Merge branch 'dht-inlinecallbacks-refactor'
This commit is contained in:
commit
31e032bb55
43 changed files with 1222 additions and 1273 deletions
|
@ -38,10 +38,8 @@ install:
|
|||
- pip install .
|
||||
|
||||
script:
|
||||
- pip install cython
|
||||
- pip install mock pylint unqlite
|
||||
- pip install mock pylint
|
||||
- pylint lbrynet
|
||||
- PYTHONPATH=. trial lbrynet.tests
|
||||
- python -m unittest discover lbrynet/tests/integration -v
|
||||
- rvm install ruby-2.3.1
|
||||
- rvm use 2.3.1 && gem install danger --version '~> 4.0' && danger
|
||||
|
|
33
CHANGELOG.md
33
CHANGELOG.md
|
@ -13,24 +13,37 @@ at anytime.
|
|||
*
|
||||
|
||||
### Fixed
|
||||
*
|
||||
*
|
||||
* handling error from dht clients with old `ping` method
|
||||
* blobs not being re-announced if no peers successfully stored, now failed announcements are re-queued
|
||||
|
||||
### Deprecated
|
||||
*
|
||||
*
|
||||
|
||||
*
|
||||
|
||||
### Changed
|
||||
*
|
||||
*
|
||||
* several internal dht functions to use inlineCallbacks
|
||||
* `DHTHashAnnouncer` and `Node` manage functions to use `LoopingCall`s instead of scheduling with `callLater`.
|
||||
* `store` kademlia rpc method to block on the call finishing and to return storing peer information
|
||||
* refactored `DHTHashAnnouncer` to longer use locks, use a `DeferredSemaphore` to limit concurrent announcers
|
||||
* decoupled `DiskBlobManager` from `DHTHashAnnouncer`
|
||||
* blob hashes to announce to be controlled by`SQLiteStorage`
|
||||
* kademlia protocol to not delay writes to the UDP socket
|
||||
* `reactor` and `callLater`, `listenUDP`, and `resolve` functions to be configurable (to allow easier testing)
|
||||
* calls to get the current time to use `reactor.seconds` (to control callLater and LoopingCall timing in tests)
|
||||
* `blob_announce` to queue the blob announcement but not block on it
|
||||
* blob completion to not `callLater` an immediate announce, let `SQLiteStorage` and the `DHTHashAnnouncer` handle it
|
||||
* raise the default number of concurrent blob announcers to 100
|
||||
* dht logging to be more verbose with errors and warnings
|
||||
* added `single_announce` and `last_announced_time` columns to the `blob` table in sqlite
|
||||
|
||||
### Added
|
||||
*
|
||||
*
|
||||
* virtual kademlia network and mock udp transport for dht integration tests
|
||||
* integration tests for bootstrapping the dht
|
||||
* configurable `concurrent_announcers` setting
|
||||
|
||||
### Removed
|
||||
*
|
||||
*
|
||||
* `announce_all` argument from `blob_announce`
|
||||
* old `blob_announce_all` command
|
||||
|
||||
|
||||
## [0.19.2] - 2018-03-28
|
||||
|
|
|
@ -40,6 +40,8 @@ ANDROID = 4
|
|||
KB = 2 ** 10
|
||||
MB = 2 ** 20
|
||||
|
||||
DEFAULT_CONCURRENT_ANNOUNCERS = 100
|
||||
|
||||
DEFAULT_DHT_NODES = [
|
||||
('lbrynet1.lbry.io', 4444),
|
||||
('lbrynet2.lbry.io', 4444),
|
||||
|
@ -263,6 +265,7 @@ ADJUSTABLE_SETTINGS = {
|
|||
'download_timeout': (int, 180),
|
||||
'is_generous_host': (bool, True),
|
||||
'announce_head_blobs_only': (bool, True),
|
||||
'concurrent_announcers': (int, DEFAULT_CONCURRENT_ANNOUNCERS),
|
||||
'known_dht_nodes': (list, DEFAULT_DHT_NODES, server_list),
|
||||
'lbryum_wallet_dir': (str, default_lbryum_dir),
|
||||
'max_connections_per_stream': (int, 5),
|
||||
|
|
|
@ -1,27 +1,23 @@
|
|||
import logging
|
||||
import os
|
||||
from sqlite3 import IntegrityError
|
||||
from twisted.internet import threads, defer, reactor, task
|
||||
from twisted.internet import threads, defer, task
|
||||
from lbrynet import conf
|
||||
from lbrynet.blob.blob_file import BlobFile
|
||||
from lbrynet.blob.creator import BlobFileCreator
|
||||
from lbrynet.core.server.DHTHashAnnouncer import DHTHashSupplier
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class DiskBlobManager(DHTHashSupplier):
|
||||
def __init__(self, hash_announcer, blob_dir, storage):
|
||||
|
||||
class DiskBlobManager(object):
|
||||
def __init__(self, blob_dir, storage):
|
||||
"""
|
||||
This class stores blobs on the hard disk,
|
||||
This class stores blobs on the hard disk
|
||||
|
||||
blob_dir - directory where blobs are stored
|
||||
db_dir - directory where sqlite database of blob information is stored
|
||||
storage - SQLiteStorage object
|
||||
"""
|
||||
|
||||
DHTHashSupplier.__init__(self, hash_announcer)
|
||||
self.storage = storage
|
||||
self.announce_head_blobs_only = conf.settings['announce_head_blobs_only']
|
||||
self.blob_dir = blob_dir
|
||||
self.blob_creator_type = BlobFileCreator
|
||||
# TODO: consider using an LRU for blobs as there could potentially
|
||||
|
@ -30,7 +26,7 @@ class DiskBlobManager(DHTHashSupplier):
|
|||
self.blob_hashes_to_delete = {} # {blob_hash: being_deleted (True/False)}
|
||||
|
||||
self.check_should_announce_lc = None
|
||||
if conf.settings['run_reflector_server']:
|
||||
if conf.settings['run_reflector_server']: # TODO: move this looping call to SQLiteStorage
|
||||
self.check_should_announce_lc = task.LoopingCall(self.storage.verify_will_announce_all_head_and_sd_blobs)
|
||||
|
||||
def setup(self):
|
||||
|
@ -62,40 +58,21 @@ class DiskBlobManager(DHTHashSupplier):
|
|||
self.blobs[blob_hash] = blob
|
||||
return defer.succeed(blob)
|
||||
|
||||
def immediate_announce(self, blob_hashes):
|
||||
if self.hash_announcer:
|
||||
return self.hash_announcer.immediate_announce(blob_hashes)
|
||||
raise Exception("Hash announcer not set")
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def blob_completed(self, blob, next_announce_time=None, should_announce=True):
|
||||
if next_announce_time is None:
|
||||
next_announce_time = self.get_next_announce_time()
|
||||
def blob_completed(self, blob, should_announce=False, next_announce_time=None):
|
||||
yield self.storage.add_completed_blob(
|
||||
blob.blob_hash, blob.length, next_announce_time, should_announce
|
||||
)
|
||||
# we announce all blobs immediately, if announce_head_blob_only is False
|
||||
# otherwise, announce only if marked as should_announce
|
||||
if not self.announce_head_blobs_only or should_announce:
|
||||
reactor.callLater(0, self.immediate_announce, [blob.blob_hash])
|
||||
|
||||
def completed_blobs(self, blobhashes_to_check):
|
||||
return self._completed_blobs(blobhashes_to_check)
|
||||
|
||||
def hashes_to_announce(self):
|
||||
return self.storage.get_blobs_to_announce(self.hash_announcer)
|
||||
|
||||
def count_should_announce_blobs(self):
|
||||
return self.storage.count_should_announce_blobs()
|
||||
|
||||
def set_should_announce(self, blob_hash, should_announce):
|
||||
if blob_hash in self.blobs:
|
||||
blob = self.blobs[blob_hash]
|
||||
if blob.get_is_verified():
|
||||
return self.storage.set_should_announce(
|
||||
blob_hash, self.get_next_announce_time(), should_announce
|
||||
)
|
||||
return defer.succeed(False)
|
||||
now = self.storage.clock.seconds()
|
||||
return self.storage.set_should_announce(blob_hash, now, should_announce)
|
||||
|
||||
def get_should_announce(self, blob_hash):
|
||||
return self.storage.should_announce(blob_hash)
|
||||
|
@ -110,13 +87,7 @@ class DiskBlobManager(DHTHashSupplier):
|
|||
raise Exception("Blob has a length of 0")
|
||||
new_blob = BlobFile(self.blob_dir, blob_creator.blob_hash, blob_creator.length)
|
||||
self.blobs[blob_creator.blob_hash] = new_blob
|
||||
next_announce_time = self.get_next_announce_time()
|
||||
return self.blob_completed(new_blob, next_announce_time, should_announce)
|
||||
|
||||
def immediate_announce_all_blobs(self):
|
||||
d = self._get_all_verified_blob_hashes()
|
||||
d.addCallback(self.immediate_announce)
|
||||
return d
|
||||
return self.blob_completed(new_blob, should_announce)
|
||||
|
||||
def get_all_verified_blobs(self):
|
||||
d = self._get_all_verified_blob_hashes()
|
||||
|
|
|
@ -1,18 +0,0 @@
|
|||
class DummyHashAnnouncer(object):
|
||||
def __init__(self, *args):
|
||||
pass
|
||||
|
||||
def run_manage_loop(self):
|
||||
pass
|
||||
|
||||
def stop(self):
|
||||
pass
|
||||
|
||||
def add_supplier(self, *args):
|
||||
pass
|
||||
|
||||
def hash_queue_size(self):
|
||||
return 0
|
||||
|
||||
def immediate_announce(self, *args):
|
||||
pass
|
|
@ -1,330 +0,0 @@
|
|||
from collections import defaultdict
|
||||
import logging
|
||||
import os
|
||||
import unqlite
|
||||
import time
|
||||
from Crypto.Hash import SHA512
|
||||
from Crypto.PublicKey import RSA
|
||||
from lbrynet.core.client.ClientRequest import ClientRequest
|
||||
from lbrynet.core.Error import RequestCanceledError
|
||||
from lbrynet.interfaces import IRequestCreator, IQueryHandlerFactory, IQueryHandler, IWallet
|
||||
from lbrynet.pointtraderclient import pointtraderclient
|
||||
from twisted.internet import defer, threads
|
||||
from zope.interface import implements
|
||||
from twisted.python.failure import Failure
|
||||
from lbrynet.core.Wallet import ReservedPoints
|
||||
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class PTCWallet(object):
|
||||
"""This class sends payments to peers and also ensures that expected payments are received.
|
||||
This class is only intended to be used for testing."""
|
||||
implements(IWallet)
|
||||
|
||||
def __init__(self, db_dir):
|
||||
self.db_dir = db_dir
|
||||
self.db = None
|
||||
self.private_key = None
|
||||
self.encoded_public_key = None
|
||||
self.peer_pub_keys = {}
|
||||
self.queued_payments = defaultdict(int)
|
||||
self.expected_payments = defaultdict(list)
|
||||
self.received_payments = defaultdict(list)
|
||||
self.next_manage_call = None
|
||||
self.payment_check_window = 3 * 60 # 3 minutes
|
||||
self.new_payments_expected_time = time.time() - self.payment_check_window
|
||||
self.known_transactions = []
|
||||
self.total_reserved_points = 0.0
|
||||
self.wallet_balance = 0.0
|
||||
|
||||
def manage(self):
|
||||
"""Send payments, ensure expected payments are received"""
|
||||
|
||||
from twisted.internet import reactor
|
||||
|
||||
if time.time() < self.new_payments_expected_time + self.payment_check_window:
|
||||
d1 = self._get_new_payments()
|
||||
else:
|
||||
d1 = defer.succeed(None)
|
||||
d1.addCallback(lambda _: self._check_good_standing())
|
||||
d2 = self._send_queued_points()
|
||||
self.next_manage_call = reactor.callLater(60, self.manage)
|
||||
dl = defer.DeferredList([d1, d2])
|
||||
dl.addCallback(lambda _: self.get_balance())
|
||||
|
||||
def set_balance(balance):
|
||||
self.wallet_balance = balance
|
||||
|
||||
dl.addCallback(set_balance)
|
||||
return dl
|
||||
|
||||
def stop(self):
|
||||
if self.next_manage_call is not None:
|
||||
self.next_manage_call.cancel()
|
||||
self.next_manage_call = None
|
||||
d = self.manage()
|
||||
self.next_manage_call.cancel()
|
||||
self.next_manage_call = None
|
||||
self.db = None
|
||||
return d
|
||||
|
||||
def start(self):
|
||||
|
||||
def save_key(success, private_key):
|
||||
if success is True:
|
||||
self._save_private_key(private_key.exportKey())
|
||||
return True
|
||||
return False
|
||||
|
||||
def register_private_key(private_key):
|
||||
self.private_key = private_key
|
||||
self.encoded_public_key = self.private_key.publickey().exportKey()
|
||||
d_r = pointtraderclient.register_new_account(private_key)
|
||||
d_r.addCallback(save_key, private_key)
|
||||
return d_r
|
||||
|
||||
def ensure_private_key_exists(encoded_private_key):
|
||||
if encoded_private_key is not None:
|
||||
self.private_key = RSA.importKey(encoded_private_key)
|
||||
self.encoded_public_key = self.private_key.publickey().exportKey()
|
||||
return True
|
||||
else:
|
||||
create_d = threads.deferToThread(RSA.generate, 4096)
|
||||
create_d.addCallback(register_private_key)
|
||||
return create_d
|
||||
|
||||
def start_manage():
|
||||
self.manage()
|
||||
return True
|
||||
d = self._open_db()
|
||||
d.addCallback(lambda _: self._get_wallet_private_key())
|
||||
d.addCallback(ensure_private_key_exists)
|
||||
d.addCallback(lambda _: start_manage())
|
||||
return d
|
||||
|
||||
def get_info_exchanger(self):
|
||||
return PointTraderKeyExchanger(self)
|
||||
|
||||
def get_wallet_info_query_handler_factory(self):
|
||||
return PointTraderKeyQueryHandlerFactory(self)
|
||||
|
||||
def reserve_points(self, peer, amount):
|
||||
"""Ensure a certain amount of points are available to be sent as
|
||||
payment, before the service is rendered
|
||||
|
||||
@param peer: The peer to which the payment will ultimately be sent
|
||||
|
||||
@param amount: The amount of points to reserve
|
||||
|
||||
@return: A ReservedPoints object which is given to send_points
|
||||
once the service has been rendered
|
||||
|
||||
"""
|
||||
if self.wallet_balance >= self.total_reserved_points + amount:
|
||||
self.total_reserved_points += amount
|
||||
return ReservedPoints(peer, amount)
|
||||
return None
|
||||
|
||||
def cancel_point_reservation(self, reserved_points):
|
||||
"""
|
||||
Return all of the points that were reserved previously for some ReservedPoints object
|
||||
|
||||
@param reserved_points: ReservedPoints previously returned by reserve_points
|
||||
|
||||
@return: None
|
||||
"""
|
||||
self.total_reserved_points -= reserved_points.amount
|
||||
|
||||
def send_points(self, reserved_points, amount):
|
||||
"""
|
||||
Schedule a payment to be sent to a peer
|
||||
|
||||
@param reserved_points: ReservedPoints object previously returned by reserve_points
|
||||
|
||||
@param amount: amount of points to actually send, must be less than or equal to the
|
||||
amount reserved in reserved_points
|
||||
|
||||
@return: Deferred which fires when the payment has been scheduled
|
||||
"""
|
||||
self.queued_payments[reserved_points.identifier] += amount
|
||||
# make any unused points available
|
||||
self.total_reserved_points -= reserved_points.amount - amount
|
||||
reserved_points.identifier.update_stats('points_sent', amount)
|
||||
d = defer.succeed(True)
|
||||
return d
|
||||
|
||||
def _send_queued_points(self):
|
||||
ds = []
|
||||
for peer, points in self.queued_payments.items():
|
||||
if peer in self.peer_pub_keys:
|
||||
d = pointtraderclient.send_points(
|
||||
self.private_key, self.peer_pub_keys[peer], points)
|
||||
self.wallet_balance -= points
|
||||
self.total_reserved_points -= points
|
||||
ds.append(d)
|
||||
del self.queued_payments[peer]
|
||||
else:
|
||||
log.warning("Don't have a payment address for peer %s. Can't send %s points.",
|
||||
str(peer), str(points))
|
||||
return defer.DeferredList(ds)
|
||||
|
||||
def get_balance(self):
|
||||
"""Return the balance of this wallet"""
|
||||
d = pointtraderclient.get_balance(self.private_key)
|
||||
return d
|
||||
|
||||
def add_expected_payment(self, peer, amount):
|
||||
"""Increase the number of points expected to be paid by a peer"""
|
||||
self.expected_payments[peer].append((amount, time.time()))
|
||||
self.new_payments_expected_time = time.time()
|
||||
peer.update_stats('expected_points', amount)
|
||||
|
||||
def set_public_key_for_peer(self, peer, pub_key):
|
||||
self.peer_pub_keys[peer] = pub_key
|
||||
|
||||
def _get_new_payments(self):
|
||||
|
||||
def add_new_transactions(transactions):
|
||||
for transaction in transactions:
|
||||
if transaction[1] == self.encoded_public_key:
|
||||
t_hash = SHA512.new()
|
||||
t_hash.update(transaction[0])
|
||||
t_hash.update(transaction[1])
|
||||
t_hash.update(str(transaction[2]))
|
||||
t_hash.update(transaction[3])
|
||||
if t_hash.hexdigest() not in self.known_transactions:
|
||||
self.known_transactions.append(t_hash.hexdigest())
|
||||
self._add_received_payment(transaction[0], transaction[2])
|
||||
|
||||
d = pointtraderclient.get_recent_transactions(self.private_key)
|
||||
d.addCallback(add_new_transactions)
|
||||
return d
|
||||
|
||||
def _add_received_payment(self, encoded_other_public_key, amount):
|
||||
self.received_payments[encoded_other_public_key].append((amount, time.time()))
|
||||
|
||||
def _check_good_standing(self):
|
||||
for peer, expected_payments in self.expected_payments.iteritems():
|
||||
expected_cutoff = time.time() - 90
|
||||
min_expected_balance = sum([a[0] for a in expected_payments if a[1] < expected_cutoff])
|
||||
received_balance = 0
|
||||
if self.peer_pub_keys[peer] in self.received_payments:
|
||||
received_balance = sum([
|
||||
a[0] for a in self.received_payments[self.peer_pub_keys[peer]]])
|
||||
if min_expected_balance > received_balance:
|
||||
log.warning(
|
||||
"Account in bad standing: %s (pub_key: %s), expected amount = %s, "
|
||||
"received_amount = %s",
|
||||
peer, self.peer_pub_keys[peer], min_expected_balance, received_balance)
|
||||
|
||||
def _open_db(self):
|
||||
def open_db():
|
||||
self.db = unqlite.UnQLite(os.path.join(self.db_dir, "ptcwallet.db"))
|
||||
return threads.deferToThread(open_db)
|
||||
|
||||
def _save_private_key(self, private_key):
|
||||
def save_key():
|
||||
self.db['private_key'] = private_key
|
||||
return threads.deferToThread(save_key)
|
||||
|
||||
def _get_wallet_private_key(self):
|
||||
def get_key():
|
||||
if 'private_key' in self.db:
|
||||
return self.db['private_key']
|
||||
return None
|
||||
return threads.deferToThread(get_key)
|
||||
|
||||
|
||||
class PointTraderKeyExchanger(object):
|
||||
implements([IRequestCreator])
|
||||
|
||||
def __init__(self, wallet):
|
||||
self.wallet = wallet
|
||||
self._protocols = []
|
||||
|
||||
######### IRequestCreator #########
|
||||
|
||||
def send_next_request(self, peer, protocol):
|
||||
if not protocol in self._protocols:
|
||||
r = ClientRequest({'public_key': self.wallet.encoded_public_key},
|
||||
'public_key')
|
||||
d = protocol.add_request(r)
|
||||
d.addCallback(self._handle_exchange_response, peer, r, protocol)
|
||||
d.addErrback(self._request_failed, peer)
|
||||
self._protocols.append(protocol)
|
||||
return defer.succeed(True)
|
||||
else:
|
||||
return defer.succeed(False)
|
||||
|
||||
######### internal calls #########
|
||||
|
||||
def _handle_exchange_response(self, response_dict, peer, request, protocol):
|
||||
assert request.response_identifier in response_dict, \
|
||||
"Expected %s in dict but did not get it" % request.response_identifier
|
||||
assert protocol in self._protocols, "Responding protocol is not in our list of protocols"
|
||||
peer_pub_key = response_dict[request.response_identifier]
|
||||
self.wallet.set_public_key_for_peer(peer, peer_pub_key)
|
||||
return True
|
||||
|
||||
def _request_failed(self, err, peer):
|
||||
if not err.check(RequestCanceledError):
|
||||
log.warning("A peer failed to send a valid public key response. Error: %s, peer: %s",
|
||||
err.getErrorMessage(), str(peer))
|
||||
return err
|
||||
|
||||
|
||||
class PointTraderKeyQueryHandlerFactory(object):
|
||||
implements(IQueryHandlerFactory)
|
||||
|
||||
def __init__(self, wallet):
|
||||
self.wallet = wallet
|
||||
|
||||
######### IQueryHandlerFactory #########
|
||||
|
||||
def build_query_handler(self):
|
||||
q_h = PointTraderKeyQueryHandler(self.wallet)
|
||||
return q_h
|
||||
|
||||
def get_primary_query_identifier(self):
|
||||
return 'public_key'
|
||||
|
||||
def get_description(self):
|
||||
return ("Point Trader Address - an address for receiving payments on the "
|
||||
"point trader testing network")
|
||||
|
||||
|
||||
class PointTraderKeyQueryHandler(object):
|
||||
implements(IQueryHandler)
|
||||
|
||||
def __init__(self, wallet):
|
||||
self.wallet = wallet
|
||||
self.query_identifiers = ['public_key']
|
||||
self.public_key = None
|
||||
self.peer = None
|
||||
|
||||
######### IQueryHandler #########
|
||||
|
||||
def register_with_request_handler(self, request_handler, peer):
|
||||
self.peer = peer
|
||||
request_handler.register_query_handler(self, self.query_identifiers)
|
||||
|
||||
def handle_queries(self, queries):
|
||||
if self.query_identifiers[0] in queries:
|
||||
new_encoded_pub_key = queries[self.query_identifiers[0]]
|
||||
try:
|
||||
RSA.importKey(new_encoded_pub_key)
|
||||
except (ValueError, TypeError, IndexError):
|
||||
log.warning("Client sent an invalid public key.")
