mirrors/core-lightning
1
0
Fork 0
mirror of https://github.com/ElementsProject/lightning.git synced 2024-11-19 09:54:16 +01:00
Code Issues Projects Releases Wiki Activity
f004952442
core-lightning/tests/test_opening.py

2419 lines
96 KiB
Python
Raw Normal View History

df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
from fixtures import * # noqa: F401,F403
from fixtures import TEST_NETWORK
schema: add 'funding_outnum' to openchannel_update schema
2021-07-08 23:23:16 +02:00
from pyln.client import RpcError, Millisatoshi
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
from utils import (
pytest: adapt all the anchor-iff-EXPERIMENTAL tests to --experimental-anchors. We use parameterization here. The old `anchor_expected()` was for non-zero-fee anchors, and have bitrotted so there are some other changes as well. Unfortunately, all the anchor accounting seems to be broken, but I cannot understand these tests at all. I had to simply disable them for now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
only_one, wait_for, sync_blockheight, first_channel_id, calc_lease_fee, check_coin_moves
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
)
pytest: Test the `mindepth` customizations of `fundchannel` and hook Just test that we can customize, and we'll add mindepth=0 support in the next couple of commits.
2022-05-20 15:45:54 +02:00
from pathlib import Path
pytest: Add a test for direct payment over zeroconf channels
2022-05-25 16:14:47 +02:00
from pprint import pprint
df-rbf: reconnection tests (init_rbf + ack_rbf)
2021-02-10 22:17:02 +01:00
import pytest
df: more reconnection tests
2021-03-16 01:50:31 +01:00
import re
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
import unittest
def find_next_feerate(node, peer):
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
chan = only_one(node.rpc.listpeerchannels(peer.info['id'])['channels'])
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
return chan['next_feerate']
rpc: add queryrates Undocumented RPC call for asking a peer what their rates are
2021-06-11 23:32:27 +02:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
@pytest.mark.openchannel('v2')
def test_queryrates(node_factory, bitcoind):
dual-fund: remove anchor assumption for all dual-funded channels Only add the anchor channel_type if it's negotiated separately!
2022-10-19 20:35:45 +02:00
pytest: make dual-funding tests using option_anchors_zero_fee_htlc_tx. They used to force option_anchor_outputs, so switch them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
opts = {'dev-no-reconnect': None,
'experimental-anchors': None}
dual-fund: remove anchor assumption for all dual-funded channels Only add the anchor channel_type if it's negotiated separately!
2022-10-19 20:35:45 +02:00
l1, l2 = node_factory.get_nodes(2, opts=opts)
rpc: add queryrates Undocumented RPC call for asking a peer what their rates are
2021-06-11 23:32:27 +02:00
amount = 10 ** 6
l1.fundwallet(amount * 10)
l2.fundwallet(amount * 10)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
channel-lease: reject if we're not currently advertising liquidity If there's no plugin currently in place, we simply won't return any funding at all, in which case we'd expect them to handle however they want. (our implementation would fail the open, as we only accept opens that have at least as much as we've requested provided)
2021-07-06 19:17:51 +02:00
with pytest.raises(RpcError, match=r'not advertising liquidity'):
queryrates: make it dev-only Since we now use 'compact_lease' to gate an open (if the rates have changed, we fail), we no longer need to rely on query rates for figuring things out, so we make it dev-only. Changelog-Changed: JSON-API: queryrates is now developer only
2021-07-06 20:48:35 +02:00
l1.rpc.dev_queryrates(l2.info['id'], amount, amount * 10)
rpc: add queryrates Undocumented RPC call for asking a peer what their rates are
2021-06-11 23:32:27 +02:00
l2.rpc.call('funderupdate', {'policy': 'match',
'policy_mod': 100,
pytest: update test_queryrates for API change. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2021-07-25 12:36:18 +02:00
'per_channel_max_msat': '1btc',
rpc: add queryrates Undocumented RPC call for asking a peer what their rates are
2021-06-11 23:32:27 +02:00
'fuzz_percent': 0,
'lease_fee_base_msat': '2sat',
'funding_weight': 1000,
'lease_fee_basis': 140,
'channel_fee_max_base_msat': '3sat',
'channel_fee_max_proportional_thousandths': 101})
queryrates: make it dev-only Since we now use 'compact_lease' to gate an open (if the rates have changed, we fail), we no longer need to rely on query rates for figuring things out, so we make it dev-only. Changelog-Changed: JSON-API: queryrates is now developer only
2021-07-06 20:48:35 +02:00
result = l1.rpc.dev_queryrates(l2.info['id'], amount, amount)
rpc: add queryrates Undocumented RPC call for asking a peer what their rates are
2021-06-11 23:32:27 +02:00
assert result['our_funding_msat'] == Millisatoshi(amount * 1000)
assert result['their_funding_msat'] == Millisatoshi(amount * 1000)
assert result['funding_weight'] == 1000
assert result['lease_fee_base_msat'] == Millisatoshi(2000)
assert result['lease_fee_basis'] == 140
assert result['channel_fee_max_base_msat'] == Millisatoshi(3000)
assert result['channel_fee_max_proportional_thousandths'] == 101
dual-funding: convert to runtime flag, --experimental-dual-fund You can now activate dual-funded channels using the `--experimental-dual-fund` flag Changelog-Changed: Config: `--experimental-dual-fund` runtime flag will enable dual-funded protocol on this node
2021-03-12 01:19:40 +01:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v1') # Mixed v1 + v2, v2 manually turned on
mfc: check that we can retry when things go haywire There's a version of this that keeps the PSBT in memory and does some fancy addition/subtraction of unuseable parts for the v2's, however it's much easier and simpler to simply error on the peer and re-start from the very beginning. This only works if we haven't gotten commitments from the peer yet (in fact either method would only work if we haven't got commitments from the peer yet), so if we've got commitments from them we simply mark them as failed an go again. In a perfect world, we'd remember what inputs we used last time, and reuse those again on the re-attempt, which would pefectly guarantee both that the failed opens (ones w/ commitments exchanged) would be canceled after this completes (and we could re-try the failed again). As it is, this is not perfect. It is, however, servicable.
2021-03-11 03:23:03 +01:00
def test_multifunding_v2_best_effort(node_factory, bitcoind):
'''
Check that best_effort flag works.
'''
disconnects = ["-WIRE_INIT",
"-WIRE_ACCEPT_CHANNEL",
"-WIRE_FUNDING_SIGNED"]
dual-funding: convert to runtime flag, --experimental-dual-fund You can now activate dual-funded channels using the `--experimental-dual-fund` flag Changelog-Changed: Config: `--experimental-dual-fund` runtime flag will enable dual-funded protocol on this node
2021-03-12 01:19:40 +01:00
l1 = node_factory.get_node(options={'experimental-dual-fund': None},
mfc: check that we can retry when things go haywire There's a version of this that keeps the PSBT in memory and does some fancy addition/subtraction of unuseable parts for the v2's, however it's much easier and simpler to simply error on the peer and re-start from the very beginning. This only works if we haven't gotten commitments from the peer yet (in fact either method would only work if we haven't got commitments from the peer yet), so if we've got commitments from them we simply mark them as failed an go again. In a perfect world, we'd remember what inputs we used last time, and reuse those again on the re-attempt, which would pefectly guarantee both that the failed opens (ones w/ commitments exchanged) would be canceled after this completes (and we could re-try the failed again). As it is, this is not perfect. It is, however, servicable.
2021-03-11 03:23:03 +01:00
allow_warning=True,
may_reconnect=True)
dual-funding: convert to runtime flag, --experimental-dual-fund You can now activate dual-funded channels using the `--experimental-dual-fund` flag Changelog-Changed: Config: `--experimental-dual-fund` runtime flag will enable dual-funded protocol on this node
2021-03-12 01:19:40 +01:00
l2 = node_factory.get_node(options={'experimental-dual-fund': None},
mfc: check that we can retry when things go haywire There's a version of this that keeps the PSBT in memory and does some fancy addition/subtraction of unuseable parts for the v2's, however it's much easier and simpler to simply error on the peer and re-start from the very beginning. This only works if we haven't gotten commitments from the peer yet (in fact either method would only work if we haven't got commitments from the peer yet), so if we've got commitments from them we simply mark them as failed an go again. In a perfect world, we'd remember what inputs we used last time, and reuse those again on the re-attempt, which would pefectly guarantee both that the failed opens (ones w/ commitments exchanged) would be canceled after this completes (and we could re-try the failed again). As it is, this is not perfect. It is, however, servicable.
2021-03-11 03:23:03 +01:00
allow_warning=True,
may_reconnect=True)
l3 = node_factory.get_node(disconnect=disconnects)
l4 = node_factory.get_node()
l1.fundwallet(2000000)
destinations = [{"id": '{}@localhost:{}'.format(l2.info['id'], l2.port),
"amount": 50000},
{"id": '{}@localhost:{}'.format(l3.info['id'], l3.port),
"amount": 50000},
{"id": '{}@localhost:{}'.format(l4.info['id'], l4.port),
"amount": 50000}]
for i, d in enumerate(disconnects):
failed_sign = d == "-WIRE_FUNDING_SIGNED"
# Should succeed due to best-effort flag.
min_channels = 1 if failed_sign else 2
l1.rpc.multifundchannel(destinations, minchannels=min_channels)
bitcoind.generate_block(6, wait_for_mempool=1)
# l3 should fail to have channels; l2 also fails on last attempt
node_list = [l1, l4] if failed_sign else [l1, l2, l4]
for node in node_list:
node.daemon.wait_for_log(r'to CHANNELD_NORMAL')
# There should be working channels to l2 and l4 for every run
# but the last
working_chans = [l4] if failed_sign else [l2, l4]
for ldest in working_chans:
inv = ldest.rpc.invoice(5000, 'i{}'.format(i), 'i{}'.format(i))['bolt11']
l1.rpc.pay(inv)
# Function to find the SCID of the channel that is
# currently open.
# Cannot use LightningNode.get_channel_scid since
# it assumes the *first* channel found is the one
# wanted, but in our case we close channels and
# open again, so multiple channels may remain
# listed.
def get_funded_channel_scid(n1, n2):
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
channels = n1.rpc.listpeerchannels(n2.info['id'])['channels']
assert channels and len(channels) != 0
mfc: check that we can retry when things go haywire There's a version of this that keeps the PSBT in memory and does some fancy addition/subtraction of unuseable parts for the v2's, however it's much easier and simpler to simply error on the peer and re-start from the very beginning. This only works if we haven't gotten commitments from the peer yet (in fact either method would only work if we haven't got commitments from the peer yet), so if we've got commitments from them we simply mark them as failed an go again. In a perfect world, we'd remember what inputs we used last time, and reuse those again on the re-attempt, which would pefectly guarantee both that the failed opens (ones w/ commitments exchanged) would be canceled after this completes (and we could re-try the failed again). As it is, this is not perfect. It is, however, servicable.
2021-03-11 03:23:03 +01:00
for c in channels:
state = c['state']
if state in ('DUALOPEND_AWAITING_LOCKIN', 'CHANNELD_AWAITING_LOCKIN', 'CHANNELD_NORMAL'):
return c['short_channel_id']
assert False
# Now close channels to l2 and l4, for the next run.
if not failed_sign:
l1.rpc.close(get_funded_channel_scid(l1, l2))
l1.rpc.close(get_funded_channel_scid(l1, l4))
for node in node_list:
node.daemon.wait_for_log(r'to CLOSINGD_COMPLETE')
# With 2 down, it will fail to fund channel
l2.stop()
l3.stop()
On Mac OS X a different error is raised on failed connection [ Typo fixed --RR ]
2021-10-09 18:27:54 +02:00
with pytest.raises(RpcError, match=r'(Connection refused|Bad file descriptor)'):
mfc: check that we can retry when things go haywire There's a version of this that keeps the PSBT in memory and does some fancy addition/subtraction of unuseable parts for the v2's, however it's much easier and simpler to simply error on the peer and re-start from the very beginning. This only works if we haven't gotten commitments from the peer yet (in fact either method would only work if we haven't got commitments from the peer yet), so if we've got commitments from them we simply mark them as failed an go again. In a perfect world, we'd remember what inputs we used last time, and reuse those again on the re-attempt, which would pefectly guarantee both that the failed opens (ones w/ commitments exchanged) would be canceled after this completes (and we could re-try the failed again). As it is, this is not perfect. It is, however, servicable.
2021-03-11 03:23:03 +01:00
l1.rpc.multifundchannel(destinations, minchannels=2)
# This works though.
l1.rpc.multifundchannel(destinations, minchannels=1)
df: more reconnection tests
2021-03-16 01:50:31 +01:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v2')
df: more reconnection tests
2021-03-16 01:50:31 +01:00
def test_v2_open_sigs_restart(node_factory, bitcoind):
disconnects_1 = ['-WIRE_TX_SIGNATURES']
disconnects_2 = ['+WIRE_TX_SIGNATURES']
l1, l2 = node_factory.get_nodes(2,
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
opts=[{'disconnect': disconnects_1,
df: more reconnection tests
2021-03-16 01:50:31 +01:00
'may_reconnect': True},
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
{'disconnect': disconnects_2,
df: more reconnection tests
2021-03-16 01:50:31 +01:00
'may_reconnect': True}])
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
# Fund the channel, should appear to finish ok even though the
# peer fails
df-tests: make bad tests a bit better, maybe We updated the "UNKNOWN TYPE channel_id" -> the actual channel id; the reason for why the error fails shouldn't be restrictive (we just know that it fails)
2021-03-18 22:22:03 +01:00
with pytest.raises(RpcError):
df: more reconnection tests
2021-03-16 01:50:31 +01:00
l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = first_channel_id(l1, l2)
df-tests: make bad tests a bit better, maybe We updated the "UNKNOWN TYPE channel_id" -> the actual channel id; the reason for why the error fails shouldn't be restrictive (we just know that it fails)
2021-03-18 22:22:03 +01:00
log = l1.daemon.is_in_log('{} psbt'.format(chan_id))
df-tests: update to take 'no channel_id' into account; use exp-df flag The logs we're looking for don't exist anymore, we got rid of them.
2021-03-18 19:49:15 +01:00
assert log
df-tests: make bad tests a bit better, maybe We updated the "UNKNOWN TYPE channel_id" -> the actual channel id; the reason for why the error fails shouldn't be restrictive (we just know that it fails)
2021-03-18 22:22:03 +01:00
psbt = re.search("psbt (.*)", log).group(1)
lightningd: split DUALOPEND_OPEN_INIT into DUALOPEND_OPEN_INIT and DUALOPEND_OPEN_COMMITTED. The latter is used when we're put in the db, the former is the uncommitted state. Currently dbid == 0 is used in addition to the state, which is unwieldy. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Experimental: JSON-RPC: added new dual-funding state `DUALOPEND_OPEN_COMMITTED`
2023-10-02 00:59:51 +02:00
l1.daemon.wait_for_log('Peer has reconnected, state DUALOPEND_OPEN_COMMITTED')
flake: permit test_v2_open_sigs_restart_while_dead to succeed/fail There's a race btw disconnecting and returning a successful RPC call for openchannel_signed; if we disconnect quickly, we get an RPC error back. Too slow and it returns w/o an error. This needs to be cleaned up on a whole, work that I'm planning to get into as part of a funder re-write. For now, let's just let the test continue whether this call succeeds or fails.
2022-07-27 00:07:39 +02:00
try:
# FIXME: why do we need to retry signed?
df: more reconnection tests
2021-03-16 01:50:31 +01:00
l1.rpc.openchannel_signed(chan_id, psbt)
flake: permit test_v2_open_sigs_restart_while_dead to succeed/fail There's a race btw disconnecting and returning a successful RPC call for openchannel_signed; if we disconnect quickly, we get an RPC error back. Too slow and it returns w/o an error. This needs to be cleaned up on a whole, work that I'm planning to get into as part of a funder re-write. For now, let's just let the test continue whether this call succeeds or fails.
2022-07-27 00:07:39 +02:00
except RpcError:
pass
df: more reconnection tests
2021-03-16 01:50:31 +01:00
l2.daemon.wait_for_log('Broadcasting funding tx')
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
txid = l2.rpc.listpeerchannels(l1.info['id'])['channels'][0]['funding_txid']
test_v2_open_sigs_restart: dont mine blocks until funding tx in mempool This was flaky because sometimes we'd generate blocks before the funding transaction reached the mempool. If we wait until it's in, it works as expected. [ Neatened to use wait_for_mempool arg --RR ]
2021-03-22 21:46:08 +01:00
bitcoind.generate_block(6, wait_for_mempool=txid)
df: more reconnection tests
2021-03-16 01:50:31 +01:00
# Make sure we're ok.
l1.daemon.wait_for_log(r'to CHANNELD_NORMAL')
df-tests: update to take 'no channel_id' into account; use exp-df flag The logs we're looking for don't exist anymore, we got rid of them.
2021-03-18 19:49:15 +01:00
l2.daemon.wait_for_log(r'to CHANNELD_NORMAL')
df: more reconnection tests
2021-03-16 01:50:31 +01:00
v2 open tests: don't drop connection when an openchannel fails Dropping the connection is bad behavior on an openchannel failure, especially given that there might be other channels currently connected. We should maintain the connection but close out the dualopend daemon for that attempt. This test partially documents the behaivor, but fails Changelog-None
2022-11-23 01:06:55 +01:00
@pytest.mark.openchannel('v2')
def test_v2_fail_second(node_factory, bitcoind):
""" Open a channel succeeds; opening a second channel
failure should not drop the connection """
l1, l2 = node_factory.line_graph(2, wait_for_announce=True)
# Should have one channel between them.
pytest: update test_v2_fail_second for listpeers changes. Now should use listpeerchannels, since listpeers['channels'] is deprecated. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-05-23 12:28:25 +02:00
only_one(l1.rpc.listpeerchannels(l2.info['id'])['channels'])
v2 open tests: don't drop connection when an openchannel fails Dropping the connection is bad behavior on an openchannel failure, especially given that there might be other channels currently connected. We should maintain the connection but close out the dualopend daemon for that attempt. This test partially documents the behaivor, but fails Changelog-None
2022-11-23 01:06:55 +01:00
amount = 2**24 - 1
l1.fundwallet(amount + 10000000)
# make sure we can generate PSBTs.
addr = l1.rpc.newaddr()['bech32']
bitcoind.rpc.sendtoaddress(addr, (amount + 1000000) / 10**8)
bitcoind.generate_block(1)
wait_for(lambda: len(l1.rpc.listfunds()["outputs"]) != 0)
# Some random (valid) psbt
psbt = l1.rpc.fundpsbt(amount, '253perkw', 250, reserve=0)['psbt']
start = l1.rpc.openchannel_init(l2.info['id'], amount, psbt)
pytest: update test_v2_fail_second for listpeers changes. Now should use listpeerchannels, since listpeers['channels'] is deprecated. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-05-23 12:28:25 +02:00
# They will both see a pair of channels
assert len(l1.rpc.listpeerchannels(l2.info['id'])['channels']) == 2
assert len(l2.rpc.listpeerchannels(l1.info['id'])['channels']) == 2
v2 open tests: don't drop connection when an openchannel fails Dropping the connection is bad behavior on an openchannel failure, especially given that there might be other channels currently connected. We should maintain the connection but close out the dualopend daemon for that attempt. This test partially documents the behaivor, but fails Changelog-None
2022-11-23 01:06:55 +01:00
# We can abort a channel
l1.rpc.openchannel_abort(start['channel_id'])
# We should have deleted the 'in-progress' channel info
pytest: update test_v2_fail_second for listpeers changes. Now should use listpeerchannels, since listpeers['channels'] is deprecated. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-05-23 12:28:25 +02:00
only_one(l1.rpc.listpeerchannels(l2.info['id'])['channels'])
v2 open tests: don't drop connection when an openchannel fails Dropping the connection is bad behavior on an openchannel failure, especially given that there might be other channels currently connected. We should maintain the connection but close out the dualopend daemon for that attempt. This test partially documents the behaivor, but fails Changelog-None
2022-11-23 01:06:55 +01:00
# FIXME: check that tx-abort was sent
# Should be able to reattempt without reconnecting
start = l1.rpc.openchannel_init(l2.info['id'], amount, psbt)
pytest: update test_v2_fail_second for listpeers changes. Now should use listpeerchannels, since listpeers['channels'] is deprecated. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-05-23 12:28:25 +02:00
assert len(l1.rpc.listpeerchannels(l2.info['id'])['channels']) == 2
v2 open tests: don't drop connection when an openchannel fails Dropping the connection is bad behavior on an openchannel failure, especially given that there might be other channels currently connected. We should maintain the connection but close out the dualopend daemon for that attempt. This test partially documents the behaivor, but fails Changelog-None
2022-11-23 01:06:55 +01:00
df: more reconnection tests
2021-03-16 01:50:31 +01:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v2')
df: more reconnection tests
2021-03-16 01:50:31 +01:00
def test_v2_open_sigs_restart_while_dead(node_factory, bitcoind):
# Same thing as above, except the transaction mines
# while we're asleep
disconnects_1 = ['-WIRE_TX_SIGNATURES']
disconnects_2 = ['+WIRE_TX_SIGNATURES']
l1, l2 = node_factory.get_nodes(2,
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
opts=[{'disconnect': disconnects_1,
df-tests: update to take 'no channel_id' into account; use exp-df flag The logs we're looking for don't exist anymore, we got rid of them.
2021-03-18 19:49:15 +01:00
'may_reconnect': True,
'may_fail': True},
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
{'disconnect': disconnects_2,
df-tests: update to take 'no channel_id' into account; use exp-df flag The logs we're looking for don't exist anymore, we got rid of them.
2021-03-18 19:49:15 +01:00
'may_reconnect': True,
'may_fail': True}])
df: more reconnection tests
2021-03-16 01:50:31 +01:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
# Fund the channel, should appear to finish ok even though the
# peer fails
df-tests: make bad tests a bit better, maybe We updated the "UNKNOWN TYPE channel_id" -> the actual channel id; the reason for why the error fails shouldn't be restrictive (we just know that it fails)
2021-03-18 22:22:03 +01:00
with pytest.raises(RpcError):
df: more reconnection tests
2021-03-16 01:50:31 +01:00
l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = first_channel_id(l1, l2)
df-tests: make bad tests a bit better, maybe We updated the "UNKNOWN TYPE channel_id" -> the actual channel id; the reason for why the error fails shouldn't be restrictive (we just know that it fails)
2021-03-18 22:22:03 +01:00
log = l1.daemon.is_in_log('{} psbt'.format(chan_id))
df-tests: update to take 'no channel_id' into account; use exp-df flag The logs we're looking for don't exist anymore, we got rid of them.
2021-03-18 19:49:15 +01:00
assert log
df-tests: make bad tests a bit better, maybe We updated the "UNKNOWN TYPE channel_id" -> the actual channel id; the reason for why the error fails shouldn't be restrictive (we just know that it fails)
2021-03-18 22:22:03 +01:00
psbt = re.search("psbt (.*)", log).group(1)
lightningd: split DUALOPEND_OPEN_INIT into DUALOPEND_OPEN_INIT and DUALOPEND_OPEN_COMMITTED. The latter is used when we're put in the db, the former is the uncommitted state. Currently dbid == 0 is used in addition to the state, which is unwieldy. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Experimental: JSON-RPC: added new dual-funding state `DUALOPEND_OPEN_COMMITTED`
2023-10-02 00:59:51 +02:00
l1.daemon.wait_for_log('Peer has reconnected, state DUALOPEND_OPEN_COMMITTED')
flake: permit test_v2_open_sigs_restart_while_dead to succeed/fail There's a race btw disconnecting and returning a successful RPC call for openchannel_signed; if we disconnect quickly, we get an RPC error back. Too slow and it returns w/o an error. This needs to be cleaned up on a whole, work that I'm planning to get into as part of a funder re-write. For now, let's just let the test continue whether this call succeeds or fails.
2022-07-27 00:07:39 +02:00
try:
# FIXME: why do we need to retry signed?
df: more reconnection tests
2021-03-16 01:50:31 +01:00
l1.rpc.openchannel_signed(chan_id, psbt)
flake: permit test_v2_open_sigs_restart_while_dead to succeed/fail There's a race btw disconnecting and returning a successful RPC call for openchannel_signed; if we disconnect quickly, we get an RPC error back. Too slow and it returns w/o an error. This needs to be cleaned up on a whole, work that I'm planning to get into as part of a funder re-write. For now, let's just let the test continue whether this call succeeds or fails.
2022-07-27 00:07:39 +02:00
except RpcError:
pass
df: more reconnection tests
2021-03-16 01:50:31 +01:00
l2.daemon.wait_for_log('Broadcasting funding tx')
pytest: fix flake in test_v2_open_sigs_restart_while_dead We didn't wait for the sendtx to complete before shutting down, and hence timed out: ``` l2.daemon.wait_for_log('Broadcasting funding tx') l1.stop() l2.stop() bitcoind.generate_block(6) l1.restart() l2.restart() # Make sure we're ok. > l2.daemon.wait_for_log(r'to CHANNELD_NORMAL') ``` Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2021-04-08 02:58:50 +02:00
l2.daemon.wait_for_log('sendrawtx exit 0')
df: more reconnection tests
2021-03-16 01:50:31 +01:00
l1.stop()
l2.stop()
bitcoind.generate_block(6)
l1.restart()
l2.restart()
# Make sure we're ok.
l2.daemon.wait_for_log(r'to CHANNELD_NORMAL')
l1.daemon.wait_for_log(r'to CHANNELD_NORMAL')
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v2')
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
def test_v2_rbf_single(node_factory, bitcoind, chainparams):
lightningd: wumbo is now the default, setting has no effect. "Patrick, I'm sorry I doubted you." Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Changed: Config: `large-channels` is now the default, wumbology for all.
2023-10-16 02:36:25 +02:00
l1, l2 = node_factory.get_nodes(2)
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
res = l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
assert(only_one(vins))
prev_utxos = ["{}:{}".format(vins[0]['txid'], vins[0]['vout'])]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
next_feerate = find_next_feerate(l1, l2)
# Check that feerate info is correct
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
info_1 = only_one(l1.rpc.listpeerchannels(l2.info['id'])['channels'])
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
assert info_1['initial_feerate'] == info_1['last_feerate']
rate = int(info_1['last_feerate'][:-5])
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
assert int(info_1['next_feerate'][:-5]) == rate * 65 // 64
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
# Initiate an RBF
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(chan_amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
# Do the bump
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
assert update['commitments_secured']
# Check that feerate info has incremented
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
info_2 = only_one(l1.rpc.listpeerchannels(l2.info['id'])['channels'])
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
assert info_1['initial_feerate'] == info_2['initial_feerate']
assert info_1['next_feerate'] == info_2['last_feerate']
rate = int(info_2['last_feerate'][:-5])
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
assert int(info_2['next_feerate'][:-5]) == rate * 65 // 64
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
# Sign our inputs, and continue
signed_psbt = l1.rpc.signpsbt(update['psbt'])['signed_psbt']
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
# Fails because we didn't put enough feerate in.
with pytest.raises(RpcError, match=r'insufficient fee'):
l1.rpc.openchannel_signed(chan_id, signed_psbt)
# Do it again, with a higher feerate
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
info_2 = only_one(l1.rpc.listpeerchannels(l2.info['id'])['channels'])
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
assert info_1['initial_feerate'] == info_2['initial_feerate']
assert info_1['next_feerate'] == info_2['last_feerate']
rate = int(info_2['last_feerate'][:-5])
assert int(info_2['next_feerate'][:-5]) == rate * 65 // 64
# We 4x the feerate to beat the min-relay fee
next_rate = '{}perkw'.format(rate * 65 // 64 * 4)
# Gotta unreserve the psbt and re-reserve with higher feerate
l1.rpc.unreserveinputs(initpsbt['psbt'])
initpsbt = l1.rpc.utxopsbt(chan_amount, next_rate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
# Do the bump+sign
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'],
funding_feerate=next_rate)
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
assert update['commitments_secured']
signed_psbt = l1.rpc.signpsbt(update['psbt'])['signed_psbt']
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
l1.rpc.openchannel_signed(chan_id, signed_psbt)
bitcoind.generate_block(1)
sync_blockheight(bitcoind, [l1])
l1.daemon.wait_for_log(' to CHANNELD_NORMAL')
# Check that feerate info is gone
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
info_1 = only_one(l1.rpc.listpeerchannels(l2.info['id'])['channels'])
df-tests: happy path rbf test I can has RBF?
2021-02-04 20:17:53 +01:00
assert 'initial_feerate' not in info_1
assert 'last_feerate' not in info_1
assert 'next_feerate' not in info_1
# Shut l2 down, force close the channel.
l2.stop()
resp = l1.rpc.close(l2.info['id'], unilateraltimeout=1)
assert resp['type'] == 'unilateral'
l1.daemon.wait_for_log(' to CHANNELD_SHUTTING_DOWN')
l1.daemon.wait_for_log('sendrawtx exit 0')
df-rbf: reconnection tests (init_rbf + ack_rbf)
2021-02-10 22:17:02 +01:00
dual-fund tests: add tests for reported incompat with Eclair Bug Report: - initiate a channel open eclair -> cln - wait for the transaction to be published - eclair initiates rbf, and cancels it by sending tx_abort before exchanging commit_sig - at that point everything looks good, cln echoes the tx_abort and stays connected - eclair initiates another RBF attempt and sends tx_init_rbf: for some unknown reason, cln answers with channel_reestablish (??) followed by an error saying "Bad reestablish message: WIRE_TX_INIT_RBF" Diagnosis: CLN is doing a reconnect after a tx-abort is sent. Extra Test: Realized that if we abort, we won't correctly advanced to NORMAL if blocks are mined while we're in hanging state. CLN should advance after block containing channel open is mined. Reported-By: @t-bast
2023-07-29 22:48:43 +02:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
@pytest.mark.openchannel('v2')
def test_v2_rbf_abort_retry(node_factory, bitcoind, chainparams):
lightningd: wumbo is now the default, setting has no effect. "Patrick, I'm sorry I doubted you." Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Changed: Config: `large-channels` is now the default, wumbology for all.
2023-10-16 02:36:25 +02:00
l1, l2 = node_factory.get_nodes(2, opts={'allow_warning': True})
dual-fund tests: add tests for reported incompat with Eclair Bug Report: - initiate a channel open eclair -> cln - wait for the transaction to be published - eclair initiates rbf, and cancels it by sending tx_abort before exchanging commit_sig - at that point everything looks good, cln echoes the tx_abort and stays connected - eclair initiates another RBF attempt and sends tx_init_rbf: for some unknown reason, cln answers with channel_reestablish (??) followed by an error saying "Bad reestablish message: WIRE_TX_INIT_RBF" Diagnosis: CLN is doing a reconnect after a tx-abort is sent. Extra Test: Realized that if we abort, we won't correctly advanced to NORMAL if blocks are mined while we're in hanging state. CLN should advance after block containing channel open is mined. Reported-By: @t-bast
2023-07-29 22:48:43 +02:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendmany("",
{l1.rpc.newaddr()['bech32']: amount / 10**8 + 0.01,
l2.rpc.newaddr()['bech32']: amount / 10**8 + 0.01})
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
wait_for(lambda: len(l2.rpc.listfunds()['outputs']) > 0)
l1_utxos = ['{}:{}'.format(utxo['txid'], utxo['output']) for utxo in l1.rpc.listfunds()['outputs']]
# setup l2 to dual-fund!
l2.rpc.call('funderupdate', {'policy': 'match',
'policy_mod': 100,
'leases_only': False})
res = l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
prev_utxos = ["{}:{}".format(vin['txid'], vin['vout']) for vin in vins if "{}:{}".format(vin['txid'], vin['vout']) in l1_utxos]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
# Check that feerate info is correct
info_1 = only_one(l1.rpc.listpeerchannels(l2.info['id'])['channels'])
next_rate = "{}perkw".format(info_1['next_feerate'][:-5])
# Initiate an RBF
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(chan_amount, next_rate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
# Do the bump
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
# We abort the channel mid-way throught the RBF
l1.rpc.openchannel_abort(chan_id)
with pytest.raises(RpcError):
l1.rpc.openchannel_update(chan_id, bump['psbt'])
# - initiate a channel open eclair -> cln
# - wait for the transaction to be published
# - eclair initiates rbf, and cancels it by sending tx_abort before exchanging commit_sig
# - at that point everything looks good, cln echoes the tx_abort and stays connected
# - eclair initiates another RBF attempt and sends tx_init_rbf: for some unknown reason, cln answers with channel_reestablish (??) followed by an error saying "Bad reestablish message: WIRE_TX_INIT_RBF"
# attempt to initiate an RBF again
info_1 = only_one(l1.rpc.listpeerchannels(l2.info['id'])['channels'])
next_rate = "{}perkw".format(int(info_1['next_feerate'][:-5]) * 2)
# Gotta unreserve the psbt and re-reserve with higher feerate
l1.rpc.unreserveinputs(initpsbt['psbt'])
initpsbt = l1.rpc.utxopsbt(chan_amount, next_rate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
# Do the bump+sign
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'],
funding_feerate=next_rate)
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
signed_psbt = l1.rpc.signpsbt(update['psbt'])['signed_psbt']
l1.rpc.openchannel_signed(chan_id, signed_psbt)
bitcoind.generate_block(1)
sync_blockheight(bitcoind, [l1])
l1.daemon.wait_for_log(' to CHANNELD_NORMAL')
assert not l1.daemon.is_in_log('WIRE_CHANNEL_REESTABLISH')
assert not l2.daemon.is_in_log('WIRE_CHANNEL_REESTABLISH')
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
@pytest.mark.openchannel('v2')
def test_v2_rbf_abort_channel_opens(node_factory, bitcoind, chainparams):
l1, l2 = node_factory.get_nodes(2, opts={'wumbo': None,
'allow_warning': True})
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendmany("",
{l1.rpc.newaddr()['bech32']: amount / 10**8 + 0.01,
l2.rpc.newaddr()['bech32']: amount / 10**8 + 0.01})
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
wait_for(lambda: len(l2.rpc.listfunds()['outputs']) > 0)
l1_utxos = ['{}:{}'.format(utxo['txid'], utxo['output']) for utxo in l1.rpc.listfunds()['outputs']]
# setup l2 to dual-fund!
l2.rpc.call('funderupdate', {'policy': 'match',
'policy_mod': 100,
'leases_only': False})
res = l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
prev_utxos = ["{}:{}".format(vin['txid'], vin['vout']) for vin in vins if "{}:{}".format(vin['txid'], vin['vout']) in l1_utxos]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
# Check that feerate info is correct
info_1 = only_one(l1.rpc.listpeerchannels(l2.info['id'])['channels'])
next_rate = "{}perkw".format(info_1['next_feerate'][:-5])
# Initiate an RBF
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(chan_amount, next_rate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
# Do the bump
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
# We abort the channel mid-way throught the RBF
l1.rpc.openchannel_abort(chan_id)
with pytest.raises(RpcError):
l1.rpc.openchannel_update(chan_id, bump['psbt'])
# When the original open tx is mined, we should still arrive at
# NORMAL channel ops
bitcoind.generate_block(1)
sync_blockheight(bitcoind, [l1])
l1.daemon.wait_for_log(' to CHANNELD_NORMAL')
tests: check that we don't re-up the leased funds when we RBF Check that the peer won't put funds into a RBF'd channel lease. FIXME: allow leases to pass through to RBFs Changelog-None.
2021-08-05 19:32:58 +02:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
@pytest.mark.openchannel('v2')
def test_v2_rbf_liquidity_ad(node_factory, bitcoind, chainparams):
opts = {'funder-policy': 'match', 'funder-policy-mod': 100,
plugin-funder: fix typos in option lease-fee-base-msat and funder-fuzz-percent Changelog-Experimental: option `--lease-fee-base-msat` renamed to `--lease-fee-base-sat` Changelog-Experimental: option `--lease-fee-base-msat` deprecated and will be removed next release
2022-01-17 14:35:40 +01:00
'lease-fee-base-sat': '100sat', 'lease-fee-basis': 100,
pytest: make dual-funding tests using option_anchors_zero_fee_htlc_tx. They used to force option_anchor_outputs, so switch them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
'experimental-anchors': None,
tests: check that we don't re-up the leased funds when we RBF Check that the peer won't put funds into a RBF'd channel lease. FIXME: allow leases to pass through to RBFs Changelog-None.
2021-08-05 19:32:58 +02:00
'may_reconnect': True}
dual-fund: remove anchor assumption for all dual-funded channels Only add the anchor channel_type if it's negotiated separately!
2022-10-19 20:35:45 +02:00
tests: check that we don't re-up the leased funds when we RBF Check that the peer won't put funds into a RBF'd channel lease. FIXME: allow leases to pass through to RBFs Changelog-None.
2021-08-05 19:32:58 +02:00
l1, l2 = node_factory.get_nodes(2, opts=opts)
# what happens when we RBF?
feerate = 2000
amount = 500000
l1.fundwallet(20000000)
l2.fundwallet(20000000)
# l1 leases a channel from l2
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
rates = l1.rpc.dev_queryrates(l2.info['id'], amount, amount)
chan_id = l1.rpc.fundchannel(l2.info['id'], amount, request_amt=amount,
feerate='{}perkw'.format(feerate),
compact_lease=rates['compact_lease'])['channel_id']
vins = [x for x in l1.rpc.listfunds()['outputs'] if x['reserved']]
assert only_one(vins)
prev_utxos = ["{}:{}".format(vins[0]['txid'], vins[0]['output'])]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
est_fees = calc_lease_fee(amount, feerate, rates)
# This should be the accepter's amount
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
fundings = only_one(l1.rpc.listpeerchannels()['channels'])['funding']
bkpr/listpeeers: add lease_fees back to funds; separate out in listpeers First, how we record "our_funds" and then apply pushes vs lease_fees (for liquidity ad buys/sales) was exactly opposite. For pushes we were reporting the total funded into the channel, with the push representing how much we'd later moved to the peer. For lease_fees we were rerporting the total in the channel, with the push representing how much was already moved to the peer. We fix this (from a view perspective) by re-adding lease fees to what's reported in the channel funding totals. Since this is now new behavior (for leased channel values), we added new fields so we can take the old field names thru a deprecation cycle. We also make it possible to differentiate btw a push and a lease_fee (before they were all the same), by adding to new fields to `listpeers`: `fee_paid_msat` and `fee_rcvd_msat`. This allows us to avoid math in the bookkeeper, instead we just pick the numbers out directly and record them. Fixes #5472 Changelog-Added: JSON-RPC: `listpeers` now has a few new fields for `funding` (`remote_funds_msat`, `local_funds_msat`, `fee_paid_msat`, `fee_rcvd_msat`). Changelog-Deprecated: JSON-RPC: `listpeers`.`funded` fields `local_msat` and `remote_msat` are now deprecated.
2022-07-29 05:44:17 +02:00
assert Millisatoshi(amount * 1000) == fundings['remote_funds_msat']
assert Millisatoshi(est_fees + amount * 1000) == fundings['local_funds_msat']
assert Millisatoshi(est_fees) == fundings['fee_paid_msat']
assert 'fee_rcvd_msat' not in fundings
tests: check that we don't re-up the leased funds when we RBF Check that the peer won't put funds into a RBF'd channel lease. FIXME: allow leases to pass through to RBFs Changelog-None.
2021-08-05 19:32:58 +02:00
# rbf the lease with a higher amount
rate = int(find_next_feerate(l1, l2)[:-5])
# We 4x the feerate to beat the min-relay fee
next_feerate = '{}perkw'.format(rate * 4)
test: restart node during rbf state isn't kept around for lease amount, so it should fail
2022-10-24 20:00:10 +02:00
# Restart the node between open + rbf; works as expected
l1.restart()
tests: check that we don't re-up the leased funds when we RBF Check that the peer won't put funds into a RBF'd channel lease. FIXME: allow leases to pass through to RBFs Changelog-None.
2021-08-05 19:32:58 +02:00
# Initiate an RBF
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)['psbt']
test: restart node during rbf state isn't kept around for lease amount, so it should fail
2022-10-24 20:00:10 +02:00
# reconnect after restart
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
tests: check that we don't re-up the leased funds when we RBF Check that the peer won't put funds into a RBF'd channel lease. FIXME: allow leases to pass through to RBFs Changelog-None.
2021-08-05 19:32:58 +02:00
# do the bump
bump = l1.rpc.openchannel_bump(chan_id, amount, initpsbt,
funding_feerate=next_feerate)
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
assert update['commitments_secured']
funder: cleanup datastore on state-change/channel failure Let's not leave old state hanging around! Note that this fires for pretty much every/any channel (even if we're not the opener).
2022-10-07 22:37:06 +02:00
tests: check that we don't re-up the leased funds when we RBF Check that the peer won't put funds into a RBF'd channel lease. FIXME: allow leases to pass through to RBFs Changelog-None.
2021-08-05 19:32:58 +02:00
# Sign our inputs, and continue
signed_psbt = l1.rpc.signpsbt(update['psbt'])['signed_psbt']
l1.rpc.openchannel_signed(chan_id, signed_psbt)
funder: cleanup datastore on state-change/channel failure Let's not leave old state hanging around! Note that this fires for pretty much every/any channel (even if we're not the opener).
2022-10-07 22:37:06 +02:00
# There's data in the datastore now (l2 only)
assert l1.rpc.listdatastore() == {'datastore': []}
only_one(l2.rpc.listdatastore("funder/{}".format(chan_id))['datastore'])
tests: check that we don't re-up the leased funds when we RBF Check that the peer won't put funds into a RBF'd channel lease. FIXME: allow leases to pass through to RBFs Changelog-None.
2021-08-05 19:32:58 +02:00
# what happens when the channel opens?
bitcoind.generate_block(6)
l1.daemon.wait_for_log('to CHANNELD_NORMAL')
funder: cleanup datastore on state-change/channel failure Let's not leave old state hanging around! Note that this fires for pretty much every/any channel (even if we're not the opener).
2022-10-07 22:37:06 +02:00
# Datastore should be cleaned up!
assert l1.rpc.listdatastore() == {'datastore': []}
test: restart node during rbf state isn't kept around for lease amount, so it should fail
2022-10-24 20:00:10 +02:00
wait_for(lambda: l2.rpc.listdatastore() == {'datastore': []})
funder: cleanup datastore on state-change/channel failure Let's not leave old state hanging around! Note that this fires for pretty much every/any channel (even if we're not the opener).
2022-10-07 22:37:06 +02:00
tests: check that we don't re-up the leased funds when we RBF Check that the peer won't put funds into a RBF'd channel lease. FIXME: allow leases to pass through to RBFs Changelog-None.
2021-08-05 19:32:58 +02:00
# This should be the accepter's amount
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
fundings = only_one(l1.rpc.listpeerchannels()['channels'])['funding']
# The is still there!
assert Millisatoshi(amount * 1000) == Millisatoshi(fundings['remote_funds_msat'])
tests: check that we don't re-up the leased funds when we RBF Check that the peer won't put funds into a RBF'd channel lease. FIXME: allow leases to pass through to RBFs Changelog-None.
2021-08-05 19:32:58 +02:00
wait_for(lambda: [c['active'] for c in l1.rpc.listchannels(l1.get_channel_scid(l2))['channels']] == [True, True])
# send some payments, mine a block or two
inv = l2.rpc.invoice(10**4, '1', 'no_1')
l1.rpc.pay(inv['bolt11'])
# l2 attempts to close a channel that it leased, should succeed
# (channel isnt leased)
l2.rpc.close(l1.get_channel_scid(l2))
l1.daemon.wait_for_log('State changed from CLOSINGD_SIGEXCHANGE to CLOSINGD_COMPLETE')
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v2')
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
def test_v2_rbf_multi(node_factory, bitcoind, chainparams):
l1, l2 = node_factory.get_nodes(2,
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
opts={'may_reconnect': True,
pytest: don't assume disconnect finished atomically, and suppress interfering redirects. In various places, we assumed that when `connected` is false, everything is finished. This is not true: we should wait for the state we expect. In addition, various places allows reconnections, which interfered with the logic; suppress them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-07-18 14:12:28 +02:00
'dev-no-reconnect': None,
df: add openchannel_abort command Allows us to clean up an in-progress open that we won't be completing Changelog-Added: EXPERIMENTAL JSON-RPC: Permit user-initiated aborting of in-progress opens. Only valid for not-yet-committed opens and RBF-attempts
2021-03-09 22:14:08 +01:00
'allow_warning': True})
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
res = l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
assert(only_one(vins))
prev_utxos = ["{}:{}".format(vins[0]['txid'], vins[0]['vout'])]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
df: add openchannel_abort command Allows us to clean up an in-progress open that we won't be completing Changelog-Added: EXPERIMENTAL JSON-RPC: Permit user-initiated aborting of in-progress opens. Only valid for not-yet-committed opens and RBF-attempts
2021-03-09 22:14:08 +01:00
# Attempt to do abort, should fail since we've
# already gotten an inflight
with pytest.raises(RpcError):
l1.rpc.openchannel_abort(chan_id)
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
rate = int(find_next_feerate(l1, l2)[:-5])
# We 4x the feerate to beat the min-relay fee
next_feerate = '{}perkw'.format(rate * 4)
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
# Initiate an RBF
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(chan_amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
# Do the bump
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
bump = l1.rpc.openchannel_bump(chan_id, chan_amount,
initpsbt['psbt'],
funding_feerate=next_feerate)
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
df: add openchannel_abort command Allows us to clean up an in-progress open that we won't be completing Changelog-Added: EXPERIMENTAL JSON-RPC: Permit user-initiated aborting of in-progress opens. Only valid for not-yet-committed opens and RBF-attempts
2021-03-09 22:14:08 +01:00
# Abort this open attempt! We will re-try
aborted = l1.rpc.openchannel_abort(chan_id)
assert not aborted['channel_canceled']
dual-open: use tx-abort instead of warning/errors When a channel open fails, we use tx-abort instead of warning/error. This means that the peer won't disconnect! And instead when a new message arrives, we'll need to rebuild the dualopend subd (if missing). Makes opens a bit easer to retry (no reconnect needed), as well as keeps the connection alive for other channels we may have with that peer. Changelog-Changed: Experimental-Dual-Fund: open failures don't disconnect, but instead fail the opening process
2022-12-01 22:36:06 +01:00
# We no longer disconnect on aborts, because magic!
assert only_one(l1.rpc.listpeers()['peers'])['connected']
df: add openchannel_abort command Allows us to clean up an in-progress open that we won't be completing Changelog-Added: EXPERIMENTAL JSON-RPC: Permit user-initiated aborting of in-progress opens. Only valid for not-yet-committed opens and RBF-attempts
2021-03-09 22:14:08 +01:00
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
# Do the bump, again, same feerate
bump = l1.rpc.openchannel_bump(chan_id, chan_amount,
initpsbt['psbt'],
funding_feerate=next_feerate)
df: add openchannel_abort command Allows us to clean up an in-progress open that we won't be completing Changelog-Added: EXPERIMENTAL JSON-RPC: Permit user-initiated aborting of in-progress opens. Only valid for not-yet-committed opens and RBF-attempts
2021-03-09 22:14:08 +01:00
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
assert update['commitments_secured']
# Sign our inputs, and continue
signed_psbt = l1.rpc.signpsbt(update['psbt'])['signed_psbt']
l1.rpc.openchannel_signed(chan_id, signed_psbt)
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
# We 2x the feerate to beat the min-relay fee
rate = int(find_next_feerate(l1, l2)[:-5])
next_feerate = '{}perkw'.format(rate * 2)
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
# Initiate another RBF, double the channel amount this time
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(chan_amount * 2, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
# Do the bump
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
bump = l1.rpc.openchannel_bump(chan_id, chan_amount * 2,
initpsbt['psbt'],
funding_feerate=next_feerate)
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
assert update['commitments_secured']
# Sign our inputs, and continue
signed_psbt = l1.rpc.signpsbt(update['psbt'])['signed_psbt']
l1.rpc.openchannel_signed(chan_id, signed_psbt)
bitcoind.generate_block(1)
sync_blockheight(bitcoind, [l1])
l1.daemon.wait_for_log(' to CHANNELD_NORMAL')
df-rbf: reconnection tests (init_rbf + ack_rbf)
2021-02-10 22:17:02 +01:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v2')
df-rbf: reconnection tests (init_rbf + ack_rbf)
2021-02-10 22:17:02 +01:00
def test_rbf_reconnect_init(node_factory, bitcoind, chainparams):
dual-fund: update extracted CSVs to latest bolt draft Changelog-None
2022-07-11 16:14:51 +02:00
disconnects = ['-WIRE_TX_INIT_RBF',
'+WIRE_TX_INIT_RBF']
df-rbf: reconnection tests (init_rbf + ack_rbf)
2021-02-10 22:17:02 +01:00
l1, l2 = node_factory.get_nodes(2,
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
opts=[{'disconnect': disconnects,
df-rbf: reconnection tests (init_rbf + ack_rbf)
2021-02-10 22:17:02 +01:00
'may_reconnect': True},
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
{'may_reconnect': True}])
df-rbf: reconnection tests (init_rbf + ack_rbf)
2021-02-10 22:17:02 +01:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
res = l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
assert(only_one(vins))
prev_utxos = ["{}:{}".format(vins[0]['txid'], vins[0]['vout'])]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
next_feerate = find_next_feerate(l1, l2)
# Initiate an RBF
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(chan_amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
# Do the bump!?
for d in disconnects:
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
with pytest.raises(RpcError):
l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
assert l1.rpc.getpeer(l2.info['id']) is not None
# This should succeed
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v2')
df-rbf: reconnection tests (init_rbf + ack_rbf)
2021-02-10 22:17:02 +01:00
def test_rbf_reconnect_ack(node_factory, bitcoind, chainparams):
dual-fund: update extracted CSVs to latest bolt draft Changelog-None
2022-07-11 16:14:51 +02:00
disconnects = ['-WIRE_TX_ACK_RBF',
'+WIRE_TX_ACK_RBF']
df-rbf: reconnection tests (init_rbf + ack_rbf)
2021-02-10 22:17:02 +01:00
l1, l2 = node_factory.get_nodes(2,
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
opts=[{'may_reconnect': True},
{'disconnect': disconnects,
df-rbf: reconnection tests (init_rbf + ack_rbf)
2021-02-10 22:17:02 +01:00
'may_reconnect': True}])
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
res = l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
assert(only_one(vins))
prev_utxos = ["{}:{}".format(vins[0]['txid'], vins[0]['vout'])]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
next_feerate = find_next_feerate(l1, l2)
# Initiate an RBF
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(chan_amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
# Do the bump!?
for d in disconnects:
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
with pytest.raises(RpcError):
l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
assert l1.rpc.getpeer(l2.info['id']) is not None
# This should succeed
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
df-rbf: test for interrupts during tx construction
2021-02-11 00:46:26 +01:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v2')
df-rbf: test for interrupts during tx construction
2021-02-11 00:46:26 +01:00
def test_rbf_reconnect_tx_construct(node_factory, bitcoind, chainparams):
disconnects = ['=WIRE_TX_ADD_INPUT', # Initial funding succeeds
'-WIRE_TX_ADD_INPUT',
'+WIRE_TX_ADD_INPUT',
'-WIRE_TX_ADD_OUTPUT',
'+WIRE_TX_ADD_OUTPUT',
'-WIRE_TX_COMPLETE',
'+WIRE_TX_COMPLETE']
l1, l2 = node_factory.get_nodes(2,
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
opts=[{'disconnect': disconnects,
pytest: disable autoreconnect in test_rbf_reconnect_tx_construct It reconnects itself, so we don't see the ['connected'] is False. ``` 2022-07-20T20:37:06.3808161Z @unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need') 2022-07-20T20:37:06.3809031Z @pytest.mark.developer("uses dev-disconnect") 2022-07-20T20:37:06.3809547Z @pytest.mark.openchannel('v2') 2022-07-20T20:37:06.3810058Z def test_rbf_reconnect_tx_construct(node_factory, bitcoind, chainparams): ... 2022-07-20T20:37:06.3864163Z # Now we finish off the completes failure check 2022-07-20T20:37:06.3864604Z for d in disconnects[-2:]: 2022-07-20T20:37:06.3865162Z l1.rpc.connect(l2.info['id'], 'localhost', l2.port) 2022-07-20T20:37:06.3865711Z bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt']) 2022-07-20T20:37:06.3866169Z with pytest.raises(RpcError): 2022-07-20T20:37:06.3866674Z update = l1.rpc.openchannel_update(chan_id, bump['psbt']) 2022-07-20T20:37:06.3867367Z > wait_for(lambda: l1.rpc.getpeer(l2.info['id'])['connected'] is False) ... 2022-07-20T20:37:06.5215961Z lightningd-1 2022-07-20T20:21:49.691Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-connectd: dev_disconnect: -WIRE_TX_COMPLETE (WIRE_TX_COMPLETE) 2022-07-20T20:37:06.5216482Z lightningd-1 2022-07-20T20:21:49.691Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-dualopend-chan#1: peer_out WIRE_TX_COMPLETE 2022-07-20T20:37:06.5216756Z lightningd-1 2022-07-20T20:21:49.691Z DEBUG connectd: drain_peer 2022-07-20T20:37:06.5217064Z lightningd-1 2022-07-20T20:21:49.692Z DEBUG connectd: drain_peer draining subd! 2022-07-20T20:37:06.5217549Z lightningd-1 2022-07-20T20:21:49.692Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-lightningd: peer_disconnect_done 2022-07-20T20:37:06.5218482Z lightningd-1 2022-07-20T20:21:49.692Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-lightningd: Reconnecting in 1 seconds 2022-07-20T20:37:06.5219110Z lightningd-1 2022-07-20T20:21:49.692Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-lightningd: Will try reconnect in 1 seconds 2022-07-20T20:37:06.5219427Z lightningd-1 2022-07-20T20:21:49.692Z DEBUG gossipd: REPLY WIRE_GOSSIPD_GET_ADDRS_REPLY with 0 fds 2022-07-20T20:37:06.5219994Z lightningd-1 2022-07-20T20:21:49.696Z DEBUG plugin-funder: Cleaning up inflights for peer id 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59 2022-07-20T20:37:06.5220305Z lightningd-1 2022-07-20T20:21:49.696Z DEBUG connectd: maybe_free_peer freeing peer! 2022-07-20T20:37:06.5220743Z lightningd-2 2022-07-20T20:21:49.699Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-gossipd: seeker: chosen as startup peer 2022-07-20T20:37:06.5221136Z lightningd-2 2022-07-20T20:21:49.699Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-hsmd: Got WIRE_HSMD_ECDH_REQ 2022-07-20T20:37:06.5221380Z lightningd-2 2022-07-20T20:21:49.699Z DEBUG connectd: drain_peer 2022-07-20T20:37:06.5221805Z lightningd-1 2022-07-20T20:21:49.700Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-hsmd: Got WIRE_HSMD_READY_CHANNEL 2022-07-20T20:37:06.5223761Z lightningd-1 2022-07-20T20:21:49.700Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-dualopend-chan#1: signature 30440220338460c0e75c08c21f4f4f96806d81426aee48e06ccc54e87892b332f55fe49e0220527573286f801a0d23317978a9aabbbbcb95c2ceb614596c0ab50bb72b96158b01 on tx 020000000105f2aa5f346f46007b9f8b128e8c6e4d40d0477aefd81250ba99adc9349aabe300000000009db0e280024a01000000000000220020be7935a77ca9ab70a4b8b1906825637767fed3c00824aa90c988983587d684881e6301000000000022002047021684129f8aa1c0b5b97dc99607f5f0850b548813a5da346fda22511284759a3ed620 using key 02324266de8403b3ab157a09f1f784d587af61831c998c151bcc21bb74c2b2314b 2022-07-20T20:37:06.5224194Z lightningd-1 2022-07-20T20:21:49.700Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-connectd: Connected out, starting crypto 2022-07-20T20:37:06.5224723Z lightningd-1 2022-07-20T20:21:49.700Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-gossipd: seeker: chosen as startup peer 2022-07-20T20:37:06.5225006Z lightningd-1 2022-07-20T20:21:49.700Z DEBUG hsmd: Client: Received message 31 from client 2022-07-20T20:37:06.5225461Z lightningd-1 2022-07-20T20:21:49.700Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-dualopend-chan#1: peer_out WIRE_COMMITMENT_SIGNED 2022-07-20T20:37:06.5225852Z lightningd-1 2022-07-20T20:21:49.700Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-connectd: Connect OUT ```
2022-07-21 06:49:14 +02:00
'may_reconnect': True,
'dev-no-reconnect': None},
{'may_reconnect': True,
'dev-no-reconnect': None}])
df-rbf: test for interrupts during tx construction
2021-02-11 00:46:26 +01:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
res = l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
assert(only_one(vins))
prev_utxos = ["{}:{}".format(vins[0]['txid'], vins[0]['vout'])]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
next_feerate = find_next_feerate(l1, l2)
# Initiate an RBF
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(chan_amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
# Run through TX_ADD wires
dev_disconnect: remove @ marker. Once connectd is doing this, we can't close as soon as we send, and in fact we can't do 'fail write' either. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2021-12-28 00:27:09 +01:00
for d in disconnects[1:-2]:
df-rbf: test for interrupts during tx construction
2021-02-11 00:46:26 +01:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
with pytest.raises(RpcError):
l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
pytest: don't assume disconnect finished atomically, and suppress interfering redirects. In various places, we assumed that when `connected` is false, everything is finished. This is not true: we should wait for the state we expect. In addition, various places allows reconnections, which interfered with the logic; suppress them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-07-18 14:12:28 +02:00
wait_for(lambda: l1.rpc.getpeer(l2.info['id'])['connected'] is False)
df-rbf: test for interrupts during tx construction
2021-02-11 00:46:26 +01:00
# Now we finish off the completes failure check
dev_disconnect: remove @ marker. Once connectd is doing this, we can't close as soon as we send, and in fact we can't do 'fail write' either. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2021-12-28 00:27:09 +01:00
for d in disconnects[-2:]:
df-rbf: test for interrupts during tx construction
2021-02-11 00:46:26 +01:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
with pytest.raises(RpcError):
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
pytest: don't assume disconnect finished atomically, and suppress interfering redirects. In various places, we assumed that when `connected` is false, everything is finished. This is not true: we should wait for the state we expect. In addition, various places allows reconnections, which interfered with the logic; suppress them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-07-18 14:12:28 +02:00
wait_for(lambda: l1.rpc.getpeer(l2.info['id'])['connected'] is False)
df-rbf: test for interrupts during tx construction
2021-02-11 00:46:26 +01:00
# Now we succeed
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
assert update['commitments_secured']
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v2')
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
def test_rbf_reconnect_tx_sigs(node_factory, bitcoind, chainparams):
disconnects = ['=WIRE_TX_SIGNATURES', # Initial funding succeeds
'-WIRE_TX_SIGNATURES', # When we send tx-sigs, RBF
'=WIRE_TX_SIGNATURES', # When we reconnect
'+WIRE_TX_SIGNATURES'] # When we RBF again
l1, l2 = node_factory.get_nodes(2,
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
opts=[{'disconnect': disconnects,
df: more reconnection tests
2021-03-16 01:50:31 +01:00
'may_reconnect': True},
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
{'may_reconnect': True}])
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
res = l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
assert(only_one(vins))
prev_utxos = ["{}:{}".format(vins[0]['txid'], vins[0]['vout'])]
# Check that we're waiting for lockin
l1.daemon.wait_for_log('Broadcasting funding tx')
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
df-tests: bump feerates to get over the 'insufficient fee' limit Naively calling bump with the next feerate doesnt add enough extra to get over the min-relay requirement for RBF.
2021-08-02 20:48:37 +02:00
rate = int(find_next_feerate(l1, l2)[:-5])
# We 4x the feerate to beat the min-relay fee
next_feerate = '{}perkw'.format(rate * 4)
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
# Initiate an RBF
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(chan_amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
df-tests: bump feerates to get over the 'insufficient fee' limit Naively calling bump with the next feerate doesnt add enough extra to get over the min-relay requirement for RBF.
2021-08-02 20:48:37 +02:00
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'],
funding_feerate=next_feerate)
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
# Sign our inputs, and continue
signed_psbt = l1.rpc.signpsbt(update['psbt'])['signed_psbt']
# First time we error when we send our sigs
with pytest.raises(RpcError, match='Owning subdaemon dualopend died'):
l1.rpc.openchannel_signed(chan_id, signed_psbt)
# We reconnect and try again. feerate should have bumped
df-tests: bump feerates to get over the 'insufficient fee' limit Naively calling bump with the next feerate doesnt add enough extra to get over the min-relay requirement for RBF.
2021-08-02 20:48:37 +02:00
rate = int(find_next_feerate(l1, l2)[:-5])
# We 4x the feerate to beat the min-relay fee
next_feerate = '{}perkw'.format(rate * 4)
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
# Initiate an RBF
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.utxopsbt(chan_amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
# l2 gets our sigs and broadcasts them
l2.daemon.wait_for_log('peer_in WIRE_CHANNEL_REESTABLISH')
l2.daemon.wait_for_log('peer_in WIRE_TX_SIGNATURES')
l2.daemon.wait_for_log('sendrawtx exit 0')
# Wait until we've done re-establish, if we try to
# RBF again too quickly, it'll fail since they haven't
# had time to process our sigs yet
l1.daemon.wait_for_log('peer_in WIRE_CHANNEL_REESTABLISH')
l1.daemon.wait_for_log('peer_in WIRE_TX_SIGNATURES')
dev_disconnect: remove @ marker. Once connectd is doing this, we can't close as soon as we send, and in fact we can't do 'fail write' either. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2021-12-28 00:27:09 +01:00
# 2nd RBF
df-tests: bump feerates to get over the 'insufficient fee' limit Naively calling bump with the next feerate doesnt add enough extra to get over the min-relay requirement for RBF.
2021-08-02 20:48:37 +02:00
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'],
funding_feerate=next_feerate)
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
signed_psbt = l1.rpc.signpsbt(update['psbt'])['signed_psbt']
# Second time we error after we send our sigs
with pytest.raises(RpcError, match='Owning subdaemon dualopend died'):
l1.rpc.openchannel_signed(chan_id, signed_psbt)
# l2 gets our sigs
l2.daemon.wait_for_log('peer_in WIRE_TX_SIGNATURES')
l2.daemon.wait_for_log('sendrawtx exit 0')
# mine a block?
bitcoind.generate_block(1)
sync_blockheight(bitcoind, [l1])
l1.daemon.wait_for_log(' to CHANNELD_NORMAL')
# Check that they have matching funding txid
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
l1_funding_txid = only_one(l1.rpc.listpeerchannels()['channels'])['funding_txid']
l2_funding_txid = only_one(l2.rpc.listpeerchannels()['channels'])['funding_txid']
df-tests: check that multiple RBFs and tx-sigs errors during RBF work
2021-02-12 00:01:07 +01:00
assert l1_funding_txid == l2_funding_txid
df-tests: check that rbf includes overlapping inputs
2021-02-12 00:14:26 +01:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v2')
df-tests: check that rbf includes overlapping inputs
2021-02-12 00:14:26 +01:00
def test_rbf_no_overlap(node_factory, bitcoind, chainparams):
l1, l2 = node_factory.get_nodes(2,
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
opts={'allow_warning': True})
df-tests: check that rbf includes overlapping inputs
2021-02-12 00:14:26 +01:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
res = l1.rpc.fundchannel(l2.info['id'], chan_amount)
chan_id = res['channel_id']
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
next_feerate = find_next_feerate(l1, l2)
df-tests: add better explainer for test logic
2021-05-07 21:50:17 +02:00
# Initiate an RBF (this grabs the non-reserved utxo, which isnt the
# one we started with)
df-tests: check that rbf includes overlapping inputs
2021-02-12 00:14:26 +01:00
startweight = 42 + 172 # base weight, funding output
initpsbt = l1.rpc.fundpsbt(chan_amount, next_feerate, startweight,
min_witness_weight=110,
excess_as_change=True)
# Do the bump
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
with pytest.raises(RpcError, match='No overlapping input present.'):
l1.rpc.openchannel_update(chan_id, bump['psbt'])
tests: add test for funder options
2021-04-27 23:12:14 +02:00
df-test: test for a failed rbf attempt, currently crashes An attempted + failed RBF results in a crashed/dead node
2021-05-07 21:31:45 +02:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
@pytest.mark.openchannel('v2')
def test_rbf_fails_to_broadcast(node_factory, bitcoind, chainparams):
l1, l2 = node_factory.get_nodes(2,
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
opts={'allow_warning': True,
'may_reconnect': True})
df-test: test for a failed rbf attempt, currently crashes An attempted + failed RBF results in a crashed/dead node
2021-05-07 21:31:45 +02:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
# Really low feerate means that the bump wont work the first time
res = l1.rpc.fundchannel(l2.info['id'], chan_amount, feerate='253perkw')
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
assert(only_one(vins))
prev_utxos = ["{}:{}".format(vins[0]['txid'], vins[0]['vout'])]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
inflights = only_one(l1.rpc.listpeerchannels()['channels'])['inflight']
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
assert inflights[-1]['funding_txid'] in bitcoind.rpc.getrawmempool()
df-test: test for a failed rbf attempt, currently crashes An attempted + failed RBF results in a crashed/dead node
2021-05-07 21:31:45 +02:00
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
def run_retry():
startweight = 42 + 173
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
rate = int(find_next_feerate(l1, l2)[:-5])
# We 2x the feerate to beat the min-relay fee
next_feerate = '{}perkw'.format(rate * 2)
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
initpsbt = l1.rpc.utxopsbt(chan_amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
df-test: test for a failed rbf attempt, currently crashes An attempted + failed RBF results in a crashed/dead node
2021-05-07 21:31:45 +02:00
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
bump = l1.rpc.openchannel_bump(chan_id, chan_amount,
initpsbt['psbt'],
funding_feerate=next_feerate)
rbf-tests: check every inflight has signed commitment produced Little check to make sure that we can produce a signed commitment tx for every inflight we've got saved.
2021-05-20 02:16:49 +02:00
# We should be able to call this with while an open is progress
# but not yet committed
l1.rpc.dev_sign_last_tx(l2.info['id'])
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
assert update['commitments_secured']
df-test: test for a failed rbf attempt, currently crashes An attempted + failed RBF results in a crashed/dead node
2021-05-07 21:31:45 +02:00
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
return l1.rpc.signpsbt(update['psbt'])['signed_psbt']
df-test: test for a failed rbf attempt, currently crashes An attempted + failed RBF results in a crashed/dead node
2021-05-07 21:31:45 +02:00
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
signed_psbt = run_retry()
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
l1.rpc.openchannel_signed(chan_id, signed_psbt)
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
inflights = only_one(l1.rpc.listpeerchannels()['channels'])['inflight']
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
assert inflights[-1]['funding_txid'] in bitcoind.rpc.getrawmempool()
df-test: test for a failed rbf attempt, currently crashes An attempted + failed RBF results in a crashed/dead node
2021-05-07 21:31:45 +02:00
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
# Restart and listpeers, used to crash
df-test: test for a failed rbf attempt, currently crashes An attempted + failed RBF results in a crashed/dead node
2021-05-07 21:31:45 +02:00
l1.restart()
l1.rpc.listpeers()
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
# We've restarted. Let's RBF
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
signed_psbt = run_retry()
l1.rpc.openchannel_signed(chan_id, signed_psbt)
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
inflights = only_one(l1.rpc.listpeerchannels()['channels'])['inflight']
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
assert len(inflights) == 3
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
assert inflights[-1]['funding_txid'] in bitcoind.rpc.getrawmempool()
l1.restart()
# Are inflights the same post restart
prev_inflights = inflights
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
inflights = only_one(l1.rpc.listpeerchannels()['channels'])['inflight']
df-rbf: test that, with enough bumps, we can eventually add to mempool We don't pass the minimum fee requirement the first few times we attempt an RBF, so it fails. Keep going until actually replaces
2021-05-19 23:51:56 +02:00
assert prev_inflights == inflights
assert inflights[-1]['funding_txid'] in bitcoind.rpc.getrawmempool()
rbf-tests: check every inflight has signed commitment produced Little check to make sure that we can produce a signed commitment tx for every inflight we've got saved.
2021-05-20 02:16:49 +02:00
# Produce a signature for every inflight
last_txs = l1.rpc.dev_sign_last_tx(l2.info['id'])
assert len(last_txs['inflights']) == len(inflights)
for last_tx, inflight in zip(last_txs['inflights'], inflights):
assert last_tx['funding_txid'] == inflight['funding_txid']
assert last_txs['tx']
df-test: test for a failed rbf attempt, currently crashes An attempted + failed RBF results in a crashed/dead node
2021-05-07 21:31:45 +02:00
open-rbf: we broadcast all the commitments for a channel when closed If you drop an rbf'd channel to chain (before any updates have been made) we should drop *all* of the inflights to chain
2021-05-20 02:18:39 +02:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
@pytest.mark.openchannel('v2')
def test_rbf_broadcast_close_inflights(node_factory, bitcoind, chainparams):
"""
Close a channel before it's mined, and the most recent transaction
hasn't made it to the mempool. Should publish all the commitment
transactions that we have.
"""
l1, l2 = node_factory.get_nodes(2,
opts={'allow_warning': True})
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
res = l1.rpc.fundchannel(l2.info['id'], chan_amount, feerate='7500perkw')
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
assert(only_one(vins))
prev_utxos = ["{}:{}".format(vins[0]['txid'], vins[0]['vout'])]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
inflights = only_one(l1.rpc.listpeerchannels()['channels'])['inflight']
open-rbf: we broadcast all the commitments for a channel when closed If you drop an rbf'd channel to chain (before any updates have been made) we should drop *all* of the inflights to chain
2021-05-20 02:18:39 +02:00
assert inflights[-1]['funding_txid'] in bitcoind.rpc.getrawmempool()
# Make it such that l1 and l2 cannot broadcast transactions
# (mimics failing to reach the miner with replacement)
def censoring_sendrawtx(r):
return {'id': r['id'], 'result': {}}
l1.daemon.rpcproxy.mock_rpc('sendrawtransaction', censoring_sendrawtx)
l2.daemon.rpcproxy.mock_rpc('sendrawtransaction', censoring_sendrawtx)
def run_retry():
startweight = 42 + 173
next_feerate = find_next_feerate(l1, l2)
initpsbt = l1.rpc.utxopsbt(chan_amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
assert update['commitments_secured']
return l1.rpc.signpsbt(update['psbt'])['signed_psbt']
signed_psbt = run_retry()
l1.rpc.openchannel_signed(chan_id, signed_psbt)
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
inflights = only_one(l1.rpc.listpeerchannels()['channels'])['inflight']
open-rbf: we broadcast all the commitments for a channel when closed If you drop an rbf'd channel to chain (before any updates have been made) we should drop *all* of the inflights to chain
2021-05-20 02:18:39 +02:00
assert inflights[-1]['funding_txid'] not in bitcoind.rpc.getrawmempool()
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
cmtmt_txid = only_one(l1.rpc.listpeerchannels()['channels'])['scratch_txid']
open-rbf: we broadcast all the commitments for a channel when closed If you drop an rbf'd channel to chain (before any updates have been made) we should drop *all* of the inflights to chain
2021-05-20 02:18:39 +02:00
assert cmtmt_txid == inflights[-1]['scratch_txid']
# l2 goes offline
l2.stop()
# l1 drops to chain.
l1.daemon.rpcproxy.mock_rpc('sendrawtransaction', None)
l1.rpc.close(chan_id, 1)
l1.daemon.wait_for_logs(['Broadcasting txid {}'.format(inflights[0]['scratch_txid']),
'Broadcasting txid {}'.format(inflights[1]['scratch_txid']),
'sendrawtx exit 0',
'sendrawtx exit 25'])
assert inflights[0]['scratch_txid'] in bitcoind.rpc.getrawmempool()
assert inflights[1]['scratch_txid'] not in bitcoind.rpc.getrawmempool()
rbf-test: confirm that rbf'd inflight is opened and closed as expected We weren't testing that a non-tip inflight would get opened correctly. Until now, that is.
2021-05-20 23:55:45 +02:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
@pytest.mark.openchannel('v2')
def test_rbf_non_last_mined(node_factory, bitcoind, chainparams):
"""
What happens if a 'non-tip' RBF transaction is mined?
"""
l1, l2 = node_factory.get_nodes(2,
opts={'allow_warning': True,
'may_reconnect': True})
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
amount = 2**24
chan_amount = 100000
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], amount / 10**8 + 0.01)
bitcoind.generate_block(1)
# Wait for it to arrive.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) > 0)
res = l1.rpc.fundchannel(l2.info['id'], chan_amount, feerate='7500perkw')
chan_id = res['channel_id']
vins = bitcoind.rpc.decoderawtransaction(res['tx'])['vin']
assert(only_one(vins))
prev_utxos = ["{}:{}".format(vins[0]['txid'], vins[0]['vout'])]
# Check that we're waiting for lockin
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
inflights = only_one(l1.rpc.listpeerchannels()['channels'])['inflight']
rbf-test: confirm that rbf'd inflight is opened and closed as expected We weren't testing that a non-tip inflight would get opened correctly. Until now, that is.
2021-05-20 23:55:45 +02:00
assert inflights[-1]['funding_txid'] in bitcoind.rpc.getrawmempool()
def run_retry():
startweight = 42 + 173
dual-funding: update fee_step to be a feerate Using a 'feestep' is more restrictive than you'd want, instead we enforce that the next feerate must be at least 1/64th more than the last, but put no upper limit on it Includes update to lnprototest changes Contributed-By: Vincenzo Palazzo <vincenzopalazzodev@gmail.com> Changelog-EXPERIMENTAL: Protocol: Replaces init_rbf's `fee_step` for RBF of v2 opens with `funding_feerate_perkw`, breaking change
2021-07-09 21:13:20 +02:00
rate = int(find_next_feerate(l1, l2)[:-5])
# We 2x the feerate to beat the min-relay fee
next_feerate = '{}perkw'.format(rate * 2)
rbf-test: confirm that rbf'd inflight is opened and closed as expected We weren't testing that a non-tip inflight would get opened correctly. Until now, that is.
2021-05-20 23:55:45 +02:00
initpsbt = l1.rpc.utxopsbt(chan_amount, next_feerate, startweight,
prev_utxos, reservedok=True,
min_witness_weight=110,
excess_as_change=True)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
bump = l1.rpc.openchannel_bump(chan_id, chan_amount, initpsbt['psbt'])
update = l1.rpc.openchannel_update(chan_id, bump['psbt'])
assert update['commitments_secured']
return l1.rpc.signpsbt(update['psbt'])['signed_psbt']
# Make a second inflight
signed_psbt = run_retry()
l1.rpc.openchannel_signed(chan_id, signed_psbt)
# Make it such that l1 and l2 cannot broadcast transactions
# (mimics failing to reach the miner with replacement)
def censoring_sendrawtx(r):
return {'id': r['id'], 'result': {}}
l1.daemon.rpcproxy.mock_rpc('sendrawtransaction', censoring_sendrawtx)
l2.daemon.rpcproxy.mock_rpc('sendrawtransaction', censoring_sendrawtx)
# Make a 3rd inflight that won't make it into the mempool
signed_psbt = run_retry()
l1.rpc.openchannel_signed(chan_id, signed_psbt)
l1.daemon.rpcproxy.mock_rpc('sendrawtransaction', None)
l2.daemon.rpcproxy.mock_rpc('sendrawtransaction', None)
# We fetch out our inflights list
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
inflights = only_one(l1.rpc.listpeerchannels()['channels'])['inflight']
rbf-test: confirm that rbf'd inflight is opened and closed as expected We weren't testing that a non-tip inflight would get opened correctly. Until now, that is.
2021-05-20 23:55:45 +02:00
# l2 goes offline
l2.stop()
# The funding transaction gets mined (should be the 2nd inflight)
bitcoind.generate_block(6, wait_for_mempool=1)
# l2 comes back up
l2.start()
# everybody's got the right things now
l1.daemon.wait_for_log(r'to CHANNELD_NORMAL')
l2.daemon.wait_for_log(r'to CHANNELD_NORMAL')
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
channel = only_one(l1.rpc.listpeerchannels()['channels'])
rbf-test: confirm that rbf'd inflight is opened and closed as expected We weren't testing that a non-tip inflight would get opened correctly. Until now, that is.
2021-05-20 23:55:45 +02:00
assert channel['funding_txid'] == inflights[1]['funding_txid']
assert channel['scratch_txid'] == inflights[1]['scratch_txid']
# We delete inflights when the channel is in normal ops
assert 'inflights' not in channel
# l2 stops, again
l2.stop()
# l1 drops to chain.
l1.rpc.close(chan_id, 1)
l1.daemon.wait_for_log('Broadcasting txid {}'.format(channel['scratch_txid']))
# The funding transaction gets mined (should be the 2nd inflight)
bitcoind.generate_block(1, wait_for_mempool=1)
l1.daemon.wait_for_log(r'to ONCHAIN')
tests: add test for funder options
2021-04-27 23:12:14 +02:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
@pytest.mark.openchannel('v2')
tests: add test for funder options
2021-04-27 23:12:14 +02:00
def test_funder_options(node_factory, bitcoind):
tests: add marker for v1/v2 channel opens Tests that will only run when !EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v1') def test_...() Tests that will only run when EXPERIMENTAL_DUAL_FUND: @pytest.marker.openchannel('v2') def test_...()
2021-04-26 21:23:40 +02:00
l1, l2, l3 = node_factory.get_nodes(3)
tests: add test for funder options
2021-04-27 23:12:14 +02:00
l1.fundwallet(10**7)
# Check the default options
funder_opts = l1.rpc.call('funderupdate')
assert funder_opts['policy'] == 'fixed'
assert funder_opts['policy_mod'] == 0
funderupdate: add documentation, update some params Documentation for the funderupdate command on the funder plugin. Realized we were missing the "leases_only" in the output; also adds '_msat' to msat denominated outputs so they're parsed correctly by our python bindings (also matches our naming conventions) Changelog-EXPERIMENTAL: funder: `funderupdate` command to view and update params for contributing our wallet funds to v2 channel openings. Provides params for enabling `option_will_fund`.
2021-07-15 18:48:36 +02:00
assert funder_opts['min_their_funding_msat'] == Millisatoshi('10000000msat')
assert funder_opts['max_their_funding_msat'] == Millisatoshi('4294967295000msat')
assert funder_opts['per_channel_min_msat'] == Millisatoshi('10000000msat')
assert funder_opts['per_channel_max_msat'] == Millisatoshi('4294967295000msat')
assert funder_opts['reserve_tank_msat'] == Millisatoshi('0msat')
funder_policy: set default to 0 As we move to advertising liquidity, set the default to not have any fuzz. This means we won't randomly reject an offer to buy liquidity.
2021-07-14 19:53:56 +02:00
assert funder_opts['fuzz_percent'] == 0
tests: add test for funder options
2021-04-27 23:12:14 +02:00
assert funder_opts['fund_probability'] == 100
funder: default to only funding leases Make the default to only lease out funds. Changelog-Changed: funder plugin defaults to leases-only
2021-08-05 19:31:49 +02:00
assert funder_opts['leases_only']
tests: add test for funder options
2021-04-27 23:12:14 +02:00
# l2 funds a chanenl with us. We don't contribute
l2.rpc.connect(l1.info['id'], 'localhost', l1.port)
l2.fundchannel(l1, 10**6)
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
chan_info = only_one(l2.rpc.listpeerchannels(l1.info['id'])['channels'])
tests: add test for funder options
2021-04-27 23:12:14 +02:00
# l1 contributed nothing
bkpr/listpeeers: add lease_fees back to funds; separate out in listpeers First, how we record "our_funds" and then apply pushes vs lease_fees (for liquidity ad buys/sales) was exactly opposite. For pushes we were reporting the total funded into the channel, with the push representing how much we'd later moved to the peer. For lease_fees we were rerporting the total in the channel, with the push representing how much was already moved to the peer. We fix this (from a view perspective) by re-adding lease fees to what's reported in the channel funding totals. Since this is now new behavior (for leased channel values), we added new fields so we can take the old field names thru a deprecation cycle. We also make it possible to differentiate btw a push and a lease_fee (before they were all the same), by adding to new fields to `listpeers`: `fee_paid_msat` and `fee_rcvd_msat`. This allows us to avoid math in the bookkeeper, instead we just pick the numbers out directly and record them. Fixes #5472 Changelog-Added: JSON-RPC: `listpeers` now has a few new fields for `funding` (`remote_funds_msat`, `local_funds_msat`, `fee_paid_msat`, `fee_rcvd_msat`). Changelog-Deprecated: JSON-RPC: `listpeers`.`funded` fields `local_msat` and `remote_msat` are now deprecated.
2022-07-29 05:44:17 +02:00
assert chan_info['funding']['remote_funds_msat'] == Millisatoshi('0msat')
assert chan_info['funding']['local_funds_msat'] != Millisatoshi('0msat')
tests: add test for funder options
2021-04-27 23:12:14 +02:00
# Change all the options
funder_opts = l1.rpc.call('funderupdate',
{'policy': 'available',
'policy_mod': 100,
funderupdate: add documentation, update some params Documentation for the funderupdate command on the funder plugin. Realized we were missing the "leases_only" in the output; also adds '_msat' to msat denominated outputs so they're parsed correctly by our python bindings (also matches our naming conventions) Changelog-EXPERIMENTAL: funder: `funderupdate` command to view and update params for contributing our wallet funds to v2 channel openings. Provides params for enabling `option_will_fund`.
2021-07-15 18:48:36 +02:00
'min_their_funding_msat': '100000msat',
'max_their_funding_msat': '2000000000msat',
'per_channel_min_msat': '8000000msat',
'per_channel_max_msat': '10000000000msat',
'reserve_tank_msat': '3000000msat',
tests: add test for funder options
2021-04-27 23:12:14 +02:00
'fund_probability': 99,
funder: default to only funding leases Make the default to only lease out funds. Changelog-Changed: funder plugin defaults to leases-only
2021-08-05 19:31:49 +02:00
'fuzz_percent': 0,
'leases_only': False})
tests: add test for funder options
2021-04-27 23:12:14 +02:00
assert funder_opts['policy'] == 'available'
assert funder_opts['policy_mod'] == 100
funderupdate: add documentation, update some params Documentation for the funderupdate command on the funder plugin. Realized we were missing the "leases_only" in the output; also adds '_msat' to msat denominated outputs so they're parsed correctly by our python bindings (also matches our naming conventions) Changelog-EXPERIMENTAL: funder: `funderupdate` command to view and update params for contributing our wallet funds to v2 channel openings. Provides params for enabling `option_will_fund`.
2021-07-15 18:48:36 +02:00
assert funder_opts['min_their_funding_msat'] == Millisatoshi('100000msat')
assert funder_opts['max_their_funding_msat'] == Millisatoshi('2000000000msat')
assert funder_opts['per_channel_min_msat'] == Millisatoshi('8000000msat')
assert funder_opts['per_channel_max_msat'] == Millisatoshi('10000000000msat')
assert funder_opts['reserve_tank_msat'] == Millisatoshi('3000000msat')
tests: add test for funder options
2021-04-27 23:12:14 +02:00
assert funder_opts['fuzz_percent'] == 0
assert funder_opts['fund_probability'] == 99
# Set the fund probability back up to 100.
funder_opts = l1.rpc.call('funderupdate',
{'fund_probability': 100})
l3.rpc.connect(l1.info['id'], 'localhost', l1.port)
l3.fundchannel(l1, 10**6)
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
chan_info = only_one(l3.rpc.listpeerchannels(l1.info['id'])['channels'])
df tests: use the amount from the logs to check for!
2022-10-19 18:26:21 +02:00
log = l1.daemon.wait_for_log(r'Policy available \(100%\) returned funding amount of')
match = re.search(r'Policy available \(100%\) returned funding amount of (\d*sat)', log)
assert match and len(match.groups()) == 1
tests: update deprecated 'funding_msat' for new output format
2021-07-09 18:47:15 +02:00
# l1 contributed all its funds!
df tests: use the amount from the logs to check for!
2022-10-19 18:26:21 +02:00
assert chan_info['funding']['remote_funds_msat'] == Millisatoshi(match.groups()[0])
bkpr/listpeeers: add lease_fees back to funds; separate out in listpeers First, how we record "our_funds" and then apply pushes vs lease_fees (for liquidity ad buys/sales) was exactly opposite. For pushes we were reporting the total funded into the channel, with the push representing how much we'd later moved to the peer. For lease_fees we were rerporting the total in the channel, with the push representing how much was already moved to the peer. We fix this (from a view perspective) by re-adding lease fees to what's reported in the channel funding totals. Since this is now new behavior (for leased channel values), we added new fields so we can take the old field names thru a deprecation cycle. We also make it possible to differentiate btw a push and a lease_fee (before they were all the same), by adding to new fields to `listpeers`: `fee_paid_msat` and `fee_rcvd_msat`. This allows us to avoid math in the bookkeeper, instead we just pick the numbers out directly and record them. Fixes #5472 Changelog-Added: JSON-RPC: `listpeers` now has a few new fields for `funding` (`remote_funds_msat`, `local_funds_msat`, `fee_paid_msat`, `fee_rcvd_msat`). Changelog-Deprecated: JSON-RPC: `listpeers`.`funded` fields `local_msat` and `remote_msat` are now deprecated.
2022-07-29 05:44:17 +02:00
assert chan_info['funding']['local_funds_msat'] == Millisatoshi('1000000000msat')
funder: test for adding utxos, calculating total available Make sure that 100% available actually puts in 100% of what we can add (note that this isn't 'all' -- we don't include uneconomic utxos)
2021-05-05 01:53:24 +02:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd doesnt yet support PSBT features we need')
def test_funder_contribution_limits(node_factory, bitcoind):
opts = {'experimental-dual-fund': None,
'feerates': (5000, 5000, 5000, 5000)}
l1, l2, l3 = node_factory.get_nodes(3, opts=opts)
pytest: speed up test_opening.py::test_funder_contribution_limits when !DEVELOPER This test takes 695 seconds, because fundwallet waits for the wallet to notice the tx, which takes 60 seconds if not DEVELOPER. Do all the waiting at once, and this speeds the test up to 153 seconds. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2021-06-04 12:47:24 +02:00
# We do a lot of these, so do them all then mine all at once.
addr, txid = l1.fundwallet(10**8, mine_block=False)
l1msgs = ['Owning output .* txid {} CONFIRMED'.format(txid)]
funder: test for adding utxos, calculating total available Make sure that 100% available actually puts in 100% of what we can add (note that this isn't 'all' -- we don't include uneconomic utxos)
2021-05-05 01:53:24 +02:00
# Give l2 lots of utxos
pytest: speed up test_opening.py::test_funder_contribution_limits when !DEVELOPER This test takes 695 seconds, because fundwallet waits for the wallet to notice the tx, which takes 60 seconds if not DEVELOPER. Do all the waiting at once, and this speeds the test up to 153 seconds. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2021-06-04 12:47:24 +02:00
l2msgs = []
for amt in (10**3, # this one is too small to add
10**5, 10**4, 10**4, 10**4, 10**4, 10**4):
addr, txid = l2.fundwallet(amt, mine_block=False)
l2msgs.append('Owning output .* txid {} CONFIRMED'.format(txid))
funder: test for adding utxos, calculating total available Make sure that 100% available actually puts in 100% of what we can add (note that this isn't 'all' -- we don't include uneconomic utxos)
2021-05-05 01:53:24 +02:00
# Give l3 lots of utxos
pytest: speed up test_opening.py::test_funder_contribution_limits when !DEVELOPER This test takes 695 seconds, because fundwallet waits for the wallet to notice the tx, which takes 60 seconds if not DEVELOPER. Do all the waiting at once, and this speeds the test up to 153 seconds. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2021-06-04 12:47:24 +02:00
l3msgs = []
for amt in (10**3, # this one is too small to add
10**4, 10**4, 10**4, 10**4, 10**4, 10**4, 10**4, 10**4, 10**4, 10**4, 10**4):
addr, txid = l3.fundwallet(amt, mine_block=False)
l3msgs.append('Owning output .* txid {} CONFIRMED'.format(txid))
bitcoind.generate_block(1)
l1.daemon.wait_for_logs(l1msgs)
l2.daemon.wait_for_logs(l2msgs)
l3.daemon.wait_for_logs(l3msgs)
funder: test for adding utxos, calculating total available Make sure that 100% available actually puts in 100% of what we can add (note that this isn't 'all' -- we don't include uneconomic utxos)
2021-05-05 01:53:24 +02:00
# Contribute 100% of available funds to l2, all 6 utxos (smallest utxo
# 10**3 is left out)
l2.rpc.call('funderupdate',
{'policy': 'available',
'policy_mod': 100,
funderupdate: add documentation, update some params Documentation for the funderupdate command on the funder plugin. Realized we were missing the "leases_only" in the output; also adds '_msat' to msat denominated outputs so they're parsed correctly by our python bindings (also matches our naming conventions) Changelog-EXPERIMENTAL: funder: `funderupdate` command to view and update params for contributing our wallet funds to v2 channel openings. Provides params for enabling `option_will_fund`.
2021-07-15 18:48:36 +02:00
'min_their_funding_msat': '1000msat',
'per_channel_min_msat': '1000000msat',
funder: test for adding utxos, calculating total available Make sure that 100% available actually puts in 100% of what we can add (note that this isn't 'all' -- we don't include uneconomic utxos)
2021-05-05 01:53:24 +02:00
'fund_probability': 100,
funder: default to only funding leases Make the default to only lease out funds. Changelog-Changed: funder plugin defaults to leases-only
2021-08-05 19:31:49 +02:00
'fuzz_percent': 0,
'leases_only': False})
funder: test for adding utxos, calculating total available Make sure that 100% available actually puts in 100% of what we can add (note that this isn't 'all' -- we don't include uneconomic utxos)
2021-05-05 01:53:24 +02:00
pyln-testing: don't default openchannel and fundwallet to p2sh-segwit, use bech32. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-01-13 01:34:22 +01:00
# Set our contribution to 50k sat, should only use 6 of 12 available utxos
funder: test for adding utxos, calculating total available Make sure that 100% available actually puts in 100% of what we can add (note that this isn't 'all' -- we don't include uneconomic utxos)
2021-05-05 01:53:24 +02:00
l3.rpc.call('funderupdate',
{'policy': 'fixed',
'policy_mod': '50000sat',
funderupdate: add documentation, update some params Documentation for the funderupdate command on the funder plugin. Realized we were missing the "leases_only" in the output; also adds '_msat' to msat denominated outputs so they're parsed correctly by our python bindings (also matches our naming conventions) Changelog-EXPERIMENTAL: funder: `funderupdate` command to view and update params for contributing our wallet funds to v2 channel openings. Provides params for enabling `option_will_fund`.
2021-07-15 18:48:36 +02:00
'min_their_funding_msat': '1000msat',
'per_channel_min_msat': '1000sat',
'per_channel_max_msat': '500000sat',
funder: test for adding utxos, calculating total available Make sure that 100% available actually puts in 100% of what we can add (note that this isn't 'all' -- we don't include uneconomic utxos)
2021-05-05 01:53:24 +02:00
'fund_probability': 100,
funder: default to only funding leases Make the default to only lease out funds. Changelog-Changed: funder plugin defaults to leases-only
2021-08-05 19:31:49 +02:00
'fuzz_percent': 0,
'leases_only': False})
funder: test for adding utxos, calculating total available Make sure that 100% available actually puts in 100% of what we can add (note that this isn't 'all' -- we don't include uneconomic utxos)
2021-05-05 01:53:24 +02:00
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l1.fundchannel(l2, 10**7)
pyln-testing: don't default openchannel and fundwallet to p2sh-segwit, use bech32. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-01-13 01:34:22 +01:00
assert l2.daemon.is_in_log('Policy .* returned funding amount of 141780sat')
funder: test for adding utxos, calculating total available Make sure that 100% available actually puts in 100% of what we can add (note that this isn't 'all' -- we don't include uneconomic utxos)
2021-05-05 01:53:24 +02:00
assert l2.daemon.is_in_log(r'calling `signpsbt` .* 6 inputs')
l1.rpc.connect(l3.info['id'], 'localhost', l3.port)
l1.fundchannel(l3, 10**7)
assert l3.daemon.is_in_log('Policy .* returned funding amount of 50000sat')
pyln-testing: don't default openchannel and fundwallet to p2sh-segwit, use bech32. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-01-13 01:34:22 +01:00
assert l3.daemon.is_in_log(r'calling `signpsbt` .* 6 inputs')
pytest: Test the `mindepth` customizations of `fundchannel` and hook Just test that we can customize, and we'll add mindepth=0 support in the next couple of commits.
2022-05-20 15:45:54 +02:00
pytest: test loading lease_chan_max_msat from channel_funding_inflights It crashes with: ``` s32 field doesn't match size: expected 4, actual 8 ``` Reported-by: @zerofeerouting Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-08-10 14:28:29 +02:00
@pytest.mark.openchannel('v2')
def test_inflight_dbload(node_factory, bitcoind):
"""Bad db field access breaks Postgresql on startup with opening leases"""
disconnects = ["@WIRE_COMMITMENT_SIGNED"]
dual-fund: remove anchor assumption for all dual-funded channels Only add the anchor channel_type if it's negotiated separately!
2022-10-19 20:35:45 +02:00
opts = [{'experimental-dual-fund': None, 'dev-no-reconnect': None,
pytest: make dual-funding tests using option_anchors_zero_fee_htlc_tx. They used to force option_anchor_outputs, so switch them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
'may_reconnect': True, 'disconnect': disconnects,
'experimental-anchors': None},
dual-fund: remove anchor assumption for all dual-funded channels Only add the anchor channel_type if it's negotiated separately!
2022-10-19 20:35:45 +02:00
{'experimental-dual-fund': None, 'dev-no-reconnect': None,
'may_reconnect': True, 'funder-policy': 'match',
'funder-policy-mod': 100, 'lease-fee-base-sat': '100sat',
pytest: make dual-funding tests using option_anchors_zero_fee_htlc_tx. They used to force option_anchor_outputs, so switch them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
'lease-fee-basis': 100,
'experimental-anchors': None}]
dual-fund: remove anchor assumption for all dual-funded channels Only add the anchor channel_type if it's negotiated separately!
2022-10-19 20:35:45 +02:00
l1, l2 = node_factory.get_nodes(2, opts=opts)
pytest: test loading lease_chan_max_msat from channel_funding_inflights It crashes with: ``` s32 field doesn't match size: expected 4, actual 8 ``` Reported-by: @zerofeerouting Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-08-10 14:28:29 +02:00
feerate = 2000
amount = 500000
l1.fundwallet(20000000)
l2.fundwallet(20000000)
# l1 leases a channel from l2
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
rates = l1.rpc.dev_queryrates(l2.info['id'], amount, amount)
l1.rpc.fundchannel(l2.info['id'], amount, request_amt=amount,
feerate='{}perkw'.format(feerate),
compact_lease=rates['compact_lease'])
l1.daemon.wait_for_log(r'dev_disconnect: @WIRE_COMMITMENT_SIGNED')
l1.restart()
pytest: Test the `mindepth` customizations of `fundchannel` and hook Just test that we can customize, and we'll add mindepth=0 support in the next couple of commits.
2022-05-20 15:45:54 +02:00
def test_zeroconf_mindepth(bitcoind, node_factory):
"""Check that funder/fundee can customize mindepth.
Zeroconf will use this to set the mindepth to 0, which coupled
with an artificial depth=0 event that will result in an immediate
`channel_ready` being sent.
"""
plugin_path = Path(__file__).parent / "plugins" / "zeroconf-selective.py"
l1, l2 = node_factory.get_nodes(2, opts=[
{},
{
'plugin': str(plugin_path),
'zeroconf-allow': '0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518',
'zeroconf-mindepth': '2',
},
])
# Try to open a mindepth=6 channel
l1.fundwallet(10**6)
l1.connect(l2)
assert (int(l1.rpc.listpeers()['peers'][0]['features'], 16) >> 50) & 0x02 != 0
# Now start the negotiation, l1 should have negotiated zeroconf,
# and use their own mindepth=6, while l2 uses mindepth=2 from the
# plugin
l1.rpc.fundchannel(l2.info['id'], 'all', mindepth=6)
assert l1.db.query('SELECT minimum_depth FROM channels') == [{'minimum_depth': 6}]
assert l2.db.query('SELECT minimum_depth FROM channels') == [{'minimum_depth': 2}]
bitcoind.generate_block(2, wait_for_mempool=1) # Confirm on the l2 side.
doc: update BOLTs to bc86304b4b0af5fd5ce9d24f74e2ebbceb7e2730 This contains the zeroconf stuff, with funding_locked renamed to channel_ready. I change that everywhere, and try to fix up the comments. Also the `alias` field is called `short_channel_id`. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Changed: Protocol: `funding_locked` is now called `channel_ready` as per latest BOLTs.
2022-09-10 04:10:31 +02:00
l2.daemon.wait_for_log(r'peer_out WIRE_CHANNEL_READY')
channeld/dualopend/lightningd: use channel_ready everywhere. This alters the billboard, but that's a human-readable thing so not noted in CHANGELOG. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Changed: JSON-RPC: `listpeers` `status` now refers to "channel ready" rather than "funding locked" (BOLT language change for zeroconf channels) Changelog-Added: JSON-RPC: `channel_opened` notification `channel_ready` flag. Changelog-Deprecated: JSON-RPC: `channel_opened` notification `funding_locked` flag (use `channel_ready`: BOLTs namechange).
2022-09-10 04:11:31 +02:00
# l1 should not be sending channel_ready yet, it is
pytest: Test the `mindepth` customizations of `fundchannel` and hook Just test that we can customize, and we'll add mindepth=0 support in the next couple of commits.
2022-05-20 15:45:54 +02:00
# configured to wait for 6 confirmations.
doc: update BOLTs to bc86304b4b0af5fd5ce9d24f74e2ebbceb7e2730 This contains the zeroconf stuff, with funding_locked renamed to channel_ready. I change that everywhere, and try to fix up the comments. Also the `alias` field is called `short_channel_id`. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Changed: Protocol: `funding_locked` is now called `channel_ready` as per latest BOLTs.
2022-09-10 04:10:31 +02:00
assert not l1.daemon.is_in_log(r'peer_out WIRE_CHANNEL_READY')
pytest: Test the `mindepth` customizations of `fundchannel` and hook Just test that we can customize, and we'll add mindepth=0 support in the next couple of commits.
2022-05-20 15:45:54 +02:00
bitcoind.generate_block(4) # Confirm on the l2 side.
doc: update BOLTs to bc86304b4b0af5fd5ce9d24f74e2ebbceb7e2730 This contains the zeroconf stuff, with funding_locked renamed to channel_ready. I change that everywhere, and try to fix up the comments. Also the `alias` field is called `short_channel_id`. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Changed: Protocol: `funding_locked` is now called `channel_ready` as per latest BOLTs.
2022-09-10 04:10:31 +02:00
l1.daemon.wait_for_log(r'peer_out WIRE_CHANNEL_READY')
pytest: Test the `mindepth` customizations of `fundchannel` and hook Just test that we can customize, and we'll add mindepth=0 support in the next couple of commits.
2022-05-20 15:45:54 +02:00
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
wait_for(lambda: only_one(l1.rpc.listpeerchannels()['channels'])['state'] == "CHANNELD_NORMAL")
wait_for(lambda: only_one(l2.rpc.listpeerchannels()['channels'])['state'] == "CHANNELD_NORMAL")
pytest: Add a test for direct payment over zeroconf channels
2022-05-25 16:14:47 +02:00
def test_zeroconf_open(bitcoind, node_factory):
"""Let's open a zeroconf channel
Just test that both parties opting in results in a channel that is
immediately usable.
"""
plugin_path = Path(__file__).parent / "plugins" / "zeroconf-selective.py"
l1, l2 = node_factory.get_nodes(2, opts=[
{},
{
'plugin': str(plugin_path),
'zeroconf-allow': '0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518'
},
])
# Try to open a mindepth=0 channel
l1.fundwallet(10**6)
l1.connect(l2)
assert (int(l1.rpc.listpeers()['peers'][0]['features'], 16) >> 50) & 0x02 != 0
# Now start the negotiation, l1 should have negotiated zeroconf,
# and use their own mindepth=6, while l2 uses mindepth=2 from the
# plugin
l1.rpc.fundchannel(l2.info['id'], 'all', mindepth=0)
assert l1.db.query('SELECT minimum_depth FROM channels') == [{'minimum_depth': 0}]
assert l2.db.query('SELECT minimum_depth FROM channels') == [{'minimum_depth': 0}]
l1.daemon.wait_for_logs([
doc: update BOLTs to bc86304b4b0af5fd5ce9d24f74e2ebbceb7e2730 This contains the zeroconf stuff, with funding_locked renamed to channel_ready. I change that everywhere, and try to fix up the comments. Also the `alias` field is called `short_channel_id`. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Changed: Protocol: `funding_locked` is now called `channel_ready` as per latest BOLTs.
2022-09-10 04:10:31 +02:00
r'peer_in WIRE_CHANNEL_READY',
pytest: Add a test for direct payment over zeroconf channels
2022-05-25 16:14:47 +02:00
r'Peer told us that they\'ll use alias=[0-9x]+ for this channel',
])
l2.daemon.wait_for_logs([
doc: update BOLTs to bc86304b4b0af5fd5ce9d24f74e2ebbceb7e2730 This contains the zeroconf stuff, with funding_locked renamed to channel_ready. I change that everywhere, and try to fix up the comments. Also the `alias` field is called `short_channel_id`. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Changed: Protocol: `funding_locked` is now called `channel_ready` as per latest BOLTs.
2022-09-10 04:10:31 +02:00
r'peer_in WIRE_CHANNEL_READY',
pytest: Add a test for direct payment over zeroconf channels
2022-05-25 16:14:47 +02:00
r'Peer told us that they\'ll use alias=[0-9x]+ for this channel',
])
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
wait_for(lambda: only_one(l1.rpc.listpeerchannels()['channels'])['state'] == 'CHANNELD_NORMAL')
wait_for(lambda: only_one(l2.rpc.listpeerchannels()['channels'])['state'] == 'CHANNELD_NORMAL')
pytest: Add a test for direct payment over zeroconf channels
2022-05-25 16:14:47 +02:00
wait_for(lambda: l2.rpc.listincoming()['incoming'] != [])
inv = l2.rpc.invoice(10**8, 'lbl', 'desc')['bolt11']
details = l1.rpc.decodepay(inv)
pprint(details)
assert('routes' in details and len(details['routes']) == 1)
hop = details['routes'][0][0] # First (and only) hop of hint 0
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
l1alias = only_one(l1.rpc.listpeerchannels()['channels'])['alias']['local']
pytest: Add a test for direct payment over zeroconf channels
2022-05-25 16:14:47 +02:00
assert(hop['pubkey'] == l1.info['id']) # l1 is the entrypoint
assert(hop['short_channel_id'] == l1alias) # Alias has to make sense to entrypoint
l1.rpc.pay(inv)
# Ensure lightningd knows about the balance change before
# attempting the other way around.
l2.daemon.wait_for_log(r'Balance [0-9]+msat -> [0-9]+msat')
# Inverse payments should work too
pprint(l2.rpc.listpeers())
inv = l1.rpc.invoice(10**5, 'lbl', 'desc')['bolt11']
l2.rpc.pay(inv)
bkpr: for zerconfs, we still wanna know you're opening a channel We need a record of the channel account before you start sending payments through it. Normally we don't start allowing payments to be sent until after the channel has locked in but zeroconf does away with this assumption. Instead we push out a "channel_proposed" event, which should only show up for zeroconfs.
2022-07-19 21:35:56 +02:00
def test_zeroconf_public(bitcoind, node_factory, chainparams):
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
"""Test that we transition correctly from zeroconf to public
The differences being that a public channel MUST use the public
scid. l1 and l2 open a zeroconf channel, then l3 learns about it
after 6 confirmations.
"""
plugin_path = Path(__file__).parent / "plugins" / "zeroconf-selective.py"
bkpr: for zerconfs, we still wanna know you're opening a channel We need a record of the channel account before you start sending payments through it. Normally we don't start allowing payments to be sent until after the channel has locked in but zeroconf does away with this assumption. Instead we push out a "channel_proposed" event, which should only show up for zeroconfs.
2022-07-19 21:35:56 +02:00
coin_mvt_plugin = Path(__file__).parent / "plugins" / "coin_movements.py"
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
l1, l2, l3 = node_factory.get_nodes(3, opts=[
bkpr: for zerconfs, we still wanna know you're opening a channel We need a record of the channel account before you start sending payments through it. Normally we don't start allowing payments to be sent until after the channel has locked in but zeroconf does away with this assumption. Instead we push out a "channel_proposed" event, which should only show up for zeroconfs.
2022-07-19 21:35:56 +02:00
{'plugin': str(coin_mvt_plugin)},
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
{
'plugin': str(plugin_path),
'zeroconf-allow': '0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518'
},
{}
])
bkpr-zeroconf: Zeroconfs will emit 'channel_proposed' event Keep the accounts as an 'append only' log, instead we move the marker for the 'channel_open' forward when a 'channel_open' comes out. We also neatly hide the 'channel_proposed' events in 'inspect' if there's a 'channel_open' for that same event. If you call inspect before the 'channel_open' is confirmed, you'll see the tag as 'channel_proposed', afterwards it shows up as 'channel_open'. However the event log rolls forward -- listaccountevents will show the correct history of the proposal then open confirming (plus any routing that happened before the channel confirmed).
2022-07-27 23:18:05 +02:00
# Advances blockheight to 102
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
l1.fundwallet(10**6)
bkpr-zeroconf: Zeroconfs will emit 'channel_proposed' event Keep the accounts as an 'append only' log, instead we move the marker for the 'channel_open' forward when a 'channel_open' comes out. We also neatly hide the 'channel_proposed' events in 'inspect' if there's a 'channel_open' for that same event. If you call inspect before the 'channel_open' is confirmed, you'll see the tag as 'channel_proposed', afterwards it shows up as 'channel_open'. However the event log rolls forward -- listaccountevents will show the correct history of the proposal then open confirming (plus any routing that happened before the channel confirmed).
2022-07-27 23:18:05 +02:00
push_msat = 20000 * 1000
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
l1.connect(l2)
bkpr-zeroconf: Zeroconfs will emit 'channel_proposed' event Keep the accounts as an 'append only' log, instead we move the marker for the 'channel_open' forward when a 'channel_open' comes out. We also neatly hide the 'channel_proposed' events in 'inspect' if there's a 'channel_open' for that same event. If you call inspect before the 'channel_open' is confirmed, you'll see the tag as 'channel_proposed', afterwards it shows up as 'channel_open'. However the event log rolls forward -- listaccountevents will show the correct history of the proposal then open confirming (plus any routing that happened before the channel confirmed).
2022-07-27 23:18:05 +02:00
l1.rpc.fundchannel(l2.info['id'], 'all', mindepth=0, push_msat=push_msat)
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
# Wait for the update to be signed (might not be the most reliable
# signal)
l1.daemon.wait_for_log(r'Got WIRE_HSMD_CUPDATE_SIG_REQ')
l2.daemon.wait_for_log(r'Got WIRE_HSMD_CUPDATE_SIG_REQ')
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
l1chan = only_one(l1.rpc.listpeerchannels()['channels'])
l2chan = only_one(l2.rpc.listpeerchannels()['channels'])
bkpr-zeroconf: Zeroconfs will emit 'channel_proposed' event Keep the accounts as an 'append only' log, instead we move the marker for the 'channel_open' forward when a 'channel_open' comes out. We also neatly hide the 'channel_proposed' events in 'inspect' if there's a 'channel_open' for that same event. If you call inspect before the 'channel_open' is confirmed, you'll see the tag as 'channel_proposed', afterwards it shows up as 'channel_open'. However the event log rolls forward -- listaccountevents will show the correct history of the proposal then open confirming (plus any routing that happened before the channel confirmed).
2022-07-27 23:18:05 +02:00
channel_id = l1chan['channel_id']
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
# We have no confirmation yet, so no `short_channel_id`
assert('short_channel_id' not in l1chan)
assert('short_channel_id' not in l2chan)
bkpr: for zerconfs, we still wanna know you're opening a channel We need a record of the channel account before you start sending payments through it. Normally we don't start allowing payments to be sent until after the channel has locked in but zeroconf does away with this assumption. Instead we push out a "channel_proposed" event, which should only show up for zeroconfs.
2022-07-19 21:35:56 +02:00
# Channel is "proposed"
pyln-testing: don't default openchannel and fundwallet to p2sh-segwit, use bech32. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-01-13 01:34:22 +01:00
chan_val = 993888000 if chainparams['elements'] else 996363000
bkpr: for zerconfs, we still wanna know you're opening a channel We need a record of the channel account before you start sending payments through it. Normally we don't start allowing payments to be sent until after the channel has locked in but zeroconf does away with this assumption. Instead we push out a "channel_proposed" event, which should only show up for zeroconfs.
2022-07-19 21:35:56 +02:00
l1_mvts = [
{'type': 'chain_mvt', 'credit_msat': chan_val, 'debit_msat': 0, 'tags': ['channel_proposed', 'opener']},
bkpr-zeroconf: Zeroconfs will emit 'channel_proposed' event Keep the accounts as an 'append only' log, instead we move the marker for the 'channel_open' forward when a 'channel_open' comes out. We also neatly hide the 'channel_proposed' events in 'inspect' if there's a 'channel_open' for that same event. If you call inspect before the 'channel_open' is confirmed, you'll see the tag as 'channel_proposed', afterwards it shows up as 'channel_open'. However the event log rolls forward -- listaccountevents will show the correct history of the proposal then open confirming (plus any routing that happened before the channel confirmed).
2022-07-27 23:18:05 +02:00
{'type': 'channel_mvt', 'credit_msat': 0, 'debit_msat': 20000000, 'tags': ['pushed'], 'fees_msat': '0msat'},
bkpr: for zerconfs, we still wanna know you're opening a channel We need a record of the channel account before you start sending payments through it. Normally we don't start allowing payments to be sent until after the channel has locked in but zeroconf does away with this assumption. Instead we push out a "channel_proposed" event, which should only show up for zeroconfs.
2022-07-19 21:35:56 +02:00
]
check_coin_moves(l1, l1chan['channel_id'], l1_mvts, chainparams)
bkpr-zeroconf: Zeroconfs will emit 'channel_proposed' event Keep the accounts as an 'append only' log, instead we move the marker for the 'channel_open' forward when a 'channel_open' comes out. We also neatly hide the 'channel_proposed' events in 'inspect' if there's a 'channel_open' for that same event. If you call inspect before the 'channel_open' is confirmed, you'll see the tag as 'channel_proposed', afterwards it shows up as 'channel_open'. However the event log rolls forward -- listaccountevents will show the correct history of the proposal then open confirming (plus any routing that happened before the channel confirmed).
2022-07-27 23:18:05 +02:00
# Check that the channel_open event has blockheight of zero
for n in [l1, l2]:
evs = n.rpc.bkpr_listaccountevents(channel_id)['events']
open_ev = only_one([e for e in evs if e['tag'] == 'channel_proposed'])
assert open_ev['blockheight'] == 0
# Call inspect, should have pending event in it
tx = only_one(n.rpc.bkpr_inspect(channel_id)['txs'])
assert 'blockheight' not in tx
assert only_one(tx['outputs'])['output_tag'] == 'channel_proposed'
# Now add 1 confirmation, we should get a `short_channel_id` (block 103)
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
bitcoind.generate_block(1)
l1.daemon.wait_for_log(r'Funding tx [a-f0-9]{64} depth 1 of 0')
l2.daemon.wait_for_log(r'Funding tx [a-f0-9]{64} depth 1 of 0')
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
l1chan = only_one(l1.rpc.listpeerchannels()['channels'])
l2chan = only_one(l2.rpc.listpeerchannels()['channels'])
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
assert('short_channel_id' in l1chan)
assert('short_channel_id' in l2chan)
bkpr-zeroconf: Zeroconfs will emit 'channel_proposed' event Keep the accounts as an 'append only' log, instead we move the marker for the 'channel_open' forward when a 'channel_open' comes out. We also neatly hide the 'channel_proposed' events in 'inspect' if there's a 'channel_open' for that same event. If you call inspect before the 'channel_open' is confirmed, you'll see the tag as 'channel_proposed', afterwards it shows up as 'channel_open'. However the event log rolls forward -- listaccountevents will show the correct history of the proposal then open confirming (plus any routing that happened before the channel confirmed).
2022-07-27 23:18:05 +02:00
# We also now have an 'open' event, the push event isn't re-recorded
bkpr: for zerconfs, we still wanna know you're opening a channel We need a record of the channel account before you start sending payments through it. Normally we don't start allowing payments to be sent until after the channel has locked in but zeroconf does away with this assumption. Instead we push out a "channel_proposed" event, which should only show up for zeroconfs.
2022-07-19 21:35:56 +02:00
l1_mvts += [
{'type': 'chain_mvt', 'credit_msat': chan_val, 'debit_msat': 0, 'tags': ['channel_open', 'opener']},
]
bkpr-zeroconf: Zeroconfs will emit 'channel_proposed' event Keep the accounts as an 'append only' log, instead we move the marker for the 'channel_open' forward when a 'channel_open' comes out. We also neatly hide the 'channel_proposed' events in 'inspect' if there's a 'channel_open' for that same event. If you call inspect before the 'channel_open' is confirmed, you'll see the tag as 'channel_proposed', afterwards it shows up as 'channel_open'. However the event log rolls forward -- listaccountevents will show the correct history of the proposal then open confirming (plus any routing that happened before the channel confirmed).
2022-07-27 23:18:05 +02:00
check_coin_moves(l1, channel_id, l1_mvts, chainparams)
# Check that there is a channel_open event w/ real blockheight
for n in [l1, l2]:
evs = n.rpc.bkpr_listaccountevents(channel_id)['events']
# Still has the channel-proposed event
only_one([e for e in evs if e['tag'] == 'channel_proposed'])
open_ev = only_one([e for e in evs if e['tag'] == 'channel_open'])
assert open_ev['blockheight'] == 103
# Call inspect, should have open event in it
tx = only_one(n.rpc.bkpr_inspect(channel_id)['txs'])
assert tx['blockheight'] == 103
assert only_one(tx['outputs'])['output_tag'] == 'channel_open'
bkpr: for zerconfs, we still wanna know you're opening a channel We need a record of the channel account before you start sending payments through it. Normally we don't start allowing payments to be sent until after the channel has locked in but zeroconf does away with this assumption. Instead we push out a "channel_proposed" event, which should only show up for zeroconfs.
2022-07-19 21:35:56 +02:00
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
# Now make it public, we should be switching over to the real
# scid.
bitcoind.generate_block(5)
# Wait for l3 to learn about the channel, it'll have checked the
# funding outpoint, scripts, etc.
zeroconf: Announce the channel with the real scid as well as aliases With zeroconf we have to duplicate the `local_channel_announcement` since we locally announce the aliased version, and then on the first confirmation we also add the funding scid version.
