Data Modelling Reverse Engineering Initial ER Digram Readability

Hi There,

I have just done a reverse engineer of our system and wondering what techniques people use to re-organise the resultant ERD that has allot of Entitles.

My ERD has approx 400 entities, I have turned off all displaying of information so that I only have Entities and their relationships and I am still overwhelmed.

What techniques do people use to organise the ERD so it is readable and printable.

Regards

Craig

Hi Craig - first suggestion would be to make use of Workspaces. Workspaces allow you to separate your model into subject areas (i.e. entities that relate to a specific function such as order processing or sales) and that makes viewing them easier since there are less entities on the screen.

Let us know how that works for you. If that’s not a workable solution, we can resort to other things.

Thanks for your suggestions Kevin, greatly appreciated.

That is how I have approached it but my concern in doing it this way is that cross subject areas may miss important relationships.

e.g.

One of the highest relationship entities in most business ERD’s I would assume is the People or Person entity. Our Person Entity will live in a works space called MHP but allot of my other workspaces will reference the People table.

We have other entities in the system that are similar to the people table as well.

How do you approach this issue, should I bring the People entity onto each of the workspaces?

regards

Craig

You can put an entity on each and every workspace. It is not limited to one.

You can even put a single entity multiple times on the same workspace.

I typically do that for entities that are referenced by many others to reduce the number of lines that cross each other.