Planning the Sun Cluster HA for SAP Web Application Server Installation and Configuration

This section contains the information that you need to plan your Sun Cluster HA for SAP Web Application Server installation and configuration.

Before you begin, consult your SAP documentation for configuration restrictions and requirements that are not stated in Sun Cluster documentation or imposed by Sun Cluster software.

Be sure to read all SAP notes that pertain to your release of SAP Web Application Server.

Configuration Restrictions

Configuration restrictions are planned and are related to design decisions. Your data service configuration might not be supported if you do not observe these restrictions.

For restrictions that apply to all data services, see the release notes for your release of Sun Cluster at http://docs.sun.com. The following configuration restriction applies only to Sun Cluster HA for SAP Web Application Server.

The following configuration restrictions apply only to Sun Cluster HA for SAP Web Application Server.

• The maximum length of the node names depends on the SAP release. Refer to the SAP installation guide for your release of SAP. This limitation is an SAP software restriction.

• Configure the database resource to be in a different resource group from the SAP enqueue server resource, if possible.

  However, if you are using Oracle as the database, and if you must configure Oracle in the same resource group as the SAP enqueue server resource, you must ensure that the value of the restart_type extension property for Sun Cluster HA for Oracle is set to RESOURCE_RESTART. This property setting means that if the response to a fault in the Oracle database is to restart the Oracle resource, only the Oracle database is restarted. If this property is set to RESOURCE_GROUP_RESTART, and if the response to a fault in the Oracle database is to restart the Oracle resource, all the resources in the resource group are restarted. However, if the SAP replica server is configured and online on another node, the SAP enqueue server cannot be restarted locally. Therefore, the restart_type property must be set to RESOURCE_RESTART.

Configuration Requirements

Configuration requirements represent actions that you must perform. If your data service configuration does not satisfy these requirements, the data service configuration might not be supported.

For requirements that apply to all data services, see Configuration Guidelines for Sun Cluster Data Services in Sun Cluster Data Services Planning and Administration Guide for Solaris OS.

The following configuration requirements apply only to Sun Cluster HA for SAP Web Application Server.

• The fault monitor probe for the resource type of the SAP message server requires the msprot program. You must download this binary from service.sap.com. Refer to the relevant SAP note for information about obtaining and using the msprot program.

• If you are using the SAP Web Application Server stack that is designated for Java only, the following requirement applies.

  The Sun Cluster HA for SAP Web Application Server data service provides a configuration file and registration script that ensure that the SAP J2EE engine is configured correctly within Sun Cluster. The configuration file is located in /opt/SUNWscswa/util/ha_sap_j2ee_config. The registration script is located in /opt/SUNWscswa/util/ha_sap_j2ee_register. This file and this script are used during the registration and configuration of Sun Cluster HA for SAP Web Application Server.

Supported Data Service Configurations

Sun Cluster HA for SAP Web Application Server supports configurations that conform to the requirements that are specified in the previous section.

All configurations of Sun Cluster HA for SAP Web Application Server have the following requirements:

• The SAP enqueue server is configured as a failover resource.

• The SAP replica server is configured as a failover resource.

• The SAP message server is configured as a failover resource.

• The SAP web application server component can be configured either as a failover or as a scalable data service resource.

• The SAP J2EE engine can be configured either as a failover or as a scalable data service resource.

• The SAP enqueue server and the SAP message server are resources in the SAP central services resource group. The SAP replica server is a resource in the SAP replica server resource group.

• The SAP replica server resource is configured to depend on the SAP enqueue server resource. This dependency ensures that the SAP replica server does not attempt to start until the SAP enqueue server is online.

• The SAP central services resource group is configured to have weak positive affinity with the SAP replica server resource group. This affinity ensures that, if a hardware or software failure occurs on the node that is mastering the SAP central services resource group, the SAP central services resource group fails over to the node where the SAP replica server resource group is currently running.

