Instrument Python Apps with Azure Monitor

Wraps azure-monitor-opentelemetry for one-line Application Insights setup with OpenTelemetry auto-instrumentation, configured via a single configure_azure_monitor() call that reads the connection string from APPLICATIONINSIGHTS_CONNECTION_STRING by default, or accepts one explicitly.

from azure.monitor.opentelemetry import configure_azure_monitor

# One-line setup - reads connection string from environment
configure_azure_monitor()

It shows drop-in setup for Flask, Django, and FastAPI (calling configure_azure_monitor() before defining routes or settings), and covers custom telemetry beyond auto-instrumentation: custom trace spans via opentelemetry.trace (tracer.start_as_current_span, setting custom attributes), custom metrics via opentelemetry.metrics (creating and incrementing a counter with dimensions), and custom logs via the standard logging module, which are captured into Application Insights including exception info. Configuration options cover sampling (sampling_ratio, e.g. 0.1 to sample 10% of requests), setting a cloud role name for the Application Map via an OpenTelemetry Resource, restricting which instrumentations are enabled (instrumentations=["flask", "requests"]), enabling the Live Metrics stream (enable_live_metrics=True), and authenticating via Azure AD with a DefaultAzureCredential.

It lists eleven libraries auto-instrumented for traces out of the box: Flask, Django, FastAPI, Requests, urllib3, httpx, aiohttp, psycopg2, pymysql, pymongo, and redis, and documents the full configure_azure_monitor() parameter table (connection_string, credential, sampling_ratio, resource, instrumentations, enable_live_metrics) with defaults. Stated best practices: call configure_azure_monitor() early, before importing the libraries it instruments; use environment variables for the connection string in production; set a cloud role name for multi-service applications; enable sampling under high traffic; use structured logging for better log-analytics queries; add custom attributes to spans for easier debugging; and use AAD authentication for production workloads.

Use it to add Application Insights observability - traces, metrics, and logs - to a Python application with minimal setup, especially Flask, Django, or FastAPI services. As with any skill output, verify results against the current environment, and stop to ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.

Input is an Application Insights connection string (or Azure AD credential) and, optionally, sampling/resource/instrumentation configuration. Output is auto-collected traces, metrics, and logs, plus any custom spans, counters, or log records, flowing into Application Insights, with the configured cloud role name shown on the Application Map.

Instruments Flask, Django, FastAPI, Requests, urllib3, httpx, aiohttp, psycopg2, pymysql, pymongo, and redis automatically, builds on the OpenTelemetry SDK (opentelemetry.trace, opentelemetry.metrics), Python's standard logging module, and azure-identity for AAD-based authentication.

Python developers who want Application Insights observability wired into a Flask, Django, or FastAPI service with a single setup call, plus the option to add custom traces, metrics, and logs.