2.4 Management elements and process configurations for Application Server

Management elements for Application Server

For systems that use Application Server, system components are managed by units called domains. A single domain administration server resides on each domain to manage the domain. A domain (managed by a domain administration server) consists of management elements called nodes and servers. Servers are managed by elements called clusters, dependency relations, and configurations.

Domain

A domain is an element that manages multiple servers making up a business system. A domain can also manage servers (Java EE servers, web servers, and performance tracers) on different machines. Domains are managed by using domain administration servers.

Domain administration server

A domain administration server is a server instance that is prepared specifically for managing a domain. A domain administration server exists for the domain and is created with the name server. We recommend that you create a domain administration server on an operation management server machine that is not the application execution environment machine.

Node

A node is an element that defines a destination host (machine) within a domain. A server is created on a host defined by a node. Multiple nodes can be defined within a domain. Multiple domains cannot share a single node.

Server

An instantiated element of various servers managed by a domain. This element represents a server entity and contains server configuration information. Server elements include Java EE servers, web servers, and performance tracers. Every server belongs to a single domain.

Server type	Overview
Java EE server	A container server that runs Java EE applications.
Web server	A service program that provides web content. Use Web Server.
Performance tracer	A software program that monitors the operating performance of servers such as Java EE servers and web servers. This program is located on the same node as the server to be monitored.

Cluster

A cluster is a group element of Java EE servers that share the same applications, resources, and configuration information. A cluster can group and manage Java EE servers that are located on different machines. A system configuration that uses a cluster to group multiple Java EE servers is called a cluster configuration. A cluster is used to ensure scalability, load balancing, and failover.

Dependency relation

An instantiated element that holds the configuration information required to correlate two servers. For servers to work with each other, this element automatically sets information about the depended-on destination server on the dependent source server. For example, this element automatically sets the listening port number of the depended-on destination server as a connection port parameter on the dependent source server. Dependency relations are classified into redirection dependency relations and performance tracer dependency relations.

Relation type	Overview
Redirection dependency relation	This relation sets up the redirection destination Java EE servers for the requests received by web servers. It enables you to set up relations between servers on different nodes. Source server: Web server Destination server: Java EE server or cluster
Performance tracer dependency relation	This relation specifies settings that enable performance tracers to collect performance analysis traces in order to check server operating performance during operations. You cannot set up a dependency relation on different nodes. Source server: Web server or Java EE server Destination server: Performance tracer

Configuration

A configuration is an instantiated element that holds configuration information for various servers. Servers and clusters obtain configuration information from this element. However, if a server is associated with a certain configuration and the value of a parameter differs between the server and the configuration, the parameter set on the server takes precedence. There are three types of configurations: Java EE server, web server, and performance tracer configurations.

Configuration type	Overview
Java EE server configuration	This configuration is associated with Java EE servers and clusters, and contains configuration information. Creating a Java EE server or cluster automatically creates a configuration for the server or cluster. If a Java EE server is registered for a cluster, the Java EE server inherits the configuration associated with the cluster.
Web server configuration	This configuration is associated with web servers and contains configuration information. Creating a web server automatically creates a configuration for the server.
Performance tracer configuration	This configuration is associated with performance tracers and contains configuration information. Creating a performance tracer automatically creates a configuration for the performance tracer.

Process configurations for Application Server

The following figure illustrates an example of process configurations for Application Server.

The following table shows a list of Application Server processes. A domain administration server automatically starts as a service when the OS starts.

Component	Process name				Process privileges	Role
Application client	java#				General user privileges	Executes an application client.
Web server control process	httpsd				System privileges (SYSTEM account privileges)	Starts a server process that handles requests, and monitors its operations.
Web server process	---	httpsd			System privileges (SYSTEM account privileges)	Sends responses to static content requests from web browsers, or transfers data to Java EE servers in response to requests from applications. For rotatelogs and rotatelogs2, the number of processes varies according to the settings. For rotatelogs, the default number of processes is two. For rotatelogs2, the default number of processes is one.
	---	rotatelogs
	---	rotatelogs2
Java EE server	java#				System privileges (SYSTEM account privileges)	Executes an application.
Performance tracer	cprfd				System privileges (SYSTEM account privileges)	Collects a series of execution logs for web servers and Java EE servers.
Domain administration server	service_ name				System privileges (SYSTEM account privileges)	Manages operations for web servers, Java EE servers, and performance tracers.
	---	cmd
	---	---	java#
	---	---	---	java#

#:

For the Windows Firewall, if you allow particular programs to communicate through the firewall, specify the process names of these programs. If you specify java as a program to be allowed to communicate through the firewall, communications are allowed for all components whose process name is java. Therefore, to prevent unintended communications, you must specify firewall settings in combination with other rules in order to allow communications with only reliable destinations.

The following table shows the current directories for the Java EE server processes.

Java EE server process	Current directory at startup
Server instance#	installation_directory_for_Java_EE_Server/glassfish/nodes/node_name/server_instance_name/config
Domain administration server#	installation_directory_for_Java_EE_Server/glassfish/domains/domain_name/config
MQ Broker	installation_directory_for_Java_EE_Server/glassfish/nodes/node_name/server_instance_name/config

#:

When a process starts or restarts, all files that were output to the current directory (excluding the log files below) are deleted. Note that, if you change the names of the log files below, the log files will also be deleted. In addition, if you execute the start-instance subcommand of the asadmin utility command with the --sync=full option specified, all files in the current directory including these log files will be deleted.

No.	Log
1	Thread dump log
2	Memory dump log
3	Java VM output message log (error report file)
4	Compiler replay file