Glossário

Data Availability Sampling (DAS)

Hard

Data Availability Sampling is a method of verifying data availability without needing to download full blocks.

What Is Data Availability Sampling?

Data Availability Sampling is a method of verifying data availability without needing to download full blocks. By using DAS, light clients can sample small, random subsets of data to probabilistically ensure the complete dataset is accessible. This provides a secure and efficient way to verify that data is available across a peer-to-peer network. 

How Does DAS Work?

Nodes sample small, random portions of erasure-coded data. If they can successfully retrieve and verify the samples, it provides confidence that the entire dataset is available. With each additional sample that is taken, the confidence level that all data is available increases as the likelihood that data is unavailable decreases. This method provides light clients the ability to sample directly from the network and provide data availability guarantees independently, without burdening any single node or relying on light clients, thereby increasing the quality and decentralization of the guarantees light nodes can provide from low-powered devices like phones and laptops.

Importance in Decentralized Applications

DAS is vital for trust-minimized applications and sovereign rollups that require a reliable method to verify data availability without reliance on full nodes or trusted entities, like a data availability committee (DAC). This ensures that transaction data remains accessible and verifiable by anyone, even when stored by a separate data availability network, providing enhanced security and verifiability for applications and users.

Example of DAS in Action

 Anyone can use DAS to verify that the blockchain is operating as it is meant to. For example, if someone decides to give themselves an extra 100 tokens, everyone can use DAS to verify that the person just made up this transaction and rejected it. This is because they all have the ability to verify the real data through data availability sampling.


Difference Between DAS and Traditional Verification

With traditional verification, the only way to be sure of a valid transaction is to download all the data. As the blockchain gets bigger this becomes harder and harder requiring massive computers. Traditional light clients attempt to solve this by having access to a trusted full node, but traditional light clients have no way of independently checking if the full node they are connected to is lying to them or not. With Data Availability Sampling, this is not the case. The light client possesses the ability to independently check and verify whether data is in fact available or not.


About Author: Anurag Arjun, Co-Founder

Anurag Arjun is the co-founder of Avail, a web3 infrastructure layer designed to allow modular execution layers to scale and interoperate in a trust-minimized way. He entered the blockchain industry in 2017, founding Matic Network, which evolved into Polygon Labs. By 2020, he launched Avail within the Polygon ecosystem, utilizing his background in research, economics, and engineering. In March 2023, he led Avail's development from Polygon to focus on developing a more modular and scalable blockchain stack for the web3 community.