Transaction pinning is a method for making fee bumping prohibitively expensive by abusing node protections against attacks that can waste bandwidth, CPU, and memory. This can make fee management more difficult in multipart contract protocols (such as LN).
Nodes such as Bitcoin Core that allow transactions to be replaced (RBF) or packaged with higher-fee child transactions (CPFP) place restrictions on those replacements in order to prevent various DoS attacks. However, when two or more people each have the ability to fee bump a transaction, this makes it possible for one of them to pin their version of a transaction at one of the limits and prevent other participants from using fee bumping.
Some of the limits that can be abused to enable transaction pinning include:
BIP125 RBF rule #3 requires a replacement transaction pay a higher absolute fee (not just feerate) than the transaction being replaced and any of its children. This can allow an attacker to attach a large and low-feerate transaction to the transaction they want to pin, forcing any fee bump to pay for the replacement of the large child transaction. E.g., with the 2019 Bitcoin Core defaults, an attacker can require an honest participant pay a minimum of 0.001 BTC to fee bump a transaction (or even greater amounts in some cases).
Maximum package size limitations prevent CPFP from being used if a transaction has more than 101,000 vbytes of children or other descendants in a mempool, or has more than 25 descendants or ancestors. This can allow an attacker to completely block fee bumping by creating the maximum amount of child transactions. If the attacker has to create those transactions for other reasons (e.g. because they operate a service paying to users), this attack can be free. For some two-party contract protocols (such as current LN), this is mitigated by CPFP carve out.
Optech newsletter and website mentions
- Using attacks such as transaction pinning against eltoo
- Discussion of attacks against LN, including transaction pinning