[Metrics] `exception_count` never available

[eiabea]

Steps to reproduce:

1. Start the operator with available execution and consensus clients
2. Stop the execution client (rpc endpoint)
3. Observe logs:

2026-04-28 16:46:43,617 ERROR    404, message='Not Found', url='https://rpc.execution.eth.example.com'
Traceback (most recent call last):
  File "/app/src/common/tasks.py", line 17, in run
    await self.process_block(interrupt_handler)
  File "/app/src/commands/start/base.py", line 128, in process_block
    await wait_execution_catch_up_consensus(
  File "/app/src/common/startup_check.py", line 258, in wait_execution_catch_up_consensus
    execution_block_number = await execution_client.eth.get_block_number()
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

This exception should increase the exception_count, but can not be seen at the /metrcis endpoint

[evgeny-stakewise]

Which release did you run? exception_count PR is not included in releases yet. Only available on master branch.

[eiabea]

@evgeny-stakewise I checked the Operator Monitoring page, but you're right I checked the master branch instead of the latest release.

[evgeny-stakewise]

@eiabea It seems the documentation is ahead of the actual release. Today, I published v4.1.11, which includes the exception_count feature in the latest release

[evgeny-stakewise]

@eiabea So, you believe this is a bug in the master branch, correct? What value did you set for the EXECUTION_ENDPOINTS parameter? If the endpoint is incorrect, the startup check will hang, and the actual tasks won't start. However, the exception in your logs is coming from a task. This is a bit confusing.

[eiabea]

I think this issue can be closed, i just thought i should get this counter because of the documentation, but didn't looked in the actual code of my running version.

Thank you for adding this metric in the newest release 💪