multiprocessing logger deadlocks if used with logging.handlers.QueueHandler

[iamdbychkov]

BPO	46163
Nosy	@iamdbychkov
Files	logger_deadlock.py

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = None
created_at = <Date 2021-12-23.13:12:46.077>
labels = ['3.9']
title = 'multiprocessing logger deadlocks if used with logging.handlers.QueueHandler'
updated_at = <Date 2021-12-23.13:12:46.077>
user = 'https://github.com/iamdbychkov'

bugs.python.org fields:

activity = <Date 2021-12-23.13:12:46.077>
actor = 'iamdbychkov'
assignee = 'none'
closed = False
closed_date = None
closer = None
components = []
creation = <Date 2021-12-23.13:12:46.077>
creator = 'iamdbychkov'
dependencies = []
files = ['50513']
hgrepos = []
issue_num = 46163
keywords = []
message_count = 1.0
messages = ['409088']
nosy_count = 1.0
nosy_names = ['iamdbychkov']
pr_nums = []
priority = 'normal'
resolution = None
stage = None
status = 'open'
superseder = None
type = None
url = 'https://bugs.python.org/issue46163'
versions = ['Python 3.9']

[iamdbychkov]

Hello!
If you use multiprocessing logger with logging.handlers.QueueHandler it will deadlock on first attempt at logging.

First attempt will initialize queue's background thread, which uses multiprocessing logger itself which will result in something what looks like deadlock :)

There are workarounds and I'm not sure if it can be classified as a bug, prehaps a little note in documentation about this behaviour will do?

I've attached example script.

[vsajip]

Should now be fixed along with #91555 via #131812.

[vsajip]

Reopened because the change in #131812 had to be reverted.

[vsajip]

Closing, now that a warning has been added to the documentation.