Data Availability Sampling

What Is Data Availability Sampling?

Data Availability Sampling is a powerful solution to the data availability problem, a major challenge in scaling blockchains. In simple terms, it provides a highly reliable guarantee that a block producer has made all transaction data public, which is essential for network security.

DAS allows network participants to verify that all transaction data for a new block has been published without needing to download the entire block themselves. This saves a massive amount of storage and bandwidth, making it a cornerstone of blockchain scalability.

This method is particularly important for light nodes—participants who don’t have the resources to download and process every single transaction. It’s also a critical component for Layer 2 rollups, which bundle transactions together off-chain and then post a compressed summary back to their main chain. DAS allows the main chain to confirm that this data is available without being overwhelmed.

How Does Data Availability Sampling Work?

DAS uses a combination of two core technologies:

Erasure Coding

First, the block producer takes the original block data and uses a technique called erasure coding to expand it. This process adds extra, redundant pieces of data (known as “parity data”). The key feature of erasure coding is that the original data can be fully reconstructed from just a fraction of these expanded pieces. For example, the data might be doubled in size, but you would only need 50% of the new, larger data set to recover 100% of the original.

Random Sampling

Once the data is encoded and published, light nodes perform the “sampling.” Instead of downloading the whole block, each light node requests a few small, randomly selected pieces of the expanded data from the network.

The security of DAS comes from statistical probability. If a malicious block producer tries to hide even a tiny portion of the original data, it would cause a large portion of the erasure-coded pieces to be unavailable. Therefore, when many light nodes perform their random checks, there is an extremely high probability that some of them will fail to retrieve their requested pieces, quickly revealing the fraud.

If all the light nodes successfully receive the random pieces they requested, the network can be mathematically certain (e.g., 99.999% confident) that the entire set of original data was published, even though no single node downloaded all of it. This process allows for much larger blocks and significantly lower costs for rollups, paving the way for massive scalability improvements on blockchains like Ethereum.