lnd/itest/lnd_onchain_test.go

846 lines
28 KiB
Go
Raw Normal View History

2019-06-04 01:12:20 +02:00
package itest
import (
"bytes"
"fmt"
"github.com/btcsuite/btcd/btcutil"
"github.com/btcsuite/btcd/chaincfg/chainhash"
2019-06-04 01:12:20 +02:00
"github.com/btcsuite/btcd/txscript"
"github.com/btcsuite/btcd/wire"
2019-06-04 01:12:20 +02:00
"github.com/lightningnetwork/lnd/lnrpc"
"github.com/lightningnetwork/lnd/lnrpc/chainrpc"
"github.com/lightningnetwork/lnd/lnrpc/signrpc"
2019-06-04 01:12:20 +02:00
"github.com/lightningnetwork/lnd/lnrpc/walletrpc"
"github.com/lightningnetwork/lnd/lntest"
"github.com/lightningnetwork/lnd/lntest/node"
"github.com/lightningnetwork/lnd/lntest/wait"
"github.com/lightningnetwork/lnd/lnwallet"
"github.com/stretchr/testify/require"
2019-06-04 01:12:20 +02:00
)
// testChainKit tests ChainKit RPC endpoints.
func testChainKit(ht *lntest.HarnessTest) {
// Test functions registered as test cases spin up separate nodes
// during execution. By calling sub-test functions as seen below we
// avoid the need to start separate nodes.
testChainKitGetBlock(ht)
testChainKitGetBlockHeader(ht)
testChainKitGetBlockHash(ht)
testChainKitSendOutputsAnchorReserve(ht)
}
// testChainKitGetBlock ensures that given a block hash, the RPC endpoint
// returns the correct target block.
func testChainKitGetBlock(ht *lntest.HarnessTest) {
// Get best block hash.
bestBlockRes := ht.Alice.RPC.GetBestBlock(nil)
var bestBlockHash chainhash.Hash
err := bestBlockHash.SetBytes(bestBlockRes.BlockHash)
require.NoError(ht, err)
// Retrieve the best block by hash.
getBlockReq := &chainrpc.GetBlockRequest{
BlockHash: bestBlockHash[:],
}
getBlockRes := ht.Alice.RPC.GetBlock(getBlockReq)
// Deserialize the block which was retrieved by hash.
msgBlock := &wire.MsgBlock{}
blockReader := bytes.NewReader(getBlockRes.RawBlock)
err = msgBlock.Deserialize(blockReader)
require.NoError(ht, err)
// Ensure best block hash is the same as retrieved block hash.
expected := bestBlockHash
actual := msgBlock.BlockHash()
require.Equal(ht, expected, actual)
}
// testChainKitGetBlockHeader ensures that given a block hash, the RPC endpoint
// returns the correct target block header.
func testChainKitGetBlockHeader(ht *lntest.HarnessTest) {
// Get best block hash.
bestBlockRes := ht.Alice.RPC.GetBestBlock(nil)
var (
bestBlockHash chainhash.Hash
bestBlockHeader wire.BlockHeader
msgBlock = &wire.MsgBlock{}
)
err := bestBlockHash.SetBytes(bestBlockRes.BlockHash)
require.NoError(ht, err)
// Retrieve the best block by hash.
getBlockReq := &chainrpc.GetBlockRequest{
BlockHash: bestBlockHash[:],
}
getBlockRes := ht.Alice.RPC.GetBlock(getBlockReq)
// Deserialize the block which was retrieved by hash.
blockReader := bytes.NewReader(getBlockRes.RawBlock)
err = msgBlock.Deserialize(blockReader)
require.NoError(ht, err)
// Retrieve the block header for the best block.
getBlockHeaderReq := &chainrpc.GetBlockHeaderRequest{
BlockHash: bestBlockHash[:],
}
getBlockHeaderRes := ht.Alice.RPC.GetBlockHeader(getBlockHeaderReq)
// Deserialize the block header which was retrieved by hash.
blockHeaderReader := bytes.NewReader(getBlockHeaderRes.RawBlockHeader)
err = bestBlockHeader.Deserialize(blockHeaderReader)
require.NoError(ht, err)
// Ensure the header of the best block is the same as retrieved block
// header.
expected := bestBlockHeader
actual := msgBlock.Header
require.Equal(ht, expected, actual)
}
// testChainKitGetBlockHash ensures that given a block height, the RPC endpoint
// returns the correct target block hash.
func testChainKitGetBlockHash(ht *lntest.HarnessTest) {
// Get best block hash.
bestBlockRes := ht.Alice.RPC.GetBestBlock(nil)
// Retrieve the block hash at best block height.
req := &chainrpc.GetBlockHashRequest{
BlockHeight: int64(bestBlockRes.BlockHeight),
}
getBlockHashRes := ht.Alice.RPC.GetBlockHash(req)
// Ensure best block hash is the same as retrieved block hash.
expected := bestBlockRes.BlockHash
actual := getBlockHashRes.BlockHash
require.Equal(ht, expected, actual)
}
// testChainKitSendOutputsAnchorReserve checks if the SendOutputs rpc prevents
// our wallet balance to drop below the required anchor channel reserve amount.
func testChainKitSendOutputsAnchorReserve(ht *lntest.HarnessTest) {
// Start two nodes supporting anchor channels.
args := lntest.NodeArgsForCommitType(lnrpc.CommitmentType_ANCHORS)
// NOTE: we cannot reuse the standby node here as the test requires the
// node to start with no UTXOs.
charlie := ht.NewNode("Charlie", args)
bob := ht.Bob
ht.RestartNodeWithExtraArgs(bob, args)
// We'll start the test by sending Charlie some coins.
fundingAmount := btcutil.Amount(100_000)
ht.FundCoins(fundingAmount, charlie)
// Before opening the channel we ensure that the nodes are connected.
ht.EnsureConnected(charlie, bob)
// We'll get the anchor reserve that is required for a single channel.
reserve := charlie.RPC.RequiredReserve(
&walletrpc.RequiredReserveRequest{
AdditionalPublicChannels: 1,
},
)
// Charlie opens an anchor channel and keeps twice the amount of the
// anchor reserve in her wallet.
chanAmt := fundingAmount - 2*btcutil.Amount(reserve.RequiredReserve)
outpoint := ht.OpenChannel(charlie, bob, lntest.OpenChannelParams{
Amt: chanAmt,
CommitmentType: lnrpc.CommitmentType_ANCHORS,
SatPerVByte: 1,
})
// Now we obtain a taproot address from bob which Charlie will use to
// send coins to him via the SendOutputs rpc.
address := bob.RPC.NewAddress(&lnrpc.NewAddressRequest{
Type: lnrpc.AddressType_TAPROOT_PUBKEY,
})
decodedAddr := ht.DecodeAddress(address.Address)
addrScript := ht.PayToAddrScript(decodedAddr)
// First she will try to send Bob an amount that would undershoot her
// reserve requirement by one satoshi.
balance := charlie.RPC.WalletBalance()
utxo := &wire.TxOut{
Value: balance.TotalBalance - reserve.RequiredReserve + 1,
PkScript: addrScript,
}
req := &walletrpc.SendOutputsRequest{
Outputs: []*signrpc.TxOut{{
Value: utxo.Value,
PkScript: utxo.PkScript,
}},
SatPerKw: 2400,
MinConfs: 1,
}
// We try to send the reserve violating transaction and expect it to
// fail.
_, err := charlie.RPC.WalletKit.SendOutputs(ht.Context(), req)
require.ErrorContains(ht, err, walletrpc.ErrInsufficientReserve.Error())
ht.MineBlocksAndAssertNumTxes(1, 0)
// Next she will try to send Bob an amount that just leaves enough
// reserves in her wallet.
utxo = &wire.TxOut{
Value: balance.TotalBalance - reserve.RequiredReserve,
PkScript: addrScript,
}
req = &walletrpc.SendOutputsRequest{
Outputs: []*signrpc.TxOut{{
Value: utxo.Value,
PkScript: utxo.PkScript,
}},
SatPerKw: 2400,
MinConfs: 1,
}
// This second transaction should be published correctly.
charlie.RPC.SendOutputs(req)
ht.MineBlocksAndAssertNumTxes(1, 1)
// Clean up our test setup.
ht.CloseChannel(charlie, outpoint)
}
// testAnchorReservedValue tests that we won't allow sending transactions when
// that would take the value we reserve for anchor fee bumping out of our
// wallet.
func testAnchorReservedValue(ht *lntest.HarnessTest) {
// Start two nodes supporting anchor channels.
args := lntest.NodeArgsForCommitType(lnrpc.CommitmentType_ANCHORS)
// NOTE: we cannot reuse the standby node here as the test requires the
// node to start with no UTXOs.
alice := ht.NewNode("Alice", args)
bob := ht.Bob
ht.RestartNodeWithExtraArgs(bob, args)
ht.ConnectNodes(alice, bob)
// Send just enough coins for Alice to open a channel without a change
// output.
const (
chanAmt = 1000000
feeEst = 8000
)
ht.FundCoins(chanAmt+feeEst, alice)
// wallet, without a change output. This should not be allowed.
ht.OpenChannelAssertErr(
alice, bob, lntest.OpenChannelParams{
Amt: chanAmt,
}, lnwallet.ErrReservedValueInvalidated,
)
// Alice opens a smaller channel. This works since it will have a
// change output.
chanPoint1 := ht.OpenChannel(
alice, bob, lntest.OpenChannelParams{Amt: chanAmt / 4},
)
// If Alice tries to open another anchor channel to Bob, Bob should not
// reject it as he is not contributing any funds.
chanPoint2 := ht.OpenChannel(
alice, bob, lntest.OpenChannelParams{Amt: chanAmt / 4},
)
// Similarly, if Alice tries to open a legacy channel to Bob, Bob
// should not reject it as he is not contributing any funds. We'll
// restart Bob to remove his support for anchors.
