Determine how to efficiently index Beacon State with proof #77

New Issue

dboreham · 2022-09-14T23:01:51Z

dboreham commented

2022-09-14 23:01:51 +00:00

Beacon state is a 60MB structured record that significantly changes every slot (12 seconds). If we want to represent this historical state in a hash-linked form, similar to our approach with ipld-eth, we need to be able to split the SSZ tree into changed and non-changed components that have their own hashes, then store the resulting pieces in a way that allows them to be found for a given block height, and also stitched together again to make a whole Beacon State record.

i-norden commented

2022-10-27 15:58:58 +00:00

Difference iteration statediffing approach (analogous to what we do in the execution layer):

Materialize SSZ serialized beacon states at n and n-1 into SSZ Merkle trees
Perform difference iteration of the two trees to collect only the nodes which form the symmetric difference set between the two trees
Index the individual nodes in the difference set

Difference iteration statediffing approach (analogous to what we do in the execution layer): 1. Materialize SSZ serialized beacon states at n and n-1 into SSZ Merkle trees 2. Perform difference iteration of the two trees to collect only the nodes which form the symmetric difference set between the two trees 3. Index the individual nodes in the difference set

Sign in to join this conversation.

No Milestone

No project

No Assignees

2 Participants

Notifications

Due Date

The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: cerc-io/ipld-eth-beacon-indexer#77