The following screenshot of the first evaluation scenario (sub-scenario basic) presents the graphical user interface of our tooling. An interactive version of the topological difference graph is shown in the center accompanied by the identified changes listed on the right. On top is the interface to specify which and how many traces should be considered for constructing the topological difference. It is also possible to specify different periods for both the baseline and canary variants.
All of the rankings presented in the screenshots of this appendix are based on the extended hybrid heuristics (HYB Ext).
Single ranking entries can be selected resulting in a graphical highlighting of the respective change in the graphs presentation.
The following screenshot shows the graphical representation including the ranking for the delayed sub-scenario of the first evaluation scenario.
Note how not only the ranking changes, but also the color coding of the ranking (which is based on the heuristic's scores) and the coding of the edges in the graph.
The following screenshots cover the second evaluation scenario.
It starts off with the basic sub-scenario followed by the scenarios including simulated performance delays.
