Partnered With:

Optimistic Rollup: Everything You Need to Know

Read this article in 16 Minutes

Discover one of the most important technologies for Ethereum in 2021: Optimistic Rollup.

The original title: "dry | about Optimistic Rollup, all you need to know ()"

Original source: Ethereum enthusiasts

Incentive mechanism of Optimistic Rollup

The Layer 2 extensibility scheme is based on the fact that we are trying to minimize the number of transactions executed on the chain. We use proof of fraud to cancel an invalid state transition that has occurred. Since proof of fraud is an on-chain transaction, we want to reduce the number of proof of fraud published on Ethereum as much as possible. In an ideal world, fraud would never occur and no fraud certificate would be issued.

We counter - incentivize fraud by introducing a Fidelity Bond. Users who want to become collators must first deposit a deposit on Ethereum. If they are proved guilty of fraud, they lose the deposit. To incentivize users to actively detect fraud, the collator's margin is rewarded to the verifier.

Good faith deposit and dispute period

In the design of fraud proof incentive mechanism, there are two parameters to be designed:

Honesty Deposit Amount: The deposit paid by the ranking person is the reward that the verifier can get. The higher the amount, the greater the incentive for the examiner, and the less the incentive for the sorter to do evil. Duration of dispute period: The window in which proof of fraud can be issued. Once this time window has passed, L2 transactions are finalized on L1. If the dispute period is longer, security is improved and censorship attacks are better protected. If the dispute period is short, the user will have a better user experience when withdrawing money from L2 to L1, because he won't have to wait long to spend the money on L1.

In my opinion, there is no correct static value for either parameter. Maybe 10 ETH deposit and 1 day dispute period is enough. The real answer is that it depends on the incentive to be a validator (the cost of running the validator software) and the difficulty of issuing a proof of fraud (the congestion on L1). These can be adjusted manually or automatically. For example, EIP 1559 introduces on Ethereum; BASEFEE Operation code. This opcode can be used to predict congestion on the chain, making the length of the dispute period programmable.

It is important that this penalty mechanism is applied correctly, or it may be used in practice. For example, here's a non-viable native implementation:

Alice pays a deposit of 1 ETH to become the sorter and Alice posts a bogus status update which Bob finds out about and disputes. If successful, Alice's 1 ETH deposit is awarded to Bob and the false status update is cancelled. Alice finds out about the controversy and raises it (challenging herself!). Alice gets back her 1 ETH and successfully escapes the punishment that would have been imposed for her evil deeds.

Alice can initiate this attack by "jumping the gun", that is, broadcasting a transaction like Bob's, but paying a higher gas price to get her own transaction executed before Bob's. This means that Alice can always do evil at a fraction of the cost.

The solution is simple: Instead of awarding the wrongdoer's entire deposit to the disputer, destroy X percent of it. In the example above, if we destroy 50% of the deposit, Alice can only get back 0.5 ETH by jumping the queue, which is enough to deter Alice from doing evil in step 2 above. Of course, margin destruction also reduces the incentive to run a verifier's software (since the reward for a successful dispute is reduced), so the amount left over should be enough to motivate the verifier.

Responses to criticisms of Optimistic Rollup

We have seen the design of Optimistic Rollup. Now, let's listen to some of the criticisms about Optimistic Rollup and respond.

Long withdrawal/dispute periods are detrimental to adoption and composability

As we mentioned above, a longer dispute period helps improve security. There seems to be an inherent trade-off here: a long dispute period is bad for OR adoption, as any user who wants to withdraw money from OR will have to wait a long time (e.g., 7 days). A shorter dispute period leads to a better user experience, but it increases the risk of failing to dispute fraud in a timely manner.

We don't see that as a problem. Due to the long withdrawal delay, we expect that market makers will provide rapid withdrawal services. This is possible because the person verifying L2 status can correctly judge whether the withdrawal is fraudulent and buy the withdrawal at a slightly lower price. Such as:

Participants:

Alice: 5 ETH on L2:4.95 ETH in the "market maker" smart contract on L1, the verifier for L2

Steps:

Alice tells Bob that she wants to make a quick withdrawal and pay him 0.05 ETH. Alice initiates a withdrawal transaction to Bob's "market maker" smart contract. Two things can happen:

(1) Bob verifies that the withdrawal is valid on L2 and agrees to Alice's request for quick withdrawal. The 4.95 ETH in the market maker contract is immediately sent to Alice's address at L1. When the dispute period is over, Bob will get 5 ETH and make a handsome profit.

