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 - 2001-09-19 : 10:13:18
|
| /nev/dull writes "i've read your part i article (with great interest) and look forward to part ii regarding unique identifiers. however, i cannot wait much longer, hence my question is this: i have to store a lot of data that is not unique yet must be individually referenced/identified. i do not want to use an "auto id" field. primarily, because i require the unique id for refernece in other dependant tables immediately. due to normalization and efficiency i populate other detail tables at the same time as my initial insert into the master table. i don't want to rely on server specific functions to insert my record then get the newly created id. are you aware of any efficient unique number generation algorithms, surely i cannot be alone in this endeavor? i am aware of guids but they seem to be a bit on the non-efficient side. besides there is some question as to weather they are "guaranteed" to be unique or not. i potentially will be inserting thousands of records a second...what do people use when they want to be ansi compliant, somewhat database neutral and cannot easily define a composite unique key without the use of some sort of database auto generation field?btw i'm on sql 6.5 ;-)" |
|
|
|
|
|
application generated unique identifiers