core-lightning/tests/test_gossip.py

1168 lines
46 KiB
Python
Raw Normal View History

from fixtures import * # noqa: F401,F403
from lightning import RpcError
from utils import wait_for, TIMEOUT, only_one
import json
import logging
import os
import pytest
import struct
import subprocess
import time
import unittest
with open('config.vars') as configfile:
config = dict([(line.rstrip().split('=', 1)) for line in configfile])
DEVELOPER = os.getenv("DEVELOPER", config['DEVELOPER']) == "1"
@unittest.skipIf(not DEVELOPER, "needs --dev-broadcast-interval, --dev-channelupdate-interval")
def test_gossip_pruning(node_factory, bitcoind):
""" Create channel and see it being updated in time before pruning
"""
opts = {'dev-channel-update-interval': 5}
l1, l2, l3 = node_factory.get_nodes(3, opts)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l2.rpc.connect(l3.info['id'], 'localhost', l3.port)
scid1 = l1.fund_channel(l2, 10**6)
scid2 = l2.fund_channel(l3, 10**6)
bitcoind.generate_block(6)
# Channels should be activated locally
wait_for(lambda: [c['active'] for c in l1.rpc.listchannels()['channels']] == [True] * 4)
wait_for(lambda: [c['active'] for c in l2.rpc.listchannels()['channels']] == [True] * 4)
wait_for(lambda: [c['active'] for c in l3.rpc.listchannels()['channels']] == [True] * 4)
# All of them should send a keepalive message
l1.daemon.wait_for_logs([
'Sending keepalive channel_update for {}'.format(scid1),
])
l2.daemon.wait_for_logs([
'Sending keepalive channel_update for {}'.format(scid1),
'Sending keepalive channel_update for {}'.format(scid2),
])
l3.daemon.wait_for_logs([
'Sending keepalive channel_update for {}'.format(scid2),
])
# Now kill l3, so that l2 and l1 can prune it from their view after 10 seconds
# FIXME: This sleep() masks a real bug: that channeld sends a
# channel_update message (to disable the channel) with same
# timestamp as the last keepalive, and thus is ignored. The minimal
# fix is to backdate the keepalives 1 second, but maybe we should
# simply have gossipd generate all updates?
time.sleep(1)
l3.stop()
l1.daemon.wait_for_log("Pruning channel {} from network view".format(scid2))
l2.daemon.wait_for_log("Pruning channel {} from network view".format(scid2))
assert scid2 not in [c['short_channel_id'] for c in l1.rpc.listchannels()['channels']]
assert scid2 not in [c['short_channel_id'] for c in l2.rpc.listchannels()['channels']]
assert l3.info['id'] not in [n['nodeid'] for n in l1.rpc.listnodes()['nodes']]
assert l3.info['id'] not in [n['nodeid'] for n in l2.rpc.listnodes()['nodes']]
@unittest.skipIf(not DEVELOPER, "needs --dev-broadcast-interval, --dev-no-reconnect")
def test_gossip_disable_channels(node_factory, bitcoind):
"""Simple test to check that channels get disabled correctly on disconnect and
reenabled upon reconnecting
"""
opts = {'dev-no-reconnect': None, 'may_reconnect': True}
l1, l2 = node_factory.get_nodes(2, opts=opts)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
scid = l1.fund_channel(l2, 10**6)
bitcoind.generate_block(5)
def count_active(node):
chans = node.rpc.listchannels()['channels']
active = [c for c in chans if c['active']]
return len(active)
l1.wait_channel_active(scid)
l2.wait_channel_active(scid)
assert(count_active(l1) == 2)
assert(count_active(l2) == 2)
l2.restart()
wait_for(lambda: count_active(l1) == 0)
assert(count_active(l2) == 0)
# Now reconnect, they should re-enable the channels
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
wait_for(lambda: count_active(l1) == 2)
wait_for(lambda: count_active(l2) == 2)
@unittest.skipIf(not DEVELOPER, "needs --dev-allow-localhost")
def test_announce_address(node_factory, bitcoind):
"""Make sure our announcements are well formed."""
# We do not allow announcement of duplicates.
opts = {'announce-addr':
['4acth47i6kxnvkewtm6q7ib2s3ufpo5sqbsnzjpbi7utijcltosqemad.onion',
'silkroad6ownowfk.onion',
'1.2.3.4:1234',
'::'],
'log-level': 'io',
'dev-allow-localhost': None}
l1, l2 = node_factory.get_nodes(2, opts=[opts, {}])
# It should warn about the collision between --addr=127.0.0.1:<ephem>
# and --announce-addr=1.2.3.4:1234 (may happen before get_nodes returns).
wait_for(lambda: l1.daemon.is_in_log('Cannot announce address 127.0.0.1:[0-9]*, already announcing 1.2.3.4:1234'))
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
scid = l1.fund_channel(l2, 10**6)
bitcoind.generate_block(5)
# Activate IO logging for l1.
subprocess.run(['kill', '-USR1', l1.subd_pid('channeld')])
l1.wait_channel_active(scid)
l2.wait_channel_active(scid)
# We should see it send node announce (257 = 0x0101)
l1.daemon.wait_for_log(r"\[OUT\] 0101.*004d010102030404d202000000000000000000000000000000002607039216a8b803f3acd758aa260704e00533f3e8f2aedaa8969b3d0fa03a96e857bbb28064dca5e147e934244b9ba50230032607'")
@unittest.skipIf(not DEVELOPER, "needs DEVELOPER=1")
def test_gossip_timestamp_filter(node_factory, bitcoind):
# Need full IO logging so we can see gossip (from gossipd and channeld)
l1, l2, l3 = node_factory.line_graph(3, opts={'log-level': 'io'}, fundchannel=False)
# Full IO logging for connectds
subprocess.run(['kill', '-USR1', l1.subd_pid('connectd')])
subprocess.run(['kill', '-USR1', l2.subd_pid('connectd')])
before_anything = int(time.time() - 1.0)
# Make a public channel.
chan12 = l1.fund_channel(l2, 10**5)
bitcoind.generate_block(5)
l3.wait_for_channel_updates([chan12])
after_12 = int(time.time())
# Full IO logging for l1's channeld
subprocess.run(['kill', '-USR1', l1.subd_pid('channeld')])
# Make another one, different timestamp.
chan23 = l2.fund_channel(l3, 10**5)
bitcoind.generate_block(5)
l1.wait_for_channel_updates([chan23])
after_23 = int(time.time())
# Make sure l1 has received all the gossip.
wait_for(lambda: ['alias' in node for node in l1.rpc.listnodes()['nodes']] == [True, True, True])
# l1 sets broad timestamp, will receive info about both channels again.
l1.rpc.dev_send_timestamp_filter(id=l2.info['id'],
first=0,
range=0xFFFFFFFF)
before_sendfilter = l1.daemon.logsearch_start
# 0x0100 = channel_announcement
# 0x0102 = channel_update
# 0x0101 = node_announcement
# The order of node_announcements relative to others is undefined.
l1.daemon.wait_for_logs([r'\[IN\] 0102',
r'\[IN\] 0102',
r'\[IN\] 0100',
r'\[IN\] 0100',
r'\[IN\] 0102',
r'\[IN\] 0102',
r'\[IN\] 0101',
r'\[IN\] 0101',
r'\[IN\] 0101'])
# Now timestamp which doesn't overlap (gives nothing).
before_sendfilter = l1.daemon.logsearch_start
l1.rpc.dev_send_timestamp_filter(id=l2.info['id'],
first=0,
range=before_anything)
time.sleep(1)
assert not l1.daemon.is_in_log(r'\[IN\] 0100', before_sendfilter)
