1. ## Defect clustering

Hello,

I know is another topic with "Defect clustering" but i really don't understand what it is .. i also checked google and wikipedia.

So if anyone have a definition, an example or a link/blog anything with "Defect clustering" please replay.

Thank You!

3. ## Re: Defect clustering

I don't think its the same thing ?

4. ## Re: Defect clustering

If you don't understand what it is how do you know?

What do you think either one is? Is there some defintion you could put to either one?

- M

5. ## Re: Defect clustering

Defect Clustering is a term that refers to the tendancy of defects to cluster in an area, meaning if you find one defect executing a particular test case, the likelihood that another defect resides in that same area is increased.

6. ## Re: Defect clustering

could sum 1 plz tell me what is difference between peanut clustering and peanut density

7. ## Re: Defect clustering

BUT SERIOUSLY FOLKS:

An observation made by a developer-deputized-as-a-tester a couple of days ago here that made me look at defect clustering a different way.

I had previously thought of defect clustering as a fruitful way to encounter buggy functionality: if a defect appears around certain functions then those functions are a good place to keep testing because it's complex enough that errors were made - we need to thoroughly exercise that area because there are likely to be multiple bugs.

The developer's point was that, after a single bug was found in that area, that area was known to be buggy and we should call of the dogs until the first defect was resolved. The first defect's resolution might solve all the other problems, or it might create new problems, but in any case the defect clustering phenomenon is something that should actually cause less testing (temporarily) in that area rather than more which is what my first instinct was.

8. ## Re: Defect clustering

peanut clustering - The tendancy of peanuts to cluster together on the vine.

peanut density - How much peanut is actually inside the shell. The physical density of the peanut to the space provided the peanut.

9. ## Re: Defect clustering

[ QUOTE ]
The developer's point was that, after a single bug was found in that area, that area was known to be buggy and we should call of the dogs until the first defect was resolved. The first defect's resolution might solve all the other problems, or it might create new problems, but in any case the defect clustering phenomenon is something that should actually cause less testing (temporarily) in that area rather than more which is what my first instinct was.

[/ QUOTE ]

See! This is why we don't let developers test!

I don't think of defect clustering that way, necessarily. Sure, you might find related issues. If there was a drop down with values 1 through 99 in it, and we select a value and it raises an exception, then we try another, and it raises an exception, then we've got 99 exceptions right there that I'm not going to bother testing. Likewise, if I then attempt to submit the parent form and get an exception, well that's a different story. I can use my logical reasoning to assume it's because of the handling of the drop down.

However, with that said, I view defect clustering as a way for black box testers to gain a better understanding of the application and the underlying code. Can we assume that in more complex areas there will be more bugs? Sure we can. Does that mean that no defects will be found in areas containing simple code? Not at all. However, if we understand the complexities of the application we can judge, based on the bug, where we might expect to see more errors within an application.

If it is completely unseable then, sure, leave it. There's not much else you CAN do. However, if you leave that area hoping that the fix will fix all bugs in that area, you can end up boxcaring your bugs (I just trademarked that term), or finding one, waiting for a fix, finding another, waiting for a fix, finding another, meaning you are stretching out your schedule.

I think that there is a point to what the developer says, but I also don't think that ignoring the area is the answer. Plus, if you just log the bugs as you find them, you may actually help to expedite the bug fix or the developer diagnostics process.

10. ## Re: Defect clustering

It could also be a Buggy Developer... :-)

(OOhhh - No - All developers are perfect writers)

Any how, I might reduce some depth in the tests of this area, but surely give it a feature wide test, to raise all main issues - at least.

And a Buggy area, will surly require one or more sets of specially heavy Regression tests sessions.

Page 1 of 3 123 Last

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•

vBulletin Optimisation provided by vB Optimise v2.6.0 Beta 4 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.