From: Brian Warner Date: Fri, 30 Mar 2007 23:50:50 +0000 (-0700) Subject: finish making the new encoder/decoder/upload/download work X-Git-Url: https://git.rkrishnan.org/readonly?a=commitdiff_plain;h=7cd9ef3bbf5fa87c661433273ab6177648f36886;p=tahoe-lafs%2Ftahoe-lafs.git finish making the new encoder/decoder/upload/download work --- diff --git a/src/allmydata/download.py b/src/allmydata/download.py index bafce789..eacc9eff 100644 --- a/src/allmydata/download.py +++ b/src/allmydata/download.py @@ -5,8 +5,7 @@ from twisted.python import log from twisted.internet import defer from twisted.application import service -from allmydata.util import idlib, bencode, mathutil -from allmydata.util.deferredutil import DeferredListShouldSucceed +from allmydata.util import idlib, mathutil from allmydata.util.assertutil import _assert from allmydata import codec from allmydata.Crypto.Cipher import AES @@ -48,6 +47,7 @@ class Output: def finish(self): return self.downloadable.finish() + class BlockDownloader: def __init__(self, bucket, blocknum, parent): self.bucket = bucket @@ -131,18 +131,23 @@ class FileDownloader: def __init__(self, client, uri, downloadable): self._client = client self._downloadable = downloadable - (codec_name, codec_params, verifierid, roothash, needed_shares, total_shares, size, segment_size) = unpack_uri(uri) + (codec_name, codec_params, tail_codec_params, verifierid, roothash, needed_shares, total_shares, size, segment_size) = unpack_uri(uri) assert isinstance(verifierid, str) assert len(verifierid) == 20 self._verifierid = verifierid self._roothash = roothash - self._decoder = codec.get_decoder_by_name(codec_name) - self._decoder.set_serialized_params(codec_params) + + self._codec = codec.get_decoder_by_name(codec_name) + self._codec.set_serialized_params(codec_params) + self._tail_codec = codec.get_decoder_by_name(codec_name) + self._tail_codec.set_serialized_params(tail_codec_params) + + self._total_segments = mathutil.div_ceil(size, segment_size) self._current_segnum = 0 self._segment_size = segment_size self._size = size - self._num_needed_shares = self._decoder.get_needed_shares() + self._num_needed_shares = self._codec.get_needed_shares() key = "\x00" * 16 self._output = Output(downloadable, key) @@ -173,14 +178,16 @@ class FileDownloader: def _get_all_shareholders(self): dl = [] for (permutedpeerid, peerid, connection) in self._client.get_permuted_peers(self._verifierid): - d = connection.callRemote("get_buckets", self._verifierid) + d = connection.callRemote("get_service", "storageserver") + d.addCallback(lambda ss: ss.callRemote("get_buckets", + self._verifierid)) d.addCallbacks(self._got_response, self._got_error, callbackArgs=(connection,)) dl.append(d) return defer.DeferredList(dl) def _got_response(self, buckets, connection): - for sharenum, bucket in buckets: + for sharenum, bucket in buckets.iteritems(): self._share_buckets.setdefault(sharenum, set()).add(bucket) def _got_error(self, f): @@ -193,30 +200,34 @@ class FileDownloader: self.active_buckets = {} self._output.open() - def _download_all_segments(self): - d = self._download_segment(self._current_segnum) + def _download_all_segments(self, res): + d = defer.succeed(None) + for segnum in range(self._total_segments-1): + d.addCallback(self._download_segment, segnum) + d.addCallback(self._download_tail_segment, self._total_segments-1) + return d + + def _download_segment(self, res, segnum): + segmentdler = SegmentDownloader(self, segnum, self._num_needed_shares) + d = segmentdler.start() + d.addCallback(lambda (shares, shareids): + self._codec.decode(shares, shareids)) def _done(res): - if self._current_segnum == self._total_segments: - return None - return self._download_segment(self._current_segnum) + for buf in res: + self._output.write(buf) d.addCallback(_done) return d - def _download_segment(self, segnum): + def _download_tail_segment(self, res, segnum): segmentdler = SegmentDownloader(self, segnum, self._num_needed_shares) d = segmentdler.start() d.addCallback(lambda (shares, shareids): - self._decoder.decode(shares, shareids)) + self._tail_codec.decode(shares, shareids)) def _done(res): - self._current_segnum += 1 - if self._current_segnum == self._total_segments: - data = ''.join(res) - padsize = mathutil.pad_size(self._size, self._segment_size) - data = data[:-padsize] - self._output.write(data) - else: - for buf in res: - self._output.write(buf) + # trim off any padding added by the upload side + data = ''.join(res) + tail_size = self._size % self._segment_size + self._output.write(data[:tail_size]) d.addCallback(_done) return d @@ -302,8 +313,7 @@ class Downloader(service.MultiService): t = IDownloadTarget(t) assert t.write assert t.close - dl = FileDownloader(self.parent, uri) - dl.set_download_target(t) + dl = FileDownloader(self.parent, uri, t) if self.debug: dl.debug = True d = dl.start() diff --git a/src/allmydata/encode_new.py b/src/allmydata/encode_new.py index 5d0458d7..dc218f18 100644 --- a/src/allmydata/encode_new.py +++ b/src/allmydata/encode_new.py @@ -5,6 +5,7 @@ from twisted.internet import defer from allmydata.chunk import HashTree, roundup_pow2 from allmydata.Crypto.Cipher import AES from allmydata.util import mathutil, hashutil +from allmydata.util.assertutil import _assert from allmydata.codec import CRSEncoder from allmydata.interfaces import IEncoder @@ -88,12 +89,32 @@ class Encoder(object): self.required_shares = self.NEEDED_SHARES self.segment_size = min(2*MiB, self.file_size) + # this must be a multiple of self.required_shares + self.segment_size = mathutil.next_multiple(self.segment_size, + self.required_shares) self.setup_codec() def setup_codec(self): + assert self.segment_size % self.required_shares == 0 self._codec = CRSEncoder() - self._codec.set_params(self.segment_size, self.required_shares, - self.num_shares) + self._codec.set_params(self.segment_size, + self.required_shares, self.num_shares) + + # the "tail" is the last segment. This segment may or may not be + # shorter than all other segments. We use the "tail codec" to handle + # it. If the tail is short, we use a different codec instance. In + # addition, the tail codec must be fed data which has been padded out + # to the right size. + self.tail_size = self.file_size % self.segment_size + if not self.tail_size: + self.tail_size = self.segment_size + + # the tail codec is responsible for encoding tail_size bytes + padded_tail_size = mathutil.next_multiple(self.tail_size, + self.required_shares) + self._tail_codec = CRSEncoder() + self._tail_codec.set_params(padded_tail_size, + self.required_shares, self.num_shares) def get_share_size(self): share_size = mathutil.div_ceil(self.file_size, self.required_shares) @@ -105,6 +126,11 @@ class Encoder(object): return self._codec.get_block_size() def set_shareholders(self, landlords): + assert isinstance(landlords, dict) + for k in landlords: + # it would be nice to: + #assert RIBucketWriter.providedBy(landlords[k]) + pass self.landlords = landlords.copy() def start(self): @@ -116,8 +142,11 @@ class Encoder(object): self.setup_encryption() self.setup_codec() d = defer.succeed(None) - for i in range(self.num_segments): + + for i in range(self.num_segments-1): d.addCallback(lambda res: self.do_segment(i)) + d.addCallback(lambda res: self.do_tail_segment(self.num_segments-1)) + d.addCallback(lambda res: self.send_all_subshare_hash_trees()) d.addCallback(lambda res: self.send_all_share_hash_trees()) d.addCallback(lambda res: self.close_all_shareholders()) @@ -137,13 +166,14 @@ class Encoder(object): def do_segment(self, segnum): chunks = [] + codec = self._codec # the ICodecEncoder API wants to receive a total of self.segment_size # bytes on each encode() call, broken up into a number of # identically-sized pieces. Due to the way the codec algorithm works, # these pieces need to be the same size as the share which the codec # will generate. Therefore we must feed it with input_piece_size that # equals the output share