| || |
Planning: E2E test approach
Currently, we are engaged in an end-to-end (E2E) test for a telco company. The scope of the E2E covers all of the business process for telco, from service order, billing, revenue-payments, supply chain, and ect...
Anyway, we came up with 2 models for E2E test
The first model will cover creating service order, completing the service order, gets billed, and then customer makes payment. This will make one E2E test (Order-to-Bill-to-Payment)
Second model is dividing the first model into three parts. There will be an E2E test for service order, an E2E test for Billing followed by another E2E test for payments.
I would like to know your oppinions about the 2 models if any of the group encountered this experience.
Re: Planning: E2E test approach
As far as approach goes, E2E just means the testing will span the entire length of the applications involved, engaging all appropriate interfaces. Either of your scenarios would work, but I would favor approach #2, as the tests are more modular and would allow more flexibility in terms of regression testing later on. It would also allow better allocation of resources (one person could work on service orders, one on billing, etc.). In addition, a serious error in one part of the system wouldn't necessarily halt testing in another; if your scripts or scenarios are "dependent", this would not be the case.