| Author |
Message
|
| kishoreraju |
 Posted: Thu Jun 23, 2005 10:26 am Post subject: testing consideration |
 |
|
Disciple
Joined: 30 Sep 2004
Posts: 156
|
Hai all
I am working as a QA on a project.can any one tell what are the main things that take into conseration while doing testing |
|
| Back to top |
|
 |
| jefflowrey |
| Posted: Thu Jun 23, 2005 10:30 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
I think the main things to take into consideration when testing are success and failure.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| zpat |
| Posted: Thu Jun 23, 2005 11:20 am Post subject: |
|
|
Jedi Council
Joined: 19 May 2001
Posts: 5866
Location: UK
|
Test message data conversion across platforms.
Test application reconnection if queue manager stopped and started.
Test correct UOW syncpoint logic (commit/rollback) if app fails.
Test reconnect of MQ client if channel stopped (pull network cable).
Test effect of MQ HA cluster failover on applications.
Test you don't have any single point of failure.
Test scalability with large message volumes.
Test response time for request/reply messages at max rate.
etc etc I could go on for hours. |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Thu Jun 23, 2005 11:25 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
All of those tests, zpat, only matter and should only be run if the business requirements need them to be.
There is no good reason to test something you don't care about.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| zpat |
| Posted: Thu Jun 23, 2005 12:31 pm Post subject: |
|
|
Jedi Council
Joined: 19 May 2001
Posts: 5866
Location: UK
|
Non-functional requirements are rarely articulated clearly by business people. But I have yet to find a system without any.
The same business people are the first to ring up when the system does down and demand to know why. One has to guide them in the area of NFRs - that's what IT architects are here for.
It's also my belief that it is as easy to develop good code as it is to develop poor code, so many of the above comes for free by having appropriate standards.
If you are using MQ, you don't want to undermine it's strengths - or you may as well not use it and find something free on the internet instead.
Why have a go at me? I was only being helpful - unlike some on here. |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Thu Jun 23, 2005 1:00 pm Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
| zpat wrote: |
| Why have a go at me? I was only being helpful - unlike some on here. |
I wasn't "having a go at you". I was responding with my opinion on things you recommended testing, and redirecting kishoreraju to think about the business requirements of the project as a means of determining what to test and what is important. In my opinion, it's not worth testing NFRs if you don't know what the FRs are, or how to test those.
And if I am unhelpful sometimes, out of choice, I think that I am also extremely helpful sometimes - again out of choice.
No worries.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
|
|
