activemq-artemis/docs/user-manual/core.adoc

= Using Core
:idprefix:
:idseparator: -

Apache ActiveMQ Artemis core is a messaging system with its own API.
We call this the _core API_.

If you don't want to use the JMS API or any of the other supported protocols you can use the core API directly.
The core API provides all the functionality of JMS but without much of the complexity.
It also provides features that are not available using JMS.

== Core Messaging Concepts

Some of the core messaging concepts are similar to JMS concepts, but core messaging concepts are also different in some ways as well.
In general the core API is simpler than the JMS API, since we remove distinctions between queues, topics and subscriptions.
We'll discuss each of the major core messaging concepts in turn, but to see the API in detail please consult the Javadoc.

Also refer to the xref:address-model.adoc#address-model[address model] chapter for a high-level overview of these concepts as well as configuration details.

=== Message

* A message is the unit of data which is sent between clients and servers.
* A message has a body which is a buffer containing convenient methods for reading and writing data into it.
* A message has a set of properties which are key-value pairs.
Each property key is a string and property values can be of type integer, long, short, byte, byte[], String, double, float or boolean.
* A message has an _address_ it is being sent to.
When the message arrives on the server it is routed to any queues that are bound to the address.
The routing semantics (i.e. anycast or multicast) are determined by the "routing type" of the address and queue.
If the queues are bound with any filter, the message will only be routed to that queue if the filter matches.
An address may have many queues bound to it or even none.
There may also be entities other than queues (e.g. _diverts_) bound to addresses.
* Messages can be either durable or non durable.
Durable messages in a durable queue will survive a server crash or restart.
Non durable messages will never survive a server crash or restart.
* Messages can be specified with a priority value between 0 and 9.
0 represents the lowest priority and 9 represents the highest.
The broker will attempt to deliver higher priority messages before lower priority ones.
* Messages can be specified with an optional expiry time.
The broker will not deliver messages after its expiry time has been exceeded.
* Messages also have an optional timestamp which represents the time the message was sent.
* Apache ActiveMQ Artemis also supports the sending/consuming of very large messages much larger than can fit in available RAM at any one time.

== Core API

=== ServerLocator

Clients use `ServerLocator` instances to create `ClientSessionFactory` instances.
`ServerLocator` instances are used to locate servers and create connections to them.

In JMS terms think of a `ServerLocator` in the same way you would a JMS Connection Factory.

`ServerLocator` instances are created using the `ActiveMQClient` factory class.

=== ClientSessionFactory

Clients use `ClientSessionFactory` instances to create `ClientSession` instances.
`ClientSessionFactory` instances are basically the connection to a server

In JMS terms think of them as JMS Connections.

`ClientSessionFactory` instances are created using the `ServerLocator` class.

=== ClientSession

A client uses a ``ClientSession``for consuming and producing messages and for grouping them in transactions.
`ClientSession` instances can support both transactional and non transactional semantics and also provide an `XAResource` interface so messaging operations can be performed as part of a http://www.oracle.com/technetwork/java/javaee/tech/jta-138684.html[JTA] transaction.

`ClientSession` instances group `ClientConsumer` instances and `ClientProducer` instances.

`ClientSession` instances can be registered with an optional `SendAcknowledgementHandler`.
This allows your client code to be notified asynchronously when sent messages have successfully reached the server.
This unique Apache ActiveMQ Artemis feature, allows you to have full guarantees that sent messages have reached the server without having to block on each message sent until a response is received.
Blocking on each messages sent is costly since it requires a network round trip for each message sent.
By not blocking and receiving send acknowledgements asynchronously you can create true end to end asynchronous systems which is not possible using the standard JMS API.
For more information on this advanced feature please see the section xref:send-guarantees.adoc#guarantees-of-sends-and-commits[Guarantees of sends and commits].

==== Identifying your session for management and debugging

Assigning IDs to your core sessions can help you with monitoring and debugging the cluster using the xref:management-console.adoc#management-console[management console].

[,java]
----
 ClientSession session;
 // ...
 session.addMetaData(ClientSession.JMS_SESSION_IDENTIFIER_PROPERTY, "jms-client-id");
 session.addMetaData("jms-client-id", "my-session");
----

Such ID will then appear in the *Client ID* column under the *Connections*, *Consumers* and *Producers* tabs.

If you are using the JMS API, the `setClientID` would give you the same effect.

=== ClientConsumer

Clients use `ClientConsumer` instances to consume messages from a queue.
Core messaging supports both synchronous and asynchronous message consumption semantics.
`ClientConsumer` instances can be configured with an optional filter expression and will only consume messages which match that expression.

=== ClientProducer

Clients create `ClientProducer` instances on `ClientSession` instances so they can send messages.
`ClientProducer` instances can specify an address to which all sent messages are routed, or they can have no specified address, and the address is specified at send time for the message.

[WARNING]
====
Please note that `ClientSession`, `ClientProducer` and `ClientConsumer` instances are _designed to be re-used_.

It's an anti-pattern to create new `ClientSession`, `ClientProducer` and `ClientConsumer` instances for each message you produce or consume.
If you do this, your application will perform very poorly.
This is  discussed further in the section on performance tuning xref:perf-tuning.adoc#performance-tuning[Performance Tuning].
====

== A simple example of using Core

Here's a very simple program using the core messaging API to send and receive a message.
Logically it's comprised of two sections: firstly setting up the producer to write a message to an _address_, and secondly, creating a _queue_ for the consumer using anycast routing, creating the consumer, and _starting_ it.

[,java]
----
ServerLocator locator = ActiveMQClient.createServerLocator("vm://0");

// In this simple example, we just use one session for both producing and receiving

ClientSessionFactory factory =  locator.createClientSessionFactory();
ClientSession session = factory.createSession();

// A producer is associated with an address ...

ClientProducer producer = session.createProducer("example");
ClientMessage message = session.createMessage(true);
message.getBodyBuffer().writeString("Hello");

// We need a queue attached to the address ...

session.createQueue("example", RoutingType.ANYCAST, "example", true);

// And a consumer attached to the queue ...

ClientConsumer consumer = session.createConsumer("example");

// Once we have a queue, we can send the message ...

producer.send(message);

// We need to start the session before we can -receive- messages ...

session.start();
ClientMessage msgReceived = consumer.receive();

System.out.println("message = " + msgReceived.getBodyBuffer().readString());

session.close();
----