django-datadog-logger

Django DataDog Logger

Django DataDog Logger integration package.

• Free software: MIT license
• Documentation: https://django-datadog-logger.readthedocs.io.

Quick start

Set up request id tracking (in front) and logging middlewares (at the end):

MIDDLEWARE = [
    "django_datadog_logger.middleware.request_id.RequestIdMiddleware",
    # ...
    "django_datadog_logger.middleware.error_log.ErrorLoggingMiddleware",
    "django_datadog_logger.middleware.request_log.RequestLoggingMiddleware",
]

Configure LOGGERS in your Django settings file:

API_LOG_ROOT = env.str("API_LOG_ROOT")
LOGGING = {
    "version": 1,
    "disable_existing_loggers": False,
    "formatters": {
        "console": {"format": "{levelname} {message}", "style": "{"},
        "json": {"()": "django_datadog_logger.formatters.datadog.DataDogJSONFormatter"},
    },
    "handlers": {
        "console": {"level": "INFO", "class": "logging.StreamHandler", "formatter": "console"},
        "application": {
            "level": API_LOG_APPLICATION_LEVEL,
            "class": "logging.FileHandler",
            "filename": os.path.join(API_LOG_ROOT, "api.application.log"),
            "formatter": "json",
        },
        "state": {
            "level": API_LOG_STATE_LEVEL,
            "class": "logging.FileHandler",
            "filename": os.path.join(API_LOG_ROOT, "api.state.log"),
            "formatter": "json",
        },
        "request": {
            "level": API_LOG_REQUEST_LEVEL,
            "class": "logging.FileHandler",
            "filename": os.path.join(API_LOG_ROOT, "api.request.log"),
            "formatter": "json",
        },
        "session": {
            "level": API_LOG_SESSION_LEVEL,
            "class": "logging.FileHandler",
            "filename": os.path.join(API_LOG_ROOT, "api.session.log"),
            "formatter": "json",
        },
        "error": {
            "level": API_LOG_ERROR_LEVEL,
            "class": "logging.FileHandler",
            "filename": os.path.join(API_LOG_ROOT, "api.error.log"),
            "formatter": "json",
        },
    },
    "loggers": {
        "": {"handlers": ["console", "error"], "level": "DEBUG", "propagate": True},
        "ddtrace": {"handlers": ["error"], "level": "ERROR", "propagate": False},
        "django.db.backends": {"handlers": ["error"], "level": "ERROR", "propagate": False},
        "twilio": {"handlers": ["error"], "level": "ERROR", "propagate": False},
        "my_project": {"handlers": ["application"], "level": "INFO", "propagate": False},
        "my_project.throttling": {"handlers": ["application"], "level": "DEBUG", "propagate": False},
        "my_project.vehicles.viewsets.state": {"handlers": ["state"], "level": "INFO", "propagate": False},
        "my_project.accounts.session": {"handlers": ["session"], "level": "DEBUG", "propagate": False},
        "my_project.session": {"handlers": ["session"], "level": "DEBUG", "propagate": False},
        "django_auth_ldap": {"level": "DEBUG", "handlers": ["session"], "propagate": False},
        "django_datadog_logger.middleware.error_log": {"handlers": ["error"], "level": "INFO", "propagate": False},
        "django_datadog_logger.middleware.request_log": {"handlers": ["request"], "level": "INFO", "propagate": False},
        "django_datadog_logger.rest_framework": {"handlers": ["application"], "level": "INFO", "propagate": False},
    },
}

If you would like to whitelist your projects for passing extra arguments to the json log record, please set the following regular expression:

DJANGO_DATADOG_LOGGER_EXTRA_INCLUDE = r"^(django_datadog_logger|my_project)(|\..+)$"

Add Celery logger configuration and request_id tracking decorator to tasks:

import logging

from celery import Celery, shared_task
from celery.result import AsyncResult
from celery.signals import after_setup_logger, after_setup_task_logger
from django.conf import settings
from django_datadog_logger.celery import store_celery_request

logger = logging.getLogger(__name__)


@after_setup_logger.connect
def on_after_setup_logger(logger, *args, **kwargs):
    from django_datadog_logger.formatters.datadog import DataDogJSONFormatter

    if settings.API_LOG_CELERY_JSON:
        formatter = DataDogJSONFormatter()
        for handler in list(logger.handlers):
            handler.setFormatter(formatter)
            handler.setLevel(settings.API_LOG_CELERY_LEVEL)


