activemq-artemis/docs/user-manual/en/address-model.md

28 KiB
Raw Blame History

Apache ActiveMQ Artemis Addressing and Queues

Apache ActiveMQ Artemis has a unique addressing model that is both powerful and flexible and that offers great performance. The addressing model comprises three main concepts: addresses, queues and routing types.

An address represents a messaging endpoint. Within the configuration, a typical address is given a unique name, 0 or more queues, and a routing type.

A queue is associated with an address. There can be multiple queues per address. Once an incoming message is matched to an address, the message will be sent on to one or more of its queues, depending on the routing type configured. Queues can be configured to be automatically created and deleted.

A routing type determines how messages are sent to the queues associated with an address. An Apache ActiveMQ Artemis address can be configured with two different routing types.

Table 1. Routing Types

If you want your messages routed to… Use this routing type …​
A single queue within the matching address, in a point-to-point manner. Anycast
Every queue within the matching address, in a publish-subscribe manner. Multicast

Note: It is possible to define more than one routing type per address, but this typically results in an anti-pattern and is therefore not recommended. If an address does use both routing types, however, and the client does not show a preference for either one, the broker typically defaults to the anycast routing type. The one exception is when the client uses the MQTT protocol. In that case, the default routing type is multicast. |

Basic Address Configuration

The following examples show how to configure basic point to point and publish subscribe addresses.

Point-to-Point Messaging

Point-to-point messaging is a common scenario in which a message sent by a producer has only one consumer. AMQP and JMS message producers and consumers can make use of point-to-point messaging queues, for example. Define an anycast routing type for an address so that its queues receive messages in a point-to-point manner.

When a message is received on an address using anycast, Apache ActiveMQ Artemis locates the queue associated with the address and routes the message to it. When consumers request to consume from the address, the broker locates the relevant queue and associates this queue with the appropriate consumers. If multiple consumers are connected to the same queue, messages are distributed amongst each consumer equally, providing the consumers are equally able to handle them.

Point to Point Figure 1. Point to Point Messaging

Configuring an Address to Use the Anycast Routing Type

Open the file /etc/broker.xml for editing.

Add an address configuration element and its associated queue if they do not exist already.

Note For normal Point to Point semantics, the queue name MUST match the address name.

<configuration ...>
  <core ...>
    ...
    <address name="orders">
      <anycast>
        <queue name="orders"/>
      </anycast>
    </address>
  </core>
</configuration>

Publish-Subscribe Messaging

In a publish-subscribe scenario, messages are sent to every consumer subscribed to an address. JMS topics and MQTT subscriptions are two examples of publish-subscribe messaging.

To configure an address with publish-subscribe semantics, create an address with the multicast routing tyoe.

Publish Subscribe Figure 2. Publish-Subscribe

Configuring an Address to Use the Multicast Routing Type

Open the file /etc/broker.xml for editing.

Add an address configuration element with multicast routing type.

<configuration ...>
  <core ...>
    ...
    <address name="pubsub.foo">
      <multicast/>
    </address>
  </core>
</configuration>

When clients connect to an address with the multicast element, a subscription queue for the client will be automatically created for the client. It is also possible to pre-configure subscription queues and connect to them directly using the queue's Fully Qualified Queue names.

Optionally add one or more queue elements to the address and wrap the multicast element around them. This step is typically not needed since the broker will automatically create a queue for each subscription requested by a client.

<configuration ...>
  <core ...>
    ...
    <address name="pubsub.foo">
      <multicast>
        <queue name="client123.pubsub.foo"/>
        <queue name="client456.pubsub.foo"/>
      </multicast>
    </address>
  </core>
</configration>

Figure 3. Point-to-Point with Two Queues

Point-to-Point Address multiple Queues

It is actually possible to define more than one queue on an address with an anycast routing type. When messages are received on such an address, they are firstly distributed evenly across all the defined queues. Using Fully Qualified Queue names, clients are able to select the queue that they would like to subscribe to. Should more than one consumer connect direct to a single queue, Apache ActiveMQ Artemis will take care of distributing messages between them, as in the example above.

Point to Point Figure 3. Point-to-Point with Two Queues


Note: This is how Apache ActiveMQ Artemis handles load balancing of queues across multiple nodes in a cluster. Configuring a Point-to-Point Address with two queues, open the file /etc/broker.xml for editing.

Add an address configuration with Anycast routing type element and its associated queues.

