Thanks for the suggestion and this is definitely one option. However, I was hoping to gain more insight into improving my unit and integration tests (which have high coverage), to cover most scenarios and break, if the integration breaks.