Azure Stream Analytics

Process and analyze streaming data in real time.

Key Concepts

Stream Analytics uses a familiar SQL-like query language (Stream Analytics Query Language - SAQL) to define the data processing logic. Here are some key concepts:

• Input: Data streams from sources like Azure IoT Hub, Azure Event Hubs, or Azure Blob Storage.
• Output: Processed data sent to various sinks such as Azure Blob Storage, Azure SQL Database, Azure Cosmos DB, Power BI, or Azure Data Lake Storage.
• Job: A Stream Analytics job encapsulates the inputs, query, and outputs for processing a data stream.
• Query: Defines the transformations and analytics to be performed on the input data streams using SAQL.
• Functions: Built-in and user-defined functions for complex data manipulation and analysis.

Getting Started

Follow these steps to start using Azure Stream Analytics:

1. Create an Azure Stream Analytics job: Navigate to the Azure portal, create a new Stream Analytics job, and configure its properties.
2. Configure Input: Add an input source to your job, specifying the event serialization format (e.g., JSON, CSV) and the connection details for your event hub or IoT hub.
3. Configure Output: Add an output sink where the processed data will be sent. Choose the appropriate output type and configure its connection settings.
4. Write a Query: Use the Stream Analytics Query Language (SAQL) to define how you want to process the incoming data. You can filter, aggregate, join, and transform data in real time.

   SELECT
       DeviceId,
       AVG(Temperature) AS AverageTemperature
   FROM
       iotHubInput
   GROUP BY
       DeviceId,
       TumblingWindow(duration = 5, size = 1) -- Analyze over 5-minute tumbling windows
   HAVING
       AVG(Temperature) > 30
   INTO
       outputSink
   

5. Start the Job: Once configured, start the Stream Analytics job to begin processing your streaming data.

Use Cases

Azure Stream Analytics is ideal for a wide range of real-time scenarios:

• IoT Telemetry Analysis: Monitor sensor data from devices, detect anomalies, and trigger alerts.
• Fraud Detection: Analyze transaction streams in real time to identify suspicious patterns.
• Real-time Dashboarding: Aggregate and visualize streaming data for business intelligence.
• Log Analysis: Process and analyze application logs for operational monitoring and troubleshooting.
• Clickstream Analysis: Understand user behavior on websites and applications by analyzing click data in real time.

Learn More

Explore the following resources for in-depth information:

• What is Azure Stream Analytics?
• Stream Analytics Query Language (SAQL) Reference
• Stream Processing with Azure Stream Analytics