Planning the Installation and Configuration

Use the following section in conjunction with your configuration worksheets as a checklist before you install and configure Sun Cluster HA for Sun ONE Web Server.

Consider the following questions before you start your installation.

• Will you run Sun Cluster HA for Sun ONE Web Server as a failover or as a scalable data service? See the Sun Cluster 3.1 Concepts Guide document for information on the two types of services. For scalable services, consider the following questions.

  • What nodes will host the scalable service? In most cases, you will want to scale across all nodes. You can, however, limit the set of nodes that host the service.

  • Will your Sun ONE Web Server instances require sticky IP? Sticky IP is a resource property setting, Load_balancing_policy, which stores the client state in memory so that return traffic from the same node always goes to the same cluster node. You can choose from several load balancing policies, as described in the table on resource properties in “Standard Properties” in Sun Cluster 3.1 Data Service Planning and Administration Guide.

    Exercise caution when you change Load_balancing_weights for an online scalable service that has Load_balancing_policy set to LB_STICKY or LB_STICKY_WILD. Changing those properties while the service is online can cause existing client affinities to be reset, and hence a different node might service a subsequent client request even if another cluster member had previously serviced the client.

    Similarly, when a new instance of the service is started on a cluster, existing client affinities might be reset.

• Where will the Web server root reside?

• Does the Web server serve data for another highly available application? If so, resource dependencies might exist between the resources so that one starts or stops before the other. See “Standard Properties” in Sun Cluster 3.1 Data Service Planning and Administration Guide for a description of the resource property Resource_dependencies that sets these dependencies.

• Determine the resource groups to use for network addresses and application resources and the dependencies between them. See “Standard Properties” in Sun Cluster 3.1 Data Service Planning and Administration Guide for a description of the resource group property RG_dependencies that sets these dependencies.

• Provide the logical hostname (for failover services) or shared address (for scalable services) for clients to use to access the data service.

• Because you can configure Sun ONE Web Server to bind to INADDR_ANY, if you plan to run multiple instances of the Sun ONE Web Server data service or multiple data services on the same node, each instance must bind to a unique network address and port number.

• Determine the entries for the Confdir_list and Port_list properties. For failover services, both of these properties can have only one entry. For scalable services, they can have multiple entries. The number of entries, however, must be the same and must map to each other in the order specified. See How to Register and Configure Sun Cluster HA for Sun ONE Web Server for details.

• Determine where to place logs, error files, and the PID file on the local file system.

• Determine where to place the contents on the cluster file system.