An API Logger for your Django Rest Framework project.
It logs all the API information for content type "application/json".
- URL
- Request Body
- Request Headers
- Request Method
- API Response
- Status Code
- API Call Time
- Server Execution Time
- Client IP Address
You can log API information into the database or listen to the logger signals for different use-cases, or you can do both.
- The logger usage a separate thread to run, so it won't affect your API response time.
Install or add drf-api-logger.
pip install drf-api-logger
Add in INSTALLED_APPS
INSTALLED_APPS = [
'django.contrib.admin',
'django.contrib.auth',
'django.contrib.contenttypes',
'django.contrib.sessions',
'django.contrib.messages',
'django.contrib.staticfiles',
'drf_api_logger', # Add here
]
Add in MIDDLEWARE
MIDDLEWARE = [
'django.middleware.security.SecurityMiddleware',
'django.contrib.sessions.middleware.SessionMiddleware',
'django.middleware.common.CommonMiddleware',
'django.middleware.csrf.CsrfViewMiddleware',
'django.contrib.auth.middleware.AuthenticationMiddleware',
'django.contrib.messages.middleware.MessageMiddleware',
'django.middleware.clickjacking.XFrameOptionsMiddleware',
'drf_api_logger.middleware.api_logger_middleware.APILoggerMiddleware', # Add here
]
Log every request into the database.
DRF_API_LOGGER_DATABASE = True # Default to False
-
Logs will be available in Django Admin Panel.
-
The search bar will search in Request Body, Response, Headers and API URL.
-
You can also filter the logs based on the "added_on" date, Status Code and Request Methods.
Note: Make sure to migrate. It will create a table for logger if "DRF_API_LOGGER_DATABASE" is True else if already exists, it will delete the table.
Listen to the signal as soon as any API is called. So you can log the API data into a file or for different use-cases.
DRF_API_LOGGER_SIGNAL = True # Default to False
Example code to listen to the API Logger Signal.
"""
Import API_LOGGER_SIGNAL
"""
from drf_api_logger import API_LOGGER_SIGNAL
"""
Create a function that is going to listen to the API logger signals.
"""
def listener_one(**kwargs):
print(kwargs)
def listener_two(**kwargs):
print(kwargs)
"""
It will listen to all the API logs whenever an API is called.
You can also listen signals in multiple functions.
"""
API_LOGGER_SIGNAL.listen += listener_one
API_LOGGER_SIGNAL.listen += listener_two
"""
Unsubscribe to signals.
"""
API_LOGGER_SIGNAL.listen -= listener_one
DRF API Logger usage queue to hold the logs before inserting into the database. Once queue is full, it bulk inserts into the database.
Specify the queue size.
DRF_LOGGER_QUEUE_MAX_SIZE = 50 # Default to 50 if not specified.
DRF API Logger also waits for a period of time. If queue is not full and there are some logs to be inserted, it inserts after interval ends.
Specify an interval (In Seconds).
DRF_LOGGER_INTERVAL = 10 # In Seconds, Default to 10 seconds if not specified.
Note: The API call time (added_on) is a timezone aware datetime object. It is actual time of API call irrespective of interval value or queue size.
You can skip the entire app to be logged into the database by specifying namespace of the app as list.
DRF_API_LOGGER_SKIP_NAMESPACE = ['APP_NAMESPACE1', 'APP_NAMESPACE2']
You can also skip any API to be logged by using url_name of the API.
DRF_API_LOGGER_SKIP_URL_NAME = ['url_name1', 'url_name2']
Note: It does not log Django Admin Panel API calls.
You may wish to hide sensitive information from being exposed in the logs.
You do this by setting DRF_API_LOGGER_EXCLUDE_KEYS
in settings.py to a list of your desired sensitive keys.
The default is
DRF_API_LOGGER_EXCLUDE_KEYS = ['password', 'token', 'access', 'refresh']
# Sensitive data will be replaced with "***FILTERED***".
DRF_API_LOGGER_DEFAULT_DATABASE = 'default' # Default to "default" if not specified
"""
Make sure to migrate the database specified in DRF_API_LOGGER_DEFAULT_DATABASE.
"""
You can also identify slow APIs by specifying DRF_API_LOGGER_SLOW_API_ABOVE
in settings.py.
A new filter (By API Performance) will be visible, and you can choose slow or fast API.
DRF_API_LOGGER_SLOW_API_ABOVE = 200 # Default to None
# Specify in milli-seconds.
You can log only selected methods by specifying DRF_API_LOGGER_METHODS
in settings.py.
DRF_API_LOGGER_METHODS = ['GET', 'POST', 'DELETE', 'PUT'] # Default to empty list (Log all the requests).
You can log only selected responses by specifying DRF_API_LOGGER_STATUS_CODES
in settings.py.
DRF_API_LOGGER_STATUS_CODES = ['200', '400', '404', '500'] # Default to empty list (Log all responses).
You can also change the timezone by specifying DRF_API_LOGGER_TIMEDELTA
in settings.py.
It won't change the Database timezone. It will still remain UTC or the timezone you have defined.
DRF_API_LOGGER_TIMEDELTA = 330 # UTC + 330 Minutes = IST (5:Hours, 30:Minutes ahead from UTC)
# Specify in minutes.
# Yoc can specify negative values for the countries behind the UTC timezone.
DRF_API_LOGGER_TIMEDELTA = -30 # Example
You can specify an endpoint of API should have absolute URI or not by setting this variable in DRF settings.py file.
DRF_API_LOGGER_PATH_TYPE = 'ABSOLUTE' # Default to ABSOLUTE if not specified
# Possible values are ABSOLUTE, FULL_PATH or RAW_URI
Considering we are accessing the following URL: http://127.0.0.1:8000/api/v1/?page=123 DRF_API_LOGGER_PATH_TYPE possible values are:
-
ABSOLUTE (Default) :
Function used
request.build_absolute_uri()
Output:
http://127.0.0.1:8000/api/v1/?page=123
-
FULL_PATH
Function used
request.get_full_path()
Output:
/api/v1/?page=123
-
RAW_URI
Function used
request.get_raw_uri()
Output:
http://127.0.0.1:8000/api/v1/?page=123
Note: Similar to ABSOLUTE but skip allowed hosts protection, so may return an insecure URI.
You can use the DRF API Logger Model to query some information.
Note: Make sure to set "DRF_API_LOGGER_DATABASE = True" in settings.py file.
from drf_api_logger.models import APILogsModel
"""
Example:
Select records for status_code 200.
"""
result_for_200_status_code = APILogsModel.objects.filter(status_code=200)
DRF API Logger Model:
class APILogsModel(Model):
id = models.BigAutoField(primary_key=True)
api = models.CharField(max_length=1024, help_text='API URL')
headers = models.TextField()
body = models.TextField()
method = models.CharField(max_length=10, db_index=True)
client_ip_address = models.CharField(max_length=50)
response = models.TextField()
status_code = models.PositiveSmallIntegerField(help_text='Response status code', db_index=True)
execution_time = models.DecimalField(decimal_places=5, max_digits=8,
help_text='Server execution time (Not complete response time.)')
added_on = models.DateTimeField()
def __str__(self):
return self.api
class Meta:
db_table = 'drf_api_logs'
verbose_name = 'API Log'
verbose_name_plural = 'API Logs'
After sometime, there will be too many data in the database. Searching and filter may get slower. If you want, you can delete or archive the older data. To improve the searching or filtering, try to add indexes in the 'drf_api_logs' table.