rucio 32.8.6__py3-none-any.whl

rucio/daemons/conveyor/submitter.py

@@ -0,0 +1,379 @@
+# -*- coding: utf-8 -*-
+# Copyright European Organization for Nuclear Research (CERN) since 2012
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""
+Conveyor transfer submitter is a daemon to manage non-tape file transfers.
+"""
+import logging
+import threading
+from collections.abc import Mapping
+from types import FrameType
+from typing import TYPE_CHECKING, Any, Optional
+
+import rucio.db.sqla.util
+from rucio.common import exception
+from rucio.common.config import config_get, config_get_bool, config_get_int, config_get_list, config_get_float
+from rucio.common.logging import setup_logging
+from rucio.common.schema import get_schema_value
+from rucio.common.stopwatch import Stopwatch
+from rucio.core.monitor import MetricManager
+from rucio.core.request import list_and_mark_transfer_requests_and_source_replicas
+from rucio.core.topology import Topology, ExpiringObjectCache
+from rucio.core.transfer import DEFAULT_MULTIHOP_TOMBSTONE_DELAY, list_transfer_admin_accounts, transfer_path_str, \
+    TRANSFERTOOL_CLASSES_BY_NAME, ProtocolFactory
+from rucio.daemons.common import db_workqueue, ProducerConsumerDaemon
+from rucio.daemons.conveyor.common import submit_transfer, get_conveyor_rses, pick_and_prepare_submission_path
+from rucio.db.sqla.constants import RequestType, RequestState
+from rucio.transfertool.fts3 import FTS3Transfertool
+from rucio.transfertool.globus import GlobusTransferTool
+
+if TYPE_CHECKING:
+    from rucio.daemons.common import HeartbeatHandler
+
+METRICS = MetricManager(module=__name__)
+GRACEFUL_STOP = threading.Event()
+DAEMON_NAME = 'conveyor-submitter'
+
+TRANSFER_TOOLS = config_get_list('conveyor', 'transfertool', False, [])  # NOTE: This should eventually be completely removed, as it can be fetched from the request
+FILTER_TRANSFERTOOL = config_get('conveyor', 'filter_transfertool', False, None)  # NOTE: TRANSFERTOOL to filter requests on
+TRANSFER_TYPE = config_get('conveyor', 'transfertype', False, 'single')
+
+
+def _fetch_requests(
+        partition_hash_var: Optional[str],
+        bulk: int,
+        activity: str,
+        rse_ids: Optional[list[str]],
+        request_type: list[RequestType],
+        ignore_availability: bool,
+        filter_transfertool: Optional[str],
+        metrics: MetricManager,
+        cached_topology,
+        set_last_processed_by: bool,
+        heartbeat_handler: "HeartbeatHandler",
+):
+    """
+    Fetches requests to be handled from the database
+    """
+    worker_number, total_workers, logger = heartbeat_handler.live()
+
+    topology = cached_topology.get() if cached_topology else Topology(ignore_availability=ignore_availability)
+    topology.configure_multihop(logger=logger)
+    stopwatch = Stopwatch()
+
+    required_source_rse_attrs = None
+    # if filter_transfertool specified, select only the source rses which are configured for this transfertool
+    if filter_transfertool:
+        # if multihop is configured, we want all possible source rses. To allow multi-hopping between transfertools
+        if not topology.multihop_enabled:
+            required_source_rse_attrs = TRANSFERTOOL_CLASSES_BY_NAME[filter_transfertool].required_rse_attrs
+
+    # retrieve (from the database) the transfer requests with their possible source replicas
+    requests_with_sources = list_and_mark_transfer_requests_and_source_replicas(
+        rse_collection=topology,
+        processed_by=heartbeat_handler.short_executable if set_last_processed_by else None,
+        total_workers=total_workers,
+        worker_number=worker_number,
+        partition_hash_var=partition_hash_var,
+        limit=bulk,
+        activity=activity,
+        older_than=None,
+        rses=rse_ids,
+        request_type=request_type,
+        request_state=RequestState.QUEUED,
+        ignore_availability=ignore_availability,
+        transfertool=filter_transfertool,
+        required_source_rse_attrs=required_source_rse_attrs,
+    )
+
+    stopwatch.stop()
+    total_transfers = len(requests_with_sources)
+
+    metrics.timer('get_transfers.time_per_transfer').observe(stopwatch.elapsed / (total_transfers or 1))
+    metrics.counter('get_transfers.total_transfers').inc(total_transfers)
+    logger(logging.INFO, 'Got %s transfers for %s in %s seconds', total_transfers, activity, stopwatch.elapsed)
+
+    must_sleep = False
+    if total_transfers < bulk:
+        must_sleep = True
+        logger(logging.DEBUG, 'Only %s transfers for %s which is less than bulk %s', total_transfers, activity, bulk)
+
+    return must_sleep, (topology, requests_with_sources)
+
+
+def _handle_requests(
+        batch,
+        *,
+        transfertools: list[str],
+        schemes: Optional[list[str]],
+        failover_schemes: Optional[list[str]],
+        max_sources: int,
+        timeout: Optional[float],
+        transfertool_kwargs,
+        metrics: MetricManager,
+        logger=logging.log,
+):
+    topology, requests_with_sources = batch
+
+    protocol_factory = ProtocolFactory()
+    default_tombstone_delay = config_get_int('transfers', 'multihop_tombstone_delay', default=DEFAULT_MULTIHOP_TOMBSTONE_DELAY, expiration_time=600)
+    admin_accounts = list_transfer_admin_accounts()
+
+    transfers = pick_and_prepare_submission_path(
+        requests_with_sources=requests_with_sources,
+        topology=topology,
+        protocol_factory=protocol_factory,
+        default_tombstone_delay=default_tombstone_delay,
+        admin_accounts=admin_accounts,
+        failover_schemes=failover_schemes,
+        schemes=schemes,
+        max_sources=max_sources,
+        transfertools=transfertools,
+        logger=logger,
+    )
+
+    for builder, transfer_paths in transfers.items():
+        # Globus Transfertool is not yet production-ready, but we need to partially activate it
+        # in all submitters if we want to enable native multi-hopping between transfertools.