|
||||
return defer.fail(Failure(ValueError("Client sent an invalid public key")))
|
||||
self.public_key = new_encoded_pub_key
|
||||
self.wallet.set_public_key_for_peer(self.peer, self.public_key)
|
||||
log.debug("Received the client's public key: %s", str(self.public_key))
|
||||
fields = {'public_key': self.wallet.encoded_public_key}
|
||||
return defer.succeed(fields)
|
||||
if self.public_key is None:
|
||||
log.warning("Expected a public key, but did not receive one")
|
||||
return defer.fail(Failure(ValueError("Expected but did not receive a public key")))
|
||||
else:
|
||||
return defer.succeed({})
|
|
@ -1,19 +0,0 @@
|
|||
from twisted.internet import defer
|
||||
|
||||
|
||||
class DummyPeerFinder(object):
|
||||
"""This class finds peers which have announced to the DHT that they have certain blobs"""
|
||||
def __init__(self):
|
||||
pass
|
||||
|
||||
def run_manage_loop(self):
|
||||
pass
|
||||
|
||||
def stop(self):
|
||||
pass
|
||||
|
||||
def find_peers_for_blob(self, blob_hash):
|
||||
return defer.succeed([])
|
||||
|
||||
def get_most_popular_hashes(self, num_to_return):
|
||||
return []
|
|
@ -1,17 +1,13 @@
|
|||
import logging
|
||||
import miniupnpc
|
||||
from twisted.internet import threads, defer
|
||||
from lbrynet.core.BlobManager import DiskBlobManager
|
||||
from lbrynet.dht import node
|
||||
from lbrynet.dht import node, hashannouncer
|
||||
from lbrynet.database.storage import SQLiteStorage
|
||||
from lbrynet.core.PeerManager import PeerManager
|
||||
from lbrynet.core.RateLimiter import RateLimiter
|
||||
from lbrynet.core.client.DHTPeerFinder import DHTPeerFinder
|
||||
from lbrynet.core.HashAnnouncer import DummyHashAnnouncer
|
||||
from lbrynet.core.server.DHTHashAnnouncer import DHTHashAnnouncer
|
||||
from lbrynet.core.utils import generate_id
|
||||
from lbrynet.core.PaymentRateManager import BasePaymentRateManager, NegotiatedPaymentRateManager
|
||||
from lbrynet.core.BlobAvailability import BlobAvailabilityTracker
|
||||
from twisted.internet import threads, defer
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
|
||||
|
@ -101,43 +97,32 @@ class Session(object):
|
|||
|
||||
"""
|
||||
self.db_dir = db_dir
|
||||
|
||||
self.node_id = node_id
|
||||
|
||||
self.peer_manager = peer_manager
|
||||
|
||||
self.peer_finder = peer_finder
|
||||
self.hash_announcer = hash_announcer
|
||||
self.dht_node_port = dht_node_port
|
||||
self.known_dht_nodes = known_dht_nodes
|
||||
if self.known_dht_nodes is None:
|
||||
self.known_dht_nodes = []
|
||||
self.peer_finder = peer_finder
|
||||
self.hash_announcer = hash_announcer
|
||||
|
||||
self.blob_dir = blob_dir
|
||||
self.blob_manager = blob_manager
|
||||
|
||||
self.blob_tracker = None
|
||||
self.blob_tracker_class = blob_tracker_class or BlobAvailabilityTracker
|
||||
|
||||
self.peer_port = peer_port
|
||||
|
||||
self.use_upnp = use_upnp
|
||||
|
||||
self.rate_limiter = rate_limiter
|
||||
|
||||
self.external_ip = external_ip
|
||||
|
||||
self.upnp_redirects = []
|
||||
|
||||
self.wallet = wallet
|
||||
self.dht_node_class = dht_node_class
|
||||
self.dht_node = None
|
||||
|
||||
self.base_payment_rate_manager = BasePaymentRateManager(blob_data_payment_rate)
|
||||
self.payment_rate_manager = None
|
||||
self.payment_rate_manager_class = payment_rate_manager_class or NegotiatedPaymentRateManager
|
||||
self.is_generous = is_generous
|
||||
self.storage = storage or SQLiteStorage(self.db_dir)
|
||||
self._join_dht_deferred = None
|
||||
|
||||
def setup(self):
|
||||
"""Create the blob directory and database if necessary, start all desired services"""
|
||||
|
@ -147,25 +132,12 @@ class Session(object):
|
|||
if self.node_id is None:
|
||||
self.node_id = generate_id()
|
||||
|
||||
if self.wallet is None:
|
||||
from lbrynet.core.PTCWallet import PTCWallet
|
||||
self.wallet = PTCWallet(self.db_dir)
|
||||
|
||||
if self.peer_manager is None:
|
||||
self.peer_manager = PeerManager()
|
||||
|
||||
if self.use_upnp is True:
|
||||
d = self._try_upnp()
|
||||
else:
|
||||
d = defer.succeed(True)
|
||||
|
||||
if self.peer_finder is None:
|
||||
d.addCallback(lambda _: self._setup_dht())
|
||||
else:
|
||||
if self.hash_announcer is None and self.peer_port is not None:
|
||||
log.warning("The server has no way to advertise its available blobs.")
|
||||
self.hash_announcer = DummyHashAnnouncer()
|
||||
|
||||
d.addCallback(lambda _: self.storage.setup())
|
||||
d.addCallback(lambda _: self._setup_dht())
|
||||
d.addCallback(lambda _: self._setup_other_components())
|
||||
return d
|
||||
|
||||
|
@ -173,16 +145,14 @@ class Session(object):
|
|||
"""Stop all services"""
|
||||
log.info('Stopping session.')
|
||||
ds = []
|
||||
if self.hash_announcer:
|
||||
self.hash_announcer.stop()
|
||||
if self.blob_tracker is not None:
|
||||
ds.append(defer.maybeDeferred(self.blob_tracker.stop))
|
||||
if self.dht_node is not None:
|
||||
ds.append(defer.maybeDeferred(self.dht_node.stop))
|
||||
if self.rate_limiter is not None:
|
||||
ds.append(defer.maybeDeferred(self.rate_limiter.stop))
|
||||
if self.peer_finder is not None:
|
||||
ds.append(defer.maybeDeferred(self.peer_finder.stop))
|
||||
if self.hash_announcer is not None:
|
||||
ds.append(defer.maybeDeferred(self.hash_announcer.stop))
|
||||
if self.wallet is not None:
|
||||
ds.append(defer.maybeDeferred(self.wallet.stop))
|
||||
if self.blob_manager is not None:
|
||||
|
@ -251,59 +221,22 @@ class Session(object):
|
|||
d.addErrback(upnp_failed)
|
||||
return d
|
||||
|
||||
# the callback, if any, will be invoked once the joining procedure
|
||||
# has terminated
|
||||
def join_dht(self, cb=None):
|
||||
from twisted.internet import reactor
|
||||
|
||||
def join_resolved_addresses(result):
|
||||
addresses = []
|
||||
for success, value in result:
|
||||
if success is True:
|
||||
addresses.append(value)
|
||||
return addresses
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def join_network(knownNodes):
|
||||
log.debug("join DHT using known nodes: " + str(knownNodes))
|
||||
result = yield self.dht_node.joinNetwork(knownNodes)
|
||||
defer.returnValue(result)
|
||||
|
||||
ds = []
|
||||
for host, port in self.known_dht_nodes:
|
||||
d = reactor.resolve(host)
|
||||
d.addCallback(lambda h: (h, port)) # match host to port
|
||||
ds.append(d)
|
||||
|
||||
dl = defer.DeferredList(ds)
|
||||
dl.addCallback(join_resolved_addresses)
|
||||
dl.addCallback(join_network)
|
||||
if cb:
|
||||
dl.addCallback(cb)
|
||||
|
||||
return dl
|
||||
|
||||
def _setup_dht(self):
|
||||
log.info("Starting DHT")
|
||||
|
||||
def start_dht(join_network_result):
|
||||
self.hash_announcer.run_manage_loop()
|
||||
return True
|
||||
|
||||
def _setup_dht(self): # does not block startup, the dht will re-attempt if necessary
|
||||
self.dht_node = self.dht_node_class(
|
||||
udpPort=self.dht_node_port,
|
||||
node_id=self.node_id,
|
||||
udpPort=self.dht_node_port,
|
||||
externalIP=self.external_ip,
|
||||
peerPort=self.peer_port
|
||||
peerPort=self.peer_port,
|
||||
peer_manager=self.peer_manager,
|
||||
peer_finder=self.peer_finder,
|
||||
)
|
||||
self.peer_finder = DHTPeerFinder(self.dht_node, self.peer_manager)
|
||||
if self.hash_announcer is None:
|
||||
self.hash_announcer = DHTHashAnnouncer(self.dht_node, self.peer_port)
|
||||
|
||||
self.dht_node.startNetwork()
|
||||
|
||||
# pass start_dht() as callback to start the remaining components after joining the DHT
|
||||
return self.join_dht(start_dht)
|
||||
if not self.hash_announcer:
|
||||
self.hash_announcer = hashannouncer.DHTHashAnnouncer(self.dht_node, self.storage)
|
||||
self.peer_manager = self.dht_node.peer_manager
|
||||
self.peer_finder = self.dht_node.peer_finder
|
||||
self._join_dht_deferred = self.dht_node.joinNetwork(self.known_dht_nodes)
|
||||
self._join_dht_deferred.addCallback(lambda _: log.info("Joined the dht"))
|
||||
self._join_dht_deferred.addCallback(lambda _: self.hash_announcer.start())
|
||||
|
||||
def _setup_other_components(self):
|
||||
log.debug("Setting up the rest of the components")
|
||||
|
@ -316,13 +249,11 @@ class Session(object):
|
|||
raise Exception(
|
||||
"TempBlobManager is no longer supported, specify BlobManager or db_dir")
|
||||
else:
|
||||
self.blob_manager = DiskBlobManager(
|
||||
self.hash_announcer, self.blob_dir, self.storage
|
||||
)
|
||||
self.blob_manager = DiskBlobManager(self.blob_dir, self.storage)
|
||||
|
||||
if self.blob_tracker is None:
|
||||
self.blob_tracker = self.blob_tracker_class(
|
||||
self.blob_manager, self.peer_finder, self.dht_node
|
||||
self.blob_manager, self.dht_node.peer_finder, self.dht_node
|
||||
)
|
||||
if self.payment_rate_manager is None:
|
||||
self.payment_rate_manager = self.payment_rate_manager_class(
|
||||
|
@ -330,8 +261,7 @@ class Session(object):
|
|||
)
|
||||
|
||||
self.rate_limiter.start()
|
||||
d = self.storage.setup()
|
||||
d.addCallback(lambda _: self.blob_manager.setup())
|
||||
d = self.blob_manager.setup()
|
||||
d.addCallback(lambda _: self.wallet.start())
|
||||
d.addCallback(lambda _: self.blob_tracker.start())
|
||||
return d
|
||||
|
|
|
@ -6,13 +6,13 @@ from twisted.internet import defer, threads, reactor
|
|||
|
||||
from lbrynet.blob import BlobFile
|
||||
from lbrynet.core.BlobManager import DiskBlobManager
|
||||
from lbrynet.core.HashAnnouncer import DummyHashAnnouncer
|
||||
from lbrynet.core.RateLimiter import DummyRateLimiter
|
||||
from lbrynet.core.PaymentRateManager import OnlyFreePaymentsManager
|
||||
from lbrynet.core.PeerFinder import DummyPeerFinder
|
||||
from lbrynet.core.client.BlobRequester import BlobRequester
|
||||
from lbrynet.core.client.StandaloneBlobDownloader import StandaloneBlobDownloader
|
||||
from lbrynet.core.client.ConnectionManager import ConnectionManager
|
||||
from lbrynet.dht.peerfinder import DummyPeerFinder
|
||||
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
|
||||
|
@ -60,7 +60,6 @@ class SingleBlobDownloadManager(object):
|
|||
class SinglePeerDownloader(object):
|
||||
def __init__(self):
|
||||
self._payment_rate_manager = OnlyFreePaymentsManager()
|
||||
self._announcer = DummyHashAnnouncer()
|
||||
self._rate_limiter = DummyRateLimiter()
|
||||
self._wallet = None
|
||||
self._blob_manager = None
|
||||
|
@ -97,7 +96,7 @@ class SinglePeerDownloader(object):
|
|||
@defer.inlineCallbacks
|
||||
def download_temp_blob_from_peer(self, peer, timeout, blob_hash):
|
||||
tmp_dir = yield threads.deferToThread(tempfile.mkdtemp)
|
||||
tmp_blob_manager = DiskBlobManager(self._announcer, tmp_dir, tmp_dir)
|
||||
tmp_blob_manager = DiskBlobManager(tmp_dir, tmp_dir)
|
||||
try:
|
||||
result = yield self.download_blob_from_peer(peer, timeout, blob_hash, tmp_blob_manager)
|
||||
finally:
|
||||
|
|
|
@ -623,6 +623,9 @@ class Wallet(object):
|
|||
d = self._get_transaction(txid)
|
||||
return d
|
||||
|
||||
def wait_for_tx_in_wallet(self, txid):
|
||||
return self._wait_for_tx_in_wallet(txid)
|
||||
|
||||
def get_balance(self):
|
||||
return self.wallet_balance - self.total_reserved_points - sum(self.queued_payments.values())
|
||||
|
||||
|
@ -681,6 +684,9 @@ class Wallet(object):
|
|||
def _get_transaction(self, txid):
|
||||
return defer.fail(NotImplementedError())
|
||||
|
||||
def _wait_for_tx_in_wallet(self, txid):
|
||||
return defer.fail(NotImplementedError())
|
||||
|
||||
def _update_balance(self):
|
||||
return defer.fail(NotImplementedError())
|
||||
|
||||
|
@ -1067,6 +1073,9 @@ class LBRYumWallet(Wallet):
|
|||
def _get_transaction(self, txid):
|
||||
return self._run_cmd_as_defer_to_thread("gettransaction", txid)
|
||||
|
||||
def _wait_for_tx_in_wallet(self, txid):
|
||||
return self._run_cmd_as_defer_to_thread("waitfortxinwallet", txid)
|
||||
|
||||
def get_name_claims(self):
|
||||
return self._run_cmd_as_defer_succeed('getnameclaims')
|
||||
|
||||
|
|
63
lbrynet/core/call_later_manager.py
Normal file
63
lbrynet/core/call_later_manager.py
Normal file
|
@ -0,0 +1,63 @@
|
|||
class CallLaterManager(object):
|
||||
_callLater = None
|
||||
_pendingCallLaters = []
|
||||
|
||||
@classmethod
|
||||
def _cancel(cls, call_later):
|
||||
"""
|
||||
:param call_later: DelayedCall
|
||||
:return: (callable) canceller function
|
||||
"""
|
||||
|
||||
def cancel(reason=None):
|
||||
"""
|
||||
:param reason: reason for cancellation, this is returned after cancelling the DelayedCall
|
||||
:return: reason
|
||||
"""
|
||||
|
||||
if call_later.active():
|
||||
call_later.cancel()
|
||||
cls._pendingCallLaters.remove(call_later)
|
||||
return reason
|
||||
return cancel
|
||||
|
||||
@classmethod
|
||||
def stop(cls):
|
||||
"""
|
||||
Cancel any callLaters that are still running
|
||||
"""
|
||||
|
||||
from twisted.internet import defer
|
||||
while cls._pendingCallLaters:
|
||||
canceller = cls._cancel(cls._pendingCallLaters[0])
|
||||
try:
|
||||
canceller()
|
||||
except (defer.CancelledError, defer.AlreadyCalledError):
|
||||
pass
|
||||
|
||||
@classmethod
|
||||
def call_later(cls, when, what, *args, **kwargs):
|
||||
"""
|
||||
Schedule a call later and get a canceller callback function
|
||||
|
||||
:param when: (float) delay in seconds
|
||||
:param what: (callable)
|
||||
:param args: (*tuple) args to be passed to the callable
|
||||
:param kwargs: (**dict) kwargs to be passed to the callable
|
||||
|
||||
:return: (tuple) twisted.internet.base.DelayedCall object, canceller function
|
||||
"""
|
||||
|
||||
call_later = cls._callLater(when, what, *args, **kwargs)
|
||||
canceller = cls._cancel(call_later)
|
||||
cls._pendingCallLaters.append(call_later)
|
||||
return call_later, canceller
|
||||
|
||||
@classmethod
|
||||
def setup(cls, callLater):
|
||||
"""
|
||||
Setup the callLater function to use, supports the real reactor as well as task.Clock
|
||||
|
||||
:param callLater: (IReactorTime.callLater)
|
||||
"""
|
||||
cls._callLater = callLater
|
|
@ -1,129 +0,0 @@
|
|||
import binascii
|
||||
import collections
|
||||
import logging
|
||||
import time
|
||||
|
||||
from twisted.internet import defer
|
||||
from lbrynet.core import utils
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class DHTHashAnnouncer(object):
|
||||
ANNOUNCE_CHECK_INTERVAL = 60
|
||||
CONCURRENT_ANNOUNCERS = 5
|
||||
|
||||
"""This class announces to the DHT that this peer has certain blobs"""
|
||||
def __init__(self, dht_node, peer_port):
|
||||
self.dht_node = dht_node
|
||||
self.peer_port = peer_port
|
||||
self.suppliers = []
|
||||
self.next_manage_call = None
|
||||
self.hash_queue = collections.deque()
|
||||
self._concurrent_announcers = 0
|
||||
|
||||
def run_manage_loop(self):
|
||||
if self.peer_port is not None:
|
||||
self._announce_available_hashes()
|
||||
self.next_manage_call = utils.call_later(self.ANNOUNCE_CHECK_INTERVAL, self.run_manage_loop)
|
||||
|
||||
def stop(self):
|
||||
log.info("Stopping DHT hash announcer.")
|
||||
if self.next_manage_call is not None:
|
||||
self.next_manage_call.cancel()
|
||||
self.next_manage_call = None
|
||||
|
||||
def add_supplier(self, supplier):
|
||||
self.suppliers.append(supplier)
|
||||
|
||||
def immediate_announce(self, blob_hashes):
|
||||
if self.peer_port is not None:
|
||||
return self._announce_hashes(blob_hashes, immediate=True)
|
||||
else:
|
||||
return defer.succeed(False)
|
||||
|
||||
def hash_queue_size(self):
|
||||
return len(self.hash_queue)
|
||||
|
||||
def _announce_available_hashes(self):
|
||||
log.debug('Announcing available hashes')
|
||||
ds = []
|
||||
for supplier in self.suppliers:
|
||||
d = supplier.hashes_to_announce()
|
||||
d.addCallback(self._announce_hashes)
|
||||
ds.append(d)
|
||||
dl = defer.DeferredList(ds)
|
||||
return dl
|
||||
|
||||
def _announce_hashes(self, hashes, immediate=False):
|
||||
if not hashes:
|
||||
return
|
||||
log.debug('Announcing %s hashes', len(hashes))
|
||||
# TODO: add a timeit decorator
|
||||
start = time.time()
|
||||
ds = []
|
||||
|
||||
for h in hashes:
|
||||
announce_deferred = defer.Deferred()
|
||||
ds.append(announce_deferred)
|
||||
if immediate:
|
||||
self.hash_queue.appendleft((h, announce_deferred))
|
||||
else:
|
||||
self.hash_queue.append((h, announce_deferred))
|
||||
log.debug('There are now %s hashes remaining to be announced', self.hash_queue_size())
|
||||
|
||||
def announce():
|
||||
if len(self.hash_queue):
|
||||
h, announce_deferred = self.hash_queue.popleft()
|
||||
log.debug('Announcing blob %s to dht', h)
|
||||
d = self.dht_node.announceHaveBlob(binascii.unhexlify(h))
|
||||
d.chainDeferred(announce_deferred)
|
||||
d.addBoth(lambda _: utils.call_later(0, announce))
|
||||
else:
|
||||
self._concurrent_announcers -= 1
|
||||
|
||||
for i in range(self._concurrent_announcers, self.CONCURRENT_ANNOUNCERS):
|
||||
self._concurrent_announcers += 1
|
||||
announce()
|
||||
d = defer.DeferredList(ds)
|
||||
d.addCallback(lambda _: log.debug('Took %s seconds to announce %s hashes',
|
||||
time.time() - start, len(hashes)))
|
||||
return d
|
||||
|
||||
|
||||
class DHTHashSupplier(object):
|
||||
# 1 hour is the min time hash will be reannounced
|
||||
MIN_HASH_REANNOUNCE_TIME = 60*60
|
||||
# conservative assumption of the time it takes to announce
|
||||
# a single hash
|
||||
SINGLE_HASH_ANNOUNCE_DURATION = 5
|
||||
|
||||
"""Classes derived from this class give hashes to a hash announcer"""
|
||||
def __init__(self, announcer):
|
||||
if announcer is not None:
|
||||
announcer.add_supplier(self)
|
||||
self.hash_announcer = announcer
|
||||
|
||||
def hashes_to_announce(self):
|
||||
pass
|
||||
|
||||
|
||||
def get_next_announce_time(self, num_hashes_to_announce=1):
|
||||
"""
|
||||
Hash reannounce time is set to current time + MIN_HASH_REANNOUNCE_TIME,
|
||||
unless we are announcing a lot of hashes at once which could cause the
|
||||
the announce queue to pile up. To prevent pile up, reannounce
|
||||
only after a conservative estimate of when it will finish
|
||||
to announce all the hashes.