size. - input_piece_size = self._codec.get_block_size() + input_piece_size = codec.get_block_size() # as a result, the number of input pieces per encode() call will be # equal to the number of required shares with which the codec was @@ -153,6 +183,21 @@ class Encoder(object): # of additional shares which can be substituted if the primary ones # are unavailable + for i in range(self.required_shares): + input_piece = self.infile.read(input_piece_size) + # non-tail segments should be the full segment size + assert len(input_piece) == input_piece_size + encrypted_piece = self.cryptor.encrypt(input_piece) + chunks.append(encrypted_piece) + d = codec.encode(chunks) + d.addCallback(self._encoded_segment, segnum) + return d + + def do_tail_segment(self, segnum): + chunks = [] + codec = self._tail_codec + input_piece_size = codec.get_block_size() + for i in range(self.required_shares): input_piece = self.infile.read(input_piece_size) if len(input_piece) < input_piece_size: @@ -160,20 +205,21 @@ class Encoder(object): input_piece += ('\x00' * (input_piece_size - len(input_piece))) encrypted_piece = self.cryptor.encrypt(input_piece) chunks.append(encrypted_piece) - d = self._codec.encode(chunks) - d.addCallback(self._encoded_segment) + d = codec.encode(chunks) + d.addCallback(self._encoded_segment, segnum) return d - def _encoded_segment(self, (shares, shareids)): + def _encoded_segment(self, (shares, shareids), segnum): + _assert(set(shareids) == set(self.landlords.keys()), + shareids=shareids, landlords=self.landlords) dl = [] for i in range(len(shares)): subshare = shares[i] shareid = shareids[i] - d = self.send_subshare(shareid, self.segment_num, subshare) + d = self.send_subshare(shareid, segnum, subshare) dl.append(d) subshare_hash = hashutil.tagged_hash("encoded subshare", subshare) self.subshare_hashes[shareid].append(subshare_hash) - self.segment_num += 1 return defer.DeferredList(dl) def send_subshare(self, shareid, segment_num, subshare): diff --git a/src/allmydata/interfaces.py b/src/allmydata/interfaces.py index cc235779..4cb69ba5 100644 --- a/src/allmydata/interfaces.py +++ b/src/allmydata/interfaces.py @@ -9,7 +9,7 @@ Hash = StringConstraint(HASH_SIZE) # binary format 32-byte SHA256 hash Nodeid = StringConstraint(20) # binary format 20-byte SHA1 hash PBURL = StringConstraint(150) Verifierid = StringConstraint(20) -URI = StringConstraint(100) # kind of arbitrary +URI = StringConstraint(200) # kind of arbitrary ShareData = StringConstraint(100000) # these six are here because Foolscap does not yet support the kind of # restriction I really want to apply to these. @@ -38,6 +38,9 @@ class RIClient(RemoteInterface): class RIBucketWriter(RemoteInterface): def put_block(segmentnum=int, data=ShareData): + """@param data: For most segments, this data will be 'blocksize' + bytes in length. The last segment might be shorter. + """ return None def put_block_hashes(blockhashes=ListOf(Hash)): @@ -68,6 +71,9 @@ class RIStorageServer(RemoteInterface): class RIBucketReader(RemoteInterface): def get_block(blocknum=int): + """Most blocks will be the same size. The last block might be shorter + than the others. + """ return ShareData def get_block_hashes(): return ListOf(Hash) diff --git a/src/allmydata/storageserver.py b/src/allmydata/storageserver.py index bb9622d1..23d3ece6 100644 --- a/src/allmydata/storageserver.py +++ b/src/allmydata/storageserver.py @@ -29,6 +29,8 @@ class BucketWriter(Referenceable): self.finalhome = finalhome self.blocksize = blocksize self.closed = False + fileutil.make_dirs(incominghome) + fileutil.make_dirs(finalhome) self._write_file('blocksize', str(blocksize)) def _write_file(self, fname, data): @@ -51,7 +53,7 @@ class BucketWriter(Referenceable): precondition(not self.closed) self._write_file('sharehashree', bencode.bencode(sharehashes)) - def close(self): + def remote_close(self): precondition(not self.closed) # TODO assert or check the completeness and consistency of the data that has been written fileutil.rename(self.incominghome, self.finalhome) @@ -72,7 +74,7 @@ class BucketReader(Referenceable): def remote_get_block(self, blocknum): f = open(os.path.join(self.home, 'data'), 'rb') f.seek(self.blocksize * blocknum) - return f.read(self.blocksize) + return f.read(self.blocksize) # this might be short for the last block def remote_get_block_hashes(self): return str2l(self._read_file('blockhashes')) @@ -101,8 +103,9 @@ class StorageServer(service.MultiService, Referenceable): alreadygot = set() bucketwriters = {} # k: shnum, v: BucketWriter for shnum in sharenums: - incominghome = os.path.join(self.incomingdir, idlib.a2b(verifierid), "%d"%shnum) - finalhome = os.path.join(self.storedir, idlib.a2b(verifierid), "%d"%shnum) + incominghome = os.path.join(self.incomingdir, + idlib.b2a(verifierid) + "%d"%shnum) + finalhome = os.path.join(self.storedir, idlib.b2a(verifierid), "%d"%shnum) if os.path.exists(incominghome) or os.path.exists(finalhome): alreadygot.add(shnum) else: diff --git a/src/allmydata/test/test_encode.py b/src/allmydata/test/test_encode.py index d3311ac9..e0383759 100644 --- a/src/allmydata/test/test_encode.py +++ b/src/allmydata/test/test_encode.py @@ -2,6 +2,7 @@ from twisted.trial import unittest from twisted.internet import defer +from foolscap import eventual from allmydata import encode_new, download from allmydata.uri import pack_uri from cStringIO import StringIO @@ -14,15 +15,39 @@ class MyEncoder(encode_new.Encoder): print " ", [i for i,h in args[0]] return defer.succeed(None) -class Encode(unittest.TestCase): - def test_1(self): - e = MyEncoder() - data = StringIO("some data to encode\n") - e.setup(data) - d = e.start() +class FakePeer: + def __init__(self, mode="good"): + self.ss = FakeStorageServer(mode) + + def callRemote(self, methname, *args, **kwargs): + def _call(): + meth = getattr(self, methname) + return meth(*args, **kwargs) + return defer.maybeDeferred(_call) + + def get_service(self, sname): + assert sname == "storageserver" + return self.ss + +class FakeStorageServer: + def __init__(self, mode): + self.mode = mode + def callRemote(self, methname, *args, **kwargs): + def _call(): + meth = getattr(self, methname) + return meth(*args, **kwargs) + d = eventual.fireEventually() + d.addCallback(lambda res: _call()) return d + def allocate_buckets(self, verifierid, sharenums, shareize, blocksize, canary): + if self.mode == "full": + return (set(), {},) + elif self.mode == "already got them": + return (set(sharenums), {},) + else: + return (set(), dict([(shnum, FakeBucketWriter(),) for shnum in sharenums]),) -class FakePeer: +class FakeBucketWriter: def __init__(self): self.blocks = {} self.block_hashes = None @@ -65,7 +90,7 @@ class FakePeer: return self.share_hashes -class UpDown(unittest.TestCase): +class Encode(unittest.TestCase): def test_send(self): e = encode_new.Encoder() data = "happy happy joy joy" * 4 @@ -79,7 +104,7 @@ class UpDown(unittest.TestCase): shareholders = {} all_shareholders = [] for shnum in range(NUM_SHARES): - peer = FakePeer() + peer = FakeBucketWriter() shareholders[shnum] = peer all_shareholders.append(peer) e.set_shareholders(shareholders) @@ -104,20 +129,24 @@ class UpDown(unittest.TestCase): return d - def test_send_and_recover(self): +class Roundtrip(unittest.TestCase): + def send_and_recover(self, NUM_SHARES, NUM_PEERS, NUM_SEGMENTS=4): e = encode_new.Encoder() data = "happy happy joy joy" * 4 e.setup(StringIO(data)) - NUM_SHARES = 100 + assert e.num_shares == NUM_SHARES # else we'll be completely confused e.segment_size = 25 # force use of multiple segments e.setup_codec() # need to rebuild the codec for that change - NUM_SEGMENTS = 4 + assert (NUM_SEGMENTS-1)*e.segment_size < len(data) <= NUM_SEGMENTS*e.segment_size shareholders = {} all_shareholders = [] + all_peers = [] + for i in range(NUM_PEERS): + all_peers.append(FakeBucketWriter()) for shnum in range(NUM_SHARES): - peer = FakePeer() + peer = all_peers[shnum % NUM_PEERS] shareholders[shnum] = peer