Python common.LOGGER类代码示例

    def add_file_to_bucket_multipart(self, bucket_name, key_name, source_path, parallel_processes=2, reduced_redundancy=True):
        """
        Parallel multipart upload.
        """
        LOGGER.info('bucket_name: {0}, key_name: {1}, filename: {2}, parallel_processes: {3}, reduced_redundancy: {4}'.format(
            bucket_name, key_name, source_path, parallel_processes, reduced_redundancy))

        source_size = os.stat(source_path).st_size
        bytes_per_chunk = 10 * 1024 * 1024
        chunk_amount = int(math.ceil(source_size / float(bytes_per_chunk)))
        if chunk_amount < 10000:
            bucket = self.get_bucket(bucket_name)

            headers = {'Content-Type': mimetypes.guess_type(key_name)[0] or 'application/octet-stream'}
            mp = bucket.initiate_multipart_upload(key_name, headers=headers, reduced_redundancy=reduced_redundancy)

            LOGGER.info('bytes_per_chunk: {0}, chunk_amount: {1}'.format(bytes_per_chunk, chunk_amount))

            # You can only upload 10,000 chunks
            pool = Pool(processes=parallel_processes)
            for i in range(chunk_amount):
                offset = i * bytes_per_chunk
                remaining_bytes = source_size - offset
                bytes_to_copy = min([bytes_per_chunk, remaining_bytes])
                part_num = i + 1
                pool.apply_async(upload_part, [self._aws_access_key_id, self._aws_secret_access_key, bucket_name, mp.id, part_num, source_path, offset, bytes_to_copy])
            pool.close()
            pool.join()

            if len(mp.get_all_parts()) == chunk_amount:
                mp.complete_upload()
            else:
                mp.cancel_upload()
        else:
            raise S3UploadException('Too many chunks')

def main():
    parser = argparse.ArgumentParser('Start a number of CLEAN servers')
    parser.add_argument('-a', '--ami_id', help='the AMI id to use')
    parser.add_argument('-i', '--instance_type', required=True, help='the instance type to use')
    parser.add_argument('-c', '--created_by', help='the username to use')
    parser.add_argument('-n', '--name', required=True, help='the instance name to use')
    parser.add_argument('-s', '--spot_price', type=float, help='the spot price to use')
    parser.add_argument('-b', '--bash_script', help='the bash script to use')
    parser.add_argument('-e', '--ebs', type=int, help='the size in GB of any EBS volume')
    parser.add_argument('bottom_frequency', help='The bottom frequency')
    parser.add_argument('frequency_range', help='the range of frequencies')
    parser.add_argument('obs_id', help='the observation id')

    args = vars(parser.parse_args())

    corrected_args = check_args(args)
    if corrected_args is None:
        LOGGER.error('The arguments are incorrect: {0}'.format(args))
    else:
        start_servers(
            corrected_args['ami_id'],
            corrected_args['user_data'],
            corrected_args['setup_disks'],
            args['instance_type'],
            make_safe_filename(args['obs_id']),
            corrected_args['created_by'],
            args['name'],
            corrected_args['instance_details'],
            corrected_args['spot_price'],
            args['ebs'],
            args['bottom_frequency'],
            args['frequency_range'])

 def _upload(retries_left=amount_of_retries):
     try:
         LOGGER.info(
             'Start uploading part: #{0}, source_path: {1}'.format(
                 part_num,
                 source_path
             )
         )
         conn = get_s3_connection(aws_access_key_id, aws_secret_access_key)
         bucket = conn.get_bucket(bucket_name)
         for mp in bucket.get_all_multipart_uploads():
             if mp.id == multipart_id:
                 with FileChunkIO(source_path, 'r', offset=offset, bytes=bytes_to_copy) as fp:
                     mp.upload_part_from_file(fp=fp, part_num=part_num)
                 break
     except Exception, exc:
         if retries_left:
             _upload(retries_left=retries_left - 1)
         else:
             LOGGER.info(
                 'Failed uploading part: #{0}, source_path: {1}'.format(
                     part_num,
                     source_path
                 )
             )
             raise exc

    def __call__(self):
        """
        Actually run the job
        """
        # Get the name of the volume
        ec2_helper = EC2Helper()
        iops = None
        if self._instance_details.iops_support:
            iops = 500

