Handling Time in Azure Stream Analytics

When you’re collecting streaming data for analysis, there’s a delay between the time when an event occurs and the time when your analytics system receives the data about that event. Usually, this delay is relatively short, but it can be much longer, such as when a system sending data temporarily loses its connection to the system receiving the data. This delay causes a surprising number of complications.

In this lesson, you’ll learn how Azure Stream Analytics handles late-arriving, early-arriving, and out-of-order events. You’ll also learn how it uses a watermark to determine when to output results.

Learning Objectives

• Describe how Azure Stream Analytics handles late-arriving, early-arriving, and out-of-order events
• Describe the function of watermarks in Azure Stream Analytics
• Describe the metrics that can be used to monitor and troubleshoot Azure Stream Analytics jobs

Intended Audience

• Azure data engineers

Prerequisites

• Basics of Azure Stream Analytics (or take Introduction to Azure Stream Analytics)
• How it ingests data
• How it uses time windows to process data