Azure stream analytics

To get real-time analytical computation on streaming data we need a fully managed event processing engine that is provided by Azure Stream Analytics. Data is gathered from devices, social media, websites, application and more.

Things you can do with Stream Analytics

You can use stream analytics to analyze a high volume of data that comes from different sources, take out information from the data stream and you can also find out patterns, trends and more. You can perform applications tasks based on the content of the data.

Working with Stream Analytics

It starts at the opening of streaming data. The data is transferred into Azure from a device. The data can also be obtained from Azure storage.

A Stream Analytics job is created that looks for the origination point of data, it is done to examine the stream. Transformation is specified by the job - how to recognize patterns, relationships, data. For this task, a SQL like a query language is used which is supported by Stream Analytics that gives you an opportunity to sort, aggregate, filter and over a time period join a streaming data.

Finally, the output is specified by job to send transformed data. On basis of analyzed information, you can give a controlled response. Like in response you can send a command to change the setting of a device.

It can be monitored by you and adjust the number of events it processes per second while a job is running. For troubleshooting, you can have the job that produces diagnostic logs.

Capabilities and Advantages

Stream Analytics is designed in such a way that it is easy to use, economical, flexible and any job size is scalable. Many inputs and outputs can be connected

For stream ingestion Stream Analytics can be directly connected to Azure IoT Hub and Azure Event Hub and to ingest historical data you can connect to Azure Blob storage service. The data is collected from event hubs, Stream Analytics can be combined with processing engine and other data sources.

Reference data can be included in Job input. This reference data can be joined to streaming data to perform lookup operation the same way as database queries.

Output can be routed in many directions. You can move to storage such as Azure SQL DB and much more. From there, via Azure HDInsight data may go for batch analytics. Output can be sent to another service to be consumed by another process, like event hubs, queue. For visualization, you can send output to power BI.

Easy to use

With no programming, you can create sophisticated analyses. Input can be taken in form of streaming data by the use of query language. You can then perform calculations, use geospatial functions, filter and sort data, aggregate values. Using syntax checking and IntelliSense you can edit queries portal. Data extracted from live streaming can be used to test queries on sample data.

Extended query language

By defining and invoking additional function you can extend capabilities of the query language. To take advantage of Azure Machine Learning Solutions function call can be defined at Azure Machine Learning Service. In order to perform a complex calculation, you can integrate JavaScript user-defined function.

Scalable

1GB incoming data can be handled by Stream Analytics. Millions of events per second can be ingested by integrating with Azure Event Hub and Azure IoT Hub.

UtterlyCloud

• admin@utterlycloud.com
• India