Lesson 1: The Azure Analytics Map, Old and New Names

Why start with a map

Azure's analytics catalogue has been renamed several times. A service you knew as SQL Data Warehouse is now documented as dedicated SQL pool (formerly SQL DW). The same warehouse job now has a newer home in Microsoft Fabric. Old names live on in Stack Overflow answers, architecture diagrams and Terraform modules. If you can't translate them, you'll design on top of something Microsoft has stopped investing in, or has already switched off.

This lesson gives you a single page with seven groups. Each service gets its current name, its former names and a status. Later lessons cover how to choose between services. This one makes sure you know what each thing is called and whether it's still alive.

Think of it like a city map after a few years of road renaming. The streets are mostly the same, but you need the old-to-new lookup before you give anyone directions.

The one-page map (status as of October 2026)

Gotcha: "Synapse" appears in two places. Azure Synapse Analytics is a PaaS workspace you deploy into a resource group. Inside Fabric, the warehouse and Spark experiences used to carry a "Synapse" prefix and have since dropped it. If a diagram says "Synapse Data Warehouse", ask which one it means.

Where the jobs overlap

The groups are not tidy silos. Several jobs can be done in more than one place:

• Spark runs in Databricks, Fabric Data Engineering, Synapse Spark pools and HDInsight.
• Pipelines run in Azure Data Factory, Synapse pipelines and Fabric Data Factory. These are largely the same idea built on three generations of the product.
• KQL over event data runs on Azure Data Explorer and on Fabric Eventhouse, which use the same Kusto engine.
• Streaming ingestion comes in through Event Hubs or Fabric Eventstream, then gets processed in Stream Analytics, ADX/Eventhouse or Spark.
• Fabric covers almost every row of the table: warehouse, lakehouse, Spark, pipelines, real-time and Power BI. That breadth is its pitch, and it's also why it overlaps with everything else.

In practice, when two teams say they need "a data platform", one may picture Databricks and the other Fabric. Both are reasonable. Spot the overlap early so you can make the choice on purpose rather than end up running both by accident.

Walkthrough: inventory what you already run

Before choosing anything new, find out what's already deployed. Azure Resource Graph lets you query every subscription you can read, all at once.

# One-time: add the Resource Graph extension to the Azure CLI
az extension add --name resource-graph

# List analytics resources by type, mapped to our seven groups
az graph query -q "
Resources
| where type in~ (
  'microsoft.synapse/workspaces',          // warehouses + query engines + pipelines
  'microsoft.sql/servers/databases',       // may hide a standalone dedicated SQL pool (formerly SQL DW)
  'microsoft.databricks/workspaces',       // big data processing
  'microsoft.datafactory/factories',       // integration and pipelines
  'microsoft.eventhub/namespaces',         // streaming and messaging
  'microsoft.streamanalytics/streamingjobs',
  'microsoft.kusto/clusters',              // Azure Data Explorer (Kusto)
  'microsoft.hdinsight/clusters',          // check the version is 5.1
  'microsoft.purview/accounts',            // governance
  'microsoft.fabric/capacities',           // Fabric F SKUs
  'microsoft.powerbidedicated/capacities') // older Power BI Embedded capacities
| project name, type, resourceGroup, location
| order by type asc" --output table

Realistic output:

Name             Type                                     ResourceGroup  Location
---------------  ---------------------------------------  -------------  ----------
fabcap-prod      microsoft.fabric/capacities              rg-data        uksouth
hdi-legacy-etl   microsoft.hdinsight/clusters             rg-legacy      westeurope
adx-telemetry    microsoft.kusto/clusters                 rg-iot         uksouth
syn-finance      microsoft.synapse/workspaces             rg-finance     uksouth

Now read the output against the map:

• hdi-legacy-etl: run az hdinsight show -n hdi-legacy-etl -g rg-legacy --query properties.clusterVersion. Anything other than 5.1 is on a retired version.
• syn-finance: this is supported and fine to keep running. Flag it, though, because new features land in Fabric.
• adx-telemetry: this is the same engine as Fabric Eventhouse, so it's a candidate for consolidation later.

Best practice: Fabric items (lakehouses, warehouses, eventhouses) live inside Fabric workspaces, not Azure resource groups. Only the capacity shows up in Resource Graph. To inventory Fabric items you need the Fabric admin portal or the Fabric REST APIs.

Spotting the dead ends before they reach a design

When you review a design, a blog post or a module, run three quick checks:

1. Is the name current? Look for SQL DW, Azure Purview, Data Lake Analytics, Azure Data Catalog, Synapse Data Explorer or HDInsight on AKS. Any of these means the source is old.
2. Is it retired? HDInsight on AKS was stopped and removed on 31 Jan 2025. Synapse Data Explorer pools retired on 7 Oct 2025. Don't design on either.
3. Is it in maintenance mode? Synapse is still supported and has no announced end-of-life date. Microsoft's guidance, though, points new real-time and Cosmos DB analytics work at Fabric. Synapse Link for Cosmos DB, for example, is no longer recommended for new projects; Cosmos DB mirroring into Fabric is the suggested path. Using Synapse is a valid choice, but write down why you picked it.

Trade-off: "Maintenance mode" is not the same as "bad". A stable Synapse dedicated pool your team knows well can beat a migration nobody has time for. The risk is a slower roadmap, not a broken service.

Key takeaways

• There are seven groups: warehouses and lakehouses, query engines, integration and pipelines, streaming and messaging, big data processing, catalogues and governance, and BI.
• The warehouse name chain is SQL Data Warehouse → Synapse dedicated SQL pool → (new work) Fabric Data Warehouse.
• Azure Purview became Microsoft Purview in April 2022. Kusto is the engine behind Azure Data Explorer and Fabric Eventhouse.
• Retired: HDInsight on AKS (31 Jan 2025), Synapse Data Explorer (7 Oct 2025), Data Lake Analytics (29 Feb 2024), Azure Data Catalog (15 May 2024).
• Synapse is supported, but new investment goes to Fabric. Choose it on purpose, not by default.
• Spark, pipelines and KQL each appear in several services. Spot the overlap before you end up paying for two of the same thing.