Compute Core Hours

compute_core_hours.py

#!/usr/bin/env python
import json
import sys
import argparse
import os
import operator


def extract_table_values(dt, column_id):
    for column_d in dt['columns']:
        if column_d['id'] == column_id:
            return column_d['values']

    raise Exception("Unable to extract column id {}".format(column_id))


def load_from_task_report_json(x):
    """
    Load from the <JOB-ROOT>/workflow/task-reports.json

    """
    with open(x) as f:
        d = json.load(f)

    dt = d['tables'][0]

    nprocs = extract_table_values(dt, "pbsmrtpipe.tasks.nproc")
    run_time_sec = extract_table_values(dt, "pbsmrtpipe.tasks.run_time_sec")
    run_time_hour = [x / 60.0 / 60.0 for x in run_time_sec]


    run_time_tasks_total = [x * y for x, y in zip(nprocs, run_time_hour)]
    return reduce(operator.add, run_time_tasks_total, 0)


def get_parser():
    desc = "Extract core hours from workflow/report-tasks.json"
    p = argparse.ArgumentParser(description = desc, version = "0.1.0")
    p.add_argument("task_report_json", help="Path to <JOB-ROOT-DIR>/workflow/report-tasks.json")
    return p


def main(argv):
    p = get_parser()
    pargs = p.parse_args(argv)
    total = load_from_task_report_json(pargs.task_report_json)
    print total
    return 0


if __name__ == '__main__':
    sys.exit(main(sys.argv[1:]))