ht.RestartNode(bob)
// Before opening the channel, make sure the nodes are connected.
ht.EnsureConnected(alice, bob)
chanPoint3 := ht.OpenChannel(
alice, bob, lntest.OpenChannelParams{Amt: chanAmt / 4},
)
chanPoints := []*lnrpc.ChannelPoint{chanPoint1, chanPoint2, chanPoint3}
// Alice tries to send all coins to an internal address. This is
// allowed, since the final wallet balance will still be above the
// reserved value.
req := &lnrpc.NewAddressRequest{
Type: lnrpc.AddressType_WITNESS_PUBKEY_HASH,
}
resp := alice.RPC.NewAddress(req)
sweepReq := &lnrpc.SendCoinsRequest{
Addr: resp.Address,
SendAll: true,
TargetConf: 6,
}
alice.RPC.SendCoins(sweepReq)
block := ht.MineBlocksAndAssertNumTxes(1, 1)[0]
assertNumTxInAndTxOut := func(tx *wire.MsgTx, in, out int) {
require.Len(ht, tx.TxIn, in, "num inputs not matched")
require.Len(ht, tx.TxOut, out, "num outputs not matched")
}
// The sweep transaction should have exactly one input, the change from
// the previous SendCoins call.
sweepTx := block.Transactions[1]
// It should have a single output.
assertNumTxInAndTxOut(sweepTx, 1, 1)
// Wait for Alice to see her balance as confirmed.
waitForConfirmedBalance := func() int64 {
var balance int64
err := wait.NoError(func() error {
resp := alice.RPC.WalletBalance()
if resp.TotalBalance == 0 {
return fmt.Errorf("no balance")
}
if resp.UnconfirmedBalance > 0 {
return fmt.Errorf("unconfirmed balance")
}
balance = resp.TotalBalance
return nil
}, defaultTimeout)
require.NoError(ht, err, "timeout checking alice's balance")
return balance
}
waitForConfirmedBalance()
// Alice tries to send all funds to an external address, the reserved
// value must stay in her wallet.
minerAddr := ht.Miner.NewMinerAddress()
sweepReq = &lnrpc.SendCoinsRequest{
Addr: minerAddr.String(),
SendAll: true,
TargetConf: 6,
}
alice.RPC.SendCoins(sweepReq)
// We'll mine a block which should include the sweep transaction we
// generated above.
block = ht.MineBlocksAndAssertNumTxes(1, 1)[0]
// The sweep transaction should have exactly one inputs as we only had
// the single output from above in the wallet.
sweepTx = block.Transactions[1]
// It should have two outputs, one being the miner address, the other
// one being the reserve going back to our wallet.
assertNumTxInAndTxOut(sweepTx, 1, 2)
// The reserved value is now back in Alice's wallet.
aliceBalance := waitForConfirmedBalance()
// Alice closes channel, should now be allowed to send everything to an
// external address.
for _, chanPoint := range chanPoints {
ht.CloseChannel(alice, chanPoint)
}
newBalance := waitForConfirmedBalance()
require.Greater(ht, newBalance, aliceBalance,
"Alice's balance did not increase after channel close")
// Assert there are no open or pending channels anymore.
ht.AssertNumWaitingClose(alice, 0)
ht.AssertNodeNumChannels(alice, 0)
// We'll wait for the balance to reflect that the channel has been
// closed and the funds are in the wallet.
sweepReq = &lnrpc.SendCoinsRequest{
Addr: minerAddr.String(),
SendAll: true,
TargetConf: 6,
}
alice.RPC.SendCoins(sweepReq)
// We'll mine a block which should include the sweep transaction we
// generated above.
block = ht.MineBlocksAndAssertNumTxes(1, 1)[0]
// The sweep transaction should have four inputs, the change output from
// the previous sweep, and the outputs from the coop closed channels.
sweepTx = block.Transactions[1]
// It should have a single output.
assertNumTxInAndTxOut(sweepTx, 4, 1)
}
// testAnchorThirdPartySpend tests that if we force close a channel, but then
// don't sweep the anchor in time and a 3rd party spends it, that we remove any
// transactions that are a descendent of that sweep.
func testAnchorThirdPartySpend(ht *lntest.HarnessTest) {
// First, we'll create two new nodes that both default to anchor
// channels.
//
// NOTE: The itests differ here as anchors is default off vs the normal
// lnd binary.
args := lntest.NodeArgsForCommitType(lnrpc.CommitmentType_ANCHORS)
alice := ht.NewNode("Alice", args)
bob := ht.NewNode("Bob", args)
ht.EnsureConnected(alice, bob)
// We'll fund our Alice with coins, as she'll be opening the channel.
