Data Availability Sampling
A technique that lets a node verify a block's full data was genuinely published by checking only a small random sample of it, rather than downloading everything.
Data availability sampling is a technique that lets a node verify a block's full underlying data was genuinely published and available, by downloading and checking only a small random sample of that data, rather than downloading the entire dataset directly.
This is a core technique behind scaling data availability (see Kaavon's "Data Availability" glossary definition) for designs like Danksharding (see Kaavon's "Danksharding" glossary definition), since it lets a lightweight node gain strong statistical confidence that data is available without the heavy bandwidth cost of downloading every byte of it directly.
Examples
- A node using data availability sampling can gain strong statistical confidence that a block's full data is available, without downloading the entire dataset itself.
- Data availability sampling is a core technique behind scaling designs like Danksharding, letting lightweight nodes verify availability efficiently.