Conducting Tests
Applies To: Windows Server 2003, Windows Server 2003 R2, Windows Server 2003 with SP1, Windows Server 2003 with SP2
After your team has designed test cases, it can begin testing. Figure 2.11 illustrates the point at which this step occurs in the overall lab development and testing cycle.
Figure 2.11 Conducting Tests
.gif "Conducting Tests")
When conducting tests, the tester must perform each test as described in the test case, evaluate the test results, escalate problems that arise until they are resolved, and document the test results. Figure 2.12 illustrates the testing process.
Figure 2.12 The Testing Process
.gif "The Testing Process")
Note
- Before testing begins, the tester might need to modify the test lab setup to meet the requirements specified in the test case.
When conducting a test, the tester needs to follow the written test cases carefully. To accurately assess results or to reproduce a test to compare results over time, team members need to know exactly which steps the tester took and the exact sequence in which those steps were performed.
When a test is complete, the tester must evaluate the results against the criteria in the test case to determine if the test passed or failed. Not all tests fail because there is a problem with your system implementation; tests can fail because there is a problem with the test itself, with the test lab setup, or with the proposed design. If one of these problems causes a test to fail, the tester should take the appropriate measures to remedy it. Table 2.2 shows some typical test problems and solutions.
Table 2.2 Test Problems and Solutions
| Problem | Potential Solution |
|---|---|
Test case problem |
Revise the test case, taking care to document all the changes made, and then rerun the test. |
Test lab setup problem |
Following the change control process for the lab, reconfigure the lab, and rerun the test. |
Design problem |
Follow the escalation procedure to notify the appropriate individuals about the problem. Prioritize outstanding problems and track them until they are resolved and the corresponding tests have been rerun. To prioritize problems, consider the potential impact and the probability that they will occur. |
Record all test results in the incident-tracking system, so you can monitor test progress.
Important
- Remember that the lab will change frequently as tests are run and new tests are begun. Make backups of baseline configurations so testers can quickly restore a computer to its prior state. Be sure to test the restore process. Document the backup files and store them in a safe, accessible place.