Prefix of system properties sent to the Launcher at server startup.

The 'server/jvm-config/jvm-option' value overrides this value.

Option to generate an exception when a thread executing ejb or jdbc operation in JEUS receives an interrupt signal.

The exception generated by an interrupted thread must be processed by the user application.

Internal queue size of thread pools used in JEUS.

Thread pools are generally used in JEUS modules. If the number of threads is insufficient to handle system load, tasks are accumulated in the queue up to the maximum queue size.

The 'server/system-thread-pool/queue-size' value in domain.xml overrides this value.

Keep-alive time for inactive threads. If the thread pool contains more than the minimum number of threads, inactive threads are automatically removed from the thread pool after the specified time period.

The 'server/system-thread-pool/keep-alive-time' value in domain.xml overrides this value.

JVM bit number.

If 64 is set to run a 64-bit JVM, '-d64' is appended.

IBM does not require '-d64'.

CPU overflow ratio when monitoring the process CPU usage.

Use a floating point number. Assuming the maximum CPU usage is 100, this is the percent of the maximum regarded as overflow.

Option to restart the server when there is a memory overflow.

Memory overflow ratio when monitoring the memory usage.

Use a floating number. Assuming the maximum memory usage is 100, this is the percent of the maximum regarded as memory overflow.

Maximum time (ms) to wait for an engine container to shut down when the Manager shuts down the engine container or the engine container shuts itself down.

If the engine container has not shut down when this time expires, it is forcibly shut down.

Number of backup domain.xml files saved during a dynamic update.

Option to create dynamic OID when using the SNMP server.

When using the SNMP server, OID can be used either statically or dynamically.

Length of dynamic OID.

Sets the number of digits for the OID component appended to the MIB based OID.

Maximum number of socket connections for deployment.

Applications can be deployed from DAS to Managed Servers(MSs) in parallel by using multiple connections to reduce the boot time. DAS is simultaneously connected to multiple MSs.

When the maximum number of connections in DAS is reached and there are application deployment requests from other MSs, the MS uses only one connection to receive the applications in order.

Maximum number of boot time deploy connections used when applications are deployed in parallel to an MS from DAS.

Note, if too many connections are configured for an MS, other MSs can be affected because the number of connections to DAS is limited.

When application deployment is complete, the connection is released and the number of available connections in DAS increments. Even if the specified number is big, the actual number of connections is automatically reduced to the number of applications.

For more information about properties, refer toJEUS Server Guide. "8.1. Overview".

Option to apply when jeus.logging.useAsync is "true". For more information about properties, refer to JEUS Server Guide. "8.1. Overview".

Interval of ping used when checking connection status in JMX connection.

Timeout of ping used when checking connection status in JMX connection.

If a response is not received within the timeout after sending a ping, then it is determined that there is a connection error and generates an IOException.

Timeout (ms) for a process, whose connection request has already been rejected, to wait for a cross connection to be established after the other server has issued a connection request.

TCP buffer size of the JEUS operating environment.

Since this value is used as the read buffer size of JEUS servers, it must be accurate in order to guarantee the best I/O performance.

TCP receive buffer size of the JEUS server environment.

TCP send buffer size of the JEUS client environment.

Maximum size of a single message that can be transmitted using the JEUS protocol.

If the message size exceeds the maximum size, an Exception occurs on the receiving end and the connection is lost.

Interval for sending a ping message.

If this value is set to 0, ping is not sent to the other node to check whether it is alive.

Total message size allowed to accumulate in the write queue until the write threads running in non-blocking mode are blocked.

If the queue size exceeds the specified limit, write threads are blocked.

Total message size accumulated in the write queue when write threads, running in non-blocking mode, are restarted. If the write queue size becomes smaller than the specified byte after a write event, all blocked write threads are restarted.

Number of write attempts allowed for a single message.

This is useful when a message in a socket is so large that the selector thread cannot write messages in other sockets. Note, if this option is enabled and all the messages in the socket are large, the response time can increase.

DNS host name and the port of the server that provides JNDI service. Set this variable to the IP address of JEUS JNSServer. When making a JNDI request to the clustered MSs, the addresses of all clustered servers must be listed using a comma (,) as a delimiter.

ex) yohan:9736, yohan1:9836

User name authenticated by the JEUS Naming Server. If the user name is not set, "anonymous" is used. The user name must be defined in the JEUS Security Domain.

Timeout for JNSLocal to connect to JNSServer.

Connection is retried during the timeout period. When the time expires, the jeus.net.ConnectException is thrown.

Period in which the connection from JNSLocal to JNSServer is maintained. If a connection is not used within the time period, the connection is automatically terminated.

The default value is zero (the connection is maintained indefinitely).

Policy for determining from which server an object is retrieved.

Input options:

If not set, the default value may vary according to the jeus.jndi.clusterlink.prefer-local-node setting.

Option to use an object on the local node first when looking up an object bound to a clustered Managed Server(MS).

The jeus.jndi.clusterlink.selection-policy value overrides this setting if it is set to a value other than 'locallinkPreference'.

Option to enable cache in JNDI clients (JNSClient).

Option to share the bound object among the clustered Managed Servers.

Option to cache an object after it is bound to JNSServer.

Object is bound only to the local JNSClient.

JNDI compatibility setting. Option to provide JNDI services between a version earlier than JEUS v7.0 Fix#2 and a JEUS v7.0 Fix#2 or later.

Option to determine how a subcoordinator registers a transaction with the root coordinator.

Option to check whether a subcoordinator successfully registered with the root coordinator.

Option to store transaction information to be used by the transaction manager for recovery.

Number of retries to recover the failed recovery resources.

Recovery retries operate as a background job.

When a transaction timeout is reached and a commit or rollback command is not received, transaction manager keeps the resource.

Period in which to enable a cleanup job for the transaction. If timeout occurs, the destroy command is internally executed. Terminating executions are internally enabled for the transaction.

Interval at which the destroy timeout process is checked.

Operation of jeus.tm.destroy-timeout is checked at the specified interval.

ProfileListener classes to be used for TM Profile.

For more information about profiles, refer to JEUS Server Guide. "7.6. Transaction Profile Function".

Input options:

TM runtime version.

Option to determine the I/O method between transaction managers.

Period in which to commit or roll back after the transaction starts.

If timeout occurs, the transaction is rolled back forcibly.

Period in which the root coordinator receives a Preparation Completed signal from a subcoordinator.

If timeout occurs, the transaction is rolled back forcibly.

Period in which a subcoordinator waits for a decision after sending a Preparation Completed signal to the root coordinator.

If timeout occurs, the subcoordinator continuously sends a Waiting for Decision signal to the root coordinator.

Recalculated time period sent when transmitting a transaction to another container. It is calculated by subtracting the time period consumed before the transaction is actually transmitted from the active timeout period.

This function is available in JEUS v6.0 Fix#9. Set to "false" for backward compatibility.

Maximum number of Backup Queue entries allowed. If the network is instable, backup fails continuously and accumulates in the Backup Queue.

Based on this value, responses are no longer sent for Flow Control.

The unit of sessions in which Backup Thread performs backup, and the size of sessions to be entered into the Backup Queue.

When sessions reach this value, a backup is not automatically triggered. Backup is performed as a batch job when load is heavy. In normal operation, one session operates and Backup Thread backs up the session.

The count unit for dumpbackup.

If the server to be backed up restarts, dumpbackup is performed. In this case, if the server has too many sessions, divide the sessions into uniform units of size and then perform backup.

Exclude DAS from clusters.

To use DAS as a manager, exclude DAS from session clusters.

This option applies to all http-session-config, session-cluster, and cluster.

Configure a distributed session server to run services even when the server is not backed up.

Distributed session servers can share sessions and respond to errors only when backup is performed by the backup server.

Set this option when running basic services is more important than keeping sessions.

Option for a restarting server to recover sessions from the server it was backed up in.

By default, recovery is not performed on a backup, but backup is supported during sequential patching in order to prevent session loss. For more information, refer to "Session Management Guide".