all_shareholders.append(peer) e.set_shareholders(shareholders) @@ -125,6 +154,7 @@ class UpDown(unittest.TestCase): def _uploaded(roothash): URI = pack_uri(e._codec.get_encoder_type(), e._codec.get_serialized_params(), + e._tail_codec.get_serialized_params(), "V" * 20, roothash, e.required_shares, @@ -138,7 +168,7 @@ class UpDown(unittest.TestCase): for shnum in range(NUM_SHARES): fd._share_buckets[shnum] = set([all_shareholders[shnum]]) fd._got_all_shareholders(None) - d2 = fd._download_all_segments() + d2 = fd._download_all_segments(None) d2.addCallback(fd._done) return d2 d.addCallback(_uploaded) @@ -147,3 +177,7 @@ class UpDown(unittest.TestCase): d.addCallback(_downloaded) return d + + def test_one_share_per_peer(self): + return self.send_and_recover(100, 100) + diff --git a/src/allmydata/test/test_storage.py b/src/allmydata/test/test_storage.py deleted file mode 100644 index 31da06ab..00000000 --- a/src/allmydata/test/test_storage.py +++ /dev/null @@ -1,156 +0,0 @@ - -import os -import random - -from twisted.trial import unittest -from twisted.application import service -from twisted.internet import defer -from foolscap import Tub, Referenceable -from foolscap.eventual import flushEventualQueue - -from allmydata import client - -class Canary(Referenceable): - pass - -class StorageTest(unittest.TestCase): - - def setUp(self): - self.svc = service.MultiService() - self.node = client.Client('') - self.node.setServiceParent(self.svc) - self.tub = Tub() - self.tub.setServiceParent(self.svc) - self.svc.startService() - return self.node.when_tub_ready() - - def test_create_bucket(self): - """ - Check that the storage server can return bucket data accurately. - """ - vid = os.urandom(20) - bnum = random.randrange(0, 256) - data = os.urandom(random.randint(1024, 16384)) - - rssd = self.tub.getReference(self.node.my_pburl) - def get_storageserver(node): - return node.callRemote('get_service', name='storageserver') - rssd.addCallback(get_storageserver) - - def create_bucket(storageserver): - return storageserver.callRemote('allocate_bucket', - verifierid=vid, - bucket_num=bnum, - size=len(data), - leaser=self.node.nodeid, - canary=Canary(), - ) - rssd.addCallback(create_bucket) - - def write_to_bucket(bucket): - def write_some(junk, bytes): - return bucket.callRemote('write', data=bytes) - def set_metadata(junk, metadata): - return bucket.callRemote('set_metadata', metadata) - def finalise(junk): - return bucket.callRemote('close') - off1 = len(data) / 2 - off2 = 3 * len(data) / 4 - d = defer.succeed(None) - d.addCallback(write_some, data[:off1]) - d.addCallback(write_some, data[off1:off2]) - d.addCallback(set_metadata, "metadata") - d.addCallback(write_some, data[off2:]) - d.addCallback(finalise) - return d - rssd.addCallback(write_to_bucket) - - def get_node_again(junk): - return self.tub.getReference(self.node.my_pburl) - rssd.addCallback(get_node_again) - rssd.addCallback(get_storageserver) - - def get_buckets(storageserver): - return storageserver.callRemote('get_buckets', verifierid=vid) - rssd.addCallback(get_buckets) - - def read_buckets(buckets): - self.failUnlessEqual(len(buckets), 1) - bucket_num, bucket = buckets[0] - self.failUnlessEqual(bucket_num, bnum) - - def check_data(bytes_read): - self.failUnlessEqual(bytes_read, data) - d = bucket.callRemote('read') - d.addCallback(check_data) - - def check_metadata(metadata): - self.failUnlessEqual(metadata, 'metadata') - d.addCallback(lambda res: bucket.callRemote('get_metadata')) - d.addCallback(check_metadata) - return d - rssd.addCallback(read_buckets) - - return rssd - - def test_overwrite(self): - """ - Check that the storage server rejects an attempt to write too much data. - """ - vid = os.urandom(20) - bnum = random.randrange(0, 256) - data = os.urandom(random.randint(1024, 16384)) - - rssd = self.tub.getReference(self.node.my_pburl) - def get_storageserver(node): - return node.callRemote('get_service', name='storageserver') - rssd.addCallback(get_storageserver) - - def