Published on: 2018-11-12T00:17:31+00:00
The email thread discusses a potential solution to address the issue of freeing up scarce resources on the network. The proposed solution focuses on making probing routes more efficient and canceling stuck probes in both the sending and receiving directions. Currently, when a probe discovers a failing node or channel, the probe HTLC gets stuck, consuming HTLC slots and memory for well-behaved nodes. This not only affects the network's resources but also poses a risk of locking up real money.To mitigate this, the proposed solution suggests allowing for canceling stuck probes from both the sending and receiving directions. The process involves generating a 128-bit random number X and calculating H by taking the bitwise-not of the SHA256 hash of X. A probe payment is then made over path P, using the hash H and amount V, and responses are awaited. These responses can include success, error, unknown hash, routing failure, or timeout.By implementing this solution, it provides an option for individuals making honest efforts to probe the network without inadvertently attacking it in case a probe fails. However, it should be noted that attackers still have the choice to not follow this approach.Another aspect discussed in the email thread is the need for reliable payment routing. To achieve this, nodes must actively probe routes regularly. However, there is a risk that if a probe discovers a failing node or channel, the probe HTLC will get stuck, consuming resources for well-behaved nodes. To address this, the suggested solution proposes allowing for canceling stuck probes from both the sending and receiving directions. In this scenario, a payment hash is generated by performing a logical XOR operation between SHA256("something") and 0xFF..FF. This allows everyone to safely drop the incoming transaction as they know it will be refunded eventually.To cancel a stuck probe, one simply needs to send a packet indicating that it was a probe and providing the verification of the "something" value. This action frees up the slot and funds from the probe, benefiting everyone except for the actual failing nodes.From the perspective of the sending node, the process involves generating a 128-bit random number X, calculating H by taking the bitwise-not of the SHA256 hash of X, making a probe payment over path P using hash H and amount V, and awaiting responses such as success, error, unknown hash, routing failure, or timeout.
Updated on: 2023-07-31T20:40:42.045403+00:00