# Now choose range which will only give first update.
l1.rpc.dev_send_timestamp_filter(id=l2.info['id'],
first=before_anything,
range=after_12 - before_anything + 1)
# 0x0100 = channel_announcement
l1.daemon.wait_for_log(r'\[IN\] 0100')
# 0x0102 = channel_update
# (Node announcement may have any timestamp)
l1.daemon.wait_for_log(r'\[IN\] 0102')
l1.daemon.wait_for_log(r'\[IN\] 0102')
# Now choose range which will only give second update.
l1.rpc.dev_send_timestamp_filter(id=l2.info['id'],
first=after_12,
range=after_23 - after_12 + 1)
# 0x0100 = channel_announcement
l1.daemon.wait_for_log(r'\[IN\] 0100')
# 0x0102 = channel_update
# (Node announcement may have any timestamp)
l1.daemon.wait_for_log(r'\[IN\] 0102')
l1.daemon.wait_for_log(r'\[IN\] 0102')
@unittest.skipIf(not DEVELOPER, "needs --dev-allow-localhost")
def test_connect_by_gossip(node_factory, bitcoind):
"""Test connecting to an unknown peer using node gossip
"""
# l1 announces a bogus addresses.
l1, l2, l3 = node_factory.get_nodes(3,
opts=[{'announce-addr':
['127.0.0.1:2',
'[::]:2',
'3fyb44wdhnd2ghhl.onion',
'vww6ybal4bd7szmgncyruucpgfkqahzddi37ktceo3ah7ngmcopnpyyd.onion'],
'dev-allow-localhost': None},
{},
{'dev-allow-localhost': None,
'log-level': 'io'}])
l2.rpc.connect(l3.info['id'], 'localhost', l3.port)
# Nodes are gossiped only if they have channels
chanid = l2.fund_channel(l3, 10**6)
bitcoind.generate_block(5)
# Let channel reach announcement depth
l2.wait_channel_active(chanid)
# Make sure l3 has given node announcement to l2.
l2.daemon.wait_for_logs(['Received node_announcement for node {}'.format(l3.info['id'])])
# Let l1 learn of l3 by node gossip
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l1.daemon.wait_for_logs(['Received node_announcement for node {}'.format(l3.info['id'])])
# Have l1 connect to l3 without explicit host and port.
l1.rpc.connect(l3.info['id'])
@unittest.skipIf(not DEVELOPER, "DEVELOPER=1 needed to speed up gossip propagation, would be too long otherwise")
def test_gossip_jsonrpc(node_factory):
l1, l2 = node_factory.line_graph(2, fundchannel=True, wait_for_announce=False)
# Shouldn't send announce signatures until 6 deep.
assert not l1.daemon.is_in_log('peer_out WIRE_ANNOUNCEMENT_SIGNATURES')
# Channels should be activated locally
wait_for(lambda: len(l1.rpc.listchannels()['channels']) == 2)
wait_for(lambda: len(l2.rpc.listchannels()['channels']) == 2)
# Make sure we can route through the channel, will raise on failure
l1.rpc.getroute(l2.info['id'], 100, 1)
# Outgoing should be active, but not public.
channels1 = l1.rpc.listchannels()['channels']
channels2 = l2.rpc.listchannels()['channels']
assert [c['active'] for c in channels1] == [True, True]
assert [c['active'] for c in channels2] == [True, True]
# The incoming direction will be considered public, hence check for out
# outgoing only
assert len([c for c in channels1 if not c['public']]) == 2
assert len([c for c in channels2 if not c['public']]) == 2
# Test listchannels-by-source
channels1 = l1.rpc.listchannels(source=l1.info['id'])['channels']
channels2 = l2.rpc.listchannels(source=l1.info['id'])['channels']
assert only_one(channels1)['source'] == l1.info['id']
assert only_one(channels1)['destination'] == l2.info['id']
assert channels1 == channels2
l2.rpc.listchannels()['channels']
# Now proceed to funding-depth and do a full gossip round
l1.bitcoin.generate_block(5)
# Could happen in either order.
l1.daemon.wait_for_logs(['peer_out WIRE_ANNOUNCEMENT_SIGNATURES',
'peer_in WIRE_ANNOUNCEMENT_SIGNATURES'])
# Just wait for the update to kick off and then check the effect
needle = "Received node_announcement for node"
l1.daemon.wait_for_log(needle)
l2.daemon.wait_for_log(needle)
# Need to increase timeout, intervals cannot be shortened with DEVELOPER=0
wait_for(lambda: len(l1.getactivechannels()) == 2, timeout=60)
wait_for(lambda: len(l2.getactivechannels()) == 2, timeout=60)
nodes = l1.rpc.listnodes()['nodes']
assert set([n['nodeid'] for n in nodes]) == set([l1.info['id'], l2.info['id']])
# Test listnodes with an arg, while we're here.
n1 = l1.rpc.listnodes(l1.info['id'])['nodes'][0]
n2 = l1.rpc.listnodes(l2.info['id'])['nodes'][0]
assert n1['nodeid'] == l1.info['id']
assert n2['nodeid'] == l2.info['id']
# Might not have seen other node-announce yet.
assert n1['alias'].startswith('JUNIORBEAM')
assert n1['color'] == '0266e4'
if 'alias' not in n2:
assert 'color' not in n2
assert 'addresses' not in n2
else:
assert n2['alias'].startswith('SILENTARTIST')
assert n2['color'] == '022d22'
assert [c['active'] for c in l1.rpc.listchannels()['channels']] == [True, True]
assert [c['public'] for c in l1.rpc.listchannels()['channels']] == [True, True]
assert [c['active'] for c in l2.rpc.listchannels()['channels']] == [True, True]
assert [c['public'] for c in l2.rpc.listchannels()['channels']] == [True, True]
@unittest.skipIf(not DEVELOPER, "needs DEVELOPER=1 for --dev-broadcast-interval")
def test_gossip_badsig(node_factory):
"""Make sure node announcement signatures are ok.
This is a smoke test to see if signatures fail. This used to be the case
occasionally before PR #276 was merged: we'd be waiting for the HSM to reply
with a signature and would then regenerate the message, which might roll the
timestamp, invalidating the signature.
"""
l1, l2, l3 = node_factory.get_nodes(3)
# l2 connects to both, so l1 can't reconnect and thus l2 drops to chain
l2.rpc.connect(l1.info['id'], 'localhost', l1.port)
l2.rpc.connect(l3.info['id'], 'localhost', l3.port)
l2.fund_channel(l1, 10**6)
l2.fund_channel(l3, 10**6)
# Wait for route propagation.
l1.bitcoin.generate_block(5)
l1.daemon.wait_for_log('Received node_announcement for node {}'
.format(l3.info['id']))
assert not l1.daemon.is_in_log('signature verification failed')
assert not l2.daemon.is_in_log('signature verification failed')
assert not l3.daemon.is_in_log('signature verification failed')
def test_gossip_weirdalias(node_factory, bitcoind):
weird_name = '\t \n \" \n \r \n \\'
normal_name = 'Normal name'
opts = [
{'alias': weird_name},
{'alias': normal_name}
]
l1, l2 = node_factory.get_nodes(2, opts=opts)
weird_name_json = json.encoder.JSONEncoder().encode(weird_name)[1:-1].replace('\\', '\\\\')
aliasline = l1.daemon.is_in_log('Server started with public key .* alias')
assert weird_name_json in str(aliasline)
assert l2.daemon.is_in_log('Server started with public key .* alias {}'
.format(normal_name))
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l2.daemon.wait_for_log('openingd-{} chan #1: Handed peer, entering loop'.format(l1.info['id']))
l2.fund_channel(l1, 10**6)
bitcoind.generate_block(6)
# They should gossip together.
l1.daemon.wait_for_log('Received node_announcement for node {}'
.format(l2.info['id']))
l2.daemon.wait_for_log('Received node_announcement for node {}'
.format(l1.info['id']))
node = l1.rpc.listnodes(l1.info['id'])['nodes'][0]
assert node['alias'] == weird_name
node = l2.rpc.listnodes(l1.info['id'])['nodes'][0]
assert node['alias'] == weird_name
@unittest.skipIf(not DEVELOPER, "needs DEVELOPER=1 for --dev-no-reconnect")
def test_gossip_persistence(node_factory, bitcoind):
"""Gossip for a while, restart and it should remember.
