Retry indexing from the same block range on handler failure

[mpetrun5]

Currently, if an event handler fails we log the error and continue to the next one, but it should
not continue and restart the loop from the same block range.

Implementation details

When an event handler raises an error loop through again.

Testing details

Unit test if possible

Acceptance Criteria

• we are resistant to RPC or unpredictable event failures and we can retry them until it works again