Sparx Systems Forum
Enterprise Architect => General Board => Topic started by: EANewbie on September 28, 2010, 07:25:40 am
-
EA Experts
I am constructing a Deployment Diagram using EA. I have created a server XYZ with a specific set of attributes, including RAM, OS, etc. These attributes are visble when constructing an instance of this in EA. Next, I have created a WebLogic Server artifact that represents my WebLogic Server container and includes attributes, such as Version. I have created a Component that represent an application EAR (ABC) that I deploy into that WebLogic Server component.
I would like to create an instance of the server XYZ, the WebLogic container, and the Component and call this combination my App Server Gold Image. I would then like to use that gold image in various flavors of my deployment diagrams. I would like to be able to set it up once where the application EAR is contained in the WLS instance, which is contained in the server instance -- with all the attributes of each displayed. I have constructed the gold image, have all attributes of each element displayed, and everything sized nicely.
So, here is the problem. When I try to use this gold image on another diagram or in another package by create a direct link to this Gold Image or copy-and-pasting into package, the Web Logic container and the Application EAR are not visible in the diagram. Instead, I have to re-drag the sub-items (WebLogic and App EAR) to the diagram. If I have a single deployment diagram that contains 50 instances of this gold image, I have to do this 50 times. What's worse? If I change one of the attributes, it causes a resize to be necessary -- you guessed it -- on each of those 50 instances.
How can this be avoided so that I can get real re-use out of the gold image?
Your help is much appreciated.
-
Have a look at UML Patterns in the help file - they may be of some use.
-
Thanks for the feedback. This has been helpful in progressing my understanding of EA.
UML Patterns looked promising initially, but here is the problem I am encountering when attempting to leverage them in this scenario.
I have created my pattern two ways.
First, I created it based on instances of elements I had defined. When creating the UML Pattern with instances, the "Preview" of the pattern looks perfect, but when I add the pattern to a diagram, it loses the names of the objects (server node, artifact, and component) and it loses the attributes of those objects. Again, the "Preview" shows the names and attributes perfectly -- just as I would like to see them.
Second, I created it using Direct Links from each of the objects. Specifically, I constructed a diagram using direct links and then saved that as a UML Pattern. When dragging that UML Pattern onto a diagram, it looks perfect -- names and attributes in-tact. However, if I then alter the pattern, the changes in the pattern do not get applied to the places where I used the pattern.
The second attempt is a solid step forward, but without being able to alter the pattern and see those changes applied (or at least asked if I want to apply them), I am back at square one.
Is there a way around this?
If not, I guess I just need my pattern to be in a final state so that revisions are no longer an issue.