L2 & rollups: how they work, why we need them & what’s next
--
Author: Roman Aliev (Strategy Marketing Director)
L2 is a collective name for solutions that solve the problems of scalability and speed on blockchain, primarily in the Ethereum network. We’ll tell you about the most popular L2 protocols, their advantages and drawbacks, and what will happen to them now that Ethereum has switched to Proof-of-Stake.
L2 and Vitalik Buterin’s trilemma
Back in 2016, Ethereum creator Vitalik Buterin first described the so-called blockchain trilemma. It boils down to the idea that an ideal blockchain should have three qualities: decentralization, scalability, and security. However, it’s very difficult to achieve all three at the same time, for several reasons.
Decentralization: the more nodes in a network, the harder it is for a malicious actor to take it over. But as you increase the number of nodes, you need more and more resources for them reach consensus, and that slows the network down.
Security: once again, when there are a lot of nodes, the network will continue to run even if some of them break down or get attacked. But scalability suffers: see the point on decentralization.
Scalability: you can create a very fast network if you delegate all transaction validation work to a small number of nodes. But such a blockchain will be centralized and easier to attack.
Ethereum is a highly decentralized and secure chain. But its scalability has long been a sore point: a single popular token sale could completely paralyze the blockchain, as it happened with the Otherside metaverse land sale in spring-2022.
The transition to Ethereum 2.0 should help solve this issue, and Vitalik Buterin even believes that the network will reach 100,000 TPS. However, this will happen only after the introduction of sharding: the division of the blockchain state into a number of shards that will process transactions in parallel and connect to each other through the central Beacon Chain. According to Buterin, it will take 64 shards to reach those coveted…