create_bucket(storageserver): - return storageserver.callRemote('allocate_bucket', - verifierid=vid, - bucket_num=bnum, - size=len(data), - leaser=self.node.nodeid, - canary=Canary(), - ) - rssd.addCallback(create_bucket) - - def write_to_bucket(bucket): - def write_some(junk, bytes): - return bucket.callRemote('write', data=bytes) - def finalise(junk): - return bucket.callRemote('close') - off1 = len(data) / 2 - off2 = 3 * len(data) / 4 - d = defer.succeed(None) - d.addCallback(write_some, data[:off1]) - d.addCallback(write_some, data[off1:off2]) - d.addCallback(write_some, data[off2:]) - # and then overwrite - d.addCallback(write_some, data[off1:off2]) - d.addCallback(finalise) - return d - rssd.addCallback(write_to_bucket) - - self.deferredShouldFail(rssd, ftype=AssertionError) - return rssd - - def deferredShouldFail(self, d, ftype=None, checker=None): - - def _worked(res): - self.fail("hey, this was supposed to fail, not return %s" % res) - if not ftype and not checker: - d.addCallbacks(_worked, - lambda f: None) - elif ftype and not checker: - d.addCallbacks(_worked, - lambda f: f.trap(ftype) or None) - else: - d.addCallbacks(_worked, - checker) - - def tearDown(self): - d = self.svc.stopService() - d.addCallback(lambda res: flushEventualQueue()) - return d diff --git a/src/allmydata/test/test_upload.py b/src/allmydata/test/test_upload.py index fbf6327a..5eb9b540 100644 --- a/src/allmydata/test/test_upload.py +++ b/src/allmydata/test/test_upload.py @@ -1,40 +1,18 @@ from twisted.trial import unittest -from twisted.python import log from twisted.python.failure import Failure -from twisted.internet import defer from cStringIO import StringIO -from foolscap import eventual - from allmydata import upload from allmydata.uri import unpack_uri from test_encode import FakePeer -class FakeStorageServer: - def __init__(self, mode): - self.mode = mode - def callRemote(self, methname, *args, **kwargs): - def _call(): - meth = getattr(self, methname) - return meth(*args, **kwargs) - d = eventual.fireEventually() - d.addCallback(lambda res: _call()) - return d - def allocate_buckets(self, verifierid, sharenums, shareize, blocksize, canary): - if self.mode == "full": - return (set(), {},) - elif self.mode == "already got them": - return (set(sharenums), {},) - else: - return (set(), dict([(shnum, FakePeer(),) for shnum in sharenums]),) - class FakeClient: def __init__(self, mode="good"): self.mode = mode def get_permuted_peers(self, verifierid): - return [ ("%20d"%fakeid, "%20d"%fakeid, FakeStorageServer(self.mode),) for fakeid in range(50) ] + return [ ("%20d"%fakeid, "%20d"%fakeid, FakePeer(self.mode),) for fakeid in range(50) ] class GoodServer(unittest.TestCase): def setUp(self): @@ -46,13 +24,11 @@ class GoodServer(unittest.TestCase): def _check(self, uri): self.failUnless(isinstance(uri, str)) self.failUnless(uri.startswith("URI:")) - codec_name, codec_params, verifierid = unpack_uri(uri) + codec_name, codec_params, tail_codec_params, verifierid, roothash, needed_shares, total_shares, size, segment_size = unpack_uri(uri) self.failUnless(isinstance(verifierid, str)) self.failUnlessEqual(len(verifierid), 20) self.failUnless(isinstance(codec_params, str)) - peers = self.node.peers - self.failUnlessEqual(peers[0].allocated_size, - len(peers[0].data)) + def testData(self): data = "This is some data to upload" d = self.u.upload_data(data) diff --git a/src/allmydata/upload.py b/src/allmydata/upload.py index 7559319d..7d9a28b0 100644 --- a/src/allmydata/upload.py +++ b/src/allmydata/upload.py @@ -5,7 +5,6 @@ from twisted.application import service from foolscap import Referenceable from allmydata.util import idlib -from allmydata.util.assertutil import _assert from allmydata import encode_new from allmydata.uri import pack_uri from allmydata.interfaces import IUploadable, IUploader @@ -28,16 +27,26 @@ class PeerTracker: def __init__(self, peerid, permutedid, connection, sharesize, blocksize, verifierid): self.peerid = peerid self.permutedid = permutedid - self.connection = connection + self.connection = connection # to an RIClient self.buckets = {} # k: shareid, v: IRemoteBucketWriter self.sharesize = sharesize self.blocksize = blocksize self.verifierid = verifierid + self._storageserver = None def query(self, sharenums): - d = self.connection.callRemote("allocate_buckets", self.verifierid, - sharenums, self.sharesize, - self.blocksize, canary=Referenceable()) + if not self._storageserver: + d = self.connection.callRemote("get_service", "storageserver") + d.addCallback(self._got_storageserver) + d.addCallback(lambda res: self._query(sharenums)) + return d + return self._query(sharenums) + def _got_storageserver(self, storageserver): + self._storageserver = storageserver + def _query(self, sharenums): + d = self._storageserver.callRemote("allocate_buckets", self.verifierid, + sharenums, self.sharesize, + self.blocksize, canary=Referenceable()) d.addCallback(self._got_reply) return d @@ -194,7 +203,7 @@ class FileUploader: self.unallocated_sharenums -= allocated if allocated: - self.usable_peers.add(peer) + self.used_peers.add(peer) if shares_we_requested - alreadygot - allocated: log.msg("%s._got_response(%s, %s, %s): self.unallocated_sharenums: %s, unhandled: %s HE'S FULL" % (self, (alreadygot, allocated), peer, shares_we_requested, self.unallocated_sharenums, shares_we_requested - alreadygot - allocated)) @@ -222,7 +231,11 @@ class FileUploader: def _compute_uri(self, roothash): codec_type = self._encoder._codec.get_encoder_type() codec_params = self._encoder._codec.get_serialized_params() - return pack_uri(codec_type, codec_params, self._verifierid, roothash, self.needed_shares, self.total_shares, self._size, self._encoder.segment_size) + tail_codec_params = self._encoder._tail_codec.get_serialized_params() + return pack_uri(codec_type, codec_params, tail_codec_params, + self._verifierid, + roothash, self.needed_shares, self.total_shares, + self._size, self._encoder.segment_size) def netstring(s): diff --git a/src/allmydata/uri.py b/src/allmydata/uri.py index ed43eb7e..a0f77fdd 100644 --- a/src/allmydata/uri.py +++ b/src/allmydata/uri.py @@ -5,26 +5,28 @@ from allmydata.util import idlib # enough information to retrieve and validate the contents. It shall be # expressed in a limited character set (namely [TODO]). -def pack_uri(codec_name, codec_params, verifierid, roothash, needed_shares, total_shares, size, segment_size): +def pack_uri(codec_name, codec_params, tail_codec_params, verifierid, roothash, needed_shares, total_shares, size, segment_size): assert isinstance(codec_name, str) assert len(codec_name) < 10 assert ":" not in codec_name assert isinstance(codec_params, str) assert ":" not in codec_params + assert isinstance(tail_codec_params, str) + assert ":" not in tail_codec_params assert isinstance(verifierid, str) assert len(verifierid) == 20 # sha1 hash - return "URI:%s:%s:%s:%s:%s:%s:%s:%s" % (codec_name, codec_params, idlib.b2a(verifierid), idlib.b2a(roothash), needed_shares, total_shares, size, segment_size) + return "URI:%s:%s:%s:%s:%s:%s:%s:%s:%s" % (codec_name, codec_params, tail_codec_params, idlib.b2a(verifierid), idlib.b2a(roothash), needed_shares, total_shares, size, segment_size) def unpack_uri(uri): assert uri.startswith("URI:") - header, codec_name, codec_params, verifierid_s, roothash_s, needed_shares_s, total_shares_s, size_s, segment_size_s = uri.split(":") + header, codec_name, codec_params, tail_codec_params, verifierid_s, roothash_s, needed_shares_s, total_shares_s, size_s, segment_size_s = uri.split(":") verifierid = idlib.a2b(verifierid_s) roothash = idlib.a2b(roothash_s) needed_shares = int(needed_shares_s) total_shares = int(total_shares_s) size = int(size_s) segment_size = int(segment_size_s) - return codec_name, codec_params, verifierid, roothash, needed_shares, total_shares, size, segment_size + return codec_name, codec_params, tail_codec_params, verifierid, roothash, needed_shares, total_shares, size, segment_size