<configuration ...>
  <core ...>
    ...
    <address name="address.foo">
      <anycast>
        <queue name="q1"/>
        <queue name="q2"/>
      </anycast>
    </address>
  </core>
</configuration>

Point-to-Point and Publish-Subscribe Addresses

It is possible to define an address with both point-to-point and publish-subscribe semantics enabled. While not typically recommend, this can be useful when you want, for example, a JMS Queue say orders and a JMS Topic named orders. The different routing types make the addresses appear to be distinct.

Using an example of JMS Clients, the messages sent by a JMS queue producer will be routed using the anycast routing type. Messages sent by a JMS topic producer will use the multicast routing type. In addition when a JMS topic consumer attaches, it will be attached to its own subscription queue. JMS queue consumer will be attached to the anycast queue.

Point to Point Figure 4. [Point-to-Point and Publish-Subscribe


Note: The behavior in this scenario is dependent on the protocol being used. For JMS there is a clear distinction between topic and queue producers and consumers, which make the logic straight forward. Other protocols like AMQP do not make this distinction. A message being sent via AMQP will be routed by both anycast and multicast and consumers will default to anycast. For more information, please check the behavior of each protocol in the sections on protocols.

The XML snippet below is an example of what the configuration for an address using both anycast and multicast would look like in /etc/broker.xml. Note that subscription queues are typically created on demand, so there is no need to list specific queue elements inside the multicast routing type.

<configuration ...>
  <core ...>
    ...
    <address name="foo.orders">
      <anycast>
        <queue name="orders"/>
      </anycast>
      <multicast/>
    </address>
  </core>
</configuration>

How to filter messages

Apache ActiveMQ Artemis supports the ability to filter messages using Apache Artemis Filter Expressions.

Filters can be applied in two places, on a queue and on a consumer.

Queue Filter

When a filter is applied to a queue, messages are filter before they sent to the queue. To add a queue filter use the filter element when configuring a queue. Open up the broker.xml and add an address with a queue, using the filter element to configure a filter on this queue.

    <address name="filter">
       <queue name="filter">
          <filter string="color='red'"/>
        </queue>
    </address>

The filter defined above ensures that only messages with an attribute "color='red'" is sent to this queue.

Consumer Filters

Consumer filters are applied after messages have reached a queue and are defined using the appropriate client APIs. The follow JMS example shows how to consumer filters work.

  1. Define an address with a single queue, with no filter applied.
    <address name="filter">
       <queue name="filter">
       </queue>
    </address>
  ...
  // Send some messages
  for (int i = 0; i < 3; i ++) {
    TextMessage redMessage = senderSession.createTextMessage("Red");
    redMessage.setStringProperty("color", "red");
    producer.send(redMessage)
        
    TextMessage greenMessage = senderSession.createTextMessage("Green");
    greenMessage.setStringProperty("color", "green");
    producer.send(greenMessage)
  }

At this point the queue would have 6 messages: red,green,red,green,red,green

  MessageConsumer redConsumer = redSession.createConsumer(queue, "color='red'");

The redConsumer has a filter that only matches "red" messages. The redConsumer will receive 3 messages.

red, red, red

The resulting queue would now be

green, green, green

Creating and Deleting Addresses and Queues Automatically

You can configure Apache ActiveMQ Artemis to automatically create addresses and queues, and then delete them when they are no longer in use. This saves you from having to preconfigure each address and queue before a client can connect to it. Automatic creation and deletion is configured on a per address basis and is controlled by following:

Parameter Description
auto-create-addresses When set to true, the broker will create the address requested by the client if it does not exist already. The default is true.
auto-delete-addresses When set to true, the broker will be delete any auto-created adddress once all of its queues have been deleted. The default is true
default-address-routing-type The routing type to use if the client does not specify one. Possible values are MULTICAST and ANYCAST. See earlier in this chapter for more information about routing types. The default value is MULTICAST.

Configuring an Address to be Automatically Created

Edit the file /etc/broker.xml and add the auto-create-addresses element to the address-setting you want the broker to automatically create.

(Optional) Add the address-setting if it does not exits. Use the match parameter and the The Apache ActiveMQ Artemis Wildcard Syntax to match more than one specific address.

Set auto-create-addresses to true

(Optional) Assign MULTICAST or ANYCAST as the default routing type for the address.

The example below configures an address-setting to be automatically created by the broker. The default routing type to be used if not specified by the client is MULTICAST. Note that wildcard syntax is used. Any address starting with /news/politics/ will be automatically created by the broker.

