原文：《DVT & Long-Term Ethereum Decentralization》by New Order
Compiled by Kate, Marsbit
As Ethereum works towards decentralization through scalability, it is becoming increasingly clear that a specialized modular blockchain approach, rather than a single blockchain approach, is gaining momentum. In this modular blockchain model, one layer can be divided into three layers: the execution layer, the data layer, and the consensus layer. In the execution layer, after several iterations, optimistic and ZK rollups have been developed (such as Optimism, Arbitrum, and Starkware). These advances have made the execution layer more mature and complete. Therefore, the focus of research has shifted to the next layer, the data layer. An important aspect driving innovation in the data layer is the use of data availability sampling methods, which randomly samples data to ensure the validity of transaction data.
Another key element of a blockchain’s existence is the consensus layer, which includes the consensus mechanism and its scalability. With the launch of the Beacon Chain on September 15, 2022, Ethereum has transitioned from a Proof-of-Work network to a Proof-of-Stake network. Over the past 8 months, the number of validators has increased from 400,000 to 598,000 at the time of writing. The focus now is on making Ethereum’s staking more decentralized, scalable, simpler, more secure, more friendly to independent stakers, and less susceptible to harsh penalties. A promising technology in this regard is Distributed Validator Technology (DVT). In this article, we will analyze DVT and explore how it contributes to Ethereum’s decentralization roadmap.
Modular blockchain ETH staking ecosystem
Distributed Validator Technology (DVT) simplifies the process of running Ethereum validators on multiple machines. Before we explore what this technology means in decentralized Ethereum validators, it is important to understand the current Ethereum staking architecture.
ETH 2.0 Staking
The above diagram depicts the key stakeholders involved in ETH 2.0 staking and their connections. In the center of the diagram, we represent the beacon chain with a cylinder. The beacon chain coordinates the validation process and generates new blocks containing acceptable transactions.
At the center, we have two types of software, commonly referred to as ETH2 clients. Consensus clients and execution clients have different roles in adding new blocks. Consensus clients validate and process transactions on the Ethereum network, ensuring consensus between all participating nodes. Execution clients, on the other hand, execute smart contracts and handle complex operations. While they collaborate with consensus clients for transaction validation and execution, their main focus is on the implementation of the Ethereum Virtual Machine (EVM), which executes smart contract code. Both client types are crucial to maintaining the network.
The circles around the cylinder represent node operators, who are responsible for running the infrastructure required to participate in the network. This includes hardware, software, and network connectivity. Node operators are responsible for the ETH2 client, maintaining the necessary hardware and software, monitoring node performance, and ensuring that the node is operating properly.
The outgroup next to the node operators consists of validators or service providers. As Ethereum holders, we interact with these service providers frequently. They come in many forms, such as centralized exchanges or staking pools, offering different product designs and capital requirements. The main value proposition of these validator service providers is convenience. They handle the complexity of staking, including private key management, node operator coordination, and even provide liquid versions of staked assets for decentralized finance (DeFi). They act as asset managers in the Ethereum staking space, bringing in funds and signing transactions when necessary.
Alternatively, individuals can choose to connect directly to the execution client and consensus client, run the required software and hardware, and become an independent staker - the most independent entity in ETH staking.
It is worth noting that the current architecture has multiple links, is susceptible to single points of failure, and lacks sufficient decentralization. Issues such as lack of client diversity, careless private key storage, or node operator failure can lead to loss of user funds. This is where Distributed Validator Technology (DVT) can add significant value. Let's delve into the details.
How does DVT work? Distributed Validator Technology
Distributed Validator Technology (DVT) allows validators to run on multiple machines, which is achieved by adopting various techniques. One key enabler is Distributed Key Generation (DKG), which is a cryptographic process that enables participants to collaboratively generate private keys without any single member having access to the full key. In DKG, each participant generates a partial shard of the private key, which are combined to derive the full private key. This approach ensures that validators do not need to store the entire private key in a single node, but can distribute the key shards across multiple nodes. Even if one shard or subset of the key is offline and cannot sign transactions, the remaining key shards still maintain integrity and can be used for signing. Publicly Verifiable Secret Sharing (PVSS) and Verifiable Secret Sharing (VSS) are common techniques for ensuring the correctness of key generation, where PVSS is combined with zero-knowledge proofs for public verification.
Once the key is shared and stored on multiple nodes, a coordination mechanism is needed to ensure secure verification and communication between nodes to ensure complete signature generation. This coordination constitutes the essence of DVT and can be achieved through various mechanisms. A common method is the threshold signature scheme. In this scheme, a specified number of validators must cooperate to sign a transaction, preventing any single validator from signing alone and enhancing the security of the verification process. In terms of consensus, the SSV network, as a DVT product, uses the Byzantine Fault Tolerance (BFT) protocol to reach a signature consensus among nodes.
Additionally, DVT must address the issue of shared infrastructure between nodes responsible for storing key shards. Shared infrastructure can lead to correlated node failures, undermining the goal of decentralization. In the SSV network, users have the flexibility to distribute key shards across different node operators based on factors such as geographic location, different data centers, different data center brands, different client implementations, etc. This emphasizes the importance of multi-client implementation as a key component of DVT. Having client diversity is critical to the long-term viability of the Ethereum network because it reduces the risk of a single client with a majority market share causing a fork or other outage that could result in penalties for validators.
Obol network is a company actively developing DVT, which is intended to serve as an infrastructure coordination layer for staking. Their non-custodial middleware, Charon, ensures that the protocol itself does not have custody of keys. Instead, only the validator client holds and manages the private keys that sign transactions. Charon acts as an intermediary between the beacon client and the validator client, intercepting communications and aggregating signatures before forwarding them back to the beacon client. This design approach eliminates the ability of the protocol to sign arbitrary data, giving the Obol network less of a control role and ensuring greater security. By integrating these technologies and approaches, DVT enhances the security, decentralization, and efficiency of the Ethereum validator ecosystem.
Who Benefits from DVT?
DVT brings benefits to almost all stakeholders in the staking ecosystem, making it an extremely attractive technology. As assets under management (AUM) continue to grow, liquidity staking pools prefer to distribute staking to multiple operators rather than relying on a single validator. DVT allows them to share a validator among multiple operators, reducing the vulnerability of assets in the event of a single operator failure.
For independent stakers, DVT provides peace of mind by mitigating the impact of internet or power outages. They can continue to validate transactions even during intermittent connectivity or power outages. Institutional staking products, such as Coinbase Stake or Blockdaemon, can benefit from DVT by reducing operational and hardware costs. By implementing a fault-tolerant solution like DVT, validators can safely increase the number of keys per node, thereby reducing hardware overhead. Additionally, the reduced risk factor associated with DVT may reduce insurance costs for service providers. Overall, the wide range of benefits of DVT makes it an attractive technology for a variety of stakeholders in the staking ecosystem.
A puzzle piece
DVT’s value proposition is shown in the figure below:
DVT’s Value Proposition
DVT brings tremendous value to ETH staking. Its technology reduces the likelihood of node failure, improves key security by distributing keys to multiple node operators, and increases client diversity through product design. This will reduce risk, reduce inactivity penalties, and in turn increase staker confidence. However, it is only part of the puzzle. Only by working together with other stakeholders in Figure 1 can the vision of ETH decentralization, scalability, and security be truly realized.
Combined with the liquidity staking pool, users are not subject to the 32 ETH requirement, thereby lowering the entry threshold for more people to participate in ETH staking.
If the staking pool also requires node operators to post collateral, it reduces the risk of node collusion, as collusion would result in their own stake being slashed.
If we add secure key storage using cold storage, multi-signature or other methods, and node operators support redundant nodes in case of failure, it will further reduce risks and improve the confidence of validators. For example, Puffer is a liquidity staking pool based on a technology called Secure-Signers. The core idea is to keep private keys in TEEs and sign blocks only after checking that slashing will not be triggered, so that operators can guarantee security to stakers. This is an improvement in node operation and can be part of DVT. This Secure-Signer supports permissionless participation through remote attestation, which means that unlike Lido, node operator login does not require DAO voting and other governance mechanisms. It also claims to reduce the staking capital requirement to 1 ETH. It also creates a "smoothing factor" that can be used to evenly distribute MEV rewards to each pool participant. There is not much research on Puffer at present. Whether it can achieve what it claims remains to be further investigated. I also look forward to future technologies that can make Ethereum more secure and reliable, and one day, nothing will keep validators awake at night.
Future challenges
I would also like to discuss the challenges that come with adopting DVT. First, DVT involves the coordination of multiple nodes, which increases the complexity of the system. As mentioned earlier, the SSV network uses a BFT consensus mechanism between nodes. However, the scaling limitation of the BFT protocol is that it is designed to handle a limited number of nodes, usually a few hundred at most. Beyond this number, the communication and consensus overhead becomes very expensive, resulting in degraded performance and reduced security. If you decide to shard your keys into 100 fragments, this makes it challenging to apply BFT consensus to large-scale networks. The current limit of key shards on the SSV network is 13 and on the Obol network is 10. While this limitation does not pose a direct risk to DVT, it can be further improved.
Similar to the trade-off between decentralization and efficiency, DVT may increase latency in the system because transactions must be signed by multiple nodes to be processed. Because the keys are sharded into many blocks, DVT requires more nodes to participate in the staking process overall, which may increase the requirements for node redundancy. This can be solved by more investment from node operators. But it also means that the cost of adopting DVT is too high for node operators.
in conclusion
In summary, the arrival of DVT represents a significant step forward for the staking ecosystem. It provides a secure, flexible, and decentralized infrastructure for staking, making it attractive to individual stakers, operators, and staking pools. DVT has the potential to change the staking landscape and become a key player in the future of Ethereum staking. As the ecosystem grows, DVT has the ability to meet the evolving needs of stakeholders and advance blockchain technology. I am excited to see the advancement of this promising technology and the exciting opportunities it will create for the broader blockchain community.