The messaging-clustering-singleton
quickstart uses a JMS topic and a queue to demonstrate clustering using WildFly messaging with MDB singleton configuration where only one node in the cluster will be active.
The messaging-clustering-singleton
quickstart demonstrates the use of clustering with integrated Apache ActiveMQ Artemis. It uses the same code as in helloworld-mdb quickstart, with only a difference in the configuration to run it as a clustered singleton. Instructions are provided to run the quickstart on either a standalone server or in a managed domain.
These are the two JMS resources contained in this quickstart:
-
A queue named
HELLOWORLDMDBQueue
bound in JNDI asjava:/queue/HELLOWORLDMDBQueue
-
A topic named
HELLOWORLDMDBTopic
bound in JNDI asjava:/topic/HELLOWORLDMDBTopic
Both contain a singleton configuration as specified in the file WEB-INF/jboss-ejb3.xml:
<c:clustering>
<ejb-name>*</ejb-name>
<c:clustered-singleton>true</c:clustered-singleton>
</c:clustering>
The wildcard (*) in the ejb-name
element indicates that all MDBs contained in the application will be clustered singleton. As a result, only one node in the cluster will have those MDBs active at a specific time. If that node shuts down, another node in the cluster will become the active node with MDBs, called the singleton provider.
Also, we can find a configuration for delivery group in the same file:
<d:delivery>
<ejb-name>HelloWorldTopicMDB</ejb-name>
<d:group>my-mdb-delivery-group</d:group>
</d:delivery>
Here, you can see that only one of the MDBs, HelloWorldTopicMDB
, is associated with a delivery group. All delivery groups used by an MDB must be declared in the ejb
subsystem configuration, and they can be enabled or disabled. If the delivery group is disabled in a cluster node, all MDBs belonging to that group will be inactive in that node. Notice that delivery groups can be used in non-clustered environments as well. In that case, the MDB will be active in the server whenever the delivery group is enabled in the server. A delivery group can be enabled using the management CLI as you will see in this quickstart.
If a delivery group is used in conjunction with singleton, as is the case of this quickstart, the MDB will be active in the singleton provider node only if that node has delivery-group
enabled. If not, the MDB will be inactive in that node and all remainder nodes of the cluster.
This quickstart demonstrates its functionalities on bare metal and using WildFly Maven plugin.
The application this project produces is designed to be run on WildFly Application Server 35 or later.
All you need to build this project is Java SE 17.0 or later, and Maven 3.6.0 or later. See Configure Maven to Build and Deploy the Quickstarts to make sure you are configured correctly for testing the quickstarts.
When deploying this quickstart to a managed domain, replace WILDFLY_HOME
with the actual path to your WildFly installation. The installation path is described in detail here: Use of WILDFLY_HOME and JBOSS_HOME Variables.
When deploying this quickstart to multiple standalone servers, this quickstart requires that you clone your WILDFLY_HOME
installation directory and run two servers. In the following instructions, replace WILDFLY_HOME_1
with the path to your first WildFly server and replace WILDFLY_HOME_2
with the path to your second cloned WildFly server.
When you see the replaceable variable QUICKSTART_HOME, replace it with the path to the root directory of all of the quickstarts.
Follow these steps to build the project without deploying it.
-
Open a terminal and navigate to the root directory of this quickstart.
-
Type this command to build the archive:
$ mvn clean package
You can choose to configure and deploy this quickstart to a managed domain or to a standalone server. The sections below describe how to configure and start the server for each configuration.
You configure the server by running the install-domain.cli
script provided in the root directory of this quickstart.
Before you begin, back up your server configuration files.
-
If it is running, stop the WildFly server.
-
Back up the following files:
WILDFLY_HOME/domain/configuration/domain.xml WILDFLY_HOME/domain/configuration/host.xml
After you have completed testing this quickstart, you can replace these file to restore the server to its original configuration.
-
Open a terminal and navigate to the root of the WildFly directory.
-
Start the WildFly managed domain by typing the following command.
$ WILDFLY_HOME/bin/domain.sh
NoteFor Windows, use the WILDFLY_HOME\bin\domain.bat
script.
