aws-samples/aws-eda-slurm-cluster

[FEATURE] Add Exostellar support

cartalla opened this issue · 0 comments

Is your feature request related to a problem? Please describe.
Exostellar provides a nested virtualization solution on EC2 that predicts spot terminations far enough in advance to live migrate the instance to another spot or on-demand instance.
This enables running long-running, stateful jobs on spot without losing job progress when a spot termination occurs.

Describe the solution you'd like
Exostellar support the Slurm scheduler.
At a minimum, add documention on how to integrate Exostellar into this Slurm cluster.
Ideally, install and configure the software.