ray: Azure Autoscaler - Permission denied connecting to Docker socket - Head Node Setup

Ray RLLib Version 2.0.0.dev0

When running ray up for an Azure deployment, the following occurs:

  [5/7] Initalizing command runner
Shared connection to 20.71.92.41 closed.
Got permission denied while trying to connect to the Docker daemon socket at unix:///var/run/docker.sock: Post http://%2Fvar%2Frun%2Fdocker.sock/v1.40/images/create?fromImage=rayproject%2Fray-ml&tag=latest-gpu: dial unix /var/run/docker.sock: connect: permission denied
Shared connection to 20.71.92.41 closed.
  New status: update-failed
  !!!
  SSH command failed.
  !!!
  
  Failed to setup head node.

About this issue

  • Original URL
  • State: closed
  • Created 3 years ago
  • Comments: 15 (11 by maintainers)

Most upvoted comments

Might be related to https://github.com/ray-project/ray/issues/13711

Update: adding the recommended sudo usermod -aG docker $USER; sleep 10 to initialization_commands worked!

Ok! Then it’s just a matter of configuring those correctly.

I got it running here now! Can definitely do a .yaml update once I am able to run an experiment then if that’s ok? Currently stuck due to #14666

@XavierGeerinck Awesome! Glad to hear that worked! I’ll be trying to fix some of these launch issues (this one and the one with setup_commands as soon as I get some Azure credentials 😃