[FREEZE] ERROR at teardown of TestsProcessPoolForkserverExecutor.test_map_chunksize #328
Labels
Comments
|
Similar problems:
_____ ERROR at teardown of TestsProcessPoolLokyExecutor.test_killed_child ______
self = <tests.test_process_executor_loky.TestsProcessPoolLokyExecutor object at 0x7fd459298a60>
method = <bound method ExecutorTest.test_killed_child of <tests.test_process_executor_loky.TestsProcessPoolLokyExecutor object at 0x7fd459298a60>>
def teardown_method(self, method):
# Make sure executor is not broken if it should not be
executor = getattr(self, 'executor', None)
if executor is not None:
expect_broken_pool = hasattr(method, "broken_pool") # old pytest
for mark in getattr(method, "pytestmark", []):
if mark.name == "broken_pool":
expect_broken_pool = True
is_actually_broken = executor._flags.broken is not None
> assert is_actually_broken == expect_broken_pool
E AssertionError
executor = <loky.process_executor.ProcessPoolExecutor object at 0x7fd456a8efb0>
expect_broken_pool = True
is_actually_broken = False
mark = Mark(name='broken_pool', args=(), kwargs={})
method = <bound method ExecutorTest.test_killed_child of <tests.test_process_executor_loky.TestsProcessPoolLokyExecutor object at 0x7fd459298a60>>
self = <tests.test_process_executor_loky.TestsProcessPoolLokyExecutor object at 0x7fd459298a60> |
|
It seems that we might have a race condition with the executor manager thread that is responsible to flag the process pool executor as broken. Let me try to fix this |
|
Actually #328 (comment) is not the same problem as the traceback in the description:
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
(randomly) observed on the CI on master:
test_loky linux-py37https://dev.azure.com/joblib/loky/_build/results?buildId=1574&view=logs&j=690d157a-d5ad-54bf-a85d-89b5d05cf663&t=35eaaa5a-db59-5698-331c-8016de4f39a3
The text was updated successfully, but these errors were encountered: