nr-log-handler 0.1.0

Send logs to New Relic via their Log REST API

nr-log-handler

A lightweight Python library for sending logs to New Relic via their Log API. No New Relic agent required — pure REST API.

Features

• Works with Python's built-in logging module as a drop-in logging.Handler
• Standalone NewRelicLogger with .info(), .error() etc.
• US and EU region support
• Sync (immediate) and async (batched, background thread) modes
• Configurable batch size and flush interval
• Exponential backoff retry — never crashes your application
• Global and per-call custom attributes

Requirements

• Python 3.10+
• requests

Installation

pip install nr-log-handler

Quick Start

Standalone logger

from newrelic_logger import NewRelicLogger

logger = NewRelicLogger(
    api_key="YOUR_NEW_RELIC_LICENSE_KEY",
    region="us",               # "us" (default) or "eu"
    mode="async",              # "sync" (default) or "async"
    attributes={"service": "my-app", "environment": "production"},
)

logger.info("Application started")
logger.error("Something went wrong", extra_attributes={"request_id": "abc-123"})

# Always call close() on shutdown to flush buffered logs
logger.close()

As a logging.Handler

import logging
from newrelic_logger import NewRelicHandler

handler = NewRelicHandler(
    api_key="YOUR_NEW_RELIC_LICENSE_KEY",
    region="us",
    mode="sync",
    attributes={"service": "my-app"},
)

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger("myapp")
logger.addHandler(handler)

logger.info("Hello from standard logging")

Configuration

Parameter	Type	Default	Description
api_key	str	required	New Relic license key
region	str	"us"	"us" or "eu"
mode	str	"sync"	"sync" or "async"
batch_size	int	100	Max logs per batch (async only)
flush_interval	float	5.0	Seconds between flushes (async only)
timeout	int	10	HTTP timeout in seconds
max_retries	int	5	Max retry attempts on transient failure
backoff_factor	float	0.5	Exponential backoff multiplier
attributes	dict	None	Global attributes on every log entry

Error Handling

The library never raises exceptions from logging calls. On failure it retries up to max_retries times with exponential backoff, then emits a warning via Python's stdlib logging to the newrelic_logger logger and silently drops the batch.

To capture these internal warnings:

import logging
logging.getLogger("newrelic_logger").setLevel(logging.WARNING)

License

MIT