Introduction to Azure Streaming Analytics Presented By: Warren Sifre
Who Am I? Email: warren.sifre@moserit.com Twitter: @WAS_SQL Professional History 20 years in the technology industry focusing on Information Technology Email: warren.sifre@moserit.com Twitter: @WAS_SQL LinkedIn: www.linkedin.com/in/wsifre
Agenda What is Streaming Analytics? How it works? Use Cases Configuration and Dependencies SAQL Demo
Typical Environment
Challenges… Real-Time Analytics Environment Scalability Many Steps in between the Source Data and the Visualization/Reporting Output. Environment Scalability Months/Years of planning is needed to plan out equipment procurement and scale out to meet increasing demand. Resource Cost Management Ideal configuration would require the purchasing of enough equipment to handle peak performance requirements. Although those peak requirements may only be for a few hours of any given day. Disaster Recovery Strategy Architecting and maintaining a DR strategy where performance, RTOs, and RPOs are met can be challenging and leave the organization with a lot of underutilized resources.
What is Streaming Analytics? A way to evaluate data before it has reached its final repository destination. Why? Hours to weeks can be the time it takes for data to be transmitted, received, processed, aggregated, then visualized in the traditional Data Warehouse architecture. Business requirements have changed and the desire to glean insights from this data sooner is now becoming a requirement, not a nice to have.
How it can work Streaming Analytics Job Event Hub Power BI Process and Deliver data to multiple end points Transmit Data Event Hub Queues data for processing Power BI Visualize real-time data stream Data Factory Gather and process data for Predictive Analytics Process Machine Learning Predictive Analytics Processing Azure SQL Store data
Reimagined Environment
Use Cases Transportation Energy Manufacturing Medical Device Reduce the need to pull vehicles from service for routine inspections by using sensors to determine when actual anomalies are occurring. Energy Monitor equipment from central locations such as Wind Turbines and Power Generators, thus reducing time spent on manual/physical inspection or replacement of parts just because of time instead of actual degraded performance. Manufacturing Monitor equipment and plant conditions for optimal performance. Medical Device Through remote monitoring expensive replacement parts can be ordered closer to the end-of-life of an equipment than by a schedule. This can reduce the cost of having an overstock of parts on-hand.
Configuration Options
SAQL - Elements
SAQL - Elements
Tumbling Window Fixed window of time with no overlap
Hopping Window Fixed window of time with a fix time of overlap
Sliding Window A Fixed window time, but a window is defined as the moment an event enters or exits an existing window.
Sliding Window A Fixed window time, but a window is defined as the moment an event enters or exits an existing window.
Sliding Window A Fixed window time, but a window is defined as the moment an event enters or exits an existing window.
Sliding Window A Fixed window time, but a window is defined as the moment an event enters or exits an existing window.
Sliding Window A Fixed window time, but a window is defined as the moment an event enters or exits an existing window.
Sliding Window A Fixed window time, but a window is defined as the moment an event enters or exits an existing window.
Sliding Window A Fixed window time, but a window is defined as the moment an event enters or exits an existing window.
Demonstration
Helpful Links More Information on Streaming Analytics SAQL https://msdn.microsoft.com/en-us/library/azure/dn834998.aspx SAQL Query Patterns https://azure.microsoft.com/en-us/documentation/articles/stream-analytics-stream-analytics-query-patterns/ Azure Portal Link Azure Portal
Thank You! Email: warren.sifre@moserit.com Twitter: @WAS_SQL LinkedIn: www.linkedin.com/in/wsifre