Transient test_multiprocessing failure (test_active_children)

[pitrou]

BPO	6661
Nosy	@birkenfeld, @pitrou, @skrah, @sandrotosi

^{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 = <Date 2013-03-28.15:25:20.531>
created_at = <Date 2009-08-06.20:10:47.130>
labels = ['tests', 'type-bug', 'library']
title = 'Transient test_multiprocessing failure (test_active_children)'
updated_at = <Date 2013-03-28.15:25:20.529>
user = 'https://github.com/pitrou'

bugs.python.org fields:

activity = <Date 2013-03-28.15:25:20.529>
actor = 'georg.brandl'
assignee = 'jnoller'
closed = True
closed_date = <Date 2013-03-28.15:25:20.531>
closer = 'georg.brandl'
components = ['Library (Lib)', 'Tests']
creation = <Date 2009-08-06.20:10:47.130>
creator = 'pitrou'
dependencies = []
files = []
hgrepos = []
issue_num = 6661
keywords = []
message_count = 8.0
messages = ['91383', '91384', '91385', '91386', '119810', '119812', '119813', '185452']
nosy_count = 5.0
nosy_names = ['georg.brandl', 'pitrou', 'jnoller', 'skrah', 'sandro.tosi']
pr_nums = []
priority = 'normal'
resolution = 'wont fix'
stage = None
status = 'closed'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue6661'
versions = ['Python 2.7']

[pitrou]

I just got the following test_multiprocessing error (cannot reproduce,
though):

test test_multiprocessing failed -- Traceback (most recent call last):
  File
"/home/antoine/cpython/seek-6629/Lib/test/test_multiprocessing.py", line
232, in test_active_children
    self.assertTrue(p in self.active_children())
AssertionError: False is not True

[jnoller]

ugh. that's not useful. Dunno how I'm going to track down seemingly random
and unreproducible issues.

[pitrou]

I know, I was posting it for completeness, and it case someone else
witnesses it.

[jnoller]

I know, just makes me :(

[sandrotosi]

Hello,
what can we do with this issue? It doesn't seem it happened again after Antoine's report (or at least no-one had followed it up) and it doesn't happen nowdays (f.e. I exec'd test_multiprocessing in a loop for some time, and no error comes).

What do you think about closing it as "sorry, unable to replicate it" and so mark it a glitch somehow?

[skrah]

I've seen that one as well, but very rarely.

[sandrotosi]

well, at least we have a confirmation it's still there :)

any idea how we can try to debug it?

[birkenfeld]

2.5 years later I don't think it's much use keeping this open.