AI-200
Azure
Back to Observe and troubleshoot apps on Azure

Instrument an app with OpenTelemetry

Exercise - Instrument an app with the OpenTelemetry SDK

Overview

This exercise walks through instrumenting a small Python AI-style API, exporting traces to Application Insights, and verifying spans in the portal and with KQL.

Goals

  • Create (or reuse) a Log Analytics workspace + Application Insights resource
  • Install and configure the Azure Monitor OpenTelemetry distro
  • Add a custom span with attributes for an “embedding” step
  • Confirm telemetry in AppRequests / AppDependencies / custom dimensions

Prerequisites

  • Azure CLI logged in (az login)
  • Python 3.10+
  • Resource group (e.g. rg-ai200)

Steps

1. Create monitoring resources

Azure CLI
az monitor log-analytics workspace create \
  --resource-group rg-ai200 \
  --workspace-name law-ai200 \
  --location eastus

az monitor app-insights component create \
  --app appi-ai200 \
  --location eastus \
  --resource-group rg-ai200 \
  --workspace law-ai200

az monitor app-insights component show \
  --app appi-ai200 \
  --resource-group rg-ai200 \
  --query connectionString -o tsv

Save the connection string.

2. Install packages

Azure CLI
pip install azure-monitor-opentelemetry flask

3. Instrument a minimal app

Python
import os
from flask import Flask, jsonify
from azure.monitor.opentelemetry import configure_azure_monitor
from opentelemetry import trace

os.environ.setdefault("OTEL_SERVICE_NAME", "ai200-demo-api")
configure_azure_monitor(
    connection_string=os.environ["APPLICATIONINSIGHTS_CONNECTION_STRING"],
)

app = Flask(__name__)
tracer = trace.get_tracer(__name__)

@app.get("/embed")
def embed():
    with tracer.start_as_current_span("generate_embedding") as span:
        span.set_attribute("gen_ai.model", "text-embedding-3-large")
        span.set_attribute("documents.count", 1)
        # pretend work
        vector = [0.1, 0.2, 0.3]
        return jsonify({"dims": len(vector)})

if __name__ == "__main__":
    app.run(port=8080)

4. Generate traffic

Azure CLI
export APPLICATIONINSIGHTS_CONNECTION_STRING="..."
python app.py
# in another shell
curl http://127.0.0.1:8080/embed

Wait 1–2 minutes for ingestion.

5. Verify in KQL

Application Insights → Logs:

Kql
AppRequests
| where TimeGenerated > ago(15m)
| where Name has "embed"
| project TimeGenerated, Name, Success, DurationMs, OperationId

AppDependencies
| where TimeGenerated > ago(15m)
| take 20

Or open Transaction search / Performance and find the generate_embedding span with attributes.

Checklist

  • Connection string set; OTEL_SERVICE_NAME is not unknown_service
  • Request appears in AppRequests
  • Custom attributes visible under customDimensions
  • You can copy OperationId and reconstruct the operation with union

Learn more