Differential D23 Diff 64 swh/storage/archiver/copier.py

Changeset View

Standalone View

swh/storage/archiver/copier.py

This file was added.

				# Copyright (C) 2015 The Software Heritage developers
				# See the AUTHORS file at the top-level directory of this distribution
				# License: GNU General Public License version 3, or any later version
				# See top-level LICENSE file for more information

				from swh.core import hashutil
				from ..objstorage.api.client import RemoteObjStorage


				class ArchiverCopier():
				""" This archiver copy some files into a remote objstorage
				in order to get a backup.

				Attributes:
				content_ids: A list of sha1 that represents the content this copier
				have to archive.
				ardumontUnsubmitted Done Inline Actions a list of sha1`s` this copier `has` ardumont: a list of sha1`s` this copier `has`
				server (RemoteArchive): The remote object storage that is used to
				backup content.
				master_storage (Storage): The master storage that contains the data
				the copier needs to archive.
				"""
				def __init__(self, destination, content, master_storage):
				""" Create a Copier for the archiver

				Args:
				destination: A tuple (archive_name, archive_url) that represents a
				remote object storage as in the 'archives' table.
				content: A list of sha1 that represents the content this copier
				have to archive.
				master_storage (Storage): The master storage of the system that
				contains the data to archive.
				"""
				_name, self.url = destination
				self.content_ids = content
				self.server = RemoteObjStorage(self.url)
				self.master_storage = master_storage

				def run(self):
				""" Do the copy on the backup storage.

				Run the archiver copier in order to copy the required content
				into the current destination.
				The content which corresponds to the sha1 in self.content_ids
				will be fetched from the master_storage and then copied into
				the backup object storage.

				Returns:
				A boolean that indicates if the whole content have been copied.
				"""
				self.content_ids = list(map(lambda x: hashutil.hex_to_hash(x[2:]),
				ardumontUnsubmitted Done Inline Actions I think it might work without needing to consume the map. Can you refresh my memory, why do you need to strip the first 2 characters (is it '\\x')? I think this might be due to how you test. ardumont: I think it might work without needing to consume the map. Can you refresh my memory, why do…
				self.content_ids))
				contents = self.master_storage.content_get(self.content_ids)
				try:
				for content in contents:
				content_data = content['data']
				self.server.content_add(content_data)
				ardumontUnsubmitted Done Inline Actions self.server.content_add(map(lambda c: c['data'], contents)) ? ardumont: ``` self.server.content_add(map(lambda c: c['data'], contents)) ``` ?
				qcamposAuthorUnsubmitted Done Inline Actions Just checked : `RemoteObjStorage::content_add` is linked to `ObjStorage::add_bytes` that only take a single content. qcampos: Just checked : `RemoteObjStorage::content_add` is linked to `ObjStorage::add_bytes` that only…
				except:
				# In case there is an error during the process, indicates that the
				ardumontUnsubmitted Done Inline Actions You may want to avoid writing your doubts in the code and use differential for that (as you also did below). I like to use, TODO (you'll need to think more and improve the code), FIXME (you saw something ugly but this is not the time to fix it) , HACK (you had to do something horribly ugly but did not see any other way around it) followed by some concise description about what this is all about. ardumont: You may want to avoid writing your doubts in the code and use differential for that (as you…
				# work may not have been done entierly.
				# TODO In case something happened during the process, some files
				# may have been copied and other not. What should be the policy?
				# consider the whole batch as undone or valdiate the already
				# copied content?
				# i.e. : Is it ok to make a database access here? Should it go
				ardumontUnsubmitted Done Inline Actions Well, it depends on how you see the asynchroneous part work... If archive-director spawns asynchroneous archive-workers which also spawns asynchroneous archive-copiers, it could be simpler to do what you propose (database access here)... But then the abstraction around the content copy state is broken, meaning we have now multiple classes which are in charge of it (first the worker initiates something that the copier must update and close). Or we could have a more complex design where the copier also states that `woohoo i'm done` (by queueing another message)... But then it's complicated... But if the async activity stops at archive-workers meaning only archive-directory spawns async archive-workers. Then what is now is ok (and simpler). ardumont: Well, it depends on how you see the asynchroneous part work... If archive-director spawns…
				qcamposAuthorUnsubmitted Done Inline Actions Now that I think about it , the specification is kinda blurry about what precisely is asynchronous. But I understood it as Director spawning async workers that execute copier's code. qcampos: Now that I think about it , the specification is kinda blurry about what precisely is…
				# up to the worker?
				return False
				qcamposAuthorUnsubmitted Not Done Inline Actions If that there is an error during the process, some files may have been copied. What should be the best option? Mark them as 'present', or just ignore them and reschedule the whole batch after the 'ongoing' maximum delay is elapsed? qcampos: If that there is an error during the process, some files may have been copied. What should be…
				ardumontUnsubmitted Not Done Inline Actions As a first approximation, considering nothing is done is reasonable. (If nothing prevents later to add again some existing content) ardumont: As a first approximation, considering nothing is done is reasonable. (If nothing prevents later…
				ardumontUnsubmitted Not Done Inline Actions (If nothing prevents later to add again some existing content) What i meant was, as long as you can write and overwrite existing contents again, i think it's ok. ardumont: > (If nothing prevents later to add again some existing content) What i meant was, as long as…
				qcamposAuthorUnsubmitted Not Done Inline Actions Shouldn't create any problem (ATM) as the object storage just write files without checking if they already exists. qcampos: Shouldn't create any problem (ATM) as the object storage just write files without checking if…

				return True