Data availability layer DA track observation: Modular blockchain is accelerating

1. What is the data availability layer?

The concept of data availability differs from decentralized file systems and databases, as articulated on the Ethereum and Celestia websites.

• **Ethereum: **Data availability means to ensure that the block released by the block proposer, and all transaction data contained in the block, are available to other network participants.
• Celestia: Data availability correlates to availability of data published in the latest block.

However, decentralized file systems and databases primarily ensure that data stored by users is accessible. They don't solve the same problems as the data availability layer.

Of course, data availability and decentralized storage also have something in common. Technically, they all use verification methods such as erasure codes and Proof of random access to verify the availability of data.

2. What is the technology for data availability?

There are currently several types of data availability layers, including:

• **Ethereum. **ETH acts as the DA (Data Availability) layer of the second layer Rollup.
• **Modular blockchain represented by Celestia. **Celestia is a purpose-designed DA layer that only handles data availability and does not execute transactions. It kicks off the modular blockchain trend in 2022. Such a modular blockchain also has projects such as Avail.
• **EigenDA and other DA products. **Data availability is ensured through committees.

Ethereum

Ethereum is already the DA layer of L2 Rollup. Ethereum Rollup packages and submits the transaction data of the second layer to the Ethereum smart contract, and Ethereum stores the L2 data and guarantees the data availability of these transaction data.

While rollups can scale the throughput of ETH through off-chain computation, their capacity is limited by the L1 ETH blockchain data throughput. Therefore, for Rollups to use Ethereum as a data availability layer, Ethereum needs to increase its data storage and processing capabilities.

Ethereum is actively promoting the upgrade of Danksharding. Danksharding can improve the ability of Ethereum to store data and make ETH a more powerful DA layer. **

• In Danksharding's scheme, data availability is delegated to individual shards, and each validator only needs to run a full node for its shard, and run other shards at light client capacity. Sharding increases Rollup's data space because the work of storing data is split across different shards.
• Proto-danksharding (EIP-4844) is an initial implementation of Dankshading and is expected to be launched in the second half of this year. Proto-danksharding has defined new transaction types for data block blobs, as well as the precompiled code required to verify blobs, implementing most of the core logic of Danksharding. Each blob is about 125kB in size, while a block is only 90kB. Currently, up to eight blobs can be mounted per chunk, resulting in an additional 1MB of storage. After sharding in Danksharding, the verifier only needs to download part of the blob data; the data in Proto-danksharding is not sharded, and the verifier still needs to download and directly verify the data availability of all blobs, so this is a step in Danksharding . After the implementation of EIP4844, Blob can store 10 times more data than Calldata under the same gas consumption. Rollup data can be stored in Blobs in the future, reducing transaction fees by an order of magnitude. Danksharding will become even cheaper once fully implemented.
• In general, Danksharding can improve the data storage capacity of Ethereum, reduce the cost of ETH being used as DA, and thus become a more powerful DA layer.

Technical principle of Danksharding:

• Now all the data on Ethereum is stored on the chain, forcing block producers to publish all transaction data to the chain and let the verification nodes download it. Danksharding introduces a data blob, which is stored off-chain and only puts the KZG commitment of the data on the chain. The storage time is shortened to 1-3 months, and the blob space is as large as ~125kB (the ETH block size is only ~90kB). Therefore, the data storage time becomes shorter and the data volume becomes larger.
• The node verifies a blob is to query whether the blob is available, that is, whether it can be downloaded from the network. Danksharding relies on data availability sampling to verify that the network has seen the full contents of the blob.
• Due to the shortened data storage time, the retrievability is reduced, that is, the ability to query historical blobs older than one month is reduced. For security, the EVM cannot access the blob data off-chain, but can only access the on-chain commitment of the blob, preventing EVM from failing due to inaccessible long-term blobs.
• In order to verify the availability of Blob data, ETH uses the verification method of data availability sampling. Multiple rounds of random sampling of a small number of samples are performed on the data to verify its usability. This allows blockchain nodes to verify their availability without downloading the full blob.
• By using erasure code technology to redundant data fragments, data recoverability is improved, and the combination of erasure code and data availability sampling can prevent nodes from hiding data.

