When starting the op-node, if the gossip broadcast of the unsafe payload triggers before the pipeline reset, there will be issues with determining whether to enter engine-sync mode. Additionally, in the new version of the engine-sync process, if geth is unable to sync due to certain reasons (e.g., zero peers), the current logic in op-node will still update the unsafe height. This PR addresses these 2 issues.
Rationale
This update adds handling logic for cases where insertUnsafe executes earlier than the reset process. It also includes a fallback mechanism to retrieve unsafe from geth. Additionally, the op-node's unsafe height is now only updated after engine-sync is completed.
Description
When starting the op-node, if the gossip broadcast of the unsafe payload triggers before the pipeline reset, there will be issues with determining whether to enter engine-sync mode. Additionally, in the new version of the engine-sync process, if geth is unable to sync due to certain reasons (e.g., zero peers), the current logic in op-node will still update the unsafe height. This PR addresses these 2 issues.
Rationale
This update adds handling logic for cases where insertUnsafe executes earlier than the reset process. It also includes a fallback mechanism to retrieve unsafe from geth. Additionally, the op-node's unsafe height is now only updated after engine-sync is completed.
Example
N/A
Changes