op-node fails to sync when --syncmode=execution-layer is enabled with op-reth backend

Hi OP-Stack Team,

We recently deployed an Optimism RPC node using the op-node + op-reth architecture.

During the deployment, we encountered a synchronization issue:

When we start op-node with the --syncmode=execution-layer flag enabled, the node completely fails to sync and gets stuck.

However, as soon as we remove the --syncmode=execution-layer flag, the node starts syncing perfectly and works as expected.

We would like to clarify the following questions:

Is the --syncmode=execution-layer parameter strictly mandatory for the op-node + op-reth setup under the latest specifications?

What are the concrete downsides or architectural impacts if we run op-node without this flag (leaving it to the default sync mode)? Does it affect the finality validation or EL/CL decoupling safety?

Is this a known compatibility/timing issue specifically between op-node and op-reth's Engine API implementation during the initial sync phase?

Looking forward to your insights!

Start by reproducing the initial sync behavior with op-node and the op-reth backend, comparing runs with and without --syncmode=execution-layer. Investigate whether the failure is a compatibility or timing issue in the Engine API interaction, and document the flag requirements, architectural impacts, and finality-validation implications.

Written by the indexing model from the issue text.