Chapter 2. Securing the Server and Its Interfaces


2.1. Building Blocks

2.1.1. Interfaces and Socket Bindings

JBoss EAP utilizes its host’s interfaces, for example inet-address and nic, as well as ports for communication for both its web applications as well as its management interfaces. These interfaces and ports are defined and configured through the interfaces and socket-binding-groups settings in the JBoss EAP.

For more information on how to define and configure interfaces and socket-binding-groups, see the Socket Bindings section of the JBoss EAP Configuration Guide.

Example: Interfaces

<interfaces>
  <interface name="management">
    <inet-address value="${jboss.bind.address.management:127.0.0.1}"/>
  </interface>
  <interface name="public">
    <inet-address value="${jboss.bind.address:127.0.0.1}"/>
  </interface>
</interfaces>

Example: Socket Binding Group

<socket-binding-group name="standard-sockets" default-interface="public" port-offset="${jboss.socket.binding.port-offset:0}">
    <socket-binding name="management-http" interface="management" port="${jboss.management.http.port:9990}"/>
    <socket-binding name="management-https" interface="management" port="${jboss.management.https.port:9993}"/>
    <socket-binding name="ajp" port="${jboss.ajp.port:8009}"/>
    <socket-binding name="http" port="${jboss.http.port:8080}"/>
    <socket-binding name="https" port="${jboss.https.port:8443}"/>
    <socket-binding name="txn-recovery-environment" port="4712"/>
    <socket-binding name="txn-status-manager" port="4713"/>
    <outbound-socket-binding name="mail-smtp">
        <remote-destination host="localhost" port="25"/>
    </outbound-socket-binding>
</socket-binding-group>

2.1.2. Elytron Subsystem

2.1.2.1. Enable Elytron Security Across the Server

There is a simple way to enable Elytron across the server. JBoss EAP 7.1 introduces an example configuration script that enables Elytron as the security provider. This script resides in the EAP_HOME/docs/examples directory in the server installation.

Execute the following command to enable Elytron security across the server.

$ EAP_HOME/bin/jboss-cli.sh --file=EAP_HOME/docs/examples/enable-elytron.cli

2.1.2.2. Create an Elytron Security Domain

Security domains in the elytron subsystem, when used in conjunction with security realms, are used for both core management authentication as well as for authentication with applications.

Important

Deployments are limited to using one Elytron security domain per deployment. Scenarios that may have required multiple legacy security domains can now be accomplished using a single Elytron security domain.

Add a Security Domain Using the Management CLI
/subsystem=elytron/security-domain=domainName:add(realms=[{realm=realmName,role-decoder=roleDecoderName}],default-realm=realmName,permission-mapper=permissionMapperName,role-mapper=roleMapperName,...)
Add a Security Domain Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Other.
  3. Click on View. Select Security Domain from the list on the SSL tab. All security domain related configurations can be done here.

2.1.2.3. Create an Elytron Security Realm

Security realms in the elytron subsystem, when used in conjunction with security domains, are used for both core management authentication as well as for authentication with applications. Security realms are also specifically typed based on their identity store, for example jdbc-realm, filesystem-realm, properties-realm, etc.

Add a Security Realm Using the Management CLI
/subsystem=elytron/type-of-realm=realmName:add(....)

Examples of adding specific realms, such as jdbc-realm, filesystem-realm, and properties-realm can be found in previous sections.

Add a Security Realm Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Security Realm / Authentication.
  3. Click on View. The Security Realm, Security Realm Mapper and Authentication tabs let you configure all the security realm and authentication related attributes.

2.1.2.4. Create an Elytron Role Decoder

A role decoder converts attributes from the identity provided by the security realm into roles. Role decoders are also specifically typed based on their functionality, for example empty-role-decoder, simple-role-decoder, and custom-role-decoder.

Add a Role Decoder Using the Management CLI
/subsystem=elytron/ROLE-DECODER-TYPE=roleDeoderName:add(....)
Add a Role Decoder Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Mapper / Decoder.
  3. Click on View. The Decoder tab lets you do all the role decoder related configurations.

2.1.2.5. Create an Elytron Role Mapper

A role mapper maps roles after they have been decoded to other roles. Examples include normalizing role names or adding and removing specific roles from principals after they have been decoded. Role mappers are also specifically typed based on their functionality, for example add-prefix-role-mapper, add-suffix-role-mapper, and constant-role-mapper.

Adding a Role Mapper Takes the General Form
/subsystem=elytron/ROLE-MAPPER-TYPE=roleMapperName:add(...)
Adding a Role Mapper Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Mapper / Decoder.
  3. Click on View. The Role Mapper tab lets you do all the role mapper related configurations.

2.1.2.6. Create an Elytron Permission Mapper

In addition to roles being assigned to a identity, permissions may also be assigned. A permission mapper assigns permissions to an identity. Permission mappers are also specifically typed based on their functionality, for example logical-permission-mapper, simple-permission-mapper, and custom-permission-mapper.

Add a Permission Mapper Using the Management CLI
/subsystem=elytron/simple-permission-mapper=PermissionMapperName:add(...)
Add a Permission Mapper Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Mapper / Decoder.
  3. Click on View. The Permission Mapper tab lets you do all the permission mapper related configurations.

2.1.2.7. Creating an Authentication Configuration

An authentication configuration contains the credentials to use when making a connection. For more information on authentication configurations, see Configure Client Authentication with Elytron Client in How to Configure Identity Management for JBoss EAP.

Note

Instead of a credential store, you can configure an Elytron security domain to use the credentials of the accessing user. For instance, a security domain can be used in conjunction with Kerberos for authenticating incoming users. Follow the instructions in Configure the Elytron Subsystem in How to Set Up SSO with Kerberos for JBoss EAP, and set obtain-kerberos-ticket=true in the Kerberos security factory.

Add an Authentication Configuration Using the Management CLI
/subsystem=elytron/authentication-configuration=AUTHENTICATION_CONFIGURATION_NAME:add(authentication-name=AUTHENTICATION_NAME, credential-reference={clear-text=PASSWORD})
Add an Authentication Configuration Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Security Realm/Authentication.
  3. Click on View. Select Authentication Configuration from the list on the Authentication tab. All authentication configuration related configurations can be done here.

For the full list of authentication-configuration attributes, see Elytron Subsystem Components Reference.

2.1.2.8. Creating an Authentication Context

An authentication context contains a set of rules and either authentication configurations or SSL contexts to use for establishing a connection. For more information on authentication contexts, see Configure Client Authentication with Elytron Client in How to Configure Identity Management for JBoss EAP.

Add an Authentication Context Using the Management CLI

An authentication context can be created using the following management CLI command.

/subsystem=elytron/authentication-context=AUTHENTICATION_CONTEXT_NAME:add()

Typically, an authentication context will contain a set of rules and either an authentication configuration or a SSL context. The following CLI command provides demonstrates defining an authentication context that only functions when the hostname is localhost.

/subsystem=elytron/authentication-context=AUTHENTICATION_CONTEXT_NAME:add(match-rules=[{authentication-configuration=AUTHENTICATION_CONFIGURATION_NAME, match-host=localhost}])
Add an Authentication Context Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Security Realm/Authentication.
  3. Click on View. Select Authentication Context from the list on the Authentication tab. All authentication context related configurations can be done here.

For the full list of authentication-context attributes, see Elytron Subsystem Components Reference.

2.1.2.9. Create an Elytron Authentication Factory

An authentication factory is an authentication policy used for specific authentication mechanisms. Authentication factories are specifically based on the authentication mechanism, for example http-authentication-factory, sasl-authentication-factory and kerberos-security-factory.

Add an Authentication Factory Using the Management CLI
/subsystem=elytron/AUTH-FACTORY-TYPE=authFactoryName:add(....)
Add an Authentication Factory Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Factory.
  3. Click on View. All Elytron settings for factories can be configured here.

2.1.2.10. Create an Elytron Keystore

A key-store is the definition of a keystore or truststore including the type of keystore, its location, and the credential for accessing it.

To generate an example keystore for use with the elytron subsystem, use the following command in Red Hat Enterprise Linux 7.

$ keytool -genkeypair -alias localhost -keyalg RSA -keysize 1024 -validity 365 -keystore keystore.jks -dname "CN=localhost" -keypass secret -storepass secret
Add a Keystore Using the Management CLI

To define a key-store in Elytron that references the newly made keystore, execute the following management CLI command. This command species the path to the keystore, relative to the file system path provided, the credential reference used for accessing the keystore, and the type of keystore.

/subsystem=elytron/key-store=newKeyStore:add(path=keystore.jks,relative-to=jboss.server.config.dir,credential-reference={clear-text=secret},type=JKS)
Note

The above command uses relative-to to reference the location of the keystore file. Alternatively, you can specify the full path to the keystore in path and omit relative-to.

Add a Keystore Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Other.
  3. Click on View. The Key Store tab lets you do all the keystore related configurations.

2.1.2.11. Create an Elytron Key Manager

A key-manager references a key-store, and is used in conjunction with an SSL context.

Add a Key Manager Using the Management CLI

The following command specifies the underlying keystore to reference, the algorithm to use when initializing the key manager, and the credential reference for accessing the entries in the underlying keystore.

/subsystem=elytron/key-manager=newKeyManager:add(key-store=KEY_STORE,algorithm="PKIX",credential-reference={clear-text=secret})
Important

If an algorithm is not specified, then it will be set to the default KeyManagerFactory algorithm name.

The available key manager algorithms are provided by the JDK in use. For example, a JDK that uses SunJSSE provides the PKIX and SunX509 algorithms.

Add a Key Manager Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Other.
  3. Click on View. The Key Manager tab lets you do all the key manager related configurations.

2.1.2.12. Create an Elytron Truststore

To create a truststore in Elytron execute the following CLI command.

/subsystem=elytron/key-store=default-trust-store:add(type=JKS, relative-to=jboss.server.config.dir, path=application.truststore, credential-reference={clear-text=password})

In order to successfully execute the command above you must have an application.truststore file inside your EAP_HOME/standalone/configuration directory. The truststore must contain the certificates associated with the endpoint or a certificate chain in case the end point’s certificate is signed by a CA.

Red Hat recommends you to avoid using self-signed certificates. Ideally, certificates should be signed by a CA and your truststore should contain a certificate chain representing your ROOT and intermediary CAs.

2.1.2.13. Create an Elytron Trust Manager

To define a trust manager in Elytron execute the following CLI command.

/subsystem=elytron/trust-manager=default-trust-manager:add(key-store=TRUST-STORE-NAME)

This sets the defined truststore as the source of the certificates that the application server trusts.

2.1.2.14. Using the Out of the Box Elytron Components

JBoss EAP provides a default set of Elytron components configured in the elytron subsystem. You can find more details on these pre-configured components in the Out of the Box section of the Security Architecture guide.

2.1.2.14.1. Securing Management Interfaces

You can find more details on the enabling JBoss EAP to use the out of the box Elytron components for securing the management interfaces in the User Authentication with Elytron section.

2.1.2.14.2. Securing Applications

The elytron subsystem provides application-http-authentication for http-authentication-factory by default, which can be used to secure applications. For more information on how to configure application-http-authentication, see the Out of the Box section of the Security Architecture guide.

To configure applications to use application-http-authentication, see Configure Web Applications to Use Elytron or Legacy Security for Authentication in How to Configure Identity Management Guide. You can also override the default behavior of all applications using the steps in the Override an Application’s Authentication Configuration section of the JBoss EAP How to Configure Identity Management Guide.

2.1.2.14.3. Using SSL/TLS

JBoss EAP does provide a default one-way SSL/TLS configuration using the legacy core management authentication, but it does not provide one in the elytron subsystem. You can find more details on configuring SSL/TLS using the elytron subsystem for both the management interfaces as well as for applications in the following sections:

2.1.2.14.4. Using Elytron with Other Subsystems

In addition to securing applications and management interfaces, Elytron also integrates with other subsystems in JBoss EAP.

batch-jberet
You can configure the batch-jberet subsystem to run batch jobs using an Elytron security domain. For more information, see Configure Security for Batch Jobs in the Configuration Guide.
datasources
You can use a credential store or an Elytron security domain to provide authentication information in a datasource definition. For more information, see Datasource Security in the Configuration Guide.
ejb3
You can create mappings for Elytron security domains in the ejb3 subsystem to be referenced by deployments. For more information, see Elytron Integration with the EJB Subsystem in Developing EJB Applications.
iiop-openjdk
You can use the elytron subsystem to configure SSL/TLS between clients and servers using the iiop-openjdk subsystem. For more information, see Configure IIOP to use SSL/TLS with the Elytron Subsystem in the Configuration Guide.
jca
You can use the elytron-enabled attribute to enable Elytron security for a work manager. For more information, see Configuring the JCA Subsystem in the Configuration Guide.
jgroups
You can configure the SYM_ENCRYPT and ASYM_ENCRYPT protocols to reference keystores or credential references defined in the elytron subsystem. For more information, see Securing a Cluster in the Configuration Guide.
mail
You can use a credential store to provide authentication information in a server definition in the mail subsystem. For more information, see Use a Credential Store for Passwords in the Configuration Guide.
messaging-activemq
You can secure remote connections to the remote connections used by the messaging-activemq subsystem. For more information, see the Using the Elytron Subsystem section of Configuring Messaging.
modcluster
You can use an Elytron client ssl-context to communicate with a load balancer using SSL/TLS. For more information, see Elytron Integration with the ModCluster Subsystem.
remoting
You can configure inbound and outbound connections in the remoting subsystem to reference authentication contexts, SASL authentication factories, and SSL contexts defined in the elytron subsystem. For full details on configuring each type of connection, see Elytron Integration with the Remoting Subsystem.
resource-adapters
You can secure connections to the resource adapter using Elytron. You can enable security inflow to establish security credentials when submitting work to be executed by the work manager. For more information, see Configure Resource Adapters to Use the Elytron Subsystem in the Configuration Guide.
undertow
You can use the elytron subsystem to configure both SSL/TLS and application authentication. For more information on configuring application authentication, see Using SSL/TLS and Configure Web Applications to Use Elytron or Legacy Security for Authentication in How to Configure Identity Management.

2.1.2.15. Elytron Audit Logging

Audit logging for the elytron subsystem enables logging of Elytron authentication and authorization events within the application server. Audit log entries are stored in either JSON or SIMPLE, human readable format. By default, audit logging is disabled in Elytron.

You can enable audit logging by configuring any of the following log handlers for Elytron, and then adding them to the desired security domain:

Important

Elytron audit logging is distinct from other audit logging, such as audit logging for the JBoss EAP management interfaces. For more information on management interface audit logging options, see the Management Audit Logging section in the JBoss EAP Configuration Guide.

File Audit Logging

File audit logging stores audit log messages in one specified file in the file system, without dividing them into multiple files.

An Elytron file audit logger, named local-audit, is defined by default. Once enabled, it will write Elytron audit logs to EAP_HOME/standalone/log/audit.log on a standalone server, or EAP_HOME/domain/log/audit.log for a managed domain host.

The attributes of a file audit logger are:

  • path and relative-to: Defines the location of the log file.
  • synchronized: Specifies whether every event should be immediately written to disk.
  • format: Use SIMPLE for human readable text format, or JSON for storing individual events in JSON.

    1. You can use a command similar to the following to create a file audit log.

      /subsystem=elytron/file-audit-log=my_audit_log:add(path="my_audit.log",relative-to="jboss.server.log.dir",format=SIMPLE,synchronized=true)
    2. Enable the defined file audit logger by adding it to a security domain.

      /subsystem=elytron/security-domain=domain-with-file-logger:write-attribute(name=security-event-listener, value=my_audit_log)
Periodic Rotating File Audit Logging

Periodic rotating file audit logging automatically rotates audit log files based on a configured schedule. It has the same basic attributes as the default file audit logger, with the following additional attribute:

  • suffix: This must be in the java.text.SimpleDateFormat format, for example .yyyy-MM-dd-HH. The period of the rotation is automatically calculated based on this suffix, and the suffix is appended to the end of the log file names.

    1. You can use a command similar to the following to create a periodic rotating file audit log.

      /subsystem=elytron/periodic-rotating-file-audit-log=my_periodic_audit_log:add(path="my_periodic_audit.log",relative-to="jboss.server.log.dir",format=SIMPLE,synchronized=false,suffix=".yyyy-MM-dd-HH")
    2. Enable the defined periodic rotating file audit logger by adding it to a security domain.

      /subsystem=elytron/security-domain=domain-with-periodic-file-logger:write-attribute(name=security-event-listener, value=my_periodic_audit_log)
Size Rotating File Audit Logging

Size rotating file audit logging automatically rotates audit log files when the log file reaches a configured file size. It has the same basic attributes as the default file audit logger, with the following additional attributes:

  • rotate-size: The maximum size that the log file can reach before being rotated. The default is 2m for 2 megabytes.
  • max-backup-index: The maximum number of files to backup when rotating.
  • rotate-on-boot: By default, a new log file is not created on server restart. You can set this to true to rotate the log on server restart.
  • suffix: This optionally adds a date suffix to a rotated log. This must be in the java.text.SimpleDateFormat format, for example .yyyy-MM-dd-HH.

When the log file size exceeds the limit defined by the rotate-size attribute, the suffix .1 is appended to the end of the current file and a new log file is created. If there are any existing log files, their suffixed number is incremented by one, for example audit_log.1 is renamed to audit_log.2. This happens until the maximum number of log files defined by max-backup-index is reached. When the max-backup-index is exceeded, the file that is over limit, for example audit_log.99, is removed.

