CODEWITHSUNDEEP
pythonpython-requestspentahogevent
user3871
user3871

Reputation: 12718

"RecursionError: maximum recursion depth exceeded" from ssl.py: `super(SSLContext, SSLContext).options.__set__(self, value)`

I am using Python 3.6.5 on the following remote server setup:

Server: Windows 10

Python: 3.6.5

Requests: 2.18.4

Pentaho: 8.0

When I run request.get against URLs in the server's command prompt, it gets the JSON as expected:

>>> import requests
>>> response = requests.get(url, headers=headers)
>>> json = response.json()
>>> print(json)
{'d': {'results': [{'_ ...

However when I run the same script in CPython for Pentaho 8.0, I get

RecursionError: maximum recursion depth exceeded

Full log:

2018/04/13 15:02:17 - Get SP Doc List.0 - ERROR (version 8.0.0.0-28, build 8.0.0.0-28 from 2017-11-05 07.27.50 by buildguy) : Unexpected error
    2018/04/13 15:02:17 - Get SP Doc List.0 - ERROR (version 8.0.0.0-28, build 8.0.0.0-28 from 2017-11-05 07.27.50 by buildguy) : org.pentaho.di.core.exception.KettleException: 
    2018/04/13 15:02:17 - Get SP Doc List.0 - Traceback (most recent call last):
      File "C:\Users\ADMINI~1\AppData\Local\Temp\2\pyServer.py", line 299, in execute_script
        exec (script, _global_env)
      File "<string>", line 16, in <module>
      File "C:\Program Files\Python36\lib\site-packages\requests\api.py", line 72, in get
        return request('get', url, params=params, **kwargs)
      File "C:\Program Files\Python36\lib\site-packages\requests\api.py", line 58, in request
        return session.request(method=method, url=url, **kwargs)
      File "C:\Program Files\Python36\lib\site-packages\requests\sessions.py", line 508, in request
        resp = self.send(prep, **send_kwargs)
      File "C:\Program Files\Python36\lib\site-packages\requests\sessions.py", line 618, in send
        r = adapter.send(request, **kwargs)
      File "C:\Program Files\Python36\lib\site-packages\requests\adapters.py", line 440, in send
        timeout=timeout
      File "C:\Program Files\Python36\lib\site-packages\urllib3\connectionpool.py", line 601, in urlopen
        chunked=chunked)
      File "C:\Program Files\Python36\lib\site-packages\urllib3\connectionpool.py", line 346, in _make_request
        self._validate_conn(conn)
      File "C:\Program Files\Python36\lib\site-packages\urllib3\connectionpool.py", line 850, in _validate_conn
        conn.connect()
      File "C:\Program Files\Python36\lib\site-packages\urllib3\connection.py", line 314, in connect
        cert_reqs=resolve_cert_reqs(self.cert_reqs),
      File "C:\Program Files\Python36\lib\site-packages\urllib3\util\ssl_.py", line 269, in create_urllib3_context
        context.options |= options
      File "C:\Program Files\Python36\lib\ssl.py", line 465, in options
        super(SSLContext, SSLContext).options.__set__(self, value)
      File "C:\Program Files\Python36\lib\ssl.py", line 465, in options
        super(SSLContext, SSLContext).options.__set__(self, value)
      File "C:\Program Files\Python36\lib\ssl.py", line 465, in options
        super(SSLContext, SSLContext).options.__set__(self, value)
      [Previous line repeated 322 more times]
    RecursionError: maximum recursion depth exceeded

Script:

import requests
import json


# By Filename
url = "https://myco.sharepoint.com/teams/dg/l/_api/web/lists/GetByTitle('eRetail%20Data%20Sources')/items?..."

authtoken = "Bearer eyJ..."

headers = {
    "Content-Type": "application/json;odata=verbose",
    "Accept": "application/json;odata=verbose",
    "Authorization": authtoken
}

response = requests.get(url, headers=headers)

json = response.json()
print('===========================')
print(json)

Upvotes: 30

Views: 24122

Answers (7)

Mu Mind
Mu Mind

Reputation: 11214

Also check for truststore usage nowadays, which also patches SSL-handling code and can also trigger these errors.

Upvotes: 0

chrisinmtown
chrisinmtown

Reputation: 4364

This max-recursion error arose in the context of a Flask app when I upgraded the requests library to version 2.32.2. The app seemed to work fine but my automated tests started failing. I took the suggestion offered above and extended my project's conftest.py file with the gevent import/patch_all() lines. Then the problem no longer occurred when running tox. Hope this helps someone out there.

Upvotes: 1

Roland
Roland

Reputation: 304

For my case, I use reqto, the solution was to put a try: except: block around it:

    peers = None

    try:
        res = reqto.get(f"https://{domain}/api/v1/instance/peers", headers=headers, timeout=5)
        peers = res.json()

        if not res.ok or res.status_code >= 400:
            print("WARNING: Cannot fetch peers:", domain)
            update_last_error(domain, res)

    except:
        print("WARNING: Some error during get():", domain)

If this doesn't work for you, you might want to cleanup your ~/.local/lib/python3/site-packages/ directory. Some old remaining files have caused a freezer and that no file was fetched from remote. Caused me a lot of hours to discover it!

Nope, not working! Maybe just pure coincidence!

For my script, I had to import reqto in fba.commands module while it was not referenced there. I mad a separate branch so you can find it out yourself: https://git.mxchange.org/?p=fba.git;a=shortlog;h=refs/heads/broken/missing-import-reqto Just put that said import line in the said file' header where the other import line exist and that code will work!

Upvotes: 0

Vladimir
Vladimir

Reputation: 145

Do not monkey-patch ssl:

from gevent import monkey
monkey.patch_all(ssl=False)

Upvotes: 5

manna
manna

Reputation: 11

from gevent import monkey as curious_george
# curious_george.patch_all(thread=False, select=False)
def stub(*args, **kwargs):  # pylint: disable=unused-argument
    pass

Upvotes: 1

luckydonald
luckydonald

Reputation: 6886

If gevent is installed, it needs to monkey-patch the python sockets to cooperate (see documentation or this github issue).

Therefore gevent.monkey.patch_all() is either missing or not called early enough.

# at the beginning of the script
import gevent.monkey
gevent.monkey.patch_all()

# all the other stuff below, like for example
import requests

Upvotes: 34

Rex Wang
Rex Wang

Reputation: 458

Update:

I figured out that my issue is because I introduced locust in another place of my application, which uses gevent==1.2.2. After I commented out the gevent import statements, this recursion issue was gone. It is worth checking if you have introduced gevent somewhere in your app.

I ran into the same error recently, and haven't found any way to solve this. I had to use http instead of https in my situation, although it's really dangerous.

Upvotes: 3

Related Questions