        zone = ec2_helper.get_cheapest_spot_price(self._instance_type, self._spot_price)
        if zone is not None:
            volume, snapshot_name = ec2_helper.create_volume(self._snapshot_id, zone, iops=iops)
            LOGGER.info('obs_id: {0}, volume_name: {1}'.format(self._obs_id, snapshot_name))
            user_data_mime = self.get_mime_encoded_user_data(volume.id)

            if self._spot_price is not None:
                ec2_helper.run_spot_instance(
                    self._ami_id,
                    self._spot_price,
                    user_data_mime,
                    self._instance_type,
                    volume.id,
                    self._created_by,
                    '{1}-{2}-{0}'.format(self._name, snapshot_name, self._counter),
                    self._instance_details,
                    zone,
                    ephemeral=True)
        else:
            LOGGER.error('Cannot get a spot instance of {0} for ${1}'.format(self._instance_type, self._spot_price))

def copy_files(processes, bottom_frequency, frequency_range):
    # Create the directory
    if not exists(DIRECTORY):
        os.makedirs(DIRECTORY)

    # Scan the bucket
    s3_helper = S3Helper()
    bucket = s3_helper.get_bucket(CHILES_BUCKET_NAME)
    LOGGER.info('Scanning bucket: {0}/CLEAN'.format(bucket))

    # Create the queue
    queue = multiprocessing.JoinableQueue()

    # Start the consumers
    for x in range(processes):
        consumer = Consumer(queue)
        consumer.start()

    for key in bucket.list(prefix='CLEAN/'):
        LOGGER.info('Checking {0}'.format(key.key))
        # Ignore the key
        if key.key.endswith('.image.tar.gz') or key.key.endswith('.image.tar'):
            # Do we need this file?
            basename_key = basename(key.key)
            if in_frequency_range(basename_key, bottom_frequency, frequency_range):
                # Queue the copy of the file
                temp_file = os.path.join(DIRECTORY, basename_key)
                queue.put(Task(key, temp_file, DIRECTORY))

    # Add a poison pill to shut things down
    for x in range(processes):
        queue.put(None)

    # Wait for the queue to terminate
    queue.join()

def copy_files(frequency_id):
    s3_helper = S3Helper()
    # Look in the output directory
    LOGGER.info('directory_data: {0}'.format(CHILES_CLEAN_OUTPUT))
    for dir_name in os.listdir(CHILES_CLEAN_OUTPUT):
        LOGGER.info('dir_name: {0}'.format(dir_name))
        result_dir = join(CHILES_CLEAN_OUTPUT, dir_name)
        if isdir(result_dir) and dir_name.startswith('cube_') and dir_name.endswith('.image'):
            LOGGER.info('dir_name: {0}'.format(dir_name))
            output_tar_filename = join(CHILES_CLEAN_OUTPUT, dir_name + '.tar')

            if can_be_multipart_tar(result_dir):
                LOGGER.info('Using add_tar_to_bucket_multipart')
                s3_helper.add_tar_to_bucket_multipart(
                    CHILES_BUCKET_NAME,
                    '/CLEAN/{0}/{1}'.format(frequency_id, basename(output_tar_filename)),
                    result_dir)
            else:
                LOGGER.info('Using make_tarfile, then adding file to bucket')
                make_tarfile(output_tar_filename, result_dir)

                s3_helper.add_file_to_bucket(
                    CHILES_BUCKET_NAME,
                    'CVEL/{0}/{1}/data.tar'.format(frequency_id, basename(output_tar_filename)),
                    output_tar_filename)

                # Clean up
                os.remove(output_tar_filename)

def main():
    parser = argparse.ArgumentParser('Start a number of CVEL servers')
    parser.add_argument('-a', '--ami_id', help='the AMI id to use')
    parser.add_argument('-i', '--instance_type', required=True, help='the instance type to use')
    parser.add_argument('-c', '--created_by', help='the username to use')
    parser.add_argument('-n', '--name', required=True, help='the instance name to use')
    parser.add_argument('-s', '--spot_price', type=float, help='the spot price to use')
    parser.add_argument('-b', '--bash_script', help='the bash script to use')
    parser.add_argument('-p', '--processes', type=int, default=1, help='the number of processes to run')
    parser.add_argument('-fc', '--frequency_channels', type=int, default=28, help='how many frequency channels per AWS instance')
    parser.add_argument('--force', action='store_true', default=False, help='proceed with a frequency band even if we already have it')

