Identifying throughput problems in performance testing
In performance testing, testers may encounter throughput or response time problems that seriously impact the application's performance. Identifying the throughput degradation problems and analyzing them to come up with the corresponding solution is an important task for WebSphere Commerce application developers and testers. With these throughput degradation problems solved, the performance of the WebSphere application improves significantly.
- Lower throughput or higher response times
This refers to throughput being lower than the required target or the response time being higher than the target. Code change, performance tuning, or hardware scalability may be able to increase the performance to your desired target.
- Throughput or response time degradation
This behavior is similar to lower throughput or higher response time, as mentioned above, except that there is a degradation or regression compared to a certain baseline established in the past. The baseline could be on another release of some product in your software stack or on another code release. Again, code change, performance tuning, or hardware scalability may be able to increase the performance to your desired target. In this case you have another tactic for troubleshooting, and that is the comparison with the previous test execution.
- Gradual throughput (or response time) degradation (GTD)
This is a systemic, and potentially complex, problem that you may encounter in performance testing. The main problem is that during the test interval (say, 3 hours), the throughput decreases gradually, or the response time gets longer and longer, and this trend does not stabalize at any acceptable level. You can easily find this trend in your test report. If the downward slope is not significant then you may have to either stress your system more or let the test run for a longer duration to observe a noticeable depreciation.