<configuration ...>
  <core ...>
    ...
    <address-settings>
       <address-setting match="/news/politics/#">
          <auto-create-addresses>true</auto-create-addresses>
          <default-address-routing-type>MULTICAST</default-address-routing-type>
       </address-setting>
    </address-settings>
    ...
  </core>
</configuration>

Configuring an Address to be Automatically Deleted

Edit the file /etc/broker.xml and add the auto-delete-addresses element to the address-setting you want the broker to automatically create.

(Optional) Add the address-setting if it does not exits. Use the match parameter and the The Apache ActiveMQ Artemis Wildcard Syntax to match more than one specific address.

Set auto-delete-addresses to true

The example below configures an address-setting to be automatically deleted by the broker. Note that wildcard syntax is used. Any address request by the client that starts with /news/politics/ is configured to be automatically deleted by the broker.

<configuration ...>
  <core ...>
    ...
    <address-settings>
       <address-setting match="/news/politics/#">
          <auto-create-addresses>true</auto-create-addresses>
          <default-address-routing-type>MULTICAST</default-address-routing-type>
       </address-setting>
    </address-settings>
    ...
  </core>
</configuration>

Fully Qualified Queue Names

Internally the broker maps a clients request for an address to specific queues. The broker decides on behalf of the client which queues to send messages to or from which queue to receive messages. However, more advanced use cases might require that the client specify a queue directly. In these situations the client and use a fully qualified queue name, by specifying both the address name and the queue name, separated by a ::.

Currently Artemis supports fully qualified queue names on Core, AMQP, JMS, OpenWire, MQTT and Stomp protocols for receiving messages only.

Specifying a Fully Qualified Queue Name

In this example, the address foo is configured with two queues q1, q2 as shown in the configuration below.

<configuration ...>
  <core ...>
    ...
    <addresses>
       <address name="foo">
          <anycast>
             <queue name="q1" />
             <queue name="q2" />
          </anycast>
       </address>
    </addresses>
  </core>
</configuration>

In the client code, use both the address name and the queue name when requesting a connection from the broker. Remember to use two colons, ::, to separate the names, as in the example Java code below.

String FQQN = "foo::q1";
Queue q1 session.createQueue(FQQN);
MessageConsumer consumer = session.createConsumer(q1);

Configuring a Prefix to Connect to a Specific Routing Type

Normally, if a Apache ActiveMQ Artemis receivs a message sent to a particular address, that has both anycast and multicast routing types enable, Apache ActiveMQ Artemis will route a copy of the message to one of the anycast queues and to all of the multicast queues.

However, clients can specify a special prefix when connecting to an address to specify whether to connect using anycast or multicast. The prefixes are custom values that are designated using the anycastPrefix and multicastPrefix parameters within the URL of an acceptor.

Configuring an Anycast Prefix

In /etc/broker.xml, add the anycastPrefix to the URL of the desired acceptor. In the example below, the acceptor is configured to use anycast:// for the anycastPrefix. Client code can specify anycast://foo/ if the client needs to send a message to only one of the anycast queues.

<configuration ...>
  <core ...>
    ...
      <acceptors>
         <acceptor name="artemis">tcp://0.0.0.0:61616?protocols=AMQP;anycastPrefix=anycast://</acceptor>
      </acceptors>
    ...
  </core>
</configuration>

Configuring a Multicast Prefix

In /etc/broker.xml, add the anycastPrefix to the URL of the desired acceptor. In the example below, the acceptor is configured to use multicast:// for the multicastPrefix. Client code can specify multicast://foo/ if the client needs the message sent to only the multicast queues of the address.

<configuration ...>
  <core ...>
    ...
      <acceptors>
         <acceptor name="artemis">tcp://0.0.0.0:61616?protocols=AMQP;multicastPrefix=multicast://</acceptor>
      </acceptors>
    ...
  </core>
</configuration>

Advanced Address Configuration

Pre-configuring subscription queue semantics

In most cases its not necessary to pre-create subscription queues. The relevant protocol managers take care of creating subscription queues when clients request to subscribe to an address. The type of subscription queue created, depends on what properties the client request. E.g. durable, non-shared, shared etc... Protocol managers uses special queue names to identify which queues below to which consumers and users need not worry about the details.

However, there are scenarios where a user may want to use broker side configuration to pre-configure a subscription. And later connect to that queue directly using a Fully Qualified Queue name.. The examples below show how to use broker side configuration to pre-configure a queue with publish subscribe behavior for shared, non-shared, durable and non-durable subscription behavior.

