University of Bath

Research Computing Team (DDaT)

Anatra HPC Documentation

Schedule Setup Anatra Access Slurm Hardware Storage Software Gaussian Jobs VNC Session GPU Jobs Apptainer Containers ORCA Jobs

Running ORCA Jobs¶

What is ORCA?¶

ORCA is a powerful quantum chemistry software package developed by the Max Planck Institute for Chemical Energy Conversion.
It is widely used for performing electronic structure calculations and supports a wide range of computational chemistry methods such as:

Density Functional Theory (DFT)
Hartree–Fock (HF) and post-HF methods (MP2, CCSD, etc.)
Geometry optimization, vibrational frequency analysis, and excited-state simulations

ORCA can efficiently utilize multiple CPU cores via MPI (Message Passing Interface) on our Anatra HPC cluster.
This submission script automates the process of running ORCA 6.1.0 using Slurm with OpenMPI, providing researchers with a reliable and reproducible workflow for large-scale quantum chemistry simulations.

ORCA Script¶

ORCA 6.1.0¶

Below is an example script for an ORCA 6.1.0 job, which should be submitted from the /scratch/projects/[project-code]/ storage area.

#!/bin/bash
set -euo pipefail

# Defaults

readonly DEFAULT_NPROCS=12
readonly DEFAULT_TIME="3-00:00:00"
readonly DEFAULT_PARTITION="chemistry"
readonly MAX_PROCS=96

nprocs=${DEFAULT_NPROCS}
time=${DEFAULT_TIME}
partition=${DEFAULT_PARTITION}

usage() 
{
    cat << EOF
ORCA 6.1.0 Job Submission Script (PMIx-Enabled)

USAGE: $(basename $0) <input_file> [input_file ...] [OPTIONS]

OPTIONS:
    -n <cores>      Number of CPU cores (default: ${DEFAULT_NPROCS}, max: ${MAX_PROCS})
    -t <time>       Time limit in dd-hh:mm:ss (default: ${DEFAULT_TIME})
    -p <partition>  SLURM partition (default: ${DEFAULT_PARTITION})
    -h              Display this help message
EOF
    exit 0
}

validate_input() 
{
    if ! [[ "$nprocs" =~ ^[0-9]+$ ]] || [ "$nprocs" -le 0 ]; then
        echo "ERROR: Number of cores must be a positive integer" >&2
        exit 1
    fi

    if [ "$nprocs" -gt "$MAX_PROCS" ]; then
        nprocs=$MAX_PROCS
    fi

    if ! [[ "$time" =~ ^[0-9]+-[0-9]{2}:[0-9]{2}:[0-9]{2}$ ]]; then
        echo "ERROR: Invalid time format. Use dd-hh:mm:ss" >&2
        exit 1
    fi
}

check_file_exists()
{
    local file=$1
    if [ ! -f "$file" ] || [ ! -r "$file" ]; then
        echo "ERROR: File not found or not readable: $file" >&2
        return 1
    fi
    return 0
}

create_job_script()
{
    local infile=$1
    local head="${infile%.*}"
    local job_script="${head}.slm"

    cat > "$job_script" <<'EOF'

#!/bin/bash
#SBATCH --job-name=HEAD_PLACEHOLDER
#SBATCH --partition=PARTITION_PLACEHOLDER
#SBATCH --output=HEAD_PLACEHOLDER.%j.out
#SBATCH --error=HEAD_PLACEHOLDER.%j.err
#SBATCH --time=TIME_PLACEHOLDER
#SBATCH --nodes=1
#SBATCH --ntasks=NPROCS_PLACEHOLDER
#SBATCH --cpus-per-task=1

set -euo pipefail

# Load modules

module purge
module load GCC/15.1.0 openmpi/4.1.6
module load orca/6.1.0

# PMIx configuration

export SLURM_MPI_TYPE=pmix
export OMPI_MCA_pmix_base_verbose=0
export ORCA_MPI_RSH="srun"

# Setup paths

logfile="$(pwd)/HEAD_PLACEHOLDER.out"
workdir="$(pwd)"
tempdir="/tmp/jobs/${SLURM_JOB_ID}/HEAD_PLACEHOLDER.files"

mkdir -p "$tempdir" || { echo "ERROR: Cannot create temp directory"; exit 1; }
cp "$workdir/HEAD_PLACEHOLDER"* "$tempdir" || { echo "ERROR: Cannot copy input files"; exit 1; }
cd "$tempdir" || { echo "ERROR: Cannot change to temp directory"; exit 1; }

# Run ORCA

$(which orca) INFILE_PLACEHOLDER 1>>"$logfile" 2>&1
orca_exit=$?

