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
max blobs will not be higher than 16 during the next year.
I think that the target for Danksharding was 128 blobs, and this is still the target for DAS. So, the max is likely targeted to 128
So for the txpool pool and blob cache, I think some implementation options are:
1. Dynamic based on engine API
2. A staged config change, e.g. increase cache multiple to 16 now and plan to increase in the future
3. Go straight for 128 and hopefully don't have to think about it again
Sounds like Fabio is in favour of (3). I think I am too, but still considering (2) for blob cache depending on effort/performance impact. It would be nice not to increase Besu's memory requirement even more in the short term.
It's likely that https://eips.ethereum.org/EIPS/eip-7691 will go in which has a max of 9 blobs, so I think we should at least increase the cache to that while we investigate other solutions
The text was updated successfully, but these errors were encountered:
EIP-7742 was replaced with EIP-7840 so these values can be read from the genesis config now instead...however, we could still preemptively oversize if we choose.
EIP-7742 changes the hardcoded max blob count from 6 to be dynamic (via engine API).
Outside of the protocol changes, I've found 3 places where this probably needs to change:
a.
besu/ethereum/eth/src/main/java/org/hyperledger/besu/ethereum/eth/transactions/BlobCache.java
Line 37 in 1671306
b.
besu/ethereum/blockcreation/src/main/java/org/hyperledger/besu/ethereum/blockcreation/txselection/selectors/BlobSizeTransactionSelector.java
Line 59 in 1671306
c.
besu/ethereum/eth/src/main/java/org/hyperledger/besu/ethereum/eth/transactions/TransactionPoolConfiguration.java
Lines 77 to 78 in 1671306
from Mikhail:
So for the txpool pool and blob cache, I think some implementation options are:
Sounds like Fabio is in favour of (3). I think I am too, but still considering (2) for blob cache depending on effort/performance impact. It would be nice not to increase Besu's memory requirement even more in the short term.
It's likely that https://eips.ethereum.org/EIPS/eip-7691 will go in which has a max of 9 blobs, so I think we should at least increase the cache to that while we investigate other solutions
The text was updated successfully, but these errors were encountered: