At 15:41 +0100 6/21/02, Bill Ferrett wrote:
>I would like to have a MySql table where each row has a unique id (using
>auto-increment) but is clustered, i.e. physically stored, in a different
>sequence, e.g. name. This is so that an sql query to order by namew will be
>I don't want the auto-increment id to be duplicated - I believe this happens
>when the auto_increment column is not the first column - so a
>of name,id seems to be out.
>Can I do this and if so how?
You have no say in how the server stores rows in tables. You can control
the order in which it retrieves by, by using an ORDER BY clause. For
fast lookups on the name column, index it.
>Accroserve Systems Ltd.
>Office: 01702 342148
>Mobile: 07855 868565