During running of test, graphs provide correct information (right curve, values, etc.). However, resulting graphs (and graphs in the report) show me much lower values.
Please, look at the screenshots taken during the running and final report. You will see, that real-time values much higher than in report.
Development team requested information which was sent about month ago. Since then they keep silence.
The difference might be due to aggregated data. The aggregation level is not the same during the runtime compared to the Results graphs.
Looking at the report statistics we can see that the maximum value for that CPU was 91.3 which makes sense with what you saw in the runtime graph.
So if you open your test result and plot that CPU counter by changing the rendering interval in the NeoLoad preferences-->Graphs section, i guess you should see a more accurate graph.
Thank you, Nouredine.
Actually, the best results can be seen if test duration is about 10 minutes and sampling interval is 1 second. Real-time and final graphs are the same and match Performance monitors of the monitored machines.
However, "real life" scenarios require much longer (30-60 minutes) duration. I'd expect to see peaks (along with statistical values in the tables) on the graphs. You know, managers look on the visual part in first place. :)