Are you trying convert float to numeric and you get "Arithmetic overflow error converting float to data type numeric" error? This could happen even if all was working before and now it fails. The reason should be in precision and scales settings.
Contact