-
Review the
install-domain.cli
file located in the root of this quickstart directory. This script creates the server group and servers and configures messaging clustering for testing this quickstart. You will note it does the following:-
Stops the servers.
-
Creates the
quickstart-messaging-clustering-singleton-group
server group to test ActiveMQ clustering. -
Enables console logging to allow you to view the quickstart output.
-
Adds two servers to the
server-group
. -
Configures ActiveMQ clustering in the
full-ha
profile. -
Creates a delivery group named
my-mdb-delivery-group
, with initial active value set totrue
. -
Deploys the
messaging-clustering-singleton.war
archive. -
Starts the servers that were added to the managed domain.
-
-
Open a new terminal, navigate to the root directory of this quickstart, and run the following command, replacing
WILDFLY_HOME
with the path to your server:$ WILDFLY_HOME/bin/jboss-cli.sh --connect --file=install-domain.cli
NoteFor Windows, use the WILDFLY_HOME\bin\jboss-cli.bat
script.You should see the following output:
{ "outcome" => "success", "result" => undefined, "server-groups" => undefined } The batch executed successfully { "outcome" => "success", "result" => "STARTED" } { "outcome" => "success", "result" => "STARTED" }
If you choose to use standalone servers rather than a managed domain, you need two instances of the application server. The second server must be started with a port offset parameter provided to the startup script as -Djboss.socket.binding.port-offset=100
.
Since both application servers must be configured in the same way, you must configure the first server and then clone it.
Before you begin, back up your server configuration file.
-
If it is running, stop the WildFly server.
-
Back up the
WILDFLY_HOME/standalone/configuration/standalone-full-ha.xml
file.
After you have completed testing this quickstart, you can replace this file to restore the server to its original configuration.
-
Open a terminal and navigate to the root of the WildFly directory.
-
Start the WildFly server with the full HA profile by typing the following command.
$ WILDFLY_HOME/bin/standalone.sh -c standalone-full-ha.xml
NoteFor Windows, use the WILDFLY_HOME\bin\standalone.bat
script.
-
Review the
install-standalone.cli
file located in the root of this quickstart directory. This script configures clustering for a standalone server. You will note it does the following:-
Because the console is disabled by default in the Full HA profile, it enables console logging to allow you to view the quickstart output.
-
Enables clustering and sets a cluster password.
-
Creates a delivery group named
my-mdb-delivery-group
, with initial active value set totrue
. -
Reloads the server configuration.
-
-
Open a new terminal, navigate to the root directory of this quickstart, and run the following command, replacing
WILDFLY_HOME
with the path to your server:$ WILDFLY_HOME_1/bin/jboss-cli.sh --connect --file=install-standalone.cli
NoteFor Windows, use the WILDFLY_HOME_1\bin\jboss-cli.bat
script.You should see the following output:
The batch executed successfully process-state: reload-required
-
Type the following command to deploy the quickstart.
$ mvn wildfly:deploy
This deploys the {archiveDir}/{archiveFileName}
to the running instance of the server.
You should see a message in the server log indicating that the archive deployed successfully.
After you have successfully configured the server, you must make a copy of this WildFly directory structure to use for the second server.
-
Stop the server.
-
Make a copy of this WildFly directory structure to use for the second server.
-
Remove the following directories from the cloned instance:
WILDFLY_HOME_2/standalone/data/activemq/bindings WILDFLY_HOME_2/standalone/data/activemq/journal WILDFLY_HOME_2/standalone/data/activemq/largemessages
Use the following commands to start the standalone servers with the full HA profile.
$ WILDFLY_HOME_1/bin/standalone.sh -c standalone-full-ha.xml
$ WILDFLY_HOME_2/bin/standalone.sh -c standalone-full-ha.xml -Djboss.socket.binding.port-offset=100
Note
|
For Windows, use the WILDFLY_HOME_1\bin\standalone.bat and WILDFLY_HOME_2\bin\standalone.bat scripts.
|
The application will be running at the following URL: http://localhost:9080/messaging-clustering-singleton/HelloWorldMDBServletClient.
