Solana’s Spam Problems Persist Despite Tech Improvements, MEV Researchers Say - Coinleaks
Current Date:September 21, 2024

Solana’s Spam Problems Persist Despite Tech Improvements, MEV Researchers Say

Despite efforts by Solana developers to discourage spammy transactions that could threaten to bog down the network, a majority of the network’s compute is still being wasted on failed trades, according to an analysis by crypto infrastructure company Jito Labs.

In one recent epoch (a time period on Solana roughly equivalent to two and a half days), arbitrage transactions took up 60% of overall compute space, per Jito. These transactions were attempts by bots to win slim margins on competitive trades – and 98% of their attempts failed.

The result is wasted blockspace for the network as well as capital burnt for no reason on losing trades, according to a blog post from Jito Foundation. It blames this on the way Solana’s infrastructure handles submitted transactions: give priority to the first in line. That creates an incentive for arbitrage bots to submit multiple duplicate transactions in the hopes they will get the winner.

Recent changes to Solana’s backend – particularly the introduction of priority fees and local fee markets – have changed the economics around spamming; and yet spam transactions will persist as long as MEV (maximal extractable value) opportunities remain, according to analysis by validator service Chorus One.

MEV refers to the methods through which traders can squeeze as much crypto as they can out of a given blockchain network, typically by previewing upcoming transactions and using this advanced information to squeeze in profitable trades.

Two popular strategies include winning arbitrage trades and fulfilling liquidation orders. On Ethereum, where MEV is a cottage industry, the network’s slower architecture and emphasis on fees combats spam transactions much more than Solana’s,

Jito Foundation is building a specialty client for the Solana network that optimizes for MEV, said the pseudonymous Buffalu, CEO of Jito Labs, which contributes to the work.