|
||||
|
||||
Args:
|
||||
num_hashes_to_announce: number of hashes that will be added to the queue
|
||||
Returns:
|
||||
timestamp for next announce time
|
||||
"""
|
||||
queue_size = self.hash_announcer.hash_queue_size()+num_hashes_to_announce
|
||||
reannounce = max(self.MIN_HASH_REANNOUNCE_TIME,
|
||||
queue_size*self.SINGLE_HASH_ANNOUNCE_DURATION)
|
||||
return time.time() + reannounce
|
||||
|
||||
|
|
@ -148,6 +148,17 @@ def json_dumps_pretty(obj, **kwargs):
|
|||
return json.dumps(obj, sort_keys=True, indent=2, separators=(',', ': '), **kwargs)
|
||||
|
||||
|
||||
class DeferredLockContextManager(object):
|
||||
def __init__(self, lock):
|
||||
self._lock = lock
|
||||
|
||||
def __enter__(self):
|
||||
yield self._lock.aquire()
|
||||
|
||||
def __exit__(self, exc_type, exc_val, exc_tb):
|
||||
yield self._lock.release()
|
||||
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def DeferredDict(d, consumeErrors=False):
|
||||
keys = []
|
||||
|
|
|
@ -25,7 +25,7 @@ from lbryschema.decode import smart_decode
|
|||
from lbrynet.core.system_info import get_lbrynet_version
|
||||
from lbrynet.database.storage import SQLiteStorage
|
||||
from lbrynet import conf
|
||||
from lbrynet.conf import LBRYCRD_WALLET, LBRYUM_WALLET, PTC_WALLET
|
||||
from lbrynet.conf import LBRYCRD_WALLET, LBRYUM_WALLET
|
||||
from lbrynet.reflector import reupload
|
||||
from lbrynet.reflector import ServerFactory as reflector_server_factory
|
||||
from lbrynet.core.log_support import configure_loggly_handler
|
||||
|
@ -198,7 +198,7 @@ class Daemon(AuthJSONRPCServer):
|
|||
self.connected_to_internet = True
|
||||
self.connection_status_code = None
|
||||
self.platform = None
|
||||
self.current_db_revision = 6
|
||||
self.current_db_revision = 7
|
||||
self.db_revision_file = conf.settings.get_db_revision_filename()
|
||||
self.session = None
|
||||
self._session_id = conf.settings.get_session_id()
|
||||
|
@ -313,7 +313,7 @@ class Daemon(AuthJSONRPCServer):
|
|||
self.session.peer_manager)
|
||||
|
||||
try:
|
||||
log.info("Daemon bound to port: %d", self.peer_port)
|
||||
log.info("Peer protocol listening on TCP %d", self.peer_port)
|
||||
self.lbry_server_port = reactor.listenTCP(self.peer_port, server_factory)
|
||||
except error.CannotListenError as e:
|
||||
import traceback
|
||||
|
@ -547,10 +547,6 @@ class Daemon(AuthJSONRPCServer):
|
|||
config['lbryum_path'] = conf.settings['lbryum_wallet_dir']
|
||||
wallet = LBRYumWallet(self.storage, config)
|
||||
return defer.succeed(wallet)
|
||||
elif self.wallet_type == PTC_WALLET:
|
||||
log.info("Using PTC wallet")
|
||||
from lbrynet.core.PTCWallet import PTCWallet
|
||||
return defer.succeed(PTCWallet(self.db_dir))
|
||||
else:
|
||||
raise ValueError('Wallet Type {} is not valid'.format(self.wallet_type))
|
||||
|
||||
|
@ -997,7 +993,7 @@ class Daemon(AuthJSONRPCServer):
|
|||
############################################################################
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def jsonrpc_status(self, session_status=False, dht_status=False):
|
||||
def jsonrpc_status(self, session_status=False):
|
||||
"""
|
||||
Get daemon status
|
||||
|
||||
|
@ -1006,7 +1002,6 @@ class Daemon(AuthJSONRPCServer):
|
|||
|
||||
Options:
|
||||
--session_status : (bool) include session status in results
|
||||
--dht_status : (bool) include dht network and peer status
|
||||
|
||||
Returns:
|
||||
(dict) lbrynet-daemon status
|
||||
|
@ -1037,18 +1032,6 @@ class Daemon(AuthJSONRPCServer):
|
|||
'announce_queue_size': number of blobs currently queued to be announced
|
||||
'should_announce_blobs': number of blobs that should be announced
|
||||
}
|
||||
|
||||
If given the dht status option:
|
||||
'dht_status': {
|
||||
'kbps_received': current kbps receiving,
|
||||
'kbps_sent': current kdps being sent,
|
||||
'total_bytes_sent': total bytes sent,
|
||||
'total_bytes_received': total bytes received,
|
||||
'queries_received': number of queries received per second,
|
||||
'queries_sent': number of queries sent per second,
|
||||
'recent_contacts': count of recently contacted peers,
|
||||
'unique_contacts': count of unique peers
|
||||
},
|
||||
}
|
||||
"""
|
||||
|
||||
|
@ -1095,8 +1078,6 @@ class Daemon(AuthJSONRPCServer):
|
|||
'announce_queue_size': announce_queue_size,
|
||||
'should_announce_blobs': should_announce_blobs,
|
||||
}
|
||||
if dht_status:
|
||||
response['dht_status'] = self.session.dht_node.get_bandwidth_stats()
|
||||
defer.returnValue(response)
|
||||
|
||||
def jsonrpc_version(self):
|
||||
|
@ -2919,17 +2900,15 @@ class Daemon(AuthJSONRPCServer):
|
|||
return d
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def jsonrpc_blob_announce(self, blob_hash=None, stream_hash=None, sd_hash=None, announce_all=None):
|
||||
def jsonrpc_blob_announce(self, blob_hash=None, stream_hash=None, sd_hash=None):
|
||||
"""
|
||||
Announce blobs to the DHT
|
||||
|
||||
Usage:
|
||||
blob_announce [<blob_hash> | --blob_hash=<blob_hash>]
|
||||
[<stream_hash> | --stream_hash=<stream_hash>] | [<sd_hash> | --sd_hash=<sd_hash>]
|
||||
[--announce_all]
|
||||
|
||||
Options:
|
||||
--announce_all : (bool) announce all the blobs possessed by user
|
||||
--blob_hash=<blob_hash> : (str) announce a blob, specified by blob_hash
|
||||
--stream_hash=<stream_hash> : (str) announce all blobs associated with
|
||||
stream_hash
|
||||
|
@ -2940,41 +2919,22 @@ class Daemon(AuthJSONRPCServer):
|
|||
(bool) true if successful
|
||||
"""
|
||||
|
||||
if announce_all:
|
||||
yield self.session.blob_manager.immediate_announce_all_blobs()
|
||||
blob_hashes = []
|
||||
if blob_hash:
|
||||
blob_hashes.append(blob_hash)
|
||||
elif stream_hash or sd_hash:
|
||||
if sd_hash and stream_hash:
|
||||
raise Exception("either the sd hash or the stream hash should be provided, not both")
|
||||
if sd_hash:
|
||||
stream_hash = yield self.storage.get_stream_hash_for_sd_hash(sd_hash)
|
||||
blobs = yield self.storage.get_blobs_for_stream(stream_hash, only_completed=True)
|
||||
blob_hashes.extend(blob.blob_hash for blob in blobs if blob.blob_hash is not None)
|
||||
else:
|
||||
blob_hashes = []
|
||||
if blob_hash:
|
||||
blob_hashes.append(blob_hash)
|
||||
elif stream_hash or sd_hash:
|
||||
if sd_hash and stream_hash:
|
||||
raise Exception("either the sd hash or the stream hash should be provided, not both")
|
||||
if sd_hash:
|
||||
stream_hash = yield self.storage.get_stream_hash_for_sd_hash(sd_hash)
|
||||
blobs = yield self.storage.get_blobs_for_stream(stream_hash, only_completed=True)
|
||||
blob_hashes.extend([blob.blob_hash for blob in blobs if blob.blob_hash is not None])
|
||||
else:
|
||||
raise Exception('single argument must be specified')
|
||||
yield self.session.blob_manager.immediate_announce(blob_hashes)
|
||||
raise Exception('single argument must be specified')
|
||||
yield self.storage.should_single_announce_blobs(blob_hashes, immediate=True)
|
||||
response = yield self._render_response(True)
|
||||
defer.returnValue(response)
|
||||
|
||||
@AuthJSONRPCServer.deprecated("blob_announce")
|
||||
def jsonrpc_blob_announce_all(self):
|
||||
"""
|
||||
Announce all blobs to the DHT
|
||||
|
||||
Usage:
|
||||
blob_announce_all
|
||||
|
||||
Options:
|
||||
None
|
||||
|
||||
Returns:
|
||||
(str) Success/fail message
|
||||
"""
|
||||
return self.jsonrpc_blob_announce(announce_all=True)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def jsonrpc_file_reflect(self, **kwargs):
|
||||
"""
|
||||
|
|
|
@ -39,6 +39,7 @@ class DaemonServer(object):
|
|||
try:
|
||||
self.server_port = reactor.listenTCP(
|
||||
conf.settings['api_port'], lbrynet_server, interface=conf.settings['api_host'])
|
||||
log.info("lbrynet API listening on TCP %s:%i", conf.settings['api_host'], conf.settings['api_port'])
|
||||
except error.CannotListenError:
|
||||
log.info('Daemon already running, exiting app')
|
||||
raise
|
||||
|
|
|
@ -6,22 +6,20 @@ def migrate_db(db_dir, start, end):
|
|||
while current < end:
|
||||
if current == 1:
|
||||
from lbrynet.database.migrator.migrate1to2 import do_migration
|
||||
do_migration(db_dir)
|
||||
elif current == 2:
|
||||
from lbrynet.database.migrator.migrate2to3 import do_migration
|
||||
do_migration(db_dir)
|
||||
elif current == 3:
|
||||
from lbrynet.database.migrator.migrate3to4 import do_migration
|
||||
do_migration(db_dir)
|
||||
elif current == 4:
|
||||
from lbrynet.database.migrator.migrate4to5 import do_migration
|
||||
do_migration(db_dir)
|
||||
elif current == 5:
|
||||
from lbrynet.database.migrator.migrate5to6 import do_migration
|
||||
do_migration(db_dir)
|
||||
elif current == 6:
|
||||
from lbrynet.database.migrator.migrate6to7 import do_migration
|
||||
else:
|
||||
raise Exception("DB migration of version {} to {} is not available".format(current,
|
||||
current+1))
|
||||
do_migration(db_dir)
|
||||
current += 1
|
||||
return None
|
||||
|
||||
|
|
13
lbrynet/database/migrator/migrate6to7.py
Normal file
13
lbrynet/database/migrator/migrate6to7.py
Normal file
|
@ -0,0 +1,13 @@
|
|||
import sqlite3
|
||||
import os
|
||||
|
||||
|
||||
def do_migration(db_dir):
|
||||
db_path = os.path.join(db_dir, "lbrynet.sqlite")
|
||||
connection = sqlite3.connect(db_path)
|
||||
cursor = connection.cursor()
|
||||
cursor.executescript("alter table blob add last_announced_time integer;")
|
||||
cursor.executescript("alter table blob add single_announce integer;")
|
||||
cursor.execute("update blob set next_announce_time=0")
|
||||
connection.commit()
|
||||
connection.close()
|
|
@ -1,6 +1,5 @@
|
|||
import logging
|
||||
import os
|
||||
import time
|
||||
import sqlite3
|
||||
import traceback
|
||||
from decimal import Decimal
|
||||
|
@ -11,6 +10,7 @@ from lbryschema.claim import ClaimDict
|
|||
from lbryschema.decode import smart_decode
|
||||
from lbrynet import conf
|
||||
from lbrynet.cryptstream.CryptBlob import CryptBlobInfo
|
||||
from lbrynet.dht.constants import dataExpireTimeout
|
||||
from lbryum.constants import COIN
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
|
@ -49,26 +49,6 @@ def open_file_for_writing(download_directory, suggested_file_name):
|
|||
return threads.deferToThread(_open_file_for_writing, download_directory, suggested_file_name)
|
||||
|
||||
|
||||
def get_next_announce_time(hash_announcer, num_hashes_to_announce=1, min_reannounce_time=60*60,
|
||||
single_announce_duration=5):
|
||||
"""
|
||||
Hash reannounce time is set to current time + MIN_HASH_REANNOUNCE_TIME,
|
||||
unless we are announcing a lot of hashes at once which could cause the
|
||||
the announce queue to pile up. To prevent pile up, reannounce
|
||||
only after a conservative estimate of when it will finish
|
||||
to announce all the hashes.
|
||||
|
||||
Args:
|
||||
num_hashes_to_announce: number of hashes that will be added to the queue
|
||||
Returns:
|
||||
timestamp for next announce time
|
||||
"""
|
||||
queue_size = hash_announcer.hash_queue_size() + num_hashes_to_announce
|
||||
reannounce = max(min_reannounce_time,
|
||||
queue_size * single_announce_duration)
|
||||
return time.time() + reannounce
|
||||
|
||||
|
||||
def rerun_if_locked(f):
|
||||
max_attempts = 3
|
||||
|
||||
|
@ -124,7 +104,9 @@ class SQLiteStorage(object):
|
|||
blob_length integer not null,
|
||||
next_announce_time integer not null,
|
||||
should_announce integer not null default 0,
|
||||
status text not null
|
||||
status text not null,
|
||||
last_announced_time integer,
|
||||
single_announce integer
|
||||
);
|
||||
|
||||
create table if not exists stream (
|
||||
|
@ -185,6 +167,7 @@ class SQLiteStorage(object):
|
|||
log.info("connecting to database: %s", self._db_path)
|
||||
self.db = SqliteConnection(self._db_path)
|
||||
self.db.set_reactor(reactor)
|
||||
self.clock = reactor
|
||||
|
||||
# used to refresh the claim attributes on a ManagedEncryptedFileDownloader when a
|
||||
# change to the associated content claim occurs. these are added by the file manager
|
||||
|
@ -229,6 +212,7 @@ class SQLiteStorage(object):
|
|||
)
|
||||
|
||||
def set_should_announce(self, blob_hash, next_announce_time, should_announce):
|
||||
next_announce_time = next_announce_time or 0
|
||||
should_announce = 1 if should_announce else 0
|
||||
return self.db.runOperation(
|
||||
"update blob set next_announce_time=?, should_announce=? where blob_hash=?",
|
||||
|
@ -250,8 +234,8 @@ class SQLiteStorage(object):
|
|||
status = yield self.get_blob_status(blob_hash)
|
||||
if status is None:
|
||||
status = "pending"
|
||||
yield self.db.runOperation("insert into blob values (?, ?, ?, ?, ?)",
|
||||
(blob_hash, length, 0, 0, status))
|
||||
yield self.db.runOperation("insert into blob values (?, ?, ?, ?, ?, ?, ?)",
|
||||
(blob_hash, length, 0, 0, status, 0, 0))
|
||||
defer.returnValue(status)
|
||||
|
||||
def should_announce(self, blob_hash):
|
||||
|
@ -269,13 +253,35 @@ class SQLiteStorage(object):
|
|||
"select blob_hash from blob where should_announce=1 and status='finished'"
|
||||
)
|
||||
|
||||
def get_blobs_to_announce(self, hash_announcer):
|
||||
def update_last_announced_blob(self, blob_hash, last_announced):
|
||||
return self.db.runOperation(
|
||||
"update blob set next_announce_time=?, last_announced_time=?, single_announce=0 where blob_hash=?",
|
||||
(int(last_announced + (dataExpireTimeout / 2)), int(last_announced), blob_hash)
|
||||
)
|
||||
|
||||
def should_single_announce_blobs(self, blob_hashes, immediate=False):
|
||||
def set_single_announce(transaction):
|
||||
now = self.clock.seconds()
|
||||
for blob_hash in blob_hashes:
|
||||
if immediate:
|
||||
transaction.execute(
|
||||
"update blob set single_announce=1, next_announce_time=? "
|
||||
"where blob_hash=? and status='finished'", (int(now), blob_hash)
|
||||
)
|
||||
else:
|
||||
transaction.execute(
|
||||
"update blob set single_announce=1 where blob_hash=? and status='finished'", (blob_hash, )
|
||||
)
|
||||
return self.db.runInteraction(set_single_announce)
|
||||
|
||||
def get_blobs_to_announce(self):
|
||||
def get_and_update(transaction):
|
||||
timestamp = time.time()
|
||||
timestamp = self.clock.seconds()
|
||||
if conf.settings['announce_head_blobs_only']:
|
||||
r = transaction.execute(
|
||||
"select blob_hash from blob "
|
||||
"where blob_hash is not null and should_announce=1 and next_announce_time<? and status='finished'",
|
||||
"where blob_hash is not null and "
|
||||
"(should_announce=1 or single_announce=1) and next_announce_time<? and status='finished'",
|
||||
(timestamp,)
|
||||
)
|
||||
else:
|
||||
|
@ -283,16 +289,8 @@ class SQLiteStorage(object):
|
|||
"select blob_hash from blob where blob_hash is not null "
|
||||
"and next_announce_time<? and status='finished'", (timestamp,)
|
||||
)
|
||||
|
||||
blobs = [b for b, in r.fetchall()]
|
||||
next_announce_time = get_next_announce_time(hash_announcer, len(blobs))
|
||||
transaction.execute(
|
||||
"update blob set next_announce_time=? where next_announce_time<?", (next_announce_time, timestamp)
|
||||
)
|
||||
log.debug("Got %s blobs to announce, next announce time is in %s seconds", len(blobs),
|
||||
next_announce_time-time.time())
|
||||
blobs = [b[0] for b in r.fetchall()]
|
||||
return blobs
|
||||
|
||||
return self.db.runInteraction(get_and_update)
|
||||
|
||||
def delete_blobs_from_db(self, blob_hashes):
|
||||
|
|
|
@ -1,22 +0,0 @@
|
|||
import time
|
||||
|
||||
|
||||
class Delay(object):
|
||||
maxToSendDelay = 10 ** -3 # 0.05
|
||||
minToSendDelay = 10 ** -5 # 0.01
|
||||
|
||||
def __init__(self, start=0):
|
||||
self._next = start
|
||||
|
||||
# TODO: explain why this logic is like it is. And add tests that
|
||||
# show that it actually does what it needs to do.
|
||||
def __call__(self):
|
||||
ts = time.time()
|
||||
delay = 0
|
||||
if ts >= self._next:
|
||||
delay = self.minToSendDelay
|
||||
self._next = ts + self.minToSendDelay
|
||||
else:
|
||||
delay = (self._next - ts) + self.maxToSendDelay
|
||||
self._next += self.maxToSendDelay
|
||||
return delay
|
22
lbrynet/dht/distance.py
Normal file
22
lbrynet/dht/distance.py
Normal file
|
@ -0,0 +1,22 @@
|
|||
class Distance(object):
|
||||
"""Calculate the XOR result between two string variables.
|
||||
|
||||
Frequently we re-use one of the points so as an optimization
|
||||
we pre-calculate the long value of that point.