  1. You can use a command similar to the following to create a size rotating file audit log.

    /subsystem=elytron/size-rotating-file-audit-log=my_size_log:add(path="my_size_audit.log",relative-to="jboss.server.log.dir",format=SIMPLE,synchronized=false,rotate-size="2m",max-backup-index=10)
  2. Enable the defined size rotating audit logger by adding it to a security domain.

    /subsystem=elytron/security-domain=domain-with-size-logger:write-attribute(name=security-event-listener, value=my_size_log)
Syslog Audit Logging

A syslog handler specifies the parameters by which audit log entries are sent to a syslog server, specifically the syslog server’s host name and port on which the syslog server is listening. Sending audit logging to a syslog server provides more security options than logging to a local file or local syslog server. Multiple syslog handlers can be defined and be active at the same time.

  1. Add a syslog handler.

    /subsystem=elytron/syslog-audit-log=syslog-logger:add(host-name=HOST_NAME, port=PORT, server-address=SERVER_ADDRESS, format=JSON, transport=UDP)
  2. Enable the defined syslog audit logger by adding it to a security domain.

    /subsystem=elytron/security-domain=domain-with-syslog-logger:write-attribute(name=security-event-listener, value=syslog-logger)
Important

To send logs to syslog server over TLS, you can add the following configuration:

/subsystem=elytron/syslog-audit-log=remote-audit:add(transport=SSL_TCP,server-address=127.0.0.1,port=9898,host-name=Elytron,ssl-context=audit-ssl)
Note

To send security events to more destinations, mainly loggers, the aggregate-security-event-listener resource is used. This delivers all events to all listeners specified in the aggregate listener definition.

2.1.2.16. Enable and Disable the Elytron Subsystem

The elytron subsystem comes pre-configured with the default JBoss EAP profiles alongside the legacy security subsystem.

If you are using a profile where the elytron subsystem has not been configured, you can add it by adding the elytron extension and enabling the elytron subsystem.

To add the elytron extension required for the elytron subsystem:

/extension=org.wildfly.extension.elytron:add()

To enable the elytron subsystem in JBoss EAP:

/subsystem=elytron:add

reload

To disable the elytron subsystem in JBoss EAP:

/subsystem=elytron:remove

reload
Important

Other subsystems within JBoss EAP may have dependencies on the elytron subsystem. If these dependencies are not resolved before disabling it, you will see errors when starting JBoss EAP.

2.1.3. Legacy Security Subsystem

2.1.3.1. Enable and Disable the Security Subsystem

To disable the security subsystem in JBoss EAP:

/subsystem=security:remove
Important

Other subsystems within JBoss EAP may have dependencies on the security subsystem. If these dependencies are not resolved before disabling it, you will see errors when starting JBoss EAP.

To enable the security subsystem in JBoss EAP:

/subsystem=security:add

2.1.4. Legacy Security Realms

JBoss EAP uses security realms to define authentication and authorization mechanisms, such as local, LDAP, properties, which can then be used by the management interfaces. For more background information on security realms, see the Security Realms section of the Red Hat JBoss Enterprise Application Platform Security Architecture guide.

Example: Security Realms

<security-realms>
  <security-realm name="ManagementRealm">
    <authentication>
      <local default-user="$local" skip-group-loading="true"/>
      <properties path="mgmt-users.properties" relative-to="jboss.server.config.dir"/>
    </authentication>
    <authorization map-groups-to-roles="false">
      <properties path="mgmt-groups.properties" relative-to="jboss.server.config.dir"/>
    </authorization>
  </security-realm>
  <security-realm name="ApplicationRealm">
    <authentication>
      <local default-user="$local" allowed-users="*" skip-group-loading="true"/>
      <properties path="application-users.properties" relative-to="jboss.server.config.dir"/>
    </authentication>
    <authorization>
      <properties path="application-roles.properties" relative-to="jboss.server.config.dir"/>
    </authorization>
  </security-realm>
</security-realms>

Note

In addition to updating the existing security realms, JBoss EAP also allows you to create new security realms. You can create new security realms via the management console as well as invoking the following command from the management CLI:

/core-service=management/security-realm=NEW-REALM-NAME:add()

If you create a new security realm and want to use a properties file for authentication or authorization, you must create a new properties file specifically for the new security domain. JBoss EAP does not reuse existing files used by other security domains nor does it automatically create new files specified in the configuration if they do not exist.

2.1.5. Using Authentication and Socket Bindings for Securing the Management Interfaces

By default, JBoss EAP defines an http-interface to connect to the management interfaces:

[standalone@localhost:9990 /] /core-service=management:read-resource(recursive=true)
{
    "outcome" => "success",
    "result" => {
        "access" => {...},
        "ldap-connection" => undefined,
        "management-interface" => {"http-interface" => {
            "allowed-origins" => undefined,
            "console-enabled" => true,
            "http-authentication-factory" => "management-http-authentication",
            "http-upgrade" => {
                "enabled" => true,
                "sasl-authentication-factory" => "management-sasl-authentication"
            },
            "http-upgrade-enabled" => true,
            "sasl-protocol" => "remote",
            "secure-socket-binding" => undefined,
            "security-realm" => undefined,
            "server-name" => undefined,
            "socket-binding" => "management-http",
            "ssl-context" => undefined
        }},
        "security-realm" => {...},
        "service" => undefined
    }
}

You can use a combination socket-binding, http-authentication-factory and http-upgrade to secure the management interfaces using the elytron subsystem. Alternatively, you can use socket-binding with security-realm to secure the management interfaces with the legacy core management authentication. You can also disable the management interfaces, and configure users of the interfaces to have various roles and access rights.

2.2. How to Secure the Management Interfaces

The following sections show how to perform various operations related to securing the JBoss EAP management interfaces and related subsystems.

Note

The management CLI commands shown assume that you are running a JBoss EAP standalone server. For more details on using the management CLI for a JBoss EAP managed domain, see the JBoss EAP Management CLI Guide.

Elytron Integration with the Management CLI

The management interfaces can be secured using resources from the elytron subsystem in the same way as it is done by the legacy security realms.

The SSL configuration for connections comes from one of these locations:

  • Any SSL configuration within the CLI specific configuration.
  • The default SSL configuration that automatically prompts users to accept the server’s certificate.
  • The java system property.

Client configuration can be modified using the wildfly-config.xml file.

Note

If you set the -Dwildfly.config.url property, any file can be used by the client for configuration.

2.2.1. Configure Networking and Ports

Depending on the configuration of the host, JBoss EAP may be configured to use various network interfaces and ports. This allows JBoss EAP to work with different host, networking, and firewall requirements.

For more information on the networking and ports used by JBoss EAP, as well as how to configure these settings, see the Network and Port Configuration section of the JBoss EAP Configuration Guide.

2.2.2. Disabling the Management Console

Other clients, such as JBoss Operations Network, operate using the HTTP interface for managing JBoss EAP. In order to continue using these services, just the web-based management console itself may be disabled. This is accomplished by setting the console-enabled attribute to false:

/core-service=management/management-interface=http-interface/:write-attribute(name=console-enabled,value=false)

2.2.3. Disabling Remote Access to JMX

Remote access to the jmx subsystem allows for JDK and application management operations to be triggered remotely. To disable remote access to JMX in JBoss EAP, remove the remoting connector in the jmx subsystem:

Removing the Remoting Connector
/subsystem=jmx/remoting-connector=jmx/:remove

For more information on JMX, see the JMX section of the Red Hat JBoss Enterprise Application Platform Security Architecture guide.

2.2.4. Silent Authentication

The default installation of JBoss EAP contains a method of silent authentication for a local management CLI user. This allows the local user the ability to access the management CLI without user name or password authentication. This functionality is enabled as a convenience, and to assist local users running the management CLI scripts without requiring authentication. It is considered a useful feature given that access to the local configuration typically also gives the user the ability to add their own user details or otherwise disable security checks.

The convenience of silent authentication for local users can be disabled where greater security control is required. This can be achieved by removing the local element within the security-realm attribute of the configuration file. This is applicable to both standalone instance as well as managed domain.

Important

The removal of the local element should only be done if the impact on the JBoss EAP instance and its configuration is fully understood.

To remove silent authentication when using the elytron subsystem:

[standalone@localhost:9990 /] /subsystem=elytron/sasl-authentication-factory=managenet-sasl-authentication:read-resource
{
    "outcome" => "success",
    "result" => {
        "mechanism-configurations" => [
            {
                "mechanism-name" => "JBOSS-LOCAL-USER",
                "realm-mapper" => "local"
            },
            {
                "mechanism-name" => "DIGEST-MD5",
                "mechanism-realm-configurations" => [{"realm-name" => "ManagementRealm"}]
            }
        ],
        "sasl-server-factory" => "configured",
        "security-domain" => "ManagementDomain"
    }
}

/subsystem=elytron/sasl-authentication-factory=temp-sasl-authentication:list-remove(name=mechanism-configurations,index=0)

reload

To remove silent authentication when using a legacy security realm:

/core-service=management/security-realm=REALM_NAME/authentication=local:remove

2.2.5. Removing Undertow Response Headers

The default JBoss EAP undertow subsystem includes two response headers that are appended to each HTTP response by the default-host:

  • Server, which is set to JBoss-EAP/7
  • X-Powered-By, which is set to Undertow/1

Although these can be useful for development and debugging purposes, you might want to remove these headers if you do not want to disclose information about the server in use.

Use the following management CLI commands to remove these response headers from the default-host:

/subsystem=undertow/server=default-server/host=default-host/filter-ref=server-header:remove

/subsystem=undertow/server=default-server/host=default-host/filter-ref=x-powered-by-header:remove

reload

2.2.6. Enable One-way SSL/TLS for the Management Interfaces Using the Elytron Subsystem

  1. Obtain or generate your keystore.

    Before enabling one-way SSL/TLS in JBoss EAP, you must obtain or generate the keystore you plan on using. To generate an example keystore in Red Hat Enterprise Linux 7, use the following command.

    $ keytool -genkeypair -alias localhost -keyalg RSA -keysize 1024 -validity 365 -keystore keystore.jks -dname "CN=localhost" -keypass secret -storepass secret
  2. Create a key-store, key-manager, and server-ssl-context.

    /subsystem=elytron/key-store=httpsKS:add(path=keystore.jks,relative-to=jboss.server.config.dir,credential-reference={clear-text=secret},type=JKS)
    
    /subsystem=elytron/key-manager=httpsKM:add(key-store=httpsKS,algorithm="SunX509",credential-reference={clear-text=secret})
    
    /subsystem=elytron/server-ssl-context=httpsSSC:add(key-manager=httpsKM,protocols=["TLSv1.2"])
    Important

    You need to know what key manager algorithms are provided by the JDK you are using. For example, a JDK that uses SunJSSE provides the PKIX and SunX509 algorithms. You also need to determine what HTTPS protocols you want to support. The example commands above use TLSv1.2. You can use the cipher-suite-filter argument to specify which cipher suites are allowed, and the use-cipher-suites-order argument to honor server cipher suite order. The use-cipher-suites-order attribute by default is set to true. This differs from the legacy security subsystem behavior, which defaults to honoring client cipher suite order.

    Note

    The above command uses relative-to to reference the location of the keystore file. Alternatively, you can specify the full path to the keystore in path and omit relative-to.

  3. Enable HTTPS on the management interface.

    /core-service=management/management-interface=http-interface:write-attribute(name=ssl-context, value=httpsSSC)
    
    /core-service=management/management-interface=http-interface:write-attribute(name=secure-socket-binding, value=management-https)
  4. Reload the JBoss EAP instance.

    reload

One-way SSL/TLS is now enabled for the management interfaces.

Important

In cases where you have both a security-realm and ssl-context defined, JBoss EAP will use the SSL/TLS configuration provided by ssl-context.

2.2.7. Enable Two-way SSL/TLS for the Management Interfaces Using the Elytron Subsystem

  1. Obtain or generate your keystore.

    Before enabling one-way SSL/TLS in JBoss EAP, you must obtain or generate the keystores, truststores and certificates you plan on using. To generate an example set of keystores, truststores, and certificates in Red Hat Enterprise Linux 7, use the following commands.

    1. Generate your server and client keystores.

      $ keytool -genkeypair -alias localhost -keyalg RSA -keysize 1024 -validity 365 -keystore server.keystore.jks -dname "CN=localhost" -keypass secret -storepass secret
      
      $ keytool -genkeypair -alias client -keyalg RSA -keysize 1024 -validity 365 -keystore client.keystore.jks -dname "CN=client" -keypass secret -storepass secret
    2. Export your server and client certificates.

      $ keytool -exportcert  -keystore server.keystore.jks -alias localhost -keypass secret -storepass secret -file server.cer
      
      $ keytool -exportcert  -keystore client.keystore.jks -alias client -keypass secret -storepass secret -file client.cer
    3. Import the server and client certificates into the opposing truststores.

      $ keytool -importcert -keystore server.truststore.jks -storepass secret -alias client -trustcacerts -file client.cer
      
      $ keytool -importcert -keystore client.truststore.jks -storepass secret -alias localhost -trustcacerts -file server.cer
  2. Configure a key-store, key-manager, trust-manager, and server-ssl-context for the server keystore and truststore.

    /subsystem=elytron/key-store=twoWayKS:add(path=server.keystore.jks,relative-to=jboss.server.config.dir,credential-reference={clear-text=secret},type=JKS)
    
    /subsystem=elytron/key-store=twoWayTS:add(path=server.truststore.jks,relative-to=jboss.server.config.dir,credential-reference={clear-text=secret},type=JKS)
    
    /subsystem=elytron/key-manager=twoWayKM:add(key-store=twoWayKS,algorithm="SunX509",credential-reference={clear-text=secret})
    
    /subsystem=elytron/trust-manager=twoWayTM:add(key-store=twoWayTS,algorithm="SunX509")
    
    /subsystem=elytron/server-ssl-context=twoWaySSC:add(key-manager=twoWayKM,protocols=["TLSv1.2"],trust-manager=twoWayTM,want-client-auth=true,need-client-auth=true)
    Important

    You need to know what key manager algorithms are provided by the JDK you are using. For example, a JDK that uses SunJSSE provides the PKIX and SunX509 algorithms. You also need to determine what HTTPS protocols you want to support. The example commands above use TLSv1.2. You can use the cipher-suite-filter argument to specify which cipher suites are allowed, and the use-cipher-suites-order argument to honor server cipher suite order. The use-cipher-suites-order attribute by default is set to true. This differs from the legacy security subsystem behavior, which defaults to honoring client cipher suite order.

    Note

    The above command uses relative-to to reference the location of the keystore file. Alternatively, you can specify the full path to the keystore in path and omit relative-to.

  3. Enable HTTPS on the management interface.

    /core-service=management/management-interface=http-interface:write-attribute(name=ssl-context, value=twoWaySSC)
    
    /core-service=management/management-interface=http-interface:write-attribute(name=secure-socket-binding, value=management-https)
  4. Reload the JBoss EAP instance.

    reload
  5. Configure your client to use the client certificate.

    You need to configure your client to present the trusted client certificate to the server to complete the two-way SSL/TLS authentication. For example, if using a browser, you need to import the trusted certificate into the browser’s trust store.

    This results in a forced two-way SSL/TLS authentication, without changing the original authentication to the server management.

    If you want to change the original authentication method, see Configure Authentication with Certificates in How to Configure Identity Management for JBoss EAP.

Two-way SSL/TLS is now enabled for the management interfaces.

Important

In cases where you have both a security-realm and ssl-context defined, JBoss EAP will use the SSL/TLS configuration provided by ssl-context.

2.2.8. Configure the Management Interfaces for One-way SSL/TLS with Legacy Core Management Authentication

Configuring the JBoss EAP management interfaces for communication only using one-way SSL/TLS provides increased security. All network traffic between the client and the management interfaces is encrypted, which reduces the risk of security attacks such as a man-in-the-middle attack.

In this procedure unencrypted communication with the JBoss EAP instance is disabled. This procedure applies to both standalone server and managed domain configurations. For a managed domain, prefix the management CLI commands with the name of the host, for example: /host=master.

Important

While performing the steps for enabling one-way SSL/TLS on the management interfaces, do not reload the configuration unless explicitly instructed. Doing so may cause you to be locked out of the management interfaces.

Create a Keystore to Secure the Management Interfaces
Note

This keystore must be in JKS format as the management interfaces are not compatible with keystores in JCEKS format.

Use the following to generate a keystore, replacing the example values for the parameters, for example alias, keypass, keystore, storepass and dname, with the correct values for the environment.

$ keytool -genkeypair -alias appserver -storetype jks -keyalg RSA -keysize 2048 -keypass password1 -keystore EAP_HOME/standalone/configuration/identity.jks -storepass password1 -dname "CN=appserver,OU=Sales,O=Systems Inc,L=Raleigh,ST=NC,C=US" -validity 730 -v
Note

The parameter validity specifies for how many days the key is valid. A value of 730 equals two years.

Ensure the Management Interfaces Bind to HTTPS

Running a Standalone Server

To ensure the management interfaces bind to HTTPS, you must add the management-https configuration and remove the management-http configuration.

Use the following CLI commands to bind the management interfaces to HTTPS:

/core-service=management/management-interface=http-interface:write-attribute(name=secure-socket-binding, value=management-https)

/core-service=management/management-interface=http-interface:undefine-attribute(name=socket-binding)

Running a Managed Domain

Change the socket element within the management-interface attribute by adding secure-port and removing port configuration.

Use the following commands to bind the management interfaces to HTTPS:

/host=master/core-service=management/management-interface=http-interface:write-attribute(name=secure-port,value=9993)

/host=master/core-service=management/management-interface=http-interface:undefine-attribute(name=port)
Optional: Implement a Custom socket-binding-group

If you want to use a custom socket-binding-group, you must ensure the management-https binding is defined, which by default is bound to port 9993. You can verify this from the socket-binding-group attribute of the server’s configuration file or using the management CLI:

/socket-binding-group=standard-sockets/socket-binding=management-https:read-resource(recursive=true)

{
    "outcome" => "success",
    "result" => {
        "client-mappings" => undefined,
        "fixed-port" => false,
        "interface" => "management",
        "multicast-address" => undefined,
        "multicast-port" => undefined,
        "name" => "management-https",
        "port" => expression "${jboss.management.https.port:9993}"
    }
}
Create a New Security Realm

In this example, the new security realm using HTTPS, ManagementRealmHTTPS, uses a properties file named https-mgmt-users.properties located in the EAP_HOME/standalone/configuration/ directory for storing user names and passwords.

  1. Create a properties file for storing user name and passwords.

    User names and passwords can be added to the file later, but for now, you need to create an empty file named https-mgmt-users.properties and save it to that location. The below example shows using the touch command, but you may also use other mechanisms, such as a text editor.

    Example: Using the touch Command to Create an Empty File

    $ touch EAP_HOME/standalone/configuration/https-mgmt-users.properties

  2. Next, use the following management CLI commands to create a new security realm named ManagementRealmHTTPS:

    /core-service=management/security-realm=ManagementRealmHTTPS:add
    
    /core-service=management/security-realm=ManagementRealmHTTPS/authentication=properties:add(path=https-mgmt-users.properties,relative-to=jboss.server.config.dir)
  3. Add users to the properties file.

    At this point, you have created a new security realm and configured it to use a properties file for authentication. You must now add users to that properties file using the add-user script, which is available in the EAP_HOME/bin/ directory. When running the add-user script, you must specify both the properties file and the security realm using the -up and -r options respectively. From there, the add-user script will interactively prompt you for the user name and password information to store in the https-mgmt-users.properties file.

    $ EAP_HOME/bin/add-user.sh -up EAP_HOME/standalone/configuration/https-mgmt-users.properties -r ManagementRealmHTTPS
    ...
    Enter the details of the new user to add.
    Using realm 'ManagementRealmHTTPS' as specified on the command line.
    ...
    Username : httpUser
    Password requirements are listed below. To modify these restrictions edit the add-user.properties configuration file.
     - The password must not be one of the following restricted values {root, admin, administrator}
     - The password must contain at least 8 characters, 1 alphabetic character(s), 1 digit(s), 1 non-alphanumeric symbol(s)
     - The password must be different from the username
    ...
    Password :
    Re-enter Password :
    About to add user 'httpUser' for realm 'ManagementRealmHTTPS'
    ...
    Is this correct yes/no? yes
    ..
    Added user 'httpUser' to file 'EAP_HOME/configuration/https-mgmt-users.properties'
    ...
    Is this new user going to be used for one AS process to connect to another AS process?
    e.g. for a slave host controller connecting to the master or for a Remoting connection for server to server EJB calls.
    yes/no? no
    Important

    When configuring security realms that use properties files to store usernames and passwords, it is recommended that each realm use a distinct properties file that is not shared with another realm.

Configure the Management Interfaces to Use the New Security Realm

Use the following management CLI command to configure the management interfaces to use the new security realm.

/core-service=management/management-interface=http-interface:write-attribute(name=security-realm,value=ManagementRealmHTTPS)
Configure the Management Interfaces to Use the Keystore

Use the below management CLI command to configure the management interfaces to use the keystore. For the parameters file, password and alias their values must be copied from the Create a Keystore to Secure the Management Interfaces step.

/core-service=management/security-realm=ManagementRealmHTTPS/server-identity=ssl:add(keystore-path=identity.jks,keystore-relative-to=jboss.server.config.dir,keystore-password=password1, alias=appserver)
Note

To update the keystore password, use the following CLI command:

/core-service=management/security-realm=ManagementRealmHTTPS/server-identity=ssl:write-attribute(name=keystore-password,value=newpassword)

At this point, you need to reload the server’s configuration:

reload

After reloading the server configuration, the log should contain the following, just before the text which states the number of services that are started:

13:50:54,160 INFO  [org.jboss.as] (Controller Boot Thread) WFLYSRV0061: Http management interface listening on https://127.0.0.1:9993/management
13:50:54,162 INFO  [org.jboss.as] (Controller Boot Thread) WFLYSRV0052: Admin console listening on https://127.0.0.1:9993

The management interfaces are now listening on port 9993, which confirms that the procedure was successful.

Important

At this point, the CLI will disconnect and will be unable to reconnect since the port bindings have changed. Proceed to the next step to update the jboss-cli.xml file to allow the management CLI to reconnect.

Update the jboss-cli.xml File

If using the management CLI to perform management actions, the following changes must to be made to the EAP_HOME/bin/jboss-cli.xml file:

  • Update the value of <default-protocol> to https-remoting.
  • In <default-controller>, update the value of <protocol> to https-remoting.
  • In <default-controller>, update the value of <port> to 9993.

Example: jboss-cli.xml

<jboss-cli xmlns="urn:jboss:cli:2.0">
    <default-protocol use-legacy-override="true">https-remoting</default-protocol>
    <!-- The default controller to connect to when 'connect' command is executed w/o arguments -->
    <default-controller>
        <protocol>https-remoting</protocol>
        <host>localhost</host>
        <port>9993</port>
    </default-controller>
...

The next time you connect to the management interface using the management CLI, you must accept the server certificate and authenticate against the ManagementRealmHTTPS security realm:

Example: Accepting Server Certificate and Authenticating

$ ./jboss-cli.sh -c
Unable to connect due to unrecognised server certificate
Subject    - CN=appserver,OU=Sales,O=Systems Inc,L=Raleigh,ST=NC,C=US
Issuer     - CN=appserver, OU=Sales, O=Systems Inc, L=Raleigh, ST=NC, C=US
Valid From - Tue Jun 28 13:38:48 CDT 2016
Valid To   - Thu Jun 28 13:38:48 CDT 2018
MD5 : 76:f4:81:8b:7e:c3:be:6d:ee:63:c1:7a:b7:b8:f0:fb
SHA1 : ea:e3:f1:eb:53:90:69:d0:c9:69:4a:5a:a3:20:8f:76:c1:e6:66:b6

Accept certificate? [N]o, [T]emporarily, [P]ermenantly : p
Authenticating against security realm: ManagementRealmHTTPS
Username: httpUser
Password:
[standalone@localhost:9993 /]

Important

In cases where you have both a security-realm and ssl-context defined, JBoss EAP will use the SSL/TLS configuration provided by ssl-context.

2.2.9. Setting up Two-way SSL/TLS for the Management Interfaces with Legacy Core Management Authentication

Two-way SSL/TLS authentication, also known as client authentication, authenticates both the client and the server using SSL/TLS certificates. This differs from the Configure the Management Interfaces for One-way SSL/TLS section in that both the client and server each have a certificate. This provides assurance that not only is the server who it says it is, but the client is also who it says it is.

In this section the following conventions are used:

HOST1
The JBoss server hostname. For example: jboss.redhat.com.
HOST2
A suitable name for the client. For example: myclient. Note this is not necessarily an actual hostname.
CA_HOST1
The DN (distinguished name) to use for the HOST1 certificate. For example: cn=jboss,dc=redhat,dc=com.
CA_HOST2
The DN (distinguished name) to use for the HOST2 certificate. For example: cn=myclient,dc=redhat,dc=com.
Prerequisites
Note

If a password vault is used to store the keystore and truststore passwords, which is recommended, the password vault should already be created. For more information on the password vault, see the Password Vault section as well as the Password Vault System section of the Red Hat JBoss Enterprise Application Platform 7 Security Architecture guide.

Warning

Red Hat recommends that SSLv2, SSLv3, and TLSv1.0 be explicitly disabled in favor of TLSv1.1 or TLSv1.2 in all affected packages.

  1. Generate the keystores.

    $ keytool -genkeypair -alias HOST1_alias -keyalg RSA -keysize 1024 -validity 365 -keystore HOST1.keystore.jks -dname "CA_HOST1" -keypass secret -storepass secret
    
    $ keytool -genkeypair -alias HOST2_alias -keyalg RSA -keysize 1024 -validity 365 -keystore HOST2.keystore.jks -dname "CA_HOST2" -keypass secret -storepass secret
  2. Export the certificates.

    $ keytool -exportcert  -keystore HOST1.keystore.jks -alias HOST1_alias -keypass secret -storepass secret -file HOST1.cer
    
    $ keytool -exportcert  -keystore HOST2.keystore.jks -alias HOST2_alias -keypass secret -storepass secret -file HOST2.cer
  3. Import the certificates into the opposing truststores.

    $ keytool -importcert -keystore HOST1.truststore.jks -storepass secret -alias HOST2_alias -trustcacerts -file HOST2.cer
    
    $ keytool -importcert -keystore HOST2.truststore.jks -storepass secret -alias HOST1_alias -trustcacerts -file HOST1.cer
  4. Define a CertificateRealm.

    Define a CertificateRealm in the configuration for the server (host.xml or standalone.xml) and point the interface to it. This can be done using the following commands:

    /core-service=management/security-realm=CertificateRealm:add()
    
    /core-service=management/security-realm=CertificateRealm/server-identity=ssl:add(keystore-path=/path/to/HOST1.keystore.jks, keystore-password=secret,alias=HOST1_alias)
    
    /core-service=management/security-realm=CertificateRealm/authentication=truststore:add(keystore-path=/path/to/HOST1.truststore.jks,keystore-password=secret)
  5. Change the security-realm of the http-interface to the new CertificateRealm.

    /core-service=management/management-interface=http-interface:write-attribute(name=security-realm,value=CertificateRealm)
  6. Add the SSL/TLS configuration for the CLI.

    Important

    In addition to adding the two-way SSL/TLS, the management interface should also be configured to bind to HTTPS. For details, see Ensure the Management Interfaces Bind to HTTPS in the section entitled Configure the Management Interfaces for One-way SSL/TLS with Legacy Core Management Authentication.

    Add the SSL/TLS configuration for the CLI, which uses EAP_HOME/bin/jboss-cli.xml as a settings file.

    To store the keystore and truststore passwords in plain text, edit EAP_HOME/bin/jboss-cli.xml and add the SSL/TLS configuration using the appropriate values for the variables:

    Example: jboss-cli.xml Storing Keystore and Truststore Passwords in Plain Text

    <ssl>
      <alias>HOST2_alias</alias>
      <key-store>/path/to/HOST2.keystore.jks</key-store>
      <key-store-password>secret</key-store-password>
      <trust-store>/path/to/HOST2.truststore.jks</trust-store>
      <trust-store-password>secret</trust-store-password>
      <modify-trust-store>true</modify-trust-store>
    </ssl>

    To use the keystore and truststore passwords stored in a password vault, you need to add the vault configuration and appropriate vault values to EAP_HOME/bin/jboss-cli.xml:

    Example: jboss-cli.xml Storing Keystore and Truststore Passwords in a Password Vault

    <ssl>
      <vault>
        <vault-option name="KEYSTORE_URL" value="path-to/vault/vault.keystore"/>
        <vault-option name="KEYSTORE_PASSWORD" value="MASK-5WNXs8oEbrs"/>
        <vault-option name="KEYSTORE_ALIAS" value="vault"/>
        <vault-option name="SALT" value="12345678"/>
        <vault-option name="ITERATION_COUNT" value="50"/>
        <vault-option name="ENC_FILE_DIR" value="EAP_HOME/vault/"/>
      </vault>
      <alias>HOST2_alias</alias>
      <key-store>/path/to/HOST2.keystore.jks</key-store>
      <key-store-password>VAULT::VB::cli_pass::1</key-store-password>
      <key-password>VAULT::VB::cli_pass::1</key-password>
      <trust-store>/path/to/HOST2.truststore.jks</trust-store>
      <trust-store-password>VAULT::VB::cli_pass::1</trust-store-password>
      <modify-trust-store>true</modify-trust-store>
    </ssl>

Important

In cases where you have both a security-realm and ssl-context defined, JBoss EAP will use the SSL/TLS configuration provided by ssl-context.

2.2.10. HTTPS Listener Reference

For a full list of attributes available for the HTTPS listener, see the Undertow Subsystem Attributes section in the JBoss EAP Configuration Guide.

2.2.10.1. About Cipher Suites

You can configure a list of the encryption ciphers which are allowed. For JSSE syntax, it must be a comma-separated list. For OpenSSL syntax, it must be a colon-separated list. Ensure that only one syntax is used. The default is the JVM default.

Important

Using weak ciphers is a significant security risk. See NIST Guidelines for NIST recommendations on cipher suites.

See the OpenSSL documentation for a list of available OpenSSL ciphers. Note that the following are not supported:

  • @SECLEVEL
  • SUITEB128
  • SUITEB128ONLY
  • SUITEB192

See the Java documentation for a list of the standard JSSE ciphers.

To update the list of enabled cipher suites, use the enabled-cipher-suites attribute of the HTTPS listener in the undertow subsystem.

Example: Management CLI Command for Updating the List of Enabled Cipher Suites

/subsystem=undertow/server=default-server/https-listener=https:write-attribute(name=enabled-cipher-suites,value="TLS_RSA_WITH_AES_128_CBC_SHA,TLS_RSA_WITH_AES_256_CBC_SHA")

Note

The example only lists two possible ciphers, but real-world examples will likely use more.

2.2.11. Enable FIPS 140-2 Cryptography for SSL/TLS on Red Hat Enterprise Linux 7

You can configure Undertow to use FIPS 140-2 compliant cryptography for SSL/TLS. The scope of this configuration example is limited to Red Hat Enterprise Linux 7, using the Mozilla NSS library in FIPS mode.

Important

Red Hat Enterprise Linux 7 must already be configured to be FIPS 140-2 compliant. For more information, see the solution titled How can I make RHEL 6 or RHEL 7 FIPS 140-2 compliant?, which is located on the Red Hat Customer Portal.

Warning

Using the TLS 1.2 protocol when running JBoss EAP in FIPS mode can cause a NoSuchAlgorithmException to occur. More details on this issue can be found in the solution titled NoSuchAlgorithmException: no such algorithm: SunTls12MasterSecret, which is located on the Red Hat Customer Portal.

Therefore, it is not possible to configure HTTP/2 in FIPS mode because HTTP/2 requires the TLS 1.2 protocol. FIPS mode (PKCS11) supports the TLS 1 and the TLS 1.1 protocols so you can use:

  • TLS 1.1 in case of Oracle/OpenJDK
  • TLS 1 in case of IBM java

To configure Undertow to use FIPS 140-2 compliant cryptography for SSL/TLS, you must do the following:

Note

The OpenSSL provider requires a private key, but it is not possible to retrieve a private key from the PKCS11 store. FIPS does not allow the export of unencrypted keys from FIPS compliant cryptographic module. Therefore, for both the elytron subsystem as well as legacy security, it is not possible to use the OpenSSL provider for TLS when in FIPS mode.

2.2.11.1. Configuring the NSS database

  1. Create a directory owned by the appropriate user to house the NSS database.

    Example Commands for Creating the NSS Database Directory

    $ mkdir -p /usr/share/jboss-as/nssdb
    $ chown jboss /usr/share/jboss-as/nssdb
    $ modutil -create -dbdir /usr/share/jboss-as/nssdb

    Note

    The jboss user is only an example. You need to replace it with a user on your operating system that you plan on using for running JBoss EAP.

  2. Create the NSS configuration file: /usr/share/jboss-as/nss_pkcsll_fips.cfg.

    It must specify:

    • a name
    • the directory where the NSS library is located
    • the directory where the NSS database was created in the previous step

      Example: nss_pkcsll_fips.cfg

      name = nss-fips
      nssLibraryDirectory=/usr/lib64
      nssSecmodDirectory=/usr/share/jboss-as/nssdb
      nssDbMode = readOnly
      nssModule = fips

      Note

      If you are not running a 64-bit version of Red Hat Enterprise Linux 6 then set nssLibraryDirectory to /usr/lib instead of /usr/lib64.

  3. Edit the $JAVA_HOME/jre/lib/security/java.security configuration file.

    Add the following line to $JAVA_HOME/jre/lib/security/java.security:

    Example: java.security

    security.provider.1=sun.security.pkcs11.SunPKCS11 /usr/share/jboss-as/nss_pkcsll_fips.cfg

    Note

    The nss_pkcsll_fips.cfg configuration file specified in the above line is the file created in the previous step.

    You also need to update the following link in $JAVA_HOME/jre/lib/security/java.security from:

    security.provider.5=com.sun.net.ssl.internal.ssl.Provider

    to

    security.provider.5=com.sun.net.ssl.internal.ssl.Provider SunPKCS11-nss-fips
    Important

    Any other security.provider.X lines in this file, for example security.provider.2, must have the value of their X increased by one to ensure that this provider is given priority.

  4. Run the modutil command on the NSS database directory you created in the previous step to enable FIPS mode.

    modutil -fips true -dbdir /usr/share/jboss-as/nssdb
    Note

    You may get a security library error at this point requiring you to regenerate the library signatures for some of the NSS shared objects.

  5. Set the password on the FIPS token.

    The name of the token must be NSS FIPS 140-2 Certificate DB.

    modutil -changepw "NSS FIPS 140-2 Certificate DB" -dbdir /usr/share/jboss-as/nssdb
    Important

    The password used for the FIPS token must be a FIPS compliant password. If the password is not strong enough, you may receive an error: ERROR: Unable to change password on token "NSS FIPS 140-2 Certificate DB".

  6. Create a certificate using the NSS tools.

    Example Command

    $ certutil -S -k rsa -n undertow -t "u,u,u" -x -s "CN=localhost, OU=MYOU, O=MYORG, L=MYCITY, ST=MYSTATE, C=MY" -d /usr/share/jboss-as/nssdb

  7. Verify that the JVM can read the private key from the PKCS11 keystore by running the following command:

    $ keytool -list -storetype pkcs11
Important

Once you have FIPS enabled, you may see the following error when starting JBoss EAP:

10:16:13,993 ERROR [org.jboss.msc.service.fail] (MSC service thread 1-1) MSC000001: Failed to start service jboss.server.controller.management.security_realm.ApplicationRealm.key-manager: org.jboss.msc.service.StartException in service jboss.server.controller.management.security_realm.ApplicationRealm.key-manager: WFLYDM0018: Unable to start service
	at org.jboss.as.domain.management.security.AbstractKeyManagerService.start(AbstractKeyManagerService.java:85)
	at org.jboss.msc.service.ServiceControllerImpl$StartTask.startService(ServiceControllerImpl.java:1963)
	at org.jboss.msc.service.ServiceControllerImpl$StartTask.run(ServiceControllerImpl.java:1896)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
	at java.lang.Thread.run(Thread.java:745)
Caused by: java.security.KeyStoreException: FIPS mode: KeyStore must be from provider SunPKCS11-nss-fips
	at sun.security.ssl.KeyManagerFactoryImpl$SunX509.engineInit(KeyManagerFactoryImpl.java:67)
	at javax.net.ssl.KeyManagerFactory.init(KeyManagerFactory.java:256)
	at org.jboss.as.domain.management.security.AbstractKeyManagerService.createKeyManagers(AbstractKeyManagerService.java:130)
	at org.jboss.as.domain.management.security.AbstractKeyManagerService.start(AbstractKeyManagerService.java:83)
	... 5 more

This message will appear if you have any existing key managers configured, such as the default key manager in legacy core management authentication, that do not use FIPS 140-2 compliant cryptography.

2.2.11.2. Configure the Management CLI for FIPS 140-2 Compliant Cryptography for SSL/TLS

You must configure the JBoss EAP management CLI to work in an environment with FIPS 140-2 compliant cryptography for SSL/TLS enabled. By default, if you try to use the management CLI in such an environment, the following exception is thrown: org.jboss.as.cli.CliInitializationException: java.security.KeyManagementException: FIPS mode: only SunJSSE TrustManagers may be used.

  • If you are using the legacy security subsystem:

    Update the javax.net.ssl.keyStore and javax.net.ssl.trustStore system properties in the jboss-cli.sh file, as shown below:

    JAVA_OPTS="$JAVA_OPTS -Djavax.net.ssl.trustStore=NONE -Djavax.net.ssl.trustStoreType=PKCS11"
    JAVA_OPTS="$JAVA_OPTS -Djavax.net.ssl.keyStore=NONE -Djavax.net.ssl.keyStoreType=PKCS11 -Djavax.net.ssl.keyStorePassword=P@ssword123"
  • If you are using the elytron subsystem:

    1. Create an XML configuration file for the management CLI with the following contents:

      Example: cli-wildfly-config.xml

      <configuration>
        <authentication-client xmlns="urn:elytron:1.0.1">
          <key-stores>
            <key-store name="truststore" type="PKCS11">
              <key-store-clear-password password="P@ssword123"/>
            </key-store>
          </key-stores>
          <ssl-contexts>
            <ssl-context name="client-cli-context">
              <trust-store key-store-name="truststore"/>
              <cipher-suite selector="${cipher.suite.filter}"/>
              <protocol names="TLSv1.1"/>
            </ssl-context>
          </ssl-contexts>
          <ssl-context-rules>
            <rule use-ssl-context="client-cli-context"/>
          </ssl-context-rules>
        </authentication-client>
      </configuration>

      Note

      If you are using the IBM JDK, see the IBM management CLI configuration example for the specific configuration required.

    2. When starting the management CLI, pass the configuration file to the management CLI script using the -Dwildfly.config.url property. For example:

      $ jboss-cli.sh -Dwildfly.config.url=cli-wildfly-config.xml

2.2.11.3. Configure the Elytron and Undertow Subsystems

  1. Add the FIPS 140-2 compliant cryptography key-store, key-manager and ssl-context.

    /subsystem=elytron/key-store=fipsKS:add(type=PKCS11,provider="SunPKCS11-nss-fips",credential-reference={clear-text="P@ssword123"})
    
    /subsystem=elytron/key-manager=fipsKM:add(key-store=fipsKS,algorithm="SunX509",provider=SunPKCS11-nss-fips,credential-reference={clear-text="P@ssword123"})
    
    /subsystem=elytron/server-ssl-context=fipsSSC:add(key-manager=fipsKM,protocols=["TLSv1.1"])
  2. Update the undertow subsystem to use the new ssl-context.

    Note

    https-listener must always have either a security-realm or ssl-context configured. When changing between the two configurations, the commands must be executed as a single batch, as shown below.

    batch
    /subsystem=undertow/server=default-server/https-listener=https:undefine-attribute(name=security-realm)
    /subsystem=undertow/server=default-server/https-listener=https:write-attribute(name=ssl-context,value=fipsSSC)
    run-batch
    
    reload

In the elytron subsystem, OpenJDK and Oracle JDK in FIPS mode restrict the usage of any advanced features that are based on providing custom KeyManager or TrustManager implementations. The following configuration attributes do not work:

  • On the client:

    • ssl-context.key-store-ssl-certificate
  • On the server:

    • server-ssl-context.security-domain
    • trust-manager.certificate-revocation-list

2.2.11.4. Configure Undertow with the Legacy Core Management Authentication

Optionally, you can still use the legacy core management authentication instead of the elytron subsystem to complete the setup of FIPS 140-2 compliant cryptography for SSL/TLS:

  1. Configure Undertow to use SSL/TLS.

    Note

    The following commands below must either be run in batch mode, or the server must be reloaded after adding the ssl server identity. The example below is shown using batch mode.

    batch
    
    /core-service=management/security-realm=HTTPSRealm:add
    
    /core-service=management/security-realm=HTTPSRealm/server-identity=ssl:add(keystore-provider=PKCS11, keystore-password="strongP@ssword1")
    
    /subsystem=undertow/server=default-server/https-listener=https:add(socket-binding=https, security-realm=HTTPSRealm, enabled-protocols="TLSv1.1")
    
    run-batch

    The basic details for configuring Undertow to SSL/TLS are covered in Setting up an SSL/TLS for Applications.

  2. Configure the cipher suites used by Undertow.

    Once you have SSL/TLS configured, you need to configure the https listener and security realm to have a specific set of cipher suites enabled:

    Required Cipher Suites

    SSL_RSA_WITH_3DES_EDE_CBC_SHA, SSL_DHE_RSA_WITH_3DES_EDE_CBC_SHA, TLS_RSA_WITH_AES_128_CBC_SHA, TLS_DHE_DSS_WITH_AES_128_CBC_SHA, TLS_DHE_RSA_WITH_AES_128_CBC_SHA, TLS_RSA_WITH_AES_256_CBC_SHA, TLS_DHE_DSS_WITH_AES_256_CBC_SHA, TLS_DHE_RSA_WITH_AES_256_CBC_SHA, TLS_ECDH_ECDSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_ECDSA_WITH_AES_128_CBC_SHA, TLS_ECDH_ECDSA_WITH_AES_256_CBC_SHA, TLS_ECDHE_ECDSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA, TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA, TLS_ECDH_RSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_RSA_WITH_AES_128_CBC_SHA, TLS_ECDH_RSA_WITH_AES_256_CBC_SHA, TLS_ECDHE_RSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA, TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA, TLS_ECDH_anon_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_anon_WITH_AES_128_CBC_SHA, TLS_ECDH_anon_WITH_AES_256_CBC_SHA

    The basics behind enabling cipher suites for the https listener are covered in About Cipher Suites. To enable cipher suites on the https listener:

    Example Command for Enabling Cipher Suites on the Https Listener

    /subsystem=undertow/server=default-server/https-listener=https:write-attribute(name=enabled-cipher-suites,value="SSL_RSA_WITH_3DES_EDE_CBC_SHA,SSL_DHE_RSA_WITH_3DES_EDE_CBC_SHA,TLS_RSA_WITH_AES_128_CBC_SHA,TLS_DHE_DSS_WITH_AES_128_CBC_SHA,TLS_DHE_RSA_WITH_AES_128_CBC_SHA,TLS_RSA_WITH_AES_256_CBC_SHA,TLS_DHE_DSS_WITH_AES_256_CBC_SHA,TLS_DHE_RSA_WITH_AES_256_CBC_SHA,TLS_ECDH_ECDSA_WITH_3DES_EDE_CBC_SHA,TLS_ECDH_ECDSA_WITH_AES_128_CBC_SHA,TLS_ECDH_ECDSA_WITH_AES_256_CBC_SHA,TLS_ECDHE_ECDSA_WITH_3DES_EDE_CBC_SHA,TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA,TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA,TLS_ECDH_RSA_WITH_3DES_EDE_CBC_SHA,TLS_ECDH_RSA_WITH_AES_128_CBC_SHA,TLS_ECDH_RSA_WITH_AES_256_CBC_SHA,TLS_ECDHE_RSA_WITH_3DES_EDE_CBC_SHA,TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA,TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA,TLS_ECDH_anon_WITH_3DES_EDE_CBC_SHA,TLS_ECDH_anon_WITH_AES_128_CBC_SHA,TLS_ECDH_anon_WITH_AES_256_CBC_SHA")

  3. Enable cipher suites on the security realm.

    Example Command for Enabling Cipher Suites on the Security Realm

    /core-service=management/security-realm=HTTPSRealm/server-identity=ssl:write-attribute(name=enabled-cipher-suites, value=[SSL_RSA_WITH_3DES_EDE_CBC_SHA, SSL_DHE_RSA_WITH_3DES_EDE_CBC_SHA, TLS_RSA_WITH_AES_128_CBC_SHA, TLS_DHE_DSS_WITH_AES_128_CBC_SHA, TLS_DHE_RSA_WITH_AES_128_CBC_SHA, TLS_RSA_WITH_AES_256_CBC_SHA, TLS_DHE_DSS_WITH_AES_256_CBC_SHA, TLS_DHE_RSA_WITH_AES_256_CBC_SHA, TLS_ECDH_ECDSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_ECDSA_WITH_AES_128_CBC_SHA, TLS_ECDH_ECDSA_WITH_AES_256_CBC_SHA, TLS_ECDHE_ECDSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA, TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA, TLS_ECDH_RSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_RSA_WITH_AES_128_CBC_SHA, TLS_ECDH_RSA_WITH_AES_256_CBC_SHA, TLS_ECDHE_RSA_WITH_3DES_EDE_CBC_SHA, TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA, TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA, TLS_ECDH_anon_WITH_3DES_EDE_CBC_SHA, TLS_ECDH_anon_WITH_AES_128_CBC_SHA, TLS_ECDH_anon_WITH_AES_256_CBC_SHA])

2.2.12. FIPS 140-2 Compliant Cryptography on IBM JDK

On the IBM JDK, the IBM Java Cryptographic Extension (JCE) IBMJCEFIPS provider and the IBM Java Secure Sockets Extension (JSSE) FIPS 140-2 Cryptographic Module (IBMJSSE2) for multi-platforms provide FIPS 140-2 compliant cryptography.

For more information on the IBMJCEFIPS provider, see the IBM Documentation for IBM JCEFIPS and NIST IBMJCEFIPS – Security Policy. For more information on IBMJSSE2, see Running IBMJSSE2 in FIPS mode.

2.2.12.1. Key Storage

The IBM JCE does not provide a keystore. The keys are stored on the computer and do not leave its physical boundary. If the keys are moved between computers they must be encrypted.

To run keytool in FIPS-compliant mode use the -providerClass option on each command like this:

keytool -list -storetype JCEKS -keystore mystore.jck -storepass mystorepass -providerClass com.ibm.crypto.fips.provider.IBMJCEFIPS

2.2.12.2. Management CLI Configuration

To configure the management CLI for FIPS 140-2 compliant cryptography on the IBM JDK, you must use a management CLI configuration file specifically for the IBM JDK, such as the following:

Example: cli-wildfly-config-ibm.xml

<configuration>
  <authentication-client xmlns="urn:elytron:1.0.1">
    <key-stores>
      <key-store name="truststore" type="JKS">
        <file name="/path/to/truststore"/>
        <key-store-clear-password password="P@ssword123"/>
      </key-store>
    </key-stores>
    <ssl-contexts>
      <ssl-context name="client-cli-context">
        <trust-store key-store-name="truststore"/>
        <cipher-suite selector="${cipher.suite.filter}"/>
        <protocol names="TLSv1"/>
      </ssl-context>
    </ssl-contexts>
    <ssl-context-rules>
      <rule use-ssl-context="client-cli-context"/>
    </ssl-context-rules>
  </authentication-client>
</configuration>

2.2.12.3. Examine FIPS Provider Information

To examine information about the IBMJCEFIPS used by the server, enable debug-level logging by adding -Djavax.net.debug=true to the standalone.conf or domain.conf files. Information about the FIPS provider is logged to the server.log file, for example:

04:22:45,685 INFO  [stdout] (http-/127.0.0.1:8443-1) JsseJCE:  Using MessageDigest SHA from provider IBMJCEFIPS version 1.7
04:22:45,689 INFO  [stdout] (http-/127.0.0.1:8443-1) DHCrypt:  DH KeyPairGenerator  from provider from init IBMJCEFIPS version 1.7
04:22:45,754 INFO  [stdout] (http-/127.0.0.1:8443-1) JsseJCE:  Using KeyFactory DiffieHellman from provider IBMJCEFIPS version 1.7
04:22:45,754 INFO  [stdout] (http-/127.0.0.1:8443-1) JsseJCE:  Using KeyAgreement DiffieHellman from provider IBMJCEFIPS version 1.7
04:22:45,754 INFO  [stdout] (http-/127.0.0.1:8443-1) DHCrypt:  DH KeyAgreement  from provider IBMJCEFIPS version 1.7
04:22:45,754 INFO  [stdout] (http-/127.0.0.1:8443-1) DHCrypt:  DH KeyAgreement  from provider from initIBMJCEFIPS version 1.7

2.2.13. Starting a Managed Domain when the JVM is Running in FIPS Mode

Important

It is assumed you have a managed domain, FIPS configured, as well as all necessary certificates configured. This includes having imported the domain controller’s certificate into each controller’s truststore. For more details on configuring a managed domain, see the Domain Management section in the JBoss EAP Configuration Guide. For more details on configuring FIPS, see Enable FIPS 140-2 Cryptography for SSL/TLS on Red Hat Enterprise Linux 7.

You need to update each host controller and the master domain controller to use SSL/TLS for communication.

Warning

Red Hat recommends that SSLv2, SSLv3, and TLSv1.0 be explicitly disabled in favor of TLSv1.1 in all affected packages.

  1. Create an SSL/TLS security realm on the master domain controller.

    You need to create an SSL/TLS security realm on the master domain controller configured to use your NSS database as a PKCS11 provider.

    Example: Security Realm on the Master Domain Controller

    <security-realm name="HTTPSRealm">
        <server-identities>
            <ssl>
                <engine enabled-protocols="TLSv1.1"/>
                <keystore provider="PKCS11" keystore-password="strongP@ssword1"/>
            </ssl>
        </server-identities>
        <authentication>
            <local default-user="\$local"/>
            <properties path="https-users.properties" relative-to="jboss.domain.config.dir"/>
        </authentication>
    </security-realm>

  2. Create an SSL/TLS security realm on each host controller.

    You need to create a security realm with an SSL/TLS truststore for authentication.

    Example: Security Realm on Each Host Controller

    <security-realm name="HTTPSRealm">
      <authentication>
        <truststore provider="PKCS11" keystore-password="strongP@ssword1"/>
      </authentication>
    </security-realm>

    Note

    You need to repeat this process on each host.

  3. Secure the native interface on the master domain controller.

    You need to ensure that the native interface on the master domain controller is secured with the security realm you just created.

    Example: Native Interface

    <management-interfaces>
      ...
      <native-interface security-realm="HTTPSRealm">
        <socket interface="management" port="${jboss.management.native.port:9999}"/>
       </native-interface>
    </management-interfaces>

  4. Use the SSL/TLS realm on each host controller to connect to the master domain controller.

    You need to update the security realm used for connecting to the master domain controller. This change must be done directly in the host controller’s configuration file, for example host.xml or host-slave.xml, while the server is not running.

    Example: Host Controller Configuration File

    <domain-controller>
      <remote security-realm="HTTPSRealm">
        <discovery-options>
          <static-discovery name="primary" protocol="${jboss.domain.master.protocol:remote}" host="${jboss.domain.master.address}" port="${jboss.domain.master.port:9999}"/>
        </discovery-options>
      </remote>
    </domain-controller>

  5. Update how each server connects back to its host controller.

    You also need to update how each server connects back to its host controller.

    Example: Server Configuration

    <server name="my-server" group="my-server-group">
      <ssl ssl-protocol="TLS" trust-manager-algorithm="SunX509" truststore-type="PKCS11" truststore-password="strongP@ssword1"/>
    </server>

  6. Configure two-way SSL/TLS in a managed domain.

    To enable two-way SSL/TLS, add a truststore authentication method to the SSL/TLS security realm for the master domain controller, execute the following management CLI commands:

    /host=master/core-service=management/security-realm=HTTPSRealm/authentication=truststore:add(keystore-provider="PKCS11",keystore-password="strongP@ssword1")
    
    reload --host=master

    You also need to update each host controller’s security realm to have an SSL server identity, execute the following management CLI commands:

    /host=host1/core-service=management/security-realm=HTTPSRealm/server-identity=ssl:add(keystore-provider=PKCS11, keystore-password="strongP@ssword1",enabled-protocols=["TLSv1.1"])
    
    reload --host=host1
    Important

    You also need to ensure that each host’s certificate is imported into the domain controller’s truststore.

2.3. Security Auditing

Security auditing refers to triggering events, such as writing to a log, in response to an event that happens within the security subsystem or the management interfaces. Auditing mechanisms are configured as part of a security domain or management interface.

Auditing uses provider modules. Both included provider modules as well as custom implementations may be used.

2.3.1. Configure Security Auditing for the Legacy Security Domains

To configure security auditing settings for a security domain, the following steps must be performed from the management console:

  1. Open the security domain’s detailed view.

    1. Click Configuration at the top of the screen.
    2. In a managed domain, select a profile to modify from the Profile selection box at the top left.
    3. Click on Subsystems, then Security.
    4. Click on the security domain to edit and click View.
  2. Navigate to the auditing configuration.

    Click on Audit on the left side of the screen.

    The configuration area is divided into two areas: Provider Modules and Details. The provider module is the basic unit of configuration. A security domain can include several provider modules each of which can include attributes and options.

  3. Add a provider module.

    Click Add and fill in the Code section with the class name of the provider module. Also fill in the Name section with the desired name.

  4. Verify that the module is working.

    The goal of an audit module is to provide a way to monitor the events in the security subsystem. This monitoring can be done by means of writing to a log file, email notifications, or any other measurable auditing mechanism.

    For example, JBoss EAP includes the org.jboss.security.audit.providers.LogAuditProvider module by default. If enabled following the steps above, this audit module writes security notifications to an audit.log file in the log subfolder within the EAP_HOME directory.

    To verify if the steps above have worked in the context of the org.jboss.security.audit.providers.LogAuditProvider, perform an action that is likely to trigger a notification and then check the audit log file.

  5. Optional: Add, edit, or remove module options.

    To add options to your module, click its entry in the Modules list, and select the Module Options tab in the Details section of the page. Click Add, and provide the key and value for the option.

    To edit an option that already exists, click Remove to remove it, and click Add to add it again with the correct options.

2.4. Configure One-way and Two-way SSL/TLS for Applications

2.4.1. Automatic Self-signed Certificate Creation for Applications

When using the legacy security realms, JBoss EAP provides automatic generation of self-signed certificate for development purposes.

Example: Server Log Showing Self-signed Certificate Creation

15:26:09,031 WARN  [org.jboss.as.domain.management.security] (MSC service thread 1-7) WFLYDM0111: Keystore /path/to/jboss/standalone/configuration/application.keystore not found, it will be auto generated on first use with a self signed certificate for host localhost
...
15:26:10,076 WARN  [org.jboss.as.domain.management.security] (MSC service thread 1-2) WFLYDM0113: Generated self signed certificate at /path/to/jboss/configuration/application.keystore. Please note that self signed certificates are not secure, and should only be used for testing purposes. Do not use this self signed certificate in production.
SHA-1 fingerprint of the generated key is 00:11:22:33:44:55:66:77:88:99:aa:bb:cc:dd:ee:ff:00:11:22:33
SHA-256 fingerprint of the generated key is 00:11:22:33:44:55:66:77:88:99:00:aa:bb:cc:dd:ee:ff:00:11:22:33:44:55:66:77:88:99:aa:bb:cc:dd:ee
...

This certificate is created for testing purposes and is assigned to the HTTPS interface used by applications. The keystore containing the certificate will be generated if the file does not exist the first time the HTTPS interface is accessed.

Example: Default ApplicationRealm Using the Self-signed Certificate

<security-realm name="ApplicationRealm">
  <server-identities>
    <ssl>
      <keystore path="application.keystore" relative-to="jboss.server.config.dir" keystore-password="password" alias="server" key-password="password" generate-self-signed-certificate-host="localhost"/>
    </ssl>
  </server-identities>
  ...
</security-realm>

Example: Default HTTPS Interface Configuration

<subsystem xmlns="urn:jboss:domain:undertow:4.0">
    ...
    <server name="default-server">
        ...
        <https-listener name="https" socket-binding="https" security-realm="ApplicationRealm" enable-http2="true"/>
        <host name="default-host" alias="localhost">
        ...

Note

If you want to disable the self-signed certificate creation, you will need to remove the generate-self-signed-certificate-host="localhost" from the server keystore configuration. The generate-self-signed-certificate-host attribute holds the host name for which the self-signed certificate should be generated.

Warning

This self-signed certificate is intended for testing purposes only and is not intended for use in production environments. For more information on configuring SSL/TLS for applications with Elytron, see the Enable One-way SSL/TLS for Applications using the Elytron Subsystem and Enable Two-way SSL/TLS for Applications using the Elytron Subsystem sections. For more information on configuring SSL/TLS for applications with legacy security, see the Enable One-way SSL/TLS for Applications Using Legacy Security Realms and Enable Two-way SSL/TLS for Applications Using Legacy Security Realms sections.

2.4.2. Using Elytron

2.4.2.1. Enable One-way SSL/TLS for Applications Using the Elytron Subsystem

In JBoss EAP, you can use the elytron subsystem, along with the undertow subsystem, to enable one-way SSL/TLS for deployed applications.

  1. Obtain or generate your keystore.

    Before enabling one-way SSL/TLS in JBoss EAP, you must obtain or generate the keystore you plan on using. To generate an example keystore in Red Hat Enterprise Linux 7:

    $ keytool -genkeypair -alias localhost -keyalg RSA -keysize 1024 -validity 365 -keystore /path/to/keystore.jks -dname "CN=localhost" -keypass secret -storepass secret
  2. Configure a key-store in JBoss EAP.

    /subsystem=elytron/key-store=httpsKS:add(path=/path/to/keystore.jks, credential-reference={clear-text=secret}, type=JKS)
    Note

    The previous command uses an absolute path to the keystore. Alternatively you can use the relative-to attribute to specify the base directory variable and path attribute to specify a relative path.

    /subsystem=elytron/key-store=httpsKS:add(path=keystore.jks, relative-to=jboss.server.config.dir, credential-reference={clear-text=secret}, type=JKS)
  3. Configure a key-manager that references your key-store.

    /subsystem=elytron/key-manager=httpsKM:add(key-store=httpsKS, algorithm="SunX509", credential-reference={clear-text=secret})
    Important

    You need to know what key manager algorithms are provided by the JDK you are using. For example, a JDK that uses SunJSSE provides the PKIX and SunX509 algorithms.

    The example command above uses SunX509 for the key manager algorithm.

  4. Configure a server-ssl-context that references your key-manager.

    /subsystem=elytron/server-ssl-context=httpsSSC:add(key-manager=httpsKM, protocols=["TLSv1.2"])
    Important

    You need to determine what SSL/TLS protocols you want to support. The example command above uses TLSv1.2. You can use the cipher-suite-filter argument to specify which cipher suites are allowed, and the use-cipher-suites-order argument to honor server cipher suite order. The use-cipher-suites-order attribute by default is set to true. This differs from the legacy security subsystem behavior, which defaults to honoring client cipher suite order.

    Warning

    Red Hat recommends that SSLv2, SSLv3, and TLSv1.0 be explicitly disabled in favor of TLSv1.1 or TLSv1.2 in all affected packages.

  5. Check and see if the https-listener is configured to use a legacy security realm for its SSL configuration.

    /subsystem=undertow/server=default-server/https-listener=https:read-attribute(name=security-realm)
    {
        "outcome" => "success",
        "result" => "ApplicationRealm"
    }

    The above command shows that the https-listener is configured to use the ApplicationRealm legacy security realm for its SSL configuration. Undertow cannot reference both a legacy security realm and an ssl-context in Elytron at the same time so you must remove the reference to the legacy security realm.

    Note

    If the result is undefined, you do not need to remove the reference to the security realm in the next step.

  6. Remove the reference to the legacy security realm, and update the https-listener to use the ssl-context from Elytron.

    Note

    https-listener must always have either a security-realm or ssl-context configured. When changing between the two configurations, the commands must be executed as a single batch, as shown below.

    batch
    /subsystem=undertow/server=default-server/https-listener=https:undefine-attribute(name=security-realm)
    /subsystem=undertow/server=default-server/https-listener=https:write-attribute(name=ssl-context, value=httpsSSC)
    run-batch
  7. Reload the server.

    reload

One-way SSL/TLS is now enabled for applications.

2.4.2.2. Enable Two-way SSL/TLS for Applications Using the Elytron Subsystem

In JBoss EAP, you can use the elytron subsystem, along with the undertow subsystem, to enable two-way SSL/TLS for deployed applications.

  1. Obtain or generate your keystore.

    Before enabling two-way SSL/TLS in JBoss EAP, you must obtain or generate the keystores, truststores and certificates you plan on using. To generate an example set of keystores, truststores, and certificates in Red Hat Enterprise Linux 7:

    Example: Server and Client Keystores:

    $ keytool -genkeypair -alias localhost -keyalg RSA -keysize 1024 -validity 365 -keystore server.keystore.jks -dname "CN=localhost" -keypass secret -storepass secret
    
    $ keytool -genkeypair -alias client -keyalg RSA -keysize 1024 -validity 365 -keystore client.keystore.jks -dname "CN=client" -keypass secret -storepass secret

    Example: Export the Server and Client Certificates:

    $ keytool -exportcert -keystore server.keystore.jks -alias localhost -keypass secret -storepass secret -file server.cer
    
    $ keytool -exportcert -keystore client.keystore.jks -alias client -keypass secret -storepass secret -file client.cer

    Example: Import the Server and Client Certificates Into the Opposing Truststores:

    $ keytool -importcert -keystore server.truststore.jks -storepass secret -alias client -trustcacerts -file client.cer
    
    $ keytool -importcert -keystore client.truststore.jks -storepass secret -alias localhost -trustcacerts -file server.cer

  2. Configure a key-store for server keystore and truststore in JBoss EAP.

    /subsystem=elytron/key-store=twoWayKS:add(path=/path/to/server.keystore.jks, credential-reference={clear-text=secret}, type=JKS)
    
    /subsystem=elytron/key-store=twoWayTS:add(path=/path/to/server.truststore.jks, credential-reference={clear-text=secret}, type=JKS)
    Note

    The previous command uses an absolute path to the keystore. Alternatively you can use the relative-to attribute to specify the base directory variable and path attribute to specify a relative path.

    /subsystem=elytron/key-store=myKS:add(path=keystore.jks, relative-to=jboss.server.config.dir, credential-reference={clear-text=secret}, type=JKS)
  3. Configure a key-manager that references your keystore key-store.

    /subsystem=elytron/key-manager=twoWayKM:add(key-store=twoWayKS, algorithm="SunX509", credential-reference={clear-text=secret})
    Important

    You need to know what key manager algorithms are provided by the JDK you are using. For example, a JDK that uses SunJSSE provides the PKIX and SunX509 algorithms.

    The example command below uses SunX509 for the key manager algorithm.

  4. Configure a trust-manager that references your truststore key-store.

    /subsystem=elytron/trust-manager=twoWayTM:add(key-store=twoWayTS, algorithm="SunX509")
    Important

    You need to know what key manager algorithms are provided by the JDK you are using. For example, a JDK that uses SunJSSE provides the PKIX and SunX509 algorithms.

    The example command above uses SunX509 for the key manager algorithm.

  5. Configure a server-ssl-context that references your key-manager, trust-manager, and enables client authentication:

    /subsystem=elytron/server-ssl-context=twoWaySSC:add(key-manager=twoWayKM, protocols=["TLSv1.2"], trust-manager=twoWayTM, need-client-auth=true)
    Important

    You need to determine what SSL/TLS protocols you want to support. The example command above uses TLSv1.2. You can use the cipher-suite-filter argument to specify which cipher suites are allowed, and the use-cipher-suites-order argument to honor server cipher suite order. The use-cipher-suites-order attribute by default is set to true. This differs from the legacy security subsystem behavior, which defaults to honoring client cipher suite order.

    Warning

    Red Hat recommends that SSLv2, SSLv3, and TLSv1.0 be explicitly disabled in favor of TLSv1.1 or TLSv1.2 in all affected packages.

  6. Check and see if the https-listener is configured to use a legacy security realm for its SSL configuration.

    /subsystem=undertow/server=default-server/https-listener=https:read-attribute(name=security-realm)
    {
        "outcome" => "success",
        "result" => "ApplicationRealm"
    }

    The above command shows that the https-listener is configured to use the ApplicationRealm legacy security realm for its SSL configuration. Undertow cannot reference both a legacy security realm and an ssl-context in the elytron subsystem at the same time. So you must remove the reference to the legacy security realm.

    Note

    If the result is undefined, you do not need to remove the reference to the security realm in the next step.

  7. Remove the reference to the legacy security realm, and update the https-listener to use the ssl-context from Elytron.

    Note

    https-listener must always have either a security-realm or ssl-context configured. When changing between the two configurations, the commands must be executed as a single batch, as shown below.

    batch
    /subsystem=undertow/server=default-server/https-listener=https:undefine-attribute(name=security-realm)
    /subsystem=undertow/server=default-server/https-listener=https:write-attribute(name=ssl-context, value=twoWaySSC)
    run-batch
  8. Reload the server.

    reload
  9. Configure your client to use the client certificate.

    You need to configure your client to present the trusted client certificate to the server to complete the two-way SSL/TLS authentication. For example, if using a browser, you need to import the trusted certificate into the browser’s trust store.

    This procedure forces a two-way SSL/TLS but it does not change the original authentication method of the application.

    If you want to change the original authentication method, see Configure Authentication with Certificates in How to Configure Identity Management for JBoss EAP.

Two-way SSL/TLS is now enabled for applications.

2.4.3. Using Legacy Security Realms

Important

As a prerequisite, an SSL/TLS encryption key and certificate should be created and placed in an accessible directory. Additionally, relevant information, such as keystore aliases and passwords, desired cipher suites, should also be accessible. For examples on generating SSL/TLS Keys and Certificates, see the first two steps in the Setting up Two-way SSL/TLS for the Management Interfaces section. For more information about the HTTPS listener, including cipher suites, see the HTTPS Listener Reference section.

2.4.3.1. Enable One-way SSL/TLS for Applications Using Legacy Security Realms

This example assumes that the keystore, identity.jks, has been copied to the server configuration directory and configured with the given properties. Administrators should substitute their own values for the example ones.

Note

The management CLI commands shown assume that you are running a JBoss EAP standalone server. For more details on using the management CLI for a JBoss EAP managed domain, see the JBoss EAP Management CLI Guide.

  1. Add and configure an HTTPS security realm first. Once the HTTPS security realm has been configured, configure an https-listener in the undertow subsystem that references the security realm:

    batch
    
    /core-service=management/security-realm=HTTPSRealm:add
    
    /core-service=management/security-realm=HTTPSRealm/server-identity=ssl:add(keystore-path=identity.jks, keystore-relative-to=jboss.server.config.dir, keystore-password=password1, alias=appserver)
    
    /subsystem=undertow/server=default-server/https-listener=https:write-attribute(name=security-realm, value=HTTPSRealm)
    
    run-batch
    Warning

    Red Hat recommends that SSLv2, SSLv3, and TLSv1.0 be explicitly disabled in favor of TLSv1.1 or TLSv1.2 in all affected packages.

  2. Restart the JBoss EAP instance for the changes to take effect.

2.4.3.2. Enable Two-way SSL/TLS for Applications Using Legacy Security Realms

Setting up two-way SSL/TLS for applications follows many of the same procedures outlined in Setting up Two-way SSL/TLS for the Management Interfaces. To set up two-way SSL/TLS for applications, you need to do the following:

  1. Generate the stores for both the client and server
  2. Export the certificates for both the client and server
  3. Import the certificates into the opposing truststores
  4. Define a security realm, for example CertificateRealm, on the server that uses the server’s keystore and truststore
  5. Update the undertow subsystem to use the security realm and require client verification

The first four steps are covered in Setting up Two-way SSL/TLS for the Management Interfaces.

Important

If the server has not been reloaded since the new security realm has been added, you must reload the server before performing the next step.

Update the Undertow Subsystem

Once the keystores, certificates, truststores, and security realms have been created and configured, you need to add an HTTPS listener to the undertow subsystem, use the security realm you created, and require client verification:

/subsystem=undertow/server=default-server/https-listener=https:write-attribute(name=security-realm, value=CertificateRealm)

/subsystem=undertow/server=default-server/https-listener=https:write-attribute(name=verify-client, value=REQUIRED)
Important

You must reload the server for these changes to take effect.

Important

Any client connecting to a JBoss EAP instance with two-way SSL/TLS enabled for applications must have access to a client certificate or keystore, in other words a client keystore whose certificate is included in the server’s truststore. If the client is using a browser to connect to the JBoss EAP instance, you need to import that certificate or keystore into the browser’s certificate manager.

Note

More details on using certificate-based authentication in applications, in addition to two-way SSL/TLS with applications, can be found in the Configuring a Security Domain to Use Certificate-based Authentication section of the JBoss EAP How to Configure Identity Management Guide.

2.5. SASL Authentication Mechanisms

Simple Authentication and Security Layer (SASL) authentication mechanisms are used for defining the mechanisms for authenticating connections to a JBoss EAP server using the elytron subsystem, and for clients connecting to servers. Clients can be other JBoss EAP instances, or Elytron Client. SASL authentication mechanisms in JBoss EAP are also significantly used in Elytron integration with the remoting subsystem.

2.5.1. Choosing SASL Authentication Mechanisms

Note

Although JBoss EAP and Elytron Client work with a variety of SASL authentication mechanisms, you must ensure that the mechanisms you use are supported. See this list for the support levels for SASL authentication mechanisms.

The authentication mechanisms you use depends on your environment and desired authentication method. The following list summarizes the use of some of the supported SASL authentication mechanisms:

ANONYMOUS
Unauthenticated guest access.
DIGEST-MD5
Uses HTTP digest authentication as a SASL mechanism.
EXTERNAL
Uses authentication credentials that are implied in the context of the request. For example, IPsec or TLS authentication.
Mechanisms beginning with GS
Authentication using Kerberos.
JBOSS-LOCAL-USER
Provides authentication by testing that the client has the same file access as the local user that is running the JBoss EAP server. This is useful for other JBoss EAP instances running on the same machine.
OAUTHBEARER
Uses authentication provided by OAuth as a SASL mechanism.
PLAIN
Plain text username and password authentication.
Mechanisms beginning with SCRAM
Salted Challenge Response Authentication Mechanism (SCRAM) that uses a specified hashing function.
Mechanisms ending with -PLUS
Indicates a channel binding variant of a particular authentication mechanism. You should use these variants when the underlying connection uses SSL/TLS.

For more information on individual SASL authentication mechanisms, see the IANA SASL mechanism list and individual mechanism memos.

2.5.2. Configuring SASL Authentication Mechanisms on the Server Side

Configuring SASL authentication mechanisms on the server side is done using SASL authentication factories.

There are two levels of configuration required:

  • A sasl-authentication-factory, where you specify authentication mechanisms.
  • A configurable-sasl-server-factory that aggregates one or more of sasl-authentication-factory, and configures mechanism properties as well as optionally applying filters to enable or disable certain authentication mechanisms.

The following example demonstrates creating a new configurable-sasl-server-factory, and a sasl-authentication-factory that uses DIGEST-MD5 as a SASL authentication mechanism for application clients.

/subsystem=elytron/configurable-sasl-server-factory=mySASLServerFactory:add(sasl-server-factory=elytron)

/subsystem=elytron/sasl-authentication-factory=mySASLAuthFactory:add(sasl-server-factory=mySASLServerFactory,security-domain=ApplicationDomain,mechanism-configurations=[{mechanism-name=DIGEST-MD5,mechanism-realm-configurations=[{realm-name=ApplicationRealm}]}])

2.5.3. Specifying SASL Authentication Mechanisms on the Client Side

SASL authentication mechanisms used by a client are specified using a sasl-mechanism-selector. You can specify any supported SASL authentication mechanisms that are exposed by the server that the client is connecting to.

A sasl-mechanism-selector is defined in Elytron configurations where authentication is configured:

  • In the elytron subsystem, this is an attribute of an authentication-configuration. For example:

    /subsystem=elytron/authentication-configuration=myAuthConfig:write-attribute(name=sasl-mechanism-selector,value="DIGEST-MD5")

    An example of using an authentication-configuration with a sasl-mechanism-selector can be seen in Configuring SSL/TLS Between Domain and Host Controllers Using Elytron.

  • For Elytron Client, it is specified under the configuration element of authentication-configurations in the client configuration file, usually named wildfly-config.xml. For example:

    <configuration>
      <authentication-client xmlns="urn:elytron:1.0.1">
        <authentication-rules>
          <rule use-configuration="default" />
        </authentication-rules>
        <authentication-configurations>
          <configuration name="default">
            <sasl-mechanism-selector selector="#ALL" />
          ...
          </configuration>
        </authentication-configurations>
      </authentication-client>
    </configuration>

See How to Configure Identity Management for more information on configuring client authentication with Elytron Client.

sasl-mechanism-selector Grammar

The selector string for sasl-mechanism-selector has a specific grammar.

In a simple form, individual mechanisms are specified by listing their names in order, separated by a spaces. For example, to specify DIGEST-MD5, SCRAM-SHA-1, and SCRAM-SHA-256 as allowed authentication mechanisms, use the following string: DIGEST-MD5 SCRAM-SHA-1 SCRAM-SHA-256.

More advanced usage of the grammar can use the following special tokens:

  • #ALL: All mechanisms.
  • #FAMILY(NAME): Mechanisms belonging to the specified mechanism family. For example, the family could be DIGEST, EAP, GS2, SCRAM, or IEC-ISO-9798.
  • #PLUS: Mechanisms that use channel binding. For example, SCRAM-SHA-XXX-PLUS or GS2-XXX-PLUS.
  • #MUTUAL: Mechanisms that authenticate the server in some way, for example making the server prove that the server knows the password. #MUTUAL includes families such as #FAMILY(SCRAM) and #FAMILY(GS2).
  • #HASH(ALGORITHM): Mechanisms that use the specified hash algorithm. For example, the algorithm could be MD5, SHA-1, SHA-256, SHA-384, or SHA-512.

The above tokens and names can also be used with the following operations and predicates:

  • -: Forbids
  • !: Inverts
  • &&: And
  • ||: Or
  • ==: Equals
  • ?: If
  • #TLS: Is true when TLS is active, otherwise false.

Below are some examples of sasl-mechanism-selector strings and their meaning:

  • #TLS && !#MUTUAL: When TLS is active, all mechanisms without mutual authentication.
  • #ALL -ANONYMOUS: All mechanisms, except for ANONYMOUS.
  • SCRAM-SHA-1 SCRAM-SHA-256: Adds those two mechanisms in that order.
  • (SCRAM-SHA-1 || SCRAM-SHA-256): Adds the two mechanisms in the order that the provider or server presents them.
  • !#HASH(MD5): Any mechanism that does not use the MD5 hashing algorithm.
  • #FAMILY(DIGEST): Any digest mechanism.

2.5.4. Configuring SASL Authentication Mechanism Properties

You can configure authentication mechanism properties on both the server side and on the client side.

  • On the server side, you define authentication mechanism properties in the configurable-sasl-server-factory. The following example defines the com.sun.security.sasl.digest.utf8 property with a value of false.

    /subsystem=elytron/configurable-sasl-server-factory=mySASLServerFactory:map-put(name=properties,key=com.sun.security.sasl.digest.utf8,value=false)
  • On the client side, you define authentication mechanisms properties in the client’s authentication configuration:

    • In the elytron subsystem, define the authentication mechanism properties in your authentication-configuration. The following example defines the wildfly.sasl.local-user.quiet-auth property with a value of true.

      /subsystem=elytron/authentication-configuration=myAuthConfig:map-put(name=mechanism-properties,key=wildfly.sasl.local-user.quiet-auth,value=true)
    • For Elytron Client, authentication mechanism properties are specified under the configuration element of authentication-configurations in the client configuration file, usually named wildfly-config.xml. For example:

      ...
      <authentication-configurations>
        <configuration name="default">
          <sasl-mechanism-selector selector="#ALL" />
          <set-mechanism-properties>
            <property key="wildfly.sasl.local-user.quiet-auth" value="true" />
          </set-mechanism-properties>
          ...
        </configuration>
      </authentication-configurations>
      ...

You can see a list of standard Java SASL authentication mechanism properties in the Java documentation. Other JBoss EAP-specific SASL authentication mechanism properties are listed in the Authentication Mechanisms Reference.

2.6. Elytron Integration with the ModCluster Subsystem

One of the security capabilities exposed by elytron subsystem is a client ssl-context that can be used to configure the modcluster subsystem to communicate with a load balancer using SSL/TLS.

When protecting the communication between the application server and the load balancer, you need to define a client ssl-context in order to:

  • Define a truststore holding the certificate chain that will be used to validate load balancer’s certificate.
  • Define a trust manager to perform validations against the load balancer’s certificate.

2.6.1. Defining a Client SSL Context and Configuring ModCluster Subsystem

The following procedure requires that a truststore and trust manager be configured. For information on creating these see Create an Elytron Truststore and Create an Elytron Trust Manager.

  1. Create the client SSL context.

    This SSL context is going to be used by the modcluster subsystem when connecting to the load balancer using SSL/TLS:

    /subsystem=elytron/client-ssl-context=modcluster-client-ssl-context:add(trust-manager=default-trust-manager)
  2. Reference the newly created client SSL context using one of the following options.

    • Configure the modcluster subsystem by setting the ssl-context.

      /subsystem=modcluster/mod-cluster-config=configuration:write-attribute(name=ssl-context, value=modcluster-client-ssl-context)
    • Configure the undertow subsystem by defining the ssl-context attribute of the mod-cluster filter.

      /subsystem=undertow/configuration=filter/mod-cluster=modcluster:write-attribute(name=ssl-context,value=modcluster-client-ssl-context)
  3. Reload the server.

    reload

For configuring the modcluster subsystem and using two-way authentication, along with the trust manager, the key manager also needs to be configured.

  1. Create the keystore.

    /subsystem=elytron/key-store=twoWayKS:add(path=/path/to/client.keystore.jks, credential-reference={clear-text=secret},type=JKS)
  2. Configure the key manager.

    /subsystem=elytron/key-manager=twoWayKM:add(key-store=twoWayKS, algorithm="SunX509", credential-reference={clear-text=secret})
  3. Create the client SSL context.

    /subsystem=elytron/client-ssl-context=modcluster-client-ssl-context:add(trust-manager=default-trust-manager, key-manager=twoWayKM)
    Note

    If you already have an existing client SSL context, you can add the key-manager to it as follows:

    /subsystem=elytron/client-ssl-context=modcluster-client-ssl-context:write-attribute(name=key-manager, value=twoWayKM)
  4. Reload the server.

    reload

2.7. Elytron Integration with the JGroups Subsystem

Components in the elytron subsystem may be referenced when defining authorization or encryption protocols in the jgroups subsystem. Full instructions on configuring these protocols are found in the Securing a Cluster section of the Configuration Guide.

2.8. Elytron Integration with the Remoting Subsystem

2.8.1. Elytron Integration with Remoting Connectors

A remoting connector is specified by a SASL authentication factory, a socket binding, and an optional SSL context. In particular, the attributes for a connector are as follows:

sasl-authentication-factory
A reference to the SASL authentication factory to use for authenticating requests to this connector. For more information on creating this factory, see Create an Elytron Authentication Factory.
socket-binding
A reference to the socket binding, detailing the interface and port where the connector should listen for incoming requests.
ssl-context
An optional reference to the server-side SSL Context to use for this connector. The SSL Context contains the server key manager and trust manager to be used, and should be defined in instances where SSL is desired.

For example, a connector can be added as follows, where SASL_FACTORY_NAME is an already defined authentication factory and SOCKET_BINDING_NAME is an existing socket binding.

/subsystem=remoting/connector=CONNECTOR_NAME:add(sasl-authentication-factory=SASL_FACTORY_NAME,socket-binding=SOCKET_BINDING_NAME)

If SSL is desired, a preconfigured server-ssl-context may be referenced using the ssl-context attribute, as seen below.

/subsystem=remoting/connector=CONNECTOR_NAME:add(sasl-authentication-factory=SASL_FACTORY_NAME,socket-binding=SOCKET_BINDING_NAME,ssl-context=SSL_CONTEXT_NAME)
Enable One-way SSL/TLS for Remoting Connectors Using the Elytron Subsystem

Before enabling one-way SSL/TLS in JBoss EAP, you must configure a key-store, key-manager, and a server-ssl-context that references the defined key-manager.

The following SASL mechanisms support channel binding to external secure channels, such as SSL/TLS:

  • GS2-KRB5-PLUS
  • SCRAM-SHA-1-PLUS
  • SCRAM-SHA-256-PLUS
  • SCRAM-SHA-384-PLUS
  • SCRAM-SHA-512-PLUS

To use any of the above mechanisms, a custom SASL factory can be configured, or one of the predefined SASL authentication factories can be modified to offer any of these mechanisms. A SASL mechanism selector can be used on the client to specify the appropriate SASL mechanism.

  1. Create a socket-binding for the connector. The following command defines the oneWayBinding binding that listens on port 11199.

    /socket-binding-group=standard-sockets/socket-binding=oneWayBinding:add(port=11199)
  2. Create a connector that references the SASL authentication factory, the previously created socket binding, and the SSL context.

    /subsystem=remoting/connector=oneWayConnector:add(sasl-authentication-factory=SASL_FACTORY,socket-binding=oneWayBinding,ssl-context=SSL_CONTEXT)
    Important

    In cases where you have both a security-realm and ssl-context defined, JBoss EAP will use the SSL/TLS configuration provided by ssl-context.

  3. Configure the client to trust the server certificate. A generic example client is found at Elytron Client Side One Way Example. This example configures an ssl-context using the client trust-store.
Enable Two-way SSL/TLS for Remoting Connectors Using the Elytron Subsystem

Before enabling two-way SSL/TLS in JBoss EAP, you must configure a separate key-store components for the client and server certificates, a key-manager for the server key-store, a trust-manager for the server trust-store, and a server-ssl-context that references the defined key-manager and trust-manager.

The following SASL mechanisms support channel binding to external secure channels, such as SSL/TLS:

  • GS2-KRB5-PLUS
  • SCRAM-SHA-1-PLUS
  • SCRAM-SHA-256-PLUS
  • SCRAM-SHA-384-PLUS
  • SCRAM-SHA-512-PLUS

To use any of the above mechanisms, a custom SASL factory can be configured, or one of the predefined SASL authentication factories can be modified to offer any of these mechanisms. A SASL mechanism selector can be used on the client to specify the appropriate SASL mechanism.

  1. Create a socket-binding for the connector. The following command defines the twoWayBinding binding that listens on port 11199.

    /socket-binding-group=standard-sockets/socket-binding=twoWayBinding:add(port=11199)
  2. Create a connector that references the SASL authentication factory, the previously created socket binding, and the SSL context.

    /subsystem=remoting/connector=twoWayConnector:add(sasl-authentication-factory=SASL_FACTORY,socket-binding=twoWayBinding,ssl-context=SSL_CONTEXT)
    Important

    In cases where you have both a security-realm and ssl-context defined, JBoss EAP will use the SSL/TLS configuration provided by ssl-context.

  3. Configure your client to trust the server certificate, and to present its certificate to the server.

    You need to configure your client to present the trusted client certificate to the server to complete the two-way SSL/TLS authentication. For example, if using a browser, you need to import the trusted certificate into the browser’s truststore. A generic example client is found at Elytron Client Side Two Way Example. This example configures an ssl-context using the client trust-store and key-store.

Two-way SSL/TLS is now enabled on the remoting connector.

2.8.2. Elytron Integration with Remoting HTTP Connectors

A remote HTTP connection is specified by referencing a connector in the undertow system and a SASL authentication factory defined in the elytron subsystem. The HTTP connector provides the configuration for the HTTP upgrade-based remoting connector, and connects to an HTTP listener specified by the connector-ref attribute.

The attributes for a connector are as follows:

connector-ref
A reference to a predefined undertow listener.
sasl-authentication-factory
A reference to the SASL authentication factory to use for authenticating requests to this connector. For more information on creating this factory, see Create an Elytron Authentication Factory.

For example, a http-connector can be added as follows, where CONNECTOR_NAME references the undertow listener, and SASL_FACTORY_NAME is an already defined authentication factory in the elytron subsystem.

/subsystem=remoting/http-connector=HTTP_CONNECTOR_NAME:add(connector-ref=CONNECTOR_NAME,sasl-authentication-factory=SASL_FACTORY_NAME)
Enable One-Way SSL on the Remoting HTTP Connector

Before enabling one-way SSL/TLS in JBoss EAP, you must configure a key-store, key-manager, and a server-ssl-context that references the defined key-manager.

The following SASL mechanisms support channel binding to external secure channels, such as SSL/TLS:

  • GS2-KRB5-PLUS
  • SCRAM-SHA-1-PLUS
  • SCRAM-SHA-256-PLUS
  • SCRAM-SHA-384-PLUS
  • SCRAM-SHA-512-PLUS

To use any of the above mechanisms, a custom SASL factory can be configured, or one of the predefined SASL authentication factories can be modified to offer any of these mechanisms. A SASL mechanism selector can be used on the client to specify the appropriate SASL mechanism.

  1. Check whether the https-listener is configured to use a legacy security realm for its SSL configuration.

    /subsystem=undertow/server=default-server/https-listener=https:read-attribute(name=security-realm)
    {
        "outcome" => "success",
        "result" => "ApplicationRealm"
    }

    The above command shows that the https-listener is configured to use the ApplicationRealm legacy security realm for its SSL configuration. Undertow cannot reference both a legacy security realm and an ssl-context in Elytron at the same time so you must remove the reference to the legacy security realm.

    Note

    If the result is undefined, you do not need to remove the reference to the security realm in the next step.

  2. Remove the reference to the legacy security realm, and update the https-listener to use the ssl-context from Elytron.

    Note

    https-listener must always have either a security-realm or ssl-context configured. When changing between the two configurations, the commands must be executed as a single batch, as shown below.

    batch
    /subsystem=undertow/server=default-server/https-listener=https:undefine-attribute(name=security-realm)
    /subsystem=undertow/server=default-server/https-listener=https:write-attribute(name=ssl-context, value=SERVER_SSL_CONTEXT)
    run-batch
  3. Create an HTTP connector that references the HTTPS listener and the SASL authentication factory.

