Question Driven to Power Designer Support Team

[Former Member]

To Power Designer Support Team.

Greetings!

I have a simple question about Power Designer Concept and Premise. I work on SQLTech consultancy, and we are Power Designer users (data administrator). We support a credit card processing company which is developing a huge project.

Scenario

We have a data model family which is composed by five (5) data models. It is usually often to define relationships between data objects of diffetent data models. So far so good.

We finished the release 1 of this project and we are now working on release 2. Recently, the data administrator that is employed by the client , and which also leads the data administrators of our consultancy, decided to keep the data models family of RELEASE 1 stored under the same ROOT FOLDER where we are also storing the data models family of RELEASE 2.

Since then, we are facing different kind o problems while working on the models. The problems vary from replication of data objects such as constraints and atributes, to issues related to loss of metadata.

Considering this short explanation, I firstly need to ask:

Is it correct to keep replicated copies of the same data model under the same ROOT FOLDER?!?

We did not have such problems before working like that!

Waiting for you answer!

Thanks in Advance.

Best Regards!

Luciano Pontes.

[GeorgeMcGeachie · Accepted solution]

Hello Luciano, welcome to the community 🙂

It sounds like you're using a configuration management tool to manage your model versions, rather than the PowerDesigner repository. If you were to use the PowerDesigner repository, your versioning requirements could be handled by using Branching. The issues that you're facing are a feature of the way that configuration management tool works - have you checked in the models as files, or in some other way?

Perhaps I've misinterpreted what you meant, and you ARE using the repository, in which case I have some questions for you:

• how did you 'replicate' the data models?
• when you talk about 'replication of data objects such as constraints and attributes' - are you using Replications in your PowerDesigner models, or are you getting model artefacts crossing from one model to another unexpectedly?
• what metadata are you losing?

[GeorgeMcGeachie]

Hello Luciano

You shouldn't be able to check a model (even two versions of the same model) into two different folders in the repository, unless you created the second set of models using the 'Save As New Model' option on the File menu. It looks like somehow you were able to check the models in to two places, which could account for the confusion. When you check in a model, PowerDesigner will look up the model's GUID in the repository to see if it's already there; once it finds the latest version of the model in the repository, all changes are applied to that model version (or a new one, depending on the current status of the model version and the 'freeze before' check in option). If there really are two instance of a model in your repository, perhaps some changes go to one instance (Release 1) and others go to the other instance (Release 2).

I can't think how this could have come about, unless the people checking in models are not able to even see the models for the other Release.

It's worth raising this as a support case with SAP.

Without actually seeing what you have in your repository, it'd be difficult to work out the exact steps needed to sort it out.

[Former Member]

Greetings George!

Thank you so much for your quick and kind answer!

We are currently using the repository on our scenario.

In summary:

1. We finished "release 1" phase of our project at the very beginning of this year; we are now working on release 2 phase;
2. We needed to keep two groups of the same models (4) within the repository; one group of models related to release 1 and another group related to release 2;
3. When release 2 has begun, the responsible for organizing the family models in the repository, defined two folders under the same ROOT FOLDER and, since then we have had the issues I tried to explain. Among the problem I can mention: We lost data objects; we lost data objects metadata and we also had data objects replication (atributes and constraints for instance);
4. From yesterday to now I had some progress trying to undertand power designer functional premise, and I am pretty convinced that using the "BRANCH" resource may bring us the proper solution, once those model families will be isolated from each other, avoiding then the conflicts we have been facing.

Did I get to explain any better?

I wanna thank you again for your time and answer, and I take this opportunity to congratulate you for you professional history which I read on linkedin.

You are also working on a very interesting company!

Best Regards!

Luciano Pontes.