Also tests for funding outpoint spends, and they should be persisted
too.
"""
opts = {'dev-no-reconnect': None, 'may_reconnect': True}
l1, l2, l3, l4 = node_factory.get_nodes(4, opts=opts)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l2.rpc.connect(l3.info['id'], 'localhost', l3.port)
l3.rpc.connect(l4.info['id'], 'localhost', l4.port)
scid12 = l1.fund_channel(l2, 10**6)
scid23 = l2.fund_channel(l3, 10**6)
# Make channels public, except for l3 -> l4, which is kept local-only for now
bitcoind.generate_block(5)
scid34 = l3.fund_channel(l4, 10**6)
bitcoind.generate_block(1)
def active(node):
chans = node.rpc.listchannels()['channels']
return sorted([c['short_channel_id'] for c in chans if c['active']])
def non_public(node):
chans = node.rpc.listchannels()['channels']
return sorted([c['short_channel_id'] for c in chans if not c['public']])
# Channels should be activated
wait_for(lambda: active(l1) == [scid12, scid12, scid23, scid23])
wait_for(lambda: active(l2) == [scid12, scid12, scid23, scid23])
# This one sees its private channel
wait_for(lambda: active(l3) == [scid12, scid12, scid23, scid23, scid34, scid34])
# l1 restarts and doesn't connect, but loads from persisted store, all
# local channels should be disabled, leaving only the two l2 <-> l3
# directions
l1.restart()
wait_for(lambda: active(l1) == [scid23, scid23])
# Now reconnect, they should re-enable the two l1 <-> l2 directions
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
wait_for(lambda: active(l1) == [scid12, scid12, scid23, scid23])
# Now spend the funding tx, generate a block and see others deleting the
# channel from their network view
l1.rpc.dev_fail(l2.info['id'])
time.sleep(1)
bitcoind.generate_block(1)
wait_for(lambda: active(l1) == [scid23, scid23])
wait_for(lambda: active(l2) == [scid23, scid23])
wait_for(lambda: active(l3) == [scid23, scid23, scid34, scid34])
# The channel l3 -> l4 should be known only to them
assert non_public(l1) == []
assert non_public(l2) == []
wait_for(lambda: non_public(l3) == [scid34, scid34])
wait_for(lambda: non_public(l4) == [scid34, scid34])
# Finally, it should also remember the deletion after a restart
l3.restart()
l4.restart()
l2.rpc.connect(l3.info['id'], 'localhost', l3.port)
l3.rpc.connect(l4.info['id'], 'localhost', l4.port)
wait_for(lambda: active(l3) == [scid23, scid23, scid34, scid34])
# Both l3 and l4 should remember their local-only channel
wait_for(lambda: non_public(l3) == [scid34, scid34])
wait_for(lambda: non_public(l4) == [scid34, scid34])
@unittest.skipIf(not DEVELOPER, "needs DEVELOPER=1")
def test_routing_gossip_reconnect(node_factory):
# Connect two peers, reconnect and then see if we resume the
# gossip.
disconnects = ['-WIRE_CHANNEL_ANNOUNCEMENT']
l1 = node_factory.get_node(disconnect=disconnects,
may_reconnect=True)
l2 = node_factory.get_node(may_reconnect=True)
l3 = node_factory.get_node()
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l1.openchannel(l2, 20000)
# Now open new channels and everybody should sync
l2.rpc.connect(l3.info['id'], 'localhost', l3.port)
l2.openchannel(l3, 20000)
# Settle the gossip
for n in [l1, l2, l3]:
wait_for(lambda: len(n.rpc.listchannels()['channels']) == 4)
@unittest.skipIf(not DEVELOPER, "needs DEVELOPER=1")
def test_gossip_no_empty_announcements(node_factory, bitcoind):
# Need full IO logging so we can see gossip
opts = {'log-level': 'io'}
l1, l2 = node_factory.get_nodes(2, opts=opts)
# l3 sends CHANNEL_ANNOUNCEMENT to l2, but not CHANNEL_UDPATE.
l3 = node_factory.get_node(disconnect=['+WIRE_CHANNEL_ANNOUNCEMENT'],
options={'dev-no-reconnect': None},
may_reconnect=True)
l4 = node_factory.get_node(may_reconnect=True)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l2.rpc.connect(l3.info['id'], 'localhost', l3.port)
l3.rpc.connect(l4.info['id'], 'localhost', l4.port)
openingd: take peer before we're opening, wait for explicit funding msg. Prior to this, lightningd would hand uninteresting peers back to connectd, which would then return it to lightningd if it sent a non-gossip msg, or if lightningd asked it to release the peer. Now connectd hands the peer to lightningd once we've done the init handshake, which hands it off to openingd. This is a deep structural change, so we do the minimum here and cleanup in the following patches. Lightningd: 1. Remove peer_nongossip handling from connect_control and peer_control. 2. Remove list of outstanding fundchannel command; it was only needed to find the race between us asking connectd to release the peer and it reconnecting. 3. We can no longer tell if the remote end has started trying to fund a channel (until it has succeeded): it's very transitory anyway so not worth fixing. 4. We now always have a struct peer, and allocate an uncommitted_channel for it, though it may never be used if neither end funds a channel. 5. We start funding on messages for openingd: we can get a funder_reply or a fundee, or an error in response to our request to fund a channel. so we handle all of them. 6. A new peer_start_openingd() is called after connectd hands us a peer. 7. json_fund_channel just looks through local peers; there are none hidden in connectd any more. 8. We sometimes start a new openingd just to send an error message. Openingd: 1. We always have information we need to accept them funding a channel (in the init message). 2. We have to listen for three fds: peer, gossip and master, so we opencode the poll. 3. We have an explicit message to start trying to fund a channel. 4. We can be told to send a message in our init message. Testing: 1. We don't handle some things gracefully yet, so two tests are disabled. 2. 'hand_back_peer .*: now local again' from connectd is no longer a message, openingd says 'Handed peer, entering loop' once its managing it. 3. peer['state'] used to be set to 'GOSSIPING' (otherwise this field doesn't exist; 'state' is now per-channel. It doesn't exist at all now. 4. Some tests now need to turn on IO logging in openingd, not connectd. 5. There's a gap between connecting on one node and having connectd on the peer hand over the connection to openingd. Our tests sometimes checked getpeers() on the peer, and didn't see anything, so line_graph needed updating. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-08-08 16:10:58 +02:00
# Turn on IO logging for openingd (make sure it's ready!)
l1.daemon.wait_for_log('openingd-.*: Handed peer, entering loop')
subprocess.run(['kill', '-USR1', l1.subd_pid('openingd')])
l2.daemon.wait_for_log('openingd-{}.*: Handed peer, entering loop'.format(l3.info['id']))
subprocess.run(['kill', '-USR1', l2.subd_pid('openingd-{}'.format(l3.info['id']))])
# Make an announced-but-not-updated channel.
l3.fund_channel(l4, 10**5)
bitcoind.generate_block(5)
# 0x0100 = channel_announcement, which goes to l2 before l3 dies.
l2.daemon.wait_for_log(r'\[IN\] 0100')
# l3 actually disconnects from l4 *and* l2! That means we never see
# the (delayed) channel_update from l4.
wait_for(lambda: not l3.rpc.listpeers(l4.info['id'])['peers'][0]['connected'])
l3.rpc.connect(l4.info['id'], 'localhost', l4.port)
# But it never goes to l1, as there's no channel_update.
time.sleep(2)
assert not l1.daemon.is_in_log(r'\[IN\] 0100')
assert len(l1.rpc.listchannels()['channels']) == 0
# If we reconnect, gossip will now flow.
l3.rpc.connect(l2.info['id'], 'localhost', l2.port)
wait_for(lambda: len(l1.rpc.listchannels()['channels']) == 2)
@unittest.skipIf(not DEVELOPER, "needs DEVELOPER=1 for --dev-broadcast-interval")
def test_routing_gossip(node_factory, bitcoind):
nodes = node_factory.get_nodes(5)
for i in range(len(nodes) - 1):
src, dst = nodes[i], nodes[i + 1]
src.rpc.connect(dst.info['id'], 'localhost', dst.port)
src.openchannel(dst, 20000)
# Allow announce messages.
bitcoind.generate_block(5)
# Deep check that all channels are in there
comb = []
for i in range(len(nodes) - 1):
comb.append((nodes[i].info['id'], nodes[i + 1].info['id']))
comb.append((nodes[i + 1].info['id'], nodes[i].info['id']))
def check_gossip(n):
seen = []
channels = n.rpc.listchannels()['channels']
for c in channels:
seen.append((c['source'], c['destination']))
missing = set(comb) - set(seen)
logging.debug("Node {id} is missing channels {chans}".format(
id=n.info['id'],
chans=missing)
)
return len(missing) == 0
for n in nodes:
wait_for(lambda: check_gossip(n))
@unittest.skipIf(not DEVELOPER, "needs DEVELOPER=1")
def test_gossip_query_channel_range(node_factory, bitcoind):
l1, l2, l3, l4 = node_factory.line_graph(4, opts={'log-level': 'io'},
fundchannel=False)
# Make public channels on consecutive blocks
l1.fundwallet(10**6)
l2.fundwallet(10**6)
num_tx = len(bitcoind.rpc.getrawmempool())
l1.rpc.fundchannel(l2.info['id'], 10**5)['tx']
wait_for(lambda: len(bitcoind.rpc.getrawmempool()) == num_tx + 1)
bitcoind.generate_block(1)
num_tx = len(bitcoind.rpc.getrawmempool())
l2.rpc.fundchannel(l3.info['id'], 10**5)['tx']
wait_for(lambda: len(bitcoind.rpc.getrawmempool()) == num_tx + 1)
bitcoind.generate_block(1)
# Get them both to gossip depth.
bitcoind.generate_block(5)
# Make sure l2 has received all the gossip.
l2.daemon.wait_for_logs(['Received node_announcement for node ' + l1.info['id'],
'Received node_announcement for node ' + l3.info['id']])
scid12 = only_one(l1.rpc.listpeers(l2.info['id'])['peers'])['channels'][0]['short_channel_id']
scid23 = only_one(l3.rpc.listpeers(l2.info['id'])['peers'])['channels'][0]['short_channel_id']
block12 = int(scid12.split('x')[0])
block23 = int(scid23.split('x')[0])
assert block23 == block12 + 1
# l1 asks for all channels, gets both.
ret = l1.rpc.dev_query_channel_range(id=l2.info['id'],
first=0,
num=1000000)
assert ret['final_first_block'] == 0
assert ret['final_num_blocks'] == 1000000
assert ret['final_complete']
assert len(ret['short_channel_ids']) == 2
assert ret['short_channel_ids'][0] == scid12
assert ret['short_channel_ids'][1] == scid23
# Does not include scid12
ret = l1.rpc.dev_query_channel_range(id=l2.info['id'],
first=0,
num=block12)
assert ret['final_first_block'] == 0
assert ret['final_num_blocks'] == block12
assert ret['final_complete']
assert len(ret['short_channel_ids']) == 0
# Does include scid12
ret = l1.rpc.dev_query_channel_range(id=l2.info['id'],
first=0,
num=block12 + 1)
assert ret['final_first_block'] == 0
assert ret['final_num_blocks'] == block12 + 1
assert ret['final_complete']
assert len(ret['short_channel_ids']) == 1
assert ret['short_channel_ids'][0] == scid12
# Doesn't include scid23
ret = l1.rpc.dev_query_channel_range(id=l2.info['id'],
first=0,
num=block23)
assert ret['final_first_block'] == 0
assert ret['final_num_blocks'] == block23
assert ret['final_complete']
assert len(ret['short_channel_ids']) == 1
assert ret['short_channel_ids'][0] == scid12
# Does include scid23
ret = l1.rpc.dev_query_channel_range(id=l2.info['id'],
first=block12,
num=block23 - block12 + 1)
assert ret['final_first_block'] == block12
assert ret['final_num_blocks'] == block23 - block12 + 1
assert ret['final_complete']
assert len(ret['short_channel_ids']) == 2
assert ret['short_channel_ids'][0] == scid12
assert ret['short_channel_ids'][1] == scid23
# Only includes scid23
ret = l1.rpc.dev_query_channel_range(id=l2.info['id'],
first=block23,
num=1)
assert ret['final_first_block'] == block23
assert ret['final_num_blocks'] == 1
assert ret['final_complete']
assert len(ret['short_channel_ids']) == 1
assert ret['short_channel_ids'][0] == scid23
# Past both
ret = l1.rpc.dev_query_channel_range(id=l2.info['id'],
first=block23 + 1,
num=1000000)
assert ret['final_first_block'] == block23 + 1
assert ret['final_num_blocks'] == 1000000
assert ret['final_complete']
assert len(ret['short_channel_ids']) == 0
# Turn on IO logging in l1 channeld.
subprocess.run(['kill', '-USR1', l1.subd_pid('channeld')])
# Make l2 split reply into two (technically async)
l2.rpc.dev_set_max_scids_encode_size(max=9)
l2.daemon.wait_for_log('Set max_scids_encode_bytes to 9')
ret = l1.rpc.dev_query_channel_range(id=l2.info['id'],
first=0,
num=1000000)
gossipd: handle overflowing query properly (avoid slow 100% CPU reports) Don't do this: (gdb) bt #0 0x00007f37ae667c40 in ?? () from /lib/x86_64-linux-gnu/libz.so.1 #1 0x00007f37ae668b38 in ?? () from /lib/x86_64-linux-gnu/libz.so.1 #2 0x00007f37ae669907 in deflate () from /lib/x86_64-linux-gnu/libz.so.1 #3 0x00007f37ae674c65 in compress2 () from /lib/x86_64-linux-gnu/libz.so.1 #4 0x000000000040cfe3 in zencode_scids (ctx=0xc1f118, scids=0x2599bc49 "\a\325{", len=176320) at gossipd/gossipd.c:218 #5 0x000000000040d0b3 in encode_short_channel_ids_end (encoded=0x7fff8f98d9f0, max_bytes=65490) at gossipd/gossipd.c:236 #6 0x000000000040dd28 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290511, number_of_blocks=8) at gossipd/gossipd.c:576 #7 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290511, number_of_blocks=16) at gossipd/gossipd.c:595 #8 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290495, number_of_blocks=32) at gossipd/gossipd.c:596 #9 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290495, number_of_blocks=64) at gossipd/gossipd.c:595 #10 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=128) at gossipd/gossipd.c:596 #11 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=256) at gossipd/gossipd.c:595 #12 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=512) at gossipd/gossipd.c:595 #13 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=1024) at gossipd/gossipd.c:595 #14 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=2047) at gossipd/gossipd.c:596 #15 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=4095) at gossipd/gossipd.c:595 #16 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=8191) at gossipd/gossipd.c:595 #17 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=16382) at gossipd/gossipd.c:595 #18 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=32764) at gossipd/gossipd.c:595 #19 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=65528) at gossipd/gossipd.c:595 #20 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=131056) at gossipd/gossipd.c:595 #21 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=262112) at gossipd/gossipd.c:595 #22 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=524225) at gossipd/gossipd.c:595 #23 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=1048450) at gossipd/gossipd.c:595 #24 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=2096900) at gossipd/gossipd.c:595 #25 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=4193801) at gossipd/gossipd.c:595 #26 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=8387603) at gossipd/gossipd.c:595 #27 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=16775207) at gossipd/gossipd.c:595 #28 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=33550414) at gossipd/gossipd.c:596 #29 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=67100829) at gossipd/gossipd.c:595 #30 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=134201659) at gossipd/gossipd.c:595 #31 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=268403318) at gossipd/gossipd.c:595 #32 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=536806636) at gossipd/gossipd.c:595 #33 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=1073613273) at gossipd/gossipd.c:595 #34 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=2147226547) at gossipd/gossipd.c:595 #35 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=4294453094) at gossipd/gossipd.c:595 #36 0x000000000040df26 in handle_query_channel_range (peer=0x3868fc8, msg=0x37e0678 "\001\ao\342\214\n\266\361\263r\301\246\242F\256c\367O\223\036\203e\341Z\b\234h\326\031") at gossipd/gossipd.c:625 The cause was that converting a block number to an scid truncates it at 24 bits. When we look through the index from (truncated number) to (real end number) we get every channel, which is too large to encode, so we iterate again. This fixes both that problem, and also the issue that we'd end up dividing into many empty sections until we get to the highest block number. Instead, we just tack the empty blocks on to then end of the final query. (My initial version requested 0xFFFFFFFE blocks, but the dev code which records what blocks were returned can't make a bitmap that big on 32 bit). Reported-by: George Vaccaro Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2019-01-15 10:37:57 +01:00
# Turns out it sends: 0+53, 53+26, 79+13, 92+7, 99+3, 102+2, 104+1, 105+999895
l1.daemon.wait_for_logs([r'\[IN\] 0108'] * 8)
