[need_help] restoring singlebackup with restorechanbackup fails

[warioishere]

Background

I am trying to restore a single channel from a single backup file. It does not work, see details below.

Your environment

• version of lnd
  lnd 0.17.4 rc1 (containerized by btcpay-docker)

• which operating system (uname -a on *Nix)
  Linux btcpay 6.5.11-7-pve SMP PREEMPT_DYNAMIC PMX 6.5.11-7 (2023-12-05T09:44Z) x86_64 x86_64 x86_64 GNU/Linux

• version of btcd, bitcoind, or other backend
  26.0

• any other relevant environment details
  Running the node on Proxmox

Steps to reproduce

Trying to export a specific channel with exportchanbackup --chan_point 9110b04f1ec66f20e3d7a8e0a8e15ce10f0a4c8729872c3a5cd036093829ef4b:1 --output_file=/root/.lnd/bitcoinsuisse.backup
and restoring with

restorechanbackup --single_file=/root/.lnd/bitcoinsuisse.backup

Expected behaviour

channel will will be restored from single backup file

Actual behaviour

Does not work:

[lncli] unable to restore chan backups: rpc error: code = Unknown desc = unable to unpack single backups: could not derive private key for legacy channel revocation root format: unable to derive private key

Additional Info: the restore was already triggered with a full restorechanbackup some months ago, but it did not trigger a forceclose on this specific channel on the other side propably due to high fee enviroment that time.

[guggero]

Are you exporting the backup from the node you initially ran restorechanbackup with the full file on? Do you still have the full file available?

[warioishere]

Are you exporting the backup from the node you initially ran restorechanbackup with the full file on? Do you still have the full file available?

Hello Oli

yes its the same node,maybe you can remember the issue I had with my SCB case we have discussed arround Christmas/Newyear

I never changed or touched the original SCB file so I guess its the original SCB file, where I export the single backup now and from which I did the full restore that time ago. I continued using the same node normally since then.

PS: ah 'exportchanbackup' does not export the single backup from the original SCB file I guess. Now I need a hint what to do because yes, this original file is still there because as I mentioned, never touched it.

Greetings

[guggero]

Yes, I remember. And I think the issue is indeed that you're exporting a SCB from an SCB which is IMO not fully supported.
What you can do instead is take the original file and use chantools filterbackup to reduce the file to one that only has the channel in question (meaning filter out all other channels). Then use that (as a multi_file but with just one entry).

[guggero]

But I'm not sure doing so would change anything. If you say the node you used the full SCB on is still running like that, then it should periodically try and request the peer to force close. Or at least that's what I think it should. Maybe it only happens on startup though. So probably still worth a try, just want to manage expectations.

[warioishere]

Yes, I remember. And I think the issue is indeed that you're exporting a SCB from an SCB which is IMO not fully supported. What you can do instead is take the original file and use chantools filterbackup to reduce the file to one that only has the channel in question (meaning filter out all other channels). Then use that (as a multi_file but with just one entry).

I understand the issue when you told me that I exported a channelback from an already restored scb channel.
I will try chantools later.

Do you think chantools triggerforce might do the trick?

I have contact with bitcoinsuisse, they cannot manually force close, they say they have the issue that says channel already closed when they try to force close again.
I dont get much information because I talk to a relation manager which passed me few infos from the IT department. I told them already to raise mempool, restart their node afterwards and do this and that. But I get an answer with almost no Infos once in a week and I dont know what they already tried and what they didnt try yet. Just saying infrastrural problems.

[guggero]

Yes, that sounds good. If they have trouble finding the "localclosing tx", it should appear in lncli listchaintxns with the full hex so they can re-broadcast manually.

[warioishere]

I will insist again ask them to send me the output of their lncli pendingchannels
and their localclosingtx as a hex from lncli listchaintxns --start_height -1 --end_height <curent_block_height>

and see what get, then I will try to tell them to republish with lncli wallet publishtx <closinghx>

I dont know what to believe, if they dont know how to operate their node, or if they really suffer from a bug.

I'll let you know if you are interessted.

Greetings

[warioishere]

Hi @guggero

Bitcoinsuisse finally made it to close this channel, guess they had no time or it had no priority. The closing is already confirmed:

https://mempool.space/tx/6d291145ee041279563f4ea94985df758775c3d66f51032c488e8cb0bb56ece0

However, the sats are not on my wallet. lncli wallet pendingsweeps gives me:

root@btcpay:~/btcpayserver-docker/docker-compose-generator/docker-fragments# bitcoin-lncli.sh wallet pendingsweeps
{
        "pending_sweeps": [
                {
                        "outpoint": "6d291145ee041279563f4ea94985df758775c3d66f51032c488e8cb0bb56ece0:1",
                        "witness_type": "COMMITMENT_TO_REMOTE_CONFIRMED",
                        "amount_sat": 1996530,
                        "sat_per_vbyte": 0,
                        "broadcast_attempts": 0,
                        "next_broadcast_height": 831428,
                        "requested_sat_per_vbyte": 0,
                        "requested_conf_target": 6,
                        "force": false
                },
                {
                        "outpoint": "6d291145ee041279563f4ea94985df758775c3d66f51032c488e8cb0bb56ece0:0",
                        "witness_type": "COMMITMENT_ANCHOR",
                        "amount_sat": 330,
                        "sat_per_vbyte": 1,
                        "broadcast_attempts": 0,
                        "next_broadcast_height": 831428,
                        "requested_sat_per_vbyte": 1,
                        "requested_conf_target": 0,
                        "force": false
                }
        ]
}

I already restarted lnd. I have sweeper.batchwindowduration=2h and this time already passed by.
I find "requested_sat_per_vbyte": 0, and "requested_sat_per_vbyte": 1, for the anchor pretty weird.

I dont have any txns in listchaintxns

This is the output of lncli pendingchannels

"pending_force_closing_channels": [
        {
            "channel": {
                "remote_node_pub": "027222caea9c1070952bc8c7a05a6f75f26f3bffc37e80e627cc651d50d50eb268",
                "channel_point": "9110b04f1ec66f20e3d7a8e0a8e15ce10f0a4c8729872c3a5cd036093829ef4b:1",
                "capacity": "2000000",
                "local_balance": "1996530",
                "remote_balance": "0",
                "local_chan_reserve_sat": "0",
                "remote_chan_reserve_sat": "0",
                "initiator": "INITIATOR_LOCAL",
                "commitment_type": "ANCHORS",
                "num_forwarding_packages": "0",
                "chan_status_flags": "",
                "private": true,
                "memo": ""
            },
            "closing_txid": "6d291145ee041279563f4ea94985df758775c3d66f51032c488e8cb0bb56ece0",
            "limbo_balance": "1996860",
            "maturity_height": 831423,
            "blocks_til_maturity": -8,
            "recovered_balance": "0",
            "pending_htlcs": [],
            "anchor": "LIMBO"
        }
    ],

blocks till maturiy is already negativ.

You got a hint for me what to do? Just wait?

Thankyou very much

Wario

[guggero]

Looks like the sweep just confirmed: https://mempool.space/tx/56258c0024fd0260d409317702cf709e4d7dd0b8e55dcff9ca6c66285cebcbbe

[warioishere]

yes saw it now, I didnt know how those sweeps and stuff worked. Learned a bit more by reading
https://docs.lightning.engineering/the-lightning-network/payment-channels/understanding-sweeping

so this topic can be finally put at acta :)
thanks again for you time and your help!

Have a nice evening Oli!