azure-functions-logging

Read this in: 한국어 | 日本語 | 简体中文

Invocation-aware observability for Azure Functions Python v2. Surfaces invocation_id, detects cold starts, warns on host.json misconfig, and outputs Application Insights-ready structured logs — without replacing Python's standard logging.

---

Part of the Azure Functions Python DX Toolkit → Bring FastAPI-like developer experience to Azure Functions

Why this exists

Azure Functions Python logging has specific failure modes that generic logging libraries don't address:

Problem	What happens	This library
host.json log level conflict	Your INFO logs silently disappear in Azure	Detects and warns at startup
No invocation_id in logs	Impossible to correlate logs to a specific execution	Auto-injects from context object
Cold start invisible	No signal when a new worker instance starts	Detects automatically on first inject_context()
Noisy third-party loggers	azure-core, urllib3 flood your Application Insights	SamplingFilter / RedactionFilter
Local vs cloud output mismatch	Colorized output breaks in production pipelines	Environment-aware formatter switching
PII leaking into logs	Sensitive fields logged in exception tracebacks	RedactionFilter with pattern matching

What it does

• Invocation context — auto-injects invocation_id, function_name, cold_start into every log
• Structured JSON output — Application Insights-ready NDJSON format for production
• Noise control — SamplingFilter rate-limits chatty third-party loggers
• PII protection — RedactionFilter masks sensitive fields before they reach log aggregation

Before / After

Without azure-functions-logging — plain print() output, no context, no structure:

With azure-functions-logging — colorized local dev output and production-ready JSON:

What this package does not do

This package does not own:

• Replacing stdlib logging — it wraps and enriches Python's standard logging, never replaces it
• Distributed tracing — use OpenTelemetry or Application Insights SDK for end-to-end trace correlation
• API documentation — use azure-functions-openapi for API documentation and spec generation

Installation

pip install azure-functions-logging

Quick Start

import azure.functions as func
from azure_functions_logging import get_logger, inject_context, setup_logging

setup_logging()
logger = get_logger(__name__)

app = func.FunctionApp()

@app.route(route="hello")
def hello(req: func.HttpRequest, context: func.Context) -> func.HttpResponse:
    inject_context(context)  # binds invocation_id, function_name, cold_start

    logger.info("Request received")
    # {"level": "INFO", "invocation_id": "abc-123", "cold_start": true, ...}

    return func.HttpResponse("OK")

Start the Functions host locally (using the e2e example app):

func start

Verify locally and on Azure

After deploying (see docs/deployment.md), the same request produces the same response in both environments.

Local

curl -s http://localhost:7071/api/logme?correlation_id=demo-123

{"logged": true, "correlation_id": "demo-123"}

Azure

curl -s "https://<your-app>.azurewebsites.net/api/logme?correlation_id=demo-123"

{"logged": true, "correlation_id": "demo-123"}

Verified against a temporary Azure Functions deployment in koreacentral (Python 3.12, Consumption plan). Response captured and URL anonymized.

Invocation Context

inject_context(context) should be the first line of every handler. It binds:

• invocation_id — unique per execution, correlates all logs for one request
• function_name — the Azure Functions function name
• trace_id — trace context from the platform
• cold_start — True on first invocation of this worker process

def my_function(req, context):
    inject_context(context)
    logger.info("handler started")
    # every log from here carries invocation_id and cold_start

Without inject_context(), these fields are None in every log line.

with_context Decorator

For less boilerplate, use the with_context decorator instead of calling inject_context() manually:

import azure.functions as func
from azure_functions_logging import get_logger, setup_logging, with_context

setup_logging()
logger = get_logger(__name__)

app = func.FunctionApp()

@app.route(route="hello")
@with_context
def hello(req: func.HttpRequest, context: func.Context) -> func.HttpResponse:
    logger.info("Request received")
    return func.HttpResponse("OK")

The decorator finds the context parameter by name, calls inject_context() before your handler runs, and resets context variables in finally after it returns.

Custom parameter name:

@with_context(param="ctx")
def hello(req: func.HttpRequest, ctx: func.Context) -> func.HttpResponse:
    ...

Both sync and async handlers are supported.

Structured JSON Output (Production)

Use JSON format when logs feed Application Insights or any aggregation system:

setup_logging(format="json")

Output per log line (NDJSON — one JSON object per line):

{"timestamp": "2024-01-15T10:30:00Z", "level": "INFO", "logger": "my_module",
 "message": "order accepted", "invocation_id": "abc-123", "function_name": "OrderHandler",
 "cold_start": false, "trace_id": "00-abc...", "exception": null,
 "extra": {"order_id": "o-999"}}

Extra fields appear in extra and are indexable in Application Insights:

logger.info("order accepted", order_id="o-999", tenant_id="t-1")

host.json Conflict Detection

If your host.json suppresses log levels that your app emits, you get this warning at startup:

WARNING: host.json logLevel.default is 'Warning'. Logs below WARNING will be suppressed in Azure.

Recommended host.json baseline:

{
  "version": "2.0",
  "logging": {
    "logLevel": {
      "default": "Information",
      "Function": "Information"
    }
  }
}

Noise Control

Suppress chatty third-party loggers without removing them:

from azure_functions_logging import SamplingFilter, setup_logging
import logging

setup_logging()

# Only log 1 in 10 azure-core messages
logging.getLogger("azure").addFilter(SamplingFilter(rate=0.1))

# Silence urllib3 completely in production
logging.getLogger("urllib3").setLevel(logging.WARNING)

PII Redaction

Strip sensitive fields before they reach Application Insights:

from azure_functions_logging import RedactionFilter, setup_logging
import logging

setup_logging()
root = logging.getLogger()
root.addFilter(RedactionFilter(patterns=["password", "token", "secret"]))

Any log record where the message or extra fields match a pattern will have those values replaced with [REDACTED].

Local vs Cloud

Environment	Format	Behavior
Local terminal	color (default)	Colorized [TIME] [LEVEL] [LOGGER] message
Azure / Core Tools	json	NDJSON, no ANSI codes, host-managed handlers
CI / pipeline	json	NDJSON, machine-parseable

setup_logging() detects FUNCTIONS_WORKER_RUNTIME and WEBSITE_INSTANCE_ID to choose the right path automatically. In Azure, it installs context filters without adding handlers (avoids duplicate output from the host pipeline).

Context Binding

Attach request-scoped metadata to every log without passing it through every call:

def process_order(order_id: str) -> None:
    order_logger = logger.bind(order_id=order_id, region="eastus")
    order_logger.info("processing started")   # includes order_id + region
    order_logger.info("processing complete")  # same metadata, new message

Create bound loggers per-invocation. Do not cache them at module level.

When to use

• You need structured, queryable logs in Application Insights
• You want invocation_id correlation across all logs for a single request
• You need cold start detection without custom instrumentation
• You want PII redaction or noise control for third-party loggers
• Your host.json config silently suppresses logs and you don't know why

Documentation

• Full docs: yeongseon.github.io/azure-functions-logging
• Configuration reference
• Troubleshooting guide
• API reference

Ecosystem

This package is part of the Azure Functions Python DX Toolkit.

Design principle: azure-functions-logging owns structured logging and invocation-aware observability. It enriches Python's standard logging — it does not replace it. Adjacent concerns belong to azure-functions-openapi (API documentation and spec generation), azure-functions-validation (request/response validation and serialization), and azure-functions-langgraph (LangGraph runtime exposure).

Package	Role
azure-functions-openapi	OpenAPI spec generation and Swagger UI
azure-functions-validation	Request/response validation and serialization
azure-functions-db	Database bindings for SQL, PostgreSQL, MySQL, SQLite, and Cosmos DB
azure-functions-langgraph	LangGraph deployment adapter for Azure Functions
azure-functions-scaffold	Project scaffolding CLI
azure-functions-logging	Structured logging and observability
azure-functions-doctor	Pre-deploy diagnostic CLI
azure-functions-durable-graph	Manifest-first graph runtime with Durable Functions (experimental)
azure-functions-python-cookbook	Recipes and examples

For AI Coding Assistants

This package provides structured logging for Azure Functions with zero modifications to stdlib logging.

LLM-Friendly Resources:

• llms.txt — Concise API reference and quick start (repo root)
• llms-full.txt — Complete API signatures, patterns, and design principles (repo root)

Key Implementation Details for Code Generation:

1. Never modify root logger — Only install filters/formatters on handlers
2. Context injection is contextvar-based — Not thread-local, works with asyncio
3. Idempotent setup — Calling setup_logging() multiple times is safe
4. Two environments, two behaviors:
   • Azure/Core Tools: Only install filters on existing root handlers (respects host.json)
   • Local dev: Add ColorFormatter or JsonFormatter handler to specified logger
5. Test-friendly:
   • inject_context() accepts any object (no hard dependency on azure.functions.Context)
   • with_context decorator works with sync and async handlers
   • Use context.reset_context_vars() in test teardown if needed

When generating code:

• Import only from azure_functions_logging public API (no underscores)
• Call setup_logging() at module level or handler startup (not per-request)
• Call inject_context(context) as first line of handler
• Use logger.bind(key=value) for per-request fields (not direct logger.extra)
• Apply RedactionFilter for PII fields, SamplingFilter for high-volume logs

Example Pattern:

from azure_functions_logging import setup_logging, get_logger, inject_context

# Module level
setup_logging()
logger = get_logger(__name__)

# Per handler
def my_function(req: func.HttpRequest, context: func.Context) -> func.HttpResponse:
    inject_context(context)
    req_logger = logger.bind(correlation_id=req.params.get("id"))
    req_logger.info("Processing")
    return func.HttpResponse("OK")

This project is an independent community project and is not affiliated with, endorsed by, or maintained by Microsoft.

Azure and Azure Functions are trademarks of Microsoft Corporation.

License

MIT