Aggregation of Logs

When using Kubernetes, we can think of two primary methods for emitting logs. The first is directly from our code to an external logging system. The second is direct output via the standard output and error streams.

To aggregate logs directly to an external logging system, we can use Azure Application Insights. It can natively instrument .NET applications (among many other supported languages) with a few code changes.

To aggregate logs written to the standard output and error streams, we can use Container Insights, which is a capability within Azure Monitor. Container Insights is a great utility for viewing and querying logs produced by multiple Kubernetes clusters, but it currently doesn’t easily support the level of structured interrelationship logging that this application will generate.

The MTAEDA application will use telemetry and logging directly to Application Insights so that we can support tracing.

Now, let’s set up an Applications Insight instance to manage the log aggregation.

Creating an Azure Application Insights instance

We head to the Azure portal and sign in with our user account to access a usable subscription.

We type and find Application Insights in the main search box.