# It should definitely have split
assert ret['final_first_block'] != 0 or ret['final_num_blocks'] != 1000000
assert ret['final_complete']
assert len(ret['short_channel_ids']) == 2
assert ret['short_channel_ids'][0] == scid12
assert ret['short_channel_ids'][1] == scid23
l2.daemon.wait_for_log('queue_channel_ranges full: splitting')
gossipd: handle overflowing query properly (avoid slow 100% CPU reports) Don't do this: (gdb) bt #0 0x00007f37ae667c40 in ?? () from /lib/x86_64-linux-gnu/libz.so.1 #1 0x00007f37ae668b38 in ?? () from /lib/x86_64-linux-gnu/libz.so.1 #2 0x00007f37ae669907 in deflate () from /lib/x86_64-linux-gnu/libz.so.1 #3 0x00007f37ae674c65 in compress2 () from /lib/x86_64-linux-gnu/libz.so.1 #4 0x000000000040cfe3 in zencode_scids (ctx=0xc1f118, scids=0x2599bc49 "\a\325{", len=176320) at gossipd/gossipd.c:218 #5 0x000000000040d0b3 in encode_short_channel_ids_end (encoded=0x7fff8f98d9f0, max_bytes=65490) at gossipd/gossipd.c:236 #6 0x000000000040dd28 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290511, number_of_blocks=8) at gossipd/gossipd.c:576 #7 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290511, number_of_blocks=16) at gossipd/gossipd.c:595 #8 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290495, number_of_blocks=32) at gossipd/gossipd.c:596 #9 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290495, number_of_blocks=64) at gossipd/gossipd.c:595 #10 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=128) at gossipd/gossipd.c:596 #11 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=256) at gossipd/gossipd.c:595 #12 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=512) at gossipd/gossipd.c:595 #13 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=1024) at gossipd/gossipd.c:595 #14 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=2047) at gossipd/gossipd.c:596 #15 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=4095) at gossipd/gossipd.c:595 #16 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=8191) at gossipd/gossipd.c:595 #17 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=16382) at gossipd/gossipd.c:595 #18 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=32764) at gossipd/gossipd.c:595 #19 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=65528) at gossipd/gossipd.c:595 #20 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=131056) at gossipd/gossipd.c:595 #21 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=262112) at gossipd/gossipd.c:595 #22 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=524225) at gossipd/gossipd.c:595 #23 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=1048450) at gossipd/gossipd.c:595 #24 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=2096900) at gossipd/gossipd.c:595 #25 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=4193801) at gossipd/gossipd.c:595 #26 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=8387603) at gossipd/gossipd.c:595 #27 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=16775207) at gossipd/gossipd.c:595 #28 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=33550414) at gossipd/gossipd.c:596 #29 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=67100829) at gossipd/gossipd.c:595 #30 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=134201659) at gossipd/gossipd.c:595 #31 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=268403318) at gossipd/gossipd.c:595 #32 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=536806636) at gossipd/gossipd.c:595 #33 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=1073613273) at gossipd/gossipd.c:595 #34 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=2147226547) at gossipd/gossipd.c:595 #35 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=4294453094) at gossipd/gossipd.c:595 #36 0x000000000040df26 in handle_query_channel_range (peer=0x3868fc8, msg=0x37e0678 "\001\ao\342\214\n\266\361\263r\301\246\242F\256c\367O\223\036\203e\341Z\b\234h\326\031") at gossipd/gossipd.c:625 The cause was that converting a block number to an scid truncates it at 24 bits. When we look through the index from (truncated number) to (real end number) we get every channel, which is too large to encode, so we iterate again. This fixes both that problem, and also the issue that we'd end up dividing into many empty sections until we get to the highest block number. Instead, we just tack the empty blocks on to then end of the final query. (My initial version requested 0xFFFFFFFE blocks, but the dev code which records what blocks were returned can't make a bitmap that big on 32 bit). Reported-by: George Vaccaro Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2019-01-15 10:37:57 +01:00
# Test overflow case doesn't split forever; should still only get 8 for this
ret = l1.rpc.dev_query_channel_range(id=l2.info['id'],
first=1,
num=429496000)
gossipd: handle overflowing query properly (avoid slow 100% CPU reports) Don't do this: (gdb) bt #0 0x00007f37ae667c40 in ?? () from /lib/x86_64-linux-gnu/libz.so.1 #1 0x00007f37ae668b38 in ?? () from /lib/x86_64-linux-gnu/libz.so.1 #2 0x00007f37ae669907 in deflate () from /lib/x86_64-linux-gnu/libz.so.1 #3 0x00007f37ae674c65 in compress2 () from /lib/x86_64-linux-gnu/libz.so.1 #4 0x000000000040cfe3 in zencode_scids (ctx=0xc1f118, scids=0x2599bc49 "\a\325{", len=176320) at gossipd/gossipd.c:218 #5 0x000000000040d0b3 in encode_short_channel_ids_end (encoded=0x7fff8f98d9f0, max_bytes=65490) at gossipd/gossipd.c:236 #6 0x000000000040dd28 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290511, number_of_blocks=8) at gossipd/gossipd.c:576 #7 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290511, number_of_blocks=16) at gossipd/gossipd.c:595 #8 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290495, number_of_blocks=32) at gossipd/gossipd.c:596 #9 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290495, number_of_blocks=64) at gossipd/gossipd.c:595 #10 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=128) at gossipd/gossipd.c:596 #11 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=256) at gossipd/gossipd.c:595 #12 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=512) at gossipd/gossipd.c:595 #13 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17290431, number_of_blocks=1024) at gossipd/gossipd.c:595 #14 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=2047) at gossipd/gossipd.c:596 #15 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=4095) at gossipd/gossipd.c:595 #16 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=8191) at gossipd/gossipd.c:595 #17 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=16382) at gossipd/gossipd.c:595 #18 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=32764) at gossipd/gossipd.c:595 #19 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=65528) at gossipd/gossipd.c:595 #20 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=131056) at gossipd/gossipd.c:595 #21 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=262112) at gossipd/gossipd.c:595 #22 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=524225) at gossipd/gossipd.c:595 #23 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=1048450) at gossipd/gossipd.c:595 #24 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=2096900) at gossipd/gossipd.c:595 #25 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=4193801) at gossipd/gossipd.c:595 #26 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=8387603) at gossipd/gossipd.c:595 #27 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=17289408, number_of_blocks=16775207) at gossipd/gossipd.c:595 #28 0x000000000040ddee in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=33550414) at gossipd/gossipd.c:596 #29 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=67100829) at gossipd/gossipd.c:595 #30 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=134201659) at gossipd/gossipd.c:595 #31 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=268403318) at gossipd/gossipd.c:595 #32 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=536806636) at gossipd/gossipd.c:595 #33 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=1073613273) at gossipd/gossipd.c:595 #34 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=2147226547) at gossipd/gossipd.c:595 #35 0x000000000040ddc6 in queue_channel_ranges (peer=0x3868fc8, first_blocknum=514201, number_of_blocks=4294453094) at gossipd/gossipd.c:595 #36 0x000000000040df26 in handle_query_channel_range (peer=0x3868fc8, msg=0x37e0678 "\001\ao\342\214\n\266\361\263r\301\246\242F\256c\367O\223\036\203e\341Z\b\234h\326\031") at gossipd/gossipd.c:625 The cause was that converting a block number to an scid truncates it at 24 bits. When we look through the index from (truncated number) to (real end number) we get every channel, which is too large to encode, so we iterate again. This fixes both that problem, and also the issue that we'd end up dividing into many empty sections until we get to the highest block number. Instead, we just tack the empty blocks on to then end of the final query. (My initial version requested 0xFFFFFFFE blocks, but the dev code which records what blocks were returned can't make a bitmap that big on 32 bit). Reported-by: George Vaccaro Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2019-01-15 10:37:57 +01:00
l1.daemon.wait_for_logs([r'\[IN\] 0108'] * 8)
