List:General Discussion« Previous MessageNext Message »
From:Paul DuBois Date:June 22 2002 6:05pm
Subject:Re: auto_increment and clustering sequence
View as plain text  
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
>quicker.
>
>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 
>multi-column index
>of name,id seems to be out.
>
>Can I do this and if so how?

No.

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.

>
>Many thanks
>
>--
>Bill Ferrett
>Accroserve Systems Ltd.
>Office: 01702 342148
>Mobile: 07855 868565

Thread
auto_increment and clustering sequenceBill Ferrett21 Jun
  • Re: auto_increment and clustering sequenceDan Nelson21 Jun
  • Re: auto_increment and clustering sequencePaul DuBois22 Jun