-
Notifications
You must be signed in to change notification settings - Fork 912
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
can't close channel in state "CHANNELD_SHUTTING_DOWN" #1865
Comments
How long have you been waiting for the closing state to resolve? In recent versions the channel should eventually fall back to a unilateral close automatically. What version are you running? |
v0.5.2-2016-11-21-2069-gb8c6365 it's old I guess... I'll install a new version and let you know. |
I now run version: v0.6-392-g898655f This is the problematic channel now: "peers": [ `
debuglog: 2018-08-24T14:43:58.306Z lightning_gossipd(20267): Received channel_update for 518494:2113:1 with old time 1526502819 2018-08-24T14:43:58.321Z lightning_gossipd(20267): Received channel_announcement for channel 518510:1692:0 2018-08-24T14:43:58.333Z lightningd(20257):jcon fd 22: Failing close: Channel is in state AWAITING_UNILATERAL |
Can you share the output of |
thanks! is this what you're after?
|
Ok, the first transaction is what I was looking for: {
"txid": "229a60c362f111fc1b9996f64104bb845adfbb4a3f5ba826ac3e5e37cd903d0c",
"hash": "cc3dbb6e32679b0993cd41b572f5e166f010b77ad8469498bda55ccaa09f0230",
"version": 2,
"size": 271,
"vsize": 106,
"locktime": 548643980,
"vin": [
{
"txid": "7dc5a110d9e890c7e2e164f1eb616fb25abfd9186b9a0aa52b9a2d7d60735bfe",
"vout": 1,
"scriptSig": {
"asm": "",
"hex": ""
},
"txinwitness": [
"",
"304402203d64688221eac6c2e22487980015ef95c5d39c1bce35895ea5fe9627e4d0d5d302204f592f9ae7123ee694ba488b015f72861c4de1b26945753b9094741f404136c501",
"30440220303b779de86407a2bba60235ffdb1bf3102047d7bd6a05e95233f2c0ce49e9bc02202c22978648d6b7c060112d7b488b205dadad5fd3949cc4cec07dc723a541c7de01",
"5221021941c80dd3899cacf8a2171385fc0129c8b6ee1e34ddac3febb0af84c3f2372d2102eba9c89d4c3a6a18b9d8b6129a900b788876ef5d61b5be0e4291aedd9608649b52ae"
],
"sequence": 2151323968
}
],
"vout": [
]
} The part that is annoying here is that the closing transaction has no output. This is a known (and fixed) issue where the fee would dominate the entire value and no funds would actually be sent back, however a 0 output transaction is invalid to bitcoin clients, and will therefore never confirm. The version you have there (b8c6365) is very old, that's why it's still producing these close transactions. We have two options here:
|
For reference: the commit that fixed this is 45e145d |
For some reason this does not work.
|
I have installed and use a new version now. How do i reset the channel state or how do i fail the channel? |
I found another problem on my node, regarding a different much more recent channel. Now using the latest version of clightning:
It seems multiple channels have the problem on my pi. How to get rid of these failed channels as dev-forget-channel does not seem to be available in v0.6-392-g898655f? (with or without recovering the funds) |
The
Then restart the node and you should have the |
Thanks. I have tried this but still no luck unfortunately. After I check the results the channel is still there. As the command suggests using 'close' crashes clighting:
|
Can you give result of |
This last command hangs clightningd with this in the log:
Or it goes like this:
With the following in the log:
|
The above is because you should use
If you will use The crash on |
I think you will also need to specify that you're using the key-value format by having the first cli argument be
|
Not the result we're looking for unfortunately.
with some logs:
And a different try:
Results in a crash with no unusual logs. The command dev-forget-channel also doesn't help out here:
The channel is still there. Logs:
I'm out of ideas here. Unless his has something to do with it:
|
OK, I see, the channels that give you I don't think we're going to be able to fix this... I checked on the remote end as well, and it also has the botched close transaction. The only thing I can offer is to hack the |
Thanks. That's good enough. |
Ok, if you remove these lines and recompile you should be able to forget with the force option: lightning/lightningd/peer_control.c Lines 1244 to 1252 in f2e0857
|
I have commented out those lines in the source and recompiled.
results in a crash:
With the peer still being there:
|
Oh, I missed that assertion. You also need to comment out these lines: lightning/lightningd/channel.c Lines 87 to 95 in 634f19a
|
Thanks a lot for all the help. This works. :-) |
Glad we could get this cleaned up, and sorry for the hoops you had to jump through, but the code that was causing this has been fixed for a while now. |
Issue and Steps to Reproduce
I cannot close a channel that is in a state "CHANNELD_SHUTTING_DOWN" for ages now. For some reason it seems as if the api does not accept this unilateral close and tries to make it bilateral which the nodes can't negotiate, bringing me to a kind of "deadlock" situation. I guess an api call like "froce close" would be good since publishing my commitment tx should always work
Steps to reproduce:
https://imgur.com/a/rYUlEu6
getinfo
outputThe text was updated successfully, but these errors were encountered: