Python Observability

Instrument Python applications with structured logs, metrics, and traces. When something breaks in production, you need to answer “what, where, and why” without deploying new code.

When to Use This Skill

• Adding structured logging to applications
• Implementing metrics collection with Prometheus
• Setting up distributed tracing across services
• Propagating correlation IDs through request chains
• Debugging production issues
• Building observability dashboards

Core Concepts

1. Structured Logging

Emit logs as JSON with consistent fields for production environments. Machine-readable logs enable powerful queries and alerts. For local development, consider human-readable formats.

2. The Four Golden Signals

Track latency, traffic, errors, and saturation for every service boundary.

3. Correlation IDs

Thread a unique ID through all logs and spans for a single request, enabling end-to-end tracing.

4. Bounded Cardinality

Keep metric label values bounded. Unbounded labels (like user IDs) explode storage costs.

Quick Start

import structlog

structlog.configure(
    processors=[
        structlog.processors.TimeStamper(fmt="iso"),
        structlog.processors.JSONRenderer(),
    ],
)

logger = structlog.get_logger()
logger.info("Request processed", user_id="123", duration_ms=45)

Fundamental Patterns

Pattern 1: Structured Logging with Structlog

Configure structlog for JSON output with consistent fields.

import logging
import structlog

def configure_logging(log_level: str = "INFO") -> None:
    """Configure structured logging for the application."""
    structlog.configure(
        processors=[
            structlog.contextvars.merge_contextvars,
            structlog.processors.add_log_level,
            structlog.processors.TimeStamper(fmt

Pattern 2: Consistent Log Fields

Every log entry should include standard fields for filtering and correlation.

import structlog
from contextvars import ContextVar

# Store correlation ID in context
correlation_id: ContextVar[str] = ContextVar("correlation_id", default="")

logger = structlog.get_logger()

def process_request(request: Request) -> Response:

Pattern 3: Semantic Log Levels

Use log levels consistently across the application.

# DEBUG: Detailed internal information
logger.debug("Cache lookup", key=cache_key, hit=cache_hit)

# INFO: Normal operational events
logger.info("Order created", order_id=order.id, total=order.total)

# WARNING: Abnormal but handled situations
logger.warning(
    "Rate limit approaching",
    current_rate=950,

Never log expected behavior at ERROR. A user entering a wrong password is INFO, not ERROR.

Pattern 4: Correlation ID Propagation

Generate a unique ID at ingress and thread it through all operations.

from contextvars import ContextVar
import uuid
import structlog

correlation_id: ContextVar[str] = ContextVar("correlation_id", default="")

def set_correlation_id(cid: str | None = None

Propagate to outbound requests:

import httpx

async def call_downstream_service(endpoint: str, data: dict) -> dict:
    """Call downstream service with correlation ID."""
    async with httpx.AsyncClient() as client:
        response = await client.post(
            endpoint,
            json=data,
            headers={"X-Correlation-ID": correlation_id.get()},

Detailed worked examples and patterns

Detailed sections (starting with ## Advanced Patterns) live in references/details.md. Read that file when the navigation summary above is insufficient.

Best Practices Summary

1. Use structured logging - JSON logs with consistent fields
2. Propagate correlation IDs - Thread through all requests and logs
3. Track the four golden signals - Latency, traffic, errors, saturation
4. Bound label cardinality - Never use unbounded values as metric labels
5. Log at appropriate levels - Don’t cry wolf with ERROR
6. Include context - User ID, request ID, operation name in logs
7. Use context managers - Consistent timing and error handling
8. Separate concerns - Observability code shouldn’t pollute business logic
9. Test your observability - Verify logs and metrics in integration tests
10. Set up alerts - Metrics are useless without alerting