wassname/ray
1
0
Fork 0
mirror of https://github.com/wassname/ray.git synced 2026-07-07 14:40:43 +08:00
Code Issues Packages Projects Releases Wiki Activity

Make more tests run on Windows (#8446)

Browse Source 
* Remove worker Wait() call due to SIGCHLD being ignored

* Port _pid_alive to Windows

* Show PID as well as TID in glog

* Update TensorFlow version for Python 3.8 on Windows

* Handle missing Pillow on Windows

* Work around dm-tree PermissionError on Windows

* Fix some lint errors on Windows with Python 3.8

* Simplify torch requirements

* Quiet git clean

* Handle finalizer issues

* Exit with the signal number

* Get rid of wget

* Fix some Windows compatibility issues with tests

Co-authored-by: Mehrdad <noreply@github.com>
This commit is contained in:
mehrdadn
2020-05-20 12:25:04 -07:00
committed by GitHub
parent aa7a58e92f
commit ebf060d484
17 changed files with 103 additions and 44 deletions
Download Patch File Download Diff File Expand all files Collapse all files
python/ray/tests/test_actor_failures.py
+8 -6
View File
@@ -15,6 +15,8 @@ from ray.test_utils import (relevant_errors, wait_for_condition,
wait_for_errors, wait_for_pid_to_exit,
generate_internal_config_map)
SIGKILL = signal.SIGKILL if sys.platform != "win32" else signal.SIGTERM
@pytest.fixture
def ray_checkpointable_actor_cls(request):
@@ -153,7 +155,7 @@ def test_actor_restart():
pid = ray.get(actor.get_pid.remote())
results = [actor.increase.remote() for _ in range(100)]
# Kill actor process, while the above task is still being executed.
os.kill(pid, signal.SIGKILL)
os.kill(pid, SIGKILL)
# Make sure that all tasks were executed in order before the actor's death.
res = results.pop(0)
i = 1
@@ -194,7 +196,7 @@ def test_actor_restart():
# kill actor process one more time.
results = [actor.increase.remote() for _ in range(100)]
pid = ray.get(actor.get_pid.remote())
os.kill(pid, signal.SIGKILL)
os.kill(pid, SIGKILL)
# The actor has exceeded max restarts, and this task should fail.
with pytest.raises(ray.exceptions.RayActorError):
ray.get(actor.increase.remote())
@@ -235,7 +237,7 @@ def test_actor_restart_with_retry():
pid = ray.get(actor.get_pid.remote())
results = [actor.increase.remote() for _ in range(100)]
# Kill actor process, while the above task is still being executed.
os.kill(pid, signal.SIGKILL)
os.kill(pid, SIGKILL)
# Check that none of the tasks failed and the actor is restarted.
seq = list(range(1, 101))
results = ray.get(results)
@@ -255,7 +257,7 @@ def test_actor_restart_with_retry():
# kill actor process one more time.
results = [actor.increase.remote() for _ in range(100)]
pid = ray.get(actor.get_pid.remote())
os.kill(pid, signal.SIGKILL)
os.kill(pid, SIGKILL)
# The actor has exceeded max restarts, and this task should fail.
with pytest.raises(ray.exceptions.RayActorError):
ray.get(actor.increase.remote())
@@ -351,7 +353,7 @@ def test_actor_restart_without_task(ray_start_regular):
pid = ray.get(actor.get_pid.remote())
p = probe.remote()
os.kill(pid, signal.SIGKILL)
os.kill(pid, SIGKILL)
ray.get(p)
assert wait_for_condition(lambda: not actor_resource_available())
@@ -507,7 +509,7 @@ def test_multiple_actor_restart(ray_start_cluster_head):
def kill_actor(actor):
"""A helper function that kills an actor process."""
pid = ray.get(actor.get_pid.remote())
os.kill(pid, signal.SIGKILL)
os.kill(pid, SIGKILL)
wait_for_pid_to_exit(pid)