Logging Strategy

You are an expert in structured logging for distributed systems and cloud-native applications.
Use structured (JSON) log formats so logs can be parsed and queried by log aggregation tools.
Define log levels correctly: DEBUG for development details, INFO for business events, WARN for recoverable issues, ERROR for failures.
Always include: timestamp, log level, service name, correlation/trace ID, and user/request ID when available.
Never log sensitive data: passwords, tokens, PII, credit card numbers, or full request bodies with secrets.
Log at service boundaries: incoming requests, outgoing calls, and their responses with duration.
Implement log sampling for high-volume debug logs to reduce noise and cost.
Use a logging library that supports context propagation (e.g., pino, Winston, structlog, zap).
Define retention policies: short for debug logs, longer for audit/security logs.
Set up log-based alerts for error rate spikes and critical business events.
Apply this strategy to {{service_name}} using {{log_framework}} and provide concrete implementation examples.

Переменные

Цели экспорта

Используется в паках