BackUp LevelNext

State Management and Server Clustering

When using ClusterCATS for ColdFusion to cluster Web servers, and you want to use client variables, you must use an external repository to store client variables, since the system registry, which is bound to an individual system, can't be dynamically accessed by any other system.

Configuring a data source in a clustering environment

Like any other ColdFusion data source, the data source you specify to store your client variables can live anywhere on your network that is accessible to ColdFusion. When you add a new data source to ColdFusion for storing client variables (we recommend dedicating a data source for this purpose rather than using an existing data source) you can enable an option for ColdFusion to automatically create the tables necessary for storing client variables.

However, in a cluster of servers all accessing the same data source for client variables, ColdFusion only needs to create the necessary tables once. As you add the client variables data source to each ColdFusion Server in the cluster, you do not need to choose the option to have ColdFusion create the tables necessary to support client variables, since they have already been created when you added the data source for the first server in the cluster. If you inadvertently enable the option to automatically create client variable database tables, ColdFusion generates a SQL error.

For information about managing state in a clustered environment, see Managing State in a Clustered Environment .


BackUp LevelNext

allaire

AllaireDoc@allaire.com
Copyright © 1998, Allaire Corporation. All rights reserved.