| || |
removing outliers from timing files
Is there a way to not include outlier data as part of the timing files? I ran a load test and had a spike due to a sudden increase in network activity, which skewed my response time data regarding mean, max, and standard deviation. I want to remove the spike so I can have truer number when creating reports.
Any help is appreciated.
Re: removing outliers from timing files
- Export data to Excel and trunc : is it true true numbers !?!
- Analyze results on time frame without network perturbation (Analyze's time range option)
- Define thresholds to help visualize that response times are not so bad when network is not perturbating
- Test when network is not overloaded
- Is network load peak related to on-going perf test ?