@must_be_writable
@synchronized
@retry_method
- def save(self, merge=True, num_retries=None):
+ def save(self, storage_classes=None, merge=True, num_retries=None):
"""Save collection to an existing collection record.
Commit pending buffer blocks to Keep, merge with remote record (if
self.update()
text = self.manifest_text(strip=False)
+ body={'manifest_text': text}
+ if storage_classes:
+ body["storage_classes_desired"] = storage_classes
+
self._remember_api_response(self._my_api().collections().update(
uuid=self._manifest_locator,
- body={'manifest_text': text}
+ body=body
).execute(
num_retries=num_retries))
self._manifest_text = self._api_response["manifest_text"]
def save_new(self, name=None,
create_collection_record=True,
owner_uuid=None,
+ storage_classes=None,
ensure_unique_name=False,
num_retries=None):
"""Save collection to a new collection record.
"replication_desired": self.replication_desired}
if owner_uuid:
body["owner_uuid"] = owner_uuid
+ if storage_classes:
+ body["storage_classes_desired"] = storage_classes
self._remember_api_response(self._my_api().collections().create(ensure_unique_name=ensure_unique_name, body=body).execute(num_retries=num_retries))
text = self._api_response["manifest_text"]
block. Default is to use the server-provided default (if any) or 2.
""")
+upload_opts.add_argument('--storage-classes', help="""
+Specify comma separated list of storage classes to be used when saving data to Keep.
+""")
+
upload_opts.add_argument('--threads', type=int, metavar='N', default=None,
help="""
Set the number of upload threads to be used. Take into account that
def __init__(self, paths, resume=True, use_cache=True, reporter=None,
name=None, owner_uuid=None, api_client=None,
ensure_unique_name=False, num_retries=None,
- put_threads=None, replication_desired=None,
- filename=None, update_time=60.0, update_collection=None,
+ put_threads=None, replication_desired=None, filename=None,
+ update_time=60.0, update_collection=None, storage_classes=None,
logger=logging.getLogger('arvados.arv_put'), dry_run=False,
follow_links=True, exclude_paths=[], exclude_names=None):
self.paths = paths
self.replication_desired = replication_desired
self.put_threads = put_threads
self.filename = filename
+ self.storage_classes = storage_classes
self._api_client = api_client
self._state_lock = threading.Lock()
self._state = None # Previous run state (file list & manifest)
else:
# The file already exist on remote collection, skip it.
pass
- self._remote_collection.save(num_retries=self.num_retries)
+ self._remote_collection.save(storage_classes=self.storage_classes,
+ num_retries=self.num_retries)
else:
+ if self.storage_classes is None:
+ self.storage_classes = ['default']
self._local_collection.save_new(
name=self.name, owner_uuid=self.owner_uuid,
+ storage_classes=self.storage_classes,
ensure_unique_name=self.ensure_unique_name,
num_retries=self.num_retries)
else:
reporter = None
+ # Split storage-classes argument
+ storage_classes = None
+ if args.storage_classes:
+ storage_classes = args.storage_classes.strip().split(',')
+ if len(storage_classes) > 1:
+ logger.error("Multiple storage classes are not supported currently.")
+ sys.exit(1)
+
+
# Setup exclude regex from all the --exclude arguments provided
name_patterns = []
exclude_paths = []
owner_uuid = project_uuid,
ensure_unique_name = True,
update_collection = args.update_collection,
+ storage_classes=storage_classes,
logger=logger,
dry_run=args.dry_run,
follow_links=args.follow_links,