+        # This "if" can be triggered in a FTS submitter if it tries to multi-hop from
+        # a globus-only RSE via a dual-stack RSE towards an FTS-only RSE.
+        #
+        # Just ignore this transfer and keep it in a queued state, so that it's picked up
+        # latter by that special submitter instance dedicated to globus transfers.
+        #
+        # TODO: remove this "if"
+        if transfertools[0] != GlobusTransferTool.external_name and builder.transfertool_class == GlobusTransferTool:
+            logger(logging.INFO, 'Skipping submission of following transfers: %s', [transfer_path_str(p) for p in transfer_paths])
+            continue
+
+        transfertool_obj = builder.make_transfertool(logger=logger, **transfertool_kwargs.get(builder.transfertool_class, {}))
+        logger(logging.DEBUG, 'Starting to group transfers %s', transfertool_obj)
+        stopwatch = Stopwatch()
+        grouped_jobs = transfertool_obj.group_into_submit_jobs(transfer_paths)
+        metrics.timer('bulk_group_transfer').observe(stopwatch.elapsed / (len(transfer_paths) or 1))
+
+        logger(logging.DEBUG, 'Starting to submit transfers for %s', transfertool_obj)
+        for job in grouped_jobs:
+            logger(logging.DEBUG, 'submitjob: transfers=%s, job_params=%s' % ([str(t) for t in job['transfers']], job['job_params']))
+            submit_transfer(transfertool_obj=transfertool_obj, transfers=job['transfers'], job_params=job['job_params'],
+                            timeout=timeout, logger=logger)
+
+
+def _get_max_time_in_queue_conf() -> dict[str, int]:
+    """
+    Retrieve and parse the max_time_in_queue configuration value into a dictionary: {"activity": int}
+    """
+    max_time_in_queue = {}
+    timelife_conf = config_get('conveyor', 'max_time_in_queue', default='', raise_exception=False)
+    if timelife_conf:
+        timelife_confs = timelife_conf.split(",")
+        for conf in timelife_confs:
+            act, timelife = conf.split(":")
+            max_time_in_queue[act.strip()] = int(timelife.strip())
+    if 'default' not in max_time_in_queue:
+        max_time_in_queue['default'] = 168
+    return max_time_in_queue
+
+
+def submitter(
+        once: bool = False,
+        rses: Optional[list[Mapping[str, Any]]] = None,
+        partition_wait_time: int = 10,
+        bulk: int = 100,
+        group_bulk: int = 1,
+        group_policy: str = 'rule',
+        source_strategy: Optional[str] = None,
+        activities: Optional[list[str]] = None,
+        sleep_time: int = 600,
+        max_sources: int = 4,
+        archive_timeout_override: Optional[int] = None,
+        filter_transfertool: Optional[str] = FILTER_TRANSFERTOOL,
+        transfertools: list[str] = TRANSFER_TOOLS,
+        transfertype: str = TRANSFER_TYPE,
+        ignore_availability: bool = False,
+        executable: str = DAEMON_NAME,
+        request_type: Optional[list[RequestType]] = None,
+        default_lifetime: int = 172800,
+        metrics: MetricManager = METRICS,
+        cached_topology=None,
+        total_threads: int = 1,
+):
+    """
+    Main loop to submit a new transfer primitive to a transfertool.
+    """
+
+    if not request_type:
+        request_type = [RequestType.TRANSFER]
+
+    partition_hash_var = config_get('conveyor', 'partition_hash_var', default=None, raise_exception=False)
+
+    schemes = config_get_list('conveyor', 'scheme', default=None, raise_exception=False)
+    failover_schemes = config_get_list('conveyor', 'failover_scheme', default=None, raise_exception=False)
+
+    timeout = config_get_float('conveyor', 'submit_timeout', default=None, raise_exception=False)
+
+    bring_online = config_get_int('conveyor', 'bring_online', default=43200, raise_exception=False)
+
+    max_time_in_queue = _get_max_time_in_queue_conf()
+    logging.debug("Maximum time in queue for different activities: %s", max_time_in_queue)
+
+    if activities:
+        activities.sort()
+        executable += '--activities ' + str(activities)
+    if filter_transfertool:
+        executable += ' --filter-transfertool ' + filter_transfertool
+    if rses:
+        rse_ids = [rse['id'] for rse in rses]
+    else:
+        rse_ids = None
+
+    transfertool_kwargs = {
+        FTS3Transfertool: {
+            'group_policy': group_policy,
+            'group_bulk': group_bulk,
+            'source_strategy': source_strategy,
+            'max_time_in_queue': max_time_in_queue,
+            'bring_online': bring_online,
+            'default_lifetime': default_lifetime,
+            'archive_timeout_override': archive_timeout_override,
+        },
+        GlobusTransferTool: {
+            'group_policy': transfertype,
+            'group_bulk': group_bulk,
+        },
+    }
+
+    @db_workqueue(
+        once=once,
+        graceful_stop=GRACEFUL_STOP,
+        executable=executable,
+        partition_wait_time=partition_wait_time,
+        sleep_time=sleep_time,
+        activities=activities)
+    def _db_producer(*, activity, heartbeat_handler):
+        return _fetch_requests(
+            bulk=bulk,
+            filter_transfertool=filter_transfertool,
+            ignore_availability=ignore_availability,
+            partition_hash_var=partition_hash_var,
+            rse_ids=rse_ids,
+            request_type=request_type,
+            metrics=metrics,
+            activity=activity,
+            cached_topology=cached_topology,
+            set_last_processed_by=not once,
+            heartbeat_handler=heartbeat_handler,
+        )
+
+    def _consumer(batch):
+        return _handle_requests(
+            batch,
+            transfertools=transfertools,
+            schemes=schemes,
+            failover_schemes=failover_schemes,
+            max_sources=max_sources,
+            timeout=timeout,
+            transfertool_kwargs=transfertool_kwargs,
+            metrics=metrics,
+        )
+
+    ProducerConsumerDaemon(
+        producers=[_db_producer],
+        consumers=[_consumer for _ in range(total_threads)],
+        graceful_stop=GRACEFUL_STOP,
+    ).run()
+
+
+def stop(signum: "Optional[int]" = None, frame: "Optional[FrameType]" = None) -> None:
+    """
+    Graceful exit.
+    """
+    GRACEFUL_STOP.set()
+
+
+def run(
+        once=False,
+        group_bulk=1,
+        group_policy='rule',
+        rses=None,
+        include_rses=None,
+        exclude_rses=None,
+        vos=None,
+        bulk=100,
+        source_strategy=None,
+        activities=None,
+        exclude_activities=None,
+        ignore_availability=False,
+        sleep_time=600,
+        max_sources=4,
+        archive_timeout_override=None,
+        total_threads=1,
+        **_kwargs
+):
+    """
+    Starts up the conveyer threads.
+    """
+    setup_logging(process_name=DAEMON_NAME)
+
+    if rucio.db.sqla.util.is_old_db():
+        raise exception.DatabaseException('Database was not updated, daemon won\'t start')
+
+    multi_vo = config_get_bool('common', 'multi_vo', raise_exception=False, default=False)
+    working_rses = None
+    if rses or include_rses or exclude_rses:
+        working_rses = get_conveyor_rses(rses, include_rses, exclude_rses, vos)
+        logging.info("RSE selection: RSEs: %s, Include: %s, Exclude: %s", rses, include_rses, exclude_rses)
+    elif multi_vo:
+        working_rses = get_conveyor_rses(rses, include_rses, exclude_rses, vos)
+        logging.info("RSE selection: automatic for relevant VOs")
+    else:
+        logging.info("RSE selection: automatic")
+
+    logging.info('starting submitter threads')
+
+    if exclude_activities:
+        if not activities:
+            if not multi_vo:
+                vos = ['def']
+            if vos and len(vos) == 1:
+                activities = get_schema_value('ACTIVITY', vos[0])
+            elif vos and len(vos) > 1:
+                logging.warning('Cannot get activity list from schema when multiple VOs given, either provide `activities` argument or run on a single VO')
+                activities = None
+            else:
+                logging.warning('Cannot get activity list from schema when no VO given, either provide `activities` argument or `vos` with a single entry')
+                activities = None
+
+        if activities is not None:
+            for activity in exclude_activities:
+                if activity in activities:
+                    activities.remove(activity)
+
+    cached_topology = ExpiringObjectCache(ttl=300, new_obj_fnc=lambda: Topology(ignore_availability=ignore_availability))
+    submitter(
+        once=once,
+        rses=working_rses,
+        bulk=bulk,
+        group_bulk=group_bulk,
+        group_policy=group_policy,
+        activities=activities,
+        ignore_availability=ignore_availability,
+        sleep_time=sleep_time,
+        max_sources=max_sources,
+        source_strategy=source_strategy,
+        archive_timeout_override=archive_timeout_override,
+        cached_topology=cached_topology,
+        total_threads=total_threads,
+    )