You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Add a boolean VM configuration option like historical-proofs which defaults to false.
In the API layer, reject calls to eth_getProof for blocks before latest - tipBufferSize, this is because with a path based trie, there is no good way to recall historical trie nodes and storing these nodes is a burden for validators. For most use cases (eg, message passing) it should suffice to verify state near the head of the chain, as often these use cases are time-sensitive.
Additional considerations / discussion topics:
Entities with an interest in maintaining historical proofs can maintain their own indexes or use the hash db and modify the config option from its default.
We can make the config option an integer which also controls tipBufferSize, which will also need to match the in-memory layers we keep at tip for the pathdb.
Should we consider an upstream PR?
The text was updated successfully, but these errors were encountered:
historical-proofs
which defaults tofalse
.eth_getProof
for blocks beforelatest - tipBufferSize
, this is because with a path based trie, there is no good way to recall historical trie nodes and storing these nodes is a burden for validators. For most use cases (eg, message passing) it should suffice to verify state near the head of the chain, as often these use cases are time-sensitive.Additional considerations / discussion topics:
tipBufferSize
, which will also need to match the in-memory layers we keep at tip for the pathdb.The text was updated successfully, but these errors were encountered: