Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat: enable fully external configuration #386

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

kyriediculous
Copy link
Contributor

@kyriediculous kyriediculous commented Dec 28, 2024

This PR allows the worker to be fully configured to use external containers only, in which case the worker no longer needs active container management.

This is primarily useful when using a fully containerised setup using e.g. docker compose, where the ai-runner containers for various pipelines/models are isolated containers.

In such case, the need for access to the Docker daemon from the ai-worker, would prevent this setup from running because by default the host's docker daemon isn't available for other containers.

E1227 10:53:39.195246       1 starter.go:1216] Error starting AI worker: Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?

While for a hybrid setup (seems uncommon), the docker socket still has to be passed as a volume or docker has to be ran as a root, having the availability for a fully containerized setup would be highly beneficial.

go-livepeer

This PR allows flexibility to how this feature can be implemented in go-livepeer, some options include

  • Explicitly define aiRunnerImage as "" as startup flag or something like "external"

  • From aiModels.json , check if all entries contain a valid URL parameter, if so, no need to start the runner container

Currently the second option is implemented in this draft PR : livepeer/go-livepeer#3335

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant