ray-project / ray

Ray is a unified framework for scaling AI and Python applications. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
https://ray.io
Apache License 2.0
33.3k stars 5.63k forks source link

[tune][kubernetes][client] cloudpickle error #16678

Open DmitriGekhtman opened 3 years ago

DmitriGekhtman commented 3 years ago

What is the problem?

Deploying on K8s following doc instructions and running a sample program leads to an error

ray.util.connect("127.0.0.1:10001")
env="CartPole-v0"
tune.run(
ppo.PPOTrainer,
config={
"env": env,
"timesteps_per_iteration": 300,
"monitor": False,
"log_level": "DEBUG",
"ignore_worker_failures": True
})
return self._call_schedule_for_task(task)
  File "/home/goranj/miniconda3/envs/rayenv/lib/python3.7/site-packages/ray/util/client/worker.py", line 318, in _call_schedule_for_task
    raise cloudpickle.loads(ticket.error)
ImportError: libgthread-2.0.so.0: cannot open shared object file: No such file or directory

Discussion: https://discuss.ray.io/t/error-running-tune-on-k8s/2640

Ray version and other system information (Python version, TensorFlow version, OS):

Reproduction (REQUIRED)

Please provide a short code snippet (less than 50 lines if possible) that can be copy-pasted to reproduce the issue. The snippet should have no external library dependencies (i.e., use fake or mock data / environments):

If the code snippet cannot be run by itself, the issue will be closed with "needs-repro-script".

DmitriGekhtman commented 3 years ago

@richardliaw -- looks like a tune package dependency issue. Can someone from Tune team take a look?

richardliaw commented 3 years ago

I think this is a rllib installation problem; can you please try running pip install ray[rllib]?

On Tue, Jul 20, 2021 at 11:43 AM Dmitri Gekhtman @.***> wrote:

@richardliaw https://github.com/richardliaw -- looks like a tune package dependency issue. Can someone from Tune team take a look?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ray-project/ray/issues/16678#issuecomment-883613063, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABCRZZLZAWD3JJ2DAIXXAPTTYW7VDANCNFSM47KKE3PQ .