Uh oh!
There was an error while loading. Please reload this page.
Store and check channel data in channel monitors - #4218
Conversation
👋 I see @valentinewallace was un-assigned. |
dda4c5c to
b00ff03Comparec3771b6 to
f387f90CompareCodecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@## main #4218 +/- ##
==========================================
- Coverage 89.38% 89.34% -0.04%
==========================================
Files 180 180 Lines 139834 139921 +87 Branches 139834 139921 +87 ==========================================
+ Hits 124985 125008 +23 - Misses 12262 12324 +62 - Partials 2587 2589 +2
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
00a95f1 to
a82719bComparesafe_channels feature flag that gates in-development work toward persisting channel monitors and channels atomically, preventing them from desynchronizing and causing force closures. This commit begins that transition by storing both pieces together and adding consistency checks during writes. These checks mirror what the channel manager currently validates only on reload, but performing them earlier increases coverage and surfaces inconsistencies sooner.safe_channels feature flag that gates in-development work toward persisting channel monitors and channels atomically, preventing them from desynchronizing and causing force closures. This commit begins that transition by storing both pieces together and adding consistency checks during writes. These checks mirror what the channel manager currently validates only on reload, but performing them earlier increases coverage and surfaces inconsistencies sooner.a82719b to
d4d0879Compare| /// The encoded channel data associated with this ChannelMonitor, if any. | ||
| #[cfg(feature = "safe_channels")] | ||
| pub encoded_channel: Option<Vec<u8>>, |
There was a problem hiding this comment.
Continuing discussion here @valentinewallace#4151 (comment)
Did you explore encoding the FundedChannel itself rather than a bag of bytes? I guess it would require propagating the signer trait parameter, which is kind of annoying?
I think it is possible, but with a few workarounds. The advantage would be that we have a clear target struct to strip all the data off that is redundant for channel monitor.
We do need to allow cloning of the channel to get a copy to store inside the monitor and monitor update. Also equality needs to be implemented. And it is the case that a channel contains a list of blocked monitor updates, that can again contain a channel copy. Not sure if we getting into something undesirable with that form of nesting.
The pending_splice field also turned out to be difficult to clone. For now, I just took the serialized version of it.
There was a problem hiding this comment.
Problems above have mostly been solved now
Tests now pass with the unencoded channel as part of Isolated repro: https://github.com/lightningdevkit/rust-lightning/compare/main...joostjager:repro-recursive-ser-issue?expand=1 |
2dc7ca2 to
17f71b0Compare44538f9 to
4c035abCompareab4f775 to
4c69502CompareAdd a new `safe_channels` feature flag that gates in-development work toward persisting channel monitors and channels atomically, preventing them from desynchronizing and causing force closures. This commit begins that transition by storing both pieces together and adding consistency checks during writes. These checks mirror what the channel manager currently validates only on reload, but performing them earlier increases coverage and surfaces inconsistencies sooner.
4c69502 to
9114c1dComparejoostjager
commented
Mar 20, 2026
Implemented safe channels via #4351 |
Add a new
safe_channelsfeature flag that gates in-development work toward persisting channel monitors and channels atomically, preventing them from desynchronizing and causing force closures.This commit begins that transition by storing both pieces together and adding consistency checks during writes. These checks mirror what the channel manager currently validates only on reload, but performing them earlier increases coverage and surfaces inconsistencies sooner.