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

WLM integration of GPU support #171

Open
wants to merge 10 commits into
base: master
Choose a base branch
from
Open

WLM integration of GPU support #171

wants to merge 10 commits into from

Conversation

rukkal
Copy link
Contributor

@rukkal rukkal commented Apr 21, 2017

This PR adds workload manager integration for GPU support + some minor changes.

We modified setupRoot to attempt to activate the Shifter's GPU support, and if the host doesn't provide GPU resources, the activation is silently skipped.

The minor changes are:

  • Removed bind mount of some NVIDIA binaries.
  • Removed execution of nvidia-modprobe to create /dev/nvidia-uvm.

rukkal and others added 10 commits April 19, 2017 11:45
Such information was no longer really useful. Removing it
also makes things easier to activate the GPU support through
setupRoot (workload manager integration).
Always attempt to automatically enable GPU support. We decided
to avoid the alternative solution of having the user to explicitely
request the GPU support activation through an extra workload manager's
option. Such alternative solution wouldn't have been consistent with
the interface of shifter when used without workload manager
integration as well as the interface of nvidia-docker.
@coveralls
Copy link

Coverage Status

Coverage increased (+0.001%) to 68.017% when pulling f4c01e9 on rukkal:gpu_support_wlm_integration into 4dc2134 on NERSC:master.

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.

3 participants