# And no more!
time.sleep(1)
assert not l1.daemon.is_in_log(r'\[IN\] 0108', start=l1.daemon.logsearch_start)
# This should actually be large enough for zlib to kick in!
l3.fund_channel(l4, 10**5)
bitcoind.generate_block(5)
l2.daemon.wait_for_log('Received node_announcement for node ' + l4.info['id'])
# Restore infinite encode size.
l2.rpc.dev_set_max_scids_encode_size(max=(2**32 - 1))
l2.daemon.wait_for_log('Set max_scids_encode_bytes to {}'
.format(2**32 - 1))
ret = l1.rpc.dev_query_channel_range(id=l2.info['id'],
first=0,
num=65535)
l1.daemon.wait_for_log(
# WIRE_REPLY_CHANNEL_RANGE
r'\[IN\] 0108'
# chain_hash
+ '................................................................'
# first_blocknum
+ '00000000'
# number_of_blocks
+ '0000ffff'
# complete
+ '01'
# length
+ '....'
# encoding
+ '01'
)
# Long test involving 4 lightningd instances.
@unittest.skipIf(not DEVELOPER, "needs DEVELOPER=1")
def test_report_routing_failure(node_factory, bitcoind):
"""Test routing failure and retrying of routing.
"""
# The setup is as follows:
# l3-->l4
# ^ / |
# | / |
# | L v
# l2<--l1
#
# l1 wants to pay to l4.
# The shortest route is l1-l4, but l1 cannot
# afford to pay to l1 because l4 has all the
# funds.
# This is a local failure.
# The next shortest route is l1-l2-l4, but
# l2 cannot afford to pay l4 for same reason.
# This is a remote failure.
# Finally the only possible path is
# l1-l2-l3-l4.
def fund_from_to_payer(lsrc, ldst, lpayer):
lsrc.rpc.connect(ldst.info['id'], 'localhost', ldst.port)
c = lsrc.fund_channel(ldst, 10000000)
bitcoind.generate_block(5)
lpayer.wait_for_channel_updates([c])
# Setup
# Construct lightningd
l1, l2, l3, l4 = node_factory.get_nodes(4)
# Wire them up
# The ordering below matters!
# Particularly, l1 is payer and we will
# wait for l1 to receive gossip for the
# channel being made.
channels = []
for src, dst in [(l1, l2), (l2, l3), (l3, l4), (l4, l1), (l4, l2)]:
src.rpc.connect(dst.info['id'], 'localhost', dst.port)
channels.append(src.fund_channel(dst, 10**6))
bitcoind.generate_block(5)
for c in channels:
l1.wait_channel_active(c)
# Test
inv = l4.rpc.invoice(1234567, 'inv', 'for testing')['bolt11']
l1.rpc.pay(inv)
@unittest.skipIf(not DEVELOPER, "needs DEVELOPER=1")
def test_query_short_channel_id(node_factory, bitcoind):
l1 = node_factory.get_node(options={'log-level': 'io'})
l2 = node_factory.get_node()
l3 = node_factory.get_node()
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l2.rpc.connect(l3.info['id'], 'localhost', l3.port)
openingd: take peer before we're opening, wait for explicit funding msg. Prior to this, lightningd would hand uninteresting peers back to connectd, which would then return it to lightningd if it sent a non-gossip msg, or if lightningd asked it to release the peer. Now connectd hands the peer to lightningd once we've done the init handshake, which hands it off to openingd. This is a deep structural change, so we do the minimum here and cleanup in the following patches. Lightningd: 1. Remove peer_nongossip handling from connect_control and peer_control. 2. Remove list of outstanding fundchannel command; it was only needed to find the race between us asking connectd to release the peer and it reconnecting. 3. We can no longer tell if the remote end has started trying to fund a channel (until it has succeeded): it's very transitory anyway so not worth fixing. 4. We now always have a struct peer, and allocate an uncommitted_channel for it, though it may never be used if neither end funds a channel. 5. We start funding on messages for openingd: we can get a funder_reply or a fundee, or an error in response to our request to fund a channel. so we handle all of them. 6. A new peer_start_openingd() is called after connectd hands us a peer. 7. json_fund_channel just looks through local peers; there are none hidden in connectd any more. 8. We sometimes start a new openingd just to send an error message. Openingd: 1. We always have information we need to accept them funding a channel (in the init message). 2. We have to listen for three fds: peer, gossip and master, so we opencode the poll. 3. We have an explicit message to start trying to fund a channel. 4. We can be told to send a message in our init message. Testing: 1. We don't handle some things gracefully yet, so two tests are disabled. 2. 'hand_back_peer .*: now local again' from connectd is no longer a message, openingd says 'Handed peer, entering loop' once its managing it. 3. peer['state'] used to be set to 'GOSSIPING' (otherwise this field doesn't exist; 'state' is now per-channel. It doesn't exist at all now. 4. Some tests now need to turn on IO logging in openingd, not connectd. 5. There's a gap between connecting on one node and having connectd on the peer hand over the connection to openingd. Our tests sometimes checked getpeers() on the peer, and didn't see anything, so line_graph needed updating. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-08-08 16:10:58 +02:00
# Need full IO logging so we can see gossip (from openingd and channeld)
l1.daemon.wait_for_log('openingd-.*: Handed peer, entering loop')
subprocess.run(['kill', '-USR1', l1.subd_pid('openingd')])
# Empty result tests.
reply = l1.rpc.dev_query_scids(l2.info['id'], ['1x1x1', '2x2x2'])
# 0x0105 = query_short_channel_ids
l1.daemon.wait_for_log(r'\[OUT\] 0105.*0000000100000100010000020000020002')
assert reply['complete']
# Make channels public.
scid12 = l1.fund_channel(l2, 10**5)
scid23 = l2.fund_channel(l3, 10**5)
bitcoind.generate_block(5)
# It will know about everything.
l1.daemon.wait_for_log('Received node_announcement for node {}'.format(l3.info['id']))
subprocess.run(['kill', '-USR1', l1.subd_pid('channeld')])
# This query should get channel announcements, channel updates, and node announcements.
reply = l1.rpc.dev_query_scids(l2.info['id'], [scid23])
# 0x0105 = query_short_channel_ids
l1.daemon.wait_for_log(r'\[OUT\] 0105')
assert reply['complete']
# 0x0100 = channel_announcement
l1.daemon.wait_for_log(r'\[IN\] 0100')
# 0x0102 = channel_update
l1.daemon.wait_for_log(r'\[IN\] 0102')
l1.daemon.wait_for_log(r'\[IN\] 0102')
# 0x0101 = node_announcement
l1.daemon.wait_for_log(r'\[IN\] 0101')
l1.daemon.wait_for_log(r'\[IN\] 0101')
reply = l1.rpc.dev_query_scids(l2.info['id'], [scid12, scid23])
assert reply['complete']
