[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[HTCondor-users] HTC-CE / Slurm and cpu request


I'm trying to set up a HTCondor-CE (3.2.2) on top of Slurm; so far we have early test pilots flowing in.

The corresponding slurm job submission requires 8 Nodes to run, however the pilot was intended to
require 8 cores (on the same node).

Apparently this "8 nodes request" happens in the /usr/libexec/condor/glite/bin/slurm_submit.sh
script here:

# Simple support for multi-cpu attributes
if [[ $bls_opt_mpinodes -gt 1 ]] ; then
 echo "#SBATCH -N $bls_opt_mpinodes" >> $bls_tmp_file

Is there someone else using HTC-CE / Slurm experience having seen a similar problem?

Thank You