Enrico Del Fante
Enrico Del Fante
In case of network instability up to a level in which it becomes difficult to sync to an accepted canonical chain, we could implement a syncing strategy that could work...
Computing attestation aggregation currently takes at least 200ms on holesky node. We could implement a strategy to trigger the aggregation in parallel with the block production flow. there is a...
on 1.9rc2 at startup teku tries to post ssz content and immediately fallback to JSON, which works ``` {"@timestamp":"2025-04-15T10:00:33,745","level":"DEBUG","thread":"OkHttp http://127.0.0.1:18550/...","class":"RestBuilderClient","message":"Failed to register validator using SSZ. Will retry using JSON (error:...
We should remove `--Xp2p-reworked-sidecar-custody-sync-enabled` once it is default for 1 or 2 releases. This also imply removal of `isEarliestAvailableDataColumnSlotSupported` in `CombinedChainDataClient` and delete `DasCustodySync` too. Than we should rename `--Xp2p-reworked-sidecar-custody-sync-*`...
to explore how difficult it is. This works!! ## Documentation - [ ] I thought about documentation and added the `doc-change-required` label to this PR if updates are required. ##...
This is just an idea, to be validated in real world, and it could explain why, before #10219, we were actually duplicating download requests instead of just rebuilding the tracker...
We do not trigger backfill because we fail to detect the change. In `DasCustodyBackfiller` constructor we initialize with the current synced custody with: ```java this.currentSyncCustodyGroupCount = custodyGroupCountManager.getCustodyGroupSyncedCount(); ``` but `custodyGroupCountManager.getCustodyGroupSyncedCount()`...
We currently write on DB `CustodyGroupCount`, but we should migrate to writing the actual column indices to custody. This way we can detect when nodeid changes (and columns reshuffle while...
We do not currently backfill columns when we do a checkpoint sync. It was implemented in the past but the feature has been "lost".