What is Data Availability?
To explain Alternative Data Availability (Alternative DA), we should start by defining what Data Availability means in the first place.
As per Ethereum.org:
Data availability refers to the confidence a user can have that the data required to verify a block is really available to all network participants.
This is important and mandatory for every chain because it ensures that each network participant (or node) securing the chain can independently verify that the information it receives in a block of data is correct, guaranteeing that all changes proposed by a given block match those that the node computes independently.
In practice, traditional "monolithic" chains like Ethereum, as well as the majority of EVM rollups today, ensure Data Availability by storing and downloading state data for each full node, downloading a copy of all the data in each block to guarantee it cannot — and will not — accept any invalid transactions.
Why Alternative Data Availability?
As mentioned, in order to ensure any invalid transactions processed on a rollup are caught and reverted, users need a way to make sure that the rollup’s block data is published and accessible. This is true for both optimistic and ZK-rollups, just as it is for all other blockchains. Right now, both Ethereum itself and the vast majority of EVM rollups solve this by simply posting and storing all of the rollup blocks on Ethereum as a Data Availability Layer (the location where transaction data is stored) using the method described above.
But, this redundant approach creates high costs and low throughput, as each node on the network is forced to store massive amounts of repeated data. As a result, over 95% of the costs associated with a rollup transaction today comes from posting data to Ethereum.
This begs the question: How can we maintain access to all relevant data without storing each entire block?
The short answer: Alternative DA.
Alternative DA solutions provide a separate layer for blockchains to handle Data Availability in a modular fashion, apart from the execution and consensus layers, using various innovative techniques to reduce costs by 100x and significantly improve throughput.
The Leading Alternative DA Solutions
Currently, the two primary Alternative DA approaches are Data Availability Sampling (DAS) and Data Availability Committees (DACs).
Data Availability Sampling
In Data Availability Sampling (DAS), rather than downloading excessive data to all nodes, each individual node is only required to download a small, random sample of the total block data. Through a clever encoding scheme, this is all that's required for light clients to verify with high confidence that all of the data in the block in available. Celestia is a popular Alternative DA solution that utilizes DAS along with Namespaced Merkle Trees to make low-cost, reliable sampling possible over a peer-to-peer network.
Data Availability Committees
Data Availability Committees (DACs) are trusted sets of nodes that store copies of data offline, providing access to that data in the event of any disputes and publishing attestations to provide that the off-chain data is readily available. Arbitrum Anytrust is a DA solution that makes use of an external DAC (rather than posting data to Ethereum), utilizing a mild trust assumption that dramatically lowers costs for chains built with it.
Build your rollup with Alternative DA
At Caldera, we're proud to offer integrations with Celestia and Arbitrum Anytrust, along with Eigenlayer and Near. as Alternative Data Availability layers for all Caldera Chains, so that teams building with us can leverage these innovative modular scaling solutions to save significant costs.
