Replies: 3 comments 11 replies
|
The lower bound for this should be at least the Frauf Proof Window + Worst case Block Reconstruction time |
1 reply
|
I would make it == unbonding period or trusting period. |
3 replies
|
I am unclear about this in-production today, are the data blobs stored indefinitely today or are they pruned? Personally I think with competitors like EIP-4844 danksharding, it is better if Celestia nodes hold the data for longer periods to be an archival back-up solution as well (after all the value of data availability sampling is supposed to enable Celestia to scale to the number of light nodes and can have much more capacity in-theory). |
7 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Currently, celestia-node storage nodes store all historical data blobs. However, the purpose of data availability protocols is not store all historical data blobs for ever. Therefore, we should discuss and define a socially-accepted network-wide constant number of days ("storage window") after which the network does not guarantee that the data will be retrievable. We intend to make it possible for rollup full nodes to easily be able to contribute to the storage of historical data blobs within their rollup (see milestone 3 in #2033).
All reactions