How EigenDA achieves ultra-large-scale data availability for Rollup

Written by: EigenLabs

Compiled by: TechFlow

EigenDA is a secure, high-throughput, and decentralized data availability (DA) service built on top of Ethereum using EigenLayer's re-staking infrastructure. Developed by EigenLabs, EigenDA will be the first Active Verification Service (AVS) to launch on EigenLayer. Once launched, re-stakers will be able to delegate their stake to node operators who perform verification tasks for EigenDA in exchange for service payments, and Rollups will be able to publish data to EigenDA in order to obtain lower transaction costs, higher transaction throughput, and secure composability in the EigenLayer ecosystem, with security and throughput designed to scale horizontally with the amount of re-staking and operators who choose to provide services to the protocol.

We hope that EigenDA will make the following contributions to the Ethereum ecosystem:

Providing innovative DA solutions for Rollup, contributing to the ultimate goal of Ethereum expansion, and gaining security and value from Ethereum stakers and validators. EigenDA is built on some core ideas and libraries that are key upgrades to Danksharding and can play a role in combat testing these technologies.

Providing high throughput and low-cost standards to promote the growth of new on-chain application cases. EigenDA will support applications such as multiplayer games, social networks, and video streaming, with a flexible cost model including variable fees and fixed fees.

Protecting the key element of decentralization. In a shared security system like EigenLayer, if every node operator needs to download and store every chain that uses the system, few node operators will be able to keep up and the system may eventually become centralized. EigenDA is designed to prevent this tendency towards centralization; it distributes work among many participating nodes, achieving high performance and requiring each operator to complete only a small amount of work.

Prove the power of programmable trust. EigenDA seeks to prove that Ethereum stakers and validators can support critical Ethereum infrastructure, and that AVS (such as EigenDA) and AVS users (such as Rollup using EigenDA) can successfully implement new business and token models on the modular foundation of the Ethereum trust network, in addition to Ethereum consensus.

We are excited to see that several teams are already planning to integrate EigenDA into their L2 infrastructure, including: Celo in its transition from L1 to Ethereum L2; Mantle and its range of complementary products in the BitDAO ecosystem; Fluent providing the zkWASM execution layer; Offshore providing the Move execution layer; Layer N providing zk-OP hybrid Rollup for financial applications, and many more.

Technology Architecture

The following diagram shows the basic flow of data in EigenDA.

The Rollup Sequencer creates blocks with transactions and sends requests to disperse the blocks.

Disperser is responsible for erasure coding data blocks into blocks, generating KZG commitments and KZG multi-reveal proofs, and sending commitments, blocks and proofs to the operating nodes of the EigenDA network.

Rollup can run its own Disperser or use a decentralized service provided by a third party (such as EigenLabs) to facilitate and share the cost of signature verification. Rollup uses decentralized services to use its own Disperser as a backup when the service is unresponsive or censored, thereby gaining the benefits of sharing without sacrificing censorship resistance.

The EigenDA node verifies the received blocks against the KZG commitment using multi-reveal proofs, persists the data, and then generates and returns signatures to the Disperser for aggregation.

Technical considerations

Now that we have a basic understanding of EigenDA’s architecture, let’s discuss the benefits and features that the system aims to achieve. Here is a brief list of some of the features we believe are necessary for a good and useful data availability layer for Rollups:

• Economical

• Throughput

• safety

• Customizability.

We will explain each feature from the perspective of EigenDA.

Economical

Today, many L2s use Ethereum as their data availability layer because of its cryptoeconomic security guarantees. This results in extremely high costs and volatility as Rollups compete with all other Ethereum users for limited block space based on congestion pricing. For example, Arbitrum and Optimism have spent tens of millions of dollars on data availability fees on Ethereum so far this year, with no consistency from month to month. One of the main value propositions of a data availability system is to significantly reduce these costs and provide Rollups with greater predictability in their cost structure.

cut costs

There are three fundamental dimensions of cost associated with operating a data availability system. Let’s analyze how EigenDA minimizes the underlying cost structure in each dimension:

Cost of staking capital. To protect the data availability layer, stakers may want to receive a certain percentage of the returns to offset their opportunity cost. EigenDA reduces the cost of staking capital by using EigenLayer, which adopts a shared security model that allows the same stake to be shared among various applications, creating economies of scale.

Operational costs. Instead of requiring each node to download and store all data, EigenDA uses erasure coding to split the data into smaller blocks and requires operators to download and store only a single block, which is a fraction of the full data block size. This reduces the cost per operator compared to storing full data blocks, allowing many nodes to operate EigenDA "lightweight". As more nodes join the EigenDA network, the resource costs incurred by each node on the network will also decrease. This allows EigenDA to provide security at a low cost by a large number of operators and achieve a gradual reduction in costs, thereby realizing the concept of abundance rather than scarcity.

Congestion costs. When the bandwidth utilization of any blockchain approaches system capacity, the cost of data begins to rise. EigenDA reduces congestion in two ways: 1. Through higher throughput, it attempts to make congestion a rare phenomenon; 2. By allowing bandwidth reservations, EigenDA can guarantee the throughput reserved by Rollup at a discounted price. To maintain flexibility, EigenDA also allows Rollup to pay for throughput on demand.

Rollup Economics

The economics of Rollup are fundamentally different from L1 because data availability costs are not only high and unpredictable, but they are paid in a non-native token. This makes it difficult for Rollup to make price promises to users and subsidize initial adoption because they must bear the "exchange rate risk" between their own Rollup tokens and the tokens that pay for data availability. In contrast, L1 pays a fixed amount of inflation and can offer a certain number of transactions per second for free to attract users.

EigenDA is exploring enabling Rollups to pay stakers in native Rollup tokens at a predictable, long-term reserve rate, on terms approved by EigenLayer stakers. This combines the inherent scale benefits of a shared security system with the inherent benefits of stable native token payments to help drive Rollup usage.

Throughput

Throughput is another fundamental value proposition of a data availability system. EigenDA is designed to scale horizontally, i.e. the more operators there are on the network, the greater the throughput of the network. In private testing using 100 nodes with standard performance characteristics, EigenDA has demonstrated throughput of up to 10 MBps, with plans to scale to 1 GBps. This opens the door to bandwidth-intensive applications based on Ethereum, such as multiplayer gaming and video streaming.

EigenDA achieves high throughput through three pillars in its design:

DA is decoupled from consensus. Existing DA systems combine the proof of data blob availability and the ordering of data blobs into a "monolithic" architecture. Proofing the availability of data can occur in parallel, as nodes can independently prove the availability of different data blocks; however, ordering requires serialization of data blocks, resulting in significant consensus lag. While this coupling may be beneficial in terms of security for systems designed to be the source of final ordering, it is neither necessary nor appropriate for DA systems designed to be used with the Ethereum blockchain. Useful because the Ethereum blockchain itself has an ordering system for settlement. By simplifying the unnecessary complexity of sorting and designing a pure DA system, EigenDA achieves significant improvements in throughput and latency.

Erasure coding. EigenDA enables rollups to break data to be published to EigenDA into smaller chunks and erasure code these chunks before storing the data. Using KZG polynomial commitments (the core mathematical scheme of ZK proofs), EigenDA only needs to download small amounts of data instead of downloading entire blocks of data. Unlike systems that use fraud proofs to detect malicious mis-encoding of data, EigenDA uses validity proofs in the form of KZG commitments, which enable nodes to verify the correct encoding of data.

Direct communication instead of P2P. Existing DA solutions use peer-to-peer (P2P) networks to transmit data chunks, where operators receive data chunks from their peers and then rebroadcast the same data chunks to others. This severely limits the achievable DA rates. In EigenDA, Disperser sends chunks of data directly to EigenDA’s operator. By relying on direct communication to decentralize data, EigenDA can delay confirmation of DA over the native network. This eliminates the significant gossip penalty that comes with P2P and results in faster data commitment times.

Safety Features

We use security as an umbrella term that covers safety and liveness, as well as decentralization and censorship resistance. The following features demonstrate the security of EigenDA:

EigenLayer. By using re-staking, EigenDA borrows two different security aspects from the EigenLayer system: 1. Economic security; 2. Decentralization. EigenDA is designed to leverage these two different trust elements in the EigenLayer and Ethereum ecosystems in a synergistic way.

Proof of Custody. A key failure mode for operators in EigenDA is for nodes to sign data items without actually storing them for the required time. To address this problem, EigenDA uses a mechanism called Proof of Custody, originally proposed by Justin Drake and Dankrad Feist of the Ethereum Foundation. With Proof of Custody, each operator must periodically compute and commit to the value of a function that can only be computed if they have stored their assigned block of data. If they attest to a data block before computing this function, anyone with access to their data item can slash the ETH held by the node.

Dual Arbitration Model. EigenDA also has a feature called dual arbitration, where two independent arbitrations can be asked to attest to the availability of data. For example, one arbitration can be composed of ETH re-stakers (ETH quorum) and the second arbitration can be composed of stakers of the rollup’s native token.

Censorship resistance. EigenDA provides higher instantaneous censorship resistance compared to coupled DA layers. This is because coupled DA architectures typically rely on a single leader or block proposer to linearly order data blocks, creating an instantaneous censorship choke point. In contrast, in EigenDA, rollup nodes can directly scatter and receive signatures from a majority of EigenDA nodes, thereby increasing censorship resistance to a majority of EigenDA nodes rather than being limited to a single leader.

Security Analysis

As discussed earlier, EigenDA is built on ETH staking via EigenLayer and uses erasure codes with configurable encoding ratios that can be set by rollups. There are three different angles to the security analysis of blockchain systems like EigenDA; we describe each angle and how it applies to EigenDA above:

Byzantine Fault Tolerance (BFT): Assume that some nodes are honest and follow the protocol exactly, while another portion of nodes are malicious and can deviate from the protocol arbitrarily.

EigenDA is secure in the sense that data can be retrieved as long as X% of the nodes are honest, where X can be anywhere from 10% to 50% depending on the encoding rate.

Nash equilibrium model: Analyzes the economic incentives of each node or small colluding node to follow the protocol, assuming that the node behaviors between different colluding nodes are independent.

As long as the size of the collusion is less than (1-X), storing and providing data to users is a Nash equilibrium: storing data is guaranteed by storage proof as an equilibrium, and the ETH of the node storing the data will be cut; providing data is guaranteed by dispersing the data among many nodes, thus triggering a competitive market to provide data.

Pure cryptoeconomic model: assumes that all shares are held by the same node and models the cost of economic corruption.

As long as the data is available, or in other words, as long as X% of the nodes are honest, the ETH staked by any node that does not host the data will be slashed. However, EigenDA does not have unconditional cryptoeconomic security; if all nodes collude and withhold data, then it may not be possible to slash them. In the dual arbitration model described earlier, when both ETH and the native rollup token are staked, the rollup can slash the native token even if ETH cannot be slashed.

As we can see, EigenDA is built on a trust model that requires not only economic trust from ETH staking, but also decentralization and independence of operators to operate safely. Fortunately, EigenLayer allows EigenDA to borrow these two trust mechanisms from Ethereum.

Customizability

Rollup developers can implement EigenDA with flexibility and adjust parameters as needed. The modular nature of EigenDA allows rollups to customize security/liveness tradeoffs, staking token models, erasure codes, accepted payment tokens, and more.

As discussed in the previous section, some of the most important flexible decisions in EigenDA are economic decisions. For example, a rollup can choose to use dual arbitration staking, where their own tokens are staked to guarantee data availability; or a rollup can choose an on-demand or reserved cost structure.

Strategic considerations

Finally, we believe EigenDA provides strategic value to rollups beyond its technical attributes.

Ethereum stakers and validators are the core force driving EigenLayer, and therefore EigenDA. By adopting EigenDA, Rollups can align with these Ethereum stakeholders who clearly value decentralization, censorship resistance, open access software, and composable, permissionless innovation.

EigenDA is planned to be one of the first many AVSs to be launched in the EigenLayer ecosystem. We foresee that as the number of AVSs grows, there will be a combination of benefits between them, which will benefit the end users of AVSs, which we expect to include various types of Rollups. For example, after EigenDA, we expect to see the launch of AVSs with use cases such as sorting, fast confirmation, monitoring networks, bridging, fair sorting, and even artificial intelligence.