• The SAP replica server resource group is configured to have strong negative affinity with the SAP central services resource group. This affinity ensures that, if the SAP central services resource group fails over, the SAP replica server resource group is off-loaded from the node where it is currently running before the SAP central services resource group is brought online on that same node. The SAP replica server resource group is started on another node, if one is available.

The following examples show the following supported configurations of Sun Cluster HA for SAP Web Application Server.

• The SAP web application server component is configured as a failover resource.

• The SAP web application server component is configured as a scalable resource.

Example 1 SAP Web Application Server Component Configured as a Failover Resource

This example shows a standard configuration of Sun Cluster HA for SAP Web Application Server with the SAP web application server component that is configured as a failover resource.

The boxes that have a solid border represent the location of the resource groups before failover.

The boxes that have a dashed border represent the location of the resource groups after a failover because of a problem on the first node. The SAP central services resource group fails over, and the SAP replica server resource group consequently moves to another node. The SAP web application server component resource group also fails over.

Example 2 SAP Web Application Server Component Configured as a Scalable Resource

This example shows a standard configuration of Sun Cluster HA for SAP Web Application Server with the SAP web application server component that is configured as a scalable resource.

The boxes that have a solid border represent the location of the resource groups before failover.

The boxes that have a dashed border represent the location of the resource groups after a failover because of a problem on the first node. As in the previous example, the SAP central services resource group fails over, and the SAP replica server resource group consequently moves to another node. Because the SAP web application server component resource group is scalable, it does not fail over, but the corresponding resource groups on the other node continue running.

Configuration Considerations

The following configuration considerations affect the installation and configuration of Sun Cluster HA for SAP Web Application Server.

• The SAP enqueue server and SAP replica server run on different cluster nodes. Therefore, the SAP application files (binary files, configuration files, and parameter files) can be installed either on the global file system or on the local file system. However, the application files for each of these applications must be accessible at all times from the nodes on which these applications are running.

  See Configuration Guidelines for Sun Cluster Data Services in Sun Cluster Data Services Planning and Administration Guide for Solaris OS. This section details the advantages and disadvantages of placing the SAP binaries on the local file system or on the cluster file system.

• The default value of the Retry_count system property for the SAP enqueue server is 0. This value ensures that the SAP enqueue server fails over after the first failure. If replication is running, do not modify this default value because enqueue locks would be lost. When replication is in operation and the Enqueue server is restarted, even on the same node, the locks are no longer valid.

• The resource group for the SAP web application server component can be configured to be a failover resource group or a scalable resource group. If the SAP web application server component is configured to be a scalable resource group, the directory /usr/sap/SAPSID/INSTANCE_NAME must be local on each node where the SAP web application server component is installed, as explained in How to Modify the Installation for a Scalable SAP Web Application Server Component. Note that SAPSID represents the SAP system identification and INSTANCE_NAME represents the name of the SAP web application server instance.

• During the installation and configuration process, track any system files that you change, because you need to copy those files to all the nodes that are to run the application.

• For each new installation, the SAP start and stop scripts are overwritten. Therefore, be sure to modify these scripts as needed at each new installation.

Configuration Planning Questions

Use the questions in this section to plan the installation and configuration of Sun Cluster HA for SAP Web Application Server. Write the answers to these questions in the space that is provided on the data service worksheets in Configuration Worksheets in Sun Cluster Data Services Planning and Administration Guide for Solaris OS.

• Will you run the SAP web application server component of the application as a failover resource or as a scalable resource?

• Will you run the SAP J2EE engine as a failover resource or as a scalable resource?

• What is the logical host name for clients that will access the data service?

• Where will the system configuration files reside?

  See Configuration Guidelines for Sun Cluster Data Services in Sun Cluster Data Services Planning and Administration Guide for Solaris OS for the advantages and disadvantages of placing the SAP binaries on the local file system or on the cluster file system.