GABC2018: Mastering IoT Hub, Azure Stream Analytics and Power BI by Matthias Gessenay

In this demo packed session we will connect a software based Smart Meter to the Azure IoT Hub, feed and transform data with Stream analytics – hot and cold data stream – and display the results in Power BI, both direct through Stream Analytics and Apache Spark. This will show you the end to end design of a typical IoT scenario.

Transcript

1. Mastering IoT Hub, Azure Stream Analytics and Power BI Global Azure Bootcamp 2018 Matthias Gessenay
2. Matthias Gessenay • Senior Consultant @ Corporate Software • MPP Data Science, Cloud Administration • Various MCSA/E • MCT • Focus on Azure • matthias.gessenay@corporatesoftware.ch • LinkedIn: https://www.linkedin.com/matthias-gessenay
3. Agenda • IoT in Azure • Technical Fundamentals • IoT Hub • IoT Central • IoT Sphere • The Scenario
4. IoT now? • Technology is not that new (device connection) • But ... • Hardware is small and cheap • Connection is available • Big Data Processing is possible
5. Technical Components involved • IoT Hub • HDInsight • Stream Analytics • Power BI • Jupyter
6. IoT Hub • Up to 10 million devices • Telemetry ingestion • Bi-directional device <-> cloud • Device Twins • Command & control • Device registry & identity • HTTP/AMQP/AMQP-WS/MQTT • Extensible protocol support • Operations Monitoring
7. HDInsight • Hadoop as PaaS • Map / Reduce • Head and Worker Nodes • Apache Spark • In Memory Processing • Used as cold data stream
8. Stream Analytics • On-demand real-time analytics service • Complex Event Processing (CEP) • Pay per Job • Interface to Power BI • SQL Like Query Language
9. Power BI • SaaS BI Dashboards • Inline Filtering • Very capable ETL Engine • Hybrid Connectors • Extensible Graphics • R Support • Support for live data streams
10. Jupyter • «OneNote for Data Scientists» • Supports Code Execution with Kernels • Interactive elements • Capable of R
11. Scenario • A client app simulates temperatur sensors • Data is passed to IoT-Hub • Data is selected via Stream Analytics • Hot Data Stream • Cold Data Stream • Hot Data Stream is passed to Power BI • Cold Data Stream is passed to HDInsight Spark • Jupyter Notebook / CSV • Common Scenario for IoT except local aggregator / edge node
12. Smart Meter IoT Hub Stream Analytics • Hot • Cold Output • Power BI • CSV
13. Questions?