LND Fix Addresses Channel Record Error After Node Restart

In a November 17, 2026 report citing CryptoSlate, ChainCatcher said developers had fixed an LND bug that could cause a Lightning node to misread channel records after restarting. It is not confirmed whether the bug was exploited or caused any loss of funds.

The problem was in how LND saved and restored channel updates. A 16-bit counter used to record the number of entries would roll back to zero at 65,536. After a restart, the node could then read those records as an empty list and omit already-settled hashed timelock contracts (HTLCs), compromising the channel’s accounting.

According to the developers, a connected peer could drive the entry count toward that limit by repeatedly sending fee updates and commitment signatures while withholding required revocation messages. LND released v0.21.8-beta and v0.2.8-beta on November 11, adding a safeguard when writing records and checks when loading older ones.