Thanks:  0
Likes:  0
Dislikes:  0

1. ## monkey testing

plz explain me what is monkey testing

2. ## Re: monkey testing

It means "testing" a system/program by having as many people (not necessarily skilled at testing) as possible use the application under test, hoping to find defects by sheer numbers and random chance. The term is derived from a popular example for explaining the concept of infinity, namely that if you give an infinite number of monkeys an ifinite number of typewriters, eventually they will write all the works of Shakespeare.

3. ## Re: monkey testing

With that said, I think some clarification is in order. What "monkey testing" really means is not the number of people, but rather essentially a "random typing" or "random action" sort of thing. (That is more in line with the example with the infinite amount of monkeys randomly typing away. However, it can be one monkey typing away for an infinite amount of time.) The idea here is that you are testing a system or an application "on the fly", as it were. (Others call this "random walk testing". Yet others add a little more structure to it and call it "ad hoc testing".)

There is also "smart monkey testing", where test inputs are generated from probability distributions that reflect actual expected usage statistics. This could be from user surrogates, user profiles, operational profiles, use cases, etc. There are different levels of "IQ" in smart monkey testing. With the "simplest" IQ level, each input is considered independent of the other inputs. That means that a given test requires what they call an input vector with five components. In the "low" IQ level, these inputs would essentially be generated independently. Then at the "high" IQ level, the correlation (e.g., the covariance) between these input distributions is taken into account.

The idea in all forms of monkey testing, as it is actually defined, is to randomly cover the input space of the unit or application under test. What "randomly" means is where the differences crop up in terms of what monkey testing means.

4. ## Re: monkey testing

My point of view on monkey testing is that - If one is testing the software without any plan and just hitting every button on the software screen or making inputs without knowing what actually input fields are intended to accept.
Like if you are not aware about the functionality of the software but you still updating the backend database(Inserting, deleting, updating records etc.) by inputting random input values... so in this you are not aware whether actual data is updated or not... so the hidden faults like these are not found...
In some articles I read that this is also mentioned as "Ad hoc testing".

Devender

5. ## Re: monkey testing

Originally posted by Jeff Nyman:
<font size="2" face="Verdana, Arial, Helvetica">Jeff,

Are you and Noel related?

6. ## Re: monkey testing

Here's an example of a semi-Smart Monkey test I am running even as we speak.

A part of my application-under-test produces industry summary reports. These reports are selected via a series of checkboxes over 4 web pages.

There are far too many checkboxes to do a comprehensive test of the reports, and the checkboxes themselves can change each month when the data is updated.

We are in the process of changing the web servers from IIS to Apache, and the processes behind these industry reports have been rewritten. The processes appear somewhat brittle now, so I needed to test them in many ways.

One of the tests I am running is an automated script which randomly selects some of the checkboxes on each of the pages and produces the reports. It then loops back and starts the process again.

If everything was working correctly, this script could run indefinately - producing a wide variety of reports.

With the current build, we have a bug. I occasionally see that the server produces only part of the expected report and simply omits the rest.

While I may have been able to detect this bug with manual testing, it would have been difficult to do so, and even more difficult to reproduce. I could easily have run 100 manual reports and not found any problem (even though the problem exists).

On the other hand, it was a simple matter to write this script, launch this Monkey Test on an unused machine or two and let it run until it found something unexpected.

7. ## Re: monkey testing

The above discussion shows why we really need to know in what context the original question was being asked. If the term was being used by a knowledgeable QA/QC person, likely Jeff's/Joe's more technical definitions apply. If it was being used by a "pointy haired boss" type (a.k.a. management), it likely was meant more along the line of getting a bunch of people to pound on the application and see what happens.

8. ## Re: monkey testing

In the businesses in which I've worked, monkey testing has unilaterally meant "banging on the keyboard like an ignorant user (monkey)" and it is not a complimentary term - in fact, it's quite insulting. Those projects I've worked on that collected basic metrics have shown such testing drives out no more than 10% of the total error; often less. Most user organizations I've worked with put more thought into their testing, but regardless, I do not use this term with clients.

- Linda

- Linda

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•
Search Engine Optimisation provided by DragonByte SEO v2.0.36 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.