direct Lake — Power BI on native Delta: refreshes nearly eliminated
  • Mirroring — the operational database replicated into the lake in near-real-time
  • Deployment pipeline — dev→test→prod per workspace: the governed release
04 · Numbers and lifecycle

The numbers that matter.

1
copy of the data for all engines: Delta on OneLake
F2-F2048
the sizes: you start small, you scale
60 s
the typical mirroring ingestion window
mensile
the new features: the platform runs fast
Fabric is operated on capacity: CU monitoring, medallion on OneLake and release pipelines — tidy analytics.
05 · Use cases

Where it really pays off.

Modern corporate BI

From scattered Excels to a certified, self-service model.

A new data warehouse

The DWH born already a lakehouse, without iron.

Operational real-time

Production and logistics observed in streaming.

United data is worth more than big data: Fabric is the platform, we put it on the ground.