[issue37721] smtpd: staling connect

cherusk report at bugs.python.org
Tue Jul 30 11:41:30 EDT 2019


New submission from cherusk <matthias.tafelmeier at gmx.net>:

>From PR:

Following script stals/hangs. According to my tracing done, in the smtpd component. The concept of the script is baseline of my unit testing for [1], so not hypothetical:

import logging
import sys
logging.basicConfig(level=logging.DEBUG, stream=sys.stdout)

from smtpd import DebuggingServer as TestMailServer
import threading
from notifiers import get_notifier
import time
import itertools as it

mail_server = threading.Thread(target=TestMailServer, args=(("localhost", 2500), None))
mail_server.daemon = True
mail_server.start()

time.sleep(2)


for msg in it.repeat('copy', 100):
    print(msg)
    get_notifier('email').notify(from_='cherusk at localhost',
                                 to='root at localhost',
                                 message=msg,
                                 subject='TST',
                                 host='localhost',
                                 port=2500,
                                 tls=False,
                                 ssl=False,
                                 html=False,
                                 username="some",
                                 password='pw'
                                 )

I evaluated that on several platforms(FED29, UBUNTU-Latest), same symptoms. Also, on network stack I saw a clean TCP handshake happening, also cleanly ACKed by the smptd. So, presumption is that it's something in the dispatcher.

[1] https://github.com/cherusk/tw_timeliness/blob/master/test/time_engine_test.py

----------
components: email
messages: 348752
nosy: barry, cherusk, r.david.murray
priority: normal
pull_requests: 14785
severity: normal
status: open
title: smtpd: staling connect
type: behavior
versions: Python 3.7

_______________________________________
Python tracker <report at bugs.python.org>
<https://bugs.python.org/issue37721>
_______________________________________


More information about the Python-bugs-list mailing list