|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
 |
|
Use cases and MQ |
« View previous topic :: View next topic » |
Author |
Message
|
twegmann |
Posted: Mon Jun 17, 2002 6:06 am Post subject: Use cases and MQ |
|
|
 Novice
Joined: 09 Aug 2001 Posts: 23 Location: New Jersey
|
Is anyone using use cases as a means to document integration requirements for MQ ? If so, would you be willing to share some examples ? I am a novice when it comes to use cases and am strugglinbg at bit with how to set them up in an MQ scenario.
An example of my struggles is what to define as the actors. Should I use the source & target applications, the 'put' and 'get' adapters, or even the source & destination queues involved ?
Any advice is most appreciated. |
|
Back to top |
|
 |
bduncan |
Posted: Wed Jun 19, 2002 9:49 am Post subject: |
|
|
Padawan
Joined: 11 Apr 2001 Posts: 1554 Location: Silicon Valley
|
I'm sorry, I don't have any examples readily available, but I've some experience with use cases involving MQSeries. Typically the actors are considered to be the applications that are doing the putting and getting. This works fine for the standard request/reply paradigm. However, if you are talking about more complex flows, where a single message may pass through the hands of lots of applications, it gets more complicated. _________________ Brandon Duncan
IBM Certified MQSeries Specialist
MQSeries.net forum moderator |
|
Back to top |
|
 |
twegmann |
Posted: Wed Jun 19, 2002 10:16 am Post subject: |
|
|
 Novice
Joined: 09 Aug 2001 Posts: 23 Location: New Jersey
|
Brandon - thanks for your help.
From an overall perspective, how did it work out ?
Did you find use cases was an effective way to approach defining integration requirements and did it help the design\construction efforts in MQ ?
Would you choose to go the use case route on future projects ? |
|
Back to top |
|
 |
bduncan |
Posted: Wed Jun 19, 2002 2:25 pm Post subject: |
|
|
Padawan
Joined: 11 Apr 2001 Posts: 1554 Location: Silicon Valley
|
This may not be the case in your firm, or anyone else's for that matter, but at least in the last few companies I worked for, use cases proved helpful for a very important reason. Typically our application designers were unfamiliar with the middleware messaging world. This makes sense, because we hired them becuase of their skills in writing a front-end application, or something which processes a chunk of data. Where that data comes from or where it goes wasn't something they focused on. That was left to other people, like myself, who administered our messaging system. Because the architects of this system were detached from those who were making use of it, it was important to communicate to the application programmers exactly what was expected when they bundled a bunch of data into a message, or what they should expect when receiving a message from another part of the system. Rather than sit down and teach all the application programmers how logical groups, syncpoint, persistence, clustering, etc., worked, we provided them with use cases which detailed exactly how their message should look when the MQPUT it, or exactly what they would receive after an MQGET. This solved quite a bit of headache, because early on, we didn't have a rigid structure, and we had people going in all sorts of different directions, because they didn't understand middleware messaging.
That being said, I would recommend using use cases in situations like the one described above. However, it the bulk of the applications you write are concerned with MQSeries operations (rather than data processing, etc.,) and the people writing these applications are quite familiar with messaging, then the time and effort taken to create the use cases might not be justified... _________________ Brandon Duncan
IBM Certified MQSeries Specialist
MQSeries.net forum moderator |
|
Back to top |
|
 |
Wjbyczek |
Posted: Fri Nov 22, 2002 5:48 pm Post subject: |
|
|
 Newbie
Joined: 08 May 2002 Posts: 4 Location: Domain Systemix
|
Since most MQ interactions are more or less transactional between application systems, creation of use cases is something of an overkill and usually not worth the effort involved. A data flow and process diagram should fill most all the needs from a design standpoint. Where I've found use cases to be the most valuable is when you're combining MQ with an MQSI environment with an operational data store (e.g. DB2 warehouse manager) at the core of the system. In this sort of environment, apps send to the ODS and that becomes the data base of record for the whole environment. That's the sort of scenario where use cases are of the greatest value- especially when augmented by sequence models. I guess what I'm saying is that if your goal is to centralize information, then the use case and object model is of great value. If your systems are inter-application transactional without any form of ODS, then it's pretty much a goldplating effort. |
|
Back to top |
|
 |
fschofer |
Posted: Fri Nov 29, 2002 4:31 am Post subject: |
|
|
 Knight
Joined: 02 Jul 2001 Posts: 524 Location: Mainz, Germany
|
|
Back to top |
|
 |
|
|
 |
|
Page 1 of 1 |
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|
|
|