// We'll fund her with *just* enough coins to open the channel and
// sweep the anchor.
const (
firstChanSize = 1_000_000
anchorFeeBuffer = 500_000
testMemo = "bob is a good peer"
)
ht.FundCoins(firstChanSize+anchorFeeBuffer, alice)
// Open the channel between the two nodes and wait for it to confirm
// fully.
aliceChanPoint1 := ht.OpenChannel(
alice, bob, lntest.OpenChannelParams{
Amt: firstChanSize,
Memo: testMemo,
},
)
// Send another UTXO if this is a neutrino backend. When sweeping
// anchors, there are two transactions created, `local_sweep_tx` for
// sweeping Alice's anchor on the local commitment, `remote_sweep_tx`
// for sweeping her anchor on the remote commitment. Whenever the force
// close transaction is published, Alice will always create these two
// transactions to sweep her anchor.
// On the other hand, when creating the sweep txes, the anchor itself
// is not able to cover the fee, so another wallet UTXO is needed. In
// our test case, there's a change output that can be used from the
// above funding process. And it's used by both sweep txes - when `lnd`
// happens to create the `remote_sweep_tx` first, it will receive an
// error since its parent tx, the remote commitment, is not known,
// hence freeing the change output to be used by `local_sweep_tx`.
// For neutrino client, however, it will consider the transaction which
// sweeps the remote anchor as an orphan tx, and it will neither send
// it to the mempool nor return an error to free the change output.
// Thus, if the change output is already used in `remote_sweep_tx`, we
// won't have UTXO to create `local_sweep_tx`.
//
// NOTE: the order of the sweep requests for the two anchors cannot be
// guaranteed. If the sweeper happens to sweep the remote anchor first,
// then the test won't pass without the extra UTXO, which is the source
// of the flakeness.
//
// TODO(yy): make a RPC server for sweeper so we can explicitly check
// and control its state.
if ht.IsNeutrinoBackend() {
ht.FundCoins(anchorFeeBuffer, alice)
}
// With the channel open, we'll actually immediately force close it. We
// don't care about network announcements here since there's no routing
// in this test.
ht.CloseChannelAssertPending(alice, aliceChanPoint1, true)
// Now that the channel has been force closed, it should show up in the
// PendingChannels RPC under the waiting close section.
waitingClose := ht.AssertChannelWaitingClose(alice, aliceChanPoint1)
// Verify that the channel Memo is returned even for channels that are
// waiting close (close TX broadcasted but not confirmed)
pendingChannelsResp := alice.RPC.PendingChannels()
require.Equal(ht, testMemo,
pendingChannelsResp.WaitingCloseChannels[0].Channel.Memo)
sweeper: relax anchor sweeping when there's no deadline pressure (#7965) * sweep: use longer variable name for clarity in `addToState` * sweeper: add more docs and debug logs * sweep: prioritize smaller inputs when adding wallet UTXOs This commit sorts wallet UTXOs by their values when using them for sweeping inputs. This way we'd avoid locking large UTXOs when sweeping inputs and also provide an opportunity to aggregate wallet UTXOs. * contractcourt+itest: relax anchor sweeping for CPFP purpose This commit changes from always sweeping anchor for a local force close to only do so when there is an actual time pressure. After this change, a forced anchor sweeping will only be attempted when the deadline is less than 144 blocks. * docs: update release notes * itest: update test `testMultiHopHtlcLocalChainClaim` to skip CPFP Since we now only perform CPFP when both the fee rate is higher and the deadline is less than 144, we need to update the test to reflect that Bob will not CPFP the force close tx for the channle Alice->Bob. * itest: fix `testMultiHopRemoteForceCloseOnChainHtlcTimeout` * itest: update related tests to reflect anchor sweeping This commit updates all related tests to reflect the latest anchor sweeping behavior. Previously, anchor sweeping is always attempted as CPFP when a force close is broadcast, while now it only happens when the deadline is less than 144. For non-CPFP purpose sweeping, it will happen after one block is mined after the force close transaction is confirmed as the anchor will be resent to the sweeper with a floor fee rate, hence making it economical to sweep.
2023-10-12 17:18:59 +02:00
// At this point, the channel is waiting close so we have the
// commitment transaction in the mempool. Alice's anchor, however,
// because there's no deadline pressure, it won't be swept.
aliceCloseTx := waitingClose.Commitments.LocalTxid
sweeper: relax anchor sweeping when there's no deadline pressure (#7965) * sweep: use longer variable name for clarity in `addToState` * sweeper: add more docs and debug logs * sweep: prioritize smaller inputs when adding wallet UTXOs This commit sorts wallet UTXOs by their values when using them for sweeping inputs. This way we'd avoid locking large UTXOs when sweeping inputs and also provide an opportunity to aggregate wallet UTXOs. * contractcourt+itest: relax anchor sweeping for CPFP purpose This commit changes from always sweeping anchor for a local force close to only do so when there is an actual time pressure. After this change, a forced anchor sweeping will only be attempted when the deadline is less than 144 blocks. * docs: update release notes * itest: update test `testMultiHopHtlcLocalChainClaim` to skip CPFP Since we now only perform CPFP when both the fee rate is higher and the deadline is less than 144, we need to update the test to reflect that Bob will not CPFP the force close tx for the channle Alice->Bob. * itest: fix `testMultiHopRemoteForceCloseOnChainHtlcTimeout` * itest: update related tests to reflect anchor sweeping This commit updates all related tests to reflect the latest anchor sweeping behavior. Previously, anchor sweeping is always attempted as CPFP when a force close is broadcast, while now it only happens when the deadline is less than 144. For non-CPFP purpose sweeping, it will happen after one block is mined after the force close transaction is confirmed as the anchor will be resent to the sweeper with a floor fee rate, hence making it economical to sweep.
