-
Notifications
You must be signed in to change notification settings - Fork 2.1k
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
[bug]: Unknown desc = edge not found #6987
Comments
Hmm.. I was also able to reproduce that a couple of times locally on a regtest setup.. Will see if I can get that error again and share some context. Although I was experimenting with 0-conf channels so could be same bug but different origin. |
I had the same problem with newly created channel today. Fixed by restarting lnd. LND version: 0.15.2-beta |
@dannydeezy and @crywolf: this was fixed by #6892. Closing this issue in the meantime, please re-open it if |
Hello. Same issue just happened to me in 0.15.5 |
Closed a channel with peer and reopened new one, that one got confirmed, but strangely fees and other channel attributes were not set at all and immediately it went inactive. Any attempt to change the parameters fail with similar "reason": "UPDATE_FAILURE_UNKNOWN", Logs are showing channel trying to re-establish, but failing due to edge not found |
Hey @Wired4ncer you could maybe try an update to 0.16 rc1. For some context, smiliar to OP, the mentioned channel is shown as active in |
Yes that exactly was happening. Channel didn't receive any params when opened (fees, htlc size, ctv etc.) and was failing in logs due to not having it when trying to reestablish, but I was not able to update it. Ended up FC'ing it and reopening. Second time it all went fine. I don't have it open anymore, so updating to 0.16 at the moment won't give us any results. |
This just happened to me too. In particular, the opening transaction was paying a very low fee (1sat/vb), so the channel has been in state pending for a very long time, for about three weeks. Then I managed to CPFP fee bump the transactions (and I had to broadcast again the opening transaction via external block explorer). The opening transaction eventually got confirmed, but the channel is now in "inactive" state in |
The remote peer will forget the channel after 2 weeks (2016 blocks). So yes, unfortunately force closing is the only option. We'll try to make this more clear with this: #7434 |
Just had the same issue on a batch open, all the channel have the same problem.
|
I can add other info about this: I chatted with one of the peers, on his side all looks fine, he tried to change the fees and worked. The channel isn't announced but isn't marked as private.
on the logs I found this error
Logs[INF] BRAR: Breach arbiter starting [INF] NTFN: New spend subscription: spend_id=16, outpoint=xxxxxxxxx:1, script=0 zzzzzzz, height_hint=yyyyyy [INF] CNCT: Close observer for ChannelPoint(xxxxxxxxx:1) active [INF] CNCT: ChannelArbitrator(xxxxxxxxx:1): starting state=StateDefault, trigger=chainTrigger, triggerHeight=yyyyyy [WRN] CHBU: Replacing disk backup for ChannelPoint(xxxxxxxxx:3) w/ newer version [INF] HSWC: Trimming open circuits for chan_id=x:y:z, start_htlc_id=695 [INF] LNWL: Marking unconfirmed transaction xxxxxxxxx mined in block block111 [INF] CRTR: Block block222 (height=block222) closed 2 channels [INF] CRTR: Block block333 (height=block333) closed 0 channels |
@bota87 - So I think we only add the channel to our local graph after receiving the |
Hi @ellemouton, all it's happened in the same day (batch channels open request and transaction confirmation). |
Here the logs from the peer, he said that didn't find any ChannelReady message |
further up in your logs, do you at all see a log starting with (also, don't worry about |
another question: when exactly did you upgrade to 0.16? did you initiate the funding and update to 0.16 afterwards or was this all done after upgrading to 0.16? |
no the only messages talking about policy is this one
I'm not sure, I migrated from 0.15.4 to 0.16 on the same day but didn't remember if I start the batch channel open before or after the update |
ok cool - I managed to reproduce your scenario. Thanks for reporting! Working on a fix asap |
That's great! The fix will let me use this channels or I have to close and reopen? I have to wait for the fix before opening other channels? |
I think the fix will let you use the channels as long as the peer doesnt force close before you apply the fix. It should be safe to open other channels in the mean time - the error resulted from the channel being opened before upgrading to 0.16, then upgrading to 0.16 before channel confirmation. |
Updated to 0.16.1.rc1, the channels works! 🥳 |
Closing this issue. Feel free to reopen if the reported problem has not been addressed. |
Still happens on most recent release.
|
i get this opening fee modal in RTL.
|
Background
Seems like my LND will randomly not be able to find active channels it has. I do
lncli listchannels
and the channel shows up as active, but then when I dolncli getchaninfo
with the id, it saysedge not found
.This has happened a lot, especially when the other node is a Tor node with only 1 total channel.
lncli version 0.15.1-beta commit=v0.15.1-beta
Your environment
lncli version 0.15.1-beta commit=v0.15.1-beta
bitcoind
backend Bitcoin Core version v22.0.0Steps to reproduce
I don't know how to reproduce, it's flaky.
Example:
lncli listchannels
and i see the following entry (among my 800+ other channels)but then doing
lncli getchaninfo 832475437786923009
immediately alfter, i get:logs from lnd:
If i restart the node, it fixes it. But I can't be restarting the node all the time.
The text was updated successfully, but these errors were encountered: