#!/usr/bin/python # # Copyright (C) 2016, 2017 Linaro Limited # Author: Matt Hart # # Copyright (C) 2017 Collabora Limited # Author: Guillaume Tucker # # This module is free software; you can redistribute it and/or modify it under # the terms of the GNU Lesser General Public License as published by the Free # Software Foundation; either version 2.1 of the License, or (at your option) # any later version. # # This library is distributed in the hope that it will be useful, but WITHOUT # ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS # FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more # details. # # You should have received a copy of the GNU Lesser General Public License # along with this library; if not, write to the Free Software Foundation, Inc., # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA # Usage ./lava-v2-submit-jobs.py --username test --token xxxx --lab lab-name --jobs jobfolder import os import xmlrpclib import yaml import subprocess import fnmatch import time import re import argparse import httplib import json import configparser from lib import utils from lib import configuration from kernelci import lava DEVICE_ONLINE_STATUS = ['idle', 'running', 'reserved'] JOBS = {} SUBMITTED = {} def submit_jobs(connection): print "Fetching all devices from LAVA" all_devices = connection.scheduler.all_devices() print "Fetching all device-types from LAVA" all_device_types = connection.scheduler.all_device_types() print "Fetching all device-type aliases from LAVA" all_aliases = [] try: for alias in connection.scheduler.aliases.list(): all_aliases.append(connection.scheduler.aliases.show(alias)) except (xmlrpclib.ProtocolError, xmlrpclib.Fault, IOError, ValueError) as e: print("Unable to retrieve device_type aliases from LAVA; continuing") print(e) result = True print "Submitting Jobs to Server..." for job in JOBS: try: with open(job, 'rb') as stream: job_data = stream.read() job_info = yaml.safe_load(job_data) # Check if request device(s) are available if 'device_type' in job_info: device_type = lava.get_device_type_by_name(job_info['device_type'], all_device_types, all_aliases) if device_type: if device_type_has_available(device_type, all_devices): print "Submitting job %s to device-type %s" % (job_info.get('job_name', 'unknown'), job_info['device_type']) job_id = connection.scheduler.submit_job(job_data) SUBMITTED[job] = job_id else: print "Found device-type %s on server, but it had no available pipeline devices" % device_type['name'] elif 'device_group' in job_info: print "Multinode Job Detected! Not supported yet :(" elif 'vm_group' in job_info: print "VMGroup Job Detected! Not supported yet :(" else: print "Should never get here - no idea what job type" print os.path.basename(job) + ' : skip' except (xmlrpclib.ProtocolError, xmlrpclib.Fault, IOError, ValueError) as e: print "Job submission error!" print job print e result = False continue return result def load_jobs(top): for root, dirnames, filenames in os.walk(top): for filename in fnmatch.filter(filenames, '*.yaml'): JOBS[os.path.join(root, filename)] = None def device_type_has_available(device_type, all_devices): for device in all_devices: if device[1] == device_type['name']: if device[2].lower() in DEVICE_ONLINE_STATUS: try: return device[4] except IndexError: print "LAVA all_devices() XMLRPC does not support pipeline" return False return False def main(args): config = configuration.get_config(args) jobs_submitted = config.get('submitted') lab = config.get('lab') if jobs_submitted: if not lab: raise Exception("Lab name required when saving submitted jobs") if os.path.exists(jobs_submitted): os.unlink(jobs_submitted) if config.get("repo"): retrieve_jobs(config.get("repo")) jobs = config.get("jobs") print("Loading jobs from {}".format(jobs)) load_jobs(jobs) if not JOBS: print("No jobs to submit") result = False else: start_time = time.time() labs_config = configparser.ConfigParser() labs_config.read('labs.ini') lava_api = labs_config[config.get("lab")]['api'] print("LAVA API: {}".format(lava_api)) url = utils.validate_input(config.get("username"), config.get("token"), lava_api) connection = utils.connect(url) result = submit_jobs(connection) if jobs_submitted and SUBMITTED: print("Saving submitted jobs data in {}".format(jobs_submitted)) data = { 'start_time': start_time, 'lab': config.get('lab'), 'jobs': {k: v for k, v in SUBMITTED.iteritems() if v}, } with open(jobs_submitted, 'w') as json_file: json.dump(data, json_file) exit(0 if result is True else 1) if __name__ == '__main__': parser = argparse.ArgumentParser() parser.add_argument("--config", help="configuration for the LAVA server") parser.add_argument("--section", default="default", help="section in the LAVA config file") parser.add_argument("--lab", help="KernelCI Lab Name") parser.add_argument("--jobs", required=True, help="path to jobs directory (default is cwd)") parser.add_argument("--username", help="username for the LAVA server") parser.add_argument("--token", help="token for LAVA server api") parser.add_argument("--repo", help="git repo for LAVA jobs") parser.add_argument("--submitted", help="path to JSON file to save submitted jobs data") args = vars(parser.parse_args()) main(args)