    /subsystem=remoting/http-connector=ssl-http-connector:add(connector-ref=https,sasl-authentication-factory=SASL_FACTORY)
  4. Reload the server.

    reload
  5. Configure the client to trust the server certificate. For example, if using a browser, you need to import the trusted certificate into the browser’s truststore.
Enable Two-way SSL/TLS on the Remoting HTTP Connectors

Before enabling two-way SSL/TLS in JBoss EAP, you must configure separate key-store components for the client and server certificates, a key-manager for the server key-store, a trust-manager for the server trust-store, and a server-ssl-context that references the defined key-manager and trust-manager.

The following SASL mechanisms support channel binding to external secure channels, such as SSL/TLS:

  • GS2-KRB5-PLUS
  • SCRAM-SHA-1-PLUS
  • SCRAM-SHA-256-PLUS
  • SCRAM-SHA-384-PLUS
  • SCRAM-SHA-512-PLUS

To use any of the above mechanisms, a custom SASL factory can be configured, or one of the predefined SASL authentication factories can be modified to offer any of these mechanisms. A SASL mechanism selector can be used on the client to specify the appropriate SASL mechanism.

  1. Check whether the https-listener is configured to use a legacy security realm for its SSL configuration.

    /subsystem=undertow/server=default-server/https-listener=https:read-attribute(name=security-realm)
    {
        "outcome" => "success",
        "result" => "ApplicationRealm"
    }

    The above command shows that the https-listener is configured to use the ApplicationRealm legacy security realm for its SSL configuration. Undertow cannot reference both a legacy security realm and an ssl-context in Elytron at the same time so you must remove the reference to the legacy security realm.

    Note

    If the result is undefined, you do not need to remove the reference to the security realm in the next step.

  2. Remove the reference to the legacy security realm, and update the https-listener to use the ssl-context from Elytron.

    Note

    https-listener must always have either a security-realm or ssl-context configured. When changing between the two configurations, the commands must be executed as a single batch, as shown below.

    batch
    /subsystem=undertow/server=default-server/https-listener=https:undefine-attribute(name=security-realm)
    /subsystem=undertow/server=default-server/https-listener=https:write-attribute(name=ssl-context, value=SERVER_SSL_CONTEXT)
    run-batch
  3. Create an HTTP connector that references the HTTPS listener and the SASL authentication factory.

    /subsystem=remoting/http-connector=ssl-http-connector:add(connector-ref=https,sasl-authentication-factory=SASL_FACTORY)
  4. Reload the server.

    reload
  5. Configure your client to trust the server certificate, and to present its certificate to the server.

    You need to configure your client to present the trusted client certificate to the server to complete the two-way SSL/TLS authentication. For example, if using a browser, you need to import the trusted certificate into the browser’s truststore.

Two-way SSL/TLS is now enabled on the remoting HTTP connector.

Important

In cases where you have both a security-realm and ssl-context defined, JBoss EAP will use the SSL/TLS configuration provided by ssl-context.

2.8.3. Elytron Integration with Remoting Outbound Connectors

A remote outbound connection is specified by an outbound socket binding and an authentication context. The authentication context provides all of the security information that is needed for the connection. In particular, the attributes for a remote-outbound-connection are as follows:

  • outbound-socket-binding-ref - The name of the outbound socket binding, which is used to determine the destination address and port for the connection.
  • authentication-context - A reference to the authentication context, which contains the authentication configuration and the defined SSL context, if one exists, required for the connection. For information on defining an authentication context, see Creating an Authentication Context.

For example, a remote-outbound-connection can be added as follows, where OUTBOUND_SOCKET_BINDING_NAME is an already defined outbound-socket-binding and AUTHENTICATION_CONTEXT_NAME is an authentication-context that has already been defined in the elytron subsystem configuration.

/subsystem=remoting/remote-outbound-connection=OUTBOUND_CONNECTION_NAME:add(authentication-context=AUTHENTICATION_CONTEXT_NAME, outbound-socket-binding-ref=OUTBOUND_SOCKET_BINDING_NAME)

2.9. Securing a Managed Domain

In addition to securing the management interfaces, you can also secure communication between JBoss EAP instances in a managed domain.

For information on concepts and general configuration for the managed domain operating mode, see the Domain Management section of the JBoss EAP Configuration Guide.

2.9.1. Configure Password Authentication Between Slaves and the Domain Controller Using Elytron

  1. Add a user on the master domain controller.

    A user needs to be added on the master domain controller for the slave controller to use for authentication. If you are using the default file based user and group authentication mechanism, this can be done by running EAP_HOME/bin/adduser.sh. Add the username, password and other configurations when prompted.

    The add-user utility can be used to manage both the users in the ManagementRealm and the users in the ApplicationRealm.

    Note

    The server caches the contents of the properties files in memory. However, the server does check the modified time of the properties files on each authentication request and reloads if the time has been updated. This means that all changes made by the add-user utility are immediately applied to any running server.

    The slave controller attemps to authenticate using the native interface. If the native interface has been secured with the ManagementRealm Elytron security realm, then you would need to add a user to ManagementRealm for the slave controller to use.

    Note

    The default name of the realm for management users is ManagementRealm. When the add-user utility prompts for the realm name, just accept the default unless you have switched to a different realm.

    The following example assumes the user slave with the password password1! has been added to ManagementRealm.

  2. Add an authentication-configuration to the slave controller.

    /host=slave/subsystem=elytron/authentication-configuration=slave:add(authentication-name=slave, credential-reference={clear-text=password1!})
  3. Add an authentication-context to the slave controller.

    /host=slave/subsystem=elytron/authentication-context=slave-context:add(match-rules=[{authentication-configuration=slave}])
  4. Specify the domain controller location and authentication-context in the slave controller.

    <domain-controller>
      <remote protocol="remote" host="localhost" port="9999" authentication-context="slave-context"/>
    </domain-controller>

2.9.2. Configure Password Authentication Between Slaves and the Domain Controller Using Legacy Core Management Authentication

When configuring a managed domain, by default, the master domain controller is configured to require authentication for each slave controller that connects to it. To configure slave controllers with the proper credentials, you must do the following:

  1. Add a user to the master domain controller

    You need to add a user to the master domain controller using the add-user script. Specifically, you will need to ensure that the user is added to the same realm the master uses to secure its management interface, which by default is ManagementRealm. You also need to ensure you answer yes to the Is this new user going to be used for one AS process to connect to another AS process? question.

    Important

    After adding the user, the script will output a <secret> element, which will be used in the next step. You must keep this value for use in the next step.

    Example: Adding a Slave User

    $ EAP_HOME/bin/add-user.sh
    
    What type of user do you wish to add?
     a) Management User (mgmt-users.properties)
     b) Application User (application-users.properties)
    (a): a
    
    Enter the details of the new user to add.
    Using realm 'ManagementRealm' as discovered from the existing property files.
    Username : slave-user
    Password recommendations are listed below. To modify these restrictions edit the add-user.properties configuration file.
     - The password should be different from the username
     - The password should not be one of the following restricted values {root, admin, administrator}
     - The password should contain at least 8 characters, 1 alphabetic character(s), 1 digit(s), 1 non-alphanumeric symbol(s)
    Password :
    Re-enter Password :
    What groups do you want this user to belong to? (Please enter a comma separated list, or leave blank for none)[  ]:
    About to add user 'slave-user' for realm 'ManagementRealm'
    Is this correct yes/no? yes
    Added user 'slave-user' to file '/home/user/EAP-7.1.0/standalone/configuration/mgmt-users.properties'
    Added user 'slave-user' to file '/home/user/EAP-7.1.0/domain/configuration/mgmt-users.properties'
    Added user 'slave-user' with groups  to file '/home/user/EAP-7.1.0/standalone/configuration/mgmt-groups.properties'
    Added user 'slave-user' with groups  to file '/home/user/EAP-7.1.0/domain/configuration/mgmt-groups.properties'
    Is this new user going to be used for one AS process to connect to another AS process?
    e.g. for a slave host controller connecting to the master or for a Remoting connection for server to server EJB calls.
    yes/no? yes
    To represent the user add the following to the server-identities definition <secret value="ABCzc3dv11Qx" />

  2. Configure the slave controllers to use the credential.

    Once you have created the user on the master domain controller, you will need to update each slave controller to use that credential in the host configuration file, for example host.xml or host-slave.xml. To do so, you need to add the user name to the <remote> element in the domain controller configuration. You will also need to add the <secret> to the server identities of the realm used to secure the <remote> element. Both the user name and <secret> were obtained when adding the user to the master domain controller in the previous step.

    Example: Configuring Slave Controllers

    ...
    <security-realm name="ManagementRealm">
        <server-identities>
            <!-- Replace this with either a base64 password of your own, or use a vault with a vault expression -->
            <secret value="ABCzc3dv11Qx"/>
        </server-identities>
    ...
    <domain-controller>
      <remote security-realm="ManagementRealm" username="slave-user">
          <discovery-options>
              <static-discovery name="primary" protocol="${jboss.domain.master.protocol:remote}" host="${jboss.domain.master.address}" port="${jboss.domain.master.port:9999}"/>
          </discovery-options>
      </remote>
    </domain-controller>

2.9.3. Configuring SSL/TLS Between Domain and Host Controllers Using Elytron

Important

When you configure SSL/TLS to be used between JBoss EAP instances in a managed domain, each instance can have a client or server role depending on the interaction. This includes all host controllers as well as domain controllers. As a result, it is recommended that you set up two-way SSL/TLS between endpoints.

You can configure JBoss EAP instances in a managed domain to use SSL/TLS when communicating with each other, in other words, between the master domain controller and host controllers. To do so using Elytron, use the following procedure.

  1. Generate and configure all necessary certificates and keystores.

    In order to set up two-way SSL/TLS between endpoints, you need to generate and configure certificates and keystores for the master domain controller as well as each host controller. You also need to import the certificate of the master domain controller into each host controller’s keystore as well as import each host controller’s certificate into the master domain controller’s keystore. The specifics of this process is covered in Enable Two-way SSL/TLS for the Management Interfaces Using the Elytron Subsystem.

  2. Add a user on the master domain controller.

    A user needs to be added on the master domain controller for the slave controller to use for authentication. If you are using the default file based user and group authentication mechanism, this can be done by running EAP_HOME/bin/adduser.sh. Add the username, password and other configurations when prompted.

    The add-user utility can be used to manage both the users in the ManagementRealm and the users in the ApplicationRealm.

    Note

    The server caches the contents of the properties files in memory. However, the server does check the modified time of the properties files on each authentication request and reloads if the time has been updated. This means that all changes made by the add-user utility are immediately applied to any running server.

    The slave controller attemps to authenticate using the native interface. If the native interface has been secured with the ManagementRealm Elytron security realm, then you would need to add a user to ManagementRealm for the slave controller to use.

    Note

    The default name of the realm for management users is ManagementRealm. When the add-user utility prompts for the realm name, just accept the default unless you have switched to a different realm.

    The following example assumes the user slave with the password password1! has been added to ManagementRealm.

  3. Configure the master domain controller to use SSL/TLS.

    The commands below configure the domain controller’s key-store, key-manager, trust-manager, and server-ssl-context for the server keystore and truststore.

    /host=master/subsystem=elytron/key-store=twoWayKS:add(path=/path/to/server.keystore.jks,credential-reference={clear-text=secret},type=JKS)
    
    /host=master/subsystem=elytron/key-store=twoWayTS:add(path=/path/to/server.truststore.jks,credential-reference={clear-text=secret},type=JKS)
    
    /host=master/subsystem=elytron/key-manager=twoWayKM:add(key-store=twoWayKS,algorithm="SunX509",credential-reference={clear-text=secret})
    
    /host=master/subsystem=elytron/trust-manager=twoWayTM:add(key-store=twoWayTS,algorithm="SunX509")
    
    /host=master/subsystem=elytron/server-ssl-context=twoWaySSC:add(key-manager=twoWayKM,protocols=["TLSv1.2"],trust-manager=twoWayTM,want-client-auth=true,need-client-auth=true)
    
    /host=master/core-service=management/management-interface=native-interface:write-attribute(name=ssl-context, value=twoWaySSC)
    Important

    You need to know what key manager algorithms are provided by the JDK you are using. For example, a JDK that uses SunJSSE provides the PKIX and SunX509 algorithms. You also need to determine what HTTPS protocols you want to support. The example commands above use TLSv1.2. You can use the cipher-suite-filter argument to specify which cipher suites are allowed, and the use-cipher-suites-order argument to honor server cipher suite order. The use-cipher-suites-order attribute by default is set to true. This differs from the legacy security subsystem behavior, which defaults to honoring client cipher suite order.

  4. Configure an authentication context and domain controller location on each slave host controller.

    The following example configuration assumes the domain controller exists on localhost. Ensure you specify the correct management user, password, and domain controller location for your environment.

    /host=slave1/subsystem=elytron/authentication-context=slaveHostSSLContext:add()
    
    /host=slave1/subsystem=elytron/authentication-configuration=slaveHostSSLConfiguration:add()
    
    /host=slave1/subsystem=elytron/authentication-configuration=slaveHostSSLConfiguration:write-attribute(name=sasl-mechanism-selector,value=DIGEST-MD5)
    
    /host=slave1/subsystem=elytron/authentication-configuration=slaveHostSSLConfiguration:write-attribute(name=authentication-name,value=slave)
    
    /host=slave1/subsystem=elytron/authentication-configuration=slaveHostSSLConfiguration:write-attribute(name=realm,value=ManagementRealm)
    
    /host=slave1/subsystem=elytron/authentication-configuration=slaveHostSSLConfiguration:write-attribute(name=credential-reference,value={clear-text=password1!})
    
    /host=slave1/subsystem=elytron/authentication-context=slaveHostSSLContext:write-attribute(name=match-rules,value=[{match-host=localhost,authentication-configuration=slaveHostSSLConfiguration}]
    
    /host=slave1:write-remote-domain-controller(host=localhost,port=9999,protocol=remote,authentication-context=slaveHostSSLContext)
  5. Configure each slave host controller to use SSL/TLS.

    The commands below configure a slave host controller’s key-store, key-manager, trust-manager, client-ssl-context for the server keystore and truststore, as well as the authentication-context.

    The following example configuration assumes the domain controller exists on localhost. Ensure you specify the correct domain controller location for your environment.

    /host=slave1/subsystem=elytron/key-store=twoWayKS:add(path=/path/to/client.keystore.jks,credential-reference={clear-text=secret},type=JKS)
    
    /host=slave1/subsystem=elytron/key-store=twoWayTS:add(path=/path/to/client.truststore.jks,credential-reference={clear-text=secret},type=JKS)
    
    /host=slave1/subsystem=elytron/key-manager=twoWayKM:add(key-store=twoWayKS,algorithm="SunX509",credential-reference={clear-text=secret})
    
    /host=slave1/subsystem=elytron/trust-manager=twoWayTM:add(key-store=twoWayTS,algorithm="SunX509")
    
    /host=slave1/subsystem=elytron/client-ssl-context=twoWayCSC:add(key-manager=twoWayKM,protocols=["TLSv1.2"],trust-manager=twoWayTM)
    
    /host=slave1/subsystem=elytron/authentication-context=slaveHostSSLContext:write-attribute(name=match-rules,value=[{match-host=localhost,authentication-configuration=slaveHostSSLConfiguration,ssl-context=twoWayCSC}])
  6. Reload all the JBoss EAP hosts in your managed domain.

2.9.4. Configuring SSL/TLS Between Domain and Host Controllers Using Legacy Core Management Authentication

Important

When you configure SSL/TLS to be used between JBoss EAP instances in a managed domain, each instance can have a client or server role depending on the interaction. This includes all host controllers as well as domain controllers. As a result, it is recommended that you set up two-way SSL/TLS between endpoints.

You can configure JBoss EAP instances in a managed domain to use SSL/TLS when communicating with each other, in other words, between the master domain controller and host controllers. To do so using legacy core management authentication, use the following procedure.

  1. Generate and configure all necessary certificates and keystores.

    In order to set up two-way SSL/TLS between endpoints, you need to generate and configure certificates and keystores for the master domain controller as well as each host controller. You also need to import the certificate of the master domain controller into each host controller’s keystore as well as import each host controller’s certificate into the master domain controller’s keystore. The specifics of this process is covered in Setting up Two-way SSL/TLS for the Management Interfaces with Legacy Core Management Authentication.

  2. Configure the master domain controller to use SSL/TLS.

    Once you have configured all certificates and keystores, you need to configure a security realm to use two-way SSL/TLS. This is done by configuring a security realm to use SSL/TLS and to require it for authentication. That security realm is then used to secure the management interface used for connecting between host controllers and the master domain controller.

    Note

    The following commands below must either be run in batch mode, or the server must be reloaded after adding the ssl server identity. The example below is shown using batch mode.

    batch
    
    /host=master/core-service=management/security-realm=CertificateRealm:add()
    
    /host=master/core-service=management/security-realm=CertificateRealm/server-identity=ssl:add(alias=domaincontroller,keystore-relative-to=jboss.domain.config.dir,keystore-path=domaincontroller.jks,keystore-password=secret)
    
    /host=master/core-service=management/security-realm=CertificateRealm/authentication=truststore:add(keystore-relative-to=jboss.domain.config.dir,keystore-path=domaincontroller.jks,keystore-password=secret)
    
    /host=master/core-service=management/security-realm=CertificateRealm/authentication=local:add(default-user=\$local)
    
    /host=master/core-service=management/security-realm=CertificateRealm/authentication=properties:add(relative-to=jboss.domain.config.dir,path=mgmt-users.properties)
    
    /host=master/core-service=management/management-interface=native-interface:write-attribute(name=security-realm,value=CertificateRealm)
    
    run-batch
  3. Configure all host controllers to use SSL/TLS.

