Overview of Defect Clustering
Defect Clustering :
When bugs are not properly distributed all over the application then some minor quantity of features causes major quality related issue in an application that is called defect clustering. A range of culprits gets indulged for the defect clustering purpose. For example legacy code prone to breaking the recent features which undergoes frequent changes and 3rd part integration also gets involved.
There are some more key indicators of defect clustering like notable number of test cases, but issues still become visible regularly. There is always two or more issue feature in which the bug seems to crop up very frequently.
Defect clustering in real shows how is the distribution of defect are not across the application evenly rather its more on the centralized side with limited section of application. Its basically a large system in which size changes, complexity, mistakes which impacts on the quality of the system and does affect a targeted module.
This concept is based on a pareto principal also know as the 80-20 rule, where its generally stated that approximately 80% of the issue occur due to 20% of the module. So, while this process of testing, most of the testers go through the phenomena i.e., where the area of code is complex and tricky. Then this information is used by the test designers in making the risk assessment planning the tests on the other hand helps in maintain track with on hotspots .
How to Minimize the defect ?
It seems general but if the organization starts hunting a round in its metrices in order to find major issue revolving around a particular application. If the product feature or code based. So, there the most gain can se made if the improvement initiative focus on a specific software. Repurposing a few extra resources and muscle can a difference in the targeted technology rather than abounding everything else in the interim .
Dealing Cluster in software Testing :
Generally, defect tends to cluster in area of the software under test where the causes can be complexity, algorithms or a higher number of integration in a few constrained segment of software. These defects cluster can be tricky to deal and find .
- Maximum number of defects are detected due to tester surrounded in the same area
- Takes less time hence, time consumption is less in the process and its also cost effective
- The initial iteration of testing are useful in identifying the defect cluster . moreover, it provides leverage to the tester as they use the gathered information while testing the application.
- At the starting stages iteration are useful but its not the base for final test case further the defects needs to be the ought and checked out in a software final conclusion cant be know through the initial iteration.
- Review Tests are done very carefully to avoid further introduction or defect .So creates confusion and wastage of time also hamper the quantity and efficiency of the software .