# Cleanup and copy results

find "$tempdir" -maxdepth 1 -type f -name "*.tmp*" -delete
cp -Rf "$tempdir"/* "$workdir/" || { echo "ERROR: Cannot copy results back"; exit 1; }

exit $orca_exit
EOF

    # Replace placeholders

    sed -i "s|HEAD_PLACEHOLDER|${head}|g" "$job_script"
    sed -i "s|PARTITION_PLACEHOLDER|${partition}|g" "$job_script"
    sed -i "s|TIME_PLACEHOLDER|${time}|g" "$job_script"
    sed -i "s|NPROCS_PLACEHOLDER|${nprocs}|g" "$job_script"
    sed -i "s|INFILE_PLACEHOLDER|${infile}|g" "$job_script"

    chmod +x "$job_script"
    echo "$job_script"
}

# Parse arguments

if [ $# -eq 0 ]; then
    usage
fi

files=()
while [[ $# -gt 0 && "$1" == *.inp ]]; do
    files+=("$1")
    shift
done

while getopts "n:t:p:h" opt; do
    case "$opt" in
        n) nprocs=$OPTARG ;;
        t) time=$OPTARG ;;
        p) partition=$OPTARG ;;
        h) usage ;;
        *) usage ;;
    esac
done

validate_input

if [ ${#files[@]} -eq 0 ]; then
    echo "ERROR: No input files provided" >&2
    usage
fi

# Process files

submitted_jobs=0
failed_jobs=0

for infile in "${files[@]}"; do
    if [[ "$infile" == *.inp ]]; then
        if ! check_file_exists "$infile"; then
            ((failed_jobs++))
            continue
        fi

        job_script=$(create_job_script "$infile")

        if job_id=$(sbatch "$job_script" 2>&1); then
            echo "Submitted: $infile -> $job_id"
            ((submitted_jobs++))
        else
            echo "Failed: $infile" >&2
            ((failed_jobs++))
        fi
    fi
done

echo "Summary: $submitted_jobs submitted, $failed_jobs failed"
exit 0

ORCA Job Submission Guide for Anatra Cluster¶

How to Run the Script¶

Follow these steps to run your ORCA jobs using this submission script on the HPC cluster.

🧩 Step 1 — Save the Script¶

Save the code above as:

submit_orca.sh

Then make it executable:

chmod +x submit_orca.sh

🧪 Step 2 — Prepare Your Input File¶

Create your ORCA input file, for example:

molecule.inp

🚀 Step 3 — Submit a Job¶

Submit a single job:

./submit_orca.sh molecule.inp -n 24 -t 2-00:00:00 -p chemistry

Or multiple jobs:

./submit_orca.sh job1.inp job2.inp -n 16 -t 1-12:00:00

📊 Step 4 — Monitor Job Progress¶

Check the status of your jobs:

squeue -u $USER

To view output logs:

cat molecule.%j.out

🧭 Step 5 — Cancel a Job (if needed)¶

Cancel a specific job:

scancel <job_id>

📂 Step 6 — Check Output Files¶

After the job finishes, the following files will appear in your working directory:

molecule.out
molecule.%j.out
molecule.%j.err
molecule.gbw

These files contain ORCA's output data and calculation logs.

📝 Notes¶

Replace <job_id> with your actual SLURM job ID when canceling jobs
The %j in filenames will be automatically replaced with your job ID number
Adjust the number of cores (-n) and time limit (-t) based on your calculation requirements
Make sure to specify the correct partition (-p) for your department (Chemistry or Lifesciences) HPC setup