2022-06-06 15:23:35 +02:00
l3.connect(l1)
pytest: Add test for zeroconf channels transitioning to be public We need to switch to the real short_channel_id at 6 confirmations, and gossip messages should reflect that in order to be valid to others.
2022-05-25 17:04:25 +02:00
wait_for(lambda: len(l3.rpc.listchannels()['channels']) == 2)
bkpr-zeroconf: Zeroconfs will emit 'channel_proposed' event Keep the accounts as an 'append only' log, instead we move the marker for the 'channel_open' forward when a 'channel_open' comes out. We also neatly hide the 'channel_proposed' events in 'inspect' if there's a 'channel_open' for that same event. If you call inspect before the 'channel_open' is confirmed, you'll see the tag as 'channel_proposed', afterwards it shows up as 'channel_open'. However the event log rolls forward -- listaccountevents will show the correct history of the proposal then open confirming (plus any routing that happened before the channel confirmed).
2022-07-27 23:18:05 +02:00
# Close the zerconf channel, check that we mark it as onchain_resolved ok
l1.rpc.close(l2.info['id'])
bitcoind.generate_block(1, wait_for_mempool=1)
# Channel should be marked resolved
for n in [l1, l2]:
wait_for(lambda: only_one([x for x in n.rpc.bkpr_listbalances()['accounts'] if x['account'] == channel_id])['account_resolved'])
pytest: Add a test for direct payment over zeroconf channels
2022-05-25 16:14:47 +02:00
pytest: Add test for forwarding over zeroconf channels
2022-05-25 17:03:31 +02:00
def test_zeroconf_forward(node_factory, bitcoind):
"""Ensure that we can use zeroconf channels in forwards.
Test that we add routehints using the zeroconf channel, and then
ensure that l2 uses the alias from the routehint to forward the
payment. Then do the inverse by sending from l3 to l1, first hop
being the zeroconf channel
"""
plugin_path = Path(__file__).parent / "plugins" / "zeroconf-selective.py"
opts = [
{},
{},
{
'plugin': str(plugin_path),
'zeroconf-allow': '022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59'
}
]
l1, l2, l3 = node_factory.get_nodes(3, opts=opts)
l1.connect(l2)
l1.fundchannel(l2, 10**6)
bitcoind.generate_block(6)
l2.connect(l3)
l2.fundwallet(10**7)
l2.rpc.fundchannel(l3.info['id'], 10**6, mindepth=0)
wait_for(lambda: l3.rpc.listincoming()['incoming'] != [])
pytest: fix flake in test_zeroconf_forward pay failed (non-DEVELOPER) because one node didn't see blocks in time: ``` inv = l3.rpc.invoice(42 * 10**6, 'inv1', 'desc')['bolt11'] > l1.rpc.pay(inv) tests/test_opening.py:1394: ... > raise RpcError(method, payload, resp['error']) E pyln.client.lightning.RpcError: RPC call failed: method: pay, payload: {'bolt11': 'lnbcrt420u1p3dnwv7sp5qnquuwndgz35ywfg3p3dtu07ywmju78r8s0379eaxjxkv5d8jueqpp5kmaxgsye02dzmdlkqkedqvrh2evdl45sz7njrm5dff42dvp4v5qsdq8v3jhxccxqyjw5qcqp9rzjqgkjyd3q5dv6gllh77kygly9c3kfy0d9xwyjyxsq2nq3c83u5vw4n0wkf0y9gwfwhgqqqqqpqqqqqzsqqc9qyysgqad2x2zv0axa3hrfz7nurw4plvspvxlld9wtcg3xxjyxqlzm773a4fkyl09gs8uskj4m7len8r4pf4rh7v9snh3grrpawhk9qsd7vwmcqa9rgxg'}, error: {'code': 210, 'message': 'Ran out of routes to try after 176 attempts: see `paystatus`', 'attempts': [{'status': 'pending', 'partid': 1, 'amount_msat': 42000000msat}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 2, 'amount_msat': 9278783msat, 'parent_partid': 1}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 10, 'amount_msat': 9278783msat, 'parent_partid': 2}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 15, 'amount_msat': 9278783msat, 'parent_partid': 10}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 18, 'amount_msat': 9278783msat, 'parent_partid': 15}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 23, 'amount_msat': 9278783msat, 'parent_partid': 18}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 29, 'amount_msat': 9278783msat, 'parent_partid': 23}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 33, 'amount_msat': 9278783msat, 'parent_partid': 29}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 39, 'amount_msat': 9278783msat, 'parent_partid': 33}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 43, 'amount_msat': 9278783msat, 'parent_partid': 39}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 48, 'amount_msat': 9278783msat, 'parent_partid': 43}, {'status': 'pending', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 54, 'amount_msat': 9278783msat, 'parent_partid': 48}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 59, 'amount_msat': 4659837msat, 'parent_partid': 54}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 69, 'amount_msat': 4659837msat, 'parent_partid': 59}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 82, 'amount_msat': 4659837msat, 'parent_partid': 69}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 92, 'amount_msat': 4659837msat, 'parent_partid': 82}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 102, 'amount_msat': 4659837msat, 'parent_partid': 92}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 112, 'amount_msat': 4659837msat, 'parent_partid': 102}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 122, 'amount_msat': 4659837msat, 'parent_partid': 112}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 131, 'amount_msat': 4659837msat, 'parent_partid': 122}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 141, 'amount_msat': 4659837msat, 'parent_partid': 131}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 147, 'amount_msat': 4659837msat, 'parent_partid': 141}, {'status': 'pending', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 158, 'amount_msat': 4659837msat, 'parent_partid': 147}, {'status': 'pending', 'partid': 175, 'amount_msat': 2250620msat, 'parent_partid': 158}, {'status': 'pending', 'partid': 176, 'amount_msat': 2409217msat, 'parent_partid': 158}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 60, 'amount_msat': 4618946msat, 'parent_partid': 54}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 70, 'amount_msat': 4618946msat, 'parent_partid': 60}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 81, 'amount_msat': 4618946msat, 'parent_partid': 70}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 91, 'amount_msat': 4618946msat, 'parent_partid': 81}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 101, 'amount_msat': 4618946msat, 'parent_partid': 91}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 111, 'amount_msat': 4618946msat, 'parent_partid': 101}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 121, 'amount_msat': 4618946msat, 'parent_partid': 111}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 132, 'amount_msat': 4618946msat, 'parent_partid': 121}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 142, 'amount_msat': 4618946msat, 'parent_partid': 132}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 148, 'amount_msat': 4618946msat, 'parent_partid': 142}, {'status': 'pending', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 157, 'amount_msat': 4618946msat, 'parent_partid': 148}, {'status': 'pending', 'partid': 168, 'amount_msat': 2320055msat, 'parent_partid': 157}, {'status': 'pending', 'partid': 169, 'amount_msat': 2298891msat, 'parent_partid': 157}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 3, 'amount_msat': 9016551msat, 'parent_partid': 1}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 7, 'amount_msat': 9016551msat, 'parent_partid': 3}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 12, 'amount_msat': 9016551msat, 'parent_partid': 7}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 17, 'amount_msat': 9016551msat, 'parent_partid': 12}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 21, 'amount_msat': 9016551msat, 'parent_partid': 17}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 25, 'amount_msat': 9016551msat, 'parent_partid': 21}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 30, 'amount_msat': 9016551msat, 'parent_partid': 25}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 36, 'amount_msat': 9016551msat, 'parent_partid': 30}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 40, 'amount_msat': 9016551msat, 'parent_partid': 36}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 47, 'amount_msat': 9016551msat, 'parent_partid': 40}, {'status': 'pending', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 51, 'amount_msat': 9016551msat, 'parent_partid': 47}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 56, 'amount_msat': 4458645msat, 'parent_partid': 51}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 63, 'amount_msat': 4458645msat, 'parent_partid': 56}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 77, 'amount_msat': 4458645msat, 'parent_partid': 63}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 87, 'amount_msat': 4458645msat, 'parent_partid': 77}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 94, 'amount_msat': 4458645msat, 'parent_partid': 87}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 105, 'amount_msat': 4458645msat, 'parent_partid': 94}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 120, 'amount_msat': 4458645msat, 'parent_partid': 105}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 124, 'amount_msat': 4458645msat, 'parent_partid': 120}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 134, 'amount_msat': 4458645msat, 'parent_partid': 124}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 143, 'amount_msat': 4458645msat, 'parent_partid': 134}, {'status': 'pending', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 154, 'amount_msat': 4458645msat, 'parent_partid': 143}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 164, 'amount_msat': 2306527msat, 'parent_partid': 154}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 165, 'amount_msat': 2152118msat, 'parent_partid': 154}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 57, 'amount_msat': 4557906msat, 'parent_partid': 51}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 68, 'amount_msat': 4557906msat, 'parent_partid': 57}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 76, 'amount_msat': 4557906msat, 'parent_partid': 68}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 86, 'amount_msat': 4557906msat, 'parent_partid': 76}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 96, 'amount_msat': 4557906msat, 'parent_partid': 86}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 103, 'amount_msat': 4557906msat, 'parent_partid': 96}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 115, 'amount_msat': 4557906msat, 'parent_partid': 103}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 125, 'amount_msat': 4557906msat, 'parent_partid': 115}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 136, 'amount_msat': 4557906msat, 'parent_partid': 125}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 146, 'amount_msat': 4557906msat, 'parent_partid': 136}, {'status': 'pending', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 161, 'amount_msat': 4557906msat, 'parent_partid': 146}, {'status': 'pending', 'partid': 173, 'amount_msat': 2208152msat, 'parent_partid': 161}, {'status': 'pending', 'partid': 174, 'amount_msat': 2349754msat, 'parent_partid': 161}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 4, 'amount_msat': 10655305msat, 'parent_partid': 1}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 9, 'amount_msat': 10655305msat, 'parent_partid': 4}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 13, 'amount_msat': 10655305msat, 'parent_partid': 9}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 19, 'amount_msat': 10655305msat, 'parent_partid': 13}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 24, 'amount_msat': 10655305msat, 'parent_partid': 19}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 28, 'amount_msat': 10655305msat, 'parent_partid': 24}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 34, 'amount_msat': 10655305msat, 'parent_partid': 28}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 38, 'amount_msat': 10655305msat, 'parent_partid': 34}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 44, 'amount_msat': 10655305msat, 'parent_partid': 38}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 49, 'amount_msat': 10655305msat, 'parent_partid': 44}, {'status': 'pending', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 53, 'amount_msat': 10655305msat, 'parent_partid': 49}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 61, 'amount_msat': 4872267msat, 'parent_partid': 53}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 71, 'amount_msat': 4872267msat, 'parent_partid': 61}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 79, 'amount_msat': 4872267msat, 'parent_partid': 71}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 89, 'amount_msat': 4872267msat, 'parent_partid': 79}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 100, 'amount_msat': 4872267msat, 'parent_partid': 89}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 109, 'amount_msat': 4872267msat, 'parent_partid': 100}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 114, 'amount_msat': 4872267msat, 'parent_partid': 109}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 126, 'amount_msat': 4872267msat, 'parent_partid': 114}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 139, 'amount_msat': 4872267msat, 'parent_partid': 126}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 149, 'amount_msat': 4872267msat, 'parent_partid': 139}, {'status': 'failed', 'failreason': 'Cannot split payment any further without exceeding the maximum number of HTLCs allowed by our channels', 'partid': 162, 'amount_msat': 4872267msat, 'parent_partid': 149}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 62, 'amount_msat': 5783038msat, 'parent_partid': 53}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 72, 'amount_msat': 5783038msat, 'parent_partid': 62}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 80, 'amount_msat': 5783038msat, 'parent_partid': 72}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 90, 'amount_msat': 5783038msat, 'parent_partid': 80}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 99, 'amount_msat': 5783038msat, 'parent_partid': 90}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 110, 'amount_msat': 5783038msat, 'parent_partid': 99}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 117, 'amount_msat': 5783038msat, 'parent_partid': 110}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 128, 'amount_msat': 5783038msat, 'parent_partid': 117}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 138, 'amount_msat': 5783038msat, 'parent_partid': 128}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 150, 'amount_msat': 5783038msat, 'parent_partid': 138}, {'status': 'pending', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 156, 'amount_msat': 5783038msat, 'parent_partid': 150}, {'status': 'pending', 'partid': 170, 'amount_msat': 2902801msat, 'parent_partid': 156}, {'status': 'pending', 'partid': 171, 'amount_msat': 2880237msat, 'parent_partid': 156}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 5, 'amount_msat': 8942693msat, 'parent_partid': 1}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 8, 'amount_msat': 8942693msat, 'parent_partid': 5}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 16, 'amount_msat': 8942693msat, 'parent_partid': 8}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 20, 'amount_msat': 8942693msat, 'parent_partid': 16}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 27, 'amount_msat': 8942693msat, 'parent_partid': 20}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 32, 'amount_msat': 8942693msat, 'parent_partid': 27}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 35, 'amount_msat': 8942693msat, 'parent_partid': 32}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 42, 'amount_msat': 8942693msat, 'parent_partid': 35}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 45, 'amount_msat': 8942693msat, 'parent_partid': 42}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 50, 'amount_msat': 8942693msat, 'parent_partid': 45}, {'status': 'pending', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 58, 'amount_msat': 8942693msat, 'parent_partid': 50}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 64, 'amount_msat': 4159394msat, 'parent_partid': 58}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 78, 'amount_msat': 4159394msat, 'parent_partid': 64}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 84, 'amount_msat': 4159394msat, 'parent_partid': 78}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 98, 'amount_msat': 4159394msat, 'parent_partid': 84}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 108, 'amount_msat': 4159394msat, 'parent_partid': 98}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 116, 'amount_msat': 4159394msat, 'parent_partid': 108}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 123, 'amount_msat': 4159394msat, 'parent_partid': 116}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 135, 'amount_msat': 4159394msat, 'parent_partid': 123}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 144, 'amount_msat': 4159394msat, 'parent_partid': 135}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 155, 'amount_msat': 4159394msat, 'parent_partid': 144}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 163, 'amount_msat': 4159394msat, 'parent_partid': 155}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 65, 'amount_msat': 4783299msat, 'parent_partid': 58}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 73, 'amount_msat': 4783299msat, 'parent_partid': 65}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 83, 'amount_msat': 4783299msat, 'parent_partid': 73}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 93, 'amount_msat': 4783299msat, 'parent_partid': 83}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 104, 'amount_msat': 4783299msat, 'parent_partid': 93}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 119, 'amount_msat': 4783299msat, 'parent_partid': 104}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 127, 'amount_msat': 4783299msat, 'parent_partid': 119}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 140, 'amount_msat': 4783299msat, 'parent_partid': 127}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 151, 'amount_msat': 4783299msat, 'parent_partid': 140}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 160, 'amount_msat': 4783299msat, 'parent_partid': 151}, {'status': 'pending', 'partid': 172, 'amount_msat': 4783299msat, 'parent_partid': 160}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 6, 'amount_msat': 4106668msat, 'parent_partid': 1}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 11, 'amount_msat': 4106668msat, 'parent_partid': 6}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 14, 'amount_msat': 4106668msat, 'parent_partid': 11}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 22, 'amount_msat': 4106668msat, 'parent_partid': 14}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 26, 'amount_msat': 4106668msat, 'parent_partid': 22}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 31, 'amount_msat': 4106668msat, 'parent_partid': 26}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 37, 'amount_msat': 4106668msat, 'parent_partid': 31}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 41, 'amount_msat': 4106668msat, 'parent_partid': 37}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 46, 'amount_msat': 4106668msat, 'parent_partid': 41}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 52, 'amount_msat': 4106668msat, 'parent_partid': 46}, {'status': 'pending', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 55, 'amount_msat': 4106668msat, 'parent_partid': 52}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 66, 'amount_msat': 2165538msat, 'parent_partid': 55}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 75, 'amount_msat': 2165538msat, 'parent_partid': 66}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 85, 'amount_msat': 2165538msat, 'parent_partid': 75}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 95, 'amount_msat': 2165538msat, 'parent_partid': 85}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 107, 'amount_msat': 2165538msat, 'parent_partid': 95}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 113, 'amount_msat': 2165538msat, 'parent_partid': 107}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 130, 'amount_msat': 2165538msat, 'parent_partid': 113}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 137, 'amount_msat': 2165538msat, 'parent_partid': 130}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 152, 'amount_msat': 2165538msat, 'parent_partid': 137}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 159, 'amount_msat': 2165538msat, 'parent_partid': 152}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 167, 'amount_msat': 2165538msat, 'parent_partid': 159}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 67, 'amount_msat': 1941130msat, 'parent_partid': 55}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 74, 'amount_msat': 1941130msat, 'parent_partid': 67}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 88, 'amount_msat': 1941130msat, 'parent_partid': 74}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 97, 'amount_msat': 1941130msat, 'parent_partid': 88}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 106, 'amount_msat': 1941130msat, 'parent_partid': 97}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 118, 'amount_msat': 1941130msat, 'parent_partid': 106}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 129, 'amount_msat': 1941130msat, 'parent_partid': 118}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 133, 'amount_msat': 1941130msat, 'parent_partid': 129}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 145, 'amount_msat': 1941130msat, 'parent_partid': 133}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 153, 'amount_msat': 1941130msat, 'parent_partid': 145}, {'status': 'failed', 'failreason': 'failed: WIRE_EXPIRY_TOO_SOON (reply from remote)', 'partid': 166, 'amount_msat': 1941130msat, 'parent_partid': 153}]} ``` Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-07-22 03:44:43 +02:00
# Make sure (esp in non-dev-mode) blockheights agree so we don't WIRE_EXPIRY_TOO_SOON...
sync_blockheight(bitcoind, [l1, l2, l3])
pytest: Add test for forwarding over zeroconf channels
2022-05-25 17:03:31 +02:00
inv = l3.rpc.invoice(42 * 10**6, 'inv1', 'desc')['bolt11']
pytest: Add a test for direct payment over zeroconf channels
2022-05-25 16:14:47 +02:00
l1.rpc.pay(inv)
pytest: Add test for forwarding over zeroconf channels
2022-05-25 17:03:31 +02:00
# And now try the other way around: zeroconf channel first
# followed by a public one.
wait_for(lambda: len(l3.rpc.listchannels()['channels']) == 4)
pytest: fix flake in test_zeroconf_forward Second pay can fail if first is not completely settled: ``` def test_zeroconf_forward(node_factory, bitcoind): """Ensure that we can use zeroconf channels in forwards. ... # Make sure (esp in non-dev-mode) blockheights agree so we don't WIRE_EXPIRY_TOO_SOON... sync_blockheight(bitcoind, [l1, l2, l3]) inv = l3.rpc.invoice(42 * 10**6, 'inv1', 'desc')['bolt11'] l1.rpc.pay(inv) # And now try the other way around: zeroconf channel first # followed by a public one. wait_for(lambda: len(l3.rpc.listchannels()['channels']) == 4) inv = l1.rpc.invoice(42, 'back1', 'desc')['bolt11'] > l3.rpc.pay(inv) ... > raise RpcError(method, payload, resp['error']) E pyln.client.lightning.RpcError: RPC call failed: method: pay, payload: {'bolt11': 'lnbcrt420p1p3junrssp588gtjzmlrr4pfj7ssmdlulzhhushrpq3rdqxjuz2m33scsvzdjlspp5fk5yhu6netc0d0sgp8es52vjk6akhd3uayr08u8max4d8rwzpjuqdq8v3jhxccxqyjw5qcqp99qyysgqcpyyugejv4nya97v6gw8fhtr0ru3vq87jjlltav99wlat436a95n0z8yzdp699p9md0zz9tmnsjpvfj622n9g9fh7r6ldhpgh9wmr4qpcru3rk'}, error: {'code': 210, 'message': 'Destination 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518 is not reachable directly and all routehints were unusable.', 'attempts': [{'status': 'failed', 'failreason': 'Destination 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518 is not reachable directly and all routehints were unusable.', 'partid': 0, 'amount_msat': 42msat}]} ``` Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-09-24 06:27:09 +02:00
# Make sure all htlcs completely settled!
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
wait_for(lambda: (p['htlcs'] == [] for p in l2.rpc.listpeerchannels()['channels']))
pytest: fix flake in test_zeroconf_forward Second pay can fail if first is not completely settled: ``` def test_zeroconf_forward(node_factory, bitcoind): """Ensure that we can use zeroconf channels in forwards. ... # Make sure (esp in non-dev-mode) blockheights agree so we don't WIRE_EXPIRY_TOO_SOON... sync_blockheight(bitcoind, [l1, l2, l3]) inv = l3.rpc.invoice(42 * 10**6, 'inv1', 'desc')['bolt11'] l1.rpc.pay(inv) # And now try the other way around: zeroconf channel first # followed by a public one. wait_for(lambda: len(l3.rpc.listchannels()['channels']) == 4) inv = l1.rpc.invoice(42, 'back1', 'desc')['bolt11'] > l3.rpc.pay(inv) ... > raise RpcError(method, payload, resp['error']) E pyln.client.lightning.RpcError: RPC call failed: method: pay, payload: {'bolt11': 'lnbcrt420p1p3junrssp588gtjzmlrr4pfj7ssmdlulzhhushrpq3rdqxjuz2m33scsvzdjlspp5fk5yhu6netc0d0sgp8es52vjk6akhd3uayr08u8max4d8rwzpjuqdq8v3jhxccxqyjw5qcqp99qyysgqcpyyugejv4nya97v6gw8fhtr0ru3vq87jjlltav99wlat436a95n0z8yzdp699p9md0zz9tmnsjpvfj622n9g9fh7r6ldhpgh9wmr4qpcru3rk'}, error: {'code': 210, 'message': 'Destination 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518 is not reachable directly and all routehints were unusable.', 'attempts': [{'status': 'failed', 'failreason': 'Destination 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518 is not reachable directly and all routehints were unusable.', 'partid': 0, 'amount_msat': 42msat}]} ``` Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-09-24 06:27:09 +02:00
pytest: Add test for forwarding over zeroconf channels
2022-05-25 17:03:31 +02:00
inv = l1.rpc.invoice(42, 'back1', 'desc')['bolt11']
l3.rpc.pay(inv)
dual-fund: Fail if you try to buy a liquidity ad w/o dualfunding on Fixes #5271 In-Collaboration-With: Base58 'n Coding Seminar Participants Changelog-Changed: `fundchannel` now errors if you try to buy a liquidity ad but dont' have `experimental-dual-fund` enabled
2022-07-09 00:07:06 +02:00
@pytest.mark.openchannel('v1')
def test_buy_liquidity_ad_no_v2(node_factory, bitcoind):
""" Test that you can't actually request amt for a
node that doesn' support v2 opens """
l1, l2, = node_factory.get_nodes(2)
amount = 500000
feerate = 2000
l1.fundwallet(amount * 100)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
# l1 leases a channel from l2
with pytest.raises(RpcError, match=r"Tried to buy a liquidity ad but we[(][?][)] don't have experimental-dual-fund enabled"):
l1.rpc.fundchannel(l2.info['id'], amount, request_amt=amount,
feerate='{}perkw'.format(feerate),
compact_lease='029a002d000000004b2003e8')
pytest: test that we implement option_scid_alias privacy. Spoiler: we don't! :( Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-08-08 04:42:52 +02:00
bkpr: confirm that replaying the open+lock-in txs at start is ok Make sure that we're not issuing duplicate lease_fee events!
2022-09-13 18:02:53 +02:00
@pytest.mark.openchannel('v2')
def test_v2_replay_bookkeeping(node_factory, bitcoind):
""" Test that your bookkeeping for a liquidity ad is good
even if we replay the opening and locking tx!
"""
opts = [{'funder-policy': 'match', 'funder-policy-mod': 100,
'lease-fee-base-sat': '100sat', 'lease-fee-basis': 100,
pytest: make dual-funding tests using option_anchors_zero_fee_htlc_tx. They used to force option_anchor_outputs, so switch them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
'rescan': 10, 'funding-confirms': 6, 'may_reconnect': True,
'experimental-anchors': None},
bkpr: confirm that replaying the open+lock-in txs at start is ok Make sure that we're not issuing duplicate lease_fee events!
2022-09-13 18:02:53 +02:00
{'funder-policy': 'match', 'funder-policy-mod': 100,
'lease-fee-base-sat': '100sat', 'lease-fee-basis': 100,
pytest: make dual-funding tests using option_anchors_zero_fee_htlc_tx. They used to force option_anchor_outputs, so switch them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
'may_reconnect': True,
'experimental-anchors': None}]
dual-fund: remove anchor assumption for all dual-funded channels Only add the anchor channel_type if it's negotiated separately!
2022-10-19 20:35:45 +02:00
bkpr: confirm that replaying the open+lock-in txs at start is ok Make sure that we're not issuing duplicate lease_fee events!
2022-09-13 18:02:53 +02:00
l1, l2, = node_factory.get_nodes(2, opts=opts)
amount = 500000
feerate = 2000
l1.fundwallet(amount * 100)
l2.fundwallet(amount * 100)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
rates = l1.rpc.dev_queryrates(l2.info['id'], amount, amount)
# l1 leases a channel from l2
l1.rpc.fundchannel(l2.info['id'], amount, request_amt=amount,
feerate='{}perkw'.format(feerate),
compact_lease=rates['compact_lease'])
# add the funding transaction
bitcoind.generate_block(4, wait_for_mempool=1)
l1.restart()
bitcoind.generate_block(2)
l1.daemon.wait_for_log('to CHANNELD_NORMAL')
chan_id = first_channel_id(l1, l2)
ev_tags = [e['tag'] for e in l1.rpc.bkpr_listaccountevents(chan_id)['events']]
assert 'lease_fee' in ev_tags
# This should work ok
l1.rpc.bkpr_listbalances()
bitcoind.generate_block(2)
sync_blockheight(bitcoind, [l1])
l1.restart()
chan_id = first_channel_id(l1, l2)
ev_tags = [e['tag'] for e in l1.rpc.bkpr_listaccountevents(chan_id)['events']]
assert 'lease_fee' in ev_tags
l1.rpc.close(l2.info['id'], 1)
bitcoind.generate_block(6, wait_for_mempool=1)
l1.daemon.wait_for_log(' to ONCHAIN')
l2.daemon.wait_for_log(' to ONCHAIN')
# This should not crash
l1.rpc.bkpr_listbalances()
bkpr: failing test for bookkeeper crash Reproduce crash for #5557! If we record the channel open because bookkeeper was added after the channel open request started but the channel confirms later, we end up with re-recording any associated push or leased fees (paid or rcvd). In the case where you've paid for these fees, your channel balance goes negative and the node crashes the next time you call `listbalances`. Reported-by: @chrisguida
2022-09-13 01:49:26 +02:00
@pytest.mark.openchannel('v2')
def test_buy_liquidity_ad_check_bookkeeping(node_factory, bitcoind):
""" Test that your bookkeeping for a liquidity ad is good."""
opts = [{'funder-policy': 'match', 'funder-policy-mod': 100,
'lease-fee-base-sat': '100sat', 'lease-fee-basis': 100,
'rescan': 10, 'disable-plugin': 'bookkeeper',
pytest: make dual-funding tests using option_anchors_zero_fee_htlc_tx. They used to force option_anchor_outputs, so switch them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
'funding-confirms': 6, 'may_reconnect': True,
'experimental-anchors': None},
bkpr: failing test for bookkeeper crash Reproduce crash for #5557! If we record the channel open because bookkeeper was added after the channel open request started but the channel confirms later, we end up with re-recording any associated push or leased fees (paid or rcvd). In the case where you've paid for these fees, your channel balance goes negative and the node crashes the next time you call `listbalances`. Reported-by: @chrisguida
2022-09-13 01:49:26 +02:00
{'funder-policy': 'match', 'funder-policy-mod': 100,
notif: dont send balance snapshot for not yet opened channel We were double counting channel lease fees because we were double firing the channel open event sequence (so to speak). If we don't report balances for unopened channels, we don't have this problem? Changelog-Changed: Plugins: `balance_snapshot` notification does not send balances for channels that aren't locked-in/opened yet
2022-09-13 16:53:46 +02:00
'lease-fee-base-sat': '100sat', 'lease-fee-basis': 100,
pytest: make dual-funding tests using option_anchors_zero_fee_htlc_tx. They used to force option_anchor_outputs, so switch them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
'may_reconnect': True,
'experimental-anchors': None}]
dual-fund: remove anchor assumption for all dual-funded channels Only add the anchor channel_type if it's negotiated separately!
2022-10-19 20:35:45 +02:00
bkpr: failing test for bookkeeper crash Reproduce crash for #5557! If we record the channel open because bookkeeper was added after the channel open request started but the channel confirms later, we end up with re-recording any associated push or leased fees (paid or rcvd). In the case where you've paid for these fees, your channel balance goes negative and the node crashes the next time you call `listbalances`. Reported-by: @chrisguida
2022-09-13 01:49:26 +02:00
l1, l2, = node_factory.get_nodes(2, opts=opts)
amount = 500000
feerate = 2000
l1.fundwallet(amount * 100)
l2.fundwallet(amount * 100)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
rates = l1.rpc.dev_queryrates(l2.info['id'], amount, amount)
# l1 leases a channel from l2
l1.rpc.fundchannel(l2.info['id'], amount, request_amt=amount,
feerate='{}perkw'.format(feerate),
compact_lease=rates['compact_lease'])
# add the funding transaction
bitcoind.generate_block(4, wait_for_mempool=1)
l1.stop()
del l1.daemon.opts['disable-plugin']
l1.start()
notif: dont send balance snapshot for not yet opened channel We were double counting channel lease fees because we were double firing the channel open event sequence (so to speak). If we don't report balances for unopened channels, we don't have this problem? Changelog-Changed: Plugins: `balance_snapshot` notification does not send balances for channels that aren't locked-in/opened yet
2022-09-13 16:53:46 +02:00
bitcoind.generate_block(2)
l1.daemon.wait_for_log('to CHANNELD_NORMAL')
bkpr: failing test for bookkeeper crash Reproduce crash for #5557! If we record the channel open because bookkeeper was added after the channel open request started but the channel confirms later, we end up with re-recording any associated push or leased fees (paid or rcvd). In the case where you've paid for these fees, your channel balance goes negative and the node crashes the next time you call `listbalances`. Reported-by: @chrisguida
2022-09-13 01:49:26 +02:00
chan_id = first_channel_id(l1, l2)
ev_tags = [e['tag'] for e in l1.rpc.bkpr_listaccountevents(chan_id)['events']]
assert 'lease_fee' in ev_tags
# This should work ok
l1.rpc.bkpr_listbalances()
l1.rpc.close(l2.info['id'], 1)
bitcoind.generate_block(6, wait_for_mempool=1)
l1.daemon.wait_for_log(' to ONCHAIN')
l2.daemon.wait_for_log(' to ONCHAIN')
notif: dont send balance snapshot for not yet opened channel We were double counting channel lease fees because we were double firing the channel open event sequence (so to speak). If we don't report balances for unopened channels, we don't have this problem? Changelog-Changed: Plugins: `balance_snapshot` notification does not send balances for channels that aren't locked-in/opened yet
2022-09-13 16:53:46 +02:00
# This should not crash
bkpr: failing test for bookkeeper crash Reproduce crash for #5557! If we record the channel open because bookkeeper was added after the channel open request started but the channel confirms later, we end up with re-recording any associated push or leased fees (paid or rcvd). In the case where you've paid for these fees, your channel balance goes negative and the node crashes the next time you call `listbalances`. Reported-by: @chrisguida
2022-09-13 01:49:26 +02:00
l1.rpc.bkpr_listbalances()
pytest: test that we implement option_scid_alias privacy. Spoiler: we don't! :( Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-08-08 04:42:52 +02:00
def test_scid_alias_private(node_factory, bitcoind):
"""Test that we don't allow use of real scid for scid_alias-type channels"""
l1, l2, l3 = node_factory.line_graph(3, fundchannel=False, opts=[{}, {},
{'log-level': 'io'}])
l2.fundwallet(5000000)
l2.rpc.fundchannel(l3.info['id'], 'all', announce=False)
bitcoind.generate_block(1, wait_for_mempool=1)
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
wait_for(lambda: only_one(l2.rpc.listpeerchannels(l3.info['id'])['channels'])['state'] == 'CHANNELD_NORMAL')
pytest: test that we implement option_scid_alias privacy. Spoiler: we don't! :( Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-08-08 04:42:52 +02:00
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
chan = only_one(l2.rpc.listpeerchannels(l3.info['id'])['channels'])
pytest: test that we implement option_scid_alias privacy. Spoiler: we don't! :( Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-08-08 04:42:52 +02:00
assert chan['private'] is True
scid23 = chan['short_channel_id']
alias23 = chan['alias']['local']
# Create l1<->l2 channel, make sure l3 sees it so it will routehint via
# l2 (otherwise it sees it as a deadend!)
l1.fundwallet(5000000)
l1.rpc.fundchannel(l2.info['id'], 'all')
bitcoind.generate_block(6, wait_for_mempool=1)
wait_for(lambda: len(l3.rpc.listchannels(source=l1.info['id'])['channels']) == 1)
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
chan = only_one(l1.rpc.listpeerchannels(l2.info['id'])['channels'])
pytest: test that we implement option_scid_alias privacy. Spoiler: we don't! :( Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-08-08 04:42:52 +02:00
assert chan['private'] is False
scid12 = chan['short_channel_id']
# Make sure it sees both sides of private channel in gossmap!
wait_for(lambda: len(l3.rpc.listchannels()['channels']) == 4)
# BOLT #2:
# - if `channel_type` has `option_scid_alias` set:
# - MUST NOT use the real `short_channel_id` in BOLT 11 `r` fields.
inv = l3.rpc.invoice(10, 'test_scid_alias_private', 'desc')
assert only_one(only_one(l1.rpc.decode(inv['bolt11'])['routes']))['short_channel_id'] == alias23
# BOLT #2:
# - if `channel_type` has `option_scid_alias` set:
# - MUST NOT allow incoming HTLCs to this channel using the real `short_channel_id`
route = [{'amount_msat': 11,
'id': l2.info['id'],
'delay': 12,
'channel': scid12},
{'amount_msat': 10,
'id': l3.info['id'],
'delay': 6,
'channel': scid23}]
l1.rpc.sendpay(route, inv['payment_hash'], payment_secret=inv['payment_secret'])
with pytest.raises(RpcError) as err:
l1.rpc.waitsendpay(inv['payment_hash'])
# PERM|10
WIRE_UNKNOWN_NEXT_PEER = 0x4000 | 10
assert err.value.error['data']['failcode'] == WIRE_UNKNOWN_NEXT_PEER
assert err.value.error['data']['erring_node'] == l2.info['id']
assert err.value.error['data']['erring_channel'] == scid23
# BOLT #2
# - MUST always recognize the `alias` as a `short_channel_id` for incoming HTLCs to this channel.
route[1]['channel'] = alias23
l1.rpc.sendpay(route, inv['payment_hash'], payment_secret=inv['payment_secret'])
l1.rpc.waitsendpay(inv['payment_hash'])
pytest: Reproduce a crash when we have multiple channels and 0conf This is a case where we assume that the `channel->scid` is set, which is no longer true with `zeroconf` Changelog-None: Zeroconf was not yet released at the time the discovery was made Reported-by: Yaacov Akiba Slama <@yaslama> Reported-by: Roei Erez <@roeierez>
2022-08-16 15:32:09 +02:00
def test_zeroconf_multichan_forward(node_factory):
"""The freedom to choose the forward channel bytes us when it is 0conf
Reported by Breez, we crashed when logging in `forward_htlc` when
the replacement channel was a zeroconf channel.
l2 -> l3 is a double channel with the zeroconf channel having a
higher spendable msat, which should cause it to be chosen instead.
"""
node_id = '022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59'
plugin_path = Path(__file__).parent / "plugins" / "zeroconf-selective.py"
l1, l2, l3 = node_factory.line_graph(3, opts=[
{},
{},
{
'plugin': str(plugin_path),
'zeroconf-allow': node_id,
}
], fundamount=10**6, wait_for_announce=True)
# Just making sure the allowlisted node_id matches.
assert l2.info['id'] == node_id
# Now create a channel that is twice as large as the real channel,
# and don't announce it.
l2.fundwallet(10**7)
lightningd: figure out optimal channel *before* forward_htlc hook. Otherwise what the hook sees is actually a lie, and if it sets it we might override it. The side effect is that we add an explicit "forward_to" field, and allow hooks to override it. This lets a *hook* control channel choice explicitly. Changelod-Added: Plugins: `htlc_accepted_hook` return can specify what channel to forward htlc to. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-09-25 15:14:12 +02:00
zeroconf_cid = l2.rpc.fundchannel(l3.info['id'], 2 * 10**6, mindepth=0)['channel_id']
pytest: Reproduce a crash when we have multiple channels and 0conf This is a case where we assume that the `channel->scid` is set, which is no longer true with `zeroconf` Changelog-None: Zeroconf was not yet released at the time the discovery was made Reported-by: Yaacov Akiba Slama <@yaslama> Reported-by: Roei Erez <@roeierez>
2022-08-16 15:32:09 +02:00
doc: update BOLTs to bc86304b4b0af5fd5ce9d24f74e2ebbceb7e2730 This contains the zeroconf stuff, with funding_locked renamed to channel_ready. I change that everywhere, and try to fix up the comments. Also the `alias` field is called `short_channel_id`. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Changed: Protocol: `funding_locked` is now called `channel_ready` as per latest BOLTs.
2022-09-10 04:10:31 +02:00
l2.daemon.wait_for_log(r'peer_in WIRE_CHANNEL_READY')
l3.daemon.wait_for_log(r'peer_in WIRE_CHANNEL_READY')
pytest: Reproduce a crash when we have multiple channels and 0conf This is a case where we assume that the `channel->scid` is set, which is no longer true with `zeroconf` Changelog-None: Zeroconf was not yet released at the time the discovery was made Reported-by: Yaacov Akiba Slama <@yaslama> Reported-by: Roei Erez <@roeierez>
2022-08-16 15:32:09 +02:00
inv = l3.rpc.invoice(amount_msat=10000, label='lbl1', description='desc')['bolt11']
l1.rpc.pay(inv)
lightningd: figure out optimal channel *before* forward_htlc hook. Otherwise what the hook sees is actually a lie, and if it sets it we might override it. The side effect is that we add an explicit "forward_to" field, and allow hooks to override it. This lets a *hook* control channel choice explicitly. Changelod-Added: Plugins: `htlc_accepted_hook` return can specify what channel to forward htlc to. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-09-25 15:14:12 +02:00
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
for c in l2.rpc.listpeerchannels(l3.info['id'])['channels']:
lightningd: figure out optimal channel *before* forward_htlc hook. Otherwise what the hook sees is actually a lie, and if it sets it we might override it. The side effect is that we add an explicit "forward_to" field, and allow hooks to override it. This lets a *hook* control channel choice explicitly. Changelod-Added: Plugins: `htlc_accepted_hook` return can specify what channel to forward htlc to. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-09-25 15:14:12 +02:00
if c['channel_id'] == zeroconf_cid:
zeroconf_scid = c['alias']['local']
else:
normal_scid = c['short_channel_id']
assert l2.daemon.is_in_log(r'Chose a better channel than {}: {}'
.format(normal_scid, zeroconf_scid))
pytest: Add a zeroreserve test
2022-06-09 17:07:03 +02:00
def test_zeroreserve(node_factory, bitcoind):
"""Ensure we can set the reserves.
3 nodes:
- l1 enforces zeroreserve
- l2 enforces default reserve
- l3 enforces sub-dust reserves
"""
plugin_path = Path(__file__).parent / "plugins" / "zeroreserve.py"
opts = [
{
'plugin': str(plugin_path),
'reserve': '0sat',
opening: Add `dev-allowdustreserve` option to opt into dust reserves Technically this is a non-conformance with the spec, hence the `dev` flag to opt-in, however I'm being told that it is also implemented in other implementations. I'll follow this up with a proposal to the spec to remove the checks we now bypass.
2022-06-24 12:58:26 +02:00
'dev-allowdustreserve': True,
},
{
'dev-allowdustreserve': True,
pytest: Add a zeroreserve test
2022-06-09 17:07:03 +02:00
},
{
'plugin': str(plugin_path),
opening: Add `dev-allowdustreserve` option to opt into dust reserves Technically this is a non-conformance with the spec, hence the `dev` flag to opt-in, however I'm being told that it is also implemented in other implementations. I'll follow this up with a proposal to the spec to remove the checks we now bypass.
2022-06-24 12:58:26 +02:00
'reserve': '123sat',
'dev-allowdustreserve': True,
pytest: Add a zeroreserve test
2022-06-09 17:07:03 +02:00
}
]
l1, l2, l3 = node_factory.get_nodes(3, opts=opts)
l1.fundwallet(10**7)
l2.fundwallet(10**7)
l3.fundwallet(10**7)
l1.connect(l2)
l2.connect(l3)
l3.connect(l1)
l1.rpc.fundchannel(l2.info['id'], 10**6, reserve='0sat')
l2.rpc.fundchannel(l3.info['id'], 10**6)
l3.rpc.fundchannel(l1.info['id'], 10**6, reserve='321sat')
bitcoind.generate_block(1, wait_for_mempool=3)
wait_for(lambda: l1.channel_state(l2) == 'CHANNELD_NORMAL')
wait_for(lambda: l2.channel_state(l3) == 'CHANNELD_NORMAL')
wait_for(lambda: l3.channel_state(l1) == 'CHANNELD_NORMAL')
# Now make sure we all agree on each others reserves
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
l1c1 = l1.rpc.listpeerchannels(l2.info['id'])['channels'][0]
l2c1 = l2.rpc.listpeerchannels(l1.info['id'])['channels'][0]
l2c2 = l2.rpc.listpeerchannels(l3.info['id'])['channels'][0]
l3c2 = l3.rpc.listpeerchannels(l2.info['id'])['channels'][0]
l3c3 = l3.rpc.listpeerchannels(l1.info['id'])['channels'][0]
l1c3 = l1.rpc.listpeerchannels(l3.info['id'])['channels'][0]
pytest: Add a zeroreserve test
2022-06-09 17:07:03 +02:00
# l1 imposed a 0sat reserve on l2, while l2 imposed the default 1% reserve on l1
opening: Add `dev-allowdustreserve` option to opt into dust reserves Technically this is a non-conformance with the spec, hence the `dev` flag to opt-in, however I'm being told that it is also implemented in other implementations. I'll follow this up with a proposal to the spec to remove the checks we now bypass.
2022-06-24 12:58:26 +02:00
assert l1c1['their_reserve_msat'] == l2c1['our_reserve_msat'] == Millisatoshi('0sat')
assert l1c1['our_reserve_msat'] == l2c1['their_reserve_msat'] == Millisatoshi('10000sat')
pytest: Add a zeroreserve test
2022-06-09 17:07:03 +02:00
# l2 imposed the default 1% on l3, while l3 imposed a custom 123sat fee on l2
opening: Add `dev-allowdustreserve` option to opt into dust reserves Technically this is a non-conformance with the spec, hence the `dev` flag to opt-in, however I'm being told that it is also implemented in other implementations. I'll follow this up with a proposal to the spec to remove the checks we now bypass.
2022-06-24 12:58:26 +02:00
assert l2c2['their_reserve_msat'] == l3c2['our_reserve_msat'] == Millisatoshi('10000sat')
assert l2c2['our_reserve_msat'] == l3c2['their_reserve_msat'] == Millisatoshi('123sat')
pytest: Add a zeroreserve test
2022-06-09 17:07:03 +02:00
# l3 imposed a custom 321sat fee on l1, while l1 imposed a custom 0sat fee on l3
opening: Add `dev-allowdustreserve` option to opt into dust reserves Technically this is a non-conformance with the spec, hence the `dev` flag to opt-in, however I'm being told that it is also implemented in other implementations. I'll follow this up with a proposal to the spec to remove the checks we now bypass.
2022-06-24 12:58:26 +02:00
assert l3c3['their_reserve_msat'] == l1c3['our_reserve_msat'] == Millisatoshi('321sat')
assert l3c3['our_reserve_msat'] == l1c3['their_reserve_msat'] == Millisatoshi('0sat')
pytest: Add a zeroreserve test
2022-06-09 17:07:03 +02:00
# Now do some drain tests on c1, as that should be drainable
# completely by l2 being the fundee
l1.rpc.keysend(l2.info['id'], 10 * 7) # Something above dust for sure
l2.drain(l1)
# Remember that this is the reserve l1 imposed on l2, so l2 can drain completely
tests: use listpeerchannels.
2023-01-12 02:25:55 +01:00
l2c1 = l2.rpc.listpeerchannels(l1.info['id'])['channels'][0]
pytest: Add a zeroreserve test
2022-06-09 17:07:03 +02:00
# And despite us briefly being above dust (with a to_us output),
# closing should result in the output being trimmed again since we
# dropped below dust again.
c = l2.rpc.close(l1.info['id'])
decoded = bitcoind.rpc.decoderawtransaction(c['tx'])
pytest: Add test for mixed zeroreserve funding Tests that we can have zeroreserve nodes accept and open new channels with non-zeroreserve nodes, without throwing errors
2022-06-24 12:42:45 +02:00
# Elements has a change output always
assert len(decoded['vout']) == 1 if TEST_NETWORK == 'regtest' else 2
def test_zeroreserve_mixed(node_factory, bitcoind):
"""l1 runs with zeroreserve, l2 and l3 without, should still work
Basically tests that l1 doesn't get upset when l2 allows us to
drop below dust.
"""
plugin_path = Path(__file__).parent / "plugins" / "zeroreserve.py"
opts = [
{
'plugin': str(plugin_path),
'reserve': '0sat',
'dev-allowdustreserve': True,
}, {
'dev-allowdustreserve': False,
}, {
'dev-allowdustreserve': False,
}
]
l1, l2, l3 = node_factory.get_nodes(3, opts=opts)
l1.fundwallet(10**7)
l3.fundwallet(10**7)
l1.connect(l2)
l3.connect(l1)
l1.rpc.fundchannel(l2.info['id'], 10**6, reserve='0sat')
l3.rpc.fundchannel(l1.info['id'], 10**6)
pytest: Exercise all dust zeroreserve case This just shows we need to add that constraint. Reported-by: Rusty Russell <@rustyrussell>
2022-08-16 17:55:34 +02:00
def test_zeroreserve_alldust(node_factory):
"""If we allow dust reserves we need larger fundings
This is because we might have up to
allhtlcs = (local.max_concurrent_htlcs + remote.max_concurrent_htlcs)
alldust = allhlcs * min(local.dust, remote.dust)
allocated to HTLCs in flight, reducing both direct outputs to
dust. This could leave us with no outs on the commitment, is
therefore invalid.
