# Copyright (c) 2019-2020, IRIS-HEP
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are met:
#
# * Redistributions of source code must retain the above copyright notice, this
# list of conditions and the following disclaimer.
#
# * Redistributions in binary form must reproduce the above copyright notice,
# this list of conditions and the following disclaimer in the documentation
# and/or other materials provided with the distribution.
#
# * Neither the name of the copyright holder nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
# AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
# DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
# SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
# CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
# OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
import asyncio
import urllib.parse
from concurrent.futures.thread import ThreadPoolExecutor
from pathlib import Path
from typing import Any, AsyncIterator, List, Optional, Dict, cast
import logging
import backoff
from backoff import on_exception
from minio import Minio, ResponseError
from minio.error import NoSuchBucket
from urllib3.exceptions import MaxRetryError
from .utils import ServiceXException, ServiceXNoFilesInCache
[docs]class MinioAdaptor:
# Threadpool on which downloads occur. This is because the current minio library
# uses blocking http requests, so we can't use asyncio to interleave them.
_download_executor = ThreadPoolExecutor(max_workers=5)
[docs] def __init__(
self,
minio_endpoint: str,
secured: bool = False,
access_key: str = "miniouser",
secretkey: str = "leftfoot1",
):
self._endpoint = minio_endpoint
self._secured = secured
self._access_key = access_key
self._secretkey = secretkey
self._client = Minio(
self._endpoint,
access_key=self._access_key,
secret_key=self._secretkey,
secure=self._secured,
)
[docs] @on_exception(backoff.constant, ResponseError, interval=0.1)
def get_files(self, request_id) -> List[str]:
try:
return [str(f.object_name) for f in self._client.list_objects(request_id)]
except MaxRetryError as e:
msg = (
f"Unable to reach Minio at {e.pool.host}:{e.pool.port}{e.url}. "
f"Max retries exceeded."
)
raise ServiceXException(msg)
except NoSuchBucket as e:
raise ServiceXNoFilesInCache(
f'No files in cache for request "{request_id}"'
) from e
[docs] def get_access_url(self, request_id: str, object_name: str) -> str:
"""Given a request ID and the file name in that request id, return a URL
that can be directly accessed to download the file.
Args:
request_id (str): The request id guid (that is the bucket in minio)
object_name (str): The file (the object in the minio bucket)
Returns:
str: A url good for some amount of time to access the bucket.
"""
return self._client.presigned_get_object(request_id, object_name)
[docs] def get_s3_uri(self, request_id: str, object_name: str) -> str:
"""Given a request ID and a file name in that request, return an S3 URI that can
be used by a correctly configured S3 client.
Args:
:param request_id: The request id guid (that is the bucket in minio)
:param object_name: The file (the object in the minio bucket)
:return:
A formatted S3 URI to that object
"""
return urllib.parse.urlunsplit(("s3", request_id, object_name, None, None))
[docs] async def download_file(
self, request_id: str, bucket_fname: str, output_file: Path
) -> None:
"""
Download a single file to a local temp file.
Arguments:
minio_client Open and authenticated minio client
request_id The id of the request we are going after
bucket_fname The fname of the bucket
output_file Filename where we should write this file.
Notes:
- Download to a temp file that is renamed at the end so that a partially
downloaded file is not mistaken as a full one
- Run with async, despite minio not being async.
"""
# Make sure the output directory exists
output_file.parent.mkdir(parents=True, exist_ok=True)
# We are going to build a temp file, and download it from there.
def do_copy() -> None:
temp_file = output_file.parent / f"{output_file.name}.temp"
try:
log = logging.getLogger(__name__)
log.debug(
f"Starting download of minio file for request {request_id}, "
f"file {bucket_fname}."
)
self._client.fget_object(request_id, bucket_fname, str(temp_file))
log.debug(
f"Finished download of minio file for request {request_id}, "
f"file {bucket_fname}."
)
temp_file.rename(output_file)
except Exception as e:
raise ServiceXException(
f"Failed to copy minio bucket {bucket_fname} from request "
f"{request_id} to {output_file}"
) from e
# If the file exists, we don't need to do anything.
if output_file.exists():
return
# Do the copy, which might take a while, on a separate thread.
return await asyncio.wrap_future(self._download_executor.submit(do_copy))
[docs]class MinioAdaptorFactory:
"""A factor that will return, when asked, the proper minio adaptor to use for a request
to get files from ServiceX.
"""
[docs] def __init__(self, always_return: Optional[MinioAdaptor] = None):
"""Create the factor with a possible adaptor to always use
Args:
always_return (Optional[MinioAdaptor], optional): The adaptor to always use. If none,
then one will be created by the best other available method. Defaults to None.
"""
# Get the defaults setup.
self._always = always_return
self._config_adaptor = None
[docs] def from_best(
self, transaction_info: Optional[Dict[str, str]] = None
) -> MinioAdaptor:
"""Using the information we have, create the proper Minio Adaptor with the correct
endpoint and login information. Order of adaptor generation:
1. The `always_return` option from the ctor
1. Use info from the `translation_info`
1. Use the info from the config handed in to the ctor
Raises:
Exception: If we do not have enough information to create an adaptor, this is raised.
Returns:
MinioAdaptor: The adaptor that can be used to extract the data from minio for this
request.
"""
if self._always is not None:
logging.getLogger(__name__).debug("Using the pre-defined minio_adaptor")
return self._always
if transaction_info is not None:
keys = [
"minio-endpoint",
"minio-secured",
"minio-access-key",
"minio-secret-key",
]
if all(k in transaction_info for k in keys):
logging.getLogger(__name__).debug(
"Using the request-specific minio_adaptor"
)
return MinioAdaptor(
transaction_info["minio-endpoint"],
bool(transaction_info["minio-secured"]),
transaction_info["minio-access-key"],
transaction_info["minio-secret-key"],
)
raise ServiceXException(
"Do not know or have enough information to create a Minio "
f"Login info ({transaction_info})"
)
[docs]async def find_new_bucket_files(
adaptor: MinioAdaptor, request_id: str, update: AsyncIterator[Any]
) -> AsyncIterator[str]:
"""
Each time we get something from the async iterator, check to see if
there are any files present.
"""
seen = []
async for _ in update:
# Sadly, this is blocking, and so may hold things up
files = cast(List[str], adaptor.get_files(request_id))
# If there are new files, pass them on
for f in files:
if f not in seen:
seen.append(f)
yield f