This fixes an issue that would lead to a flake during intergration
tests. Carol would start up with a outdated state and attempt to force
close the channel. At the same time she would connect to Dave,
triggering the dataloss protection. Dave would respond by force closing
the channel, and Dave transaction would in some cases have a higher fee,
resulting Carol's tx being replaced.
We fix this by suspending Dave until Carol's close tx is mined.
In this commit, we update the `AbandonChannel` method to also remove the
state from the countract court as well as the channel graph. Abandoning
a channel is now a three step process: remove from the open channel
state, remove from the graph, remove from the contract court. Between
any step it's possible that the users restarts the process all over
again. As a result, each of the steps below are intended to be
idempotent.
We also update the integration test to assert that no channel is found
in the graph any longer. Before this commit, this test would fail as the
channel was still found in the graph, which can cause other issues for
an operational daemon.
Fixes#3716.
This commit beings the process of deprecating unsafe-disconnect. Many
moons ago this was disallowed to prevent concurency bugs surrounding
reconnect. Despite the name, it has been safe to enable this feature for
well over a year, as several PRs have been merged that addressed the
possible issues that existed when the feature was added.
In this commit, we refactor the testSingleHopSendToRoute test to support
table driven tests for various endpoints and payment types. Currently
only the main rpcserver's SendToRoute is tested, so we also add
support the SendToRouteSync and the routerrpc's SendToRoute.
The tests are also modified to have each endpoint perform a single-hop,
single-shot MPP payment. This asserts that the Hop messages are being
properly unmarshalled and that setting correctly yields a successful
payment. At the momemnt the receiver does not actually verify or use the
MPP fields presented in the onion, though this test will be expanded
later as those pieces are assembled.
We add a wait predicate to make sure the node's on-chain balance is
restored before continuing the restore test case.
This is needed since the DLP test scenario includes several restarts of
the node, and if the node isn't done scanning for on-chain balance
before the restart happens, it would be unlocked without a recovery
window, causing funds to be left undiscovered.
Since the ErrorCodes are not part of the spec, they cannot be read by
other implementations.
Instead of only sending the error code we therefore send the complete
error message. This will have the same effect at the client, as it will
just get the full error instead of the code indicating which error it
is. It will also be compatible with other impls.
Note that the GRPC error codes will change, since we don't set them
anymore.
This changes the defer function in the test for channel backups to
correctly close over the 'dave' variable.
Without this closure, the shutdownAndAssert call would attempt to
shutdown the original (non-restored) dave instead of the most recently
created (restored) dave, causing a leak of a node during tests.
The test assumed that transactions would be broadcast and confirmed at
incorrect heights. Due to timing issues, it was possible for the test to
still succeed, resulting in a flake.
The test assumes that Bob will sweep a pending outgoing HTLC and commit
output back to their wallet. This commit ensures that these operations
are done when expected, i.e.:
1. Bob force closes the channel due to the HTLC timing out.
2. Once the channel is confirmed, Bob broadcasts their HTLC timeout
transaction.
3. Bob broadcasts their commit output sweep transaction once its CSV
expires.
4. Bob broadcasts their second layer sweep transaction for the timed out
HTLC once its CSV expires.
Alice and Dave don't need to be connected in order to receive the node
announcement as we assume that she can receive it from Bob because they
are connected at the beginning of every test.
In this commit, we force Dave to use the legacy onion payload for the
multi-hop test to ensure that we're able to properly mix the old and new
formats, and have all nodes properly decode+forward the HTLC.
htlcs
config: Adding RejectHTLC field in config struct
This commit adds a RejectHTLC field in the config struct in config.go.
This allows the user to run lnd as a node that does not accept onward
HTLCs.
htlcswitch/switch: Adding a field RejectHTLC to the switch config
This commit adds a field RejectHTLC to the switch config. When the
switch receives an HTLC it will check this flag and if the HTLC is not
from the source hop, the HTLC will be rejected.
htlcswitch/switch: adding check for RejectHTLC flag and incomingChanID
This commit adds a check when receiving UpdateAddHTLC. The check looks
for the RejectHTLC flag set and whether the HTLC is from the sourceHop
(the local switch). If the HTLC is not from the sourceHop, then we
reject the HTLC and return a FailChannelDisabled error.
server: adding RejectHTLC field to initialization of switch
lnd_test: adding test for RejectHTLC
This commit adds a test which tests that a node with the --rejecthtlc
flag will reject any onward HTLCs but still can receive direct HTLCs and
can send HTLCs.
Previously a temporary channel failure was returning for unexpected
malformed htlc failures. This is not what we want to communicate to the
sender, because the sender may apply a penalty to us only.
Returning the temporary channel failure is especially problematic if we
ourselves are the sender and the malformed htlc failure comes from our
direct peer. When interpretating the failure, we aren't able to
distinguish anymore between our channel not having enough balance and
our peer sending an unexpected failure back.
Debug invoices are rarely used nowadays, but keep asking for maintenance
every time refactoring in primarily the invoice registry occurs. We have
passed the cost/benefit tipping point, so therefore the debug invoice
concept is removed in this commit.
Previously the debughtlc flag also controlled whether hodl masks were
active. It is safe to remove that additional condition because the hodl
masks are still guarded by the dev build tag.
Previously mission control tracked failures on a per node, per channel basis.
This commit changes this to tracking on the level of directed node pairs. The goal
of moving to this coarser-grained level is to reduce the number of required
payment attempts without compromising payment reliability.
Align naming better with the lightning spec. Not the full name of the
failure (FailIncorrectOrUnknownPaymentDetails) is used, because this
would cause too many long lines in the code.
With the introduction of the WatchtowerClient RPC subserver, the lnd
configuration flag to specify private watchtowers for the client is no
longer needed and can lead to confusion upon users. Therefore, we remove
the flag completely, and only rely on the watchtower client being active
through a new --wtclient.active flag.
This commit makes the outgoing link pipeline the settle to the
switch as soon as it receives it. Previously, it would wait for a
revocation before sending it, which caused increased latency on
payments as well as possibly never settling on the incoming link.
A duplicate settle is still sent to the switch, but it is handled
gracefully. A new AckEventTicker was added to the switch which
acknowledges any pending settle / fail entries in an outgoing
link's fwd pkgs in batch. This was needed in order to reduce the
number of db txn's which would have been incurred by acking whenever
we receive a duplicate settle without batching.
This flake was caused by the rpcserver receiving a CloseChannel request
before Alice's channel event subscription request, causing Alice to miss one
notification. As a result, we move Alice's subscription to the beginning of the
test.
Additionally, we add a check to ensure the opening notifications are
received in the right order.
Earlier versions of ListPayments only included completed payments. We
return to this behavior by ignore all other payments if the nonSucceeded
boolean is not set in the request.
testHoldInvoicePersistence tests that a sender to a hold-invoice, can be
restarted before the payment gets settled, and still be able to receive
the preimage.
Previously every payment had its own local mission control state which
was in effect only for that payment. In this commit most of the local
state is removed and payments all tap into the global mission control
probability estimator.
Furthermore the decay time of pruned edges and nodes is extended, so
that observations about the network can better benefit future payment
processes.
Last, the probability function is transformed from a binary output to a
gradual curve, allowing for a better trade off between candidate routes.
