search

radical-cybertools / radical.saga

A Light-Weight Access Layer for Distributed Computing Infrastructure and Reference Implementation of the SAGA Python Language Bindings.
http://radical-cybertools.github.io/saga-python/
Other
83 stars 34 forks source link

fixed `core-spec` with consideration of `threads-per-core` (Slurm) #875

Closed mtitov closed 1 year ago

codecov[bot] commented 1 year ago

Codecov Report

Merging #875 (f43c1ca) into devel (1e21463) will not change coverage. The diff coverage is 100.00%.

@@           Coverage Diff           @@
##            devel     #875   +/-   ##
=======================================
  Coverage   43.38%   43.38%           
=======================================
  Files          97       97           
  Lines       12736    12736           
=======================================
  Hits         5526     5526           
  Misses       7210     7210
Impacted Files Coverage Δ
src/radical/saga/adaptors/slurm/slurm_job.py 37.25% <100.00%> (ø)

:mega: We’re building smart automated test selection to slash your CI/CD build times. Learn more

mtitov commented 1 year ago

Generated SLURM batch script for 1 node

$ cat /lustre/orion/scratch/matitov/csc449/radical.pilot.sandbox/rp.session.login06.matitov.019472.0005/pilot.0000/tmp_204hw17z.slurm 
#!/bin/sh

#SBATCH -N 1
#SBATCH -J "pilot.0000"
#SBATCH -D "/lustre/orion/scratch/matitov/csc449/radical.pilot.sandbox/rp.session.login06.matitov.019472.0005/pilot.0000/"
#SBATCH --output "bootstrap_0.out"
#SBATCH --error "bootstrap_0.err"
#SBATCH --account "CSC449"
#SBATCH --partition "batch"
#SBATCH --time 00:15:00
#SBATCH --core-spec=8
#SBATCH --threads-per-core=2

## ENVIRONMENT
export RADICAL_PROFILE="TRUE"
export RP_PILOT_SANDBOX="/lustre/orion/scratch/matitov/csc449/radical.pilot.sandbox/rp.session.login06.matitov.019472.0005/pilot.0000/"
export RADICAL_BASE="/lustre/orion/scratch/matitov/csc449/radical.pilot.sandbox"
export RADICAL_SMT="2"

## EXEC
/bin/bash "-l" "/lustre/orion/scratch/matitov/csc449/radical.pilot.sandbox/rp.session.login06.matitov.019472.0005/pilot.0000//bootstrap_0.sh" "-p" "pilot.0000" "-s" "rp.session.login06.matitov.019472.0005" "-m" "local" "-r" "installed" "-b" "default" "-v" "/autofs/nccs-svm1_home1/matitov/frontier/ve.rp" "-y" "15" "-z" "rp.session.login06.matitov.019472.0005.pmgr_launching.0000.tgz" "-e" "module load cray-python"

p.s. info from PMGLaunching component (smt=2 and 16 blocked cores/threads -> 8 blocked physical cores) -> nodes: 1 [128 8], cores: 112, gpus: 8 (with applied RP PR #2917)