Configuring a shared durable subscription queue with up to 10 concurrent consumers

The default behavior for queues is to not limit the number connected queue consumers. The max-consumers parameter of the queue element can be used to limit the number of connected consumers allowed at any one time.

Open the file /etc/broker.xml for editing.

<configuration ...>
  <core ...>
    ...
    <address name="durable.foo">
      <multicast>
        <!-- pre-configured shared durable subscription queue -->
        <queue name="q1" max-consumers="10">
          <durable>true</durable>
        </queue>
      </multicast>
    </address>
  </core>
</configuration>

Configuring a non-shared durable subscription

The broker can be configured to prevent more than one consumer from connecting to a queue at any one time. The subscriptions to queues configured this way are therefore "non-shared". To do this simply set the max-consumers parameter to "1"

<configuration ...>
  <core ...>
    ...
    <address name="durable.foo">
      <multicast>
        <!-- pre-configured non shared durable subscription queue -->
        <queue name="q1" max-consumers="1">
          <durable>true</durable>
        </queue>
      </multicast>
    </address>
  </core>
</configuration>

Pre-configuring a queue as a non-durable subscription queue

Non-durable subscriptions are again usually managed by the relevant protocol manager, by creating and deleting temporary queues.

If a user requires to pre-create a queue that behaves like a non-durable subscription queue the purge-on-no-consumers flag can be enabled on the queue. When purge-on-no-consumers is set to true. The queue will not start receiving messages until a consumer is attached. When the last consumer is detached from the queue. The queue is purged (it's messages are removed) and will not receive any more messages until a new consumer is attached.

Open the file /etc/broker.xml for editing.

<configuration ...>
  <core ...>
    ...
    <address name="non.shared.durable.foo">
      <multicast>
        <queue name="orders1" purge-on-no-consumers="true"/>
      </multicast>
    </address>
  </core>
</configuration>

Additional Information: Protocol Managers, Address

A protocol manager maps protocol specific concepts down to the Apache ActiveMQ Artemis core model of addresses, queues and routing types. For example, when a client sends a MQTT subscription packet with the addresses

/house/room1/lights
/house/room2/lights

The MQTT protocol manager understands that the two addresses require multicast semantics. The protocol manager will therefore first look to ensure that multicast is enabled for both addresses. If not, it will attempt to dynamically create them. If successful, the protocol manager will then create special subscription queues with special names, for each subscription requested by the client.

The special name allows the protocol manager to quickly identify the required client subscription queues should the client disconnect and reconnect at a later date. If the subscription is temporary the protocol manager will delete the queue once the client disconnects.

When a client requests to subscribe to a point to point address. The protocol manager will look up the queue associated with the point to point address. This queue should have the same name as the addresss.

Note: If the queue is auto created, it will be auto deleted once there are no consumers and no messages in it. For more information on auto create see the next section Configuring Addresses and Queues via Address Settings

Configuring Addresses and Queues via Address Settings

There are some attributes that are defined against an address wildcard rather than a specific address/queue. Here an example of an address-setting entry that would be found in the broker.xml file.

<address-settings>
   <address-setting match="order.foo">
      <dead-letter-address>DLA</dead-letter-address>
      <max-delivery-attempts>3</max-delivery-attempts>
      <redelivery-delay>5000</redelivery-delay>
      <expiry-address>ExpiryQueue</expiry-address>
      <last-value-queue>true</last-value-queue>
      <max-size-bytes>100000</max-size-bytes>
      <page-size-bytes>20000</page-size-bytes>
      <redistribution-delay>0</redistribution-delay>
      <send-to-dla-on-no-route>true</send-to-dla-on-no-route>
      <address-full-policy>PAGE</address-full-policy>
      <slow-consumer-threshold>-1</slow-consumer-threshold>
      <slow-consumer-policy>NOTIFY</slow-consumer-policy>
      <slow-consumer-check-period>5</slow-consumer-check-period>
   </address-setting>
</address-settings>

The idea with address settings, is you can provide a block of settings which will be applied against any addresses that match the string in the match attribute. In the above example the settings would only be applied to the address "order.foo" address but you can also use wildcards to apply settings. See: The chapter on the wild card syntax.

For example, if you used the match string jms.queue.# the settings would be applied to all addresses which start with jms.queue. which would be all JMS queues.

The meaning of the specific settings are explained fully throughout the user manual, however here is a brief description with a link to the appropriate chapter if available.

max-delivery-attempts defines how many time a cancelled message can be redelivered before sending to the dead-letter-address. A full explanation can be found here.

redelivery-delay defines how long to wait before attempting redelivery of a cancelled message. see here.

expiry-address defines where to send a message that has expired. see here.

expiry-delay defines the expiration time that will be used for messages which are using the default expiration time (i.e. 0). For example, if expiry-delay is set to "10" and a message which is using the default expiration time (i.e. 0) arrives then its expiration time of "0" will be changed to "10." However, if a message which is using an expiration time of "20" arrives then its expiration time will remain unchanged. Setting expiry-delay to "-1" will disable this feature. The default is "-1".

last-value-queue defines whether a queue only uses last values or not. see here.

max-size-bytes and page-size-bytes are used to set paging on an address. This is explained here.

redistribution-delay defines how long to wait when the last consumer is closed on a queue before redistributing any messages. see here.

send-to-dla-on-no-route. If a message is sent to an address, but the server does not route it to any queues, for example, there might be no queues bound to that address, or none of the queues have filters that match, then normally that message would be discarded. However if this parameter is set to true for that address, if the message is not routed to any queues it will instead be sent to the dead letter address (DLA) for that address, if it exists.

address-full-policy. This attribute can have one of the following values: PAGE, DROP, FAIL or BLOCK and determines what happens when an address where max-size-bytes is specified becomes full. The default value is PAGE. If the value is PAGE then further messages will be paged to disk. If the value is DROP then further messages will be silently dropped. If the value is FAIL then further messages will be dropped and an exception will be thrown on the client-side. If the value is BLOCK then client message producers will block when they try and send further messages. See the following chapters for more info Flow Control, Paging.

slow-consumer-threshold. The minimum rate of message consumption allowed before a consumer is considered "slow." Measured in messages-per-second. Default is -1 (i.e. disabled); any other valid value must be greater than 0.

slow-consumer-policy. What should happen when a slow consumer is detected. KILL will kill the consumer's connection (which will obviously impact any other client threads using that same connection). NOTIFY will send a CONSUMER_SLOW management notification which an application could receive and take action with. See slow consumers for more details on this notification.

slow-consumer-check-period. How often to check for slow consumers on a particular queue. Measured in seconds. Default is 5. See slow consumers for more information about slow consumer detection.

auto-create-jms-queues. Whether or not the broker should automatically create a JMS queue when a JMS message is sent to a queue whose name fits the address match (remember, a JMS queue is just a core queue which has the same address and queue name) or a JMS consumer tries to connect to a queue whose name fits the address match. Queues which are auto-created are durable, non-temporary, and non-transient. Default is true. This is DEPRECATED. See auto-create-queues.

auto-delete-jms-queues. Whether or not the broker should automatically delete auto-created JMS queues when they have both 0 consumers and 0 messages. Default is true. This is DEPRECATED. See auto-delete-queues.

auto-create-jms-topics. Whether or not the broker should automatically create a JMS topic when a JMS message is sent to a topic whose name fits the address match (remember, a JMS topic is just a core address which has one or more core queues mapped to it) or a JMS consumer tries to subscribe to a topic whose name fits the address match. Default is true. This is DEPRECATED. See auto-create-addresses.

auto-delete-jms-topics. Whether or not the broker should automatically delete auto-created JMS topics once the last subscription on the topic has been closed. Default is true. This is DEPRECATED. See auto-delete-addresses.

auto-create-queues. Whether or not the broker should automatically create a queue when a message is sent or a consumer tries to connect to a queue whose name fits the address match. Queues which are auto-created are durable, non-temporary, and non-transient. Default is true.

auto-delete-queues. Whether or not the broker should automatically delete auto-created queues when they have both 0 consumers and 0 messages. Default is true.

config-delete-queues. How the broker should handle queues deleted on config reload, by delete policy: OFF or FORCE. See config-reload for more details. Default is OFF.

auto-create-addresses. Whether or not the broker should automatically create an address when a message is sent to or a consumer tries to consume from a queue which is mapped to an address whose name fits the address match. Default is true.

auto-delete-addresses. Whether or not the broker should automatically delete auto-created addresses once the address no longer has any queues. Default is true.

config-delete-addresses. How the broker should handle addresses deleted on config reload, by delete policy: OFF or FORCE. See config-reload for more details. Default is OFF.