Peer Drivers on a Network

Ray Core
1

1. Severity of the issue: (select one)
[X ] None: I’m just curious or want clarification.
Low: Annoying but doesn’t hinder my work.
Medium: Significantly affects my productivity but can find a workaround.
High: Completely blocks me.

2. Environment:

  • Ray version: 2.44.1
  • Python version: 3.12
  • OS: Linux
  • Cloud/Infrastructure: Runpod
  • Other libs/tools (if relevant): LiveKit

3. What happened vs. what you expected:

  • Expected: Connect to drivers on different pods. Add an event bus and/or shared datastore (Redis). Check shared state.
  • Actual: …?

While there are other solutions for this. I’m wondering if Ray is a good platform to:

  • Discover peer drivers on other pods
  • Use a shared Redis DB or another Fast & Consistent network data store
  • Perform a remote call on an actor in another Pod.

Any good docs or guidelines to set this up?

2

Hi @btak ,

What are you asking for?

3

I was wondering if you could have different machines with different ray drivers share data. I suppose you would need to have a master ray driver for the new machines to connect to?

What happens if the master driver is taken offline?