Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
AskSQLTeam
Ask SQLTeam Question
0 Posts |
 Posted - 2000-11-07 : 11:06:29
|
Olle writes "I am construction a web service which will possibly have thousands of users divided into several subsystems. Each subsystem will allow different kinds of information attached to each user. For example, SystemA might connect information about what cars (with color, type, age ...) a user owns and SystemB perhaps what colors she likes. The problem is that the types of information that in the future might be interesting is not known in advance. I need to have a quite flexible method that is not to complex or slow. As I can see it, there are 3 different solutions to the problem:
1) Create separate tables for each subsystem (like tblCars, tblColors).
2) Create a number of standardized tables that are shared across the subsystems. New tables must be added when needed.
3) Create a table with field type definitions and have one table with all data where each row contains a single field (userid, fieldtype and data/value).
For each of these solutions there will be stored procedures for searching and retrieving data that must not be too slow.
Do you have any alternative options? Which one of these is the best with respect to complexity and speed?
Really hope you can help me with this one!
SQL Server 7.0 + NT 4.0 Server SP6" |
|
|
|
|
|
Flexible database design that is fast and not to complex?