Data Analytics
Microsoft Fabric · Synapse · Databricks · Data Explorer · Data Lake
Practice questions for this topic →
Pick the platform
| Service | What it is | Pick when… |
|---|---|---|
| Microsoft Fabric | SaaS analytics platform; OneLake shared storage; Data Engineering, Data Factory, Data Science, Real-Time Intelligence, Data Warehouse | New unified analytics with Power BI; minimal infrastructure |
| Azure Synapse Analytics | SQL (dedicated and serverless), Spark, pipelines; Synapse Data Explorer was retired Oct 7, 2025 | Existing tuned dedicated SQL pool estates |
| Azure Databricks | Unified lakehouse: Spark, Delta Lake, MLflow, Unity Catalog | Spark engineering, ML, multi-cloud lakehouse |
| Azure Data Explorer / Fabric Eventhouse | Fully managed near-real-time analytics with KQL over logs, time series and IoT | Interactive analysis of high-volume telemetry; also powers Azure Monitor Logs and Application Insights |
| Azure Stream Analytics | Fully managed stream processing with submillisecond latency and a SQL-like language | Real-time windowed aggregations from Event Hubs or IoT Hub; also runs on IoT Edge |
| Data Lake Storage Gen2 | A set of capabilities on Blob Storage unlocked by the hierarchical namespace | Foundation for data lakes: directories, ACLs, Hadoop-compatible (ABFS) access via the dfs endpoint |
Synapse Link for Cosmos DB is no longer supported for new projects: use Cosmos DB mirroring in Fabric. Data Lake Storage isn't a separate account type.
Exam Tips
Tick each tip once you can answer it without looking.
“New unified analytics platform with minimal infrastructure management and Power BI integration” → Microsoft Fabric.
“Interactive analysis of billions of log/telemetry records with KQL” → Azure Data Explorer / Fabric Eventhouse. Trap: Synapse Data Explorer is retired.
“Large-scale Spark ML pipelines, collaborative notebooks” → Azure Databricks.
“Existing heavily tuned MPP warehouse in dedicated SQL pools” → Keep Synapse (plan Fabric migration on own timeline).
“Data lake storage with directories and fine-grained ACLs” → ADLS Gen2 (hierarchical namespace).
“Ad hoc SQL over files in the lake without provisioning” → Serverless SQL (Synapse) / Fabric SQL analytics endpoint.
“Azure Data Explorer” → Fully managed near-real-time analytics with KQL over logs, time series and IoT data. Trap: It also underpins Azure Monitor Logs and Application Insights.
“Azure Stream Analytics” → Fully managed stream processing with submillisecond latency and a SQL-like language. Trap: Inputs include Event Hubs and IoT Hub; runs on IoT Edge too.
“Azure Databricks” → Unified lakehouse platform: Spark, Delta Lake, MLflow, Unity Catalog. Trap: Fabric is the SaaS alternative.
“Data Lake Storage Gen2” → Capabilities on Blob Storage unlocked by the hierarchical namespace; Hadoop-compatible ABFS access via the dfs endpoint. Trap: Not a separate account type.