-
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]: "edge not found" for local channel #7261
Comments
Something weird happened during channel open. It looks like the second open attempt worked?
The open transaction is confirmed and, according to mempool.space, opened two other channels to other peers. |
I also get this:
The node_announcement is extremely long, filling roughly one screen. |
Discussed offline, will post more details here after proper investigation. |
I think this is distinct from #7229 since this was not a zero-conf channel |
if edge for local channel cannot be found see lightningnetwork/lnd#7261
i get this issue too for a few channels |
@dannydeezy with what version of lnd? @C-Otto does this still happen with lnd 0.16.1? |
I'm running lnd 0.16.0 with some WT related commits, not 0.16.1. Currently, I don't see this issue. The last time the error message appeared in my logs was 2023-04-09. I updated to lnd 0.16.0 a few days before, so I think this is not resolved, yet. |
We're running into this also, on nodes running 0.16.0 |
For those who recently experienced this bug, pls can you try updated to 0.16.1-beta.rc1 to see if that fixes your issue? It might be that your issue is the same as this user's issue which is distinct from this issue even though it throws the same error message. That user's issue was fixed in 0.16.1-beta.rc1 |
Same here after upgrading to
Edit: After restarting only LND node the problem did not go away. However, after restarting the other node (CLN) everything started to work as expected. |
Thanks for the info @theborakompanioni! Question to everyone else experiencing this issue - is your channel peer for the problem channel a CLN node? |
No, both sides of the channel were LND in our case (we saw this a few times) |
I have same case with one of my peers. My node: LND 0.16.0-beta commit=v0.16.0-beta We both had "[lncli] rpc error: code = Unknown desc = edge not found" (!) The issue was resolved on my peer's side by upgrading LND 0.16.0 -> 0.16.1-beta. Immediately after that the problem disappeared and the channel became active. On my side the issue is currently active, obviously I need to upgrade my LND to 0.16.1 too, but I haven't done it yet. Hope this information is helpful. Addition logs from my side:
|
hi @SunnySarahNode - i think your specific case was solved in 0.16.1-beta (in #7613) . Pls can you update & see if that works? your peer managing to solve the problem on their side by upgrading makes me think it is the same as this user: #6987 (comment) |
Closing this issue, as the fix has been provided with 0.16.1. Feel free to reopen if the problem persists. |
@saubyk |
Hello, If I run, I get the following error lncli getchaninfo 9133577121506xxxxx Version: 0.17.4-beta I have attached the log trace of the opening transaction till I received the error. Maybe you can make some sense out of it. Would be nice to be able to use the channel. Big thanks!! |
From the logs, the policy was not found even in
which is originated here when the edge cannot be found: Lines 1024 to 1026 in 98c52df
These four lines are also weird,
As we'd expect lnd/netann/chan_status_manager.go Lines 353 to 359 in 98c52df
There's also a disconnect/connect attempt here, I assume it's triggered manually?
@miephos I've noticed the chan was forced closed - do you by chance still have the logs? What about the channel type, like is it a public or private chan? |
The channel was a public channel. I did a manual disconnect and connect to peer. It still seemed stuck. So I force closed the the channel. I still have the logs. Let me know how to provide them, if needed. |
@miephos it'd be great to provide some logs! You can find me via the handle |
Yeah really strange behaviour would it make sense to bring this log to the Error level, so that we at least know when the edge failed to be added to the graph, otherwise there are some situations where we basically shadow the err which makes it really hard to tell what happened if node runners don't run the gossiper in the debug level ? Lines 2532 to 2533 in 71ba355
|
Background
A peer opened a channel to my node. This channel is listed in
lncli listchannels
. However,lncli getchaninfo X
shows an error (with X being the channel ID):According to
lncli listchannels
, this appears to be a regular (inactive) channel. It's not private, it has "remote_constraints", there's an ID in "alias_scids". It's not a zero-conf channel. Strangely, there are two incoming HTLCs already, where "forwarding_channel" is set to 0. Currently, the channel is inactive, but it has been active before.When reconnecting to the peer, I get:
In my logs I also see:
I'd be happy to provide further information.
Your environment
The text was updated successfully, but these errors were encountered: