JBoss EAP 7 embeds a JBoss EAP server instance inside the management CLI process. By using the embedded JBoss EAP 7 server, you can access the management CLI remote administration without connecting to a running server.

The main goal of using this approach is to enable the local administration of JBoss EAP without requiring a socket-based connection. This can be very useful when, for example, for security reasons, the 9990 management port is not available.

Another example of how this is useful is that you can perform some configuration offline and apply in production only when the update can be made.

To use this feature, it is necessary to start the management CLI without providing the connection parameter:

# $JBOSS_HOME/bin/jboss-cli.sh

Then, you can start the embedded server:

[disconnected /] embed-server --server-config=standalone-ha.xml

In the previous example, every change made by the management CLI persists on the standalone-ha.xml file, which is available in the JBOSS_HOME/standalone/configuration directory. If you do not specify the --server-config parameter, then the standalone.xml contains the changes.

It is possible to start with an empty configuration using the parameter --empty-config

[disconnected /] embed-server --server-config=production-ha.xml --empty-config

The production-ha.xml file is created in the JBOSS_HOME/standalone/configuration directory. This file only contains the XML tags related to configurations performed in the management CLI. If the file already exists, then the command fails. This condition prevents the accidental deletion of the configuration file.

Operations are a low level way to manage the JBoss EAP server.

An operation in the management CLI has the following format:

[node] : operation_name [ parameters ] [ headers ]

The node represents the target resource address or the node where the operation is invoked. It is a key/value pair with the node type and the node name. An example of a node is /subsystem=datasources.

After specifying the desired node, you must define the operation name with an optional list of parameters. The colon is required before the operation name because it serves as a separator between the node and the operation. A colon is required even if the node is empty. In this case, the operation is executed in the current node level.

The following operations are very common:

The commands contains a user friendly syntax and translate into operation requests.

The following commands are the basic supported commands for the management CLI:

Tab completion shows all possible commands available at any point in a current command. For example, enter /. Then, press Tab to view all possible values after the / character.

[standalone@localhost:9990 /] /
core-service          deployment-overlay    interface             socket-binding-group  system-property
deployment            extension             path                  subsystem

Start typing interface after the / character, and press Tab. The management CLI not only completes the interface sub-level, but it adds an equals sign, because an equals sign is the only possible value after /interface.

Press Tab again and all the interfaces display:

[standalone@localhost:9990 /] /interface=
management  public

JBoss EAP supports using a text file for the management CLI as a script. Using this approach, you can create scripts for repetitive tasks. For example, you can create a script file that configures a data source and test a connection from the pool:

/subsystem=datasources/data-source=appDs:add\
(jndi-name=java:jboss/datasources/appDS,driver-name=h2,user-name=jb248, \
password=jb248,connection-url="jdbc:h2:mem:app;DB_CLOSE_DELAY=-1;DB_CLOSE_ON_EXIT=FALSE")
/subsystem=datasources/data-source=appDs:test-connection-in-pool

You can execute the previous content from a /home/student/create-datasource.cli text file:

# $JBOSS_HOME/jboss-cli.sh --connect --controller=localhost:9990 \
--file=/home/student/create-datasource.cli

The management CLI provides a batch command that supports multiple commands in one atomic unit. If one of the commands or operations fails, then all the other successfully commands and operations in the batch are rolled back.

In the previous example, it as good idea to create the data source using the batch command. If testing a connection from the pool fails, then the data source creation is rolled back. To execute a batch, use the run-batch command:

batch
/subsystem=datasources/data-source=appDs:add\
(jndi-name=java:jboss/datasources/appDS,driver-name=h2,user-name=jb248,\
password=jb248,connection-url="jdbc:h2:mem:app;DB_CLOSE_DELAY=-1;DB_CLOSE_ON_EXIT=FALSE")
/subsystem=datasources/data-source=appDs:test-connection-in-pool
run-batch

The management CLI enables integration with external scripts. You can pass a command or an operation to the CLI by using the --command attribute:

# $JBOSS_HOME/jboss-cli.sh -c --controller=localhost:9990 \
--command="/subsystem=datasources:read-resource"

The result of the operation returns, in DMR syntax, to the script that invoked the CLI. You can specify a set of commands by using the commands attribute. This attribute specifies a comma-separated list (the list must not contain whitespace) of commands and operations that should be executed in the management CLI session.

# ./jboss-cli.sh -c --controller=localhost:9990 --commands="cd /subsystem=datasources,ls"