@after_setup_task_logger.connect
def on_after_setup_task_logger(logger, *args, **kwargs):
    from django_datadog_logger.formatters.datadog import DataDogJSONFormatter

    if settings.API_LOG_CELERY_JSON:
        formatter = DataDogJSONFormatter()
        for handler in list(logger.handlers):
            handler.setFormatter(formatter)
            handler.setLevel(settings.API_LOG_CELERY_LEVEL)


app = Celery("my_project")

# Using a string here means the worker will not have to
# pickle the object when using Windows.
app.config_from_object("django.conf:settings", namespace="CELERY")
app.autodiscover_tasks(lambda: settings.INSTALLED_APPS)


@shared_task(bind=True)
@store_celery_request
def debug_task(self):
    print("Request: {0!r}".format(self.request))
    logger.critical("CRITICAL", extra={"level": "CRITICAL"})
    logger.error("ERROR", extra={"level": "ERROR"})
    logger.warning("WARNING", extra={"level": "WARNING"})
    logger.info("INFO", extra={"level": "INFO"})
    logger.debug("DEBUG", extra={"level": "DEBUG"})
    return 42

ddtrace

The ddtrace library has an option to inject tracing context data into log records: https://ddtrace.readthedocs.io/en/stable/advanced_usage.html#logs-injection

There is a helper to look for those attributes and add them automatically to the log entry created by this library.

# log.py

# Patch logging library to inject dd.* attributes on log records
import ddtrace
ddtrace.patch(logging=True)

# Configure logger with DataDogJSONFormatter
import logging
from django_datadog_logger.formatters.datadog import DataDogJSONFormatter

logger = logging.root

handler = logging.StreamHandler()
handler.formatter = DataDogJSONFormatter()
logger.addHandler(handler)
logger.setLevel(logging.INFO)


# Log a test message
logger.info("test")

$ DD_SERVICE=django DD_ENV=test DD_VERSION=1234 python log.py
{"message": "test", "logger.name": "root", "logger.thread_name": "MainThread", "logger.method_name": "<module>", "syslog.timestamp": "2021-08-23T18:26:10.391099+00:00", "syslog.severity": "INFO", "dd.version": "1234", "dd.env": "test", "dd.service": "django", "dd.trace_id": "0", "dd.span_id": "0"}

If you remove the call to datadog.patch(logging=True) you end up with:

$ python test.py
{"message": "test", "logger.name": "root", "logger.thread_name": "MainThread", "logger.method_name": "<module>", "syslog.timestamp": "2021-08-23T18:27:47.951461+00:00", "syslog.severity": "INFO"}

Credits

This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.

History

0.6.2 (2023-04-27)

• Fixed case where accessing request.auth may raise errors

0.6.1 (2023-04-27)

• Removed dependency on Celery package, fixed import error

0.6.0 (2023-04-27)

• Removed dependency on Celery package

0.5.6 (2023-01-17)

• Datadog formatter: avoid recursion loop when accessing WSGI request auth attribute

0.5.5 (2023-01-16)

• Improved support for request version reporting in http.request_version
• Add support for http.url_details.view_name

0.5.4 (2023-01-16)

• Added support for HTTP Accept header as http.accept

0.5.3 (2022-12-19)

• Added support for JWT cid claim

0.5.2 (2022-11-23)

• Fixed: don't let the logger throw a DisallowedHost error

0.5.1 (2022-08-09)

• Fixed: ActionLoginMixin class methods perform_create and perform_update call super(). Remove atomic transaction

0.5.0 (2021-10-20)

• Added support for Celery v5+

0.4.0 (2021-08-27)

• Enhancement: Updated formatting in README.rst #5
• Enhancement: Extract and add dd.* attributes from log record to log entry dict #6
• Fixed: KeyError because a dict appears where a list is expected #7

0.3.5 (2021-06-14)

• Prevent recursion when warnings are logged whilst accessing WSGI request.user

0.3.4 (2021-06-14)

• Fixed import error for future package

0.3.3 (2020-11-04)

• Added support for incoming HTTP X-Request-ID header values

0.3.2 (2020-04-24)

• Respect User.USERNAME_FIELD

0.3.1 (2020-04-24)

• Removed API_LOG_REQUEST_DURATION_WARN_SECONDS

0.3.0 (2020-04-15)

• Improved Celery task received messages logging.
• Removed RequestIdFilter (not needed anymore).

0.2.0 (2020-04-14)

• Added Celery request local.

0.1.0 (2020-02-17)

• First release on PyPI.