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

define engine API/rpc interface (make engine callable) #1709

Open
Tracked by #1696
mwartell opened this issue Oct 25, 2022 · 1 comment
Open
Tracked by #1696

define engine API/rpc interface (make engine callable) #1709

mwartell opened this issue Oct 25, 2022 · 1 comment

Comments

@mwartell
Copy link
Collaborator

mwartell commented Oct 25, 2022

Currently the launcher/loader

  1. reads an experiment
  2. gathers all the necessary resources
  3. launches an image with resource mounts
  4. passes the experiment to the container as a shell command
  5. retrieves the standard out from the container

Because the mounts in (3) assume a local unix platform, this means that the container must be run on localhost. Also,
the narrow stdin/stdout communication with the container forces us into a batch oriented run. That is, niceties like interacting with the armory run is impossible.

We are looking at the way Jupyterlab connects to its compute kernel to see if we can break the batch-orientation.

@mwartell
Copy link
Collaborator Author

Per today's stand-up, @christopherwoodall will be looking at the Jupiterlab/kernel interface protocol for ideas.

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

No branches or pull requests

2 participants