Saturday, February 18, 2012

The State of Data Modeling Tools (Conclusion)

After reviewing a bunch of the data modeling tools out there, I've come to the conclusion that there are some great tools out there, some workable tools and some tools to avoid.

Data Modeling Tools Result Summary
  • CA ERwin 
    • This is a toolset that has been in the decline for while. CA is trying to pull it out of the nose-dive it's in but it has a legacy code-base without a lot of the original team around to explain it. As a long term system I can't recommend it. It will continue to go downhill.
  • Visio 2010 Professional 
    • Fine for small models but it will quickly become to unwieldy with a significant model. Microsoft doesn't seem interested in this aspect of the tool market and are not improving Visio for it.
  • MySQL Workbench
    • Stability is important. It's free but my time is not. I don't recommend this tool.
  • SQL Power Architect (Community Edition)
    • This tool needs better integration with the actual database. It has the most potential of any open source modeling tool I've seen.
  • Embarcadero's ER/Studio Data Architect (Trial)
    • This tool works flawlessly. If you can afford it, buy it. They've been doing this for over 15 years and have built an entire product line around modeling tools.
  • Toad Data Modeler (Trial)
    • I admit that the lack of stability irritated me greatly but I don't think that this product has improved much over the last few years and it may be on life support from Quest. Not recommended. 
  • ModelRight Professional (Trial)
    • For the price (roughly 25% of ER/Studio) this tool works amazingly well. The only thing it lacks is logical modeling, if you have one platform (like most folks) this is a great choice.
I hope that some of you found my research helpful. Your great comments help me remember to post.  :-)

1 comment:

  1. Thanks for the comment Michael!

    I haven't used Enterprise Architect in a while, and I agree, it is a terrifically solid tool for UML documentation. The only problem was that it lacked a way to forward-engineer the data model changes. Part of my goals were to avoid the extra book-keeping of managing the schema and documentation independently.