L2 Scaling: The Existing Solutions Review

If you’ve been into cryptocurrency for a while, you’ve definitely heard the scaling debate buzzing in the community. Pretty much since the Ethereum launch, there have been talks on possible solutions to increase the speed and the throughput of the network.

The spike in discussions always comes after major congestions. The first one was in 2017 when Crypto Kitties clogged the network to a point never seen before. In 2020, the issue was even bigger, when DeFi and yield farming peaked. The bigger the congestion, the higher are gas fees. So when more users began farming yield, we were seeing gas prices as high as 1400GWEI per tx. Crazy, huh?

Let me reinstate the scaling issues at hand:

• Clogged network
• Extremely high gas fees
• Slower than expected transaction processing that holds back some uses-cases, for instance, blockchain games
• Low transaction throughput

This is where L2 solutions started to come into play.

The solution

Before Ethereum 2.0 introduces its own solutions to the scaling problem — like sharding and Proof Of Stake consensus, the level 2 services took upon the role. If to explain the situation in a few words, consider Ethereum as Layer 1 — the main network. Every solution that is built upon it, and is connected to it is Layer 2.

Since the Ethereum blockchain is currently processing a huge amount of transactions, delegating some of the tasks to the layer 2 services doesn’t seem like a bad idea. L2 can provide a possibility to process transactions off-chain, and thus increasing the capabilities of L1.

Such solutions are built on top of the existing network and don’t require any changes to it. This means a fast scaling solution that doesn’t require the whole network to be rebuilt. Layer 2 solutions are connected to Layer 1 in a manner that allows the service to use the ‘mother’ smart contacts and leverage its security.

Currently, Ethereum processes 15 tx/second; I have seen experts stating that with the help of Layer 2 solutions that number can plummet as high as 2–4K/second. A very drastic increase.

The options

Right now you can find a few types of L2 solutions. Let’s take a look at the main groups, and define their pros & cons.

State channels

Channels allow exchanging of multiple transactions off-chain while broadcasting only a few to the main chain.

You probably heard about payment channels, allowing to make payments off-chain. Well, state channels can be considered a more generalized version of them: they allow for any state updates of the blockchain to be made outside of it, and not only payments.

Since only the final state is getting submitted to the blockchain, the participants only pay fees for 1 transaction and don’t clog the networks with multiple interactions. However efficient the state channels can be, they allow for no open participation, since the channel has to know the addresses beforehand. The addresses participating in the channel have to lock their funds into the multi-sig contract, and the final state only gets broadcasted to layer 1 once all the participants sign the transaction.

Plasma

The solution called Plasma was introduced by the Ethereum founders themselves, as one of the scaling campaigns. It is a framework allowing for the creation of a child-chain connected to the Ethereum blockchain.

Just as channels, it allows for off-loading some transactions by delegating them to the sub-chains, which can further create their own sub-chain, and so on. In such a system, the transactions do not get replicated across the Ethereum blockchain, thus the speed increases while fees decrease.

You can see this solution in use by looking into OMG. They are using a modified version of Plasma called More Viableplasma.

Sidechains

Sidechains are very close to the Plasma solution. The significant difference is that sidechains are using the same EVM that Ethereum uses to ensure interoperability between the networks.

One of the biggest drawbacks of sidechains is the security side. They have less computing power to maintain consensus, which makes it possible for wrong-doers to attack the sidechain. Consequently, it is better to stick to another more reliable solution.

Rollups

Rollups are a solution that essentially allows the process of the transactions off-chain, while the network only stores the data about it.

There are two types of roll-ups: Zero-knowledge and Optimistic rollups.

Zero-knowledge roll-ups allow to make all the processing off-chain: all the funds are held on layer 1, while all the data storage and computing are performed on layer 2. When the processes are finished, the rollup generates a cryptographic proof, known as a SNARK (succinct non-interactive argument of knowledge), or a validity proof, that gets posted on layer 1.

“Validity of the side chains is ensured by ZK proofs, which enables the protocol to compress the aggregated transactions and reduces the amount of information that is written in the main chain.”

Drawback: Zero-knowledge rollups require a lot of computational power and thus are not a viable solution for blockchains that are low on transactions.

Optimistic rollups use a chain parallel to Ethereum. They don’t do any computation. After the transaction is completed, they propose a new state to the mainnet, optimistically assuming that the transaction is legit, if no one states the contrary.

Drawback: If someone notices a fraudulent transaction, the rollup will execute a fraud-proof and run the transaction’s computation, using the available state data. This means you may have longer wait times for transaction confirmation than a ZK-rollup because it could be challenged.

The wrap up:

Zero-knowledge runs computation off-chain and submits a validity proof to the chain.

Optimistic assumes transactions are valid by default and only runs computation, via a fraud-proof, in the event of a challenge.

Conclusion

Overall there are a lot of challenges that L2 solutions are facing, like the composability between apps, or the fact that not every service can switch to using L2. But as of today, we are seeing quite the progress and I can confidently assume that the future of the crypto space is multi-chain. The main advantage of the L2 solutions is room for experiments. Apps and users can explore new opportunities with low commissions. Which will definitely birth new types of solutions pretty soon.