Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

CuPy 7.2 breaks CI #241

Closed
quasiben opened this issue Feb 21, 2020 · 5 comments · Fixed by #248
Closed

CuPy 7.2 breaks CI #241

quasiben opened this issue Feb 21, 2020 · 5 comments · Fixed by #248

Comments

@quasiben
Copy link
Member

This is probably any issue with CuPy but wanted note here that we have seen breaking issues with CuPy 7.2

E   cupy.cuda.curand.CURANDError: CURAND_STATUS_INITIALIZATION_FAILED

cupy/cuda/curand.pyx:86: CURANDError
``` params = {'device_memory_limit': 1000000000.0, 'host_pause': None, 'host_spill': 0.0, 'host_target': 0.0, ...}
@pytest.mark.parametrize(
    "params",
    [
        {
            "device_memory_limit": 1e9,
            "memory_limit": 4e9,
            "host_target": 0.0,
            "host_spill": 0.0,
            "host_pause": None,
            "spills_to_disk": False,
        },
        {
            "device_memory_limit": 1e9,
            "memory_limit": 1e9,
            "host_target": 0.0,
            "host_spill": 0.0,
            "host_pause": None,
            "spills_to_disk": True,
        },
    ],
)
@pytest.mark.asyncio
async def test_cupy_cluster_device_spill(params):
    cupy = pytest.importorskip("cupy")
    with dask.config.set({"distributed.worker.memory.terminate": False}):
        async with LocalCUDACluster(
            1,
            scheduler_port=0,
            processes=True,
            silence_logs=False,
            dashboard_address=None,
            asynchronous=True,
            death_timeout=60,
            device_memory_limit=params["device_memory_limit"],
            memory_limit=params["memory_limit"],
            memory_target_fraction=params["host_target"],
            memory_spill_fraction=params["host_spill"],
            memory_pause_fraction=params["host_pause"],
        ) as cluster:
            async with Client(cluster, asynchronous=True) as client:

                rs = da.random.RandomState(RandomState=cupy.random.RandomState)
              x = rs.random(int(250e6), chunks=10e6)

dask_cuda/tests/test_spill.py:181:


/conda/envs/gdf/lib/python3.6/site-packages/dask/array/random.py:387: in random_sample
return self._wrap("random_sample", size=size, chunks=chunks)
/conda/envs/gdf/lib/python3.6/site-packages/dask/array/random.py:188: in _wrap
small_kwargs,
/conda/envs/gdf/lib/python3.6/site-packages/dask/array/random.py:453: in _apply_random
state = RandomState(state_data)
/conda/envs/gdf/lib/python3.6/site-packages/cupy/random/generator.py:51: in init
self._generator = curand.createGenerator(method)
cupy/cuda/curand.pyx:93: in cupy.cuda.curand.createGenerator
???
cupy/cuda/curand.pyx:97: in cupy.cuda.curand.createGenerator
???


???
E cupy.cuda.curand.CURANDError: CURAND_STATUS_INITIALIZATION_FAILED

cupy/cuda/curand.pyx:86: CURANDError

</details>
@jakirkham
Copy link
Member

Does this go away when setting cupy.cuda.cub_enabled = False?

@pentschev
Copy link
Member

Didn't we just pin a couple days ago an older version of CuPy because of that?

@quasiben
Copy link
Member Author

Yes, we pinned. I was raising this issue to track that we are not working with 7.2

@pentschev
Copy link
Member

Oh, I see. Sorry, I thought we were seeing that again, my mistake.

@jakirkham
Copy link
Member

Added PR ( #248 ) to relax the pin since we've rebuilt the cupy package with a fix.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants