How severe does this issue affect your experience of using Ray?
High: It blocks me to complete my task.
I am new to ray and I have a project to work on so I am trying to use ray
I am running a ubuntu 16.0.4 lts and another laptop with windows os and head node is running on ubuntu and ray starts normally when i go into worker laptop (windows ) and type ray start --address=“192.168.xxx.xxxx:xxxx”
it says local node ip: 127.0.0.1 and gives me no error but when i check ray dashboard this laptop is not available there.
I googled and tried this
ray start --address=‘127.0.0.1’ --node-ip-address=‘192.168.xx.xxx:xxx’ then it says unable to connect to gcs server check firewall or gcs address is correct.
Do your machines have bidirectional networking access? In particular, for the ports listed here? Configuring Ray — Ray 1.12.1
Windows multi-node support is in beta. If you’ve confirmed that the networking access is set up correctly, do you mind filling a github issue including more information such as the ray version, whether they’re on the same network, and the relevant logs from /tmp/ray/session_latest/logs on both machines?
I have very little idea of networking, I don’t know what is happening with the ports.
I am just starting ray cluster manually, first in the host machine i type ray start --head
and then in worker machine ray start --address=“192.168.x.xx:6379” ← address of head node.
I am trying to trouble shoot this since last month and now I am worried as this is not at all working. can somebody help me I am so badly in need of using disrtibuted computing for my project I am googling a lot and watching youtube videos but nothing works.
@cade The issue still exists in ray1.13 it is not resolved. It would have been better if such a powerful library had support to windows because here in the schools at UK most of our systems are running on windows.