Parameters are as follows:
- Regtest:
- max_concurrent_htlcs = 483
- dust = 546sat
- minfunding = (483 * 2 + 2) * 546sat = 528528sat
- Mainnet:
- max_concurrent_htlcs = 30
- dust = 546sat
- minfunding = (30 * 2 + 2) * 546sat = 33852s
"""
plugin_path = Path(__file__).parent / "plugins" / "zeroreserve.py"
l1, l2 = node_factory.get_nodes(2, opts=[{
'plugin': plugin_path,
'reserve': '0sat',
'dev-allowdustreserve': True
}] * 2)
maxhtlc = 483
mindust = 546
minfunding = (maxhtlc * 2 + 2) * mindust
l1.fundwallet(10**6)
error = (f'channel funding {minfunding}sat too small for chosen parameters: '
f'a total of {maxhtlc * 2} HTLCs with dust value {mindust}sat would '
f'result in a commitment_transaction without outputs')
# This is right on the edge, and should fail
with pytest.raises(RpcError, match=error):
l1.connect(l2)
l1.rpc.fundchannel(l2.info['id'], minfunding)
# Now try with just a bit more
l1.connect(l2)
l1.rpc.fundchannel(l2.info['id'], minfunding + 1)
tests: test for coinbase wallet spend. Attempt to spend a coinbase transaction, expected to fail.
2022-10-17 20:10:12 +02:00
def test_coinbase_unspendable(node_factory, bitcoind):
""" A node should not be able to spend a coinbase output
before it's mature """
[l1] = node_factory.get_nodes(1)
addr = l1.rpc.newaddr()["bech32"]
bitcoind.rpc.generatetoaddress(1, addr)
addr2 = l1.rpc.newaddr()["bech32"]
# Wait til money in wallet
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) == 1)
wallet: mark coinbase outputs as 'immature' until spendable Changelog-Changed: JSON-RPC: `listfunds` now lists coinbase outputs as 'immature' until they're spendable Changelog-Changed: JSON-RPC: UTXOs aren't spendable while immature
2022-10-17 21:59:57 +02:00
out = only_one(l1.rpc.listfunds()['outputs'])
assert out['status'] == 'immature'
with pytest.raises(RpcError, match='Could not afford all using all 0 available UTXOs'):
l1.rpc.withdraw(addr2, "all")
tests: test for coinbase wallet spend. Attempt to spend a coinbase transaction, expected to fail.
2022-10-17 20:10:12 +02:00
# Nothing sent to the mempool!
assert len(bitcoind.rpc.getrawmempool()) == 0
wallet: mark coinbase outputs as 'immature' until spendable Changelog-Changed: JSON-RPC: `listfunds` now lists coinbase outputs as 'immature' until they're spendable Changelog-Changed: JSON-RPC: UTXOs aren't spendable while immature
2022-10-17 21:59:57 +02:00
# Mine 98 blocks
bitcoind.rpc.generatetoaddress(98, l1.rpc.newaddr()['bech32'])
assert len([out for out in l1.rpc.listfunds()['outputs'] if out['status'] == 'confirmed']) == 0
with pytest.raises(RpcError, match='Could not afford all using all 0 available UTXOs'):
l1.rpc.withdraw(addr2, "all")
# One more and the first coinbase unlocks
bitcoind.rpc.generatetoaddress(1, l1.rpc.newaddr()['bech32'])
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) == 100)
assert len([out for out in l1.rpc.listfunds()['outputs'] if out['status'] == 'confirmed']) == 1
l1.rpc.withdraw(addr2, "all")
# One tx in the mempool now!
assert len(bitcoind.rpc.getrawmempool()) == 1
# Mine one block, assert one more is spendable
bitcoind.rpc.generatetoaddress(1, l1.rpc.newaddr()['bech32'])
assert len([out for out in l1.rpc.listfunds()['outputs'] if out['status'] == 'confirmed']) == 1
tests: check that non-anchor opens can't use liquidity ads liquidity ads (as proposed) rely on the CSV addition to the to_remote output that anchors introduced.
2022-10-19 22:45:29 +02:00
test (df): check 'require-confirmed-inputs' for v2 opens
2023-01-13 02:30:20 +01:00
@pytest.mark.openchannel('v2')
def test_openchannel_no_confirmed_inputs_opener(node_factory, bitcoind):
""" If the opener flags 'require-confirmed-inputs' for an open,
and accepter sends unconfirmed inputs check that the
accepter aborts the open """
l1_opts = {'funder-policy': 'match', 'funder-policy-mod': 100,
'lease-fee-base-sat': '100sat', 'lease-fee-basis': 100,
'may_reconnect': True, 'funder-lease-requests-only': False,
'allow_warning': True}
l2_opts = l1_opts.copy()
l1_opts['require-confirmed-inputs'] = True
l1, l2 = node_factory.get_nodes(2, opts=[l1_opts, l2_opts])
pytest: use modern listconfigs. Use the configs object, as the others are about to be deprecated. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-02 04:36:04 +02:00
assert l1.rpc.listconfigs()['configs']['require-confirmed-inputs']['value_bool'] is True
test (df): check 'require-confirmed-inputs' for v2 opens
2023-01-13 02:30:20 +01:00
amount = 500000
l1.fundwallet(20000000)
l2.fundwallet(20000000)
utxo_lookups = set()
def _no_utxo_response(r):
utxo_lookups.add(tuple(r['params']))
return {'id': r['id'], 'result': None}
# We mock l1 out such that it thinks no inputs are confirmed
l1.daemon.rpcproxy.mock_rpc('gettxout', _no_utxo_response)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
# l1 should return an error + abort the open as it thinks it's
# sending unconfirmed inputs to a peer that's requested only
# confirmed inputs
with pytest.raises(RpcError, match=r'Input .* is not confirmed'):
l1.rpc.fundchannel(l2.info['id'], amount)
assert l1.daemon.is_in_log('validating psbt for role: accepter')
# Verify that the looked up utxo is l2's
# Build a set of outpoints for node (l2)
outs = {(out['txid'], out['output']) for out in l2.rpc.listfunds()['outputs']}
# Confirm that seen utxo lookups are a subset of l2's outpoints
assert utxo_lookups <= outs
@pytest.mark.openchannel('v2')
def test_openchannel_no_unconfirmed_inputs_accepter(node_factory, bitcoind):
""" If the accepter flags 'require-confirmed-inputs' for an open,
and opener send unconfirmed inputs check that the
accepter aborts the open """
l1_opts = {'funder-policy': 'match', 'funder-policy-mod': 100,
'lease-fee-base-sat': '100sat', 'lease-fee-basis': 100,
'may_reconnect': True, 'funder-lease-requests-only': False,
'allow_warning': True}
l2_opts = l1_opts.copy()
l2_opts['require-confirmed-inputs'] = True
l1, l2 = node_factory.get_nodes(2, opts=[l1_opts, l2_opts])
pytest: use modern listconfigs. Use the configs object, as the others are about to be deprecated. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-02 04:36:04 +02:00
assert l2.rpc.listconfigs()['configs']['require-confirmed-inputs']['value_bool'] is True
test (df): check 'require-confirmed-inputs' for v2 opens
2023-01-13 02:30:20 +01:00
amount = 500000
l1.fundwallet(20000000)
l1.fundwallet(20000000)
l2.fundwallet(20000000)
utxo_lookups = set()
def _verify_utxos(n, lookedup):
# Build a set of outpoints for node (l2)
outs = {(out['txid'], out['output']) for out in n.rpc.listfunds()['outputs']}
# Confirm that seen utxo lookups are a subset of l2's outpoints
assert lookedup <= outs
lookedup.clear()
def _no_utxo_response(r):
utxo_lookups.add(tuple(r['params']))
# Check that the utxo belongs to l2
return {'id': r['id'], 'result': None}
l1.daemon.rpcproxy.mock_rpc('gettxout', _no_utxo_response)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
# l1 should return an error + abort the open as it thinks it's
# sending unconfirmed inputs to a peer that's requested only
# confirmed inputs
with pytest.raises(RpcError, match=r'Input .* is not confirmed'):
l1.rpc.fundchannel(l2.info['id'], amount)
_verify_utxos(l1, utxo_lookups)
l1.daemon.rpcproxy.mock_rpc('gettxout', None)
l2.daemon.rpcproxy.mock_rpc('gettxout', _no_utxo_response)
# l2 should return an error + abort the open
with pytest.raises(RpcError, match=r'Input .* is not confirmed'):
l1.rpc.fundchannel(l2.info['id'], amount)
_verify_utxos(l1, utxo_lookups)
# Let's negotiate the open, remove option from l2, and then RBF
# Turn the txout unconfirmed off, so we can open a channel
l2.daemon.rpcproxy.mock_rpc('gettxout', None)
res = l1.rpc.fundchannel(l2.info['id'], amount)
l1.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
l2.daemon.wait_for_log(' to DUALOPEND_AWAITING_LOCKIN')
# Remove option from l2
l2.stop()
del l2.daemon.opts['require-confirmed-inputs']
l2.start()
pytest: use modern listconfigs. Use the configs object, as the others are about to be deprecated. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-02 04:36:04 +02:00
assert l2.rpc.listconfigs()['configs']['require-confirmed-inputs']['value_bool'] is False
test (df): check 'require-confirmed-inputs' for v2 opens
2023-01-13 02:30:20 +01:00
# Turn the mock back on so we pretend everything l1 sends is unconf
l2.daemon.rpcproxy.mock_rpc('gettxout', _no_utxo_response)
# Prep for RBF
startweight = 42 + 172 # base weight, funding output
next_feerate = find_next_feerate(l1, l2)
psbt = l1.rpc.fundpsbt(amount, next_feerate, startweight,
min_witness_weight=110,
excess_as_change=True)['psbt']
# Attempt bump, fail. L2 should remember required-confirmed-inputs
# from original channel negotiation, despite node-wide setting
# being flagged off
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
bump = l1.rpc.openchannel_bump(res['channel_id'], amount, psbt)
with pytest.raises(RpcError, match=r'Input .* is not confirmed'):
l1.rpc.openchannel_update(res['channel_id'], bump['psbt'])
_verify_utxos(l1, utxo_lookups)
pytest: remove EXPERIMENTAL_FEATURES as a consideration. This currently means anchors tests are disabled, awaiting the PR which implements zero-fee-htlc anchors to reenable them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-05-22 02:51:44 +02:00
@unittest.skip("anchors not available")
tests: check that non-anchor opens can't use liquidity ads liquidity ads (as proposed) rely on the CSV addition to the to_remote output that anchors introduced.
2022-10-19 22:45:29 +02:00
@pytest.mark.openchannel('v2')
def test_no_anchor_liquidity_ads(node_factory, bitcoind):
""" Liquidity ads requires anchors, which are no longer a
requirement for dual-funded channels. """
pytest: make dual-funding tests using option_anchors_zero_fee_htlc_tx. They used to force option_anchor_outputs, so switch them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
l2_opts = {'funder-policy': 'match', 'funder-policy-mod': 100,
tests: check that non-anchor opens can't use liquidity ads liquidity ads (as proposed) rely on the CSV addition to the to_remote output that anchors introduced.
2022-10-19 22:45:29 +02:00
'lease-fee-base-sat': '100sat', 'lease-fee-basis': 100,
'may_reconnect': True, 'funder-lease-requests-only': False}
pytest: make dual-funding tests using option_anchors_zero_fee_htlc_tx. They used to force option_anchor_outputs, so switch them. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
l1_opts = l2_opts.copy()
l1_opts['experimental-anchors'] = None
tests: check that non-anchor opens can't use liquidity ads liquidity ads (as proposed) rely on the CSV addition to the to_remote output that anchors introduced.
2022-10-19 22:45:29 +02:00
l1, l2 = node_factory.get_nodes(2, opts=[l1_opts, l2_opts])
feerate = 2000
amount = 10**6
l1.fundwallet(10**8)
l2.fundwallet(10**8)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
with pytest.raises(RpcError, match=r'liquidity ads not supported, no anchors.'):
l1.rpc.fundchannel(l2.info['id'], amount, request_amt=amount,
feerate='{}perkw'.format(feerate),
compact_lease='029a002d000000004b2003e8')
# But you can make it work without the liquidity ad request
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l1.rpc.fundchannel(l2.info['id'], amount,
feerate='{}perkw'.format(feerate))
# Confirm that we used the DUAL_FUND flow
chan = only_one(only_one(l1.rpc.listpeers()['peers'])['channels'])
assert chan['state'] == 'DUALOPEND_AWAITING_LOCKIN'
assert chan['funding']['local_funds_msat'] == chan['funding']['remote_funds_msat']
assert 'option_anchor_outputs' not in chan['features']
pytest: test for commitment feerate. Interesting we didn't actually test that the feerate we use is actually delivered. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-26 01:03:21 +02:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd has different feerates')
pytest: adapt all the anchor-iff-EXPERIMENTAL tests to --experimental-anchors. We use parameterization here. The old `anchor_expected()` was for non-zero-fee anchors, and have bitrotted so there are some other changes as well. Unfortunately, all the anchor accounting seems to be broken, but I cannot understand these tests at all. I had to simply disable them for now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
@pytest.mark.parametrize("anchors", [False, True])
def test_commitment_feerate(bitcoind, node_factory, anchors):
opts = {}
if anchors:
opts['experimental-anchors'] = None
l1, l2 = node_factory.get_nodes(2, opts=opts)
pytest: test for commitment feerate. Interesting we didn't actually test that the feerate we use is actually delivered. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-26 01:03:21 +02:00
openingd: use a lower feerate for first commitment tx for anchors. We need to know both, because in theory we could negotiate a non-anchor channel even if they support it. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-26 01:11:21 +02:00
opening_feerate = 2000
pytest: adapt all the anchor-iff-EXPERIMENTAL tests to --experimental-anchors. We use parameterization here. The old `anchor_expected()` was for non-zero-fee anchors, and have bitrotted so there are some other changes as well. Unfortunately, all the anchor accounting seems to be broken, but I cannot understand these tests at all. I had to simply disable them for now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
if anchors:
openingd: use a lower feerate for first commitment tx for anchors. We need to know both, because in theory we could negotiate a non-anchor channel even if they support it. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-26 01:11:21 +02:00
# anchors use lowball fees
commitment_feerate = 3750
else:
commitment_feerate = opening_feerate
pytest: test for commitment feerate. Interesting we didn't actually test that the feerate we use is actually delivered. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-26 01:03:21 +02:00
l1.fundwallet(10**8)
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l1.rpc.fundchannel(l2.info['id'], 10**6,
feerate=f'{opening_feerate}perkw')
wait_for(lambda: bitcoind.rpc.getrawmempool() != [])
tx = only_one([t for t in bitcoind.rpc.getrawmempool(True).values()])
feerate_perkw = int(tx['fees']['base'] * 100_000_000) / (tx['weight'] / 1000)
assert opening_feerate - 10 < feerate_perkw < opening_feerate + 10
bitcoind.generate_block(1)
l2.stop()
l1.rpc.close(l2.info['id'], unilateraltimeout=1)
# feerate for this will be the same
wait_for(lambda: bitcoind.rpc.getrawmempool() != [])
tx = only_one([t for t in bitcoind.rpc.getrawmempool(True).values()])
fee = int(tx['fees']['base'] * 100_000_000)
# Weight is idealized worst case, and we don't meet it!
pytest: adapt all the anchor-iff-EXPERIMENTAL tests to --experimental-anchors. We use parameterization here. The old `anchor_expected()` was for non-zero-fee anchors, and have bitrotted so there are some other changes as well. Unfortunately, all the anchor accounting seems to be broken, but I cannot understand these tests at all. I had to simply disable them for now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
if anchors:
pytest: test for commitment feerate. Interesting we didn't actually test that the feerate we use is actually delivered. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-26 01:03:21 +02:00
# 200 is the approximate cost estimate used for anchor outputs.
assert tx['weight'] < 1124 - 200
else:
assert tx['weight'] < 724
pytest: adapt all the anchor-iff-EXPERIMENTAL tests to --experimental-anchors. We use parameterization here. The old `anchor_expected()` was for non-zero-fee anchors, and have bitrotted so there are some other changes as well. Unfortunately, all the anchor accounting seems to be broken, but I cannot understand these tests at all. I had to simply disable them for now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:09 +02:00
if anchors:
pytest: test for commitment feerate. Interesting we didn't actually test that the feerate we use is actually delivered. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-26 01:03:21 +02:00
# We pay for two anchors, but only produce one.
fee -= 330
weight = 1124
else:
weight = 724
feerate_perkw = fee / (weight / 1000)
assert commitment_feerate - 10 < feerate_perkw < commitment_feerate + 10
pytest: test for emergency minimum fund violations. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:10 +02:00
@unittest.skipIf(TEST_NETWORK != 'regtest', 'elementsd has different tx costs')
def test_anchor_min_emergency(bitcoind, node_factory):
l1, l2 = node_factory.line_graph(2, opts={'experimental-anchors': None},
fundchannel=False)
addr = l1.rpc.newaddr()['bech32']
bitcoind.rpc.sendtoaddress(addr, 5000000 / 10**8)
bitcoind.generate_block(1, wait_for_mempool=1)
wait_for(lambda: l1.rpc.listfunds()['outputs'] != [])
# Cost of tx itself is 3637.
with pytest.raises(RpcError, match=r'We would not have enough left for min-emergency-msat 25000sat'):
l1.rpc.fundchannel(l2.info['id'], f'{5000000 - 3637}sat')
l1.rpc.fundchannel(l2.info['id'], 'all')
bitcoind.generate_block(1, wait_for_mempool=1)
# Wait for l1 to see that spend.
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) == 1)
# Default is 25000 sats.
assert only_one(l1.rpc.listfunds()['outputs'])['amount_msat'] == Millisatoshi('25000sat')
# And we can't spend it, either!
addr2 = l2.rpc.newaddr()['bech32']
with pytest.raises(RpcError, match=r'We would not have enough left for min-emergency-msat 25000sat'):
l1.rpc.withdraw(addr2, '500sat')
with pytest.raises(RpcError, match=r'We would not have enough left for min-emergency-msat 25000sat'):
l1.rpc.withdraw(addr2, 'all')
pytest: fix test_anchor_min_emergency "bad gossip" messages. l1 shuts down too fast, channeld doesn't get to tell gossipd about the channel, and l2 sends (private) update_channel and we complain: ``` lightningd-2 2023-07-20T03:42:37.744Z DEBUG gossipd: received private channel announcement from channeld for 103x1x0 lightningd-2 2023-07-20T03:42:37.791Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-hsmd: Got WIRE_HSMD_CUPDATE_SIG_REQ lightningd-2 2023-07-20T03:42:37.796Z DEBUG hsmd: Client: Received message 3 from client lightningd-1 2023-07-20T03:42:37.857Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-gossipd: Bad gossip order: WIRE_CHANNEL_UPDATE before announcement 103x1x0/0 lightningd-1 2023-07-20T03:42:37.864Z DEBUG 022d223620a359a47ff7f7ac447c85c46c923da53389221a0054c11c1e3ca31d59-gossipd: Bad gossip order: WIRE_CHANNEL_UPDATE before announcement 103x1x0/0 ``` Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-07-20 23:46:48 +02:00
# Make sure channeld tells gossipd about channel before we close, otherwise
# we get spurious "bad gossip" complaints if l2 sends channel_updates.
l1.daemon.wait_for_log("received private channel announcement from channeld")
pytest: test for emergency minimum fund violations. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-06-29 02:14:10 +02:00
# Even with onchain anchor channel, it still keeps reserve (just in case!).
l1.rpc.close(l2.info['id'])
bitcoind.generate_block(1, wait_for_mempool=1)
sync_blockheight(bitcoind, [l1])
# This workse, but will leave the emergency funds as change.
l1.rpc.withdraw(addr2, 'all')
bitcoind.generate_block(1, wait_for_mempool=1)
sync_blockheight(bitcoind, [l1])
wait_for(lambda: [(o['amount_msat'], o['status']) for o in l1.rpc.listfunds()['outputs']] == [(Millisatoshi('25000sat'), 'confirmed')])
# Can't spend it!
with pytest.raises(RpcError, match=r'We would not have enough left for min-emergency-msat 25000sat'):
l1.rpc.withdraw(addr2, 'all')
# Once it's totally forgotten, we can spend that!
bitcoind.generate_block(99)
wait_for(lambda: l1.rpc.listpeerchannels()['channels'] == [])
# And it's *all* gone!
l1.rpc.withdraw(addr2, 'all')
bitcoind.generate_block(1, wait_for_mempool=1)
wait_for(lambda: l1.rpc.listfunds()['outputs'] == [])
tests: test for fundchannel fail with tiny utxo Tony Giorgio <tonygiorgio@protonmail.com> says: Reproduce: 1. Add 1 600 sat UTXO to a fresh node 2. Verify the fundchannel command fails with a low fee rate: ``` ./lightning-cli fundchannel 0366abc8eb4da61e31a8d2c4520d31cabdf58cc5250f855657397f3dd62493938a 100000 1000 { "code": 301, "message": "Could not afford 100000sat using all 1 available UTXOs: 99522sat short" } ``` 3. Now do the command again, but with a higher fee rate, making the 600 sat UTXO uneconomical: ``` ./lightning-cli fundchannel 0366abc8eb4da61e31a8d2c4520d31cabdf58cc5250f855657397f3dd62493938a 100000 10000 ``` 4. Observe the RPC call and the logs. The RPC call will never return, and the logs will stop after this: ``` 2023-04-16T10:58:45.839Z DEBUG plugin-spenderp: mfc 34: multiconnect done. 2023-04-16T10:58:45.839Z DEBUG plugin-spenderp: mfc 34: 'parsefeerate' done 2023-04-16T10:58:45.839Z DEBUG plugin-spenderp: mfc 34: fundpsbt. ``` 5. Keep CLN running long enough and you'll eventually run OOM.
2023-09-13 02:25:38 +02:00
def test_fundchannel_utxo_too_small(bitcoind, node_factory):
l1, l2 = node_factory.get_nodes(2)
# Add 1 600 sat UTXO to a fresh node
bitcoind.rpc.sendtoaddress(l1.rpc.newaddr()['bech32'], 0.00000600)
bitcoind.generate_block(1, wait_for_mempool=1)
wait_for(lambda: len(l1.rpc.listfunds()['outputs']) == 1)
# a higher fee rate, making the 600 sat UTXO uneconomical:
l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
with pytest.raises(RpcError, match=r'Could not afford 100000sat using all 0 available UTXOs'):
l1.rpc.fundchannel(l2.info['id'], 100000, 10000)
Reference in New Issue Copy Permalink