04 - Logging and Monitoring (Consumption)¶

Capture structured logs, connect Application Insights, and validate telemetry queries.

Prerequisites¶

Tool	Version	Purpose
Node.js	20+	Local runtime and package execution
Azure Functions Core Tools	v4	Local host and publishing
Azure CLI	2.61+	Azure resource provisioning and management

Consumption plan basics

Consumption (Y1) is serverless with scale-to-zero, up to 200 instances, 1.5 GB memory per instance, and a default 5-minute timeout (max 10 minutes).

What You'll Build¶

You will emit structured logs from Node.js v4 handlers, connect Application Insights, and verify telemetry queries return data.

Infrastructure Context

Plan: Consumption (Y1) | Network: Public internet only | VNet: ❌ Not supported

This tutorial adds Application Insights monitoring to your deployed function app.

flowchart TD
    FA[Function App\nConsumption Y1] -->|telemetry| AI[Application Insights]
    AI -->|query| KQL["KQL Queries\n• traces\n• requests\n• dependencies"]
    CLI["Azure CLI"] -->|"az monitor app-insights query"| AI

    style FA fill:#0078d4,color:#fff
    style AI fill:#68217A,color:#fff

flowchart LR
    A[Log with context.log] --> B[Create App Insights]
    B --> C[Set connection string]
    C --> D[Invoke function]
    D --> E[Query traces via KQL]

Steps¶

Step 1 - Set variables (if not already set)¶

export RG="rg-func-node-consumption-demo"
export APP_NAME="<your-function-app-name>"
export LOCATION="koreacentral"

Step 2 - Understand structured logging¶

The Node.js v4 model uses context.log() for structured logging:

const { app } = require('@azure/functions');

app.http('status', {
    methods: ['GET'],
    handler: async (_request, context) => {
        context.log('status endpoint called');
        context.warn('this is a warning');
        context.error('this is an error');
        return { status: 200, jsonBody: { status: 'ok' } };
    }
});

Log levels

context.log() maps to Information level, context.warn() to Warning, and context.error() to Error in Application Insights. These appear in the traces table.

Step 3 - Create Application Insights (if not auto-created)¶

az monitor app-insights component create \
  --app "$APP_NAME-ai" \
  --location "$LOCATION" \
  --resource-group "$RG" \
  --kind web

Auto-created by az functionapp create

Since Azure CLI 2.80+, az functionapp create automatically provisions an Application Insights resource. Check with az functionapp config appsettings list for APPLICATIONINSIGHTS_CONNECTION_STRING. If it already exists, skip this step.

Step 4 - Set the connection string (if needed)¶

AI_CONN=$(az monitor app-insights component show \
  --app "$APP_NAME-ai" \
  --resource-group "$RG" \
  --query "connectionString" \
  --output tsv)

az functionapp config appsettings set \
  --name "$APP_NAME" \
  --resource-group "$RG" \
  --settings "APPLICATIONINSIGHTS_CONNECTION_STRING=$AI_CONN"

Step 5 - Generate telemetry¶

Invoke the function to generate log entries:

curl --request GET "https://$APP_NAME.azurewebsites.net/api/health"
curl --request GET "https://$APP_NAME.azurewebsites.net/api/hello/TelemetryTest"
curl --request GET "https://$APP_NAME.azurewebsites.net/api/requests/log-levels"

Telemetry ingestion delay

Application Insights typically has a 2-5 minute ingestion delay for new telemetry. Wait at least 3 minutes after invoking functions before querying traces.

Step 6 - Query traces¶

az monitor app-insights query \
  --app "$APP_NAME-ai" \
  --resource-group "$RG" \
  --analytics-query "traces | take 5" \
  --output json

Step 7 - Review Consumption-specific notes¶

Use --consumption-plan-location for app creation and expect cold starts under idle periods.
Use long-form CLI flags for maintainable runbooks.
Keep FUNCTIONS_WORKER_RUNTIME=node across all environments.

Verification¶

The traces | take 5 query should return rows with log messages:

{
  "tables": [
    {
      "name": "PrimaryResult",
      "columns": [
        { "name": "timestamp", "type": "datetime" },
        { "name": "message", "type": "string" },
        { "name": "severityLevel", "type": "int" }
      ],
      "rows": [
        ["2026-04-10T00:20:00.0000000Z", "Handled hello for TelemetryTest", 1],
        ["2026-04-10T00:20:00.0000000Z", "Health check requested", 1]
      ]
    }
  ]
}

Severity levels

0 = Verbose
1 = Information (context.log)
2 = Warning (context.warn)
3 = Error (context.error)

Next Steps¶

Next: 05 - Infrastructure as Code