Description
Describe how you confirmed the issue is with the library, and not with the API itself, or a server-side issue of some other kind.
This issue appears during the client management of a 429 response.
Is is not reproducible with clients other that the library.
Debugging the code line my line brings to the request function on the RestSession class (file meraki/rest_session.py)
Python version installed
3.12.6
Meraki library version installed
1.5.0
Have you reproduced the issue with the latest version of this library? And with the latest version of Python?
No but the code for managing the 429 responses is the same across versions
OS Platform
Building in W11 but typically running in python:3.12-slim-bullseye (debian:11-slim) container
Describe the bug
The API Client waits the amount of seconds specified in the Retry-After header after a 429 response from the Meraki API even if the wait_on_rate_limit variable is set to False.
While it is normal for a client to respect the Retry-After header, it would make sense for the API Client to respect the configuration it was given to it.
How can we replicate the problem you're reporting?
Requesting any API endpoint while being rate limited by the Meraki API will make your script run the part of the code that should manage this type of responses and it will ignore the wait_on_rate_limit variable.
Expected behavior
If the wait_on_rate_limit variable is set to False it should ignore the Retry-After header value and the nginx_429_retry_wait_time variable and immediatly try again or raise the APIError exception.
Code snippets
This is the code in the request function of the RestSession class that I think it should be modified to check the value of the wait_on_rate_limit variable.
# Rate limit 429 errors
elif status == 429:
if 'Retry-After' in response.headers:
wait = int(response.headers['Retry-After'])
else:
wait = random.randint(1, self._nginx_429_retry_wait_time)
if self._logger:
self._logger.warning(f'{tag}, {operation} - {status} {reason}, retrying in {wait} seconds')
time.sleep(wait)
retries -= 1
if retries == 0:
raise APIError(metadata, response)
I would simply add if wait_on_rate_limit: after the elif status == 429: line and skip directly to retries -= 1 if False:
# Rate limit 429 errors
elif status == 429:
if wait_on_rate_limit:
if 'Retry-After' in response.headers:
wait = int(response.headers['Retry-After'])
else:
wait = random.randint(1, self._nginx_429_retry_wait_time)
if self._logger:
self._logger.warning(f'{tag}, {operation} - {status} {reason}, retrying in {wait} seconds')
time.sleep(wait)
retries -= 1
if retries == 0:
raise APIError(metadata, response)
Possibly other pieces of the code that I don't know needs to be changed too.
Additional context
The issue is being discussed in the Meraki Forum too: Meraki Python SDK Library wait_on_rate_limit=False but it waits anyway