Skip to content
Pydantic Docs

Structlog

Add Logfire’s structlog processor and your existing structlog output shows up in Logfire as structured logs (individual timestamped records of something that happened), next to the traces (the full journey of one request through your app) from the rest of your code. Every event you already log, with all its structured fields, becomes a Logfire log.

What you’ll capture

  • Each structlog event message
  • Its severity level (info, warning, error, and so on)
  • Its timestamp
  • The structured key/value fields you attach to the event

Before you start

You’ll need a Logfire project. Open Add data in your project (top navigation) and follow the setup for your language: it signs your machine in with logfire auth (a browser sign-in, no token to copy) and, for production or other languages, creates a write token (the credential your app uses to send data). New to Logfire? Start with Getting Started.

Installation

Install logfire:

Terminal
pip install logfire

This works with your existing structlog package: nothing extra to install. If you don’t have it yet, run pip install structlog in your terminal.

Usage

Call logfire.configure() to connect to your project, then add logfire.StructlogProcessor() to your structlog processor chain so every event is also sent to Logfire.

main.py
from dataclasses import dataclass

import structlog

import logfire

logfire.configure()

structlog.configure(
    processors=[
        structlog.contextvars.merge_contextvars,
        structlog.processors.add_log_level,
        structlog.processors.StackInfoRenderer(),
        structlog.dev.set_exc_info,
        structlog.processors.TimeStamper(fmt='%Y-%m-%d %H:%M:%S', utc=False),
        logfire.StructlogProcessor(),
        structlog.dev.ConsoleRenderer(),
    ],
)
logger = structlog.get_logger()


@dataclass
class User:
    id: int
    name: str


logger.info('Login', user=User(id=42, name='Fred'))
#> 2024-03-22 12:57:33 [info     ] Login                          user=User(id=42, name='Fred')

The Logfire processor MUST come before the last processor that renders the logs in the structlog configuration.

Verify it worked

Run your program, then open the Live view in the Logfire web app. Within a few seconds you’ll see your event as a record, with its fields as attributes.

Advanced

Console logging

By default, LogfireProcessor shown above disables console logging by Logfire so you can use the existing logger you have configured for structlog. If you want to log with Logfire’s console output as well, use LogfireProcessor(console_log=True).

Positional arguments

Troubleshooting

Not seeing your events in Logfire? Check that logfire.configure() ran first, your write token is set, and logfire.StructlogProcessor() is in your processor chain before the final rendering processor.

Reference