It will send some messages to the queue.
To send messages to the topic, use the following URL: http://localhost:9080/messaging-clustering-singleton/HelloWorldMDBServletClient?topic
The application will be running at the following URL: http://localhost:8080/messaging-clustering-singleton/HelloWorldMDBServletClient.
It will send some messages to the queue.
To send messages to the topic, use the following URL: http://localhost:8080/messaging-clustering-singleton/HelloWorldMDBServletClient?topic
Review the messages in both WildFly server consoles or logs.
The following messages are sent to the queue:
INFO [class org.jboss.as.quickstarts.mdb.HelloWorldQueueMDB] (Thread-0 (ActiveMQ-client-global-threads)) Received Message from queue: This is message 1
INFO [class org.jboss.as.quickstarts.mdb.HelloWorldQueueMDB] (Thread-2 (ActiveMQ-client-global-threads)) Received Message from queue: This is message 3
INFO [class org.jboss.as.quickstarts.mdb.HelloWorldQueueMDB] (Thread-4 (ActiveMQ-client-global-threads)) Received Message from queue: This is message 5
INFO [class org.jboss.as.quickstarts.mdb.HelloWorldQueueMDB] (Thread-3 (ActiveMQ-client-global-threads)) Received Message from queue: This is message 4
INFO [class org.jboss.as.quickstarts.mdb.HelloWorldQueueMDB] (Thread-1 (ActiveMQ-client-global-threads)) Received Message from queue: This is message 2
The following messages are sent to the topic:
INFO [class org.jboss.as.quickstarts.mdb.HelloWorldTopicMDB] (Thread-5 (ActiveMQ-client-global-threads)) Received Message from topic: This is message 1
INFO [class org.jboss.as.quickstarts.mdb.HelloWorldTopicMDB] (Thread-6 (ActiveMQ-client-global-threads)) Received Message from topic: This is message 2
INFO [class org.jboss.as.quickstarts.mdb.HelloWorldTopicMDB] (Thread-8 (ActiveMQ-client-global-threads)) Received Message from topic: This is message 4
INFO [class org.jboss.as.quickstarts.mdb.HelloWorldTopicMDB] (Thread-7 (ActiveMQ-client-global-threads)) Received Message from topic: This is message 3
INFO [class org.jboss.as.quickstarts.mdb.HelloWorldTopicMDB] (Thread-9 (ActiveMQ-client-global-threads)) Received Message from topic: This is message 5
You will notice that only one of the nodes, elected as the singleton provider node, will be receiving the messages. For that, check both servers, only one will contain the received message log entries.
You will see the following warnings in the server logs. You can ignore these warnings as they are intended for production servers.
WARNING [org.jgroups.protocols.UDP] (Thread-0 (ActiveMQ-server-ActiveMQServerImpl::serverUUID=c79278db-56e6-11e5-af50-69dd76236ee8-1573164340)) JGRP000015: the send buffer of socket DatagramSocket was set to 1MB, but the OS only allocated 212.99KB. This might lead to performance problems. Please set your max send buffer in the OS correctly (e.g. net.core.wmem_max on Linux)
WARNING [org.jgroups.protocols.UDP] (Thread-0 (ActiveMQ-server-ActiveMQServerImpl::serverUUID=c79278db-56e6-11e5-af50-69dd76236ee8-1573164340)) JGRP000015: the receive buffer of socket DatagramSocket was set to 20MB, but the OS only allocated 212.99KB. This might lead to performance problems. Please set your max receive buffer in the OS correctly (e.g. net.core.rmem_max on Linux)
WARNING [org.jgroups.protocols.UDP] (Thread-0 (ActiveMQ-server-ActiveMQServerImpl::serverUUID=c79278db-56e6-11e5-af50-69dd76236ee8-1573164340)) JGRP000015: the send buffer of socket MulticastSocket was set to 1MB, but the OS only allocated 212.99KB. This might lead to performance problems. Please set your max send buffer in the OS correctly (e.g. net.core.wmem_max on Linux)
WARNING [org.jgroups.protocols.UDP] (Thread-0 (ActiveMQ-server-ActiveMQServerImpl::serverUUID=c79278db-56e6-11e5-af50-69dd76236ee8-1573164340)) JGRP000015: the receive buffer of socket MulticastSocket was set to 25MB, but the OS only allocated 212.99KB. This might lead to performance problems. Please set your max receive buffer in the OS correctly (e.g. net.core.rmem_max on Linux)
If you reboot the singleton server node, the other node will be elected the new singleton provider, and will start receiving the MDB messages instead.
