From: Ben Clewett
Date: May 13 2008 7:44am
Subject: Re: Data Modelling
List-Archive: http://lists.mysql.com/mysql/212766
Message-Id: <482946F3.2050406@clewett.org.uk>
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

If you use InnoDB you should not have a problem as it used row-level 
locking and isolated transitions.

Other than that you can split your tables into smaller ones using either 
partitioning or the federated engine...

Ben




Krishna Chandra Prajapati wrote:
> Hi,
> 
> Generally, in data modelling there are some independent table, user related
> tables, master tables and one main table which is related with most of the
> tables. Almost in all the queries main table is involved to fetch the `id`.
> In this way main table is used maximum. some times the main table gets
> locked due to the other query. When the hits on database server increases
> the locking time will increase.
> 
> Is there any way to reduce the locking time of main table.
>  for eg main_dup can be created. To reflect all the changes from main table
> to main_dup trigger can be used. So main_dup can be used to reduce the
> locking time of table.
> 
> The above is my idea. Is there any other way to reduce the locking period.
>