Data availability (DA) ensures that transaction data is accessible to anyone who needs to verify the blockchain state, reconstruct history, or detect fraud. This matters for rollup security: even with valid proofs, if the underlying transaction data is hidden, users can't independently verify state or prove fraud. The data availability problem asks: how do we know data was published without downloading all of it? Naive solutions don't scale. Ethereum's current approach has rollups post data directly as calldata, which is expensive but guarantees availability through the Ethereum validator set. EIP-4844 (Proto-Danksharding) introduces 'blobs', temporary data storage that's cheaper than calldata but still secured by Ethereum. Dedicated data availability layers like Celestia and EigenDA offer even cheaper alternatives by specializing in DA rather than general computation. These use techniques like data availability sampling (DAS), where nodes verify availability by checking random samples rather than downloading everything. Validiums take the extreme approach: post proofs to Ethereum but store data off-chain, greatly reducing costs but weakening security guarantees. The DA layer choice involves fundamental security-cost tradeoffs that define different categories of Layer 2 solutions.
Back to Glossary