Open trianglesphere opened 8 months ago
@trianglesphere during backlog grooming today we moved this issue out of the current cycle to the next one.
The rationale is that Base will be merging upstream geth changes this cycle (https://github.com/ethereum-optimism/protocol-quest/issues/186) and we want to see if that helps address the snapshot corruption issue first before debugging further.
We are sometimes seeing logs like
Failed to journal state snapshot
on shutdown. This does not appear to be due to a hard shutdown, but a race condition. If this occurs, the snapshot has to be regenerated on op-geth which takes 6-12 hours on mainnet and ruins performance in the mean time.