The Three Pillars of Observability

1. Introduction

Observability is crucial for understanding system behavior, performance, and reliability. The three pillars of observability are:

• Metrics
• Logs
• Traces

2. Metrics

Metrics are numerical values that represent the state of a system. They provide quantifiable data that helps track performance over time.

Key Metrics Examples

• CPU Usage
• Memory Usage
• Request Latency

To collect metrics, you can use monitoring tools like Prometheus. Here’s a simple example:

import prometheus_client

# Create a metric to track request latency
latency_metric = prometheus_client.Histogram('request_latency_seconds', 'Request latency in seconds')

@latency_metric.time()
def handle_request():
    # Handle request logic
    pass
        

3. Logs

Logs provide a detailed record of events that happen within a system. They are invaluable for debugging and understanding system behavior.

Log Levels

• INFO
• DEBUG
• ERROR

Here's an example of logging in Python:

import logging

logging.basicConfig(level=logging.INFO)

def do_something():
    logging.info("Doing something important")
    # More code...
        

4. Traces

Tracing allows you to follow the path of a request through various services, helping you understand the flow and performance bottlenecks.

Example of Tracing with OpenTracing

from opentracing import tracer

def some_service():
    with tracer.start_span('some_service') as span:
        # Service logic
        span.set_tag('service', 'example')
        span.log_event('event_name')
        pass
        # More code...
        

5. FAQ