    parser.add_argument('obs_ids', nargs='+', help='the ids of the observation')

    args = vars(parser.parse_args())

    corrected_args = check_args(args)
    if corrected_args is None:
        LOGGER.error('The arguments are incorrect: {0}'.format(args))
    else:
        start_servers(
            args['processes'],
            corrected_args['ami_id'],
            corrected_args['user_data'],
            corrected_args['setup_disks'],
            args['instance_type'],
            corrected_args['obs_ids'],
            corrected_args['created_by'],
            args['name'],
            corrected_args['instance_details'],
            corrected_args['spot_price'],
            args['frequency_channels'],
            args['force'])

def main():
    parser = argparse.ArgumentParser("Start a number of CLEAN servers")
    parser.add_argument("-a", "--ami_id", help="the AMI id to use")
    parser.add_argument("-i", "--instance_type", required=True, help="the instance type to use")
    parser.add_argument("-c", "--created_by", help="the username to use")
    parser.add_argument("-n", "--name", required=True, help="the instance name to use")
    parser.add_argument("-s", "--spot_price", type=float, help="the spot price to use")
    parser.add_argument("-b", "--bash_script", help="the bash script to use")
    parser.add_argument("-p", "--processes", type=int, default=1, help="the number of processes to run")
    parser.add_argument("snapshots", nargs="+", help="the snapshots to use")

    args = vars(parser.parse_args())

    corrected_args = check_args(args)
    if corrected_args is None:
        LOGGER.error("The arguments are incorrect: {0}".format(args))
    else:
        start_server(
            args["processes"],
            corrected_args["ami_id"],
            corrected_args["user_data"],
            corrected_args["setup_disks"],
            args["instance_type"],
            args["snapshots"],
            corrected_args["created_by"],
            args["name"],
            corrected_args["instance_details"],
            corrected_args["spot_price"],
        )

def copy_files(frequency_id, processes, days):
    s3_helper = S3Helper()
    bucket = s3_helper.get_bucket(CHILES_BUCKET_NAME)
    LOGGER.info('Scanning bucket: {0}, frequency_id: {1}'.format(bucket, frequency_id))

    # Create the queue
    queue = multiprocessing.JoinableQueue()

    # Start the consumers
    for x in range(processes):
        consumer = Consumer(queue)
        consumer.start()

    for key in bucket.list(prefix='CVEL/{0}'.format(frequency_id)):
        LOGGER.info('Checking {0}'.format(key.key))
        # Ignore the key
        if key.key.endswith('/data.tar.gz') or key.key.endswith('/data.tar'):
            elements = key.key.split('/')
            if elements[2] in days:
                directory = '/mnt/output/Chiles/split_vis/{0}/'.format(elements[2])

                # Queue the copy of the file
                temp_file = os.path.join(directory, 'data.tar.gz' if key.key.endswith('/data.tar.gz') else 'data.tar')
                queue.put(Task(key, temp_file, directory, frequency_id))

    # Add a poison pill to shut things down
    for x in range(processes):
        queue.put(None)

    # Wait for the queue to terminate
    queue.join()

def copy_files(args):
    # Create the queue
    queue = multiprocessing.JoinableQueue()
    # Start the consumers
    for x in range(PROCESSES):
        consumer = Consumer(queue)
        consumer.start()

    # Look in the output directory
    for root, dir_names, filenames in os.walk(args.product_dir):
        LOGGER.debug('root: {0}, dir_names: {1}, filenames: {2}'.format(root, dir_names, filenames))
        for match in fnmatch.filter(dir_names, '13B-266*calibrated_deepfield.ms'):
            result_dir = join(root, match)
            LOGGER.info('Queuing result_dir: {0}'.format(result_dir))

            queue.put(
                CopyTask(
                    args.bucket,
                    match,
                    result_dir,
                    args.aws_access_key_id,
                    args.aws_secret_access_key
                )
            )

