I need to change activemq persistence adapter from AMQ to KahaDB, but it's not acceptable to loose all undelivered messages stored via AMQAdapter. Is there any way to automatically use old adapter to send undelivered messages and then switch to KahaDB store?
The general solution to this is to create a new Broker instance and network the old broker and the new one. Once the brokers are networked you create a consumer on the new broker for the destinations and allow the demand to drain the mesasge's from old to new Broker.
Refer to this thread for more:
Related
I am trying to use a queue in activemq to dequeue messages from a virtual topic. I tried sending some messages and it is showing up in the topic under "message enqueued" but it is not able to be consumed.
The virtual topic name that i created was VirtualTopic.AA and the consumer is called Consumer.client1.VirtualTopic.AA.
In the consumer.client1.VirtualTopic.AA, i can see that there is a consumer but it is just not able to dequeue messages from the virtual topic.
Anyone knows why this is happening ? Do i need to change some settings in the configuration in the xml file ?
When you publish to a virtual topic using Spring's JmsTemplate, you need to configure it for a topic by setting the pubSubDomain property to "true".
From the JmsTemplate documentation:
If you want to use dynamic destination creation, you must specify the type of JMS destination to create, using the "pubSubDomain" property. For other operations, this is not necessary. Point-to-Point (Queues) is the default domain.
And in JmsDestinationAccessor#setPubSubDomain:
pubSubDomain - "true" for the Publish/Subscribe domain (Topics), "false" for the Point-to-Point domain (Queues)
I am new to JMS and currently developing a simple Chat application explained in Oreilly 'Java Message service'. I've configured a TopicConnectionFactory in ActiveMQ that receives chat messages from TopicPublishers and dispatch that to TopicSubscribers.
My question is 'why do we need to create more than one TopicConnectionFactory' in any JMS application? Since Connectionfactory instances are not tied up with a Topic/Queue, why can't we use one instance of ConnectionFactory for creating connections to all the Topics (or Queues) configured in an application?
Technically speaking you are right. You may be able just to use one ConnectionFactory.
However it is a better design to use multiple ConnectionFactories depending on your requirements so the traffic would be spread out evenly and you do not run out of connections.
So if you know about a JMS Client application that may be problematic (the logic does not allow proper connection handling open/close), you may isolate it to use its own connection factory.
Also some connection factories allow a pool of 10 default active connections at the same time (it depends on the implementation/ settings) if you will need more you may use more than one connection factory.
I've configured a TopicConnectionFactory in ActiveMQ that receives chat messages from TopicPublishers and dispatch that to TopicSubscribers.
Very ambiguous statement. TopicConnectionFactory does not receive or send any messages. It is just one of the admin Objects used to create Connection which in turn creates Session which in turn creates your publishers and subscribers which publish and subscribe the messages.
why can't we use one instance of ConnectionFactory for creating connections to all the Topics (or Queues) configured in an application?
You definitely can. There is no one stopping you from doing that.
As per specs
A connection factory object encapsulates a set of connection configuration
parameters that has been defined by an administrator. A client uses it to
create a connection with a JMS provider.
So unless you have different configuration requirements you can use same ConnectionFactory to create multiple Connections. And yes as otc has mentioned above number of connections is one of the configuration parameter.
I am using a distributed jms queue and weblogic is my app server. There are three jms servers deployed in my clustered enviroment. The producers just send the message using name of queue jndi lookup 'udq' for example. Now I have associated a consumer for each jms server and I was able to consume the message, no problem so far.
Here is question, can I have a single consumer to consume the messages from the 3 jms servers. The weblogic allows jndi naming for destination lookup with following syntax #
qsession1 = qcon1.createQueueSession(false, Session.AUTO_ACKNOWLEDGE);
qsession2 = qcon2.createQueueSession(false, Session.AUTO_ACKNOWLEDGE);
qsession3 = qcon3.createQueueSession(false, Session.AUTO_ACKNOWLEDGE);
queue1 = (Queue)ctx.lookup("JMSServer-1#UDQ");
queue2 = (Queue)ctx.lookup("JMSServer-2#UDQ");
queue3 = (Queue)ctx.lookup("JMSServer-3#UDQ");
qreceiver1 = qsession1.createReceiver(queue1);
qreceiver2 = qsession2.createReceiver(queue2);
qreceiver3 = qsession3.createReceiver(queue3);
qreceiver1.setMessageListener(this);
qreceiver2.setMessageListener(this);
qreceiver3.setMessageListener(this);
qcon1.start();
qcon2.start();
qcon3.start();
I have only one OnMessage implemented for the above consumer. This does not work. Any suggestions please..
No, you can't have a consumer receiving messages from more than one JMS server. A consumer can receive messages from only one JMS server. You need to create multiple consumers to receive messages from multiple JMS servers.
you can use "Queue Forwarding" function.
From online documentation of wls 10.3:
"Queue members can forward messages to other queue members by configuring the Forward Delay attribute in the Administration Console, which is disabled by default. This attribute defines the amount of time, in seconds, that a distributed queue member with messages, but which has no consumers, will wait before forwarding its messages to other queue members that do have consumers."
here a link: http://docs.oracle.com/cd/E13222_01/wls/docs103/jms/dds.html#wp1260816
You can configure this feature from administration of each single queue.
Hi I am new to the Websphere JMS. It is bit different configuration i.e we need to create BUS first. I also seen that Connection factory has service bus reference and same service reference is given in Queue. I have following understanding till now
We have enterprise BUS to special needs. i.e BUS has control over the messages before MDB.
We can filter,alter the message in the BUS.
What is relation between Connection factory and Queue.I just seen that Connection factory has same reference of BUS.
Why we refer the same BUS name in Connection factory and Queue.
Please correct my understanding and help me in finding answers of above questions.
SIBus is what holds the messaging engine(s) which contains destinations (point-to-point or publish/subscribe) and actually dispatch messages
In the JMS Part you got the Default JMS Provider which provides you JMS resources (Connection factories, Queues, Topics) which allows you to interact with the SIBus
In the SIBus you define a Destination on a messaging engine with some type (point-to-point or publish/subscribe)
A JMS Connection Factory creates connections to the messaging engine so you can send/receive messages, this is why you need to tell it which Bus you want it to create connections to
A JMS Queue referes to a point-to-point Destination in the SIBus.
Now for your questions:
The Bus itself is not intended to modify or filter messages
Filtering can be done by using a message selector in the MDB, when the MDB attach to the destination it will provide the message selector and the messaging engine will filter messages by it.Regarding modification I am not sure
As I stated before,
ConnectionFactory is what creates connections to the messaging engine while the Queue is an abstraction to the SIB Destination. This is why you need to tell each one which Bus it is interacting with.
Why we refer the same BUS name in Connection factory and Queue.
One reason is that the buses configured in the connection factory and the queue are not necessarily the same. That is the case in a topology with buses that are interconnected using SIBus links. In that case, you can connect to a messaging engine in one bus to send messages to a destination in another bus.
In other queue managers it is possible to setup a bridge between a queue to a topic so that publishers and subscribers does not need to know that they are using a topic:
Example in TIBCO EMS
create Bridge source=queue:QName
target=topic:TName
create Bridge
source=topic:TName target=queue:QName
How do I do this in WMQ?
See WebSphere MQ V7.0 Features and Enhancements page 47, it mentions the use of Alias queues and administrative subscriptions
WebSphere MQ V7.0 introduces an
extension to the alias queue object
that allows it to be resolved to the
new topic object. This is useful for
migrating point-to-point messaging
applications to the Publish/Subscribe
model. A traditional point-to-point
application that puts messages into
WebSphere MQ can operate as a
publisher without any code changes by
utilizing an alias queue that resolves
to a topic object. This is implemented
administratively by defining a topic
object that maps to an appropriate
topic string on which the messages are
to be published. The original local
queue is deleted and replaced by an
alias queue of the same name that
resolves to the topic object. Also
note that a point-to-point application
that gets messages from WebSphere MQ
can operate as a subscriber without
any code changes by defining an
administrative subscription to a
topic.
Example 9.3 on page 220 shows how to setup a administrative subscription from a Topic to a destination queue
DEFINE SUB(SUB.RETAIL.CAT) TOPICOBJ(MATT.RETAIL.CAT) DESTCLAS(MANAGED)
DEFINE SUB(SUB.MATTRETCAT) TOPICSTR(‘matt/retail/cat’) DESTCLAS(MANAGED)
DEFINE SUB(SUB.PROVCAT) TOPICSTR(‘matt/retail/cat’) DEST(SUB.PROVCAT.DESTQ)