logger

Overview ¶

Package logger provides a structured logging utility built on top of zerolog, designed specifically for AWS Lambda workloads.

The logger automatically:

• Configures log level from environment variables
• Emits JSON structured logs
• Enriches logs with AWS Lambda context (request ID, ARN, memory, etc.)
• Supports contextual and formatted logging

Log Level Configuration ¶

The log level is controlled using the environment variable:

POWERTOOLS_LOG_LEVEL

Supported values:

ERROR, WARN, INFO, DEBUG, TRACE

If the variable is not set, the default log level is DEBUG.

AWS Lambda Context Injection ¶

When running inside AWS Lambda, the logger can inject runtime metadata from the Lambda execution context:

• function_name
• function_memory_size
• function_arn
• function_request_id

This is achieved by calling InjectContext with the Lambda context.

Usage ¶

Basic usage:

log := logger.New(types.LogConfig{})
log.Info("service started")

Injecting Lambda context:

func handler(ctx context.Context) {
    log := logger.New(types.LogConfig{})
    log.InjectContext(ctx)
    log.Info("request received")
}

Structured Properties ¶

Static key-value properties can be attached to every log entry using LogConfig.Properties:

log := logger.New(types.LogConfig{
    Properties: map[string]string{
        "service": "orders",
        "env":     "prod",
    },
})

Output ¶

Logs are emitted in JSON format and written to stdout by default, making them compatible with AWS CloudWatch Logs and log aggregators.

This package is inspired by AWS Powertools logging patterns and follows best practices for observability in serverless environments.