2023-10-12 17:18:59 +02:00
ht.MineBlocksAndAssertNumTxes(1, 1)
forceCloseTxID, _ := chainhash.NewHashFromStr(aliceCloseTx)
sweeper: relax anchor sweeping when there's no deadline pressure (#7965) * sweep: use longer variable name for clarity in `addToState` * sweeper: add more docs and debug logs * sweep: prioritize smaller inputs when adding wallet UTXOs This commit sorts wallet UTXOs by their values when using them for sweeping inputs. This way we'd avoid locking large UTXOs when sweeping inputs and also provide an opportunity to aggregate wallet UTXOs. * contractcourt+itest: relax anchor sweeping for CPFP purpose This commit changes from always sweeping anchor for a local force close to only do so when there is an actual time pressure. After this change, a forced anchor sweeping will only be attempted when the deadline is less than 144 blocks. * docs: update release notes * itest: update test `testMultiHopHtlcLocalChainClaim` to skip CPFP Since we now only perform CPFP when both the fee rate is higher and the deadline is less than 144, we need to update the test to reflect that Bob will not CPFP the force close tx for the channle Alice->Bob. * itest: fix `testMultiHopRemoteForceCloseOnChainHtlcTimeout` * itest: update related tests to reflect anchor sweeping This commit updates all related tests to reflect the latest anchor sweeping behavior. Previously, anchor sweeping is always attempted as CPFP when a force close is broadcast, while now it only happens when the deadline is less than 144. For non-CPFP purpose sweeping, it will happen after one block is mined after the force close transaction is confirmed as the anchor will be resent to the sweeper with a floor fee rate, hence making it economical to sweep.
2023-10-12 17:18:59 +02:00
2024-04-08 10:48:43 +02:00
// Alice's should have the anchor sweep request.
ht.AssertNumPendingSweeps(alice, 1)
// Mine 3 blocks so Alice will sweep her commit output.
forceClose := ht.AssertChannelPendingForceClose(alice, aliceChanPoint1)
ht.MineEmptyBlocks(int(forceClose.BlocksTilMaturity) - 1)
// Alice's should have two sweep request - one for anchor output, the
// other for commit output.
ht.AssertNumPendingSweeps(alice, 2)
sweeper: relax anchor sweeping when there's no deadline pressure (#7965) * sweep: use longer variable name for clarity in `addToState` * sweeper: add more docs and debug logs * sweep: prioritize smaller inputs when adding wallet UTXOs This commit sorts wallet UTXOs by their values when using them for sweeping inputs. This way we'd avoid locking large UTXOs when sweeping inputs and also provide an opportunity to aggregate wallet UTXOs. * contractcourt+itest: relax anchor sweeping for CPFP purpose This commit changes from always sweeping anchor for a local force close to only do so when there is an actual time pressure. After this change, a forced anchor sweeping will only be attempted when the deadline is less than 144 blocks. * docs: update release notes * itest: update test `testMultiHopHtlcLocalChainClaim` to skip CPFP Since we now only perform CPFP when both the fee rate is higher and the deadline is less than 144, we need to update the test to reflect that Bob will not CPFP the force close tx for the channle Alice->Bob. * itest: fix `testMultiHopRemoteForceCloseOnChainHtlcTimeout` * itest: update related tests to reflect anchor sweeping This commit updates all related tests to reflect the latest anchor sweeping behavior. Previously, anchor sweeping is always attempted as CPFP when a force close is broadcast, while now it only happens when the deadline is less than 144. For non-CPFP purpose sweeping, it will happen after one block is mined after the force close transaction is confirmed as the anchor will be resent to the sweeper with a floor fee rate, hence making it economical to sweep.
2023-10-12 17:18:59 +02:00
// Mine one block to trigger Alice's sweeper to reconsider the anchor
2024-04-08 10:48:43 +02:00
// sweeping - it will be swept with her commit output together in one
// tx.
sweeper: relax anchor sweeping when there's no deadline pressure (#7965) * sweep: use longer variable name for clarity in `addToState` * sweeper: add more docs and debug logs * sweep: prioritize smaller inputs when adding wallet UTXOs This commit sorts wallet UTXOs by their values when using them for sweeping inputs. This way we'd avoid locking large UTXOs when sweeping inputs and also provide an opportunity to aggregate wallet UTXOs. * contractcourt+itest: relax anchor sweeping for CPFP purpose This commit changes from always sweeping anchor for a local force close to only do so when there is an actual time pressure. After this change, a forced anchor sweeping will only be attempted when the deadline is less than 144 blocks. * docs: update release notes * itest: update test `testMultiHopHtlcLocalChainClaim` to skip CPFP Since we now only perform CPFP when both the fee rate is higher and the deadline is less than 144, we need to update the test to reflect that Bob will not CPFP the force close tx for the channle Alice->Bob. * itest: fix `testMultiHopRemoteForceCloseOnChainHtlcTimeout` * itest: update related tests to reflect anchor sweeping This commit updates all related tests to reflect the latest anchor sweeping behavior. Previously, anchor sweeping is always attempted as CPFP when a force close is broadcast, while now it only happens when the deadline is less than 144. For non-CPFP purpose sweeping, it will happen after one block is mined after the force close transaction is confirmed as the anchor will be resent to the sweeper with a floor fee rate, hence making it economical to sweep.
2023-10-12 17:18:59 +02:00
ht.MineEmptyBlocks(1)
sweepTxns := ht.Miner.GetNumTxsFromMempool(1)
_, aliceAnchor := ht.FindCommitAndAnchor(sweepTxns, aliceCloseTx)
// Assert that the channel is now in PendingForceClose.
//
// NOTE: We must do this check to make sure `lnd` node has updated its
// internal state regarding the closing transaction, otherwise the
// `SendCoins` below might fail since it involves a reserved value
// check, which requires a certain amount of coins to be reserved based
// on the number of anchor channels.
ht.AssertChannelPendingForceClose(alice, aliceChanPoint1)
// Verify that the channel Memo is returned even for channels that are
// pending force close (close TX confirmed but sweep hasn't happened)
pendingChannelsResp = alice.RPC.PendingChannels()
require.Equal(ht, testMemo,
pendingChannelsResp.PendingForceClosingChannels[0].Channel.Memo)
// With the anchor output located, and the main commitment mined we'll
// instruct the wallet to send all coins in the wallet to a new address
// (to the miner), including unconfirmed change.
minerAddr := ht.Miner.NewMinerAddress()
sweepReq := &lnrpc.SendCoinsRequest{
Addr: minerAddr.String(),
SendAll: true,
TargetConf: 6,
MinConfs: 0,
SpendUnconfirmed: true,
}
sweepAllResp := alice.RPC.SendCoins(sweepReq)
// Both the original anchor sweep transaction, as well as the
// transaction we created to sweep all the coins from Alice's wallet
// should be found in her transaction store.
sweepAllTxID, _ := chainhash.NewHashFromStr(sweepAllResp.Txid)
ht.AssertTransactionInWallet(alice, aliceAnchor.SweepTx.TxHash())
ht.AssertTransactionInWallet(alice, *sweepAllTxID)
2024-04-08 10:48:43 +02:00
// Next, we mine enough blocks to pass so that the anchor output can be
// swept by anyone. Rather than use the normal API call, we'll generate
// a series of _empty_ blocks here.
//
// TODO(yy): also check the restart behavior of Alice.
const anchorCsv = 16
2024-04-08 10:48:43 +02:00
ht.MineEmptyBlocks(anchorCsv - defaultCSV - 1)
// Now that the channel has been closed, and Alice has an unconfirmed
// transaction spending the output produced by her anchor sweep, we'll
// mine a transaction that double spends the output.
thirdPartyAnchorSweep := genAnchorSweep(ht, aliceAnchor, anchorCsv)
2024-04-08 10:48:43 +02:00
ht.Logf("Third party tx=%v", thirdPartyAnchorSweep.TxHash())
ht.Miner.MineBlockWithTx(thirdPartyAnchorSweep)
// At this point, we should no longer find Alice's transaction that
// tried to sweep the anchor in her wallet.
ht.AssertTransactionNotInWallet(alice, aliceAnchor.SweepTx.TxHash())
// In addition, the transaction she sent to sweep all her coins to the
// miner also should no longer be found.
ht.AssertTransactionNotInWallet(alice, *sweepAllTxID)
// The anchor should now show as being "lost", while the force close
// response is still present.
assertAnchorOutputLost(ht, alice, aliceChanPoint1)
2024-04-08 10:48:43 +02:00
// We now one block so Alice's commit output will be re-offered to her
// sweeper again.
ht.MineEmptyBlocks(1)
ht.AssertNumPendingSweeps(alice, 1)
// At this point Alice's CSV output should already be fully spent and
// the channel marked as being resolved. We mine a block first, as so
// far we've been generating custom blocks this whole time.
commitSweepOp := wire.OutPoint{
Hash: *forceCloseTxID,
Index: 1,
}
ht.Miner.AssertOutpointInMempool(commitSweepOp)
ht.MineBlocks(1)
ht.AssertNumWaitingClose(alice, 0)
}
// assertAnchorOutputLost asserts that the anchor output for the given channel
// has the state of being lost.
func assertAnchorOutputLost(ht *lntest.HarnessTest, hn *node.HarnessNode,
chanPoint *lnrpc.ChannelPoint) {
cp := ht.OutPointFromChannelPoint(chanPoint)
expected := lnrpc.PendingChannelsResponse_ForceClosedChannel_LOST
err := wait.NoError(func() error {
resp := hn.RPC.PendingChannels()
channels := resp.PendingForceClosingChannels
for _, c := range channels {
// Not the wanted channel, skipped.
if c.Channel.ChannelPoint != cp.String() {
continue
}
// Found the channel, check the anchor state.
if c.Anchor == expected {
return nil
}
return fmt.Errorf("unexpected anchor state, want %v, "+
"got %v", expected, c.Anchor)
}
return fmt.Errorf("channel not found using cp=%v", cp)
}, defaultTimeout)
require.NoError(ht, err, "anchor doesn't show as being lost")
}
// genAnchorSweep generates a "3rd party" anchor sweeping from an existing one.
// In practice, we just re-use the existing witness, and track on our own
// output producing a 1-in-1-out transaction.
func genAnchorSweep(ht *lntest.HarnessTest,
2024-04-08 10:48:43 +02:00
aliceAnchor *lntest.SweptOutput, anchorCsv uint32) *wire.MsgTx {
// At this point, we have the transaction that Alice used to try to
// sweep her anchor. As this is actually just something anyone can
// spend, just need to find the input spending the anchor output, then
// we can swap the output address.
aliceAnchorTxIn := func() wire.TxIn {
sweepCopy := aliceAnchor.SweepTx.Copy()
for _, txIn := range sweepCopy.TxIn {
if txIn.PreviousOutPoint == aliceAnchor.OutPoint {
return *txIn
}
}
require.FailNow(ht, "anchor op not found")
return wire.TxIn{}
}()
// We'll set the signature on the input to nil, and then set the
// sequence to 16 (the anchor CSV period).
aliceAnchorTxIn.Witness[0] = nil
aliceAnchorTxIn.Sequence = anchorCsv
minerAddr := ht.Miner.NewMinerAddress()
addrScript, err := txscript.PayToAddrScript(minerAddr)
require.NoError(ht, err, "unable to gen addr script")
// Now that we have the txIn, we can just make a new transaction that
// uses a different script for the output.
tx := wire.NewMsgTx(2)
tx.AddTxIn(&aliceAnchorTxIn)
tx.AddTxOut(&wire.TxOut{
PkScript: addrScript,
Value: anchorSize - 1,
})
2024-04-08 10:48:43 +02:00
return tx
}
// testRemoveTx tests that we are able to remove an unconfirmed transaction
// from the internal wallet as long as the tx is still unconfirmed. This test
// also verifies that after the tx is removed (while unconfirmed) it will show
// up as confirmed as soon as the original transaction is mined.
func testRemoveTx(ht *lntest.HarnessTest) {
// Create a new node so that we start with no funds on the internal
// wallet.
alice := ht.NewNode("Alice", nil)
const initialWalletAmt = btcutil.SatoshiPerBitcoin
// Funding the node with an initial balance.
ht.FundCoins(initialWalletAmt, alice)
// Create an address for Alice to send the coins to.
req := &lnrpc.NewAddressRequest{
Type: lnrpc.AddressType_WITNESS_PUBKEY_HASH,
}
resp := alice.RPC.NewAddress(req)
// We send half the amount to that address generating two unconfirmed
// outpoints in our internal wallet.
sendReq := &lnrpc.SendCoinsRequest{
Addr: resp.Address,
Amount: initialWalletAmt / 2,
TargetConf: 6,
}
alice.RPC.SendCoins(sendReq)
txID := ht.Miner.AssertNumTxsInMempool(1)[0]
// Make sure the unspent number of utxos is 2 and the unconfirmed
// balances add up.
unconfirmed := ht.GetUTXOsUnconfirmed(
alice, lnwallet.DefaultAccountName,
)
require.Lenf(ht, unconfirmed, 2, "number of unconfirmed tx")
// Get the raw transaction to calculate the exact fee.
tx := ht.Miner.GetNumTxsFromMempool(1)[0]
// Calculate the tx fee so we can compare the end amounts. We are
// sending from the internal wallet to the internal wallet so only
// the tx fee applies when calucalting the final amount of the wallet.
txFee := ht.CalculateTxFee(tx)
// All of alice's balance is unconfirmed and equals the initial amount
// minus the tx fee.
aliceBalResp := alice.RPC.WalletBalance()
expectedAmt := btcutil.Amount(initialWalletAmt) - txFee
require.EqualValues(ht, expectedAmt, aliceBalResp.UnconfirmedBalance)
// Now remove the transaction. We should see that the wallet state
// equals the amount prior to sending the transaction. It is important
// to understand that we do not remove any transaction from the mempool
// (thats not possible in reality) we just remove it from our local
// store.
var buf bytes.Buffer
require.NoError(ht, tx.Serialize(&buf))
alice.RPC.RemoveTransaction(&walletrpc.GetTransactionRequest{
Txid: txID.String(),
})
// Verify that the balance equals the initial state.
confirmed := ht.GetUTXOsConfirmed(
alice, lnwallet.DefaultAccountName,
)
require.Lenf(ht, confirmed, 1, "number confirmed tx")
// Alice's balance should be the initial balance now because all the
// unconfirmed tx got removed.
aliceBalResp = alice.RPC.WalletBalance()
expectedAmt = btcutil.Amount(initialWalletAmt)
require.EqualValues(ht, expectedAmt, aliceBalResp.ConfirmedBalance)
// Mine a block and make sure the transaction previously broadcasted
// shows up in alice's wallet although we removed the transaction from
// the wallet when it was unconfirmed.
block := ht.Miner.MineBlocks(1)[0]
ht.Miner.AssertTxInBlock(block, txID)
// Verify that alice has 2 confirmed unspent utxos in her default
// wallet.
err := wait.NoError(func() error {
confirmed = ht.GetUTXOsConfirmed(
alice, lnwallet.DefaultAccountName,
)
if len(confirmed) != 2 {
return fmt.Errorf("expected 2 confirmed tx, "+
" got %v", len(confirmed))
}
return nil
}, lntest.DefaultTimeout)
require.NoError(ht, err, "timeout checking for confirmed utxos")
// The remaining balance should equal alice's starting balance minus the
// tx fee.
aliceBalResp = alice.RPC.WalletBalance()
expectedAmt = btcutil.Amount(initialWalletAmt) - txFee
require.EqualValues(ht, expectedAmt, aliceBalResp.ConfirmedBalance)
}
// testListSweeps tests that we are able to:
// - list only those sweeps that are currently in the mempool,
// - list sweeps given a starting block height,
// - list all sweeps.
func testListSweeps(ht *lntest.HarnessTest) {
// Skip this test for neutrino, as it's not aware of mempool
// transactions.
if ht.IsNeutrinoBackend() {
ht.Skipf("skipping ListSweeps test for neutrino backend")
}
// Create nodes so that we start with no funds on the internal wallet.
alice := ht.NewNode("Alice", nil)
bob := ht.NewNode("Bob", nil)
const initialWalletAmt = btcutil.SatoshiPerBitcoin
// Fund Alice with an initial balance.
ht.FundCoins(initialWalletAmt, alice)
// Connect Alice to Bob.
ht.ConnectNodes(alice, bob)
// Open a few channels between Alice and Bob.
var chanPoints []*lnrpc.ChannelPoint
for i := 0; i < 3; i++ {
chanPoint := ht.OpenChannel(
alice, bob, lntest.OpenChannelParams{
Amt: 1e6,
PushAmt: 5e5,
},
)
chanPoints = append(chanPoints, chanPoint)
}
ht.Shutdown(bob)
// Close the first channel and sweep the funds.
ht.ForceCloseChannel(alice, chanPoints[0])
// Jump a block.
ht.MineBlocks(1)
// Get the current block height.
bestBlockRes := ht.Alice.RPC.GetBestBlock(nil)
blockHeight := bestBlockRes.BlockHeight
// Close the second channel and also sweep the funds.
ht.ForceCloseChannel(alice, chanPoints[1])
// Now close the third channel but don't sweep the funds just yet.
closeStream, _ := ht.CloseChannelAssertPending(
alice, chanPoints[2], true,
)
ht.AssertStreamChannelForceClosed(
alice, chanPoints[2], false, closeStream,
)
// Mine enough blocks for the node to sweep its funds from the force
// closed channel. The commit sweep resolver offers the outputs to the
// sweeper up to one block before the CSV elapses, so wait until
// defaulCSV-1.
ht.MineEmptyBlocks(node.DefaultCSV - 1)
ht.AssertNumPendingSweeps(alice, 1)
// Mine a block to trigger the sweep.
2024-04-08 10:48:43 +02:00
ht.MineEmptyBlocks(1)
// Now we can expect that the sweep has been broadcast.
pendingTxHash := ht.Miner.AssertNumTxsInMempool(1)
// List all unconfirmed sweeps that alice's node had broadcast.
sweepResp := alice.RPC.ListSweeps(false, -1)
txIDs := sweepResp.GetTransactionIds().TransactionIds
require.Lenf(ht, txIDs, 1, "number of pending sweeps, starting from "+
"height -1")
require.Equal(ht, pendingTxHash[0].String(), txIDs[0])
// Now list sweeps from the closing of the first channel. We should
// only see the sweep from the second channel and the pending one.
sweepResp = alice.RPC.ListSweeps(false, blockHeight)
txIDs = sweepResp.GetTransactionIds().TransactionIds
require.Lenf(ht, txIDs, 2, "number of sweeps, starting from height %d",
blockHeight)
// Finally list all sweeps from the closing of the second channel. We
// should see all sweeps, including the pending one.
sweepResp = alice.RPC.ListSweeps(false, 0)
txIDs = sweepResp.GetTransactionIds().TransactionIds
require.Lenf(ht, txIDs, 3, "number of sweeps, starting from height 0")
// Mine the pending sweep and make sure it is no longer returned.
2024-04-08 10:48:43 +02:00
ht.MineBlocksAndAssertNumTxes(1, 1)
sweepResp = alice.RPC.ListSweeps(false, -1)
txIDs = sweepResp.GetTransactionIds().TransactionIds
require.Empty(ht, txIDs, "pending sweep should not be returned")
}