Public preview: New metrics capabilities in OpenTelemetry-based Application Insights
Дата публикации: 26 октября, 2022
Azure Monitor Application Insights is a cloud native application monitoring offering which enables customers to observe failures, bottlenecks, and usage patterns to resolve incidents faster and reduce downtime.
With the latest release of Azure Monitor OpenTelemetry packages for .NET, Node.js, and Python, we continue to build on OpenTelemetry’s vendor-neutral APIs/SDKs, introducing new capabilities in four areas:
OpenTelemetry-based metrics now flow to Application Insights. This includes metrics emitted by dozens of available OpenTelemetry Instrumentation Libraries or custom metrics you create using OpenTelemetry APIs. you can alert on user behavior that matters to your business, such as an “add to shopping cart” operation.
Sampling empowers you to better optimize cost. The custom sampler uses a sampling algorithm that populates an “itemCount” field that corrects for sampled out events. You can use it alongside existing Application Insights SDKs and traces will be preserved.
Exception capture and correlation is available both when automatically collected by OpenTelemetry Instrumentation Libraries and when added manually to a span as a span event. For example, if your app calls out to database and fails, we capture the details and associate it to the relevant trace for end-to-end troubleshooting.
Azure Monitor Exporters now include offline storage and automatic retries to minimize data loss in the event your service loses its connection with Application Insights, or the Application Insights service is temporarily unavailable.
This release brings our .NET, Node.js, and Python instrumentation closer to parity with Application Insights SDKs. Over the next several months we will continue to add new capabilities to Microsoft’s OpenTelemetry-based offerings.