You should see the following output in the new singleton provider server:
WFLYCLSV0003: master:quickstart-messagingcluster-nodeX elected as the singleton provider of the org.wildfly.ejb3.clustered.singleton service
Where nodeX
will be either node1
or node2
, depending on which node is the new singleton provider.
If you now try to access the servlet urls, you will see that the new provider is receiving all new messages.
You will see the following warnings in the log of the server that is not the singleton provider. These messages show that the other node went down unexpectedly, which is exactly the scenario we are reproducing in this quickstart. For that reason, those warnings can be ignored.
WARN [org.apache.activemq.artemis.core.client] (Thread-2 (ActiveMQ-client-global-threads)) AMQ212037: Connection failure has been detected: AMQ119015: The connection was disconnected because of server shutdown [code=DISCONNECTED]
WARN [org.apache.activemq.artemis.core.server] (Thread-2 (ActiveMQ-client-global-threads)) AMQ222095: Connection failed with failedOver=false
You may see the following log message as well. When a server is restarted, it may broadcast that it is up and running (with its nodeID) while other nodes still reference the previous server instance for the same nodeID. Eventually, the cluster will be informed of the new instance representing the given nodeID but as the warning explains, it is possible to see this log (once or more) when a server is restarted.
WARN [org.apache.activemq.artemis.core.client] (activemq-discovery-group-thread-dg-group1) AMQ212034: There are more than one servers on the network broadcasting the same node id. You will see this message exactly once (per node) if a node is restarted, in which case it can be safely ignored. But if it is logged continuously it means you really do have more than one node on the same network active concurrently with the same node id. This could occur if you have a backup node active at the same time as its live node. nodeID=a114b652-689e-11e7-a2f4-54ee751c6182
The next error message is a known issue. You can ignore it, as it does not affect the scenario that this quickstart reproduces:
ERROR [org.apache.activemq.artemis.core.server] (Thread-3 (ActiveMQ-client-global-threads)) AMQ224037: cluster connection Failed to handle message: java.lang.IllegalStateException: Cannot find binding for jms.queue.HelloWorldMDBQueuedea3e995-713c-11e7-85f2-b8f6b112daf7 on ClusterConnectionImpl@1129705701[nodeUUID=dabaa1fa-713c-11e7-8f3a-b8f6b112daf7, connector=TransportConfiguration(name=http-connector, factory=org-apache-activemq-artemis-core-remoting-impl-netty-NettyConnectorFactory) ?httpUpgradeEndpoint=http-acceptor&activemqServerName=default&httpUpgradeEnabled=true&port=9080&host=localhost, address=jms, server=ActiveMQServerImpl::serverUUID=dabaa1fa-713c-11e7-8f3a-b8f6b112daf7]
at org.apache.activemq.artemis.core.server.cluster.impl.ClusterConnectionImpl$MessageFlowRecordImpl.doConsumerCreated(ClusterConnectionImpl.java:1294)
at org.apache.activemq.artemis.core.server.cluster.impl.ClusterConnectionImpl$MessageFlowRecordImpl.handleNotificationMessage(ClusterConnectionImpl.java:1029)
at org.apache.activemq.artemis.core.server.cluster.impl.ClusterConnectionImpl$MessageFlowRecordImpl.onMessage(ClusterConnectionImpl.java:1004)
at org.apache.activemq.artemis.core.client.impl.ClientConsumerImpl.callOnMessage(ClientConsumerImpl.java:1001)
at org.apache.activemq.artemis.core.client.impl.ClientConsumerImpl.access$400(ClientConsumerImpl.java:49)
at org.apache.activemq.artemis.core.client.impl.ClientConsumerImpl$Runner.run(ClientConsumerImpl.java:1124)
at org.apache.activemq.artemis.utils.OrderedExecutorFactory$OrderedExecutor$ExecutorTask.run(OrderedExecutorFactory.java:101)
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)
Run the following command, replacing WILDFLY_HOME
with the path to your server, and replacing NODE_X
in the script name with either node1
or node2
, depending on whether the current singleton provider is node1
or node2
.
