You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Intermittent iOS buildbot failure in test_sock_client_racing #148793
Thanks - I had seen that; I wasn't overly concerned because (a) the failure-on-repeat rate is very low, and (b) from what I can make out, the test that is failing is a explicitly testing performance under extreme network load. Obviously it would be preferable to pass the test consistently, but it's also going to be difficult to reproduce (and I haven't had any luck to date reproducing it locally)
FWIW - there are at least two failure modes that I've seen. They both occur on the sock_send_racing portion of the test. The sock_recv and sock_recv_into portions of the test seem to be stable. It does make me wonder if the problem is some sort of time-based reset after the first sequence of the test.
Timeout on the second socket test:
Traceback (most recent call last):
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/5F33C29B-26A9-433F-8EF2-24292B3B9DD8/iOSTestbed.app/python/lib/python3.15/asyncio/tasks.py", line 488, in wait_for
return await fut
^^^^^^^^^
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/5F33C29B-26A9-433F-8EF2-24292B3B9DD8/iOSTestbed.app/python/lib/python3.15/test/test_asyncio/test_sock_lowlevel.py", line 216, in _basetest_sock_send_racing
await self.loop.sock_sendall(sock, b'world')
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/5F33C29B-26A9-433F-8EF2-24292B3B9DD8/iOSTestbed.app/python/lib/python3.15/asyncio/selector_events.py", line 562, in sock_sendall
return await fut
^^^^^^^^^
asyncio.exceptions.CancelledError
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/5F33C29B-26A9-433F-8EF2-24292B3B9DD8/iOSTestbed.app/python/lib/python3.15/test/test_asyncio/test_sock_lowlevel.py", line 273, in test_sock_client_racing
self.loop.run_until_complete(asyncio.wait_for(
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^
self._basetest_sock_send_racing(listener, sock), 10))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/5F33C29B-26A9-433F-8EF2-24292B3B9DD8/iOSTestbed.app/python/lib/python3.15/asyncio/base_events.py", line 724, in run_until_complete
return future.result()
~~~~~~~~~~~~~^^
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/5F33C29B-26A9-433F-8EF2-24292B3B9DD8/iOSTestbed.app/python/lib/python3.15/asyncio/tasks.py", line 487, in wait_for
async with timeouts.timeout(timeout):
~~~~~~~~~~~~~~~~^^^^^^^^^
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/5F33C29B-26A9-433F-8EF2-24292B3B9DD8/iOSTestbed.app/python/lib/python3.15/asyncio/timeouts.py", line 114, in __aexit__
raise TimeoutError from exc_val
TimeoutError
A connection reset error
Traceback (most recent call last):
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/3DAA3568-C09B-454E-9E2B-4749EEB89567/iOSTestbed.app/python/lib/python3.15/test/test_asyncio/test_sock_lowlevel.py", line 273, in test_sock_client_racing
self.loop.run_until_complete(asyncio.wait_for(
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^
self._basetest_sock_send_racing(listener, sock), 10))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/3DAA3568-C09B-454E-9E2B-4749EEB89567/iOSTestbed.app/python/lib/python3.15/asyncio/base_events.py", line 724, in run_until_complete
return future.result()
~~~~~~~~~~~~~^^
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/3DAA3568-C09B-454E-9E2B-4749EEB89567/iOSTestbed.app/python/lib/python3.15/asyncio/tasks.py", line 488, in wait_for
return await fut
^^^^^^^^^
File "/Users/buildbot/Library/Developer/CoreSimulator/Devices/4F9AFDE7-A49A-46AD-B161-FCA22E48EA55/data/Containers/Bundle/Application/3DAA3568-C09B-454E-9E2B-4749EEB89567/iOSTestbed.app/python/lib/python3.15/test/test_asyncio/test_sock_lowlevel.py", line 196, in _basetest_sock_send_racing
sock.send(b' ' * size)
~~~~~~~~~^^^^^^^^^^^^^
ConnectionResetError: [Errno 54] Connection reset by peer
Activity
freakboy3742 commentedon Apr 20, 2026
Thanks - I had seen that; I wasn't overly concerned because (a) the failure-on-repeat rate is very low, and (b) from what I can make out, the test that is failing is a explicitly testing performance under extreme network load. Obviously it would be preferable to pass the test consistently, but it's also going to be difficult to reproduce (and I haven't had any luck to date reproducing it locally)
FWIW - there are at least two failure modes that I've seen. They both occur on the
sock_send_racingportion of the test. Thesock_recvandsock_recv_intoportions of the test seem to be stable. It does make me wonder if the problem is some sort of time-based reset after the first sequence of the test.mhsmith commentedon Apr 21, 2026
Also, the GitHub buildbot comments aren't including anything useful. This might be because every line in the test log has an extra space at the start.
kumaraditya303 commentedon Jul 19, 2026
What's the status of this?
freakboy3742 commentedon Jul 19, 2026
It's still happening - but it's a lot less frequent than it was. Looking at the last 200 builds on Buildbot, there are:
test_sock_client_racingfailures).