# Technically, this order could be different, but this matches code.
# 0x0100 = channel_announcement
l1.daemon.wait_for_log(r'\[IN\] 0100')
# 0x0102 = channel_update
l1.daemon.wait_for_log(r'\[IN\] 0102')
l1.daemon.wait_for_log(r'\[IN\] 0102')
# 0x0100 = channel_announcement
l1.daemon.wait_for_log(r'\[IN\] 0100')
# 0x0102 = channel_update
l1.daemon.wait_for_log(r'\[IN\] 0102')
l1.daemon.wait_for_log(r'\[IN\] 0102')
# 0x0101 = node_announcement
l1.daemon.wait_for_log(r'\[IN\] 0101')
l1.daemon.wait_for_log(r'\[IN\] 0101')
def test_gossip_addresses(node_factory, bitcoind):
l1 = node_factory.get_node(options={'announce-addr': [
'[::]:3',
'127.0.0.1:2',
'vww6ybal4bd7szmgncyruucpgfkqahzddi37ktceo3ah7ngmcopnpyyd.onion',
'3fyb44wdhnd2ghhl.onion:1234'
]})
l2 = node_factory.get_node()
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l1.fund_channel(l2, 100000)
bitcoind.generate_block(6)
l2.daemon.wait_for_log('Received node_announcement for node {}'
.format(l1.info['id']))
nodes = l2.rpc.listnodes(l1.info['id'])['nodes']
assert len(nodes) == 1 and nodes[0]['addresses'] == [
{'type': 'ipv4', 'address': '127.0.0.1', 'port': 2},
{'type': 'ipv6', 'address': '::', 'port': 3},
{'type': 'torv2', 'address': '3fyb44wdhnd2ghhl.onion', 'port': 1234},
{'type': 'torv3', 'address': 'vww6ybal4bd7szmgncyruucpgfkqahzddi37ktceo3ah7ngmcopnpyyd.onion', 'port': 9735}
]
def test_gossip_store_load(node_factory):
"""Make sure we can read canned gossip store"""
l1 = node_factory.get_node(start=False)
with open(os.path.join(l1.daemon.lightning_dir, 'gossip_store'), 'wb') as f:
f.write(bytearray.fromhex("07" # GOSSIP_STORE_VERSION
"000001b0" # len
"fea676e8" # csum
"5b8d9b44" # timestamp
"0100" # WIRE_CHANNEL_ANNOUNCEMENT
"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"
"0000000a" # len
"99dc98b4" # csum
"00000000" # timestamp
"1005" # WIRE_GOSSIP_STORE_CHANNEL_AMOUNT
"0000000001000000"
"00000082" # len
"fd421aeb" # csum
"5b8d9b44" # timestamp
"0102" # WIRE_CHANNEL_UPDATE
"1ea7c2eadf8a29eb8690511a519b5656e29aa0a853771c4e38e65c5abf43d907295a915e69e451f4c7a0c3dc13dd943cfbe3ae88c0b96667cd7d58955dbfedcf43497fd7f826957108f4a30fd9cec3aeba79972084e90ead01ea33090000000013a63c0000b500015b8d9b440000009000000000000003e8000003e800000001"
"00000095" # len
"f036515e" # csum
"5aab817c" # timestamp
"0101" # WIRE_NODE_ANNOUNCEMENT
"cf5d870bc7ecabcb7cd16898ef66891e5f0c6c5851bd85b670f03d325bc44d7544d367cd852e18ec03f7f4ff369b06860a3b12b07b29f36fb318ca11348bf8ec00005aab817c03f113414ebdc6c1fb0f33c99cd5a1d09dd79e7fdf2468cf1fe1af6674361695d23974b250757a7a6c6549544300000000000000000000000000000000000000000000000007010566933e2607"))
l1.start()
# May preceed the Started msg waited for in 'start'.
wait_for(lambda: l1.daemon.is_in_log(r'gossip_store: Read 1/1/1/0 cannounce/cupdate/nannounce/cdelete from store \(0 deleted\) in 770 bytes'))
assert not l1.daemon.is_in_log('gossip_store.*truncating')
@unittest.skipIf(not DEVELOPER, "Needs fast gossip propagation")
def test_node_reannounce(node_factory, bitcoind):
"Test that we reannounce a node when parameters change"
l1, l2 = node_factory.line_graph(2, opts={'may_reconnect': True,
'log_all_io': True})
bitcoind.generate_block(5)
# Wait for node_announcement for l1.
l2.daemon.wait_for_log(r'\[IN\] 0101.*{}'.format(l1.info['id']))
# Wait for it to process it.
wait_for(lambda: l2.rpc.listnodes(l1.info['id'])['nodes'] != [])
wait_for(lambda: 'alias' in only_one(l2.rpc.listnodes(l1.info['id'])['nodes']))
assert only_one(l2.rpc.listnodes(l1.info['id'])['nodes'])['alias'].startswith('JUNIORBEAM')
l1.stop()
l1.daemon.opts['alias'] = 'SENIORBEAM'
l1.start()
# Wait for l1 to send us its own node_announcement.
nannouncement = l2.daemon.wait_for_log(r'{}.*\[IN\] 0101.*{}'.format(l1.info['id'], l1.info['id'])).split('[IN] ')[1]
wait_for(lambda: only_one(l2.rpc.listnodes(l1.info['id'])['nodes'])['alias'] == 'SENIORBEAM')
# Restart should re-xmit exact same update on reconnect.
l1.restart()
# l1 should retransmit it exactly the same (no timestamp change!)
l2.daemon.wait_for_log(r'{}.*\[IN\] {}'.format(l1.info['id'], nannouncement))
def test_gossipwith(node_factory):
l1, l2 = node_factory.line_graph(2, wait_for_announce=True)
out = subprocess.run(['devtools/gossipwith',
'--initial-sync',
'--max-messages=5',
'{}@localhost:{}'.format(l1.info['id'], l1.port)],
check=True,
timeout=TIMEOUT, stdout=subprocess.PIPE).stdout
num_msgs = 0
while len(out):
l, t = struct.unpack('>HH', out[0:4])
# channel_announcement node_announcement or channel_update
assert t == 256 or t == 257 or t == 258
out = out[2 + l:]
num_msgs += 1
# one channel announcement, two channel_updates, two node announcements.
assert num_msgs == 5
def test_gossip_notices_close(node_factory, bitcoind):
# We want IO logging so we can replay a channel_announce to l1.
l1 = node_factory.get_node(options={'log-level': 'io'})
l2, l3 = node_factory.line_graph(2)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
# FIXME: sending SIGUSR1 immediately may kill it before handler installed.
l1.daemon.wait_for_log('Handed peer, entering loop')
subprocess.run(['kill', '-USR1', l1.subd_pid('openingd')])
bitcoind.generate_block(5)
