Dawid
February 6, 2022, 12:19pm
2
This post was hidden so I created new topic in Ray Cluster’s category. Apparently this problem arises when Ray Cluster is for few hours. Here is 2nd thread:
Hey guys,
creation an actor on worker node after it’s been running a while takes 5-7 minutes. There is no problem with creation of actor on head node.
python-core-worker log:
[2022-02-06 07:02:12,887 I 11033 11033] core_worker.cc:2320: Creating actor: b0f9db98654cd3b2d462f7d630000000
[2022-02-06 07:09:35,118 I 11033 11080] actor_manager.cc:228: received notification on actor, state: ALIVE, actor_id: a6c31f8498b256bc927bad0518000000, ip address: 10.0.1.157, port: 10112, worker_id: 9ad6581ff8f2…