Apologies for the disruption. Part of the fun of being early in Lightning is discovering new, undocumented issues, that nobody has found before. We're working with the best minds in Lightning currently to get the channels back online. It's a networking issue specific to an "unfortunate" interaction of two implementations Lightning protocol, that (as far as we know) was a theoretical issue that we realized could possibly happen -- but never actually seen "in the wild" before about 12 hours ago... it's not a failure of the LSP node, a hardware issue, or anything like that. Alby Hub is a pioneer and we're the biggest LSP on Alby Hub, so we have the joy of seeing the interesting/hard problems before anyone else does.

Reply to this note

Please Login to reply.

Discussion

Thanks to a fix suggested provided by Matt Corallo, I believe that this issue is now fixed and all of the Alby nodes will see that their Megalith LSP channels come back online over the next hour or so. Apologies again for the disruption, but you can remember you were around for an actual discovery of a potential problem in Lightning that had never been discovered before....

not a fan of taproot assets, so i force closed

wouldn't have known that if you had tested updating your node without ensuring it wasn't gonna cause interop problems

i'm glad i know, i don't want to be part of funding lightning shitcoinery, sorry, not sorry

What was the issue that was fixed exactly?

We discovered a vulnerability in the interaction of two different Lightning implementations. It creates a situation close to DOS (denial of service), so we're not publicly disclosing how it works -- if we did, we could enable a bad actor to attempt to bring down other services. But the implementations involved are both aware of the situation and will be patching it in an upcoming release. We also remediated the situation on our end so we're no longer vulnerable.

Wow! Amazing work