Support for Multiple Outputs, Job Health Alerts, Improved Query Testing


The Stream Analytics team is pleased to announce the release of several new features today.

 

Multiple outputs per job

As a response to strong feedback from ASA users, we have added support for multiple outputs in a single job.  This enables scenarios like alerting over incoming events while simultaneously archiving logs to long term storage.  You can access this feature by using the SELECT… INTO clause.

Alerting over job metrics

We know that alerting on job health is a critical feature for maintaining production pipelines over time, so we added functionality to create alerts over ASA job metrics.  When the value of a specified metric violates the threshold assigned for the alert rule, an email notification is sent.  Learn more about configuring alerts from How to: Receive Alert Notifications and Manage Alert Rules in Azure.

Improved query testing with actual input events

We have extended the existing in-browser query validation experience to enable query testing over actual job data.  This is exposed via the Sample Data feature, which downloads events from your job’s input source in  JSON. The data can be used to test the query defined for the job in the Query Tab.

PartitionKey property for Event Hub output

You can now define an Event Hub PartitionKey field for your output events to determine which Event Hub partition the event will be sent to. 

Updated event ordering policy

In addition to the exiting options for events received by the Stream Analytics job out of order (Out of Order Tolerance Window), you can now specify how to handle events that are delayed in reaching the input source from the client via the Acceptable Input Source Arrival Delay.  Both settings are available on the Configure tab of your ASA job.

 

Thanks!

Thank you for your continued feedback on Stream Analytics. We look forward to hearing from you on the Stream Analytics Forum and Azure Feedback Forum. Stay tuned for more product updates coming soon!


Comments (0)

Skip to main content