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

Automatic detection of the system configuration #3320

Draft
wants to merge 15 commits into
base: develop
Choose a base branch
from

Conversation

Blanca-Fuentes
Copy link
Contributor

@Blanca-Fuentes Blanca-Fuentes commented Nov 18, 2024

This PR introduces the option of detecting the system configuration automatically and writing the corresponding ReFrame config file.

  • The system's partitions are created based on the node features detected with Slurm .
  • It is only implemented for Slurm
  • ReFrame must be run with the new action option: reframe --detect-configuration=autodetect_config.py (an example of the autodetect_config.py is included). In the detect_config() the optional arguments are
  1. detect_containers: submits jobs to each partition to detect container platforms
  2. sched_options: additional scheduler access options that must be included for all partitions
  3. time_limit: time limit for the remote detection jobs (container platforms)
  4. exclude_feats: node features to be ignored in the filtering of the nodes to determine the partitions, patterns like row* are accepted
  5. filename: name of the configuration file that will be generated
  • For the job submissions, a shared directory with the compute nodes must/can be specified using the RFM_REMOTE_WORKDIR

Ideas/todo's

  • Move the detection of container platforms in the remote partitions to --detect-topology
  • Combine the configuration detection with reframe --detect-topology
  • Fix tests
  • Add documentation
  • Add unit tests

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
Status: Todo
Development

Successfully merging this pull request may close these issues.

2 participants