4.2.2 Effective Logging Strategy

Implement an Effective Logging Strategy

Instrumentation

Instrumentation refers to the measure of a product’s performance to:

• Write trace information
• Diagnose errors
• Record application behavior and state
• Feed data into operational dashboards
• Add real-time metrics for visibility

Centralized & Structured Logging

Centralized logging standardizes application logging cho:

• Operational information về workload state
• User behavior tracking
• System state monitoring
• Anomalous events detection

Structured logging libraries ensure:

• ✅ Configurable logging levels
• ✅ API consistency
• ✅ Common output formats (JSON)
• ✅ Easy parsing and analysis

Lambda Powertools Logging

from aws_lambda_powertools import Logger

logger = Logger(service="order-service")

@logger.inject_lambda_context
def handler(event, context):
    logger.info("Processing order", extra={"order_id": event['orderId']})
    # Automatically includes: request_id, function_name, cold_start

Lợi ích:

• Structured JSON logs
• Automatic context injection
• Correlation IDs
• Sampling support

Logging Levels

What to Log

CloudWatch Logs Insights for Lambda Performance

-- Find Lambda performance data from default log events
fields @timestamp, @duration, @billedDuration, @maxMemoryUsed
| filter @type = "REPORT"
| stats avg(@duration) as avg_duration,
        max(@duration) as max_duration,
        pct(@duration, 99) as p99_duration
  by bin(5m)

Lambda Logging Best Practice

import logging
import json

logger = logging.getLogger()
logger.setLevel(logging.INFO)

def handler(event, context):
    logger.info(json.dumps({
        "action": "process_order",
        "order_id": event.get("orderId"),
        "remaining_ms": context.get_remaining_time_in_millis()
    }))

Log Retention Strategy

Multi-layered Architectures & Distributed Systems

Understanding multi-layered architectures và distributed systems requires more than traditional logs and infrastructure metrics. Cần kết hợp:

• Centralized logging
• Structured logging
• Distributed tracing
• Custom metrics