Bitcoin Stack Exchange is a question and answer site for Bitcoin crypto-currency enthusiasts. It only takes a minute to sign up.
Sign up to join this community
What would happen, if I would create a large number of large transactions with a very high nLockTime?
Would something stop me from filling not only the miners', but also regular nodes' mempools?
Would miners drop transactions with unreasonably high nLockTime and low transaction fee?
What would BitcoinCore do?
There is no size limit to the mempool from what I can find. It might be possible to flood it but rather difficult and expensive / slow. I see two approaches:
A) minimum relay fee
It would take a while to generate that many addresses and tx and get them broadcasted. Also it would cost at least 3200BTC.
B) 0 fee txs with minimum output of 0.01 so they still get relayed
Mempool is gated by a rate limiter for free tx:
// Continuously rate-limit free transactions
// This mitigates 'penny-flooding' -- sending thousands of free transactions just to
// be annoying or make others' transactions take longer to confirm.
// Use an exponentially decaying ~10-minute window:
// -limitfreerelay unit is thousand-bytes-per-minute
// At default rate it would take over a month to fill 1GB
So this will be at least very slow. Also you need to have 320 000 BTC for 8GB. :)
There might be ways to artificially make txs larger. Alternate clients might do things differently.
nLockTime will not be relayed until the blockheight one before they can be included is reached or until the timestamp in the nLockTime has been reached. Therefore, a mempool flooding attack with locktimed transactions is not possible.
nLocktime transactions that are not valid yet are not relayed and possibly even dropped. It's the responsibility of the sender and/or receiver to store the transactions until they are valid and then broadcast them.
Transactions with a nLockTime set to a date or blockheight in the future will not be considered valid by Bitcoin Core until the timestamp has passed, or the block height one before they become valid has been reached. Only at this point, full nodes will keep them in their mempools and forward them.
Therefore, a mempool flooding attack with locktimed transactions is not possible.
