Observe a container app
Observability in Azure Container Apps starts with what is happening now, then widens to history and trends. Use the live log stream to see current container activity, Log Analytics to query historical logs with KQL, and Azure Monitor or the Grafana dashboard to track metrics over time. OpenTelemetry exports telemetry to the backend you choose, while the Activity log keeps a separate audit trail for operations on the resource.
Tail stdout/stderr and platform events from a running container, right now.
Query historical app and system logs with KQL — filter, correlate, aggregate.
Track requests, replicas, CPU and memory. Alert or split by revision.
Pipe logs, metrics and traces to the backend of your choice.
Control-plane record of create, update, scale, restart and delete — who, when, and the result.
App and system log streaming
Stream logs in near real time from a running container. The stream includes application logs (stdout/stderr from your code) and system logs from platform lifecycle events such as image pulls, scaling, and health probes. Use it when you need to watch container behavior while you debug. Log streaming in Container Apps.
Log Analytics
When your environment sends logs to a Log Analytics workspace, you can run KQL queries across historical application and system logs. Filter by revision, correlate errors, aggregate over time, and compare behavior across deployments. Use Log Analytics for analysis and troubleshooting beyond the live stream. Query logs with Log Analytics.
Azure Monitor and the Grafana dashboard
Azure Monitor metrics track compute and network usage, including request volume, replica count, CPU, and memory. You can alert on these metrics or split them by revision. For an at-a-glance visual view, point Azure Managed Grafana at the same Azure Monitor data and use the prebuilt Container Apps dashboard to watch traffic and scaling trends across your apps. Container Apps metrics · Azure Managed Grafana.
OpenTelemetry collector
For OpenTelemetry-instrumented apps, the environment can run a managed OpenTelemetry agent. Microsoft provisions and operates the collector inside your environment at no extra compute cost. Instrument your code with the OpenTelemetry SDK, then configure the agent once at the environment level to send logs, metrics, and traces to the destinations you choose: Azure Monitor Application Insights, Datadog, or any OTLP-compatible endpoint (New Relic, Honeycomb, and others). You can switch destinations later without changing your app. Collect and read OpenTelemetry data.
Azure Activity log
The Activity log records what's happening with the resource itself: control-plane operations such as creating the app, updating configuration, deploying a new revision, restarting, scaling, or deleting. Each entry captures who did it, when, and the result. Start here when something changed unexpectedly. Open it from the app's Activity log blade. Azure Monitor activity log.