What is a Sequencer

A Sequencer is an important component in Ethereum’s Layer2 Rollup. It receives unordered transactions and generates ordered transaction batches. Then, these transaction batches can be put into blocks and sent to the data availability layer. After receiving a user’s transaction, the sequencer provides an almost instant receipt as a “soft confirmation”, while the “hard confirmation” is received after the transaction is sent to the data availability layer.

Mainstream Layer2 Rollups, such as Arbitrum, Optimism, zkSync, StarkNet, and Polygon zkEVM, all run centralized sequencers operated by operators. However, centralized sequencers have the following problems:

1. Transaction censorship: the sequencer controls transaction sequencing, so it may not include user transactions;
2. MEV extraction: the sequencer can extract the maximum extractable value, which may be disadvantageous to users;
3. Single point of failure: if the sequencer fails, the entire Rollup will be affected.

In response to the problem of sequencer centralization, all major Layer2 Rollups have included the decentralization of sequencers in their roadmaps, but they are all still in the planning stage and have not been implemented in the short to medium term. The reason for this is that the decentralization of sequencers is not a technical issue, but an economic one.

For those interested in the technical principles of Layer2 Rollup, you can refer to my previous article “Ethereum Layer2: The Rollup War”.

Economics of Layer2 Rollup

Layer2 Rollup has three main roles: User, Rollup Operator, and Base Layer. The main process they run is roughly: when users trade on L2, the Rollup Operator acts as the interface between the users and the Base Layer, and eventually publishes the data to the base layer, as shown:

1. Users: They send their transactions on the Layer2 network, deploy their assets on Layer2 on Rollup for contract interactions, and pay fees to the Rollup operator.
2. Rollup Operators: They represent all the infrastructure needed to handle transactions on the Layer2 network. In addition to sequencers, there are also executors and validators. For Optimistic Rollup, there are challengers who report fraud proofs.
3. Base Layer: This can also be understood as a full node. Its purpose is to protect the data protocol of the Rollup, process and verify all transactions, ensure the correct state of Rollup, and ensure the validity of each transaction. If an erroneous transaction is detected, it will be deleted.

The costs for Rollup operators include maintaining a transaction pool, serial batch processing, calculating state roots/state differences/validity proofs, etc. These costs are related to sorting, transaction verification, block generation, and other issues involved in bulk transaction processing. The main sources of income are token rewards and Maximal Extractable Value (MEV).

MEV here is particularly important. MEV refers to the value extracted beyond the standard token reward by manipulating transactions within the block, that is, by including, excluding, and changing the order of transactions. Common forms of extracting MEV include front-running and sandwich attacks.

Therefore, maintaining a centralized sequencer might be the most economically wise choice for Rollup operators. However, there are also Layer2 Rollups that use decentralized sequencers, such as Metis.

Metis promotes its decentralized PoS sequencer as a major selling point. Both the sequencer operators and consensus layer node operators of Metis need to stake a certain number of tokens as collateral. If a sequencer shows dishonest behavior during transaction processing, others can challenge it. If the challenge is successful, the sequencer will lose part of its stakes.

Given that the mainstream Layer2 Rollups are all running a single centralized sequencer, if these sequencers have problems, the entire Rollup program will be adversely affected. The whole point of users using Layer2 Rollup is to save transaction costs, and one of the basic ideas behind blockchain is to prevent reliance on a single centralized provider. The centralization of sequencers is obviously a pressing issue that needs to be resolved.

Shared Sequencer

The solution to the aforementioned problems is a decentralized shared sequencer. “Shared” implies that multiple different Layer2 Rollups can use the same sequencer network, meaning that transactions from multiple Layer2 Rollups are aggregated in a mempool before being sorted.

The shared sequencer aims to alleviate MEV extraction issues, provide censorship resistance, and enhance the efficacy guarantees of Layer2 Rollups. In addition, there are two points worth noting:

1. Modularity: Just like data availability, the sequencer is a modular capability. Regarding data availability, you can refer to my previous article “Data Availability: Who Moved Ethereum’s Cheese?”. The need for a decentralized sequencer can be abstracted into a type of “commodity”. This crafty approach based on modular thinking, due to its low cost and high efficiency, is bound to attract a large number of “one-click chain launch” Layer2 Rollup demands. And all these Layer2 Rollups will benefit from the censorship resistance and real-time capabilities that a decentralized network can provide, without the need to establish such network themselves.
2. Cross Rollup Composability: Since these shared sequencer solutions are intended to handle the transaction sorting of multiple Rollups, they can thus offer unique interoperability guarantees that are currently unavailable. For instance, users should be able to specify that a transaction on Rollup 1 can only be included in a block if and only if a different transaction on Rollup 2 is also included in the same block. By enabling this conditional transaction inclusion, shared sequencers can unlock new possibilities, including atomic cross Rollup arbitrage.

Projects such as Espresso, Astria, SUAVE, and Radius are all focused on decentralized sequencer solutions, each with different implementation paths.

Espresso

Espresso Systems, initially a company focused on privacy solutions, announced in March 2022 that it had raised nearly $30 million in Series A funding, with investments from Electric Capital, Sequoia, and Blockchain Capital. Espresso Systems has largely transitioned to developing the Espresso Sequencer, dedicated to providing decentralized sorting services for Rollups.

Espresso has resolved the problem of sequencer rotation and the determination of “finality” of transactions based on HotShot consensus, and introduced EigenLayer to address the problem of sequencer admission.

The re-staking mechanism of EigenLayer has made it possible for Ethereum stakers to also become Espresso Sequencers, providing security guarantee for the HotShot consensus. In simple terms, Ethereum node stakers can become Espresso Sequencers (ESQ) through the re-staking mechanism of EigenLayer. Ethereum stakers, while earning rewards from PoS nodes, also capture the value of L2 MEV.

As a universal solution for decentralized sequencers, Espresso’s ecological cooperation projects include not only EigenLayer, but also a host of star modular projects such as Arbitrum, OP Stack, Caldera, and AltLayer.

Astria

Astria is positioned as a universal, permissionless decentralized sequencer, providing out-of-the-box shared sequencing services for different Rollups. In terms of financing, Astria announced in April 2023 that it had completed a US$5.5 million seed round of funding led by Maven 11, with participating investment institutions including 1k(x), Delphi Digital, Lemniscap, Robot Ventures, etc. Although the financing is small, the institutional line-up is magnificent.

The operating mechanism of Astria’s decentralized sequencer is similar to that of Espresso Sequencer. The purpose is to weaken the privileges of the sequencer by decentralizing transaction sequencing rights. Specifically:

For sequencer rotation, Astria proposes 2 rotation mechanisms:

1. Simple Leader Rotation: A set is formed through elected sequencers, and the set of sequencers takes turns to sort Rollup transactions. This method prevents a single sequencer from monopolizing the sequencing rights of transactions for a long time, and solves the problem of continuous censorship of users to a certain extent;
2. Byzantine Fault Tolerance (BFT) consensus algorithm: Similar to the leader rotation mechanism, the sequencer on its turn has the right to be responsible for transaction ordering, but 2/3 or more of the members in the sequencer set must reach a consensus on this ordering.

SUAVE

SUAVE is a decentralized, plug-and-play shared sequencer solution built by Flashbots. As a universal solution, SUAVE can provide a memory pool and decentralized block construction for any L1/L2. What sets SUAVE apart from aforementioned shared sequencer designs is that the SUAVE Chain itself is an EVM-compatible chain that achieves transaction sequencing through block “bidding”.

SUAVE’s architecture consists of three core components:

1. Universal Preference Environment: Preferences cover a wide range, from simple transactions to complex events. Users’ preferences are reflected in the mempool in the form of transactions. The preference environment, as a public mempool, brings together preferences. SUAVE’s universal preference environment makes the preferences of multiple chains publicly transparent, eliminates information asymmetry, and to some extent solves the problem of cross-chain MEV.
2. Best Execution Market: The execution market is a network participated by executors, who are responsible for monitoring the SUAVE memepool and competing with each other. Competition drives them to provide the best execution for user preferences. It can be understood that all executors implement user preferences through a “bidding” process, returning as much of the MEV generated by user transactions to the user as possible.
3. Decentralized Block Construction: Finally, based on the collected preferences and the best execution path, the decentralized block construction network includes them in the block. This completes the entire process of transaction discovery, transaction sorting, and block production.

Radius

The goal of Radius is to serve as a trustless shared sequencing layer. Unlike the implementation mechanisms of the aforementioned solutions, Radius ensures that Rollup transactions are sequenced without needing to trust anyone by enabling an encrypted mempool, thus eliminating effective MEV and user transaction censorship.

In terms of funding, Radius announced in June 2023 that it completed a Pre-seed round of financing of 1.7 million USD, led by Hashed, with participation from investment firms including Superscrypt, LambdaClass, and Crypto.com.

Espresso, Astria, and other consensus-based decentralized sequencers reduce MEV and censorship risk to some extent, but this comes at the cost of network scalability and time efficiency, leading to a certain transaction confirmation delay (as consensus on transaction sequencing is needed). Moreover, although transaction sequencing occurs in a decentralized environment, due to the transparency of the related transactions in the mempool, sequencers still have room for malicious MEV extraction. By using an encrypted mempool, Radius aims to nip in the bud sequencers’ malicious MEV extraction and transaction censorship by making related transaction information invisible to sequencers.

Radius uses a zero-knowledge proof-based encryption scheme, “Practical Verifiable Delay Encryption” (PVDE), to create an encrypted mempool. The encrypted mempool ensures the trustlessness of the sequencer, but the risk of single-point failure still exists. If a single sequencer + encrypted mempool is run, a sequencer failure will cause a network outage. To address this issue, Radius proposes several decentralized sequencer implementation solutions, including a secret leader election mechanism and sequencer group sharding mechanism.

Shared Sequencers and MEV

Astria, Espresso, and SUAVE all incorporate MEV as part of their ecosystem, even providing some conveniences to facilitate MEV in a PBS architecture. Radius, on the other hand, avoids MEV by encrypting transactions, but it also acknowledges that MEV is a significant incentive for builders and cannot be completely eliminated.

MEV is a significant market incentive. Mainstream Layer2 Rollups earn a considerable amount by monopolizing MEV through running centralized sequencers. Decentralizing the right to sequence transactions inevitably involves revenue sharing, creating a paradox.

Unless shared sequencers can find a balance between protecting user interests and maintaining ecosystem incentives - in other words, passing the profits to users themselves - this could potentially be a significant traffic entry point to attract users. This could create a demonstrative effect in the Layer2 Rollup niche, thereby forcing other Layer2 Rollups to adopt shared sequencers.

Summary

The modular paradigm of shared sequencers lowers the technical threshold for the entire Layer2 Rollup ecosystem, and cross-Rollup composability opens up more possibilities. However, it’s different from outsourcing the data availability layer. No Layer2 Rollup project would want to outsource their core “tax-collecting” capability.

At its core, Layer2 Rollup lacks an independent decentralized economic framework. Giving up the sequencer makes it difficult for its market ecosystem to take off. Even if they were to adopt a decentralized sequencer, it’s highly likely that they would either be decentralized in name only, keeping control in the hands of operators or a small group of people, or require staking with their own tokens, giving real utility to the tokens to support their price.

Regardless, there will be many new participants in both the Layer2 Rollup field and shared sequencer field. Some mainstream Layer2 Rollups may prefer to create their proprietary solutions rather than use shared sequencing networks, while some well-established shared sequencers may attract many non-mainstream Layer2 Rollups, achieving powerful network effects. There will undoubtedly be new changes in the future.

Disclaimer：

1. This article is reprinted from [PiggyWeb3], All copyrights belong to the original author [web3 ZhuDaDan]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.