This section describes an overview about the settings for integrating with the cluster software.

The cluster software is a program that aims to improve the reliability and operating rate of the system, and implements switching of the systems containing server programs. In Application Server, you can integrate the system with the following cluster software:

• Windows Server Failover Cluster
  This is the cluster software used in Windows Server 2012 and Windows Server 2008. You can use the following operating systems:
  
  • Microsoft Windows Server 2012 Standard
  • Microsoft Windows Server 2012 Datacenter
  • Microsoft Windows Server 2008 Enterprise 32-bit
  • Microsoft Windows Server 2008 Enterprise
  • Microsoft Windows Server 2008 R2 Enterprise
• HA monitor
  This is the cluster software used in AIX/ HP-UX/ Linux.
  

Note that in Solaris, you cannot operate a system linked with cluster software.

When you operate a system that is linked with the cluster software, if a failure occurs in the Application Server, the application server switches automatically. Therefore, you can execute the recovery process of the Application Server wherein the failure has occurred using the standby recovery server. Also, when failure occurs in the Management Server, you can switch to the Management Server that is set up in the standby mode. As a result, you can reduce the system downtime, and can improve the reliability and operating rate of the system.

Note

When monitoring the failure of Administration Agent and Management Server, and then switching the systems linked with the cluster software, do not set up automatic restart in Administration Agent and Management Server.

In the systems of the Application Server, you can operate the systems linked with the cluster software for reducing the system downtime as per the failure. You can continue the operations even when failure occurs. In the Application Server, the following node switching functionality is used based on the linkage with cluster software:

• 1-to-1 node switching system
  
  • 1-to-1 node switching of the Management Server
    Using a configuration in which the Management Server of the executing node and the Management Server of the standby node have the 1-to-1 ratio, you can continue the operations by switching the nodes when a failure occurs in the Management Server of the executing node.
    
  • Mutual node switching
    In this system, the Application Server of the executing node and the Application Server of the standby node are specified in the 1-to-1 ratio, and while each Application Server operates as an executing node, each Application Server functions as the standby node for the other Application Server. This enables the execution of economical operations using a few Application Server machines.
    
• Node switching system for the host unit management model
  In this system, deploy multiple (1-to-N) executing node Application Server machines (hosts) and one standby node Application Server machine (host), and then deploy Management Server and Administration Agent in each Application Server machine. When a failure occurs in the Application Server machine of the executing node, you can continue the operations by performing node switching with the Application Server machine of the standby node.
  

  Reference note

  You use the following cluster software to build and operate the 1-to-1 node switching system and node switching system for the host unit management model in Application Server:

  • Windows Server Failover Cluster (In Windows)
  • HA monitor (In AIX, HP-UX, or Linux)
    In Solaris, you cannot operate a system linked with the cluster software.
    

  You must understand the following terminology when you link a system with the cluster software.

  In a cluster software, the generic name for the entire system that contains programs and communication equipment apart from the hardware required for the business processing is called a node. The names for distinguishing the nodes running in systems are the executing node and the standby node.

  • Executing node
    A node that has a server running currently. If node switching occurs, this node becomes the standby node.
    
  • Standby node
    A node that has a server currently in the standby mode. If node switching occurs, this node becomes the executing node.
    

  Furthermore, the names for distinguishing the nodes in the system settings when the system is not running are the active node and the spare node.

  • Active node
    A node that is initially started as an executing node.
    
  • Spare node
    A node that is initially started as a standby node.
    

In Application Server, by integrating the systems with the cluster software, you can operate the system using the node-switching configuration of the J2EE application execution environment.

The following table lists and describes the references of the node switching functionality according to the linkage with the cluster software.

Table 16-2 Node switching functionality based on linkage with cluster software

Functionality		Reference
1-to-1 node switching system	1-to-1 node switching of Management Server	17.2
	Mutual node switching	18.2
Node switching system for the host unit management model		20.2

Organization of this section

16.2.1 Systems that operate the executing node and the standby node in the 1-to-1 ratio (1-to-1 node switching system)

16.2.2 System operating on mutual standby (mutual node switching systems)

16.2.3 System that operates the standby node as the recovery server (N-to-1 recovery system)

16.2.4 System in which Application Server machines (host) of the executing node and the standby node operate in a N-to-1 ratio (Node switching system for the host unit management model)