Jump to content

Talk:System testing

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Limitations

[edit]

What are the Limitations of system testing??

Well, the way I see it, system testing and functional testing are really two different things, and the article should from a neutrality point of view at least make the statement that the wording around testing is not very fixed as of now. System testing to me occurs at the system level, testing functional, as well as non-functional aspects of the software. Tprosser 12:19, 15 May 2007 (UTC)[reply]
System tests the whole system, functionally and non-functionally. But functional testing can apply to anything from a whole system, down to an individual code class. So system testing is about scope, and functional testing is more of a design technique. DRogers 19:54, 15 May 2007 (UTC)[reply]

Redirect from Functional Testing

[edit]

I disagree with the redirect from Functional Testing to this page about System Testing. They are two different things. Functional Testing is all about systematic test-generation from specifications, and applies to units and systems; and contrasts with Structural Testing, which is code-based. System Testing applies to any large-scale integration testing, using any approach which observes how the various units interact. AJHSimons (talk) 16:01, 1 April 2008

  • Okay, the book "Testing Computer Software" has the following definition:
"Functional testing is a type of black box testing. Functions are tested by feeding them input and examining the output. Internal program structure is rarely considered." Kaner, Falk, Nguyen. Testing Computer Software. Wiley Computer Publishing, 1999, p. 42. ISBN 0-471-35846-0.
They further differentiate between functional testing and system testing by saying functional testing "*verif[ies]* a program by checking it against ... design document(s) or specification(s)", while system testing "*validate[s]* a program by checking it against the published user or system requirements".
Based on this I believe that functional testing and system testing are two different things, and should have their own articles. Kaner, Falk, and Nguyen seem to be saying that functional testing checks whether parts of the program behave correctly, while system testing operates on the entire structure, and determines if the thing does what people want. This article event currently begins with "System testing ... is testing conducted on a complete, integrated system to evaluate the system's compliance with its specified requirements", so I think that agrees.
Furthermore, splitting up functional and system testing would make the wikipedia testing articles consistent - Black-box testing currently reads "This method of test can be applied to all levels of software testing: unit, integration, functional, system and acceptance", thereby indicating its belief that function and system testing are two different things.
So, I am going to be bold and move Functional testing back to its own article. There are some things on the talk page there that also need to be cleared up, but let's move one step at a time. Please post here if you agree or disagree. --Culix (talk) 05:21, 1 December 2010 (UTC)[reply]

Types of System Level Test

[edit]

I disagree that "Exploratory Testing" is a type of "System Level Testing". I think Exploratory Testing is in a category all on its own. The purpose of System Level Testing is to verify some "requirement". It is to do Verification and Validation. — Preceding unsigned comment added by Lordneeko (talkcontribs) 18:32, 11 July 2014 (UTC)[reply]

Here are some suggestions we could add to a section describing Types of System Level Test:

  • Parallel testing: the process of feeding data into two systems—the modified system and an alternate system—and computing the results in parallel. In this approach, the old and new systems operate concurrently for a period of time and perform the same processing functions.
  • Pilot testing: takes place first at one location and is then extended to other locations. The purpose is to see whether the new system operates satisfactorily in one place before implementing it at other locations.
  • Sociability testing: ensures that a new or modified system can operate in its target environment without adversely impacting existing systems.
  • System integration testing: evaluates the connection of two or more components that pass information from one area to another. The objective is to take unit-tested modules and build an integrated structure. The Wiki article states that this occurs after integration of units, which by definition is a system even if it is a subsystem or incomplete system.

I need to find a good reference for these terms. The definitions here are from an ISACA CISA exam study guide.Stephen Charles Thompson (talk) 21:44, 21 October 2018 (UTC)[reply]

Removing stress testing and load testing from "Types of tests to include in system testing" suggestion

[edit]

Hello, Wikipedia users, I suggest to remove stress testing and Load testing from "Types of tests to include in system testing" section on the page because they are already part of software performance testing (linked there too) so to me, it is a duplicity and confuses the readers.

Thanks Walter Görlitz for reverting my change, it was actually not intended as Test edit but real edit to improve quality of the page, but as it is my very first edit of Wiki page, I am not sure if I did it right. Originally, I made an edit without leaving edit comment, than I realized that I cannot add edit comment later so I have Undoed my original edit and Re-added the same change, this time with a comment. Please feel free to fix whatever I may have broken unwillingly ;)

(Walter wrote on my page: "Hello, I'm Walter Görlitz. An edit that you recently made to System testing seemed to be a test and has been removed. If you want more practice editing, please use the sandbox. If you think I made a mistake, or if you have any questions, you can leave me a message on my talk page. Thanks! Walter Görlitz (talk) 14:07, 13 June 2016 (UTC)" ) — Preceding unsigned comment added by JetyCZ (talkcontribs) 14:22, 13 June 2016 (UTC)[reply]

Why do you think that there are separate articles for those topics? Walter Görlitz (talk) 14:24, 13 June 2016 (UTC)[reply]
@JetyCZ: Dear fellow editor,
Thank you for your interest in the present article. As a reader, it is clear to me that 'Stress testing', 'Load testing' and 'Performance testing' are different test methods, each of which requires specific approaches. 'Software performance testing' is a variant of 'Performance testing' applied to the specific component of 'Software'. The other component is obviously 'Hardware' which, when combined with 'Software', becomes a complete, end-to-end 'System' (one 'end' being a network terminal user, and the other 'end' being a data record on disk, for example). To be complete (and a little pedantic ), one could argue that our encyclopedia should also feature an article on Hardware performance testing, although the performance of hardware components is often measured as part of 'System testing' anyway.
Please forgive me for over-simplifying the above somewhat, but I daresay an article on the all-encompassing 'System testing' ought to include a summary overview of all potential system 'components' (all the 'hardware' and 'software' parts put together) and also of the main test 'methods'; the latter include 'Load testing' ("Will the system and/or component absorb the increased load generated during incremental repeats of this test?"), 'Performance testing' ("What are the tuning parameters required (in various components) to deliver the greater number of transactions per minute?") and 'Stress testing' ("Under what load conditions does the system and/or component(s) break or stop operating properly?"). An article's See also section is one of Wikipedia's ways of providing this 'summary' function, with wikilinks pointing to the relevant articles.
So, in conclusion, I would suggest that the present article is complete as it includes a mention of these various test methods, and separate articles are warranted on all of these also, where each topic can be expanded in greater detail. I hope the above helped in giving you another viewpoint about the current version of the article. Thank you once again for your interest in the present article, and for opening the present discussion on its talk page.
With kind regards;
Patrick. ツ Pdebee.(talk)(guestbook) 15:34, 13 June 2016 (UTC)[reply]
Hi, thank you Patrick for your complex explanation, it makes sense and I am happy now to leave the page as it is. Walter, I think we misunderstood a bit each other. I agree that Load and Stress testing are two different testing types, I actually did not want to merge them or delete one of them, but only I wanted only remove the links from System testing page (kind of to keep the tree of Top to bottom division of testing types without duplicites). Hope it makes sense to you. Anyway, I am impressed by the speed and quality of peer review process of my edit, glad that collaborative effort works so nice :) thanks to you both. — Preceding unsigned comment added by JetyCZ (talkcontribs) 21:28, 13 June 2016 (UTC)[reply]
We obviously need to establish what the intended scope and focus of this article is. See Talk:System_testing#What_is_the_intended_scope_of_this_article?Stephen Charles Thompson (talk) 22:19, 19 October 2018 (UTC)[reply]

Difference between this and System Integration Testing

[edit]

What's the difference between this and System integration testing? It feels like these two pages should be combined. kriskhaira (talk) 07:18, 13 February 2017 (UTC)[reply]

The other article discusses a subject of testing that is at a lightly different level than. System integration testing is concerned with only where two system components integrate while system testing is focused on all functional elements of a system. Walter Görlitz (talk) 07:52, 13 February 2017 (UTC)[reply]
Judging by semantics alone, System integration testing sounds like a form or type of System testing. If this article is supposed to be a summary intro to system testing in general it should reference and link to the more detailed sub-topic article Stephen Charles Thompson (talk) 22:17, 19 October 2018 (UTC)[reply]
It sounds to me like system testing is performed on subsystems which are then combined and undergo system integration testing. But clearly this applies only to very complex systems, and I don't think this terminology is universal. I'd call these different levels of integration testing. Beland (talk) 05:38, 20 October 2018 (UTC)[reply]

Incorrect definition in summary?

[edit]

I believe this definition is inaccurate, "System testing falls within the scope of black-box testing, and as such, should require no knowledge of the inner design of the code or logic." I believe (from my experience) the reverse is true, that black box testing is one form of system testing. The cited source of this definition is obsolete. Will someone with expertise or access to authoritative standards please verify or update the summary definition? Stephen Charles Thompson (talk) 22:06, 19 October 2018 (UTC)[reply]

System testing is not necessarily black box testing ... although in practice it usually is done that way. ... There are a zillion ways to categorize testing and most are independent of each other. A particular test can fall into multiple categories. Stevebroshar (talk) 04:28, 27 April 2024 (UTC)[reply]

Black box testing can be done at any level, including unit, integration, and system testing. It simply requires testing to specification without knowing implementation details. Beland (talk) 05:31, 20 October 2018 (UTC)[reply]

Many web sources seem to think system testing is generally done in a black box fashion, but I doubt that's universally so. I'd say neither is a type of the other n but that black box techniques are typically used when system testing. Beland (talk) 05:33, 20 October 2018 (UTC)[reply]

What is the intended scope of this article?

[edit]

I believe that mistakes in the summary definition, lack of cited current and authoritative sources, and haphazard inclusion of a few certain kinds of testing have cast doubt on the intended focus / scope of this article. Stephen Charles Thompson (talk) 22:33, 19 October 2018 (UTC)[reply]

What is the intended scope of this article?

  • Is it intended to be a summary introduction to all types of system testing in general, regardless of (non-specific to) the type of system?
  • Is this article intended to be a taxonomic top-level or root of all types of system testing?
  • Would this article have subsections with introductions and links to articles describing testing terms that are specific to a certain kind of system (example, Software testing)?
  • Would this article have subsections with introductions and links to articles describing specific kinds of tests (example, Destructive testing)?

I hope so, because that would be a good starting point for anyone exploring all the various kinds of system testing methods. What do you think? Stephen Charles Thompson (talk)

Those all sound like useful subtopics of system testing which can or could be summarized in or linked to from this overview article. -- Beland (talk) 05:42, 20 October 2018 (UTC)[reply]

I went ahead and re-focused the article from being exclusively about software to a general overview of system testing concepts. All of content describing software testing that was removed was already in the well-formed main article Software testing. Stephen Charles Thompson (talk) 19:54, 20 October 2018 (UTC)[reply]

I think that system testing implies testing a software system. Or if the term also applies to non-software systems (which may be true since my expertise is software) then there is a meaning to system testing that is particular to software system testing, and that a non-software system meaning might be significantly different. ... thinking of the destructive testing section: that clearly does not apply to software. hmmm Stevebroshar (talk) 04:33, 27 April 2024 (UTC)[reply]