📝 Summary: Eclair, Core-Lightning, LND, and LDK use aggressive rebroadcasting and local-mempool preimage monitoring to prevent pinning attacks on Lightning Network channels. A vulnerability allows nodes to cheat and claim funds from HTLC transactions. No pinning attacks have been observed, but monitoring mempool logs and fee bumping are suggested as mitigation strategies. Mempool scanning, transaction re-signing/re-broadcasting, and bumping CLTV delta are discussed as potential fixes, but their effectiveness is uncertain. Stratum v2 deployment and a replacement buffer at the mempool level may make the attack harder. Fees are also a challenge.
👥 Authors: • ziggie1984 ( ziggie1984 [ARCHIVE] (npub1taj…8jyw) ) • Matt Corallo ( Matt Corallo [ARCHIVE] (npub1e46…xmcu) ) • Antoine Riard ( Antoine Riard [ARCHIVE] (npub1vjz…x8dd) ) • ZmnSCPxj ( ZmnSCPxj [ARCHIVE] (npub1g5z…ms3l) )
📅 Messages Date Range: 2023-10-16 to 2023-10-17
✉️ Message Count: 8
📚 Total Characters in Messages: 126555
Messages Summaries
✉️ Message by Antoine Riard on 16/10/2023: Cross-posting mempool issues have exposed lightning channels to the risk of loss of funds, potentially affecting other multi-party bitcoin applications. Mitigations have been implemented, but their effectiveness against advanced replacement cycling attacks is still uncertain.
✉️ Message by ziggie1984 on 17/10/2023: Aggressive rebroadcasting and local-mempool preimage monitoring are deployed by Eclair, Core-Lightning, LND, and LDK to mitigate pinning attacks on Lightning Network channels.
✉️ Message by ZmnSCPxj on 17/10/2023: The summary of the text is that there is a vulnerability in the Lightning Network where a node can cheat and claim funds from an HTLC transaction.
✉️ Message by Antoine Riard on 17/10/2023: At block height 100, node B forces the B-C channel onchain because the B-C HTLC timelock has expired without node C claiming it. However, the onchain feerates have risen and the transactions do not confirm. At block height 144, node A drops the A-B channel onchain and is able to recover the HTLC funds. Node C then broadcasts an HTLC-success transaction with high feerates that replaces the HTLC-timeout transaction, allowing C to get the value of the HTLC. Node B is no longer able to use the knowledge of the preimage.
✉️ Message by Antoine Riard on 17/10/2023: No pinning attacks have been observed on the mainnet. Monitoring mempool logs can help detect suspicious behavior. Feebumping could be considered as a mitigation strategy.
✉️ Message by Antoine Riard on 17/10/2023: No experiments have been conducted yet due to limited experts and other pending security issues.
✉️ Message by Matt Corallo on 18/10/2023: The deployed mitigations for the transaction-relay jamming attack on lightning channels are not expected to fix the issue. Mempool scanning and transaction re-signing/re-broadcasting are discussed as possible mitigations, but they have limitations and may not effectively prevent the attack. The only real fix would be for miners to keep a history of transactions and try them again after a potential attack.
✉️ Message by Antoine Riard on 18/10/2023: The disclosed mitigations for lightning attacks include mempool scanning, transaction re-signing/re-broadcasting, and bumping CLTV delta. However, these mitigations may not be effective in fixing the issue. Additional measures such as stratum v2 deployment and a replacement buffer at the mempool level may make the attack harder. The issue of fees is also a challenge.
Follow Lightning Mailing List (npub1j3t…4gll) for full threads