(2) When verifying, Bob finds that the withdrawal is invalid. Bob disputes the state transition of the transaction, cancels the state transition, and is rewarded with a deposit from the sorter who tried to do evil.

If Alice is honest, she can complete the withdrawal immediately; If she tried to do evil, she would be punished. We expect that if there is a real demand for fast transaction services, the fees paid to market makers will gradually fall to the point where users will not feel the process at all.

The most important impact of the rapid transaction service is the ability to achieve composability with the L1 contract without waiting for the entire dispute period to end.

Note: This technique first appeared in Simple, Quick Things.

Miners take bribes to review withdrawals and destroy the security of OR

A near-zero-cost attack scenario on Optimistic Rollup points out that sequesters can easily bribe ethereum miners to review controversial issues. This can be fatal for Optimistic Rollup systems because the security of the entire system is derived from the dispute mechanism.

We do not share this view. We think the honest side will pay as much or more to bribe the miners as the bad side. Moreover, miners incur additional costs each time they help. This will affect the value of Ethereum, and therefore the miners' own interests.

In fact, there is academic literature that has studied this scenario, and it turns out that "the threat of such a counterattack produces a seed-game refined Nash equilibrium, so that the attack never happens in the first place."

Thanks to Hasu for recommending this article.

The dilemmas of the verifier can bring about adverse incentives and reduce the security of the OR

Ed Felten has analyzed the verifier's dilemma and proposed a solution in his great essay. We summarize it as follows:

If the system's incentives work as expected, no one's going to do anything bad if no one's going to do anything bad, there's no point in running the validator software, because there's no payoff if no one's going to run the validator software, then the sequencer is going to have the opportunity to do something bad and if the sequencer is going to do something bad, then the system is not going to work as expected.

This question seems so important, and seems like a paradox! Assuming that the total reward amount is fixed, the more verifiers, the lower the expected revenue per verifier. In addition, if the number of verifiers increases, the total reward amount is likely to decrease because there will be less fraud, resulting in lower revenue for the verifier. In the following analysis, Felten presents a solution to the verifier's dilemma.

I would argue that the validator dilemma is not as bad as critics make it out to be. In fact, verifiers do not rely on financial incentives. Let's say you're building a large app on Rollup, or you're a coin-holder. If this system is compromised, your app won't work and your tokens will lose value. Moreover, the need for quick withdrawals leads to market maker services (as we mentioned above), regardless of whether there is fraud or not. As a practical example, Bitcoin doesn't provide financial incentives for node operators to store a complete blockchain transaction history or provide local data to their peers, but people do it selflessly anyway.

Even if not providing financial incentives to run verifiers is not incentive compatible, it provides security for the system, which is important to the entity investing in the system. Therefore, we believe that Optimistic Layer2 system does not need to solve the verifier's dilemma problem through design mechanism.

conclusion

As the title of the article suggests, we analyze one of the most important technologies for Ethereum in 2021: Optimistic Rollup.

The benefits of OR include: it is an extension to Ethereum, integrating Ethereum's security and composability, and developer benefits. At the same time, it would improve Ethereum's performance with little additional cost or trust requirements for Ethereum users. We explore the incentives that make OR viable and offer a rebuttal to common criticisms.

We want to emphasize that the upper performance limit for OR is the amount of data that can be carried on L1. Therefore, it is best to 1) compress the data you publish on L1 as much as possible (e.g., aggregated by BLS signature) and 2) have a large and low-cost data layer (e.g., ETH 2.0).

For additional reading, we recommend Vitalik's Incomplete Guide to Rollup (The Chinese version) and Patterns of Trust (The Chinese version). We also recommend that you look into another rollup solution, the ZK rollup. Our friend StarkWare is building the ZK Rollup solution. Finally, there are other ways to achieve decentralized scalability, such as sharding and state channels. They all have their advantages and disadvantages.

In the next article, we'll dive into the mechanics and code base of the first EVM-compatible OR scheme built by Optimism.

We would like to thank Hasu, Patrick McCorry, Liam Horne, Ben Jones, Kobi Gurkan, and Dave White for their valuable feedback on this article.