|
||||
"""
|
||||
|
||||
def __init__(self, key):
|
||||
self.key = key
|
||||
self.val_key_one = long(key.encode('hex'), 16)
|
||||
|
||||
def __call__(self, key_two):
|
||||
val_key_two = long(key_two.encode('hex'), 16)
|
||||
return self.val_key_one ^ val_key_two
|
||||
|
||||
def is_closer(self, a, b):
|
||||
"""Returns true is `a` is closer to `key` than `b` is"""
|
||||
return self(a) < self(b)
|
||||
|
||||
def to_contact(self, contact):
|
||||
"""A convenience function for calculating the distance to a contact"""
|
||||
return self(contact.id)
|
76
lbrynet/dht/hashannouncer.py
Normal file
76
lbrynet/dht/hashannouncer.py
Normal file
|
@ -0,0 +1,76 @@
|
|||
import binascii
|
||||
import logging
|
||||
|
||||
from twisted.internet import defer, task
|
||||
from lbrynet.core import utils
|
||||
from lbrynet import conf
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class DHTHashAnnouncer(object):
|
||||
def __init__(self, dht_node, storage, concurrent_announcers=None):
|
||||
self.dht_node = dht_node
|
||||
self.storage = storage
|
||||
self.clock = dht_node.clock
|
||||
self.peer_port = dht_node.peerPort
|
||||
self.hash_queue = []
|
||||
self.concurrent_announcers = concurrent_announcers or conf.settings['concurrent_announcers']
|
||||
self._manage_lc = task.LoopingCall(self.manage)
|
||||
self._manage_lc.clock = self.clock
|
||||
|
||||
def start(self):
|
||||
self._manage_lc.start(30)
|
||||
|
||||
def stop(self):
|
||||
if self._manage_lc.running:
|
||||
self._manage_lc.stop()
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def do_store(self, blob_hash):
|
||||
storing_node_ids = yield self.dht_node.announceHaveBlob(binascii.unhexlify(blob_hash))
|
||||
now = self.clock.seconds()
|
||||
if storing_node_ids:
|
||||
result = (now, storing_node_ids)
|
||||
yield self.storage.update_last_announced_blob(blob_hash, now)
|
||||
log.debug("Stored %s to %i peers", blob_hash[:16], len(storing_node_ids))
|
||||
else:
|
||||
result = (None, [])
|
||||
self.hash_queue.remove(blob_hash)
|
||||
defer.returnValue(result)
|
||||
|
||||
def _show_announce_progress(self, size, start):
|
||||
queue_size = len(self.hash_queue)
|
||||
average_blobs_per_second = float(size - queue_size) / (self.clock.seconds() - start)
|
||||
log.info("Announced %i/%i blobs, %f blobs per second", size - queue_size, size, average_blobs_per_second)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def immediate_announce(self, blob_hashes):
|
||||
self.hash_queue.extend(b for b in blob_hashes if b not in self.hash_queue)
|
||||
|
||||
log.info("Announcing %i blobs", len(self.hash_queue))
|
||||
start = self.clock.seconds()
|
||||
progress_lc = task.LoopingCall(self._show_announce_progress, len(self.hash_queue), start)
|
||||
progress_lc.start(60, now=False)
|
||||
s = defer.DeferredSemaphore(self.concurrent_announcers)
|
||||
results = yield utils.DeferredDict({blob_hash: s.run(self.do_store, blob_hash) for blob_hash in blob_hashes})
|
||||
now = self.clock.seconds()
|
||||
|
||||
progress_lc.stop()
|
||||
|
||||
announced_to = [blob_hash for blob_hash in results if results[blob_hash][0]]
|
||||
if len(announced_to) != len(results):
|
||||
log.debug("Failed to announce %i blobs", len(results) - len(announced_to))
|
||||
if announced_to:
|
||||
log.info('Took %s seconds to announce %i of %i attempted hashes (%f hashes per second)',
|
||||
now - start, len(announced_to), len(blob_hashes),
|
||||
int(float(len(blob_hashes)) / float(now - start)))
|
||||
defer.returnValue(results)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def manage(self):
|
||||
need_reannouncement = yield self.storage.get_blobs_to_announce()
|
||||
if need_reannouncement:
|
||||
yield self.immediate_announce(need_reannouncement)
|
||||
else:
|
||||
log.debug("Nothing to announce")
|
|
@ -1,24 +1,22 @@
|
|||
from collections import Counter
|
||||
import datetime
|
||||
from twisted.internet import task
|
||||
|
||||
|
||||
class HashWatcher(object):
|
||||
def __init__(self):
|
||||
def __init__(self, clock=None):
|
||||
if not clock:
|
||||
from twisted.internet import reactor as clock
|
||||
self.ttl = 600
|
||||
self.hashes = []
|
||||
self.next_tick = None
|
||||
self.lc = task.LoopingCall(self._remove_old_hashes)
|
||||
self.lc.clock = clock
|
||||
|
||||
def tick(self):
|
||||
|
||||
from twisted.internet import reactor
|
||||
|
||||
self._remove_old_hashes()
|
||||
self.next_tick = reactor.callLater(10, self.tick)
|
||||
def start(self):
|
||||
return self.lc.start(10)
|
||||
|
||||
def stop(self):
|
||||
if self.next_tick is not None:
|
||||
self.next_tick.cancel()
|
||||
self.next_tick = None
|
||||
return self.lc.stop()
|
||||
|
||||
def add_requested_hash(self, hashsum, contact):
|
||||
from_ip = contact.compact_ip
|
||||
|
|
|
@ -11,19 +11,23 @@ import hashlib
|
|||
import operator
|
||||
import struct
|
||||
import time
|
||||
import logging
|
||||
from twisted.internet import defer, error, task
|
||||
|
||||
from twisted.internet import defer, error, reactor, threads, task
|
||||
from lbrynet.core.utils import generate_id
|
||||
from lbrynet.core.call_later_manager import CallLaterManager
|
||||
from lbrynet.core.PeerManager import PeerManager
|
||||
|
||||
import constants
|
||||
import routingtable
|
||||
import datastore
|
||||
import protocol
|
||||
|
||||
from error import TimeoutError
|
||||
from peerfinder import DHTPeerFinder
|
||||
from contact import Contact
|
||||
from hashwatcher import HashWatcher
|
||||
import logging
|
||||
from distance import Distance
|
||||
|
||||
from lbrynet.core.utils import generate_id
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
|
||||
|
@ -51,7 +55,9 @@ class Node(object):
|
|||
|
||||
def __init__(self, node_id=None, udpPort=4000, dataStore=None,
|
||||
routingTableClass=None, networkProtocol=None,
|
||||
externalIP=None, peerPort=None):
|
||||
externalIP=None, peerPort=None, listenUDP=None,
|
||||
callLater=None, resolve=None, clock=None, peer_finder=None,
|
||||
peer_manager=None):
|
||||
"""
|
||||
@param dataStore: The data store to use. This must be class inheriting
|
||||
from the C{DataStore} interface (or providing the
|
||||
|
@ -76,6 +82,18 @@ class Node(object):
|
|||
@param externalIP: the IP at which this node can be contacted
|
||||
@param peerPort: the port at which this node announces it has a blob for
|
||||
"""
|
||||
|
||||
if not listenUDP or not resolve or not callLater or not clock:
|
||||
from twisted.internet import reactor
|
||||
listenUDP = listenUDP or reactor.listenUDP
|
||||
resolve = resolve or reactor.resolve
|
||||
callLater = callLater or reactor.callLater
|
||||
clock = clock or reactor
|
||||
self.clock = clock
|
||||
CallLaterManager.setup(callLater)
|
||||
self.reactor_resolve = resolve
|
||||
self.reactor_listenUDP = listenUDP
|
||||
self.reactor_callLater = CallLaterManager.call_later
|
||||
self.node_id = node_id or self._generateID()
|
||||
self.port = udpPort
|
||||
self._listeningPort = None # object implementing Twisted
|
||||
|
@ -84,14 +102,17 @@ class Node(object):
|
|||
# information from the DHT as soon as the node is part of the
|
||||
# network (add callbacks to this deferred if scheduling such
|
||||
# operations before the node has finished joining the network)
|
||||
self._joinDeferred = None
|
||||
self.next_refresh_call = None
|
||||
self._joinDeferred = defer.Deferred(None)
|
||||
self.change_token_lc = task.LoopingCall(self.change_token)
|
||||
self.change_token_lc.clock = self.clock
|
||||
self.refresh_node_lc = task.LoopingCall(self._refreshNode)
|
||||
self.refresh_node_lc.clock = self.clock
|
||||
|
||||
# Create k-buckets (for storing contacts)
|
||||
if routingTableClass is None:
|
||||
self._routingTable = routingtable.OptimizedTreeRoutingTable(self.node_id)
|
||||
self._routingTable = routingtable.OptimizedTreeRoutingTable(self.node_id, self.clock.seconds)
|
||||
else:
|
||||
self._routingTable = routingTableClass(self.node_id)
|
||||
self._routingTable = routingTableClass(self.node_id, self.clock.seconds)
|
||||
|
||||
# Initialize this node's network access mechanisms
|
||||
if networkProtocol is None:
|
||||
|
@ -115,75 +136,95 @@ class Node(object):
|
|||
self._routingTable.addContact(contact)
|
||||
self.externalIP = externalIP
|
||||
self.peerPort = peerPort
|
||||
self.hash_watcher = HashWatcher()
|
||||
self.hash_watcher = HashWatcher(self.clock)
|
||||
|
||||
self.peer_manager = peer_manager or PeerManager()
|
||||
self.peer_finder = peer_finder or DHTPeerFinder(self, self.peer_manager)
|
||||
|
||||
def __del__(self):
|
||||
log.warning("unclean shutdown of the dht node")
|
||||
if self._listeningPort is not None:
|
||||
self._listeningPort.stopListening()
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def stop(self):
|
||||
# cancel callLaters:
|
||||
if self.next_refresh_call is not None:
|
||||
self.next_refresh_call.cancel()
|
||||
self.next_refresh_call = None
|
||||
# stop LoopingCalls:
|
||||
if self.refresh_node_lc.running:
|
||||
yield self.refresh_node_lc.stop()
|
||||
if self.change_token_lc.running:
|
||||
self.change_token_lc.stop()
|
||||
yield self.change_token_lc.stop()
|
||||
if self._listeningPort is not None:
|
||||
self._listeningPort.stopListening()
|
||||
self.hash_watcher.stop()
|
||||
yield self._listeningPort.stopListening()
|
||||
if self.hash_watcher.lc.running:
|
||||
yield self.hash_watcher.stop()
|
||||
|
||||
def startNetwork(self):
|
||||
""" Causes the Node to start all the underlying components needed for the DHT
|
||||
to work. This should be called before any other DHT operations.
|
||||
"""
|
||||
log.info("Starting DHT underlying components")
|
||||
|
||||
# Prepare the underlying Kademlia protocol
|
||||
if self.port is not None:
|
||||
def start_listening(self):
|
||||
if not self._listeningPort:
|
||||
try:
|
||||
self._listeningPort = reactor.listenUDP(self.port, self._protocol)
|
||||
self._listeningPort = self.reactor_listenUDP(self.port, self._protocol)
|
||||
except error.CannotListenError as e:
|
||||
import traceback
|
||||
log.error("Couldn't bind to port %d. %s", self.port, traceback.format_exc())
|
||||
raise ValueError("%s lbrynet may already be running." % str(e))
|
||||
else:
|
||||
log.warning("Already bound to port %d", self._listeningPort.port)
|
||||
|
||||
# Start the token looping call
|
||||
self.change_token_lc.start(constants.tokenSecretChangeInterval)
|
||||
# #TODO: Refresh all k-buckets further away than this node's closest neighbour
|
||||
# Start refreshing k-buckets periodically, if necessary
|
||||
self.next_refresh_call = reactor.callLater(constants.checkRefreshInterval,
|
||||
self._refreshNode)
|
||||
self.hash_watcher.tick()
|
||||
def bootstrap_join(self, known_node_addresses, finished_d):
|
||||
"""
|
||||
Attempt to join the dht, retry every 30 seconds if unsuccessful
|
||||
:param known_node_addresses: [(str, int)] list of hostnames and ports for known dht seed nodes
|
||||
:param finished_d: (defer.Deferred) called when join succeeds
|
||||
"""
|
||||
@defer.inlineCallbacks
|
||||
def _resolve_seeds():
|
||||
bootstrap_contacts = []
|
||||
for node_address, port in known_node_addresses:
|
||||
host = yield self.reactor_resolve(node_address)
|
||||
# Create temporary contact information for the list of addresses of known nodes
|
||||
contact = Contact(self._generateID(), host, port, self._protocol)
|
||||
bootstrap_contacts.append(contact)
|
||||
if not bootstrap_contacts:
|
||||
if not self.hasContacts():
|
||||
log.warning("No known contacts!")
|
||||
else:
|
||||
log.info("found contacts")
|
||||
bootstrap_contacts = self.contacts
|
||||
defer.returnValue(bootstrap_contacts)
|
||||
|
||||
def _rerun(closest_nodes):
|
||||
if not closest_nodes:
|
||||
log.info("Failed to join the dht, re-attempting in 30 seconds")
|
||||
self.reactor_callLater(30, self.bootstrap_join, known_node_addresses, finished_d)
|
||||
elif not finished_d.called:
|
||||
finished_d.callback(closest_nodes)
|
||||
|
||||
log.info("Attempting to join the DHT network")
|
||||
d = _resolve_seeds()
|
||||
# Initiate the Kademlia joining sequence - perform a search for this node's own ID
|
||||
d.addCallback(lambda contacts: self._iterativeFind(self.node_id, contacts))
|
||||
d.addCallback(_rerun)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def joinNetwork(self, knownNodeAddresses=None):
|
||||
def joinNetwork(self, known_node_addresses=None):
|
||||
""" Causes the Node to attempt to join the DHT network by contacting the
|
||||
known DHT nodes. This can be called multiple times if the previous attempt
|
||||
has failed or if the Node has lost all the contacts.
|
||||
|
||||
@param knownNodeAddresses: A sequence of tuples containing IP address
|
||||
@param known_node_addresses: A sequence of tuples containing IP address
|
||||
information for existing nodes on the
|
||||
Kademlia network, in the format:
|
||||
C{(<ip address>, (udp port>)}
|
||||
@type knownNodeAddresses: tuple
|
||||
@type known_node_addresses: list
|
||||
"""
|
||||
log.info("Attempting to join the DHT network")
|
||||
|
||||
# IGNORE:E1101
|
||||
# Create temporary contact information for the list of addresses of known nodes
|
||||
if knownNodeAddresses != None:
|
||||
bootstrapContacts = []
|
||||
for address, port in knownNodeAddresses:
|
||||
contact = Contact(self._generateID(), address, port, self._protocol)
|
||||
bootstrapContacts.append(contact)
|
||||
else:
|
||||
bootstrapContacts = None
|
||||
|
||||
# Initiate the Kademlia joining sequence - perform a search for this node's own ID
|
||||
self._joinDeferred = self._iterativeFind(self.node_id, bootstrapContacts)
|
||||
|
||||
result = yield self._joinDeferred
|
||||
defer.returnValue(result)
|
||||
self.start_listening()
|
||||
# #TODO: Refresh all k-buckets further away than this node's closest neighbour
|
||||
# Start refreshing k-buckets periodically, if necessary
|
||||
self.bootstrap_join(known_node_addresses or [], self._joinDeferred)
|
||||
yield self._joinDeferred
|
||||
self.hash_watcher.start()
|
||||
self.change_token_lc.start(constants.tokenSecretChangeInterval)
|
||||
self.refresh_node_lc.start(constants.checkRefreshInterval)
|
||||
|
||||
@property
|
||||
def contacts(self):
|
||||
|
@ -193,41 +234,19 @@ class Node(object):
|
|||
yield contact
|
||||
return list(_inner())
|
||||
|
||||
def printContacts(self, *args):
|
||||
print '\n\nNODE CONTACTS\n==============='
|
||||
for i in range(len(self._routingTable._buckets)):
|
||||
print "bucket %i" % i
|
||||
for contact in self._routingTable._buckets[i]._contacts:
|
||||
print " %s:%i" % (contact.address, contact.port)
|
||||
print '=================================='
|
||||
|
||||
def hasContacts(self):
|
||||
for bucket in self._routingTable._buckets:
|
||||
if bucket._contacts:
|
||||
return True
|
||||
return False
|
||||
|
||||
def getApproximateTotalDHTNodes(self):
|
||||
# get the deepest bucket and the number of contacts in that bucket and multiply it
|
||||
# by the number of equivalently deep buckets in the whole DHT to get a really bad
|
||||
# estimate!
|
||||
bucket = self._routingTable._buckets[self._routingTable._kbucketIndex(self.node_id)]
|
||||
num_in_bucket = len(bucket._contacts)
|
||||
factor = (2 ** constants.key_bits) / (bucket.rangeMax - bucket.rangeMin)
|
||||
return num_in_bucket * factor
|
||||
|
||||
def getApproximateTotalHashes(self):
|
||||
# Divide the number of hashes we know about by k to get a really, really, really
|
||||
# bad estimate of the average number of hashes per node, then multiply by the
|
||||
# approximate number of nodes to get a horrendous estimate of the total number
|
||||
# of hashes in the DHT
|
||||
num_in_data_store = len(self._dataStore._dict)
|
||||
if num_in_data_store == 0:
|
||||
return 0
|
||||
return num_in_data_store * self.getApproximateTotalDHTNodes() / 8
|
||||
|
||||
def announceHaveBlob(self, key):
|
||||
return self.iterativeAnnounceHaveBlob(key, {'port': self.peerPort, 'lbryid': self.node_id})
|
||||
return self.iterativeAnnounceHaveBlob(
|
||||
key, {
|
||||
'port': self.peerPort,
|
||||
'lbryid': self.node_id,
|
||||
}
|
||||
)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def getPeersForBlob(self, blob_hash):
|
||||
|
@ -245,69 +264,57 @@ class Node(object):
|
|||
def get_most_popular_hashes(self, num_to_return):
|
||||
return self.hash_watcher.most_popular_hashes(num_to_return)
|
||||
|
||||
def get_bandwidth_stats(self):
|
||||
return self._protocol.bandwidth_stats
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def iterativeAnnounceHaveBlob(self, blob_hash, value):
|
||||
known_nodes = {}
|
||||
contacts = yield self.iterativeFindNode(blob_hash)
|
||||
# store locally if we're the closest node and there are less than k contacts to try storing to
|
||||
if self.externalIP is not None and contacts and len(contacts) < constants.k:
|
||||
is_closer = Distance(blob_hash).is_closer(self.node_id, contacts[-1].id)
|
||||
if is_closer:
|
||||
contacts.pop()
|
||||
yield self.store(blob_hash, value, originalPublisherID=self.node_id,
|
||||
self_store=True)
|
||||
elif self.externalIP is not None:
|
||||
pass
|
||||
else:
|
||||
raise Exception("Cannot determine external IP: %s" % self.externalIP)
|
||||
|
||||
def log_error(err, n):
|
||||
if err.check(protocol.TimeoutError):
|
||||
log.debug(
|
||||
"Timeout while storing blob_hash %s at %s",
|
||||
binascii.hexlify(blob_hash), n)
|
||||
else:
|
||||
log.error(
|
||||
"Unexpected error while storing blob_hash %s at %s: %s",
|
||||
binascii.hexlify(blob_hash), n, err.getErrorMessage())
|
||||
contacted = []
|
||||
|
||||
def log_success(res):
|
||||
log.debug("Response to store request: %s", str(res))
|
||||
return res
|
||||
@defer.inlineCallbacks
|
||||
def announce_to_contact(contact):
|
||||
known_nodes[contact.id] = contact
|
||||
try:
|
||||
responseMsg, originAddress = yield contact.findValue(blob_hash, rawResponse=True)
|
||||
if responseMsg.nodeID != contact.id:
|
||||
raise Exception("node id mismatch")
|
||||
value['token'] = responseMsg.response['token']
|
||||
res = yield contact.store(blob_hash, value)
|
||||
if res != "OK":
|
||||
raise ValueError(res)
|
||||
contacted.append(contact)
|
||||
log.debug("Stored %s to %s (%s)", blob_hash.encode('hex'), contact.id.encode('hex'), originAddress[0])
|
||||
except protocol.TimeoutError:
|
||||
log.debug("Timeout while storing blob_hash %s at %s",
|
||||
blob_hash.encode('hex')[:16], contact.id.encode('hex'))
|
||||
except ValueError as err:
|
||||
log.error("Unexpected response: %s" % err.message)
|
||||
except Exception as err:
|
||||
log.error("Unexpected error while storing blob_hash %s at %s: %s",
|
||||
binascii.hexlify(blob_hash), contact, err)
|
||||
|
||||
def announce_to_peer(responseTuple):
|
||||
""" @type responseMsg: kademlia.msgtypes.ResponseMessage """
|
||||
# The "raw response" tuple contains the response message,
|
||||
# and the originating address info
|
||||
responseMsg = responseTuple[0]
|
||||
originAddress = responseTuple[1] # tuple: (ip adress, udp port)
|
||||
# Make sure the responding node is valid, and abort the operation if it isn't
|
||||
if not responseMsg.nodeID in known_nodes:
|
||||
return responseMsg.nodeID
|
||||
dl = []
|
||||
for c in contacts:
|
||||
dl.append(announce_to_contact(c))
|
||||
|
||||
n = known_nodes[responseMsg.nodeID]
|
||||
yield defer.DeferredList(dl)
|
||||
|
||||
result = responseMsg.response
|
||||
if 'token' in result:
|
||||
value['token'] = result['token']
|
||||
d = n.store(blob_hash, value, self.node_id, 0)
|
||||
d.addCallback(log_success)
|
||||
d.addErrback(log_error, n)
|
||||
else:
|
||||
d = defer.succeed(False)
|
||||
return d
|
||||
log.debug("Stored %s to %i of %i attempted peers", blob_hash.encode('hex')[:16],
|
||||
len(contacted), len(contacts))
|
||||
|
||||
def requestPeers(contacts):
|
||||
if self.externalIP is not None and len(contacts) >= constants.k:
|
||||
is_closer = Distance(blob_hash).is_closer(self.node_id, contacts[-1].id)
|
||||
if is_closer:
|
||||
contacts.pop()
|
||||
self.store(blob_hash, value, self_store=True, originalPublisherID=self.node_id)
|
||||
elif self.externalIP is not None:
|
||||
self.store(blob_hash, value, self_store=True, originalPublisherID=self.node_id)
|
||||
ds = []
|
||||
for contact in contacts:
|
||||
known_nodes[contact.id] = contact
|
||||
rpcMethod = getattr(contact, "findValue")
|
||||
df = rpcMethod(blob_hash, rawResponse=True)
|
||||
df.addCallback(announce_to_peer)
|
||||
df.addErrback(log_error, contact)
|
||||
ds.append(df)
|
||||
return defer.DeferredList(ds)
|
||||
|
||||
d = self.iterativeFindNode(blob_hash)
|
||||
d.addCallbacks(requestPeers)
|
||||
return d
|
||||
contacted_node_ids = [c.id.encode('hex') for c in contacted]
|
||||
defer.returnValue(contacted_node_ids)
|
||||
|
||||
def change_token(self):
|
||||
self.old_token_secret = self.token_secret
|
||||
|
@ -365,33 +372,26 @@ class Node(object):
|
|||
to the specified key
|
||||
@rtype: twisted.internet.defer.Deferred
|
||||
"""
|
||||
# Prepare a callback for this operation
|
||||
outerDf = defer.Deferred()
|
||||
|
||||
def checkResult(result):
|
||||
if isinstance(result, dict):
