View Single Post
Old 07-25-2014, 06:28 AM   #17
jdraw
Super Moderator
 
jdraw's Avatar
 
Join Date: Jan 2006
Location: Ottawa, Ontario, Canada;West Palm Beach, Florida
Posts: 12,151
Thanks: 83
Thanked 2,011 Times in 1,959 Posts
jdraw is a glorious beacon of light jdraw is a glorious beacon of light jdraw is a glorious beacon of light jdraw is a glorious beacon of light jdraw is a glorious beacon of light
Re: How to normalize and optimize this database?

Since you are in design mode, and I am supposing you have some test data or at least some specifications/requirements from the Engineers, I recommend you work with the latest version of your evolving model and vet the model with your test data. We used to play what we called "stump the model". This involved people working with the design, users and support personnel. A picture of the model was on display; business rules were "actioned" against the model; "What-if scenarios" from users and support personnel were also 'tested' against the model. This was a test of the model, not the modeller. If any scenario could not be proven/answered with the model, things stopped and a reconciliation/rationalization occurred. This included identifying WHAT was "wrong" - was it the model, was it a business rule, was there a hidden entity/attribute/relationship, was the scenario itself not logical for the business; was it bad test data? This had to be resolved. This was repeated- often involving a few different representatives from the various participant groups. A revised model, or new data etc would be prepared and another "stump the model" session would take place. All anomalies were reconciled. At some point there would be agreement/acceptance of design. To make this work efficiently, you should have preliminary specifications and any modifications;definitions of tables and fields; keep samples of model versions (to show changes), a current list of business rules (with any modifications) and a summary of the "stump the model" sessions.
It is interesting to see what happens in such sessions. Users who know the process intimately are often surprised with what happens next or how X relates to Y-- you get the"I've been doing this for 5 years and I didn't know it did that"; or "I wasn't aware that they used that data I thought it was local to our group.."

You and the engineers etc know the situation better than any reader here. Readers will help as best they can if you post specific questions.

Good luck.
__________________

To view links or images in signatures your post count must be 10 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 10 or greater. You currently have 0 posts.
jdraw is offline   Reply With Quote
The Following 4 Users Say Thank You to jdraw For This Useful Post:
EternalMyrtle (03-17-2015), pheidlauf (07-25-2014), Riter19 (04-28-2016), spikepl (07-26-2014)