 |
 |
Archives of the TeradataForum
Message Posted: Fri, 11 Apr 2003 @ 18:16:21 GMT
Subj: | | Re: How to implement the sequence concept in Teradata |
|
From: | | Ganga Palakattil |
Sequence Number is available with V2R5.
It is very useful in some cases.
For example, if you need to keep a log table (process log), it is a good idea to generate an automatic sequence number to identify the
process. Latest run of the process will have the maximum value for the Sequence Number. Form a visual perspective, it is very easy to
distinguish from a timestamp. Though you can always create one programmatically, it is much more convenient to depend on a system generated
one. That is the reason Oracle/UDB(DB2)/SQL Server are providing it. In very rare situations, if you cannot guarantee uniqueness of a row
and still you need one, this is the most convenient one from my experience with Oracle/DB2/SQl Server
-Ganga Palakattil
| |