    # Add a poison pill to shut things down
    for x in range(PROCESSES):
        queue.put(None)

    # Wait for the queue to terminate
    queue.join()

def main():
    parser = argparse.ArgumentParser('Start a number of CLEAN servers')
    parser.add_argument('-a', '--ami_id', help='the AMI id to use')
    parser.add_argument('-i', '--instance_type', required=True, help='the instance type to use')
    parser.add_argument('-c', '--created_by', help='the username to use')
    parser.add_argument('-n', '--name', required=True, help='the instance name to use')
    parser.add_argument('-s', '--spot_price', type=float, help='the spot price to use')
    parser.add_argument('-b', '--bash_script', help='the bash script to use')
    parser.add_argument('-p', '--processes', type=int, default=1, help='the number of processes to run')
    parser.add_argument('frequencies', nargs='+', help='the frequencies to use (vis_14XX~14YY')

    args = vars(parser.parse_args())

    corrected_args = check_args(args)
    if corrected_args is None:
        LOGGER.error('The arguments are incorrect: {0}'.format(args))
    else:
        start_servers(
            args['processes'],
            corrected_args['ami_id'],
            corrected_args['user_data'],
            corrected_args['setup_disks'],
            args['instance_type'],
            args['frequencies'],
            corrected_args['created_by'],
            args['name'],
            corrected_args['instance_details'],
            corrected_args['spot_price'])

def start_servers(
        processes,
        ami_id,
        user_data,
        setup_disks,
        instance_type,
        obs_ids,
        created_by,
        name,
        instance_details,
        spot_price,
        frequency_channels,
        force):
    cvel_data = get_cvel()

    # Create the queue
    tasks = multiprocessing.JoinableQueue()

    # Start the consumers
    for x in range(processes):
        consumer = Consumer(tasks)
        consumer.start()

    counter = 1
    for obs_id in obs_ids:
        snapshot_id = OBS_IDS.get(obs_id)
        if snapshot_id is None:
            LOGGER.warning('The obs-id: {0} does not exist in the settings file')
        else:
            obs_id_dashes = obs_id.replace('_', '-')
            for frequency_groups in get_frequency_groups(frequency_channels, obs_id_dashes, cvel_data, force):
                tasks.put(
                    Task(
                        ami_id,
                        user_data,
                        setup_disks,
                        instance_type,
                        obs_id,
                        snapshot_id,
                        created_by,
                        name,
                        spot_price,
                        instance_details,
                        frequency_groups,
                        counter
                    )
                )
                counter += 1

        # Add a poison pill to shut things down
    for x in range(processes):
        tasks.put(None)

    # Wait for the queue to terminate
    tasks.join()

 def __call__(self):
     # noinspection PyBroadException
     try:
         LOGGER.info('Copying {0} to s3:{1}'.format(self._filename, self._bucket_location))
         s3_helper = S3Helper()
         s3_helper.add_file_to_bucket(
             CHILES_BUCKET_NAME,
             self._bucket_location,
             self._filename)
     except Exception:
         LOGGER.exception('CopyTask died')

def get_mime_encoded_user_data(volume_id, setup_disks, in_user_data, now):
    """
    AWS allows for a multipart m
    """
    user_data = MIMEMultipart()
    user_data.attach(get_cloud_init())

    data_formatted = in_user_data.format(volume_id, now, PIP_PACKAGES)
    LOGGER.info(data_formatted)
    user_data.attach(MIMEText(setup_disks + data_formatted))
    return user_data.as_string()

def copy_files(date, vis_file):
    s3_helper = S3Helper()
    # Look in the output directory
    for root, dir_names, filenames in os.walk(CHILES_CVEL_OUTPUT):
        LOGGER.info('root: {0}, dir_names: {1}, filenames: {2}'.format(root, dir_names, filenames))
        for match in fnmatch.filter(dir_names, vis_file):
            result_dir = join(root, match)
            LOGGER.info('Working on: {0}'.format(result_dir))

            if can_be_multipart_tar(result_dir):
                LOGGER.info('Using add_tar_to_bucket_multipart')
                s3_helper.add_tar_to_bucket_multipart(
                    CHILES_BUCKET_NAME,
                    'CVEL/{0}/{1}/data.tar'.format(vis_file, date),
                    result_dir)
            else:
                LOGGER.info('Using make_tarfile, then adding file to bucket')
                output_tar_filename = join(root, match + '.tar')
                make_tarfile(output_tar_filename, result_dir)

                s3_helper.add_file_to_bucket(
                    CHILES_BUCKET_NAME,
                    'CVEL/{0}/{1}/data.tar'.format(vis_file, date),
                    output_tar_filename)

                # Clean up
                os.remove(output_tar_filename)

            shutil.rmtree(result_dir, ignore_errors=True)