PyTraceKit

PyTraceKit, is a low code solution designed to simplify the integration of distributed tracing into Python microservices through the power of OpenTelemetry. This package provides straightforward approach to enrich your logging with trace IDs and allowing detailed tracing with minimal setup.

The package comes with inbuilt support for Falcon, FastAPI , Flask ,redis , MongoDB

Distributed Tracing Design

Pytracekit provides the funcitonality to send the trace to the trace collector. From there on you could employ a tempo , grafana as shown in the diagram or you could use a observality tool of your choice.

Soon we plan to add details of how to setup otel collector ,tempo and grafana.

Key Features

• Simplified Tracing: Easily integrate distributed tracing into FastAPI, Flask, or Falcon applications with just one line of code.
• Function-Level Tracing with Customization: Add detailed tracing to any function, specifying custom operation names and span attributes for in-depth observability.
• Advanced Logging: Each log message includes a trace ID for easy correlation, with customizable logging settings such as file paths, sizes, and rotation.
• Framework Agnostic: Automatically detects and integrates with your Python web framework, offering out-of-the-box support for FastAPI, Flask, and Falcon.
• Extensibility: Adapt and extend PyTraceKit to suit your specific tracing and logging needs.

Installation

To install PyTraceKit, run the following command in your terminal:

pip install pytracekit

Getting Started

Setup the environment variables (These environment varaibles are mandatory)

export TRACE_SERVICE_NAME=your service name
export TRACING_ENDPOINT=http://localhost:4317/api/traces # your collector endpoint
export ENABLE_TRACING=true

For FastAPI or Flask:

from pytracekit import instrument

# Assuming `app` is your FastAPI or Flask application instance
instrument(app)

For Falcon

from pytracekit import instrument

# Instrument before defining your Falcon app
instrument()

app = falcon.App()  # Now define your Falcon app

Add Function-Level Tracing

PyTraceKit allows you to add tracing to specific functions using the @instrument_with_tracing decorator. You can specify a custom operation name and additional span attributes for detailed tracing:

Simple Function Level Tracing

from pytracekit import instrument_with_tracing

@instrument_with_tracing()
def process_data():
    # Your function logic here

Function level tracing with custom operation name and custom span attributes

from pytracekit import instrument_with_tracing

@instrument_with_tracing(operation_name="custom_operation", span_attrs={"key": "value"})
def process_data():
    # Your function logic here

Setup Enhanced Logging

To take advantage of PyTraceKit's enhanced logging capabilities, use the get_logger function. This allows you to create a logger that includes trace IDs in log messages, making it easier to correlate logs with traces:

Standard Logger

   from pytracekit import get_logger

    logger = get_logger()
    # Now you can use `logger` to log messages that include trace IDs
    logger.info("This log entry includes a trace ID.")

Logger with User Arguments

from pytracekit import get_logger

logger = get_logger(name='my_service_logger', log_file_path='logs/my_service.log', max_bytes=10*1024*1024, backup_count=5)
# Now you can use `logger` to log messages that include trace IDs
logger.info("This will push the logs into logs/my_service.log")