    Once you have the master domain controller configured to use two-way SSL/TLS, you need to configure each host controller to use it as well. The process is very much the same as the master domain controller configuration, except you will need to use the keystore specific to each host.

    Note

    The following commands below must either be run in batch mode, or the server must be reloaded after adding the ssl server identity. The example below is shown using batch mode.

    batch
    
    /host=instance1/core-service=management/security-realm=CertificateRealm:add()
    
    /host=instance1/core-service=management/security-realm=CertificateRealm/server-identity=ssl:add(alias=instance1,keystore-relative-to=jboss.domain.config.dir,keystore-path=instance1.jks,keystore-password=secret)
    
    /host=instance1/core-service=management/security-realm=CertificateRealm/authentication=truststore:add(keystore-relative-to=jboss.domain.config.dir,keystore-path=instance1.jks,keystore-password=secret)
    
    /host=instance1/core-service=management/security-realm=CertificateRealm/authentication=local:add(default-user="\$local")
    
    /host=instance1/core-service=management/security-realm=CertificateRealm/authentication=properties:add(relative-to=jboss.domain.config.dir,path=mgmt-users.properties)
    
    /host=instance1/core-service=management/management-interface=native-interface:write-attribute(name=security-realm,value=CertificateRealm)
    
    run-batch

    Additionally, you will need to update the security realm used when connecting the master domain controller. This change must be done directly in the host controller’s configuration file, for example host.xml or host-slave.xml, while the server is not running.

    Example: Host Controller Configuration File

    <domain-controller>
      <remote security-realm="CertificateRealm" username="slave-user">
        <discovery-options>
          <static-discovery name="primary" protocol="${jboss.domain.master.protocol:remote}" host="${jboss.domain.master.address}" port="${jboss.domain.master.port:9999}"/>
        </discovery-options>
      </remote>
    </domain-controller>

    Warning

    Red Hat recommends that SSLv2, SSLv3, and TLSv1.0 be explicitly disabled in favor of TLSv1.1 or TLSv1.2 in all affected packages.

2.10. Additional Elytron Components for SSL/TLS

The basic concepts behind configuring one-way SSL/TLS and two-way SSL/TLS are covered in the following:

Elytron also offers some additional components for configuring SSL/TLS.

2.10.1. Using an ldap-key-store

An ldap-key-store allows you to use a keystore stored in an LDAP server. You can use an ldap-key-store in the same way as you use a key-store.

Note

It is not possible to use a JMX ObjectName to decrypt the LDAP credentials. Instead, credentials can be secured by using a credential store.

To create and use an ldap-key-store:

  1. Configure a dir-context.

    To connect to the LDAP server from JBoss EAP, you need to configure a dir-context that provides the URL as well as the principal used to connect to the server.

    Example: dir-context

    /subsystem=elytron/dir-context=exampleDC:add(url="ldap://127.0.0.1:10389", principal="uid=admin,ou=system", credential-reference={clear-text="secret"})

  2. Configure an ldap-key-store.

    When you configure an ldap-key-store, you need to specify both the dir-context used to connect to the LDAP server as well as how to locate the keystore stored in the LDAP server. At a minimum, this requires you to specify a search-path.

    Example: ldap-key-store

    /subsystem=elytron/ldap-key-store=ldapKS:add(dir-context=exampleDC, search-path="ou=Keystores,dc=wildfly,dc=org")

  3. Use the ldap-key-store.

    Once you have defined your ldap-key-store, you can use it in the same places where a key-store could be used. For example, you could use an ldap-key-store when configuring One-way SSL/TLS and Two-way SSL/TLS for applications.

For the full list of attributes for ldap-key-store as well as other Elytron components, see Elytron Subsystem Components Reference.

2.10.2. Using a filtering-key-store

A filtering-key-store allows you to expose a subset of aliases from an existing key-store, and use it in the same places you could use a key-store. For example, if a keystore contained alias1, alias2, and alias3, but you only wanted to expose alias1 and alias3, a filtering-key-store provides you several ways to do that.

To create a filtering-key-store:

  1. Configure a key-store.

    /subsystem=elytron/key-store=myKS:add(path=keystore.jks, relative-to=jboss.server.config.dir, credential-reference={clear-text=secret}, type=JKS)
  2. Configure a filtering-key-store.

    When you configure a filtering-key-store, you specify which key-store you want to filter and the alias-filter for filtering aliases from the key-store. The filter can be specified in one of the following formats:

    • alias1,alias3, which is a comma-delimited list of aliases to expose.
    • ALL:-alias2, which exposes all aliases in the keystore except the ones listed.
    • NONE:+alias1:+alias3, which exposes no aliases in the keystore except the ones listed.

      This example uses a comma-delimted list to expose alias1 and alias3.

      /subsystem=elytron/filtering-key-store=filterKS:add(key-store=myKS, alias-filter="alias1,alias3")
      Note

      The alias-filter attribute is case sensitive. Because the use of mixed-case or uppercase aliases, such as elytronAppServer, might not be recognized by some keystore providers, it is recommended to use lowercase aliases, such as elytronappserver.

  3. Use the filtering-key-store.

    Once you have defined your filtering-key-store, you can use it in the same places where a key-store could be used. For example, you could use a filtering-key-store when configuring One-way SSL/TLS and Two-way SSL/TLS for applications.

For the full list of attributes for filtering-key-store as well as other Elytron components, see Elytron Subsystem Components Reference.

2.10.3. Reload a Keystore

You can reload a keystore configured in JBoss EAP from the management CLI. This is useful in cases where you have made changes to certificates referenced by a keystore.

To reload a keystore:

/subsystem=elytron/key-store=httpsKS:load

2.10.4. Keystore Alias

The alias denotes the stored secret or credential in the store. If you add a keystore to the elytron subsystem using the key-store component, you can check the keystore’s contents using the alias related key-store operations.

The different operations for alias manipulation are:

  • read-alias - Read an alias from a keystore.
  • read-aliases - Read aliases from a keystore.
  • remove-alias - Remove an alias from a keystore.

For example, to read an alias:

/subsystem=elytron/key-store=httpsKS/:read-alias(alias=localhost)

2.10.5. Using a client-ssl-context

A client-ssl-context is used for providing an SSL context when the JBoss EAP instance creates an SSL connection as a client, such as using SSL in remoting.

To create a client-ssl-context:

  1. Create key-store, key-manager, and trust-manager components as needed.

    If establishing a two-way SSL/TLS connection, you need to create separate key-store components for the client and server certificates, a key-manager for the client key-store, and a trust-manager for the server key-store. Alternatively, if you are doing a one-way SSL/TLS connection, you need to create a key-store for the server certificate and a trust-manager that references it. Examples on creating keystores and truststores are available in the Enable Two-way SSL/TLS for Applications using the Elytron Subsystem section.

  2. Create a client-ssl-context.

    Create a client-ssl-context referencing keystores, truststores, as well as any other necessary configuration options.

    Example: client-ssl-context

    /subsystem=elytron/client-ssl-context=exampleCSC:add(key-manager=clientKM, trust-manager=clientTM, protocols=["TLSv1.2"])

  3. Reference the client-ssl-context.

For the full list of attributes for client-ssl-context as well as other Elytron components, see Elytron Subsystem Components Reference.

2.10.6. Using a server-ssl-context

A server-ssl-context is used for providing a server-side SSL context. In addition to the usual configuration for an SSL context, it is possible to configure additional items such as cipher suites and protocols. The SSL context will wrap any additional items that are configured.

  1. Create key-store, key-manager, and trust-manager components as needed.

    If establishing a two-way SSL/TLS connection, you need to create separate key-store components for the client and server certificates, a key-manager for the server key-store, and a trust-manager for the server trust-store. Alternatively, if you are doing a one-way SSL/TLS connection, you need to create a key-store for the server certificate and a key-manager that references it. Examples on creating keystores and truststores are available in the Enable Two-way SSL/TLS for Applications Using the Elytron Subsystem section.

  2. Create a server-ssl-context.

    Create a server-ssl-context that references the key manager, trust manager, or any other desired configuration options using one of the options outlined below.

Add a Server SSL Context Using the Management CLI
/subsystem=elytron/server-ssl-context=newServerSSLContext:add(key-manager=KEY_MANAGER,protocols=["TLSv1.2"])
Important

You need to determine what HTTPS protocols will be supported. The example commands above use TLSv1.2. You can use the cipher-suite-filter argument to specify which cipher suites are allowed, and the use-cipher-suites-order argument to honor server cipher suite order. The use-cipher-suites-order attribute by default is set to true. This differs from the legacy security subsystem behavior, which defaults to honoring client cipher suite order.

Add a Server SSL Context Using the Management Console
  1. Access the management console. For more information, see the Management Console section in the JBoss EAP Configuration Guide.
  2. Navigate to Configuration Subsystems Security - Elytron Other.
  3. Click on View. The Server SSL Context tab lets you do all the server SSL context related configurations.

For the full list of attributes for server-ssl-context as well as other Elytron components, see Elytron Subsystem Components Reference.

2.10.7. Custom Components

When configuring SSL/TLS in the elytron subsystem, you can provide and use custom implementations of the following components:

  • key-store
  • key-manager
  • trust-manager
  • client-ssl-context
  • server-ssl-context
Warning

It is not recommended to provide custom implementations of any component outside of the trust-manager without an intimate knowledge of the Java Secure Socket Extension (JSSE).

Important

When using FIPS it is not possible to utilize a custom trust manager or key manager, as FIPS requires these managers be embedded in the JDK for security reasons. Similar behavior can be accomplished by implementing a SecurityRealm that validates X509 evidences.

When creating custom implementations of Elytron components, they must present the appropriate capabilities and requirements. For more details on capabilities and requirements, see the Capabilities and Requirements section of the JBoss EAP Security Architecture guide. Implementation details for each component are provided by the JDK vendor.

2.10.7.1. Add a Custom Component to Elytron

The following steps describe adding a custom component within Elytron.

  1. Add the JAR containing the provider for the custom component as a module into JBoss EAP, declaring any required dependencies, such as javax.api:

    module add --name=MODULE_NAME --resources=FACTORY_JAR --dependencies=javax.api,DEPENDENCY_LIST
    Important

    Using the module management CLI command to add and remove modules is provided as Technology Preview only. This command is not appropriate for use in a managed domain or when connecting to the management CLI remotely. Modules should be added and removed manually in a production environment. For more information, see the Create a Custom Module Manually and Remove a Custom Module Manually sections of the JBoss EAP Configuration Guide.

    Technology Preview features are not supported with Red Hat production service level agreements (SLAs), might not be functionally complete, and Red Hat does not recommend to use them for production. These features provide early access to upcoming product features, enabling customers to test functionality and provide feedback during the development process.

    See Technology Preview Features Support Scope on the Red Hat Customer Portal for information about the support scope for Technology Preview features.

  2. When the component is added to the elytron subsystem the java.util.ServiceLoader will be used to discover the provider. Alternatively, a reference to the provider can be provided by defining a provider-loader. There are two methods of creating the loader, and only one should be implemented for each component.

    • Reference the provider directly when defining the provider-loader:

      /subsystem=elytron/provider-loader=LOADER_NAME:add(class-names=[CLASS_NAME],module=MODULE_NAME)
    • Include a reference to the provider in META-INF/services/java.security.Provider. This reference is automatically created when using the @MetaInfServices annotation in org.kohsuke.metainf-services. When using this method only the module needs to be referenced by the provider-loader, as seen below:

      /subsystem=elytron/provider-loader=LOADER_NAME:add(module=MODULE_NAME)
  3. Add the custom component into Elytron’s configuration, using the appropriate element for the type to be added and referencing any defined providers.

    /subsystem=elytron/COMPONENT_NAME=NEW_COMPONENT:add(providers=LOADER_NAME,...)

    For instance, to define a trust manager, the trust-manager element would be used, as seen in the following command:

    Example: Adding a Custom Trust Manager

    /subsystem=elytron/trust-manager=newTrustManager:add(algorithm=MyX509,providers=customProvider,key-store=sampleKeystore)

  4. Once defined, the component can be referenced from other elements.

2.10.7.2. Using Custom Trust Managers with Elytron

By implementing a custom trust manager, it is possible to extend the validation of certificates when using HTTPS in Undertow, LDAPS in a dir-context, or any place where Elytron is used for SSL connections. This component is responsible for making trust decisions for the server, and it is strongly recommended that these be implemented if a custom trust manager is used.

Important

When using FIPS it is not possible to utilize a custom trust manager, as FIPS requires this manager be embedded in the JDK for security reasons. Similar behavior can be accomplished by implementing a SecurityRealm that validates X509 evidences.

Requirements for Implementing a Custom Trust Manager

When using a custom trust manager, the following must be implemented:

  • A trust manager that implements the X509ExtendedTrustManager interface.
  • A trust manager factory that extends TrustManagerFactorySpi.
  • The provider of the trust manager factory.

The provider must be included in the JAR file to be added into JBoss EAP. Any implemented classes must be included in JBoss EAP as a module. Classes are not required to be in one module, and can be loaded from module dependencies.

Example Implementations

The following example demonstrates a provider that registers the custom trust manager factory as a service.

Example: Provider

import org.kohsuke.MetaInfServices;
import javax.net.ssl.TrustManagerFactory;
import java.security.Provider;
import java.util.Collections;
import java.util.List;
import java.util.Map;

@MetaInfServices(Provider.class)
public class CustomProvider extends Provider {

    public CustomProvider() {
        super("CustomProvider", 1.0, "Demo provider");

        System.out.println("CustomProvider initialization.");

        final List<String> emptyList = Collections.emptyList();
        final Map<String, String> emptyMap = Collections.emptyMap();

        putService(new Service(this, TrustManagerFactory.class.getSimpleName(),"CustomAlgorithm", CustomTrustManagerFactorySpi.class.getName(), emptyList, emptyMap));
    }

}

The following example demonstrates a custom trust manager. This trust manager contains overloaded methods on checking if a client or server is trusted.

Example: TrustManager

import javax.net.ssl.SSLEngine;
import javax.net.ssl.X509ExtendedTrustManager;
import java.net.Socket;
import java.security.cert.CertificateException;
import java.security.cert.X509Certificate;

public class CustomTrustManager extends X509ExtendedTrustManager {

    public void checkClientTrusted(X509Certificate[] x509Certificates, String s, Socket socket) throws CertificateException {
        // Insert your code here
    }

    public void checkServerTrusted(X509Certificate[] x509Certificates, String s, Socket socket) throws CertificateException {
        // Insert your code here
    }

    public void checkClientTrusted(X509Certificate[] x509Certificates, String s, SSLEngine sslEngine) throws CertificateException {
        // Insert your code here
    }

    public void checkServerTrusted(X509Certificate[] x509Certificates, String s, SSLEngine sslEngine) throws CertificateException {
        // Insert your code here
    }

    public void checkClientTrusted(X509Certificate[] x509Certificates, String s) throws CertificateException {
        // Insert your code here
    }

    public void checkServerTrusted(X509Certificate[] x509Certificates, String s) throws CertificateException {
        // Insert your code here
    }

    public X509Certificate[] getAcceptedIssuers() {
        // Insert your code here
    }

}

The following example is a factory used to return instances of the trust manager.

Example: TrustManagerFactorySpi

import javax.net.ssl.ManagerFactoryParameters;
import javax.net.ssl.TrustManager;
import javax.net.ssl.TrustManagerFactorySpi;
import java.security.InvalidAlgorithmParameterException;
import java.security.KeyStore;
import java.security.KeyStoreException;

public class CustomTrustManagerFactorySpi extends TrustManagerFactorySpi {

    protected void engineInit(KeyStore keyStore) throws KeyStoreException {
        // Insert your code here
    }

    protected void engineInit(ManagerFactoryParameters managerFactoryParameters) throws InvalidAlgorithmParameterException {
        // Insert your code here
    }

    protected CustomTrustManager[] engineGetTrustManagers() {
        // Insert your code here
    }

}

Adding the Custom Trust Manager

Once the provider and trust manager have been created, add them to the elytron subsystem by using the steps outlined in Add a Custom Component to Elytron.

2.10.8. Using a Certificate Revocation List

If you want to validate a certificate against a certificate revocation list (CRL), you can configure this using the certificate-revocation-list attribute for a trust manager in the elytron subsystem. For example:

/subsystem=elytron/trust-manager=TRUST_MANAGER:write-attribute(name=certificate-revocation-list,value={path=/path/to/CRL_FILE.crl.pem}

For more information on the available attributes for a trust manager, see the trust-manager Attributes table.

Note

Your truststore must contain the certificate chain in order to check the validity of both the certification revocation list and the certificate. The truststore should not contain end-entity certificates, just certificate authority and intermediate certificates.

You can instruct the trust manager to reload the certificate revocation list by using the reload-certificate-revocation-list operation.

/subsystem=elytron/trust-manager=TRUST_MANAGER:reload-certificate-revocation-list
Red Hat logoGithubRedditYoutubeTwitter

Learn

Try, buy, & sell

Communities

About Red Hat Documentation

We help Red Hat users innovate and achieve their goals with our products and services with content they can trust.

Making open source more inclusive

Red Hat is committed to replacing problematic language in our code, documentation, and web properties. For more details, see the Red Hat Blog.

About Red Hat

We deliver hardened solutions that make it easier for enterprises to work across platforms and environments, from the core datacenter to the network edge.

© 2024 Red Hat, Inc.