|
||||
# We have found the value; now see who was the closest contact without it...
|
||||
# ...and store the key/value pair
|
||||
outerDf.callback(result)
|
||||
else:
|
||||
# The value wasn't found, but a list of contacts was returned
|
||||
# Now, see if we have the value (it might seem wasteful to search on the network
|
||||
# first, but it ensures that all values are properly propagated through the
|
||||
# network
|
||||
if self._dataStore.hasPeersForBlob(key):
|
||||
# Ok, we have the value locally, so use that
|
||||
peers = self._dataStore.getPeersForBlob(key)
|
||||
# Send this value to the closest node without it
|
||||
outerDf.callback({key: peers})
|
||||
else:
|
||||
# Ok, value does not exist in DHT at all
|
||||
outerDf.callback(result)
|
||||
|
||||
# Execute the search
|
||||
iterative_find_result = yield self._iterativeFind(key, rpc='findValue')
|
||||
checkResult(iterative_find_result)
|
||||
result = yield outerDf
|
||||
defer.returnValue(result)
|
||||
if isinstance(iterative_find_result, dict):
|
||||
# We have found the value; now see who was the closest contact without it...
|
||||
# ...and store the key/value pair
|
||||
defer.returnValue(iterative_find_result)
|
||||
else:
|
||||
# The value wasn't found, but a list of contacts was returned
|
||||
# Now, see if we have the value (it might seem wasteful to search on the network
|
||||
# first, but it ensures that all values are properly propagated through the
|
||||
# network
|
||||
if self._dataStore.hasPeersForBlob(key):
|
||||
# Ok, we have the value locally, so use that
|
||||
# Send this value to the closest node without it
|
||||
peers = self._dataStore.getPeersForBlob(key)
|
||||
defer.returnValue({key: peers})
|
||||
else:
|
||||
# Ok, value does not exist in DHT at all
|
||||
defer.returnValue(iterative_find_result)
|
||||
|
||||
def addContact(self, contact):
|
||||
""" Add/update the given contact; simple wrapper for the same method
|
||||
|
@ -486,18 +486,20 @@ class Node(object):
|
|||
else:
|
||||
raise TypeError, 'No contact info available'
|
||||
|
||||
if ((self_store is False) and
|
||||
('token' not in value or not self.verify_token(value['token'], compact_ip))):
|
||||
raise ValueError('Invalid or missing token')
|
||||
if not self_store:
|
||||
if 'token' not in value:
|
||||
raise ValueError("Missing token")
|
||||
if not self.verify_token(value['token'], compact_ip):
|
||||
raise ValueError("Invalid token")
|
||||
|
||||
if 'port' in value:
|
||||
port = int(value['port'])
|
||||
if 0 <= port <= 65536:
|
||||
compact_port = str(struct.pack('>H', port))
|
||||
else:
|
||||
raise TypeError, 'Invalid port'
|
||||
raise TypeError('Invalid port')
|
||||
else:
|
||||
raise TypeError, 'No port available'
|
||||
raise TypeError('No port available')
|
||||
|
||||
if 'lbryid' in value:
|
||||
if len(value['lbryid']) != constants.key_bits / 8:
|
||||
|
@ -506,7 +508,7 @@ class Node(object):
|
|||
else:
|
||||
compact_address = compact_ip + compact_port + value['lbryid']
|
||||
else:
|
||||
raise TypeError, 'No lbryid given'
|
||||
raise TypeError('No lbryid given')
|
||||
|
||||
now = int(time.time())
|
||||
originallyPublished = now # - age
|
||||
|
@ -628,39 +630,22 @@ class Node(object):
|
|||
result = yield outerDf
|
||||
defer.returnValue(result)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def _refreshNode(self):
|
||||
""" Periodically called to perform k-bucket refreshes and data
|
||||
replication/republishing as necessary """
|
||||
|
||||
df = self._refreshRoutingTable()
|
||||
df.addCallback(self._removeExpiredPeers)
|
||||
df.addCallback(self._scheduleNextNodeRefresh)
|
||||
yield self._refreshRoutingTable()
|
||||
self._dataStore.removeExpiredPeers()
|
||||
defer.returnValue(None)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def _refreshRoutingTable(self):
|
||||
nodeIDs = self._routingTable.getRefreshList(0, False)
|
||||
outerDf = defer.Deferred()
|
||||
|
||||
def searchForNextNodeID(dfResult=None):
|
||||
if len(nodeIDs) > 0:
|
||||
searchID = nodeIDs.pop()
|
||||
df = self.iterativeFindNode(searchID)
|
||||
df.addCallback(searchForNextNodeID)
|
||||
else:
|
||||
# If this is reached, we have finished refreshing the routing table
|
||||
outerDf.callback(None)
|
||||
|
||||
# Start the refreshing cycle
|
||||
searchForNextNodeID()
|
||||
return outerDf
|
||||
|
||||
def _scheduleNextNodeRefresh(self, *args):
|
||||
self.next_refresh_call = reactor.callLater(constants.checkRefreshInterval,
|
||||
self._refreshNode)
|
||||
|
||||
# args put here because _refreshRoutingTable does outerDF.callback(None)
|
||||
def _removeExpiredPeers(self, *args):
|
||||
df = threads.deferToThread(self._dataStore.removeExpiredPeers)
|
||||
return df
|
||||
while nodeIDs:
|
||||
searchID = nodeIDs.pop()
|
||||
yield self.iterativeFindNode(searchID)
|
||||
defer.returnValue(None)
|
||||
|
||||
|
||||
# This was originally a set of nested methods in _iterativeFind
|
||||
|
@ -770,7 +755,7 @@ class _IterativeFindHelper(object):
|
|||
|
||||
def removeFromShortlist(self, failure, deadContactID):
|
||||
""" @type failure: twisted.python.failure.Failure """
|
||||
failure.trap(protocol.TimeoutError)
|
||||
failure.trap(TimeoutError, defer.CancelledError, TypeError)
|
||||
if len(deadContactID) != constants.key_bits / 8:
|
||||
raise ValueError("invalid lbry id")
|
||||
if deadContactID in self.shortlist:
|
||||
|
@ -827,7 +812,7 @@ class _IterativeFindHelper(object):
|
|||
if self._should_lookup_active_calls():
|
||||
# Schedule the next iteration if there are any active
|
||||
# calls (Kademlia uses loose parallelism)
|
||||
call = reactor.callLater(constants.iterativeLookupDelay, self.searchIteration)
|
||||
call, _ = self.node.reactor_callLater(constants.iterativeLookupDelay, self.searchIteration)
|
||||
self.pending_iteration_calls.append(call)
|
||||
# Check for a quick contact response that made an update to the shortList
|
||||
elif prevShortlistLength < len(self.shortlist):
|
||||
|
@ -867,30 +852,6 @@ class _IterativeFindHelper(object):
|
|||
)
|
||||
|
||||
|
||||
class Distance(object):
|
||||
"""Calculate the XOR result between two string variables.
|
||||
|
||||
Frequently we re-use one of the points so as an optimization
|
||||
we pre-calculate the long value of that point.
|
||||
"""
|
||||
|
||||
def __init__(self, key):
|
||||
self.key = key
|
||||
self.val_key_one = long(key.encode('hex'), 16)
|
||||
|
||||
def __call__(self, key_two):
|
||||
val_key_two = long(key_two.encode('hex'), 16)
|
||||
return self.val_key_one ^ val_key_two
|
||||
|
||||
def is_closer(self, a, b):
|
||||
"""Returns true is `a` is closer to `key` than `b` is"""
|
||||
return self(a) < self(b)
|
||||
|
||||
def to_contact(self, contact):
|
||||
"""A convenience function for calculating the distance to a contact"""
|
||||
return self(contact.id)
|
||||
|
||||
|
||||
class ExpensiveSort(object):
|
||||
"""Sort a list in place.
|
||||
|
||||
|
|
|
@ -2,7 +2,7 @@ import binascii
|
|||
import logging
|
||||
|
||||
from zope.interface import implements
|
||||
from twisted.internet import defer, reactor
|
||||
from twisted.internet import defer
|
||||
from lbrynet.interfaces import IPeerFinder
|
||||
from lbrynet.core.utils import short_hash
|
||||
|
||||
|
@ -10,7 +10,23 @@ from lbrynet.core.utils import short_hash
|
|||
log = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class DHTPeerFinder(object):
|
||||
class DummyPeerFinder(object):
|
||||
"""This class finds peers which have announced to the DHT that they have certain blobs"""
|
||||
|
||||
def run_manage_loop(self):
|
||||
pass
|
||||
|
||||
def stop(self):
|
||||
pass
|
||||
|
||||
def find_peers_for_blob(self, blob_hash):
|
||||
return defer.succeed([])
|
||||
|
||||
def get_most_popular_hashes(self, num_to_return):
|
||||
return []
|
||||
|
||||
|
||||
class DHTPeerFinder(DummyPeerFinder):
|
||||
"""This class finds peers which have announced to the DHT that they have certain blobs"""
|
||||
implements(IPeerFinder)
|
||||
|
||||
|
@ -47,7 +63,7 @@ class DHTPeerFinder(object):
|
|||
finished_deferred = self.dht_node.getPeersForBlob(bin_hash)
|
||||
|
||||
if timeout is not None:
|
||||
reactor.callLater(timeout, _trigger_timeout)
|
||||
self.dht_node.reactor_callLater(timeout, _trigger_timeout)
|
||||
|
||||
try:
|
||||
peer_list = yield finished_deferred
|
|
@ -1,9 +1,9 @@
|
|||
import logging
|
||||
import time
|
||||
import socket
|
||||
import errno
|
||||
|
||||
from twisted.internet import protocol, defer, error, reactor, task
|
||||
from twisted.internet import protocol, defer
|
||||
from lbrynet.core.call_later_manager import CallLaterManager
|
||||
|
||||
import constants
|
||||
import encoding
|
||||
|
@ -11,7 +11,6 @@ import msgtypes
|
|||
import msgformat
|
||||
from contact import Contact
|
||||
from error import BUILTIN_EXCEPTIONS, UnknownRemoteException, TimeoutError
|
||||
from delay import Delay
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
|
||||
|
@ -28,104 +27,6 @@ class KademliaProtocol(protocol.DatagramProtocol):
|
|||
self._sentMessages = {}
|
||||
self._partialMessages = {}
|
||||
self._partialMessagesProgress = {}
|
||||
self._delay = Delay()
|
||||
# keep track of outstanding writes so that they
|
||||
# can be cancelled on shutdown
|
||||
self._call_later_list = {}
|
||||
|
||||
# keep track of bandwidth usage by peer
|
||||
self._history_rx = {}
|
||||
self._history_tx = {}
|
||||
self._bytes_rx = {}
|
||||
self._bytes_tx = {}
|
||||
self._unique_contacts = []
|
||||
self._queries_rx_per_second = 0
|
||||
self._queries_tx_per_second = 0
|
||||
self._kbps_tx = 0
|
||||
self._kbps_rx = 0
|
||||
self._recent_contact_count = 0
|
||||
self._total_bytes_tx = 0
|
||||
self._total_bytes_rx = 0
|
||||
self._bandwidth_stats_update_lc = task.LoopingCall(self._update_bandwidth_stats)
|
||||
|
||||
def _update_bandwidth_stats(self):
|
||||
recent_rx_history = {}
|
||||
now = time.time()
|
||||
for address, history in self._history_rx.iteritems():
|
||||
recent_rx_history[address] = [(s, t) for (s, t) in history if now - t < 1.0]
|
||||
qps_rx = sum(len(v) for (k, v) in recent_rx_history.iteritems())
|
||||
bps_rx = sum(sum([x[0] for x in v]) for (k, v) in recent_rx_history.iteritems())
|
||||
kbps_rx = round(float(bps_rx) / 1024.0, 2)
|
||||
|
||||
recent_tx_history = {}
|
||||
now = time.time()
|
||||
for address, history in self._history_tx.iteritems():
|
||||
recent_tx_history[address] = [(s, t) for (s, t) in history if now - t < 1.0]
|
||||
qps_tx = sum(len(v) for (k, v) in recent_tx_history.iteritems())
|
||||
bps_tx = sum(sum([x[0] for x in v]) for (k, v) in recent_tx_history.iteritems())
|
||||
kbps_tx = round(float(bps_tx) / 1024.0, 2)
|
||||
|
||||
recent_contacts = []
|
||||
for k, v in recent_rx_history.iteritems():
|
||||
if v:
|
||||
recent_contacts.append(k)
|
||||
for k, v in recent_tx_history.iteritems():
|
||||
if v and k not in recent_contacts:
|
||||
recent_contacts.append(k)
|
||||
|
||||
self._queries_rx_per_second = qps_rx
|
||||
self._queries_tx_per_second = qps_tx
|
||||
self._kbps_tx = kbps_tx
|
||||
self._kbps_rx = kbps_rx
|
||||
self._recent_contact_count = len(recent_contacts)
|
||||
self._total_bytes_tx = sum(v for (k, v) in self._bytes_tx.iteritems())
|
||||
self._total_bytes_rx = sum(v for (k, v) in self._bytes_rx.iteritems())
|
||||
|
||||
@property
|
||||
def unique_contacts(self):
|
||||
return self._unique_contacts
|
||||
|
||||
@property
|
||||
def queries_rx_per_second(self):
|
||||
return self._queries_rx_per_second
|
||||
|
||||
@property
|
||||
def queries_tx_per_second(self):
|
||||
return self._queries_tx_per_second
|
||||
|
||||
@property
|
||||
def kbps_tx(self):
|
||||
return self._kbps_tx
|
||||
|
||||
@property
|
||||
def kbps_rx(self):
|
||||
return self._kbps_rx
|
||||
|
||||
@property
|
||||
def recent_contact_count(self):
|
||||
return self._recent_contact_count
|
||||
|
||||
@property
|
||||
def total_bytes_tx(self):
|
||||
return self._total_bytes_tx
|
||||
|
||||
@property
|
||||
def total_bytes_rx(self):
|
||||
return self._total_bytes_rx
|
||||
|
||||
@property
|
||||
def bandwidth_stats(self):
|
||||
response = {
|
||||
"kbps_received": self.kbps_rx,
|
||||
"kbps_sent": self.kbps_tx,
|
||||
"total_bytes_sent": self.total_bytes_tx,
|
||||
"total_bytes_received": self.total_bytes_rx,
|
||||
"queries_received": self.queries_rx_per_second,
|
||||
"queries_sent": self.queries_tx_per_second,
|
||||
"recent_contacts": self.recent_contact_count,
|
||||
"unique_contacts": len(self.unique_contacts)
|
||||
}
|
||||
return response
|
||||
|
||||
def sendRPC(self, contact, method, args, rawResponse=False):
|
||||
""" Sends an RPC to the specified contact
|
||||
|
@ -168,16 +69,15 @@ class KademliaProtocol(protocol.DatagramProtocol):
|
|||
df._rpcRawResponse = True
|
||||
|
||||
# Set the RPC timeout timer
|
||||
timeoutCall = reactor.callLater(constants.rpcTimeout, self._msgTimeout, msg.id)
|
||||
timeoutCall, cancelTimeout = self._node.reactor_callLater(constants.rpcTimeout, self._msgTimeout, msg.id)
|
||||
# Transmit the data
|
||||
self._send(encodedMsg, msg.id, (contact.address, contact.port))
|
||||
self._sentMessages[msg.id] = (contact.id, df, timeoutCall, method, args)
|
||||
df.addErrback(cancelTimeout)
|
||||
return df
|
||||
|
||||
def startProtocol(self):
|
||||
log.info("DHT listening on UDP %i", self._node.port)
|
||||
if not self._bandwidth_stats_update_lc.running:
|
||||
self._bandwidth_stats_update_lc.start(1)
|
||||
|
||||
def datagramReceived(self, datagram, address):
|
||||
""" Handles and parses incoming RPC messages (and responses)
|
||||
|
@ -206,8 +106,9 @@ class KademliaProtocol(protocol.DatagramProtocol):
|
|||
try:
|
||||
msgPrimitive = self._encoder.decode(datagram)
|
||||
message = self._translator.fromPrimitive(msgPrimitive)
|
||||
except (encoding.DecodeError, ValueError):
|
||||
except (encoding.DecodeError, ValueError) as err:
|
||||
# We received some rubbish here
|
||||
log.exception("Error decoding datagram from %s:%i - %s", address[0], address[1], err)
|
||||
return
|
||||
except (IndexError, KeyError):
|
||||
log.warning("Couldn't decode dht datagram from %s", address)
|
||||
|
@ -215,18 +116,6 @@ class KademliaProtocol(protocol.DatagramProtocol):
|
|||
|
||||
remoteContact = Contact(message.nodeID, address[0], address[1], self)
|
||||
|
||||
now = time.time()
|
||||
contact_history = self._history_rx.get(address, [])
|
||||
if len(contact_history) > 1000:
|
||||
contact_history = [x for x in contact_history if now - x[1] < 1.0]
|
||||
contact_history.append((len(datagram), time.time()))
|
||||
self._history_rx[address] = contact_history
|
||||
bytes_rx = self._bytes_rx.get(address, 0)
|
||||
bytes_rx += len(datagram)
|
||||
self._bytes_rx[address] = bytes_rx
|
||||
if address not in self.unique_contacts:
|
||||
self._unique_contacts.append(address)
|
||||
|
||||
# Refresh the remote node's details in the local node's k-buckets
|
||||
self._node.addContact(remoteContact)
|
||||
if isinstance(message, msgtypes.RequestMessage):
|
||||
|
@ -253,8 +142,18 @@ class KademliaProtocol(protocol.DatagramProtocol):
|
|||
else:
|
||||
exception_type = UnknownRemoteException
|
||||
remoteException = exception_type(message.response)
|
||||
log.error("Remote exception (%s): %s", address, remoteException)
|
||||
df.errback(remoteException)
|
||||
# this error is returned by nodes that can be contacted but have an old
|
||||
# and broken version of the ping command, if they return it the node can
|
||||
# be contacted, so we'll treat it as a successful ping
|
||||
old_ping_error = "ping() got an unexpected keyword argument '_rpcNodeContact'"
|
||||
if isinstance(remoteException, TypeError) and \
|
||||
remoteException.message == old_ping_error:
|
||||
log.debug("old pong error")
|
||||
df.callback('pong')
|
||||
else:
|
||||
log.error("DHT RECV REMOTE EXCEPTION FROM %s:%i: %s", address[0],
|
||||
address[1], remoteException)
|
||||
df.errback(remoteException)
|
||||
else:
|
||||
# We got a result from the RPC
|
||||
df.callback(message.response)
|
||||
|
@ -281,18 +180,6 @@ class KademliaProtocol(protocol.DatagramProtocol):
|
|||
class (see C{kademlia.msgformat} and C{kademlia.encoding}).
|
||||
"""
|
||||
|
||||
now = time.time()
|
||||
contact_history = self._history_tx.get(address, [])
|
||||
if len(contact_history) > 1000:
|
||||
contact_history = [x for x in contact_history if now - x[1] < 1.0]
|
||||
contact_history.append((len(data), time.time()))
|
||||
self._history_tx[address] = contact_history
|
||||
bytes_tx = self._bytes_tx.get(address, 0)
|
||||
bytes_tx += len(data)
|
||||
self._bytes_tx[address] = bytes_tx
|
||||
if address not in self.unique_contacts:
|
||||
self._unique_contacts.append(address)
|
||||
|
||||
if len(data) > self.msgSizeLimit:
|
||||
# We have to spread the data over multiple UDP datagrams,
|
||||
# and provide sequencing information
|
||||
|
@ -319,13 +206,9 @@ class KademliaProtocol(protocol.DatagramProtocol):
|
|||
|
||||
def _scheduleSendNext(self, txData, address):
|
||||
"""Schedule the sending of the next UDP packet """
|
||||
delay = self._delay()
|
||||
key = object()
|
||||
delayed_call = reactor.callLater(delay, self._write_and_remove, key, txData, address)
|
||||
self._call_later_list[key] = delayed_call
|
||||
delayed_call, _ = self._node.reactor_callLater(0, self._write, txData, address)
|
||||
|
||||
def _write_and_remove(self, key, txData, address):
|
||||
del self._call_later_list[key]
|
||||
def _write(self, txData, address):
|
||||
if self.transport:
|
||||
try:
|
||||
self.transport.write(txData, address)
|
||||
|
@ -333,13 +216,15 @@ class KademliaProtocol(protocol.DatagramProtocol):
|
|||
if err.errno == errno.EWOULDBLOCK:
|
||||
# i'm scared this may swallow important errors, but i get a million of these
|
||||
# on Linux and it doesnt seem to affect anything -grin
|
||||
log.debug("Can't send data to dht: EWOULDBLOCK")
|
||||
log.warning("Can't send data to dht: EWOULDBLOCK")
|
||||
elif err.errno == errno.ENETUNREACH:
|
||||
# this should probably try to retransmit when the network connection is back
|
||||
log.error("Network is unreachable")
|
||||
else:
|
||||
log.error("DHT socket error: %s (%i)", err.message, err.errno)
|
||||
raise err
|
||||
else:
|
||||
log.warning("transport not connected!")
|
||||
|
||||
def _sendResponse(self, contact, rpcID, response):
|
||||
""" Send a RPC response to the specified contact
|
||||
|
@ -418,7 +303,7 @@ class KademliaProtocol(protocol.DatagramProtocol):
|
|||
# See if any progress has been made; if not, kill the message
|
||||
if self._hasProgressBeenMade(messageID):
|
||||
# Reset the RPC timeout timer
|
||||
timeoutCall = reactor.callLater(constants.rpcTimeout, self._msgTimeout, messageID)
|
||||
timeoutCall, _ = self._node.reactor_callLater(constants.rpcTimeout, self._msgTimeout, messageID)
|
||||
self._sentMessages[messageID] = (remoteContactID, df, timeoutCall, method, args)
|
||||
else:
|
||||
# No progress has been made
|
||||
|
@ -443,19 +328,5 @@ class KademliaProtocol(protocol.DatagramProtocol):
|
|||
Will only be called once, after all ports are disconnected.
|
||||
"""
|
||||
log.info('Stopping DHT')
|
||||
|
||||
if self._bandwidth_stats_update_lc.running:
|
||||
self._bandwidth_stats_update_lc.stop()
|
||||
|
||||
for delayed_call in self._call_later_list.values():
|
||||
try:
|
||||
delayed_call.cancel()
|
||||
except (error.AlreadyCalled, error.AlreadyCancelled):
|
||||
log.debug('Attempted to cancel a DelayedCall that was not active')
|
||||
except Exception:
|
||||
log.exception('Failed to cancel a DelayedCall')
|
||||
# not sure why this is needed, but taking this out sometimes causes
|
||||
# exceptions.AttributeError: 'Port' object has no attribute 'socket'
|
||||
# to happen on shutdown
|
||||
# reactor.iterate()
|
||||
CallLaterManager.stop()
|
||||
log.info('DHT stopped')
|
||||
|
|
|
@ -5,7 +5,6 @@
|
|||
# The docstrings in this module contain epytext markup; API documentation
|
||||
# may be created by processing this file with epydoc: http://epydoc.sf.net
|
||||
|
||||
import time
|
||||
import random
|
||||
from zope.interface import implements
|
||||
import constants
|
||||
|
@ -20,7 +19,7 @@ log = logging.getLogger(__name__)
|
|||
class TreeRoutingTable(object):
|
||||
""" This class implements a routing table used by a Node class.
|
||||
|
||||
The Kademlia routing table is a binary tree whose leaves are k-buckets,
|
||||
The Kademlia routing table is a binary tree whFose leaves are k-buckets,
|
||||
where each k-bucket contains nodes with some common prefix of their IDs.
|
||||
This prefix is the k-bucket's position in the binary tree; it therefore
|
||||
covers some range of ID values, and together all of the k-buckets cover
|
||||
|
@ -34,7 +33,7 @@ class TreeRoutingTable(object):
|
|||
"""
|
||||
implements(IRoutingTable)
|
||||
|
||||
def __init__(self, parentNodeID):
|
||||
def __init__(self, parentNodeID, getTime=None):
|
||||
"""
|
||||
@param parentNodeID: The n-bit node ID of the node to which this
|
||||
routing table belongs
|
||||
|
@ -43,6 +42,9 @@ class TreeRoutingTable(object):
|
|||
# Create the initial (single) k-bucket covering the range of the entire n-bit ID space
|
||||
self._buckets = [kbucket.KBucket(rangeMin=0, rangeMax=2 ** constants.key_bits)]
|
||||
self._parentNodeID = parentNodeID
|
||||
if not getTime:
|
||||
from time import time as getTime
|
||||
self._getTime = getTime
|
||||
|
||||
def addContact(self, contact):
|
||||
""" Add the given contact to the correct k-bucket; if it already
|
||||
|
@ -194,7 +196,7 @@ class TreeRoutingTable(object):
|
|||
bucketIndex = startIndex
|
||||
refreshIDs = []
|
||||
for bucket in self._buckets[startIndex:]:
|
||||
if force or (int(time.time()) - bucket.lastAccessed >= constants.refreshTimeout):
|
||||
if force or (int(self._getTime()) - bucket.lastAccessed >= constants.refreshTimeout):
|
||||
searchID = self._randomIDInBucketRange(bucketIndex)
|
||||
refreshIDs.append(searchID)
|
||||
bucketIndex += 1
|
||||
|
@ -221,7 +223,7 @@ class TreeRoutingTable(object):
|
|||
@type key: str
|
||||
"""
|
||||
bucketIndex = self._kbucketIndex(key)
|
||||
self._buckets[bucketIndex].lastAccessed = int(time.time())
|
||||
self._buckets[bucketIndex].lastAccessed = int(self._getTime())
|
||||
|
||||
def _kbucketIndex(self, key):
|
||||
""" Calculate the index of the k-bucket which is responsible for the
|
||||
|
@ -289,8 +291,8 @@ class OptimizedTreeRoutingTable(TreeRoutingTable):
|
|||
of the 13-page version of the Kademlia paper.
|
||||
"""
|
||||
|
||||
def __init__(self, parentNodeID):
|
||||
TreeRoutingTable.__init__(self, parentNodeID)
|
||||
def __init__(self, parentNodeID, getTime=None):
|
||||
TreeRoutingTable.__init__(self, parentNodeID, getTime)
|
||||
# Cache containing nodes eligible to replace stale k-bucket entries
|
||||
self._replacementCache = {}
|
||||
|
||||
|
@ -301,6 +303,7 @@ class OptimizedTreeRoutingTable(TreeRoutingTable):
|
|||
@param contact: The contact to add to this node's k-buckets
|
||||
@type contact: kademlia.contact.Contact
|
||||
"""
|
||||
|
||||
if contact.id == self._parentNodeID:
|
||||
return
|
||||
|
||||
|
|
274
lbrynet/tests/functional/test_dht.py
Normal file
274
lbrynet/tests/functional/test_dht.py
Normal file
|
@ -0,0 +1,274 @@
|
|||
import time
|
||||
import logging
|
||||
from twisted.trial import unittest
|
||||
from twisted.internet import defer, threads, task
|
||||
from lbrynet.dht.node import Node
|
||||
from lbrynet.tests import mocks
|
||||
from lbrynet.core.utils import generate_id
|
||||
|
||||
log = logging.getLogger("lbrynet.tests.util")
|
||||
# log.addHandler(logging.StreamHandler())
|
||||
# log.setLevel(logging.DEBUG)
|
||||
|
||||
|
||||
class TestKademliaBase(unittest.TestCase):
|
||||
timeout = 300.0 # timeout for each test
|
||||
network_size = 0 # plus lbrynet1, lbrynet2, and lbrynet3 seed nodes
|
||||
node_ids = None
|
||||
seed_dns = mocks.MOCK_DHT_SEED_DNS
|
||||
|
||||
def _add_next_node(self):
|
||||
node_id, node_ip = self.mock_node_generator.next()
|
||||
node = Node(node_id=node_id.decode('hex'), udpPort=4444, peerPort=3333, externalIP=node_ip,
|
||||
resolve=mocks.resolve, listenUDP=mocks.listenUDP, callLater=self.clock.callLater, clock=self.clock)
|
||||
self.nodes.append(node)
|
||||
return node
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def add_node(self):
|
||||
node = self._add_next_node()
|
||||
yield node.joinNetwork(
|
||||
[
|
||||
("lbrynet1.lbry.io", self._seeds[0].port),
|
||||
("lbrynet2.lbry.io", self._seeds[1].port),
|
||||
("lbrynet3.lbry.io", self._seeds[2].port),
|
||||
]
|
||||
)
|
||||
defer.returnValue(node)
|
||||
|
||||
def get_node(self, node_id):
|
||||
for node in self.nodes:
|
||||
if node.node_id == node_id:
|
||||
return node
|
||||
raise KeyError(node_id)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def pop_node(self):
|
||||
node = self.nodes.pop()
|
||||
yield node.stop()
|
||||
|
||||
def pump_clock(self, n, step=0.01):
|
||||
"""
|
||||
:param n: seconds to run the reactor for
|
||||
:param step: reactor tick rate (in seconds)
|
||||
"""
|
||||
for _ in range(n * 100):
|
||||
self.clock.advance(step)
|
||||
|
||||
def run_reactor(self, seconds, *deferreds):
|
||||
dl = [threads.deferToThread(self.pump_clock, seconds)]
|
||||
for d in deferreds:
|
||||
dl.append(d)
|
||||
return defer.DeferredList(dl)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def setUp(self):
|
||||
self.nodes = []
|
||||
self._seeds = []
|
||||
self.clock = task.Clock()
|
||||
self.mock_node_generator = mocks.mock_node_generator(mock_node_ids=self.node_ids)
|
||||
|
||||
join_dl = []
|
||||
for seed_dns in self.seed_dns:
|
||||
other_seeds = list(self.seed_dns.keys())
|
||||
other_seeds.remove(seed_dns)
|
||||
|
||||
self._add_next_node()
|
||||
seed = self.nodes.pop()
|
||||
self._seeds.append(seed)
|
||||
join_dl.append(
|
||||
seed.joinNetwork([(other_seed_dns, 4444) for other_seed_dns in other_seeds])
|
||||
)
|
||||
|
||||
if self.network_size:
|
||||
for _ in range(self.network_size):
|
||||
join_dl.append(self.add_node())
|
||||
yield self.run_reactor(1, *tuple(join_dl))
|
||||
self.verify_all_nodes_are_routable()
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def tearDown(self):
|
||||
dl = []
|
||||
while self.nodes:
|
||||
dl.append(self.pop_node()) # stop all of the nodes
|
||||
while self._seeds:
|
||||
dl.append(self._seeds.pop().stop()) # and the seeds
|
||||
yield defer.DeferredList(dl)
|
||||
|
||||
def verify_all_nodes_are_routable(self):
|
||||
routable = set()
|
||||
node_addresses = {node.externalIP for node in self.nodes}
|
||||
node_addresses = node_addresses.union({node.externalIP for node in self._seeds})
|
||||
for node in self._seeds:
|
||||
contact_addresses = {contact.address for contact in node.contacts}
|
||||
routable.update(contact_addresses)
|
||||
for node in self.nodes:
|
||||
contact_addresses = {contact.address for contact in node.contacts}
|
||||
routable.update(contact_addresses)
|
||||
self.assertSetEqual(routable, node_addresses)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def verify_all_nodes_are_pingable(self):
|
||||
ping_replies = {}
|
||||
ping_dl = []
|
||||
contacted = set()
|
||||
|
||||
def _ping_cb(result, node, replies):
|
||||
replies[node] = result
|
||||
|
||||
for node in self._seeds:
|
||||
contact_addresses = set()
|
||||
for contact in node.contacts:
|
||||
contact_addresses.add(contact.address)
|
||||
d = contact.ping()
|
||||
d.addCallback(_ping_cb, contact.address, ping_replies)
|
||||
contacted.add(contact.address)
|
||||
ping_dl.append(d)
|
||||
for node in self.nodes:
|
||||
contact_addresses = set()
|
||||
for contact in node.contacts:
|
||||
contact_addresses.add(contact.address)
|
||||
d = contact.ping()
|
||||
d.addCallback(_ping_cb, contact.address, ping_replies)
|
||||
contacted.add(contact.address)
|
||||
ping_dl.append(d)
|
||||
self.run_reactor(2, *ping_dl)
|
||||
yield threads.deferToThread(time.sleep, 0.1)
|
||||
node_addresses = {node.externalIP for node in self.nodes}.union({seed.externalIP for seed in self._seeds})
|
||||
self.assertSetEqual(node_addresses, contacted)
|
||||
self.assertDictEqual(ping_replies, {node: "pong" for node in contacted})
|
||||
|
||||
|
||||
class TestKademliaBootstrap(TestKademliaBase):
|
||||
"""
|
||||
Test initializing the network / connecting the seed nodes
|
||||
"""
|
||||
|
||||
def test_bootstrap_network(self): # simulates the real network, which has three seeds
|
||||
self.assertEqual(len(self._seeds[0].contacts), 2)
|
||||
self.assertEqual(len(self._seeds[1].contacts), 2)
|
||||
self.assertEqual(len(self._seeds[2].contacts), 2)
|
||||
|
||||
self.assertSetEqual(
|
||||
{self._seeds[0].contacts[0].address, self._seeds[0].contacts[1].address},
|
||||
{self._seeds[1].externalIP, self._seeds[2].externalIP}
|
||||
)
|
||||
|
||||
self.assertSetEqual(
|
||||
{self._seeds[1].contacts[0].address, self._seeds[1].contacts[1].address},
|
||||
{self._seeds[0].externalIP, self._seeds[2].externalIP}
|
||||
)
|
||||
|
||||
self.assertSetEqual(
|
||||
{self._seeds[2].contacts[0].address, self._seeds[2].contacts[1].address},
|
||||
{self._seeds[0].externalIP, self._seeds[1].externalIP}
|
||||
)
|
||||
|
||||
def test_all_nodes_are_pingable(self):
|
||||
return self.verify_all_nodes_are_pingable()
|
||||
|
||||
|
||||
class TestKademliaBootstrapSixteenSeeds(TestKademliaBase):
|
||||
node_ids = [
|
||||
'000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000',
|
||||
'111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111',
|
||||
'222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222',
|
||||
'333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333',
|
||||
'444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444',
|
||||
'555555555555555555555555555555555555555555555555555555555555555555555555555555555555555555555555',
|
||||
'666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666',
|
||||
'777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777',
|
||||
'888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888',
|
||||
'999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999',
|
||||
'aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa',
|
||||
'bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb',
|
||||
'cccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc',
|
||||
'dddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddd',
|
||||
'eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee',
|
||||
'ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff'
|
||||
]
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def setUp(self):
|
||||
self.seed_dns.update(
|
||||
{
|
||||
"lbrynet4.lbry.io": "10.42.42.4",
|
||||
"lbrynet5.lbry.io": "10.42.42.5",
|
||||
"lbrynet6.lbry.io": "10.42.42.6",
|
||||
"lbrynet7.lbry.io": "10.42.42.7",
|
||||
"lbrynet8.lbry.io": "10.42.42.8",
|
||||
"lbrynet9.lbry.io": "10.42.42.9",
|
||||
"lbrynet10.lbry.io": "10.42.42.10",
|
||||
"lbrynet11.lbry.io": "10.42.42.11",
|
||||
"lbrynet12.lbry.io": "10.42.42.12",
|
||||
"lbrynet13.lbry.io": "10.42.42.13",
|
||||
"lbrynet14.lbry.io": "10.42.42.14",
|
||||
"lbrynet15.lbry.io": "10.42.42.15",
|
||||
"lbrynet16.lbry.io": "10.42.42.16",
|
||||
}
|
||||
)
|
||||
yield TestKademliaBase.setUp(self)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def tearDown(self):
|
||||
yield TestKademliaBase.tearDown(self)
|
||||
|
||||
def test_bootstrap_network(self):
|
||||
pass
|
||||
|
||||
def _test_all_nodes_are_pingable(self):
|
||||
return self.verify_all_nodes_are_pingable()
|
||||
|
||||
|
||||
class Test250NodeNetwork(TestKademliaBase):
|
||||
network_size = 250
|
||||
|
||||
def test_setup_network_and_verify_connectivity(self):
|
||||
pass
|
||||
|
||||
def update_network(self):
|
||||
import random
|
||||
dl = []
|
||||
announced_blobs = []
|
||||
|
||||
for node in self.nodes: # random events
|
||||
if random.randint(0, 10000) < 75 and announced_blobs: # get peers for a blob
|
||||
log.info('find blob')
|
||||
blob_hash = random.choice(announced_blobs)
|
||||
dl.append(node.getPeersForBlob(blob_hash))
|
||||
if random.randint(0, 10000) < 25: # announce a blob
|
||||
log.info('announce blob')
|
||||
blob_hash = generate_id()
|
||||
announced_blobs.append((blob_hash, node.node_id))
|
||||
dl.append(node.announceHaveBlob(blob_hash))
|
||||
|
||||
random.shuffle(self.nodes)
|
||||
|
||||
# kill nodes
|
||||
while random.randint(0, 100) > 95:
|
||||
dl.append(self.pop_node())
|
||||
log.info('pop node')
|
||||
|
||||
# add nodes
|
||||
while random.randint(0, 100) > 95:
|
||||
dl.append(self.add_node())
|
||||
log.info('add node')
|
||||
return tuple(dl), announced_blobs
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def _test_simulate_network(self):
|
||||
total_blobs = []
|
||||
for i in range(100):
|
||||
d, blobs = self.update_network()
|
||||
total_blobs.extend(blobs)
|
||||
self.run_reactor(1, *d)
|
||||
yield threads.deferToThread(time.sleep, 0.1)
|
||||
routable = set()
|
||||
node_addresses = {node.externalIP for node in self.nodes}
|
||||
for node in self.nodes:
|
||||
contact_addresses = {contact.address for contact in node.contacts}
|
||||
routable.update(contact_addresses)
|
||||
log.warning("difference: %i", len(node_addresses.difference(routable)))
|
||||
log.info("blobs %i", len(total_blobs))
|
||||
log.info("step %i, %i nodes", i, len(self.nodes))
|
||||
self.pump_clock(100)
|
|
@ -4,8 +4,7 @@ from twisted.trial import unittest
|
|||
from lbrynet import conf
|
||||
from lbrynet.core.StreamDescriptor import get_sd_info
|
||||
from lbrynet import reflector
|
||||
from lbrynet.core import BlobManager
|
||||
from lbrynet.core import PeerManager
|
||||
from lbrynet.core import BlobManager, PeerManager
|
||||
from lbrynet.core import Session
|
||||
from lbrynet.core import StreamDescriptor
|
||||
from lbrynet.lbry_file.client import EncryptedFileOptions
|
||||
|
@ -15,6 +14,8 @@ from lbrynet.file_manager import EncryptedFileManager
|
|||
from lbrynet.tests import mocks
|
||||
from lbrynet.tests.util import mk_db_and_blob_dir, rm_db_and_blob_dir
|
||||
|
||||
from lbrynet.tests.mocks import Node
|
||||
|
||||
|
||||
class TestReflector(unittest.TestCase):
|
||||
def setUp(self):
|
||||
|
@ -28,7 +29,6 @@ class TestReflector(unittest.TestCase):
|
|||
wallet = mocks.Wallet()
|
||||
peer_manager = PeerManager.PeerManager()
|
||||
peer_finder = mocks.PeerFinder(5553, peer_manager, 2)
|
||||
hash_announcer = mocks.Announcer()
|
||||
sd_identifier = StreamDescriptor.StreamDescriptorIdentifier()
|
||||
|
||||
self.expected_blobs = [
|
||||
|
@ -55,13 +55,14 @@ class TestReflector(unittest.TestCase):
|
|||
db_dir=self.db_dir,
|
||||
node_id="abcd",
|
||||
peer_finder=peer_finder,
|
||||
hash_announcer=hash_announcer,
|
||||
blob_dir=self.blob_dir,
|
||||
peer_port=5553,
|
||||
use_upnp=False,
|
||||
wallet=wallet,
|
||||
blob_tracker_class=mocks.BlobAvailabilityTracker,
|
||||
external_ip="127.0.0.1"
|
||||
external_ip="127.0.0.1",
|
||||
dht_node_class=Node,
|
||||
hash_announcer=mocks.Announcer()
|
||||
)
|
||||
|
||||
self.lbry_file_manager = EncryptedFileManager.EncryptedFileManager(self.session,
|
||||
|
@ -74,17 +75,17 @@ class TestReflector(unittest.TestCase):
|
|||
db_dir=self.server_db_dir,
|
||||
node_id="abcd",
|
||||
peer_finder=peer_finder,
|
||||
hash_announcer=hash_announcer,
|
||||
blob_dir=self.server_blob_dir,
|
||||
peer_port=5553,
|
||||
use_upnp=False,
|
||||
wallet=wallet,
|
||||
blob_tracker_class=mocks.BlobAvailabilityTracker,
|
||||
external_ip="127.0.0.1"
|
||||
external_ip="127.0.0.1",
|
||||
dht_node_class=Node,
|
||||
hash_announcer=mocks.Announcer()
|
||||
)
|
||||
|
||||
self.server_blob_manager = BlobManager.DiskBlobManager(hash_announcer,
|
||||
self.server_blob_dir,
|
||||
self.server_blob_manager = BlobManager.DiskBlobManager(self.server_blob_dir,
|
||||
self.server_session.storage)
|
||||
|
||||
self.server_lbry_file_manager = EncryptedFileManager.EncryptedFileManager(
|
||||
|
@ -364,6 +365,7 @@ class TestReflector(unittest.TestCase):
|
|||
d.addCallback(lambda _: verify_stream_on_reflector())
|
||||
return d
|
||||
|
||||
|
||||
def iv_generator():
|
||||
iv = 0
|
||||
while True:
|
||||
|
|
|
@ -67,7 +67,7 @@ class TestStreamify(TestCase):
|
|||
blob_dir=self.blob_dir, peer_port=5553,
|
||||
use_upnp=False, rate_limiter=rate_limiter, wallet=wallet,
|
||||
blob_tracker_class=DummyBlobAvailabilityTracker,
|
||||
is_generous=self.is_generous, external_ip="127.0.0.1"
|
||||
is_generous=self.is_generous, external_ip="127.0.0.1", dht_node_class=mocks.Node
|
||||
)
|
||||
|
||||
self.lbry_file_manager = EncryptedFileManager(self.session, sd_identifier)
|
||||
|
@ -112,7 +112,7 @@ class TestStreamify(TestCase):
|
|||
self.session = Session(
|
||||
conf.ADJUSTABLE_SETTINGS['data_rate'][1], db_dir=self.db_dir, node_id="abcd",
|
||||
peer_finder=peer_finder, hash_announcer=hash_announcer,
|
||||
blob_dir=self.blob_dir, peer_port=5553,
|
||||
blob_dir=self.blob_dir, peer_port=5553, dht_node_class=mocks.Node,
|
||||
use_upnp=False, rate_limiter=rate_limiter, wallet=wallet,
|
||||
blob_tracker_class=DummyBlobAvailabilityTracker, external_ip="127.0.0.1"
|
||||
)
|
||||
|
|
|
@ -1,12 +1,17 @@
|
|||
import struct
|
||||
import io
|
||||
|
||||
from Crypto.PublicKey import RSA
|
||||
from twisted.internet import defer
|
||||
from twisted.internet import defer, error
|
||||
from twisted.python.failure import Failure
|
||||
|
||||
from lbrynet.core import PTCWallet
|
||||
from lbrynet.core.client.ClientRequest import ClientRequest
|
||||
from lbrynet.core.Error import RequestCanceledError
|
||||
from lbrynet.core import BlobAvailability
|
||||
from lbrynet.core.utils import generate_id
|
||||
from lbrynet.daemon import ExchangeRateManager as ERM
|
||||
from lbrynet import conf
|
||||
from util import debug_kademlia_packet
|
||||
|
||||
KB = 2**10
|
||||
|
||||
|
@ -18,15 +23,18 @@ class FakeLBRYFile(object):
|
|||
self.stream_hash = stream_hash
|
||||
self.file_name = 'fake_lbry_file'
|
||||
|
||||
|
||||
class Node(object):
|
||||
def __init__(self, *args, **kwargs):
|
||||
pass
|
||||
def __init__(self, peer_finder=None, peer_manager=None, **kwargs):
|
||||
self.peer_finder = peer_finder
|
||||
self.peer_manager = peer_manager
|
||||
self.peerPort = 3333
|
||||
|
||||
def joinNetwork(self, *args):
|
||||
pass
|
||||
return defer.succeed(True)
|
||||
|
||||
def stop(self):
|
||||
pass
|
||||
return defer.succeed(None)
|
||||
|
||||
|
||||
class FakeNetwork(object):
|
||||
|
@ -69,6 +77,82 @@ class ExchangeRateManager(ERM.ExchangeRateManager):
|
|||
feed.market, rates[feed.market]['spot'], rates[feed.market]['ts'])
|
||||
|
||||
|
||||
class PointTraderKeyExchanger(object):
|
||||
|
||||
def __init__(self, wallet):
|
||||
self.wallet = wallet
|
||||
self._protocols = []
|
||||
|
||||
def send_next_request(self, peer, protocol):
|
||||
if not protocol in self._protocols:
|
||||
r = ClientRequest({'public_key': self.wallet.encoded_public_key},
|
||||
'public_key')
|
||||
d = protocol.add_request(r)
|
||||
d.addCallback(self._handle_exchange_response, peer, r, protocol)
|
||||
d.addErrback(self._request_failed, peer)
|
||||
self._protocols.append(protocol)
|
||||
return defer.succeed(True)
|
||||
else:
|
||||
return defer.succeed(False)
|
||||
|
||||
def _handle_exchange_response(self, response_dict, peer, request, protocol):
|
||||
assert request.response_identifier in response_dict, \
|
||||
"Expected %s in dict but did not get it" % request.response_identifier
|
||||
assert protocol in self._protocols, "Responding protocol is not in our list of protocols"
|
||||
peer_pub_key = response_dict[request.response_identifier]
|
||||
self.wallet.set_public_key_for_peer(peer, peer_pub_key)
|
||||
return True
|
||||
|
||||
def _request_failed(self, err, peer):
|
||||
if not err.check(RequestCanceledError):
|
||||
return err
|
||||
|
||||
|
||||
class PointTraderKeyQueryHandlerFactory(object):
|
||||
|
||||
def __init__(self, wallet):
|
||||
self.wallet = wallet
|
||||
|
||||
def build_query_handler(self):
|
||||
q_h = PointTraderKeyQueryHandler(self.wallet)
|
||||
return q_h
|
||||
|
||||
def get_primary_query_identifier(self):
|
||||
return 'public_key'
|
||||
|
||||
def get_description(self):
|
||||
return ("Point Trader Address - an address for receiving payments on the "
|
||||
"point trader testing network")
|
||||
|
||||
|
||||
class PointTraderKeyQueryHandler(object):
|
||||
|
||||
def __init__(self, wallet):
|
||||
self.wallet = wallet
|
||||
self.query_identifiers = ['public_key']
|
||||
self.public_key = None
|
||||
self.peer = None
|
||||
|
||||
def register_with_request_handler(self, request_handler, peer):
|
||||
self.peer = peer
|
||||
request_handler.register_query_handler(self, self.query_identifiers)
|
||||
|
||||
def handle_queries(self, queries):
|
||||
if self.query_identifiers[0] in queries:
|
||||
new_encoded_pub_key = queries[self.query_identifiers[0]]
|
||||
try:
|
||||
RSA.importKey(new_encoded_pub_key)
|
||||
except (ValueError, TypeError, IndexError):
|
||||
return defer.fail(Failure(ValueError("Client sent an invalid public key")))
|
||||
self.public_key = new_encoded_pub_key
|
||||
self.wallet.set_public_key_for_peer(self.peer, self.public_key)
|
||||
fields = {'public_key': self.wallet.encoded_public_key}
|
||||
return defer.succeed(fields)
|
||||
if self.public_key is None:
|
||||
return defer.fail(Failure(ValueError("Expected but did not receive a public key")))
|
||||
else:
|
||||
return defer.succeed({})
|
||||
|
||||
|
||||
class Wallet(object):
|
||||
def __init__(self):
|
||||
|
@ -94,10 +178,10 @@ class Wallet(object):
|
|||
return defer.succeed(True)
|
||||
|
||||
def get_info_exchanger(self):
|
||||
return PTCWallet.PointTraderKeyExchanger(self)
|
||||
return PointTraderKeyExchanger(self)
|
||||
|
||||
def get_wallet_info_query_handler_factory(self):
|
||||
return PTCWallet.PointTraderKeyQueryHandlerFactory(self)
|
||||
return PointTraderKeyQueryHandlerFactory(self)
|
||||
|
||||
def reserve_points(self, *args):
|
||||
return True
|
||||
|
@ -164,6 +248,9 @@ class Announcer(object):
|
|||
def stop(self):
|
||||
pass
|
||||
|
||||
def get_next_announce_time(self):
|
||||
return 0
|
||||
|
||||
|
||||
class GenFile(io.RawIOBase):
|
||||
def __init__(self, size, pattern):
|
||||
|
@ -313,4 +400,87 @@ def mock_conf_settings(obj, settings={}):
|
|||
obj.addCleanup(_reset_settings)
|
||||
|
||||
|
||||
MOCK_DHT_NODES = [
|
||||
"000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000",
|
||||
"FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF",
|
||||
"DEADBEEFDEADBEEFDEADBEEFDEADBEEFDEADBEEFDEADBEEFDEADBEEFDEADBEEFDEADBEEFDEADBEEFDEADBEEFDEADBEEF",
|
||||
]
|
||||
|
||||
MOCK_DHT_SEED_DNS = { # these map to mock nodes 0, 1, and 2
|
||||
"lbrynet1.lbry.io": "10.42.42.1",
|
||||
"lbrynet2.lbry.io": "10.42.42.2",
|
||||
"lbrynet3.lbry.io": "10.42.42.3",
|
||||
}
|
||||
|
||||
|
||||
def resolve(name, timeout=(1, 3, 11, 45)):
|
||||
if name not in MOCK_DHT_SEED_DNS:
|
||||
return defer.fail(error.DNSLookupError(name))
|
||||
return defer.succeed(MOCK_DHT_SEED_DNS[name])
|
||||
|
||||
|
||||
class MockUDPTransport(object):
|
||||
def __init__(self, address, port, max_packet_size, protocol):
|
||||
self.address = address
|
||||
self.port = port
|
||||
self.max_packet_size = max_packet_size
|
||||
self._node = protocol._node
|
||||
|
||||
def write(self, data, address):
|
||||
dest = MockNetwork.peers[address][0]
|
||||
debug_kademlia_packet(data, (self.address, self.port), address, self._node)
|
||||
dest.datagramReceived(data, (self.address, self.port))
|
||||
|
||||
|
||||
class MockUDPPort(object):
|
||||
def __init__(self, protocol):
|
||||
self.protocol = protocol
|
||||
|
||||
def startListening(self, reason=None):
|
||||
return self.protocol.startProtocol()
|
||||
|
||||
def stopListening(self, reason=None):
|
||||
return self.protocol.stopProtocol()
|
||||
|
||||
|
||||
class MockNetwork(object):
|
||||
peers = {} # (interface, port): (protocol, max_packet_size)
|
||||
|
||||
@classmethod
|
||||
def add_peer(cls, port, protocol, interface, maxPacketSize):
|
||||
interface = protocol._node.externalIP
|
||||
protocol.transport = MockUDPTransport(interface, port, maxPacketSize, protocol)
|
||||
cls.peers[(interface, port)] = (protocol, maxPacketSize)
|
||||
|
||||
|
||||
def listenUDP(port, protocol, interface='', maxPacketSize=8192):
|
||||
MockNetwork.add_peer(port, protocol, interface, maxPacketSize)
|
||||
return MockUDPPort(protocol)
|
||||
|
||||
|
||||
def address_generator(address=(10, 42, 42, 1)):
|
||||
def increment(addr):
|
||||
value = struct.unpack("I", "".join([chr(x) for x in list(addr)[::-1]]))[0] + 1
|
||||
new_addr = []
|
||||
for i in range(4):
|
||||
new_addr.append(value % 256)
|
||||
value >>= 8
|
||||
return tuple(new_addr[::-1])
|
||||
|
||||
while True:
|
||||
yield "{}.{}.{}.{}".format(*address)
|
||||
address = increment(address)
|
||||
|
||||
|
||||
def mock_node_generator(count=None, mock_node_ids=MOCK_DHT_NODES):
|
||||
if mock_node_ids is None:
|
||||
mock_node_ids = MOCK_DHT_NODES
|
||||
|
||||
for num, node_ip in enumerate(address_generator()):
|
||||
if count and num >= count:
|
||||
break
|
||||
if num >= len(mock_node_ids):
|
||||
node_id = generate_id().encode('hex')
|
||||
else:
|
||||
node_id = mock_node_ids[num]
|
||||
yield (node_id, node_ip)
|
||||
|
|
|
@ -1,55 +1,82 @@
|
|||
import tempfile
|
||||
import shutil
|
||||
from twisted.trial import unittest
|
||||
from twisted.internet import defer, task
|
||||
from twisted.internet import defer, reactor, threads
|
||||
|
||||
from lbrynet.core import utils
|
||||
from lbrynet.tests.util import random_lbry_hash
|
||||
from lbrynet.dht.hashannouncer import DHTHashAnnouncer
|
||||
from lbrynet.core.call_later_manager import CallLaterManager
|
||||
from lbrynet.database.storage import SQLiteStorage
|
||||
|
||||
|
||||
class MocDHTNode(object):
|
||||
def __init__(self):
|
||||
def __init__(self, announce_will_fail=False):
|
||||
# if announce_will_fail is True,
|
||||
# announceHaveBlob will return empty dict
|
||||
self.call_later_manager = CallLaterManager
|
||||
self.call_later_manager.setup(reactor.callLater)
|
||||
self.blobs_announced = 0
|
||||
self.announce_will_fail = announce_will_fail
|
||||
|
||||
def announceHaveBlob(self, blob):
|
||||
self.blobs_announced += 1
|
||||
return defer.succeed(True)
|
||||
|
||||
class MocSupplier(object):
|
||||
def __init__(self, blobs_to_announce):
|
||||
self.blobs_to_announce = blobs_to_announce
|
||||
self.announced = False
|
||||
def hashes_to_announce(self):
|
||||
if not self.announced:
|
||||
self.announced = True
|
||||
return defer.succeed(self.blobs_to_announce)
|
||||
if self.announce_will_fail:
|
||||
return_val = {}
|
||||
else:
|
||||
return defer.succeed([])
|
||||
return_val = {blob: ["ab"*48]}
|
||||
|
||||
self.blobs_announced += 1
|
||||
d = defer.Deferred()
|
||||
self.call_later_manager.call_later(1, d.callback, return_val)
|
||||
return d
|
||||
|
||||
|
||||
class DHTHashAnnouncerTest(unittest.TestCase):
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def setUp(self):
|
||||
from lbrynet.conf import initialize_settings
|
||||
initialize_settings()
|
||||
self.num_blobs = 10
|
||||
self.blobs_to_announce = []
|
||||
for i in range(0, self.num_blobs):
|
||||
self.blobs_to_announce.append(random_lbry_hash())
|
||||
self.clock = task.Clock()
|
||||
self.dht_node = MocDHTNode()
|
||||
utils.call_later = self.clock.callLater
|
||||
from lbrynet.core.server.DHTHashAnnouncer import DHTHashAnnouncer
|
||||
self.announcer = DHTHashAnnouncer(self.dht_node, peer_port=3333)
|
||||
self.supplier = MocSupplier(self.blobs_to_announce)
|
||||
self.announcer.add_supplier(self.supplier)
|
||||
self.dht_node.peerPort = 3333
|
||||
self.dht_node.clock = reactor
|
||||
self.db_dir = tempfile.mkdtemp()
|
||||
self.storage = SQLiteStorage(self.db_dir)
|
||||
yield self.storage.setup()
|
||||
self.announcer = DHTHashAnnouncer(self.dht_node, self.storage, 10)
|
||||
for blob_hash in self.blobs_to_announce:
|
||||
yield self.storage.add_completed_blob(blob_hash, 100, 0, 1)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def tearDown(self):
|
||||
self.dht_node.call_later_manager.stop()
|
||||
yield self.storage.stop()
|
||||
yield threads.deferToThread(shutil.rmtree, self.db_dir)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def test_announce_fail(self):
|
||||
# test what happens when node.announceHaveBlob() returns empty dict
|
||||
self.dht_node.announce_will_fail = True
|
||||
d = yield self.announcer.manage()
|
||||
yield d
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def test_basic(self):
|
||||
self.announcer._announce_available_hashes()
|
||||
self.assertEqual(self.announcer.hash_queue_size(), self.announcer.CONCURRENT_ANNOUNCERS)
|
||||
self.clock.advance(1)
|
||||
d = self.announcer.immediate_announce(self.blobs_to_announce)
|
||||
self.assertEqual(len(self.announcer.hash_queue), self.num_blobs)
|
||||
yield d
|
||||
self.assertEqual(self.dht_node.blobs_announced, self.num_blobs)
|
||||
self.assertEqual(self.announcer.hash_queue_size(), 0)
|
||||
self.assertEqual(len(self.announcer.hash_queue), 0)
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def test_immediate_announce(self):
|
||||
# Test that immediate announce puts a hash at the front of the queue
|
||||
self.announcer._announce_available_hashes()
|
||||
d = self.announcer.immediate_announce(self.blobs_to_announce)
|
||||
self.assertEqual(len(self.announcer.hash_queue), self.num_blobs)
|
||||
blob_hash = random_lbry_hash()
|
||||
self.announcer.immediate_announce([blob_hash])
|
||||
self.assertEqual(self.announcer.hash_queue_size(), self.announcer.CONCURRENT_ANNOUNCERS+1)
|
||||
self.assertEqual(blob_hash, self.announcer.hash_queue[0][0])
|
||||
|
||||
self.assertEqual(len(self.announcer.hash_queue), self.num_blobs+1)
|
||||
self.assertEqual(blob_hash, self.announcer.hash_queue[-1])
|
||||
yield d
|
||||
|
|
|
@ -8,7 +8,6 @@ from twisted.internet import defer, threads
|
|||
|
||||
from lbrynet.tests.util import random_lbry_hash
|
||||
from lbrynet.core.BlobManager import DiskBlobManager
|
||||
from lbrynet.core.HashAnnouncer import DummyHashAnnouncer
|
||||
from lbrynet.database.storage import SQLiteStorage
|
||||
from lbrynet.core.Peer import Peer
|
||||
from lbrynet import conf
|
||||
|
@ -21,8 +20,7 @@ class BlobManagerTest(unittest.TestCase):
|
|||
conf.initialize_settings()
|
||||
self.blob_dir = tempfile.mkdtemp()
|
||||
self.db_dir = tempfile.mkdtemp()
|
||||
hash_announcer = DummyHashAnnouncer()
|
||||
self.bm = DiskBlobManager(hash_announcer, self.blob_dir, SQLiteStorage(self.db_dir))
|
||||
self.bm = DiskBlobManager(self.blob_dir, SQLiteStorage(self.db_dir))
|
||||
self.peer = Peer('somehost', 22)
|
||||
yield self.bm.storage.setup()
|
||||
|
||||
|
|
|
@ -195,7 +195,7 @@ class StreamStorageTests(StorageTest):
|
|||
|
||||
should_announce_count = yield self.storage.count_should_announce_blobs()
|
||||
self.assertEqual(should_announce_count, 2)
|
||||
should_announce_hashes = yield self.storage.get_blobs_to_announce(FakeAnnouncer())
|
||||
should_announce_hashes = yield self.storage.get_blobs_to_announce()
|
||||
self.assertSetEqual(set(should_announce_hashes), {sd_hash, blob1})
|
||||
|
||||
stream_hashes = yield self.storage.get_all_streams()
|
||||
|
|
|
@ -198,7 +198,7 @@ class NodeLookupTest(unittest.TestCase):
|
|||
h = hashlib.sha384()
|
||||
h.update('node1')
|
||||
node_id = str(h.digest())
|
||||
self.node = lbrynet.dht.node.Node(node_id, 4000, None, None, self._protocol)
|
||||
self.node = lbrynet.dht.node.Node(node_id=node_id, udpPort=4000, networkProtocol=self._protocol)
|
||||
self.updPort = 81173
|
||||
self.contactsAmount = 80
|
||||
# Reinitialise the routing table
|
||||
|
|
|
@ -1,32 +1,55 @@
|
|||
import time
|
||||
import unittest
|
||||
import twisted.internet.selectreactor
|
||||
|
||||
from twisted.internet.task import Clock
|
||||
from twisted.internet import defer
|
||||
import lbrynet.dht.protocol
|
||||
import lbrynet.dht.contact
|
||||
import lbrynet.dht.constants
|
||||
import lbrynet.dht.msgtypes
|
||||
from lbrynet.dht.error import TimeoutError
|
||||
from lbrynet.dht.node import Node, rpcmethod
|
||||
from lbrynet.tests.mocks import listenUDP, resolve
|
||||
from lbrynet.core.call_later_manager import CallLaterManager
|
||||
|
||||
import logging
|
||||
|
||||
log = logging.getLogger()
|
||||
|
||||
|
||||
class KademliaProtocolTest(unittest.TestCase):
|
||||
""" Test case for the Protocol class """
|
||||
|
||||
def setUp(self):
|
||||
del lbrynet.dht.protocol.reactor
|
||||
lbrynet.dht.protocol.reactor = twisted.internet.selectreactor.SelectReactor()
|
||||
self.node = Node(node_id='1' * 48, udpPort=9182, externalIP="127.0.0.1")
|
||||
self.protocol = lbrynet.dht.protocol.KademliaProtocol(self.node)
|
||||
udpPort = 9182
|
||||
|
||||
def setUp(self):
|
||||
self._reactor = Clock()
|
||||
CallLaterManager.setup(self._reactor.callLater)
|
||||
self.node = Node(node_id='1' * 48, udpPort=self.udpPort, externalIP="127.0.0.1", listenUDP=listenUDP,
|
||||
resolve=resolve, clock=self._reactor, callLater=self._reactor.callLater)
|
||||
|
||||
def tearDown(self):
|
||||
CallLaterManager.stop()
|
||||
del self._reactor
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def testReactor(self):
|
||||
""" Tests if the reactor can start/stop the protocol correctly """
|
||||
lbrynet.dht.protocol.reactor.listenUDP(0, self.protocol)
|
||||
lbrynet.dht.protocol.reactor.callLater(0, lbrynet.dht.protocol.reactor.stop)
|
||||
lbrynet.dht.protocol.reactor.run()
|
||||
|
||||
d = defer.Deferred()
|
||||
self._reactor.callLater(1, d.callback, True)
|
||||
self._reactor.advance(1)
|
||||
result = yield d
|
||||
self.assertTrue(result)
|
||||
|
||||
def testRPCTimeout(self):
|
||||
""" Tests if a RPC message sent to a dead remote node times out correctly """
|
||||
dead_node = Node(node_id='2' * 48, udpPort=self.udpPort, externalIP="127.0.0.2", listenUDP=listenUDP,
|
||||
resolve=resolve, clock=self._reactor, callLater=self._reactor.callLater)
|
||||
dead_node.start_listening()
|
||||
dead_node.stop()
|
||||
self._reactor.pump([1 for _ in range(10)])
|
||||
dead_contact = lbrynet.dht.contact.Contact('2' * 48, '127.0.0.2', 9182, self.node._protocol)
|
||||
self.node.addContact(dead_contact)
|
||||
|
||||
@rpcmethod
|
||||
def fake_ping(*args, **kwargs):
|
||||
|
@ -38,19 +61,18 @@ class KademliaProtocolTest(unittest.TestCase):
|
|||
real_attempts = lbrynet.dht.constants.rpcAttempts
|
||||
lbrynet.dht.constants.rpcAttempts = 1
|
||||
lbrynet.dht.constants.rpcTimeout = 1
|
||||
|
||||
self.node.ping = fake_ping
|
||||
deadContact = lbrynet.dht.contact.Contact('2' * 48, '127.0.0.1', 9182, self.protocol)
|
||||
self.node.addContact(deadContact)
|
||||
# Make sure the contact was added
|
||||
self.failIf(deadContact not in self.node.contacts,
|
||||
self.failIf(dead_contact not in self.node.contacts,
|
||||
'Contact not added to fake node (error in test code)')
|
||||
lbrynet.dht.protocol.reactor.listenUDP(9182, self.protocol)
|
||||
self.node.start_listening()
|
||||
|
||||
# Run the PING RPC (which should raise a timeout error)
|
||||
df = self.protocol.sendRPC(deadContact, 'ping', {})
|
||||
df = self.node._protocol.sendRPC(dead_contact, 'ping', {})
|
||||
|
||||
def check_timeout(err):
|
||||
self.assertEqual(type(err), TimeoutError)
|
||||
self.assertEqual(err.type, TimeoutError)
|
||||
|
||||
df.addErrback(check_timeout)
|
||||
|
||||
|
@ -61,20 +83,24 @@ class KademliaProtocolTest(unittest.TestCase):
|
|||
|
||||
# See if the contact was removed due to the timeout
|
||||
def check_removed_contact():
|
||||
self.failIf(deadContact in self.node.contacts,
|
||||
self.failIf(dead_contact in self.node.contacts,
|
||||
'Contact was not removed after RPC timeout; check exception types.')
|
||||
|
||||
df.addCallback(lambda _: reset_values())
|
||||
|
||||
# Stop the reactor if a result arrives (timeout or not)
|
||||
df.addBoth(lambda _: lbrynet.dht.protocol.reactor.stop())
|
||||
df.addCallback(lambda _: check_removed_contact())
|
||||
lbrynet.dht.protocol.reactor.run()
|
||||
self._reactor.pump([1 for _ in range(20)])
|
||||
|
||||
def testRPCRequest(self):
|
||||
""" Tests if a valid RPC request is executed and responded to correctly """
|
||||
remoteContact = lbrynet.dht.contact.Contact('2' * 48, '127.0.0.1', 9182, self.protocol)
|
||||
|
||||
remote_node = Node(node_id='2' * 48, udpPort=self.udpPort, externalIP="127.0.0.2", listenUDP=listenUDP,
|
||||
resolve=resolve, clock=self._reactor, callLater=self._reactor.callLater)
|
||||
remote_node.start_listening()
|
||||
remoteContact = lbrynet.dht.contact.Contact('2' * 48, '127.0.0.2', 9182, self.node._protocol)
|
||||
self.node.addContact(remoteContact)
|
||||
|
||||
self.error = None
|
||||
|
||||
def handleError(f):
|
||||
|
@ -87,16 +113,18 @@ class KademliaProtocolTest(unittest.TestCase):
|
|||
% (expectedResult, result)
|
||||
|
||||
# Publish the "local" node on the network
|
||||
lbrynet.dht.protocol.reactor.listenUDP(9182, self.protocol)
|
||||
self.node.start_listening()
|
||||
# Simulate the RPC
|
||||
df = remoteContact.ping()
|
||||
df.addCallback(handleResult)
|
||||
df.addErrback(handleError)
|
||||
df.addBoth(lambda _: lbrynet.dht.protocol.reactor.stop())
|
||||
lbrynet.dht.protocol.reactor.run()
|
||||
|
||||
for _ in range(10):
|
||||
self._reactor.advance(1)
|
||||
|
||||
self.failIf(self.error, self.error)
|
||||
# The list of sent RPC messages should be empty at this stage
|
||||
self.failUnlessEqual(len(self.protocol._sentMessages), 0,
|
||||
self.failUnlessEqual(len(self.node._protocol._sentMessages), 0,
|
||||
'The protocol is still waiting for a RPC result, '
|
||||
'but the transaction is already done!')
|
||||
|
||||
|
@ -105,8 +133,12 @@ class KademliaProtocolTest(unittest.TestCase):
|
|||
Verifies that a RPC request for an existing but unpublished
|
||||
method is denied, and that the associated (remote) exception gets
|
||||
raised locally """
|
||||
remoteContact = lbrynet.dht.contact.Contact('2' * 48, '127.0.0.1', 9182, self.protocol)
|
||||
self.node.addContact(remoteContact)
|
||||
remote_node = Node(node_id='2' * 48, udpPort=self.udpPort, externalIP="127.0.0.2", listenUDP=listenUDP,
|
||||
resolve=resolve, clock=self._reactor, callLater=self._reactor.callLater)
|
||||
remote_node.start_listening()
|
||||
remote_contact = lbrynet.dht.contact.Contact('2' * 48, '127.0.0.2', 9182, self.node._protocol)
|
||||
self.node.addContact(remote_contact)
|
||||
|
||||
self.error = None
|
||||
|
||||
def handleError(f):
|
||||
|
@ -123,24 +155,26 @@ class KademliaProtocolTest(unittest.TestCase):
|
|||
self.error = 'The remote method executed successfully, returning: "%s"; ' \
|
||||
'this RPC should not have been allowed.' % result
|
||||
|
||||
# Publish the "local" node on the network
|
||||
lbrynet.dht.protocol.reactor.listenUDP(9182, self.protocol)
|
||||
self.node.start_listening()
|
||||
self._reactor.pump([1 for _ in range(10)])
|
||||
# Simulate the RPC
|
||||
df = remoteContact.not_a_rpc_function()
|
||||
df = remote_contact.not_a_rpc_function()
|
||||
df.addCallback(handleResult)
|
||||
df.addErrback(handleError)
|
||||
df.addBoth(lambda _: lbrynet.dht.protocol.reactor.stop())
|
||||
lbrynet.dht.protocol.reactor.run()
|
||||
self._reactor.pump([1 for _ in range(10)])
|
||||
self.failIf(self.error, self.error)
|
||||
# The list of sent RPC messages should be empty at this stage
|
||||
self.failUnlessEqual(len(self.protocol._sentMessages), 0,
|
||||
self.failUnlessEqual(len(self.node._protocol._sentMessages), 0,
|
||||
'The protocol is still waiting for a RPC result, '
|
||||
'but the transaction is already done!')
|
||||
|
||||
def testRPCRequestArgs(self):
|
||||
""" Tests if an RPC requiring arguments is executed correctly """
|
||||
remoteContact = lbrynet.dht.contact.Contact('2' * 48, '127.0.0.1', 9182, self.protocol)
|
||||
self.node.addContact(remoteContact)
|
||||
remote_node = Node(node_id='2' * 48, udpPort=self.udpPort, externalIP="127.0.0.2", listenUDP=listenUDP,
|
||||
resolve=resolve, clock=self._reactor, callLater=self._reactor.callLater)
|
||||
remote_node.start_listening()
|
||||
remote_contact = lbrynet.dht.contact.Contact('2' * 48, '127.0.0.2', 9182, self.node._protocol)
|
||||
self.node.addContact(remote_contact)
|
||||
self.error = None
|
||||
|
||||
def handleError(f):
|
||||
|
@ -153,15 +187,14 @@ class KademliaProtocolTest(unittest.TestCase):
|
|||
(expectedResult, result)
|
||||
|
||||
# Publish the "local" node on the network
|
||||
lbrynet.dht.protocol.reactor.listenUDP(9182, self.protocol)
|
||||
self.node.start_listening()
|
||||
# Simulate the RPC
|
||||
df = remoteContact.ping()
|
||||
df = remote_contact.ping()
|
||||
df.addCallback(handleResult)
|
||||
df.addErrback(handleError)
|
||||
df.addBoth(lambda _: lbrynet.dht.protocol.reactor.stop())
|
||||
lbrynet.dht.protocol.reactor.run()
|
||||
self._reactor.pump([1 for _ in range(10)])
|
||||
self.failIf(self.error, self.error)
|
||||
# The list of sent RPC messages should be empty at this stage
|
||||
self.failUnlessEqual(len(self.protocol._sentMessages), 0,
|
||||
self.failUnlessEqual(len(self.node._protocol._sentMessages), 0,
|
||||
'The protocol is still waiting for a RPC result, '
|
||||
'but the transaction is already done!')
|
||||
|
|
|
@ -1,12 +1,12 @@
|
|||
import hashlib
|
||||
import unittest
|
||||
|
||||
#from lbrynet.dht import contact, routingtable, constants
|
||||
|
||||
import lbrynet.dht.constants
|
||||
import lbrynet.dht.routingtable
|
||||
import lbrynet.dht.contact
|
||||
import lbrynet.dht.node
|
||||
import lbrynet.dht.distance
|
||||
|
||||
|
||||
class FakeRPCProtocol(object):
|
||||
""" Fake RPC protocol; allows lbrynet.dht.contact.Contact objects to "send" RPCs """
|
||||
|
@ -42,15 +42,15 @@ class TreeRoutingTableTest(unittest.TestCase):
|
|||
basicTestList = [('123456789', '123456789', 0L), ('12345', '98765', 34527773184L)]
|
||||
|
||||
for test in basicTestList:
|
||||
result = lbrynet.dht.node.Distance(test[0])(test[1])
|
||||
result = lbrynet.dht.distance.Distance(test[0])(test[1])
|
||||
self.failIf(result != test[2], 'Result of _distance() should be %s but %s returned' %
|
||||
(test[2], result))
|
||||
|
||||
baseIp = '146.64.19.111'
|
||||
ipTestList = ['146.64.29.222', '192.68.19.333']
|
||||
|
||||
distanceOne = lbrynet.dht.node.Distance(baseIp)(ipTestList[0])
|
||||
distanceTwo = lbrynet.dht.node.Distance(baseIp)(ipTestList[1])
|
||||
distanceOne = lbrynet.dht.distance.Distance(baseIp)(ipTestList[0])
|
||||
distanceTwo = lbrynet.dht.distance.Distance(baseIp)(ipTestList[1])
|
||||
|
||||
self.failIf(distanceOne > distanceTwo, '%s should be closer to the base ip %s than %s' %
|
||||
(ipTestList[0], baseIp, ipTestList[1]))
|
||||
|
@ -184,10 +184,6 @@ class TreeRoutingTableTest(unittest.TestCase):
|
|||
'New contact should have been discarded (since RPC is faked in this test)')
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
class KeyErrorFixedTest(unittest.TestCase):
|
||||
""" Basic tests case for boolean operators on the Contact class """
|
||||
|
||||
|
|
|
@ -8,7 +8,6 @@ from lbrynet.database.storage import SQLiteStorage
|
|||
from lbrynet.core.StreamDescriptor import get_sd_info, BlobStreamDescriptorReader
|
||||
from lbrynet.core import BlobManager
|
||||
from lbrynet.core import Session
|
||||
from lbrynet.core.server import DHTHashAnnouncer
|
||||
from lbrynet.file_manager import EncryptedFileCreator
|
||||
from lbrynet.file_manager import EncryptedFileManager
|
||||
from lbrynet.tests import mocks
|
||||
|
@ -32,10 +31,7 @@ class CreateEncryptedFileTest(unittest.TestCase):
|
|||
|
||||
self.session = mock.Mock(spec=Session.Session)(None, None)
|
||||
self.session.payment_rate_manager.min_blob_data_payment_rate = 0
|
||||
|
||||
hash_announcer = DHTHashAnnouncer.DHTHashAnnouncer(None, None)
|
||||
self.blob_manager = BlobManager.DiskBlobManager(
|
||||
hash_announcer, self.tmp_blob_dir, SQLiteStorage(self.tmp_db_dir))
|
||||
self.blob_manager = BlobManager.DiskBlobManager(self.tmp_blob_dir, SQLiteStorage(self.tmp_db_dir))
|
||||
self.session.blob_manager = self.blob_manager
|
||||
self.session.storage = self.session.blob_manager.storage
|
||||
self.file_manager = EncryptedFileManager.EncryptedFileManager(self.session, object())
|
||||
|
@ -74,6 +70,7 @@ class CreateEncryptedFileTest(unittest.TestCase):
|
|||
# this comes from the database, the blobs returned are sorted
|
||||
sd_info = yield get_sd_info(self.session.storage, lbry_file.stream_hash, include_blobs=True)
|
||||
self.assertDictEqual(sd_info, sd_file_info)
|
||||
self.assertListEqual(sd_info['blobs'], sd_file_info['blobs'])
|
||||
self.assertEqual(sd_info['stream_hash'], expected_stream_hash)
|
||||
self.assertEqual(len(sd_info['blobs']), 3)
|
||||
self.assertNotEqual(sd_info['blobs'][0]['length'], 0)
|
||||
|
|
|
@ -5,24 +5,36 @@ import os
|
|||
import tempfile
|
||||
import shutil
|
||||
import mock
|
||||
import logging
|
||||
|
||||
from lbrynet.dht.encoding import Bencode
|
||||
from lbrynet.dht.error import DecodeError
|
||||
from lbrynet.dht.msgformat import DefaultFormat
|
||||
from lbrynet.dht.msgtypes import ResponseMessage, RequestMessage, ErrorMessage
|
||||
|
||||
_encode = Bencode()
|
||||
_datagram_formatter = DefaultFormat()
|
||||
DEFAULT_TIMESTAMP = datetime.datetime(2016, 1, 1)
|
||||
DEFAULT_ISO_TIME = time.mktime(DEFAULT_TIMESTAMP.timetuple())
|
||||
|
||||
log = logging.getLogger("lbrynet.tests.util")
|
||||
|
||||
|
||||
def mk_db_and_blob_dir():
|
||||
db_dir = tempfile.mkdtemp()
|
||||
blob_dir = tempfile.mkdtemp()
|
||||
return db_dir, blob_dir
|
||||
|
||||
|
||||
def rm_db_and_blob_dir(db_dir, blob_dir):
|
||||
shutil.rmtree(db_dir, ignore_errors=True)
|
||||
shutil.rmtree(blob_dir, ignore_errors=True)
|
||||
|
||||
|
||||
def random_lbry_hash():
|
||||
return binascii.b2a_hex(os.urandom(48))
|
||||
|
||||
|
||||
def resetTime(test_case, timestamp=DEFAULT_TIMESTAMP):
|
||||
iso_time = time.mktime(timestamp.timetuple())
|
||||
patcher = mock.patch('time.time')
|
||||
|
@ -37,5 +49,28 @@ def resetTime(test_case, timestamp=DEFAULT_TIMESTAMP):
|
|||
patcher.start().return_value = timestamp
|
||||
test_case.addCleanup(patcher.stop)
|
||||
|
||||
|
||||
def is_android():
|
||||
return 'ANDROID_ARGUMENT' in os.environ # detect Android using the Kivy way
|
||||
|
||||
|
||||
def debug_kademlia_packet(data, source, destination, node):
|
||||
if log.level != logging.DEBUG:
|
||||
return
|
||||
try:
|
||||
packet = _datagram_formatter.fromPrimitive(_encode.decode(data))
|
||||
if isinstance(packet, RequestMessage):
|
||||
log.debug("request %s --> %s %s (node time %s)", source[0], destination[0], packet.request,
|
||||
node.clock.seconds())
|
||||
elif isinstance(packet, ResponseMessage):
|
||||
if isinstance(packet.response, (str, unicode)):
|
||||
log.debug("response %s <-- %s %s (node time %s)", destination[0], source[0], packet.response,
|
||||
node.clock.seconds())
|
||||
else:
|
||||
log.debug("response %s <-- %s %i contacts (node time %s)", destination[0], source[0],
|
||||
len(packet.response), node.clock.seconds())
|
||||
elif isinstance(packet, ErrorMessage):
|
||||
log.error("error %s <-- %s %s (node time %s)", destination[0], source[0], packet.exceptionType,
|
||||
node.clock.seconds())
|
||||
except DecodeError:
|
||||
log.exception("decode error %s --> %s (node time %s)", source[0], destination[0], node.clock.seconds())
|
||||
|
|
0
requirements_testing.txt
Normal file
0
requirements_testing.txt
Normal file
|
@ -73,11 +73,33 @@ def connect(port=None):
|
|||
yield reactor.stop()
|
||||
|
||||
|
||||
def getApproximateTotalDHTNodes(node):
|
||||
from lbrynet.dht import constants
|
||||
# get the deepest bucket and the number of contacts in that bucket and multiply it
|
||||
# by the number of equivalently deep buckets in the whole DHT to get a really bad
|
||||
# estimate!
|
||||
bucket = node._routingTable._buckets[node._routingTable._kbucketIndex(node.node_id)]
|
||||
num_in_bucket = len(bucket._contacts)
|
||||
factor = (2 ** constants.key_bits) / (bucket.rangeMax - bucket.rangeMin)
|
||||
return num_in_bucket * factor
|
||||
|
||||
|
||||
def getApproximateTotalHashes(node):
|
||||
# Divide the number of hashes we know about by k to get a really, really, really
|
||||
# bad estimate of the average number of hashes per node, then multiply by the
|
||||
# approximate number of nodes to get a horrendous estimate of the total number
|
||||
# of hashes in the DHT
|
||||
num_in_data_store = len(node._dataStore._dict)
|
||||
if num_in_data_store == 0:
|
||||
return 0
|
||||
return num_in_data_store * getApproximateTotalDHTNodes(node) / 8
|
||||
|
||||
|
||||
@defer.inlineCallbacks
|
||||
def find(node):
|
||||
try:
|
||||
log.info("Approximate number of nodes in DHT: %s", str(node.getApproximateTotalDHTNodes()))
|
||||
log.info("Approximate number of blobs in DHT: %s", str(node.getApproximateTotalHashes()))
|
||||
log.info("Approximate number of nodes in DHT: %s", str(getApproximateTotalDHTNodes(node)))
|
||||
log.info("Approximate number of blobs in DHT: %s", str(getApproximateTotalHashes(node)))
|
||||
|
||||
h = "578f5e82da7db97bfe0677826d452cc0c65406a8e986c9caa126af4ecdbf4913daad2f7f5d1fb0ffec17d0bf8f187f5a"
|
||||
peersFake = yield node.getPeersForBlob(h.decode("hex"))
|
||||
|
|
|
@ -14,7 +14,7 @@ from lbrynet.core import log_support, Wallet, Peer
|
|||
from lbrynet.core.SinglePeerDownloader import SinglePeerDownloader
|
||||
from lbrynet.core.StreamDescriptor import BlobStreamDescriptorReader
|
||||
from lbrynet.core.BlobManager import DiskBlobManager
|
||||
from lbrynet.core.HashAnnouncer import DummyHashAnnouncer
|
||||
from lbrynet.dht.hashannouncer import DummyHashAnnouncer
|
||||
|
||||
log = logging.getLogger()
|
||||
|
||||
|
|
|
@ -5,12 +5,11 @@ import sys
|
|||
|
||||
from twisted.internet import defer
|
||||
from twisted.internet import reactor
|
||||
from twisted.protocols import basic
|
||||
from twisted.web.client import FileBodyProducer
|
||||
|
||||
from lbrynet import conf
|
||||
from lbrynet.core import log_support
|
||||
from lbrynet.core.HashAnnouncer import DummyHashAnnouncer
|
||||
from lbrynet.dht.hashannouncer import DummyHashAnnouncer
|
||||
from lbrynet.core.BlobManager import DiskBlobManager
|
||||
from lbrynet.cryptstream.CryptStreamCreator import CryptStreamCreator
|
||||
|
||||
|
|
|
@ -17,7 +17,7 @@ from twisted.protocols import basic
|
|||
|
||||
from lbrynet import conf
|
||||
from lbrynet.core import BlobManager
|
||||
from lbrynet.core import HashAnnouncer
|
||||
from lbrynet.dht import hashannouncer
|
||||
from lbrynet.core import log_support
|
||||
from lbrynet.cryptstream import CryptStreamCreator
|
||||
|
||||
|
@ -52,7 +52,7 @@ def reseed_file(input_file, sd_blob):
|
|||
sd_blob = SdBlob.new_instance(sd_blob)
|
||||
db_dir = conf.settings['data_dir']
|
||||
blobfile_dir = os.path.join(db_dir, "blobfiles")
|
||||
announcer = HashAnnouncer.DummyHashAnnouncer()
|
||||
announcer = hashannouncer.DummyHashAnnouncer()
|
||||
blob_manager = BlobManager.DiskBlobManager(announcer, blobfile_dir, db_dir)
|
||||
yield blob_manager.setup()
|
||||
creator = CryptStreamCreator.CryptStreamCreator(
|
||||
|
|
Loading…
Reference in a new issue