Loop In by joostjager · Pull Request #34 · lightninglabs/loop

joostjager · 2019-04-01T22:15:33Z

This PR adds loop in functionality to the Lightning Loop client.

Loop in allows a lightning channel to be 'refilled' with on-chain coins. It allows a channel to remain open and offers an alternative over closing and reopening the channel.

High level flow

User generates a secret preimage.
User publishes an on-chain htlc tied to the hash of the secret preimage.
Server picks up confirmation of the on-chain htlc.
Server sends an off-chain lightning payment to the user.
User settles the payment by revealing the secret preimage.
Server learns the preimage and uses this to sweep the on-chain htlc.

Usage

Similar to Loop Out. A new command has been added to the loop command line tool:

loop in <amt>

Also the --external flag can be supplied. In that case step 2 of the flow above is skipped and the user is given the opportunity to publish the htlc manually. An example use of this is to withdraw from an exchange into a lightning channel by looping in.

There is also the --channel flag that allows selection of the channel to loop in to. This functionality is not implemented yet.

Testnet only

Loop in is currently only supported on testnet. There are still several changes required to lnd for it to be used on mainnet.

This commit adds the required code to persist loop in swaps. It also introduces the file loop.go to which shared code is moved. Sharing of contract serialization/deserialization code has been reverted. The prepay fields do not apply to loop in, but were part of the shared contract struct. Without also adding a migration, it wouldn't be possible to keep the shared code. In general it is probably more flexible to keep the contract serialization code separated between in and out swaps.

Allow user to specify that htlc will be published by an external source. In that case no internal htlc tx will be published. To facilitate sending to the htlc address, the swap initiation response is extended with that address.

Roasbeef

Excited to eventually roll this out on mainnet! Comments from the prior earlier review look to be addressed. After running a few more Loops I noticed two things:

Loop In quotes still show "loop out"
We mark the Loop In as complete after the HTLC is swept, rather than as soon as the invoice has been settled

Roasbeef · 2019-04-02T00:04:46Z

+		// TODO: Add miner fee of timeout tx to swap cost balance.
+	}
+
+	// Wait for a final state of the swap invoice. It should either be


I think we should move this to be above waiting for the HTLC spend, or be concurrent even. I tried another Loop in on testnet now and realized that things are shown as pending until the HTLC is swept on chain. However, from the PoV of the client, the Loop In process is complete once its invoice is settled, as it now has more outbound capacity.

Everywhere in loop swaps are only marked complete when all information needed for accounting is there. For loop in on the client, this means the htlc sweep tx needs to be confirmed and the swap invoice either canceled or settled. Also after a restart, a swap that is completed from the user PoV, still needs to be resumed to complete the accounting info.

An intermediate state that the swap transitions to when the invoice is paid could be a solution here.

Added a commit that adds this intermediate state.

Roasbeef · 2019-04-02T00:12:05Z

The README should also be updated to instruct users to build with the invoicerpc tag as well.

joostjager · 2019-04-04T13:52:09Z

@Roasbeef ptal

Roasbeef · 2019-04-05T00:03:22Z

Panic on loop monitor w/ the latest attempting a Loop In:

oroutine 68 [running]:
main.(*swapClientServer).marshallSwap(0xc0001c2290, 0xc00024f530, 0xc0006fe420, 0xb51840, 0xc000343dc0)
        /root/loop/cmd/loopd/swapclient_server.go:122 +0x134
main.(*swapClientServer).Monitor.func1(0x6d5a1c0, 0xed42e0803, 0x14cc500, 0xb680a30070e9430c, 0xaa124c441abd7ade, 0xc887a080307fdaae, 0xd00a71501b250270, 0x102, 0x511c5b616b2f239f, 0x7ef9a4a62c974031, ...)
        /root/loop/cmd/loopd/swapclient_server.go:134 +0x4a
main.(*swapClientServer).Monitor(0xc0001c2290, 0xc000343d20, 0xef9960, 0xc000174a30, 0x0, 0x0)
        /root/loop/cmd/loopd/swapclient_server.go:197 +0x877
github.com/lightninglabs/loop/looprpc._SwapClient_Monitor_Handler(0xc3b860, 0xc0001c2290, 0xef84c0, 0xc0001049c0, 0x14eaa10, 0xc000102c00)
        /root/loop/looprpc/client.pb.go:975 +0x109
google.golang.org/grpc.(*Server).processStreamingRPC(0xc000200180, 0xefc2a0, 0xc00024b800, 0xc000102c00, 0xc0001bf0b0, 0x14b49e0, 0x0, 0x0, 0x0)
        /root/gocode/pkg/mod/google.golang.org/grpc@v1.19.0/server.go:1175 +0xb12
google.golang.org/grpc.(*Server).handleStream(0xc000200180, 0xefc2a0, 0xc00024b800, 0xc000102c00, 0x0)
        /root/gocode/pkg/mod/google.golang.org/grpc@v1.19.0/server.go:1254 +0xcbe
google.golang.org/grpc.(*Server).serveStreams.func1.1(0xc000332030, 0xc000200180, 0xefc2a0, 0xc00024b800, 0xc000102c00)
        /root/gocode/pkg/mod/google.golang.org/grpc@v1.19.0/server.go:690 +0x9f
created by google.golang.org/grpc.(*Server).serveStreams.func1
        /root/gocode/pkg/mod/google.golang.org/grpc@v1.19.0/server.go:688 +0xa1

Roasbeef · 2019-04-05T00:04:45Z

Nvm, it was due to me having issued a swap using the old DB format.

Roasbeef · 2019-04-05T00:11:24Z

 	return s.sendUpdate(ctx)
 }
+
+// setState updates the swap state and last update timestamp.


I think we should also roll in persistState into this method as well. Otherwise, I see a few instances above, where we modify the swap state, but don't persist it to the database.

Scratch that, I see it's persisted for the first thing that causes it to break out of the loop.

Roasbeef · 2019-04-05T00:17:00Z

+	return nil
+}
+
+func viewOut(swapClient *loop.Client, chainParams *chaincfg.Params) error {


I think we should instead make this an command for the loop binary, but it isn't blocking and can be done in a follow up.

Roasbeef

LGTM 🐝

joostjager added 5 commits March 28, 2019 11:28

loopdb: move code to shared code file

e81298c

multi: loop in swap

3e960b8

loopd: disable mainnet loop in

8b2da64

loopd: loop in from external address

2e48ead

Allow user to specify that htlc will be published by an external source. In that case no internal htlc tx will be published. To facilitate sending to the htlc address, the swap initiation response is extended with that address.

Roasbeef reviewed Apr 2, 2019

View reviewed changes

joostjager added 4 commits April 2, 2019 10:54

loop: hide loop in channel selection flag

f7f9751

loop: show correct swap type

f7676c3

README: add invoicesrpc tag

a7f4feb

loopd: add loop in state InvoiceSettled

b5d2fb3

joostjager force-pushed the loopin-merge branch 3 times, most recently from b2e7bf4 to db39b37 Compare April 4, 2019 12:08

loopd: check loop in timeout already before a new block comes in

e765baa

joostjager force-pushed the loopin-merge branch 2 times, most recently from a25b081 to 907a118 Compare April 4, 2019 13:40

use np2wsh for loop in htlc

30c7d71

joostjager force-pushed the loopin-merge branch from 907a118 to 30c7d71 Compare April 4, 2019 13:50

Roasbeef reviewed Apr 5, 2019

View reviewed changes

Roasbeef approved these changes Apr 5, 2019

View reviewed changes

Roasbeef merged commit e8005d0 into lightninglabs:master Apr 5, 2019

Conversation

joostjager commented Apr 1, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

High level flow

Usage

Testnet only

Uh oh!

Roasbeef left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Roasbeef Apr 2, 2019

Choose a reason for hiding this comment

Uh oh!

joostjager Apr 2, 2019

Choose a reason for hiding this comment

Uh oh!

joostjager Apr 2, 2019

Choose a reason for hiding this comment

Uh oh!

Roasbeef commented Apr 2, 2019

Uh oh!

joostjager commented Apr 4, 2019

Uh oh!

Roasbeef commented Apr 5, 2019

Uh oh!

Roasbeef commented Apr 5, 2019

Uh oh!

Roasbeef Apr 5, 2019

Choose a reason for hiding this comment

Uh oh!

Roasbeef Apr 5, 2019

Choose a reason for hiding this comment

Uh oh!

Roasbeef Apr 5, 2019

Choose a reason for hiding this comment

Uh oh!

Roasbeef left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

joostjager commented Apr 1, 2019 •

edited

Loading