Three useful terminologies are:
1. Useful but not the best practice, based on specific test case technologies;
For example
1) the system test case design has passed the review
2) completed the system test according to the system test plan
3) The functional coverage rate of the system test reaches 100%.
4) The system functions and performance meet the requirements of the Product Specification.
5) errors found during system tests have been modified and the defect repair rates at all levels have reached the standard.
6) There are no class A, B, and C defects after the system test.
7) Class D defects may exist, not exceeding 5% of the total Defects
8) E-type defects can exist, and cannot exceed 10% of the total Defects
2. The most valuable criterion is to describe the conditions for the end of the test based on the exact quantity, and to predict the error discovery period.
Involved:
Defect rate (defect rate): the ratio of defects found after the full test case is executed;
Defect density (defect density): number of defects detected per unit of length (generally 1000 rows) code or FP (Function Point)
Test Coverage: the extent to which the Code or use case is covered;
3. record the number of errors found per unit time during the testing process. Check the shape of the statistical curve to determine whether to continue or terminate the testing of this phase.
Defect trend: whether the trend of the defect is gentle; cost;
The best ending principle may be the combination of the above three types. Unit Tests use the first type, while other tests use the combination of the last two types. This is a good strategy.
You can also consider the legacy risks (residual risk) defined in the schedule)
Exit criteria of System Test