Modular blockchain DA layer represented by Celestia

Celestia is a modular blockchain that only orders and publishes transactions, not executes them. By decoupling the consensus layer and application execution layer, Celestia modularizes the blockchain technology stack and opens up new possibilities for decentralized application builders.

• Celestia is responsible for the DA layer, ETH and other networks are responsible for consensus and settlement, and the application chain is responsible for execution.
• Celestia is responsible for the DA layer and the consensus layer, while settlement and execution are handled by AppChain. Alternatively, settlement can use Cevmos, while execution remains with AppChain.

**Celestia's verification method for data availability is similar to that of Ethereum. **Using two-dimensional Reed-Solomon erasure codes, a set of random sampling is designed to verify data and restore data to ensure data availability.

Celestia is also very different from **ETH. **

• Celestia only concentrates on the DA layer and consensus layer of Rollup, while ETH also makes the settlement layer for Rollup.
• Celestia only does DA layer and consensus layer, does not have a Turing-complete smart contract virtual machine, and does not support smart contracts.
• Ethereum’s Rollup cannot be forked, but Celestia’s Sovereign Rollup can be forked.
• Rollup communicates assets with Celestia through the bridge. Since Celestia does not support smart contracts, the cross-chain bridge has limited functions and may only transfer DA Token.

Celestia's ecology is also developing rapidly.

Off-chain DA layer

The off-chain DA layer mainly includes two types

• Data Availability Committee
• Blockchain producers and trusted parties form the Data Availability Committee (DAC), which stores transaction data off-chain and does not store it on the blockchain. DAC needs to be provided in case of dispute. DAC members will also publish on-chain proofs to prove that the above data is indeed available.
• For example, validiums and plasma use off-chain storage to reduce on-chain data.
• Data Availability Committee for PoS
• The committee deposited a deposit and added a reward and punishment mechanism to enhance the credibility of DAC. Other aspects are the same as the above-mentioned DAC.
• For example, EigenDA is built on the second pledge protocol EigenLayer, attracting Ethereum validators to participate in EigenDA's data availability guarantee work.

3. What data availability layers are available now?

Currently, there are several data availability projects, including:

• ETH: Currently, ETH is used as the data availability layer for L2 optimistic rollups and zk rollups. The launch of EIP4844 (Proto-Danksharding) in the second half of the year will benefit the development of L2. It will improve the ability of Ethereum to store data and reduce storage costs.
• **Celestia: **Celestia is primarily designed for consensus and data availability layers. Celestia's test network will be launched in June 2022, and Celestia has brought fire to blockchain modularization. Celestia needs to build its own ecosystem and compete with Ethereum. At present, some ecological projects have been developed.
• **Avail:**Avail was originally published by Polygon in June 2022. However, after its founders left Polygon, Avail became an independent modular blockchain project and released a testnet. Similar to Celestia Avail is a separate consensus and DA layer, just like Celestia. Avail had planned to bridge the mainnet to Polygon and use MATIC as the base currency. MATIC is a more mature token compared to Celestia token.
• EigenDA: EigenDA is an Ethereum-based DA layer that incentivizes validators to maintain the DA network through ETH re-mortgaging, eliminating the startup burden required by Celestia.
• **Other off-chain DA: **Validium stores transaction data off-chain, and uses ETH for consensus and settlement. With the widespread adoption of Celestia and Danksharding, the adoption of Validium may gradually decrease or even exit.

Summarize

• A data availability layer is a promising and important way to scale blockchains. Ethereum is the data availability layer for Layer 2 Rollup. Celestia wants to build a purer data availability layer than Ethereum. There are also some off-chain data availability layer solutions such as EigenDA.
• The launch of EIP4844 in the second half of the year makes ETH a more powerful DA layer, and Celestia's main network may also be launched in the second half of this year. The current data availability offerings each have strengths that deserve our close attention this year.
• Celestia's technology still needs market verification, and ETH and Celestia may also approach each other technically in the future.