JSON Logging¶

This example demonstrates how to configure Logly for structured JSON output, ideal for log aggregation systems like ELK stack, Splunk, or CloudWatch.

Code Example¶

from logly import logger
import json

# Configure for JSON output
logger.configure(level="INFO", json=True)

# Add console sink without colors for JSON
logger.add("console", colorize=False)

# Log structured data
logger.info("User authentication successful", user_id=12345, ip="192.168.1.100")
logger.warning("Rate limit exceeded", user_id=12345, requests_per_minute=150)
logger.error("Database connection failed", error_code="ECONNREFUSED", retry_count=3)

# Log with complex nested data
user_data = {
    "id": 12345,
    "name": "John Doe",
    "preferences": {
        "theme": "dark",
        "notifications": True
    }
}
logger.info("User profile updated", user=user_data)

Expected Output¶

The JSON logging produces structured, machine-readable log entries:

{"timestamp": "2025-01-15T10:30:45.123456Z", "level": "INFO", "message": "User authentication successful", "user_id": 12345, "ip": "192.168.1.100"}
{"timestamp": "2025-01-15T10:30:46.234567Z", "level": "WARNING", "message": "Rate limit exceeded", "user_id": 12345, "requests_per_minute": 150}
{"timestamp": "2025-01-15T10:30:47.345678Z", "level": "ERROR", "message": "Database connection failed", "error_code": "ECONNREFUSED", "retry_count": 3}
{"timestamp": "2025-01-15T10:30:48.456789Z", "level": "INFO", "message": "User profile updated", "user": {"id": 12345, "name": "John Doe", "preferences": {"theme": "dark", "notifications": true}}}

What happens: Each log entry is serialized as a single JSON object, making it easy to parse and analyze with log aggregation tools.

Per-Sink JSON Configuration¶

You can also enable JSON logging for individual sinks using the json=True parameter:

from logly import logger

# Regular text console output
logger.add("console")

# JSON file output (structured logging)
logger.add("logs/app.jsonl", json=True)

# JSON console output for debugging
logger.add("console", json=True, filter_min_level="DEBUG")

# Log some events
logger.info("Application started", version="1.2.3", environment="production")
logger.debug("Database connection established", pool_size=10, timeout=30)
logger.warning("High memory usage detected", memory_mb=850, threshold_mb=800)

Pretty JSON for Development¶

For development and debugging, you can enable pretty printing:

logger.configure(level="INFO", json=True, pretty_json=True)
logger.add("console")

logger.info("Server started", port=8080, workers=4)

Expected Output:

{
  "timestamp": "2025-01-15T10:30:45.123456Z",
  "level": "INFO",
  "message": "Server started",
  "port": 8080,
  "workers": 4
}

What happens: Pretty JSON formatting makes logs easier to read during development, with proper indentation and line breaks.

Integration Examples¶

ELK Stack (Elasticsearch, Logstash, Kibana)¶

logger.configure(level="INFO", json=True)
logger.add("/var/log/app.jsonl")

CloudWatch Logs¶

logger.configure(level="INFO", json=True)
logger.add("console", colorize=False)

Key Features Demonstrated¶

Structured logging: Consistent JSON format for all logs
Extra fields: Automatically include additional data in JSON
Machine readable: Perfect for log aggregation and analysis
Flexible formatting: Customize JSON structure as needed