|
As a software developer, to deal with numerous bugs every day is common. Before developing the system, each task will be carefully analyzed and classified; it provides different processes to different functions. However, unpredictable bugs still can not be avoided after launching the actual operation of the system, no matter how careful planning of each unit testing and integrated testing. This thesis will take a log file created by a software company's development system as an example, systematical classifying the bugs, and then building a log file based software bug automatic tracking method.
There are 524 log data collected in this thesis and 159 in total after deducting repeated ones. Among the 159 data, 63 are bugs; 45 are data issues; 12 are internet issues; 39 are other issues. Experimental results indicate that the method can correctly track 80% of bugs. In sum, it can greatly enhance the software developers to track the efficiency of bugs.
|