Anti DoS for tx replacement

Summary:

In a thread discussing the shutdown of services dependent on replacement, Mike Hearn explains that good users of replacement would still have priority over constantly churning DoS replacements. He adds that if there is no form of tx replacement at all, then the "tried and doesn't work" state is the same as "never tried", which is not a win. However, he warns that an attack is more subtle than John Dillon realizes. Assuming that nodes prioritize transactions with the fewest total replacements first, it becomes a multiplier on the standard attack of just broadcasting transactions. While it is not bad for non-replacement users, an attack still shuts down useful tx replacement. For instance, in the adjusting payments example, an attacker sets up a legitimate adjusting payment channel, does a bunch of adjustments, then broadcasts enough adjustments that their adjustment session looks like part of an attack, and doesn't have to pay for the full adjusted amount.Regarding proposals for alternatives, Hearn notes that prioritization of work is still necessary. He also suggests that one way to do replacement anti-DoS would be to only allow each input a given number of chances to do a replacement. In the current transaction format, it is difficult to determine which input is responsible for a replacement. Hearn also thinks that $500 is rather low for the amount of work required to implement the proposal and suggests adding a zero onto that to attract more takers.

Updated on: 2023-06-06T15:09:08.749055+00:00