Home » RDBMS Server » Performance Tuning » Industry standard best practice for Primary keys..should they be meaningful or meaningless?
Industry standard best practice for Primary keys..should they be meaningful or meaningless? [message #215681] Tue, 23 January 2007 08:38 Go to next message
orausern
Messages: 826
Registered: December 2005
Senior Member
Hello,

I was a little surprised by the debate on the above at:
http://jonathanlewis.wordpress.com/2006/12/29/meaningless-keys/

I remember having read somewhere on tomkyte that surrogate keys are better....can you please help summarise all the advantage/disadvantage of each approach...I still feel meaningless keys are better, but shall be thankful for your sharing of views..
nirav
Re: Industry standard best practice for Primary keys..should they be meaningful or meaningless? [message #215741 is a reply to message #215681] Tue, 23 January 2007 13:35 Go to previous messageGo to next message
smartin
Messages: 1803
Registered: March 2005
Location: Jacksonville, Florida
Senior Member
It is a somewhat contraversial topic with wide-ranging debate and opinion. Everyone has their own opinion. Pros and cons have been listed in many forums, including a long one I remember on Howard Rogers' site.

To me, the primary (no pun intended) advantage to "meaningless" keys is that it isolates your design from the data/app (and the changes that often happen to those in life). There are several disadvantages.

As far as which is better? It depends.
Re: Industry standard best practice for Primary keys..should they be meaningful or meaningless? [message #216200 is a reply to message #215741] Thu, 25 January 2007 09:55 Go to previous message
orausern
Messages: 826
Registered: December 2005
Senior Member
Thank you very much for sharing your views! I think I have to reexamine my understanding...
Previous Topic: users not logged on last 30 days
Next Topic: Adding ORDER BY increases parse time dramatically
Goto Forum:
  


Current Time: Thu May 16 02:35:45 CDT 2024