Azure Stream Analytics is a big data analytics service for the Internet of Things (IoT) that provides data analytics services over streaming data. It enables developers to combine streaming data with historical data to derive business insights. It is a fully managed service that has low latency and provides scalable infrastructure for complex data processing.
How Azure Stream Analytics is used?
Azure Stream Analytics primarily empowers organizations to extract and derive useful business intelligence, insights and information from streaming data.
- Provides a simple, declarative query model for describing transformations.
- Implement temporal functions, including temporal-based joins, windowed aggregates, temporal filters, and other common operations such as joins, aggregates, projections, and filters.
- Enables in going back in time and investigate computations for scenarios like root-cause analysis and what-if analysis.
- Provides users the ability to specify and use reference data
What to monitor in Azure Stream Analytics?
Here are the metrics that must be monitored for monitoring usage and performance of Stream Analytics jobs:
- SU % Utilization – An indicator of the relative event processing capacity for one or more of the query steps.
- Errors – Number of error messages incurred by a Stream Analytics job.
- Input events – Amount of data received by the Stream Analytics job, in terms of event count.
- Output events – Amount of data sent by the Stream Analytics job to the output target, in terms of event count.
- Out-of-order events – Number of events received out of order that were either dropped or given an adjusted timestamp, based on the out-of-order policy.
- Data conversion errors – Number of data conversion errors incurred by a Stream Analytics job.