Discuss the status quo of Ethereum and the pros and cons of Layer 2 solutions

Source: Messari

Ecosystem of Ethereum Scaling Solutions

Translator: Babbitt Kyle

The problem with Ethereum today is that the more users there are, the slower Ethereum becomes and the more expensive it becomes to use. The deteriorating user experience has attracted major players such as Binance smart chain BSC, Solana and Polkadot to compete for the smart contract platform position. Ethereum's continued dominance depends on its ability to increase throughput and reduce costs.

Even if Ethereum's power users aren't interested in words like zk-rollups, Sharding or Plasma, the future of the world's most active blockchain depends on them. In this report, we will address the current expansion issues of Ethereum and potential solutions for Ethereum.

The status quo

Ethereum supports a huge amount of economic activity every day. It clears billions of dollars of transactions every day and runs thousands of decentralized applications (dApps) in the booming Decentalized Finance (DEFI) and Non-homogeneous Token (NFT) sectors. But Ethereum has limited block space and all these applications are fighting for the same resources, which means transactions become more expensive and more delayed when the network is congested. In fact, a single smart contract transaction on Ethereum, such as a Uniswap transaction, can cost more than a few hundred dollars, depending on how congesced the network is, making interactive transactions prohibitively expensive for users. But it's great for miners who benefit from such high transaction fees, which can currently increase their earnings by 50 percent from the bull market's high in 2017:

Fortunately, the developer team is working on several extension solutions, each with its own unique optimizations and trade-offs. Strictly speaking, "scalability" is the volume of transactions handled by a single node, and "throughput" is the total volume of transactions the network can handle. For the sake of discussion, we are talking about scalability in terms of all transactions processed.

Ethereum's evolution toward this goal looks like several separate paths, sometimes linked together, leading to the same destination. Each extension solution looks similar to a tree or hub radiation model, but each solution has its own nuances that address a specific use case. There are two main types of solutions:

Layer 2 -- Off-chain scaling (transactions and computing outside the Ethereum base chain). There are four main approaches to implementing solutions for various projects. In fact, each project uses a hybrid approach that combines different types of technologies. Layer 1 -- An extended solution for scaling up the chain, or keeping all transactions on Ethereum.

1. State channel

The status channel allows users to make multiple de-linking transactions, while only submitting two transactions to the Ethereum network -- once on an open channel and once on a closed channel. This relieves the main network of the burden of validating many transactions, but provides the same level of security. Participants in the network need to deposit deposits into a multi-signature contract, which needs to perform a threshold number of signatures (for example, 3 out of 5). Once funds are deposited into the channel, participants can trade as many times as needed. When a participant no longer needs a channel, they can submit a result, wait a while to ensure there are no challenges to the result, and then submit a final transaction on the chain and unlock their funds.

Take a simplified version of the status channel example: a payment channel between two or more parties. This is useful when there are many transactions (such as micropayments) between a known number of participants. Transactions are instantaneous and greatly reduce the cost of processing transactions on the Ethereum blockchain.

The downside is that it takes time to set up the channel and actively monitor it to make sure there are no malicious actors. In addition, funds are locked up during the life of the channel and smart contracts are not yet supported.

Projects related to this include Raiden Network, Celer, Connext, Statechannels and Perun. For example, Raiden has implemented a Bitcoin-like Lightning Network system on Ethereum, including support for the ERC20 token. Celer focuses on state channel technology, but also offers several complementary side chains that are compatible with Ethereum, Polkadot and NEO.

2.Plasma

Plasma is a replica framework for the Ethereum blockchain, called Childchain. Thousands of transactions can be processed in these Layer2 blocks and bundled into a single transaction. Plasma can have many layers, so you can have an infinite grandson and great-grandson chain. Subchains are untrusted and unmanaged chains whose funds the user can control. This means that if an error occurs or is hacked, the user can refer to the latest correct snapshot of the Plasma chain to restore the state and retrieve their token. There is a challenge period where users will pay a transaction fee on the Ethereum mainchain after withdrawing money from the Plasma chain.

