Over time your SQL Server transaction log will fill up and if you don't watch it it'll consume all your server's disk space.

So to cure that problem it's best adding a weekly job to SQL Agent with these commands

Truncate log...

USE [databasename]
GO
BACKUP LOG databasename WITH TRUNCATE_ONLY
GO

Shrink database...

USE [databasename]
GO
DBCC SHRINKDATABASE(N'databasename' )
GO

(truncating only clears the logfile, to reclaim that space you have to shrink the database)

Important!

Do remember to schedule a full database backup before this so you don't lose any important transactions; critical in a live environment.

Clustering in effect provides a very high-level of availability and performance to your application.

In layman’s terms it involves setting up a series of nodes within the data warehouse, with each one containing a live concurrent copy of the same data. In the event of failure of the primary node, the connection is automatically routed to the secondary node (which then becomes the primary), once the original node is available again connection shifts back to the first node. Thus keeping your application online and available with very little downtime (at most 5 minutes).

Not just a component of MySQL, this can also be found in SQL 2005 and Oracle 10 it allows you to keep your system up and performance balanced across your data network; and moving these nodes to different locations / buildings allows greater redundancy in the event of major physical damage (e.g. a nuclear strike).

Worth the effort if your keen about keeping your job, here’s some links to get you started…

Guides

• Wiki: MySql Cluster

• MySQl AB: Official Info

• Alex Davies How-To

Videos

• Google MySQL Cluster Talk

• MySQL Cluster in 5 Minutes

• MySQL Performace Tuning