Data availability

Deterministic ordering of data commitments is part of the onchain consensus, but actual data storage, synchronization, and function/task execution is performed offchain. Like most existing data availability protocols, Recall relies on a combination of random sampling and erasure coding.

Alpha entanglement

Unlike many systems which use Reed-Solomon erasure codes, Recall employs alpha entanglement (AE) codes, a specialized erasure coding scheme designed for distributed systems. AE codes create redundancy with less storage overhead, enhance data integrity through block entanglement, and enable efficient recovery of missing data blocks without reconstructing large portions of data. This reduces repair costs and improves scalability, even in dynamic environments.

Alpha entanglement codes are constructed by "entangling" data and redundant blocks in a multi-dimensional lattice structure. This entanglement process propagates redundant information across the mesh-like structure. This process allows for decentralized and scalable repairs, supports random access to data, and optimizes storage by grouping data based on retention periods (TTL). The result is a highly efficient, decentralized system that maintains data availability and integrity even under challenging conditions.