For example, the Ethereum blockchain is like the Supreme Court, and the Plasma chain is like a lower court.

The advantage is that Plasma can sustain up to 1,000 transactions per second (TPS) at a low cost per transaction, while the gas cost per Plasma block is fixed. Unlike state channels, Plasma can also handle a flexible number of users, with no need to set the number in the first place. Funds are also highly secure and retrievable.

The drawback is that Plasma does not support the flexibility of smart contract execution, as it only supports basic functions such as a transfer or swap. In addition, while users can withdraw funds, they also need to check the Plasma chain regularly for any errors to prevent them from being exploited by hackers. Similar to the status channel, a "watch tower" is required to maintain network snapshots. For these reasons, Plasma is no longer the preferred solution.

Projects associated with this include Polygon (formerly known as Matic), OMG Network, Gluon, LEAPDAO, and Gazelle. Ploygon is a mixed Plasma and proof of stake side chain, which we'll cover in the next section.

3. The side chain

Plasma subchains and side chains are similar to each other, but differ from a security perspective. The Plasma subchain relies on Ethereum's security mechanisms in a trustless environment and is optimized for payments due to its high throughput and security guarantees. But side chains are separate blockchains that run alongside and communicate with Ethereum. It uses another token to hook up to Ethereum, creating a two-way bridge. A side chain is a completely independent blockchain with its own consensus mechanism and security guarantees.

Just as they underpin Ethereum, these side chains are independent of the blockchain, and can also support other base layers, such as Cardano, by creating hooks using Ada instead of ETH. They interact with the main chain only when the ledger status is updated. Just as side chains can be friendly with other blockchains, users and operators can coexist and maintain multiple side chains under the chain without relying on Ethereum.

The advantage is that the side chain is a blockchain with its own tokens that can support smart contracts (not on the main chain) and is therefore optimized for flexibility and has multiple use cases. Relatively, this technology was built as an extended solution that can deliver speeds of around 10,000 TPS depending on the design of the side chain.

The downside is that it is not a untrusted environment, as the user needs to transfer funds custody to the side chain. Security is also another concern, as side chains are less mature and less decentralized than Ethereum, which has benefited from decentralization over the past few years. In addition, two-way linking means that side chain tokens also need to maintain value and be economically viable, usually by charging fees on that side chain and having effective token economics.

Related projects include XDAI, Polygon, POA networks, Liquid networks, and Skale networks. XDAI Chain is a payment blockchain designed to enable fast, inexpensive and stable transactions. XDAI is used for transactions, payments, and fees. Matic (now Polygon) partnered with Circle to launch USDC Stablecoin and teamed up with ChainLink to provide support for Ethereum games. Matic changed its name to Polygon and focused on starting the Layer2 aggregation framework. The team will continue to host Matic Network's current solution, the POS Ethereum side chain using the Plasma framework. But they now consider this existing side chain an "undesirable Layer2 solution". Polygon's upcoming SDK will reportedly enable developers to create multiple Layer2 solutions, such as Optimistic Rollups, ZK Rollups, and Validium, which is why the team calls it a Layer2 aggregator.

4.Rollup

Rollup allows thousands of transactions to be bundled into a single Rollup block. It may provide a hundredfold throughput because the data digest published for transmission outside Layer1 has less storage and computation burden and is cheaper than Layer1. Abstract Data is still protected on Layer1 (Ethereum) without the need for complete computation and storage on the Ethereum chain.

Unlike the state channel, the money in the Rollup is held by a smart contract in which the operator invests the money in the Layer1 smart contract. All transactions take place on layer2, and if the user thinks layer2 operations are malicious, they can be performed on Layer1. Bad actors will have their pledges reduced, and whistleblowers will get a portion of the reduced pledges as a reward.

Rollup is useful for reducing costs, providing faster transaction throughput, and opening up user participation. There are two main types: Optimistic Rollup and ZK Rollup.

4a Optimistic Rollup

Optimistic Rollup uses a side chain that runs parallel to the Ethereum main chain. After completing a batch of transactions, Rollup presents a new state to the mainnet. For example, they brokered and notarized the deal. They can handle about 300 smart contract calls per second or about 2,000 base transfers per second. Since it is compatible with the Ethereum Virtual Machine (EVM), anything done on Ethereum can also be done on Optimistic Rollup. This is a solution for extending common smart contracts and is an easy way to migrate decentralized applications (dApps) with a reasonable degree of security.

But the compromise is that funding could be threatened by a potential agent of attack. If an incorrect state transition is published, the user will be able to undo the incorrect block and significantly reduce the amount of money pledged by the bad actor. Withdrawals are also slow and can take several days to allow for a challenge or dispute period.

Projects focused on this include Optimism, Arbitrum, Fuel Network and Cartesi. Uniswap V3 is starting on Optimism, which currently validates this solution as the preferred Layer2 solution. Optimism's mainnet had a soft launch in January 2021 using whitelist protocols such as Synthetix, although it is now ready to be fully operational some time after July 2021. Users of these whitelist agreements have saved more than $10 million in fees. However, it is still not possible to determine the actual user adoption because it requires changing user behavior, such as having to rely on a new wallet that supports this new side-chain enhancement.

4b ZK Rollup

While Optimism assumes that the transaction is valid by default and only runs the computation when challenged, ZK Rollups runs all the computation out of the chain and submits the proof of validity to be stored on Ethereum. Zero Knowledge (ZK) refers to the cryptographic proof that issuers must provide in order to record a series of transactions on the Ethereum blockchain.

This proof is sometimes called a SNARK (Concise Non-Interactive Knowledge Argument). Operators need to generate proof for each state transition, which is validated by the Rollup contract on Ethereum. The SNARK proves that a sequence of transactions from one state to another exists. This is similar to showing the results of calculations to the Ethereum blockchain without showing them all the transaction data or how the results were arrived at. Validating blocks is therefore faster and cheaper because less data is involved.

ZK Snarks (interchangeable with ZK Proofs) are often used by privacy coins such as Z Cash. But ZK Rollups uses zero-knowledge techniques to improve efficiency, rather than privacy, by taking thousands of transactions off the chain and submitting them to Ethereum as a single Rollup transaction. ZK Rollup will reportedly be able to process approximately 3,000 TPS on ETH1 and 20,000 TPS on ETH2 at 1% of Layer1's cost. By comparison, credit cards typically handle 5,000 TPS, although they are said to be able to handle much higher throughput (65,000 TPS).

The advantage is that there is no delay in transferring funds between Layer1 and Layer2 because the funds have already been validated by the proof of validity accepted by the contract. Therefore, if the user decides to withdraw his money from Layer2 to Layer1, settlement is faster than Optimistic Rollups. However, because the ZK proof needs to be generated, a Layer 2 to Layer 2 transaction takes about 10 to 30 minutes. Another drawback is that smart contract support is still in development, so the ZK Rollup can only do simple transaction transfers. It is also currently impossible to encapsulate EVM in zero-knowledge proof, so DApps cannot yet be migrated to ZK Rollup.

Projects on ZK Rollup include Loopring, ZKSync (Matter Labs), Aztec and Hermez. Loopring applied the technology to building order decentralised trading centres by moving data and calculations down the chain while using zero-knowledge proof to maintain the property of minimal trust. On Loopring 3.0, executing transactions costs less than a penny of Ethereum gas. Aztec recently deployed ZK. Money, a proprietary Rollup service for ETH and future ERC-20 tokens. Users can submit the ETH to an Aztec Rollup contract to block and send private transactions.

Take a step back. Just as ZK and Optimistic Rollups are similar in that the data is on a chain, Plasma and Optimistic Rollups are similar in that they are based on fraud-proof systems. There are four quadrants, representing different types of systems. The second type of validation system is Validium.

5. Validium

Validium uses proof of validity (e.g. ZK Rollup), but the data is not stored in the Ethereum chain. Because each transfer must be authorized, the money in Validium is secure. Keeping data outside the main chain, each Validium chain can achieve up to 20,000 TPS, which can run in parallel with each other.

The benefits are that there is no withdrawal delay and that it is not vulnerable to the kind of attacks that fraud-based systems face. The tradeoff is that it has limited support for smart contracts and requires a lot of computing power and 10 to 30 minutes of time to generate ZK proofs, which is neither time saving nor cost effective for low-throughput applications. As a result, it may have a chicken-and-egg problem when it comes to getting started.

StarkWare and Matter Labs use Validium. Deversifi was the first platform to use StarkWare batch technology to trade and transfer tokens without having to pay gas fees, rollback hazards or transaction failures. Matter Labs developed ZKPorter.

Capacity expansion solution: hybrid solution

Despite the neatness of the scaling solutions described above, the reality is that technologies can be combined, each with its own tradeoffs. For example, Matic turned to Polygon and started the Layer2 aggregator. In addition, Celer combines a state channel with a side chain solution.

Disadvantages of the Layer2 extension solution

The tradeoff of these heterogeneous scaling solutions is that there is no single global state that supports composable smart contracts. Currently, most users rely on a simple single system, Layer1. The Layer2 extension requires significant changes in user behavior, wallets, predictors, and DApps. Developers building smart contracts may not want to deal with Layer2, or state across Layer2, or security models in side chains, or flow routes in a network of state channels. They may also not want to use ZK proof to deal with how to run a computation down the chain.

There is also a lack of communication between the various Layer2 projects. This means that transfers across Layer2 are not seamless, or side chains need to be bridged. Applications may also need to consider exogenous states, which are still under investigation.

These difficulties, as well as competition from other smart contract platforms, forced Ethereum to prioritize the development of Layer2 solutions, especially Rollup. The ETH2 developers have changed the roadmap to illustrate the rise of Rollup, and Ethereum plans to make ETH2 sharding and Rollup work together.

Extending ETH 1.x to ETH 2.0

ETH2 is the first large-scale blockchain to be fully operational and rebuilt under the new consensus mechanism. This is Ethereum's most ambitious system-wide upgrade, such as keeping going while trying to build a new ship. A key aspect of ETH2 is Ethereum's transition from a proof-of-work (PoW) to a proof-of-stake (POS) mechanism, which helps process transactions faster.

* PoW: In this consensus approach, all transactions must be hash confirmed by the miner using brute force calculation over the block. This keeps the network safe, but limits the speed at which miners can quickly propagate blocks. Each new Ethereum block must be sequential and cannot be processed in parallel.

* POS: In the POS consensus model, ETH2 enables any number of validators (at least 32 ETH) to run the validator without expensive mining equipment or maintenance requirements. The upgrade has had no effect on miners' trading fees, which have recently accounted for about 50% of their revenues (see chart on miners' earnings). PoS decentralizes Ethereum by making it easier for participants to access. Block times will be more predictable because, as with the case on the PoW chain, there will be no fluctuations associated with the process. This is despite lower end-user fees due to sharding (or splitting the blockchain). At the time of this writing, 3.6 million Ethereum has been deposited into ETH2, worth $7 billion, or 3% of the total supply.

Sharding

Sharding simply means that the network can be divided into multiple tracks to process transactions in parallel. This is horizontal scaling, similar to distributing computing and storage capacity across multiple servers. Each shard has its own separate status and transaction record. Dedicated nodes process some sharded transactions, resulting in higher overall throughput.

The beacon chain links POS and sharding together. It is the coordinator of all system-level activities, stores and manages the validators' registry, selects block producers, applies consensus rules, and stores data in a shard state. The beacon chain was put into operation in December 2020, but there is currently no activity other than its existence. This is the first step in allowing Ethereum to scale without sacrificing security.

Unlike Plasma, sharding is the antenna of the main chain and periodically commits the state root hash of the transaction from its shard. The application will also access the data in the shard directly. In contrast to subchains and side chains, sharding ensures that the entire system is an aggregate -- with the same availability and access to data. Sharding also doesn't have any deposits or funds because it's part of the main chain.

Not everyone has to run every shard, which means that not everyone needs to run the full ETH2 node. Also known as a "light client" or lightweight node, this refers to a trusted blockchain copy of the full node, but does not require downloading the entire blockchain copy. Light clients can play an important role in Ethereum sharding to enable validators to quickly validate and synchronize different shards.

The initial idea was that ETH1 was just one of many shards. However, this adds to the complexity of cross-shard transactions. As Ethereum's evolution continues to respond to market conditions and new technologies, the technology is still evolving and may or may not be a central feature of the migration to ETH2. The final phase of the migration to ETH2 is designed to unlock smart contract execution in the shard in an estimated time of 2023.

The evolution of Ethereum

In December 2020, we witnessed the first step towards ETH2 with the activation of the beacon chain. Then there is sharding, then the abstract execution engine, and finally you see ETH1 and ETH2 merged. However, the evolution of ETH is dynamic, and given the urgent need to address the issues of high gas costs, expansion and intense competition, consolidation was given priority.

The merger will be preceded by two other technology upgrades: "Berlin" (April 2021) and "London" (July 2021) hard forks. The Ethereum Improvement Proposal (EIP) doesn't usually get much attention, but EIP 1559 will be included on the London hard fork along with the other EIPs.

EIP 1559 led to increased tension between miners and developers. ETH mining has been very profitable recently, reaching a record $1.3 billion in revenue in February 2021. Understandably, the mining industry is opposed to the change because mining is an enterprise-level, capital-intensive business and therefore EIP 1559 will have a negative impact on revenues. But the network will evolve according to the needs of its users, and miners are not particularly important in the equation. If miners threaten to use hard forks to oppose it, then their new forks will be of low value and won't be adopted by users. This is especially true for DEFI for wrapped assets and NFTs that cannot be replicated under the new network.

EIP1559 will help alleviate cost pain by making costs more predictable. According to EIP1559, a block requires a minimum price (the "Basic Fee"), which will be set dynamically based on demand. While users can cap their fees at twice the base fee and tip, this eliminates any guesswork when estimating the transaction costs that users are incurring today. The main difference with EIP1559 is that the difference between the user's maximum fee and the final base fee will be refunded. Currently, users who grossly overestimate their fees will not receive a refund. According to EIP1559, if the block is full, the fee market will re-auction based on the highest tip and will not be refundable.

These upgrades do not include the updates required for the combination of eth1.x and ETH2, but the developers are considering speeding up the schedule. A quick merge recommendation is outlined here, with only a few changes needed to migrate from PoW to PoS. However, it's debatable whether Ethereum should consider another fork upgrade (the "Shanghai" upgrade is about 3-6 months after the "London" upgrade) or look at a merger. Often, it can take months or even years for the governance process to be included. EIP 1559 itself was championed by Vitalik Buterin in April 2019, but was implemented only two years later.

conclusion

Ethereum is the most valuable network under system-level reconstruction. Understandably, the developers are implementing the changes cautiously. While Ethereum is working on ETH2, the project offers multiple technologies simultaneously to provide the best possible scaling solution. The order in which various extension solutions are delivered is irrelevant. Defi's prosperity proves that composability is the key to success and enjoys network effects. Equally, expanding Ethereum is not a winner-takes-all contest. The scalability benefits from Layer1 and Layer2 improvements will add up, especially if the project can work together for a composable extension mechanism. Victory will be a smooth transition to POS and Ethereum, maintaining its primacy as the de facto smart contract platform for decentralized applications.

The original link

Link to the translation

欢迎加入律动 BlockBeats 官方社群：

Telegram 订阅群：https://t.me/theblockbeats

Telegram 交流群：https://t.me/BlockBeats_App

Twitter 官方账号：https://twitter.com/BlockBeatsAsia