$ WILDFLY_HOME/bin/jboss-cli.sh --connect --file=restart-NODE_X-domain.cli
Note
|
For Windows, use the WILDFLY_HOME\bin\jboss-cli.bat script.
|
To disable the delivery group "my-mdb-delivery-group" to which the topic belongs, run the disable-delivery-group-domain.cli
or disable-delivery-group-standalone.cli
script, located in the root directory of this quickstart. Follow the instructions in the next sections, depending on the server configuration you choose to run.
After disabling the delivery group, try sending messages to the topic, You should notice that the topic messages are not delivered when the delivery group is inactive.
Next, enable the delivery group using the appropriate enable-delivery-group-domain.cli
or enable-delivery-group-standalone.cli
script, also located in the root directory of this quickstart, so that the topic messages can be delivered again.
To disable the delivery group named "my-mdb-delivery-group" to which the topic belongs, run the disable-delivery-group-domain.cli
script, replacing WILDFLY_HOME
with the path to your server:
$ WILDFLY_HOME/bin/jboss-cli.sh --connect --file=disable-delivery-group-domain.cli
Note
|
For Windows, use the WILDFLY_HOME\bin\jboss-cli.bat script.
|
Similarly, to enable the delivery group, run the enable-delivery-group-domain.cli
script:
$ WILDFLY_HOME/bin/jboss-cli.sh --connect --file=enable-delivery-group-domain.cli
Note
|
For Windows, use the WILDFLY_HOME\bin\jboss-cli.bat script.
|
To disable the delivery group named "my-mdb-delivery-group" to which the topic belongs, run the disable-delivery-group-standalone.cli
script on both servers, replacing WILDFLY_HOME
with the path to your server:
$ WILDFLY_HOME_1/bin/jboss-cli.sh --connect --file==disable-delivery-group-standalone.cli
$ WILDFLY_HOME_2/bin/jboss-cli.sh --connect controller=localhost:10090 --file=disable-delivery-group-standalone.cli
Note
|
For Windows, use the WILDFLY_HOME_1\bin\jboss-cli.bat and WILDFLY_HOME_2\bin\jboss-cli.bat scripts.
|
Similarly, to enable the delivery group, run the enable-delivery-group-standalone.cli
script in both servers:
$ WILDFLY_HOME_1/bin/jboss-cli.sh --connect --file==enable-delivery-group-standalone.cli
$ WILDFLY_HOME_2/bin/jboss-cli.sh --connect controller=localhost:10090 --file=enable-delivery-group-standalone.cli
Note
|
For Windows, use the WILDFLY_HOME_1\bin\jboss-cli.bat and WILDFLY_HOME_2\bin\jboss-cli.bat scripts.
|
When you are finished testing, use the following instructions to undeploy the quickstart.
-
Start the WildFly managed domain as described above.
-
Open a new terminal, navigate to the root directory of this quickstart, and run the
undeploy-domain.cli
script, replacingWILDFLY_HOME
with the path to your server:$ WILDFLY_HOME/bin/jboss-cli.sh --connect --file==undeploy-domain.cli
Note
|
For Windows, use the WILDFLY_HOME\bin\jboss-cli.bat script.
|
When you are finished testing the quickstart, follow these steps to undeploy the archive.
-
Make sure WildFly server is started.
-
Open a terminal and navigate to the root directory of this quickstart.
-
Type this command to undeploy the archive:
$ mvn wildfly:undeploy
Repeat the last step for the second server:
$ mvn wildfly:undeploy -Dwildfly.port=10090
You can restore the domain configuration by manually restoring the backup configuration files or by running the management CLI script.
Important
|
This method ensures the server is restored to its prior configuration. |
-
If it is running, stop the WildFly server.
-
Restore the
WILDFLY_HOME/domain/configuration/domain.xml
andWILDFLY_HOME/domain/configuration/host.xml
files with the back-up copies of the files. Make sure you replaceWILDFLY_HOME
with the path to your server.
Important
|
This script returns the server to a default configuration and the result might not match the server configuration prior to testing this quickstart. If you were not running with the default configuration before testing this quickstart, you should follow the intructions above to manually restore the configuration to its previous state. |
-
Start the WildFly managed domain as described above.
-
Open a new terminal, navigate to the root directory of this quickstart, and run the
remove-domain.cli
script, replacingWILDFLY_HOME
with the path to your server.$ WILDFLY_HOME/bin/jboss-cli.sh --connect --file=remove-domain.cli
NoteFor Windows, use the WILDFLY_HOME\bin\jboss-cli.bat
script.This script removes the server configuration that was done by the
install-domain.cli
script. You should see the following result following the script commands:The batch executed successfully
You can Restore the standalone configuration by manually restoring the back-up copies the configuration files or by running the management CLI script.
Important
|
This method ensures the server is restored to its prior configuration. |
-
If they are running, stop both WildFly servers.
-
Restore the
WILDFLY_HOME_1/standalone/configuration/standalone-full-ha.xml
file with the back-up copies of the file. Make sure you replaceWILDFLY_HOME_1
with the path to your server.
Important
|
This script returns the server to the default configuration, which might not match the server configuration that existed prior to testing this quickstart. If you were not running with the default configuration before testing this quickstart, you should follow the intructions above to manually restore the standalone configuration to its previous state. |
-
Start the WildFly server by typing the following, replacing
WILDFLY_HOME_1
with the path to your first server:$ WILDFLY_HOME_1/bin/standalone.sh -c standalone-full-ha.xml
NoteFor Windows, use the WILDFLY_HOME_1\bin\standalone.bat
script. -
Open a new terminal, navigate to the root directory of this quickstart, and run the
remove-standalone.cli
script, replacingWILDFLY_HOME_1
with the path to your first server.$ WILDFLY_HOME_1/bin/jboss-cli.sh --connect --file=remove-standalone.cli
NoteFor Windows, use the WILDFLY_HOME_1\bin\jboss-cli.bat
script.This script removes the server configuration that was done by the install-standalone.cli script. You should see the following result following the script commands:
The batch executed successfully
Instead of using a standard WildFly server distribution, the two WildFly servers to deploy and run the quickstart can be alternatively provisioned by activating the Maven profile named provisioned-server
when building the quickstart:
$ mvn clean package -Pprovisioned-server
The provisioned WildFly servers, with the quickstart deployed, can then be found in the target
directory, and their usage is similar to a standard server distribution, with the simplification that there is never the need to specify the server configuration to be started.
The server provisioning functionality is provided by the WildFly Maven Plugin, and you may find its configuration in the pom.xml files of the quickstart.
The integration tests included with this quickstart, which verify that the quickstart runs correctly, may also be run with provisioned server.
Follow these steps to run the integration tests.
-
Make sure the servers are provisioned by running the commands reported in Building and running the quickstart application with provisioned WildFly server
-
Start the WildFly provisioned servers in three distinct terminals, this time using the WildFly Maven Plugin, which is recommended for testing due to simpler automation.
$ mvn wildfly:start -Djboss-as.home=target/server
$ mvn wildfly:start -Djboss-as.home=target/server2 \ -Dwildfly.port=10090 \ -Dwildfly.javaOpts="-Djboss.socket.binding.port-offset=100"
-
Type the following command to run the
verify
goal with theintegration-testing
profile activated, and specifying the quickstart’s URL using theserver.host
system property.$ mvn verify -Pintegration-testing -Dserver.host="http://localhost:8080"
$ mvn verify -Pintegration-testing -Dserver.host="http://localhost:8180"
-
To shut down the WildFly provisioned servers using the WildFly Maven Plugin:
$ mvn wildfly:shutdown
$ mvn wildfly:shutdown -Dwildfly.port=10090