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
We noticed that the consensus in our testnet occasionally stops for several minutes, causing no new checkpoints to be created. From the metrics, it’s clear that the number of missing blocks is higher than normal during these periods when consensus stops.
The following warning appears in the logs during this time::
WARN consensus_core::synchronizer: No results returned while requesting missing blocks
We just encountered another temporary consensus stop again
2024-12-20T09:23:36.466720Z WARN consensus_core::authority_service: Errored while trying to fetch missing ancestors via synchronizer: Synchronizer for fetching blocks directly from I is saturated
2024-12-20T09:23:36.468730Z WARN consensus_core::authority_service: Errored while trying to fetch missing ancestors via synchronizer: Synchronizer for fetching blocks directly from P is saturated
2024-12-20T09:23:36.728751Z WARN consensus_core::authority_service: Errored while trying to fetch missing ancestors via synchronizer: Synchronizer for fetching blocks directly from F is saturated
2024-12-20T09:23:37.241722Z WARN consensus_core::authority_service: Errored while trying to fetch missing ancestors via synchronizer: Synchronizer for fetching blocks directly from D is saturated
2024-12-20T09:23:40.128437Z WARN consensus_core::synchronizer: No results returned while requesting missing blocks
2024-12-20T09:23:46.420191Z WARN consensus_core::synchronizer: Max retries 6 reached while trying to fetch blocks from peer F.
And this one, although the consensus didn't stop, but the consensus latency is still getting very high, you can notice that this happened after the node came back online from being offline.
WARN consensus_core::synchronizer: No results returned while requesting missing blocks
Bug description
We noticed that the consensus in our testnet occasionally stops for several minutes, causing no new checkpoints to be created. From the metrics, it’s clear that the number of missing blocks is higher than normal during these periods when consensus stops.
The following warning appears in the logs during this time::
Explore-logs-2024-12-17 18_21_17.txt
The text was updated successfully, but these errors were encountered: