ray-project / ray

Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
https://ray.io
Apache License 2.0
33.5k stars 5.69k forks source link

Support Availability Zone Deployment in Azure #39966

Open ben-directai opened 1 year ago

ben-directai commented 1 year ago

Description

Ability to specify Azure availability zone in cluster deployment. Note that it's possible on AWS - see Cluster YAML Configuration.

Use case

I'm unable to deploy A10 and A100 Ray Workers on Azure even though I have the quota to support them in a given region. Ray's inability to support deployment in a specific Azure Region's Availability Zone makes it impossible for me to transfer deployment from AWS to Azure.

avi-directai commented 1 year ago

It looks like this is possible with Azure's SDK: https://stackoverflow.com/questions/77291804/how-to-specify-availability-zone-when-creating-vm-through-azure-sdk.

anyscalesam commented 5 months ago

@ben-directai @avi-directai were you able to get this working?