# Make sure l1 learns about channel and nodes.
wait_for(lambda: len(l1.rpc.listchannels()['channels']) == 2)
wait_for(lambda: ['alias' in n for n in l1.rpc.listnodes()['nodes']] == [True, True])
l1.rpc.disconnect(l2.info['id'])
# Grab channel_announcement from io logs (ends in ')
channel_announcement = l1.daemon.is_in_log(r'\[IN\] 0100').split(' ')[-1][:-1]
channel_update = l1.daemon.is_in_log(r'\[IN\] 0102').split(' ')[-1][:-1]
node_announcement = l1.daemon.is_in_log(r'\[IN\] 0101').split(' ')[-1][:-1]
txid = l2.rpc.close(l3.info['id'])['txid']
wait_for(lambda: only_one(l2.rpc.listpeers(l3.info['id'])['peers'])['channels'][0]['state'] == 'CLOSINGD_COMPLETE')
bitcoind.generate_block(1, txid)
wait_for(lambda: l1.rpc.listchannels()['channels'] == [])
wait_for(lambda: l1.rpc.listnodes()['nodes'] == [])
# FIXME: This is a hack: we should have a framework for canned conversations
# This doesn't naturally terminate, so we give it 5 seconds.
try:
subprocess.run(['devtools/gossipwith',
'{}@localhost:{}'.format(l1.info['id'], l1.port),
channel_announcement,
channel_update,
node_announcement],
timeout=5, stdout=subprocess.PIPE)
except subprocess.TimeoutExpired:
pass
# l1 should reject it.
assert(l1.rpc.listchannels()['channels'] == [])
assert(l1.rpc.listnodes()['nodes'] == [])
l1.stop()
l1.start()
assert(l1.rpc.listchannels()['channels'] == [])
assert(l1.rpc.listnodes()['nodes'] == [])
@unittest.skipIf(not DEVELOPER, "gossip propagation is slow without DEVELOPER=1")
def test_getroute_exclude(node_factory, bitcoind):
"""Test getroute's exclude argument"""
l1, l2, l3, l4 = node_factory.line_graph(4, wait_for_announce=True)
# This should work
route = l1.rpc.getroute(l4.info['id'], 1, 1)['route']
# l1 id is > l2 id, so 1 means l1->l2
chan_l1l2 = route[0]['channel'] + '/1'
chan_l2l1 = route[0]['channel'] + '/0'
# This should not
with pytest.raises(RpcError):
l1.rpc.getroute(l4.info['id'], 1, 1, exclude=[chan_l1l2])
# Blocking the wrong way should be fine.
l1.rpc.getroute(l4.info['id'], 1, 1, exclude=[chan_l2l1])
# Now, create an alternate (better) route.
l2.rpc.connect(l4.info['id'], 'localhost', l4.port)
scid = l2.fund_channel(l4, 1000000, wait_for_active=False)
bitcoind.generate_block(5)
# We don't wait above, because we care about it hitting l1.
l1.daemon.wait_for_logs([r'update for channel {}/0 now ACTIVE'
.format(scid),
r'update for channel {}/1 now ACTIVE'
.format(scid)])
# l3 id is > l2 id, so 1 means l3->l2
# chan_l3l2 = route[1]['channel'] + '/1'
chan_l2l3 = route[1]['channel'] + '/0'
# l4 is > l2
# chan_l4l2 = scid + '/1'
chan_l2l4 = scid + '/0'
# This works
l1.rpc.getroute(l4.info['id'], 1, 1, exclude=[chan_l2l3])
# This works
l1.rpc.getroute(l4.info['id'], 1, 1, exclude=[chan_l2l4])
# This doesn't
with pytest.raises(RpcError):
l1.rpc.getroute(l4.info['id'], 1, 1, exclude=[chan_l2l3, chan_l2l4])
@unittest.skipIf(not DEVELOPER, "need dev-compact-gossip-store")
def test_gossip_store_local_channels(node_factory, bitcoind):
l1, l2 = node_factory.line_graph(2, wait_for_announce=False)
# We see this channel, even though it's not announced, because it's local.
wait_for(lambda: len(l1.rpc.listchannels()['channels']) == 2)
l2.stop()
l1.restart()
# We should still see local channels!
time.sleep(3) # Make sure store is loaded
chans = l1.rpc.listchannels()['channels']
assert len(chans) == 2
# Now compact store
l1.rpc.call('dev-compact-gossip-store')
l1.restart()
time.sleep(3) # Make sure store is loaded
# We should still see local channels!
chans = l1.rpc.listchannels()['channels']
assert len(chans) == 2
@unittest.skipIf(not DEVELOPER, "need dev-compact-gossip-store")
def test_gossip_store_private_channels(node_factory, bitcoind):
l1, l2 = node_factory.line_graph(2, announce_channels=False)
# We see this channel, even though it's not announced, because it's local.
wait_for(lambda: len(l1.rpc.listchannels()['channels']) == 2)
l2.stop()
l1.restart()
# We should still see local channels!
time.sleep(3) # Make sure store is loaded
chans = l1.rpc.listchannels()['channels']
assert len(chans) == 2
# Now compact store
l1.rpc.call('dev-compact-gossip-store')
l1.restart()
time.sleep(3) # Make sure store is loaded
# We should still see local channels!
chans = l1.rpc.listchannels()['channels']
assert len(chans) == 2
def setup_gossip_store_test(node_factory, bitcoind):
l1, l2, l3 = node_factory.line_graph(3, fundchannel=False)
# Create channel.
scid23 = l2.fund_channel(l3, 10**6)
# Have that channel announced.
bitcoind.generate_block(5)
# Make sure we've got node_announcements
wait_for(lambda: ['alias' in n for n in l2.rpc.listnodes()['nodes']] == [True, True])
# Now, replace the one channel_update, so it's past the node announcements.
l2.rpc.setchannelfee(l3.info['id'], 20, 1000)
# Old base feerate is 1.
wait_for(lambda: sum([c['base_fee_millisatoshi'] for c in l2.rpc.listchannels()['channels']]) == 21)
# Create another channel, which will stay private.
scid12 = l1.fund_channel(l2, 10**6)
# Now insert channel_update for previous channel; now they're both past the
# node announcements.
l3.rpc.setchannelfee(l2.info['id'], 20, 1000)
wait_for(lambda: [c['base_fee_millisatoshi'] for c in l2.rpc.listchannels(scid23)['channels']] == [20, 20])
# Replace both (private) updates for scid12.
l1.rpc.setchannelfee(l2.info['id'], 20, 1000)
l2.rpc.setchannelfee(l1.info['id'], 20, 1000)
wait_for(lambda: [c['base_fee_millisatoshi'] for c in l2.rpc.listchannels(scid12)['channels']] == [20, 20])
# Active records in store now looks like:
# channel_announcement (scid23)
# channel_amount
# node_announcement
# node_announcement
# channel_update (scid23)
# local_add_channel (scid12)
# channel_update (scid23)
# private_channel_update (scid12)
# private_channel_update (scid12)
return l2
def test_gossip_store_load_complex(node_factory, bitcoind):
l2 = setup_gossip_store_test(node_factory, bitcoind)
l2.restart()
wait_for(lambda: l2.daemon.is_in_log('gossip_store: Read '))
@unittest.skipIf(not DEVELOPER, "need dev-compact-gossip-store")
def test_gossip_store_compact(node_factory, bitcoind):
l2 = setup_gossip_store_test(node_factory, bitcoind)
# Now compact store.
l2.rpc.call('dev-compact-gossip-store')
# Should still be connected.
time.sleep(1)
assert len(l2.rpc.listpeers()['peers']) == 2
# Should restart ok.
l2.restart()
wait_for(lambda: l2.daemon.is_in_log('gossip_store: Read '))
@unittest.skipIf(not DEVELOPER, "need dev-compact-gossip-store")
def test_gossip_store_compact_restart(node_factory, bitcoind):
l2 = setup_gossip_store_test(node_factory, bitcoind)
# Should restart ok.
l2.restart()
wait_for(lambda: l2.daemon.is_in_log('gossip_store: Read '))
# Now compact store.
l2.rpc.call('dev-compact-gossip-store')