The online community for software testing & quality assurance professionals
 
 
Calendar   Today's Topics
Sponsors:




Lost Password?

Home
BetaSoft
Blogs
Jobs
Training
News
Links
Downloads



Quality Engineering >> Requirements and Design

Pages: 1
ThomasD
Newbie


Reged: 07/25/12
Posts: 7
Fault/Error/Failure terminology question
      #712965 - 07/25/12 01:19 AM

I am trying to understand the following and deleted my previuos question as it was not proper. I would like to understand what is meant by "an error is the part of the total state of the system"?

Failure: A failure is said to occur whenever the external behavior of a system does not conform to that prescribed in the system specification.
Error: An error is a state of the system. In the absence of any corrective action by the system, an error state could lead to a failure which would not be attributed to any event subsequent to the error. For this reason the definition of an error is: the part of the total state of the system that may lead to its subsequent service failure.
Fault: A fault is the adjudged cause of an error.

An example: Requirement: Do not let the customer to withdraw any money if the balance is 100 or less. The actual implementation:

if (Balance<100) //developer did not use =< as he should
{
return false;
}
Also it is clear that there is a fault as the operator < is incorrect. But what will be error (and failure)?


Post Extras: Print Post   Remind Me!   Notify Moderator  
Richard_Hunter
Member


Reged: 08/16/04
Posts: 388
Re: Fault/Error/Failure terminology question [Re: ThomasD]
      #713394 - 07/31/12 05:02 AM

so by the definition provided, the error is the fact that the operator is incorrect. The fault is not clear, but let's assume human error on the coders part. The failure is when that error appears.

So if I try and take out 80, the error is still in the code but I have not seen it and therefore I do not have a failure. I only get to see the error (as a failure) when I try and take out 100.

--------------------
Richard Hunter


Post Extras: Print Post   Remind Me!   Notify Moderator  
ThomasD
Newbie


Reged: 07/25/12
Posts: 7
Re: Fault/Error/Failure terminology question [Re: Richard_Hunter]
      #713814 - 08/06/12 02:25 AM

Richard, no - the fault is the defect in the code (coding error), while error is the state of program caused by that fault - incorrect computation leading to e.g. incorrect values=>incorrect state that can cause failures.

A quote from fundamental literature:
fault in the written software (faulty instruction(s) or data); upon
activation (invoking the component where the fault resides and triggering the faulty instruction,
instruction sequence or data by an appropriate input pattern) the fault becomes active and
produces an error; if and when the error affects the delivered service (in information content
and/or in the timing of delivery), a failure occurs.




Edited by ThomasD (08/06/12 02:34 AM)


Post Extras: Print Post   Remind Me!   Notify Moderator  
Richard_Hunter
Member


Reged: 08/16/04
Posts: 388
Re: Fault/Error/Failure terminology question [Re: ThomasD]
      #713914 - 08/07/12 03:04 AM

ThomasD - what fundamental literature? I have been testing for more years than I care to count and I have yet come across fundamental literature that everybody agrees with and follows.

Case in point the question defines a fault in a different way to your fundamental literature.

Personally splitting hairs between the definition of a fault, error and failure is rather besides the point of testing, but there you go

--------------------
Richard Hunter


Post Extras: Print Post   Remind Me!   Notify Moderator  
Walen
Super Member


Reged: 05/09/01
Posts: 1254
Re: Fault/Error/Failure terminology question [Re: Richard_Hunter]
      #713924 - 08/07/12 05:39 AM

I suspect ThomasD is citing:
http://www.cs.cmu.edu/~garlan/17811/Readings/avizienis01_fund_concp_depend.pdf
or
http://drum.lib.umd.edu/bitstream/1903/6459/1/TR_2004-47.pdf
or
http://alexandria.tue.nl/extra2/201010332.pdf
or possibly
http://books.google.com/books?id=YdUsviY...on"&f=false
which appears to be a citation of the above works

It is generally considered good form, however, to credit the people whose work you are citing - particularly if it is in an academic context, as these three researchers obviously were working.

--------------------
P. Walen

My Blog: http://rhythmoftesting.blogspot.com/


Post Extras: Print Post   Remind Me!   Notify Moderator  
Richard_Hunter
Member


Reged: 08/16/04
Posts: 388
Re: Fault/Error/Failure terminology question [Re: Walen]
      #714121 - 08/09/12 01:47 AM

Good googling Walen, thanks, I will have a read through, but I think my point that it depends what your software house/shop/business calls a fault, error, failure is how you should define it.

(as an aside, as a young man having come back from a course (I think ISEB) I was trying to explain to my Boss the difference between an error, defect etc as they define it. He said 'I don't care if is an error, defect, fault or bug. Fix it', which I thought was a good point :-) )

--------------------
Richard Hunter


Post Extras: Print Post   Remind Me!   Notify Moderator  
ThomasD
Newbie


Reged: 07/25/12
Posts: 7
Re: Fault/Error/Failure terminology question [Re: Richard_Hunter]
      #714233 - 08/10/12 04:46 AM

Richard, being a good tester definitely does not mean to know these definitions Its more about theoretical and academical point of view - these terms are especially used in dependability concept what should be the base line for testing books.

Post Extras: Print Post   Remind Me!   Notify Moderator  
Walen
Super Member


Reged: 05/09/01
Posts: 1254
Re: Fault/Error/Failure terminology question [Re: ThomasD]
      #714260 - 08/10/12 08:36 AM

"should be the base line for testing books"

Interesting. Can you explain why?

--------------------
P. Walen

My Blog: http://rhythmoftesting.blogspot.com/


Post Extras: Print Post   Remind Me!   Notify Moderator  
ThomasD
Newbie


Reged: 07/25/12
Posts: 7
Re: Fault/Error/Failure terminology question [Re: Walen]
      #714272 - 08/10/12 10:05 AM

Walen: Because dependability is the one of the main features of the system and thus definitons used by authors describing that would finally unify these - often significantly - differently defined terms. Also for its analogy with hardware - fault is the actual defect, error is the result of defect, failure is the deviation from specs...

Post Extras: Print Post   Remind Me!   Notify Moderator  
FredMan
Super Member


Reged: 10/29/07
Posts: 1226
Loc: Lenexa, Kansas
Re: Fault/Error/Failure terminology question [Re: ThomasD]
      #714295 - 08/10/12 01:12 PM

I say a bug is a bug.

--------------------
Onward thru the fog...


Post Extras: Print Post   Remind Me!   Notify Moderator  
Walen
Super Member


Reged: 05/09/01
Posts: 1254
Re: Fault/Error/Failure terminology question [Re: ThomasD]
      #714303 - 08/10/12 07:48 PM

So, what is "dependability"? Is it always one of the main features of the system? Is this always true? In every instance?

Sorry - This sounds like an academic series of definitions and explanations. This appears to be a search to define a fundamental root cause for some variance from expectaions.

--------------------
P. Walen

My Blog: http://rhythmoftesting.blogspot.com/


Post Extras: Print Post   Remind Me!   Notify Moderator  
ThomasD
Newbie


Reged: 07/25/12
Posts: 7
Re: Fault/Error/Failure terminology question [Re: Walen]
      #714310 - 08/11/12 01:11 AM

Walen: actually, it is Dependability is how you can rely on the system to work properly. That is something valid for every software system, from games to banking systems. And if there are faults (flaws in the code), there will be errors and possibly failures.

I am sorry I did not want to stir a discussion about these, I am sure that in common use, we do not use these terms as defined but we know what we are speaking about.

PS: Having spoken with a professon in computer science, my original question has a following answer:
The fault in the code is programmer's error, incorrect operator in this case. The error state is the value returned (false/true) based on the input - program should not be in that state when value of Balance is 100. The failure will be immediate if this is the only condition - specified behaviour is different.


Post Extras: Print Post   Remind Me!   Notify Moderator  
Pages: 1



Extra information
0 registered and 2 anonymous users are browsing this forum.

Moderator:  blueinatl, swt88, AJ, Daniel_S 

Print Topic

Forum Permissions
      You cannot start new topics
      You cannot reply to topics
      HTML is disabled
      UBBCode is enabled

Rating:
Topic views: 6106

Rate this topic

Jump to

Contact Us | Privacy statement SQAForums

Powered by UBB.threads™ 6.5.5